The microbiome can significantly impact host phenotypes and serve as an additional source of heritable genetic variation. While patterns across eukaryotes are consistent with a role for symbiotic microbes in host macroevolution, few studies have examined symbiont-driven host evolution or the ecological implications of a dynamic microbiome across temporal, spatial or ecological scales. The pea aphid, Acyrthosiphon pisum, and its eight heritable bacterial endosymbionts have served as a model for studies on symbiosis and its potential contributions to host ecology and evolution. But we know little about the natural dynamics or ecological impacts of the heritable microbiome of this cosmopolitan insect pest. Here we report seasonal shifts in the frequencies of heritable defensive bacteria from natural pea aphid populations across two host races and geographic regions. Microbiome dynamics were consistent with symbiont responses to host-level selection and findings from one population suggested symbiont-driven adaptation to seasonally changing parasitoid pressures. Conversely, symbiont levels were negatively correlated with enemy-driven mortality when measured across host races, suggesting important ecological impacts of host race microbiome divergence. Rapid drops in symbiont frequencies following seasonal peaks suggest microbiome instability in several populations, with potentially large costs of 'superinfection' under certain environmental conditions. In summary, the realization of several laboratory-derived, a priori expectations suggests important natural impacts of defensive symbionts in host-enemy eco-evolutionary feedbacks. Yet negative findings and unanticipated correlations suggest complexities within this system may limit or obscure symbiont-driven contemporary evolution, a finding of broad significance given the widespread nature of defensive microbes across plants and animals.

Most asexual species of fungi have either lost sexuality recently, or they experience recombination by cryptic sexual reproduction. Verticillium dahliae is a plant-pathogenic, ascomycete fungus with no known sexual stage, even though related genera have well-described sexual reproduction. V. dahliae reproduces mitotically and its population structure is highly clonal. However, previously described discrepancies in phylogenetic relationships among clonal lineages may be explained more parsimoniously by recombination than mutation; therefore, we looked for evidence of recombination within and between clonal lineages. Genotyping by sequencing was performed on 141 V. dahliae isolates from diverse geographic and host origins, resulting in 26,748 single-nucleotide polymorphisms (SNPs). We found a strongly clonal population structure with the same lineages as described previously by vegetative compatibility groups (VCGs) and molecular markers. We detected 443 recombination events, evenly distributed throughout the genome. Most recombination events detected were between clonal lineages, with relatively few recombinant haplotypes detected within lineages. The only three isolates with mating type MAT1-1 had recombinant SNP haplotypes; all other isolates had mating type MAT1-2. We found homologs of eight meiosis-specific genes in the V. dahliae genome, all with conserved or partially conserved protein domains. The extent of recombination and molecular signs of sex in (mating-type and meiosis-specific genes) suggest that V. dahliae clonal lineages arose by recombination, even though the current population structure is markedly clonal. Moreover, the detection of new lineages may be evidence that sexual reproduction has occurred recently and may potentially occur under some circumstances. We speculate that the current clonal population structure, despite the sexual origin of lineages, has arisen, in part, as a consequence of agriculture and selection for adaptation to agricultural cropping systems.

Fungi are prolific producers of natural products, compounds which have had a large societal impact as pharmaceuticals, mycotoxins, and agrochemicals. Despite the availability of over 1,000 fungal genomes and several decades of compound discovery efforts from fungi, the biosynthetic gene clusters (BGCs) encoded by these genomes and the associated chemical space have yet to be analyzed systematically. Here, we provide detailed annotation and analyses of fungal biosynthetic and chemical space to enable genome mining and discovery of fungal natural products. Using 1,037 genomes from species across the fungal kingdom (e.g., Ascomycota, Basidiomycota, and non-Dikarya taxa), 36,399 predicted BGCs were organized into a network of 12,067 gene cluster families (GCFs). Anchoring these GCFs with reference BGCs enabled automated annotation of 2,026 BGCs with predicted metabolite scaffolds. We performed parallel analyses of the chemical repertoire of fungi, organizing 15,213 fungal compounds into 2,945 molecular families (MFs). The taxonomic landscape of fungal GCFs is largely species specific, though select families such as the equisetin GCF are present across vast phylogenetic distances with parallel diversifications in the GCF and MF. We compare these fungal datasets with a set of 5,453 bacterial genomes and their BGCs and 9,382 bacterial compounds, revealing dramatic differences between bacterial and fungal biosynthetic logic and chemical space. These genomics and cheminformatics analyses reveal the large extent to which fungal and bacterial sources represent distinct compound reservoirs. With a >10-fold increase in the number of interpreted strains and annotated BGCs, this work better regularizes the biosynthetic potential of fungi for rational compound discovery.

The apparent rarity of sex in many fungal species has raised questions about how much sex is needed to purge deleterious mutations and how differences in frequency of sex impact fungal evolution. We sought to determine how differences in the extent of recombination between populations of Aspergillus flavus impact the evolution of genes associated with the synthesis of aflatoxin, a notoriously potent carcinogen. We sequenced the genomes of, and quantified aflatoxin production in, 94 isolates of A. flavus sampled from seven states in eastern and central latitudinal transects of the United States. The overall population is subdivided into three genetically differentiated populations (A, B, and C) that differ greatly in their extent of recombination, diversity, and aflatoxin-producing ability. Estimates of the number of recombination events and linkage disequilibrium decay suggest relatively frequent sex only in population A. Population B is sympatric with population A but produces significantly less aflatoxin and is the only population where the inability of nonaflatoxigenic isolates to produce aflatoxin was explained by multiple gene deletions. Population expansion evident in population B suggests a recent introduction or range expansion. Population C is largely nonaflatoxigenic and restricted mainly to northern sampling locations through restricted migration and/or selection. Despite differences in the number and type of mutations in the aflatoxin gene cluster, codon optimization and site frequency differences in synonymous and nonsynonymous mutations suggest that low levels of recombination in some A. flavus populations are sufficient to purge deleterious mutations. IMPORTANCE Differences in the relative frequencies of sexual and asexual reproduction have profound implications for the accumulation of deleterious mutations (Muller’s ratchet), but little is known about how these differences impact the evolution of ecologically important phenotypes. Aspergillus flavus is the main producer of aflatoxin, a notoriously potent carcinogen that often contaminates food. We investigated if differences in the levels of production of aflatoxin by A. flavus could be explained by the accumulation of deleterious mutations due to a lack of recombination. Despite differences in the extent of recombination, variation in aflatoxin production is better explained by the demography and history of specific populations and may suggest important differences in the ecological roles of aflatoxin among populations. Furthermore, the association of aflatoxin production and populations provides a means of predicting the risk of aflatoxin contamination by determining the frequencies of isolates from low- and high-production populations.