A similar 20E regulatory pathway is utilized in the mosquito Aedes aegypti during egg maturation. An increase in the 20E concentration takes place shortly after blood feeding (Hagedorn, 1989). Yolk protein precursors (YPPs) are synthesized in the fat body, a metabolic tissue functionally analogous to the vertebrate liver, under the control of 20E; they are eventually deposited in the developing oocytes. The major YPP gene, Vitellogenin (Vg), is induced about 10,000-fold at its peak expression 20-24 h post-blood meal (PBM), but declines to a background level by 36 h PBM. Sequence analysis of the 5′ regulatory region of Vg reveals the presence of binding sites for EcR/USP, E74, E75 and BR (Kokoza et al., 2001; Chen et al., 2004). Genes encoding these transcriptional factors are upregulated after blood feeding, albeit with distinct patterns, suggesting the direct participation of both the EcR complex and early gene products in regulation of Vg (Pierceall et al., 1999; Wang et al., 2000, 2002; Sun et al., 2002; Chen et al., 2004). Indeed, we have demonstrated that the EcR/USP dimer binds to the Vg promoter shortly after blood feeding and acts synergistically with E74 to activate the Vg expression (Zhu et al., 2003a; Sun et al., 2005).

Much less is known about the role of br in adult insects. In adult Drosophila, genetic analysis has shown that br is a key control gene during oogenesis (Deng and Bownes, 1997). The different expression patterns of BR isoforms regulate the number of eggs produced by determining whether the cells in each individual egg chamber become committed to apoptosis or to progress through normal development (Terashima and Bownes, 2004). Moreover, Z1 and Z3 overexpression were demonstrated to suppress yp gene expression in egg chambers, while Z2 and Z4 had no effect (Terashima and Bownes, 2004). Our study revealed that the mosquito BR proteins modulate stage-specific expression of Vg, a 20E target gene, during vitellogenesis in the mosquito female fat body, providing additional evidence that br has evolutionarily conserved roles in the insect ecdysone response.

Our functional analyses of br by means of isoform-specific RNAi in adult female mosquitoes demonstrated that BR is required for the robust and precisely timed expression of the Vg gene after a blood meal. Moreover, individual BR isoforms were shown to play distinct roles in such regulation. We observed Z2 to serve as a transcription activator, essential for the optimal Vg expression at 8 and 24 h PBM. Z3 appeared to be functionally similar to Z2, but its effect was limited to 24 h PBM. On the contrary, Z1 and Z4 acted as negative regulators to attenuate the Vg expression at 24 and 36 h, and even at 48 h PBM once the 20E titers in the female mosquitoes had reached the maximum at about 20 h PBM. The results of in vivo studies are in line with effects of BR on the Vg promoter in the cell transfection assays. Importantly, these functions of BR isoforms also appear to be consistent with their expression patterns in the mosquito fat body. Transcription of Z2 exhibited a sharp rise after the onset of vitellogenesis and reached two peaks, one at 8 h and one at 24 h PBM. Conversely, the Z3 transcripts remained at a relatively low level throughout the entire vitellogenic period, increasingly slightly at 24 h PBM. mRNA levels of both Z1 and Z4 were relatively low during the early vitellogenic stage, but started to rise after 16 h PBM and peaked at 24 h (Chen et al., 2004). In summary, individual BR isoforms participated in the regulation of Vg in response to the 20E pulse after blood ingestion through a coordinated and well-orchestrated effort.

It has been shown in D. melanogaster that different BR isoforms can have opposite regulatory functions on the same gene. For example, the larval fat-body-specific expression of the Fbp-1 gene is activated by the Z3 isoform, but repressed by the Z2 protein (Mugat et al., 2000). Interestingly, functions of an individual isoform are promoter dependent. While Z3 is required for activation of the Fbp-1 gene, it acts as a repressor for the transcription of a salivary-gland-specific gene Sgs4 (Bayer et al., 1996). A similar phenomenon is also observed in the mosquito 20E response. Our preliminary study suggested that both Z1 and Z4 isoforms are required to upregulate expression of the fat-body-specific lipophorin receptor gene, which is activated at 30-36 h PBM when the 20E titers in the female adults drop to the basal level (Jun and Raikhel, unpublished result). Therefore, it appears that mere existence of the BR proteins cannot explain their diverse properties in terms of transcription regulation. The activities of the BR isoforms probably rely on their binding sequences in the context of native target promoters and/or proteins associated with BR at different scenarios.

The presence of zinc-finger DNA binding domains in BR proteins suggested that they work as transcription regulators and interact directly with cis-acting regulatory elements in target genes. Indeed, there are numerous binding sites for BR isoforms in the mosquito Vg promoter. EMSA assays confirmed that BR can bind these sites in vitro. When binding sites for Z1 and Z2 were mutagenized in the pVg-Luc reporter construct, the BR proteins lost their influences on the Vg promoter in cell transfection assays. Similarly, mutations in the Z4 zinc-finger domain also relieved repression of the Vg promoter by this isoform. Thus, these data collectively imply that the BR proteins act through their binding sites in the Vg promoter, and that direct DNA-protein interactions are indispensable for appropriate activation of this ecdysone effector gene. It is worth noting that the upregulation of usp gene after blood feeding is also modulated by the BR isoforms. This observation not only indicates that BR has profound effects on the vitellogenic 20E response, but also raises the possibility that Vg is under direct and indirect control of BR.

Intriguingly, BR by itself had no marked effect on the Vg promoter in the cell transfection assay. Instead, the BR isoforms either repressed or enhanced transcription of the Vg promoter activated by the EcR complex (Chen et al., 2004). The functional interaction between BR and EcR/USP may stem from their physical association that we found in vitro. Proteins harboring the BTB domain have been shown to play crucial roles in specifying the overall transcriptional activity of nuclear receptors in response to their cognate ligands. The human promyelocytic leukemia zinc finger (PLZF) protein is a transcriptional repressor belonging to the BTB family. PLZF interacts directly with the Vitamin D3 receptor (VDR) and represses VDR-dependent gene activation (Ward et al., 2001). The ligand-independent physical contact involves the DNA-binding domain of VDR and the BTB domain of PLZF. Likewise, PLZF has been found to bind the retinoic acid receptor (RAR) and to interfere with the transcriptional activity of the RAR-RXR heterodimer (Martin et al., 2003). The zinc finger domain of PLZF and the ligand binding domain of RAR are sufficient for the interaction of these two proteins, while participation of RXR is not required. Unlike BR, DNA binding of PLZF in both cases is not required for its repressive functions. Interestingly, a chromosomal translocation joining the PLZF gene to RARα is associated with a rare syndrome of acute promyelocytic leukemia (Chen et al., 1994). The PLZF-RARα binds to retinoic acid response element as a heterodimeric partner of RXR, interfering with RARα functions by exerting a dominant negative effect (Licht et al., 1996). The functional interactions between members of the nuclear receptor and BTB families thus represent a regulatory mechanism well-conserved in vertebrates and invertebrates.