Sequence capture identifies fastidious chytrid fungi directly from host tissue.

• Target enrichment can directly sequence host-associated fungal pathogens. • Fungal pathogen load is positively associated with enrichment success. • Resulting genetic data can identify phylogenetic placements of fastidious strains. • Salamander at US zoo was infected with Asian-origin Batrachochytrium strain. • Method can be applied to other unculturable fungal samples to identify cryptic strains. The chytrid fungus Batrachochytrium dendrobatidis (Bd) was discovered in 1998 as the cause of chytridiomycosis, an emerging infectious disease causing mass declines in amphibian populations worldwide. The rapid population declines of the 1970s-1990s were likely caused by the spread of a highly virulent lineage belonging to the Bd-GPL clade that was introduced to naïve susceptible populations. Multiple genetically distinct and regional lineages of Bd have since been isolated and sequenced, greatly expanding the known biological diversity within this fungal pathogen. To date, most Bd research has been restricted to the limited number of samples that could be isolated using culturing techniques, potentially causing a selection bias for strains that can grow on media and missing other unculturable or fastidious strains that are also present on amphibians. We thus attempted to characterize potentially non-culturable genetic lineages of Bd from distinct amphibian taxa using sequence capture technology on DNA extracted from host tissue and swabs. We focused our efforts on host taxa from two different regions that likely harbored distinct Bd clades: (1) wild-caught leopard frogs (Rana) from North America, and (2) a Japanese Giant Salamander (Andrias japonicus) at the Smithsonian Institution's National Zoological Park that exhibited signs of disease and tested positive for Bd using qPCR, but multiple attempts failed to isolate and culture the strain for physiological and genetic characterization. We successfully enriched for and sequenced thousands of fungal genes from both host clades, and Bd load was positively associated with number of recovered Bd sequences. Phylogenetic reconstruction placed all the Rana -derived strains in the Bd-GPL clade. In contrast, the A. japonicus strain fell within the Bd-Asia3 clade, expanding the range of this clade and generating additional genomic data to confirm its placement. The retrieved ITS locus matched public barcoding data from wild A. japonicus and Bd infections found on other amphibians in India and China, suggesting that this uncultured clade is widespread across Asia. Our study underscores the importance of recognizing and characterizing the hidden diversity of fastidious strains in order to reconstruct the spatiotemporal and evolutionary history of Bd. The success of the sequence capture approach highlights the utility of directly sequencing pathogen DNA from host tissue to characterize cryptic diversity that is missed by culture-reliant approaches. [ABSTRACT FROM AUTHOR]