A Glycine Insertion in the Estrogen-Related Receptor (ERR) Is Associated with Enhanced Expression of Three Cytochrome P450 Genes in Transgenic Drosophila melanogaster

Insecticide-resistant Drosophila melanogaster strains represent a resource for the discovery of the underlying molecular mechanisms of cytochrome P450 constitutive over-expression, even if some of these P450s are not directly involved in the resistance phenotype. For example, in select 4,4’-dichlorodiphenyltrichloroethane (DDT) resistant strains the glucocorticoid receptor-like (GR-like) potential transcription factor binding motifs (TFBMs) have previously been shown to be associated with constitutively differentially-expressed cytochrome P450s, Cyp12d1, Cyp6g2 and Cyp9c1. However, insects are not known to have glucocorticoids. The only ortholog to the mammalian glucocorticoid receptor (GR) in D. melanogaster is an estrogen-related receptor (ERR) gene, which has two predicted alternative splice isoforms (ERRa and ERRb). Sequencing of ERRa and ERRb in select DDT susceptible and resistant D. melanogaster strains has revealed a glycine (G) codon insertion which was only observed in the ligand binding domain of ERR from the resistant strains tested (ERR-G). Transgenic flies, expressing the ERRa-G allele, constitutively over-expressed Cyp12d1, Cyp6g2 and Cyp9c1. Only Cyp12d1 and Cyp6g2 were over-expressed in the ERRb-G transgenic flies. Phylogenetic studies show that the G-insertion appeared to be located in a less conserved domain in ERR and this insertion is found in multiple species across the Sophophora subgenera.


Introduction
Resistance to insecticide is a major challenge for pest control and also an important man-made micro-evolutionary force [1,2].Resistance is also a system for studying microevolution and environmental adaptation [2,3].DDT (4,4'-dichlorodiphenyltrichloroethane) resistance in Drosophila melanogaster has been used for the study of the evolution of insecticide resistance.Metabolic insecticide resistance is an important form of resistance and is often associated with constitutive over-expression of (or, in some cases, structural changes in) cytochrome P450s as well as glutathione-S-transferases (GSTs), esterases, or a combination of these genes [4][5][6].
Much of the emphasis in the pesticide resistance community has been on the discovery of which genes confer the resistant phenotype, however, pesticide resistant strains also represent a resource to investigate the molecular mechanisms by which transcription of P450s occurs, regardless if these genes play any role in resistance.For example, moderate-to high-levels of metabolic DDT resistance in D. melanogaster is generally considered to be polygenic [5][6][7][8][9][10].However, metabolic DDT resistance is not a single phenotype; the lethal concentration 50 (LC 50 ) for DDT varies considerably across pesticide-resistant D. melanogaster strains [10].The two metabolically insecticide-resistant fly strains, Wisconsin and 91-R, show very different LC 50 s when bioassayed with DDT; 91-R is far more resistant to DDT than Wisconsin.Several reports in the literature have shown that constitutive over-expression of Cyp12d1, Cyp6a2 and Cyp6g1 occurs in some DDT-resistant strains [6,11,12].However, a recent toxicokinetic analysis of 91-R revealed that oxidative P450s likely cause little direct metabolic resistance, with reduced penetration, increased reductive dechlorination and enhanced excretion playing more dominant roles [13].It is unlikely, therefore, that these P450s play a critical role in moderateto high-levels of DDT resistance through DDT metabolism, however, they are constitutively over-transcribed and little is known about the regulation of the expression.Plapp [14] hypothesized that mutated trans-regulatory factors might be involved in the constitutive overexpression of detoxification genes.Recently, it has been shown that there is constitutive activation of the Nrf2/Keap1 pathway in DTT-resistant strains and that this might cause about 20% of the genes (including several P450s) to be differentially expressed in 91-R strain [15].To date, only the constitutively over-expression of Cyp6a2 has been associated with an allele containing an intact Nrf2/Maf-binding-site [16].The mechanism(s) activating this pathway is not fully understood and it is difficult to address because there are multiple steps that can be regulated.
Recently, Qiu and collaborators [12] used a potential transcription factor binding motif (TFBM) analysis of the genes differentially transcribed in the two DDT-resistant strains, Wisconsin and 91-R found that for all the genes coding for detoxification enzymes a glucocorticoid receptor-like (GR-like) motif was present.The glucocorticoid receptor (GR) is a versatile nuclear receptor (NR) of the family NR3 that in mammalians regulates genes controlling metabolism, development and immunity.Specifically, GR has been implicated in xenobiotic-induced expression of several cytochrome P450 enzymes [17].Insects are not known to have GRs, however, the GR's closest NR3 member in D. melanogaster is the estrogen-related receptor (ERR, CG7404, FBgn0035849), of which there are two known transcripts, ERRa and ERRb.
In the current study, we demonstrated that ERRa and ERRb isoforms, containing a G codon insertion caused constitutively over-expression of Cyp12d1, Cyp6g2 and Cyp9c1 as transgenes.The region in the ERR gene where the G insertion occurs appeared to be variable across D. melanogaster strains and this insertion was observed in other Drosophila species in the Sophophora subgenera.

Drosophila melanogaster strains
Three D. melanogaster strains were used for the initial sequencing of ERRa and ERRb cDNAs: Canton-S (susceptible to DDT), Wisconsin (moderately resistant to DDT), and 91-R (highly resistant to DDT).Detailed descriptions of these fly strains, as well as the 91-C (susceptible to DDT) strain, rearing conditions, and sample preparation were given in [18] and [6].The LC 50 s for Canton-S, Wisconsin and 91-R are given in Festucci-Buselli et al [10].Populations of 91-C and 91-R strains have been maintained in the Pittendrigh laboratory for over a dozen years.For the transgenic lines w 1118 was used.In addition to the above strains, another 53 wild-type lines of D. melanogaster were obtained from the Bloomington Drosophila Stock Center and were genotyped for ERR.A detailed description of these fly strains is given in [19].

ERR transcript sequencing and cloning
Total RNA from three strains (Canton-S, Wisconsin and 91-R) were used to synthesize fulllength cDNA with oligo dT by using a cDNA Cycle kit (Invitrogen, CA).Gene-specific primers were used to amplify the full-length cDNA for all three strains (S1 Table ).The amplified cDNA-specific products were purified using a PCR clean-up kit (Qiagen, CA) and directly sequenced with three forward and three reversing primers to cover the full length.All sequences were assembled and compared by using Vector NTI (Invitrogen, CA).

Absolute real time quantitative (RT-qPCR) for transcript abundance of ERRa and ERRb
In order to evaluate the differences in transcript abundances between ERRa and ERRb in the Canton-S, Wisconsin, and 91-R lines, absolute quantitative real-time PCR (RT-qPCR) was performed.Gene-specific primers were designed at the UTR of ERRa and ERRb (S1 Table ).The single-stranded cDNA was synthesized using the GoScript reverse transcription system from Promega.Real-time PCR was performed at an annealing temperature of 58°C, which had been optimized by gradient RT-qPCR.For each gene, one single amplification product was confirmed by a single peak melting curve and a single band on an agarose gel.The PCR amplicon for each gene was purified using a Qiagen PCR clean-up kit and sequenced to insure correct amplification.Amplicon sizes were 125 bp and 131 bp for ERRa and ERRb, respectively.The purified amplicons were quantified using a Nanodrop 1000 spectrophotometer and tenfold dilution series (from 100 pM to 0.1 fM) were prepared for each as standards for RT-PCR analysis.Six biological replicates of adult flies (3 days old) of mixed sexes for Canton-S, Wisconsin, and 91-R strains were tested.The absolute expression level difference between ERRa and ERRb was compared within each fly strain by using paired t-test since ERRa and ERRb transcripts were from the same cDNA reaction.

Transgenic flies
Full length ERRa and ERRb cDNA from the 91-R strain was cloned into a pCR2.1 Topo vector (Invitrogen, CA).Individual clones were sequenced to find an open reading frame and correct amino acid sequences for ERRa and ERRb.The selected clones for both ERR transcripts were sub cloned into a pUAST vector.Transgenic flies were generated by the Best Gene Inc (Chino Hills, CA) using w 1118 strain.The driver strain (y 1 w Ã ; P{Act5C-Gal4-w}E1/CyO) was obtained from Bloomington Drosophila Stock Center.

Relative quantitative real time PCR (RT-qPCR)
To compare the effects of cloned ERRa and ERRb genes on P450 transcription patterns from the Cantons-S and 91-R strains, we used transgenic flies before crossing (F 0 ) as controls.For each gene and/or treatment combination, three biological replicates (RNA samples) were prepared.Three-day-old flies were collected for RNA extraction.RNA was extracted from 30 flies (1:1 male/female ratio) using the Qiagen RNeasy kit (Qiagen, Valencia, CA, USA) with an "on-column" DNase digestion procedure.The first-strand cDNA was synthesized by using 0.5 μg of total RNA with GoScript Reverse Transcription System from Promega (Madison, WI) in a 20 μl reaction volume with random primers.A StepOne plus real-time PCR system from Applied Biosystems (Carlsbad, CA) were used to run qRT-PCR with GoTaq qPCR master mix from Promega (Madison, WI).For each cDNA, three RT-qPCR reactions were performed.All cDNA samples were equalized with rp49 gene as reference before target genes were tested.The average threshold cycle (Ct) for each cDNA was calculated by the StepOne software.The relative expression levels were calculated as given in [20] by using rp49 as the reference gene.Statistical analysis was done on target gene delta Ct after adjusted by rp49 Ct in SAS software (SAS Institute Inc., Cary, NC, USA) with a GLM model and contrast statements.All the primers for the assayed genes are listed in S1 Table.

The structure analyses of ERRa and ERRb proteins
The tertiary structures of the isoforms ERRa and ERRb with and without the G insertion were predicted by SWISS-MODEL Workplace (http://swissmodel.expasy.org/workspace/)automated mode [21].The tertiary structures with and without the G insertion were superimposed for both ERRa and ERRb respectively by inputting Protein Data Bank (PDB) files into the server (http://mspc.bii.a-star.edu.sg/minhn/click.html;[22]).

Analysis of ERR across D. melanogaster strains
Genomic DNA was extracted from 10 to 15 adult flies using the DNeasy Blood and Tissue Kit (Qiagen, Valencia, CA).DNA was quantified by spectrophotometry using Nanodrop 1000 (Thermo Scientific, Wilmington, DE).Based on D. melanogaster reference genome (NT_037436.3),primers were designed to amplify a region of the ligand binding domain (LBD) of ERR gene (S1 Table ).The LBD region was amplified from 100 ng of genomic DNA, 1.25 U of Go Taq (Promega Corporation, Madison, WI) with 1 × Promega Mg 2+ -free buffer, 1.5 mM of MgCl 2 , 200 μM of each dNTP and a final concentration of 200 nM of primers in a 25 μl volume.Thermal cycling started at 94°C for 5 min; followed by 35 cycles of 94°C for 30 s, annealing at 59°C for 30 s and extension at 72°C for 1 min and finished with a final extension at 72°C for 7 min.The PCR products were purified with Qiaquick PCR purification kit (Qiagen, Valencia, CA) and sequenced at the Core Sequencing Facility at University of Illinois at Urbana-Champaign.

Results
Sequencing and expression of ERR transcripts from Canton-S, Wisconsin, and 91-R The sequences for ERRa and ERRb transcripts were identical except for an extra 12 amino acids at 5' end of ERRb (S1 Fig. ).The ERRa and ERRb sequences from Canton-S strain were identical to the sequences in the GenBank.Sequencing of ERRa and ERRb from 91-R revealed an insertion of three nucleotides that resulted in a G amino acid insertion at position 282 of the protein that was not found in Canton-S or the sequences in the GenBank.The sequencing for Wisconsin revealed heterozygosity for the G codon insertion in both ERRa and ERRb.As shown in S1 Fig., the insertion was in the same relative location for both ERRa and ERRb.Absolute RT-qPCR of ERRa and ERRb from the Canton-S, Wisconsin and 91-R strains revealed that, in all three aforementioned strains, ERRa transcripts were 3.4 to 5.4 times more abundant than that found for ERRb (Table 1).

ERR expression in three different strains
As shown in Fig. 1A, ERRa transcript level was not different between Canton-S and Wisconsin.The ERRa transcript level was significantly higher in 91-R compared to the transcripts either in Canton-S or Wisconsin.For ERRb, the transcript abundance levels were significantly higher in both Wisconsin and 91-R as compared to Canton-S and no difference was observed between them (Fig. 1B).However, the highest relative difference for either ERRa or ERRb was around 50%.

ERR transgenic flies over-transcribe selected P450s
Both the ERRa and ERRb cDNAs, from 91-R, contained the G codon insertion and caused a constitutive 3.5-to 8-fold over-expression of multiple P450 genes when expressed in a transgenic insect.In the transgenic flies expressing the ERRa cDNA from 91-R, transcriptional levels of Cyp12d1, Cyp6g2 and Cyp9c1 were significantly increased (Fig. 2A).In transgenic flies,  expression of the ERRb cDNA from 91-R, significantly increased expression levels of Cyp12d1 and Cyp6g2, but not Cyp9c1 (Fig. 2B).However, the ERRa and ERRb transcripts from Canton-S (wt) that do not have the G codon insertion did not cause significant increases in P450 expression in any of the P450s tested when expressed in transgenic insects (Fig. 2C and 2D; Cyp12d1 was not significantly over-transcribed after a Bonferroni correction in Fig. 2D).

Analysis of ERR gene across D. melanogaster populations
In order to determine the frequency of the G insertion allele across different D. melanogaster populations, a total of 57 strains (including 91-R and 91-C) were genotyped for a portion of the LBD of ERR, which was 160 amino acids in length.Our analysis revealed that the insertion of the G codon is a common event, occurring in 61.4% (35/57) of the D. melanogaster strains; with 54.4% (31/57) of the strains showing this allele in homozygosity and 7% (4/57) of them showing the allele in heterozygosity.In four strains (CO 7, Florida-9, pi 2 and Wild 5C), we found a third allele that had a 21 nucleotide deletion in combination with the G codon insertion that resulted a deletion of seven amino acids upstream of the insertion.This third combined allele (deletion + G) was present in homozygosity in 3.5% (2/57) of the strains tested.Strains pi 2 and Wild 5C were heterozygous for del+G and wt+G (Fig. 3).

The structural analyses of ERRa and ERRb proteins
Using NCBI PSI-BLAST and DELTA-BLAST, we were able to predict both the putative DNA binding domain (DBD) and LBD for the ERRa and ERRb proteins.For ERRa, the predicted zinc finger DBD occurred at the amino acid sequences from 109 to 205 in both Canton-S and 91-R proteins.Additionally, the amino acid sequences from 225 to 482 were predicted to be the LBD, for both the Canton-S and 91-R ERRa proteins.For ERRb, the DBD domain was predicted to be from amino acid 121 to 217 for both the Canton-S and 91-R proteins.The LBD domain, however, was predicted to be from amino acid 237 to 493 for the Canton-S protein and from 237 to 494 for the 91-R protein; the G insertion for 91-R ERR was near the amino-end of the LBD domain (270 for ERRa and 282 for ERRb).Pairwise structure alignment using the server on http://ekhidna.biocenter.helsinki.fi/dali_lite/start [25] indicates that the G insertion has caused structurally misalignment of about 40 amino acids at the amino-end of the insertion and about 12 amino acids at the carboxyl end of insertion in both ERRa and ERRb (S2 Fig. and S3 Fig.).The effect of this shift in the secondary structure was observed in the tertiary structure model generated by SWISS Model automated mode and superimposed by CLICK [22].Superimposing of the tertiary structures of the wt ERRa and G ERRa revealed two loops not overlapping due to the G insertion (Fig. 4A).Similarly, unmatched loops were likewise observed when wt ERRb and G ERRb tertiary structures were superimposed (Fig. 4B).

Analysis of ERR mutation across Drosophila species
To investigate if the G insertion was present in other Drosophila species, we made comparisons among the twelve available Drosophila species ERR sequences.The region in ERR, where the G insertion occurred in D. melanogaster, appears to be highly variable across the Drosophila species we investigated (Fig. 5).Additionally, insertion of the G codon occurs in multiple species across the Sophophora subgenera (Fig. 6).

Discussion
The NR superfamily of genes codes for ligand-inducible transcription factors, in which are important for signaling pathways and coordination of the transcriptional response to specific ligands.Most NR proteins share a similar structure being composed of several modular domains, which are differentially conserved, including a highly conserved DBD, a variable hinge region and a less conserved LBD.The ERRs are a subfamily of the NR for which ligands have yet to be identified (so called orphan receptors).In humans, there are three ERRs (α, β and γ) that were identified through homology searches for genes coding for proteins related to estrogen receptors (ERs).In Drosophila, there is only a single ERR gene [26], which is orthologous to human ERRγ, and there is no Drosophila orthologous for human ERs.This finding suggests a common ancestor but the phylogeny of these genes remains controversial [27].
Studies, mainly in mammalians, have identified common mechanisms and divergent functions of ERRs [28].To date, no natural ligand for ERRs has been described and the research approaches focus mostly on the physiological process under transcriptional control of ERRs.Recent research provides evidence that ERRs are master regulators of the mitochondrial biogenesis and function [29].The fact that mammalian ERRs are expressed in tissues with high metabolic needs gave the first clues in determining the role of the ERRs in the regulation of energy metabolism [30].In Drosophila, a recent study demonstrated the role of ERR in carbohydrate metabolism during larva stages [31] and a very recent paper shows an emerging role of ERR in hypoxic networks [32].
Here we described that an amino acid insertion in the hinge close to the 5'end of LBD of both Drosophila ERRa and ERRb in the 91-R resistant strain and that this amino acid insertion caused constitutive over-expression of multiple cytochrome P450s when transformed into the w 1118 strain.This represents, to the authors' knowledge, the first reported study of a naturallyoccurring allele in ERR influencing the expression of the three xenobiotic responsive P450s, Cyp12d1, Cyp6g2 and Cyp9c1, in adult Drosophila.The G insertion allele was also common across D. melanogaster strains as well as other Drosophila species.The G insertion and flanking deletion alleles were also observed in some strains of D. melanogaster.Previous analyses of the expression of several P450s in the larva of the ERR double mutant showed that Cyp6a8 and Cyp9c1 are the most up-regulated P450 genes [31].
Interestingly, Cyp6g2 over-expression has been shown to be induced by nitenpyram and diazinon treatment [33].Cyp12d1 over-expression has been shown to be induced by DDT [18,33], dicyclanil [33], phenobarbital, hydrogen peroxide, heat (37°C) and starvation [19].To date, and to the authors' knowledge, there are no existing reports demonstrating any xenobiotics causing differential transcription of Cyp9c1.These results indicate that expression of at least two of these aforementioned P450s, Cyp6g2 and Cyp12d1, are responsive to environmental or xenobiotic stresses or both.
The position of the G insertion is at amino acid 282 relative to ERRb, in the hinge close to the 5'end of LBD.An engineered triple mutation, Y295A/T333I/Y365L, in the ERR binding domain enabled the binding of a known ERR inverse agonists and suppressed the transcriptional activity of the receptor [34].It is possible that the insertion of G close to this general region may alter the affinity of ERR to putative ligands and/or may make the DNA binding domain more accessible; a hypothesis that remains to be tested.
It has previously been shown that ERR is able to interact with either single or repeated response elements in vitro [34].Our in silico search within 10 kb of 5 prime region (10 Kb upstream of the start codon) of the over-expressed P450s, Cyp12d1, Cyp6g2 and Cyp9c1, revealed a putative ERR binding motif for all three genes that is conserved across the 12 Drosophila species with known genome (S2 Table ).In contrast, the same approach was used in the search of As a nuclear receptor protein, ERR may exert its effect through multiple regulatory pathways.It has been shown that the nuclear receptor, DHR96, which is orthologous to two human NR, Steroid and Xenobiotic Receptor (SXR) and Constitutive Androstane Receptor (CAR), controls only a small portion of the phenobarbital (PB) response [35].The Nrf2/Keap1 pathway is also thought to regulate xenobiotic responsive genes [15].Recent results from the same research group show that the constitutive activation of the Nrf2/Keap1 pathway can contribute to the overexpression of detoxifying genes in resistant strains but the mechanism(s) remain unknown [36].The regulation of NRF2 pathway by ERs and ERRs is well documented in mammalians [37][38][39][40].Specifically, it has been shown that ERRβ (particularly the short isoform SFhERRβ) interacts with NRF2 and inhibits its transcriptional activity [41].
Our results show that the pattern of expression of both ERR isoforms is different between Drosophila strains (Table 1), particularly the levels of ERRa expression in the 91-R resistant strain, which is more than two times that of the level of expression in the susceptible Canton-S strain.It has been described an association between the constitutive over-expression of Cyp6a2 with an allele containing an intact Nrf2/Maf-binding-site [16].We were not able to identify a highly conserved Nrf2 binding site close to the locations of the putative ERR binding site described among 12 Drosophila species (S2 Table ).Finally, it has been postulated that mutations in repressor gene(s) on the third chromosome may be one reason for the constitutive over- The evolutionary history was inferred by using the Maximum Likelihood method based on the JTT matrix-based model [23].The tree with the highest log likelihood (-2145.5939) is shown.Initial tree(s) for the heuristic search were obtained automatically by applying Neighbor-Join and BioNJ algorithms to a matrix of pairwise distances estimated using a JTT model, and then selecting the topology with superior log likelihood value.The tree is drawn to scale, with branch lengths measured in the number of substitutions per site.All positions containing gaps and missing data were eliminated.There were a total of 455 positions in the final dataset.Evolutionary analyses were conducted in MEGA5.2 [24].doi:10.1371/journal.pone.0118779.g006expression of Cyp6a2 and Cyp6a8 in resistant Drosophila strains [42].Similar to the and dKeap1 genes, ERR is also on the third chromosome.It remains to be determined, however, if ERR is a regulator of Nrf2/Keap1pathway.Further experimentation will need to be performed to test this hypothesis.
Future studies, using antibodies, might be able to determine protein levels of ERR between the wt and G insertion EER alleles.We speculate that the level of ERR protein would not be different between wt and G insertion ERRs.The 3D protein modeling (Fig. 4 indicates that the G insertion causes tertiary structural changes to the ERR proteins.These changes might have increased the ERRs DNA binding to these three CYPs transcription factor binding motifs after binding its putative ligand.As of now, no known physiological ligand has been identified to bind either mammalian or insect ERR proteins.

Conclusion
In the present study, we observed that transgenic flies expressing an ERR gene with the amino acid G insertion in the resultant ERR protein caused the constitutive over-expression of the P450s, Cyp12d1, Cyp6g2 and Cyp9c1, all known to be involved in xenobiotic responses.The mechanism of how this ERR G insertion influences Cyp12d1, Cyp6g2 and Cyp9c1 expression, potentially through the change of ERR protein structure, remains to be experimentally determined as does the role of native copies of this allele across D. melanogaster strains.The fact that the G insertion can be found across many other Drosophila species in the Sophophora subgenera suggests that either this insertion is very old evolutionally, or it has occurred independently multiple times.However, the precise role or selective advantage, if any, that this allele may play in populations of D. melanogaster, and other species of Drosophila, remains to be determined.

Fig 2 .
Fig 2. Expression of P450s in transgenic flies.Expression of five P450 genes in w 1118 transgenic flies expressing ERRa or ERRb cloned from 91-R (G insertion) or Canton-S (wt) strains not crossed with a driver (light gray) or the progeny of a cross with the driver strain (darker gray).(A) ERRa from 91-R resulted in three P450s being significantly over-transcribed after a Bonferonni correction.(B) ERRb from 91-R resulted in two P450s being significantly overtranscribed after a Bonferonni correction.(C, D) Respectively ERRa and ERRb from Canton-S did not result in any of the P450s being significantly overtranscribed after a Bonferonni correction.doi:10.1371/journal.pone.0118779.g002

Fig 3 .
Fig 3. Comparison of ERR genomic sequences.Sequencing of ERR, corresponding to the amino acid range 251 to 310, revealed three ERR variants across 57 tested Drosophila melanogaster strains for alleles: (1) without the G insertion, (2) with the G insertion, and (3) with a deletion of seven codons plus the G insertion.Additionally, strains heterozygous for G insertion are designated with # and compound heterozygous (del+G / wt+G) strains are shown with *. doi:10.1371/journal.pone.0118779.g003

Fig 4 .
Fig 4. Protein structure comparisons.Predicted structures of ERRa and ERRb proteins superimposed between wt and G insertion.(A) Superimposed ERRa structures.(B) Superimposed ERRb structures.The orange color represents the wt ERR and green color represents ERR with a G insertion.The arrow points to the G insertion.The protein structure and Protein Data Bank (PDB) files were generated by using Swiss model First Approach mode (http:// swissmodel.expasy.org/workspace/).The PDB files were used as input at http://mspc.bii.a-star.edu.sg/minhn/click.html to generate the superimposed structure images [22].doi:10.1371/journal.pone.0118779.g004

Fig 5 .
Fig 5. Clustal Omega Alignment of ERR.The ligand binding domain region of Drosophila species were compared.The sequences were downloaded from Hierarchical Catalog of Orthologs.Only the wt allele was included for Drosophila melanogaster.doi:10.1371/journal.pone.0118779.g005

Fig 6 .
Fig 6.Molecular Phylogenetic analysis.The evolutionary history was inferred by using the Maximum Likelihood method based on the JTT matrix-based model[23].The tree with the highest log likelihood (-2145.5939) is shown.Initial tree(s) for the heuristic search were obtained automatically by applying Neighbor-Join and BioNJ algorithms to a matrix of pairwise distances estimated using a JTT model, and then selecting the topology with superior log likelihood value.The tree is drawn to scale, with branch lengths measured in the number of substitutions per site.All positions containing gaps and missing data were eliminated.There were a total of 455 positions in the final dataset.Evolutionary analyses were conducted in MEGA5.2[24].

Table 1 .
Expression levels of ERRa and ERRb in three Drosophila strains by absolute RT-qPCR.