Comparison between imidacloprid effects on AChE and nAChRα1 in target Aphis craccivora and non-target Apis mellifera: experimental and theoretical approaches

Abstract

Background

Neonicotinoids are widespread insecticides because of their potent effects against aphids and other piercing-sucking insects in addition to having high selectivity toward insects rather than vertebrates. However, they affect severely some non-target insects, mainly honeybee in a phenomenon called colony collapse disorder (CCD).

Results

Effects of imidacloprid (IMI), most used neonicotinoids, on aphid acetylcholinesterase (AChE), in vivo and in vitro were examined; besides, molecular modeling was used to investigate similarities and differences of AChE and nicotinic acetylcholine receptors α1-subunit (nAChRα1) in aphids, target insect, and honeybees, non-target insect. Results showed that aphid AChE was inhibited in vitro, with IC50 108.6 mg/L but not affected in vivo while the mortality was concentration-dependent with high toxicity (LC50 9.50 mg/L); in addition, aphid AChE was more inhibited, in vitro, but with much less effects, in vivo, than that of honeybees. These results indicate that AChE is not the main cause of the observed mortality, but it still has a role in insect resistance system with different responses in both insects. Molecular modeling showed high similarity in primary and secondary structures of AChE indicated by high identity (67%) and low gaps (1%); besides, the same template for both enzymes was auto-selected for homology. In addition, similar positions of the triad amino acids were found in AChE of both insects indicating high similarity. Conversely, the similarity in nAChRα1 in both insects is lower (50% identity and 9% gaps). These gaps (50 amino acids) are found in the intracellular large loop between TM3 and TM4 and account for the observed differences in the nAChRα1 binding sites of in both insects.

Conclusion

These observed variations in nAChRα1 structures and binding sites in different insect species can be used as good bases in designing new neonicotinoids that express high effects on target insects with better selectivity to minimize adverse effects on non-target organisms.

Graphical Abstract

Introduction

The use of chemicals in pest control increases continuously worldwide. Despite their great impact on enhancing crop production, they exert adverse effects on human and various environmental spheres e.g., soils, water and animals. Therefore, new chemicals are constantly introduced to the world market for the sake of finding out effective pesticides with minimum harmful effects. Neonicotinoids are group of insecticides recently emerged (1990s) and since then, their use has been expanded steadily to reach of 24% of the total used insecticides [22]. They are systematic insecticides effective against piercing-sucking insects at low concentrations compared to other insecticides, in addition to being less effective on vertebrates [14, 23] where their selectivity factor reaches 456, much higher than those of organophosphates (33), methylcarbamates (16) and organochlorines (91) [14]. Structurally, neonicotinoids are related to nicotine, which has also some insecticidal activity. Although they do not belong to one chemical group, they have the same toxicity mechanism and common structural features i.e., aromatic heterocycle, flexible methylene group, hydroheterocycle, guanidine or amidine group, and an electron-withdrawing group (e.w.g.) e.g., nitro or cyano group [5]. They are water-soluble compounds stable at physiological pH (5–7), but stability decreases out of this range. The e.w.g. presents in the form of nitromine (e.g., imidacloprid) or cyanomine (e.g., acetamiprid) showed more photo-stability than that of nitromethylene group which absorbs light at 290–400 nm [23].

Neonicotinoids’ mode of action includes nicotinic acetylcholine receptors (nAChRs) as the main target bioreceptors in the central nervous system where neonicotinoids block the receptors causing nervous stimulation, paralysis and finally insect death [14]. nAChRs functions as both ligand gated-ion channels for Na⁺, K⁺ and Ca²⁺ ion transmission and fast cholinergic synaptic neurotransmission. They are activated by binding the neurotransmitter acetylcholine (ACh) and deactivated when ACh departs. ACh is then hydrolyzed by acetylcholinesterase (AChE) to free choline and acetyl enzyme, which is hydrolyzed later, assisted by the histidine, to give back the free enzyme.

Each nicotinic acetylcholine receptor unit is a combination of five subunits (α, β, γ, δ or ε) circularly assembled around the gate channel; at least two of the subunits are α-subunits while others are called non-α-subunits. There are 17 types of nAChRs in vertebrates and 10–12 types in insects depending on their structures (combinations) and functions. ACh binds at the surface between two subunits, one of them at least is α-subunit. Each subunit composed of extracellular β-sheet loops called A, B and C in the α-subunit and D, E and F in the second binding subunit. The A loop has the N peptide terminal while the C loop is characterized by two cysteines separated by 13 amino acids. There is also Cys-loop distinguished by two adjacent cysteine amino acids. The intracellular portion consists of four α-helix transmembrane segments (TM1–TM4) contain large loop between TM3 and TM4 and the C-terminal [5, 22, 23, 30].

On the other hand, acetylcholinesterase responses to neonicotinoids are controversial. Imidacloprid (IMI) is the most widespread neonicotinoids [35]. Its effects on AChE vary depending on insect species, concentration and exposure period where the enzyme was reported to be inhibited [20, 26], stimulated [4] or not affected [35]. Inhibition of AChE results in over-binding of ACh and subsequently over-stimulation of the postsynaptic neurons [6]. Acetylcholinesterase consists of two similar units, each of them contains active site which consists of two subunits, the esteratic site (ES) and the anionic site (AS). The esteratic site is located in a narrow cavity called “aromatic gorge” since it contains 14 aromatic amino acids. More importantly, it contains the catalytic triad (CT) responsible for the catalytic activity and hydrolysis of acetylcholine (ACh). The triad amino acids are serine, glutamate and histidine; the nucleophilic serine hydroxyl group attacks the acetylcholine carbonyl group and liberates a choline molecule while the enzyme is acetylated, which is later hydrolyzed through nucleophilic attack by water molecule, assisted by the imidazole ring of histidine as a general base catalysts, to regenerate he free enzyme. Proteases also, as serine enzymes, contain the same CT but with opposite stereochemistry and aspartate instead of glutamate. The anionic site, in human, has hydrophobic nature of aromatic amino acids and binds the quaternary ammonium of ACh during the catalytic activity. There is also a second “peripheral” aromatic anionic site appears at a distance from the esteratic site (Sussman et al., 1991, [30, 32]).

Although neonicotinoids are used in much lower concentrations than other insecticide groups e.g., organophosphorus and carbamates, they are found to have great dangerous effects on pollinators in a phenomenon called colony collapse disorder (CCD) where the number of honeybee insects in colonies decreased dramatically as a result of the negative effects on coordinated movement, tremors, convulsions, forage capacity, insect growth, queen fertility and remembering e.g., finding the route back to their hives [5, 10, 11, 23]. This phenomenon affects not only honeybee production but also pollination process and thus has harmful effects on agricultural economics. Therefore, studying the mechanistic differences of neonicotinoids on nAChR in pests and pollinators is crucial to achieve selectivity of this group of insecticides [11].

The role of acetylcholinesterase in the detoxification and resistance of various species against neonicotinoids is still unclear, especially in target (e.g., aphids) and non-target (honeybees) insects. In addition, up till now, only few theoretical studies were reported on comparing honeybee nAChR-binding sites with those of harmful insects [8]. More importantly, structural variations of AChE and nAChR in target (aphid) and non-target (honeybee) insects have not yet been documented. Therefore, this study aims to examine the effects of IMI on the target sucking insect Aphis craccivora (Cowpea aphid) in vivo and in vitro as well as comparing these results with those obtained from the non-target insect Apis mellifera (Honeybee) [1]. In addition, molecular modeling was used to compare the IMI interaction with both AChE and nAChRα1 of aphid and honeybee to find out the similarity and/or the differences in structures and binding sites of the two IMI receptors in both insects. This information helps more deep understanding of IMI toxicity and resistance mechanisms in both insects including finding the differences between the two insects that can assist selecting or designing new neonicotinoids with better selectivity.

Materials and methods

Chemicals and instrumentations

Commercial formulation (35% SC, chinook) of imidacloprid (1-(6-chloro-3-pyridylmethyl)-N-nitroimidazolidin-2-ylideneamine), was gifted from Shora chemicals, Cairo, Egypt. Triton X-100 (t-octylphenoxypolyethoxyethanol) was purchased from Advent Chembio Private Limited, Mumbai, India. Acetylthiocholine iodide (ATChI) was obtained from Sigma Chemical Company while 5,5′-dithio-bis (2-nitrobenzoic acid) (DTNB) was purchased from E. Merck Darmstadt, Germany. The UV–Vis instrument was Evolution 300 UV–Vis Thermo Fisher Scientific, Madison, WI, USA. The cooling centrifuge was CRU-5000, International Equipment Company, MA, USA.

Aphid rearing and treatments

The colony of cowpea aphid, Aphis craccivora Koch, used in this study was obtained from Syngenta laboratory in Qalyubia government, Egypt. Insects were reared on faba bean plants, Vicia faba. Plants were cultivated in plastic pots (11 cm diameter) and maintained in a growth chamber at 23 ± 2 °C, 65 ± 5% humidity (R.H.) and photoperiod of 16:8 (L:D) h. The spray method [21] was used to treat plants with different IMI concentrations (0–350 mg/L). Plants were then left to dry before transferring aphid adults to treated plants with the aid of a soft brush [7]. Insects were transferred at the rate of 90 insects for each pot containing two plants (45 insect/plant). Each pot served as a replicate while in each experiment, concentration was performed in triplicate. Corrected mortality was evaluated 24 h after exposure to imidacloprid. The research project was reviewed by the Committee of Experimental Plant Care and Research Ethics at Ain Shams University, Agriculture Sector Committee and issued “Approval No.14-2024-11”.

AChE extraction and activity determination

Acetylcholinesterase (EC 3.1.1.7) activity was measured using the method of Ellman et al. [9] with some modifications as reported by Ali et al. [3]. Adult cowpea aphids (270 insects) were chilled at 4 °C for 3 min for anesthetization [4]. Insects were crushed in 1 mL iced phosphate buffer (0.1 M pH 7.1) containing 0.5% Triton X-100 and 0.25 M sucrose, then centrifuged at 2000 rpm for 15 min at 4 °C. In a cuvette, 2580 µL buffer, containing the chromogen 5,5′-dithio-bis(2-nitrobenzoic acid) (DTNB), 200 µL enzyme extract and 220 µL, the substrate acetylthiocholine iodide (ATChI, 155.61 mM), were mixed. In blank, buffer substituted the enzyme solution. Change in absorbance in 1 min was recorded at 405 nm. Temperature was kept at 4 °C during the determination process.

Aphid AChE activity in vitro

Adult cowpea aphids (640 insects) collected from the rearing pots were extracted in 2 mL phosphate buffer (0.1 M pH 7.1) and treated as described above. Stock solution of the imidacloprid was prepared by dissolving 700 µL formulation in 100 mL phosphate buffer (0.1 M pH 7.1) containing 0.62 mM DTNB. The cuvette enclosed 2323 µL phosphate buffer containing DTNB, 78 µL enzyme extract, 90 µL ATChI (155.61 mM) and 500 µL IMI solution. The final IMI concentrations in the cuvette were 40.84, 81.67, 163.33, 245.0, 326.66 and 408.33 mg/L. AChE activity was determined as described above.

Molecular modeling of aphid AChE

Acetylcholinesterase EC:3.1.1.7 (Carboxylic ester hydrolase) sequence of Aphis craccivora (Cowpea aphid) was extracted from UniProt database (681 Amino acids, 77,077 Da, Accession A0A6G0ZJ06). It includes the active site. The sequence was placed in the Swiss-Model server and a template search was run. Analysis results led to choosing the crystal structure of an insecticide-resistant acetylcholinesterase mutant from the malaria vector Anopheles gambiae in the ligand-free state (6arx.1.A), which has the highest quality and the matching parameters (GMQE 0.73, QSQE 0.91, QMEANDisCo Global 0.85, sequence identity 64.68%, X-ray 2.3 Å, homo-dimer) as a template. The model was then built and downloaded as PDB file.

Molecular modeling of aphid nAChRα1

The amino acid (AA) sequence of nicotinic acetylcholine receptor subunit alpha-L1 (nAChRα1) of Aphis craccivora (cowpea aphid) was obtained from the GenBank ID: KAF0770983.1 (508 AA, Mass 57,403) or UniProt database Accession A0A6G0ZJF0 (CC 178–179) with function property (gene name FWK35_00001485). The sequence was loaded into the Swiss Model server to run a template search. After analyzing resulted templates, acetylcholine receptor subunit alpha, Cryo-EM structure of Torpedo acetylcholine receptor in complex with d-tubocurarine and carbachol (7smt.1.A) that has the highest quality and matching parameters (GMQE 0.52, QMEANDisCo Global 0.58, sequence identity 40.61%, EM 2.6 Å, monomer), was chosen and downloaded (PDB file).

Sequence similarity between AChE or nAChRα1 in aphid and honeybee

Sequence similarity between each of AChE or nAChRα1 in the target insect, Aphis craccivora, and their respective proteins in the non-target insect, Apis mellifera (honeybee), was checked by the Protein BLAST server (https://blast.ncbi.nlm.nih.gov/Blast.cgi, last access 5/9/2023).

Molecular docking of IMI

Imidacloprid was docked into each of Aphid AChE and nAChRα1 online employing the CB-Dock2 server (https://cadd.labshare.cn/cb-dock2/php/index.php, last access 15/9/2023) [18]. Ligand (IMI) was energy-minimized using Gaussian 09 with DFT calculation and dataset. IMI file (sdf) and protein file (pdb) were uploaded. Then search cavity was performed to find the highest five cavities in free energy. All ion and ligands were removed. Then cavity volume, center and size were identified for each pocket. Blind docking was executed in the five pockets to calculate the binding free energy (∆G). Visualization and identifying protein–IMI interactions were performed by Discovery Studio 2019. For each docking, models of the five binding sites with the highest free energy were downloaded and then reordered and named (C1–C5) in descending order according to the binding free energy.

Statistical analysis

Nonlinear sigmoidal curve fitting (Origin 2019b) was used to calculate the lethal concentrations (LC20, LC50 and LC80) and inhibitory concentrations (IC20, IC50 and IC80). One-way ANOVA and Duncan test, in IBM SPSS Statistics 25, were used to determine significant differences among treatment means at significance p < 0.05.

Results and discussion

Aphid treatments and AChE activities

To examine the IMI effects on aphids, cowpea aphid insects (Aphis craccivora) were brought up on faba bean plants sprayed with a series of imidacloprid solutions (0–350 mg/L) including the manufacturer recommended dose (175 mg/L), and then mortality levels and AChE activity were determined and presented in Table S1 and Fig. 1A. Corrected mortality ranged from 39.25% (5 mg/L) to 97.70% (350 mg/L) with LC50 as low as 9.50 mg/L (Table 1). A reported oral LC50 of Aphis craccivora on IMI thin film is 0.063 mg/L [24]. The variations observed in literature in the reported values of LC50 were attributed to several factors e.g., insect species, geographical region and insecticide formulation [12]. Oppositely, AChE activity was not affected even at the highest concentration (350 ppm), which is double the recommended concentration (175 ppm). In the meantime, the mortality level reached 84.01% for the IMI recommended dose and 97.70% for double concentration (350 mg/L), indicating that resistance of aphid AChE does not prevent the dependent of mortality on IMI concentration as presented in Fig. 1A. Aphid AChE result differs from that obtained from honeybee [1] where AChE in the latter insect was sharply inhibited after IMI treatment, then recovered by 84% in 4.0 h accompanied with high toxicity expressed by LC50 2.92 ppm. These strict results are in agreement with the fact that AChE is not the main cause of the observed high IMI toxicity toward both insects but reveals different AChE responses in both insects. It should be noted that an attempt to compare the LC50 of honeybee and aphid is not accurate since the LC50 of the latter is based on the imidacloprid concentration sprayed of plant leaves while that of honeybee is based on the insecticide concentration directly fed to the insect.

Fig. 1

(A) Aphid mortality and AChE activity upon exposure to IMI in vivo. B In vitro aphid and honeybee AChE inhibition. The in vitro bee AChE curve was adopted from [1] with publisher permission

Table 1 Lethal and inhibitory concentrations of Aphid AChE

Aphid AChE inhibition, in vitro, was also determined with various imidacloprid concentrations (Fig. 1B). Inhibitory concentrations (IC20, IC50 and IC80) are included in Table 1. Results showed that imidacloprid inhibits the enzyme in vitro from 38.91 to 74.24% at concentration range 40.84–408.33 mg/L respectively (Table S2). The IC50 of aphid AChE, in vitro, was 108.6 mg/L, compared to that of honeybee, 719.1 mg/L [1], which indicates another difference between AChE responses in both insects since while the aphid enzyme was more stable (not affected) in vivo. It was more sensitive to IMI in vitro than honeybee enzyme. In other words, the aphid enzyme is inhibited by IMI in vitro but not affected in vivo, in addition to the previously observed recovered enzyme activity in honeybee [1], which reflects the involvement of AChE in resistant system. It was reported that the enzyme has a role in insecticide detoxification since aphids develop resistance by over-producing AChE upon exposure to IMI or other neonicotinoid where the enzyme level was found higher in resistant strains [2, 28, 33].

Molecular modeling

Homology of aphid AChE and nAChRα1, and their similarities to respective honeybee proteins

The sequence structures of AChE of aphid (681 amino acids) were obtained from the UniProt database (Fig. S1). Sequence similarity between honeybee and aphid AChE revealed that the coverage reached 84% with identity 66.96% (Fig. 2) and 1% gaps, only indicating high similarity between the enzyme sequences (primary structure) in the two insects. Figure 2 shows also that there is high similarity in location and distance between the triad amino acids in Aphis craccivora (Ser306, Glu432 and His546, Uniprot, A0A6G0ZJ06) and Apis mellifera (Ser263, Glu389 and His504, Uniprot, A0A087ZWF0) where the distance between Ser and Glu is 126 amino acids in both insects, but between Glu and His is 114 in aphid, but 115 in honeybee because of a vacancy in the aphid AChE against honeybee Gly406 enzyme (Fig. 2). The differences also include two vacant regions in honeybee AChE, four amino acids each, near the two terminals and at a distance from the triad amino acids. Since the crystal structures of either acetylcholinesterase or acetylcholine receptors of A. craccivora insect are not available to data, 3D structures of these target proteins were obtained by performing homology modeling for both aphid proteins. Homology modeling of aphid AChE resulted in choosing 6arx.1.A protein as an alignment template of the two AChE chains (A and B). The same template protein (6arx.1.A) was also chosen for honeybee AChE homology confirming also similarity in the secondary structure between the enzymes in both insects. It should be noted that despite this similarity, the observed enzyme activity is much affected, in vivo, in honeybee compared to that of aphid, which might refer to stronger defense system including IMI detoxification or not directly targeting AChE in aphid insects.

Fig. 2

Sequence similarity between AChE in Apis mellifera (black) and Aphis craccivora (Red). Blue boxes indicate the triad amino acids of aphid (Ser306, Glu432 and His546) and honeybee (Ser263, Glu389 and His504). Black box indicates the honeybee Gly406 and its absence in aphid AChE. The dashed red boxes are for the two vacancy regions (5 and 4 amino acids) in honeybee enzyme

Since insect nicotinic acetylcholine receptors (nAChRs) are not as well-understood as those of vertebrates and it is known that the alpha subunit is a common factor in all nAChR-binding sites [16, 34], nAChRα1 of Aphis craccivora (508 AA) was downloaded (Fig. S2). Homology of aphid nAChRα1, as in AChE, resulted in choosing the same alignment template of honeybee nAChRα1 i.e., Torpedo acetylcholine receptor (7smt.1.A), suggesting still good similarity between the primary and secondary structures of the two receptor proteins. Similarity inspection indicates also that the cysteine amino acids in the C loop in aphid (Cys178 and Cys179 match those of honeybee C loop (Cys218 and Cys219 while the two cysteines of Cys-loop in aphid (Cys105 and 119 match their respective ones in honeybee receptor (Cys145 and Cys159). Each couple is separated by 13 amino acids. However, sequence similarity between aphid and honeybee receptors (Fig. 3) showed a high coverage region (89%) but lower identity (50.00%) and much higher gaps (9%) compared to that found between AChE in the two insects (66.96% and 1% respectively). The extracellular loops end at Leu197 (aphid) and Leu237 (honeybee) as indicated in Fig. 3. It can be seen in Fig. 3 that the number of different amino acids in the two receptors increases in the intracellular portions of receptors. Besides, there are six vacant regions in aphid receptor starts after Thr349 and ends before Asn408; these regions are located in the large loop between TM3 and TM4 and thus the aphid large loop (Met308-Glu421) is much smaller than that of honeybee (Leu350-Glu511) i.e., less 50 amino acids accounts for the 9% gaps. Therefore, it can be concluded that most differences between the receptors of both insects are found in the intracellular transmembranes and large loop.

Fig. 3

Sequence similarity between nAChRα1 in Apis mellifera (black) and Aphis craccivora (Red). Blue boxes indicate the two adjacent cysteine amino acids in the C loop in aphid (Cys178 and Cys179) and honeybee (Cys218 and Cys219), and the two cysteines of the Cys loop in aphid (Cys105 and 119) and honeybee (Cys145 and Cys159), each is separated by 14 amino acids. The dashed red boxes are for the six vacant regions (start after Thr 349 and end before Asn408) in aphid large loop. Black box highlights the start and the end of the large loop in aphid (Leu310-Glu421) and honeybee (Leu350-Glu511). Green box marks the Leu197 (aphid) and Leu237 at the end of the extracellular loops

Imidacloprid-AChE interactions

The two acetylcholinesterase units of aphid built model are presented in Fig. 4. In the examined model, the triad amino acids are Ser306, Glu432 and His546 (Uniprot, Accession A0A6G0ZJ06). Docking IMI into the enzyme suggested the highest five free energy binding sites (C1–C5); their volume, location and binding amino acids are listed in Table S3. Results revealed that C1 and C2 with the highest binding energy (7.1 and 7.0 kcal/mol respectively) are located in the vicinity of the active site Fig. 4 and Table S3. Site C1 as presented in Fig. 4 shows the vicinity of the IMI to the triad amino acids and the formation of seven H bonds in addition to other interactions e.g., pi-sulfur of Met190 with aromatic pyridine ring and attractive charge between acidic amino acid Asp548 and the formal positive charge of the nitro nitrogen atom. Site C2 displays also various interactions including alkyl interactions with the chlorine atom, H bond with the nitro oxygen and pi–pi interaction between the aromatic of Phe395 and pyridine as well as the involvement of the triad amino acids Ser306 and His546 (Fig. 4 and Table S3). These diverse interactions display the role of each IMI function group in binding with active site amino acids. Binding sites C3–C5 (Fig. S3) are found far from the enzyme active site but still have good binding energy (−6.3–5.7 kcal/mol).

Fig. 4

Aphid AChE enzyme prepared by homology on insecticide-resistant acetylcholinesterase mutant from the malaria vector Anopheles gambiae in the ligand-free state (6arx.1.A), and C1 and C2 IMI-binding sites, presenting the triad of the active site (Ser306, E432 and His546)

Imidacloprid–nAChRα1 interactions

The location, size, binding free energy and contacting amino acids are listed in Table S4. Binding sites C1–C5 have free energy (−6.4 and −4.9 kcal/mol respectively) very close to those of respective five sites of docking IMI into α1–β2 subunits (−6.25–5.67 kcal/mol) of fruit fly nAChRs [16]. Binding site C1 is found in a cavity within the extracellular loops (Fig. 5 forming H bonds with Ser154 and Ala155. The interaction is also characterized by forming pi–sigma interaction between the aromatic pyridine ring and Ile151, in addition to charge attraction between the nitro nitrogen and Asp152; these amino acids are found before the C loop (Val160-Thr196) while IMI binds also with the A loop through the pi–alkyl hydrophobic interaction (Val31 and van der Waals interaction) (Ser16 and Trp32).

Fig. 5

2D and 3D of C1 and C2 of Aphid nAChRα1 prepared by homology on Torpedo acetylcholine receptor in complex with d-tubocurarine and carbachol (7smt.1.A), and resonance mechanism of IMI

Different from all other honeybee and aphid nAChRα1-binding sites, the second aphid site, C2 (∆G –6.2 kcal/mol), is positioned in a cavity among the transmembrane segments where IMI is connected to TM1 (Ser211, Tyr212 and Ser214), TM3 (Thr266, Leu269 and Cys273) and TM4 (Ile463, Phe464, Ala467 and Ser468) as depicted in Fig. 5. It can be noticed also in the 2D model of C2 that this is the only binding site where the conformation of the IMI methylene group has been changed to place pyridine ring away from the imidazolidine substituent to bridge the cavity among the three transmembrane segments. The other aphid binding sites (C3–C5) are placed in Fig. S4. Binding site C3 includes amino acids in loops A (Leu20, Asn21, Leu22 and Lys23) and C (Glu161, Trp162 and Arg194). The only aphid binding site found in the TM3–TM4 large loop (Tyr285-Leu422) is C4; it is bonded to Ser345-Leu350, Ala391, Ala392, Arg394 and Leu395. Finally, the binding site C5 is located entirely in the Cys-loop where all the contacting amino acids shown in Fig. S4 are part of the Cys loop.

The present results revealed the structure–binding activity relationship and the role of each IMI function group in binding with aphid in addition to the type of amino acids in nAChRα1 that can interact with each function group. The aromatic pyridine ring exhibited various hydrophobic binding interactions e.g., Pi-Sigma and pi-alkyl with Ala, Leu, Ile and Val (C1, C2), pi–pi T-shaped with Phe (C2) and pi–sulfur with Cys (C2). Pyridine nitrogen forms H bond with Cys (C2) and Gln (C5). Ethylene group of the imidazolidine ring and the methylene group form van der Waals interaction with Ser, Tyr and Leu (C1, C2 and C3). The chlorine atom can give pi–alkyl interaction with Ile, Phe (C1), Cys, Leu (C2), Lys (C3) and Tyr (C5), van der Waals interaction with Leu (C1), Ile (C2), Glu (C3), Ala, Thr (C4) and Asp (C5). Chlorine atom can also form a halogen bond. It is a noncovalent bond formed between a partially positive part of a covalently bonded halogen and an electronically rich atom (N, O, S) or group (aromatic ring) usually found in protein molecules [29]. Chlorine atom could form a halogen bond with Phe (C5). The importance of the nitro group that it can participate in protein interaction through its formal positive nitrogen atom by attractive charge with the acidic or polar amino acids e.g., Asp (C1), Glu (C3) and Asp (C5). In addition, the nitro negative oxygen atom, formed by resonance through the imidazolidine nitrogen atoms to the nitro oxygen as depicted in Fig. 5 [14], forms H bond with Ala, Ser (C1), Thr (C2), Arg (C3), Ser, Gly (C4), Glyn and Thr (C5). The imidazolidine NH and imine nitrogen can also participate in interaction by forming H bonds with Thr (C2), Tyr, Glu (C3), Ala, Ser (C4), and Thr (C5).

The observed variations in the structure of nAChRα1 in aphid and honeybee, discussed above, are expressed in having different binding sites (location and size), binding energy and contacting amino acids in both insects. The shortening in the aphid intracellular large loop between TM3 and TM4 by 50 amino acids also contributes to the observed differences in the binding sites of nAChRα1 in both insects where in honeybee receptor, four (C1, C3, C4 and C5) of the highest five binding sites are located in the intracellular large loop and one (C2) in the extracellular loops [1] while in aphid nAChRα1, one only (C4) is found in the intracellular TM3–TM4 large loop, one (C2) in a cavity between TM1, TM3 and TM4 segments, and three binding sites (C1, C3 and C5) were found in the extracellular loops. In addition, the more complicated honeybee receptor may also express the higher sense characters, in this insect, as in distinguishing among different flower smells.

It can be noted that binding neonicotinoids, especially IMI, to nAChRs as well as the role of nAChRs in insect resistance were proved experimentally. The radio-ligand binding assay, using [³H] imidacloprid, was used to confirm IMI binding to nAChR; some cholinergic ligands e.g., nicotine and cysteine were then used to displace [³H] imidacloprid [27]. The radio-ligand binding assay was also used to determine the binding affinity of nAChR toward IMI in different species [15, 17]. Strains possessing a R81T mutation in the nAChR β1 subunit were found to express reduced sensitivity toward imidacloprid in in Aphis sp. [13, 25]. In addition, Li et al. [19] indicated that Y/S mutation found in loop B is related to neonicotinoid resistance in Drosophila and Musca; while docking and molecular stimulation showed the importance of loops A, B, C and D in the interaction with the neonicotinoids. Rocher and Marchand-Geneste [31] showed that compound with high toxicity expresses higher number of hydrogen bonds with nAChR of Apis mellifera.

Therefore, nAChRα1 not only plays an important role in imidacloprid, and generally in neonicotinoid, toxicity but also is a source of variation in different insect species that could be useful in designing new neonicotinoids with better selectivity toward harmful insects and less effects on agro-friendly species.

Conclusion

Effects of imidacloprid, as a representative of neonicotinoid family (Fig. S5), on AChE and nAChRα1 in cowpea aphids, were examined. Despite AChE was inhibited, in vitro, with IC50 108.6 mg/L, it is not affected, in vivo. The mortality was concentration-dependent with LC50 9.50 mg/L. In addition, AChE was more inhibited, in vitro, in aphid (IC50 108.6 mg/L) than in honeybee (IC50 719.1 mg/L). However, in vivo, it was not affected even at double recommended IMI concentration (350 mg/L), but previous work indicated that the enzyme in honeybee is sharply inhibited then recovered by only 84% in 4.0 h with IC50 5.63. These results indicate that the enzyme is not the main cause of the observed mortality, but it still has a role in insect resistance system with different responses in both insects.

Molecular modeling showed high similarity in the primary and secondary structures of AChE in honeybees and aphids manifested by high identity (67%) and low gaps (1%) as well as choosing the same alignment template for both enzyme homology and similar position of the triad amino acids. This high similarity suggests difficulty in depending on AChE for searching better insecticide selectivity. Despite this similarity, the honeybee enzyme, in vivo, is much affected, compared to that of aphid, suggesting stronger defense system in aphid insects while IMI might not directly interact with the enzyme. On the other hand, the similarity in nAChRα1 of both insects is lower as the identity decreased to 50% and gaps increased to 9%. These gaps (50 amino acids) are found in the intracellular large loop between TM3 and TM4 and caused the aphid loop to be much smaller and thus accounts for the observed differences in the nAChRα1-binding sites of in both insects where in honeybee receptor, four of the highest five binding sites are located in the large loop while in aphid receptor, only one site (C4) was found entirely in the large loop. Other aphid nAChRα1 sites are three (C1, C3 and C5) located in the extracellular loops and one (C2) found in a cavity between TM1, TM3 and TM4 segments. These variations in nAChRα1 structures and binding sites in different insect species can assist in designing new neonicotinoids that express better selectivity. It can be noted that similar approach can be used to understand the source of sensitivity or resistance of different species by examining the molecular modeling of nAChR and AChE in these species to find out differences in the structural and binding sites.