Optimized Enzymatic Extraction of Phenolic Compounds from Verbascum nigrum L.: A Sustainable Approach for Enhanced Extraction of Bioactive Compounds

Abstract

Verbascum nigrum, commonly known as black mullein, is widely used in traditional medicine for its expectorant, mucolytic, sedative, and diuretic properties. This study aimed to develop and optimize a standardized method for extracting phenolic compounds from V. nigrum using enzymatic pretreatment followed by solvent extraction. Enzymatic treatment does not rely on harmful solvents and is a low energy-intensive process, making it a suitable green technology for the food, cosmetic, and pharmaceutical industries. The research explored the use of different lignocellulolytic enzymes, including pectinase, cellulase, α-amylase, and xylanase, to break down plant cell walls, enhancing the release and bioaccessibility of active compounds. The two-step extraction process proposed combined enzymatic pretreatment and hydroalcoholic extraction, resulting in a considerably improved yield of phenolic compounds (24 mg/g DM). Analytical characterization using a high-performance liquid chromatography (HPLC) system coupled with a diode-array-detector (DAD) and ultra-high-performance liquid chromatography (UHPLC) coupled with DAD and tandem mass spectrometry (MS/MS) revealed a higher concentration of target bioactive compounds in enzymatically treated extracts compared to traditional methods, including phenolic derivatives (e.g., caffeic acid, p-coumaric acid, and verbascoside), and flavonoids (e.g., luteolin). Up to 22 phenolic and flavonoid compounds were characterized. This study provides new insight into the potential of enzymatic extraction as a green and efficient alternative to conventional extraction methods, for the production of high-quality herbal products richer in (poly)phenolic compounds, highlighting its potential for industrial applications.

1. Introduction

The practice of herbal medicine, intended as the use of plants or plant materials to produce biologically active natural products possibly useful to manage various diseases, is the oldest form of healthcare and is relied upon by approximately 80% of the population, particularly in developing countries [1,2]. Despite the great advances achieved in modern medicine, plants still make an important contribution to healthcare due to their diverse chemical constituents, which may offer poly-pharmacological benefits. Indeed, plants can be considered as “living factories”, synthesizing a wide range of chemical compounds. These compounds include primary metabolites, essential for plant growth, such as amino acids, proteins, and carbohydrates, as well as secondary metabolites, including alkaloids, terpenes, and (poly)phenols [3,4]. The interplay among these components can result in a synergistic effect in the final product [5,6,7].

The reliability and consistency in the quality of herbal products are fundamental to their efficacy and safety [8]. Given the inherently variable and complex nature of plant-derived products, which often contain numerous biologically active components that are not fully identified, health-promoting outcomes and safety profiles can vary significantly among products, even within the same class. This underscores the need for standardized extraction processes to ensure consistent efficacy and safety of herbal products.

Verbascum species are used externally for wound desiccation, and as antiseptic, astringent, demulcent, emollient, narcotic, antimalarial agent, and for the treatment of inflammation, migraine, asthma, and spasmodic cough [9,10,11]. Among Verbascum genus, V. nigrum L., the black mullein or dark mullein, also exhibits expectorant, mucolytic, sudorific, sedative, and diuretic activities [12]. Traditionally, Verbascum species contain different classes of biologically active compounds, including flavonoids, which may be responsible for its various health benefits [13,14]. Apigenin, one of the most abundant flavonoids in Verbascum species, is reported to be a potent anti-inflammatory compound, which may explain the traditional use of Verbascum as a herbal drug against inflammation. Verbascoside, a phenylethanoid glycoside, exhibits strong antioxidant activity and has been identified as a major contributor to the antioxidant capacity of Verbascum extracts [15,16]. In addition, verbascoside is a compound scarcely present in human diet, which makes it extremely interesting as a nutraceutical compound. For these reasons, different approaches are employed to extract and characterize the active molecules present within V. nigrum targeting high extraction yields and efficacy [12,13,17,18].

Lignocellulolytic enzymes play a crucial role in breaking down the components of the plant cell wall, thereby releasing the active compounds contained within the cells. These enzymes are capable of degrading the major polymers in lignocellulose, namely cellulose and hemicellulose [19], thereby increasing the bioaccessibility of plant bioactives and making them available for downstream applications. Enzymatic treatment is considered a ’green’ technology because it avoids the use of harmful solvents and energy-intensive reaction conditions that are often associated with traditional chemical processes. Thus, the use of enzymes results in safer processes and more environmentally friendly products, which is particularly important in industries where the final products are intended for human consumption or contact, such as the food, cosmetics, and pharmaceutical sectors [20].

The aim of the present work was to develop and optimize a standard method for the extraction of (poly)phenolic compounds from V. nigrum L. that entails both an enzymatic pretreatment and a solvent extraction. In addition, to explore the possibility of further improving the release of the phytocompounds, the enzymatic cocktail was optimized by studying the role of each enzyme and the potential synergies between them.

2. Materials and Methods

2.1. Plant Material Processing

Plant biomasses of V. nigrum were cultured in a Vertical Farm in optimal agro-climatic conditions, kindly supplied by Perfect srl (Milan, Italy), and stored at −20 °C. To carry out the experiments, a portion of each biomass was pre-dried in a ventilated oven at 45 °C for 24 h. The dried biomass was then crushed and sieved to obtain particles of approximately 1 mm in diameter (Figure 1). The dry matter (DM) content of each biomass was then determined [21].

2.2. Enzyme-Assisted Hydroalcoholic Extraction

2.2.1. Development of the Two-Step Extraction Method

A two-step extraction method, including a first enzymatic treatment and a subsequent solvent (hydroalcoholic) extraction, was developed based on studies previously published [22,23].

The enzyme cocktail used for pre-treatment consists of four enzymes: cellulase (EC 3.2.1.4), xylanase (EC 3.2.1.8), pectinase (EC 3.2.1.15), and amylase (EC 3.2.1.1). Such enzymes are often used in combination to promote extraction of bioactive components from lignocellulosic biomasses [22,23,24,25,26]. The following enzymes were purchased from Creative Enzymes (Shirley, NY, USA):

• Pectinase (Cat No: DIS-1030; Lot No: ECD0021415; activity: 30,240 U/g)
• Cellulase (Cat No: DIS-1017; Lot No: ECD7011006; activity: 20,340 U/g)
• Mid-temperature refining α-amylase for beer (Cat No: BER-1513; Lot No: ECB3081602; Activity: 4230 U/g)
• Xylanase (Cat No: DIS-1032; Lot No: ECD2022310; activity: 20,280 U/g)

All the enzymes were mixed to obtain an enzyme cocktail wherein they all had equal activity. The assays for the development of the two-step extraction method were carried out under the following conditions, established based on previous works [22,23,26]:

• Enzyme treatment: biomass concentration of 40 g DM/L. Temperature of 47.5 °C. Citrate buffer (0.1 M) at pH 5.5. Treatment time of 24 h. Enzyme/biomass mixture ratios of 15% and 30% were tested (corresponding to 100 and 200 U/g DM of biomass). Notably, high enzyme loads were studied compared to the 5–20 U/g of dry biomass previously reported [22,23,25] to ensure that enzyme concentration would not be limiting the outcome of the pretreatment (Figure 1).
• Extraction after enzyme treatment: biomass concentration of 16 g DM/L. Temperature of 25 or 35 °C. Solvent: ethanol/water (60:40 v/v) or pure ethanol. Extraction time of 2, 5, or 24 h.

Enzymatic digestion and solvent extraction were carried out in 15 mL tubes, employing a mixing wheel to mix the contents. All experiments were carried out in duplicate (Figure 1).

2.2.2. Optimization of the Enzymatic Reaction Conditions

In order to determine the optimal conditions for enzymatic digestion, an experimental design approach was applied using MODDE^® 13 software (Sartorius AG, Göttingen, Germany). This software allows the development of mathematical models to identify the influence of experimental parameters (pH, temperature, time, and enzyme/substrate ratio) on the variables of interest (extraction yield of total phenolic compounds and total flavonoids).

Experimental tests to optimize the enzymatic digestion process were carried out under the following conditions: Biomass concentration of 20 g DM/L; temperature of 40 °C, 47.5 °C or 55 °C; citrate buffer (0.1 M) at pH 4.5, 5.5 or 6.5; treatment time of 2, 5, or 24 h; enzyme/substrate mixture ratio of 0, 7.5, or 15 wt% (corresponding to 0, 50, and 100 U/g DM of biomass).

The objective of this study was set as “screening” with the primary model focusing on the interaction effects. Specifically, a D-optimal design (

Table 1. Experimental design developed with MODDE^®.

Exp No	pH	Temperature (°C)	Time (h)	Enz. (wt%)
1	6.5	40	2	0
2	5.5	55	2	0
3	4.5	40	24	0
4	6.5	55	24	0
5	4.5	55	2	15
6	6.5	55	2	15
7	5.5	40	2	15
8	4.5	40	24	15
9	6.5	40	24	15
10	5.5	55	24	15
11	4.5	40	2	7.5
12	6.5	40	24	7.5
13	4.5	55	24	7.5
14	6.5	47.5	5	7.5
15	5.5	47.5	5	7.5
16	6.5	40	2	0
17	5.5	55	2	0
18	4.5	40	24	0
19	6.5	55	24	0
20	4.5	55	2	15
21	6.5	55	2	15
22	5.5	40	2	15
23	4.5	40	24	15
24	6.5	40	24	15
25	5.5	55	24	15
26	4.5	40	2	7.5
27	6.5	40	24	7.5
28	4.5	55	24	7.5
29	6.5	47.5	5	7.5
30	5.5	47.5	5	7.5

) was used, consisting of 30 experimental runs. D-optimal designs are computer-generated and maximize the determinant of the X’X matrix, ensuring optimal distribution of experimental points. The selected design was chosen for its high statistical “power”, representing the DOE’s relative ability to detect significant effects, which was 98 for this study, indicating excellence reliability. The model was then fitted with Multiple linear regression (MLR) model.

In this case, the solvent extraction step following the enzymatic step consisted of a biomass concentration of 8 g DM/L, temperature 25 °C, first solvent step with ethanol/water (60:40 v/v) and second solvent step with pure ethanol, and an extraction time of 2 h.

2.2.3. Optimization of the Enzymatic Cocktail Composition

Enzymatic pretreatment (conditions referred to as EH) was conducted in citrate buffer (0.1 M, pH 4.5, with sodium citrate and citric acid purchased from Sigma Aldrich, Milan, Italy. The DM concentration used for each test was 40 g/L. Enzymes were added in amounts of 100 or 200 U/g DM (units per gram of DM). In addition to this, control experiments without enzyme addition were also performed. Pretreatments were performed using a mixing wheel at a temperature of 55 °C for 24 h.

Different combinations of enzymes were investigated for the pretreatments: cellulase (C); xylanase (X); pectinase (P); amylase (A); cellulase and xylanase (C + X); cellulase and amylase (C + A); cellulase and pectinase (C + P); xylanase and amylase (X + A); xylanase and pectinase (X + P); amylase and pectinase (A + P); cellulase, xylanase and amylase (C + X + A); cellulase, xylanase and pectinase (C + X + P); xylanase, amylase and pectinase (X + A + P); cellulase, amylase and pectinase (C + A + P); and cellulase, xylanase, amylase and pectinase (C + X + A + P).

2.2.4. Hydroalcoholic Extraction

After the optimization of the enzymatic pretreatment, hydroalcoholic extraction was performed. Pure ethyl alcohol (Carlo Erba, Milan, Italy) was added to the product obtained by enzymatic treatment, until a final alcohol concentration of 60 vol% was reached. Hydroalcoholic extractions were conducted using a mixing wheel, at room temperature, for 2 h.

To highlight the role of enzymatic pretreatment, for some experimental conditions (referred to as EH) hydroalcoholic extraction was not conducted. In addition, to highlight the role of hydroalcoholic extraction, experiments of only hydroalcoholic extraction (referred to as HA) were also conducted.

2.3. Determining the Total Phenolic Content (TPC) and Total Flavonoid Content (TFC)

TPC was determined using Folin–Ciocalteu’s assay [27,28]. Calibration curves were constructed using gallic acid (Sigma Aldrich). Briefly, 20 μL of sample (either standard or blank) was introduced into the well of a plate, followed by dilution with 60 μL of demineralized water, and of 20 μL of Folin–Ciocalteu reagent. The reaction was achieved in 5 min, after which 100 μL of sodium carbonate (80 g/L) was added to stop the reaction. It was diluted with 100 μL of demineralized water. The mixture was left to stand in the dark for 30 min. Finally, a spectrophotometer (Infinite M Nano⁺, TECAN, Milan, Italy) was used to determine the absorbance of the mixture at 765 nm. This value was corrected by subtracting the absorbance of the blank (i.e., the pure solvent). The absorbance values were correlated with the concentration of phenolic compounds based on a calibration line made with gallic acid at different concentrations.

TFC was determined by reaction with aluminum chloride, using a method widely known in the literature [28]. Calibration curves were constructed using rutin (Sigma Aldrich). Then, 50 μL of sample (either standard or blank) was added to 150 μL of aluminum chloride (20 g/L). After stirring, the reaction was achieved in 60 min. Finally, a spectrophotometer was used to determine the absorbance of the mixture at 420 nm. The absorbance values were correlated with the concentration of phenolic compounds based on a calibration line made with rutin standards at different concentrations.

The extraction yield of the active compounds was calculated in relation to the dry weight of the biomass, according to the following equation:

$$Yield(\mathrm{m}\mathrm{g}/\mathrm{g}\mathrm{D}\mathrm{M})={\displaystyle \frac{m_{product}\left(\mathrm{m}\mathrm{g}\right)}{m_{DMbiomass}\left(\mathrm{g}\mathrm{D}\mathrm{M}\right)}}$$

(1)

2.4. Chromatography

2.4.1. HPLC-DAD

For the most promising experimental conditions in terms of TPC and TFC, the chemical species in the extracts were identified and quantified to understand the role of enzymatic treatment with respect to the preferential release of specific compounds. The presence of common phenolic acids, flavonoids, and sesquiterpenes in mullein extracts was evaluated using the following analytical standards: gallic acid, chlorogenic acid, vanillic acid, ferulic acid, p-coumaric acid, syringic acid, caffeic acid, 3,5-dicaffeoylquinic acid, quercetin, rutin, apigenin, naringin, and verbascoside. (Sigma Aldrich, Milan, Italy).

The preliminary identification and quantification of such chemical species present in the extracts was conducted using HPLC-DAD (Shimadzu Italia, Milan, Italy). A LiChroCART 250-4 column, Purosphere STAR RP-18e (250 × 4.6 mm, 5 μm), was used for HPLC-DAD analysis. The mobile phases were formic acid (0.2 vol% in water, (Phase A) and acetonitrile with 0.2 vol% formic acid (Phase B). The flow rate in the column was 1 mL/min. The starting gradient consisted of 90% phase A, maintained for 2 min and then decreased to 35% at 15 min. From 15 to 19 min phase A was reduced to 0% until 35 min; then the starting gradient at 90% of A was restored. Phenolic compounds were analyzed using a wavelength of 254 nm, while flavonoids were analyzed using a wavelength of 350 nm [29].

2.4.2. HPLC-DAD/MS

Further chemical characterization was carried out by HPLC-DAD/MS analysis, using a UHPLC JASCO X-LC system (Lecco, Italy) coupled with a HESI-LC/MS (Heated Eletctrospray Ionization-Liquid Chromatoraphy/Mass Spectrometry). The analyses were conducted using a Restek Raptor ARC-18 column (1.8 μm, 100 × 2.1 mm), coupled with a diode detector (detection wavelength: 200–800 nm) and an ion trap mass spectrometer (MS) with a HESI source (LTQ XL by Thermo Scientific, Waltham, MA, USA). The mobile phases and the gradients used during the analyses was the same reported in the previous section. Mass spectrometer settings were as follows. HESI Probe: Gas = N₂, T = 95 °C, Voltage = 3.2 kV; Capillary T = 275 °C, Voltage = 48 V, Tube Lens = 72 V. Tune Settings: Multipole 00 Offset = 2.4 V, Lens 0 = −4.26 V, Multipole 0 Offset =-5.18 V, Lens 1 = −8.95 V, Gate Lens = −65.1 V, Multipole 1 Offset = −6.3 V, Multipole RF Amplitude (p-p)= 400 V, Front Lens = −6.1 V. Where possible, the identity of each peak present in the obtained chromatograms was confirmed by comparison with commercial standards. Alternatively, peak identification was conducted on the basis of a comparison of experimentally obtained mass spectra with those available in the literature [17,29,30]. Semiquantitative analysis was performed by comparing peak areas in the chromatograms of the extracts with corresponding peak areas in the chromatograms of the control samples (to which no enzymes were added). Specifically, the reference wavelength of 254 nm was chosen for phenolic acids, while the reference wavelength of 330 nm was chosen for flavonoids [31,32,33].

2.4.3. UHPLC-DAD/MS/MS

After the chemical characterization of the extracts, seven promising samples, particularly rich in target bioactive compounds (

Table 2. Samples of V. nigrum investigated to comparative analysis.

Sample Code	Biomass	Pretreatment
VE45E	Verbascum nigrum	EH + HA
VE45F	Verbascum nigrum	EH + HA
63	Verbascum nigrum	EH + HA
65	Verbascum nigrum	EH
VN45E	Verbascum nigrum	H + HA
VN45F	Verbascum nigrum	H + HA
29	Verbascum nigrum	HA

) were selected, and further analyzed (as described below) to further inspect the phytochemical profiles. Samples VE45E and VE45F, as well as samples VN45E and VN45F, were treated as duplicates.

Each extract reported in Table 2 was 5-fold diluted with a mixture of water/methanol 1:1 (v/v) acidified with 0.1% formic acid. After vortexing, each extract was centrifuged at 7378 g for 5 min at 10 °C, before UHPLC-DAD-MS/MS analysis.

For the mobile phase, a binary system of water (phase A) and acetonitrile (phase B) was used, both acidified with 0.01% formic acid. The starting gradient consisted of 99% phase A, maintained for 0.5 min and then decreased to 85% at 3 min. From 3 to 6 min phase A was reduced to 50%, then dropped rapidly to 5% at 9 min to clean the column. At 11 min the starting gradient at 99% of A was restored and maintained until 14 min to guarantee a complete equilibration of the column for the next analysis. The mobile phase pumped at 0.4 mL/min. The column compartment was set at 40 °C, while the autosampler temperature was set at 10 °C. An Acquity UPLC HSS T3 (2.1 × 100 mm) equipped with a VanGuard Acquity UPLC HSS T3 (2.1 × 5 mm) guard column was used as the chromatographic column (Waters, Milford, MA, USA). The photodiode array detector scanned in a range of 220–500 nm. The injection volume was 2 μL.

Different spectrometric modalities were used to characterize the phenolic components in the various extracts. Initially, extracts 29, 63, 65, and VN45E were analyzed in MS Scan mode both in ESI+ and ESI−, scanning from m/z 100 to 1500. In negative mode, the capillary voltage was set to 2.3 kV, the cone voltage was 60 V, with the source and desolvation T set at 150 °C and 600 °C, respectively. Nitrogen has been used as nebulizer gas at P equal to 7 bar. In positive mode, 1 kV was applied on the capillary with a cone voltage set at 80 V, while the other MS parameters were the same used for negative mode. Ultra-pure argon has been used as collision gas. Following the MS Scan analyses, both the most abundant ions and the ions inherent in chromatographic peaks with increased/decreased ion abundance between the various samples were fragmented in MS/MS mode (collision energy set to 30 eV). This led to obtaining the peak identification or at least to ascribe each chromatographic peak with a specific (poly)phenolic subclass. In addition, the extracts 29, 63, 65, and VN45E were analyzed in ESI by monitoring the precursor ions of the fragment ions with m/z 163, 179, and 193, corresponding to coumaric, caffeic, and ferulic acids, respectively. This analytical approach was used since these hydroxycinnamic acids are particularly abundant in Verbascum spp. as glycosides and as esters of iridoids and phenylethanoids [34,35].

As a result of the comparative analyses mentioned above, a final method in Multiple Reaction Monitoring (MRM) mode was developed, which is the most sensitive and selective method to perform a comparison of the main phenolic compounds in the extracts reported in Table 2. Each diluted extract was analyzed twice.

2.5. Statistical Analysis

TPC and TFC were statistically analyzed by analysis of variance (ANOVA) by Tukey’s multiple comparisons test, through the GraphPad Prism 10 software (Boston, MA, USA).

Ion abundance levels of components analyzed through LC-MS/MS were reported as mean values ± SD. Analysis of variance (ANOVA) was carried out through Tukey’s HSD test using the IBM SPSS Statistics 19 software package (SPSS Inc., Chicago, IL, USA).

3. Results

Enzymatic treatment is commonly adopted for enhancing the degradation of lignocellulose, making the active compounds present in plant matrices accessible and therefore more readily extractable [36]. Since most (poly)phenols are poorly soluble in aqueous media, an extraction step may be applied after the enzymatic treatment [37], to maximize their recovery. In the first phase of this work, the effect of alcoholic or hydroalcoholic extraction after enzymatic treatment of V. nigrum was studied with the aim to establish the optimal extraction sequence for maximizing the extraction yields of phenolic and flavonoid components.

3.1. Development of the Extraction Method

3.1.1. Preliminary Analyses of the Products Extracted

To preliminarily investigate the effect of enzymatic treatment, extracts obtained by hydroalcoholic extraction alone and those obtained by enzymatic treatment followed by hydroalcoholic extraction were analyzed by HPLC. Chromatograms comparison showed that the extract obtained by enzymatic treatment appeared richer in specific phenolic compounds compared to the extract obtained from hydroalcoholic extraction alone (Figure S1).

Such observation may be explained by the accelerated release of phenolic compounds thanks to the enzymatic action. Indeed, when the enzymatic treatment was applied, the yield of TPC reached its maximum value at 2 h after the initiation of the hydroalcoholic extraction (Figure 2A,B). Conversely, when the hydroalcoholic extraction alone was applied, the same yield was not achieved even after 24 h of extraction. Therefore, a 2 h hydroalcoholic extraction after enzymatic treatment was sufficient to maximize the extraction yield of the active compounds. A similar behavior was found for flavonoid extraction: the TFC yield was improved when the hydroalcoholic extraction was coupled with enzymatic treatment, for all the hydroalcoholic extraction durations tested.

Moreover, by assaying two different enzymatic concentrations (15 and 30 wt% relative to the weight of dry mass), no remarkable differences were observed, indicating that the lower concentration is sufficient for an effective enzymatic treatment.

3.1.2. Evaluation of the Best Process Sequence

Since hydrophilicity of phenolic and flavonoid compounds is known to vary substantially from species to species the next part of our work was designed to optimize the extraction sequence after the enzymatic treatment. A series of extraction steps were applied to the biomass after enzymatic treatment, monitoring the extraction yield after each step.

The graphs in Figure 3 show that the enzymatic treatment (referred as to EH(H₂O)) of V. nigrum, followed by hydroalcoholic extraction (referred as to HA (EtOH + H₂O)) and one cycle of alcoholic extraction (referred as to EE I (EtOH)), which allowed a recovery of phenolic compounds (Figure 3A) and flavonoids (Figure 3C) equal to 24 mg/g DM and 25 mg/g DM, respectively. Indeed, additional alcoholic extraction steps (referred as to EE II and EE III) did not result in substantial increase in extraction yields, as further shown in Figure 3B,D. Therefore, all subsequent experiments used this identified optimal sequence for recovery of active compounds: enzymatic treatment followed by hydroalcoholic extraction and subsequent alcoholic extraction.

3.2. Optimization of the Enzymatic Treatment

The following results highlight the best conditions for the enzymatic treatment in terms of conditions of the reaction and of enzymatic cocktail composition.

3.2.1. Optimization of the Enzymatic Reaction Conditions

The optimization phase of the enzyme treatment aimed at identifying the pH, temperature, duration, and enzyme concentration conditions that would maximize the subsequent extraction yield of phenolic compounds and flavonoids. For each experimental condition (Table 1), the results are shown in Figure 4 and detailed in Table S1. Analysis of the experimental results performed with the MODDE^® software led to the identification of the trends for the extraction yield of TPC and TFC: the release of TPC was favored by an enzymatic treatment carried out at pH 4.5, 55 °C, 2 or 24 h, and with an enzyme concentration of up to 15 wt%. The release of TFC was enhanced by enzymatic treatments carried out at pH 6.5, 55 °C, 24 h, and 15 wt% of enzymes.

3.2.2. Optimization of the Enzymatic Cocktail Composition

The optimization of the enzymatic cocktail was performed to identify the best combination for an improved extraction. TPC and TFC were determined after the application of different enzymatic combinations, as well as at different enzymatic concentrations. The yield of TPC and TFC was analyzed to select the best conditions (Table S2). By coupling the enzymatic pretreatment and the hydroalcoholic extraction (EH + HA), the extraction yield seemed higher than the single extractants (EH or HA). No differences were appreciated among enzymes at different concentrations (100 and 200 U/g DM). This suggested that the concentration of 100 U/g DM could be sufficient to pretreat the plant material Table S2. All the enzyme combinations seemed to display similar performances, highlighting that all four enzymes could have an action on the structural breakdown of plant matrices. Nevertheless, the enzymes contributing most to the extraction yield of phenolics and flavonoids appeared to be mainly cellulase and pectinase, suggesting that the combined use of cellulase and pectinase could lead to an effective enzyme cocktail for extraction of bioactive compounds from V. nigrum (Table S2).

Noteworthy higher extraction yields have been reported by other authors using conventional extraction techniques such as homogenizer-assisted extraction (TPC up to 42 mg/g DM and TFC up to 39 mg/g DM for an extraction in methanol [38]) or maceration (TPC up to 61 mg/g DM and TFC up to 50 mg/g DM for maceration in methanol [39]), with most studies reporting an extremely high variability in extraction yields from different Verbascum species, highlighting the effect of biomass composition [31,40]. Importantly, several works exploring the use of other green methods for extraction of bioactive compounds, such as supercritical fluid extraction, pulsed electric field extraction, microwave-assisted extraction or ultrasound-assisted extraction, showed extraction yields in the range of 2–120 mg/g DM [40], suggesting that the tandem enzymatic and hydroalcoholic extraction proposed in the present work can be a valuable methodology.

The extracts obtained for each condition were then analyzed through HPLC-DAD/MS, to preliminarily highlight potential compound enrichment. Five peaks were observed with or without the preliminary enzymatic treatment. Quantification of verbascoside was performed for different enzyme combinations, as a proxy for extraction efficiency. The extraction yield of the analyzed compound could generally be improved by coupling the enzymatic treatment with the hydroalcoholic extraction, if compared to the enzymatic treatment or the hydroalcoholic extraction alone.

3.3. Characterization of the (Poly)Phenolic Compounds of V. nigrum Extracts

Table 3. LC-DAD-MS/MS characteristics of the main (poly)phenolic/phytochemical compounds identified in Verbascum nigrum.

Compound	R.T.	λmax (nm)	ESI Mode	[M-H]−/[M]+ (m/z)	MS2 Fragment Ions (m/z)
p-Coumaric acid *	4.83	309	-	163	119
Caffeic acid *	4.28	323	-	179	135, 107
Unknown I	0.68	220	-	278	113, 193, 158, 160
Luteolin *	5.90	345	-	285	133, 151, 175, 199, 217, 213
Caffeic acid-hexoside-rhamnoside	3.53	328	-	487	179, 135, 161
Ferulic acid-hexoside-rhamnoside	4.20	303	-	501	193, 134, 149
Unknown III	2.09	220	-	555	129, 157, 139, 365
Caffeic acid derivative I	3.50	330	-	619	179, 135
Caffeic acid derivative II	4.97	324	-	621	161, 179, 135
Caffeic acid derivative III	5.07	324	-	621	161, 179, 135
Verbascoside *	4.86	315	-	623	161, 461, 135, 179, 133
Caffeic acid derivative IV	3.15	328	-	635	473, 291, 309, 141, 179, 135
Caffeic acid derivative V	4.56	328	-	667	179, 343, 161, 135
Unknown IV	6.73	241, 361	-	673	249, 209, 307, 231, 221, 163
Coumaroyl-acetylrhamnosyl-aucubin	5.34	266	-	679	187, 145, 163, 205, 309, 119, 499, 161
Caffeoyl-acetylrhamnosyl-aucubin	5.08	333	-	695	161, 203, 135, 179, 325, 653, 515
Forsythoside B	4.61	329	-	755	593, 161, 133, 179, 135, 447, 125
Ferulic acid derivative	5.21	323	-	783	175, 193, 149, 125, 160, 134
Unknown V	2.41	220	-	929	455, 353, 689, 809, 293, 851, 767, 191, 113
Unknown VI	2.86	340	-	987	467, 365, 701, 323, 779, 353, 305, 191, 941
Unknown VII	3.81	252	+	557	131, 203, 395

* Unambiguous identification by comparison with the corresponding standard compound. The fragment ions have been reported in decreasing order of ion abundance. The base MS² ions used in the MRM analysis are shown in bold.

showed the LC-DAD-MS/MS characteristics of the 22 compounds detected. In detail, (poly)phenolic compounds mostly belonged to the phenolic acid subclass of hydroxycinnamic acids, with the flavone luteolin as the only flavonoid identified in the V. nigrum extracts. Luteolin, verbascoside, caffeic acid, and p-coumaric acid were unambiguously identified by comparison with their corresponding standard compounds. The exact structure of several hydroxycinnamic acid derivatives could not be assigned, but the fragment ions of caffeoyl and feruloyl groups led to their tentative identification as caffeic acid or ferulic acid derivatives. The identification of hydroxycinnamic acid derivatives was helped by DAD, as their maximum absorption peaks were close to 325 nm, a common feature for hydroxycinnamic acids. Besides (poly)phenolic compounds, other unknown compounds were detected. We argue that these species are unlikely to be (poly)phenols because of their UV absorption below 260 nm.

The ion abundance of relevant compounds mentioned above is reported in Figure 5. The LC-MS/MS traces of some (poly)phenols are shown in Figure S2 and ion abundances of the unknown compounds in Figure S3.

The data reported in Figure 5 show that verbascoside levels were lower in all samples compared to the control sample 29, while luteolin remained at similar levels in all the samples except for sample 65, where a drastic reduction was observed, indicating that a subsequent hydroalcoholic extraction could be beneficial in increasing the extraction yield. Both forsythoside B and caffeoyl-acetylrhamnosyl-aucubin showed lower concentrations in all samples in comparison to control 29 (HA extraction alone), suggesting that the enzymatic treatment could negatively affect the recovery of such compounds. Most of the other hydroxycinnamic acids showed notable higher values in enzyme-treated samples than in both control sample 29 and samples VN45E-F. The highest abundance of these compounds was observed in particular in sample 65.

4. Discussion

Conventional methods for extracting bioactive compounds from plants often involve the use of organic solvents, high temperatures, and long extraction times. While effective, these methods can have significant environmental impacts and high energy costs. Among conventional methods there are solvent extraction, ultrasound-assisted extraction, microwave-assisted extraction, and supercritical CO₂ extraction [40,41,42,43]. On the other hand, enzymatic extraction represents a “green” method for extracting active compounds from plant materials.

The enzymatic approach utilizes specific enzymes to degrade plant cell walls allowing the release of the active compounds contained within the cells. The enzymes used in the present work include cellulases, xylanases, pectinases, and amylases, which may work synergistically. Indeed, enzymatic combinations can improve extraction yields of phenolics due to their wide spectrum of action, allowing the hydrolysis of different components of the cell wall and increasing the soluble phenolic fraction, and favoring the release of free phenolic acids [44]. Enzymatic extraction is considered environmentally friendly as it avoids the use of toxic solvents and reduces energy consumption, making the process safer for the environment and end-users. For these reasons, several researchers are currently working to optimize enzymatic-assisted extraction protocols from different plant sources [45,46].

In the specific case of Verbascum, conventional extraction methods often involve the use of hydroalcoholic solvents for prolonged periods of time to maximize the recovery of phenolic compounds and flavonoids [39,47,48]. However, these methods can be improved by integrating a preliminary enzymatic treatment. In this work we demonstrated the effective and increased extraction of (poly)phenols, flavonoids, and other phytocomplexes from V. nigrum using a novel enzymatic treatment followed by traditional solvent extraction. The application of an enzymatic treatment followed by hydroalcoholic extraction accelerated the release of phenolic compounds, achieving maximum yields in only 2 h compared to the >24 h required for hydroalcoholic extraction without enzymatic treatment. These results confirmed the fact that enzymatic extraction improves the degradation of lignocellulose, making active compounds more accessible and therefore easier to extract [26]. The two-step method takes advantage of enzymatic extraction to produce high-quality, bioactive extracts in a sustainable manner, breaking down the plant cell walls, enhancing the release of active compounds; in the second step, hydroalcoholic extraction isolates the target compounds. By combining these steps, the process maximizes the efficiency and yield of valuable compounds. Another benefit associated with the application of enzymatic pretreatment is related to the plethora of bioactives that can be obtained. Enzymes allow for the breakdown of complex plant matrices and release entrapped compounds more effectively. In our work, enzymatic extraction was shown to yield higher concentrations of some bioactive compounds such as p-coumaric acid, caffeic acid, and some other hydroxycinnamate derivatives. Although it affected the recovery of some compounds, it clearly increased the amount of other species, enriching the extracts into compounds that are usually not present in many industrial products.

The molecules extracted using enzymatic methods often exhibit enhanced bioactivity and stability compared to those obtained through conventional methods [49]. This may be related to the absence of harsh conditions that can degrade sensitive compounds. This preservation is crucial for maintaining the bioefficacy of the extracts. For instance, phenolic compounds, such as caffeic acid, p-coumaric acid, and luteolin, which could have relevant health benefits [50], were generally found in higher and more stable concentrations in enzymatic extracts compared to those obtained by solvent extraction alone. These results are consistent with a previous study on finger millet where certain caffeic acid derivatives were recovered only after xylanase incubation [51]. The fact that some phenolic components decreased after enzymatic treatment (i.e., verbascoside and forsythoside B) is not surprising, nor is the fact that some phenolic acids appeared only in the samples incubated with enzymes. Indeed, this behavior was previously observed in sweet orange peel after a treatment with a commercial mixture of β-glucanases, pectinases, hemicellulases, and xylanases [52].

5. Conclusions

In conclusion, enzymatic extraction offers significant advancements over conventional methods, combining efficiency, environmental sustainability, and improved product quality. Applied to V. nigrum, this approach optimized the recovery of (poly)phenolic compounds, yielding extracts with an enhanced phytochemical profile. This study highlights enzymatic extraction as a green and efficient alternative for producing high-quality herbal products. Future research should evaluate the bioactivity of these extracts in relevant physiological models.

6. Patents

These results have been used to file a patent application (Italian patent, No. 102023000022293).