Can ITS sequence data identify fungal endophytes from cultures? A case study from Rhizophora apiculata

Abstract

© Guizhou Academy of Agricultural Sciences. Culture-based studies have recovered fungal endophytes from numerous plant hosts, while direct examination of sporulating cultures has enabled identification. However, many endophytes cannot be identified due to the fact that they only form mycelia sterilia in culture. Although next generation sequencing (NGS), as well as ITS sequence analyses have been used to identify endophytes, identification is still rudimentary. In this study, we isolated fungal endophytes from Rhizophora apiculata in Thailand and established how many can be identified to species level based on ITS sequence data. Endophytic fungi were isolated from leaves, petioles and aerial roots of R. apiculata in four provinces of Thailand. One hundred and fifty four isolates were obtained and initially grouped into 20 morphotypes based on cultural characteristics. Nine were sporulating morphotypes, which were assigned to seven genera (Colletotrichum, Diaporthe, Hypoxylon, Neopestalotiopsis, Neodevriesia, Pestalotiopsis and Phyllosticta), and eleven morphotypes were non-sporulating mycelia sterilia. Sequence similarity comparison and phylogenetic analysis of the ITS regions were further used to identify taxa. While ITS sequence data is reliable to assign isolates at the generic rank, and can be useful to identify taxa to species level in a small number of fungal genera, it cannot generally be used to determine specific species in most genera. ITS analysis classified 30 representative isolates into 20 taxonomic units residing in 15 known genera: Allophoma sp., Colletotrichum spp., Diaporthe spp., Hortaea werneckii, Hypoxylon griseobrunneum, Hypoxylon sp., Pestalotiopsis sp., Phanerochaete sp., Phyllosticta spp., Pseudopithomyces maydicus, Preussia sp., Nemania sp., Neodevriesia sp., Neopestalotiopsis sp., Rigidoporus vinctus, Schizophyllum sp. and one unidentified genus. Of the morphotypes, four were identified to species. The results from Blast searches and ITS phylogeny of the 15 genera and the one unidentified genus are discussed. Twenty-five of the 30 isolates could not be identified and thus an estimated 20 isolates are likely to be new species and one a new genus. This is remarkable, as endophytes of a single host in Thailand, may yield 75% or more of new species. This is not akin to the OTUs generated in NGS platforms, which are generally short sequences with high possibility to not match with sequences in GenBank, not because they are new, but because of the quality of the sequence data.