Chemosystematics of Penicillium based on profiles of volatile metabolites

Abstract

A total of 132 isolates of 25 different terverticillate Penicillium taxa have been investigated for the production of volatile metabolites on yeast extract sucrose agar (SYES) under controlled conditions. Fungal volatiles were collected from Petri dishes by diffusive sampling onto tubes with Tenax TA between day 4 and day 11 of incubation, and were analysed by gas chromatography (GC) with flame ionization detection (FID) after thermal desorption (TD). Identity of volatile metabolites from some taxa was confirmed by TD-GC mass spectrometry (MS). Relative amounts of detected volatiles collected from each of the 132 isolates have been treated by multivariate statistical methods. Thus, the 132 × 131 matrix was analysed by average linking clustering (UPGMA) with different distance coefficients. A clear separation was obtained for practically all taxa investigated, especially when using the Jaccard distance coefficient on the binary type matrix. Despite the fact that volatile metabolites were only collected from cultivation on one medium, our results perfectly agree with previous classifications of Penicillium , based on chemotaxonomy of biosynthetic families of non-volatile secondary metabolites. All taxa previously described as varieties, for example P. roqueforti var. carneum and P. hirsutum var. venetum , clustered independently, supporting a separate taxonomic status. The present study is the first to demonstrate that fungal volatile metabolites can be used in detection and classification of fungi at the species level, using more than one isolate of each taxon.