Genome-Wide Analysis of the Odorant Receptor Gene Family in Solenopsis invicta, Ooceraea biroi, and Monomorium pharaonis (Hymenoptera: Formicidae)

Olfactory systems in eusocial insects play a vital role in the discrimination of various chemical cues. Odorant receptors (ORs) are critical for odorant detection, and this family has undergone extensive expansion in ants. In this study, we re-annotated the OR genes from the most destructive invasive ant species Solenopsis invicta and 2 other Formicidae species, Ooceraea biroi and Monomorium pharaonis, with the aim of systematically comparing and analyzing the evolution and the functions of the ORs in ant species, identifying 356, 298, and 306 potential functional ORs, respectively. The evolutionary analysis of these ORs showed that ants had undergone chromosomal rearrangements and that tandem duplication may be the main contributor to the expansion of the OR gene family in S. invicta. Our further analysis revealed that 9-exon ORs had biased chromosome localization patterns in all three ant species and that a 9-exon OR cluster (SinvOR4–8) in S. invicta was under strong positive selection (Ka/Ks = 1.32). Moreover, we identified 5 S. invicta OR genes, namely SinvOR89, SinvOR102, SinvOR352, SinvOR327, and SinvOR135, with high sequence similarity (>70%) to the orthologs in O. biroi and M. pharaonis. An RT-PCR analysis was used to verify the antennal expression levels of these ORs, which showed caste-specific expression. The subsequent analysis of the antennal expression profiles of the ORs of the S. invicta workers from the polygyne and monogyne social forms indicated that SinvOR35 and SinvOR252 were expressed at much higher levels in the monogyne workers than in the polygyne workers and that SinvOR21 was expressed at higher levels in polygyne workers. Our study has contributed to the identification and analysis of the OR gene family in ants and expanded the understanding of the evolution and functions of the ORs in Formicidae species.


Introduction
The formation of new genes, which has contributed tremendously to the evolution of developmental programs in various organisms, is a major reason for the generation of biodiversity [1,2]. New genes can be formed in many ways, including through gene duplication, exon-shuffling, gene fission-fusion, and de novo origination [3]. The genes derived from the same ancestor, called homologous genes, are divided into three types: orthologs, xenologs, and paralogs. Orthologs are genes that have evolved as a result of 3 of 17 genotype Gp-9(Bb). The workers carrying the Gp-9(Bb) genotype recognize a specific CHC secreted by a queen with a b-allele at the Gp-9 locus. With the assistance of Gp-9, this unknown compound activates OR-expressed OSNs and guides the Bb workers to kill the BB queens, while Gp-9(Bb) queens are spared [30]. Hence, the workers bearing the allele Gp-9(b) may distinguish odorants from BB and Bb queens [31].
Although genome-sequencing technology has greatly improved, it is difficult to identify all ant ORs because the gene family is so large. In this study, we re-annotated the ORs in S. invicta, and O. biroi, and annotated the ORs in M. pharaonis for the first time using publicly available genome data with the aim of further understanding and comparing the evolution and the functions of the OR genes in ants [26]. We constructed a phylogenetic tree of the ORs in these three ants and analyzed their chromosomal localization and evolution. Moreover, we analyzed the antennal expression profiles of the ORs in the workers, using S. invicta transcriptome data [26]. Our systematic analysis of the ORs in Formicidae species will assist in future efforts to functionally characterize the ORs in order to elucidate the olfactory mechanisms.

Annotation of the ORs in the Genomes of Three Formicidae Species
Three Formicidae species were selected for a comparative analysis of the OR gene family because they had high-quality genome assemblies. The genome size of S. invicta (365 Mb) was larger than that of O. biroi (216 Mb) and M. pharaonis (314 Mb) ( Figure S1, Table S1). Of the BUSCOs, 99.1% (n = 1367) were identified in S. invicta; 97.4% of the BUSCOs were complete with a single copy, 1.7% were complete and duplicated, and only 0.3% and 0.6% were fragmented and missing, respectively (Table S1). In the O. biroi and M. pharaonis genomes, 98.5% and 98.9% of BUSCO genes were present, respectively. The scaffold N50 values were 26.22 Mb in S. invicta, 16.89 Mb in O. biroi, and 29.66 Mb in M. pharaonis, indicating that all three genomes were well assembled (Table S1). The high integrity of the assembly and the high accuracy of the annotation were beneficial for the gene family analysis.
We used A. mellifera as an outgroup to construct a phylogenetic tree of the three Formicidae species using all the longest transcripts of each species. According to the tree, S. invicta and M. pharaonis, which belong to the Myrmicinae subfamily, diverged from each other around 31.8-44.1 MYA. O. biroi, which belongs to the Dorylinae subfamily, diverged from the Myrmicinae around 74.7-100.8 MYA, indicating that it was distantly related to the other 2 ant species ( Figure 1).

Genomic Organization of ORs
Almost all the ORs could be mapped to chromosomes because of the high quality of the S. invicta, O. biroi, and M. pharaonis genomes. We renamed these ORs according to their order on the chromosomes. The ORs that could not be mapped to a chromosome (Sin-vOR355, SinvOR356, and MphaOR299-MphaOR306) were named according to the order in which they appeared in the general feature format file. The chromosomal mapping of the OR genes from S. invicta showed that the 354 ORs were located on all chromosomes except for Chromosome 9 (Chr9) (Figure 2a). Similarly, in O. biroi OR genes were present on all chromosomes except Chr13 (Figure 2b). However, OR genes were present on all chromosomes in M. pharaonis (Figure 2c).
We found that most of the ORs in the 3 Formicidae species were found in clusters We obtained a set of intact ORs that were putatively functional by filtering out the pseudogenes. We annotated more intact OR genes, as compared to previous studies [32,33]. For example, we annotated 356 intact ORs with 496 alternative spliced variants in S. invicta, which was significantly more than the 297 intact ORs annotated in the first publication of the S. invicta genome [32]. As compared to the 503 alternatively spliced OR transcripts annotated in the first version of the O. biroi genome [33], we identified 1 additional putatively intact OR. We classified them as 298 intact OR genes with 504 alternative spliced variants, which slightly improved the previous annotation. We annotated 306 intact ORs with 356 alternative spliced variants in the M. pharaonis genome, which had not previously been annotated.

Genomic Organization of ORs
Almost all the ORs could be mapped to chromosomes because of the high quality of the S. invicta, O. biroi, and M. pharaonis genomes. We renamed these ORs according to their order on the chromosomes. The ORs that could not be mapped to a chromosome (SinvOR355, SinvOR356, and MphaOR299-MphaOR306) were named according to the order in which they appeared in the general feature format file. The chromosomal mapping of the OR genes from S. invicta showed that the 354 ORs were located on all chromosomes except for Chromosome 9 (Chr9) (Figure 2a). Similarly, in O. biroi OR genes were present on all chromosomes except Chr13 ( Figure 2b). However, OR genes were present on all chromosomes in M. pharaonis (Figure 2c).
We found that most of the ORs in the 3 Formicidae species were found in clusters ( Figure 2): 335 ORs were present in 28 clusters and 19 ORs were present as singletons in S. invicta; 277 ORs were present in 26 clusters and 21 ORs were present as singletons in M. pharaonis; and 266 ORs were present in 34 clusters and 32 ORs were present as singletons in O. biroi. Next, we counted the number of the exons in all the ORs in S. invicta to determine whether the ORs were derived from transposon-mediated duplication. There was no evidence of recent intron loss, indicating that there had been no transposon-mediated OR gene replication. An MCscanX analysis showed that 71 pairs of the ORs were tandem duplicates (Table S2).

The 9-Exon OR Subfamily Expanded in Formicidae Species
Next, we analyzed the expansion of the 9-exon OR gene family in the Formicidae species. In total, 89, 118, and 35 9-exon ORs were annotated in S. invicta, O. biroi, and M. pharaonis, respectively. There was more expansion of the 9-exon OR subfamily in S. invicta and O. biroi than in M. pharaonis ( Figure 2). We found that more than 85% of the 9-exon ORs in S. invicta were present in gene clusters and were found on specific chromosomes, rather than being equally distributed across the genome. A large number of the 9-exon ORs were located on Chr1, 5, 6, 7, and 8 of S. invicta, whereas none were found on Chr4, 9, 10, 11, or 14 ( Figure 2a). In O. biroi, the 9-exon ORs were located on Chr1, 2, 6, and 10 ( Figure 2b).
Next, we calculated the selective evolutionary pressure experienced by the 9-exon OR clusters in order to determine the potential reasons for the formation of these clusters, using S. invicta as an example. There were 2 tandem clusters of 9-exon ORs (SinvOR4-8, SinvOR18-22) on Chr1. The Ka/Ks value of the cluster SinvOR4-8 was 1.32, which was greater than 1 (p < 0.05, Table 1). This result suggested that this cluster was under strong positive selection. There was no evidence for the positive selection of the other clusters on Chr1, 3, 6, 7, 8, and 16. The Ka/Ks value of the cluster (SinvOR65-84) on Chr5 was even higher (3.10) than that of SinvOR4-8, but the p-value was greater than 0.05, indicating that this could have been a false positive (Table 1).

Genome Synteny Analyses
The analysis of the synteny of the OR genes in the three Formicidae species revealed that S. invicta and M. pharaonis had the most collinear blocks (e.g., a high level of conservation between the first chromosome of S. invicta and the second chromosome of M. pharaonis), while S. invicta and O. biroi had the fewest. The analysis of the collinear blocks showed that the chromosomal rearrangements had occurred during the evolution of these species (Figure 3a). We counted the ORs in the collinear blocks between these ant species. There were 129 ORs in the collinear blocks between S. invicta and M. pharaonis, and 45 ORs in those between S. invicta and O. biroi (Figure 3a). In total, 36 ORs in 3 ant species were collinear. These collinear genes included six 9-exon ORs in S. invicta, seven 9-exon ORs in O. biroi, and only one 9-exon OR in M. pharaonis (Figure 3b).

Phylogenetic Analysis of ORs
A phylogenetic tree of the ORs from the three Formicidae species was constructed. A large clade of the OR genes in the phylogenetic tree revealed the expansions of the OR gene family, especially in O. biroi and S. invicta ( Figure 4). We found that most of the ORs in this clade were 9-exon ORs, suggesting that the 9-exon ORs in ants evolved along with

Phylogenetic Analysis of ORs
A phylogenetic tree of the ORs from the three Formicidae species was constructed. A large clade of the OR genes in the phylogenetic tree revealed the expansions of the OR gene family, especially in O. biroi and S. invicta ( Figure 4). We found that most of the ORs in this clade were 9-exon ORs, suggesting that the 9-exon ORs in ants evolved along with lineage-specific expansions of the OR subfamily.  In addition, 17 clades (A-Q) with single-copy orthologs of the ORs in the three species were analyzed ( Figure S2). The ORs in 5 of these clades (D, E, G, O, and Q), were significantly conserved, with amino acid similarities greater than 70%. A MEME analysis of the S. invicta proteins SinvOR89, SinvOR102, SinvOR352, SinvOR327, and SinvOR135 showed that the motifs in these proteins had been conserved, suggesting they may have important functions ( Figure 5).

Analysis of Differentially Expressed OR Genes in S. invicta Workers from Different Societies
Using published data [26], we constructed the antennal expression profiles of the OR genes in the worker ants from monogyne SB/SB and polygyne SB/SB colonies ( Figure 6, Table S3). There were differences in the expression levels of the ORs in the antennae of the workers from the monogyne and polygyne S. invicta colonies. SinvOR35 and SinvOR252 were expressed at much higher levels in the monogyne workers than in the polygyne workers. SinvOR21 was expressed at higher levels in polygyne workers, but the expression level was still very low. (e). Motif analysis was carried out using MEME software (version 5.3.0). The colored boxes represent conserved motifs in each OR.

Analysis of Differentially Expressed OR Genes in S. invicta Workers from Different Societies
Using published data [26], we constructed the antennal expression profiles of the OR genes in the worker ants from monogyne SB/SB and polygyne SB/SB colonies ( Figure 6, Table S3). There were differences in the expression levels of the ORs in the antennae of the workers from the monogyne and polygyne S. invicta colonies. SinvOR35 and SinvOR252 were expressed at much higher levels in the monogyne workers than in the polygyne workers. SinvOR21 was expressed at higher levels in polygyne workers, but the expression level was still very low.  Three OR genes from the conserved clades, SinvOR89, SinvOR102, and SinvOR352, were expressed at a relatively high level in the antennae, while the expression levels of SinvOR327 and SinvOR135 were low. The expression levels of all five of these ORs in S. invicta were analyzed by an RT-PCR analysis of the antennae and legs of five castes of ants: WF, WR, WN, AVQ, and M. All five ORs were expressed in the antennae of all five castes, consistent with the data from the transcriptome analysis. SinvOR102 was expressed at a relatively high level in the antennae of the WR and the WF; SinvOR89 was expressed at a relatively high level in the antennae of the workers and the AVQ; and SinvOR327 and SinvOR352 were highly expressed in the antennae of the WR. The expression levels of SinvOR135 were consistently low (Figure 7).
Three OR genes from the conserved clades, SinvOR89, SinvOR102, and SinvOR352, were expressed at a relatively high level in the antennae, while the expression levels of SinvOR327 and SinvOR135 were low. The expression levels of all five of these ORs in S. invicta were analyzed by an RT-PCR analysis of the antennae and legs of five castes of ants: WF, WR, WN, AVQ, and M. All five ORs were expressed in the antennae of all five castes, consistent with the data from the transcriptome analysis. SinvOR102 was expressed at a relatively high level in the antennae of the WR and the WF; SinvOR89 was expressed at a relatively high level in the antennae of the workers and the AVQ; and SinvOR327 and SinvOR352 were highly expressed in the antennae of the WR. The expression levels of SinvOR135 were consistently low (Figure 7). At the same time, we also evaluated the 9-exon ORs expressed at high levels in the antennae of the workers. We found that SinvOR78, SinvOR120, and SinvOR50 were expressed at much higher levels than the other 9-exon ORs, and these highly expressed ORs may have important functions in distinguishing nest mates and distinguishing between the queen and the workers.

Discussion
The OR gene family in ants has undergone rapid expansion, but little is known about the evolutionary mechanisms of OR expansion. A variety of molecular mechanisms lead to gene duplication; the resulting duplicates can be located in tandem clusters or interspersed across the genome. Furthermore, the genomic rearrangement and transposition can disrupt and scatter tandem duplicated genes [5]. Our analysis showed that there was almost no segmental duplication or transposon-mediated duplication during the expansion of the OR gene family in S. invicta. In addition, no recent intron losses were observed in any ant OR, but a large number of tandem repeats were observed, which suggested that the expansion of the ORs in S. invicta may have occurred via tandem replication. Interestingly, a large number of OR tandem repeats have also been found in other Hymenoptera species. For example, A. mellifera ORs were found to be mostly located in tandem duplicate arrays, one of which even contained 60 OR genes [34]. The discovery of tandem duplicate arrays of the ORs in both ants and bees suggests that tandem replication may be the main mechanism of OR amplification in the Hymenoptera. At the same time, we also evaluated the 9-exon ORs expressed at high levels in the antennae of the workers. We found that SinvOR78, SinvOR120, and SinvOR50 were expressed at much higher levels than the other 9-exon ORs, and these highly expressed ORs may have important functions in distinguishing nest mates and distinguishing between the queen and the workers.

Discussion
The OR gene family in ants has undergone rapid expansion, but little is known about the evolutionary mechanisms of OR expansion. A variety of molecular mechanisms lead to gene duplication; the resulting duplicates can be located in tandem clusters or interspersed across the genome. Furthermore, the genomic rearrangement and transposition can disrupt and scatter tandem duplicated genes [5]. Our analysis showed that there was almost no segmental duplication or transposon-mediated duplication during the expansion of the OR gene family in S. invicta. In addition, no recent intron losses were observed in any ant OR, but a large number of tandem repeats were observed, which suggested that the expansion of the ORs in S. invicta may have occurred via tandem replication. Interestingly, a large number of OR tandem repeats have also been found in other Hymenoptera species. For example, A. mellifera ORs were found to be mostly located in tandem duplicate arrays, one of which even contained 60 OR genes [34]. The discovery of tandem duplicate arrays of the ORs in both ants and bees suggests that tandem replication may be the main mechanism of OR amplification in the Hymenoptera.
In this study, we observed that some ORs in S. invicta, O. biroi, and M. pharaonis were clustered on specific chromosomes, and the chromosomes with these clusters differed between species. However, the OR clusters appeared almost exclusively in the Hymenoptera and were not common in other orders [35,36]. Moreover, the ORs of three Drosophila species, D. melanogaster, D. yakuba, and D. pseudoobscura, were found to be scattered across multiple chromosomes, without obvious clustering [37]. However, mice had a total of 1391 ORs, most of which were found within 69 clusters [38]. Hence, the chromosomal distribution of the ORs in ants is similar to that in vertebrates; the similar patterns of the genetic evolution involved in olfactory functions (i.e., tandem duplication) may be related to the common eusocial behaviors of ants and vertebrates.
In total, 89, 118, and 35 nine-exon ORs were identified in S. invicta, O. biroi, and M. pharaonis, respectively. The 9-exon ORs in S. invicta and O. biroi accounted for approximately one-third of the total ORs. The expansion of the 9-exon OR subfamily is a common phenomenon in the Hymenoptera. For example, a large number of 9-exon OR subfamily members were found in the genome of the army ant Eciton burchellii, despite having a reduced gene complement relative to other ants [39]. The 9-exon OR subfamily was also large in the paper wasp, accounting for almost half of the Polistes ORs [20]. The shaping of the 9-exon ORs was partly related to evolutionary pressure [20]. We identified a 9-exon OR cluster SinvOR4-8 that was subject to strong positive selection, which was in accordance with previous studies that had found that the 9-exon ORs had experienced slower but continued expansion in modern ant lineages [12,13]. We observed that 9-exon OR clusters were located on specific chromosomes in all three ant species, which is a widespread pattern in ant OR evolution.
Chemical communication is vital for social insects, particularly via chemical cues, such as a variety of pheromones, mediated foraging, recognition, and reproductive behaviors [40]. For example, trail pheromones are important for foraging behaviors in ant species. Recently, four trail pheromones, Z, E-α-farnesene (ZEF); E, E-α-farnesene (EEF); Z, E-αhomofarnesene (ZEHF); and E, E-α-homofarnesene (EEHF), were identified in S. invicta, and two of them, ZEF and EEF, mediated the mutualism between ants and aphids [41]. In addition, the trail pheromones in Myrmicine and Formicine species were identified as farnesene isomers [42,43], suggesting that the mechanism for recognizing trail pheromones may be similar between species. In this study, we identified 5 S. invicta ORs, SinvOR89, SinvOR327, SinvOR135, SinvOR327, and SinvOR135, that were highly conserved singlecopy orthologs in all 3 ant species examined. These ORs may be involved in interspecific recognition of specific chemicals, such as trail pheromones. Further functional studies are needed to test whether these ORs detect farnesene isomers.
The components of the alarm pheromones in ant species have also been identified. For example, 2-ethyl-3,6-dimethyl pyrazine, which had been reported to warn against the predator Pseudacteon tricuspis, was isolated from S. invicta [44]. A mixture of 4-methyl-3heptanone and 4-methyl-3-heptanol was identified as an alarm pheromone in O. biroi and the close relatives of Eciton army ants [45][46][47]. These results indicated that the components of the alarm pheromones in ants were variable. However, little is known about the molecular basis of the interspecific and intraspecific alarm pheromone detection in ants. The ORs involved in detecting alarm pheromones in O. biroi and Eciton army ants may be characterized by searching for conserved clades and comparing the expression levels of the ORs in specific tissues of ants from different castes, because ants are more sensitive to alarm pheromones.
Eusocial ants use colony-or species-specific blends of CHCs to convey social information, including nestmate and caste recognition [48] CHCs are detected by OSNs, which are housed in the female-specific basiconic sensilla in the ant antennae [49][50][51], and they are detected by specific members of the expanded 9-exon OR subfamily [48]. Several studies have shown that 9-exon ORs could detect numerous CHCs, including nestmate and non-nestmate CHC blends and queen pheromones, which are necessary for normal nesting behavior and the recognition of nestmates. In general, the ORs that function in response to CHCs have been shown to be highly expressed, suggesting that high expression could be used as a criterion to identify these ORs in other species. For example, highly expressed OR HsOR263 in H. saltator could respond strongly to the candidate queen pheromone component 13, 23-dimethylheptatriacontane (13,23-DiMeC37), and 15-methylnonacosane  in cuticular extracts of workers and gamergates, suggesting that this receptor could play an important role in the detection of reproductive individuals [19]. Another study showed that the 9-exon H. saltator ORs HsOR271 and HsOR259-L2 were also responsible for the detection of 13,23-DiMeC37 [52]. In addition, non-9-exon ORs were found to detect CHCs with long-chain alkanes. In this study, we identified 3 highly expressed 9-exon ORs in S. invicta, namely SinvOR78, SinvOR50, and SinvOR120, by constructing S. invicta antenna expression profiles of the workers. SinvOR78 exists in a 9-exon OR cluster that is rapidly evolving. This result, combined with previous findings for other ants, suggested that S. invicta could sense CHCs, such as nestmate and queen pheromones, through these highly expressed 9-exon ORs.
The social types of S. invicta were determined by assessing the Gp-9 genotypes of both the workers and the queens. In this study, we analyzed the expression levels of the ORs in the antennae of the S. invicta workers from monogyne and polygyne colonies and found that the ORs SinvOR35, SinvOR252, and SinvOR21 were differentially expressed among these workers. SinvOR35 and SinvOR252 were much more highly expressed in the antennae of the monogyne workers than in the polygyne workers, suggesting that these genes could have key functions in the olfactory detection of workers in the monogyne social type. However, SinvOR21 was more highly expressed in the polygyne workers, suggesting that this gene could be involved in detecting BB-queen-derived odorants and assist in maintaining the polygyne social type [53].

Data Sources
We searched the genomes of all Formicidae species in the NCBI database. Three Formicidae species, namely M. pharaonis (GCF_013373865.1; from NCBI), O. biroi (GCF_003672135.1; from NCBI), and S. invicta (GCF_016802725.1; from NCBI) were selected for further analysis because they had the best chromosome-scale genome assemblies and the most complete gene annotations. Benchmarking Universal Single-Copy Orthologs (BUSCOs, version 5.2.2) was used to assess the quality of the genome assemblies for these three ants; 1367 universal single-copy orthologous genes from insecta_odb10 were used as queries in searches against the assembly using the default parameters [54]. The scaffold N50 length was calculated using Python scripts.

Phylogenetic Analysis
Three Formicidae species, M. pharaonis, O. biroi, and S. invicta, and one outgroup species, Apis mellifera (GCF_003254395.2; from NCBI), were selected to construct a phylogenetic tree. The phylogenetic tree was constructed using Orthofinder2 (version 2.5.4, Oxford University, Oxford, United Kingdom) with default parameters [55] and using the longest transcripts for each gene. To estimate species-divergence times, we first queried the divergence time between A. mellifera and S. invicta on the Time Tree of Life (http://www.timetree.org/, accessed on 10 November 2022) divergence time between 114.9-163.5 MYA) as the calibration point, and then we used MCMCTREE from the PAML package (version 4.8a) to calculate divergence times for each species [56]. The phylogenetic tree was visualized with Evolview (version v2, Northeastern University, Boston, MA, USA) [57].

Analysis of the OR Gene Family
First, all known OR protein sequences of Hymenoptera were downloaded from the NCBI database (https://www.ncbi.nlm.nih.gov/, accessed on 28 March 2022). Pseudogenes were removed manually. Then, the remaining sequences were used as queries in BLASTP searches against the transcripts of M. pharaonis, O. biroi, and S. invicta (BLAST version 2.9.0+, E-value = 1 × 10 −5 ) [58]. HMMER (version 3.3.2) searches against the Pfam database were performed to confirm the structural domains of the candidate OR genes [59,60]. In brief, for each species, the HMM profile 7tm_6 (PF02949) was used for an hmmsearch against candidate OR genes identified by BLAST. Next, the OR genes obtained by the preliminary HMMER screen were used to construct separate hidden Markov models for each ant species, and a secondary hmmsearch was conducted to ensure that no ORs were missing. Protein sequences identified by the above steps were then used as queries in a BLASTP search against the non-redundant protein sequence database (e-value = 1 × 10 −5 ), and CD-HIT (version 4.8.1) was used to filter out redundant genes (C = 0.98) [61].
For phylogenetic analysis of gene families, the protein sequences of the OR family in each species were aligned and trimmed by MAFFT (version 7.487) and TRIMAL (v1.4. rev15) with default options [62,63]. IQ-TREE (version 2.1.4-beta) was used to construct the tree, which was formatted in Evolview [57,64]. Motif analysis was carried out using MEME software (version 5.3.0) with default parameters [65].

Tandem Replication of the ORs in S. invicta
The first transcript of each gene in the genome annotation file was selected as the representative gene sequence and was used as a query in BLAST searches to identify ORs derived from tandem duplication and segmental duplication (parameters: blastp, e-value = 1 × 10 −5 ). Next, we used MCscanX (version 2.1, Michigan State University, USA) to search for all tandem repeats while using scripts to extract those of OR genes [66]. MapChart (version 2.32) was used to construct the chromosomal maps of M. pharaonis, O. biroi, and S. invicta. A cluster was defined as two or more neighboring OR genes separated by a distance of less than 100 kb, and the numbers of clusters were counted.

Genome Synteny Analysis
Whole-genome synteny among M. pharaonis, O. biroi, and S. invicta was detected and plotted with jcvi, which was a package in Python, with default parameters [67]. The OR synteny blocks were colored by appending the specified color code to the OR gene in the collinear block file.

Analysis of Evolutionary Selective Pressure
The OR protein sequences on each chromosome were aligned by MAFFT, and MEGA [68] was used to construct a neighbor-joining tree. Codon sequence alignments were built using Python scripts. To assess the role of Darwinian positive selection on the evolution of OR genes in ants, we used the CodeML program from the PAML (version 4.8a) package to perform positive selection tests using the branch model [56]. The nonsynonymous/synonymous rate ratio (ω = Ka/Ks) was calculated for clusters of 9-exon OR genes on each single chromosome in S. invicta. Command settings for the null model were "model = 0, NSsites = 0, fix_omega = 0, omega = 0.8", and those for the alternative model were "model = 2, NSsites = 0, fix_omega = 1, omega = 0.8". A Chi-squared test was performed to identify the significant differences between the two models (p-value < 0.05).

Analysis of Differential Expression of the ORs in the Antennae of S. invicta Workers from Different Societies
First, we downloaded raw antennae transcriptome data for S. invicta workers of monogyne SB/SB and polygyne SB/SB ant colonies from NCBI (GenBank accession number: GSE126684). Next, Trimmomatic (version 0.38, Cambridge University, UK) and Multiqc (version 1.10.1, Cambridge University, UK) were used to review the quality of every sample [69,70]. S. invicta OR transcripts were used to construct the OR database. Specifically, Salmon with default parameters was used to quantify OR transcripts for each sample [71]. Transcripts-per-million (TPM) values of all ORs in each sample were extracted from the total quantitative results. Finally, the R package DEGseq2 was used to standardize the data and identify differentially expressed genes (DEGs) using the criteria p < 0.05, FDR < 0.001, and |log2FoldChange| > 2. A base-mean TPM value greater than 80 was used as the criterion for the high expression of the 9-exon ORs. The results were visualized as a heatmap generated with Evolview (version v2, Northeastern University, Boston, MA, USA) [57].

Expression Analysis Using Semi-Quantitative RT-PCR
Semi-quantitative RT-PCR was performed to verify the expression of OR genes. Antennae (500 individual) and legs (300 individual) were collected from monogyne individuals from five castes of the polygyne-type S. invicta, namely workers for foragers (WF), reserves (WR), and nurses (WN), alate virgin queen (AVQ), and males (M) [18,72]. Total RNA was extracted as described in [73]. The cDNA was synthesized from total RNA using the Color Reverse Transcription Kit (with gDNA Remover) (EZBioscience, Roseville, California, USA). Gene-specific primers were designed using Primer Premier 5.0 software (Table S4) and synthesized by Tsingke Biotechnology Co., Ltd. (Guangzhou, China). Green Tap Mix (Vazyme Biotech Co., Ltd., Nanjing, China) was used for PCR under the following conditions: 95 • C for 5 min; 30 cycles of 95 • C for 15 s, 54 • C for 15 s, and 72 • C for 10 s; and extension at 72 • C for 5 min. The 20 µL reaction system consisted of 10 µL Green Tap Mix, 1 µL Primer F, 1 µL Primer R, 2 µL cDNA, and 6 µL ddH 2 O. RT-PCR products were separated on 1% agarose gels, stained by ethidium bromide, and photographed under a UV light in a Gel Doc XR+ Gel Documentation System with Image Lab Software (version 6.1, Bio-Rad, Hercules, CA, USA).

Conclusions
In this study, 356, 298, and 306 potentially functional ORs were identified in the 3 Formicidae species S. invicta, O. biroi, and M. pharaonis, respectively. We observed that the S. invicta OR gene family had expanded via tandem replication and that there was a biased localization of 9-exon ORs on the chromosomes. This indicated that tandem replication could be the main mechanism of ant OR amplification. The phylogenetic analysis of the ORs of the three Formicidae species indicated that there were highly conserved ORs and species-specific 9-exon OR subfamily members in these species. The antennal expression profiles of the ORs of the S. invicta workers of different social types showed that three SinvOR genes had been differentially expressed among the social types and that three 9-exon ORs in S. invicta were highly expressed in the antennae of the workers. Five SinvORs conserved in the Formicidae species had been differentially expressed between castes. When considered altogether, our findings provided new insights into the function and the evolution of OR genes in Formicidae ants.