Chromosome-End Knockoff Strategy to Reshape Alkaloid Profiles of a Fungal Endophyte

Abstract

Abstract Molecular genetic techniques to precisely eliminate genes in asexual filamentous fungi require the introduction of a marker gene into the target genome. We developed a novel strategy to eliminate genes or gene clusters located in subterminal regions of chromosomes, and then eliminate the marker gene and vector backbone used in the transformation procedure. Because many toxin gene clusters are subterminal, this method is particularly suited to generating nontoxic fungal strains. We tested this technique on Epichloë coenophiala, a seed-transmissible symbiotic fungus (endophyte) of the important forage grass, tall fescue (Lolium arundinaceum). The endophyte is necessary for maximal productivity and sustainability of this grass but can produce ergot alkaloids such as ergovaline, which are toxic to livestock. The genome sequence of E. coenophiala strain e19 revealed two paralogous ergot alkaloid biosynthesis gene clusters, designated EAS1 and EAS2. EAS1 was apparently subterminal, and the lpsB copy in EAS2 had a frame-shift mutation. We designed a vector with a fungal-active hygromycin phosphotransferase gene (hph), an lpsA1 gene fragment for homologous recombination at the telomere-distal end of EAS1, and a telomere repeat array positioned to drive spontaneous loss of hph and other vector sequences, and to stabilize the new chromosome end. We transformed E. coenophiala with this vector, then selected “knockoff” endophyte strains, confirmed by genome sequencing to lack 162 kb of a chromosome end including most of EAS1, and also to lack vector sequences. These ∆EAS1 knockoff strains produced no detectable ergovaline, whereas complementation with functional lpsB restored ergovaline production.