Trace amines inhibit insect odorant receptor function through antagonism of the co-receptor subunit

Many insect behaviors are driven by olfaction, making insect olfactory receptors (ORs) appealing targets for insect control. Insect ORs are odorant-gated ion channels, with each receptor thought to be composed of a representative from a large, variable family of odorant binding subunits and a highly conserved co-receptor subunit (Orco), assembled in an unknown stoichiometry. Synthetic Orco directed agonists and antagonists have recently been identified. Several Orco antagonists have been shown to act via an allosteric mechanism to inhibit OR activation by odorants. The high degree of conservation of Orco across insect species results in Orco antagonists having broad activity at ORs from a variety of insect species and suggests that the binding site for Orco ligands may serve as a modulatory site for compounds endogenous to insects or may be a target of exogenous compounds, such as those produced by plants. To test this idea, we screened a series of biogenic and trace amines, identifying several as Orco antagonists. Of particular interest were tryptamine, a plant-produced amine, and tyramine, an amine endogenous to the insect nervous system. Tryptamine was found to be a potent antagonist of Orco, able to block Orco activation by an Orco agonist and to allosterically inhibit activation of ORs by odorants. Tyramine had effects similar to those of tryptamine, but was less potent. Importantly, both tryptamine and tyramine displayed broad activity, inhibiting odorant activation of ORs of species from three different insect orders (Diptera, Lepidoptera and Coleoptera), as well as odorant activation of six diverse ORs from a single species (the human malaria vector mosquito, Anopheles gambiae). Our results suggest that endogenous and exogenous natural compounds serve as Orco ligands modulating insect olfaction and that Orco can be an important target for the development of novel insect repellants.


Introduction
Insects have positive and negative impacts on humans, in terms of health, economy, and food stores. Insects pollinate plants to increase global food production, with 35% of global production of crops depending on animal pollinators 1,2 . Insects also cause significant destruction of crops and food stores [3][4][5] . Insects can also transmit fatal diseases such as dengue fever 6 , malaria 7 , yellow fever and epidemic typhus 8 . Insects use olfaction to sense their surroundings and to guide important activities, including feeding, mating and oviposition. This makes the insect olfactory system receptors an attractive target for the chemical control of deleterious insect species.
Insects use odorant receptors (ORs) to recognize and distinguish a diverse range of odorants 9,10 . Each OR is composed of two functionally essential parts: a highly conserved co-receptor subunit (Orco) and one of a large number of variable odorant-binding (or "tuning") subunits 11-17 . These subunits associate in an unknown stoichiometry to form an odorant-gated ion channel 18,19 . ORs have also been proposed to initiate, or be modified by, second messenger cascades 13,18 . While the odorant-binding subunit is responsible for interacting with odorants 9,20,21 , both the odorant-binding subunits and Orco are involved in forming the ion channel pore 21,22 . Insect ORs are not related to the receptors and channels of humans and other tetrapods 15 , suggesting that control of detrimental insect activity may be possible through the development of insect OR selective compounds. A current approach to developing these compounds is to identify the particular odorant binding subunits that recognize behaviorally important odorants 10,23-26 and then conduct large scale ligand screens 27,28 , but high diversity among the odorant binding subunit repertoires of different species makes this approach exceptionally labor intensive 29,30 .
The recent identification of the synthetic compound VUAA1 as a novel OR agonist that acts directly on Orco 27 , suggests that manipulation of insect behavior might be achieved by targeting Orco. Based on the VUAA1 structure, several additional synthetic Orco agonists and a larger, more diverse series of synthetic Orco antagonists have been identified [31][32][33] . Importantly, several of these Orco antagonists were shown to inhibit odorant activation of ORs through a noncompetitive mechanism 31-33 . These findings suggest that Orco antagonists might be useful in altering insect behavior.
Orco subunits are highly conserved across insect species, suggesting that Orco serves an essential function common to all insect ORs 15,34 . This high conservation underlies observations that Orco subunits from different species are functionally interchangeable; an Orco subunit from one species can form functional ORs with an odorantbinding subunit from a different species 21,22 . As the "pharmacology" of synthetic Orco agonists and antagonists has expanded, it has also become clear that Orco subunits from disparate insect species have very similar sensitivities to known Orco ligands 27,[31][32][33]35 . This suggested to us that the binding site for Orco ligands may serve as a modulatory site for compounds endogenous to the insects or may be a target of exogenous compounds, such as those generated by plants. Insects use a variety of amines as neurotransmitters and neuromodulators [36][37][38][39] . Plants also generate a variety of amines that may play a role in resistance to insect herbivores [40][41][42] . For these reasons, we screened a panel of biogenic and trace amines for agonist and antagonist activity at insect Orco subunits. We found tryptamine to be a potent Orco antagonist with broad activity at Orco subunits from different species. Tyramine and phenethylamine also function as Orco antagonists, but were substantially less potent than tryptamine. Importantly, we found that tryptamine, acting through Orco, could inhibit odorant activation of a wide range of ORs from a variety of insect species. Our findings suggest a role for Orco as a modulatory site common to all insect ORs and support the development of Orco-directed compounds that can be used to manipulate insect behavior.

Expression of insect ORs in Xenopus oocytes
Mature Xenopus laevis frogs were anesthetized by submersion in 0.1% 3-aminobenzoic acid ethyl ester. Depth of anesthesia was judged by loss of nasal flare and swallow reflexes. Oocytes were surgically removed. The incision was treated with gentamicin sulfate (two subcutaneous injections of 0.1 mL 10 mg/mL gentamycin at the surgical site) and sutured. Immediately following surgery (and before recovery from anesthesia), as an analgesia agent, one subcutaneous injection of Meloxicam solution (0.1 mg/mL) (0.1 mg/kg body weight) was administered to the dorsal lymph sac of the frogs. The frogs were allowed to recover from surgery in a humid chamber before being placed back in the holding tank. Surgeries were performed on individual frogs no more often than once every 3 months. Following the fourth surgery, frogs were anesthetized as described above and then pithed.

Experimental protocols and data analysis
To screen for agonist activity, oocytes were exposed to 30 sec applications of candidate compounds with 5 min washes between applications ( Figure 2A). For the concentration-response protocol (Table 1), applications were for 20 sec at a flow rate of 1.65 ml/min. Electrophysiology and data capture Odorant and Orco ligand induced currents were recorded under two-electrode voltage clamp, using an automated parallel electrophysiology system (OpusExpress 6000A, Molecular Devices). Oocytes were perfused with ND96 (in mM: 96 NaCl, 2 KCl, 1 CaCl 2 , 1 MgCl 2 , 5 HEPES, pH 7.5). Orco ligands were prepared as 50 or 100 mM stock solutions in DMSO and then diluted into ND96 on the day of the experiment. Odorants were prepared as 100 mM stock solutions in DMSO and then diluted into ND96. Unless otherwise noted, applications were for 60 sec at a flow rate of 1.0 ml/min, with extensive washing in ND96 at 4.6 ml/min between applications. Micropipettes were filled with 3 M KCl and had resistances of 0.2-2.0 MΩ. The holding potential was -70 mV. Current responses, filtered (4-pole, Bessel, low pass) at 20 Hz (-3 db) and sampled at 100 Hz, were captured and stored using OpusXpress 1.1 software (Molecular Devices).

Figure 2. Tryptamine and several other amines are antagonists of Cqui\Orco. A)
The tested amines do not display Orco agonist activity. Oocytes expressing Cqui\Orco were challenged with 30 sec applications of 100µM gramine, tyramine, tryptamine and melatonin (top trace), phenethylamine, serotonin, octopamine and dopamine (middle trace), or histamine, epinephrine and norepinephrine (bottom trace), with 5 min washes between applications. 30µM OLC12 (Orco agonist) was applied at the end of each trace. B) Tryptamine and tyramine are antagonists of Cqui\Orco. Oocytes expressing Cqui\Orco were exposed to 60 sec applications of 30µM OLC12 with 4 min washes between applications. 100µM tryptamine (top trace), tyramine (middle trace), or octopamine (bottom trace) were applied and incubated for 90 sec preceding the third application of OLC12 and then co-applied during the OLC12 application. C) Screen of 11 amines for Orco antagonism. Responses of Cqui\Orco to 30µM OLC12 (~EC 5 ) in the presence of 100µM of each compound are presented as a percentage of the average of two preceding responses to OLC12 alone (mean ± SEM, n = 3-10). Statistical significance was assessed by one-way ANOVA, followed by Dunnett's post-test comparing to sham treated oocytes (*p<0.01; **p<0.001).
To measure antagonist activity at Orco ( Figure 2B, 2C, Figure 3, Figure 4A and Figure 5A), oocytes were exposed to two 60 sec applications of the synthetic Orco agonist OLC12 (2-((4-Ethyl-5-(4-pyridinyl)-4H-1,2,4-triazol-3-yl)sulfanyl)-N-(4-isopropylphenyl)acetamide) with 4 min washes between applications. Oocytes were then exposed to a 90 sec application of antagonist candidate, immediately followed by a 60 sec co-application of antagonist candidate and OLC12. The current response in the presence of antagonist candidate was compared to the mean of the preceding two responses to OLC12 alone and is presented as a percentage.
To measure inhibition of odorant activation of heteromeric ORs (Figures 4B, 4C, Figure 5B and Figure 6), oocytes were exposed to a 30 sec application of odorant followed by a 10 min wash. Oocytes were then exposed to a 90 sec application of tryptamine or tyramine, immediately followed by a 30 sec co-application of tryptamine or tyramine and odorant. The current response in the presence of antagonist candidate was compared to the preceding response to odorant alone and expressed as a percentage. In our previous work, we found that repeated odorant applications to some ORs could cause a progressive decrease in response amplitude 31,33 . For this reason, we then re-normalized antagonism data to the value obtained when the assay was run in the absence of antagonist candidate (sham). In the "sham" assay, oocytes were exposed to a 30 sec application of odorant followed by a 10 min wash and then exposed to a 90 sec application of ND96 (no antagonist candidate), immediately followed by a 30 sec application of odorant. The second odorant response was compared to the first response and expressed as a percentage. In Figure 4B, 4C, Figure 5B and Figure 6, the sham value for 100nM eugenol was 57 ± 3% (mean ± SEM, n = 3). In Figure 5B, the sham value for 10nM eugenol was 93 ± 4% (n = 4). In Figure 4B and C, the sham value for 1µM Z11-14:OAc was 82 ± 6% (n = 6) and the sham value for 150µM 2-phenylethanol was 92 ± 2% (n = 3). In Figure 6, the sham value for 3µM l-fenchone was 83 ± 1% (n = 3), the sham value for 40µM acetophenone was 92 ± 1% (n = 3), the sham value for 70µM geranyl acetate was 97 ± 1% (n = 3), the sham value for 10µM 6-methyl-5-hepten-2-one was 94 ± 2% (n = 3) and the sham value for 3µM 2-nonanone was 81 ± 1% (n = 3).    Initial analysis of electrophysiological data was done using Clampfit 9.1 software (Molecular Devices). Curve fitting and statistical analyses were done using Prism 5 (Graphpad). Concentrationinhibition data were fit to the equation: I = I max /(1+ (X/IC 50 ) n ) where I represents the current response at a given concentration of inhibitor, X; I max is the maximal response in the absence of inhibitor; IC 50 is the concentration of inhibitor present that still allows a half maximal response from odorant; n is the apparent Hill coefficient. Concentration-response data were fit to the equation: I = I max /(1+(EC 50 /X) n ) To screen a panel of biogenic and trace amines (Figure 1), we expressed Orco from Culex quinquefasciatus (Southern House Mosquito) in Xenopus oocytes and recorded ligand-induced current responses using two-electrode voltage clamp electrophysiology (see Methods). Orco subunits from several species, including Cqui\ Orco, have been shown to form homomeric channels when heterologously expressed in the absence of odorant-binding subunits 27,33 . This convenient property of Orco allowed us to perform the initial screen without potentially confounding interactions with odorantbinding subunits. Successful functional expression of Cqui\Orco was confirmed by application of OLC12, a previously identified Orco specific agonist 31 . While OLC12 elicited robust current responses, none of the amines displayed agonist activity at Cqui\ Orco ( Figure 2A). Next we screened the amines for antagonist activity by applying 30µM OLC12 (~EC 5 ) to activate Cqui\Orco and co-applying 100µM of each amine ( Figure 2B, C). Several amines were able to inhibit OLC12 activation of Cqui\Orco. Tryptamine was the most effective antagonist, blocking more than 90% of the OLC12 response (92 ± 2% inhibition). Highly significant inhibition (p<0.001) was also observed for phenethylamine (41 ± 1%), tyramine (40 ± 5%), gramine (30 ± 4%) and serotonin 23 ± 3%), but the extent of inhibition was less than 50%, suggesting relatively low affinity interactions. Histamine (16 ± 8%), melatonin (13 ± 1%) and epinephrine (9 ± 3%) also displayed significant (p<0.01), but modest, inhibition of the OLC12 current. Octopamine, dopamine and norepinephrine were inactive in this assay.
In Figure 3A, we constructed concentration-inhibition curves for block of Cqui\Orco activity in order to quantitatively evaluate the inhibitory potency of tryptamine, as well as phenethylamine and tyramine, representing the less effective amines. Tryptamine was clearly the most potent of these antagonists, inhibiting Cqui\Orco with an IC 50 of 4.7 ± 0.7µM, a value similar to that of the most potent synthetic Orco antagonists that we identified in our previous work 33 . Phenethylamine (IC 50 = 117 ± 12µM) and tyramine (IC 50 = 157 ± 22µM) were substantially less potent than tryptamine (25-fold and 33-fold, respectively). Previously identified Orco antagonists inhibited OLC12 activation of Orco through a competitive mechanism 31,33 . To determine whether tryptamine was also a competitive antagonist of Orco, we measured blockade of Cqui\ Orco achieved by tryptamine when the OLC12 concentration was increased from 30µM to 100µM ( Figure 3B). Tryptamine was significantly less effective at inhibiting responses to 100µM OLC12 (IC 50 = 143 ± 18µM, p<0.0001, F-test), indicating that tryptamine is a competitive antagonist of Cqui\Orco.
We next asked whether tryptamine could also inhibit Orco from other insect species. In addition to Cqui\Orco, we tested Agam\Orco from An. gambiae (human malaria vector mosquito) and Dmel\ Orco from D. melanogaster. Co-application of 10µM tryptamine inhibited OLC12 activation of Orco from each of these three insect species ( Figure 4A). We then wondered whether tryptamine could also inhibit odorant activation of heteromeric  where I represents the current response at a given concentration of odorant, X; I max is the maximal response; EC 50 is the concentration of agonist yielding a half maximal response; n is the apparent Hill coefficient. Statistical significance (p<0.05) was assessed using a two-tailed unpaired t test, an F test, or a one-way analysis of variance followed by the Dunnett's post-test, as appropriate.
Recent reports 27,32 , together with our previous findings 31,33 , have revealed the existence of a ligand-binding site on the Orco subunit and that inhibition of odorant activation through a non-competitive mechanism may be a general property of Orco-directed antagonists.
Our current results suggest that endogenous and exogenous natural compounds serve as Orco ligands and modulate insect olfaction. While tyramine is a major neurotransmitter in insects 53 , its low potency in our assay ( Figure 3) suggests that it might not serve as an endogenous OR modulator. However, the function of an endogenous Orco antagonist is unlikely to be the complete block of OR function. Rather, an endogenous Orco antagonist might be used to diminish olfactory sensitivity by inhibiting a fraction of the available receptors. For tyramine, such inhibition could occur at concentrations ranging from 10µM to 30µM. Alternatively, there may be additional, more potent, but as yet uncharacterized, endogenous Orco antagonists that can decrease olfactory sensitivity at lower concentrations.
In contrast to the low potency of tyramine, we found tryptamine to be a high potency Orco antagonist. Tryptamine inhibited odorant activation of an OR with an IC 50 in the low micromolar range ( Figure 5). While it is currently unclear whether tryptamine is endogenous to insects, tryptamine and similar compounds, such as gramine, are produced by a variety of plants and are thought to serve as a defense against insect herbivores 42,59 . Various tryptamine analogs have been proposed as larvicides 60 and when tryptamine is caused to accumulate in poplar and tobacco, through ectopic expression of tryptophan decarboxylase, the feeding behavior of insects that target these plants is altered 41 . Tryptamine-based structures also act on various receptors and transporters, particularly those involved in serotonergic neurotransmission, exerting psychedelic effects in humans. Indeed, many plant derived and synthetic hallucinogens are based on the tryptamine and phenethylamine scaffolds [61][62][63] . Interestingly, the potency that we observed for tryptamine inhibition of odorant activation of an insect OR ( Figure 5) is similar to the potency for tryptamine inhibition of the D. melanogaster serotonin transporter 64 .
Might there also be natural endogenous or exogenous Orco agonists? An endogenous Orco agonist could serve to increase olfactory sensitivity, perhaps in a circadian fashion, to alter behavior during critical foraging or mating periods. An exogenous, plantderived Orco agonist would, by activating all ORs through Orco, serve as an olfactory "confusant" and might alter the feeding behavior of insect herbivores. The limited screen of 11 compounds Mcar\Orco+Mcar\Or5 from M. caryae (Long-Horned Beetle, Order Coleoptera) that responds to 2-phenylethanol 44 . We chose to proceed with an OR from An. gambiae instead of Cx. quinquefasciatus for two reasons. The best characterized of the Cqui\Or subunits respond to indoles 23,24 , which are structurally related to tryptamine and might confound our experiments. Also, the Agam\Or subunit family has been more extensively characterized 10,25 , offering more options for OR expression (see below). Each odorant was applied at or near the EC 50 concentration (Table 1, 44,45 ). Co-application of 10µM tryptamine resulted in substantial inhibition of each receptor ( Figure 4B). We also examined tyramine. While tyramine is a low potency Orco antagonist ( Figure 3A), it is a major neurotransmitter in insects 37 . Tyramine was also able to reduce odorant activation of these ORs, but was less effective than tryptamine ( Figure 4C). These results suggest that tryptamine and tyramine are broadly active antagonists of insect ORs.
Several previously identified Orco antagonists have been shown to inhibit odorant activation of insect ORs through a non-competitive mechanism 31-33 . To determine whether the tryptamine inhibition of odorant activation that we observed in Figure 4 was also noncompetitive, we examined the effect of tryptamine on activation of the heteromeric Agam\Orco+Agam\Or65 in more detail ( Figure 5). When the concentration of Orco directed agonist (OLC12) was increased, the tryptamine inhibition curve was significantly shifted to the right ( Figure 5A). However, when the concentration of odorant agonist (eugenol) was increased, the tryptamine inhibition curve did not shift ( Figure 5B). These results indicate that, similar to previously identified synthetic Orco antagonist compounds, tryptamine is a competitive antagonist of direct activation of Orco and a non-competitive antagonist of odorant activation of the OR.
The ability of tryptamine to interact with Orco and exert a noncompetitive inhibitory effect on odorant activation of a heteromeric OR ( Figure 5) suggests that tryptamine should be able to inhibit activation of a variety of ORs activated by diverse odorants. To examine this possibility, we tested the ability of tryptamine to inhibit odorant activation of ORs formed by Agam\Orco and each of six different odorant-binding subunits chosen from across the An. gambiae OR gene family 46 . We activated each OR with a previously identified cognate odorant 25 at a concentration at or near the EC 50 (Table 1, 25 ). In addition to Agam\Orco+Agam\Or65 (activated by eugenol), we tested Agam\Orco+Agam\Or27 (activated by L-fenchone), Agam\Orco+Agam\Or28 (activated acetophenone), Agam\Orco+Agam\Or31 (activated by geranyl acetate), Agam\ Orco+Agam\Or39 (activated by 6-methyl-5-hepten-2-one) and Agam\Orco+Agam\Or48 (activated by 2-nonanone). With the exception of Agam\Or39 and Agam\Or48, which display overlapping odorant specificities at 4 odorants, there is little or no similarity among the odorant specificities of these six odorant-binding subunits 25 . In each case, 10µM tryptamine was able to inhibit odorant activation of the receptor, despite the disparate odorantbinding subunits and diverse odorant structures ( Figure 6). Tyramine was also able to inhibit odorant activation of each of these receptors, but was less effective than tryptamine (note that tyramine is applied at 100µM). We conclude that tryptamine and tyramine are general antagonists of insect ORs.

Competing interests
No competing interests were disclosed.

Grant information
This work was supported by a grant from the National Institutes of Health (RO1 DC011091 to CWL).
The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Acknowledgements
We thank B. Sherman and A. Castro for Xenopus care and oocyte preparation.
that we conducted here did not identify any Orco agonists, but more extensive screening is clearly warranted.
Several synthetic Orco antagonists have been shown to inhibit odorant activation of ORs through an allosteric mechanism 31-33 . The ability of these compounds to inhibit multiple ORs from a variety of species is likely due to the high conservation of Orco across the insects 12 . Similarly, we found that tryptamine and tyramine, acting as Orco antagonists, could inhibit odorant activation of ORs from insect species chosen from three different orders: Diptera (An. gambiae), Lepidoptera (O. nubilalis) and Coleoptera (M. caryae). Furthermore, when we examined multiple ORs from a single species (An. gambiae), we found that tryptamine and tyramine blocked odorant activation of each receptor. The action of these compounds through Orco allowed blockade to occur despite the highly diverse odorant-binding subunits used to form the receptors and the different odorant structures used to activate the receptors. Interestingly, while all six receptors were inhibited, the extent of inhibition varied depending on the odorant-binding subunit present and the pattern of variation was similar for tryptamine and tyramine. This suggests differences in allosteric coupling between Orco and the various odorant-binding subunits. Also, while we showed that tryptamine is a potent inhibitor of odorant activation of Agam\Or65+Agam\Orco, the results we present in Figure 6 suggest that tryptamine is even more potent at other ORs, such as those formed by Agam\Or27, Agam\Or31 and Agam\Or39. Our current results with naturally occurring amines, together with previous reports with synthetic compounds 27,31-33,35 strongly suggest that: 1) allosteric antagonism of odorant activation of ORs is a general property of Orco antagonists; 2) Orco antagonists are broadly active at ORs of many insect species; and 3) Orco is an important target for the development of novel insect repellants. The broad activity of Orco directed compounds across many insect species that has been observed to date