Functional characterization of a novel promoter element required for an innate immune response in Drosophila

Abstract

Innate immune reactions are crucial processes of metazoans to protect the organism against overgrowth of faster replicating microorganisms. Drosophila melanogaster is a precious model for genetic and molecular studies of the innate immune system. In response to infection, the concerted action of a battery of antimicrobial peptides ensures efficient killing of the microbes. The induced gene expression relies on translocation of the Drosophila Rel transcription factors Relish, Dif, and Dorsal to the nucleus where they bind to kappaB-like motifs in the promoters of the inducible genes. We have identified another putative promoter element, called region 1 (R1), in a number of antimicrobial peptide genes. Site-directed mutagenesis of the R1 site diminished Cecropin A1 (CecA1) expression in transgenic Drosophila larvae and flies. Infection of flies induced a nuclear R1-binding activity that was unrelated to the kappaB-binding activity in the same extracts. Although the R1 motif was required for Rel protein-mediated CecA1 expression in cotransfection experiments, our data argue against it being a direct target for the Drosophila Rel proteins. We propose that the R1 and kappaB motifs are targets for distinct regulatory complexes that act in concert to promote high levels of antimicrobial peptide gene expression in response to infection.

FIG. 1.

(A) Nested R1 and cecκB sequences found in the 5′ region of Cecropin and Defensin genes. Capital letters indicate conserved nucleotides. Numbers refer to positions relative to the transcriptional start site. (B) The RBA is specific for the R1 sequence. EMSA with nuclear extracts from LPS-treated Drosophila mbn-2 cells in the absence (−) or presence of increasing concentrations of competing oligonucleotides. For sequences, see Materials and Methods.

FIG. 2.

The RBA and κBA display different binding specificities and do not interact with the same motifs. EMSA with nuclear extracts from mbn-2 cells incubated with ³²P-labeled R1 probe (A) or cecκB probe (B) in the absence (−) or presence of increasing concentrations of unlabeled R1 or cecκB oligonucleotides.

FIG. 3.

Schematic representations of the CecA1-lacZ fusion genes. The constructs contain 760 bp of upstream region and 62 bp of untranslated region (open box) of the CecA1 gene fused to a simian virus 40 leader (filled box), providing a translational start site in frame with the Escherichia coli lacZ coding sequence (cross-hatched box). Numbers refer to positions relative to the transcriptional start site. Plasmids A18 and A19 carry mutations in the R1 and κB sequences, respectively (for sequences, see Materials and Methods). The A15 plasmid has a 40-bp deletion (from −105 to −69) including the R1 and κB sites.

FIG. 4.

Both the R1 and the cecκB sites are required for full CecA1-lacZ expression in transgenic larvae and flies. (A to E) Transgenic larvae were not injected (−) or injected (+) with a mixture of E. cloacae and M. luteus and assayed for β-Gal activity in situ. The micrographs show dissected fat body (fb) and salivary glands (sg) from noninjected A10 larvae (WT promoter) (A), bacterium-injected A10 larvae with strong staining in the fat body (B), A15 larvae (R1 and cecκB deletion) (C), A18 larvae (R1 mutant) (D), and A19 larvae (cecκB mutant) (E). (F) Quantitative measurements of β-Gal activity in extracts of noninjected (gray bars) or bacterium-injected (black bars) transgenic flies. The β-Gal activity was measured in extracts from decapitated flies and calculated in relation to the protein concentration. Three independent strains of pA18 (a to c) and pA19 (a to c) were analyzed. The graph presents the results from one representative experiment in a series of four independent experiments.

FIG. 5.

The RBA and κBA constitute independent DNA-binding activities that are induced with different kinetics in response to infection. EMSA with nuclear extracts from flies incubated with ³²P-labeled R1 probe (A) or cecκB probe (B). Flies were injected with a suspension of live bacteria or S. cerevisiae and incubated for the time indicated before making nuclear extracts.

FIG. 6.

Both the R1 and the cecκB sites are necessary for full trans-activation of CecA1 by Relish, Dif, and Dorsal. Relative β-Gal activity in mbn-2 cells after cotransfection of WT and mutant CecA1-lacZ constructs and the expression plasmid pAct-Relish (A), pAct-Dif (B), or pAct-Dorsal (C) is shown. The cells were treated with (+) or without (−) LPS 4 h prior to harvest. The results shown are the mean values of the results from two independent experiments with standard deviations indicated by error bars.

FIG. 7.

Neither Relish, Dif, or Dorsal is a component of RBA in mbn-2 nuclear extracts, whereas Relish, but not Dif or Dorsal, is a component of the κBA in the same extracts. (A) EMSA with nuclear extracts from mbn-2 cells transfected with Relish-FLAG, Dif-V5, or Dorsal-V5 incubated with ³²P-labeled cecκB probe. (B) Cotransfections of Relish-FLAG-Dif-V5 and Relish-FLAG-Dorsal-V5. Mock refers to cells transfected with an empty pAct5C vector, showing that the FLAG (F) and V5 antibodies do not react with any endogenous protein in these cells. −, absent; +, present.

FIG. 8.

Nuclear κBA is induced upon infection in WT flies, but absent in Relish mutant flies, while nuclear RBA is unchanged in the same extracts. WT (Or^r) and homozygous mutant flies (Rel mutant, Dif mutant, and dl mutant) were injected with a suspension of growing E. cloacae (G−) or M. luteus (G+) and incubated for 1 h. Nuclear extracts made from uninfected (−) or infected (G− or G+) flies were subjected to EMSA with ³²P-labeled cecκB probe (A) or R1 probe (B).