Inhibitor of apoptosis 2 and TAK1-binding protein are components of the Drosophila Imd pathway

Abstract

The Imd signaling cascade, similar to the mammalian TNF-receptor pathway, controls antimicrobial peptide expression in Drosophila. We performed a large-scale RNAi screen to identify novel components of the Imd pathway in Drosophila S2 cells. In all, 6713 dsRNAs from an S2 cell-derived cDNA library were analyzed for their effect on Attacin promoter activity in response to Escherichia coli. We identified seven gene products required for the Attacin response in vitro, including two novel Imd pathway components: inhibitor of apoptosis 2 (Iap2) and transforming growth factor-activated kinase 1 (TAK1)-binding protein (TAB). Iap2 is required for antimicrobial peptide response also by the fat body in vivo. Both these factors function downstream of Imd. Neither TAB nor Iap2 is required for Relish cleavage, but may be involved in Relish nuclear localization in vitro, suggesting a novel mode of regulation of the Imd pathway. Our results show that an RNAi-based approach is suitable to identify genes in conserved signaling cascades.

Figure 1

RNAi-based screen to identify components of the Imd signaling cascade in S2 cells. (A) RNAi-mediated gene silencing effectively and specifically decreases the mRNA level of the targeted gene in S2 cells. A total of 2.5 × 10⁶ S2 cells in six-well plates containing 3 ml of medium were incubated with or without 10 μg of CG5210 dsRNA for 48 h. The expression level of more than 13 500 genes was measured using Affymetrix Drosophila genechips. Relative expression levels of all genes are shown. Each point represents the average mRNA expression level of three independent CG5210 RNAi treatments compared pairwise to untreated S2 cells. Arrowhead indicates the dot representing the CG5210 mRNA level. (B) dsRNA treatments targeting ribosomal proteins moderately affect the translation rate of Act5C in S2 cells. Expression of 40 genes coding for ribosomal proteins was silenced using RNAi, and translation rate of Act5C-β-gal was measured in relation to GFP RNAi-treated controls. Gene-specific β-galactosidase activities were thereafter compared to z-scores reported by Boutros et al (2004). There was a significant correlation between Act5C-β-gal activities and z-scores (see also Supplementary Table SI). (C) Relish RNAi blocks the Imd pathway activity in a dose-dependent manner in S2 cells. Att-luc plasmid was used to measure the activity of the Imd pathway and Act5C-β-gal to normalize the results. A total of 5.0 × 10⁵ S2 cells in 500 μl of medium were treated with the total of 2 μg of the indicated dsRNA(s). GFP dsRNA was used as a negative control. Imd pathway was induced with heat-killed E. coli, and luciferase and β-galactosidase activities were measured 24 h afterwards. Data are shown as mean±s.d., N⩾3.

Figure 2

Targeted RNAi of Iap2 or TAB reduces the Imd pathway activity. (A) The effect of targeted Iap2 or TAB RNAi on Imd and Toll pathways. S2 cells were transfected either with Att-luc (the Imd pathway) or with Drs-luc reporter plasmid together with Toll10b (the Toll pathway). Act5C-β-gal reporter was included to control transfection efficiency and viability. dsRNA (2 μg) was added directly to transfection medium. For Imd pathway, the transfected cells were incubated with ecdysone (day 3) prior to the addition of heat-killed E. coli (day 4), after which the Att-luc activity was measured on day 5. The Drs-luc activity (Toll pathway) was measured on day 4 after transfection. Data are shown as mean±s.d., N⩾3. (B) CecA1 induction is inhibited by dsRNA treatments targeting either TAB or Iap2. S2 cells were treated with 20 μg of indicated dsRNA and ecdysone was added 48 h later. After 24 h, the cells were exposed to heat-killed E. coli for 6 h, after which the total RNA was extracted. RT–PCR analysis was used to monitor the expression level of CecA1. Each treatment was carried out in duplicate. (C) Iap2 regulates the expression of antibacterial peptide gene Diptericin in Drosophila adults. Quantitative RT–PCR analysis was performed with total RNA extracts from wild-type (C564/+) flies and flies overexpressing the UAS-Iap2-IR or UAS-BG4-IR females with C564. Diptericin expression was monitored in flies collected 6 h after septic injury with E. carotovora, while Drs was monitored in flies collected 24 h after septic injury with M. luteus. Number of experiments (N) is shown.

Figure 3

Iap2 and TAB are located below Imd in the Imd signaling cascade. (A) Iap2 and TAB are located downstream of Imd. Overexpression of Relish constructs or Imd caused an activation of Att response in S2 cells. Att induction caused by expression of either Rel ΔS29–S45 or wild-type Relish could be blocked by RNAi targeting Rel but not RNAi targeting either imd, TAB or Iap2, indicating that both TAB and Iap2 are located upstream of Relish. Att induction caused by overexpression of Imd was blocked by RNAi targeting either Rel, imd, TAB or Iap2, indicating that both TAB and Iap2 are downstream of Imd. (B) Iap2 lies below TAK1 in the Imd signaling cascade. Overexpression of Iap2 caused a minimal but reproducible induction of Att expression, which was blocked by dsRNAs targeting the known components of the Imd pathway except dsRNA targeting either imd or TAK1. This indicates that Iap2 lies downstream of TAK1 in the Imd signaling pathway. ^*P<0.05, ^**P<0.01 and ^***P<0.001, when the Att expression levels are compared to GFP RNAi controls. Data are shown as mean±s.d. of three to six independent dsRNA treatments.

Figure 4

Neither Iap2 nor TAB is needed for Relish cleavage, but both appear to be involved in Relish nuclear localization. (A) Iap2 and TAB are not required for Relish cleavage. Western blots of protein extracts (25 μg protein/lane) from LPS-induced mbn-2 cells, using α-C Relish antibody. Mbn-2 cells were treated with 20 μg of indicated dsRNA for 72 h and thereafter incubated for 30 min with LPS. (B) Quantification of REL-110 bands from Western blots using Bio-Rad Quantity one software (version 4.5.2.). Mbn-2 or S2 cells were treated with 20 μg of indicated dsRNA for 72 h and thereafter incubated for 30 min with LPS. ^*P=0.03; Dredd RNAi inhibits LPS induced cleavage of Relish compared to GFP RNAi-treated control cells. Iap2 and TAB RNAi treatments have no effect. Data are shown as mean±s.d. of at least three independent experiments. (C) Iap2 and TAB RNAi affect the nuclear localization of Relish. S2 cells were treated with indicated dsRNA for 72 h followed by 10-min exposure to LPS. Thereafter, immunostaining of Relish with α-RHD antibody was performed.

Figure 5

Iap2 and TAB are conserved throughout phylogeny. (A) Comparison of the deduced amino-acid sequences of Iap2 homologs from rat (Rattus norvegicus, AAH62055.1), mouse (Mus musculus, AAC53532.1), human (Homo sapiens, NP_001157.1) and Drosophila melanogaster (AAF58095.1). The three conserved BIR domains are boxed and shown in bold; CARD domain (missing from Drosophila) is shown in lower case and italicized; the RING domain in the C-terminus is boxed and italicized. (B) Alignment of Drosophila TAB with human TAK1-binding protein 3 (AAQ88279.1) and mouse TAK1-binding protein 2 (NP_619608.1). CUE (97–139) and ZnF (765–789) domains are shown boxed and in bold. The low complexity region (157–720) showing no homology with other TABs is not shown. Alignments were created using ClustalW program.

Figure 6

Schematic representation of the Imd signaling pathway in Drosophila S2 cells. The asterisk (^*) represents a physical interaction of PGRP-LC and Imd detected by immunoprecipitation assay (Choe et al, 2005). Two asterisks (^**) represents the interaction of TAB and TAK1 detected in a yeast two-hybrid screen (Giot et al, 2003). As neither TAB, TAK1 nor Iap2 dsRNA treatment affected Relish cleavage, we conclude that these factors affect antimicrobial peptide release via a cleavage independent mechanism. We speculate that the mechanism may involve Relish nuclear transportation, JNK pathway activation or possibly the activation of another yet unidentified transcriptional cofactor for Relish (question marks).