Quit bugging me: phorid fly parasitoids affect expression of an immune gene in foraging fire ant workers

Host-parasitoid interactions are complex. A parasitoid can change its host’s behavior by direct infection or simply by its presence in the shared environment. In the red imported fire ant (RIFA, Solenopsis invicta), workers display defensive postures to avoid potential parasitism when fire ant decapitating flies (Pseudacteon spp.) hover above them. In addition to changes in the behavior of individual ants, RIFA colonies limit their foraging when decapitating flies are present. To further understand these interactions, we investigated the molecular basis of changes in foraging behavior at the colony level. By quantifying expression of the S. invicta foraging gene (Sifor), odorant binding protein 11, and immune genes of workers of unknown infection status in colonies exposed to decapitating flies and control colonies over a 48-h period, we aimed to provide information on the molecular processes related to change in foraging behavior of RIFA. We found upregulation of an immune gene, defensin-2, in colonies exposed to decapitating fly parasitoids. However, we did not detect significant differences in expression levels of three other genes related to immune function: abaecin, cytochrome P450 4C1-like, and hymenoptaecin. Although colonies exposed to parasitoids foraged less than control colonies (flies absent), they did not experience significant changes in Sifor, a gene correlated with worker caste performance, or odorant binding protein 11, a gene upregulated in foragers. These results show that molecular processes associated with diminished foraging activity are set into place by the mere presence of the flies or by early stages of phorid fly infection.


Introduction
Interactions between a host and their parasites are often complex (de Bekker et al. 2018;Thomas et al. 2010;Moore 2002). Parasites can have strong or subtle effects on their hosts by affecting their behavior, growth, fecundity, and/or mortality (Marcogliese 2004). A parasite can change its host's behavior by direct infection or simply by its presence in the shared environment (Kavaliers et al. 2020). A notable example of direct infection is that of Nematomorpha horsehair worms which manipulate their terrestrial insect hosts into searching for water into which they then enter and drown to promote the reproduction of the adult nematomorph (Biron et al. 2005;Libersat et al. 2009;Schmidt-Rhaesa and Ehrmann 2001). Parasite threat can also alter behavior before infection takes place (Kavaliers et al. 2020). Host behavioral adaptations are a response of the host to avoid infection and are considered the host's first line of defense in response to parasite threat (Holmes and Johnston 2022;Vale et al. 2018;Hart 1994;Moore 2002;Daly and Johnson 2011). There are many examples of parasites affecting behavior of their hosts. For example, cattle have various fly-repelling behaviors and selectively graze to avoid parasitic flies (Hart 1994). Another example is that a variety of ant species elicit unique defensive postures or movements to avoid parasitism by fly parasitoids in the family Phoridae (Feener Jr and Brown 1997).
Parasites of particular interest are parasitoids, which consume and ultimately kill their host (Godfray 1994;Bjørnstad 2018). This is in contrast to a parasite that lives in or on, weakens and/or manipulates their host, but does not typically kill its host (Kortet et al. 2010). Interactions between host and a parasitoid are dominated by distinctive 1 3 behaviors associated with oviposition attack and parasitoid development (Feener and Brown 1997). The presence of parasitoids can lead to strong selection and rapid evolution of their hosts. For example, a mutant phenotype, flatwing, in the wild field cricket, Teleogryllus oceanicus, rapidly spread throughout several populations in response to pressure from the acoustically orienting parasitoid tachinid fly, Ormia ochracea (Pascoal et al. 2016;Zuk et al. 2006). Flatwing interferes with the ability of male crickets to sing, which is a strong sexual signal of crickets (Zuk et al. 2006).
Certain fly parasitoid genera in the family Phoridae, commonly referred to as decapitating flies or ant-decapitating flies, elicit defensive behavioral responses from their hosts (Feener and Brown 1997). Leaf-cutting ants, Atta spp. have individual defensive responses and may engage in collaborative defensive behaviors (Alma et al. 2019;Elizalde and Folgarait 2012). In response to a parasitoid phorid attack, workers will drop their load, retreat to the nest, move appendages (e.g., antennae, legs, mandibles) or display a body posture (such as a "C posture") to avoid oviposition and thus becoming parasitized (Alma et al. 2019;Bragança et al. 2009;Elizalde and Folgarait 2012;Feener and Brown 1993;Orr 1992;Tonhasca 1996;Feener and Moss 1990). Another defensive behavior is that small workers can defend larger workers from phorid attack by riding (also referred to as "hitchhiking") on leaf fragments which are carried by the larger foragers (Alma et al. 2019;Vieira-Neto et al. 2006). These behavioral responses to parasitoids can occur at the individual level or social level where workers engage in a collaborative response to fend off parasitoids (Elizalde and Folgarait 2012;Alma et al. 2019).
Fire ants (certain members of the genus Solenopsis) have a number of phorid fly parasites (Chen and Fadamiro 2018;Briano et al. 2012). In this group, Pseudacteon phorid parasitoid flies modify the foraging behavior of their ant hosts (Chen and Fadamiro 2018). The presence of just a few phorid flies triggers specific defense responses, including, but not limited to, reduced foraging, curled defensive postures, and general immobility of workers (Porter et al. 1995). These behaviors are thought to reduce phorid parasitoid infection (Orr et al. 1995;Porter et al. 1995;Feener 1987;Feener and Brown 1992). The extreme behavioral responses are interesting, given that parasitism rates of fire ant workers in field colonies of North America (where S. invicta is an invasive pest) is very low, at 0-2.42% per colony (Morrison and Porter 2005;Morrison et al. 1997) and that the success rate of oviposition attempts by female phorids is only 8-35% (Porter 1998). However, such parasitism pressure on fire ants reduces their interspecific competitive ability, resulting in a strong positive indirect effect of phorid flies on some nonhost ant species (i.e., the avoidance behavior by fire ants in response to phorid flies can favor other ant species) and thus were suggested to be a viable biological control agent of S. invicta in its invasive range (Morrison 1999).
It is reasonable to expect that changes in gene expression underlie the behavioral response of fire ants to phorid flies. However, little is known about the genes underlying the response of fire ants to parasitoids. There are several classes of genes that could potentially be involved. The foraging (for) gene (cGMP-activated protein kinase gene (PKG)) is an important behavioral gene in diverse taxa of animals (Ingram et al. 2011;Reaume and Sokolowski 2009;Ben-Shahar et al. 2002). In ants and other insects, for has been shown to have a direct relationship with foraging behavior (Ingram et al. 2011;Tobback et al. 2011;Ben-Shahar et al. 2002;Ingram et al. 2005;Wenseleers et al. 2008;Osborne et al. 1997;Ben-Shahar et al. 2003;Lucas and Sokolowski 2009). Foraging has been shown to have a high level of expression during daylight hours in foraging ants when they are active outside their nest (Ingram et al. 2016). Previous research has shown that colonies of the red imported fire ant, Solenopsis invicta, with relatively high levels of foraging activity also showed elevated levels of expression of the red imported fire ant for gene (Sifor) (Bockoven et al. 2017). Environmental factors have also been shown to influence the expression of Sifor. In foragers, Sifor is downregulated under low temperature and food deprivation (Zhou et al. 2020). However, it is not known whether the presence of phorid parasitoid flies affects the expression level of Sifor in S. invicta.
In addition to the Sifor gene being associated with task behaviors of worker ants, olfaction is critical for mediating various behaviors of social insects such as S. invicta (Du and Chen 2021). Odorant binding proteins (OBPs) are thought to play a role in olfactory perception (Pelosi 1994). In ants, OBPs are expressed in the antennae and serve in olfactory functions (McKenzie et al. 2014). Notably, OBP11 is thought to play essential roles in the response of foragers of S. invicta to encounters with natural enemies and food recruitment (Qiu et al. 2017).
Since oviposition by phorid parasitoid flies presents a risk of infection to S. invicta, there may also be upregulation of immune genes in response to their presence. After the penetration of the cuticle by a parasite or pathogen, innate immunity protects insects from microorganisms and is the first line of defense against infections (Tian et al. 2004;Uvell and Engström 2007). Antimicrobial peptides (AMPs) are an important part of the innate immune system and are the primary line of host immune defense against pathogens (Luiz et al. 2017;Huan et al. 2020;Uvell and Engström 2007). Mechanical stress and predator exposure can increase gene transcription for some AMPs in insects (Adamo 2017;Adamo et al. 2017). Hymenoptaecin (Hym) and defensin-2 (Def-2) have been shown to be significantly upregulated in S. invicta foragers compared to nurses (Qiu et al. 2017). Qiu et al. (2017 further point out that increased expression of Def-2 in foragers aligns with the observation that foraging workers contact more microbes when they are outside the nest. Moreover, Cytochrome P450s may play a role in immunity of insects. Cytochrome P450 genes encode enzymes involved in immune defense and the detoxification process in insects (Scharf et al. 2021;Yocum et al. 2018;Zhang et al. 2016b). A variety of cytochrome P450 genes are found in ants (Zhang et al. 2016b;Scharf et al. 2021). In S. invicta and other ants, cytochrome P450 genes are thought to play a role in immunity, and the detoxification of insecticides (Zhang et al. 2016a;Zhang et al. 2016b).
There is no information on how phorid parasitoid flies impact the molecular basis of host behavior change in S. invicta. Therefore, we aimed to elucidate some of the molecular mechanisms associated with the change in foraging behavior of S. invicta workers exposed to the fire ant decapitating fly, Pseudacteon curvatus. In the laboratory, we showed that fire ants reduced foraging activity when exposed to P. curvatus. We then investigated changes in gene expression of several genes related to foraging and immunity. These genes were Sifor, OBP11, Hym, Def-2, abaecin, and cytochrome P450 4C1-like (CYP4C1-like).

Fire ant collection and maintenance
Five S. invicta colonies were collected from three locations in College Station, Texas on 2 April 2021 (Table 1) by excavating their mounds and the soil directly underneath the mounds and placing the contents in a 20 L bucket. Colonies collected from the same area were located at least 300 m apart. Colonies were noted to be mature by the presence of alates (Tschinkel 1998). Within 24 h of collection, colonies were extracted from the soil by slowly dripping water into the bucket and the rafting fire ants were removed (Banks et al. 1981). To ensure that colonies had a low chance of workers infected with phorid parasitoid flies from the field environment, experiments did not begin until 33 days after S. invicta collection. This is sufficient time for many of the potential phorids present inside worker ants to complete their development at temperatures above 25 °C (Morrison et al. 1997;Porter et al. 1997;Porter 1998). Phorid flies were not observed on the days of S. invicta colony collection in the field. It was not possible to ensure low parasitism rates after experiments began; however, parasitized workers do not forage (Porter 1998), thus it is highly unlikely that they would be present in the foraging arena. Each colony was split into two sub-colonies to be used as the control and treatment, which consisted of at least three queens and ample brood. The colonies were then placed into a box (15-5/8" × 13-1/8" × 6-3/4" h) with the insides coated with Fluon and containing artificial nests. Each nest was a polystyrene Petri dish with plaster bottoms and darkened lids with two entry holes. Colonies were maintained at 26-27 °C and LD 14:10. They were fed an artificial diet (Dussutour and Simpson 2008) and freshly frozen cockroaches (Nauphoeta cinerea or Periplaneta americana). An artificial diet and insects were used because they are both important for fire ant growth (rather than a diet with a sole food source) (Porter 1989). We also determined the social form of each colony. There are two distinct social forms of the fire ant, S. invicta, monogyne and polygyne (Gotzek and Ross 2007;Tschinkel 2006), and they differ in many life-history traits (Tschinkel 2006;Ross and Keller 1995). To determine the social form of each nest, we extracted DNA from 10 foragers per nest using a DNeasy blood and tissue kit (Qiagen). We checked each pooled sample for the presence of the Gp-9 b allele, which is only present in polygyne colonies (Arsenault et al. 2020;Ross and Keller 1998;Krieger and Ross 2002). A Gp-9 b PCR assay was used on each pooled sample with the specific primer pair 24bS and 25bAS, which corresponds to social form as only polygyne colonies successfully amplify the b allele (Valles and Porter 2003). Visualization occurred on a 1% agarose gel. All our colonies were determined to have the polygyne social form.

Phorid parasitoid fly collection
For our experiments we used only P. curvatus flies to expose to the fire ants. Pseudacteon curvatus parasitoids for the behavioral assays were collected from the field by disturbing fire ant nests with a long apparatus. Each nest was re-disturbed every 15-20 min (depending on P. curvatus activity and presence of S. invicta workers). Once a fire ant colony is disturbed, alarm pheromones are released. Pseudacteon curvatus cues in on these pheromones (Ngumbi and Fadamiro 2015) and approaches the disturbed nest. For

Behavioral assay arena
The arenas were designed so that decapitating flies would remain contained and visible during the experiment (Fig. 1). Sixty-five L containers were flipped upside down, their bottoms were removed, and replaced with a clear plastic material (Fig. 1a). Inside, there was a box that contained the colony's nest and a smaller box that sat on top of the main colony box that was connected via a metal bridge (Fig. 1b).
The smaller box served as the foraging arena. A screw cap was also added so that S. invicta foragers could be collected with an aspirator during the different time points of the experiment. This design trapped phorid flies inside the enclosure, while allowing for the collection of ant foragers. Quikrete premium play sand was used to seal the bottom of the containers to prevent flies from escaping from the bottom of the enclosure.

Behavioral assays
Behavioral assays were conducted in a climate-controlled laboratory, started between 13:30 and 14:20 pm, and performed over 48 h. Each experiment lasted 48 h due to the lifespan of the phorid flies (P. curvatus). The cockroaches were removed 48 h prior to the behavioral assays. Fresh roaches were placed in foraging trays on the day of each assay. The experiment was conducted five times between 5 May and 6 June 2021. Before each assay, 50 P. curvatus (males and females) were collected (see above under Phorid parasitoid fly collection). The ants were provided a settling period of at least four days after moving them into the behavioral assay arenas and before beginning each assay. The dates of each assay are as follow: 10-12 May 2021, 17-19 May 2021, 25-27 May 2021, 1-3 June 2021, and 8-11 June 2021. The artificial diet was removed from each colony prior to introducing the phorid flies and replaced with a weighed quantity of artificial diet and fresh cockroaches, minutes before adding flies. Both were weighed after the experiment (at 48 h). Ten foragers were collected and snap frozen on dry ice at hours 0, 6, 24, 30, and 48. Foragers were considered any S. invicta within the foraging box. The foragers were then stored in − 80 °C prior to RNA and DNA extraction.

RNA extraction
Ten independent samples of fire ant foragers were evenly distributed from the five experiment dates from between 5 May 2021 and 11 June 2021. For each time point, total RNA from pooled samples of 10 whole body foragers was isolated using phenol-chloroform extraction with TRIzol reagent (Thermo Scientific™) and standard protocols (with an extra ethanol wash). SUPERase In™ RNase Inhibitor was used to inhibit RNases and to provide protection against RNA degradation. Total RNA at 40 ng/μl was treated using the TURBO DNA-free™ Kit (Invitrogen™) and then reverse A smaller container sat on top of a larger container. They were connected via a metal bridge. Each box contains Fluon on the sides so that ants cannot crawl up. The partitioning of the main colony box from the food source box allows for separation of workers (i.e., foragers are more likely to be in the food source box) transcribed into cDNA with the Verso cDNA Synthesis Kit (Thermo Scientific™).

Primers
A list of the target gene primers can be seen on Table 2. Primers for the target genes abaecin and Def-2 were designed with the NCBI primer-BLAST tool. Gene expression data in S. invicta were available for abaecin from Tian et al. (2004) and Def-2 from Qiu et al. (2017). Primers were purchased from Millipore-Sigma from these primer sequences. Ribosomal Protein 18 (RPL18) was used as a reference gene as it has been shown to exhibit stable expression among different developmental stages, castes and tissues in S. invicta (Cheng et al. 2013;Hawkings et al. 2019

Quantification of gene expression
We assayed relative expression of Sifor, OBP11, Hym, CYP4C1-like, abaecin, and Def-2, mRNA from pooled whole body fire ant foragers relative to an endogenous control gene (RPL18) using real-time quantitative reverse transcription-polymerase chain reaction (qPCR) techniques on a QuantStudio™ 6 Pro System with PowerUp™ SYBR™ Green Master Mix and manufacturer's protocols. Each 10 μl of qPCR reactions contained 2 μl of template cDNA, 0.25 μl of each primer (0.25 μl for each forward and reverse), 5 μl of PowerUp™ SYBR™ Green Master Mix, and 2.5 μl of nuclease-free water). A negative control (nuclease-free water from each cDNA run) was used. To calculate the relative fold change of gene expression of the samples, the delta-delta Ct (Δ ΔC t ) method was used (Schmittgen and Livak 2008) and transcript level for each target gene was normalized to the internal control gene, RPL18 (at hour 0 flies absent). There were 10 independent biological replicates of each sample type and two technical replicates for each sample (the mean of two samples was calculated) for each of the set of target gene qPCR reactions. The qPCRs were divided into two batches due to space limitations on the PCR plates: 1) Sifor, OBP11, Hym, and RPL18, and 2) CYP4C1-like, abaecin, Def-2, and RPL18.

Statistical analysis
There were 25 paired sample t-tests performed to compare food consumption between treatment and control colonies for both artificial diet and roaches. Five pairs of replicates were conducted during each experimental run (5 replicates × 5 runs).
Statistical analyses were conducted using paired t-tests of the of ΔC t scores to compare differences of gene expression levels in foragers at each sampling time point in the treatment (flies present) and control (flies absent). Due to multiple testing, the Bonferroni correction was applied, using p value < 0.01 for designating a significant difference in the presence of phorid flies for each gene (i.e., five time points for each of the six genes). All analyses were performed using the Excel Data Analysis package.

Effect of phorid fly presence on food consumption
The presence of phorids in the experimental arenas significantly affected food consumption by the fire ants. Based on the 25 paired-samples t-tests (5 replicates × 5 experimental runs), the experimental groups consumed about 16% less cockroach mass than the controls over the 48-h experimental  period (mean consumption (g) ± SD = 1.69 ± 0.31 and 1.42 ± SD = 0.54 for the treatment and control groups, respectively; paired-samples t-test: t 24 = 2.298, p = 0.006) (Fig. 2a). The treatment groups also consumed less artificial diet (mean consumption (g) ± SD = 2.487 ± 0.2485 and 2.775 ± 0.3311, for treatment and controls, respectively, paired-samples t-test: t 24 = 4.729, p < 0.001) (Fig. 2b). These results demonstrated that the presence of the phorid flies reduced the foraging activity of the fire ant colonies.

Effect of phorid fly presence on gene transcript levels
The mean distributions of threshold cycle (Ct) values for the RPL18 gene did not significantly differ in S. invicta samples from different treatments (flies present vs. flies absent) for either gene expression assay (first set of genes and additional immune genes), as determined by a paired-samples t-test (paired t-test: t 49 = 0.3799, p = 0.3923), confirming this gene was a suitable control for assessing changes in expression levels of the six target genes studied. There was an immune gene significantly upregulated by the presence of the parasitoid phorid flies. Def-2 showed increased expression in the presence of the phorid flies. Def-2 expression was significantly higher in foragers exposed to flies than in control colonies (flies absent) at 48 h (t 9 = 10.9086, p = 0.0039) (Fig. 3d). Def-2 expression was not significantly different at any of the other time points: t 9 < 0.6467, p > 0.0966 (Fig. 3d). Hym expression was not significantly different at any of the other time points: t 9 < 0.5460, p > 0.039 (Fig. 3a). Hymenoptaecin, Abaecin, and CYP4C1-like expression in foragers was not significantly different at any of the time points between the flies absent and flies present colonies as determined by a paired t-test: t 9 < 1.5597, p > 0.039 (Fig. 3a, e, and f). These results suggest that the presence of P. curvatus leads to an upregulation Def-2 over a time of 48 h, while Hym, abaecin, and CYP4C1-like expression remains unaffected by the presence of the phorid flies. There was a slight increase in CYP4C1like and Hym expression in colonies exposed to phorid flies at 48 h, however this increase was nonsignificant. Sifor and OBP11 expression in foragers was not significantly different at any of the time points between the flies absent and flies present colonies as determined by a paired t-test: p > 0.3018 ( Fig. 3b and c). These results suggest that in the presence of P. curvatus, Sifor and OBP11 expression are not significantly changed by the presence of the phorid parasitoid flies.

Discussion
Unravelling the complex relationship between behavior and its genetic mechanisms is an important area of behavioral research. The results of this study represent a step in increasing our understanding of this relationship. We found that the AMP gene Def-2 increased in expression in foragers at 48 h after exposure to decapitating flies (P. curvatus). It is important to note that the workers were exposed to parasitoid flies and not necessarily parasitized, and that the 48-h cut off was primarily based on the lifespan of active adult Pseudacteon phorid flies. The lifespan is 3-7 days for inactive phorid flies, but much shorter when the phorid flies are active and attacking the fire ants (Porter 1998). This contrasts with other studies investigating immune response to a parasite because we investigated gene expression related to exposure, not parasitization. To the best of our knowledge, other studies have primarily investigated gene expression in relation to infection.
The increase of Def-2 expression in colonies exposed to decapitating flies adds to our understanding of the mechanisms of parasite-host interactions. Defensins are a family of small cationic, variable antibiotic peptides found in the hemolymph of insects (Ganz and Lehrer 1995;Wu et al. 2018). They are ancient innate antibiotics with strong antimicrobial activity against a variety of microorganisms (Wu et al. 2018;Ganz 2003). Defensins are regulated by the interaction of Toll and Imd signaling pathways and have a broad spectrum of antimicrobial action (Ilyasov et al. 2012). In the northern blow fly, Phormia terranovae, Def has been shown to increase within 8 h after the blow flies are injected with low doses of bacteria (Dimarcq et al. 1990). Further, even after a sterile injury, expression of Def increased (Dimarcq et al. 1990). This indicates that Defensin genes can increase in expression without direct infection of a pathogen or parasite. In the kissing bug, Triatoma brasiliensis, Def3 and Def4 were induced after feeding on a blood meal and the expression of Def4 was highly up-regulated in the stomach and fat-body tissue three and five days after feeding (Waniek et al. 2009). Def1 has also been shown to be expressed in the cardia and stomach tissue of T. brasiliensis, Fig. 3 Effect of phorid parasitoid fly presence on gene transcript levels in Solenopsis invicta foragers. Representative genes are as follows: a hymenoptaecin, b odorant-binding protein 11, c Sole-nopsis invicta foraging gene (Sifor), d defensin-2, e abaecin, and f cytochrome P450 4C1-like. Bars indicate mean ± SE levels of relative gene expression. The * symbol denotes significance presumably to help control luminal symbionts (Araújo et al. 2006). Perhaps the upregulation of these AMP genes is associated with stress. Stress is known to lead to changes in gene expression patterns (Spriggs et al. 2010) and has been shown to trigger downregulation of epidermal AMP expression in mice (Aberg et al. 2007). In the moth, Ostrinia furnacalis, low temperatures led to an upregulation of immunity related genes such as AMPs (Chen et al. 2019). Such an immune response is thought to be involved in resistance to a cold environment for the O. furnacalis larvae (Chen et al. 2019).
Insects can produce a variety of AMPs which are synthesized as inactive precursor proteins or pro-proteins, and active peptides (Yi et al. 2014). After the defensive protection provided by the cuticle (Tian et al. 2004), innate immunity is the first line of defense against infection caused by microorganisms for many different animals, including insects (Uvell and Engström 2007). For hymenopteran insects there are multiple AMPs of interest. In the honey bee, Apis mellifera, and the buff-tailed bumblebee, Bombus terrestris, abaecin and hymenoptaecin (Hym) are two AMPs that may respond to infection by a pathogen (Saltykova et al. 2005;Erler et al. 2011;Casteels et al. 1993). Abaecin is a major broad-spectrum antibacterial proline-enriched cationic peptide (Luiz et al. 2017) and Hym is a cationic polypeptide demonstrated to have broad-spectrum antibacterial defense in honey bees (Casteels et al. 1993). In honey bees, Hym has been shown to be regulated by Imd pathway and abaecin is regulated by both the Imd and Toll pathways (Lourenço et al. 2018). Abaecin and Hym have been shown to be up-regulated in S. invicta dealate queens with no infection status or apparent injury, suggesting that the immune system is activated during the transition from alate to dealate (i.e., ripping off their wings after mating) (Tian et al. 2004). In feral honey bee colonies, Hym expression and other immune genes have been shown to increase in expression and it is thought that higher levels of immune expression is a response to high pathogen burdens (Hinshaw et al. 2021). In honey bees, Hym and Def1 have been shown to increase in mRNA expression in honey bees infected with the parasitic mite, Tropilaelaps mercedesae (Wu et al. 2020).
Although the presence of the phorid parasitoid flies changes the foraging behavior of its fire ant host, the underlying genetic mechanisms suggests a more complicated story: the expression of Sifor and OBP11 remain unchanged in their presence. These two initial findings alone are fundamentally interesting because Sifor is known to experience changes in its expression due to environmental factors (Zhou et al. 2020;Lei et al. 2019) and OBP11 is thought to play a role in the response of foragers to natural enemies (Qiu et al. 2017). Although OBP11 expression levels did not change within 48 h of exposure to phorid flies, it is possible that other OBP genes may undergo changes in expression and play a role in the behavioral response of fire ants to phorid parasitoid flies within this time frame. OBPs and chemosensory binding proteins (CSPs), as well as deactivation enzymes such as cytochrome P450s, belong to large gene families that are involved in a variety of chemosensory and non-chemosensory function, which further complicates elucidating their potential roles (Shah and Renthal 2020). It is possible that expression levels of some OBPs may change in foragers in response to environmental stimuli given that there are 24 OBPs identified in the S. invicta genome (Shah and Renthal 2020;Gotzek et al. 2011;Pracana et al. 2017). Looking at the gene expression of CSPs and OBPs of infected S. invicta workers may be interesting as CSPs and SiOBP15 have been shown to have increased expression in workers infected with the fungus, Beauveria bassiana (Zhang et al. 2021). SiCSPs and SiOBPs increased in expression at 24-and 48-h post-infection (there were mixed responses at 72 h) (Zhang et al. 2021). This broad host range insect pathogen can parasitize and kill S. invicta fire ants because of high population density within the nest; thus, a number of social strategies (i.e., individual and social immunity) exist for managing pathogens (Zhang et al. 2021;Fan et al. 2012). The results from Zhang et al. (2021) allude to the mechanisms for detection and response to pathogens as CSP and OBP genes experience change in expression due to microbial infection.
There was an increase in CYP4C1-like and Hym expression in the colonies exposed to phorid flies at 48 h, however both changes were nonsignificant. It is possible that CYP4C1-like and Hym may be upregulated in fire ants exposed to decapitating flies at later times; however, we did not test this hypothesis, as active P. curvatus has a short lifespan of a few days (Portilla et al. 2007). At 48 h, most of the P. curvatus were alive in our behavioral arenas, but by 72 h, more than half perished. Perhaps this is why there is a change in expression at 48 h and not at previous time points. It may take a more prolonged exposure period than provided here for gene expression levels to change. In future studies, it may be possible to add additional flies throughout the assay, but it is difficult to count remaining flies without removing the container and disrupting the experimental arenas. Interestingly, in our currently unpublished RNA-Seq data from another study, we see decreased expression of CYP4C1-like in fire ant workers infected with P. curvatus decapitating flies. Therefore, CYP4C1-like may play a role in preventing parasitization of P. curvatus and thus needs to be successfully downregulated by the parasitoid to bypass the host's immune defense. However, more experimentation is needed to disentangle the specific roles of cytochrome P450 genes in fire ants.
No significant change in gene expression of abaecin was found between treatment and control S. invicta foragers, therefore phorid parasitoid fly presence does not appear to affect its gene expression. In honey bees (A. mellifera), abaecin levels have been shown to respond to infection by a pathogen (Saltykova et al. 2005;Erler et al. 2011;Casteels et al. 1993), however, this is different from our study because we are looking at exposure and not direct infection. There are some examples in A. mellifera worker brood infected with certain viruses that show no significant differences in the transcription levels of abaecin (Khongphinitbunjong et al. 2015). This is perhaps because this gene (along with Hym) might have multiple functions (Khongphinitbunjong et al. 2015;Evans 2006). Insect innate immune responses are similar among insect taxa (Viljakainen et al. 2009;Gillespie et al. 1997). However, a robust immune response is especially important to eusocial insects such as bees and ants because nestmate proximity and density increases exposure to pathogens and their transmission (Holmes and Johnston 2022). The immune defense system protects against infections of various origins, such as through wounds (Uvell and Engström 2007). In the case of fire ants, the phorid parasitoid flies stab their host with a sharp ovipositor to lay an egg, thus creating a wound. Just the mere presence of phorid parasitoids or a stab inflicted by their oviposition may initiate an immune response, but this should be tested by investigating infected, exposed, and uninfected individual worker ants. In addition, the immune genes tested here should be explored beyond 48 h. It is possible that a further increase in upregulation may occur. Due to the pooled nature of the samples, we were not able to determine which S. invicta foragers, if any, were parasitized in this study. It may be that some individuals are responding strongly, and many individuals are responding moderately. There may have been a few parasitized foragers in the pooled samples that had a strong immune response to P. curvatus infection, and we are observing only the mean response of all the individuals. Whether or not infection was successful, it is highly likely that many of the foragers sampled were attacked by a female phorid parasitoid fly as a female will attempt to oviposit 30-120 times within a period of an hour (Morrison et al. 1997;Farnum and Loftin 2011). Whether or not successful, wounds greatly increase the risk of infection, which increases the need for immune protection, especially in the presence of predators (Dhabhar 2014;Adamo et al. 2017). When a parasitoid lays an egg, the host resists infection by a haemocyte-mediated immune response that surrounds the egg with capsules (i.e., cellular structures that form around objects larger than bacteria) (Ratcliffe and Rowley 1979;Fellowes and Godfray 2000;Carton and Nappi 2001;Lavine and Strand 2002;Theopold et al. 2004). In hymenopterous parasitoids, venom and associated constituents (e.g., polydnaviruses) prevent cellular encapsulation of injected eggs of their hosts (Feener and Brown 1997;). It is still not known how the larvae of the phorid fly parasitoids avoid the immune system of their hosts (Feener and Brown 1997;Elizalde et al. 2019). In Pseudacteon spp., once an egg hatches inside the thorax of a fire ant worker, the newly hatched maggot migrates into the head of the ant host (Porter et al. 1995). Perhaps quickly moving to the host's head is a way to escape the host's immune response. They may have associations with viruses, bacteria, or other microorganisms, but little is known about the possible involvement of microbes. Elizalde et al. (2019) investigated immune response of leaf-cutting ants to their phorid parasitoids and found that various leaf-cutting ant species differed in their immune responses. Particularly, there was interspecific variation in encapsulation response, phenoloxidase (PO) and prophenoloxidae (PPO) levels, and number of haemocytes . It is possible that PO and PPO play a role in S. invicta immune defense against their phorid parasitoids, however our attempts to use the sequences obtained from a blast search of the S. invicta genome did not amplify so we were not able to investigate their possible role in this study. Future work should focus on exploring the role of immunity in S. invicta in response to phorid fly parasitoids, specifically with the aim of disentangling mere exposure to the presence of the flies from response to actual parasitism by the flies.

Conclusions
Our study shows that fire ant, Solenopsis invicta, foragers experience an upregulation of Def-2 expression after 48 h when exposed to the phorid parasitoid fly, Pseudacteon curvatus. The Hym, abaecin, and CYP4C1-like genes did not experience significant changes between the treatment and control groups. Although the fire ants showed a change in their foraging behavior by consuming less food, it is not reflected at the level of Sifor or OBP11 expression. These results provide insights into changes in gene expression in S. invicta exposed to phorid parasitoid flies and stress the need for further information on how fire ants are impacted by the presence of phorid flies (Pseudacteon spp.).

Supplementary Information
The online version contains supplementary material available at https:// doi. org/ 10. 1007/ s00040-023-00930-7. adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/.