Fresh ascocarps of two truffle species Tuber Aestivum (eight samples) and Tuber Borchii (eleven samples) were harvested from natural truffle zones in Greece (Prefectures of Ioannina, Achaia, Evoia, and Chalkidiki) during April and July 2021, when truffles mature in this region, and were kindly provided by local suppliers between May and June 2021. As it is known that the expression of vioaltile species is also depending on the maturity of the truffles (23), they were investigated in comparable states of maturity. The truffles were rinsed with tap water and left to dry in an air hood according to Culleré et al. (6). They were stored at −20°C until analysis.

The determination of the volatile compounds in truffle samples was conducted using an Agilent 6890 N Gas Chromatograph coupled to an Agilent 5973 K Quadrupole Mass Spectrometric Detector (Palo Alto, CA). The separation of the compounds was performed on a DB-WAX capillary column (60 m × 0.32 mm, 0.25 μm film thickness, Agilent Technologies, Santa Clara, CA) using helium (99.999%) as mobile phase delivered at a flow rate of 1.2 ml min⁻¹. The initial oven temperature was 40°C and it was held constant for 5 min. Then the temperature was raised to 240°C (rate 5°C min⁻¹) and it was held constant for 5 min. Using this oven program, the separation of the volatile compounds was completed within 50 min. Splitless injection was performed. The following temperatures were adopted: injector temperature; 270°C, MS source; 250°C and MS Quad; 130°C. The compounds were determined in scan mode recording ions with an m/z ratio of 35–350.

An 85 μm Carboxen/Polydimethylsiloxane (CAR/PDMS) fiber attached to a manual SPME fiber holder (Supelco, Bellefonte, PA) was used for the HS-SPME procedure. Conditioning of the fiber was conducted prior to the analysis based on the instructions of the manufacturer. For the extraction of the volatile compounds, an aliquot of truffle (100 mg) was weighted into a 15 ml-glass vial. Extraction of the volatile compounds took place at 50°C within 45 min. After this time span, the fiber was removed from the vial containing the sample and the analytes were desorbed in the injector of the GC-MS instrument for 5 min which previously had been shown to be sufficient for complete desorption, and for the avoidance of sample carry-over. Each experiment was performed in triplicate.

Prior to the analysis of the truffle samples, optimization of different parameters that could potentially affect the performance of the HS-SPME method was undertaken. For this purpose, different sample masses (100–500 mg), extraction time spans (15–45 min) and extraction temperatures (30–50°C) were investigated using bulk samples. During optimization the signals of representative compounds belonging to different compound classes (i.e, 2-butanol, 2-butanone, 4-hydroxy-3-hexanone and 1-octen-3-ol) were monitored (1, 5, 7). The peak areas obtained under the individual conditions were normalized with regard to the peak area obtained for each compound under the optimum/selected conditions, and were expressed as percentage.

Clustering analysis was employed to divide the samples into clusters according to their similarity criteria. Agglomerative Hierarchical Clustering (HCA) was used to build a hierarchy of clusters, group the samples in different clusters, and visualize the clusters of the samples (24). Partial least squares discriminant analysis (PLS-DA) is a supervised pattern recognition technique used to find the appropriate class for each sample (20). In PLS-DA a mathematical model is built to establish a correlation and classify the samples, knowing the label of each class (25). A PLS-DA prediction model was developed using the MetaboAnalyst 5.0 platform (26) to discover patterns in the chromatographic data of the truffles species, establishing the most important volatile markers responsible for their discrimination. The data table was normalized by dividing the peak area of each individual peak with the sum of all the peak areas of each chromatogram before exporting to HCA and PLS-DA. Pareto scaling was used and the variables were mean-centered (21, 22).

The optimum extraction parameters were adopted in the subsequent analysis of Tuber Melanosporum and Tuber Aestivum samples. Overall, 45 compounds were identified using the NIST library (Version 2.0 g, 2011) on the basis of the agreement of the mass spectra as well as their respective retention indices (RI). The peak areas of the identified compounds were normalized after assuming that the sum of the peak areas of all the identified compounds in each sample is equal to 100%. Table 1 presents a list of the identified compounds, their corresponding retention times (RTs), retention index (RI) and their normalized peak areas presented as percentages.

The characteristic volatiles that are responsible for the unique aroma of the truffles belonged mainly to the classes of aldehydes, alcohols, and ketones (32). According to Table 1, 3-methyl-butanal (2–36%) was the most abundant compound and is responsible, according to the descriptions in the literature, for the “sulfurous” and “animal” odor, and has been reported as a characteristic volatile of black truffle species and characteristic marker of truffle degradation and product spoilage produced by axenic cultures of truffle mycelium, fungal and bacterial phyla (5, 8). The second most abundant compound to be identified in Tuber Aestivum was 2-methyl-1-butanol at a percentage of up to 32%. It is a naturally occurring alcohol responsible for the ethereal type odor that has already been determined in black truffles (5). Furthermore, 2-butanone was identified and determined within the range between 7 and 30% which has been previously reported to vary among Tuber Aestivum ascorcaps (8). Octen-3-ol, imparting “mushroom, earthy herbal, woody, green” odor notes (4, 33), was determined over the range 5–22%. Its presence in black truffle species has been associated with the earthy and dusty aroma of the product. In addition, 3-octanone (<11.2%) and ethanol (2–7%) that are responsible for “mushroom/herbal” and “alcoholic” notes (34), respectively, and their high abundance in black truffle species has already been reported (5, 8, 35) were also detected.

In Tuber Borchii, the most abundant compound identified was 2-methyl-butanoic acid within the range 33–53%, and this has already been reported to contribute to truffle aroma (18, 36). 2-methyl-butanoic acid is responsible for the pervasive, cheesy, sweaty odor. The percentages of 2-methyl-1-propanol and 2-methyl-2-butenal were also high up to 22%, for both compounds. 2-methyl-1-propanol is associated with descriptors like “sweet” and “musty”, and 2-methyl-butenal is responsible for the “green” and “fruity” odor. The presence of 2-methyl-1-propanol and 2-methyl-2-butenal in Tuber Borchii has been previously reported by D'Auria et al. (37), as well.

As regards the rest of the identified compounds, 2-methyl-1-butanol and 2-butanol, have been identified in black truffle species (5). The presence of dodecanol, phenylethanol, 2-methyl-butanoic acid, 2-octen-1-ol, 1-octanol, benzaldehyde, 1-hexanol, heptanal, hexanal, and 2,3-butanedione, in black truffles has been reported by Choo et al. (4). Nonanal, decanal, 2-nonenal, 2-methyl furan, 2-octenal, 2-ethyl-1-hexanol, 6-methyl-5-hepten-2-one, 3-octanone, and 3-octanol have been identified in the white truffle species Tuber Magnatum Pico originating from Italy, as well (18, 36).