A genetic screen targeting the tumor necrosis factor/Eiger signaling pathway: identification of Drosophila TAB2 as a functionally conserved component

Abstract

Signaling by tumor necrosis factors (TNFs) plays a prominent role in mammalian development and disease. To fully understand this complex signaling pathway it is important to identify all regulators and transduction components. A single TNF family member, Eiger, is encoded in the Drosophila genome, offering the possibility of applying genetic approaches for pursuing this goal. Here we present a screen for the isolation of novel genes involved in the TNF/Eiger pathway. On the basis of Eiger's ability to potently activate Jun-N-terminal kinase (JNK) and trigger apoptosis, we used the Drosophila eye to establish an assay for dominant suppressors of this activity. In a large-scale screen the Drosophila homolog of TAB2/3 (dTAB2) was identified as an essential component of the Eiger-JNK pathway. Genetic epistasis and biochemical protein-protein interaction assays assign an adaptor role to dTAB2, linking dTRAF1 to the JNKKK dTAK1, demonstrating a conserved mechanism of TNF signal transduction in mammals and Drosophila. Thus, in contrast to morphogenetic processes, such as dorsal closure of the embryo, in which the JNK pathway is activated by the JNKKK Slipper, Eiger uses the dTAB2-dTAK1 module to induce JNK signaling activity.

Figure 1.

The small eye phenotype caused by eiger (egr) overexpression in the eye provides a sensitized system to screen for new components. (A) GMR-Gal4 UAS-egr/+. Overexpression of egr in the Drosophila compound eye leads to a small eye phenotype due to massive induction of apoptosis. (B) GMR-Gal4/+ control eye. (C) GMR-Gal4 UAS-egr/bsk¹. Removing a single copy of the Drosophila JNK gene bsk dominantly suppresses the small eye phenotype. (D) dTAK1⁴/+ ; GMR-Gal4 UAS-egr/+. dTAK1⁴ is a dominant-negative allele of dTAK1, which fully suppresses the small eye phenotype in a dominant fashion. (E) Schematic representation of the current model of Eiger signaling.

Figure 2.

A dominant modifier screen to identify new components of the Eiger pathway in Drosophila. (A) Crossing scheme. Males carrying a GMR-Gal4 transgene are mutagenized with EMS and crossed to virgins carrying a UAS-egr transgene. Suppressors were rescreened, balanced, and mapped to a chromosome by virtue of visible dominant markers. The second chromosome is marked by the w⁺ (dark red) included in the GMR-Gal4 insertion. The w⁺ of the UAS-egr insertion displays an orange eye color. The third chromosome is marked by the absence of a P[y⁺] insertion present on the non-EMS-treated chromosome. Lethal mutations on the X chromosome escaped our detecting system. Since the X chromosome was not marked, viable mutations, which were obtained only when female suppressors were selected, could not be followed and therefore were obtained only rarely. (B) Numbers of animals that were screened and rescreened and stocks that were established, with the number of alleles identified for positive controls. C–I are in a GMR-Gal4 UAS-egr/+ background. (C) GMR-Gal4 UAS-egr/G56. An example of a suppressor belonging to the “intermediate” class is shown. (D) GMR-Gal4 UAS-egr/G87. An example of a suppressor belonging to the “weak” class is shown. (E) GMR-Gal4 UAS-egr/bsk^G258. G258, one of the new bsk alleles identified in the screen, dominantly suppresses the small eye phenotype. (F) dTAK1^G14/+; GMR-Gal4 UAS-egr/+. G14, the dTAK1 allele identified in the screen, dominantly suppresses the small eye phenotype. (G) dTAK1^G14/+; GMR-Gal4 UAS-egr; P(dTAK1)/+. (H) dTAK1⁴/+; GMR-Gal4 UAS-egr/+; P(dTAK1)/+. (I) GMR-Gal4 UAS-egr/G56; P(dTAK1)/+. A dTAK1 genomic rescue transgene partially reverts the suppression of the small eye phenotype caused by heterozygous loss of dTAK1^G14 or dTAK1⁴ but not the suppression caused by the unrelated suppressor mutation G56. The suppression of the small eye phenotype is not completely reverted by the dTAK1 rescue construct due to the dominant-negative nature of the dTAK1^G14 and dTAK1⁴ alleles. Both carry amino acid substitutions in the kinase domain of dTAK1 (Vidal et al. 2001), which leads to dominant-negative proteins. (J) Molecular lesions found in the 10 novel bsk mutants.

Figure 3.

Identification of dTAB2 (CG7417). (A) The two overlapping deficiencies Df(2R)Exel6069 and Df(2R)P34 narrow down the region of interest to 11 genes. The distal breakpoint of Df(2R)P34 was placed between CG11906 and ribbon on the basis of the fact that Df(2R)P34 and ribbon fail to complement each other (Bradley and Andrew 2001). The CG7417 coding region has a mutation in 39 of the established suppressor stocks. (B) GMR-Gal4 UAS-egr/Df(2R)Exel6069. Df(2R)Exel6069 dominantly suppresses the small eye phenotype whereas Df(2R)P34 does not. (C) Domain architecture and mutations identified in the dTAB2 protein. Red, Gln → Stop. Blue, Arg → Stop. Green, Trp → Stop. Gold, deletions. Black, splice sites mutated. Gray, amino acid substitutions.

Figure 4.

dTAB2 functions upstream of Hep and dTAK1. A–G are in GMR-Gal4 UAS-egr/+ background. (A) GMR-Gal4 UAS-egr/dTAB2^G609. dTAB2 alleles dominantly suppress the small eye phenotype. (B) GMR-Gal4 UAS-egr dTAB2^G609/dTAB2^G71. The small eye phenotype is not completely suppressed in dTAB2 homozygous mutant flies. (C) GMR-Gal4 UAS-egr/dTAB2^G609; tub-dTAB2. (D) GMR-Gal4 UAS-egr/G56; tub-dTAB2. A tub-dTAB2 rescue transgene reverts the suppression of the small eye phenotype caused by heterozygous loss of dTAB2, but not the suppression brought about by the unrelated suppressor mutation G56. (E) GMR-Gal4 UAS-egr/UAS-dTAB2. Overexpression of wild-type dTAB2 has a dominant-negative effect on Eiger signal transduction. (F) dTAK1²/Y; GMR-Gal4 UAS-egr/+. A hemizygous null allele of dTAK1 completely suppresses the small eye phenotype. dTAK1² has an early stop mutation (Vidal et al. 2001). (G) dTAK1²/Y; GMR-Gal4 UAS-egr/+; P(dTAK1)/+. In contrast to dominant-negative alleles of dTAK1 (G14, 4), the suppression of the small eye phenotype brought about by a null allele of dTAK1 can be completely reverted by introducing a dTAK1 genomic rescue transgene. (H) GMR-Gal4 UAS-hep^CA/+. (I) GMR-Gal4 UAS-hep^CA/bsk^G258. (J) GMR-Gal4 UAS-hep^CA/dTAB2^G71. The small eye phenotype caused by overexpression of a constitutive active form of Hep is dominantly suppressed by removing one copy of bsk but not by removing one copy of dTAB2. (K) RNAi against bsk but not against msn or dTAB2 suppresses the dTAK1-FLAG-induced activation of the AP1-luc-reporter in S2 cells. (L) Overexpression of UAS-HA-dTAB2 does not activate the AP1-luc-reporter. Wild-type dTAB2 does not exert a dominant-negative effect on dTAK1-mediated activation of the JNK pathway. Numbers in parentheses indicate amounts of plasmid in micrograms.

Figure 5.

dTAB2 is in a complex with dTRAF1/2 and dTAK1. (A) dTAB2 interacts with dTAK1. S2 cells were transfected with plasmids encoding UAS-HA-dTAB2 and UAS-dTAK1-FLAG together with ptub-GAL4. Samples that immunoprecipitated with anti-FLAG antibody were immunoblotted with anti-HA antibody and vice versa. (B) dTAB2 interacts with dTAK1 most likely through its coiled-coil domain. S2 cells were transfected with plasmids encoding UAS-dTAK1-FLAG and UAS-HA-dTAB2 or UAS-HA-dTAB2-N or UAS-HA-dTAB2-C or UAS-HA-dTAB2-Δcc together with ptub-GAL4. Anti-FLAG immunoprecipitated samples were immunoblotted with anti-HA antibody or vice versa. (C) dTAB2 associates with both dTRAF1 and dTRAF2. S2 cells were transfected with plasmids encoding UAS-FLAG-dTAB2 and UAS-HA-dTRAF1 or UAS-HA-dTRAF2 together with ptub-GAL4. Samples immunoprecipitated with the anti-HA antibody were detected with anti-FLAG antibody. (D) The presumptive Eiger receptor Wengen interacts with both dTRAF1 and dTRAF2. S2 cells were transfected with plasmids encoding Wengen and UAS-HA-dTRAF1 or UAS-HA-dTRAF2 together with ptub-GAL4. Anti-HA antibody immunoprecipitated samples were immunoblotted with an anti-Wengen antibody. (E) dTAB2 links dTRAF1 and dTRAF2 to dTAK1 and forms a triple complex. S2 cells were transfected with plasmids encoding UAS-FLAG-dTAB2, UAS-dTAK1-FLAG, and UAS-HA-dTRAF1 or UAS-HA-dTRAF2 together with ptub-GAL4. Anti-HA antibody-immunoprecipitated samples were immunoblotted with anti-FLAG antibody. (C–E) Asterisks (*) indicate nonspecific bands in the HA-dTRAF2 lysates. (F) dTAB2 is also required for LPS-induced JNK activation. RNAi against dTAK1 or dTAB2 but not against GFP abolishes LPS-mediated phosphorylation of JNK in S2 cells.

Figure 6.

Proposed model for Eiger signaling in Drosophila. Following the binding of Eiger to Wengen a signaling complex consisting of Msn-dTRAF1-dTAB2-dTAK1 is stabilized, which allows the phosphorylation and activation of dTAK1 by Msn. Subsequently dTAK1 activates the core JNKK-JNK module, consisting of the Drosophila homologs Hep and Bsk.