Odorant-Binding Proteins and Chemosensory Proteins in Spodoptera frugiperda: From Genome-Wide Identification and Developmental Stage-Related Expression Analysis to the Perception of Host Plant Odors, Sex Pheromones, and Insecticides

Spodoptera frugiperda is a worldwide generalist pest with remarkable adaptations to environments and stresses, including developmental stage-related behavioral and physiological adaptations, such as diverse feeding preferences, mate seeking, and pesticide resistance. Insects’ odorant-binding proteins (OBPs) and chemosensory proteins (CSPs) are essential for the chemical recognition during behavioral responses or other physiological processes. The genome-wide identification and the gene expression patterns of all these identified OBPs and CSPs across developmental stage-related S. frugiperda have not been reported. Here, we screened for genome-wide SfruOBPs and SfruCSPs, and analyzed the gene expression patterns of SfruOBPs and SfruCSPs repertoires across all developmental stages and sexes. We found 33 OBPs and 22 CSPs in the S. frugiperda genome. The majority of the SfruOBP genes were most highly expressed in the adult male or female stages, while more SfruCSP genes were highly expressed in the larval or egg stages, indicating their function complementation. The gene expression patterns of SfruOBPs and SfruCSPs revealed strong correlations with their respective phylogenic trees, indicating a correlation between function and evolution. In addition, we analyzed the chemical-competitive binding of a widely expressed protein, SfruOBP31, to host plant odorants, sex pheromones, and insecticides. Further ligands binding assay revealed a broad functional related binding spectrum of SfruOBP31 to host plant odorants, sex pheromones, and insecticides, suggesting its potential function in food, mate seeking, and pesticide resistance. These results provide guidance for future research on the development of behavioral regulators of S. frugiperda or other environmentally friendly pest-control strategies.


Introduction
Insects are among the organisms with robust behavioral and physiological adaptations to complex and competitive ecosystems [1,2]. Among their various sensory capacities, insects use chemosensory perception to initiate several behavioral and physiological responses [3]. For sensing chemicals, several chemosensory neurons house specific sensilla

Identification of OBP and CSP Genes in S. frugiperda
We identified 33 OBP and 22 CSP genes in the S. frugiperda sequence (Tables 1 and 2). The OBP sequence multiple alignment performed with DNAMAN was 12.43%, while the CSP was 24.91%. All the OBPs we have found have conserved cysteines C3 and C4, except for OBP18 and OBP23. Twenty-five OBPs belong to the classical OBP group (SfruOBP1- 14, 16, 19-21, 24, 26-28, 30, 31, and 33), provided with six conserved cysteine residues. Among them, SfruOBP10, 16,19,20,27,28, and 30 miss the conserved C1, and SfruOBP9 misses the conserved C2, although classified as a classical OBP given the conservation of the other five conserved cysteines. Three of the OBPs we have identified belong to the minus-C OBP group (SfruOBP17, 22, and 32), provided with four conserved cysteines, missing C2 and C5. SfruOBP15, 25, and 29 have six conserved cysteines, and they also carry an extra cysteine downstream of C6, being classified as Plus-C OBPs. For SfruOBP18, provided with only the conserved C5, and for SfruOBP23, missing conserved cysteines, we cannot determine which OBP group they belong to, although we have identified OBP genes with a high homology with them in NCBI (Supplemental Table S1). Contrary to SfruOBPs, SfruCSPs are highly conserved, provided with four conserved cysteines except for SfruCSP1, which had no conserved cysteines; however, blast in NCBI revealed a CSP with a high homology, so it was identified as a putative CSP (Supplemental Table S1).   Phylogenetic analysis of the CSPs (Figure 2) demonstrated the existence of four clusters based on the homology among the different representatives. Clusters 2 and 4 contain most of the SfruCSPs (12 + 7); among them, most CSPs of Cluster 2 belong to lepidopterans. Interestingly, Cluster 3 contains only CSPs from S. invicta (SinvCSP1-5/8/10/13 -15/17-20). Cluster 1 contains the remaining three SfruCSPs (SfruCSP1/8/21) and four LmigCSPs (LmigCSP10/26/27/31). Like SfruOBPs, most SfruCSPs that are in the same subclade with SlitCSPs show a close evolutionary relationship to S. litura CSPs (Supplemental Table S3

Chromosomal Distribution of OBP and CSP Genes
The chromosomal location map of OBP and CSP genes was plotted with tbtools based on the information of the genome annotation file of S. frugiperda (Figure 3). The identified OBP genes of S. frugiperda are located on 10 different chromosomes and on contig ctg319_4 (which is a different contig not yet defined, belonging to a specific chromosome), and the majority of the SfruOBPs are distributed in clusters on chromosomes. For instance, nine genes from Cluster 1 (SfruOBP5/6/11/14/21/25/26/29/33) are located in proximity on chro-

Chromosomal Distribution of OBP and CSP Genes
The chromosomal location map of OBP and CSP genes was plotted with tbtools based on the information of the genome annotation file of S. frugiperda ( Figure 3). The identified OBP genes of S. frugiperda are located on 10 different chromosomes and on contig ctg319_4 (which is a different contig not yet defined, belonging to a specific chromosome), and the majority of the SfruOBPs are distributed in clusters on chromosomes. For instance, nine genes from Cluster 1 (SfruOBP5/6/11/14/21/25/26/29/33) are located in proximity on chromosome 10, while the remaining two genes of Cluster 1 are instead distributed on chromosome 20 (SfruOBP10) and chromosome 22 (SfruOBP19). One SfruOBP gene from Cluster 2 is located on chromosome 2 (SfruOBP23), and two SfruOBP genes from the same  Contrary to SfruOBPs, the distribution of SfruCSPs seems to be more centralized, having identified loci on only two chromosomes: two genes of Cluster 2 (SfruCSP2/20) are located on adjacent loci of chromosome 19, while the remaining 21 genes of Clusters 1, 2, 3, and 4 are all located on chromosome 8 ( Figure 4). On this chromosome, most of the genes fall in proximal loci, with the exception of SfruCSP1 and SfruCSP12, located at 3.5 Mbp and 6.6 Mbp, respectively, on the chromosome.

Intron/Exon Organization of OBP and CSP Genes of S. frugiperda
The exon/intron structure varies greatly within SfruOBP ( Figure 5) and S ure 6) genes from different clusters.

Gene Expression Analysis of OBPs and CSPs of S. frugiperda
Our transcriptomic dataset was used to create a heatmap revealing expression levels of SfruOBPs and SfruCSPs across the developmental stages of egg, first to sixth instars (identified by the size of the larvae), and pupae, as well as adult males and females ( Figure 9). Transcriptomic analysis revealed that the expression of almost all of the genes of Cluster 1 was higher in the first instar and decreased in the second to sixth instars and the pupal stage, but recovered to a higher level in the adult stage. However, for SfruOBP19, expression appeared stable in both the larval and adult stages, and SfruOBP11 showed high expression only in the first instar and maintained a low expression for all the other developmental stages. Two genes in Cluster 2 (SfruOBP17/18) were stably expressed at all stages except eggs and the first instar, while SfruOBP23 was expressed at instar one through five and showed a decreasing trend and absent to low expression from the pupal to adult stages. The expression of SfruOBP7 and SfruOBP31 from Cluster 3 was higher in the first instar. Three other genes (SfruOBP8/22/32) were highly expressed in the pupae and adults. Two genes of Cluster 4 (SfruOBP20/27) and two genes of Cluster 5 (SfruOBP16/30) were highly expressed in both the male and female adults. SfruOBP20 was also highly expressed in the first and fourth instars, and SfruOBP1 in Cluster4 was highly expressed in the first to third instars. Some of the genes from Cluster 5 were highly expressed in the egg stage, while all genes from this cluster were highly expressed in adults. Genes of Cluster 5 were also moderately to highly expressed in several earlier or later larval stages: SfruOBP3/13/15 were highly expressed in earlier larval stages (instars 1-3), SfruOBP4/12 were highly expressed in later larval stages (instars 4-5), while SfruOBP2 was expressed in all instars, and SfruOBP9/24/28 were also expressed in most stages of the larva. It is worth mentioning that SfruOBP9 was highly expressed in males but low in females.

Gene Expression Analysis of OBPs and CSPs of S. frugiperda
Our transcriptomic dataset was used to create a heatmap revealing expression levels of SfruOBPs and SfruCSPs across the developmental stages of egg, first to sixth instars (identified by the size of the larvae), and pupae, as well as adult males and females ( Figure  9). Transcriptomic analysis revealed that the expression of almost all of the genes of Cluster 1 was higher in the first instar and decreased in the second to sixth instars and the pupal stage, but recovered to a higher level in the adult stage. However, for SfruOBP19, expression appeared stable in both the larval and adult stages, and SfruOBP11 showed high expression only in the first instar and maintained a low expression for all the other developmental stages. Two genes in Cluster 2 (SfruOBP17/18) were stably expressed at all stages except eggs and the first instar, while SfruOBP23 was expressed at instar one through five and showed a decreasing trend and absent to low expression from the pupal to adult stages. The expression of SfruOBP7 and SfruOBP31 from Cluster 3 was higher in the first instar. Three other genes (SfruOBP8/22/32) were highly expressed in the pupae and adults. Two genes of Cluster 4 (SfruOBP20/27) and two genes of Cluster 5 (Sfru-OBP16/30) were highly expressed in both the male and female adults. SfruOBP20 was also highly expressed in the first and fourth instars, and SfruOBP1 in Cluster4 was highly expressed in the first to third instars. Some of the genes from Cluster 5 were highly expressed in the egg stage, while all genes from this cluster were highly expressed in adults. Genes of Cluster 5 were also moderately to highly expressed in several earlier or later larval The majority of the SfruCSPs showed a more complex expression tendency across the developmental stages and sexes ( Figure 10) when compared with SfruOBPs. In Cluster 1, SfruCSP21 gradually decreased in expression level from eggs to the sixth instar but was highly expressed in pupae and adults, SfruCSP8 was highly expressed only in adults, and SfruCSP1, from a low to absent expression in eggs, maintained a moderate expression in all developmental stages. All the genes from Cluster 2 were moderately or highly expressed from the egg to pupal stages. However, SfruCSP11/16 showed low expression in the sixth instar, SfruCSP17 showed low expression in the fourth to sixth instars and in the egg stage, SfruCSP5 showed low expression in the fourth instar, and SfruCSP4 showed a moderate expression for all larval stages except for the sixth instar. Two genes showed a relatively low expression in pupae (SfruCSP6/13), and only three genes of this cluster exhibited high expression levels in adult males and females (SfruCSP4/6/17). Within the same cluster, SfruCSP2, after its high expression in eggs, gradually increased expression levels from the first instar to the pupal stage, dropping to low levels in adults. Within the same cluster, SfruCSP6 showed an overall high expression except for the first instar and the pupae, while SfruCSP20 showed a sort of off-on expression pattern throughout all the developmental stages of the insect, from eggs to adults. Except for SfruCSP15/22, all genes from Cluster 4 were highly expressed from the first instar to the fifth instar, where SfruCSP18 showed a reduced expression. Expression of the genes from Cluster 4 seems to be consistent in both adult males and females. Interestingly, there was low expression of SfruCSP7/9/14 in pupae and of Sfru3/7/18 in the sixth instar. SfruCSP22 was highly expressed in adult females and showed a lower expression across the other developmental stages.     Among the examined genes, the expression patterns of SfruOBP32 ( Figure S1A), SfruCSP3 ( Figure S1B), SfruCSP19 (Figure S1B), and SfruCSP20 ( Figure S1B) were similar across all developmental stages comparing the qRT-PCR data with the transcriptomic data. The expression of SfruOBP24 ( Figure S1A) in four stages (fifth instar, sixth instar, pupae, and female adult) was consistent between the qRT-PCR and transcriptomic data, but the qRT-PCR analysis also revealed moderate expression of SfruOBP24 at the fourth instar, third instar, first instar, and egg stages and high expression of SfruOBP24 at the second instar and the male adult stages. The qRT-PCR results revealed an adult male biased expression pattern of SfruOBP22. The expression of five stages (L2, L3, L4, L6, and AM) of SfruOBP17 was also consistent between the qRT-PCR and transcriptomic data ( Figure S1A).

Discussion
S. frugiperda is an important and invasive agricultural pest whose destructiveness occurs across life stages, behavior patterns, and physiological changes [48]. In this paper, by performing a genome-wide screening of S. frugiperda, we have unveiled its asset of OBPs and CSPs. Based on a phylogenetic analysis of SfruOBPs and SfruCSPs, we compared the gene structures, polypeptide sequence motif patterns, and patterns of gene expression across all developmental stages of the insect, from eggs to adults. We chose one of the highest expressed binding proteins of S. frugiperda, SfruOBP31, that we heterologously expressed and purified from transgenic E. coli ( Figure S2) to assess chemical-competitive binding assays demonstrating binding to several ligands among odors emitted from the host plant Z. mays, sex pheromones, and some of the main pesticides for this insect.
In the last decades, most of the investigation of lepidopterans targeted their chemoreceptors [49][50][51][52]. Among various reports, some provided in-depth analysis of specific OBPs [53,54], CSPs [18,55], or both [30,56]. However, to our knowledge, this is among the few [57], or may be the sole study in lepidopterans, in which a systematic investigation of soluble proteins merges with a variegate analysis including genomic loci, gene and polypeptide structures, expression patterns across every developmental stage, and functional evidence of semiochemical/pesticide binding.
By mediating the ligand transport through the sensillar endolymph to the activation

Discussion
S. frugiperda is an important and invasive agricultural pest whose destructiveness occurs across life stages, behavior patterns, and physiological changes [48]. In this paper, by performing a genome-wide screening of S. frugiperda, we have unveiled its asset of OBPs and CSPs. Based on a phylogenetic analysis of SfruOBPs and SfruCSPs, we compared the gene structures, polypeptide sequence motif patterns, and patterns of gene expression across all developmental stages of the insect, from eggs to adults. We chose one of the highest expressed binding proteins of S. frugiperda, SfruOBP31, that we heterologously expressed and purified from transgenic E. coli ( Figure S2) to assess chemical-competitive binding assays demonstrating binding to several ligands among odors emitted from the host plant Z. mays, sex pheromones, and some of the main pesticides for this insect.
In the last decades, most of the investigation of lepidopterans targeted their chemoreceptors [49][50][51][52]. Among various reports, some provided in-depth analysis of specific OBPs [53,54], CSPs [18,55], or both [30,56]. However, to our knowledge, this is among the few [57], or may be the sole study in lepidopterans, in which a systematic investiga-tion of soluble proteins merges with a variegate analysis including genomic loci, gene and polypeptide structures, expression patterns across every developmental stage, and functional evidence of semiochemical/pesticide binding.
By mediating the ligand transport through the sensillar endolymph to the activation of chemosensory transmembrane proteins [7,8], OBPs and CSPs play a crucial role in insect behavioral and physiological adaptation, including food seeking, reproduction, and pesticide resistance [58].
Belonging to two multigene families, OBPs and CSPs generally differ in numbers and functionalities across insect species [21]. To our knowledge, this is among the first studies identifying OBPs and CSPs with the in-sight/in-depth view of the genome of S. frugiperda, wherein we demonstrate the existence of 33 genes for OBPs and 22 genes for CSPs. Our findings are consistent with a previous study conducted on S. frugiperda where 36 OBPs and 21 CSPs were identified [45], with 22 SfruOBPs and 19 SfruCSPs identical with our study (Supplemental Table S4) and comparable with the number of the respective genes for S. exigua (34 OBPs, 20 CSPs) [59].
Taking all binding proteins together, S. frugiperda expresses fewer OBPs and CSPs than B. mori (44 OBPs, 20 CSPs) [33,57] and S. litura (45 OBPs, 23 CSPs) [60]. The SfruOBPs were classified according to Hekmat-Scafe et al. [61]. SfruOBPs with Cys at only six conserved sites were classified as classical OBP, while SfruOBPs in the absence of C2 and C5 were classified as minus-C OBP. SfruOBPs with more than six conserved Cys were classified as plus-C OBPs. Among the 33 identified SfruOBP genes, 22 belonged to the classical group; only three SfruOBPs, which lacked C2 and C5, belonged to the minus-C group (SfruOBP22/32); and three genes belonged to the plus-C group (SfruOBP15/25/29). This is comparable to S. litura, in which only classical, minus-C, and plus-C OBPs were identified [62,63]. However, one dimer OBP was annotated in Danaus plexippus [53]. The majority of the SfruCSPs contained four conserved Cys, and only three CSPs contained three conserved Cys.
Our phylogenetic analysis demonstrates evolution relationships of OBPs and CSPs between S. frugiperda and other species of Lepidoptera; specifically, between S. frugiperda and S. litura (Figure 1). According to the phylogenetic tree, we found five clusters of SfruOBPs, among which some OBPs from the same cluster have similar gene structures and expression patterns, while some OBPs from different clusters present different gene structures and expression patterns. The phylogenetic tree of CSPs was divided into four clusters, of which only three out of four contain CSPs from S. frugiperda. Among these, Cluster 2 and Cluster 4 contain most of the SfruCSPs, as a possible indication of their similarity in evolution and possibly as evidence of differences in foraging behavior or adaptation to the environment between S. frugiperda and S. litura [64].
Chromosomal location, gene structure, and protein motif analysis may help unveil potentials from gene evolution such as duplication, reversal, or skipping [62,65] to functional conservation. In general, about half of the SfruOBPs are distributed on multiple chromosomes, whereas the SfruCSPs are distributed on only two chromosomes, and most of them are clustered on adjacent loci of one chromosome, which may be due to the high conservation of the gene family in S. frugiperda. In addition, the SfruOBP genes display divergent patterns of intron/exon organization ( Figure 5), which, although speculative, may widen a potential for alternative splicing, as known for a different class of receptors in lepidopterans [66,67], and motif patterns that are only partially conserved among the representatives belonging to the same clusters (Figure 7). The evolution of the OBP gene family is suggested to follow the birth-and-death model through the pseudogenization or the functional divergence of the duplicate gene during duplication [68,69]. However, the adjacent genes on chromosomes could be involved in analogous functions. For instance, 10 of 24 OBPs located in the social chromosome of S. invicta may participate in the behavioral modulation between the monogyne colony and the polygyne colony [70]. Unlike SfruOBPs, the majority of SfruCSPs from four clusters are instead distributed in proximity on the same chromosome (chromosome 8), evidence of their possible origins from gene duplication.
In addition, several SfruCSP genes display similar intron/exon organization (Figure 6), and analysis of their expressed amino acid sequences demonstrates the existence of an identical motif pattern, 4-3-1-2 ( Figure 8). SfruCSP2 and SfruCSP20 are located in a separate chromosome, suggesting their divergent functions. As a homolog to SfruCSP20, SlitCSP3 was revealed to function in plant defensive metabolites recognition [71]. The functions of the chromosomal location close and far SfruOBPs/SfruCSPs are worthy of further study to verify their function evolution.
Transcriptomic analysis unveiled that six genes of OBPs of the SfruOBP-Cluster 5 (SfruOBP2/3/4/12/13/15), five genes of OBPs of the SfruOBP-Cluster 3 (SfruOBP7/8/22/31/ 32), two genes of OBPs of the SfruOBP-Cluster 4 (SfruOBP20/27), two genes of OBPs of the SfruOBP-Cluster 5 (SfruOBP16/30), and two genes of OBPs of the SfruOBP-Cluster 1 (SfruOBP10/19) showed high expression in adults, suggesting that they have important functions in the adult stage ( Figure 9). RNA interference of highly expressed OBP1 in the adult antennae of Culex quinquefasciatus led to significantly reduced electrophysiological responses to egg-laying attractants, indicating that CquiOBP1 may mediate spawning behavior [72]. Although speculative, this may suggest potential olfactory functions of the OBPs that are highly expressed in adults' SfruOBPs and their involvement in behaviors such as spawning but also courtship or reproduction. The majority of SfruOBPs showed low expression at the sixth instar, whereas we demonstrated higher expression for representatives of Cluster 2. We assume that the lower expression of SfruOBPs at this larval stage may be associated with delayed feeding as a prepupal physiological feature from older larvae. As expected, most SfruOBPs showed low or moderate expression in the egg stage, with the sole exception of SfruOBP12, which may be associated with some sort of role in embryo development like in Galeruca daurica, where GdauOBP28 also showed an exclusively high expression in the egg stage [73]. Contrary to our expectations, the expression of several, if not the majority, of SfruOBPs was higher in the earlier instars than in the older ones. Among these, SfruOBP11 was the OBP with the highest expression in the first instar, which, hypothetically, may suggest its role in conspecific recognition or defensive mechanisms at the larval stage. Indeed, studies have shown the role of OBP in binding chemical ligands and regulating behavior, like the larval-specific OBP of S. exigua, SexiOBP13, which showed a high binding ability to the sex pheromone, Z9, and E12-14:Ac [74]. In the pea aphid Acyrthosiphon pisum, ApisOBP3 is expressed in old larvae and in apterous adults [75] to function in short-term defensive responses like feeding cessation and dropping from host plants in a phase of the behavioral response to the alarm pheromone (E)-β-farnesene [76]. In addition, OBPs from the same clusters have similar gene structures, which can help us to classify OBPs that may have similar functions for further study. For example, SfruOBP2/3/4/12/13/15 from Cluster 5 present the 4-1-5-3 motif pattern (Figure 7), and all contain loci clustered on the same chromosome, organizing three exons, as an indication of their similarity in function.
Analyzing CSP expression, we demonstrated that most of the CSPs we identified were expressed throughout the whole life cycle of S. frugiperda, especially among the various instars ( Figure 10). Compared with OBPs, CSPs were more abundantly expressed at different developmental stages, which may be associated with their broader physiological functions and tissue distribution than OBPs, being involved in leg regeneration [77] and sucking [20]. The expression of some of the genes from Cluster 2 (SfruCSP2/5/10/11/12/13/16/19/20) was lower in the adult stages, but in comparison, it was higher in the later larval stages. These genes may be related to the developmental-stage-dependent behavioral or physiological adaptations in older larvae. Genes of Cluster 4 (SfruCSP15/22) and SfruCSP8 were highly expressed during adulthood, indicating a function in behaviors such as courtship, oviposition, or reproduction [78]. For example, the adult expression with a female bias of SfruCSP22 suggests this gene's involvement in ligand binding to the neuronal modulation for behaviors forming the basis of choosing suitable sites for egg-laying or other femalespecific activities [79]. Several SfruCSPs were expressed in the egg stage, among which the highest expression belonged to the CSPs from Cluster 2 (SfruCSP2/6/11/13). Studies have shown that CSP has odor recognition functions, such as the AlinCSP1-3 in the alfalfa plant bug, Adelphocoris lineolatus, binding to several host-related ligands including (Z)-3hexen-1-ol, (E)-2-hexen-1-al, and valeraldehyde, as a possible involvement of these CSPs in host recognition [80]. Beyond chemosensation, these CSPs may have functional properties forming the basis of embryonic development, physiologically or morphologically [81]. For example, the honey bee gene AmelCSP5 is associated with the formation of the embryonic covering membrane [82].
Previous reports on Spodoptera OBPs provides a blueprint for the prediction of SlitOBP ligands based on the interaction of phylogeny and chemical structure. For example, larval SlitGOBP2 of S. litura have been reported to function in sex pheromone recognition [83]. In the same species, OBP11 displays strong binding with sex pheromones [84] and CSP8 can bind with the oviposition deterrent chemical rhodojaponin III [85]. CSP18 binds directly to chlorpyrifos/fipronil and CSP6 to chlorpyrifos, emamectin benzoate, and fipronil [86]. In addition, three CSPs of S. litura (SlitCSP11, 3, and 8) that are highly expressed in the midgut of the insect were shown to bind to host plant chemicals, and their expression levels varies depending on the host plants [80]. In this study, we verified the function of a widely expressed OBP, SfruOBP31, using a chemical-competitive binding assay based on measurements of 1-NPN-associated fluorescence ( Figure 11). The reason we chose SfruOBP31 for binding assays is its high expression in the first and fifth larval instars and in adults ( Figure 9). Indeed, these developmental stages were reported to be involved in some remarkable physiological and behavioral adaptions, including the dispersing behavior of the first instar [87], higher pesticide resistance in elder larvae [88], and the egg laying/mate seeking behavior of adults [89,90]. Our phylogenetic analysis clustered SfruOBP31 in Cluster 3, where, according to our transcriptomic analysis (Figure 9), it is represented by OBPs of S. frugiperda with the high expression patterns in the first instar or in the fifth instar and adults, suggesting its potential multiple functions, including the dispersing behavior in the first instar, higher pesticide resistance in elder instars, and the egg laying/mate seeking behavior of adults. Among other developmental stages, SfruOBP31 was also low to moderately expressed. In addition, SfruOBP31 is the ortholog of DmelOBP69a in Drosophila, sharing 81% query cover and 28.57% identity, and is involved in social interactions by the detection of contact sex pheromones [47]. Among lepidopterans, sex pheromones are key to mating behavior in adults, and previous studies demonstrated that the sex pheromone mixture of S. frugiperda is composed of (Z)-9-tetradecen-1-yl acetate (Z9-14:Ac), (Z)-7-dodecen-1-yl acetate (Z7-12:Ac), (Z)-9-dodecen-1-yl acetate (Z9-12:Ac), and (Z)-11-hexadecen-1-yl acetate (Z11-16:Ac) in an 81:0.5:0.5:18 ratio [91]. Here, we tested the binding ability of SfruOBP31 to these four sex pheromones. Competitive binding assays with 1-NPN demonstrated that SfruOBP31 selectively binds with two sex pheromones ((Z)-11-hexadecenyl acetate and (Z)-9-tetradecenyl acetate); this may be due to the high content of these two compounds in the sex pheromone mixture of S. frugiperda. Binding of a larval-specific OBP to odorant pheromones is not surprising in the genus Spodoptera. Indeed, a larval-specific OBP of S. exigua, SexiOBP13, shows strong binding capacity to the female sex pheromone Z9,E12-14:Ac that can induce the attraction behavior in larvae [74]. However, evidence of the expression of this OBP also in adults with a similar level between males and females ( Figure 9) may suggest involvement of this binding protein to different ecological mechanisms. The insect olfactory system adapts to the environment by detecting specific chemical volatiles [92]. S. frugiperda can locate their host plants through plant odors. Previous studies have revealed that many plant odors attract S. frugiperda, such as linalool [93]. Therefore, we measured the binding capacity of SfruOBP31 to five host plant volatiles ( Figure 11). Our findings demonstrate that our host plant odorants linalool, cis-3-hexenyl acetate, 1-nonanol, and decanal associate fluorescent decrement in competitive assays conducted with 1-NPN with the following Ki values: 4.04, 6.38, 14.17, and 19.56 µM, respectively. These findings provide evidence of the ability of SfruOBP31 to recognize odors emitted from host plants while the neonate larvae are emerging from the egg mass.
OBP has been shown to bind to insecticide components and to be involved in the increment of resistance. For instance, GOBP2 is responsible for chlorpyrifos tolerance in S. litura [94]. In other insects like Nilaparvata lugens, OBP3 is involved in the resistance to nitenpyram and sulfoxaflor [15]. In our binding trials we also tested the binding capacity of SfruOBP31 to five pesticides commonly used in the control of S. frugiperda, demonstrating that SfruOBP31 had the strongest binding capacity to lambda-cyhalothrin (Ki = 2.78 µM) and chlorfenapyr (Ki = 8.1 µM). Evidence of binding to pheromones, plant odorants, and pesticides suggests multiple functions for SfruOBP31 in host plant recognition and pesticide resistance, and as demonstrated for its ortholog in Drosophila, DmelOBP69a, involved in social interaction based on binding with sex pheromones [47].
In conclusion, we performed a systematic identification of SfruOBPs and SfruCSPs based on the in-sight/in-depth view of the genome and analyzed their expression patterns across all different developmental stages and adult males and females. The different expression patterns of some SfruOBPs and SfruCSPs have developmental stage specificity or sex biases, suggesting the potential functions of these genes. These results lay the groundwork for future investigation of their in vivo functions with methods such as CRISPR/Cas9 or modified RNAi [95,96]. The in vitro chemical competitive binding assay of SfruOBP31 further confirmed the involvement of some OBPs or CSPs in multiple functions such as host plant plus pheromone binding or pesticide resistance. Our study opens investigations on behavioral and physiological adaptation mechanisms of the Fall armyworm S. frugiperda for the final target to develop alternative control strategies to interfere with the behavior or other physiological functions of this pest.

Gene Identification
The amino acid sequences of OBPs and CSPs from B. mori, S. invicta, D. melanogaster, L. migratoria, and S. litura were selected as templates to search against the genome of S. frugiperda. Homology searching was conducted using the tBLASTn programs with the E-value cut-off of 0.0001 to retrieve alternative protein sequences of S. frugiperda OBPs and CSPs. HMMER software (version 3.0) [97] was used in search of unique domains of the corresponding gene family in the alternative protein sequences. Genes that contain a conserved PBP_GOBP domain for the OBP genes (accession number pfam01395) and a conserved OS-D domain for the CSP genes (accession number pfam03392) were maintained. These genes were then manually verified based on the conserved Cys residue characteristics of lepidopterans' OBP genes (C1-X 25-30 -C2-X 3 -C3-X 36-42 -C4-X 8-14 -C5-X 8 -C6) and CSP genes (C1-X 6 -C2-X 18 -C3-X 2 -C4) [30]. After filtering the sequences without characteristic domains, the protein sequences belonging to this gene family were kept for further analysis (Table S5).

Sequence Alignment and Phylogenetic Tree Construction
The phylogenetic tree of S. frugiperda, L. migratoria, D. melanogaster, S. litura, B. mori, and S. invicta OBP and CSP genes was constructed to reveal the evolutionary relationship of SfruOBPs and SfruCSPs among species. Amino acid motifs were identified using MEME [98]. The amino acid multiple sequence alignment (MSA) was made using the muscle method with MEGA 11.0, from which we estimated the best-fitting model of amino acid substitution [99]. Phylogenetic trees were built using MEGA 11.0, where branch lengths were optimized, and branch supports were calculated using bootstrapping with 1000 replicates of neighbor joining (NJ). To facilitate future research, the OBP and CSP genes of S. frugiperda were named based on the homology of genes from other species found using blastx in NCBI.

Chromosomal Distribution, Gene Structure, and Motif Analysis of OBPs and CSPs
Chromosomal location information of SfruOBP and SfruCSP genes was downloaded from genome data [44]. We visualized the data using TBtools [100], including the location and length of the genes on the chromosome and the visualization of the exon-intron structure of the SfruOBP and SfruCSP genes. To discover the motif pattern of SfruOBPs and SfruCSPs, 33 OBPs and 22 CSPs were submitted to the MEME (version 5.5.1) online tool (https://meme-suite.org/meme/, accessed on 9 December 2022) server and a Zero-or-One Occurrence per Sequence (zoops) distribution pattern was adopted.
4.4. Samples Preparation, RNA Extraction, and RNA-seq S. frugiperda were reared individually in plastic boxes with fresh corn leaves. Insects were maintained at 28 ± 3 • C under 70-75% relative humidity and a photoperiod of 16 h light, 8 h dark cycles. Adults of S. frugiperda were reared in 30 cm × 40 cm mesh cages supplied with fresh 10% honey water and corn seedlings for egg laying. Eggs were collected in 24 h after they were laid. Synchronized 1st-to-6th instars were collected according to their size. Three-day pupae and unmated adults were also collected. For each replication, RNA was extracted from 300 eggs and the whole body of 200, 100, 50, 10, 5, and 5 larvae at their 1st, 2nd, 3rd, 4th, 5th, and 6th stages, respectively; 5 pupae; 5 adult females; and 5 adult males. All stages were collected with three replications.
Trizol Reagent (Invitrogen, Waltham, MA, USA) was used for total RNA extraction according to the manufacturer's instructions. Genome DNA was extracted using Turbo DNase (Thermofifisher, Waltham, MA, USA). Gel extraction was performed, and a Nanodrop ND-1000 spectrophotometer (LabTech, Hopkinton, MA, USA) was used to assess the quality and quantity of the RNA samples. RNA libraries were constructed and sequenced as described previously [101]. RNA-seq was performed by a commercial company (Lianchuan, Hangzhou, China) with a next-generation sequencing platform (Illumina NovaseqTM 6000). Differentially expressed genes across all developmental stages were analyzed using the DESeq R package (1.10.1).

q-RT-PCR Verification of Selected SfruOBPs and SfruCSPs
Beacon Designer Software (Palo Alto, CA, USA) was used to design primers (Table S6), with RPL32 (Ribosomal Protein L32) as the reference gene. Q-RT-PCR experiments were performed in accordance with the Minimum Information Required for Publication of Quantitative Real-Time PCR guidelines. A Premix Ex TaqTM II (Tli RNaseH Plus) Kit (Takara, Shiga, Japan) was used as a reagent for qRT-PCR analysis. The reaction conditions were set as follows: 95 • C for 3 min, 40 cycles of 95 • C for 15 s, and 60 • C for 30 s. Reactions were carried out in triplicate, followed by dissociation in the iCycler iQ™ Real Time PCR Detection System (Bio-Rad, Hercules, CA, USA). All developmental stages and sexes were performed with three separate biological replicates. The data were analyzed according to 2 −∆∆CT [∆∆Ct = ∆Ct (test) − ∆Ct (calibrator)]. Four SfruOBPs (32/24/22/17) and three SfruCSPs (3/19/20) were selected for q-RT-PCR assay as representatives that are specifically expressed at different development stages.

Heterologous Expression and Purification of SfruOBP31 in Bacteria
We expressed SfruOBP31 from S. frugiperda in bacteria using a prokaryotic expression system. The predicted signal peptide was removed to produce a properly folded protein.
The ORF (Open Reading Frame) of SfruOBP31 was cloned into a pET28a vector (Novagen, Darmstadt, Germany) and transformed into an Escherichia coli BL21 strain. Individual positive colonies were incubated until the OD (600 nm) reached 0.6-0.8. Protein expression was induced with 0.5 mM IPTG (isopropyl-β-D-1-thiogalactopyranoside) at 16 • C with overnight shaking at 220 rpm/min. Cells were harvested using centrifugation at 3000× g and resuspended in HEPES buffer (10 mM HEPES, 100 mM NaCl, pH 7.5). The supernatant and pellet were then separated after sonication and centrifugation at 14,000× g and 4 • C for 30 min. SDS-PAGE was used to demonstrate that the recombinant protein was soluble. The solution was next applied to his-trap affinity columns (Cobalt Chelating Resin, GBiosciences, St. Louis, MO, USA). Bound protein was eluted with HEPES buffer containing progressively increasing amounts of imidazole from 50 mM to 500 mM. The fractions with the recombinant protein were pooled and dialyzed three times against 3 L of HEPES buffer at 4 • C overnight after electrophoretic analysis. To confirm the size of the recombinant protein, his-tag antibody (Thermofisher, Waltham, MA, USA) was used to perform Western blot. The recombinant protein was stored at −80 • C until use.