Ants are the most numerous of the social insects and one of the most successful animals on earth (Holldobler and Wilson, 1990). The red imported fire ant, Solenopsis invicta (Buren) was originally endemic to South America (Northern Argentina, Southern Brazil, and parts of Paraguay). Sometime between 1933 and 1945, this ant likely found its way aboard ships carrying exotic fruits, lumber, and other goods from South America apparently establishing somewhere around Mobile, Alabama in the United States, from where it has spread to many other parts of the world (Tschinkel, 2006; Shoemaker et al., 2011). In this respect, S. invicta has demonstrated a remarkable plasticity in environmental adaptations making it one of the most successful invasive animal species on our planet. It is an important pest throughout the southern United States and has recently invaded and established itself in California, South East Asia, and even parts of Europe. S. invicta form complex societies that are territorial, and whose members display strong nest-mate recognition, and elaborate task specialization (Tschinkel, 2006). Their colonies are considered to form a multi-tiered “caste” system consisting of non-reproductive workers, that display task-differentiation, and reproductive winged males and females. Male and female reproductive forms are produced seasonally as the colony matures, a process that can take up to 5 years. Depending upon their genetic background S. invicta colonies can have one (monogyny) or multiple queens (polygyny) (Ross and Fletcher, 1985; Gotzek and Ross, 2008). These two differing colony organizations are genetically distinct and have consequent important differences in the biology and responses of S. invicta. Originally linked to a single gene, termed Gp-9 (renamed OBP3), mono- versus polygyne colony organization appears to be linked to a larger 13 Mb non-recombining “social” chromosomal fragment (∼55% of the chromosome) containing at least 616 open reading frames (Wurm et al., 2011; Wang et al., 2013).

Chemosensory proteins (CSPs) are a protein family defined by amino acid homology and sequence motifs that include four cysteine residues (two disulfide bridges) with conserved spacing and a set of α-helices that form a hydrophobic binding cavity as part of a compact, small molecular weight (10–15 kDa) structure (Lartigue et al., 2002; Zhou et al., 2006; Pelosi et al., 2014). CSPs were originally identified as a protein accumulating during leg regeneration in nymphal stages of the cockroach Periplaneta americana (Nomura et al., 1992), other CSPs were subsequently found enriched in sensory organs including the antennae, and hence presumed to function within the context of chemoreception or olfaction by acting as odorant (ligand) carrier proteins (Angeli et al., 1999; Picimbon et al., 2000; Ban et al., 2003). CSPs in the alfalfa plant bug (Adelphocoris lineolatus) have been implicated in mediating host recognition (Gu et al., 2012), and CSP/Takeout genes have been shown to be involved in olfactory-based behaviors including repulsion and attraction in the migratory locust (Locusta migratoria) (Guo et al., 2011). However, it has also been recognized that many CSPs are expressed in a variety of other tissues, e.g., in the pheromone glands of the cabbage moth, Mamestra brassicae and the silk moth, Bombyx mori, and in male and female reproductive organs, e.g., ejaculatory bulb of Drosophila melanogaster and female organs of L. migratoria manilensis (Zhou et al., 2013), where CSPs potentially function in pheromone detection and release (Jacquin-Joly et al., 2001; Dani et al., 2011). More recently, various CSPs have been implicated in a wide range of physiological processes beyond olfaction and/or chemical communication. RNAi knockdown of the honeybee (Apis mellifera) CSP5 indicated a role for this protein in embryonic integument development (Maleszka et al., 2007) and CSP3 (designed ASP3c), also from the honeybee, has been shown to act as a brood pheromone carrier protein (Briand et al., 2002). CSPs, therefore, should be considered as general ligand carrier proteins, some of which may function within the scope of chemical perception mediating binding of volatile and/or hydrophobic odorants, whereas others may ferry hormones and/or other chemical compounds participating in organismal homeostasis and/or developmental process. Participation in pheromone/semiochemical sequestration and/or release would still link CSPs to chemical communication, albeit in the broad sense, namely outside of sensillar mediated olfaction.

Chemosensory proteins are considered to have at least some parallels in function to another class of small molecular weight (∼130–150 amino acids), soluble proteins known as odorant binding proteins (OBPs). OBPs, like CSPs, are thought to mediate signal transduction in insects by shuttling hydrophobic compounds (e.g., odorants) to odorant receptors expressed on the dendritic membrane of the olfactory neurons. Insects OBPs do not share homology to the vertebrate OBPs and are sometimes annotated as pheromone binding proteins (PBPs) or general odorant binding proteins (GOBPs) (Zhang et al., 2016). Expression profiling of the S. invicta 17-gene member OBP family revealed both antennal and non-antennal expression specific SiOBPs, as well as caste and tissue specificity in the expression of these proteins (Zhang et al., 2016). Intriguingly, the S. invicta “social” chromosomal fragment contains multiple members of both the CSP and OBP gene families amongst other genes.

Antenna-specific CSPs, enriched in specialized chemosensory sensilla, have been identified in the Argentine ant (Linepithema humile) and the Japanese carpenter ant (Camponotus japonicus), with CjCSP1 mediating cuticular hydrocarbon recognition and transport, a critical process for nestmate/non-nestmate discrimination in C. japonicus (Ishida et al., 2002; Ozaki et al., 2005). Comparative transcriptomics in a variety of ant species also indicated expression of CSPs in non-antennal tissues, further supporting their functioning beyond olfaction (McKenzie et al., 2014). Initially, fourteen CSP sequences were identified in S. invicta expressed sequence tag (EST) libraries, with one, designated SiCSP1(SiCSP19 in our study), as well as an apolipophorin-like protein, found to be highly expressed in the worker antennae (Guntur et al., 2004; Gonzalez et al., 2009; Xu et al., 2009). Subsequent genome mining and phylogenetic analyses revealed intriguing CSP gene expansions in the available ant genomes with the apparent evolution of ant-specific CSP clades (Kulmuni et al., 2013). Within these analyses 21 CSPs in total were identified in the S. invicta genome, representing one of the largest CSP sets annotated in eusocial Hymenoptera to date. Most Drosophila spp., by contrast, have only four CSP genes (Vieira and Rozas, 2011).

Here, the expression pattern of the 21 CSP genes identified in S. invicta was examined by quantitative reverse transcriptase polymerase chain reaction (qRT-PCR). SiCSP gene expression patterns were examined as a function of caste, tissue distribution, and developmental stage. Developmental specificity of SiCSP was examined in eggs, early instar larvae, late instar larvae, and pupae. For adult ant (workers, male and female alates) samples were further subdivided into specific anatomical tissues that included the antenna, head, thorax, and abdomen. High to robust expression of most SiCSPs was found in the antennae of workers and male alates, and to a lesser extent, in female alate antennae. SiCSP expression was significantly lower in the developmental stages as compared to the adults, although robust expression of specific SiCSPs were found in early-instar larvae and pupae. These data revealed dynamic and differential SiCSP expression patterns among developmental stages and adult tissues with little correlation to their phylogenetic relatedness, although sub-groups of phylogenetically related SiCSPs showed similar patterns of expression. Both divergence between more closely related SiCSPs and convergence between more remotely related SiCSPs was observed. This study provides a basis for a further systematic characterization of the functions of the different CSPs in S. invicta.

Chemosensory proteins represent one branch of at least three phylogenetically un-related insect protein families, that are however, considered partially functionally related. The other two members are the arthropod OBPs and the lipid transporters classified as Niemann-Pick type C₂ proteins (NPC₂) (Pelosi et al., 2014). All three classes of proteins are small (∼12–20 kDa), soluble, ligand-binding proteins, that are often considered capable of solubilizing hydrophobic compounds including odorants, volatiles, and lipids within the aqueous milieu of the organism and/or cell. These molecules can encompass pheromones, semio-chemicals, hydrocarbons, fatty acids, amongst other hydrophobic compounds, but can also potentially include soluble substrates such as carbohydrates and other molecules. In arthropods, CSPs exist as multigene families, and have been associated with chemoreception of both olfactory and gustatory cues. Aside from numerous studies showing expression of CSPs in antennae (for recent reports see Antony et al., 2016; Hu et al., 2016; Xue et al., 2016; Jing et al., 2020), these proteins have also been found in many other tissues some of which contain specialized sensilla, e.g., in tarsal chemosensilla that respond to non-volatile surface chemicals and are involved in gustatory detection (Ma et al., 2016). In the latter case, transcriptomic analyses in the tea geometrid, Ectropis obliqua, revealed sex and leg-specific expression of CSPs, suggesting their involvement in taste detection. From these and other studies, it is now well established that many CSPs are expressed outside olfactory organs, functioning in chemosensory (e.g., taste) and non-chemosensory roles (e.g., development). Transcriptomic analyses in several ant species revealed that many OBPs and CSPs are expressed primarily in non-olfactory tissues (McKenzie et al., 2014), strongly suggesting non-olfactory roles for many of these proteins in ants. In A. mellifera, AmCSP5 is not expressed in the antennae, but instead has been implicated in embryonic development, with high expression noted in queen ovaries and eggs (Maleszka et al., 2007). In addition, proteomic analyses have indicated the presence of CSPs in insect hemolymph (Hou et al., 2016).

Our data indicate caste-specific expression patterns of the S. invicta CSPs. Transcripts for most of the SiCSPs could be detected at robust and high levels in worker and male antennae. Levels of SiCSP expression in female alate antennae were notably lower, with the expression of at least four SiCSPs, i.e., 22, 2, 6, and 14, being low in female alate antennae but robustly expressed in workers and male alate antennae. This suggests that in S. invicta, similar to H. saltator, olfactory responses depend on reproductive status (Ghaninia et al., 2017). Pheromonal inhibition by the reproductive queen(s) and/or ecological adaptations - since female alates remain most of the time in the nest and are not involved in colony maintenance – might account for S. invicta female alates (i.e., virgin females) being less responsive to olfactory cues than male alates and workers. Generalized robust expression of CSPs in male alates, on the other hand, might reflect the need for male alates to respond to olfactory signals in the context of mating (Ghaninia et al., 2018).

In workers, SiCSP19 was the most highly expressed CSP, consistent with its original discovery in S. invicta antennae (Leal and Ishida, 2008; Gonzalez et al., 2009). The protein product of SiCSP19 has been localized to the tip (A9-A10 segments) of the antennal club (Leal and Ishida, 2008). As the porous sensilla contain self-enclosed neuronal compartments located in the S. invicta antennal club that are not contiguous with the cuticle and hemolymph of the rest of the antennae (Renthal et al., 2003), it was concluded that SiCSP19 is most likely found in the olfactory sensilla (Gonzalez et al., 2009). Intriguingly, SiCSP19 is not an ortholog of the C. japonicus CSP1 (CjCSP1) that has been shown to be localized to chemosensory sensilla and implicated as critical for binding of hydrocarbons implicated in nestmate/non-nestmate discrimination (Ozaki et al., 2005). The ortholog of CjCSP1 appears to be SiCSP1, which, while equally highly expressed across female alate, male and worker antennae, is more preferentially expressed in the male head. Proteomic analyses of S. invicta worker antennae also revealed the presence of SiCSP12 (Gonzalez et al., 2009), that was not only highly expressed in worker antennae in our dataset, but also in male and female alate antennae. In addition, robust to high expression of SiCSPs 8 and 11 were noted in worker, male and female alate antennae. Whether these proteins (and/or other SiCSPs) are made in the antennae will require further deeper probing of the proteomic content of these structures.

A limited number of downstream behavioral effects have been attributed to CSPs involved in odorant detection. The B. dorsalis BdorCSP2 has been implicated in insect responses to rhodojaponin-III, a non-volatile diterpene which shows antifeedant and anti-ovipositing activities against insects (Yi et al., 2013). BdorCSP2 was shown to be capable of binding rhodojaponin-III in vitro, and RNAi mediated knockdown of BdorCSP2 decreased the effects of the chemical compound on the insect. SiCSP4 is the ortholog of BdorCSP2 and is robustly expressed in the antennae of fire ant males and workers and in much higher levels in the worker abdomen. Several CSPs, highly expressed in A. cerana antennae, have been shown to bind various floral chemicals (Li H.L. et al., 2016) and two CSPs in the mite Tyrophagus putrescentiae have been implicated in host recognition (Li X.M. et al., 2016), although these studies lack experimental designs that would directly test the role(s) of these CSPs in mediating specific behaviors. Differential male and female-specific expression of olfaction genes, has been reported in a number of insects including the oriental fruit fly (Bactrocera dorsalis) and the hover fly Scaeva pyrastri (Li X.M. et al., 2016; Liu et al., 2016). Caste and sex-specific expression of chemosensory genes has also been reported in termites and honeybees (Mitaka et al., 2016; Jain and Brockmann, 2020). When the absolute number of SiCSP transcripts is examined by tissue distribution, clear caste- and sex-specific differences are apparent. In workers, SiCSP transcripts are distributed mainly in the antennae and the abdomen. This suggests these SiCSPs may function as ligand carriers in antennal chemosensation, and that in the abdomen these proteins may act to bind/sequester endogenous hydrophobic compounds that can include pheromones (for release), hydrocarbons and lipids (for delivery to the cuticle), and/or endocrine molecules (endogenous hormones and other compounds). In contrast, in female alates, SiCSP transcripts appear to be widely distributed in the thorax, and to a lesser extent, the male thorax appeared also enriched in SiCSP expression. These tissues include the legs and the wings that contain tarsal and wing chemosensory sensilla, some of which have gustatory roles (Fleischer and Krieger, 2018; He et al., 2019). As reveal by the clustering analyses, similar patterns of CSP expression in S. invicta male and female thoraces could indicate roles for these CSPs in gustatory and/or taste functions in this ant species. Gustatory pathways in Drosophila have been reported to be activated by pheromonal cuticular hydrocarbons (CHCs) (Fleischer and Krieger, 2018). In ants, select CHCs also mediate social interaction between colony members and can act as mating pheromones (Kather and Martin, 2015; Slone et al., 2017). In Drosophila, some pheromonal CHCs are primarily detected via taste organs on legs and wings (Fleischer and Krieger, 2018). It is intriguing to speculate that in male and female S. invicta alates, some SiCSPs, (particularly, S. invicta-specific CSPs) expressed in the thorax may function to bind mating CHCs, and that sensilla located in the male and female thorax participate in mate recognition.

SiCSP expression levels and tissue distribution appear to also reflect the task distribution of the various castes. Workers are “pluripotent” in task and are hence exposed to a significantly larger range of environmental compounds than female or male alates. Female alates need to mate, establish the colony, and subsequently act as the reproductive unit of the colony, and thus antennal chemosensation is likely to be far less important than in workers. In addition, queen pheromones act to regulate offspring behavior, but the queen herself may be largely non-responsive to these chemicals. Finally, male alates have only one task, and that is to mate. Therefore comparatively, in males, the requirements for and functions of the CSPs may be more limited, consistent with their lower expression and scattered tissue distribution.

A lack of congruent patterns expression of orthologous CSPs was generally seen when our data are compared to CSP expression analyses in C. japonicus. For example, the most highly expressed CSPs in C. japonicus were CjCSPs 1, 3, 7, 12, and 13, with only CjCSPs1, 3, and 7 having true orthologs in S. invicta (Hojo et al., 2015). Of the corresponding SiCSPs only SiCSP1 was highly expressed in worker antennae. These data suggest little cross-species conservation of the expression patterns of orthologous CSPs.

In general, the SiCSPs were poorly expressed in the developmental stages examined and were especially low in eggs and pupae. Some expression was seen in the various larval stages, and when compared across the developmental stages, SiCSP expression appeared to peak in 1st/2nd instar larvae. These data suggest the potential for a role for some SiCSPs as ligand carriers in larvae, potentially for distribution and delivery of cuticular hydrocarbons and lipids and/or of developmental hormones and other molecules. Apparent co-expression of some CSPs has been reported in the antennae of C. japonicus (Hojo et al., 2015). Hierarchical clustering analyses examining potential co-regulation of phylogenetically related SiCSPs revealed patterns of co-expression of SiCSPs expression within caste tissues and between developmental stages. However, notable divergences were also seen. No robust co-expression consistent with phylogeny was observed among SiCSPs across worker, male and female alate antennae. In contrast, in the head, thorax, abdomen and developmental stages co-expression of phylogenetic sub-groups could be discerned. Our data revealed potential examples of co-regulation of both divergent and convergent lines of SiCSPs.

These data, together with the characterization of the expression profiles of the S. invicta OBPs (Zhang et al., 2016), indicate that both OBPs and CSPs show discrete expression profiles within S. invicta. These profiles extended between workers, male and female alates, and across various developmental stages and supports an emerging model of these proteins as general ligand carriers involved in a wide range of physiological processes beyond olfaction and/or chemosensation. Comparison between SiOBP expression distribution and SiCSP expression revealed a number of interesting insights. Similar to the SiCSPs, only a subset of the SiOBPs were highly expressed in the worker, male and female alate antennae. However, SiCSP expression was more enriched in the antennae than SiOBP expression (Zhang et al., 2016). In the abdomen, SiCSP expression was enhanced in the worker but not in male and female alates, while SiOBP expression showed the opposite trend (more highly expressed in the male and female abdomen than in the worker abdomen). The fire ant head and thorax tissues showed lower expression of the SiCSPs than of the SiOBPs. In contrast, SiCSP expression was much higher than SiOBP expression in the various developmental stages (Zhang et al., 2016). These comparisons indicate a general inverse correlation between SiCSP and SiOBP expression. This may suggest compensatory functions and/or that these proteins might perform unique tasks within different tissues and, in some instance, may complement each other. Thus, the differential regulation of SiCSPs and SiOBPs may be linked to the substrates they may recognize.

Motif analyses of the SiCSPs revealed a core set of three motifs (i.e., motifs 4, 1, and 2) found in almost all of the SiCSPs as well as two slightly different conserved signal peptide sequences consisting on motif 5 alone and a combination of motifs 7 and 5. Additional motifs (i.e., 3, 6, and 8) were found at the C-terminal end in subsets of the SiCSPs. Essentially, three different C-terminal ends were observed in SiCSPs, which yielded three distinct groups of SiCSPs. Group I, which included one of the fire ant-specific CSP gene expansion (SiCSPs 9, 22, 10, 19, 20,13, and 12) and the shortest C-terminus, containing motif 8. Group II containing motif 3 and including SiCSPs 1, 14, 18 and the second fire ant-specific CSP gene expansion (i.e., SiCSPs 21, 15, 11, 17, and 16). And Group III (SiCSPs 8, 6, 2, 3, 4, and 7) with the longest C-terminal end, which comprised motifs 3 and 6. Within Group III, SiCSP8 and 6 exhibited a degenerate motif 6 and SiCSP2 a much shorter C-terminus as motif 6 appeared shuffled toward the N-terminus, where it replaced motif 4. The similarities and differences observed amongst SiCSPs within the core motifs, the N-terminal and C-terminal ends suggest that SiCSP functional diversification was mainly driven by gradual changes occurring at the periphery of the protein core via point mutations, indels and domain shuffling (Bagowski et al., 2010). Nevertheless, the consequences of divergences and convergences of SiCSP expression, as well as the motif analyses that likely reflects structural and functional properties, will only be understood as more information concerning their ligand binding affinities and downstream interacting partners are uncovered.

Our data provide a foundation for future exploration of the functional roles of insect CSPs and suggest that like OBPs, CSPs participate in diverse biological processes as ligand carriers. The systematic expression profiles of both families of SiCSPs and SiOBPs are now available and can be used to identify targets that may participate in specific physiological processes. In addition, these data can be used as a basis for exploring factors that may contribute to ant colony organization and development, self/non-self-recognition and communication, and/or identify potential targets for ant control, i.e., attractants and/or repellents. Combining these expression data with the ligand binding profiles of these proteins is a much-needed future step in their functional characterization (e.g., behavioral studies combined with RNAi). Such ligands may be derived from the environment during chemosensation or endogenously within the organism. In addition, the differential expression of the CSPs and OBPs in ant castes may be one defining feature or signature that can account for the different behavioral responsiveness of each caste in terms of reacting to chemical cues that lead to task allocation, mating, and other features of the social nature of this ant species.

All the data supporting the findings is contained within the article and/or Supplementary Materials.

NK, AW, WZ, AO-U, and JB initiated the project, conceived and designed the study. WZ, AW, AO-U, and JB performed the samples collection, library constructions, q-RT-PCR, data processing, bioinformatic analyses, data processing, and contributed to revisions of the manuscript. JB and YX provided resources and/or materials for the study. NK, AO-U, and YX wrote the manuscript, contributed to bioinformatic analyses, data processing, and data interpretation. All authors contributed to the article and approved the submitted version.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.