Magnetic field-assisted blanching enhances sulforaphane synthesis in broccoli florets, revealed by UPLC-ESI-QTRAP-MS/MS

Abstract

Sulforaphane (SF) is generated from glucoraphanin hydrolysis catalyzed by endogenous myrosinase in Brassicaceae vegetables such as broccoli, with blanching critical to maximize SF formation. However, the effect of magnetic field (MF) treatment on glucoraphanin hydrolysis remains unclear. This study investigated MF-assisted blanching of broccoli florets under 0–10 mT, monitoring myrosinase activity and quantifying SF, sulforaphane nitrile (SFN), and precursor glucosinolates using UPLC-ESI-QTRAP-MS/MS. MF-assisted blanching significantly enhanced myrosinase activity and SF formation, with the greatest effect at 1.0 mT, where myrosinase activity increased 3.11-fold and SF content reached 61.14 ± 1.21 μmol/100 g fresh weight, a 1.89-fold increase relative to the control (blanching only). SFN increased but remained at trace levels, reaching 3.10 ± 0.10 μmol/100 g fresh weight at 1.0 mT versus 0.24 ± 0.09 μmol/100 g in the control. Enhanced hydrolysis was supported by residual glucosinolate levels. These findings suggest a promising strategy to improve SF content in broccoli.

Highlights

• •Magnetic field-assisted blanching enhanced sulforaphane synthesis.
• •The greatest enhancement occurred at 1.0 mT magnetic field.
• •Myrosinase activity increased 3.11-fold under optimal conditions.
• •Sulforaphane content was 1.89-fold higher than the control.

1. Introduction

Sulforaphane (4-methylsulfinylbutyl isothiocyanate) is one of the most potent natural compounds with anticancer activity (Khan et al., 2022). It also exhibits multiple biological functions, including prevention of metabolic diseases (Zhao et al., 2025), cardioprotective effects (Angeloni et al., 2009), and anti-inflammatory activity (Zhao et al., 2018), making it highly valuable for applications in functional foods and medicine (Zhang et al., 2024). In Brassicaceae vegetables such as broccoli, cabbage, and radish, sulforaphane (SF) occurs in its inactive precursor form, glucoraphanin (GRA). Broccoli is particularly notable as a rich and convenient dietary source of GRA, the dominant glucosinolate (GSL) in this vegetable (Wu et al., 2021). When plant tissues are damaged, endogenous myrosinase is released, catalyzing the hydrolysis of GRA to yield SF and other breakdown products. Although chemical synthesis and microbial biotransformation of SF have been explored, these approaches are limited by high costs, low efficiency, and environmental concerns. Consequently, enzymatic hydrolysis remains the predominant strategy for SF production (Zhang et al., 2021).

However, enzymatic production of sulforaphane (SF) is typically limited by intrinsic plant factors, such as substrate availability and low myrosinase activity (Luo et al., 2022; Piekarska et al., 2013), as well as external conditions, including pH, temperature, and metal ions (Wu et al., 2021), which often result in low SF yields and the formation of undesired nitriles. The enzymatic pathways of GSL in Brassicaceae plants such as broccoli are illustrated in Fig. 1. GSL undergoes myrosinase-catalyzed hydrolysis to form an unstable intermediate (thiohydroxamate-O-sulfonate) and d-glucose. Under neutral pH, this intermediate spontaneously undergoes a Lossen-type rearrangement to form isothiocyanates. In addition, nitriles are another type of hydrolysis product, as well as thiocyanates and epithionitriles formed to a lesser extent. Nitriles are considered undesirable due to their lack of demonstrated health benefits and potential toxicity. The presence of epithiospecifier protein (ESP) promotes nitrile formation under physiological pH, while nitriles are predominantly formed under acidic pH conditions (Brindisi et al., 2023; Wu et al., 2021). Ferrous ions further enhance ESP activity, particularly under mildly acidic conditions, strongly inhibiting isothiocyanate formation (Wu et al., 2021). Thus, the predominant GRA in broccoli produces both SF and sulforaphane nitrile (SFN) upon hydrolysis.

Fig. 1.

Enzymatic hydrolysis pathway of glucosinolates in Brassicaceae vegetables and the chemical structures of glucoraphanin (GRA) and its major degradation products: sulforaphane (SF) and sulforaphane nitrile (SFN). TFP, thiocyanate-forming protein; ESP, epithiospecifier protein.

To enhance SF yields, broccoli processing typically involves two sequential steps. The first step is a blanching treatment to inactivate ESP while preserving myrosinase activity, based on the fact that ESP is much more thermolabile than myrosinase, thereby suppressing nitrile formation (Perez et al., 2014). Second, the broccoli matrix is incubated at temperatures below 40 °C to optimize SF formation and minimize product loss, considering that SF is a volatile and heat-sensitive compound (Mahn & Perez, 2016). For example, blanching broccoli florets at 50–60 °C for 5–15 min was shown to specifically inactivate ESP while maintaining myrosinase activity. Optimal conditions (57 °C for 13 min) preserved peak myrosinase activity in broccoli floret and resulted in a 237% increase in SF content compared to fresh sample (Perez et al., 2014). Subsequent incubation at 38 °C for 1 h further increased SF levels by 10-fold, achieving up to 94% conversion of GRA to SF (Gonzalez et al., 2021; Mahn & Perez, 2016). Similarly, pretreatment methods such as short-duration microwave heating (Zheng et al., 2023) and ultrasound-assisted blanching (Mahn et al., 2020) prior to incubation have also been shown to modulate myrosinase activity and significantly enhance SF production.

Magnetic field (MF) technology has emerged as a novel non-thermal approach for modulating enzymatic and physicochemical properties in food systems. Studies demonstrate that static MF treatment can regulate pectinase activity in a dose-dependent manner: intensities or exposure times below a specific threshold suppress activity, whereas exceeding this threshold markedly enhances it (Sun et al., 2023). Myrosinase belongs to the glycosidase family. Although no studies have yet reported the regulation of myrosinase activity by MF treatment, other glycosidases, such as cellulase and α-amylase, have shown positive responses to MF. For instance, cellulase activity peaked after 20 min under a 2.2 mT alternating MF treatment (Zhang et al., 2012), while α-amylase activity was significantly increased by static MF intensities of 0.15–0.45 mT (Jia et al., 2009). Similarly, peroxidase activity in horseradish (Brassicaceae plant) reached its maximum after 10 min under a 52 mT static MF (Emamdadi et al., 2021). Spectroscopic analyses have confirmed that these activity changes are associated with modifications in amino acid residues and alterations in the secondary and tertiary enzyme structures (Sun et al., 2023).

Accurate quantification of SF and its related metabolites in the plant matrix requires analytical methods with high sensitivity and selectivity. Gas chromatography–mass spectrometry (GC–MS) has been widely used to analyze GSL hydrolysis products (Bell et al., 2021); however, it is unsuitable for SF because of the thermal degradation of SF during injection (Chiang et al., 1998). High-performance liquid chromatography with UV detection (HPLC–UV) offers greater stability and lower cost and is therefore the most commonly applied technique for SF quantification (Zheng et al., 2023). Nevertheless, it is limited by susceptibility to matrix interferences, insufficient sensitivity, and the absence of molecular mass information, which is critical for accurate compound identification (Ali Redha et al., 2023). By contrast, ultra-performance liquid chromatography coupled with electrospray ionization triple quadrupole tandem mass spectrometry (UPLC-ESI-QTRAP-MS/MS) offers rapid separation, high sensitivity, and multiple reaction monitoring (MRM) for targeted detection, enabling precise measurement of both GSLs and their degradation products (Andernach et al., 2023). These advantages make UPLC-ESI-QTRAP-MS/MS a preferred tool for studying GSL metabolism in Brassicaceae vegetables.

Given that blanching is a critical step for the enzymatic formation of SF and a common pretreatment in vegetable processing, integrating MF assistance into this step holds both theoretical significance and practical potential for improving SF yield. This study investigated the effects of static MF assistance during blanching on myrosinase activity and subsequent SF accumulation in broccoli florets. For this purpose, a rapid and robust UPLC-ESI-QTRAP-MS/MS method was established to simultaneously quantify SF and SFN. To the best of our knowledge, this study is the first to demonstrate MF modulation of myrosinase activity and SF production within a Brassicaceae vegetable matrix.

2. Materials and methods

2.1. Materials

Fresh broccoli (Brassica oleracea L. var. italica) heads were obtained from a fixed local supplier in Wuhan, China. All samples originated from Yunnan Province, China, and were harvested at commercial maturity. The broccoli heads were from the same cultivar and the same production batch, transported to the market within 48 h after harvest, and used for experiments within 24 h of purchase to minimize variations in GSL composition.

SF (98.87%), sinigrin (99.92%), and GRA (99.88%) standards were purchased from MedChem Express (Monmouth Junction, NJ, USA). The glucose content assay kit was purchased from Solarbio Technology Co., Ltd. (Beijing, China). Chromatographic-grade methanol and formic acid were purchased from Sinopharm Chemical Reagent Co., Ltd. (Shanghai, China).

2.2. Sample handling and blanching process

Broccoli florets were cut into lengths of approximately 5 cm and widths of 0.7–0.9 cm (measured at the stem) before blanching. Blanching was performed under the optimal conditions reported by Perez et al. (2014), in which broccoli florets were immersed in water at 57 °C for 13 min. These conditions were selected because they maximize myrosinase activity and minimize residual glucoraphanin, resulting in the highest sulforaphane content after subsequent incubation, thereby providing an ideal baseline for evaluating the effect of magnetic field (MF) treatment. In addition, these blanching conditions are commonly used in the vegetable processing industry, ensuring practical feasibility.

Specifically, 1.0 L of deionized water was preheated to 57 °C, transferred to a beaker, and placed in a MF constant-temperature chamber (MFL10, INDUC Scientific Co., Ltd., Wuxi, China) preset to 57 °C to ensure thermal equilibration. After temperature stabilization, 200 g of broccoli florets were immersed in the water and blanched for 13 min. During blanching, static MF treatments at different intensities (0, 0.5, 1, 3, 5, and 10 mT) were applied. This intensity range was selected based on preliminary experiments (Table S1) evaluating the effect of MF intensities (0–20 mT) on myrosinase activity in broccoli. The MF treatment did not affect the set blanching temperature, as a static rather than alternating magnetic field was applied; thus, no magnetic-induced heating of the diamagnetic food matrix occurred at present intensities, which was confirmed in preliminary tests using temperature data loggers. After blanching, surface moisture was immediately removed, and samples were pulverized in liquid nitrogen. The powdered samples were packed in food-grade aluminum foil self-sealing bags (PET/aluminum foil/PE) to minimize light and oxygen exposure and stored at −80 °C until analysis. All treatments were conducted in triplicate.

2.3. MF device and characterization

The static MF was generated using a Helmholtz coil system integrated into the MFL10 constant-temperature and humidity chamber (Fig. 2A). The system consisted of a magnetic field generator, sample chamber, temperature control unit, humidity control unit, air circulation fan, and control panel. The chamber had an internal volume of 50 L. The internal temperature was regulated via an air heater and refrigeration compressor, allowing a range of 4–60 °C with a deviation of ±0.3 °C, while the humidity was kept at the ambient indoor level on the day of the experiment, ensuring stable environmental conditions during MF treatment.

Fig. 2.

Configuration and characterization of the static magnetic field (MF) system used for MF-assisted blanching. (A) MF-assisted constant temperature and humidity chamber (MFL10) diagram. (B) Simulated distribution and orientation of the static magnetic field within the sample chamber, showing a homogeneous magnetic field aligned with the axial (horizontal) direction of the Helmholtz coils. (C) Experimental setup for magnetic field uniformity calibration, consisting of a Gaussian meter coupled with a precision motorized positioning platform for spatial field mapping in the effective treatment zone. (D) Three-dimensional numerical simulation of magnetic field intensity distribution inside the chamber, confirming the uniformity of the static MF at the sample position.

The magnetic field generator comprised a pair of square Helmholtz coils (80 cm × 80 cm, 400 turns each) connected to a regulated power supply, producing a uniform magnetic field in the range of 0–10 mT at an excitation current of 0–8 A. The system was operated in direct-current (DC) mode, generating a static magnetic field with no temporal variation in field intensity. In a standard Helmholtz coil configuration, MF lines in the central region are parallel to the coil axis. Accordingly, the static MF applied in this study was oriented along the axial direction of the Helmholtz coils, corresponding to the horizontal direction of the chamber (Fig. 2B).

During treatment, broccoli samples were positioned at the geometric center of the Helmholtz coils, approximately 20 cm from the MF generator, where a highly uniform MF region was established. MF uniformity within the effective treatment zone was maintained within ±5% of the target field intensity. Field calibration and spatial uniformity assessment were performed using a calibrated gaussmeter (BLD-1030, Bolandun Co., Ltd., Beijing, China) in combination with a precision motorized positioning platform (Fig. 2C). In addition, COMSOL Multiphysics 6.2 software was used to numerically simulate the magnetic field distribution within the sample chamber, further confirming the uniformity of the static MF in the central treatment region (Fig. 2D). This configuration ensured that all samples were exposed to a homogeneous static MF during blanching.

2.4. Myrosinase activity after blanching

Myrosinase activity was determined according to the method of Zheng et al. (2023), with minor modifications. Pulverized broccoli samples (1.0 g) were mixed with 2.0 mL of phosphate buffer (0.1 mmol/L, pH 7.0) in an ice bath and vortexed to homogenize. The mixture was centrifuged at 10,000 ×g for 10 min at 4 °C, and the resulting supernatant was used as the crude enzyme solution. A 200 μL aliquot of the crude enzyme solution was combined with 100 μL of sinigrin solution (2 mmol/L) and incubated at 37 °C for 15 min. The incubation time was chosen to ensure that glucose formation occurred within the linear range of the reaction, as verified in preliminary experiments (Fig. S1). The reaction was terminated by placing the mixture in a boiling water bath for 5 min. Glucose content was determined using a glucose assay kit. One unit (U) of myrosinase activity corresponded to the production of 1 nmol of glucose per min. The activity was expressed as U per gram of sample.

2.5. SF content determination

2.5.1. Broccoli matrix incubation and sample preparation

The incubation procedure followed the optimal conditions reported by Gonzalez et al. (2021) to promote GSL hydrolysis at the optimal temperature for myrosinase activity, thereby maximizing SF content in the broccoli matrix. One gram of the pulverized, blanched sample was transferred into a 15 mL glass centrifuge tube with a cap and incubated at 38 °C for 60 min.

Following incubation, 10 mL of dichloromethane was added, and the mixture was vortexed for 1 min. The sample was centrifuged at 10,000 rpm for 10 min at 4 °C. The dichloromethane layer was carefully removed using a rubber-tipped dropper and passed through anhydrous sodium sulfate for dehydration. The residue was extracted once more using the same procedure. The combined dichloromethane extracts were evaporated under nitrogen, and the residue was resuspended in 5 mL of acetonitrile. The final solution was filtered through a 0.22 μm nylon membrane, transferred into dark glass chromatography vials, sealed, and stored at −80 °C until UPLC-ESI-QTRAP-MS/MS analysis.

2.5.2. UPLC-ESI-QTRAP-MS/MS conditions

The hydrolysis products of GRA, including SF and SFN, were quantitatively analyzed using a SCIEX ExionLC AD LC system equipped with an Agilent ZORBAX RRHD Eclipse Plus C18 column (1.8 μm, 50 × 2.1 mm; Agilent Technologies, Santa Clara, CA, USA), coupled to a SCIEX 6500+ QTRAP mass spectrometer (SCIEX, Framingham, MA, USA). The column oven temperature was maintained at 25 °C. The mobile phase consisted of 0.1% formic acid in water (A) and 0.1% formic acid in acetonitrile (B) at a flow rate of 0.2 mL/min with a 2 μL injection volume. The linear gradient program for solvent B (10 min) was as follows: initially 5%, increased to 20% within 2 min; ramped to 70% from 2 to 7 min; further increased to 100% from 7 to 7.5 min; held for 0.5 min; returned to 5% at 9 min; maintained for 1 min (Ali Redha et al., 2023).

The QTRAP was set up for IonSpray operation, and compounds were detected using MRM in positive ion mode. Additional QTRAP parameters were as follows: curtain gas (CUR) = 35; collision gas (CAD) = medium; ion spray voltage (IS) = 4500; temperature (TEM) = 450 °C; ion source gas 1 (GS1) = 40; ion source gas 2 (GS2) = 40. The precursor/product transitions of SF and SFN are listed in Table 1, along with declustering potential (DP), entrance potential (EP), collision energy (CE), and collision cell exit potential (CXP) values. These parameters were optimized by direct infusion of authentic standards into the 6500+ QTRAP to achieve maximal response. The dwell time for each analyte was 20 ms. There were typically 14 points across all chromatographic peaks with a total cycle time of 0.3 s. Quantification of SF and SFN was based on the SF standard calibration curve. SFN was quantified using the SF calibration curve due to the lack of an authentic SFN standard. Given the structural similarity and comparable chromatographic behavior of SF and SFN, this approach was considered suitable for comparative analysis. Nevertheless, potential differences in ionization efficiency between SF and SFN may introduce systematic bias; therefore, SFN data are interpreted as semi-quantitative and are primarily used to assess relative changes among treatments rather than absolute concentrations. Data acquisition and processing were performed using Analyst software 1.6.2 (AB Sciex, Foster City, CA, USA) (D'Urso et al., 2020).

Table 1.

Precursor/product transitions and mass spectral parameters of sulphoraphane (SF) and sulphoraphane nitrile (SFN) measured using UPLC-ESI-QTRAP-MS/MS in positive ion MRM mode.

Compounds	PI	DI	DP	EP	CE	CXP
SF	178	114	24	10	30	11
SFN	146	55	26	10	20	11

PI, product ion; DI, daughter ion; DP, declustering potential; EP, entrance potential; CE, collision energy; CXP, collision cell exit potential.

2.5.3. Method validation of SF quantification

The developed method was validated for sensitivity, linearity (R²), intra- and inter-day precision, and accuracy according to the US Food and Drug Administration guidelines, with slight modifications (U.S. Department of Health and Human Services, 2018).

Sensitivity was determined by calculating the limit of detection (LOD, S/N = 3) and the limit of quantitation (LOQ, S/N = 10). Linearity was evaluated from the correlation coefficients of calibration curves. Precision and accuracy were assessed at three SF concentration levels: low (0.005 μmol/L), medium (0.025 μmol/L), and high (0.1 μmol/L). Intra-day precision was determined from the relative standard deviation (RSD) of five replicate measurements per concentration in a single analytical run for low and medium concentrations, and three replicates for the high concentration. Inter-day precision was determined from three replicate measurements per concentration across three different days. Accuracy was evaluated by analyzing spiked samples at low, medium, and high concentrations, with five replicates per level in a single run. The medium concentration was chosen to approximate the SF content of the control group, ensuring that method performance was evaluated under realistic sample conditions. For spiking, broccoli florets were first incubated following the procedure described in Section 2.5.1, after which 1.0 g of treated sample was spiked with 10 μL of an SF standard solution prior to extraction and analysis (Alvarez-Jubete et al., 2014; Andernach et al., 2023).

2.6. Residual GSL content determination

Following MF-assisted blanching and incubation, the residual GSL (including GRA) in the broccoli matrix was determined using the same LC–MS system to confirm the complete hydrolysis of the precursor. Owing to their high polarity, GSLs were extracted from 1.0 g of treated sample with 3.0 mL of preheated methanol–water (75% v/v, 80 °C). After vortexing, the mixture was incubated in a water bath at 75 °C for 20 min, immediately cooled, and centrifuged at 12,000 rpm for 10 min to collect the supernatant. The extraction was repeated twice, and all supernatants were combined, evaporated to dryness under a nitrogen stream, and reconstituted in 5 mL of ultrapure water. The reconstituted solution was filtered through a 0.22 μm nylon membrane and analyzed by UPLC-ESI-QTRAP-MS/MS in MRM mode.

Chromatographic separation was performed using the same column and column temperature as described in Section 2.4.2, with a flow rate of 0.2 mL/min. The mobile phase consisted of (A) 0.1% formic acid in water and (B) 0.1% formic acid in methanol. The gradient program for solvent B was as follows: 10% to 90% within 3 min, held at 90% for 3 min, returned to 10% at 6.17 min, and maintained at 10% until 9 min. The injection volume was 2 μL. The 6500+ QTRAP mass spectrometer was operated in negative ion mode using an IonSpray source with the following parameters: CUR = 35; CAD = medium; IS = −4500 V; TEM = 450 °C; GS1 = 40; GS2 = 40. MRM transitions and mass spectrometric parameters (DP, EP, CE, and CXP) for each GSL were optimized using authentic standards and literature data (Maldini et al., 2017), as summarized in Table 2. The dwell time for each analyte was 20 ms, and GSL concentrations were calculated based on the standard curve of GRA.

Table 2.

Precursor/product transitions and mass spectral parameters of glucosinolates measured using UPLC-ESI-QTRAP-MS/MS in negative MRM mode.

Glucosinolate	Abbreviation	PI	DI	DP	EP	CE	CXP
Glucoraphanin	GRA	436	97	−60	−10	−54	−11
Glucoiberin	GIB	422	97	−60	−10	−54	−11
Glucoerucin	GER	420	97	−60	−10	−54	−11
Glucobrassicin	GBC	447	97	−85	−10	−53	−11
4-Methoxyglucobrassicin	4MGB	477	97	−100	−10	−84	−11
Neoglucobrassicin	NGB	477	97	−100	−10	−54	−11

PI, product ion; DI, daughter ion; DP, declustering potential; EP, entrance potential; CE, collision energy; CXP, collision cell exit potential.

2.7. Statistical analysis

Results are expressed as mean ± standard deviation (n = 3 independent experiments). All experiments were conducted using broccoli samples from the same batch to minimize biological variability. Analysis of variance was conducted using SPSS 27.0 software (IBM Corp., Armonk, NY, USA) to assess significant differences (p < 0.05). Graphics were prepared using Origin 2021 (OriginLab, Northampton, MA, USA).

3. Results and discussion

3.1. Myrosinase activity analysis

Myrosinase catalyzes the hydrolysis of GSLs to produce d-glucose and an unstable aglycone intermediate, which subsequently undergoes molecular rearrangement to form hydrolysis products (Fig. 1). The enzyme activity during the initial hydrolysis step directly determines the generation of intermediates and ultimately affects the yield of products such as SF. Fig. 3 illustrates the myrosinase activity in broccoli florets subjected to optimal blanching (57 °C, 13 min) combined with MF treatment at intensities ranging from 0 to 10 mT. All MF-assisted blanching treatments significantly enhanced myrosinase activity compared to the control (0 mT, blanching only) (p < 0.05), indicating that MF application effectively enhances catalytic activity in broccoli matrices. The maximum activity, 86.32 ± 3.08 U/g, was observed at 1.0 mT, approximately 3.11 times higher than the control. It should be noted that myrosinase activity was measured after tissue disruption and enzyme extraction; therefore, the observed enhancement reflects changes in enzymatic properties rather than MF-induced alterations in tissue permeability. Within the tested range, the activity first increased and then decreased with increasing field intensity, suggesting that intensities above the optimal value may impair enzyme function. A similar pattern was reported for cellulase exposed to MFs of varying intensities for 20 min, with activity following a normal distribution and peaking at 2.2 mT (Zhang et al., 2012). Spectroscopic studies, including UV absorption, fluorescence emission, circular dichroism, and FT-IR, have shown that MF exposure can alter enzyme secondary and tertiary structures and intramolecular bonding (Emamdadi et al., 2021; Jia et al., 2009). Moderate structural adjustments may enhance catalytic efficiency by improving substrate binding or aligning catalytic residues. In contrast, stronger MF intensities may cause excessive structural disruption, exposing hydrophobic residues or breaking key hydrogen bonds, ultimately reducing activity. Although studies on other enzymes support these mechanisms, direct evidence for myrosinase is not yet available, and further targeted investigation is needed.

Fig. 3.

Myrosinase activity in broccoli florets after blanching at different magnetic field intensities. Data are expressed as mean ± SD (n = 3 independent experiments). Different letters indicate that values are significantly different at the p < 0.05 level.

Although myrosinase activity differs among broccoli cultivars, MF-assisted blanching consistently yielded higher activity than microwave pretreatment under identical assay conditions (Zheng et al., 2023). Given that both blanching and low-power microwave treatment are thermal processes that significantly enhance myrosinase activity (Perez et al., 2014), the present findings suggest that integrating MF with thermal blanching results in a greater enhancement in myrosinase activity in broccoli.

3.2. Method validation for SF analysis using LC-MS MRM

The results of the method validation for SF analysis are presented in Table 3. The LOD and LOQ were 0.56 and 2.82 nmol/L, respectively. These results are consistent with the value range of GSL degradation products detected by 6500 QTRAP+ mass spectrometry (Andernach et al., 2023). The calibration curve was constructed using eight calibration points covering the range of 2.82 nmol/L to 2.82 mmol/L. Calibration curves were calculated by linear regression. The coefficient of determination (R²) from three repeated measurements all exceeded 0.998. Intra-day precisions at the three concentration levels (low, medium and high) ranged from 1.89 to 3.59% for SF, whereas inter-day precisions ranged from 3.35 to 8.54%. The accuracy of the spiked samples was 108.7%, 98.73%, and 95.33% for the low, medium, and high concentrations, respectively.

Table 3.

Final MRM parameters and method validation results for the analysis of sulphoraphane (SF).

Compound	Retention time[min]	MRM transition quantifier	MRM transitions qualifier	Calibration curve equation	R2	LOD [nmol/L]	LOQ [nmol/L]	Level	Intra-day Precision RSD [%]	Inter-day Precision RSD [%]	Accuracy (%)
SF	4.78	178 → 114	178 → 114	y = 1.83e6x-2.53e3	0.9987	0.56	2.82	Low	1.89	3.35	108.7
			178 → 55					Medium	2.02	8.54	98.73
								High	3.59	4.08	95.33

MRM: multiple reaction monitoring; LOD: limit of detection; LOQ: limit of quantitation; RSD: relative standard deviation.

Due to the unavailability of a commercial SFN standard, SFN was relatively quantified using the SF standard curve. The MS parameters for SFN (Table 1) were optimized based on previously reported parameters and fragmentations (Alvarez-Jubete et al., 2014). The quantification of SFN was utilized to evaluate the effect of MF-assisted blanching on SFN formation and verify the conversion efficiency of SF.

3.3. SF content

Broccoli florets were subjected to blanching with the simultaneous application of MFs of varying intensities (0–10 mT), followed by hydrolysis under reported optimal conditions (38 °C, 1 h). Fig. 4A presents the SF content determined by UPLC-ESI-QTRAP-MS/MS. In the control sample (blanching and incubation only), SF content was 32.43 ± 2.28 μmol/100 g FW, comparable to previously reported values for optimally blanched broccoli florets (Perez et al., 2014). Consistent with the enhanced myrosinase activity, MF-assisted blanching significantly increased SF content at all tested intensities compared with the control, with low-intensity fields (0.5–3 mT) producing greater enhancement than higher intensities (5–10 mT). The maximum SF content (61.14 ± 1.21 μmol/100 g FW) was observed at 1 mT, corresponding to the intensity that maximized myrosinase activity, representing a 1.89-fold increase over the control.

Fig. 4.

Effect of magnetic field-assisted blanching at different intensities on the contents of SF (A) and SFN (B) in broccoli florets after subsequent incubation. Data are expressed as mean ± SD (n = 3 independent experiments). Different letters indicate that values are significantly different at the p < 0.05 level.

The undesired hydrolysis product SFN remained at low levels across all treatments (Fig. 4B). The control sample showed the lowest SFN content (0.24 ± 0.09 μmol/100 g FW), indicating that blanching was effective in suppressing nitrile formation. Although MF-assisted blanching led to a statistically significant increase in SFN at all tested field intensities, the maximum level observed at 1.0 mT (3.10 ± 0.10 μmol/100 g FW) was still approximately two orders of magnitude lower than the corresponding SF concentration. Accordingly, SFN remained a minor hydrolysis product relative to SF, and its contribution to the overall hydrolysis profile was limited.

Nitriles generally lack demonstrated health benefits and are considered undesirable due to their potential toxicity. Notably, SFN remained at low levels across all treatments (Fig. 4B). The control sample showed the lowest SFN content (0.24 ± 0.09 μmol/100 g FW), indicating that blanching was effective in suppressing nitrile formation (Matusheski et al., 2004). Although MF-assisted blanching led to a statistically significant increase in SFN at all tested field intensities, the maximum level observed at 1.0 mT (3.10 ± 0.10 μmol/100 g FW) was still approximately two orders of magnitude lower than the corresponding SF concentration. Accordingly, SFN remained a minor hydrolysis product relative to SF, and its contribution to the overall hydrolysis profile was limited. Moreover, Toxicological studies further indicate that adverse effects of SFN, such as impaired detoxification, oxidative stress, and DNA damage, occur only at concentrations far exceeding normal dietary exposure (Kupke et al., 2016; Matusheski & Jeffery, 2001). Therefore, the trace amounts detected in this study are unlikely to pose any meaningful safety concern for human consumption.

The observed increase in SFN under MF-assisted blanching may be attributed to several non-mutually exclusive biochemical factors. Although blanching effectively inactivates ESP, incomplete inactivation under mild thermal conditions cannot be fully excluded, allowing residual ESP activity to redirect a small fraction of the unstable aglycone toward nitrile formation (Matusheski et al., 2004). In addition, MF exposure may subtly influence residual ESP behavior or the local hydrolysis microenvironment, such as transient changes in pH or Fe²⁺ availability, both of which are known to favor nitrile formation (Fig. 1). However, these effects appear limited, as the overall hydrolysis pathway remained strongly biased toward isothiocyanate formation. It should be noted that the present study focused on product outcomes within the vegetable matrix, and the underlying molecular mechanisms cannot be conclusively resolved without targeted studies using purified enzymes or controlled model systems. Overall, blanching remained the dominant factor suppressing SFN formation, while MF-assisted blanching primarily enhanced GSL hydrolysis toward SF production with only a minor increase in nitrile byproducts.

3.4. Insights into the potential mechanism of MF-enhanced SF formation

The enhancement of SF formation by blanching can be primarily explained by two established factors. First, short-term heat treatment disrupts plant cell structures, facilitating the release of GSLs and myrosinase and thereby increasing the extent of enzymatic hydrolysis. Blanching at 50–70 °C has been reported to significantly increase GSL content in broccoli; for example, blanching at 70 °C for 15 min increased glucoraphanin levels by 93% compared with fresh samples (Gonzalez et al., 2021). Similarly, higher GRA contents have been reported in steamed, boiled, and oven-steamed broccoli relative to fresh samples (Pellegrini et al., 2010), and microwave-assisted treatments have also been shown to enrich GSL levels (Zheng et al., 2023). Second, although ESP redirects GSL hydrolysis toward biologically inactive nitriles, ESP and myrosinase exhibit markedly different thermal sensitivities. Myrosinase activity increases with temperature and reaches a maximum at approximately 60 °C (Perez et al., 2014), whereas ESP is more thermolabile and is substantially inactivated at around 50 °C (Matusheski et al., 2004). Under the blanching conditions applied in the present study, GSL availability and myrosinase activity are therefore favored, while ESP activity is largely suppressed, which collectively promotes SF formation. In contrast, temperatures above 70 °C or prolonged heating have been reported to reduce SF yield due to myrosinase deactivation and SF thermal instability (Perez et al., 2014).

The increase in SF yield observed after MF-assisted blanching may be attributed to factors related to enzyme molecular structure. Previous studies have reported that MF exposure can alter enzyme conformation. Spectroscopic analyses have revealed changes in amino acid residues as well as in secondary and tertiary structures, which are accompanied by alterations in the activity, reaction kinetics, and stability of enzymes such as peroxidase and pectinase (Emamdadi et al., 2021; Sun et al., 2023). Based on these reports, it is hypothesized that MF treatment may influence myrosinase structures during blanching, although this effect was not directly examined in the present study.

MF treatment has also been reported to influence the aqueous environment by modifying hydrogen-bond networks and water polarization, which may indirectly affect laccase–substrate interactions (Wasak et al., 2019). During GSL hydrolysis, substrate recognition by myrosinase involves hydrogen-bond-mediated interactions between the active site and the thiohydroximate moiety of GSLs, often bridged by water molecules (Bourderioux et al., 2005). It is therefore conceivable that MF-induced alterations in water structure could modulate these interactions and consequently influence hydrolysis efficiency. However, it should be emphasized that the present study did not directly examine MF-induced changes in water structure or their potential effects on myrosinase catalysis. Accordingly, this consideration is proposed solely as a speculative hypothesis rather than a demonstrated mechanism, and further dedicated physicochemical and enzymatic investigations are required to clarify its relevance.

Moreover, MF as a non-thermal physical processing method has been reported to induce morphological changes and alter membrane permeability in biological systems, thereby affecting mass transfer processes (Wang et al., 2024). By analogy, MF-assisted blanching may enhance the disruption of broccoli tissue microstructures, facilitating contact between enzymes and substrates and improving their interaction. Such structural effects could also contribute to improved SF extraction efficiency prior to quantitative analysis.

3.5. Residual GSLs in broccoli florets after blanching and incubation

GSL profiles in broccoli have been extensively reported, including aliphatic GSLs such as GRA, gluconasturtiin (GST), glucoerucin (GER), and glucoiberin (GIB), as well as indole GSLs such as 4-methoxy-glucobrassicin (4MGB), 4-hydroxyglucobrassicin (4HGB), neoglucobrassicin (NGB), and glucobrassicin (GBC). Among these, GRA is the predominant GSL, accounting for approximately 80–90% of the total GSL content (Zheng et al., 2023). Following thermal processing, low-abundance GSLs are often completely lost, and typically 5–6 GSLs can be quantified in processed Brassicaceae vegetables (Luo et al., 2022). In the present study, UPLC–ESI–QTRAP–MS/MS analysis quantified six residual GSLs in broccoli florets after blanching and incubation, including GRA, GBC, 4MGB, NGB, GIB, and GER. Their concentrations under different intensities of MF assisted blanching are shown in Fig. 5. The control sample exhibited the highest levels of all GSLs except for GIB, in agreement with previous reports that substantial amounts of GSLs can remain in broccoli after blanching and incubation steps (Gonzalez et al., 2021; Perez et al., 2014). Previous studies have indicated that supplementing exogenous myrosinase can enhance GSL hydrolysis, but this approach inevitably increases production costs (Shen et al., 2010). In contrast, MF-assisted treatment markedly reduced GRA to trace levels after incubation, with the lowest residual contents observed at 0.5–3 mT (no significant differences within this range), corresponding to GRA reductions of approximately 97.5–98.3% relative to the control (0 mT). At higher MF intensities (5–10 mT), GRA degradation remained substantial (92.5–93.7%) but was less pronounced than that observed under low-field conditions, consistent with the MF intensity-dependent trends observed for myrosinase activity and SF formation. Other GSLs showed comparable reduction patterns.

Fig. 5.

Effect of magnetic field-assisted blanching at different intensities on the contents of residual glucosinolates in broccoli florets after subsequent incubation: GRA (A), GBC (B), 4MGB (C), NGB (D), GIB (E), GER (F). Data are expressed as mean ± SD (n = 3 independent experiments). Different letters indicate that values are significantly different at the p < 0.05 level.

Although MF treatment clearly enhanced GSL degradation beyond the effect of blanching alone, a complete mass balance between GRA depletion and the quantified formation of SF and SFN was not achieved. The combined concentrations of SF, SFN, and residual GRA varied across MF intensities, indicating that additional hydrolysis pathways may contribute to GSL conversion. Notably, the control sample without MF assistance exhibited the lowest combined levels of residual GRA, SF, and SFN and showed a pronounced discrepancy compared with MF-treated samples, suggesting that GRA was diverted to other hydrolysis products to a considerable extent under conventional blanching conditions. In Brassicaceae vegetables, GRA hydrolysis can also yield thiocyanates or epithionitriles (Fig. 1), particularly when ESP and thiocyanate-forming protein (TFP) were still existed or when the enzymatic microenvironment is altered (Burow et al., 2007; Wang et al., 2019). These products were not included in the present analytical scope and may partly account for the observed discrepancy. In contrast, MF-assisted blanching appeared to suppress the diversion of GRA toward these alternative pathways, thereby favoring SF formation and resulting in higher SF yields.

4. Conclusion

Blanching remains a critical step for enzymatic synthesis of SF in broccoli, and MF-assisted blanching was shown to significantly enhance both myrosinase activity and SF content, with the greatest effect observed at 1.0 mT. Under this condition, myrosinase activity increased 3.11-fold, and SF content reached 61.14 ± 1.21 μmol/100 g FW after incubation, representing a 1.89-fold increase relative to the control as accurately quantified by the proposed UPLC-ESI-QTRAP-MS/MS method. These findings highlight the potential of MF application to modulate enzyme–substrate interactions and enhance SF formation, thereby improving the nutritional and functional value of broccoli products. Key limitations of this study include the absence of direct ESP activity measurements and detailed kinetic or structural analyses of myrosinase, which restrict a definitive mechanistic interpretation of MF effects at the molecular level. As a result, mechanistic interpretations regarding the observed changes in SFN formation under MF-assisted blanching remain indirect and are inferred from product distribution patterns rather than from direct enzymatic evidence. Nevertheless, SFN consistently remained a minor hydrolysis product relative to SF across all treatments, indicating that blanching remained the dominant factor governing SF/SFN partitioning under the investigated conditions. Future studies will systematically optimize MF-assisted blanching conditions, investigate tissue-level morphological changes via microscopy, perform enzyme kinetic analyses, assess ESP activity, and explore structural dynamics of myrosinase to further elucidate the mechanisms underlying MF-enhanced SF formation.

CRediT authorship contribution statement

Jiawei Cheng: Validation, Investigation. Muci Wu: Writing – review & editing, Writing – original draft, Supervision, Funding acquisition, Data curation, Conceptualization. Xin Liu: Validation, Methodology. Jingyi Wang: Supervision, Resources. Wangting Zhou: Methodology. Rui Zhang: Resources, Methodology. Jingren He: Supervision, Resources.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Appendix A. Supplementary data

Supplementary material 1

Supplementary material 2

Data availability

Data will be made available on request.