The translational inhibitor Thor (Drosophila 4E-BP) was first identified in a screen for immune-compromised phenotypes (Bernal et al, 2000). 4E-BP inhibits the initiation factor eIF4E, which is required for the recognition of the m7G cap of transcripts to load ribosomes for translation. Most cellular transcripts are translated via a cap-dependent mechanism requiring eIF4E. However, subsets of transcripts can be translated cap-independently via special structures in their 5’UTR. While 4E-BP has drawn interest due to its role in lifespan extension, how it is induced during infection and its functional relevance remain unexplained. In this work, we show that 4E-BP is induced by the integrated stress response pathway and is required for the efficient translation of anti-microbial peptides (AMPs) involved in pathogen clearance.
Following pathogen recognition during infection, a signaling cascade is activated culminating in the transcriptional induction and synthesis of AMPs. AMP induction in response to the pathogen Ecc15 was unaltered in 4E-BP-/- suggesting that AMP translation may be affected. To test if AMPs are translated cap-independently in the presence of 4E-BP, we designed a bicistronic assay with the 5’UTR of AMPs sandwiched between GFP and dsRed. Thus dsRed will only be synthesized if the 5’UTR permits cap-independent translation. The 5’UTR of the AMP drosomycin not only scored positively in this assay, but enhanced the synthesis of dsRed when the cells were challenged with Ecc15. These data suggest that while AMPs may be translated cap-independently, their efficient translation during infection requires the inhibition of cap-dependent translation by 4E-BP.
We find that the transcriptional induction of 4E-BP during infection is mediated by the integrated stress response transcription factor, ATF4. ATF4 mutants are immune-compromised similar to 4E-BP-/- and analysis of the 4E-BP genomic region reveals several ATF4 binding sites in an intronic element (4E-BPi). A dsRed reporter driven by 4E-BPi responds specifically to ATF4 but not the canonical 4E-BP transcription factor, FOXO. Mutating the putative ATF4 binding sites in 4E-BPi resulted in loss of reporter activity. Reporter expression is elevated in the gut of larvae orally infected with Ecc15 in accordance with our other data. ATF4 can be activated by one of two upstream kinases, PERK, which responds to stress in the endoplasmic reticulum and GCN2, which responds to amino acid deprivation. Knockdown of GCN2 but not PERK resulted in reduced ability to clear Ecc15 in larvae. These results indicate a role for GCN2/ATF4/4E-BP in mounting an immune response.