WHILE THE ATTINE ants are not well-studied, I posted this week about a study into their evolution that revealed the history of innovations in their cultivation of fungus. The attine ants are part of a symbiotic network between the ants, their fungal cultivars, the parasitic fungus Escovopus, and actinomycete bacteria that serve to suppress this parasite. These bacteria are members of Pseudonocardia and grow in filamentous mycelia on the insects’ integument, where the ants have evolved cuticular crypts to house the bacteria and glandular secretions that support their growth.
The actinomycetes are an order of bacteria that are know to produce a wide range of biologically active molecules, many of which are active against other bacteria and against fungi. Some of these natural products are now used clinically, such as the antibacterial antibiotics streptomycin, erythromycin, and tetracycline, anticancer drugs daunorubicin and doxorubicin, and antifungal drugs amphotericin B and natamycin. Actinomycetes inhabit a variety of environments, but many are ubiquitous soil bacteria.
The evolution of symbiotic relationships often involves extreme specialization. Attine ants are dependent upon their fungal gardens for food, while many of their cultivars are so adapted to being farmed by the ants that they cannot survive outside a colony. Fungal cultivars descend vertically with the ants, as each queen that leaves the nest takes a tiny fragment of fungus with her to seed a new garden.
Escovopsis fungi are parasitic upon this association. While Escovopsis has coevolved with this mutualistic association such that it is not found outside attine colonies, the fungus is not passed down vertically. It is transmitted laterally from colony to colony, perhaps by hitching a ride upon the other arthropods that occasionally infiltrate ant colonies. Because of this lateral transmission Escovopsis preserves more flexibility than the ants and their fungal cultivars. While each species of Escovopsis has a preferred host fungus, it can also parasitize alternative cultivars.
It turns out the the actinomycetes that grow upon attine ants’ cuticles is even less specialized. Kost and coworkers report collecting ants from a variety of fungus-cultivating and non-fungus-cultivating colonies and culturing the bacteria living on their cuticles. The attine ants (Acromyrmex octospinosus) were collected from seven different colonies, three from one area and four from another area 5 km away. Non-attine ants (Lasius flavus and Myrmica ruguluosa) were collected from three colonies in one area in Germany. They found that among the attine ants most ants from the same colony carried the same strain of bacteria, but some individual ants carried up to three strains. Each colony could contain up to seven different strains. Statistical studies showed that the number of bacteria strains collected was limited by sample size, so the actual bacteria diversity in attine ant colonies is actually greater than that revealed by this study. While it is usually thought that Pseudonocardia is the only strain mutualistic with attine ants, random analysis of two strains isolated showed that they were Streptomyces, another actinomycete.
When they carried out a study of Escovopsis growth inhibition by these bacteria, they found strains both from attine ants and non-attine ants suppressed parasite growth, although there was considerable variation in effectiveness among the different strains. However, the strains on the attine ants were considerably more effective, sometimes completely suppressing fungal growth, which none of the bacteria strains from non-attine ants could achieve. While attine ants carried the most effective strains they also carried three strains that did not suppress Escovopsis at all, although the non-attine ants carried ineffective bacterial strains at twice this frequency.
This great diversity in bacterial strains suggests that they are not primarily passed down vertically from parent colony to daughter colony, but acquired laterally from the environment. The prevalence of a single major strain in most attine colonies suggests positive selection within that colony, but the wide variety of possible bacteria strains suggests this selection varies with local conditions. While Escovopsis is the major pest of attine fungal gardens, it is not the only cultivar pathogen the ants have to combat. The variety of possible cultivar pathogens may encourage variety in domesticated actinomycetes.
The additional presence of actinomycetes on non-attine ants suggests an evolutionary path for recruitment of actinomycete bacteria for suppression of Escovopsis. Ants live in crowded, damp colonies, communicate by touching antenna, and often groom each other. These conditions are ideal for growth and transmission of fungal pathogens. Yet fungal diseases are relatively rare. The presence of actinomycetes with antifungal activity on non-farming ants suggests the growth of these bacteria on the cuticle may be a widespread way of preventing fungal diseases among ants. Among the attine ants this preexisting relationship was modified by natural selection for more active culture of antifungal bacteria.
Future routes of research will include determining if actinomycetes help protect non-attine ants from fungal pathogens and exploring whether attine ants can modify their bacteria load, perhaps by adaptive changes in exocrine gland secretions in response to local conditions.
Kost, C., Lakatos, T., Böttcher, I., Arendholz, W., Redenbach, M., Wirth, R. (2007). Non-specific association between filamentous bacteria and fungus-growing ants. Naturwissenschaften, 94(10), 821-828. DOI: 10.1007/s00114-007-0262-y
Nimravid's Weblog 
3 comments
Comments feed for this article
March 29, 2008 at 7:32 pm
myrmecos
The attine/fungus system is turning out to be an extraordinarily rich vein of research. Thanks for blogging it- I expect we’ll continue to see plenty more, especially with regards to the microbes.
April 7, 2008 at 12:52 pm
Ford
“natural selection for more active culture of antifungal bacteria”
I always wonder how this works. Why aren’t bacteria that use their resources to produce antifungal compounds out-competed by strains that use those resources for their own reproduction, instead, assuming that any benefits to the bacteria from killing Escovopsis are shared (tragedy of the commons)? If “individual ants carried up to three strains”, maybe the bacteria benefit from producing chemicals that kill other bacteria with which they compete directly (on the same ant), with any effects on fungi being a side-effect? This is analogous to a hypothesis we propose to explain why some rhizosphere bacteria make toxins that (happen to?) kill plant pathogens (Kiers and Denison, Annual Review of Ecology, in press). But, under this hypothesis, you might expect the ant bacteria to be as likely to kill the food fungus as the Escovopsis.
October 4, 2008 at 11:12 pm
tony
How do you kill these mother fuckers