Species-Specific Antioxidant and Metabolic Responses to Selenium in Brassica Microgreens

Abstract

Selenium (Se) biofortification is a promising approach to improve the nutritional value and functional quality of microgreens, although species-specific responses to Se remain insufficiently understood. This study investigated the effects of Se biofortification on physiological status, antioxidant responses, phenolic composition, and molecular changes in four Brassica microgreens: broccoli, kohlrabi, pak choi, and kale, using biochemical analyses, HPLC, and FTIR spectroscopy. The indicators of nutritional quality and stress-related metabolism in Brassica microgreens showed species-specific responses due to selenium treatment. Kohlrabi showed coordinated osmotic and metabolic adjustment involving osmolyte accumulation and enhanced antioxidant response, although moderate membrane sensitivity was observed at the highest selenium concentration. Pak choi maintained tolerance through balanced metabolic adjustment and enzymatic defense, while broccoli responded predominantly through enzymatic antioxidant mechanisms. Kale exhibited pronounced non-enzymatic responses, including anthocyanin accumulation and enhanced radical scavenging capacity. PCA confirmed species-specific response strategies and differential associations among biochemical parameters. Changes in antioxidant functionality were associated with both metabolite accumulation and structural reorganization of phenolic-related compounds. Overall, Se biofortification improved functional and nutritional traits of the investigated Brassica microgreens, although higher selenium concentrations induced moderate oxidative and membrane-related stress in certain Brassica microgreens. These findings highlight the importance of species-specific optimization of Se application to maximize crop quality while minimizing potential effects of Se toxicity.

1. Introduction

In recent years, interest in microgreens cultivation has grown due to increased awareness of their multiple health-promoting effects. In addition to their intense flavor, various microgreens are notable for their high content of bioactive compounds, including vitamins, antioxidants, and minerals [1]. One of the main advantages of microgreens production is the short growth cycle (7–21 days) and the minimal space required for cultivation. Compared to mature plants, microgreens contain lower levels of phytates—a nutrient that reduces mineral absorption, resulting in improved mineral bioavailability and absorption efficiency [2,3].

Beyond improved mineral bioavailability, microgreens often exhibit a more diverse and higher concentration of bioactive or secondary metabolite compounds than mature vegetables [3]. Microgreens belonging to the Brassicaceae family possess an especially rich profile of bioactive compounds. Sun et al. [4] identified a total of 164 polyphenols in five microgreen species, including 30 anthocyanins, 105 flavanol glycosides, and 29 hydroxycinnamic acids. Kyriacou et al. [5] reported that pak choi and tatsoi exhibited the highest chlorophyll content among the analyzed microgreen species. Additionally, microgreens of kale, red cabbage, kohlrabi, and red radish contain not only high levels of bioavailable polyphenols but also significant concentrations of glucosinolates [6], compounds known for their anti-inflammatory, antitumor, antimicrobial, and antidiabetic activities [7]. It has also been demonstrated that microgreens of kale, radish, mustard, and broccoli reduce the viability of colon cancer cells, indicating their antiproliferative potential [8]. The same authors reported that mustard and kale showed stronger effects than radish and broccoli, likely due to higher levels of ascorbic acid, total carotenoids, and isothiocyanates. Differences in bioactive compound content among species are attributed to genotypic characteristics, including variations in photosynthetic and metabolic activity [9].

One promising strategy for enhancing the nutritional value of plants is biofortification, which includes conventional breeding, agronomic biofortification, and transgenic approaches [10]. The present study focuses on agronomic biofortification, in which nutrient solutions are applied to increase selenium (Se) concentration in edible plant tissues [11,12]. Selenium plays an important role in maintaining antioxidant status in humans, and its biofortification represents an effective way for improving the nutritional quality of sprouts and microgreens [13]. As a component of selenoproteins, selenium contributes to cellular redox balance, thyroid hormone metabolism, immune function, fertility, aging processes, and protection against tumors and viral infections [14,15]. The recommended dietary allowance (RDA) for selenium in adults aged 19–55 years is 55 µg per day [16]. Previous studies on microgreens have demonstrated that optimal selenium concentrations depend on plant species and selenium dose, with beneficial effects on bioactive compound accumulation occurring within relatively narrow concentration ranges [17]. Selenium biofortification is more effective when selenium is applied as selenate rather than selenite, since selenate is more readily translocated to the aerial edible plant tissues, whereas selenite is predominantly retained in the roots [18,19]. Several studies demonstrated that selenite treatment increases selenium accumulation and may improve the content of bioactive compounds, antioxidant capacity, and mineral composition in various microgreens, including basil, coriander, tatsoi, mizuna, arugula, cress, and radish [17,20,21]. In addition, in commonly cultivated Brassica microgreens, selenite application was shown to maintain glucosinolate content without significant reductions [22]. Although selenium is essential for humans and animals, at low concentrations it may promote growth, development, antioxidant activity, and physiological performance [12,23]. Our previous investigation also showed the beneficial effect of Se biofortification on photosynthetic performance in Brassica microgreens [24]. Similarly, Viltres-Portales et al. [25] reported that Se uptake in kale and kohlrabi microgreens did not induce pronounced oxidative damage. Members of the Brassicaceae family exhibit high selenium uptake potential with a reduced risk of toxicity, and are thus considered natural selenium bioaccumulators. However, excessive selenium concentrations may still induce toxic effects. Therefore, given the narrow range between the beneficial and toxic effects of selenium, it is important to investigate its impact on plant metabolism [26].

Given that in previous research [24], we determined that kale received the most selenium and proved to be the best accumulator, our goal in this study was to identify which of the four Brassica microgreens, in addition to selenium accumulation, has the most favorable balance between metabolic and antioxidant profile for potential cultivation and selenium biofortification while ensuring consumer safety. In this study, we evaluated the effects of selenium biofortification with sodium selenate on the metabolic profiles and species-specific stress responses of four Brassica microgreens: pak choi, kale, broccoli, and kohlrabi.

2. Materials and Methods

2.1. Microgreens Cultivation and Experimental Setup

Four species from the Brassicaceae family: broccoli (Brassica oleracea var. italica), kohlrabi (Brassica oleracea var. gongylodes), pak choi (Brassica rapa var. chinensis), and kale (Brassica napus var. pabularia) were grown from ecologically certified seeds for microgreen cultivation, purchased from a local supplier, Lokvina d.o.o. (Savska Ves, Čakovec, Croatia). Microgreens were cultivated hydroponically in plastic containers using tap water (Figure 1). Six-day-old microgreens were transferred to aqueous solutions of sodium selenate (Na₂SeO_4; Sigma Aldrich, Darmstadt, Germany) at concentrations of 2 mg L⁻¹ (Se2), 5 mg L⁻¹ (Se5), and 10 mg L⁻¹ (Se10), selected from the literature to balance nutritional enhancement and physiological safety [19,20,27,28,29]. The electrical conductivity (EC) of the solutions was measured using a pH/conductivity meter (SevenCompact Duo; Mettler-Toledo GmbH, Greifensee, Switzerland) and ranged from 0.929 ± 0.008 dS m⁻¹ in the control (tap water) to 0.943 ± 0.005 dS m⁻¹ in Se aqueous solutions, proving that no salinity stress affected results [24].

The treatment was carried out over the following four days in a growth chamber (light intensity 320 µmol m⁻² s⁻¹; photoperiod 16/8 light/dark; temperature 20 ± 2 °C) under violet LED light (VGD Lumia, Mouans-Sartoux, France) with peak intensities at 450 and 630 nm. The control plants were grown in tap water throughout the experiment. The hypocotyls (stems and cotyledons) of the ten-day-old microgreen plantlets were harvested and stored at −80 °C for further analyses. For measurements of antioxidant parameters, composite plant tissue was powdered using liquid nitrogen and used immediately for analyses.

2.2. Physiological Stress Indicators

Lipid peroxidation intensity was determined as the amount of thiobarbituric acid reactive substances (TBARSs) determined by the TBA reaction [30]. The absorbance was measured spectrophotometrically (Specord 40, Analytik Jena, Jena, Germany) at 532 nm. The value for non-specific absorption at 600 nm was subtracted. The concentration of TBARS was expressed as nmol g⁻¹ fresh weight (FW) by using an extinction coefficient of 155 mM⁻¹ cm⁻¹.

For relative water content (RWC) and electrolyte leakage (EL) determination, whole cotyledons of each Brassica microgreen were cut at the base and used immediately after cutting. The RWC represents the proportion of water present in the tissue relative to its maximum water-holding capacity. For RWC [31,32], approximately 100 mg of the fresh cotyledons’ weight (FW) was determined after cutting, followed by immersing them in distilled water for 3 h at room temperature (RT) under dim light in order to determine turgid weight (TW). Finally, dry weight (DW) was determined after 24 h of oven-drying at 105 °C. RWC was calculated as the ratio of the actual water content to the fully hydrated water content and expressed as % RWC.

Damage to cell membranes was determined by measuring the rate of EL [33]. To remove residual nutrient solution and surface-adhered ions, 40 cotyledons per replicate were washed thoroughly with deionized water and then placed in closed glass tubes filled with 20 mL of deionized water. Initial conductivity of the solution (L1) was measured after 24 h of incubation at RT. The final conductivity was measured after the samples were autoclaved at 121 °C for 20 min and cooled to RT. The % of EL was calculated as EL (%) = (L1/L2) × 100, where L1 represents the initial conductivity and L2 the final conductivity after samples were autoclaved.

2.3. Biochemical and Antioxidant Analyses

For total soluble phenols, about 500 mg of powdered plant tissue was extracted with 2.5 mL of 70% methanol (Sigma Aldrich, Germany) overnight at −20 °C. The reaction mixture for the determination of total soluble phenols (Phe) contained 10 µL of supernatant, 190 µL of deionized H₂O, 25 µL of Folin−Ciocalteau reagent (Sigma Aldrich, Germany), and 75 µL of saturated Na₂CO₃ (Kemika, Croatia) solution [34,35]. The samples were incubated at 37 °C for 1 h and then cooled to RT. The absorbance was measured at 765 nm on a microplate reader (Tecan, Spark, Männedorf, Switzerland). The total phenolic content was expressed as equivalents of gallic acid (GAE, Sigma Aldrich, Germany) g⁻¹ fresh weight (FW).

The DPPH scavenging activity was determined according to the Brand–Williams method [36], as modified by Bibi Sadeer et al. [13], using the same extract prepared for PHE determination. The reaction mixture was prepared with 20 µL of microgreen extract and 180 µL of 0.04‰ DPPH (2,2-diphenyl-1-picrylhydrazyl, Sigma Aldrich, Germany). The absorbance was measured at 517 nm on a microplate reader after 30 min of incubation in the dark at RT with occasional shaking. The total antioxidant activity was determined from the standard curve and expressed as equivalents of Trolox (6-hydroxy-2,5,7,8-tetramethylchroman-2-carboxylic acid, Sigma Aldrich, Germany) g⁻¹ FW.

For the ferric reducing antioxidant power (FRAP) assay [37], the same extract prepared for the Phe determination was used. The reaction mixture contained 5 µL of microgreen extract and 180 µL of FRAP reaction mixture. The FRAP reaction mixture was prepared with 0.3 M acetate buffer (pH = 3.6), 10 mM TPTZ (2,4,6-Tris (2-pyridyl)-1,3,5-triazine, Sigma Aldrich, Germany) solution, and 20 mM FeCl₃ × 6H₂O (10:1:1 v/v, Kemika, Croatia). The reaction was carried out in the dark for 15 min at 37 °C. The absorbance was measured at 593 nm using a microplate reader, and the antioxidant activity was expressed as equivalents of Trolox (µmol g⁻¹ FW).

To determine the relative amount of anthocyanins (AC), 30−50 mg of powdered plant tissue was extracted in 600 µL methanol (Kemika, Croatia) +1% HCl (Kemika, Croatia) at −20 °C overnight. To the extracts, 400 µL ultrapure H₂O and 1 mL chloroform were added. The upper supernatant layer was separated into new tubes, and an equal volume of a mixture of methanol 1% HCl:ultrapure H₂O (60:40) was added. Finally, 200 µL of each sample was added to the wells of microtiter plates, and the absorbance was measured at 530 and 657 nm using a microplate reader. Total AC was calculated using the formula of Neff and Chory [38] and expressed as OD (optical density) g⁻¹ FW.

For proline (PRO) content [39], approximately 100 mg of powdered plant tissue was extracted with 400 µL in ethanol (Kemika, Croatia):water (40:60, v/v) overnight at 4 °C. The reaction mixture was prepared with 1% ninhydrin (Sigma Aldrich, Germany) in 60% acetic acid (Kemika, Croatia) and 20% ethanol. The mixture of 50 μL of extract and 300 μL of reaction mixture was mixed and heated to 95 °C for 20 min. The absorbance was measured at 520 nm using a microplate reader. The proline content was calculated using a standard curve with proline (Sigma Aldrich, Germany) as the standard and expressed in µmol g⁻¹ FW.

To determine the total soluble protein content (TSP) [40], approximately 500 mg of powdered plant tissue was extracted with a total of 2 mL of 100 mM potassium phosphate buffer (pH 7.0) with 2% polyvinylpyrrolidone (PVP, Sigma Aldrich, Germany). The reaction mixture contained 5 µL of extract and 250 µL of Bradford reagent (Sigma Aldrich, Germany). The absorbance was measured using a microplate reader at 595 nm, with bovine serum albumin (BSA, Sigma Aldrich, Germany) as the standard. Protein concentrations were expressed in mg g⁻¹ of FW.

For the determination of total soluble sugars (TSS) [41], approximately 20 mg of dry tissue was extracted with 4 mL of acetone (Kemika, Croatia), and the extraction was carried out at 4 °C overnight. To the precipitate, 1 mL of 80% ethanol was added, and the extraction was carried out in a water bath at 80 °C for 30 min. The extraction was repeated twice, and 500 µL of the combined supernatants was then added to 2 mL tubes and evaporated until dry in a water bath at 85 °C. To the residue, 500 µL of ultrapure H₂O was added, and the mixture was then used for the measurements. The reaction mixture consisted of 20 µL of sample, 80 µL of purified water, and 200 µL of anthrone (Sigma Aldrich, Germany) dissolved in 95% H₂SO₄ (Kemika, Croatia). The mixture was incubated for 30 min at 80 °C and cooled to RT [41]. The absorbance was measured on a microplate reader at 635 nm in polyethylene microplates, using glucose dilutions as a standard. The concentration of total soluble sugars was expressed in mg g⁻¹ of DW.

For guaiacol peroxidase (GPOD, EC 1.11.1.7) activity, the method described by Siegel and Galston [18] for GPOD determination was adapted for use with a microplate reader. Approximately 200 mg of powdered plant tissue was extracted with a total of 2 mL of 100 mM potassium phosphate buffer (pH = 7.0) with 2% polyvinylpyrrolidone (PVP). The reaction mixture was prepared using 8 mM H₂O₂ (Kemika, Croatia) and 90 mM guaiacol (Sigma Aldrich, Germany) at a 1:1 (v/v) ratio. All measurements were taken in triplicate in 96-well plates using a microplate reader by adding 150 µL of phosphate buffer, 40 µL guaiacol/H₂O₂ mix, and 10 µL of extract. Absorbance was read at 436 nm using an extinction coefficient of 25.5 mM⁻¹ cm⁻¹, and results were expressed as U_GPOD g⁻¹ FW.

2.4. Fourier Transform Infrared Spectroscopy (FT-IR) Measurements

To screen four Brassica microgreens exposed to Se treatments, FT-IR spectroscopy was used. Approximately 10 g of powdered leaves was extracted with 50 mL of ethanol and vigorously stirred on a mechanical shaker for 3 h. The supernatants were collected and dried, and then mixed with KBr. Sixteen absorbance spectra were measured between 400 and 4000 cm⁻¹ spectrophotometrically (FTIR−8400S, Shimadzu, Tokyo, Japan). The spectra were baseline-corrected. Peak maxima for each Brassica microgreen at control and three Se concentrations.

2.5. HPLC Analysis of Polyphenolic Compounds

Before HPLC analysis, the crushed samples were prepared by extracting 1 g of dry plant material with 20 mL of solvent (70% EtOH) with an ultrasonically assisted extractor (UAE UP400ST/Sonotrode S24d3-Hielscher Ultrasonics, Teltow, Germany) for 3 min. The extract was then subjected to pre-filtration through a 0.22 µm syringe filter for analysis.

Determination of bioactive components for Se-treated microgreens extracts was performed by the HPLC method with UV detection [42] on Cosmosil 5C18−MS−II column (Nacalai Tesque, Inc., Kyoto, Japan), 250 mm long with an internal diameter of 4.6 mm, filled with 5 μm particles. The HPLC system used for the analysis (Agilent, 1260 Infinity II series, Waldbronn, Germany) consisted of a quaternary pump (G7111A), a DAD (photo-diode array) detector (G7115A), a column heater with a temperature range from 10 to 85 °C (G7116A), an autosampler with a capacity of 36.6 mL vials (G7157A), and a fraction collector (G1364E). The system was run using an Open Lab CDS ChemStation Edition program. The separation of the analyzed compounds was carried out by gradient elution at a flow rate of 1 mL min⁻¹ for 65 min. As phases A and B, 0.1% formic acid (Sigma Aldrich, Germany) in Nirosta Ultrapure systems water (Nirosta, Osijek, Croatia) and 0.1% formic acid in methanol, respectively, were used. The gradient was set as follows: 0.00−8.00 min 90% A; 8.00−16.00 min 75% A; 16.00−28.00 min 55% A; 28.00−55.00 min 20% A; from 55.00−65.00 min 10% A followed by a period of 15 min where the analysis conditions returned to the initial value. The injection volume was 10 µL; the temperature of the column compartment was 50 °C, and the UV detection wavelengths were 240, 250, 260, 270, 280, 330, and 360 nm. Purity of peaks is calculated through ChemStation software (Agilent OpenLAB CDS ChemStation Edition software (version B.01.03.092, Agilent Technologies, Santa Clara, CA, USA)), and all analytes are verified through purity factor, spectra threshold, and noise threshold. Determined LOD is 0.2 mg L⁻¹ for most components (LOQ = 0.66 mg L⁻¹), and just for myrcetin, LOD is 1.2 mg L^−1, and calculated LOQ is 3.6 mg L⁻¹. All standard concentrations are calculated to 100% of purity regardless of whether some have lower purity. All solvents were of HPLC grade.

2.6. Statistical Analysis

Statistical analyses for four selenium-treated Brassica microgreens were performed using Statistica (v.14.0.1.25; Tibco Software Inc., Palo Alto, CA, USA). To minimize bias in comparisons among Brassica microgreens differing in their baseline responses to Se biofortification, all data were normalized to their respective control values (100%). Absolute mean values with standard deviations, as well as factorial ANOVA p-values for species, selenium treatment, and their interaction, are presented in Supplementary Tables S4 and S5. Between-treatment differences in measured parameters of the four Brassica microgreens were evaluated using factorial analysis of variance (ANOVA) followed by the Tukey Honest Significant Difference (HSD) post hoc test. Significance of difference was considered at p ≤ 0.05. Factorial ANOVA was conducted considering species, selenium treatment, and their interaction as fixed factors. The corresponding p-values for factor effects and interaction terms are provided in Supplementary Table S3. The results are presented as mean ± standard deviation (SD) of six replicates (n = 6), except for RWC and EL, where four replicates (n = 4) were taken due to methodological and sample handling limitations associated with these measurements.

Correlation between measured parameters of the control and Se-treated Brassica microgreens was revealed by principal component analysis (PCA). It was performed using a correlation matrix of average values after autoscaling, with XLSTAT software (ver. 2025.2.0) implemented in Microsoft Office Excel (version 1808). Linear correlations between Brassica microgreens and treatments were assessed using Pearson correlation coefficients, and differences were considered significant at p ≤ 0.05. The factor loadings for the parameters are presented in Supplementary Table S2. All graphical presentations were created with Microsoft Excel.

3. Results

3.1. Oxidative Stress and Membrane Integrity

The level of lipid peroxidation, expressed as relative TBARS content (Figure 2a), varied significantly among Brassica microgreens and selenium treatments. While broccoli exhibited higher TBARS values across treatments, kale consistently had lower TBARS values than the control. In pak choi, TBARS decreased under higher selenium concentrations compared to the control. Kohlrabi, however, showed no significant change in response to Se treatments compared with the control.

The electrolyte leakage (EL) showed noticeable variability among both Brassica microgreens and treatments (Figure 2b). In broccoli, EL increased at Se5 and Se10 compared to the control. A similar response, an increase in EL at Se5 and Se10, was observed in kale, with Se5 inducing the highest EL value. While in pak choi, EL values remained relatively stable across treatments, with no significant difference between Se treatments, in kale, Se10 induced the highest increase relative to the control.

Relative water content (RWC) remained largely stable across all Brassica microgreens and treatments (Figure 2c). Minor variations were observed, with a slight decrease at Se5 in pak choi and a slight increase at Se10 in kale; however, these differences were not significant compared to the control.

Overall, the investigated Brassica microgreens differed mainly in membrane-related responses to selenium treatment, with broccoli and kale showing greater EL sensitivity, while pak choi maintained the most stable physiological response across treatments.

3.2. Metabolic and Antioxidant Responses

The content of total soluble sugars (TSS; Figure 3a) significantly increased in kohlrabi and pak choi with increasing selenium concentration, reaching the highest values at Se10. In contrast, broccoli exhibited consistently low TSS values across all treatments compared to the control. Kale, however, exhibited a relatively stable TSS content, with no clear dose-dependent trend and no significant changes relative to the control.

The proline content (PRO) increased with selenium concentration in most Brassica microgreens (Figure 3b). The highest accumulation was observed in broccoli and pak choi at Se10, whereas kohlrabi showed no significant increase. Kale showed elevated PRO levels across all treatments, with a significant increase at Se5 and Se10 when compared to the corresponding control.

The content of total soluble proteins (TSP) varied in response to Se treatment among Brassica microgreens (Figure 3c). Broccoli, kohlrabi, and kale exhibited the highest protein content at Se2 and Se10. While kohlrabi showed no significant difference at Se5, broccoli and kale showed an increase in TSP at Se5 compared to the control; however, the values were significantly lower than those at Se2 and Se10. Pak choi revealed a significant increase in TSP exposed to Se5 and Se10 compared to the control.

The activity of guaiacol peroxidase (GPOD) increased across all Brassica microgreens at higher Se concentrations, namely Se5 and Se10 (Figure 3d). The most pronounced increase was observed in kohlrabi, where all three Se concentrations induced significant increases. Broccoli and pak choi showed no significant difference at Se2, whereas kale showed a significant decrease at the same concentration compared with the control.

Overall, selenium treatments induced pronounced species-specific changes in osmolyte accumulation, protein content, and antioxidant enzyme activity, with kohlrabi and pak choi showing stronger osmotic adjustment responses, while kale and broccoli exhibited more pronounced PRO- and GPOD-related responses at higher selenium concentrations.

3.3. Phenolic Content and Antioxidant Capacity

Total phenolic content (PHE) showed clear species-dependent responses to selenium (Figure 4a). The most pronounced increase was observed in broccoli, with a significant increase in Se concentration, while kohlrabi showed a significant increase when exposed to Se5 and Se10 compared to the control. Pak choi and kale, however, showed no significant difference compared to the control at any applied Se concentration.

The changes in the relative number of anthocyanins (AC) are shown in Figure 4b. The most stable AC content was observed in kohlrabi, showing no significant difference between control and Se treatments. While broccoli and kale showed an increase in AC when exposed to all Se concentrations, pak choi showed a significant increase at higher Se concentrations (Se5 and Se10) compared to their corresponding control.

Total antioxidant capacity was estimated by DPPH and FRAP assays (Figure 4c,d) and species- and treatment-dependent responses to selenium. DPPH radical-scavenging activity (Figure 4c) in broccoli and pak choi showed significantly lower values when exposed to Se2 and Se5, respectively, whereas Se10 induced no significant change relative to the control. In contrast, kohlrabi showed a significant increase only with the Se10 treatment compared to the corresponding control, whereas kale showed increased DPPH activity across all Se concentrations, with the highest value observed at Se5. The ferric reducing antioxidant power (FRAP) showed a response similar to DPPH activity in broccoli and kale, while Se treatment significantly decreased FRAP values in broccoli; in kale, all Se treatments induced higher values compared to control. However, both kohlrabi and pak choi showed no significant difference at any Se treatment compared to the control.

Overall, selenium treatments induced pronounced species-specific changes in phenolic compounds and antioxidant capacity, with broccoli and kale showing stronger antioxidant-related responses, while pak choi and kohlrabi exhibited comparatively more moderate or selective changes depending on the antioxidant parameter evaluated.

3.4. Individual Phenolic Profiles

The profile of individual phenolic compounds showed pronounced species- and treatment-dependent variation (Figure 5).

Caffeic acid content increased with selenium concentration across all species, with the highest accumulation observed in broccoli and pak choi at Se10, where it increased up to 5 times compared to the corresponding controls. Kohlrabi showed a moderate increase, up to two times compared to control, while kale exhibited consistently low levels across treatments. Trans-ferulic acid showed a different trend, with the highest values (1.7 times higher than the control) recorded in broccoli at Se5, followed by a slight decrease at Se10 (1.3 times higher than the control). In pak choi, ferulic acid was stable at Se2 and Se5, but decreased almost 3.5 times at Se10. Kohlrabi and kale showed relatively stable values. Sinapic acid was the dominant phenolic compound in broccoli and kohlrabi, whereas in pak choi and kale, its initial levels were lower. In broccoli, its content increased with selenium concentration, reaching a two times higher value at Se10, the same as in pak choi. In contrast, kohlrabi and kale showed relatively stable or slightly decreasing levels at higher selenium concentrations, respectively.

Among flavonoids, myricetin was present in all species except kale in the control, and its levels were relatively stable across treatments. Quercetin was detected in broccoli, kohlrabi, and kale, but not in pak choi, and generally showed minor fluctuations at lower Se concentrations (Se2 and Se5). However, Se10 decreased by almost 2-fold in kohlrabi. In contrast, kale exhibited 1.8 times higher values at Se10 compared to the control. Apigenin and kaempferol were detected only in selected species and treatments, indicating species-specific metabolic profiles. In general, selenium treatment altered both the composition and concentration of individual phenolic compounds, with the most pronounced effects observed in broccoli and pak choi.

3.5. FTIR Spectral Analysis

FTIR analysis revealed species-specific shifts in peak positions across key spectral regions following selenium treatment (

Table 1. Species-specific shifts in FTIR peak positions (cm⁻¹) induced by selenium biofortification in Brassica microgreens.

Region (cm−1)	Assignment	Brassica Microgreen
		Broccoli	Kohlrabi	Pak Choi	Kale
~3300	O−H	−19 ↓	+79 ↑	−15 ↓	+16 ↑
~2900	C−H	0	−2 ↓	0	+2 ↑
~1730	C=O	−1 ↓	+23 ↑	n.d.	−171 ↓
~1630	Amide I	+24 ↑	+21 ↑	+27 ↑	−19 ↓
~1450	Proteins	+41 ↑	−6 ↓	+124 ↑	−4 ↓
~1020	Carbohydrates	+27 ↑	+21 ↑	−2 ↓	+4 ↑
~860	Aromatics	0	n.d.	+2 ↑	+2 ↑
650−540	Fingerprint	−36 ↓	+37 ↑	+2 ↑	+39 ↑

Shifts in characteristic spectral regions indicate Se-related molecular reorganization of specific molecules. Positive (↑) and negative (↓) values indicate shifts toward higher or lower wavenumbers, respectively; n.d.—not detected; blue—strong negative shift; yellow—weak negative shift; pink—strong positive shift; white—weak positive shift.

and Table S1). Maxima peak shift differences were calculated as the difference between the specific treatment and its corresponding control. Observed shifts in FTIR peak positions may be associated with selenium-induced changes in functional group-related spectral regions and/or molecular composition. Therefore, our results showed that in the O–H stretching region (~3300 cm⁻¹), kohlrabi exhibited a strong positive shift, while kale showed a moderate increase. In contrast, broccoli and pak choi displayed negative shifts. Minimal changes were observed in the C–H region (~2900 cm⁻¹), with shifts ranging from −2 cm⁻¹ in kohlrabi to +2 cm⁻¹ in kale, indicating relatively stable lipid-associated vibrations. In the carbonyl region (~1730 cm⁻¹), kohlrabi showed a positive shift, whereas kale exhibited a pronounced negative shift. Broccoli showed only minor variation, while no data were detected for pak choi in this region. The amide I region (~1630 cm⁻¹) showed consistent positive shifts in broccoli, kohlrabi, and pak choi, whereas kale exhibited a negative shift. In the protein-associated region (~1450 cm⁻¹), strong positive shifts were observed in pak choi and broccoli, while kohlrabi and kale showed negative ones. The carbohydrate region (~1020 cm⁻¹) showed moderate positive shifts in broccoli, kohlrabi, and kale, while pak choi exhibited a slight negative shift. In the aromatic region (~860 cm⁻¹), no change was observed in broccoli, small positive shifts were recorded in pak choi and kale, while in kohlrabi, the aromatic region was not detected. Finally, the fingerprint region (650–540 cm⁻¹) showed contrasting responses, with positive shifts in kohlrabi, pak choi, and kale, while broccoli exhibited a notable negative shift. Overall, selenium treatment induced distinct structural changes across species, particularly in regions associated with hydroxyl groups, proteins, and low-frequency molecular vibrations.

3.6. PCA

Principal component analysis (PCA) highlights distinct metabolic strategies among species in response to selenium treatment. It explained 68.84% of the total variance, with the first principal component (F1) accounting for 49.11% and the second component (F2) for 19.72% (Figure 6).

The F1 axis clearly separated the samples based on antioxidant-related parameters. Positive loadings on F1 were associated with total soluble sugars (TSS), proline (PRO), relative anthocyanin content (AC), DPPH, and FRAP, indicating that these variables contributed strongly to antioxidant potential. In contrast, negative loadings were associated with GPOD activity, lipid peroxidation (TBARS), electrolyte leakage (EL), and relative water content (RWC). The F2 axis was mainly influenced by phenolic content (PHE) and total soluble proteins (TSP), which showed strong positive loadings, whereas RWC contributed negatively to this axis.

Sample distribution revealed clear species- and treatment-dependent clustering. Kohlrabi samples, particularly at higher selenium concentrations (Se5 and Se10), were positioned on the positive side of F1, reflecting high antioxidant activity and osmolyte accumulation. In contrast, control samples and lower selenium treatment were generally grouped on the negative side of F1. Pak choi samples were more dispersed along the F2 axis, indicating variability associated with phenolic content and protein-related responses. Selenium-treated broccoli samples were positioned closer to stress-related variables (TBARS, EL, GPOD), while Se-treated kale samples showed separation toward antioxidant-related variables, particularly DPPH and FRAP.

4. Discussion

Recently, selenium (Se) biofortification has been shown to be a vital practice for improving both the nutritional value and stress resistance of crops. Our results highlight substantial interspecific variability in selenium-responsive antioxidant strategies among Brassica microgreens. Selenium was applied as sodium selenate at concentrations previously shown to induce measurable selenium accumulation without severe impairment of photosynthetic performance in Brassica microgreens [24]. Selenium content increased noticeably with increasing Se concentration in all investigated Brassica microgreens, with kale and kohlrabi exhibiting the highest accumulation potential. Moreover, preliminary results clearly demonstrated that Se biofortification enhanced photosynthetic performance and promoted acclimation responses, particularly in broccoli, kale, and kohlrabi, thereby implying increased photoprotection. In addition, pak choi, kale, and kohlrabi revealed well-regulated energy distribution throughout the photosynthetic apparatus, contributing to the high overall Se tolerance. Beyond stress-related responses, the observed changes in total soluble sugars (TSS), total soluble proteins (TSP), phenolic compounds (PHE), anthocyanins (AC), and antioxidant capacity additionally suggest that selenium biofortification influenced important nutritional and health-promoting traits of Brassica microgreens. These findings support the potential of selenium enrichment not only for improving stress resilience but also for enhancing the functional quality of microgreens intended for human consumption. Our further interest included investigation of oxidative response in the same four Brassica microgreens after Se biofortification, revealing that the efficiency of these responses depended on the metabolic organization of specific Brassica microgreens, as well as on the balance between different antioxidant pathways. Recent advances in single-cell and spatial transcriptomic approaches further suggest that stress adaptation involves substantial cell-type-specific regulatory heterogeneity, which may contribute to the species-specific selenium responses observed in the present study. Such approaches could provide valuable future insight into selenium-responsive antioxidant regulation and metabolic coordination at cellular resolution, beyond the whole-tissue responses investigated here [43]. Recent studies additionally suggest that ROS/RNS signaling interactions play an important role in coordinating antioxidant regulation and redox-mediated stress responses in plants under environmental stress conditions [44]. FTIR spectral shifts observed in regions associated with proteins, carbonyls, and O–H groups may reflect selenium-related changes in antioxidant metabolism, membrane-associated compounds, and phenolic interactions [45,46]. In particular, shifts in protein-related regions may be associated with stress-related protein remodeling and antioxidant enzyme activation, while changes in aromatic and fingerprint regions may indicate modifications in phenolic composition and secondary metabolite interactions. However, direct linkage to specific metabolic pathways would require targeted molecular and metabolomic analyses.

4.1. Effect of Se Biofortification on Broccoli

Broccoli microgreens showed relatively high TBARS levels compared with the control, accompanied by increased electrolyte leakage (EL), despite increased guaiacol peroxidase (GPOD) activity, suggesting increased susceptibility to Se-induced oxidative stress. Such results are consistent with the dual effect of Se, which at lower and moderate Se concentrations may stimulate antioxidant defense mechanisms while at higher concentrations it can promote the formation of reactive oxygen species (ROS), thereby acting as a pro-oxidant [47,48]. In that case, Se can induce a redox imbalance and stimulate ROS production, leading to oxidative stress and membrane damage [48,49]. This suggests that protective responses induced by selenium were not always sufficient to fully prevent oxidative and membrane-related damage under elevated Se exposure. Increased EL further confirms membrane destabilization under Se treatment since loss of membrane integrity is a common consequence of oxidative stress [50,51]. Such results suggest the presence of oxidative stress and impaired membrane integrity, thus indicating inadequate protection from oxidative stress [52]. An increase in GPOD activity confirms activation of the enzymatic antioxidant defense, while an increased TBARS level suggests that the oxidative stress is still present. While selenium enhances the antioxidant activity, including peroxidases [53,54], increased ROS production often triggers an enzymatic response; however, this does not necessarily induce complete protection [55,56]. Such a response is further supported by positive shifts in the Amide I region (~1630 cm⁻¹), indicating certain modifications in protein structure and increased involvement of stress-responsive enzymes, such as GPOD. This kind of spectral change might be associated with modifications of protein conformation caused by stressful conditions [57,58]. The PCA showed broccoli was mainly positioned on the negative side of the F1 axis, clustering closer to the variables related to oxidative stress and enzymatic defense. The shift toward GPOD suggests that in broccoli, the primary response to selenium treatments is activation of the enzymatic antioxidant mechanism [56].

Broccoli microgreens showed a pronounced increase in antioxidant-related parameters, such as total soluble proteins (TSP), total soluble phenolics (PHE), and anthocyanins (AC), as well as proline (PRO), at the highest Se concentration, suggesting strong activation of both osmotic and non-enzymatic antioxidant responses. Increased TSP content might reflect enhanced synthesis of stress-related enzymes, while increased PRO accumulation is recognized as a response to oxidative stress, which can function both as an osmoprotectant and a ROS scavenger [56]. Interestingly, despite increases in PHE and AC, total antioxidant capacity, as determined by the DPPH and FRAP methods, and total soluble sugar content (TSS) did not increase proportionally. Moreover, DPPH, FRAP, and TSS levels decreased after Se treatment, suggesting that antioxidant capacity depends more on compound composition and structure rather than on total concentration. This is corroborated by previous findings showing that antioxidant activity is influenced not only by total PHE content but also by chemical structure, degree of hydroxylation, and specific interactions among phenolic compounds [59,60]. Observed increase in specific phenolic acids, particularly caffeic and sinapic acids, indicates that antioxidant activity depends on phenolic structure and their interaction with other metabolites [61]. This is further supported by FTIR analysis since the weak negative shift in the O−H region (~3000 cm⁻¹) indicates modifications in phenolic interactions, suggesting possible selenium-associated changes in functional group-related spectral regions rather than just increased accumulation [62,63]. Since the antioxidant capacity of phenolic compounds strongly depends on the number, position, and availability of −OH groups, such alterations might change the accessibility of hydroxyl groups responsible for ROS scavenging and, in that way, affect DPPH and FRAP response [61]. PCA revealed a separation of the higher Se treatments (Se5 and Se10) from the control and Se2 toward stress-related variables, which suggests that elevated Se concentrations enhance metabolic activity. Despite this, their distance from non-enzymatic antioxidant variables suggests that the contribution of phenolic-driven antioxidant activity is insufficient. Unlike other Brassica microgreens, broccoli’s response to selenium biofortification relied primarily on enzymatic antioxidant activation.

4.2. Effect of Se Biofortification on Kohlrabi

Kohlrabi microgreens showed a relatively stable response to Se biofortification with only a few significant changes compared to other Brassica microgreens. Despite this, an increase in GPOD, TSP, and TSS suggested activation of both enzymatic defense and primary metabolism. The marked increase in GPOD activity across all applied Se concentrations, accompanied by stable TBARS and RWC levels relative to the control, suggests efficient enzymatic antioxidant activity. This is consistent with recent studies showing that Se enhances peroxidase activity as part of competent ROS detoxification systems [64,65]. However, a significant increase in EL at the highest Se concentration indicates that membrane integrity remains compromised despite activation of the antioxidant enzyme system, suggesting that oxidative stress was not fully alleviated at this concentration.

It was recently shown that Se treatment can increase TSS, which can serve as a key osmoprotectant that maintains cellular osmotic balance under stressful conditions [66,67,68]. Based on this, a significant increase in TSS with increasing Se concentration in kohlrabi microgreens suggested osmotic adjustment and carbohydrate-mediated metabolic regulation. At the same time, PRO content showed no significant change, suggesting that osmotic adjustment is primarily driven by carbohydrate accumulation rather than by proline accumulation. Proline accumulation is often variable and may depend on stress intensity and plant species [69]. Moreover, selenium-induced enhancement of sugar metabolism was shown to be closely linked to antioxidant responses and improved stress tolerance [70], especially at lower Se concentrations (Se2 and Se5). However, the observed accumulation of TSS and TSP at Se10 was insufficient to completely prevent membrane damage. This is further supported by FTIR results, where predominantly positive shifts across spectral regions indicate reorganization in functional group-related spectral regions and changes in molecular arrangement, especially in molecules associated with membranes, such as lipids and proteins. Such deviations might be connected with increased membrane permeability and oxidative damage at the highest Se concentration. Recent studies have revealed that Se can enhance antioxidant activity and osmoprotectant accumulation, but at higher concentrations, EL and membrane instability may still occur, reflecting inadequate protection against oxidative stress [71,72]. However, an increase in EL despite increases in TSS and GPOD suggests that antioxidant and osmoprotective responses were not sufficient to fully maintain membrane integrity at higher Se concentrations, possibly due to incomplete ROS scavenging [52,73].

Unlike other Brassica microgreens, PCA revealed that kohlrabi microgreens were separated and clearly clustered on the positive side of the F1 axis, which is strongly associated with DPPH, FRAP, AC, PRO, and TSS. A switch among osmoprotectants suggests species–specific adaptation strategies [69]. Further, Se10 increased DPPH, while FRAP remained unchanged, suggesting an enhancement of radical-scavenging capacity rather than of overall reducing power. This indicates variations in antioxidant compounds composition since different phenolics and other metabolites contribute differentially to DPPH and FRAP responses [74]. Moreover, Se altered the accumulation of individual phenolic acids rather than evenly increasing all compounds. Specifically, changes in caffeic and sinapic acids usually contribute more to increased radical activity (DPPH), whereas ferric reducing capacity, measured by FRAP, may not have been altered [61]. Shifts in the O−H stretching region (~3300 cm⁻¹) might suggest reorganization in spectral changes associated with biochemical responses rather than simple accumulation [63], thereby influencing the DPPH response more strongly than the FRAP response, which corroborates our results. Overall, kohlrabi microgreens exhibited coordinated antioxidant and osmoprotective responses with moderate membrane sensitivity at higher Se concentrations.

4.3. Effect of Se Biofortification on Pak Choi

Selenium biofortification in pak choi microgreens induced a more coordinated protective response than the one observed in broccoli or kale. It is characterized by reduced TBARS levels at higher Se concentrations (Se5 and Se10) without significant changes in EL and RWC. Such results indicate enhanced protection against oxidative damage while maintaining membrane stability [56] without triggering stress response, which is associated with effective oxidative stress mitigation [75]. This is corroborated with minimal shifts in FTIR spectra, especially in the carbonyl and C−H regions, suggesting protection of membrane structures [76]. Moreover, an increase in TSS in pak choi across all Se concentrations, together with increases in TSP at Se5 and Se10, and in PRO at Se10, suggests that osmotic and metabolic responses contributed considerably to stress tolerance [69]. Therefore, soluble sugars acted most likely as primary osmoprotectants, while proline acts as an additional osmoprotectant only under higher stress conditions. Soluble carbohydrates often function as major osmoprotectants by scavenging ROS and thus protecting membrane structure and proteins [77], while proline accumulation is commonly increased only at intense stress conditions [78]. Increased GPOD activity at Se5 and Se10 further supports the protection of membrane structures by activating the enzymatic detoxification of ROS [64].

Further, phenolic content and FRAP remained unchanged, while DPPH decreased at Se2 and Se5, followed by recovery to the control level at Se10. Such a response additionally suggests that the antioxidant response in pak choi was mainly driven by the accumulation of osmolytes and the activation of enzymatic defense rather than by non-enzymatic radical scavenging. It was recently shown that osmoprotectants and antioxidant enzymes may act as primary protective components, even without activation of the phenolic-based antioxidant response [38,51], especially at lower Se concentrations, where reduced oxidative pressure may diminish the requirement for non-enzymatic radical scavenging. Additionally, recovery of DPPH at Se10, together with increased proline and anthocyanins (AC), suggests activation of an additional protective response under the highest Se concentration. Together with the reduced DPPH response at lower Se concentrations, this response may additionally indicate lower oxidative pressure due to efficient ROS scavenging, thus confirming the reduced need for non-enzymatic radical scavenging [52]. This is further supported by positive shifts in Amide I (~1630 cm⁻¹) and protein-related (~1450 cm⁻¹) regions, suggesting changes in functional group-associated regions, structural reorganization of stress-responsive proteins, while a minor shift at ~1020 cm⁻¹, designated to the carbohydrate region, corroborated osmotic adjustment related to the soluble sugars. Such spectral changes can be related to adaptive molecular responses [57,58]. The large positive shift observed in the ~1450 cm⁻¹ region indicates a shift from Amide I to Amide II/aromatic region, which might suggest possible protein−phenolic interactions [62]. Such interactions may increase the stabilization of proteins or modulate phenolic role under Se exposure [79], which could explain the coordinated biochemical exposure at Se10. Moreover, such a pronounced shift toward ~1541 cm⁻¹ might reflect an increased contribution of aromatic/flavonoid-related vibrations [33,34], suggesting that anthocyanins may act as a secondary protective response in pak choi under the highest Se concentration. Alterations in the phenolic acid profiles, particularly caffeic and sinapic acids, despite the fact that total phenolic content remained unchanged, might affect radical scavenging activity [61], thus explaining why DPPH and FRAP response did not follow total PHE content. This is supported with shifts in the O−H region, suggesting modifications in phenolic interactions rather than its functionality [62].

Such a response of pak choi microgreens is supported by PCA, in which pak choi samples were distributed along the F2 axis, particularly at Se5 and Se10, which are primarily associated with phenolic- and protein-related variables. This suggests a coordinated response that involves protein remodeling, osmotic adjustment, and activation of secondary metabolite-related protection rather than dependence on a single dominant defense pathway [80].

4.4. Effect of Se Biofortification on Kale

Kale microgreens revealed a distinct response to selenium biofortification that includes simultaneous activation of the enzymatic and non-enzymatic antioxidant defense system, followed by reduced TBARS level across all Se concentrations. Decline in lipid peroxidation intensity was accompanied by increased GPOD activity, suggesting efficient ROS scavenging. At the same time, a significant increase in AC, unlike in kohlrabi and pak choi, together with increased FRAP and DPPH, indicates activation of complementary non-enzymatic antioxidant protection. This coordinated response supports previously reported findings that Se enhances both metabolic and antioxidant pathways in plants, thus operating synergistically rather than individually [52,54]. The results also revealed stable levels of phenolic compounds across all three Se concentrations, indicating that the antioxidant response was driven by changes in specific phenolic compounds rather than by overall PHE accumulation. A significant increase in AC, accompanied by variations in trans-ferulic acid, might contribute to enhanced radical scavenging [81] due to phenolic composition [61] and alterations in phenolic interactions [62,63]. Moreover, noticeable shifts in fingerprint (650−540 cm⁻¹) and carbonyl region suggest molecular reorganization related to secondary metabolites. In particular, remodeling of cell wall phenolics [58] and membrane structures suggested that Se biofortification in kale simulated changes in antioxidant response at both biochemical and molecular levels. Based on this, increased EL despite reduced lipid peroxidation suggests that membrane stability may still be partially affected under elevated selenium exposure [82] due to the dual pro-oxidant role of Se under excessive accumulation [52]. Moreover, increased EL might reflect the physiological cost of adaptation [28,45,46]. Changes in trans-ferulic acid, followed by shifts in the fingerprint region, suggest cell wall phenolics remodeling and changes in membrane to cell wall interactions, which can disturb ion permeability. However, controlled ion efflux has been proposed to contribute to metabolic adjustment and stress acclimation under moderate stress conditions [83].

The accumulation of PRO and TSP, while TSS remained stable, further supported the hypothesis that metabolic and molecular composition adjustments partially confer stress tolerance. Unlike other Brassica microgreens investigated here, osmotic adjustment in kale depends primarily on proline accumulation, thereby maintaining membrane stability [50], whereas increased TSP likely contributes to the induction of stress-responsive proteins and enhanced protective metabolism [84]. PCA additionally supported a distinct response strategy to Se biofortification. Higher Se concentrations, namely Se5 and Se10, were separated from the control along the positive side of the F1 axis and positioned closer to variables related to antioxidant response, such as DPPH, FRAP, AC, and PRO. This clustering indicates that responses in kale revealed more pronounced non-enzymatic antioxidant capacity compared to the other three Brassica microgreens, followed by metabolic adjustment, and not only the enzymatic defense [80]. This is consistent with increased AC, radical-scavenging activity, and PRO accumulation, suggesting coordinated activation of antioxidant response [85] and protective secondary metabolite accumulation [86] under Se biofortification.

5. Conclusions

Based on the obtained results of selenium biofortification on four Brassica microgreens, it can be concluded that selenium induced a moderate response to stress by triggering pronounced, species-specific physiological, metabolic, and biochemical responses, revealing different adaptive strategies. While the response in broccoli depended mostly on enzymatic defense, pak choi maintained stress tolerance through balanced metabolic adjustment. Kohlrabi showed the most efficient response between osmotic adjustment, antioxidant capacity, and redox regulation, whereas kale displayed pronounced non-enzymatic antioxidant response accompanied by indications of physiological adjustment at higher selenium concentrations. Combined biochemical, FTIR, and PCA analyses demonstrated that selenium effects were driven by changes in metabolite accumulation as well as molecular reorganization and modifications in antioxidant functionality. Our results highlight the potential of selenium biofortification to enhance the nutritional quality and stress resilience of Brassica microgreens. Our findings provide valuable insight into species-specific mechanisms involved in selenium responses in Brassica microgreens, including osmolyte accumulation, antioxidant enzyme activation, non-enzymatic antioxidant responses, and metabolic adjustment. Although Se biofortification improved several functional and nutritional traits, higher selenium concentrations induced oxidative and membrane-related stress responses in certain species, highlighting the importance of species-specific optimization of Se application to maximize crop quality while minimizing potential effects of excess Se accumulation. Future studies, including single-cell and spatial transcriptomic approaches, direct ROS quantification, would provide additional mechanistic insight into selenium-induced oxidative signaling and antioxidant regulation in Brassica microgreens.