Improvement of ecdysone receptor gene switch for applications in plants: Locusta migratoria retinoid X receptor (LmRXR) mutagenesis and optimization of translation start site

Abstract

Gene switches have potential applications for the regulation of transgene expression in plants and animals. Recently, we have developed a two-hybrid ecdysone receptor (EcR) gene switch using chimera 9 [CH9, a chimera between helices 1–8 of Homo sapiens retinoid X receptor (HsRXR) and helices 9–12 of Locusta migratoria RXR (LmRXR)] as a partner for Choristoneura fumiferana EcR (CfEcR). As CH9 includes a region of human RXR, public acceptance of this gene switch for use in genetically modified crops may be an issue. The current studies were conducted to identify an LmRXR mutant that could replace CH9 as a partner for CfEcR. The amino acid identity between LmRXR and HsRXR is fairly high. However, there are a few amino acid residues that are different between these two proteins. LmRXR mutants were produced by changing the amino acids in the helices 1–8 that are different between LmRXR and HsRXR to HsRXR residues. Screening of these mutants in tobacco protoplasts identified a triple mutant, A62S:T81H:V123I (SHILmRXR), that performed as well as CH9. The performance of the EcR gene switch was further improved by optimizing the translational start site (Kozak sequence, AACAATGG) of the transgene. The EcR gene switch containing SHILmRXR and the modified translation start site supported very low background activity in the absence of a ligand and a higher induced activity in the presence of a ligand in tobacco protoplasts, as well as Arabidopsis thaliana transgenic plants. At 16–80 nm methoxyfenozide, the induction of luciferase activity was better than that observed with the CfEcR:CH9 switch.

Abbreviations

AD

activation domain

CfEcR

Choristoneura fumiferana ecdysone receptor

CH9

chimera 9

DBD

DNA-binding domain

EcR

ecdysone receptor

FMV

figwort mosaic virus

HsRXR

Homo sapiens retinoid X receptor

LBD

ligand-binding domain

LmRXR

Locusta migratoria retinoid X receptor

MMV

Mirabilis mosaic virus

Introduction

The conditional regulation of transgene expression by the interaction of a chemical inducer with a designed receptor or transcription factor is a powerful tool for the regulation of transgenes in genetically modified crops, as well as for plant biotechnological applications [1–4]. Several chemically inducible gene regulation systems or gene switches that respond to a variety of chemicals have been developed [1,4–15]. Most of the gene switches developed to date use compounds that are not suitable for large-scale applications in the field because of environmental concerns. To overcome these problems, a more versatile chemically inducible gene regulation system has been developed [3,4,16]. The ecdysone receptor (EcR)-based gene regulation system is one of the best gene switches available, because the chemical ligands required for its regulation, tubufenozide and methoxyfenozide, are already registered for field use [17,18]. The advantages and limitations associated with various chemically inducible gene switches that have been developed to date have been discussed in recent reviews [4,18–20].

A chemically inducible gene regulation system that specifically regulates transgene expression in response to an exogenous inducer at a particular stage of plant development, or in a specific organ, is valuable for the expression of transgenes whose constitutive overexpression is likely to compromise plant viability or fertility. In addition, gene switches are useful to reduce environmental concerns, such as gene pollution and antibiotic resistance development, associated with genetically modified crops [17,18,21].

The properties of an ideal chemically inducible gene regulation system vary with each application; in general, the gene switch should show an undetectable level of transgene expression in the absence of a chemical ligand, followed by rapid and robust induction of transgene expression in the presence of a low concentration (nanomolar) of a chemical ligand. In an effort to find an ideal chemically inducible gene regulation system, several approaches have been tried [1,10,14,22–25]. An EcR gene switch with a potential for use in large-scale field applications and applicability to a variety of plant species has been developed by adopting a two-hybrid format [26]. This two-hybrid gene switch uses chimera 9 [CH9, a chimera between helices 1–8 of Homo sapiens retinoid X receptor (HsRXR) and helices 9–12 of Locusta migratoria RXR (LmRXR)] as a partner of Choristoneura fumiferana ecdysone receptor (CfEcR), and shows low background activity in the absence of ligand and high induction of luciferase reporter gene in the presence of nanomolar concentrations of methoxyfenozide ligand [27].

In a two-hybrid switch format, the GAL4 DNA-binding domain (GAL4 DBD) is fused to the ligand-binding domain (LBD) of CfEcR, and the VP16 activation domain (VP16 AD) is fused to LBD of LmRXR. On application of methoxyfenozide, the heterodimer of these two fusion proteins transactivates the luciferase reporter gene placed under the control of multiple copies of GAL4 response elements and the −46 35S minimal promoter (that includes −46 to the TATAA box of the 35S promoter [27]).

To develop an EcR gene switch that is highly sensitive and exhibits low background activity in the absence of ligand, we compared the amino acid residues in helices 1–8 of LmRXR and HsRXR. Based on the amino acid sequence differences in helices 1–8 of LmRXR and HsRXR, several LmRXR mutants were created by employing the site-directed mutagenesis method. We identified a triple mutant of LmRXR, SHILmRXR, which is as efficient as CH9 as a partner of CfEcR. The performance of the EcR gene switch with the SHILmRXR mutant was compared with that of the gene switches containing wild-type LmRXR and CH9, as a partner of CfEcR, in inducing reporter gene expression in tobacco protoplasts. The results observed in protoplasts were confirmed in Arabidopsis transgenic plants. To further improve the performance of the EcR gene switch containing the SHILmRXR mutant as a partner of CfEcR, Kozak sequences were incorporated upstream of the coding sequence of the reporter gene and compared with the standard reporter construct. A Kozak sequence was identified that worked well in combination with the CfEcR gene switch.

Results

Transient expression studies with tobacco protoplasts

RXR mutagenesis

The amino acid identity between LmRXR and HsRXR is fairly high. However, there are a few amino acid residues that are different between these two proteins. Site-directed mutagenesis was carried out to change the amino acid residues that were different between LmRXR and HsRXR in helices 1–8 to HsRXR residues in LmRXR (Fig. 1A, B). The performance of the LmRXR mutants in inducing luciferase reporter gene activity in a two-hybrid format was tested by co-electroporation of CfEcR and LmRXR and the luciferase reporter constructs into tobacco protoplasts. Among the LmRXR mutants screened, the T81H mutant showed low background activity in the absence of ligand, and the A62S mutant showed high induction of luciferase in the presence of ligand, when compared with wild-type LmRXR (Fig. 2A). In order to combine the desirable properties of these two LmRXR mutants, we incorporated these two mutations together. The A62S:T81H mutant (SHLmRXR) showed low background activity in the absence of ligand when compared with wild-type LmRXR, but lower induced luciferase activity. With a goal to improve the induction levels of the reporter gene in the presence of ligand, we introduced two additional mutations (V68I and V123I) into A62S:T81H LmRXR. Screening of these mutants identified a triple mutant, A62S:T81H:V123I (SHILmRXR), that showed low background activity in the absence of ligand and high induced activity in the presence of ligand when compared with the levels observed in either wild-type LmRXR or the double mutant SHLmRXR (Fig. 2B).

Figure 1.  (A) Schematic diagram of EcR two-hybrid switch. CfEcR plasmid was constructed by cloning a fusion of the GAL4 DBD (DBD) and CfEcR LBD (EcR LBD) cloned under the control of the 35S promoter (35SP) in the pKYLX80 vector [29] containing the terminator sequence (T). LmRXR plasmid was constructed by cloning the fusion of the VP16 activation domain (AD) and LmRXR LBD (RXR LBD) under the control of the 35S promoter in the pKYLX80 vector. The reporter vector −4635SLuc was constructed by cloning the luciferase reporter under the control of 5× GAL4 response elements and a 35S minimal promoter containing −46 to the TATAA box [27]. The ligand used in this switch is the ecdysone agonist, methoxyfenozide (M). (B) The alignment of the amino acid sequences of HsRXR and LmRXR helices 1–8. The amino acids that are identical between HsRXR and LmRXR are marked with a shaded background. Five amino acids selected for the first round of site-directed mutagenesis are boxed. Arrowheads point to the two amino acids tested as triple mutants.

Figure 2.  (A) Screening of LmRXR mutants S122A, A105S, T94A, T81H and A62S in tobacco protoplasts. Tobacco protoplasts were electroporated with receptor and the reporter constructs. Electroporated protoplasts were exposed to various concentrations of methoxyfenozide. The luciferase activity was measured 24 h after the addition of methoxyfenozide. The luciferase values are expressed as relative light units (RLU) per microgram of protein. The data shown are the average of three replicates ± standard deviation. (B) Comparison of performance of LmRXR, LmRXR double mutant A62S:T81H (SHLmRXR) and LmRXR triple mutant A62S:T81H:V123I (SHILmRXR) as a partner of CfEcR. The receptor and reporter constructs were electroporated into protoplasts. The transfected protoplasts were exposed to various concentrations of methoxyfenozide. Luciferase activity was measured 24 h after the addition of methoxyfenozide. The luciferase values are expressed as RLU per microgram of protein. The data shown are the average of three replicates ± standard deviation.

Optimization of translational start site

Several Kozak sequences placed upstream of the luciferase reporter gene translational start site were screened to determine whether they improved the transgene expression in plants. Among several Kozak sequences tested, AACAATGG performed the best in enhancing the induction of luciferase activity (Fig. 3). When this Kozak sequence was placed upstream of the luciferase reporter gene, the induction of luciferase increased significantly when compared with the induction of luciferase supported by the start site present in the commercial reporter vector (Promega Corporation, Madison, WI, USA).

Figure 3.  Improvement of Kozak sequence for the expression of transgenes regulated by the EcR gene switch. CfEcR, LmRXR and reporter constructs containing three modified versions of Kozak sequences were electroporated into protoplasts. The transfected protoplasts were exposed to various concentrations of methoxyfenozide. Luciferase activity was measured 24 h after the addition of methoxyfenozide. The luciferase values are expressed as relative light units (RLU) per microgram of protein. The data shown are the average of three replicates ± standard deviation.

Comparison of performance of CfEcR:SHILmRXR, CfEcR:CH9 and CfEcR:LmRXR gene switches in tobacco protoplasts

We compared the performance of CfEcR:SHILmRXR, CfEcR:CH9 and CfEcR:LmRXR gene switches in combination with the newly identified Kozak sequence by transfection of receptor and luciferase reporter constructs into tobacco protoplasts. The transfected protoplasts were exposed to various concentrations of methoxyfenozide and the luciferase activity was quantified. As shown in Fig. 4, both SHILmRXR and CH9 showed very low background activity in the absence of ligand when compared with the background activity exhibited by the LmRXR wild-type receptor. The reporter gene placed under the control of the EcR gene with either SHILmRXR, CH9 or LmRXR as a partner increased after the protoplasts were exposed to as low as 0.64 nm methoxyfenozide. The reporter activity increased steadily as the dose of methoxyfenozide increased, and reached maximum levels in protoplasts exposed to 80 nm methoxyfenozide. At this concentration of ligand, the CfEcR:SHILmRXR switch supported the highest induction of reporter activity. The CfEcR:CH9 switch supported a reporter activity that was lower than that observed for the CfEcR:SHILmRXR switch, but higher than that observed for the CfEcR:LmRXR switch. These transient expression studies in tobacco protoplasts showed that SHILmRXR was an excellent partner for CfEcR because of low background activity in the absence of ligand and higher induced activity in the presence of ligand, the two most desirable properties of a successful gene switch. A time-course experiment was conducted to determine the ‘on’ and ‘off’ properties of the CfEcR:SHILmRXR gene switch. The luciferase gene expression regulated by CfEcR:SHILmRXR began to increase at 3 h after addition of the ligand, and reached maximum levels 24 h after the addition of the ligand (Fig. 4B). The luciferase activity began to decrease 24 h after withdrawal of the ligand, and showed a continuous decrease until 42 h after ligand withdrawal. By this time, about 75% of the induced luciferase activity had disappeared (Fig. 4B). The performances of CfEcR:SHILmRXR, CfEcR:CH9 and CfEcR:LmRXR switches were further compared in transgenic plants as described in the next section.

Figure 4.  (A) Comparison of performance of LmRXR, CH9 and LmRXR triple mutant A62S:T81H:V123I (SHILmRXR) as a partner of CfEcR. The receptor and reporter constructs were electroporated into protoplasts. The transfected protoplasts were exposed to various concentrations of methoxyfenozide. Luciferase activity was measured 24 h after the addition of methoxyfenozide. The luciferase values are expressed as relative light units (RLU) per microgram of protein. The data shown are the average of three replicates ± standard deviation. DMSO, dimethylsulfoxide. (B) Time course of ‘turn on’ and ‘turn off’ of CfEcR:SHILmRXR switch. The receptor and reporter constructs were electroporated into protoplasts. The transfected protoplasts were exposed to 80 nm methoxyfenozide for 0, 3, 6, 12 and 24 h. The protoplasts were collected at the end of each time point and the luciferase activity was measured; the values are expressed as RLU per microgram of protein. In ligand withdrawal experiments, the transfected protoplasts were exposed to 80 nm methoxyfenozide for 24 h, followed by incubation in ligand-free medium for 6, 12, 24, 36 and 42 h. The protoplasts were collected at the end of each time point and the luciferase activity was measured. The data shown are the average of three replicates ± standard deviation.

Comparison of performance of CfEcR:SHILmRXR, CfEcR:CH9 and CfEcR:LmRXR gene switches in Arabidopsis transgenic plants

For stable transformation, CfEcR:SHILmRXR, CfEcR:CH9 and CfEcR:LmRXR gene switches and the luciferase reporter constructs were cloned into the T-DNA region of the pCAMBIA2300 binary vector. The Arabidopsis transgenic plants were developed and evaluated for their dose–response and time course of induction of luciferase activity in the presence of methoxyfenozide.

Dose–response study with T2 Arabidopsis plants

In order to analyze the dose–response of methoxyfenozide, four Arabidopsis lines were selected for each construct. The T2 seeds were germinated on agar medium supplemented with 50 mg·L⁻¹ kanamycin and 0 (dimethylsulfoxide), 0.64, 3.2, 16, 80, 400, 2000 and 10 000 nm methoxyfenozide. After 20 days, three seedlings from each plate were collected and assayed separately for luciferase activity. The plants containing the CfEcR:SHILmRXR switch exhibited low background activity of luciferase in the absence of ligand. On application of 16–80 nm methoxyfenozide, the luciferase activity in the CfEcR:SHILmRXR plants increased in a dose-dependent manner (Fig. 5). The induction of the luciferase reporter supported by this switch was initiated at a concentration as low as 0.64 nm methoxyfenozide and reached maximum levels at 16–80 nm methoxyfenozide. The methoxyfenozide dose–response of the CfEcR:SHILmRXR switch in combination with the new Kozak sequence was similar to the dose–response described above for the CfEcR:SHILmRXR switch and the original Kozak sequence. However, the luciferase activity at each dose of methoxyfenozide was higher in plants containing the new Kozak sequence. The transgenic plants containing the CfEcR:LmRXR switch showed higher background luciferase activity in the absence of ligand and lower induced luciferase activity in the presence of ligand. The transgenic plants containing the CfEcR:LmRXR switch showed a similar methoxyfenozide dose–response to that observed in plants containing the CfEcR:SHILmRXR switch (Fig. 5).

Figure 5.  Comparison of performance of LmRXR, CH9 and LmRXR triple mutant T81H:A62S:V123 (SHILmRXR) as a partner of CfEcR in T2 Arabidopsis plants. Seeds collected from T1 Arabidopsis lines were plated on agar medium containing 50 mg·L⁻¹ kanamycin and different concentrations (0, 0.64, 3.2, 16, 80, 400, 2000, 10 000 nm) of methoxyfenozide. Three seedlings from each plate were collected separately and ground in a volume of 100 μL of 1× passive lysis buffer, and the luciferase activity was measured and expressed in terms of relative light units (RLU) per microgram of protein. Data shown are the average of three replicates ± standard deviation.

Time course studies in Arabidopsis plants

In order to evaluate the time course of the methoxyfenozide response by the three EcR gene switches, T2 seeds collected from Arabidopsis transgenic lines were germinated on agar medium containing 50 mg·L⁻¹ kanamycin without methoxyfenozide. After 20 days, the seedlings were transferred to the glasshouse and different concentrations of methoxyfenozide ligand were applied to soil. The luciferase reporter gene expression was quantified at 0, 1, 2 and 4 days after application of methoxyfenozide. As shown in Fig. 6, luciferase activity began to increase 24 h after application of the ligand to the soil, and continued to increase up to 4 days. At most of the time points tested, the luciferase induction values observed were higher in seedlings exposed to 16–80 nm methoxyfenozide (Fig. 6A). The time course of activity of methoxyfenozide was similar for all three gene switches tested.

Figure 6.  (A) Comparison of performance of LmRXR, CH9 and LmRXR triple mutant A62S:T81H:V123I (SHILmRXR) as a partner of CfEcR in T2 Arabidopsis plants growing in soil. Seedlings grown on agar medium without added methoxyfenozide were transferred to the glasshouse; 80 nm methoxyfenozide was applied to the soil. Three leaf disks were collected at 0, 1, 2 and 4 days after the addition of ligand, and the luciferase activity was measured and expressed in terms of relative light units (RLU) per microgram of protein. Data shown are the average of three replicates ± standard deviation. (B) Time course of ‘turn on’ and ‘turn off’ of CfEcR:SHILmRXR switch in transgenic plants. Seedlings grown on agar medium without added methoxyfenozide were transferred to the glasshouse; 80 nm methoxyfenozide was applied to the soil. Three leaf disks were collected at 0, 6, 12, 24 and 96 h after the addition of ligand, and the luciferase activity was measured and expressed in terms of RLU per microgram of protein. In ligand withdrawal experiments, transgenic plants grown in soil containing 80 nm methoxyfenozide for 96 h were transferred to ligand-free soil, and samples were collected at 24, 48, 72, 96, 120 and 144 h after transfer. The luciferase activity was measured in leaf disks collected at each time point. The data shown are the average of four replicates ± standard deviation.

Time course studies were also conducted using T3 Arabidopsis transgenic plants expressing CfEcR and SHILmRXR to determine the ‘on’ and ‘off’ properties of this gene switch. As shown in Fig. 6B, luciferase activity began to increase 6 h after the addition of ligand, and reached maximum levels by 96 h after application of the ligand. The luciferase activity began to decrease 48 h after withdrawal of the ligand, and reached less than 25% of the induced levels by 144 h after withdrawal of the ligand.

Discussion

The major contribution of this study was the identification of the SHILmRXR mutant, which performs better than CH9 as a partner for CfEcR for application as a chemically inducible gene regulation system in plants. The CfEcR:SHILmRXR switch showed a low background activity of the reporter gene in the absence of ligand and a high induced activity of the luciferase gene in the presence of a concentration as low as 16 nm methoxyfenozide. Several versions of EcR gene switches have been developed for use in plants [3,4,13,16]. The EcR gene switch developed by Tavva et al. [26] uses CH9 as a partner of CfEcR. As CH9 is a region of HsRXR, this may not be accepted by the public for applications in genetically modified crops. Therefore, the current study was conducted to find an alternative partner for CfEcR.

Previous studies on comparisons between EcR:HsRXR and EcR:LmRXR gene switches have shown that the EcR:HsRXR switch exhibits a low background activity of the reporter in the absence of ligand, but the ligand sensitivity of this switch is lower than that of the EcR:LmRXR gene switch [27]. In contrast, the EcR:LmRXR switch shows higher ligand sensitivity when compared with the EcR:HsRXR switch, but this switch shows higher background activity of the reporter gene in the absence of ligand. We hypothesized that the differences in ligand sensitivity and background expression levels of luciferase reporter gene activity between LmRXR and HsRXR containing EcR gene switches might be a result of differences in the amino acid sequences in helices 9–11, because CH9, which was created by fusing helices 1–8 of HsRXR LBD and helices 9–12 of LmRXR LBD, performed very well by showing the desirable properties of both EcR:HsRXR and EcR:LmRXR switches. To test this hypothesis, seven LmRXR mutants were created and tested by transfection of the LmRXR mutants, together with CfEcR and reporter constructs, into tobacco protoplasts. These screening assays identified a triple mutant of LmRXR, SHILmRXR, that showed low background activity in the absence of ligand when used as a partner of CfEcR. The activity of this triple mutant of LmRXR was different from that of wild-type LmRXR in this respect, as wild-type LmRXR as a partner of EcR showed high background activity in the absence of ligand. In addition, the CfEcR:SHILmRXR gene switch also supported a high induced activity of reporter gene in the presence of a ligand concentration as low as 16 nm methoxyfenozide. This induced activity supported by SHILmRXR as a partner of CfEcR was better than that observed when wild-type LmRXR was used as a partner of CfEcR. Thus, this newly identified triple mutant of LmRXR as a partner of CfEcR possesses two of the most desirable properties of receptors used in gene switches. Comparative studies in tobacco protoplasts, as well as in Arabidopsis transgenic plants, showed that the background activity of the CfEcR:SHILmRXR switch was similar to that supported by the CfEcR:CH9 switch, but the induced reporter activity supported by the CfEcR:SHILmRXR switch in the presence of 16–80 nm methoxyfenozide was better than that observed for the CfEcR:CH9 switch. The differences observed in the background activity in the absence of ligand and induced reporter activity in the presence of ligand between wild-type LmRXR and its triple mutant are most probably caused by differences in their ability to heterodimerize with CfEcR [17,18]. Previous studies in mammalian cells on the heterodimerization of EcR and RXR have suggested that SHILmRXR does not heterodimerize with EcR in the absence of ligand [28]. This mutant behaves like HsRXR in this respect. Palli et al. [28] reported that HsRXR does not heterodimerize with CfEcR in the absence of ligand. In the presence of a nanomolar concentration of methoxyfenozide, SHILmRXR heterodimerizes well with CfEcR and transactivates genes placed under the control of methoxyfenozide response promoters. In this respect, SHILmRXR behaves like CH9. Palli et al. [18] showed that, when compared with either HsRXR or LmRXR, the chimeras between HsRXR and LmRXR heterodimerize well with CfEcR at nanomolar concentrations of ecdysteroid ligands.

We screened several Kozak sequences to identify a sequence that supports better transgene expression in plants. We observed a significant increase in luciferase reporter gene activity in the presence of ligand when a Kozak sequence containing AACAATGG was used to replace the native luciferase reporter gene Kozak sequence. Interestingly, an increase in the ligand-induced activity of this switch containing the improved Kozak sequence was not accompanied by an increase in background activity in the absence of ligand. Therefore, this newly identified Kozak sequence could be used in EcR gene switches, as well as other gene switches, for the regulation of transgenes in plants.

The data presented here clearly demonstrate that luciferase reporter gene expression is tightly regulated by the CfEcR:SHILmRXR gene switch. This gene switch showed negligible levels of background reporter gene activity in the absence of ligand and the highest levels of induced reporter gene activity in the presence of a concentration as low as 16 nm of methoxyfenozide. Although the amino acid identity between SHILmRXR and HsRXR was increased, when compared with the amino acid identity between wild-type receptors, SHILmRXR performed much better than wild-type LmRXR as a partner for EcR, and SHILmRXR is an insect receptor. Therefore, the improvement made to the EcR gene switch in this study by the manipulation of LmRXR as a partner of CfEcR may lead to the widespread use of this gene switch for applications in plants.

Experimental procedures

Ligand

Technical grade (99% pure) methoxyfenozide (a gift from Rohm and Haas Company, Spring House, PA, USA) was dissolved in dimethylsulfoxide to prepare 100 mm solution. Various dilutions of this ligand in dimethylsulfoxide were prepared from this stock solution.

LmRXR mutants

Site-directed mutagenesis was carried out using the Quick Change Site-Directed Mutagenesis Kit (Stratagene, La Jolla, CA, USA). Mutations were verified by sequencing. The primers used for mutagenesis were as follows: LmS122A (F), CTTGGCTGCTTGCGAGCTGTTATTCTTTTCAATCC; LmS122A (R), GGATTGAAAAGAATAACAGCTCGCAAGCAGCCAAG; LmA105S (F), TTGACAGAACTGGTATCAAAGATGAGAGAAATG; LmA105S (R), CATTTCTCTCATCTTTGATACCAGTTCTGTCAA; LmT94A (F), CAAGCTGGAGTCGGCGCAATATTTGACAGAGTTTTG; LmT94A (R), CAAAACTCTGTCAAATATTGCGCCGACTCCAGCTTG; LmT81H (F), CTTGCCACTGGTCTCCACGTGCATCGAAATTCTGCC; LmT81H (R), GGCAGAATTTCGATGCACGTGGAGACCAGTGGCAA; Lm A62S (F), GAACTGCTAATTGCATCATTTTCACATCGATCTG; Lm A62S (R), CAGATCGATGTGAAAATGATGCAATTAGCAGTTC; Lm V123I (F), TGGCTGCTTGCGATCTATTATTCTTTTCAATCC; Lm V123I (R), GGATTGAAAAGAATAGATCGCAAGCAGCCA.

Screening of LmRXR mutants in tobacco protoplasts

The LmRXR mutants S122A, A105S, T94A, T81H, A62S, A62S:T81H and A62S:T81H:V123I were cloned downstream of the VP16 AD sequence in the pVP16 vector (BD Biosciences Clonetech, San Jose, CA, USA). DNA sequences coding for the fusion protein of VP16 AD and LmRXR mutants were transferred from pVP16LmRXR to the pKYLX80 vector [29] using NheI and XbaI restriction endonucleases. GAL4 and CfEcRDEF fusion protein construct (CfEcR) and a reporter construct containing the gene coding for luciferase under the control of the −46 35S minimal promoter and GAL4 response elements (−4635SLuc) were cloned into the pKYLX80 vector (Table 1). Transient expression studies were carried out by isolating protoplasts from cell suspension cultures of tobacco (Nicotiana tabacum cv. Xanthi-Brad). A detailed description of the isolation and electroporation of protoplasts has been given previously by Tavva et al. [26].

Table 1.   Constructs used in the experiments.

Name of the construct	Description
CfEcR	A fusion of GAL4 DNA-binding domain with CfEcR ligand-binding domain cloned under the control of the 35S promoter in the pKYLX80 vector
LmRXR	A fusion of VP16 activation domain and LmRXR ligand-binding domain cloned under the control of the 35S promoter in the pKYLX80 vector
CH9	Chimera 9, a chimera between helices 1–8 of Homo sapiens retinoid X receptor (HsRXR) and helices 9–12 of Locusta migratoria retinoid X receptor (LmRXR)
SHLmRXR	A double mutant A62S:T81H of LmRXR
SHILmRXR	A triple mutant A62S:T81H:V123I of LmRXR
−46 35SLuc	Luciferase regulated by 5× GAL4 response elements and −46 to TATAA 35S promoter
−46 35SLuc	Luciferase containing modified Kozak sequence and regulated by 5× GAL4 response elements and −46 to TATAA 35S promoter

Dose–response and time course studies in tobacco protoplasts

The methoxyfenozide dose-dependent performance of different LmRXR mutants in inducing luciferase reporter gene activity in a two-hybrid gene switch was tested by co-electroporation of pK80-46 35S:luc, pK80GCfE and mutant LmRXR constructs. The electroporated protoplasts were incubated in growth medium containing 0, 0.64, 3.2, 16, 80, 400, 2000 and 10 000 nm methoxyfenozide. Twenty-four hours after the addition of ligand, the protoplasts were assayed for luciferase reporter gene activity using a Fluoroscan FL plate reader (Fluoroscan Ascent FL, Thermo Biosystems, Milford, MA, USA) as described previously [26]. In time course experiments, the protoplasts transfected with CfEcR and SHILmRXR and the −46 35SLuc reporter were exposed to 80 nm methoxyfenozide for 0–24 h, protoplasts were collected and the luciferase activity was measured. In ligand withdrawal experiments, the transfected protoplasts were exposed to 80 nm methoxyfenozide for 24 h, and the protoplasts were then transferred to ligand-free medium and incubated for an additional 6–42 h. The protoplasts were collected at the end of each time point and the luciferase activity was measured.

Optimization of translational start site using Kozak sequences

Kozak sequences (AAAAATGG, AACCATGG and AACAATGG) were placed upstream of the luciferase reporter gene and screened for transgene expression in plants. −46 and −31 35S minimal promoters were cloned into the pKYLX80 vector as described by Tavva et al. [26]. The luciferase reporter gene was PCR amplified from the pFRLuc vector (Stratagene) using several sets of primers containing different Kozak sequences, and cloned into pGEM-T Easy vector to verify the integration of the Kozak sequences. The PCR primers used were as follows: KZKLUC2 (F), CTCGAGAAAAATGGAAGACGCCAAAAACATAAAG; KZKLUC4 (F), CTCGAGAACCATGGAAGACGCCAAAAACATAAAG; KZKLUC1 (F), CTCGAGAACAATGGAAGACGCCAAAAACATAAAG. The bold letters in the primers show Kozak sequence.

The luciferase reporter gene with the Kozak sequence was then excised from the pGEM-T Easy vector and cloned into the XhoI/SacI sites downstream of the −46 35S minimal promoter in the modified pKYLX80 vector. The reporter gene expression cassette with the pKYLX80 background was designated as pK80-46 35S:KLuc. The full-length luciferase gene was cloned into the pKYLX80 vector under the control of the cauliflower mosaic virus (CaMV) 35S promoter and used as a positive control in transient transfection studies.

Binary vectors for stable transformation of Arabidopsis

Binary vectors for stable transformation of Arabidopsis thaliana were constructed in pCAMBIA2300 vectors (CAMBIA, Canberra, Australia). To construct the binary vector for plant transformation, the GAL4DBD:CfEcRDEF fusion gene was cloned under the control of the figwort mosaic virus (FMV) promoter and ubiquitin 2 (Ubi) terminator sequence, and the VP16AD:LmRXR fusion gene was cloned under the control of the Mirabilis mosaic virus (MMV) promoter and Agrobacterium tumefaciens octopine synthase (Ocs) poly A sequence. The FMV- and MMV-driven expression cassettes were assembled into the pSL301 vector. The reporter and receptor expression cassettes were excised with appropriate restriction enzymes and assembled into the pCAMBIA2300 vector for plant transformation. The map of the binary vector used is shown in Fig. 7.

Figure 7.  p2300:Lm7:KLuc: T-DNA region of the pCAMBIA2300 binary vector showing the assembly of receptor and reporter expression cassettes. FMVP, figwort mosaic virus promoter; MMVP, Mirabilis mosaic virus promoter; OCST, Agrobacterium tumefaciens octopine synthase poly A; rbcST, poly A sequence; UbiT, ubiquitin 3 terminator; 35S P, CaMV 35S promoter with double enhancer sequence.

Production of transgenic plants

Arabidopsis thaliana (L.) Heynth. ecotype Columbia ER was used for plant transformation experiments. The binary vectors were mobilized into Agrobacterium tumefaciens strain GV3850 by the freeze–thaw method. Arabidopsis plants were transformed using the whole-plant dip method [30]. Transgenic Arabidopsis plants were selected by germinating the seeds collected from the infiltrated plants on medium containing 50 mg·L⁻¹ kanamycin. The analysis of transgenic plants for luciferase induction level was carried out on T2 and T3 seeds plated on kanamycin-containing medium.

Dose–response studies in T2 Arabidopsis plants

Seeds collected from four T1 Arabidopsis lines were plated on agar medium containing 50 mg·L⁻¹ kanamycin and different concentrations of methoxyfenozide (0, 0.64, 3.2, 16, 80, 400, 2000, 10 000 nm). The seeds were allowed to germinate and were grown on the induced medium for 20 days at 25 °C, 16 h light/8 h dark. Three seedlings from each plate were collected separately and ground in a volume of 100 μL of 1× passive lysis buffer (Promega Corporation), and the luciferase activity was measured. To study the dose–response in soil-grown plants, T2 Arabidopsis plants were transferred to soil in pots placed in a glasshouse, and the plants were allowed to grow in the glasshouse until they had developed a complete whorl of rosette leaves. Different doses of methoxyfenozide (0, 0.64, 3.2, 16, 80, 400, 2000, 10 000 nm) were applied to the pots three times at 2-day intervals. Leaf disks were collected on day 4 and the luciferase activity was measured.

Time course studies in soil-grown plants

T2 Arabidopsis plants were transferred to a glasshouse, and a time course study was conducted after application of 0, 0.64, 3.2, 16, 80, 400, 2000 or 10 000 nm methoxyfenozide to the soil. Care was taken not to leach out any excess solution. Leaf disks were collected at 0, 1, 2 and 4 days after application of methoxyfenozide, and the luciferase activity was measured. To determine the ‘on’ and ‘off’ properties of the CfEcR:SHILmRXR switch, time course studies were conducted using T3 Arabidopsis plants. The transgenic T3 Arabidopsis plants growing in the soil were exposed to 80 nm methoxyfenozide for 0–96 h. The luciferase activity was determined in leaf disks collected at 0, 6, 12, 24 and 96 h after application of the ligand. For ligand withdrawal experiments, T3 Arabidopsis plants were exposed to 80 nm methoxyfenozide for 96 h, followed by transfer to ligand-free soil. The luciferase activity was determined in leaf disks collected at 24, 48, 72, 96, 120 and 144 h after withdrawal of the ligand.

Acknowledgements

This work was supported by Consortium for Plant Biotechnology Research. This is contribution number 10-08-112 from the Kentucky Agricultural Experimental Station.