Characterization of the Key Aroma Compounds in Three Truffle Varieties from China by Flavoromics Approach

The volatile compounds of three different fresh-picked truffle varieties (Tuber sinensis, T1, Tuber sinoalbidum, T2 and Tuber sinoexcavatum, T3) were extracted by headspace solid-phase microextraction (HS-SPME). Separation and identification of volatile components and sulfur compounds were investigated by gas chromatography-olfactometry (GC-O), gas chromatography-mass spectrometry (GC–MS) and gas chromatography with flame photometric detection (GC-FPD). The results showed that 44, 43 and 44 volatile compounds were detected in T1, T2 and T3 samples, respectively. In addition, 9, 10 and 9 sulfur compounds were identified in three samples by GC-FPD, respectively. Combining physicochemical and sensory properties, T1 presented fatty, green and rotten cabbage odor; T2 exhibited mushroom, sulfuric and musty odor notes; T3 had nutty, floral and roasted potato odor. Dimethyl sulfide, 3-methylbutanal, dimethyl disulfide, 3-octanone, bis(methylthio) methane, octanal, 1-octen-3-one, 1-octen-3-ol and benzeneacetaldehyde played indispensable roles in the overall aroma of three truffles. Finally, based on quantitative concentration in T1, odorous compounds (OAV) > 1 were mixed to recombine aroma, demonstrating that these key aroma compounds based on OAV can successfully recombine pretty similar aroma of each variety.


Introduction
Truffles (Tuber spp.) are ascomycete symbiotic fungi that strictly depend on other organisms to complete their life cycle [1]. High-quality truffles are mainly produced in limestone topographic areas, such as Alba (Italy), Perigo (France), Yunnan (China) etc. Especially the white Alba truffle (WAT, Tuber pignatum pico), is regarded as the noblest among the truffle species, because it elicits the most intense and pleasant aroma [2,3]. Truffles are very fastidious for the growth environment. As long as sunlight, humidity, soil pH value and the surrounding flora and fauna change slightly, they can't grow. This is why the production of truffles is rare and the price is pretty expensive. Moreover, truffle also has many physiological activities, such as anti-virus, bacteriostasis, anti-inflammation, anti-cancer, anti-oxidation, corresponding to T1, T2 and T3 truffle samples, respectively. Three samples were characterized by comparison with their retention indices (RIs), aroma description and mass spectra with authentic standards. The aroma intensities (AIs) of volatile components ranged from 1.2 to 8.7 for T1, 1.2 to 9.2 for T2, 1.1 to 8.8 for T3.
From Table 1, the three samples differed greatly in sensory olfaction. Dimethyl sulfide and dimethyl disulfide had high AIs in T1 sample, which brought more decayed cabbage odor and sweet smell of popcorn ( Table 1). Both of the compounds were also considered to represent the aroma of black truffles [23]. Eight carbon volatile compounds accounted for large proportion of aroma compounds in T2 samples, including 3-octanone, 2-octanone, 1-octen-3-one, 3-octanol, 1-octen-3-ol. Eight carbon alcohol and ketone mainly provided aroma of mushroom, earthy and herbal. These compounds had been found in most fungi, which gave samples typical mushroom odors [24,25]. Bis(methylthio)methane has always been considered as the most important sulfide in Italian white truffle (Tuber magnatum Pico) research [26] and it has also been found to play an important role in T2 samples. In addition to sulfur compounds, 2-methyl-butanal and 3-methyl-butanal were studied as important aroma compounds in different truffles [27,28], however, T3 samples had much less rough sulfide odor and more focused on 2-methyl-butanal and 3-methyl-butanal, which provided strong nutty and grain aroma. The AI of phenylacetaldehyde was also very high in T3 sample.
The compounds had high OAVs showed that the aroma of the three varieties of Chinese truffles were less strong, pungent sulfurous odor, plenty of sweetness of flowers and fatty notes. This conclusion also had many similarities with other related Chinese truffle aroma studies [15,26]. The unique aroma of Chinese truffles might also be related to the plant environment, soil conditions, such as host tree and complex bacterial colonies in soils [36,37]. GC-O sensory evaluation combined with OAVs could provide a better assessment of key aroma compounds.

Sensory Analysis
After consensus session of sensory evaluation, seven notes were selected to evaluate the aroma of three truffle samples by well-trained panelists, following "mushroom", "nutty and malty", "fatty and green", "floral and sweet", "sulfuric and musty", "roasted potato", "rotten cabbage and corn". ANOVA statistical analysis showed that "mushroom", "nutty and malty', "fatty and green", "floral and sweet", "roasted potato" notes were significantly different in three truffle samples (p < 0.05) (Figure 1) through sensory evaluation scores (Table 4).   PLSR was applied to certify the correlation between the GC-O data and flavor notes by the panelists. The 44 odor-active compounds characterized by GC-O were used as X-matrix, and the 7 flavor notes obtained by sensory evaluation were Y-matrix, which generated the correlation load diagram of PLSR as shown in Figure 2. The two ellipses represent the variance contribution rate of 50% (small) and 100% (large), respectively. The volatiles between the two ellipses had a good correlation with the seven flavor notes. As could be seen from Figure 2, the contribution rate of X variable was 85%, and that of Y variable was 62%, indicating that the PLSR model can further verify the correlation between volatile compounds and flavor notes. The specific analysis was as the following:  PLSR was applied to certify the correlation between the GC-O data and flavor notes by the panelists. The 44 odor-active compounds characterized by GC-O were used as X-matrix, and the 7 flavor notes obtained by sensory evaluation were Y-matrix, which generated the correlation load diagram of PLSR as shown in Figure 2. The two ellipses represent the variance contribution rate of 50% (small) and 100% (large), respectively. The volatiles between the two ellipses had a good correlation with the seven flavor notes. As could be seen from Figure 2, the contribution rate of X variable was 85%, and that of Y variable was 62%, indicating that the PLSR model can further verify the correlation between volatile compounds and flavor notes. The specific analysis was as the following:  Table 1. T1 sample was thought to have much more fatty, green and rotten cabbage, corn aroma than other two samples. From Table1, AIs of hexanal (A5), octanal (A19), nonanal (A22) in T1 were higher than those of T2, T3 and flavor note of these compounds were closely related to the "green, fatty, and waxy"; Dimethyl sulfide and 2-acetylthiazole had high AIs and concentrations in T1, which contributed to the rotten cabbage, corn notes ( Figure 2).

Aroma Recombination
Aroma recombination model was performed to verify the quantitative data and selected key compounds (OAV > 1). Model was evaluated by panelists with original sample T1 according to seven flavor notes. As was shown in Figure 3, "nutty and malty", "roasted potato" of model presented lower score than T1 sample, while "mushroom" note of model was higher than that of T1 sample. From the general flavor profile, model and T1 sample did not have significant differences by seven notes, demonstrating that these key aroma compounds based on OAV could successfully recombine pretty similar aroma.  Table 1. T1 sample was thought to have much more fatty, green and rotten cabbage, corn aroma than other two samples. From Table 1, AIs of hexanal (A5), octanal (A19), nonanal (A22) in T1 were higher than those of T2, T3 and flavor note of these compounds were closely related to the "green, fatty, and waxy"; Dimethyl sulfide and 2-acetylthiazole had high AIs and concentrations in T1, which contributed to the rotten cabbage, corn notes ( Figure 2).

Aroma Recombination
Aroma recombination model was performed to verify the quantitative data and selected key compounds (OAV > 1). Model was evaluated by panelists with original sample T1 according to seven flavor notes. As was shown in Figure 3, "nutty and malty", "roasted potato" of model presented lower score than T1 sample, while "mushroom" note of model was higher than that of T1 sample. From the general flavor profile, model and T1 sample did not have significant differences by seven notes, demonstrating that these key aroma compounds based on OAV could successfully recombine pretty similar aroma.

Materials
Three varieties of ripe truffle namely black truffle Tuber sinensis (T1), white truffle Tuber sinoalbidum (T2) and Tuber sinoexcavatum (T3), were collected from Nanhua County, Yi Autonomous Prefecture, Yunnan Province of China at December 19 th and 20 th , 2018. After collection, they were wrapped in non-woven fabrics and transported to the laboratory with the ice bag within 24h. The fresh truffle samples were washed with Milli-Q water and crushed into truffle purees via JYL-C051 type blender (Joyoung, Shandong, China), and kept in the −18 °C refrigerator for further study.

Solid Phase Microextraction (SPME) Absorption of Aroma Compounds
5.0 g fresh truffle purees were accurately weighed in 20 mL vials, Teflon covers and added 5 μL internal standard solutions (100 mg/L 1, 2-dichlorobenzene or 100 mg/L 2-methyl-3 -tetrahydrofuran thiol for sulfide). Samples were kept at 45 °C in a water bath with 10 min of equilibration time.
A 50/30 μm divinylbenzene-Carboxen-polydimethylsiloxane (DVB-CAR-PDMS) fiber (Supelco, Bellefonte, PA, USA) with a 1 cm length was used. The extraction time was 45 min. Before chemical absorption, the fiber was preconditioned for 30 min on an Agilent 7890 gas chromatograph (Agilent Technologies, Santa Clara, CA, USA) with the injector temperature of 250 °C.

Materials
Three varieties of ripe truffle namely black truffle Tuber sinensis (T1), white truffle Tuber sinoalbidum (T2) and Tuber sinoexcavatum (T3), were collected from Nanhua County, Yi Autonomous Prefecture, Yunnan Province of China at December 19th and 20th, 2018. After collection, they were wrapped in non-woven fabrics and transported to the laboratory with the ice bag within 24h. The fresh truffle samples were washed with Milli-Q water and crushed into truffle purees via JYL-C051 type blender (Joyoung, Shandong, China), and kept in the −18 • C refrigerator for further study.

Solid Phase Microextraction (SPME) Absorption of Aroma Compounds
5.0 g fresh truffle purees were accurately weighed in 20 mL vials, Teflon covers and added 5 µL internal standard solutions (100 mg/L 1, 2-dichlorobenzene or 100 mg/L 2-methyl-3 -tetrahydrofuran thiol for sulfide). Samples were kept at 45 • C in a water bath with 10 min of equilibration time.
A 50/30 µm divinylbenzene-Carboxen-polydimethylsiloxane (DVB-CAR-PDMS) fiber (Supelco, Bellefonte, PA, USA) with a 1 cm length was used. The extraction time was 45 min. Before chemical absorption, the fiber was preconditioned for 30 min on an Agilent 7890 gas chromatograph (Agilent Technologies, Santa Clara, CA, USA) with the injector temperature of 250 • C.

SPME-GC-FID-O Analysis of Truffle
The Agilent 7890A gas chromatograph was used for GC-O analysis. The gas chromatograph was equipped with flame ionization detector (FID) and ODP-2 olfactory port (Gerstel, Mulheim an der Ruhr, Germany). GC effluent was split into 1:1 between the FID and sniffing port. Purified, moist air flowing with odorant eluting were carried to the individual olfactory assessor via an insulated stainless steel tube at 40 mL/min. Samples were conducted using a HP-Innowax and a DB-5 analytical fused silica capillary column (both columns: 60 m × 0.25 mm × 0.25 µm; Agilent Technologies, Santa Clara, CA, USA). GC-FID-O analysis conditions were as the following: the flow rate of carrier gas (nitrogen) was 1.8 mL/min, the oven temperature was held at 50 • C, ramped with a rate of 10 • C/min to 100 • C, and then ramped to 140 • C at a rate of 3 • C/min, finally reached at 200 • C with a rate of 2 • C/min and kept for 10 min. The injection mode was set in splitless for 3 min at 250 • C. The desorption time was 5 min.
The olfactory experiment was performed by 10 trained panelists (six females and four males). Panelists were very sensitive to aroma identification by training olfactory characteristics of reference compounds and truffle sample matrices in sniffing bottles. Aroma characteristics, aroma intensity and frequency of occurrence were written down by the assessors with 50 min of sniffing time. The intensity was calculated as the average of all panelists' scores for an identified aroma. The odor intensities were evaluated on a 10-point intensity scale, where 0 meant a compound had no odor, 5 represented a moderate intensity and 10 stood for an extreme strong odor. Each sample was performed in triplicate by each panelist.

Calibration of Standard Curves
Similar truffle matrix was prepared by adding 2.4 mg/g glucose, 48 mg/g mannitol, 5.2 mg/g malic acid, 2.4 mg/g alanine, 3.7 mg/g glutamate and 1.8 mg/g glycine in Milli-Q deionized water before external standard quantification [38,39]. A recombination containing all volatile compounds was diluted with methanol to 1:5, 1:10, 1:20, 1:50, 1:100 and 1:200 strengths. Then, 5 µL 1,2-dichlorobenzene (100 mg/L) was introduced to the 5 g of model matrix in a 20 mL vial to establish the calibration curves. Equally, 5 µL 2-methyltetrahydrofuran-3-thiol (100 mg/L) was added to establish the calibration curves for sulfur compounds. These mixture models were extracted by HS−SPME. The standard curves, coefficient of determination (r 2 ) and validated linear range for the volatile compounds were set up. All experiments were repeated three times. The calculation formula was as the following: (A x /A i ) equaled peak area of volatile compounds standard/peak area of internal standard; (C x /C i ) represented concentration of volatile compounds standard/concentration of internal standard; a is the slope and b is the intercept on y axis of the standard curve.

SPME-GC-MS of Volatile Compounds in Truffle
The volatile compounds were analyzed by an Agilent 6890 gas chromatography with SPME and a 5975 mass selective detector (MSD) (Agilent Technologies, Santa Clara, CA, USA), HP-Innowax and DB-5 analytical fused silica capillary column (both columns: 60 m × 0.25 mm × 0.25 µm; Agilent, Santa Clara, CA, USA). Conditions for GC-MS analysis were as the following: the injection port was set in a splitless mode, and the desorption time was 5 min and the desorption temperature was 250 • C, the temperature program referred as that of GC-O, the carrier gas was helium with a constant flow rate of 1 mL/min. Chemical identification was performed by MSD. Its electron ionization energy was 70 eV. The temperature of ion source was set at 230 • C. The compounds were identified by matching retention time of authentic standards, retention indices (RIs), and mass spectra in the NIST 11 database. RIs in the literature that matched the column condition (60 m × 0.25 mm × 0.25 µm) and temperature ramp were compared, according to the Van Den Dool and Kratz RI. The RIs of unknown compounds were determined by pure n-alkanes mixture (C 5 −C 30 , Sigma-Aldrich, St. Louis, MO, USA). The calculation formula was as the following: t x represented the retention time of volatile compounds; t z was the retention time of n-alkanes which had same carbon atoms of volatile compounds; Z was the number of carbon atoms of volatile compounds. The GC-MS chromatograms are given in the Supplementary Materials.

SPME-GC-FPD Detection of Sulfur Containing Volatile Compounds in Truffle
The Agilent 7890A gas chromatography equipped with FPD was used to detect sulfur compounds in truffle samples. The oven temperature and heating procedure were consistent with the setting of GC-MS. The FPD temperature was set at 250 • C, the PMT voltage was set at 500 V. The desorption time was 5 min and the injection mode was splitless. The sulfur-containing compounds were identified by retention time of authentic standards and retention index on both columns.

Odor Activity Value (OAV)
By using the formula of olfactory activity value, OAV = C/T, in which OAV represents the olfactory activity value of the flavor compound, C represents the concentration of each compound, T represents the detection threshold in air. Available threshold values were from literature reference. It is generally believed that aromatic compounds with high OAV are most likely to be the key contributors to the overall aroma. OAV > 1 indicates that the compound has a direct impact on the aroma [40].

Sensory Evaluation
According to the guidelines and conditions of ISO 8589-2007, sensory evaluation was carried out in the sensory laboratory. Based on the previous studies [2,41], the method of sensory analysis was generic descriptive analysis. 20 g truffle puree in the 100 mL plastic cup was prepared with a Teflon cover for evaluation. At the beginning, the aroma of truffles was evaluated by a well-trained panel of 10 members (4 males and 6 females). Then through the three preliminary consensus training (each 2 h), eventually the panelists made the final agreement about the aroma description of truffle ("sulfuric, musty", "rotten cabbage, corn", "nutty, malty", "roasted potato", "fatty, green", "mushroom-like" and "floral, sweet" note). Each sensory attribute was defined as following reference compounds: bis(methylthio)methane for "sulfuric, musty" note, dimethyl sulfide for "rotten cabbage, corn" note, 3-methylbutanal for "nutty, malty" note, 3-(methylthio)propanal for "roasted potato" note, octanal for "fatty, green" note, 1-octen-3-ol for "mushroom-like" note, benzeneacetaldehyde for "floral, sweet" note.
A 0-10 linear scale, from 0 (not perceivable) to 3.0, 4.0, 5.0 (moderately perceivable) to 10.0 (very strongly perceivable) were given to the intensities of the respective aroma qualities. The sensory evaluation experiment of each sample was repeated three times to find the average value.

Aroma Recombination of Truffle
Black truffle T1 sample is very popular and has high sales share in China market, so it was chosen to make aroma recombination. Accordingly, in order to confirm high OAV compounds play important roles in T1 s aroma, a total of 21 (OAV ≥ 1) volatile compounds were added in an aqueous solution aroma model at their actual quantitative concentration as determined in T1. The sensory panelists scored the aroma recombination model and truffle samples through seven sensory notes discussed in generic descriptive analysis. Each sample was evaluated in triplicate by each member.

Statistical Analysis
In this study, relative standard deviation (RSD) well reflected the precision of GC-MS and GC-O test data. Aroma intensity of GC-O and quantitative of volatile compounds were performed by analysis of variance (ANOVA). When there were significant differences between samples, Duncan's multiple range tests were used at the level of significance set at p < 0.05. Both ANOVA and Duncan's multiple range tests were conducted by Predictive Analytics Software (PASW) Statistics 18 (IBM, Chicago, IL, USA).
Partial least squares regression (PLSR) analysis was carried out by the Unscrambler 9.8 (CAMO ASA, Oslo, Norway). Odor-active compounds detected with GC-O as X-variables and the odor attributes from the descriptive profile as Y-variables. The correlation between GC-O data and sensory attributes was illustrated by PLS2. All regression models were validated using full cross-validation.

Conclusions
In summary, the volatile constituents of truffles from three different varieties from Yunnan Province of China were studied by flavoromics approach through SPME extraction combined with GC-O, GC-MS, GC-FPD and aroma recombination. 44, 43 and 44 volatile compounds were detected in T1, T2 and T3 samples, respectively. Among them, 9, 10, and 9 sulfur-containing compounds were authenticated. Dimethyl sulfide, 3-methylbutanal, 2-methylbutanol, 3-octanone, 1-octen-3-one, 3-octanol and1-octen-3-ol showed higher level in external standard quantitative analysis. Finally, 24 key aroma-active compounds screened out by GC-O and OAV were considered to make important contribution to the overall aroma of three truffles.
Although there were some differences in the aroma of the three Chinese truffles, in general, the aroma of Chinese truffles was more floral, mushroom and sweet, which was also the unique aroma of Chinese Yunnan truffles. Through a flavoromics study on the fragrance of domestic truffles, further development of truffle products with Chinese characteristics, and transformation of the Chinese truffle market from original excavation to intensive processing industry.