Dual requirement for the EcR/USP nuclear receptor and the dGATAb factor in an ecdysone response in Drosophila melanogaster

Abstract

The EcR/USP nuclear receptor controls Drosophila metamorphosis by activating complex cascades of gene transcription in response to pulses of the steroid hormone ecdysone at the end of larval development. Ecdysone release provides a ubiquitous signal for the activation of the receptor, but a number of its target genes are induced in a tissue- and stage-specific manner. Little is known about the molecular mechanisms involved in this developmental modulation of the EcR/USP-mediated pathway. Fbp1 is a good model of primary ecdysone response gene expressed in the fat body for addressing this question. We show here that the dGATAb factor binds to three target sites flanking an EcR/USP binding site in a 70-bp enhancer that controls the tissue and stage specificity of Fbp1 transcription. We demonstrate that one of these sites and proper expression of dGATAb are required for specific activation of the enhancer in the fat body. In addition, we provide further evidence that EcR/USP plays an essential role as a hormonal timer. Our study provides a striking example of the integration of molecular pathways at the level of a tissue-specific hormone response unit.

FIG. 1

Structures of lacZ reporter transgenes. Fbp1 sequences between −194 and +80, including the −138/−69 enhancer, were fused to the lacZ reporter gene, giving rise to the AE reporter construct. Replacement of the EcR/USP binding site (EBS) with the UAS site gave rise to the AE(UAS) construct. The AEα, AEβ, and AEββ′ constructs are mutated at the indicated GATA sites. The 5UAS-hsp70-lacZ construct consists of five multimerized UAS sites fused to the minimal hsp70 promoter and the lacZ reporter gene.

FIG. 2

Pattern of expression conferred by the wild-type or mutated Fbp1 enhancers on the lacZ reporter transgene. (A) Late-third-instar larval tissues from transgenic lines with the indicated genotypes were dissected and histochemically stained for determination of β-galactosidase activity. (B) Transgenic larvae with the indicated genotypes were synchronized at eclosion and recovered during the third larval instar at the indicated times. β-Galactosidase activity was measured in extracts from whole larvae. Error bars represent the standard error of the mean.

FIG. 3

Structure of the Fbp1 enhancer. The sequence of the Fbp1 enhancer between positions −138 and −69 relative to the Fbp1 transcription start site is in capital letters. Three GATAb binding sites (boxed; GBS1 to GBS3) were found in sequences flanking the pseudopalindromic EcR/USP binding site (horizontal arrows). GBS1 perfectly fits the (A/T)GATA(A/G) consensus sequence for a GBS (18). GBS2 (as read on the lower strand) and GBS3 do not fit this consensus but are efficient binding sites for vertebrate GATA transcription factors (32). Dashed lines indicate the extent of the GATAb footprints (see Fig. 5A) on both DNA strands. The positions and lengths of the competitor oligonucleotides used in gel shift assays are indicated in the lower part of the scheme. Positions of mutations α, β, and β′ are marked by bold lettering in the Fbp1 enhancer, and sequence substitutions are indicated for each mutated competitor, Aα, Bβ, or Fβ′.

FIG. 4

Expression of GATAb in third-instar larvae. (A) RT-PCR analysis. Total expression of mRNAs for GATAb, Fbp1, and the ribosomal protein L17A was analyzed at specific developmental stages by quantitative RT-PCR using specific primers. (B) Distribution of GATAb in late-third-instar tissues. Tissues were dissected from late-third-instar larvae and stained with the #Srp anti-GATAb antibody. Nuclear staining was detected in the gut (g), lymph glands (lg), pericardial cells (pc), gonads (go), and fat body (fb). No staining was detected in the other tissues (not shown). (C) Western blot analysis. GATAb-specific bands were identified with the #Srp antibody in a Western blot analysis by comparing the profiles obtained with fat bodies from w1118 or UAS-srp/hs-GAL4 larvae at 25°C with those obtained with fat bodies from UAS-srp/hs-GAL4 larvae after a 1-h heat shock at 37°C (left). On the right is the temporal profile of GATAb protein expression in isolated fat bodies as detected by Western blotting. The protein quantity loaded in each lane was estimated by detection of myosin. The values on the left are molecular sizes in kilodaltons.

FIG. 5

Bacterially produced or in vitro-translated GATAb binds to the Fbp1 enhancer. (A) An Fbp1 promoter fragment 5′ end labelled on the upper or lower strand as indicated was incubated in the absence (lane F) or in the presence of a GST-GATAb fusion protein in crude bacterial extract (lane crude) or after GST purification (lane GST-purified). Samples were then treated with DNase I and analyzed on a sequencing gel. Sequencing reactions (lanes A, T, G, and C) performed with the free probe were run in parallel in order to locate GATAb-specific footprints (black lines). Nucleotide positions are shown on the right. (B) A gel shift assay was performed with the labelled double-stranded oligonucleotide ADH as a probe in the presence of unprogrammed rabbit reticulocyte lysate (lane 1, UL) or in vitro-translated GATAb protein (lanes 2 to 11). A 200-fold molar excess of competitor oligonucleotides or anti-GATAb antibodies was added as indicated (Fig. 3 shows the positions and sequences of the competitors used).

FIG. 6

GATAb in third-instar fat body nuclear extracts binds to the Fbp1 enhancer. (A) Binding of proteins in a nuclear extract from a late-instar fat body was analyzed by a gel shift assay using the Fbp1 enhancer (−138/−69) as a probe, in the presence or absence of 2B8 antibody and protein A, as indicated. The EcR/USP complex and faster-migrating complexes X and Y, whose identities remain unknown, have been characterized previously by using mass-prepared fat body nuclear extracts (4). Control experiments indicated that protein A alone has no effect on the retardation pattern (not shown). The nuclear extract used in lane 5 was prepared from heat-shocked hs-GAL4/5UAS-srp transgenic third-instar larvae overexpressing GATAb. Complexes a and b, as well as complexes X, Y, and EcR/USP, were markedly decreased or absent from the retardation pattern obtained with this extract, indicating that expression or stability of the factors responsible for their formation was reduced upon heat shock. Both the major retarded band and the upper minor band were supershifted with the #2B8 antibody (not shown). This minor band may correspond to the formation of GATAb dimers under conditions of high GATAb concentrations. UB, unspecific binding. (B) Oligonucleotides A, B, and F were used as radioactive probes in a gel shift assay with fat body nuclear extract. Free probes were run out from the gel.

FIG. 7

GATAb is involved in both the timing and tissue specificity of Fbp1 gene expression. (A) Larvae heterozygous for the AE-lacZ, hs-GAL4, and UAS-srp transgenes were heat shocked for 1 h at 37°C at various developmental stages and allowed to recover for 7 h at 25°C (bottom; +HS). Immunostaining of heat-shocked larvae indicated that the Srp protein was ubiquitously overexpressed (data not shown). Tissues were dissected and histochemically stained for β-galactosidase activity. The times indicated correspond to the dissection of larvae. The images at the top show histochemical staining of control larvae that were not heat shocked (−HS). Abbreviations: P, proventiculus; FB, fat body; SG, salivary glands. (B) β-Galactosidase (β-gal) activity in extracts from AE/hs-GAL4/UAS-srp larvae treated as described for panel A with (black bars) or without (open bars) heat shock treatment. The times indicated at the bottom in hours correspond to times after egg laying at which larvae were dissected. (C) β-Galactosidase activity in extracts from heat-shocked AE[UAS]/hs-GAL4/UAS-srp larvae treated as described for panel A.