Oyster mushrooms are an interesting source of biologically active glucans and other polysaccharides. This work is devoted to the isolation and structural characterization of polysaccharides from basidiocarps of the cultivated oyster mushroom, Pleurotus ostreatus. Five polysaccharidic fractions were obtained by subsequent extraction with cold water, hot water and two subsequent extractions with 1 m sodium hydroxide. Branched partially methoxylated mannogalactan and slightly branched (1→6)-β-d-glucan predominated in cold- and hot-water-soluble fractions, respectively. Alternatively, these polysaccharides were obtained by only hot water extraction and subsequent two-stage chromatographic separation. The alkali-soluble parts originating from the first alkali extraction were then fractionated by dissolution in dimethyl sulfoxide (DMSO). The polysaccharide insoluble in DMSO was identified as linear (1→3)-α-d-glucan, while branched (1→3)(1→6)-β-d-glucans were found to be soluble in DMSO. The second alkaline extract contained the mentioned branched β-d-glucan together with some proteins. Finally, the alkali insoluble part was a cell wall complex of chitin and β-d-glucans.

Fruiting bodies (basidiocarps) of the cultivated mushroom Pleurotus ostreatus (16 strains) were characterized by vibration spectroscopy and chemometrics. According to organic elemental analysis and Megazyme assay, the basidiocarps contained ~6.2–17.5% protein and ~18.8–58.2% total glucans. The neutral sugar analysis confirmed that glucose predominated in all the samples (~71.3–94.4 mol%). Fourier-transformed (FT) mid- and near-infrared (FT MIR, FT NIR) and FT Raman spectra of the basidiocarps were recorded, and the characteristic bands of proteins, glucans and chitin were assigned. The samples were discriminated based on principal component analysis (PCA) of the spectroscopic data in terms of biopolymeric composition. The partial least squares regression (PLSR) models based on first derivatives of the vibration spectra were obtained for the prediction of the macromolecular components, and the regression coefficients R2 and root mean square errors (RMSE) were calculated for the calibration (cal) of proteins (R2cal 0.981–0.994, RMSEcal ~0.3–0.5) and total glucans (R2cal 0.908–0.996, RMSEcal ~0.6–3.0). According to cross-validation (CV) diagnosis, the protein models were more precise and accurate (R2cv 0.901–0.970, RMSEcv ~0.6–1.1) than the corresponding total glucan models (R2cv 0.370–0.804, RMSEcv ~4.7–8.5) because of the wide structural diversity of these polysaccharides. Otherwise, the Raman band of phenylalanine ring breathing vibration at 1004 cm−1 was used for direct quantification of proteins in P. ostreatus basidiocarps (R ~0.953). This study showed that the combination of vibration spectroscopy with chemometrics is a powerful tool for the evaluation of culinary and medicinal mushrooms, and this approach can be proposed as an alternative to common analytical methods.