Antifungal Activity against Botryosphaeriaceae Fungi of the Hydro-Methanolic Extract of Silybum marianum Capitula Conjugated with Stevioside

Silybum marianum (L.) Gaertn, viz. milk thistle, has been the focus of research efforts in the past few years, albeit almost exclusively restricted to the medicinal properties of its fruits (achenes). Given that other milk thistle plant organs and tissues have been scarcely investigated for the presence of bioactive compounds, in this study, we present a phytochemical analysis of the extracts of S. marianum capitula during the flowering phenological stage (stage 67). Gas chromatography–mass spectroscopy results evidenced the presence of high contents of coniferyl alcohol (47.4%), and secondarily of ferulic acid ester, opening a new valorization strategy of this plant based on the former high-added-value component. Moreover, the application of the hydro-methanolic extracts as an antifungal agent has been also explored. Specifically, their activity against three fungal species responsible for the so-called Botryosphaeria dieback of grapevine (Neofusicoccum parvum, Dothiorella viticola and Diplodia seriata) has been assayed both in vitro and in vivo. From the mycelial growth inhibition assays, the best results (EC90 values of 303, 366, and 355 μg·mL−1 for N. parvum, D. viticola, and D. seriata, respectively) were not obtained for the hydroalcoholic extract alone, but after its conjugation with stevioside, which resulted in a strong synergistic behavior. Greenhouse experiments confirmed the efficacy of the conjugated complexes, pointing to the potential of the combination of milk thistle extracts with stevioside as a promising plant protection product in organic Viticulture.


Introduction
Silybum marianum (L.) Gaertn (syn. Carduus marianus L.), commonly known as milk thistle, St. Mary's Thistle, or wild artichoke, is an herbaceous plant of the Asteraceae family. Native to the Mediterranean area, it is nowadays grown in many countries as a medicinal plant, due to the variety of biological activities-mostly linked to the hepatoprotective properties and anti-carcinogenic capacity-associated with the main pharmacological active ingredient extracted from its achenes (fruits): silymarin [1,2].
The standardized extract obtained from the dried fruits of S. marianum contains 70-80% of silymarin and 20-30% of polymeric and oxidized polyphenolic compounds [3]. Silymarin is a flavonolignan complex of polyphenolic molecules, which includes diasterereoisomers silybin A and silybin B (whose mixture in a 1:1 ratio is named silibinin), constituent of Stevia rebaudiana (Bertoni) Bertoni extract), which has antifungal properties, has been chosen to form such conjugate complexes, aiming at an enhancement of activity through synergism.

Plant Material, Reagents, and Fungal Isolates
The specimens of S. marianum under study were collected in the banks of Carrión river as it passes through the town of Palencia (Spain) during stage 67 (or 6N7) according to the extended BBCH scale [34]. This stage was chosen because silybins precursors ( Figure S1) should not have yet been consumed. The capitula of S. marianum were shade-dried and pulverized to fine powder in a mechanical grinder. Different specimens (n = 25) were thoroughly mixed to obtain a composite sample.
The three fungal pathogens (Table 1) were supplied as lyophilized vials (later reconstituted and refreshed as PDA subcultures) by the Agricultural Technological Institute of Castilla and Leon (ITACYL, Valladolid, Spain) [35].

Preparation of Plant Extracts
Silybum marianum capitula samples were mixed (1:20, w/v) with a methanol/water solution (1:1 v/v) and heated in a water bath at 50 • C for 30 min, followed by sonication for 5 min in pulse mode with a 1 min stop for each 2.5 min, using a probe-type ultrasonicator model UIP1000hdT (Hielscher Ultrasonics, Teltow, Germany; 1000 W, 20 kHz). The solution was then centrifuged at 9000 rpm for 15 min and the supernatant was filtered through Whatman No. 1 paper. Finally, aliquots of the extract were lyophilized for infrared spectroscopy analyses.

Physicochemical Characterization of S. marianum Extracts
The infrared vibrational spectra were registered using a Thermo Scientific (Waltham, MA, USA) Nicolet iS50 Fourier-transform infrared spectrometer, equipped with an inbuilt diamond attenuated total reflection (ATR) system. The spectra were collected with a 1 cm −1 spectral resolution over the 400-4000 cm −1 range, taking the interferograms that resulted from co-adding 64 scans. The spectra were then corrected using the advanced ATR correction algorithm [36] available in OMNIC TM software suite.
The hydroalcoholic plant extracts were studied by gas chromatography-mass spectrometry (GC-MS) at the Research Support Services (STI) at Universidad de Alicante (Alicante, Spain), using a gas chromatograph model 7890A coupled to a quadrupole mass spectrometer model 5975C (both from Agilent Technologies). The chromatographic conditions were: injection volume = 1 µL; injector temperature = 280 • C, in splitless mode; initial oven temperature = 60 • C, 2 min, followed by ramp of 10 • C/min up to a final temperature of 300 • C, 15 min. The chromatographic column used for the separation of the compounds was an Agilent Technologies HP-5MS UI of 30 m length, 0.250 mm diameter and 0.25 µm film. The mass spectrometer conditions were: temperature of the electron impact source of the mass spectrometer = 230 • C and of the quadrupole = 150 • C; ionization energy = 70 eV. NIST11 library and Adams [37] were used for compound identification.

Preparation of Bioactive Formulations
The stevioside-S. marianum, stevioside-coniferyl alcohol, and stevioside-ferulic acid conjugate complexes were obtained by mixing of the respective solutions in a 1:1 (v/v) ratio. The mixture was then sonicated for 15 min in five 3-min periods (so that the temperature did not exceed 60 • C) using a probe-type ultrasonicator [38].

In Vitro Tests of Mycelial Growth Inhibition
The antifungal activity of the different treatments was determined using the agar dilution method according to EUCAST standard antifungal susceptibility testing procedures [39], by incorporating aliquots of stock solutions onto the PDA medium to obtain concentrations in the 62.5-1500 µg·mL −1 range. The solutions were added to the PDA after being sterilized in an autoclave, when the temperature of the medium was close to that of polymerization (over 60 • C), in the same way that antibiotics are usually incorporated into these synthetic media. Mycelial plugs (∅ = 5 mm), from the margin of 1-week-old PDA cultures of N. parvum, D. viticola or D. seriata, were transferred to the center of plates incorporating the above-mentioned concentrations for each treatment (3 plates per treatment/concentration, with 2 replicates). Plates were then incubated at 25 • C in the dark for a week. PDA medium without any amendment was used as control. Mycelial growth inhibition was estimated according to the formula: ((d c − d t )/d c ) × 100, where d c and d t represent the average diameters of the fungal colony of the control and of the treated fungal colony, respectively. Effective concentrations (EC 50 and EC 90 ) were estimated using PROBIT analysis in IBM SPSS Statistics v.25 (IBM; Armonk, NY, USA) software. The level of interaction, i.e., synergy factors, were determined according to Wadley's method [40].

Greenhouse Bioassays on Grafted Plants
Together with the experiments of mycelial growth inhibition in vitro, bioassays with stevioside-S. marianum conjugate complexes were performed in living grapevine plants in order to scale the protective capabilities of these compounds against certain selected Botryosphaeriaceae species usually associated with GTD symptoms on young grapevine plants. Especially, N. parvum, D. viticola, and D. seriata were selected for the in vivo assay due to their significant presence as part of the contingent of fungi associated with decay problems in young vine plants [41] in Spain and other viticultural areas internationally. Plant material consisted of 30 plants each of cultivars 'Tempranillo' (CL. 32 clone) (2-years old) and 'Garnacha' (VCR3 clone) (one year old) grafted on 775P and 110R rootstocks, respectively (60 plants in total). Each grapevine plant was grown on a 3.5 L plastic pot containing a mixed substrate of moss peat and sterilized natural soil (75:25), incorporating slow release fertilizer when needed. Plants were kept in the greenhouse with drip irrigation and anti-weed ground cover for six months (June-December 2020). One week after placing them in the pots, grapevine plants were artificially inoculated with the mentioned three Botryosphaeriaceae taxa along with the stevioside-S. marianum treatment. Five repetitions were arranged for each pathogen/control product combination and variety ('Tempranillo' and 'Garnacha'), together with 8 positive controls (4 per grapevine variety) inoculated only with the pathogens, plus 6 negative controls (incorporating only the bioactive product), 3 for each variety (Table S1). Artificial inoculations of the pathogens and the control product were carried out directly on the trunk of the living plants at two sites per plant stand (separated a minimum of 5 cm among them) below the grafting point and not reaching the root crown. For the pathogens, agar plugs coming from 5-days-old fresh PDA cultures of each species were used as fungal inoculum. In the mentioned two inoculation points of each grapevine plant, slits (made up with a scalpel) of approx. 15 mm in diameter and 5 mm deep were done. After this, 5 mm diameter agar plugs were inoculated in contact with vascular tissues in the stem. Calcium alginate beads served as a dispersal matrix for the different control products and conjugates assayed, and were placed at both sides of the agar plug. For this, beads were prepared as follows: the control product was added to a 3% sodium alginate solution in a 2:8 ratio (20 mL compound/80 mL sodium alginate). Then, this solution was dispensed drop by drop onto a 3% calcium carbonate solution resulting in beads of 4-6 mm diameter containing the different control treatments. Finally, both agar plugs and beads were covered with cotton soaked in sterile bi-distilled water and sealed with Parafilm TM tape. During the assay period, application of copper to control powdery mildew outbreaks was performed in mid-July, together with a first sprouting (followed by periodic sprouting). Grapevine plants were visually examined weekly during the whole assay period, and the presence of foliar symptoms-including both internerval and nerval necroses-was scored to establish correlations between these and vascular symptoms at the end of bioassay. After six months in the greenhouse, plants were removed and two sections of the inoculated stems between the grafting point and the root crown were prepared, sectioned longitudinally, and the length of the vascular necroses caused by the different pathogens evaluated. Thus, the length of the vascular necroses was measured longitudinally on upper and lower directions from the inoculation point for both halves of the longitudinal cut, and the averages were statistically analyzed and compared depending on the type of pathogen and product formulation employed. All the data were compared with controls. Finally, grapevine plants removed and measured at the end of the assay were also processed (after taking measures) to re-isolate the different pathogens previously inoculated. Then, wood chips (0.5 cm long) exhibiting vascular necroses (1-2 cm around the wounds) were washed, surface sterilized, placed on PDA plates amended with streptomycin sulphate (to prevent bacterial contamination) and incubated at 26 • C in the dark in a culture chamber for 2-3 days to fulfil Koch's postulates.

Statistical Analyses
The results of the in vitro inhibition of mycelial growth of the three phytopathogens by the different concentrations of the treatments were statistically analyzed using oneway analysis of variance (ANOVA), followed by post hoc comparison of means through Tukey's test at p < 0.05 (provided that the homogeneity and homoscedasticity requirements were satisfied, according to the Shapiro-Wilk and Levene tests). In the case of the greenhouse assay results, since the normality and homoscedasticity requirements were not met, Kruskal-Wallis non-parametric test was used instead, with Conover-Iman test for post hoc multiple pairwise comparisons. R statistical software was used [42].

Vibrational Characterization
The assignments of the major absorption IR bands in S. marianum extract spectrum ( Figure S3) are presented in Table 2. The most prominent bands occurred at 3335, 1651-1602, 1457, 1313, 1242, and 1029 cm −1 . The band at 3335 cm −1 is attributed to phenolic (OH) vibrations; the multi-peak band at 1651 cm −1 to mixed (C=O) amide and (C=C) vibrations; the peak at 1515 cm −1 (typical of ferulic acid and vanillin) to >C=C< aromatic; the peak at 1457 cm −1 to symmetric aromatic ring stretching vibration (C=C ring); and the peaks at 1030 cm −1 and 779 cm −1 to C-O stretching and C=C, respectively (both vanillin-related peaks).

In Vitro Growth Inhibition Tests
The results of the mycelial growth inhibition tests are summarized in Figure 1. When tested alone, a higher efficacy of coniferyl alcohol as compared to ferulic acid could be observed: full inhibition was only reached for the former in the case of N. parvum and D. viticola. In the case of D. seriata, for which both treatments resulted in full inhibition, it was attained at a lower dose for coniferyl alcohol (1000 vs. 1500 µg·mL −1 ). Hence, the antifungal efficacy found for the extracts should be mostly ascribed to its main constituent. On the other hand, upon conjugation with stevioside, a clear enhancement in terms of efficacy was found in all cases, which was particularly evident in the case of the extracts, for which even higher inhibition than that of the coniferyl alcohol conjugates was attained at almost all concentrations against the three pathogens.
If effective concentrations are compared (Table 4), differences in the efficacy of the treatments as a function of the pathogen could be observed for some of the treatments more clearly: for instance, a slightly higher efficacy of stevioside and stevioside-S. marianum conjugate complex was found against N. parvum, and D. seriata seemed to be particularly sensitive to ferulic acid and its conjugate. On the other hand, the response of the three fungi to the coniferyl alcohol-based treatments was very similar. In concordance with the above statements, the calculation of synergy factors (Table 5) indicated a strong synergistic behavior for the stevioside-S. marianum conjugate, with SF values in the 2.7-5.1 range.    If effective concentrations are compared (Table 4), differences in the efficacy of the treatments as a function of the pathogen could be observed for some of the treatments more clearly: for instance, a slightly higher efficacy of stevioside and stevioside-S. marianum conjugate complex was found against N. parvum, and D. seriata seemed to be particularly sensitive to ferulic acid and its conjugate. On the other hand, the response of the three fungi to the coniferyl alcohol-based treatments was very similar.

Greenhouse Bioassays
Protective tests conducted on grafted plants with the stevioside-milk thistle treatment confirmed its efficacy in more realistic conditions (i.e., closer to field ones): the application of the conjugate complex resulted in statistically significant differences as compared to the positive (pathogen) controls in all cases (Table 6). It is worth noting that the median lengths of the vascular necroses were higher in the case of treated plants artificially inoculated with N. parvum than for treated plants artificially inoculated with the other two taxa (for which the effectiveness would be similar), which may be regarded as an unexpected result, provided that the associated EC 90 value was the lowest in the in vitro tests. This point was confirmed by including the fungus taxon in the statistical analysis as a second independent variable (Table S2). However, no statistically significant differences were observed among the three fungi in terms of necrosis lengths in the positive controls. Interestingly, in the two-factor analysis, the lengths of the necroses for the treated plants artificially inoculated with D. viticola were not significantly different from those of the negative controls, pointing to a particularly high inhibition of this pathogen. Table 6. Kruskal-Wallis test and multiple pairwise comparisons using the Conover-Iman procedure for the lengths of the vascular necroses for the three phytopathogen in greenhouse in vivo assays.

Valorization of Coniferyl Alcohol and Ferulic Acid
As expected from the phenological stage in which the plants were collected (flowering, before fruit ripening) and taking into consideration that the entire capitula were used for the extraction (not only the fruits), the panel of extracted components was different from those present in the commercially available milk thistle seed extract: instead of silybin (A and B) and isosilybin (A and B), coniferyl alcohol and other eugenol analogues were identified; and instead of vanillin, the quantitative presence of its precursor (ferulic acid methyl ester) was evidenced.
Coniferyl alcohol is a valuable chemical, which reaches 350 USD·g −1 when bought from commercial suppliers such as Sigma-Aldrich. Current approaches to obtain coniferyl alcohol are either inefficient, harmful (Penicillium simplicissimum (Oudemans) Thom vanillyl alcohol oxidase (PsVAO) can be used to produce it, but it intrinsically produces harmful byproduct H 2 O 2 ), or expensive (its synthesis involves expensive substrates and catalyst and harsh reaction conditions) [43,44]. These limitations can be overcome with the ultrasonic-assisted hydro-methanolic extraction of the capitula, reported in this paper, which may allow for the obtainment of the phenylpropanoid coniferyl alcohol with a yield of 50-80%. Alternative extractive approaches, such as the use of ionic liquid analogs (deep eutectic solvents) as extractive solvents [45], microwave-assisted extraction, dynamic maceration process [46], negative pressure cavitation-assisted extraction with macroporous resin enrichment [47], etc., should nonetheless be explored in order to optimize the yield.
In the case that the production of silymarin-based drugs is desired, the biotransformation of eugenol and coniferyl alcohol to silybin and isosilybin can be efficiently attained by the oxidation of the precursors by milk thistle ascorbate peroxidase (APX1), as shown in Figure 2a.
In the same way, the finding of a 10:1 ratio for the ferulic acid-vanillin pair confirms that, for S. marianum capitula during the flowering phenological stage in a hydromethanolic medium, the presence of the ferulic acid precursor is enhanced. Should vanillin be the desired chemical to obtain, the quantitative conversion of ferulic acid into vanillic acid could be feasible in presence of Pseudomonas spp. [48] (Figure 2b). The polypore species Pycnoporus cinnabarinus (Jacq.) P. Karst. has also been proposed for the production of vanillin from ferulic acid [49], although the vanillin produced is either rapidly converted to other products or utilized by the fungus as a source of carbon and energy. Genetic engineering has been applied to produce vanillin from ferulic acid using metabolically engineered Escherichia coli (Migula, 1895) Castellani and Chalmers, 1919 [50,51]. Another alternative would be the use of packed bed-stirred fermenters using Bacillus subtilis (Ehrenberg, 1835) Cohn, 1872 [52]. 21, 10, x FOR PEER REVIEW 11 of 16 [50,51]. Another alternative would be the use of packed bed-stirred fermenters using Bacillus subtilis (Ehrenberg, 1835) Cohn, 1872 [52]. pathway for conversion of ferulic acid into vanillin, according to [51].
It should be noted that the extraction of coniferyl alcohol and ferulic acid would not preclude the valorization of the rest of the biomass as a feedstock for bioenergy production [53][54][55].

Efficacy of the Treatments
Stevioside, a terpene glycoside obtained from Stevia rebaudiana (Bertoni) Bertoni, showed a high inhibitory activity, comparable to that of coniferyl alcohol. Since -to the best of the authors' knowledge-this is the first time that this compound is assayed against Botryosphaeriaceae fungi, no comparisons with similar taxa in terms of MIC values are available. However, the detected antifungal activity would be in good agreement with the results presented by other authors, who reported an inhibitory effect against other fungi (Alternaria solani Sorauer, Helminthosporium solani Durieu & Montagne, Aspergillus spp., Fusarium spp., Penicillium chrysogenum Thom, or Botrytis cinerea Pers., among others) [56][57][58][59][60], with MIC values varying over a wide range (from 250 to 3000 μg·mL −1 ).
With regard to the activity of S. marianum-derived phytochemicals, the antifungal activity of silymarin/silibinin against Candida spp. and its underlying mechanism has been studied by Yun and Lee [18,61] and Janeczko and Kochanowicz [62]. Fernández, et al. [63] found significant inhibition against Fusarium graminearum Schwabe for four flower defensins from milk thistle. Safarpoor, et al. [64] reported moderate antifungal activities of ethanolic extracts of milk thistle against C. albicans and Aspergillus oryzae (Ahlb.) Cohn. Some antifungal activity was also reported for leaf and flower ethanolic extracts by Keskin, et al. [65] against C. albicans. Nonetheless, in these two latter studies no details were provided about the phenological stage in which the plants were collected, and effective concentrations were not reported.
In relation to ferulic acid, it has been assayed against GTDs, and, according to Lambert, et al. [28], it is the phenolic acid with the strongest activity against D. seriata, N. parvum, E. lata, and P. chlamydospora. The same group, in a different study, found inhibition percentages in the 23-35% range for ferulic acid at a concentration of 500 μM Adapted from [44]. (b) Schematic representation of the non-β-oxidative pathway for conversion of ferulic acid into vanillin, according to [51].
It should be noted that the extraction of coniferyl alcohol and ferulic acid would not preclude the valorization of the rest of the biomass as a feedstock for bioenergy production [53][54][55].

Efficacy of the Treatments
Stevioside, a terpene glycoside obtained from Stevia rebaudiana (Bertoni) Bertoni, showed a high inhibitory activity, comparable to that of coniferyl alcohol. Since-to the best of the authors' knowledge-this is the first time that this compound is assayed against Botryosphaeriaceae fungi, no comparisons with similar taxa in terms of MIC values are available. However, the detected antifungal activity would be in good agreement with the results presented by other authors, who reported an inhibitory effect against other fungi (Alternaria solani Sorauer, Helminthosporium solani Durieu & Montagne, Aspergillus spp., Fusarium spp., Penicillium chrysogenum Thom, or Botrytis cinerea Pers., among others) [56][57][58][59][60], with MIC values varying over a wide range (from 250 to 3000 µg·mL −1 ).
With regard to the activity of S. marianum-derived phytochemicals, the antifungal activity of silymarin/silibinin against Candida spp. and its underlying mechanism has been studied by Yun and Lee [18,61] and Janeczko and Kochanowicz [62]. Fernández, et al. [63] found significant inhibition against Fusarium graminearum Schwabe for four flower defensins from milk thistle. Safarpoor, et al. [64] reported moderate antifungal activities of ethanolic extracts of milk thistle against C. albicans and Aspergillus oryzae (Ahlb.) Cohn. Some antifungal activity was also reported for leaf and flower ethanolic extracts by Keskin, et al. [65] against C. albicans. Nonetheless, in these two latter studies no details were provided about the phenological stage in which the plants were collected, and effective concentrations were not reported.
In relation to ferulic acid, it has been assayed against GTDs, and, according to Lambert, et al. [28], it is the phenolic acid with the strongest activity against D. seriata, N. parvum, E. lata, and P. chlamydospora. The same group, in a different study, found inhibition percentages in the 23-35% range for ferulic acid at a concentration of 500 µM (97 µg·mL −1 ) against different N. parvum isolates [69]. Gómez, et al. [29] reported half maximal effective concentrations of 3530 and 4740 µg·mL −1 against Botryosphaeriaceae sp. and P. minimum, and Dekker, et al. [30] found an EC 50 value of 15 mM (2913 µg·mL −1 ) against Botryosphaeria sp. Srivastava, et al. [70] found that ferulic acid at 25 mM (4855 µg·mL −1 ) resulted in ca. Regarding the conjugate complexes with stevioside, no data is available against GTDs. The most similar assayed product would be the stevioside:silymarin conjugate complexes (in a 1:1 molar ratio) tested against Fusarium culmorum (Wm.G. Sm.) Sacc., for which an EC 90 value of 160 µg·mL −1 and a synergy factor of 1.43 were reported [71]. No antifungal efficacy data is available for stevioside-coniferyl alcohol conjugate complexes, but steviosideferulic acid inclusion compounds (with different molar ratios to the one assayed herein, and involving a more complex preparation procedure) have been tested against F. culmorum and Phytophthora cinnamomi de Bary. In the former case, composites based on stevioside:ferulic acid inclusion compounds (in a 5:1 molar ratio), combined with chitosan oligomers in hydroalcoholic solution or in choline chloride:urea deep eutectic solvent media, led to EC 90 values in the 377-713 µg·mL −1 range against F. culmorum [72], depending on the dispersion medium. In the case of P. cinnamomi, inclusion compounds from stevioside and ferulic acid in 6:1 ratio, dispersed in a hydroalcoholic solution of chitosan oligomers, resulted in EC 90 values of 446-450 µg·mL −1 (depending on the presence/absence of silver nanoparticles) [73,74].

Conclusions
In the hydromethanolic extract of Silybum marianum capitula, during the flowering stage, high contents of coniferyl alcohol derivatives and ferulic acid esters were found, instead of other chemical species such as the silymarin complex or vanillin. Given the high price of coniferyl alcohol, this may pose an alternative valorization strategy for this weed, compatible with a subsequent valorization for bioenergy purposes. Concerning the antifungal activity of the hydroalcoholic extract, the EC 50 and EC 90 values obtained against the three studied Botryospheriaceous grapevine pathogens (N. parvum, D. viticola and D. seriata) were in the 557-1088 and 1461-9942 µg·mL −1 range, respectively. However, a significant efficacy enhancement (with EC 50 and EC 90 values in the 87-148 and 303-596 µg·mL −1 , respectively) was obtained by formation of conjugate complexes of the hydrometanolic extract of S. marianum with stevioside, evidencing a clear synergistic behavior (with synergy factor values of up to 5.1) as a result of the solubility and bioavailabity improvement. The efficacy of the stevioside-milk thistle conjugate complexes was further assessed in artificially inoculated grafted plants, obtaining significant differences in the vascular necroses lengths vs. the positive controls in all cases. The presented results support the possibility of extending the applications of milk thistle to agriculture as an antifungal agent, in particular for the protection of grapevines against certain fungal trunk diseases.
Supplementary Materials: The following are available online at https://www.mdpi.com/article/10 .3390/plants10071363/s1, Table S1. Repetitions for each of the plant/treatment/pathogen combinations in the greenhouse bioassay; Table S2. Kruskal-Wallis test and multiple pairwise comparisons using the Conover-Iman procedure for the lengths of the vascular necroses in greenhouse in vivo assays considering two independent variables (treatment and taxa); Figure S1. Biosynthesis of silybins from taxifolin and coniferyl alcohol; Figure S2. Formation of coniferyl alcohol; Figure S3. Infrared spectrum of S. marianum extract (after lyophilization); Figure S4 Funding: This research was funded by Junta de Castilla y León under project VA258P18, with FEDER co-funding; by Cátedra Agrobank under "IV Convocatoria de Ayudas de la Cátedra AgroBank para la transferencia del conocimiento al sector agroalimentario" program; and by Fundación Ibercaja-Universidad de Zaragoza under "Convocatoria Fundación Ibercaja-Universidad de Zaragoza de proyectos de investigación, desarrollo e innovación para jóvenes investigadores" program.

Institutional Review Board Statement: Not applicable.
Informed Consent Statement: Not applicable.

Data Availability Statement:
The data presented in this study are available on request from the corresponding author. The data are not publicly available due to their relevance to be part of an ongoing Ph.D. Thesis.