Recombination, balancing selection and adaptive evolution in the aflatoxin gene cluster of Aspergillus parasiticus

Abstract

Aflatoxins are toxic and carcinogenic polyketides produced by several Aspergillus species that are known to contaminate agricultural commodities, posing a serious threat to animal and human health. Aflatoxin (AF) biosynthesis is almost fully characterized and involves the coordinated expression of approximately 25 genes clustered in a 70-kb DNA region. Aspergillus parasiticus is an economically important and common agent of AF contamination. Naturally occurring nonaflatoxigenic strains of A. parasiticus are rarely found and generally produce O-methylsterigmatocystin (OMST), the immediate precursor of AF. To elucidate the evolutionary forces acting to retain AF and OMST pathway extrolites (chemotypes), we sequenced 21 intergenic regions spanning the entire cluster in 24 A. parasiticus isolates chosen to represent the genetic diversity within a single Georgia field population. Linkage disequilibrium analyses revealed five distinct recombination blocks in the A. parasiticus cluster. Phylogenetic network analyses showed a history of recombination between chemotype-specific haplotypes, as well as evidence of contemporary recombination. We performed coalescent simulations of variation in recombination blocks and found an approximately twofold deeper coalescence for cluster genealogies compared to noncluster genealogies, our internal standard of neutral evolution. Significantly deeper cluster genealogies are indicative of balancing selection in the AF cluster of A. parasiticus and are further corroborated by the existence of trans-species polymorphisms and common haplotypes in the cluster for several closely related species. Estimates of Ka/Ks for representative cluster genes provide evidence of selection for OMST and AF chemotypes, and indicate a possible role of chemotypes in ecological adaptation and speciation.

Global population structure and adaptive evolution of aflatoxin‐producing fungi

“…Other haplotypes contain isolates that associate with Geiser's Group I (H10 and H39) and Group II (H78). There is also evidence of trans‐speciation in aflM/aflN , which has been reported previously (Carbone et al., 2007). For example, in haplotype H39, an A. parasiticus isolate from Argentina groups with global A. flavus isolates.…”

“…First, we see widespread evidence of balancing selection acting on G 1 ‐ and B 1 ‐dominant A. parasiticus chemotypes in both the aflM/aflN and aflW/aflX regions. Previously this was detected only in the aflM/aflN (hypE) locus for the Georgia A. parasiticus population (Carbone et al., 2007). Similarly, there is global evidence for Lineage IB (predominantly nonaflatoxigenic) in A. flavus that was first inferred from this same Georgia field (Horn & Greene, 1995; Moore et al., 2009).…”

“…Seven distinct genetic clusters were more strongly structured by G chemotype diversity than geography. Although there can be other factors that underlie these seven clusters, chemotype‐specific evolutionary lineages were observed in previous studies (Carbone et al., 2007; Moore et al., 2009). Whether these chemotype‐specific lineages have increased pathogenicity or enhanced fitness for some other trait or ecological condition is unknown and warrants further study.…”

“…Other evolutionary mechanisms that may result in a signature of trans‐speciation include hybridization and introgression, or incomplete lineage sorting, at specific loci (Sun, Corcoran et al., 2012). Trans‐speciation has been observed for chemotype‐specific alleles in A. parasiticus (Carbone et al., 2007) and A. flavus (Moore et al., 2009). In the aflatoxin gene cluster, selection may act to maintain the nonaflatoxigenic phenotype in A. flavus , and G 1 or B 1 dominance in A. parasiticus (Carbone et al., 2007; Moore et al., 2009).…”

“…Trans‐speciation has been observed for chemotype‐specific alleles in A. parasiticus (Carbone et al., 2007) and A. flavus (Moore et al., 2009). In the aflatoxin gene cluster, selection may act to maintain the nonaflatoxigenic phenotype in A. flavus , and G 1 or B 1 dominance in A. parasiticus (Carbone et al., 2007; Moore et al., 2009). Because balancing selection could be acting on chemotype or morphological differences (Moore et al., 2009; Nei, 2007), species delimitation using cluster regions alone is tenuous, and a more holistic approach is warranted (Samson & Varga, 2009).…”