Highly Flexible Ligand Binding Pocket of Ecdysone Receptor A SINGLE AMINO ACID CHANGE LEADS TO DISCRIMINATION BETWEEN TWO GROUPS OF NONSTEROIDAL ECDYSONE AGONISTS*

The insect steroid hormone 20-hydroxyecdysone works through a ligand-activated nuclear receptor, the ecdysone receptor (EcR), which plays critical roles in insect development and reproduction. The EcR has been exploited to develop insecticides to control pests and gene switches for gene regulation. Recently reported crystal structures of the EcR protein show different but partially overlapping binding cavities for ecdysteroid (ECD) and diacylhydrazine (DAH) ligands, providing an explanation for the differential activity of DAH ligands in insects. 1-Aroyl-4- (arylamino)-1,2,3,4-tetrahydro-quinoline (THQ) ligands were recently discovered as ecdysone agonists. Mutagenesis of the EcR (from Choristoneura fumiferana , CfEcR) ligand binding domain followed by screening in a reporter assay led to the identification of CfEcR mutants, which responded well to THQ ligands but poorly to both ECD and DAH ligands. These mutants were further improved by introducing a second mutation, A110P, which was previously reported

The insect steroid hormone, 20-hydroxyecdysone (20E) 1 plays critical roles in insect development and reproduction (1). The effects of 20E are mediated through a ligand-activated nuclear receptor, the Ecdysone receptor (EcR) (2). Like many nuclear receptors EcR has five modular domains, the A/B (ligand-independent transactivation), C (DNA binding and heterodimerization), D (hinge, heterodimerization), E (ligand binding, heterodimerization, and transactivation), and F (transactivation) (2). The EcR heterodimerizes with another nuclear receptor, ultraspiracle (USP), a homologue of vertebrate RXR (3)(4)(5). The EcR/USP heterodimers are able to bind to ecdysone response elements (EcRE) present in the promoters of 20E-responsive genes. The binding of ligands to EcR and the subsequent activation of 20E-responsive genes are greatly stimulated by the formation of a functional heterodimer. The EcR gene encodes three isoforms, EcR-A, EcR-B1, and EcR-B2, that contribute to the tissue specificity of 20E response (6). In addition to 20E, several phytoecdysteroids such as ponasterone (PonA) and muristerone A bind to EcR with high affinity.
The crystal structure of USP has been elucidated by two groups (7,8). The structure of USP is similar to its mammalian homologue RXR, except that USP structures showed a long H1-H3 loop and an insert between H5 and H6. These structures appear to lock USP in an inactive conformation by displacing helix 12 from the agonist conformation. In addition, in both crystal structures USP has a large hydrophobic cavity that contains phospholipid ligands (7,8). The crystal structures of the EcR protein were determined recently and showed different but partially overlapping binding cavities for ecdysteroid (ECD) and diacylhydrazine (DAH) ligands, providing an explanation for the differential activity of DAH ligands in insects (9). Similar to vertebrate nuclear receptors, the EcR is presumed to recruit coregulators such as corepressors and coactivators in the process of transcriptional activation of 20E-responsive genes. EcR-specific coactivator Taiman protein was shown to colocalize with EcR and USP in the border cells of Drosophila melanogaster and to augment transcriptional activity by EcR in cultured cells (10). An EcR-interacting protein, SMRTER, which is functionally similar to vertebrate corepressors of nuclear receptors, SMRT and N-CoR, was identified in D. melanogaster (11).
The diacylhydrazine class of synthetic nonsteroidal ligands discovered by the Rohm & Haas Co. as potential insecticides are capable of binding to EcR and activating 20E-responsive genes (12). Some of these compounds bind to EcR with high affinity and initiate a premature molt that leads to mortality of larvae in lepidopteran and coleopteran insects (13). These compounds are not toxic to other insects and appear to be harmless to vertebrates (13). The differences in the activity of DAH ligands in insects belonging to various orders has been attributed to the differential uptake of compounds in to the cells and differences in EcR sequences (12,14). Homology models of EcR predicted that ECD and DAH ligands will occupy different but substantially overlapping regions of a ligand binding cavity (15)(16)(17). Recent determination of crystal structures of EcR showed different but only partially overlapping ligand binding cavities for ECD and DAH ligands (9).
EcR has been used in several inducible gene regulation systems or gene switches to control transgene expression in mammalian cells, transgenic animals, and plants. Several EcRbased gene switches that function through steroid and nonsteroidal ligands have been developed (22)(23)(24). Improvements that increase the sensitivity and magnitude of induction of cognate genes have been made to the EcR-based gene switch. An improved EcR-based gene switch capable of lower background, increased sensitivity, and higher magnitudes of induction was developed (25).
Through homology modeling, ligand docking, and mutagenesis followed by screening we have identified two mutants (V128F and V128Y) of EcR from Choristoneura fumiferana (CfEcR) that responded well to THQ but not to ECD or DAH ligands. These mutants were further improved by introducing a second mutation, A110P, which was previously reported to cause ECD insensitivity (17). Other tests showed that the V128Y/A110P double mutant responded to THQ ligand but not to ECD or DAH ligands in vivo in mice. The V128F/A110P mutant recruited the EcR coactivator, Taiman, in the presence of THQ ligand but not ECD or DAH ligands, and this mutant EcR induced the expression of the 20E-responsive hormone receptor 3 (DHR3) gene in D. melanogaster L57 cells only in the presence of THQ ligand but not ECD or DAH ligands. The WTCfEcR and its V128F/A110P mutant were used to demonstrate simultaneous regulation of two reporter genes using THQ and DAH ligands.

Site-directed Mutagenesis
Site-directed mutagenesis was carried out using the QuikChange® site-directed mutagenesis kit (Stratagene, La Jolla, CA). Mutations were verified by sequencing.

Reporter Gene Assays
3T3 cells were transfected with G:CfE(DEF), VP:Hs-LmR(EF), and pFRLUC plasmids in 96-well plates using SuperFect lipid reagent (Qiagen), and ligands were administered 4 h post-transfection. The cells were harvested at 40 h after transfection and analyzed for reporter gene activity. L57 cells were transfected using LipofectAMINE 2000 reagent (Invitrogen). Luciferase was measured using the Dual-luciferase TM reporter assay system from Promega Corp. (Madison, WI). ␤-Galactosidase was measured using the Gal-Screen® system from Applied Biosystems (Bresford, MA). All the transactivation assays were performed in triplicate and were repeated three times.

Coactivator Interaction Assay
3T3 cells were transfected with G:CfE(DEF) or its mutants VP:D-mTaiman and pFRLUC. The transfected cells were exposed to 1 M ligands, and the reporter activity was quantified at 48 h after transfection.

Testing the CfEcR V128Y/A110P Mutant in Vivo
The gene switch plasmids were electroporated into the quadriceps of C57BL/6 mice. The animals were anesthetized and shaved, and DNA vector was injected into the muscle. Electrode conductivity gel was applied, and an electrode (1 ϫ 1 cm) was placed on the hind leg and electroporated with 200 V/cm 8 times for 20 ms/pulse at 1-s time intervals. The transverse electrical field direction was reversed after the animals received half of the pulses. Animals were treated with 0 -4-mg eq (0 -10.5 mol/mice) of RG-120499, PonA, or RG-102240 in 50 l of Me 2 SO by intraperitoneal injection at 3 days after electroporation. The SEAP activity in mouse sera was evaluated at 3 days after ligand administration using Great Escape Chemiluminescence kit (Clontech).

Activation of a Reporter Gene and an Endogenous DHR3 Gene in Insect Cells
The L57 cells from Drosophila were grown in 25-cm 2 cell culture flasks using HyQ CCM3 medium (HyClone, Logan, UT). For transfections into L57 cells, cells were seeded into 5-ml culture flasks at a density of 100,000-cells/ml medium. The following day, 3 g of VP: CfE(CDEF) or its V128F/A110P mutant DNA was transfected with LipofectAMINE 2000 reagent. At 24 h after transfection, a 1 M concentration of 20E or RG-102240 or RG-120768 was added to the culture medium. At 6 h after addition of ligands, the cells were harvested, quick-frozen in liquid nitrogen, and stored at Ϫ80°C.
Total RNA was isolated from L57 cells using Trizol reagent (Invitrogen) and treated with DNase I (Ambion, Austin, TX) to remove contaminating genomic DNA. RNA was reverse-transcribed in a total volume of 20 l with iScript cDNA synthesis kit (Bio-Rad). Aliquots (2.5 l) of cDNA were added to a 20-l reaction mixture containing 10 l of iQSYBR Green supermix (Bio-Rad), 0.25 M primers. The real-time PCR was performed using DHR3 primers (5Ј-cgccggcgggacaaaacaact-3Ј and 5Ј-actgcgcggctcgtaggtggtg-3Ј) and a MyiQ single color real-time PCR detection system (Bio-Rad). The PCR settings were initial incubation at 95°C for 3 min followed by 45 cycles of denaturation at 95°C for 10s, annealing at 60°C for 30s, and elongation at 72°C for 60s.

THQ Ligand Design and Synthesis-
The synthesis of some THQs by benzoylation of 1 (Fig. 1A) with various monosubstituted benzoyl chlorides, and their activity as agonists of A. aegypti EcR has been reported previously (20). Synthesis of additional analogs bearing one or two substituents on the benzoyl ring led to the discovery of the new compounds RG-120499, RG-120768, and RG-120929, which exhibited improved potency against A. aegypti EcR. These compounds were selected as candidates for screening against CfEcR mutants. In transactivation assays all three compounds induced the reporter activity through V128F or V128Y mutant but not wild type versions of CfEcR. RG-120499 and RG-120929 were synthesized first and used in the initial experiments showed in Figs. 2, A and B, and 3. RG-120768 was synthesized latter and was found to be better than RG-120499 and RG-120929 in inducing reporter activity through V128F or V128Y mutants. Therefore, subsequent experiments reported in Figs. 2C, 4, and 5 used this compound.
Identification of THQ Ligand-specific Val-128 Mutants-Homology modeling of the CfEcR ligand binding domain (LBD) followed by docking of THQ ligand into the binding pocket indicated that valine at position 128 (Val-128) may play an important role in differentiating the binding of THQ as compared with the DAH and ECD chemotypes. We used random primers and site-directed mutagenesis to produce a library of G:CfE(DEF) mutants where the three nucleotides coding for Val-128 have been randomly changed. We screened this library of CfEcR mutants and identified two mutants that responded well to THQ ligands, RG-120499 and RG-120929, but not to the DAH ligand, RG-102240, and the ECD, PonA. Sequencing of the two identified mutants showed that the valine at the128 position was changed to phenylalanine (V128F) and tyrosine (V128Y). This valine residue at position 128 is present in EcRs from lepidopteran insects but not in EcRs from the other insects (Fig. 1B). The two mutants and WT EcRs were tested in the 3T3 cell assay. As shown in Fig. 2A, THQ ligands RG-120499 and RG-120929 did not induce reporter activity with WT CfEcR, whereas the DAH RG-102240 and the ECD PonA did ( Fig. 2A). Both RG-120499 and RG-120929 induced reporter activity through V128F and V128Y mutant CfEcRs. RG-102240 and PonA did not induce detectable levels of reporter activity with V128F and V128Y mutant CfEcRs (Fig. 2A). These results from analysis of two mutants and WT CfECR with four ligands showed that the two CfEcR mutants, V128F and V128Y, respond well to THQ ligands but not to ECD and DAH ligands, and they have potential uses in simultaneous regulation of multiple genes.
Our previous work showed that the A110P mutation in CfEcR LBD reduced the ECD response, although its DAH activity was unaffected (17). To determine the affect of this A110P mutation on the THQ-responsive Val-128 CfEcR mutant, we introduced the A110P mutation into the V120F and tested its activity in the 3T3 cell reporter assay. As shown in Fig. 2B, both RG-102240 and PonA but not RG-120929 induced reporter activity with WTCfEcR. On the other hand, only RG-120929 but not RG-102240 or PonA induced reporter activity through the V128F/A110P mutant CfEcR (Fig. 2B). These results demonstrate that the V128F/A110P mutant of CfEcR responds well to THQ ligand and that it is insensitive to both ECD and DAH ligands.  (43); Ct, Chironomus tentans, (44); Tm, Tenebrio molitor, (45); Am, Amblyomma americanum, (46); Up, Uca pugilator, (47). Amino acid residues that match with those in CfEcR are shaded.
The mutagenesis, screening, and confirmation of EcR mutations were performed using the G:CfE(DEF) construct. To ascertain that the V128F response to THQ ligands is not an artifact of the GAL4 fusion or the truncation of EcR, the V128F mutation was introduced into the full-length CfEcR receptor. This mutant was tested in 3T3 cells using a reporter regulated by ecdysone response element. A 1 M concentration of PonA and RG-102240 but not RG-120768 (THQ ligand) induced re- porter activity through WTCfEcR(A-F) (Fig. 2C). On the other hand, only RG-120768, but not PonA or RG-102240 induced reporter activity with the V128F/A110P mutant of CfEcR(A-F) (Fig. 2C).
In Vivo Testing of V128Y/A110P Mutant-The V128Y/A110P EcR mutant was tested in a C57BL/6 mouse model to determine whether the mutant behaves in the same fashion in vivo as that seen in cultured mammalian cells. As shown in Fig. 3A, RG-102240 induced dose-dependent reporter gene activity with WTCfEcR. The PonA-induced activity was relatively weak (Fig.  3B) and THQs, failed to induce significant reporter gene activity over the background with WTCfEcR (Fig. 3C). These observations are consistent with the results observed in cultured cells. The V128Y/A110P mutant on the other hand failed to induce reporter gene activity in the presence of RG-102240 or PonA ( Fig. 3D and E) but induced dose-dependent reporter gene activity when a THQ compound, RG-120499, was administered to the mice (Fig. 3F). These results confirm our in vitro data and establish that the V128Y/A110P mutant EcR as a genuine THQ-sensitive mutant possessing orthogonal properties toward THQ, DAH, and ECD ligands. (10) to the THQ-specific mutant was explored using a coactivator interaction assay. We hypothesized that WTEcR or its mutant would recruit Taiman only in the presence of complementary ligands that bind with high affinity. G:CfE(DEF) or its V128F/A110P mutant, VP:DmTaiman, and pFRLuc were transfected into 3T3 cells. The transfected cells were exposed to PonA, RG-102240, or RG-120768. As shown in Fig. 4, WTEcR and Taiman interaction resulted in an increase in reporter activity in the presence of PonA or RG-102240 but not RG-120768. In contrast, V128F/A110P and Taiman interacted only in the presence of RG-120768 but not PonA or RG-102240. These results suggest that PonA and RG-102240 bind to WTCfEcR with high affinity and help EcR in recruiting the coactivator, Taiman. On the other hand, only THQ ligand, RG-120768, but not PonA or RG-102240 binds to V128F/A110P mutant EcR with high affinity and help in recruiting Taiman.

Ligand-dependent Interaction of EcR Coactivator Taiman with WTEcR and Mutants-The recruitment of insect EcR coactivator Taiman
Activation of a Reporter Gene and an Endogenous DHR3 Gene in Insect Cells-To determine the ligand specificity of the V128F/A110P mutant EcR in insect cells, we performed reporter assays and studied the induction of a ␤-galactosidase reporter regulated by EcRE as well as 20E-responsive Drosophila DHR3 (an ecdysone induced delayed-early gene directly regulated by 20E (26)) in EcR-deficient L57 cells. L57 cells were developed from Kc cells by parahomologous gene targeting to inactivate EcRB1 and EcRB2 isoforms (27). These cells retain only 10% of the EcR binding capability due to the presence of EcRA isoform. However, complete 20E response can be restored by transfecting with a DNA construct where EcR is expressed under a strong promoter (28). VP:CfE(CDEF) or its mutant, V128F/A110P, and EcRE reporter, pMK43.2 (where ␤-galactosidase is expressed under the control of 6XEcRE (2)), were transfected into L57 cells, and the transfected cells were exposed to PonA, RG-102240, or RG-120768. PonA and RG-102240 but not RG-120768 induced reporter activity through WTCfEcR (Fig. 5A). Only RG-120768, but not PonA or RG-102240, induced reporter activity with V128F/A110P mutant EcR.
To determine the specificity of the V128F/A110P mutant in regulating the expression of endogenous genes containing EcRE in their promoter, we transfected VP:CfE(CDEF) or its mutant into L57 cells, and the transfected cells were exposed to 1 M 20E, RG-102240, or RG-120768. At 24 h after the addition of ligands, the cells were harvested, and the DHR3 mRNA was quantified using quantitative real-time PCR. As shown in Fig.  5B, DHR3 mRNA was induced by 20E and RG-102240 but not by RG-120768 when WTEcR construct was transfected into the cells. On the other hand only RG-120768 but not 20E or RG-102240 induced DHR3 mRNA when the V128F/A110P mutant EcR was transfected into the cells. These data demonstrate the specificity of the V128F/A110P mutant to THQ ligand in insect cells through regulation of EcRE-driven reporter and 20E-responsive endogenous gene.
WTCfEcR and Its V128F/A110P Mutant Can Regulate Two Genes Simultaneously-As shown above, WTCfEcR and its V128F/A110P mutant showed orthogonal sensitivity to DAH for 24 h. The cells were harvested, and total RNA was isolated and quantitative real-time PCR was performed using DHR3 gene-specific primers. RNA isolated from L57 cells transfected with WTVP:CfE(CDEF) and exposed to 20E was used as a standard, and relative levels of DHR3 mRNA were calculated. The values presented are the mean Ϯ S.D. (n ϭ 3). C, development of EcR-based dual switch for simultaneous regulation of two genes in mammalian cells. 3T3 cells were transfected with plasmids expressing L:CfE-(DEF) (LexA:CfEcR(DEF)), G:CfE(DEF)-V128F/A110P mutant, VP:Hs-LmR(EF), pFRLacZ, and p8OPFRLUC. The cells were exposed to either THQ (RG-120768), DAH (RG-102240), or both ligands for 48 h. The cells were harvested, and the activities of luciferase and ␤-galactosidase were determined. The values presented are the mean Ϯ S.D. (n ϭ 4).
(e.g. RG-102240) and THQ (e.g. RG-120768) ligands. To determine whether these orthogonal LBDs and ligand pairs can be used for simultaneous regulation of two genes, we transfected G:CfE(DEF)V128F/A110P mutant, L:CfE(DEF) (LexA: CfEcR(DEF)), VP:Hs-LmR(EF), pFRLacZ, and p8OPLUC into 3T3 cells, and the transfected cells were exposed to RG-120768, RG-102240, or both ligands simultaneously. As shown in Fig.  5C, cells treated with RG-120768 showed a substantially preferential increase in ␤-galactosidase relative to luciferase activity. On the other hand, cells treated with RG-102240 showed a preferential increase in luciferase relative to ␤-galactosidase activity. However, cells treated with both compounds showed an increase in both luciferase and ␤-galactosidase activities, suggesting that it is possible to use WTCfEcR and its V128F/ A110P mutant for simultaneous regulation of two genes. DISCUSSION The EcR is being exploited to develop insecticides to control pests as well as to develop gene switches for regulation of transgenes in agriculture and medicine. The DAH ecdysone agonist insecticides are active in moths and beetles but not in other insects (12). Similarly, in mammalian cells, EcR-based gene switches that used EcR from D. melanogaster worked better with the ecdysteroid PonA as opposed to the DAH ecdysone agonists such as methoxyfenozide and GS TM E (gene switch ecdysone) (22,25). In contrast, the EcR-based gene switches that used EcR from Bombyx mori or C. fumiferana worked better with methoxyfenozide or GS TM E than with PonA (24,25). Recent crystal structures of EcR from H. virescens that bound to ECD and DAH ligands showed that these two classes of ligands have distinct but overlapping regions of the ligand binding pocket (9).
Tetrahydroquinoline compounds were originally identified in cell assays using EcR from D. melanogaster or EcR from H. virescens (21). These ligands showed only weak insecticidal activity against H. virescens (21). In mammalian cell transactivation assays that we developed, the THQ ligands showed very weak activity with CfEcR 2 but showed a stronger activity with EcR from A. aegypti (18). Because WTCfEcR responds to DAH ligands very well but poorly to THQ ligands, we hypothesized that a CfEcR mutant that responds well to THQ ligands but poorly to DAH ligands would be useful for development of multiplexed gene switches.
Precise regulation of transgene expression is very important for the applications in functional genomics, proteomics, protein expression, and gene therapy. Several gene switch systems are available for regulation of a single gene in mammalian cells and transgenic animals (29). For the purposes of functional genomics and proteomics, where attempts are being made to understand the function of several proteins in either a single cascade or multiple cascades, a system needs to be developed to regulate more than one gene. In the process of developing such a system we attempted to generate orthogonal mutants of EcRs, where a mutant responds to ligands belonging to one chemotype while being non-responsive to ligands from other chemotypes. This would allow different transgenes to be regulated by different ligands without interference from each other.
We have previously reported identification of an EcR mutant that is non-responsive to ECD but that is sensitive to DAH chemotype (17). Here, we report the identification of another mutant of CfEcR that fails to respond to DAH and ECD ligands but is sensitive to the THQ class of ligands. We constructed several CfEcR homology models, docked the THQ, ECD, and DAH ligands, and predicted that the ligands from these three chemotypes occupy different but overlapping binding pockets.
Val-128 was identified as a potential amino acid residue that plays a critical role in discriminating THQ, DAH, or ECD ligand binding. Recently reported EcR from H. virescens crystal structures showed that ECD and DAH ligands occupy distinct but overlapping binding cavities and that Val-128 (Val-133 in EcR from H. virescens) is a proximal residue to both PonA and the DAH BYI06830 (9). The models have led to a mutation (V128F) that showed decreased ECD and DAH sensitivity and increased THQ sensitivity. In view of the proximal position of Val-128 to both PonA and the DAH BYI06830 in the EcR from H. virescens crystal structures and the increased excluding volume of the phenylalanine-for-valine substitution, the inability of the ECD and DAH to bind to the V128F mutant is readily rationalized. An explanation for the THQ sensitivity is, on the other hand, more speculative. Perhaps Phe-128 enables favorable -stacking interactions. Entropic factors may improve due to a more confining cavity that nonetheless remains in a THQcomplementary shape. Alternatively, more significant cavity remodeling may occur as a result of mutual ligand-proteininduced fitting. Whatever factors may be involved in causing the reversed THQ-DAH/ECD discrimination, it is clear that the EcR ligand binding pocket can have remarkable plasticity and sensitivity to single point mutations. This is not without precedence since point mutations have led to changes in ligand specificity for nuclear receptors including EcR (30 -33).
Several lines of evidence support the conclusion that the V128F/A110P mutant is THQ ligand-specific. In both mammalian and insect cell transactivation assays, THQ ligands but not ECD or DAH ligands induced reporter gene activity with the V128F/A110P mutant. Similar results were observed when full-length CfEcR V128F/A110P mutant and EcRE-driven reporter constructs were used in the mammalian cell transactivation assay. In vivo in mice only the THQ ligand was able induce reporter activity with V128Y/A110P EcR. In EcR-deficient L57 cells only the THQ ligand through the V128F/A110P mutant EcR induced the expression of mRNA for 20E-responsive DHR3 gene. Taken together these data clearly show the THQ ligand specificity of the V128F/A110P mutant EcR.
Because labeled ECD ([ 3 H]PonA) and DAH ([ 3 H]RH-2485) ligands only poorly bind to the V128F mutant 2 and because the radiolabeled THQ ligands were not available, we were not able to test ligand binding specificity of this mutant. Instead, we used a coactivator interaction assay as an indirect determinant of ligand binding to V128F/A110P mutant EcR. This is similar to coactivator-dependent receptor ligand assay (developed by Krey et al. (34) except that the interaction was assayed in cells. They used coactivator-dependent receptor ligand assay to identify ligands for peroxisome proliferator-activated receptors. Coactivator-dependent receptor ligand assay was also used to identify ligands that interact with pregnane X receptor (35). Using coactivator interaction assay we showed that the V128F/ A110P mutant recruited EcR coactivator, Taiman, only in the presence of THQ ligand but not in the presence of ECD or DAH ligands, showing ligand binding specificity of V128F/A110P mutant EcR.
The identification of this mutant permits the development of a new generation of gene switches for simultaneous regulation of multiple genes. We tested the possibility of regulating two genes simultaneously using WTCfEcR, the V128F/A110P mutant, and their respective ligands. The results clearly showed that it is possible to regulate two genes using these two receptors (Fig. 5C). The expression levels of luciferase and ␤-galactosidase were lower in cells where both genes were induced when compared with the levels where either luciferase or ␤-galactosidase was induced. This may be due to the competition of both receptors for heterodimeric partner or coactivators. Further fine-tuning is nec-essary for designing an efficient dual switch that functions through WTCfEcR and its V128F/A110P mutant.
Several versions of gene switches are available for regulation of single genes. Estrogen receptor has been exploited for development of gene switches for regulation of two genes (36,37). Kramer et al. (38) proposed a multilevel transgene control system using tetracycline-, streptogramin-, and macrolide-repressible gene regulation systems. Because of the availability of natural variation in both EcRs and their ligands and remarkable plasticity shown by EcR LBD (9,17,33), EcR-based switches are very attractive for development of multiple gene regulation systems. Identification of ECD-insensitive A110P (17) mutant and THQ-sensitive V128F mutant (this study) are two critical steps in this direction. Work is in progress to identify an EcR mutant that is capable of responding to ECD but not to DAH or THQ ligands. Successful completion of these studies will provide three EcR LBD-ligand pairs for simultaneous regulation of three genes.
It is remarkable that changing a single amino acid in the ligand binding domain of EcR resulted in complete discrimination in the activity of nonsteroidal ligands belonging to two different chemotypes. These studies along with the published reports (9,17,33) show the extreme flexibility and adaptability in the ligand binding pocket of EcRs. The studies also underscore the expectation that different insect pests should be able to discriminate among compounds from different as yet undiscovered chemotypes, thereby providing an avenue for development of target-specific insecticides.