Leaf-cutting ants depend on mutualisticfungi to survive. An infection that massively affects the workers compromising the proper maintenance of the fungus, or that can attack the fungus garden, can be fatal to the colony. Thus, leaf-cutting ants have evolved a complex defense system composed of both innate individual immunity and collective immunity to protect the colony against potential threats. To characterize the collective and individual immunity of Atta cephalotes workers to Metarizhium anisopliae we assessed the hygienic behavior and the expression of antimicrobial peptides of A. cephalotes workers triggered by Metarizhium anisopliae spores. As a control challenge, workers were treated with water. Regardless of whether the challenge was with water or spore suspension, A. cephalotes workers displayed an immediate response characterized by an increase in time spent both self-grooming and collective grooming along with a reduction in time spent fungus-grooming. The individual immunity triggered the expression of abaecin as early as 24 hours post-infection, exclusively in workers challenged with M. anisopliae. In contrast, the level of expression of defensin remained constant. These results suggest that upon being challenged with a suspension of M. anisopliae spores, A. cephalotes workers deploy both collective and individual immunity to produce a response against the invader. However, when the spores of M. anisopliae are applied as liquid suspension collective immunity deploys a generic strategy, while individual immunity shows a specific response against this entomopathogen.

abaecin, allogroming, defensin, expression, fungus grooming, gene, self-grooming

Attine ants (Formicidae: Myrmicinae: Attini: Attina) comprise approximately 250 species that establish associations with mutualisticfungi as a source of nutrition (Ward et al. 2015; Branstetter et al. 2017). Leaf-cutting ants of genera Atta and Acromyrmex are among the most derived of the tribes, in which they cultivate Leucoagaricusfungi using fresh plant material as a substrate (Chapela et al. 1994; Mueller et al. 2001, 2005). In return, mutualistfungi produce gongylidia—specialized structures used to feed the queen and the brood (Hölldobler and Wilson 2010). To assure fungus garden productivity, the workers select, harvest, and process large quantities of suitable plant material (Littledyke and Cherrett 1978; Folgarait et al. 1996; Estrada et al. 2013; Rocha et al. 2017) and build complex nests underground to provide an optimum environment for fungus growth (Kleineidam and Roces 2000; Bollazzi and Roces 2010; Verza et al. 2017). For instance, the nest can have up to thousands of chambers inhabited by millions of individuals in species of Atta (Jonkman 1980; Moreira et al. 2004).

The social lifestyle of leaf-cutting ants could, in principle, make them prone to infections with hazardous microorganisms for workers or the fungus garden (Hughes et al. 2002) due to enhanced transmission between frequently interacting individuals with a similar genetic background that possibly are susceptible to the same pathogens (Wilson et al. 2003; Pie et al. 2004; Cremer et al. 2007). In this sense, workers are in permanent contact withfungi that can overgrow the mutualist (e.g., Syncephalastrum racemosum and Trichoderma harzianum) that inhabit the soil and within the plant tissue they cut (Rodrigues et al. 2009; Rocha et al. 2017). Furthermore, infection of the fungus garden by Escovopsis, a specialized parasite of the mutualist (Muchovej and Della Lucia 1990; Seifert et al. 1995; Currie et al. 1999; Currie and Stuart 2001), is frequently detected in the garden of leaf-cutting ants, but it only significantly compromises the garden if the ants cannot groom the mutualist or properly dispose of the waste (Heine et al. 2018).

To protect the colonies against infection, leaf-cutting ants have evolved a complex defense system composed of the innate immunity of individuals and collective immunity. Individual immunity involves physiologic mechanisms to clear potential threats, including the production of reactive oxygen species, encapsulation, and the production of antimicrobial peptides (Hoffmann 1995). In contrast, collective immunity is based on altruistic behaviors that result in avoidance, control, or elimination of parasitic infections (Siva-Jothy et al. 2005; Masri and Cremer 2014). Once a potential hazard is detected, the workers deploy hygienic behaviors of collective immunity including self-grooming, allogrooming (Fernández-Marín et al. 2003; Reber et al. 2011), and fungus grooming (Currie and Stuart 2001; Little et al. 2006; Cremer et al. 2007). The particles removed via hygienic behaviors are compacted, possibly sterilized in the infrabucal pocket, and dumped as pellets (Little et al. 2003, 2006). This hygienic behavior is complemented by chemical defenses mediated by secretions of exocrine glands and the metapleural gland that contain fungal and bacterial inhibitors (Do Nascimento et al. 1996; Poulsen et al. 2003; Rodrigues et al. 2008; Fernández-Marín et al. 2015).

In addition to these mechanisms, Acromyrmex workers maintain a symbiosis with Pseudonocardia—an antibiotic-producing actinobacterium maintained in the cuticle. This association protects the fungus garden against infection with parasitic fungus Escovopsis (Currie et al. 2003; Poulsen et al. 2010; Cafaro et al. 2011). In contrast, Atta species have lost or reduced the symbiosis with Pseudonocardia (Currie et al. 2006; Mueller et al. 2008) suggesting that they rely more on collective immunity and chemical disinfection to protect the colony against invaders (Fernández-Marín et al. 2009).

The first barrier of individual immunity in leaf-cutting ants is a hard exoskeleton reinforced with a biomineral armor that protects workers from invaders in most cases (Siva-Jothy et al. 2005; Li et al. 2020). However, specific pathogens such as entomopathogenic fungus Metarizhium anisopliae can penetrate the worker’s cuticle and reach the hemocoel (Hajek and St. Leger 1994; Moino Jr et al. 2002). In insects, constitutive mechanisms of innate immunity such as profenol oxidase cascade, phagocytosis, and reactive oxygen species production act to neutralize the microorganism as the first line of defense (Nappi and Christensen 2005; Evans et al. 2006; Buchon et al. 2014). The expression of antimicrobial peptides (AMPs) is induced after this generic response. These molecules eliminate the microorganisms via pore formation or metabolism disruption (Bulet et al. 2004; Siva-Jothy et al. 2005; Haine et al. 2008). The genome of leaf-cutting ants, Atta cephalotes and Acromyrmex equinatior contains sequences that code to abaecins, hymenoptaecins, and defensins (Zhang and Zhu 2012). Furthermore, abaecin and hymenoptaecin are expressed as a part of the immune response of Ac. equinatior workers against M. anisopliae (Yek et al. 2013).

The collective immunity of Acromyrmex against M. anisopliae has been previously described (Richard and Errard 2009; Walker and Hughes 2009; Morelos-Juárez et al. 2010; Abramowski et al. 2011; Yek et al. 2013; Tranter et al. 2015; Nilsson-Møller et al. 2018; Calheiros et al. 2019). In contrast, the collective immunity of Atta has been less explored (Fernández-Marín et al. 2006; Fernández-Marín et al. 2009; Walker and Hughes 2011) and the research has been primarily centered on describing metapleural gland grooming. While this evidence highlights the relevance of chemical disinfection in Atta species, it does not assess the full spectrum of workers’ behavior upon a challenge with an entomopathogen. Furthermore, studies that simultaneously assess innate and collective immunity responses against an entomopathogen in leaf-cutting ants are scarce (Yek et al. 2013). To characterize the collective and individual immunity of A. cephalotes workers to M. anisopliae, we assessed workers’ hygienic behavior before and after a challenge with M. anisopliae. We evaluated the time spent by workers executing self-grooming, allogrooming, metapleural gland grooming, fecal fluid grooming, and the production of infrabuccal pellets. In addition, we evaluated the expression of antimicrobial peptides abaecin and defensin as effectors of individual immunity. We hypothesized that upon a challenge with M. anisopliae, workers deploy collective and individual immunity to clear the entomopathogen.

The ICC was 0.76, thus suggesting that the behavioral observations were consistent between observers. Self-grooming was a behavior in which workers invested more time before and after the application of challenges. In the basal status, workers spent a median of 1000 seconds dedicated to self-grooming; the investment in fungus grooming was seen for a median of 160 s; the time invested in allogrooming and fecal fluid grooming was under 100 s (Fig. 2 A–D). The metapleural gland grooming was detected only once in the 5440 s of footage; hence, the time invested in this behavior is negligible.

Figure 2. 

Time invested by Atta cephalotes workers in prophylactic behavior during the basal period (0 h) the first hour, and the second hour after challenge with water (white boxes) or Metarizhium anisopliae spores (gray boxes) A self-grooming B allogrooming C fungus grooming, and D fecal fluid grooming. Each box represents the sum of the time in seconds invested in each behavior from 10 workers (n = 5 nests, 60 workers per nest). Different letters indicate significant differences (p < 0.05).

Workers reacted to the treatments by increasing their investment in self-grooming: However, the time after challenge (F = 49.54, p < 0.0001, Df 9)—but not the nature of the treatment itself (F = 0.66, p = 0.4, Df 9)—influenced their behavior. In the first hour after the challenge, workers duplicated the time investment in self-grooming behavior (Tukey test p < 0.0001). The time investment in self grooming decreased in the second hour, but it was higher than in the basal status (Tukey test p < 0.0001) (Fig. 2A).

A similar tendency was observed for the time invested in allogrooming. Workers increased the time investment in this behavior depending on the time after the treatment (F = 6.32, p = 0.0028, Df 9) no matter whether they were challenged with water or M. anisopliae conidia (F = 0.16; p = 0.68, Df 9) (Fig. 2B). Workers almost duplicated the time invested in allogrooming in the first hour after challenges. This variable returned to basal levels in the second hour (Tukey test p = 0.83).

Workers showed a significant reduction in the time invested in fungus grooming after challenge (F = 6.4, p < 0.0026, Df 9) independent of whether they were challenged with water or spores (F = 0.78, p < 0 .37, Df 9) (Fig. 2C). In the first hour after the challenge, workers reduced the time investment in fungus grooming by approximately 60% (Tukey test p = 0.0228). This reduction was maintained in the second hour after treatment (Tukey test p = 0.010). In contrast, workers did not modify the time invested in fecal fluid grooming after the challenge neither in response to the time after treatment (F = 2.02, p = 0.16, Df 9) nor to nature of the treatment (F = 0.54, p = 0.58, Df 9) (Fig. 2D). Finally, workers treated with M. anisopliae showed ten-fold more infrabuccal pellets than control workers treated with water 24 h after the challenge (F = 124.80, p < 0.0001) (Fig. 3).

Figure 3. 

Production of infrabuccal pellets 24 hours after control challenge with water or spores of Metarhizium anisopliae. (n = 5 nests, 60 workers assessed per nest). Significant differences are marked with asterisks (* p < 0.05, ** p < 0.01, *** p < 0.001).

The workers’ survival decreased progressively in control-treated workers and workers treated with spores over the ten-day observation period. However, the workers treated with M. anisopliae spores showed higher mortality from day four than workers treated with water (z = 2.80, p = 0 .005) (Fig. 4). Furthermore, the colonization assay showed that 80% of worker’s corpses were effectively colonized by M. anisopliae; no signs of colonization were observed in water-treated workers (data not shown).

Figure 4. 

The survival probability of Atta cephalotes workers upon control challenge with water (blue line) or spores of Metarhizium anisopliae (black line). Blue and black dashed lines represent the 95% confidence interval. n = 4 nests with 60 workers assessed per nest.

The expression of abaecin was affected by treatment (F = 6.2, p = 0.03). Workers treated with M. anisopliae showed an increase in the expression of abaecin as early as 24 h after the challenge. The expression of abaecin increased by nearly 200-fold at 48 h versus immediately after the challenge (Fig. 5A). In contrast, the expression of abaecin remained similar across time points in workers treated with water. The expression of defensin increased at 48 h after the challenge in workers treated with a spore suspension; however, no significant differences were detected versus expression before treatment (Fig. 5B).

Figure 5. 

Expression of antimicrobial peptides in workers of Atta cephalotes upon challenge with water (white boxes) or spores of Metharhizium anisopliae (gray boxes) A abaecin B defensin. n = 3 nests; 20 workers assessed per time point. Significative differences are marked with asterisks (* p < 0.05, ** p < 0.01, *** p < 0.001).

The results indicated that under laboratory conditions, A. cephalotes workers reacted to a potential hazard by increasing their time investment in self-grooming and allogrooming and reducing their interaction with the fungus. These hygienic behaviors were complemented by the production of infrabuccal pellets. However, the results suggest that workers could not discriminate an innocuous challenge from an actual hazard. Hence, the time invested in hygienic behaviors was similar between workers treated with water and workers treated with spores of M. anisopliae. Furthermore, workers were unable to clear the spores completely under this experimental setup—the invader reached the hemocoel triggering a rise in the expression of abaecin. Finally, infection with M. anisopliae decreased the workers’ survival.

Workers of A. cephalotes combined self-grooming and allogrooming to remove the contaminants applied directly to their bodies, thus suggesting that this is the first line of defense against a potential hazard. Self-grooming has been identified as a proactive behavior widely extended in social insects including ants, wasps (Sumana and Starks 2004), bees (Morfin et al. 2020), and termites (Yanagawa et al. 2008). In leafcutter ants, this behavior is deployed by several species in response to challenges applied directly to the workers and to those applied to the mutualistfungi (Wilson 1976; Wheeler 1984; Morelos-Juárez et al. 2010; Reber et al. 2011; Walker and Hughes 2011; Bos et al. 2012; Nilsson-Møller et al. 2018; Bos et al. 2019), thus suggesting the relevance of this behavior to remove contaminants.

Allogrooming complements self-grooming hence allowing two or more individuals to clean the body of a third party in places that this one cannot reach, thus clearing potential hazards (Hughes et al. 2002; Reber et al. 2011). However, the time investment in self-grooming was 10-fold higher than the time invested in allogrooming. The difference in the time investment between those behaviors is probably explained because allogrooming depends on the proportion of clean and contaminated workers. In carpenter and leaf-cutting ants, if the treatment is applied simultaneously to the majority of individuals in a group, then the investment in this behavior decreases versus the investment observed when there is a proportion of clean workers (Bos et al. 2012; Da Silva Camargo et al. 2017). In this situation, clean workers dedicate time to clean other workers in the group.

The workers did not alter the investment in fecal fluid grooming, thus suggesting that this behavior is not deployed in response to challenges with water or M. anisopliae. In agreement with those findings, Ac. echinatior workers perform fecal fluid grooming to prepare plant material for degradation in (Kooij et al. 2016). Similarly, fecal fluid grooming is performed by Ac. octospinosus minor workers when they incorporate plant material into the mutualist garden (Forti et al. 2020), and also by workers that make contact with the mutualist (Nilsson-Møller et al. 2018). Finally, it has been observed in foundress queens of Ac. octospinosus during the preparation of plant material to incorporate into the fungus garden (Fernández-Marín et al. 2003).

The deployment of hygienic behaviors led to the production of infrabuccal granules. An increase in infrabuccal pellet production has been previously described in Acromyrmex and Tachymirmex workers after exposure to Escovopsis and Penicillium (Fernández-Marín et al. 2006; Little et al. 2006). In line with these findings, the infrabuccal pocket has been described as a structure that compacts and possibly sterilizes the material removed by the workers during hygiene (Little et al. 2006). Furthermore, the final disposition of the pellets in the waste chambers helps isolate the potential hazard, thus protecting the workers and the fungus garden from further contamination (Currie and Stuart 2001; Little et al. 2003).

Atta cephalotes workers respond to the challenges applied in this study by modifying the time invested in the hygienic behaviors, but they could not produce a differential response between an innocuous challenge and an actual threat with an entomopathogen. This finding contrasts with previous reports, which show that leaf-cutting ants deploy a differential response against innocuous substances like talc and an actual threat (Fernández-Marín et al. 2006; Morelos-Juárez et al. 2010; Yek et al. 2013; Tranter et al. 2015; Nilsson-Møller et al. 2018). This finding cannot be attributed to the innocuity of the strain of M. anisopliae because the colonization assay and the survival curve showed that the strain of M. anisopliae has a virulence mechanism that allows it to reach the hemocoel of its host and significantly impacts its viability.

It is also possible that the application of the entomopathogen as a suspension of spores hinders the detection of specific cues by the workers. Although the mechanism that mediates the recognition of hazards in leaf-cutting is not well understood, microorganisms might release semiochemicals, volatile compounds detected by ants as cues of danger, thus triggering a prophylactic response (Davis et al. 2013; Goes et al. 2020). Thus, it is possible that those semiochemicals are diluted in suspension, thus hindering workers from detecting the threat. In this line of evidence, similar to the findings reported here, workers of Atta colombica did not generate a differential response to a suspension of spores of M. anisopliae and a challenge with a solution of detergent in water (Walker and Hughes 2011). Furthermore, workers of Formica selysi challenged with different concentrations of a suspension of Metarhizium bruneum spores at a concentration of 10⁷ equivalent to the one used in this study workers produced a response similar to the one observed in control. In contrast, they deployed a differential response when they were challenged with a concentration of 10⁸ spores/ml (Reber et al. 2011), thus supporting the idea that the concentration of the microorganisms is essential to its recognition and the scaling of the behavioral response against it.

Workers of A. cephalotes did not deploy the grooming of the metapleural gland in response to the challenge with M. anisopliae. This evidence is in contrast with a previous report showing that A. cephalotes workers increase the frequency of this behavior up to 150 times in response to a challenge with dry spores of Penicillium. This indicates that this is the primary mechanism of response against fungal threats among A. cephalotes (Fernández-Marín et al. 2006). The differences between these findings can be explained by the strategy used to deliver the challenge to the workers. Dry spores may elicit metapleural gland grooming because they are easily detected by ants whereas the cues released by the microorganisms are diluted in suspension, thus hindering recognition. Nevertheless, among workers of Ac. echinatior, the inoculation of the fungus garden with dry Escovopsis spores did not increase the frequency of metapleural grooming thus indicating that another factor is necessary to trigger this behavior (Nilsson-Møller et al. 2018). Recently, Tachymirmex workers have been shown to increase the frequency of metapleural gland grooming upon challenge with germinated spores of Metarhizium and Escovopsis versus the challenge with ungerminated conidia, thus suggesting that the status of the microorganisms also influences the workers’ response (Bonadies et al. 2019).

The evidence shows that although treatment with M. anisopliae spores, significantly impacts workers’ viability, the treatment with water also caused a reduction in this parameter. Previously, it has been shown that A. cephalotes workers lose viability when isolated on Petri dishes, even when left untreated (Valencia-Giraldo et al. 2021), possibly because they do not consume the agar diet. In contrast, colonies maintain their viability under laboratory conditions six weeks after the extraction (Valencia-Giraldo, personal observation). These findings suggest that experimental conditions may alter the health status of workers. Hence it cannot be ruled out that this factor may influence workers’ behavior avoiding the discrimination between a sham challenge and the challenge with an entomopathogen.

In terms of individual immunity, evidence shows that workers of A. cephalotes increase the expression of the antimicrobial peptide abaecin. The peak in the expression of abaecin at 48 h is consistent with the dynamics of invasion of M. anisopliae, which reaches the hemocoel between 24 and 48 h after hosts exposure (Moino Jr et al. 2002). In insect hosts, abaecin has shown antibacterial activity (Casteels et al. 1990; Otvos 2002; Rahnamaeian et al. 2015). However, the expression of this AMP in response to fungal infection has also been reported in Apis mellifera (Bull et al. 2012) and Ac. echinatior (Yek et al. 2013) suggesting that this peptide might protect against M. anisopliae.

In contrast, the expression of defensin was not altered in response to infection with M. anisopliae. This finding contrasts with evidence showing that expression of defensin in response to infection with M. anisopliae is increased in social insects including A. mellifera (Bull et al. 2012) and Lasius neglectus (Konrad et al. 2012) as well as in solitary insects like Aedes aegypti and Locusta migratoria (Cabral et al. 2020; Jiang et al. 2020). This finding may be explained by reducing genes related to individual immunity observed in A. cephalotes (Suen et al. 2011). However, the IMD and toll pathway genes responsible for the signaling transduction that lead to defensin expression and are functional in this species (Schlüns and Crozier 2007; Suen et al. 2011). Therefore, the explanation for the low defensin expression during the Metarhizium challenge remains unclear.

The results show for the first time that A. cephalotes workers deploy mechanisms of collective and individual immunity upon challenge with M. anisopliae spores. Under this particular experimental setup A. cephalotes workers cannot discriminate between a hazardous stimulus and an innocuous one; hence, they deploy a generic behavioral response independent of the level of threat posed by the challenge. In contrast, once M. anisopliae reaches the hemocoel, individual immunity recognizes the danger and triggers the expression of abaecin, possibly as a defense mechanism against the invader.

We would like to thank Vicerrectoria de Investigaciones of the Universidad del Valle for financial support through project grants number CI 71153. We thank Dr. James Montoya- Lerma and Inge Armbrecht and the members of the Group on Ecology of Agroecosystems and Natural Habitats (GEAHNA). We are grateful to Andrea López Peña and the DAGMA-Cali team, especially Diana Ortiz and Elsy Alvear, for support throughout the study.

This work was financed by Vicerrectoría de Investigaciones- Universidad del Valle grant number 53111: Defense strategies against bio-controllingfungi in leaf cutter ant Atta cephalotes (Hymenoptera: myrmicinae): synergism between individual and social immunity. Juan Sebastián Gómez Días was funded by Vicerrectoría de Investigaciones- Universidad del Valle Grant 71202.