Fatty Acids Predominantly Affect Anti-Hydroxyl Radical Activity and FRAP Value: The Case Study of Two Edible Mushrooms

Compared to plants, nowadays mushrooms attract more attention as functional foods, due to a number of advantages in manipulating them. This study aimed to screen the chemical composition (fatty acids and phenolics) and antioxidant potential (OH•, 2,2-diphenyl-1-picrylhydrazyl (DPPH•) and ferric reducing ability of plasma (FRAP)) of two edible mushrooms, Coprinus comatus and Coprinellus truncorum, collected from nature and submerged cultivation. Partial least square regression analysis has pointed out the importance of some fatty acids—more precisely, unsaturated fatty acids (UFAs) followed by fatty acids possessing both short (C6:0 and C8:0) and long (C23:0 and C24:0) saturated chains—and phenolic compounds (such as protocatechuic acid, daidzein, p-hydroxybenzoic acid, genistein and vanillic acid) for promising anti-OH•, FRAP and anti-DPPH• activities, respectively. However, other fatty acids (C16:0, C18:0 and C18:3n3) along with the flavonol isorhamnetin are actually suspected to negatively affect (by acting pro-oxidative) the aforementioned parameters, respectively. Taken together, design of new food supplements targeting oxidative stress might be predominantly based on the various UFAs combinations (C18:2n6, C20:1, C20:2, C20:4n6, C22:2, C22:1n9, etc.), particularly if OH• is suspected to play an important role.


Introduction
Antioxidant activity (AO) is considered to be an important feature of edible mushrooms in the prevention of oxidative stress [1][2][3]. Their antioxidants may well contribute to maintaining their physiological balance by neutralisation of free radicals (with stress on reactive oxygen species), without toxic or mutagenic effects, as opposed to synthetic antioxidants [3,4]. Generally speaking, it is believed that phenolics are primarily responsible for AO [1,5,6]. These compounds may originate either from fruiting bodies (generative structures) or from mycelia and extracellular broth (i.e., from the filtrate RSC (DPPH) (%) = (1 − A sample /A control ) × 100% (1) where A sample and A control stand for the absorbance of the tested and control samples, respectively. A lower IC 50 value corresponds to higher AO of the sample.

OH Assay
Anti-OH radical activity was determined according to a modified method of Halliwell and Gutteridge [19]. The reaction mixture contained 100 µL H 2 O 2 , 100 µL FeSO 4 , 100 µL 2-deoxyribose-D-ribose, 2.7 mL of phosphate buffer pH 7.4 and 10 µL of the each extract. After incubation (60 min at 37 • C), 0.2 mL of EDTA (ethylenediaminetetraacetic Acid) and 2 mL of TBA reagent (5.2 mL perchloric acid, 1.5 g thiobarbituric acid and 60 g of trichloroacetic acid) were added. Afterwards, the absorbance of a characteristic pink complex was measured at 532 nm. Finally, the results were expressed as IC 50 values ± standard deviations (µg/mL).

FRAP Assay
Ferric reducing ability of plasma (FRAP) was evaluated by a spectrophotometric assay as previously described [20]. The reaction mixture contained 225 µL of FRAP reagent (10 mmol/L TPTZ solution in 40 mmol/L HCl, 0.02 mmol/L FeCl 3 ·6H 2 O and acetate buffer (pH 3.6), in a ratio 10:1:1), 22.5 µL of distilled water and 10 µL of each extract. After 6 min of incubation, the absorbance was measured, while reduction potential was calculated as milligrams of ascorbic acid equivalents (AAE) per gram of dry weight (mg AAE/g d.w.), calculated according to the standard calibration curve of ascorbic acid solution.

Determination of Phenolic Compounds by HPLC-MS/MS Analysis
This chemical analysis was done applying the method of Orčić et al. [21]. All extracts were diluted with mobile phase solvents A (water) and B (methanol), premixed in 1:1 ratio, to obtain a final concentration of 2 mg/mL. Reference standards of the phenolic compounds were obtained from Sigma-Aldrich Chem (Steinheim, Germany), Fluka Chemie gmbh (Buchs, Switzerland) or from ChromaDex (ChromaDex Corp., Santa Ana, CA, USA). HPLC gradient grade methanol was purchased from J. T. Baker (Deventer, The Netherlands), and p.a. formic acid and DMSO from Merck (Darmstadt, Germany). A total of 15 working standards, ranging from 1.53 ng/mL to 2.50 × 10 4 ng/mL, were prepared by serial 1:1 dilutions of the standard mixture with solvents A and B (1:1). Samples and standards were analysed using Agilent Technologies 1200 Series high-performance liquid chromatograph coupled with Agilent Technologies 6410A Triple Quad tandem mass spectrometer with electrospray ion source (Agilent Technologies, Inc., Santa Clara, CA, USA), and controlled by Agilent Technologies MassHunter Workstation software -Data Acquisition (ver. B.03.01, Agilent Technologies, Inc., Santa Clara, CA, USA). First, 5 µL were injected into the system. Afterwards, compounds were separated on Zorbax Eclipse XDB-C18 (50 mm × 4.6 mm, 1.8 µm, Agilent Technologies, Inc., Santa Clara, CA, USA) rapid resolution column held at 50 • C. Mobile phase was delivered at flow rate of 1 mL/min in gradient mode (0 min 30% B, 6 min 70% B, 9 min 100% B, 12 min 100% B, re-equilibration time 3 min). Eluted compounds were detected by ESI-MS, using the ion source parameters as follows: nebulisation gas (N 2 ) pressure 40 psi, drying gas (N 2 ) flow 9 L/min and temperature 350 • C, capillary voltage 4 kV, negative polarity. Data were acquired in dynamic MRM mode, using the optimised compound-specific parameters (retention time, precursor ion, product ion, fragmentor voltage, collision voltage) as reported by Orčić et al. [21]. For all the compounds, peak areas were determined using Agilent MassHunter Workstation software -Qualitative Analysis (ver. B.04.00, Agilent Technologies, Inc., Santa Clara, CA, USA). Briefly, calibration curves were plotted in the OriginLabs Origin Pro (ver. 8.0) software (Northampton, MA, USA). Limit of detection (LoD) was estimated as the lowest concentration resulting in well-defined peak [21].

GC-MS Identification and Quantification of Fatty Acids
This chemical analysis was also performed as previously described [22]. As a solvent, n-heptane was used, along with the evaporation in the nitrogen stream. The prepared samples were analysed on a GC Agilent 7890A system (Agilent Technologies, Santa Clara, CA, USA) equipped with a Flame Ionisation Detector (FID) and an auto-injecting liquid system on a capillary column of mixed silica (Supelco SP-2560 Capillary GC Column, 100 m × 0.25 mm, d = 0.20 µm, Merck KGaA, Darmstadt, Germany). The gas carrier was helium of purity of 99.9997%, at a flow rate of 1.5 mL/min and a pressure of 1.092 bar. The samples were injected in a column in split mode in the ratio 30:1. The applied temperatures ranged from 40 to 230 • C. Total time of analysis was 41.311 min. Fatty acid methyl ester peaks were identified by comparing retention times (RI) from RI samples of the Supelco 37 component fatty acid methyl ester mix standard as well as by the internal data obtained in the pre-assay of fatty acids in a GC with a mass detector. The obtained results were expressed as the mass of the individual fatty acid or group of fatty acids (g) in 100 g of fatty acids from the biological material.

Statistical Analysis
All measurements were performed in triplicate. The results were expressed as the mean values ± standard deviations. IC 50 values were obtained by interpolation from a linear regression analysis using OriginLabs Origin Pro (ver. 8.0) software. One-way analysis of variance (ANOVA) with Tukey's test was used to determine the statistically significant difference between the analysed extracts (p < 0.01). The strength of association between pairs of variables was measured with the Pearson product moment correlation at a 5% level of significance (p < 0.05). Partial Least Squares Regression (PLSR) was applied for multivariate analysis (XLSTAT statistical and data analysis solution, Addinsoft 2019, Boston, MA, USA).

Anti-OH Radical Activity
C. truncorum FB extract, the sample displaying the most potent anti-OH radical activity, was followed by C. comatus FB extract (Table 1). In comparison, C. comatus FB extract (collected in China) had much lower activity (3.23 ± 0.28 mg/mL) [5].

FRAP Value
Once again FB extracts were more effective. The most profound FRAP value was recorded for C. comatus FB extract. A number of other mushroom species including Xylaria polymorpha (3.25 ± 0.04 mg AAE/g d.w.), Meripilus giganteus (10.45 ± 0.44 mg AAE/g d.w.) and Agrocybe aegerita (10.74 ± 0.09 mg AAE/g d.w.) exhibited lower FRAP values [1]. Generally speaking, differences in the antioxidant potential of different samples were clearly observed; some of the tested samples were actually proven to be more effective compared to the methanolic extracts of some previously analysed species [1,2,5,14,23].

HPLC-MS/MS Determination of Phenolic Compounds
Following HPLC-MS/MS procedure optimised for the quantification of 45 phenolics, 28 compounds were identified in the tested samples (Table 2). p-Hydroxybenzoic acid was the most abundant compound, followed by quinic acid. Both phenolics are known as good antioxidants, due to their reducing properties (depending on hydrogen or electron donors) and ability to stabilise the unpaired electron [24,25]. In addition to this, protocatechuic acid was also found in all the samples, with notably greater amount in the submerged extracts, with stress on C. comatus F extract. Both in vitro and in vivo designed studies have clearly pointed out that protocatechuic acid may be considered as effective antioxidant, even more potent than trolox, a synthetic vitamin E analog [26][27][28]. Furthermore, cinnamic acid was detected in the most of the extracts (except C. truncorum M extract). On the other hand, the isoflavonoids daidzein and genistein, also proven antioxidants [29][30][31], were detected only in the submerged cultures. Finally, vanillic acid, another antioxidant of natural origin [32,33], was identified only in the submerged extracts.

GC-MS Analysis
Gas Chromatography-Mass Spectrometry (GC-MS) was used to analyse the contents of fatty acids. A total of 28 fatty acids were identified in the screened extracts ( Table 3). The content of total unsaturated fatty acids (UFAs), mono-unsaturated fatty acids (MUFAs), poly-unsaturated fatty acids (PUFAs) and saturated fatty acids (SFAs) revealed that UFAs were most abundant ones. The same fatty acids are recommended as high-quality ingredients of a healthy diet, inter alia, capable of decreasing blood lipids [13,16,17]. In the majority of the analysed samples (except C. truncorum F extract), linoleic acid (C18:2n6c) was the most common one. Additionally, oleic fatty acid (C18:1n9c) was present in all the samples. Table 3. The content of fatty acid compounds in the analysed samples (relative %).

Fatty Acid Carbon Numbers
Common Names (Acid)  These findings are in a good agreement with literature data [13][14][15]34,35]. However, no one has previously reported C. truncorum FB fatty acid profile, to the best of our knowledge. The aforementioned profile is somewhat similar to C. micaceus fatty acid profile [36].
Also, these are real pioneering data for the both mushrooms samples developed in the submerged cultivation. Furthermore, it's noteworthy to mention that FB extracts contained more UFAs and PUFAs, compared to the rest of samples (Table 3). Thus far, UFA content has been linked with AO increase [37]. On the other hand, PUFAs have been claimed to modulate the activity of antioxidant enzymes. However, their AO cannot be easily predicted, since it doesn't depend on the length of the carbon chain and/or degree of unsaturation [17].

Partial Least Squares Regression (PLSR) Analysis
Partial least squares regression (PLSR) analysis was used to define the possible interrelationships between the chemical composition (based on their phenolic (Table 2) and fatty acid (Table 3) profiles) (independent variables, X) and AO activity (DPPH•, OH• and FRAP, dependent variables, Y; reciprocal values of IC 50 for anti-DPPH and anti-OH radical activities, Table 1) of the analysed methanolic extracts. Firstly, PLSR was performed for all three AO measures, resulting in the correlation circles between the extracts, their fatty acid ( Figure 1A) or phenolic ( Figure 1B) profile and AO measures, with first two PLSR components (t 1 , t 2 ). Although the global R 2 between Y and (t 1 , t 2 ) (which gives an upper bound of how well the model explains the data and predicts new observations) is slightly higher for the phenolic profile (0.962), compared to the fatty acid one (0.934), the quality of the former regression is lower since R 2 resulting from the cross-validation (Q 2 cum), that defines the stability of the model and sets the lower bound of how well the model explains the data [38], is 0.389, compared to 0.686 for the fatty acid profile. Figure 1 shows that all dependent variables are located at the periphery of the correlation circle meaning that can be explained by the concentrations of fatty acid or phenolic compounds located either in their vicinity (e.g., C22:2 or cinnamic acid, in the case of OH•) exhibiting a positive (antioxidative) influence, or opposite to them (e.g., C18:3n3 or vanillic acid, in the case of OH•) displaying a negative (pro-oxidative) one.
cross-validation (Q cum), that defines the stability of the model and sets the lower bound of how well the model explains the data [38], is 0.389, compared to 0.686 for the fatty acid profile. Figure 1 shows that all dependent variables are located at the periphery of the correlation circle meaning that can be explained by the concentrations of fatty acid or phenolic compounds located either in their vicinity (e.g., C22:2 or cinnamic acid, in the case of OH • ) exhibiting a positive (antioxidative) influence, or opposite to them (e.g., C18:3n3 or vanillic acid, in the case of OH • ) displaying a negative (pro-oxidative) one. Although the correlation circle is useful for gaining the overall picture, it actually doesn't indicate (specify) which combination of fatty acid or phenolic compounds has statistically significant influence on the specific measure of AO activity of the analysed mushrooms extracts. In order to estimate this, a separate one-component PLSR model was built for each dependent variable. The models were then pruned until all variables with insignificant standardised regression coefficients (confidence intervals include 0) were deleted. Bar graphs of the regression coefficients for all three models are shown in Figure 2. Upwards and downwards pointing bars indicate positive (antioxidative) and negative (pro-oxidative) influences, respectively. Comparison of R 2 and Q 2 cum values clearly pointed out that AO activity of the extracts measured by FRAP assay was much better explained by PLSR model based on their fatty acid profiles ( Figure 2E, R 2 = 0.927, Q 2 cum = 0.535) versus the phenolic ones ( Figure 2F, R 2 = 0.567, Q 2 cum = −0.393). Practically, entire variability (92.7%) in the AO activity of the extracts measured by FRAP assay can be explained by coordinated antioxidative potential of unsaturated fatty acids (UFAs) (C18:2n6, C20:1, C20:2, C20:4n6, C22:2 and C22:1n9) followed by short (C6:0 and C8:0) or long (C23:0 and C24:0) chain saturated fatty acids (SFAs). Since palmitic and stearic acids were by far most abundant SFAs, SFAs generally may be linked with pro-oxidative action, unlike UFAs and PUFAs ( Figure 2E). Similarly to FRAP, anti-OH radical activity of the extracts is also much better explained by PLSR model based on their fatty acid profiles ( Figure 2C, R 2 = 0.854, Q 2 cum = 0.443), compared to the phenolic ones ( Figure 2D, R 2 = 0.671, Q 2 cum = 0.274). Such a trend is actually expected due to a strong correlation between FRAP and OH• values (R = 0.932, p = 0.007), that is confirmed by their close position at the correlation circle ( Figure 1A), too. In fact, due to such a tight correlation, OH• PLSR model is essentially based on the identical fatty acids as FRAP PLSR model ( Figure 2C). Contrary to FRAP and OH assays, the PLSR model based on the phenolic profiles of the extracts ( Figure 2B, R 2 = 0.960, Q 2 cum = 0.876) much better explains their anti-DPPH radical activity, compared to the model based on the fatty acid profiles (Figure 2A, R 2 = 0.825, Q 2 cum = −0.608). Protocatechuic acid, daidzein, p-hydroxybenzoic acid, genistein and vanillic acid are phenolics suspected to primarily contribute to anti-DPPH radical activity, unlike the flavonol isorhamnetin that is, indeed, likely to display pro-oxidative activity ( Figure 2B).

Conclusions
Taken together, submerged C. comatus F extract was most effective in neutralising DPPH radicals, while C. truncorum & C. comatus FB extracts were most effective in neutralising OH radicals. The aforementioned FB extracts also displayed potent FRAP values. According to PLSR analysis, fatty acid chemistry is suspected to predominantly affect anti-OH radical activity and FRAP value, while phenolic chemistry is likely to be the key one for the observed anti-DPPH radical activity. Consequently, design of new food supplements targeting OH radicals might be predominantly based on the various UFAs combinations (C18:2n6, C20:1, C20:2, C20:4n6, C22:2, C22:1n9, etc.)

Conclusions
Taken together, submerged C. comatus F extract was most effective in neutralising DPPH radicals, while C. truncorum & C. comatus FB extracts were most effective in neutralising OH radicals. The aforementioned FB extracts also displayed potent FRAP values. According to PLSR analysis, fatty acid chemistry is suspected to predominantly affect anti-OH radical activity and FRAP value, while phenolic chemistry is likely to be the key one for the observed anti-DPPH radical activity. Consequently, design of new food supplements targeting OH radicals might be predominantly based on the various UFAs combinations (C18:2n6, C20:1, C20:2, C20:4n6, C22:2, C22:1n9, etc.) Author Contributions: M.K., B.P. and M.Ž. designed this study, while K.A. performed most of the experimental work, kindly supported by A.N. and F.Š. K.S. was the key author responsible for statistics. M.K., K.A., B.P. and M.Ž. predominantly discussed the obtained data. Finally, K.A., M.K. and B.P. wrote this manuscript, while M.Ž. critically read it providing fruitful insights.

Conflicts of Interest:
No potential conflict of interest was reported by the authors.