Structural and Functional Characterization of One Unclassified Glutathione S-Transferase in Xenobiotic Adaptation of Leptinotarsa decemlineata

Arthropod Glutathione S-transferases (GSTs) constitute a large family of multifunctional enzymes that are mainly associated with xenobiotic or stress adaptation. GST-mediated xenobiotic adaptation takes place through direct metabolism or sequestration of xenobiotics, and/or indirectly by providing protection against oxidative stress induced by xenobiotic exposure. To date, the roles of GSTs in xenobiotic adaptation in the Colorado potato beetle (CPB), a notorious agricultural pest of plants within Solanaceae, have not been well studied. Here, we functionally expressed and characterized an unclassified-class GST, LdGSTu1. The three-dimensional structure of the LdGSTu1 was solved with a resolution up to 1.8 Å by X-ray crystallography. The signature motif VSDGPPSL was identified in the “G-site”, and it contains the catalytically active residue Ser14. Recombinant LdGSTu1 was used to determine enzyme activity and kinetic parameters using 1-chloro-2, 4-dinitrobenzene (CDNB), GSH, p-nitrophenyl acetate (PNA) as substrates. The enzyme kinetic parameters and enzyme-substrate interaction studies demonstrated that LdGSTu1 could catalyze the conjugation of GSH to both CDNB and PNA, with a higher turnover number for CDNB than PNA. The LdGSTu1 enzyme inhibition assays demonstrated that the enzymatic conjugation of GSH to CDNB was inhibited by multiple pesticides, suggesting a potential function of LdGSTu1 in xenobiotic adaptation.


Introduction
Glutathione S-transferases (GSTs) constitute a large superfamily of multifunctional enzymes that are ubiquitously present in both prokaryotes and eukaryotes [1][2][3][4]. In general, GSTs catalyze the conjugation of the reduced glutathione (GSH)-a nucleophilic tripeptide comprised of three amino acids: cysteine, glutamic acid, and glycine-to a wide range of substrates that have an electrophilic carbon, nitrogen, or sulfur atom [5,6]. The GST substrates can be natural or artificial compounds including cancer chemotherapeutic agents, carcinogens, pesticides, environmental pollutants, and byproducts of oxidative stress [4,6]. In addition, GSTs are capable of binding numerous endogenous and exogenous compounds by non-catalytic interactions, which are associated with their functions in sequestration, storage, or transportation [3,6,7]. cal adaptation pathways will help us understand mechanisms of stress adaptation in the global agricultural pest CPB [15,31,36,37].
With about 30 crystal structures of insect GSTs having been solved, there is no crystal structure of a beetle GST available (Table S1). Here, we solved the LdGSTu1 co-crystal structure with its nucleophilic substrate GSH by using X-ray diffraction. The three-dimensional structure of the LdGSTu1 was solved with a resolution up to 1.8 Å by X-ray crystallography. A typical GST global fold and an active site composed of two substrate binding sites, the "G-site" and the "H-site" were identified. The LdGSTu1 enzyme kinetic parameters and enzyme-substrate interaction studies demonstrated that the conjugation of GSH to CDNB could be inhibited by multiple pesticides, suggesting a potential function of LdGSTu1 in pesticide adaptation.

Phylogenetic Relationship of LdGSTu1 with Other Insect GSTs
The LdGSTu1 gene was cloned from the L. decemlineata susceptible and resistant strains and shared 100% sequence similarity with the gene XP_023027125.1 in NCBI database. Sequence analysis showed that the ORF was 693 bp, encoding a deduced polypeptide of 230 amino acids. The predicted molecular weight of LdGSTu1 was 26.4 kDa and isoelectric point was 5.36. The phylogenetic tree was constructed by the maximum likelihood method using the deduced amino acid sequences to investigate the evolutionary relationships of LdGSTu1 and 31 GSTs from L. decemlineata and other insect species (Figure 1, Table S2). Phylogenetic tree showed that the GSTs from the same class were grouped together ( Figure 1). As expected, LdGSTu1 was clustered in the unclassified clade with other five unclassified GSTs identified from Anoplophora glabripennis, D. mauritiana, D. mojavensis, B. mori and Sitophilus oryzae (Figure 1, Table S2). LdGSTu1 originated from the same evolutionary root with SoGST1-X2 from S. oryzae with the bootstrap value of 75 ( Figure 1).

X-ray Crystal Structure of LdGSTu1 in Complex with GSH
LdGSTu1 crystalized in space group P2 with a unit cell of a = 58.45, b = 46.44, and c = 87. 19. The LdGSTu1 structure was refined to a resolution of 1.80 Å. Two monomers of LdGSTu1 were in the crystal asymmetric unit. One monomer, Chain A exhibited glutathione (GSH) bound to the active site ( Figure 2). Whereas the other monomer (Chain B) had no bound GSH molecule ( Figure S1). Data collection and refinement statistics are listed in Table 1.

X-Ray Crystal Structure of LdGSTu1 in Complex with GSH
LdGSTu1 crystalized in space group P2 with a unit cell of a = 58.45, b = 46.44, and c = 87.19. The LdGSTu1 structure was refined to a resolution of 1.80 Å. Two monomers of LdGSTu1 were in the crystal asymmetric unit. One monomer, Chain A exhibited glutathione (GSH) bound to the active site ( Figure 2). Whereas the other monomer (Chain B) had no bound GSH molecule ( Figure S1). Data collection and refinement statistics are listed in Table 1. Ag, Anopheles gambiae/Anoplophora glabripennis; Am, Apis mellifera; Bm, Bombyx mori; Dm, Drosophila mauritiana/Drosophila mojavensis/Drosophila melanogaster; Ld, Leptinotarsa decemlineata; Lm, Locusta migratoria; Nv, Nasonia vitripennis; Sl, Spodoptera litura; So, Sitophilus oryzae. The red star indicates LdGSTu1. 3 10 -2 (residues 43-48), β3 (residues 56-59), β4 (residues 62-64), and α2 (residues 67-77). Then the linker coil (residues 78-88) connects the N-term domain to the C-term domain. The C-term domain consists of five helices, α3 (residues 89-119), α4 (residues 126-146), α5 (residues 158-173), α6 (residues 181-193), and α7 (residues 195-211) (Figures 2a and 3). A substrate binding pocket is located between the N-term domain and the C-term domain. GSTs have been previously described as having two binding sites within the substrate binding pocket, the "G-site" which binds GSH and "H-site" which binds a hydrophobic co-substrate [42]. A molecular lipophilicity potential surface reveals that the LdGSTu1 substrate binding pocket has a more hydrophilic region on the N-term domain side of the binding pocket, which is bound with GSH and makes up the "G-site", and a more lipophilic region in the substrate binding pocket adjacent to the bound GSH on the C-term domain side of the pocket making up the "H-site" (Figure 2b).  The N-term domain exhibits the thioredoxin-like fold (residues 1-77) and was complexed with GSH in chain A of the crystal structure. Helices are depicted in cornflower blue, coils in powder blue, and β-strands in yellow. The linker coil (residues 78-88) connecting the polypeptide of the N-term domain to the polypeptide of the C-term domain is depicted in orange. The C-term domain is mainly helical in nature, consisting of five α-helices (medium blue) with coils depicted in powder blue; (b) MLP surface diagram of LdGSTu1 monomer with bound GSH in active site. The color scale is cyan for most hydrophilic and passes through white to golden rod for most hydrophobic. In the active site the most hydrophilic location was the G-site with GSH bound and the most hydrophobic region of the active site was the H-site on the C-term side of the active site. The red and purple dashed lines outline the G-site and H-site, respectively.    . Structural sequence alignment of LdGSTu1 with highest sequence identity insect GSTs deposited in PDB. Chimera was used to superpose LdGSTu1 crystal structure with 5ZFG, 1PN9, 1JLV, and 3WYW to generate a multiple structure sequence alignment. Sequence amino acids are highlighted black for 80% identity and gray for 80% similarity. Secondary structural element positions are mapped and shown with yellow arrows for β-strands, and blue cylinders for helices. Domain distribution is depicted as boxed green for the N-term domain and boxed in red for the C-term domain. The linker region is mapped with orange equal signs. G-site residues are highlighted green and marked with a G. H-site residues are highlighted blue and marked with an H.

Active Site of LdGSTu1
The active site of LdGSTu1 is formed by a pocket consisting of a "G-site", which binds GSH and an "H-site" which binds a hydrophobic substrate, consistent with previously insect GST studies ( Figure 2b) [40,42]. The "G-site" is depicted in Figure 4a with all side chains of all amino acids within 4.5 Å. The GSH is bound to LdGSTu1 via an extensive hydrogen bonding network that includes amino acid side chains, backbone carbonyls, backbone amides and crystallographic water. Ile54 and Ser64 form direct hydrogen bonding interactions with GSH ( Figure 4a). Ser14, Pro55, Asn68, and Glu66 form hydrogen bonds bridged by crystallographic water molecules with GSH. The presumptive catalytic Ser14 formed a hydrogen bond with the thiol group of GSH bridged via a water molecule. Additionally, the side chain OE1 and OE2 atoms of Glu66 formed an ionic interaction with N1 atom of the bound GSH. The "H-site" of LdGSTu1 is largely hydrophobic (Figure 2b) and consists of Tyr107, Ser111, Leu115, Ala116, Phe119, Phe120, Leu208, Ala210, and Phe211 ( Figure 4b). . Structural sequence alignment of LdGSTu1 with highest sequence identity insect GSTs deposited in PDB. Chimera was used to superpose LdGSTu1 crystal structure with 5ZFG, 1PN9, 1JLV, and 3WYW to generate a multiple structure sequence alignment. Sequence amino acids are highlighted black for 80% identity and gray for 80% similarity. Secondary structural element positions are mapped and shown with yellow arrows for β-strands, and blue cylinders for helices. Domain distribution is depicted as boxed green for the N-term domain and boxed in red for the C-term domain. The linker region is mapped with orange equal signs. G-site residues are highlighted green and marked with a G. H-site residues are highlighted blue and marked with an H.

Active Site of LdGSTu1
The active site of LdGSTu1 is formed by a pocket consisting of a "G-site", which binds GSH and an "H-site" which binds a hydrophobic substrate, consistent with previously insect GST studies ( Figure 2b) [40,42]. The "G-site" is depicted in Figure 4a with all side chains of all amino acids within 4.5 Å. The GSH is bound to LdGSTu1 via an extensive hydrogen bonding network that includes amino acid side chains, backbone carbonyls, backbone amides and crystallographic water. Ile54 and Ser64 form direct hydrogen bonding interactions with GSH ( Figure 4a). Ser14, Pro55, Asn68, and Glu66 form hydrogen bonds bridged by crystallographic water molecules with GSH. The presumptive catalytic Ser14 formed a hydrogen bond with the thiol group of GSH bridged via a water molecule. Additionally, the side chain OE1 and OE2 atoms of Glu66 formed an ionic interaction with N1 atom of the bound GSH. The "H-site" of LdGSTu1 is largely hydrophobic (Figure 2b) and consists of Tyr107, Ser111, Leu115, Ala116, Phe119, Phe120, Leu208, Ala210, and Phe211 ( Figure 4b).

LdGSTu1 Enzyme Inhibition Assay
To test the interaction of LdGSTu1 with the known GST inhibitor ethacrynic acid (EA) and multiple pesticides (carbaryl, diazinon, imidacloprid, acetamiprid, chlorpyrifos, and thiamethoxam), inhibition assays were conducted by measuring change to the rate of GSH conjugation with CDNB ( Figure 6, Table 2). In the case of LdGSTu1, ethacrynic acid acted as inhibitor of GSH enzyme catalyzed conjugation to CDNB at µM concentrations, consistent with previous GST inhibition studies [44]. At a concentration of 40 µM EA, LdGSTu1 residual activity was 88.8%; at 200 µM EA, LdGSTu1 residual activity was 49.6%; and at 1mM EA, LdGSTu1 residual activity was 0.0%. Compared to EA, the inhibitory effect of the pesticides screened was relatively lower. At 40 µM, none of the pesticides showed significant inhibitory effect on the enzymatic conjugation of GSH to CDNB. However, at increasing concentrations of pesticides, the inhibitory effects became significant ( Figure 6). For the LdGSTu1, GSH catalyzed conjugation of CDNB in the presence of 1 mM acetamiprid, 1 mM carbaryl, 1 mM diazinon, 1 mM chlorpyrifos, 1 mM imidacloprid, and 1 mM thiamethoxam, the residual enzyme activity fell to 81.0%, 88.5%, 88.5%, 89.9%, 93.7%, 95.0%, respectively. In the presence of 5 mM acetamiprid, 5 mM diazinon, 5 mM chlorpyrifos, 5 mM imidacloprid, and 5 mM thiamethoxam, the residual enzyme activity was 39.1%, 75.3%, 70.5%, 66.4%, and 72.3%, respectively. Carbaryl was not included in 5 mM grouping due to insolubility and EA was not included in 5 mM grouping because at 1 mM EA residual activity already fell to 0%. These results demonstrated that the enzymatic conjugation of GSH to CDNB could be inhibited by multiple pesticides, suggesting these pesticides are potential substrates of LdGSTu1.

LdGSTu1 Ligand Docking
LdGSTu1 was docked with the ligands CDNB, EA, carbaryl, diazinon, imidacloprid, acetamiprid, and thiamethoxam to test their binding affinities. Binding poses for all ligands screened were found and ranked according to affinity dG scores. The highest ranked poses for each ligand were used for further analysis. All of the ligands docked into the presumptive hydrophobic binding site "H" of LdGSTu1 adjacent to the GSH binding site "G" and the co-crystal complexed GSH molecule with favorable calculated binding energies ( Figure S2). For CDNB, EA, carbaryl, diazinon, imidacloprid, acetamiprid, and thiamethoxam, the calculated binding energies for top poses by the affinity dG scoring function were −4.4 kcal/mol, −5.5 kcal/mol, −4.6 kcal/mol, −4.9 kcal/mol, −4.6 kcal/mol, −5.4 kcal/mol, and −5.2 kcal/mol, respectively. For the docked ligand poses with LdGSTu1, closest atom distances for ligand to GSH are listed. The CDNB ligand docked with carbon 3 positioned 3.77 Å from the glutathione sulfur atom. EA docked with carbon 2 and 3 at 4.39 Å from the glutathione sulfur atom. For carbaryl, the ligand docked with its carbonyl carbon positioned 4.01 Å from the glutathione sulfur atom. Diazinon docked with the pyrimidine ester carbon located 2.74 Å from the glutathione sulfur atom. Imidacloprid docked with C3 on the pyridine ring located 3.9 Å from the sulfur atom of glutathione. Docked acetamiprid had C7 positioned at 3.7 Å from the sulfur of GSH. Lastly, thiamethoxam docked with its sulfur atom 3.8 Å from the sulfur of GSH. Molecular docking with the crystal structure of LdGSTu1 gave favorable binding poses for all the ligands screened. Additionally, the predicted binding locations for all the highest ranked ligand poses were localized the hydrophobic binding pocket in the active site of LdGSTu1 and adjacent to the co-crystalized position of GSH ( Figure S2). The LdGSTu1 ligand docking results suggested that LdGSTu1 is capable of binding pesticides tested, further suggesting these pesticides are potential substrates of LdGSTu1.
LdGSTu1 residual activity was 88.8%; at 200 μM EA, LdGSTu1 residual activity was 49.6%; and at 1mM EA, LdGSTu1 residual activity was 0.0%. Compared to EA, the inhibitory effect of the pesticides screened was relatively lower. At 40 μM, none of the pesticides showed significant inhibitory effect on the enzymatic conjugation of GSH to CDNB. However, at increasing concentrations of pesticides, the inhibitory effects became significant ( Figure 6). For the LdGSTu1, GSH catalyzed conjugation of CDNB in the presence of 1 mM acetamiprid, 1 mM carbaryl, 1 mM diazinon, 1 mM chlorpyrifos, 1 mM imidacloprid, and 1 mM thiamethoxam, the residual enzyme activity fell to 81.0%, 88.5%, 88.5%, 89.9%, 93.7%, 95.0%, respectively. In the presence of 5 mM acetamiprid, 5 mM diazinon, 5 mM chlorpyrifos, 5 mM imidacloprid, and 5 mM thiamethoxam, the residual enzyme activity was 39.1%, 75.3%, 70.5%, 66.4%, and 72.3%, respectively. Carbaryl was not included in 5 mM grouping due to insolubility and EA was not included in 5 mM grouping because at 1 mM EA residual activity already fell to 0%. These results demonstrated that the enzymatic conjugation of GSH to CDNB could be inhibited by multiple pesticides, suggesting these pesticides are potential substrates of LdGSTu1. Statistical significance of residual enzyme activity among treatments was calculated using one-way ANOVA, followed by a Turkey's HSD multiple comparation test. There was no significant difference among treatments with the same alphabetic letters (e.g., a, b, and c).

mRNA Expression Patterns of LdGSTu1
The LdGSTu1 expression levels in an insecticide resistant strain and the susceptible strain were examined. As shown in Figure 7, the relative expression level of LdGSTu1 was significantly higher in the resistant strain than the susceptible one. We then further investigated the temporal and spatial expression patterns of LdGSTu1 in both insecticide resistant and susceptible strains. The results in Figure 8a showed that during different developmental stages, LdGSTu1 was expressed at the highest level during the adults (both males and females) in the insecticide resistant strain than other stages. There were significant differences in adult expression between resistant strain and susceptible strain. The tissue expression profiles showed that LdGST1 had the highest expression levels in the head of resistant strain, followed by midgut, Malpighian tubule, and ovary in the resistant strain. There were significantly differences in head and midgut expression between resistant strain and susceptible strain (Figure 8b).    Each value is the mean ± SE of three collections. Differences among multiple treatments were analyzed using one-way ANOVA, followed by a Tukey's HSD multiple comparison test. There was no significant difference among treatments with the same alphabetic letters (e.g., a, b, and c).

Discussion
LdGSTu1 crystalizes with a dimer in the asymmetric unit ( Figure S1). The dimer is a result of crystal packing and not expected to be a biologically relevant dimer assembly as the active sites are solvent exposed and not internal to the dimer interface as reported for biological dimer assemblies in previous GST crystal structure studies [17,40,42,45] ( Figure  S1). However, the differences between the two monomers making up the crystallographic dimer are interesting ( Figure S1). The active sites of chain A has the bound GST cofactor GSH, but chain B did not have a bound GSH molecule. Chain A complexed with GSH has an open active site similar to previously published GST structures, whereas the chain B monomer vacant of GSH has a more closed active site, suggesting flexibility in loop-helixloop region in the active site of the N-term domain of GSTs.
Our data showed that the crystal structure of LdGSTu1 exhibited a bound GSH ligand in the "G-site" of chain A. That bound GSH revealed the hydroxyl of Ser14 to be hydrogen bonded to the thiol of GSH via a water bridge ( Figure 4A), suggesting that Ser14 is a residue responsible for catalytically activating GSH in LdGSTu1. The only other unclassified insect GST with a published crystal structure in the PDB (5ZFG) was in apoform but also posed a crystallographic water hydrogen bonded the hydroxyl of Ser14 in BmGSTu2 [40]. Previously, characterized and classified GSTs have been shown to poses a catalytically active serine, tyrosine, or cysteine in their active sites [46]. The catalytic residue activates the glutathione thiol group through hydrogen bonding. Moreover, the Each value is the mean ± SE of three collections. Differences among multiple treatments were analyzed using one-way ANOVA, followed by a Tukey's HSD multiple comparison test. There was no significant difference among treatments with the same alphabetic letters (e.g., a, b, and c).

Discussion
LdGSTu1 crystalizes with a dimer in the asymmetric unit ( Figure S1). The dimer is a result of crystal packing and not expected to be a biologically relevant dimer assembly as the active sites are solvent exposed and not internal to the dimer interface as reported for biological dimer assemblies in previous GST crystal structure studies [17,40,42,45] (Figure S1). However, the differences between the two monomers making up the crystallographic dimer are interesting ( Figure S1). The active sites of chain A has the bound GST cofactor GSH, but chain B did not have a bound GSH molecule. Chain A complexed with GSH has an open active site similar to previously published GST structures, whereas the chain B monomer vacant of GSH has a more closed active site, suggesting flexibility in loop-helix-loop region in the active site of the N-term domain of GSTs.
Our data showed that the crystal structure of LdGSTu1 exhibited a bound GSH ligand in the "G-site" of chain A. That bound GSH revealed the hydroxyl of Ser14 to be hydrogen bonded to the thiol of GSH via a water bridge (Figure 4a), suggesting that Ser14 is a residue responsible for catalytically activating GSH in LdGSTu1. The only other unclassified insect GST with a published crystal structure in the PDB (5ZFG) was in apo-form but also posed a crystallographic water hydrogen bonded the hydroxyl of Ser14 in BmGSTu2 [40]. Previously, characterized and classified GSTs have been shown to poses a catalytically active serine, tyrosine, or cysteine in their active sites [46]. The catalytic residue activates the glutathione thiol group through hydrogen bonding. Moreover, the unclassified insect GSTs display the sequence motif VSDGPPSL in the "G-site", which contains Ser14 (Figure 9). In the study of BmGSTu2 (PDB: 5ZFG), the authors created a mutant P13A swapping Pro13 for an Ala [40]. It was found that the P13A mutant exhibited decreased specific activities for both CDNB and diazinon [40]. As proline residues are frequently found to be conserved in and around enzyme active sites, it can be inferred that proline residues are important for maintaining positional orientation of catalytically relevant residues. As the mutant P13A residue in BmGSTu2 directly precedes Ser14, we suggest the mutation may have caused a local structural positioning shift for Ser14, thus effecting the ability of BmGSTu2 to activate the thiol of GSH, further suggesting the catalytic importance of Ser14 in unclassified insect GSTs. frequently found to be conserved in and around enzyme active sites, it can be inferred that proline residues are important for maintaining positional orientation of catalytically relevant residues. As the mutant P13A residue in BmGSTu2 directly precedes Ser14, we suggest the mutation may have caused a local structural positioning shift for Ser14, thus effecting the ability of BmGSTu2 to activate the thiol of GSH, further suggesting the catalytic importance of Ser14 in unclassified insect GSTs.  (Table  S3). Sequence logo was generated with WebLogo 3 (http://weblogo.threeplusone.com/manual.html, accessed on 2 November 2021).
GST-mediated xenobiotic adaptation is through direct metabolism or sequestration of xenobiotics, and/or indirectly by providing protection against oxidative stress induced by xenobiotic exposure [15,40,41]. The GSTs being upregulated in insecticide-resistant insects have been previously reported in BmGSTu2, in which transcript expression was induced 1.7-fold in a resistant strain of B. mori [47]. Additionally, at the protein level, increased GST activity was observed in insecticide resistant insects, such as an imidaclopridresistant Nilaparvata lugens [48] and an abamectin-resistant Liriomyza sativae [49]. Based on previous reports on overexpression of GSTs in insecticide resistance and increased GST activity, it was inferred that LdGSTu1 may play a role in insecticide resistance in CPB, as it is overexpressed in the insecticide resistant strain (Figure 7). Tissue expression profile analysis showed that LdGSTu1 expressed at the highest level in the head of resistant strain ( Figure 8B). Since the head or central nervous system is critical organ for insect survival and serves as the target for numerous neurotoxic pesticides [50,51], the high expression of LdGSTu1 implies its potential primary functions in xenobiotic adaptation. Figure 9. Sequence logo of amino acids in the substrate binding pocket of unclassified insect GSTs. Sequence identity is represented by the letter size. Amino acid number and type label on x-axis is representative of the LdGSTu1 sequence. The sequence logo is overlayed on a surface representation of LdGSTu1. The G-site is colored in cornflower blue and the H-site is colored in red. Residues in the sequence logo are separated by signature motifs identified in the substrate binding pocket region of LdGSTu1 upon multiple alignment with other unclassified insect GSTs. Sequences used for the multiple sequence alignment are the unclassified GST sequences from phylogenic analysis (Table S3). Sequence logo was generated with WebLogo 3 (http://weblogo.threeplusone.com/manual.html, accessed on 2 October 2021).
GST-mediated xenobiotic adaptation is through direct metabolism or sequestration of xenobiotics, and/or indirectly by providing protection against oxidative stress induced by xenobiotic exposure [15,40,41]. The GSTs being upregulated in insecticide-resistant insects have been previously reported in BmGSTu2, in which transcript expression was induced 1.7-fold in a resistant strain of B. mori [47]. Additionally, at the protein level, increased GST activity was observed in insecticide resistant insects, such as an imidacloprid-resistant Nilaparvata lugens [48] and an abamectin-resistant Liriomyza sativae [49]. Based on previous reports on overexpression of GSTs in insecticide resistance and increased GST activity, it was inferred that LdGSTu1 may play a role in insecticide resistance in CPB, as it is overexpressed in the insecticide resistant strain (Figure 7). Tissue expression profile analysis showed that LdGSTu1 expressed at the highest level in the head of resistant strain (Figure 8b). Since the head or central nervous system is critical organ for insect survival and serves as the target for numerous neurotoxic pesticides [50,51], the high expression of LdGSTu1 implies its potential primary functions in xenobiotic adaptation.
Our LdGSTu1 kinetic enzyme studies showed that LdGSTu1 displayed a higher catalytic efficiency for CDNB than PNA (Table 2). LdGSTu1 enzyme inhibition assay showed that ethacrynic acid and the pesticides carbaryl, diazinon, imidacloprid, acetamiprid, chlorpyrifos, and thiamethoxam acted as inhibitors of the enzyme catalyzed conjugation of GSH to CDNB ( Figure 6, Table 2). Functional studies have previously shown insect GSTs to be associated with adaptation to plant allelochemicals and insecticides by means of direct metabolism or defense against reactive oxygen species (ROS) [15,47]. In our study, neither HNE nor T2H were conjugated to GSH enzymatically by LdGSTu1 (Table 2). This result is consistent with bmGSTu2 and pxGSTu1, unclassified GSTs identified in silkworm [47] and diamondback moth [52], respectively.
In summary, we identified a beetle GST, LdGSTu1 belonging to the unclassified class of insect GSTs and characterized the structure and function of LdGSTu1 through X-ray crystallography, enzyme activity and binding studies. LdGSTu1 crystal structure exhibits a typical GST global fold and an active site composed of two substrate binding sites, the "G-site" and the "H-site". The signature motif VSDGPPSL was identified, and it contained the catalytically active residue Ser14. The enzyme kinetic parameters and enzyme-substrate interaction studies demonstrated that LdGSTu1 could be inhibited by multiple pesticides tested; thus, it can be potentially involved in Colorado potato beetle resistance to insecticides. Further investigation is on the way to identify putative catalytic active residues through site-directed mutagenesis along with continuing enzyme activity and binding studies to identify if LdGSTu1 only binds pesticides or whether it can also metabolize them.

Insects
The susceptible CPB was purchased from French Ag Research, Inc.

LdGSTu1 Cloning, Bioinformatics, and Phylogenetic Analysis
The LdGSTu1 cloning was performed with a ligation-independent cloning strategy following a previous protocol [53]. Briefly, the full-length LdGSTu1 was amplified from resistant CPB cDNA using PCR with the primers containing ligation independent cloning F and R sites (Table S2), T4 polymerase treated, and then annealed with T4 polymerase treated pET-9Bc vector. The pET C-terminal TEV His6 cloning vector with BioBrick polycistronic restriction sites (9Bc) was a gift from Scott Gradia (Addgene plasmid #48285; http://n2t.net/addgene:48285, accessed on 2 October 2021; RRID:Addgene 48285). Then the products were transformed into DH 5α competent cells. Positive colonies were verified using T7 primers, then cultured in liquid LB overnight at 37°C. The plasmids were extracted and identified, and sequenced by Functional Biosciences, Inc. The cloned sequences were submitted to the website of National Center for Biotechnology Information (NCBI) (https://www.ncbi.nlm.nih.gov/, accessed on 2 October 2021). The conserved domains were detected using bioinformatics tools on the NCBI server. The theoretical isoelectric point (pI) and molecular weight (MW) were computed using the Compute pI/Mw tool (https://web.expasy.org/compute_pi/, accessed on 2 October 2021). To classify the GST gene, the phylogenetic tree was constructed with Muscle and MEGA X using the maximum likelihood, LG model, gamma distributed method with 1000 bootstrap replicates [54,55]. The available amino acid sequences of GSTs used in the phylogenetic analysis were downloaded from the NCBI database [56]. Multiple alignment analysis was also conducted with several GTSs from different insects by DANMAN v. 6.03 (Lynnon BioSoft, Vaudreuil, Quebec, CA, USA).

LdGSTu1 Protein Expression and Purification
The pET-9Bc-LdGSTu1 plasmids were transformed into Rosetta TM II (DE3) pLysS, positive colonies were verified with PCR. The successful inserts were grown in 50 mL LB cultures at 37 • C in a MaxQ 6000 Incubated Stackable Floor Shaker (Thermo Scientific, Marietta, OH, USA) for induction testing. Cell stocks positive for LdGSTu1 expression were frozen at −80 • C for later use. For expression, overnight 50 mL cultures grown at 37 • C in terrific broth with 1× ampicillin and chloramphenicol, and then used to inoculate a 1.2 L Terrific Broth (TB) culture, incubated at 37 • C until OD 600 reached 0.4~0.6. Then the culture was cooled and induced with 0.5 mM IPTG and incubated at 20 • C for an additional 20 h. Cells were harvested by centrifugation at 4000 rpm and 4 • C to acquire cell pellet in a tabletop centrifuge (Thermo Sorvall Legend XTR refrigerated centrifuge, Langenselbold Germany). The cells were lysed with buffer containing 50 mM NaPi, 300mM NaCl, 20 mM Imidazole, and 1 mM PMSF (pH: 7.6) for 5 cycles of 30 s at 70% power 5-7 times using a sonicator (Branson Digital Sonifier SFX 150, Emerson Electric Co., St. Louis, MO, USA) on ice. After cell lysis, the homogenate was centrifuged at 18,000 g to separate soluble protein from the insoluble fractions. The soluble fraction was added to a Ni-NTA column. The Ni-NTA column was washed with 50 mM NaPi, 300 mM NaCl, 20 mM Imidazole (pH: 7.6). Next, LdGSTu1 protein was eluted with buffer containing 20 mM NaPi, 300 mM NaCl, and 250 mM Imidazole (pH: 8.0). Protein was then 100 × fold buffer exchanged into buffer containing 5 mM NaPi, 5 mM HEPES (pH: 7.2). Next, protein was further purified by a Hydroxyapatite column (HA) connected to an NGC Medium-Pressure Liquid Chromatography System (Bio-Rad Laboratories, Hercules, CA, USA). A gradient form 5 mM NaPi, 5 mM HEPES (pH: 7.2) to 500 mM NaPi (pH: 7.2) was used to wash and elute LdGSTu1.
Fractions containing LdGSTu1 were visualized with SDS-PAGE, combined, and concentrated. Then concentrated fractions were applied to SEC Enrich 650 (Bio-Rad Laboratories, Hercules, CA, USA) connected to the NGC. Size exclusion buffer was 20 mM HEPES, 1 mM DTT, and 1 mM EDTA, 20 mM GSH pH 7.2. Fractions were checked with SDS-PAGE. Purified LdGSTu1 protein concentrations were calculated using the Bradford assay or UV280 methods on NanoDrop One (Thermo Scientific, Madison, WI, USA) and Spark®multi-mode plate reader (Tecan Austria GmbH, Untersbergstr, Austria). Protein was concentrated and used directly for crystallization or buffer exchanged for enzyme assays.

Crystal Data Collection, Refinement, and Structural Analysis
LdGSTu1 crystals were grown by sitting drop vapor diffusion at 18 • C. High-purity LdGSTu1 at 20mg/mL was mixed 1:1 with reservoir solution in sitting drop well and incubated against mother liquor reservoir solution (100 mM MES pH 6.5, 100 mM NaCl, 25% PEG 4K). LdGSTu1 crystal data were collected at the Macromolecular X-ray science at the Cornell High Energy Synchrotron Source (MacCHESS) beamline 7B2. The software package XDS was used for data processing [57]. Phasing was done by using BmGSTu1 (PDB: 5ZFG) as a search model in PHENIX Phaser [58]. Refinement and model building were performed by using PHENIX and Coot [58][59][60]. Search models for molecular replacement were identified by a NCBI blastp with the LdGSTu1 amino acid sequence search against the Protein Data Bank (PDB) database [38,39]. Structural analysis and figures of LdGSTu1 were conducted by using UCSF Chimera, UCSF Chimera X, and Coot [61][62][63]. The coordinates and structure factors for the final model of LdGSTu1 and GSH was deposited in the PDB under accession code 7RKA.

Enzyme Assay
The kinetic analysis of LdGSTu1 was conducted by steady state with varied concentrations of substrates 1-chloro-2,4-dinitrobenzene (CDNB) from 0.05 to 3 mM and p-nitrophenyl acetate (PNA) from 0.2 to 3.2 mM, while holding the GSH concentration constant at 5 mM, and for varied concentrations of GSH at 0.125 to 5mM while holding CDNB at a constant concentration of 2 mM. The reaction buffer was 100 mM KPi (pH: 6.5). Reactions were carried out in 96-well Greiner Bio-One UV-Star ® microplates (Sigma-Aldrich, St. Louis, MO, USA). Assays were run on a Spark ® multi-mode plate reader (Tecan Austria GmbH, Untersbergstr, Austria) in the kinetic mode, for a continuous read assay for 3 min. Product concentrations were calculated by path-length corrected molar attenuation coefficient from Habig et al. [64,65]. Kinetic parameters and plots were calculated and generated with GraphPad Prism (GraphPad, San Diego, CA, USA).

Enzyme Inhibition Assay
Residual enzyme activity in the presence of inhibitors was measured by individually incubating LdGSTu1 (50 µg) with inhibitors ethacrynic acid (EA) and carbaryl, diazinon, imidacloprid, acetamiprid, chlorpyrifos, and thiamethoxam at 40 µM, 200 µM, 1 mM, and 5 mM for 10 min at 30 • C followed by addition of GSH (0.5 mM) and CDNB (0.5 mM), for total reaction volume of 200 µL in 96-well Greiner Bio-One UV-Star ® microplates (Sigma-Aldrich, St. Louis, MO, USA). After addition of GSH and CDNB the change in absorbance at a wavelength of 340 nm was immediately measured on a Spark ® multimode plate reader (Tecan Austria GmbH, Untersbergstr, Austria) at 30 • C in kinetic mode for 70 s with 10 s reads. All inhibition and control reactions were run in triplicate. Residual activity was calculated as precent of enzyme activity retained in reaction in presence of inhibitor relative to control reaction without inhibitor and an equivalent amount of acetone. Additionally, controls without enzyme were run to control for non-enzymatic reaction contribution. Inhibitor stock solutions were prepared in acetone. For carbaryl reactions were only run at 40 µM, 200 µM, and 1 mM due to insolubility for the reaction conditions at 5 mM. Additionally, EA reactions were only run at 40 µM, 200 µM, and 1 mM due to having reached 100% inhibition at 1mM. The reaction buffer was 100 mM KPi at a pH of 6.5.

Docking of LdGSTu1 Crystal Structure with Xenobiotics
The LdGSTu1 crystal structure was prepared for docking studies with Molecular Operating Environment (MOE), 2020.9 (Chemical Computing Group, Montreal, QC, Canada). The refined LdGSTu1 crystal structure was prepared with the Structure Preparation, Protonate 3D, and Partial Charge applications. The molecular mechanics forcefield used was Amber10:EHT [66]. The ligand structures of CDNB, EA and carbaryl, diazinon, imidacloprid, acetamiprid, chlorpyrifos, and thiamethoxam were downloaded from PubChem ® database in SDF format [67]. Ligands were docked with the MOE Dock application using the prepared LdGSTu1 chain A structure containing complexed GSH as the receptor and the substrate binding pocket selected as ligand binding site. The method parameters selected were Triangle Matcher for placement with scoring function London dG, 30 poses, and refinement of Induced Fit with scoring function Affinity dG, 5 poses. Ligand docking runs gave 5 final poses for each ligand with favorable binding energy. The highest ranked docked ligand poses were selected for further analysis. Figure with docked ligands was generated with ChimeraX [62].

RNA Extraction, cDNA Synthesis and qRT-PCR Analysis
Total RNA was isolated from insect samples using TRIzol reagent (Thermo Fisher Scientific, Carlsbad, CA, USA). The total RNA was treated with DNase I (Ambion Inc., Austin, TX, USA) to remove contaminating genomic DNA. Approximately 3 µg of DNase Itreated RNA was reverse transcribed to cDNA using 10 µM of oligo (dT) primer and M-MLV reverse transcriptase (Promega, Madison, WI, USA) in a 20 µL reaction. NanoDrop One (Thermo Scientific, Madison, WI, USA) was used to perform spectroscopic quantification. The ef1α and rpl4 were used as reference genes for the qRT-PCR [31]. PCR conditions included 3 min at 95 • C followed by 39 cycles of 10 s at 95 • C, and 55 • C for 30 s. qRT-PCR was conducted with 1 µL cDNA, 5 µL FastStart SYBR Green Master (Roche Diagnostics, Indianapolis, IN, USA), 0.4 µL qRT-PCR primers (Table S2), and 3.6 µL ddH 2 O in a 10 µL total reaction volume using Bio-Rad CFX Connect™ Real-Time PCR System (Bio-Rad Laboratories, Hercules, CA, USA). The 2 −∆∆Ct method was used for the quantitative analysis. Three biological replications were conducted independently.

Developmental and Spatial Expression of LdGSTu1
Different developmental stages including eggs (first day and fourth day); larvae (1st-4th instar); pupae; and female and male adults were collected. At least 10 individuals were used for each stage. Different tissues including head (with antenna), × midgut, Malpighian tubule, fat body, and ovary from adult females were dissected in ice-cold 1× phosphate-buffered saline (PBS) solution. Each tissue, collected from at least 20 adults (sex ratio = 1:1), was pooled as one sample for RNA extraction and qRT-PCR analysis. Three independent replicates were performed.

Statistical Analyses
All statistical analyses were conducted using SPSS 20.0 (SPSS Inc., Chicago, IL, USA). Data were expressed as the mean ± standard error (SE) of three independent replicates. Differences in gene expression between two experimental treatments were analyzed using independent Student's t-test. Differences among multiple treatments were analyzed by oneway ANOVA, followed by Tukey's HSD multiple comparison test (p < 0.05). In Student's t-test, significance levels were denoted by * (0.01 < p < 0.05) and ** (p < 0.01).