maydis life cycle [5, 6]. Additionally, O-glycosylation may play an important role in the regulation of enzymatic activity,

as has been shown for the Aspergillus awamori Gluco-amylase, which has a Ser/Thr-rich domain that carries several O-linked oligomannose structures necessary for the activity of the enzyme against raw, but not against dissolved, starch [7]. In metazoans, mucin-type O-glycosylation sites are found grouped in clusters in protein regions rich in Ser and Thr residues [8]. AR-13324 proteins containing mucin-like O-glycosylation are often found bound to the plasma membrane constituting the glycocalyx, or in the extracellular medium contributing to the formation of the extracellular matrix or the gel-like mucus in the mucosal

surfaces. Mucins seem to be restricted to metazoans, learn more where they appeared soon in evolution [9], and in silico analysis has been applied to the identification of mucins in animal species with sequenced genomes [9, 10]. To our knowledge, selleckchem a similar approach has never been used in fungi despite the fact that fungal secretory proteins are frequently highly glycosylated and contain Ser/Thr-rich regions predicted to be the site of high density O-glycosylation of the polypeptide chains [11]. Here we have analyzed in silico the presence and distribution of such regions among the putatively secretory proteins coded by the genomes of S. cerevisiae, four plant-pathogenic filamentous

fungi (Botrytis cinerea, Magnaporthe grisea, Sclerotinia sclerotiorum and Ustilago maydis) and three non-pathogenic filamentous fungi (Aspergillus nidulans, Neurospora crassa and Trichoderma reesei). The results show a high frequency of Ser/Thr rich regions in the secretory proteins for all the fungi studied, as well as the prediction of regions highly O-glycosylated for about 25% of them. Results NetOGlyc 3.1 can predict regions with a selleck kinase inhibitor high density of O-glycosylation in fungal proteins Part of the results presented here relies on the prediction of O-glycosylation by the web-based server NetOGlyc 3.1 [12, 13]. This tool consists of a Neural Network trained on mucin-type mammalian O-glycosylation sites (O-N-acetylgalactosamine) and thus has not been designed to predict fungal O-glycosylation sites (mainly O-mannose). In order to check the usefulness of NetOGlyc for fungal proteins, we used all the available fungal proteins with experimentally confirmed O-glycosylation sites that were produced in their natural host, only 30 to our knowledge (Additional file 1), and compared them with the predictions of NetOGlyc for the same group of proteins. NetOGlyc predicted a total of 288 O-glycosylation sites for the whole set, while the number of experimentally-determined O-glycosylation sites was 197. The number of sites predicted by NetOGlyc that were actually found experimentally was 106.