Molecular Basis of CO2 Sensing in Hyphantria cunea

Carbon dioxide (CO2) released by plants can serve as a cue for regulating insect behaviors. Hyphantria cunea is a widely distributed forestry pest that may use CO2 as a cue for foraging and oviposition. However, the molecular mechanism underlying its ability to sense CO2 has not been elucidated. Our initial study showed that CO2 is significantly attractive to H. cunea adults. Subsequently, 44 H. cunea gustatory receptors (GRs) were identified using transcriptome data, and 3 candidate CO2 receptors that are specifically expressed in the labial palps were identified. In vivo electrophysiological assays revealed that the labial palp is the primary organ for CO2 perception in H. cunea, which is similar to findings in other lepidopteran species. By using the Xenopus oocyte expression system, we showed that the HcunGR1 and HcunGR3 co-expressions produced a robust response to CO2, but HcunGR2 had an inhibitory effect on CO2 perception. Finally, immunohistochemical staining revealed sexual dimorphism in the CO2-sensitive labial pit organ glomerulus (LPOG). Taken together, our results clarified the mechanism by which H. cunea sense CO2, laying the foundation for further investigations into the role of CO2 in the rapid spread of H. cunea.


Introduction
CO 2 , as a ubiquitous gas in the natural atmosphere, is produced by almost all organisms when they obtain energy via respiration.Insects possess the ability to detect subtle changes in the concentration of carbon dioxide (CO 2 ) in their environment [1].CO 2 not only serves as a danger signal for insects, but in social insects such as honeybees, CO 2 could also cause an imminent increase in the nest temperature; thus, they expel CO 2 by flapping their wings to lower the temperature of their nest [2].CO 2 is also important in host-seeking and oviposition in insects; e.g., blood-sucking mosquitoes use CO 2 as a cue to locate their vertebrate hosts [3,4]; in phytophagous insects, evidence has shown that CO 2 can guide Elasmopalpus lignosellus larvae to locate the freshest parts of the plant [5], while Manduca sexta and Cactoblastis cactorum use CO 2 to locate suitable places to lay eggs [6,7].However, the CO 2 -sensing mechanism of most other lepidopteran pests, such as the major forestry pest Hyphantria cunea, remains unclear.H. cunea is a highly polyphagous and fertile pest; one female can lay up to 500 eggs with a hatch rate of over 95% [8], and can forage on more than 400 plant species, resulting in great damage to forest ecosystems, which is known as "smokeless fires" locally [9].One of the predominant reasons for their rapid spread is their ability to choose suitable foraging and oviposition sites [10], suggesting that CO 2 cues might contribute to their adaptation to the environment.
The CO 2 sensing pathway in insects varies among different taxa.Two CO 2 receptor genes were found in Drosophila melanogaster, and only DmelGr21a and DmelGr63a were co-expressed in response to CO 2 [11,12], suggesting that a heterodimer is needed in D. melanogaster to carry out CO 2 sensing.However, CO 2 receptor homologs have not been identified in honeybees, ants, or blacklegged ticks [13], suggesting that these species utilize distinct mechanisms for CO 2 sensing [14,15].Three CO 2 receptors have been found in lepidopterans, such as Danaus Plexippus, Bombyx mori, Heliconius Melpomene, and Helicoverpa armigera [16,17].A functional study of H. armigera showed that although HarmGr1, HarmGr3, and HarmGr2 were expressed in the same neuron of the labial palps, only the co-expression of HarmGr1+HarmGr3 or HarmGr1+HarmGr2+HarmGr3 resulted in a robust response to Sodium bicarbonate (NaHCO 3 ) [18], indicating that HarmGr1 and HarmGr3 were necessary for CO 2 perception in H. armigera, whereas the role of HarmGr2 is still unclear.To date, the molecular CO 2 sensing pathway in H. cunea is unclear, and whether the CO 2 sensing mechanism in H. cunea is conserved among lepidopteran species remains unknown.
At the central nervous system (CNS) level, the location and projection modes of CO 2sensing glomeruli also vary among different insects.In D. melanogaster, CO 2 -sensitive neurons in the antennae project to a ventrally distributed glomerulus called the V glomerulus [19,20]; in Aedes aegypti and in several other mosquitos, CO 2 -sensitive neurons in the maxillary palps are linked to the dorsomedial glomerulus in the antennal lobe (AL) [21][22][23].These two species both have an ipsilateral projection; i.e., CO 2 neurons project to only one side of the AL, which is also called the "single side projection mode".In Lepidoptera, the well-accepted hypothesis is that labial palp pit organs (LPO) are used for CO 2 sensing [24], and LPO neurons project to a specific glomerulus located on the most ventral side of the antennal lobe; this type of glomerulus is referred to as the "labial pit organ glomerulus" (LPOG) [24,25].The LPO neurons in lepidopterans project to both the ipsilateral and contralateral antennal lobes via bilateral projection.LPOG is considered a specific CO 2 glomerulus, and no other olfactory neurons project to this glomerulus; therefore, an assessment of its projection mechanisms and olfactory glomerulus volume could provide further motivation to assess the importance of CO 2 sensing in H. cunea through higherlevel mechanisms.In conclusion, although the rapid spread of H. cunea may be related to herbivore-induced plant volatiles (HIPVs) such as pinene [26], its ability to sense CO 2 also plays a major role.But to our knowledge, no studies have investigated the role of CO 2 in the spread of H. cunea using molecular-level approaches.Our study will contribute to a better understanding of the important role that CO 2 plays in the dispersal of H. cunea, which will provide a theoretical basis for pest management.

Effect of CO 2 on H. cunea Behavior
To verify the effect of CO 2 on the behavior of female H. cunea, Y-tube olfactometer tests were conducted with 60 H. cunea at different CO 2 concentrations.Compared with those of the negative control (pure air without CO 2 , Figure 1), 32 moths preferred 1% CO 2 .As the CO 2 concentration increased, 34 moths were attracted to the side with CO 2 at a 3% concentration, 37 moths were attracted to 5% CO 2 , 38 moths were attracted to 8% CO 2, and 41 moths were attracted to 10% CO 2 .These results showed that CO 2 concentrations ranging from 8%-10% had an attractive effect on H. cunea adults; within this range, the attractive effect of CO 2 increased with increasing CO 2 concentration.

Identification and Homology Analysis of HcunGRs
To identify CO 2 receptors in H. cunea, 44 candidate gustatory receptors (GRs) were identified from H. cunea transcriptome data; 24 GRs were full-length sequences, and 20 were truncated sequences.Blast results revealed that the similarity between HcunGRs and those of other species ranged from 35% to 91% (Table S1).HcunGR16 was the homolog of the fructose receptors of D. melanogaster, B. mori, H. armigera, and H. melpomene and was also hypothesized to be a candidate fructose receptor in H. cunea (Figure 2) [16].HcunGR4-10 and HcunGR12-14 clustered together with the conserved sugar receptor DmelGR64 [27], suggesting that these GRs have a sugar receptor (non-fructose) sensing role.HcunGR44 is a candidate for bitter receptors in H. cunea.Most importantly, we found that HcunGR1, DmelGR21a, and HarmGR1; HcunGR2 and HarmGR2; and HcunGR3, DmelGR63a, and HarmGR3 were grouped together, indicating that these three HcunGRs might play a role in CO 2 sensing in H. cunea.

Identification and Homology Analysis of HcunGRs
To identify CO2 receptors in H. cunea, 44 candidate gustatory receptors (GRs) were identified from H. cunea transcriptome data; 24 GRs were full-length sequences, and 20 were truncated sequences.Blast results revealed that the similarity between HcunGRs and those of other species ranged from 35% to 91% (Table S1).HcunGR16 was the homolog of the fructose receptors of D. melanogaster, B. mori, H. armigera, and H. melpomene and was also hypothesized to be a candidate fructose receptor in H. cunea (Figure 2) [16].HcunGR4-10 and HcunGR12-14 clustered together with the conserved sugar receptor DmelGR64 [27], suggesting that these GRs have a sugar receptor (non-fructose) sensing role.HcunGR44 is a candidate for bitter receptors in H. cunea.Most importantly, we found that HcunGR1, DmelGR21a, and HarmGR1; HcunGR2 and HarmGR2; and HcunGR3, DmelGR63a, and HarmGR3 were grouped together, indicating that these three HcunGRs might play a role in CO2 sensing in H. cunea.Notably, the expression levels of the GRs showed that HcunGR1, HcunGR2, and HcunGR3 were highly abundant in the labial palps (Figure 3a).The expression patterns of HcunGR1, HcunGR2, and HcunGR3 were further verified by q-PCR (Figure 3b), and the qPCR results were consistent with the heatmap.These results indicate that the main organ Notably, the expression levels of the GRs showed that HcunGR1, HcunGR2, and HcunGR3 were highly abundant in the labial palps (Figure 3a).The expression patterns of HcunGR1, HcunGR2, and HcunGR3 were further verified by q-PCR (Figure 3b), and the qPCR results were consistent with the heatmap.These results indicate that the main organ responsible for CO 2 sensing in H. cunea is the labial palp, which is similar to that in other lepidopteran species.

Electrophysiological Response of the Antennae and Labial Palp to CO2
The results of electrolabialpalpography (ELPG) and electroantennogram (EAG) showed that the response to CO2 was significantly higher in the labial palp of female H. cunea than in males (Figure 4a) (p = 0.0313), but there was no significant difference in the antennae response to CO2 (Figure 4b).In addition, the response of the labial palp to CO2 was consistently higher than that of the antennae at 1%-10% concentrations (Figure 4).These findings suggest that the labial palp is the main organ involved in CO2 sensing in H. cunea and that female moths have stronger CO2-sensing abilities than males.

Electrophysiological Response of the Antennae and Labial Palp to CO 2
The results of electrolabialpalpography (ELPG) and electroantennogram (EAG) showed that the response to CO 2 was significantly higher in the labial palp of female H. cunea than in males (Figure 4a) (p = 0.0313), but there was no significant difference in the antennae response to CO 2 (Figure 4b).In addition, the response of the labial palp to CO 2 was consistently higher than that of the antennae at 1%-10% concentrations (Figure 4).These findings suggest that the labial palp is the main organ involved in CO 2 sensing in H. cunea and that female moths have stronger CO 2 -sensing abilities than males.

Two-Electrode Voltage Clamp (TEVC) Response of HcunGr1, HcunGr2, HcunGr3, and Their Combinations
By analyzing the dissolved CO 2 concentration in NaHCO 3 solution, a direct correlation between the concentration of NaHCO 3 and the dissolved CO 2 concentration was identified (Figure S1).Moreover, we found that Na + in NaCl solution also elicits a channel current (Figures 5a-g and S2).Therefore, eliminating the Na + effect from the NaCl solution is necessary when calculating the "real response"; we used full response minus NaCl response to look at the real response of CO 2 [28].Oocytes expressing HcunGR1, HcunGR2, or HcunGR3 alone or with co-expressions of HcunGR1+HcunGR2 or HcunGR2+HcunGR3 did not respond to CO 2 after excluding the effect of Na + (Figure 5c-g).However, the HcunGR1+HcunGR3 and HcunGR1+HcunGR2+HcunGR3 expression sets significantly responded to CO 2 beginning at a concentration of 100 mM (equivalent to 51 ± 3 ppm CO 2 ) (Figure S2), and the response increased gradually with increasing CO 2 concentration (Figure 5a,b).In the range of 100-300 mM, the response of HcunGR1+HcunGR3 ranged from 257 ± 87 nA to 1387.67 ± 162.7 nA, and the response of HcunGR1+HcunGR2+HcunGR3 ranged from 245.67 ± 114.33 nA to 575.3 ± 89.3 nA (Figure 5h).The response of the HcunGR1+HcunGR2+HcunGR3 set was significantly lower than that of the HcunGR1+HcunGR3 set from a concentration of 200 mM or greater (Figure 5i).

Electrophysiological Response of the Antennae and Labial Palp to CO2
The results of electrolabialpalpography (ELPG) and electroantennogram (EAG) showed that the response to CO2 was significantly higher in the labial palp of female H. cunea than in males (Figure 4a) (p = 0.0313), but there was no significant difference in the antennae response to CO2 (Figure 4b).In addition, the response of the labial palp to CO2 was consistently higher than that of the antennae at 1%-10% concentrations (Figure 4).These findings suggest that the labial palp is the main organ involved in CO2 sensing in H. cunea and that female moths have stronger CO2-sensing abilities than males.S4 and S5).S4 and S5).By analyzing the dissolved CO2 concentration in NaHCO3 solution, a direct correlation between the concentration of NaHCO3 and the dissolved CO2 concentration was identified (Figure S1).Moreover, we found that Na + in NaCl solution also elicits a channel current (Figures 5a-g and S2).Therefore, eliminating the Na + effect from the NaCl solution is necessary when calculating the "real response"; we used full response minus NaCl response to look at the real response of CO2 [28].Oocytes expressing HcunGR1, HcunGR2, or HcunGR3 alone or with co-expressions of HcunGR1+HcunGR2 or HcunGR2+HcunGR3 did not respond to CO2 after excluding the effect of Na + (Figure 5c-g).However, the HcunGR1+HcunGR3 and HcunGR1+HcunGR2+HcunGR3 expression sets significantly responded to CO2 beginning at a concentration of 100 mM (equivalent to 51 ± 3 ppm CO2) (Figure S2), and the response increased gradually with increasing CO2 concentration (Figure 5a,b).In the range of 100-300 mM, the response of HcunGR1+HcunGR3 ranged from 257 ± 87 nA to 1387.67 ± 162.7 nA, and the response of HcunGR1+HcunGR2+HcunGR3 ranged from 245.67 ± 114.33 nA to 575.3 ± 89.3 nA (Figure 5h).The response of the HcunGR1+HcunGR2+HcunGR3 set was significantly lower than that of the HcunGR1+HcunGR3 set from a concentration of 200 mM or greater (Figure 5i).S6); The data are presented as the mean ± SEM.

Anterograde Dye Filling of Labial Palps in H. cunea
After we concluded that the labial palp is the main organ for sensing CO2, an anterograde dye-filling experiment on the labial palps was conducted to further explore the transmission of CO2 signals to the central nervous system.We observed the dye's entry from the labial palp nerves, passing through the gnathal ganglion (GNG), then dividing  S6); The data are presented as the mean ± SEM.

Anterograde Dye Filling of Labial Palps in H. cunea
After we concluded that the labial palp is the main organ for sensing CO 2 , an anterograde dye-filling experiment on the labial palps was conducted to further explore the transmission of CO 2 signals to the central nervous system.We observed the dye's entry from the labial palp nerves, passing through the gnathal ganglion (GNG), then dividing into two bundles, and finally arriving at the LPOG (Figure 6a-f).The LPOG is located in the ventral region of the ALs, where it is similar to the DP region in D. melanogaster; we named this region DP1.The projections displayed a clear boundary, and the neurons exhibited a bilateral projection pattern.
Int. J. Mol.Sci.2024, 25, x FOR PEER REVIEW 7 of named this region DP1.The projections displayed a clear boundary, and the neurons e hibited a bilateral projection pattern.
To precisely locate and calculate the volume of the LPOG in the AL, we, based on th nine individuals (five females and four males), performed two-dimensional and thre dimensional reconstructions of all the glomeruli (Figure 6g-j) (Figure S3).It was foun that female H. cunea possess 81 glomeruli, whereas male moths have only 74; most of th glomeruli were roughly spherical, while a few were irregularly shaped, forming a centr fiber nucleus.After computing the surface area and volume of DP1 in both male and f male H. cunea, we found that the total surface area of DP1 in females (5741.25 μm 2 ) w nearly the same as that in males (5728.94μm 2 ); however, the average volume of DP1 females (34,388.3μm 3 ) was significantly (p = 0.0115) greater than that in males (28,852 μm 3 ) (Figure 6k), suggesting that sexual dimorphism existed in the DP1 glomerulus.To precisely locate and calculate the volume of the LPOG in the AL, we, based on the nine individuals (five females and four males), performed two-dimensional and threedimensional reconstructions of all the glomeruli (Figure 6g-j) (Figure S3).It was found that female H. cunea possess 81 glomeruli, whereas male moths have only 74; most of the glomeruli were roughly spherical, while a few were irregularly shaped, forming a central fiber nucleus.After computing the surface area and volume of DP1 in both male and female H. cunea, we found that the total surface area of DP1 in females (5741.25 µm 2 ) was nearly the same as that in males (5728.94µm 2 ); however, the average volume of DP1 in females (34,388.3µm 3 ) was significantly (p = 0.0115) greater than that in males (28,852.4µm 3 ) (Figure 6k), suggesting that sexual dimorphism existed in the DP1 glomerulus.

Discussion
Initially, we discovered that H. cunea adults exhibit a preference for CO 2 concentrations ranging from 1% to 10%; this preference might be linked to the foraging and oviposition behavior of H. cunea.Studies have shown that many plants, such as Nepenthes, release in excess of 5% concentration of CO 2 [29], which is consistent with the CO 2 in our behavioral test.In addition, researchers have found that adults H. cunea prefer to forage and oviposition at night [30], which is the peak time of CO 2 release, which partially explains why the high concentration of CO 2 could also attract H. cunea females.Previous studies have shown that the level of respiratory metabolism in plants is a main indicator reflecting plant quality.A high respiration rate indicates the presence of more nutrients, such as carbohydrates [7]; therefore, H. cunea can choose strong nectar as food by sensing the CO 2 released by plants [31].On the other hand, it has been shown that insect spawning increases with increasing CO 2 concentration [6], suggesting that H. cunea can also choose oviposition sites by CO 2 cues to maximize the survival of their offspring.Field observations have also shown that H. cunea prefer to lay eggs on the dorsal sides of leaves [32], which may be because the undersides of terrestrial plant leaves have more stomata and can generate higher CO 2 concentrations [33].H. cunea can use this CO 2 cue to lay eggs on the dorsal side of the leaf to prevent damage from direct sunlight.In summary, our behavioral results showed that CO 2 can attract adult H. cunea, which may be related to their oviposition behavior and its adaptation.And the possibility of further study about CO 2 baited traps such as mosquitos will be an exciting topic in the future [34].
Homology analysis revealed that three CO 2 receptor homologs exist in H. cunea, which is consistent with findings in other lepidopteran insects [16,17], indicating that the CO 2 sensing pathway might be relatively conserved in Lepidoptera.The expression levels of the three candidate CO 2 receptors demonstrated that they were all specifically expressed in labial palps; however, their expression was low in the antennae.And the response of the labial palp is significantly greater than that of the antennae, suggesting that the labial palp is the primary organ for sensing CO 2 in H. cunea, as is the response of other lepidopteran species, such as H. armigera and M. sexta [18,25].The fact that females produce stronger action potentials than males in the labial palp also suggests that CO 2 may play an important role in female-specific behaviors such as spawn selection in H. cunea.In D. melanogaster, the molecular mechanisms for CO 2 sensing in taste and olfaction are mutually independent [35], and ionotropic receptors (IRs) are involved in the detection of CO 2 [36][37][38].In addition, the response of D. melanogaster to CO 2 is regulated by two different neural pathways, one for CO 2 attraction and the other for avoidance [39].Therefore, it is necessary to further investigate whether there are other CO 2 -sensing pathways in H. cunea.
In vitro expression of HcunGR1, HcunGR2, and HcunGR3 revealed that CO 2 responses occurred only when HcunGR1 was combined with HcunGR3 or when HcunGR1 was combined with HcunGR2+HcunGR3.These results support the crucial role of the HcunGR1+HcunGR3 coexpression in CO 2 sensing in H. cunea, which is consistent with the findings in H. armigera [18].Notably, the response of the HcunGR1+HcunGR2+HcunGR3 ternary complex was significantly lower than that of the co-expression at concentrations greater than 200 mm.The role of HcunGR2 in CO 2 sensing remains unknown.Our results could be due to two possible explanations.One possibility is that the limited number of ion channels on the oocyte surface results in a reduction in the expression ratio of the HcunGR1+HcunGR3 co-expression when HcunGR2 is expressed [40], therefore suppressing the TEVC response of HcunGR1+HcunGR3; another possibility is that HcunGR2 may serve as a modulator, playing an inhibitory role in the CO 2 sensing process in H. cunea.In conclusion, we have shown that the co-expression formed by HcunGR1+HcunGR3 plays a primary role in CO 2 sensing, but the function of HcunGR2 requires further exploration.
At the CNS level, a bilateral projection pattern of labial palps was observed in H. cunea; this result is consistent with what has been observed in H. armigera and Anopheles gambiae (A.gambiae) but differs from the unilateral projection pattern observed in D. melanogaster and A. aegypti.Our heatmap result showed that GR1, GR2, and GR3 were the top three highest expressing receptors in labial palps, which were responsible for CO 2 sensing; thus, the projection from the labial palp into the brain was mainly for CO 2 sensing.But we also found that there are other GRs also expressed in the labial palp, which may be related to wider sensing of different cues [41].And previous research has suggested that bilateral projections enable olfactory receptor neurons (ORNs) to release an asymmetric amount of neurotransmitters on both sides of ALs [42], which leads to a stronger signal in one projection neuron (PN) than in the other, enhancing the contrast of odor concentration gradients between the two brain hemispheres [43].This asymmetrical projection pattern may help H. cunea detect the differences in CO 2 concentrations between the two labial palps and rapidly locate the CO 2 source.Moreover, the average volume of DP1 in females was significantly higher than that of males; however, the total surface area of DP1 was nearly the same in both genders.This indicates that the shape of DP1 is different between males and females, which is consistent with our observations that the male's DP1 had an irregularly shaped (Figure S3).And the sexual dimorphism was observed in the morphology of LPOGs; the average volume of LPOGs was significantly greater in females than in males.It is well accepted that a larger glomerulus indicates greater odor sensitivity due to the greater number of synaptic connections [44,45].An enlarged LPOG may reflect the crucial role of CO 2 sensing in H. cunea females; in contrast, no significant sexual dimorphism was observed on the LPOG in M. sexta [46].In summary, the identified sexual dimorphism of labial palp projections in H. cunea may somewhat explain the differences in CO 2 function between males and females, which might be linked to female-specific behavior such as oviposition.

Insect
A single fall webworm egg mass was collected from a Manchurian ash (Fraxinus mandschurica) tree in Animal and Plant Park, Jinlin Province, China (43 • 86 96.71 N, 125 • 33 29.82 E), and was reared in an artificial incubator (BIC-300 artificial incubator, Boxun, Shanghai, China) at 26 • C, 80% RH and a 19:5 light: dark cycle beginning in 2019.After each 12 generations, the wild population was collected again from the same place and crossed with the laboratory colony for rejuvenation for more than 30 generations.The larvae were fed on the leaves of mulberry trees (Morus alba), and the adults were given a 10% sucrose solution for energy supplementation.

Binary Choice Assay
The airflow of CO 2 (Juyang Company, Changchun, China) was controlled by a flow meter and mixed with Zero Air (21% O 2 and 79% N 2 , CO 2 free, Juyang Company, China) to ensure the delivery of 1%, 3%, 5%, 8%, and 10% CO 2 .A Y-tube olfactometer (1.6 cm in diameter, 7.2 cm in base and arm length) was used for the binary choice assay, and pure air without any CO 2 was used as a negative control.At 7 p.m. (peak spawning), day 3 after emergence, females were selected.A moth was first placed at the entrance of the main arm, and a "choice" was recorded when it entered an arm and stayed for more than 30 s.If no choice was made within 5 min, the data were recorded as "no choice".A total of 60 H. cunea were tested.The Y-tube was cleaned with hexane, and the position of the stimulus was exchanged before and after each test.All the assays were performed in a dark room with red light (Intelligent LED solutions, Berkshire, UK) to avoid light interference.

Electrophysiological Recording
EAG and ELPG were used to detect the electrophysiological responses of the antennae and labial palp to CO 2 in H. cunea.Three days after emergence, females and males were selected at night (oviposition peak time).Glass electrodes were pulled by using a micropipette PC-10 (Narishige, Tokyo, Japan) and then filled with 1 M potassium chloride containing 1% polyvinylpyrrolidone.The reference electrode was subsequently inserted into one eye of the insects, while the recording electrode was placed in contact with the tips of the antennae and labial palp using the micromanipulator MP-12 (Syntech, Kirchzarten, Germany).The obtained signals were amplified by a high-impedance ac/dc preamplifier (Syntech, Kirchzarten, Germany).CO 2 stimuli ranging from 1% to 10% were injected into a carbon-filtered and humidified airflow for 0.2 s to deliver the stimulus to the antenna and labial palp at 500 mL/min generated by an air stimulus controller CS-55 (Syntech, Kirchzarten, Germany).A minimum of 3 individuals were tested, and 3 puffs were performed for each antenna or labial palps.The EAG and ELPG data were acquired with EAG Pro version 2.0 software (Syntech, Kirchzarten, Germany) and normalized by the response of negative control (21% O 2 and 79% N 2 , CO 2 free) (Juyang Company, China) by the equation "relative EAG response = EAG response of CO 2 / EAG response of negative control".The data were subsequently analyzed with GraphPad Prism 6.0 (GraphPad Software, San Diego, CA, USA).

RNA Extraction, Expression Pattern Analysis, Quantitative PCR, and Cloning
Previous studies have shown that some GRs are differently expressed in the olfactory sensing organ (antennae) between males and females, but in non-olfactory organs (such as the head, chest, abdomen, and legs), no sex-based different expressions were detected; thus, we use mix samples of these body parts as a negative control [50].The head, thorax, abdomen, leg, and labial palp carefully dissected from 15 individuals (female:male = 1:1) with DEPC-treated forceps under a stereomicroscope (Motic, Hong Kong, China).Then, female and male antennae were dissected from 15 individuals in the same way.Total RNA was isolated from homogenized body parts with TRIzol reagent (Invitrogen, Carlsbad, CA, USA) following the manufacturer's protocol.After extraction, the total RNA concentration was assessed with a NanoDrop 2000 spectrophotometer (Thermo Fisher Scientific, Waltham, MA, USA) and 1% agarose gel (Sangon Biotech, Shanghai, China) electrophoresis.
For the qPCR assay, 1 µg of total RNA was transcribed into cDNA by using TransScript One-Step gDNA Removal and cDNA Synthesis SuperMix (TransGen Biotech, Beijing, China).qPCR was performed with a LightCycler 480 II Detection System (Roche, Shanghai, China) and TransStar Tip Top Green qPCR Supermix (TransGen Biotech, Beijing, China) under the following conditions: 94 • C for 30 s; 45 cycles of 94 • C for 5 s, 55 • C for 15 s, and 72 • C for 10 s; the β-actin gene was used as an internal control.The primers used in this study are listed in Table S2.The qPCR results were analyzed via the 2 −∆∆CT method [51].The data were subsequently analyzed with GraphPad Prism 6.0 (GraphPad Software, CA, USA).
The primer was designed by Primer 3 (https://bioinfo.ut.ee/primer3-0.4.0, accessed on 29 June 2023) (Table S3), the full ORF of GRs containing 5 UTR and 3 UTR were obtained by PCR and ligated to pUCm-T Vector (Sangon Biotech, Shanghai, China) for sequencing.Subsequently, the ORF of GRs were amplified by a specific primer and subcloned to pGEMHE using pEASY-Uni Seamless Cloning and Assembly Kit (TransGen Biotech, Beijing, China) with BamHI and HindIII restriction sites (New England Biolabs, Ipswich, MA, USA).The recombinant plasmids were transformed in DH5α (TransGen Biotech, Beijing, China) competent cell, and then plasmids were extracted with a SanPrep Column Plasmid Mini-Preps Kit (Sangon Bio, Shanghai, China).After transformation, the inserts were verified via DNA sequencing (Sangon Biotech, Shanghai, China).

CO 2 Quantification and TEVC
Since a TEVC requires a liquid environment, it is not possible to directly measure the response of GRs to CO 2 in air; we chose to quantify the function of GRs by their response to dissolved CO 2 in NaHCO 3 solution, which was previously described by Xu et al. [28].In brief, the concentration of dissolved CO 2 in NaHCO 3 solution can be calculated by the following equation; thus, we can obtain different concentrations of dissolved CO 2 by controlling the pH and the concentration of bicarbonate.
The TEVC technique was used to record the channel currents in Xenopus oocytes at a holding potential of −80 Mv [52].Signals were amplified with an Axonclamp 900A amplifier (Molecular Devices, San Jose, CA, USA) and 50-Hz low-pass filters and digitized at 1 kHz.Data acquisition and analysis were performed using Axon Digi 1550B and pCLAMP10 software (Molecular Devices, San Jose, CA, USA).We used the same concentration of Na + in NaCl solution as negative control and then used full response minus NaCl response to look at the real response of CO 2 .The data were subsequently analyzed with GraphPad Prism 6.0 (GraphPad Software, San Diego, CA, USA).

Anterograde Dye Filling and Immunohistochemical Staining of the Labial Palps
In order to furthermore explore labial palps projection on AL, the anterograde dye filling was performed on labial palp with both genders.According to previous studies [53], the adult insects were fixed in a plastic tube with dental wax so that their heads were exposed.The base of the labial palp was then cut off, and the fluorescent dye, tetramethylrhodamine dextran (MicroRuby, Molecular Probes; Invitrogen, Eugene, OR, USA), was applied at the cutting surface by using a needle.After staining, the insects were placed in a refrigerator with moist filter paper overnight, allowing transportation of the dye to the sensory axons.The next day, the and ventral nerve cord were dissected in Ringer's saline, fixed in 4% paraformaldehyde (PFA) for 1 h, dehydrated via an EtOH gradient, cleared with methyl salicylate, and mounted in Permount on perforated aluminum slides with coverslips.
For immunohistochemical staining, the brains of H. cunea were carefully removed with forceps, fixed with 4% PFA at 4 • C overnight, and then rinsed with PBST three times for 45 min.The fixed brain body part was then transferred to 5% normal goat serum (NGS; Thermo Fisher Scientific, Waltham, MA, USA) and preincubated at 4 • C for 15 h.After preincubation, the primary antibody 3C11 (anti-SYNORF1, 1:100 dilution with 5% NGS and PBST) (DSHB, University of Iowa, Johnson County, IA, USA) was applied and incubated at 4 • C for 5 days.Afterward, the brain was rinsed with PBST again and treated with the secondary antibody Cy2 coupled to an Alexa FluorTM 488 (1:300 dilution with 1% NGS and PBST) (Invitrogen, Eugene, OR, USA) at 4 • C for 3 days.Finally, the brain was dehydrated using an alcohol gradient, cleared with methyl salicylate, stored at 4 • C, and mounted with 1 mm aluminum slides for confocal laser microscopy imaging.
The slides were observed under a laser scanning confocal microscope (LSM880, Carl Zeiss, Jena, Germany) with an excitation wavelength of 488 nm and collected between 490 nm and 560 nm.A clear image was captured with ZEN v2.6.The identified neuropils within the brain and ventral nerve cords were reconstructed by using the 3D reconstruction software Amira 5.4.3 (FEI, Hillsboro, OR, USA) [54].

Statistical Analysis
One-way ANOVA followed by Tukey's test or Dunnett's test was used for multigroup comparisons (Figure 3b).Wilcoxon signed-rank test was used as a comparison between two curves (Figure 4).When two sets of data were compared, the data were first evaluated by the Shapiro-Wilk normality test.If p ≤ 0.05, the data sets will be applied for the Mann-Whitney test.If p > 0.05, the data will subsequently apply to the F-test to check the equal variances.If the data pass the F-test, it will apply to unpaired t-tests (Figures 6k and S2).If they do not pass the F-test, the data will apply to the unpaired t-tests with Welch's correction.The data were analyzed with SPSS 27.0 (IBM Corp., Armonk, NY, USA) and GraphPad Prism 6.0 (GraphPad Software, San Diego, CA, USA).And for electrophysiological and TEVC experiments, the value of each biological replicate is the average of its three technical replicates, and three technical replicates mean that when conducting an antenna and labial palp or oocyte preparation, the same stimulus was applied three times.

Conclusions
Although the detailed biological functions of CO 2 in foraging and oviposition in H. cunea could not be fully clarified, some conclusions were reached at this stage.First, CO 2 (ranging from 1% to 10%) strongly affects H. cunea adults; second, the main organ involved in CO 2 sensing in H. cunea is the labial palp, and female moths have a more sensitive ability to perceive CO 2 than male moths, whereas HcunGR1 and HcunGR3 are indispensable elements in the CO 2 sensing process; and third, sexual dimorphism is observed in the volume of the LPOG, the main CO 2 projection region in the antennal lobe.In summary, these results showed that CO 2 has an attractive effect on H. cunea, which may be related to their oviposition behaviors; by identifying the CO 2 receptors in H. cunea, the olfactory sensing pathway was further clarified.These results would benefit the development of a CO 2 -based trap for the monitoring and control of H. cunea, meanwhile providing more evidence on the biological function of CO 2 among varied insects, especially those pests that take serious damage to the agriculture and forest.

Figure 3 .
Figure 3. Expression profiles of HcunGRs in different body parts of adult H. cunea.(a) Characteristic expression patterns of 44 HcunGRs in different body parts based on FPKM (normalization by row).FPKM: Fragments per kilobase of transcript per million fragments mapped.(b) Expression patterns of three candidate CO2 GRs in different body parts of adult H. cunea.A different lowercase indicates a significant difference based on one-way ANOVA followed by Tukey's multiple comparison test (p < 0.05).The data are presented as the means ± standard errors of the means (SEMs), N = 3.

Figure 3 .
Figure 3. Expression profiles of HcunGRs in different body parts of adult H. cunea.(a) Characteristic expression patterns of 44 HcunGRs in different body parts based on FPKM (normalization by row).FPKM: Fragments per kilobase of transcript per million fragments mapped.(b) Expression patterns of three candidate CO 2 GRs in different body parts of adult H. cunea.A different lowercase indicates a significant difference based on one-way ANOVA followed by Tukey's multiple comparison test (p < 0.05).The data are presented as the means ± standard errors of the means (SEMs), N = 3.

Figure 3 .
Figure 3. Expression profiles of HcunGRs in different body parts of adult H. cunea.(a) Characteristic expression patterns of 44 HcunGRs in different body parts based on FPKM (normalization by row).FPKM: Fragments per kilobase of transcript per million fragments mapped.(b) Expression patterns of three candidate CO2 GRs in different body parts of adult H. cunea.A different lowercase indicates a significant difference based on one-way ANOVA followed by Tukey's multiple comparison test (p < 0.05).The data are presented as the means ± standard errors of the means (SEMs), N = 3.

Figure 4 .
Figure 4. Electrolabialpalpography (ELPG) and electroantennogram (EAG) nonlinear regression curve for CO2.(a) Response of the labial palps of the H. cunea to CO2; (b) response of the antennae of the H. cunea to CO2.Statistical differences were evaluated by the Wilcoxon signed-rank test.* p < 0.05, ns: no significance.The data are presented as the means ± SEMs, N = 3 biological replicates (TablesS4 and S5).

Figure 4 .
Figure 4. Electrolabialpalpography (ELPG) and electroantennogram (EAG) nonlinear regression curve for CO 2 .(a) Response of the labial palps of the H. cunea to CO 2 ; (b) response of the antennae of the H. cunea to CO 2 .Statistical differences were evaluated by the Wilcoxon signed-rank test.* p < 0.05, ns: no significance.The data are presented as the means ± SEMs, N = 3 biological replicates (TablesS4 and S5).

Figure 5 .
Figure 5. Two-electrode voltage clamp recording (TEVC) responses of HcunGR1, HcunGR2, and HcunGR3 alone and in combination with different concentrations of NaCl and NaHCO3.(a) Response of HcunGR1+HcunGR3; (b) response of HcunGR1+HcunGR2+HcunGR3; (c) response of HcunGR1; (d) response of HcunGR2; (e) response of HcunGR3; (f) response of HcunGR1+HcunGR2; (g) response of HcunGR2+HcunGR3; green traces represent the response in NaCl solution; red traces represent the response in NaHCO3 solution; the number involved in (a-g) indicates concentration NaCl and NaHCO3; (h) nonlinear regression curve of HcunGR1, HcunGR2, and HcunGR3 alone and in combination after excluding the influence of Na + .N = 3 biological replicates (TableS6); The data are presented as the mean ± SEM.

Figure 5 .
Figure 5. Two-electrode voltage clamp recording (TEVC) responses of HcunGR1, HcunGR2, and HcunGR3 alone and in combination with different concentrations of NaCl and NaHCO 3 .(a) Response of HcunGR1+HcunGR3; (b) response of HcunGR1+HcunGR2+HcunGR3; (c) response of HcunGR1; (d) response of HcunGR2; (e) response of HcunGR3; (f) response of HcunGR1+HcunGR2; (g) response of HcunGR2+HcunGR3; green traces represent the response in NaCl solution; red traces represent the response in NaHCO 3 solution; the number involved in (a-g) indicates concentration NaCl and NaHCO 3 ; (h) nonlinear regression curve of HcunGR1, HcunGR2, and HcunGR3 alone and in combination after excluding the influence of Na + .N = 3 biological replicates (TableS6); The data are presented as the mean ± SEM.

Figure 6 .
Figure 6.Anterograde dye-filling of labial palps and two-dimensional reconstructions of the antennal lobe (AL) in the DP1 region of H. cunea.(a-c) The central projections of female labial pit organ sensory neurons passing through the gnathal ganglion (GNG) to DP1; (d-f) the central pro jections of male labial pit organ sensory neurons passing through the gnathal ganglion (GNG) to DP1; (g-h) confocal images of the male H. cunea AL glomeruli seen from the ventral view.The

Figure 6 .
Figure 6.Anterograde dye-filling of labial palps and two-dimensional reconstructions of the antennal lobe (AL) in the DP1 region of H. cunea.(a-c) The central projections of female labial pit organ sensory neurons passing through the gnathal ganglion (GNG) to DP1; (d-f) the central projections of male labial pit organ sensory neurons passing through the gnathal ganglion (GNG) to DP1; (g,h) confocal images of the male H. cunea AL glomeruli seen from the ventral view.The sections are from anterior to posterior at a depth of 156 µm.Scale bar = 50 µm; (i,j) confocal images of female H. cunea AL glomeruli taken from the ventral view.The sections are from anterior to posterior at a depth of 142 µm.Scale bar = 50 µm; (k) the volume compares male and female DP1s.The data are presented as the mean ± SEM, N ≥ 4. Statistical differences were evaluated by unpaired t tests.* p < 0.05.