Abstract
This study examined the volatile organic compounds (VOCs) in Stropharia rugosoannulata Farlow that were cultivated on four substrates formulated with agricultural and forestry wastes. The VOCs were analyzed by an electronic nose (E-nose), gas chromatography-ion mobility spectroscopy (GC-IMS), principal component analysis (PCA), and an orthogonal partial least squares discriminant analysis (OPLS-DA). A4(40% sawdust, 30% camellia shells, 20% rice husk, 8% bran, and 2% lime) was the most effective overall at determining the quality of flavor. The E-nose showed that there were similar profiles of aromas for A2(100% Eleusine coracana (L.) Gaertn straw) and A3(70% bamboo chips, 20% rice husk, 8% bran, and 2% lime). A total of 91 VOCs, including 82 known compounds, such as formaldehyde, alcohols, esters, and ketones, and 9 unknown compounds, were detected in each sample by GC-IMS. The relative contents of formaldehyde, ketones, alcohols, and esters in the samples was more than 80%. Among the 29 VOCs with variable importance in projection (VIP) values > 1 and P < 0.05, formaldehyde, heptagonal(dimer), 2-methyl-E-2-butenal-M”, 3-methyl-2-butenal-M(dimer), 1-octen-3-ol, butyl acetate(dimer), ethyl 3-methylbutanoate, and 2-pentylfuran were the markers that distinguished the volatiles in S. rugosoannulata cultivated with different groups of raw substrate materials.
Download PDF
Full Article
Analysis of Volatile Substances in Stropharia rugosoannulata Farlow Cultivated Under Forest Canopy with Four Different Culture Substrates by Electronic Nose and GC-IMS
Hui Wang,b,¶ Ying Rao,a,¶ Juanjuan Chen,a,& Fei Zhou,a,& Jinping Zhang,a,* Hongxia Xu,b,& and Jianbin Xu b,&
This study examined the volatile organic compounds (VOCs) in Stropharia rugosoannulata Farlow that were cultivated on four substrates formulated with agricultural and forestry wastes. The VOCs were analyzed by an electronic nose (E-nose), gas chromatography-ion mobility spectroscopy (GC-IMS), principal component analysis (PCA), and an orthogonal partial least squares discriminant analysis (OPLS-DA). A4(40% sawdust, 30% camellia shells, 20% rice husk, 8% bran, and 2% lime) was the most effective overall at determining the quality of flavor. The E-nose showed that there were similar profiles of aromas for A2(100% Eleusine coracana (L.) Gaertn straw) and A3(70% bamboo chips, 20% rice husk, 8% bran, and 2% lime). A total of 91 VOCs, including 82 known compounds, such as formaldehyde, alcohols, esters, and ketones, and 9 unknown compounds, were detected in each sample by GC-IMS. The relative contents of formaldehyde, ketones, alcohols, and esters in the samples was more than 80%. Among the 29 VOCs with variable importance in projection (VIP) values > 1 and P < 0.05, formaldehyde, heptagonal(dimer), 2-methyl-E-2-butenal-M”, 3-methyl-2-butenal-M(dimer), 1-octen-3-ol, butyl acetate(dimer), ethyl 3-methylbutanoate, and 2-pentylfuran were the markers that distinguished the volatiles in S. rugosoannulata cultivated with different groups of raw substrate materials.
DOI: 10.15376/biores.20.1.1365-1383
Keywords: Stropharia rugosoannulata; Volatile organic compounds; Electronic nose; Gas Chromatography-ion mobility spectroscopy
Contact information: a: Research Institute of Subtropical Forestry, Chinese Academy of Forestry, Hangzhou, Fuyang, China; b: Chun ‘an County Thousand-island Lake Forestry Farm, Hangzhou, Zhejiang, China; ¶: These authors contributed equally to this work; &: These authors also contributed equally to this work.
* Corresponding author: jinpingzhang@126.com
GRAPHICAL ABSTRACT
INTRODUCTION
Edible mushrooms are a class of nutritious and healthy foods that have high amounts of protein, dietary fiber, and trace minerals (Roupas et al. 2012; Roncero-Ramos and Delgado-Andrade 2017). Stropharia rugosoannulata Farlow, commonly known as wine cap stropharia, burgundy mushroom and king stropharia, is a rare edible fungus in the Strophariaceae family (Wu et al. 2013; Chen et al. 2020). This mushroom is brightly colored and has a smooth cap. In addition, the meat and stems are crispy. S. rugosoannulata contains abundant proteins and various mineral elements that are healthy to humans. It is also a good source of various biologically active substances. Therefore, S. rugosoannulata is one of the top 10 traded mushrooms in the international edible fungus market (Hu et al. 2020). It is highly promoted in China as a high-quality and rare edible fungus that is rich in nutrients, an antioxidant and has anti-tumor effects (Liu et al. 2020). It is recommended by the Food and Agriculture Organization (FAO) of the United Nations as one of the characteristic varieties for cultivation in developing countries (Yang et al. 2021). S. rugosoannulata is highly resistant to stress and bacterial contamination. It is strongly adaptable, and it has a substantial potential for degrading environmental pollutants (Castellet-Rovira et al. 2018).
Its volatile components are among the key factors that determine its quality and consumer perception; thus, they can be used to evaluate the nutritional value and freshness of food to some extent (Fang et al. 2017). Volatile flavor compounds are usually qualified by headspace solid-phase microextraction coupled with gas chromatography mass spectrometry (HS-SPME-GC-MS), gas chromatography-olfactometry-mass spectrometry (GC-O-MS) and headspace-gas-chromatography ion-mobility spectrometry (HS-GC-IMS), and an electronic nose (E-nose) is usually used as an auxiliary tool for the qualitative analysis (Chen et al. 2020; Adelina et al. 2021; Shen et al. 2021). Gas chromatography-ion mobility spectroscopy (GC-IMS), an emerging analytical technique of volatile organic compounds (VOCs). It does not require the concentration of VOCs from the sample by SPME, but rather, it directly extracts and analyzes some volume of gas from the sample (Zhang et al. 2020). This type of analysis not only can reflect the true aroma components of a sample more accurately, but is also quicker, more sensitive, and cheaper than gas chromatography-mass spectrometry (GC-MS) (Gerhardt et al. 2018; Wang et al. 2019). The high moisture content and thin skin of fresh S. rugosoannulata makes it likely to rot soon after harvest, which seriously affects its value as a commodity (Mahajan et al. 2008). Drying is the most common convenient method to control the moisture and effectively prolong the shelf life of these fresh mushrooms. However, high-temperature (≥ 100 ℃) drying tends to cause losses of flavor of the fresh mushrooms (Yang et al. 2021). Therefore, S. rugosoannulata samples were usually pre-dried at 40 °C for 20 h to reduce the moisture content from 98% to 7%, but the flavor remains unchanged to ensure that the flavor remains as fresh as possible.
S. rugosoannulata is a straw-rot fungus, and it can be cultivated on a variety of substrates. With the development of the under-canopy economy in China, the under-canopy cultivation technique of S. rugosoannulata has become increasingly mature, and various materials, such as straw, awn stalk, bamboo, sawdust and agricultural and forestry wastes, have been used as substrates to cultivate this fungus. Currently, the studies on S. rugosoannulata have primarily focused on the pretreatment of raw cultivation materials, formulation of cultivation substrates, cultivation pattern, antioxidant capacity, protoplasts, and breeding among others (Qin et al. 2022). In this study, the volatile flavor compounds in S. rugosoannulata were analyzed by the E-nose and GC-IMS cultivated with four different formulas of substrates which were primarily composed of wood chips, straw, bamboo chips, and camellia shells. The obtained GC-IMS data were subjected to an orthogonal partial least squares discriminant analysis (OLPS-DA) using SIMCAP14 and a principal component analysis (PCA) to determine the effects of the substrates on the volatile flavor compounds. The VOC fingerprints and evaluation criteria of S. rugosoannulata cultivated with four types of substrates were established to provide a scientific basis for improving substrate processing and formulating and screening substrates capable of cultivating Agaricus blazei mushrooms with better flavors.
MATERIALS AND METHODS
The study was conducted in Chun’an County, Zhejiang Province, China, (29°11′-30°02′ N and 118°20 ‘-119°20’ E). The area has a mid-subtropical monsoon warm and humid climate with mean annual precipitation of 1,430±309.00 mm and a mean annual temperature of 17 ℃±4.80 ℃. The mushrooms were cultivated under a canopy of a mixed broad-needle forest with a canopy density of 0.7. The study plot is flat and drained well. The soil pH was 5 to 7. The cultivation substrate formulas are shown in Table 1. There were 12 mm particles of sawdust, bamboo chips and camellia shells.
Table 1. The Cultivation Substrate Formulas
The ingredients in formula A1, A3, and A4 were added with 2% quicklime, moisturized to 70 to 75%, and fermented in piles for 1 month. The piles were turned once in the middle. A total of 20% rice husk and 8% bran were added sequentially, moisturized to a content of 65 to 75%, and piled in trapezoidal heaps for fermentation. The pile was first turned 3 days after the temperature reached 50 °C. The pile was turned again approximately 2 to 3 days after the temperature had reached 65 ºC, and water was added to rebuild the heap of the same size with a moisture content of 75%. The pile was turned a third time after 3 to 4 days. The fungus was only sown on the substrate when the temperature dropped below 28 °C. No fermentation was conducted on formula A2. Finger millet (Eleusine coracana) straw was air-dried and used directly as the substrate. The strain was first sown on Nov 1, 2022. An 8-10 cm thick substrate layer was first spread on a cultivation bed that was 50 to 60 cm wide. The strain was broken into sizes of approximately 2.5 cm, sown on the substrate with spaces of 10 cm, and gently pressed tight. The second layer of substrate was spread 10 to 12 cm thick, and the strain was sown by the same method. Finally, the strain was covered with a third layer of substrate that was 3 to 5 cm thick. The cultivation piles were gently compacted, and the pile surfaces were shaped to resemble the back of a turtle.
On March 15, 2023, 3 kg of well-mixed samples were removed from each cultivation formulation. S. rugosoannulata specimens were observed to have a height of 1 to 2 centimeters, stem cap diameter of 4 to 5 centimeters, stem diameter of 1 to 2 centimeters, and stem length of 3 to 4 centimeters. The test samples of three triplicate were dried for 20 h at 40 °C after the surface dirt had been removed and stored at room temperature for the E-nose and GC-IMS analyses (Sun et al. 2023).
E-nose Sensing
The PEN3 E-nose (Airsense Analytics Co. Ltd., Schwerin, Germany) contains 10 different metal oxide sensors. The individual sensors of the electronic nose exhibit different selectivities for classes of compounds, such as W1C (aromatic components and benzenes), W5S (nitrogen oxides), W3C (ammonia and aromatic components), W6S (hydrogen), W5C (alkane aromatic components), W1S (short chain alkanes, such as methane), W1W (inorganic sulfides), W2S (alcohol, ethers, aldehydes and ketones), W2W (aromatic components and organic sulfides) and W3S (long-chain alkanes) (Shen et al. 2021). The preparation for the E-nose to detect the aroma profiles included placing each 0.5 g sample in a 15 mL headspace bottle and incubating it in a 26 °C water bath for 30 min before the profiles of the aroma were measured using the following conditions: cleaning time, 120 s; reset time, 5 s; pre-injection time, 5 s; flow rate of carrier gas, 400 mL/min; and measurement time, 60 s. The data were analyzed by a linear discriminant analysis (LDA) using the E-nose software. Three replicates were established for each group of the S. rugosoannulata samples.
GC-IMS Analysis
The GC-IMS analysis was conducted on a Flavour Spec® flavor analyzer [FlavourSpec®, GAS (Adelina et al. 2021). Gesellschaft fuir analytische Sensorsysteme GmbH, Dortmund, Germany] that consisted of a syringe and an automatic headspace sampling unit. The headspace was sampled by transferring 3.0 g of S. rugosoannulata powder to a 20 mL headspace vial and incubating it at 80 °C at 500 rpm for 15 min. A syringe at 85 °C was used to inject a 200 μL headspace sample into a MXT-WAX gas chromatography column (30 m, 0.53 mm ID, 1.0 μm df) (Restek Corporation, Bellefonte, PA, USA) for pre-separation followed by IMS detection (45 °C). The column temperature was 60 °C, and 99.999% N2 was used as the carrier gas. The initial flow rate (E1) of the carrier gas was 150 mL/min. The E2 was held at 2 mL/min for 5 min, increased to 10 mL/min, held at 10 mL/min for 20 min, increased to 100 mL/min, and then held at 100 mL/min until 30 min (Xi et al. 2024). There were three replicates of each group of the S. rugosoannulata samples.
Calculation of the ROAV
The contribution of each compound to the flavor of S. rugosoannulata was evaluated by the relative odor activity value (ROAV). The OAVi and ROAVi were calculated as described by Eqs. 1 and 2,
OAVi=Ci/OTi (1)
ROVAi=OAVi/OAVmax×100% (2)
where Ci is the content of a compound /(mg/kg) in the sample; OTi is the odor threshold (mg/kg) of the compound, and OAVmax is the maximum value of OAV of all the compounds in each sample. The compounds of ROAV 0.1 contributed to the overall flavor (Wei et al. 2019), while the other compounds contributed less. A larger ROVA represents a greater contribution of the flavor compounds to the overall flavor in the sample (Fan et al. 2019).
Statistical Analysis
The data were processed using Microsoft Excel 2019 (Redmond, WA, USA) and SPSS 21.0 (IBM, Inc., Armonk, NY, USA). The LSD analysis method was used for multiple comparison analysis. The OPLS-DA was conducted in the SIMCAP14.1 software. VOCal was used to examine the spectra and qualitatively and quantitatively analyze the data. The compounds detected were identified by searches against the built-in NIST and IMS databases. The spectral differences among the four samples were determined using the Reporter plug-in and visualized as a three-dimensional (3-D) graph, two-dimensional (2-D) topography and a difference graph. The flavor fingerprints were compared using the Gallery Plot plug-in. The dynamic PCA was conducted using the Dynamic PCA plug-in. The contents of volatile substances were determined as the corresponding normalized relative peak areas (%).
RESULTS AND DISCUSSION
E-nose Detection
A linear discriminant analysis (LDA) maximizes the interclass variance and minimizes the intraclass variance, i.e., it reduces the differences within the class and enhances the differences between different classes (Gerhardt et al. 2018; Xu et al. 2021). Therefore, an LDA was used to distinguish the four groups of the S. rugosoannulata samples. As shown in Fig. 1I, the variance contribution rates of LD1 and LD2 were 92.27% and 6.26%, respectively, and there were relatively large distances among the four samples. In particular, the longest distances were observed between A1 and the other three groups. A2 and A3 were the closest. This may be related to the fact that substrate A1 was primarily composed of sawdust that originated from woody trees, while substrate A4 contained 30% Camellia oleifera shells and 40% sawdust. Substrates A2 and A3 were composed of finger millet (Eleusine coracana) and bamboo, respectively, which are both members of the Gramineae family. Therefore, the smallest difference was observed between A2 and A3.
As shown in Fig. 1II, W1W was more sensitive to A4 than to A2, A3, and A1, and its responses to A4, A2, and A3 were 2.52-, 2.08-, and 1.65-fold higher than that of A1, respectively. The significantly differential responses suggest that the sensor can clearly distinguish the substrates used to culture S. rugosoannulata samples by their odors. The W1W, W5S and W2W sensors showed stronger responses, which indicated that the S. rugosoannulata samples, particularly A4, may contain high amounts of organosulfur compounds and nitrogen oxides. A1 had the lowest responses, which suggested that there was a low content of organic sulfide in A1. This was consistent with the low amount of dimethyl trisulfide detected by the GC-IMS (Table 1). There were weak responses from the other sensors, such as the W2S sensor, which is sensitive to alcohols, ethers, aldehydes and ketones. This suggested that the E-nose is only sensitive to a limited number of volatiles in S. rugosoannulata, and it can be used as a tool to supplement the GC-IMS analysis.
Fig. 1. Linear discriminant analysis (LDA) (I) and E-nose response radar map (II)
Note: The horizontal coordinate in I is the first principal component contribution, and the vertical coordinate is the second principal component contribution.
Characterization of VOCs
Qualitative analysis
The VOCs in the four groups of S. rugosoannulata samples were analyzed by GC-IMS. Figure 2 shows the 3-D graph, 2-D topography and difference graph of the ion mobility mass spectra. The qualitative analysis was based on the 2-D separation. Topography that consists of many 2-D maps can serve as the VOC fingerprint of a sample (Gerhardt et al. 2018).
Fig. 2. GC-IMS analysis of volatile flavor substances in the four groups of S. rugosoannulata samples. Note: I: Top view of the map; Ⅱ: Difference map. For Ⅰ/Ⅱ, the ordinate, abscissa and vertical lines on the abscissa represent the retention time RT (s), ion migration time Dt (ms), and RIP, respectively.
The four groups showed a similar composition of the VOCs with the differences only observed in the peak intensities of a few components. Their topographies were compared to facilitate the convenience of their observations (Fig. 2Ⅰ). The ordinate and abscissa axes in Fig. 2Ⅰ represent the retention time (RT, s) of the GC and ion migration time (Dt, ms). The red line on the abscissa axis is the reaction ion peak (RIP). Each dot on the two sides of the RIP represents a VOC, and a darker dot indicates a higher concentration. The retention and ion migration times of most of the VOCs ranged from 200 to 800 s and 1.00 to 1.75 ms for all the groups, respectively. The differences in their spectra were visualized more directly using the spectrum of A1 as a reference, and the spectra of the other groups were subtracted from the reference to obtain the difference graph as shown in Fig. 2Ⅱ. Identical contents of a VOC in the two groups resulted in the appearance of the subtraction results in the white background. A red dot indicated that the concentration of VOC was higher than that in the reference, and a blue dot indicated that the VOC concentration was lower than that of the reference. Overall, the composition of the four groups of VOCs in S. rugosoannulata were generally similar, and only a few compounds had different peak intensities. A1 differed significantly from the other three groups, and A2 and A 3 were close, which was consistent with the results of the E-nose.
Quantitative Analysis
The VOCs in the headspace samples of the four groups were analyzed by GC-IMS and matched against the built-in NIST and IMS databases. A total of 91 VOCs are detected in each group, which included 82 known VOCs (3 acids, 25 aldehydes, 17 alcohols, 17 esters, 17 ketones, 3 hydrocarbons, 1 furan, and 2 organosulfurs), and 9 unknown compounds (Table 2).
The VOCs in the samples of the four groups were primarily aldehydes, ketones, alcohols and esters, such as propanal#, 2-methyl propanal, 2-methylbutanal, hexanal#, acetone, butan-2-one, 2-pentanone, butan-1-ol#, 1-propanol#, methyl 3-methylbutanoate#, and ethyl acetate among others. These compounds were followed by unknown ones, acids, hydrocarbons, furans, and organosulfurs. A1 contained the highest contents of alcohols, acids, hydrocarbons and organosulfurs and the lowest amounts of unknown compounds. A2 had the highest contents of aldehydes, ketones and furans and the lowest contents of alcohols, esters, and hydrocarbons. A3 had the highest contents of esters, while the lowest contents of organosulfurs. A4 had the highest contents of unknown compounds and the lowest contents of aldehydes, ketones, acids and furans. The cellulose content of oil tea husk is 18.6%, hemicellulose content is 49.3%, lignin is 29.7%, saponin content is 4.8%, and tannin content is 2.3% (Zhang et al. 2018). The content of its main components and secondary metabolites differed greatly from that of oak (46.0% cellulose, 25.4% hemicellulose, 17.3% lignin) (Li et al. 2023) and bamboo (42.5% cellulose, 27.0% hemicellulose, 23.1% lignin) (Zhang et al. 2011), and there were few secondary metabolites of saponin and tannin in oak and bamboo. It can be seen that the addition of an appropriate proportion of oil tea husk to the cultivation substrate can affect the growth and metabolism of S. rugosoannulata by influencing the production and release of VOCs.
Table 2. GC-IMS Identification Results of VOCs in Four S. rugosoannulata Groups
