Dedicated acetic acid preference coded by broad spectrum ionotropic receptors in a moth species

Acetic acid as one of the food related odorant cues attracts many insect species. In the moth Mythimna separata, the olfaction of acid was coded via multiple pathways including 3 sensilla types on the antennae and 3 glomeruli in the antennal lobes. Among, suitable dosages of acetic acid exclusively activated DC3 glomerulus that receives integrated projections across sensilla types, which drives attractiveness and feeding attempts of the moth. This circuit encodes broad spectrum ionotropic receptors 8a, 75q1 and 75q2 which were sufficient to confer acid responsiveness in Xenopus oocytes. Ir75q2 was expressed in vivo with Ir75q1 and it enhanced sensitivity of the receptor functional group toward acids. Furthermore, Ir75q1 and Ir75q2 are both necessary for the moth to conduct acetic acid induced reactions of sensilla, DC3 glomerulus as well as attractiveness. Together, it reveals that an indispensable tetramer IR-based unit is employed to fulfill acetic acid specialized preference under suitable dosages through balancing of transcription and peripheral coding. Understanding of the Ir75q1/2 olfactory pathway provides insights into investigations on acid sensory process in insects. Author Summary The preference to acids are common in various organisms, and it may involve both olfactory and gustatory reception. In particular, airborne acidity volatiles can be sensed through antennae of insects and later assessed to help locating foraging, mating, and egg laying sites. However, these stimulatory processes can only be delivered by suitable dosages of acids, as we all know that, high acidity could be fatal in most circumstances. To date, avoidance to acids has been well explained in insects, but attractiveness and its basis remain uncharted. In the brain of oriental armyworm Mythimna separata, we have located 3 olfactory pathways which may play roles in acetic acid reception. Fortunately, when acetic acid was applied at attractive dosages, it only activated 1 dedicated pathway among the three. Later we found that this attractiveness pathway employed two ionotropic receptor genes namely Ir75q1 and Ir75q2, to successfully deliver this trait. Both genes were necessary for the moth to conduct acetic acid preference, but their roles are different. Ir75q1 recognized the acetic acid ligand and Ir75q2 later amplify the sensitivity. By comparing with evidences from electrophysiology and brain imaging tests, we found that the expression bias of either of the two genes has caused the separation of the pathways. It has been revealed in this moth that a smart decision system for olfactory reception exists, and this system may extrapolate to other insect species, as Ir75q1 and Ir75q2 are commonly expressed in many insect families.

exclusively activated DC3 glomerulus that receives integrated projections across sensilla types, 23 which drives attractiveness and feeding attempts of the moth. This circuit encodes broad spectrum 24 ionotropic receptors 8a, 75q1 and 75q2 which were sufficient to confer acid responsiveness in 25 Xenopus oocytes. Ir75q2 was expressed in vivo with Ir75q1 and it enhanced sensitivity of the 26 receptor functional group toward acids. Furthermore, Ir75q1 and Ir75q2 are both necessary for the 27 moth to conduct acetic acid induced reactions of sensilla, DC3 glomerulus as well as attractiveness. 28 Together, it reveals that an indispensable tetramer IR-based unit is employed to fulfill acetic acid 29 specialized preference under suitable dosages through balancing of transcription and peripheral 30 coding. Understanding of the Ir75q1/2 olfactory pathway provides insights into investigations on 31 acid sensory process in insects.

32
Author Summary 33 The preference to acids are common in various organisms, and it may involve both olfactory and 34 gustatory reception. In particular, airborne acidity volatiles can be sensed through antennae of 35 insects and later assessed to help locating foraging, mating, and egg laying sites. However, these 36 stimulatory processes can only be delivered by suitable dosages of acids, as we all know that, high 37 acidity could be fatal in most circumstances. To date, avoidance to acids has been well explained 38 in insects, but attractiveness and its basis remain uncharted. In the brain of oriental armyworm 39 Mythimna separata, we have located 3 olfactory pathways which may play roles in acetic acid 40 reception. Fortunately, when acetic acid was applied at attractive dosages, it only activated 1 41 dedicated pathway among the three. Later we found that this attractiveness pathway employed two Introduction 52 Insects are precisely developed and possess a complex chemosensory system to 53 communicate with the outside world [1]. Olfaction is critical to this system and conveys crucial 54 information to enable survival in response to multiple chemical cues, including pheromones, host 55 volatiles, or enemy smells [2][3][4]. Many of these compounds are acids, which exist widely in 56 nature as common products of plants [5,6]. In Drosophila, acetic acid is sensed to assess food, Ir64a is expressed together with a co-receptor Ir8a and they together form a dimer-dimer 68 complex to perform as an acid sensing ion channel [20]. Furthermore, Ir75a is also used by D. 69 melanogaster and D. sechellia to smell acetic acid and propionic acid [29,30]. Ir76b mediates 70 oviposition preference in female Drosophila by sensing acetic acid and citric acid through 71 gustatory reception [10]. Meanwhile, reports on lepidopteran species have also revealed 72 glomeruli based coding for feeding behaviors which were elicited by acidity components [31]. 73 However, the attractiveness olfactory pathways encoded by candidate IR genes to acetic acid 74 itself are still unclear, especially in insects who employ acetic acid as a food related odorant [32]. 75 The Oriental armyworm moth, Mythimna separata Walker (Lepidoptera: Noctuidae), 76 also known as Pseudaletia separata, is a key pest of crops in Asia and it has been monitored with 77 acetic acid rich lures for decades [33]. In particular, sweet vinegar solution luring is one of the 78 most cost-effective methods of trapping this pest and the recipe for the solution is now a national proportion of acetic acid in its luring recipe, we believe that M. separata is a suitable target for 84 exploring the olfactory mechanism on acetic acid attractiveness. We hypothesize that acetic acid 130 Acetic acid stimulates attractiveness and feeding attempt in M. separata in a dosage manner 131 In wind tunnel tests, both the sweet vinegar solution and 10 -1000 g acetic acids 132 elicited comparable attractiveness to M. separata, excepted for landing behavior ( Figure 3C).

133
Dosages of 10 g to 1000 g acetic acids all caused significant upwind flights comparing to the 134 water control. Moreover, we found that 12 h fasted moths adults conducted significant 135 attractiveness behaviors to acetic acid than fed moths ( Figure 3B, 3C). We asked whether this 136 feature is specific for acetic acid by comparing it to two non-food related odorant acid chemicals 137 propionic acid (related to acetic acid in Drosophila IR sensory) and enanthic acid (performed 138 very well in SSR tests). Interestingly, the result revealed that acetic acid was the only one which 139 elicited significant upwind flights to the moth among acids ( Figure 3C and Figure S3).

140
Furthermore, an overwhelming of 5000 g acetic acid terminated attractiveness in the moths.

141
As feeding status significantly influenced behavioral outputs, we then tested the fasted 142 moths in an adopted proboscis extension response (PER) assay with olfactory stimuli [40] ( Figure   143 3D). Results showed that sweet vinegar solution and acetic acid both elicited significant PER 144 behaviors comparing to water control ( Figure 3E). This clearly revealed that acetic acid can serve 145 both at range and closely as a food related odorant source (Movie S1  150 We then traced the acetic acid signaling in the antennal lobe, the first olfactory neuropil 151 of the insect's brain. In in vivo optical imaging tests, a total three different regions of interest 152 (ROIs) were evoked by multiple acidity components ( Figure 4A). A pattern change was observed 153 in acetic acid stimuli: only 1 area was constantly evoked during the dosages of 10 to 1000 g, 154 while 3 areas were evoked when applied at 5000 g (Movie S2). Enanthic acid evoked only 1 155 area with weak intensity regardless of dosages (Movie S3). Propionic acid, butyric acid, and 156 valeric acid all evoked 3 areas of interests ( Figure 4A). Other tested acids did not show a 157 significant evoking patterns under the scope of this study ( Figure 4C and Figure S4).

158
To identify correspondence glomeruli, we merged ROIs from the above 5 active acids.

159
Result showed that the ROIs from all acids were highly overlapped, indicating 1 glomerulus was 160 involved in each area ( Figure 4B).  Figure 4C). Reactions of DC3 also started at a dosage of 100 g and it 169 increased with dosages of acetic acid, but this glomerulus did not respond to increasing of 170 enanthic acid dosage ( Figure 4D). It was obvious that DC3 received integrated axon projections 171 from multiple sensilla types as per the SSR test results.

172
To better address the relationships of antennal lobe evoking patterns with behavioral 173 consequences, we cross-correlated these acids in a pair wise comparison along 40 frames in DC3 174 areas ( Figure 4E). During stimuli, results showed that 1000 g acetic acid did not correlate well 175 with 1000 g enanthic acid (blue square, Figure 4E), which is consistent with different behavioral 176 outputs in wind tunnel tests for these acids. On the other, a shallow yet significant correlation was 177 observed between 10 g acetic acid and 1000 g enanthic acid, indicating that evoking intensity 178 of DC3 has contributed in close search behaviors (orange square, Figure 4E). To sum up, so far 179 evidences showed that antennal lobe reaction patterns could be the main driving force for 180 determination of behavioral outputs to acids, and DC3 is highly likely to be the main glomerulus 181 involved in acetic acid stimulated attractiveness during 100 to 1000 g dosages. With results have were used to assess off-target effect. Injection of siGFPs did not cause down-regulation of either 250 Ir75q1 or Ir75q2 ( Figure 7A). Attractiveness of siGFP moths to acetic acid and sweet vinegar 251 solution was not significantly influenced in wind tunnel ( Figure 7B). The results from RNAi tests 252 not only confirmed the existence of Ir75q1/2-as-DC3 pathway but also clearly indicated that this 253 pathway acts as an indispensable stimulatory olfactory pathway which drives acetic acid 254 attractiveness behaviors in M. separata adults. 255 We then established a schematic to show sensing mechanism of acids in M. separata 256 ( Figure 8A, 8B). To assess this model, we knocked-down each gene one by one and tested moth 257 adults with an EAG approach. Results showed that, single injection of siIr8a, siIr75q1, or 258 siIr75q2 all significantly decreased reactions to acetic acid ( Figure 8C). While for enanthic acid, 259 only siIr75q2 strain showed a significant decrease compared to wild type ( Figure 8C). 282 responses yet at a low reaction rate, while Ir75q2 + Ir8a alone had an unexpected response to 283 enanthic acid yet no preference to acetic acid was shown. Ir75q1 + Ir75q2 + Ir8a had an 284 amplified response to acetic acid yet maintained the response to enanthic acid ( Figure 8A). The 285 response to enanthic acid could be an evolutionary accident when M. separata employ Ir75q2 to 286 enhance acetic acid sensing in the Ir75q1 pathway ( Figure S8E).

287
To sum up, the response pattern of oocytes expressing Ir75q1/q2/8a correlated with as1A 288 ORNs, and the as1A neuron projected to DC3 together with integration of other sensilla types 289 which can also be reversely proved by RNAi tests. The responses to enanthic acid caused by New features in IR-involved olfactory reception 303 We revealed a phenomenon that three IRs may form a functional complex.  346 We managed to minimize this disadvantage by applying acids at low concentrations (10 -6 mol/L) 347 in order not to change the pH levels. Another thought was to analyze Ir75q1/q2 with reference to 348 neutralized acid solutions at pH 7.5. In this way, we not only successfully characterized Ir75q1/q2 in terms of its ligand recognition traits, but also further identified necessary genes for 350 forming the functional ion channel complex.    The authors declare no conflict of interests.           pair of subjects for all tested chemicals including water control, acetic acid, propionic acid, 741 enanthic acid, butyric acid, valeric acid, lactic acid, citric acid, and hydrochloric acid. (C) 742 Representative spike patterns of as1 (n = 12), as2 (n = 10), and as3 (n = 16) type sensilla.   Table S3. ★ indicate predicted conserved amino acid residues which are 924 essential for ligand recognitions within binding domain of each IR.