Molecular and Pharmacological Characterization of β-Adrenergic-like Octopamine Receptors in the Endoparasitoid Cotesia chilonis (Hymenoptera: Braconidae)

Octopamine (OA) is structurally and functionally similar to adrenaline/noradrenaline in vertebrates, and OA modulates diverse physiological and behavioral processes in invertebrates. OA exerts its actions by binding to specific octopamine receptors (OARs). Functional and pharmacological characterization of OARs have been investigated in several insects. However, the literature on OARs is scarce for parasitoids. Here we cloned three β-adrenergic-like OARs (CcOctβRs) from Cotesia chilonis. CcOctβRs share high similarity with their own orthologous receptors. The transcript levels of CcOctβRs were varied in different tissues. When heterologously expressed in CHO-K1 cells, CcOctβRs induced cAMP production, and were dose-dependently activated by OA, TA and putative octopaminergic agonists. Their activities were inhibited by potential antagonists and were most efficiently blocked by epinastine. Our study offers important information about the molecular and pharmacological properties of β-adrenergic-like OARs from C. chilonis that will provide the basis to reveal the contribution of individual receptors to the physiological processes and behaviors in parasitoids.

OA exerts its activities via binding to membrane proteins OA receptors (OARs) that belong to the class A (rhodopsin-like) G protein-coupled receptor (GPCR) family [31]. Insect OARs can be classified into three major groups based on their structural, pharmacological and biochemical properties, similar to vertebrate α-adrenergic or β-adrenergic receptors: α 1 -adrenergic-like OARs (Octα1R, also referred to as OAMB), β-adrenergic-like OARs (OctβR), and α 2 -adrenergic-like OARs (Octα2R) [32][33][34]. Activation of Octα1R can elevate intracellular Ca 2+ levels at lower concentrations of OA as compared to increasing cyclic adenosine monophosphate (cAMP) levels, whereas activation of OctβRs selectively stimulate intracellular cAMP production [35,36]. Octα2R has two alternative splicing forms, and they are both activated by OA, resulting in the inhibition of forskolin-simulated intracellular cAMP levels [33].

Cloning and Sequence Analysis of CcOctβ1R, CcOctβ2R, and CcOctβ3R
The fragments of three putative β-adrenergic-like OARs were identified from the transcriptome data of C. chilonis. The full lengths of cDNA of CcOctβ1R, CcOctβ2R, and CcOctβ3R were amplified by RT-PCR and further verified via DNA sequencing. The sequences of CcOctβRs were submitted to the GenBank with the following accession numbers: CcOctβ1R (OP422531), CcOctβ2R (OP422532), CcOctβ3R (OP422533). The deduced amino-acid sequences contain 417, 435, and 358 residues for CcOctβ1R, CcOctβ2R, and CcOctβ3R, respectively.

Expression Profiles of CcOctβRs
The expression profiles of CcOctβ1R, CcOctβ2R, and CcOctβ3R were investigated in different tissues of C. chilonis, including head, thorax, abdomen, antenna, and leg. qRT-PCR results showed that the relative expression levels of CcOctβ1R, CcOctβ2R, and CcOctβ3R were varied in various tissues. CcOctβ1R and CcOctβ2R were mostly expressed in the head, while CcOctβ3R was more highly expressed in the antenna than in the head ( Figure 2).
CcOctβ2R displayed a relatively higher expression in the antenna than abdomen, leg and thorax ( Figure 2B), and the transcript level of CcOctβ3R in the thorax and leg was higher than in the abdomen ( Figure 2C).

Expression Profiles of CcOctβRs
The expression profiles of CcOctβ1R, CcOctβ2R, and CcOctβ3R were investigated in different tissues of C. chilonis, including head, thorax, abdomen, antenna, and leg. qRT-PCR results showed that the relative expression levels of CcOctβ1R, CcOctβ2R, and CcOctβ3R were varied in various tissues. CcOctβ1R and CcOctβ2R were mostly expressed in the head, while CcOctβ3R was more highly expressed in the antenna than in the head ( Figure 2). CcOctβ2R displayed a relatively higher expression in the antenna than abdomen, leg and thorax ( Figure 2B), and the transcript level of CcOctβ3R in the thorax and leg was higher than in the abdomen ( Figure 2C).

Ligand Specificity of CcOctβRs
To unravel second messenger coupling and pharmacological profiles of CcOctβ1R, CcOctβ2R, and CcOctβ3R, we obtained CHO-K1 cells stably expressing each receptor, respectively. The cAMP levels were determined in the cells when incubated in the presence of 1 μM concentrations of various biogenic amines and putative synthetic agonists including OA, TA, DA, 5-HT, HA, naphazoline, clonidine, tolazoline, medetomidine, and lisuride. Neither DA, 5-HT nor HA caused the elevation of cAMP production. In contrast, the cell lines of each receptor showed significantly increasing cAMP levels after application of OA, TA, naphazoline, clonidine, tolazoline, medetomidine and lisuride ( Figure 3). Thus, we concluded that CcOctβ1R, CcOctβ2R, and CcOctβ3R couple to Gs protein.
Further, the increasing concentrations of OA, TA and seven potent synthetic agonists were applied to the cells, and dose-response curves for each receptor were displayed (Figure 4). The rank order of agonist potency for CcOctβ1R was naphazoline > DMPF > amitraz > OA > lisuride > medetomidine > clonidine > tolazoline > TA ( Figure 4A, Table 1). Similarly, half-maximal stimulation (EC50) values for tested compounds were calculated from concentration-response curves for CcOctβ2R and CcOctβ3R ( Figure 4B-C), and were summarized in Table 1. Half-maximal activations of CcOctβRs with OA, naphazoline, DMPF, amitraz and lisuride were in the low nanomolar range. OA was about two orders of magnitude more efficient than TA for CcOctβRs (Table 1).

Ligand Specificity of CcOctβRs
To unravel second messenger coupling and pharmacological profiles of CcOctβ1R, CcOctβ2R, and CcOctβ3R, we obtained CHO-K1 cells stably expressing each receptor, respectively. The cAMP levels were determined in the cells when incubated in the presence of 1 µM concentrations of various biogenic amines and putative synthetic agonists including OA, TA, DA, 5-HT, HA, naphazoline, clonidine, tolazoline, medetomidine, and lisuride. Neither DA, 5-HT nor HA caused the elevation of cAMP production. In contrast, the cell lines of each receptor showed significantly increasing cAMP levels after application of OA, TA, naphazoline, clonidine, tolazoline, medetomidine and lisuride ( Figure 3). Thus, we concluded that CcOctβ1R, CcOctβ2R, and CcOctβ3R couple to G s protein.
Further, the increasing concentrations of OA, TA and seven potent synthetic agonists were applied to the cells, and dose-response curves for each receptor were displayed ( Figure 4). The rank order of agonist potency for CcOctβ1R was naphazoline > DMPF > amitraz > OA > lisuride > medetomidine > clonidine > tolazoline > TA ( Figure 4A, Table 1). Similarly, half-maximal stimulation (EC 50 ) values for tested compounds were calculated from concentration-response curves for CcOctβ2R and CcOctβ3R ( Figure 4B-C), and were summarized in Table 1. Half-maximal activations of CcOctβRs with OA, naphazoline, DMPF, amitraz and lisuride were in the low nanomolar range. OA was about two orders of magnitude more efficient than TA for CcOctβRs (Table 1).

Pharmacological Properties of CcOctβRs
To examine the ability of potential antagonists to impair OA-activated signaling, measurements were carried out with the antagonists (10 µM) on a non-saturating concentration of OA in the CHO-K1 cells stably expressing each receptor. The reduction in cAMP level was quantified and normalized to the value achieved in the absence of antagonists (=100%). Chlorpromazine, cyproheptadine, metoclopramine, mianserin, phentolamine, methiothepin, epinastine, spiperone, SCH-23390, clozapine, asenapine, amitriptyline and doxepin displayed the antagonistic effects for each receptor, butaclamol, cinanerin and chlorprothixene inhibited the activation for CcOctβ2R and CcOctβ3R ( Figure 5).    Next, the effects of 12 antagonists were measured with increasing concentrations on a non-saturating concentration of OA for each receptor ( Figure 6). Ligand concentrations that resulted in the half-maximal inhibition of each receptor (IC 50 ) were determined from the dose-response curves and shown in Table 2. However, IC 50 values of methiothepin, clozapine and chlorprothixene for CcOctβ1R were not calculated ( Table 2). Taking into account the effects of the antagonists on the basal activity, the rank order of antagonist potency was also different for each receptor. For CcOctβ1R, it was epinastine > mianserin > asenapine > phentolamine > metoclopramine > doxepin > cyproheptadine > amitriptyline > chlorpromazine. For CcOctβ2R, it was epinastine > asenapine > mianserin > phentolamine > cyproheptadine > clozapine > methiothepin > chlorprothixene > amitriptyline > metoclopramine > doxepin > chlorpromazine. For CcOctβ3R, it was epinastine > asenapine > phentolamine > mianserin > doxepin > cyproheptadine > metoclopramine > clozapine > chlorprothixene > amitriptyline > methiothepin > chlorpromazine. The most efficient antagonist on OA-induced CcOctβRs was epinastine, with IC 50 s ranging from~7.34 × 10 −9 M (CcOctβ3R) to~3.76 × 10 −8 M (CcOctβ1R) ( Table 2). Asenapine and amitriptyline reduced OA-stimulated receptor activity by approximately 90% at the high concentrations for each receptor ( Figure 6). For each antagonist, the order of receptor antagonistic efficacy was CcOctβ3R > CcOctβ2R > CcOctβ1R (Table 2).
CcOctβRs belong to the class A GPCR family, which are supported by D 3.49 RY motif at the cytoplasmic of TM3 and N 7.49 PLIY in TM7 (Supplementary Figures S1-S3). Most class A GPCRs are activated by ligands docking to specific amino acid residues in the receptor binding pocket close to the extracellular side [58], and functionally important residues of β-adrenergic-like OARs are highly conserved in CcOctβRs. The D 3.32 residue in TM3 (D 121 in CcOctβ1R, D 127 in CcOctβ2R, D 75 in CcOctβ3R), acts as an important residue that interacts with the protonated amino group of OA or TA. One of the closely grouped Ser residues (S 5.42 / 5.46 ) in TM5 (S 210 SSVS 214 in CcOctβ1R, S 216 SSIS 220 in CcOctβ2R, S 168 SSVS 172 in CcOctβ3R) can interact with the hydroxyl group of OA or TA [63,64]. The motif F 6.44 -X-X-C 6.47 -W 6.48 -X-P 6.50 -F 6.51 -F 6.52 is highly conserved in TM6 of CcOctβRs. W 6.48 and F 6.52 residues were reported to interact with the ligand aromatic ring, which modulates the bend angle of TM6 near the well conserved kink at P 6.50 , prompting the cytoplasmic end of TM6 to move upon activation [65].
Our phylogenetic tree included major insect biogenic amine receptors and vertebrate adrenergic receptors. CcOctβRs were closely related to their orthologous receptors, and they clustered nicely with D1-like DARs and human β-adrenergic receptors (Figure 1), highlighting the concept of 'ligand-hopping' in the evolution of aminergic GPCRs [66]. In addition, pharmacological properties are similar in all these receptors, and activation of them induce an increase in intracellular cAMP production, suggesting that a close pharmacological and phylogenetic relationship occurs between OARs and DARs. Interestingly, they are four β-adrenergic-like OARs (Octβ1R-Octβ4R) reported in A. mellifera [41], and phylogenetic tree indicated that CcOctβ3R is more closely related to AmOctβ3R than AmOctβ4R ( Figure 1). However, AmOctβ3R and AmOctβ4R were originated by alternative splicing of the same gene [41]. Additionally, the latest research showed that AmOctβ2R expressed multiple isoforms due to differential splicing [67]. Thus, these studies supported that β-adrenergic-like OAR family has only three genes and the different alternative splicing events may occur in the β-adrenergic-like OARs.
The transcriptional profiles of CcOctβ1R, CcOctβ2R and CcOctβ3R varied in different tissues (Figure 2). CcOctβ1R and CcOctβ2R were highly expressed in the head. Similar results were found in PxOctβ1R [62], MsOctβ2R [68], DmOctβRs [69], and NlOctβRs [42], which agreed with OA abundant in the central nervous system. CcOctβ3R was predominantly expressed in the antenna, and the CcOctβ2R transcript in the antenna was second only to that in the head. BmOctβ2R [35] and NcOctβ2R [70] were also expressed in the antenna. Interestingly, octopaminergic modulation in the antenna of honeybee workers could be enhanced not only by increasing the OA level, but also by upregulation of OARs [71], and OA has been reported to increase the olfactory responses of Heliothis virescens to female sex pheromones [72]. CcOctβRs were all expressed in the leg. The transcript levels of NcOctβ2R [70] and TcOctβ2R [50] were found to exhibit high expressions in the leg. There are large muscles in insect legs, and the expressions of DmOctβRs were all detected in the adult muscles [28]. OA regulates neuromuscular transmission, lipid oxidation and glycolysis for motor behavior [73].
As a formamidine acaricide, amitraz is particularly effective against mites and ticks, as well as hemipteran and lepidopteran insects. Amitraz and its main metabolite DPMF act as high-affinity agonists of OARs [35]. For CcOctβRs, the rank order of the potency was DPMF > amitraz > OA (Table 1). Amitraz and DPMF could potently activate αand β-adrenergic-like OARs of B. mori, and DPMF was more potent than amitraz and OA [35]. For PxOctβ1R and NcOctβ2R, amitraz and DPMF were less potent than OA [62,70]. The recent study indicated that amitraz and DPMF potently activated all four OARs from Varroa destructor, and VdOctβ2R was more sensitive to amitraz and DPMF than AmOctβ2R. Furthermore, behavioral assays using D. melanogaster OAR mutants identified Octβ2R as the sole target of amitraz [51].
Here, the interactions of three CcOctβRs with a series of potential antagonists were studied. Epinastine is an antagonist of vertebrate histamine H1 receptor, and also acts as an OAR antagonist in insects. Our results indicated epinastine was the most potent on CcOctβRs (Table 2), and similar observation was reported in NcOctβ2R and PaOctβ2R [58,70]. Mianserin is known as a non-selective 5-HT 2 serotonin receptor antagonist in mammals but also a potent antagonist of OARs, TARs and DARs in insects [74,75]. Mianserin was highly potent to CcOctβRs (Table 2), and acted as the most efficient inhibitor for DmOctβRs and AmOctβRs [40,41]. In A. mellifera, the physiological and behavioral effects of mianserin have been studied, suggesting that it could affect gustatory and olfactory learning abilities [76,77]. Spiperone, SCH-23390, asenapine, amitriptyline and doxepin are generally known as markedly effective antagonists of DARs [78,79], and our results indicated that they also effectively blocked CcOctβRs ( Figure 5 and Table 2), which may suggest that overlapping antagonistic interactions exist between DARs and OARs. Asenapine, amitriptyline and doxepin exhibited highly potent characteristics to DOP2, with nanomolar IC 50 values for CqDOP2 and AaDOP2 [79]. The previous studies showed that phentolamine could efficiently block Octα1R, OctβR and Octα2R [40,57,60], and our results indicated that CcOctβRs were significantly attenuated by phentolamine (Figure 6), and also weakly activated by phentolamine (Supplementary Figure S4). Phentolamine is capable of showing agonist activity on OctβRs, and its dual actions were present in DmOctβRs [40], CsOctβ2R [52], and NcOctβ2R [70].
Overall, our study supports the proposition that OA signaling is highly complex in insects. Therefore, establishing pharmacological profiles of molecularly identified and heterologously expressed receptors is a valuable challenge, and by uncovering substances acting through individual receptors, it is possible to facilitate the design of novel in vivo pharmacological experiments which accurately address a specific physiological feature. This detailed pharmacological characterization of three CcOctβRs marks an important step towards precisely understanding the role of OA in this parasitoid.

Insects
Laboratory colony of C. chilonis has been continuously reared on its host C. suppressalis. They were initially obtained from rice fields of the experimental farm (Hangzhou, China). They were kept at a temperature of 28 ± 1 • C, a photoperiod of 16:8 h (L:D) and relative humidity of 75 ± 5%. After eclosion, the wasp adults were maintained in the glass containers and fed on 20% (v/v) honey solution [38,80].
The C. chilonis adults were sterilized with 75% ethanol, and their heads were then dissected into the TRIzol reagent (Invitrogen, Carlsbad, CA, USA). Total RNA was isolated according to the manufacturer's protocol. The NanoDrop 2000 spectrophotometer (Thermo Fisher Scientific, Waltham, MA, USA) was used to determine the purities and concentrations of RNA samples. cDNA was synthesized with 1 µg RNA via TransScript One-Step gDNA Removal and cDNA Synthesis SuperMix kit (Transgen, Beijing, China). The full lengths of CcOctβRs were cloned by RT-PCR (Supplementary Table S2).

Sequence Alignment and Phylogenetic Analysis
The amino acid sequences of β-adrenergic-like OARs from A. mellifera, D. melanogaster, C. suppressalis, N. lugens, and T. castaneum, were obtained in the NCBI databases. Multiple sequence alignments were carried out using ClustalX2 and consequently edited with GeneDoc software. The putative transmembrane regions were predicted using TMHMM-2.0 (https://services.healthtech.dtu.dk/service.php?TMHMM-2.0, accessed on 15 August 2022). The putative PKC phosphorylation sites were identified by the NetPhos 3.1 Server (http://www.cbs.dtu.dk/services/NetPhos/, accessed on 15 August 2022). The potential N-glycosylation sites were predicted by the NetNGlyc 1.0 Server (http://www. cbs.dtu.dk/services/NetNGlyc/, accessed on 15 August 2022). To identify the potential orthologues of the cloned C. chilonis β-adrenergic-like OARs, the phylogenetic analysis of these receptors was performed with other insect biogenic amine receptors and human adrenergic receptors. The phylogenetic tree was constructed with MEGA 7.0 using the neighbor-joining method and 1000 bootstrap tests [81] and edited by the interactive tree of life (iTOL version 5) tool [82].

Expression Profiles of CcOctβRs
To investigate the relative expression levels of CcOctβ1R, CcOctβ2R and CcOctβ3R in various tissues of wasp adults, total RNAs from the heads without antennae, thorax, abdomen, antennae, and legs were extracted using the TRIzol reagent (Invitrogen, Carlsbad Table S2). qRT-PCR analyses were conducted on a CFX 96 TM Real-Time Detection System (Bio-Rad, Hercules, CA, USA) with the following procedure: 95 • C for 30 s, and then 40 cycles of 95 • C for 5 s and 60 • C for 30 s with the melt curve dissociation step. qRT-PCR was done in a 25 µL volume containing 12.5 µL TB Green ® Premix Ex Taq ™ II (Tli RNaseH Plus) (TaKaRa, Dalian, China), 1 µL each primer (10 µM), 2 µL cDNA, and 8.5 µL sterile H 2 O. Reactions for each sample were performed in triplicate. The 28S rRNA was used as the reference gene to calculate the expression levels of CcOctβ1R, CcOctβ2R and CcOctβ3R [80,83], and the relative quantification was determined using the 2 −∆∆CT method [84].

Construction of Expression Vectors
The eukaryotic expression vectors of CcOctβ1R, CcOctβ2R and CcOctβ3R were generated by RT-PCR using the specific primers (Supplementary Table S2), with the insertion of a Kozak consensus motif [85]. The KpnI and EcoRI were used to digest the pcDNA3.0 vector with hemagglutinin A tag (Invitrogen, Carlsbad, CA, USA), and the PCR products were then homologously cloned into pcDNA3.0 vector to obtain pcDNA3.0-CcOctβ1R, pcDNA3.0-CcOctβ2R and pcDNA3.0-CcOctβ3R with the ClonExpress ® II One Step Cloning Kit (Vazyme, Nanjing, China). All expression constructs were finally verified by DNA sequencing (Sangon Biotech, Shanghai, China). The high-quality expression plasmids were prepared with the EndoFree Plasmid Midi Kit (Qiagen, Hilden, Germany).

Heterologous Expression
Chinese hamster ovary K1 (CHO-K1) cells were cultured at 37 • C with 5% CO 2 in an incubator. The culture medium was composed of DMEM/F-12 medium (Invitrogen, Carlsbad, CA, USA) and 10% fetal bovine serum (Invitrogen, Carlsbad, CA, USA). To prevent the contaminations of gram-positive and gram-negative bacteria, 100 units/mL penicillin and 100 mg/mL streptomycin (Invitrogen, Carlsbad, CA, USA) were added to the culture medium, respectively. The plasmids of pcDNA3.0-CcOctβ1R, pcDNA3.0-CcOctβ2R and pcDNA3.0-CcOctβ3R were transfected into CHO-K1 cells at the confluency of 70-90% in Petri dishes (3.5 cm). The Opti-MEM ® Reduced Serum Medium (1 mL) (Invitrogen, Carlsbad, CA, USA) containing 6 µL Lipofectamine ® 2000 (Invitrogen, Carlsbad, CA, USA) and 3 µg plasmid was prepared as transfection medium. After incubation for 20 min at room temperature, the transfection medium was supplemented after removing the growth medium. After 4-6 h in the incubator, the transfection medium was substituted with the fresh growth medium. Transfected cells were selected with the antibiotics G418 (800 µg/mL, Amresco, Solon, OH, USA) for 2-3 weeks, and G418-resistant cell clones were trypsin digested and transferred to cell culture plates (12-well) for expansion. The individual cell lines were monitored by immunocytochemistry for receptor expression, and then the stably transfected cells were cultured in the growth medium supplemented with 250 µg/mL G418.

cAMP Assays
Ligand stimulation of intracellular cAMP production was measured as described previously [33,70,78]. The cells were grown in 12-well plates (Nunc, Rochester, NY, USA) and cultured in the incubator. After 20 min of preincubation in the presence of 100 µM IBMX (Sigma-Aldrich, St. Louis, MO, USA) in D-PBS at 25 • C, the cells were added with 50 µL D-PBS (Invitrogen, Carlsbad, CA, USA) containing different concentrations of agonists, and they were then incubated for 20 min at 25 • C. The effects of antagonists were examined with 10 µM antagonists in the presence of the following OA concentrations: CcOctβ1R (100 nM), CcOctβ2R (30 nM) and CcOctβ3R (300 nM). After removing the solutions, the reactions were stopped by quickly adding 250 µL cell lysis buffer. The cell lysates were collected into the tubes and centrifuged at 12,000× g for 15 min. The cAMP Parameter Assay Kit (R&D Systems, Minneapolis, MN, USA) was used to determine cAMP in the supernatant. All tested compounds were dissolved in dimethyl sulfoxide, and further serially diluted with D-PBS.

Statistical Analysis
GraphPad Prism version 9.1.1 (GraphPad Software, San Diego, CA, USA) was used to analyze and display the data. Differences in transcript levels were analyzed by a one-way ANOVA followed by Tukey's honestly significant difference (HSD) test. To examine whether statistical differences occurred in pharmacological comparisons, a oneway ANOVA followed by Dunnett's multiple comparison test was performed (* p < 0.05, ** p < 0.01, *** p < 0.001). The EC 50 and IC 50 values were calculated using a special doseresponse, nonlinear regression model.

Conflicts of Interest:
The authors declare no conflict of interest.