Genome wide screening of RNAi factors of Sf21 cells reveal several novel pathway associated proteins

Abstract

Background: RNA interference (RNAi) leads to sequence specific knock-down of gene expression and has emerged as an important tool to analyse gene functions, pathway analysis and gene therapy. Although RNAi is a conserved cellular process involving common elements and factors, species-specific differences have been observed among different eukaryotes. Identification of components for RNAi pathway is pursued intensively and successful genome-wide screens have been performed for components of RNAi pathways in various organisms. Functional comparative genomics analysis offers evolutionary insight that forms basis of discoveries of novel RNAi-factors within related organisms. Keeping in view the academic and commercial utility of insect derived cell-line from Spodoptera frugiperda, we pursued the identification and functional analysis of components of RNAi-machinery of Sf21 cell-line using genome-wide application.

Results: The genome and transcriptome of Sf21 was assembled and annotated. In silico application of comparative genome analysis among insects allowed us to identify several RNAi factors in Sf21 line. The candidate RNAi factors from assembled genome were validated by knockdown analysis of candidate factors using the siRNA screens on the Sf21-gfp reporter cell-line. Forty two (42) potential factors were identified using the cell based assay. These include core RNAi elements including Dicer-2, Argonaute-1, Drosha, Aubergine and auxiliary modules like chromatin factors, RNA helicases, RNA processing module, signalling allied proteins and others. Phylogenetic analyses and domain architecture revealed that Spodoptera frugiperda homologs retained identity with Lepidoptera (Bombyx mori) or Coleoptera (Tribolium castaneum) sustaining an evolutionary conserved scaffold in post-transcriptional gene silencing paradigm within insects.

Conclusion: The database of RNAi-factors generated by whole genome association survey offers comprehensive outlook about conservation as well as specific differences of the proteins of RNAi machinery. Understanding the interior involved in different phases of gene silencing also offers impending tool for RNAi-based applications.

Figure 1

A genome-wide screen for identification of Sf21 RNAi factors. Schematic representation of experimental design of the siRNA mediated reversion assay in Sf21cell line for RNAi screening.

Figure 2

Reversion Assay for putative RNAi factor. (A) (i) Fluorescence imaging of Sf21-gfp reporter cell line (ii) Sf21-gfp reporter cell line transfected with gfp siRNA (iii) Sf21-gfp reporter cell line co-transfected with putative RNAi factor Sf-Dcr-2 specific siRNA with gfp siRNA. (B) Histogram overlay plot shows different gfp expression in transfected Sf21 cell line. Plots depict number of cells (counts) on y axis vs. expression of the gfp reporter (FLH-1) on x axis. Black trace, normal Sf21 cells; green trace, Sf21-gfp reporter cell line; blue trace, siRNA transfected gfp silenced Sf21-gfp cells; red trace, gfp reverted Sf21-gfp cells with putative RNAi factor transfection (a) Sf-Ago-1 (b) Sf-Dcr-2. (C) The bar graph representation of the FACS results for (i) putative core RNAi factors Dicer-2, Argonaute-1, Aubergine, Drosha and (ii) auxiliary RNAi candidates RNA helicase DDX18/HAS1 subfamily, Integrator complex subunit (Int11), Multi Drug Resistance MDR1A transporter and Histone-3 Lysine-4 N-Methyltransferase in Sf21-gfp reporter cell line represented on the x axis, with the percentage of cells expressing gfp on the y axis. Data shown are mean ± SD of three independent experiments.

Figure 3

Domain analysis and phylogenetic tree of Dicer and Ribonuclease III protein family. (A) (i) Domain organisation clearly indicates the presence of an amino terminal DEAD domain with significant similarity with other Dicer homologs of Bombyx, Tribolium and Drosophila that helps to classify conservation of Sf21 RNAi factor Dicer-2 gene. (ii) The other Ribonuclease III component Drosha of Tribolium is devoid of N-terminal dsRBM. While Sf21-Drosha shows similarity with Drosophila and Bombyx. (B) Neighbor-joining tree is based on CLUSTALW alignment of full length proteins Dicer-1, Dicer-2 and Drosha as an out group. The phylogenetic tree distributes three different clusters where Sf-Dicer-2 is in the same clade with Bm-Dicer-2 and Sf-Drosha with a Bm-Ribonuclease III. (*) indicates a Bm-Ribonuclease III as a putative Drosha.

Figure 4

Domain analysis and phylogenetic tree of dsRBM effector proteins. (A) Tc, Bm and Sf21 has three tandem dsRBM in Loquacious. (B) The domain architecture of R2D2 is much conserved while all referring insects have two RNA binding domains. (C) Phylogenecic analysis revealed the dsRBM effector components of Sf21 Pasha, R2D2 and Loquacious are closest with respective Bombyx homologs.

Figure 5

Alignment Domain analysis and phylogenetic tree of Argonaute family of proteins. (A) Amino acid alignment of Argonaute-1 protein of Bombyx mori, Tribolium castaneum, Drosophila melanogaster and Sf21. Indicated sequences were aligned using CLUSTALX. The conserved amino acid residues Asp, Asp and His (DDH) triad residues are marked with red. (B) Domain architecture of Argonaute proteins. The similarity in the domain architectures of Bm-Ago-1 and Sf21-Ago-1 suggests that they are orthologous and the organisation of PAZ/Piwi domain is highly conserved within the insects. (C) The phylogenetic tree was based on consideration of two different class of Argonaute proteins, miRNA class (Argonaute-1) and siRNA class (Argonaute-2). Sf21 genome contains Ago-1 while Tribolium, Bombyx and Drosophila have both Ago-1 and Ago-2.

Figure 6

Domain structure and phylogeny of piRNA class components. (A) and (B) domain search finds both Sf21 Argonaute-3 and Aubergine like protein have conserved lineage of PAZ and Piwi domains. (C) A phylogenetic tree was created using the full-length sequences of piRNA class proteins. Siwi represents silkmoth Piwi.

Figure 7

Domain analysis and phylogenetic tree of DEAD RNA helicase domain proteins. (A) Conserved Domains Database search shows both Dbp45A and Vasa RNA helicases has DEAD domain and a C-terminal Helicase domain while DDX18/Has1 subfamily of helicase has an unknown domain which is present in all concerned helicase proteins of this family indicating clear orthology. (B) The neighbour-joining tree is based on the alignment of the conserved DEAD domain containing RNA helicases. The DEAD helicases forms three distinct clusters categorized into three subfamilies Dbp45A, Vasa and DDX18/Has1. Sf-DDX18 like proteins is closest with Bombyx mori but similar with Pitchoune of Drosophila and Tribolium too. Vasa subgroup family of proteins being important for a piRNA pathway suggests Sf-Vasa might participate in Piwi associated RNA silencing.

Figure 8

Sil protein phylogenetic analysis. The neighbour-joining tree is based on the alignment of sil gene family of proteins in different organisms. Sf-Sil-2 sub-clusters with the silkmoth Bm-Sil-2, while Tc-SirA and Tc-SirB compose a distinct cluster. Orthology of these insect is clear from this phylogenetic analysis that Lepidoptera Sid like proteins are evolutionary conserved. Ce-Sid1-3 and Hs-SidT1 and SidT2 diverge out separately and Tc-SirC joins Hs cluster.

Figure 9

Interactome Map of Sf21 RNAi factors. Protein-protein interaction data for 42 functionally validated RNAi factors identified in the siRNA mediated high through-put screening were constructed by homology based search in Tribolium. The protein interaction data shows 345 nodes with 1257 edges. The nodes in blue highlighted cluster identifies core RNAi components. The red nodes indicate Sf21 RNAi factor homologous to corresponding Tribolium protein. The pink node identifies the interactors of putative candidates postulated from Tribolium network.