A mixed community of actinomycetes produce multiple antibiotics for the fungus farming ant Acromyrmex octospinosus

Abstract

Background: Attine ants live in an intensely studied tripartite mutualism with the fungus Leucoagaricus gongylophorus, which provides food to the ants, and with antibiotic-producing actinomycete bacteria. One hypothesis suggests that bacteria from the genus Pseudonocardia are the sole, co-evolved mutualists of attine ants and are transmitted vertically by the queens. A recent study identified a Pseudonocardia-produced antifungal, named dentigerumycin, associated with the lower attine Apterostigma dentigerum consistent with the idea that co-evolved Pseudonocardia make novel antibiotics. An alternative possibility is that attine ants sample actinomycete bacteria from the soil, selecting and maintaining those species that make useful antibiotics. Consistent with this idea, a Streptomyces species associated with the higher attine Acromyrmex octospinosus was recently shown to produce the well-known antifungal candicidin. Candicidin production is widespread in environmental isolates of Streptomyces, so this could either be an environmental contaminant or evidence of recruitment of useful actinomycetes from the environment. It should be noted that the two possibilities for actinomycete acquisition are not necessarily mutually exclusive.

Results: In order to test these possibilities we isolated bacteria from a geographically distinct population of A. octospinosus and identified a candicidin-producing Streptomyces species, which suggests that they are common mutualists of attine ants, most probably recruited from the environment. We also identified a Pseudonocardia species in the same ant colony that produces an unusual polyene antifungal, providing evidence for co-evolution of Pseudonocardia with A. octospinosus.

Conclusions: Our results show that a combination of co-evolution and environmental sampling results in the diversity of actinomycete symbionts and antibiotics associated with attine ants.

Figure 1

Actinomycete species isolated from attine ants. Actinomycete species isolated from Acromyrmex octospinosus worker ants viewed under a light microscope at 40 × magnification. Streptomyces strains are numbered S1-S9 and Pseudonocardia strains P1-P2.

Figure 2

Antifungal bioassays against Escovopsis. Bioassays against the fungal garden parasite Escovopsis weberi. The actinomycete strains S3, S4, S5, S9 and P1 formed clear inhibition zones while the control strain, Streptomyes lividans, produced no zone of inhibition and was overgrown by the nest parasite.

Figure 3

Antifungal bioassays against Candida. Bioassays against the human pathogen Candida albicans. S4, S5 and, to a lesser extent, P1 all inhibit the growth of C. albicans whereas the control strain Streptomyes lividans is overgrown.

Figure 4

Identification of nystatin P1. Identification of a nystatin-like compound by liquid chromatography-tandem mass spectrometry. For each panel the lower tier corresponds to the nystatin standard and the upper tier to the Pseudonocardia P1 extract. (a) Extracted ion chromatograms for m/z 926.5 (nystatin standard) and m/z 1088.6 (nystatin P1). (b) Mass spectra averaged across the retention times indicated in panel A. (c) MS² analysis of the molecular ions identified above. The main mode of fragmentation is the loss of water molecules (m/z 18). (d) Enlarged region of the MS² spectra. These product ions arise from loss of the carbohydrate portion plus one to seven water molecules and most are found in the nystatin standard as well as in nystatin P1. (e) Ultraviolet spectra extracted at the retention times indicated in panel A.