Summary:

Fungus-growing termites (subfamily Macrotermitinae, family Termitidae) live in an obligate mutualistic symbiosis with the fungus Termitomyces. All other termites rely on gut microbes for the breakdown of plant material and other forage, and it has been generally assumed that the association with Termitomyces has reduced the need for fermentative gut microbes after the Macrotermitinae became fungus-farmers. Only few studies have explored this in any detail and the identities, levels of interaction-specificity with the termite host, and consistency in bacterial communities between host species have remained largely unknown. Here, we employ bacterial 16S rRNA 454 high-throughput pyrosequencing to identify a potential core microbiome in the fungus-growing termites - i.e., a distinct set of bacteria present across lineages in the termite phylogeny. Comparative analysis of 9 fungus-growing termite species from 5 genera suggests that a core gut microbiome indeed exists, as all bacterial taxa of high abundance were present in all termite species examined. However, quantitative differences in microbiome composition between termite species and genera were also noticed, possibly associated with differences in substrate use and Termitomyces lineage reared. Our results are consistent with major changes in gut microbiomes having occurred when fungus farming evolved 30 MYA, followed by relatively modest elaborations in response to ecological conditions. This might help explain why neither the termites nor Termitomyces ever abandoned the symbiosis or teamed-up with another termite of fungal partner lineage.

Summary:

Multicellularity is a highly cooperative state prone to invasion by cheating genotypes that use the resources provided by the multicellular organism without contributing their fair share to non-reproductive functions of the organism (e.g. cancers). Kin selection, often realised through regular single -celled bottlenecks (and in some organisms by an early germline separation), is a solution to prevent selection for cheating. In fungi, the lack of an early germline separation and the potential to fuse with other individuals make cheating a realistic threat. However a genetic allorecognition mechanism that limits fusion to almost only clonally related individuals, seems to effectively protect fungi against cheating genotypes. In order to test the hypothesis that cheating is a realistic threat to multicellular growth in fungi, we used an experimental evolution approach with Neurospora crassa, that maximised the potential for cheating genotypes by selecting under low relatedness and completely local competition (i.e. under a high inoculation density of spores, in the absence of genetic allorecognition). Within less than 300 generations all eight replicate lines contained genotypes that matched our criteria for cheating: they had increased relative fitness (measured as proportion of spores produced) when in competition with a cooperative ancestral type, but spore production in monoculture was significantly decreased. So there is a clear trade-off between competitive fitness and production of asexual spores when grown alone. Contrary to predictions about the evolution of social behaviour that cheating genotypes will completely eradicate the social behaviour (the tragedy of the commons), we found a stable polymorphism in all evolved lines: a relatively cooperative type producing many spores when grown in monoculture, and the cheating type described above. We are currently studying the conditions leading to this apparently balanced polymorphism in our evolving lines.

Summary:

Multicellularity is a highly cooperative state prone to invasion by cheating genotypes that use the resources provided by the multicellular organism without contributing their fair share to non-reproductive functions of the organism (e.g. cancers). Kin selection, often realised through regular single-celled bottlenecks (and in some organisms by an early germline separation), is a solution to protect against cheating. In fungi, the lack of an early germline separation and the potential to fuse with other individuals make cheating a realistic threat. However a genetic allorecognition mechanism that limits fusion to almost only clonally related individuals, seems to effectively protect fungi against cheating genotypes. We have shown earlier that in the absence of cheating genotypes, the fitness advantage of fusion selects against allotype diversity in the fungus Neurospora crassa. Individuals with more common allotypes have a higher fitness because they fuse more frequently and gain a larger average size. Studies that model evolution of allorecognition in fungi show that cheating can cause a stable polymorphism for genetic allorecognition loci. Using cheating genotypes generated during an experimental evolution experiment, we empirically test the hypothesis that genetic allorecognition in fungi can be stabilized by the presence of cheating.