Effect of Sodium Sulfate Treatment on the Modulation of Aliphatic Glucosinolates in Eruca sativa Mill Organs at Flowering Stage

Featured Application

A simple method to elicit the synthesis of glucosinolates in different organs of Eruca sativa can have multiple applications. Increasing their concentration in leaves can make this species a more powerful functional food but boosting them in organs not usually used for food consumption can make them new matrices that can be used for the development of bioproducts, in the areas of health, agriculture, and food innovation.

Abstract

Glucosinolates are secondary metabolites of the Brassicales, playing a role in plant protection and as health-promoting compounds. Here, Na₂SO₄ was used to modulate the aliphatic glucosinolate content in different organs of Eruca sativa Mill. In flowers, which accumulate the highest amount of glucosinolates, Na₂SO₄ increased the concentration of glucoraphanin, in roots of glucoerucin and in apical leaves it doubled the amount of dimeric 4-mercaptobutyl glucosinolate. The biosynthetic gene Branched-Chain Aminotransferase 4 was also induced in roots at the highest salt concentration, while in leaves all tested genes biosynthetic genes were downregulated or unaffected. Cytochromes P450 83A1 monooxygenase was downregulated at the highest salt concentration in all organs. Overall, E. sativa is a reliable source of glucosinolates, which can be modulated with Na₂SO₄.

1. Introduction

Glucosinolates (GSLs) constitute a group of sulfur-rich secondary plant metabolites characterized by a variable side chain (R group) derived from amino acids, and a fixed core formed by a S-β-D-glucopyrano unit linked to an O-sulfated (Z)-thiohydroximate function [1]. Their presence has been reported in various tissues of plants belonging to the Brassicales order, with the Brassicaceae family having the highest level of biodiversity with respect to plant species synthesizing GSLs and GSL structures. Recent findings have also revealed their presence in orders other than Brassicales, such as in Malpighiales [2]. GSLs can be classified according to their precursor amino acids in (i) aliphatic GSLs (AGs), deriving from methionine, alanine, leucine, isoleucine, valine, or glutamate, (ii) benzenic GSLs from phenylalanine and tyrosine, and (iii) indole GSLs from tryptophan [3].

AGs represent the most diverse group of GSLs in Arabidopsis thaliana and the main class found in the species of the order Brassicales, especially in Brassicaceae, Capparaceae and Cleomaceae [4]. AGs are constitutively synthesized in the epidermal cells and along the vascular bundles where they contribute to the specialized defense strategies: when activated by the enzymes myrosinase upon tissue damage due to the attack of herbivores or other biotic agents (bacterial and fungal pests), they are hydrolysed to isothiocyanates, nitriles and other active molecules. Despite indolic GSLs being usually described as the main GSLs involved in pattern triggered immunity and effector triggered immunity [5], modulations in AGs have also been described in response to biotic [6] and abiotic stress, such as saline and temperature [7,8,9]. Moreover, recent evidence depicts GSLs, particularly AGs, as part of the natural leaf recruitment of non-pathogenic bacteria, which may protect plants against disease-causing microorganisms, and as a sulfur reservoir able to counteract sulfur deficiency in the sulfur cycling in plants [10,11].

AGs show a high structural diversity, since the precursor amino acid undergoes a series of chain elongation steps, giving rise to compounds with different side chain length, known as short- and long-chained, depending on the number of methylene groups. The biosynthetic pathway leading to AGs accumulation has been elucidated first in the model plant A. thaliana, followed by other Brassicaceae crops (for a review refer to Qin et al., 2023 [12]), and it can be summarized into three major steps: the chloroplast located amino acid chain elongation (step 1), the core structure assembly (step 2, cytoplasmatic) and the side chain modification (step 3, cytoplasmatic). A complex interplay exists between GSL synthesis, primary metabolism and phytohormones, which affects normal plant growth [12]. As for other secondary metabolisms, GSL biosynthesis is costly for plant cells, and it must, therefore, be tightly regulated, activated and repressed in response to environmental dynamics. Transcription factors (TFs) belonging to the R2R3 class of MYBs have been recognized as both positive and negative transcriptional regulators of genes encoding for GSL-related enzymes [12,13,14]. The A. thaliana MYB28, MYB29 and MYB76 were identified as AGs regulatory TFs, MYB28 being the master regulator, while MYB29 showing a minor role, since its expression did not affect biosynthetic gene profiles, and MYB76 exerting a role in transport regulation of AGs, thereby controlling their spatial distribution. Interestingly, the myb28-1 myb29-1 double mutant plant does not accumulate AGs. Moreover, a complex interplay and a both positive and negative combinatory effect among these regulators was shown in A. thaliana [15,16].

Additionally, members of the bHLH family of TFs, namely MYC2, MYC3 and MYC4, were also identified, with the triple knock-out mutant A. thaliana plants (myc234) being strongly affected in both indolic and aliphatic GSLs, thereby altering the herbivore susceptibility [17]. Interestingly, nor MYB28-29 nor MYB76 were shown to be modulated at transcriptional level by the MYC2-4 expression, suggesting independent roles in GSL regulation [17].

AG structure-related bioactivity and their differential synthesis and accumulation depending on developmental stages, growth conditions, type of organs and tissues have been the subject of many studies [18]. AGs have also been exploited in the development of bioproducts for diverse applications, ranging from agriculture (e.g., amendments, bioherbicides, and biopesticides for pest and pathogen control both in open field and in post-harvest), to animal and human health [19,20,21,22,23,24]. For this reason, there is a growing interest in strategies aimed at developing ideotype crops with the desired GSL profile. It is well known that, despite their biological diversity and in planta function, the modulation of secondary metabolites biosynthesis such as GSLs can be achieved by the application of diverse exogenous elicitors [25]. Chemical, physical and hormone-like treatments can be applied to plant tissues and organs to mimic biotic and abiotic cues, thereby triggering a defense response and promoting the accumulation of specific bioactive compounds. Elicitation strategies can be tested under various experimental designs (in vitro, in vivo establishing trials in pot, etc.) to identify optimal conditions, such as agronomic practices, to direct the metabolites flux towards the selective accumulation of phytochemicals. These strategies can be used to both increase the production of specific compounds in plant cells and stimulate the synthesis of novel metabolites [18,26,27,28]. Indeed, over the past years several attempts at modulating GSLs biosynthesis and accumulation through molecular strategies [29] and treatments with elicitors have been carried out [30,31]. In particular, exposure to salinity influences the synthesis of GSLs, with variations in both AGs and indolic GSLs, that depend also on the ionic composition of the salts [31,32].

Among the crops belonging to the Brassicaceae family, there is one species particularly interesting for its abundant methionine-derived AGs: Eruca sativa Mill [33], synonym of Eruca vesicaria subsp. sativa (Mill.) Hegi (wfoplantlist.org (accessed on 24 June 2025)), commonly known as rocket or arugula. It is an annual allogamous species having a diploid genome whose sequence was recently released [34], known since ancient times for both food and diverse industrial uses [35,36].

In the current study, with the intention to modulate AGs in E. sativa to then exploit them for the development of new bioproducts, plants were treated with Na₂SO₄, selected as a simple elicitor, which had already proven to be able to modulate GSL content in different organ of Brassica rapa [31]. Moreover, benefitting from the knowledge on the GSL pathway from other Brassicaceae and the available information about E. sativa GSL orthologous genes identified in silico [37], as well as putative MYBs encoding genes with regulatory function [34], the expression profiles of selected biosynthetic and regulatory genes that could be involved in AG synthesis were correlated with the GSL content [31]. Four key members of E. sativa genes families involved in the first and second step of AG biosynthesis and encoding for the enzymes Branched-Chain Aminotransferase 4 (BCAT4), Methylthioalkylmalate Synthase 1 (MAM1), Superroot 1 (SUR1) and Cytochromes P450 83A1 monooxygenase (CYP83A1) were selected. Indeed, BCAT4 and MAM1 have a central function in the chain elongation steps of methionine-derived AG biosynthesis, also determining their structural diversity (Figure 1). The C-S lyase SUR1, involved in the core formation step of both AG and indole GSL biosynthesis, and CYP83A1, which in A. thaliana was shown to catalyze the initial step in the conversion of chain-elongated homologs of methionine to thiohydroximate and to be associated together with MAM1 to aliphatic glucosinolates biosynthesis in leaves, were also included [16,37,38]. It was shown that cyp83a1 mutants have reduced levels of aliphatic glucosinolates and a lower susceptibility to the parasitic fungus Erysiphe cruciferarum [39]. Interestingly, Katsarou and coworkers [37] found that in E. sativa MAM1, SUR1 and CYP83A1 genes resulted to be downregulated under lower N and S conditions thereby lowering AG content. Orthologs of regulatory TF encoding genes MYB28a and b, MYC3, MYC4, whose sequences were retrieved from E. sativa genome assembly [34], were also assessed for their expression profile.

2. Materials and Methods

2.1. Plant Material and Elicitation Treatment

Seeds from E. sativa Mill. sel. Nemat (Nutrien Italia S.p.A., Livorno (LI), Italy), selection that is particularly rich in AGs, already in use as trap crop against nematode and under study for the development of new bioproducts for agriculture, were sown in spring 2021 in square pots using a randomized block design (7 pots per treatment with 4 plants per pot). Seven-day old seedlings were treated with Na₂SO₄ (Supelco, Merck KGaA, Darmstadt, Germany), which was gradually increased in a 3-day dose course till the final concentrations of 25 and 50 mM. Throughout plant growth, no macroscopic signs of toxicity or growth impairment were observed in control nor in treated plants. Plants were then grown until flowering stage (sampling time). Samples from roots, basal and apical leaves and flowers from control and treated plants at the two different salt concentrations were collected and immediately flash-frozen for further analysis. Each biological replicate consisted of a pool of samples collected from five, randomly chosen, individual plants.

2.2. Glucosinolate Extraction and Analysis

GSLs were extracted from finely powdered freeze-dried samples, i.e., basal and apical leaves (350 mg), flowers (200 mg) and roots (500 mg) (biological replicates n = 3) using boiling 70% ethanol. The determination of GSLs was performed after desulfation following the ISO 9167:2019(en) [40] method by using a HPLC-UV system (HP 1100 Series Agilent HPLC System, Agilent Technologies, Santa Clara, CA, USA) equipped with a diode array detector and a Pursuit RXs5 C18 (250 × 3.0 mm, 5 μm) (Agilent Technologies, Santa Clara, CA, USA) [41]. GRA and GER were identified with respect to retention times and UV spectra according to Franco et al., 2016 [42], while 4-(-β-D-glucopyranosyldisulfanyl)butyl GSL and dimeric 4-mercaptobutyl GSL (dimeric 4-MBGSL) were first isolated as desulfated GSLs in preparative HPLC and then recognized by nuclear magnetic resonance (NMR) and HPLC-MS analyses, according to the following procedures: 100 g of finely powdered freeze-dried E. sativa leaves were extracted in 1 L boiling 30% aqueous ethanol. The extract was centrifuged, filtered and finally concentrated in rotavapor Laborota 4002-digital (Heidolph Instruments GmbH & Co. KG, Schwabach, Germany) to 100 mL final volume. 50 mL of extract were divided among eight DEAE Sephadex A-25 mini-columns (6 mL DEAE Sephadex A-25 for each column) conditioned with 25 mM sodium acetate buffer (pH 5.6). After washing with the same buffer (10 mL), 0.2 U purified sulfatase (0.26 U/mL) was loaded onto each mini-column, which were left overnight at 20 °C. The desulfated GSLs were eluted in three steps of 2.5 mL water each, for each mini-column, pooled and tested for the presence of desulfated GSLs. As desulfated GSLs were detected only in the first two eluates, these were combined and freeze-dried for successive purification. The obtained powder was dissolved in 2 mL of water and filtered. For preparative HPLC, it was used a Pursuits XRs C18 column (250 × 10.0 mm, 5 µm) with a Pursuits XRs C18 guard column (50 × 10.0 mm, 5 µm) (Agilent Technologies, Santa Clara, CA, USA). The column was eluted at a flow rate of 3.5 mL/min with aqueous acetonitrile (ACN) (solvent A: water; solvent B: ACN) at 30 °C following the following program: 0–1 min isocratic 1% B; 1–28 min linear gradient 1–22% B; 28–37 min linear gradient 22–1% B; 37–40 min isocratic 1% B. The desulfated GSLs were detected by monitoring their absorbance at 229 nm and the fractions corresponding to the major AGs in E. sativa leaves were collected from 10 runs (injection volume = 100 µL). After removal of ACN by nitrogen evaporation the combined fractions were freeze-dried and subjected to HPLC-UV, HPLC-MS and to NMR spectroscopy for structural identification.

The NMR spectra were recorded in D₂O at 25 °C on a Varian Inova 600 spectrometer (Varian, Palo Alto, CA, USA) operating at 600 MHz (for 1H-NMR) and chemical shifts were measured in δ (ppm) with reference to the solvent (δ = 4.79 ppm). MS data were acquired using an Orbitrap Exploris 120 mass spectrometer (Thermo Fisher Scientific Inc., Waltham, MA, USA) equipped with a Raptor Biphenyl column (2.1 × 100 mm, 2.7 µm) (Restek, Bellfonte, PA, USA). The column was eluted with solvent A: 4 mM ammonium formate in water and solvent B: 4 mM ammonium formate in methanol, both containing 0.1% (v/v) formic acid, following the program: 0–2 min isocratic 2% B; 2–28 min linear gradient 2–40% B; 28–32 min linear gradient 2% B. MS spectra were acquired operating in full scan, positive polarity. A full MS data dependent MS/MS (Full MS/dd-MS2) acquisition mode was used, with full MS scan resolution at 60,000 FWHM (scan range m/z 100–1000), and dd-MS2 resolution at 15,000 FWHM. A stepped collision energy HCD (10, 30 and 50%) was applied for fragmentation.

2.3. Bioinformatic Analysis of Target Sequences

A panel of candidate genes for E. sativa AG synthesis encoding for both structural BCAT4, MAM1, SUR1, CYP83A1, and regulatory proteins, MYB28a and MYB28b, MYC3 and MYC4, was identified based on literature [34]. Selected gene sequences were verified in silico querying the E. sativa genome assembly (unpublished data).

2.4. Transcriptional Profiling of Glucosinolate Candidate Genes

For RNA isolation, 100 mg of frozen samples (biological replicates n = 3) were finely ground and extracted using the Spectrum Plant Total RNA Kit (Sigma-Aldrich, St. Louis, MO, USA) with on-column treatment with DNase I Amplification Grade (Sigma-Aldrich, St. Louis, MO, USA), according to technical bulletin. Total RNA was eluted in 50 μL of DEPC-treated water and spectrophotometrically quantified. One thousand ng of total RNA was retrotranscribed with the High-Capacity RNA to cDNA kit (Thermo Fisher Scientific, Inc., Waltham, MA, USA) according to manufacturer’s instructions. The transcriptional levels of target genes were measured by reverse transcription quantitative PCR (RT-qPCR) using a QS3 instrument (Thermo Fisher Scientific, Inc., Waltham, MA, USA) and SYBR Green chemistry. Each reaction contained 2 μL of a 1:15 dilution of cDNA, 5 μL of Power Up^® SYBR master mix (Thermo Fisher Scientific, Inc., Waltham, MA, USA) or PerfeCTa SYBR Green FastMix master mix (Quantabio, Beverly, MA, USA) depending on the target, highly specific primers, and RNA-free water to a final volume of 10 μL. Primer sequences, master mix and amplicon length for each target are listed in Table S1. Specificity of primer pairs designed for this work was assessed in silico by BLAST +2.15.0 alignment with available E. sativa sequences, with melt curve analysis and Sanger sequencing of PCR products. A standard curve was added in all assays, both for target and reference genes. Standard curves were made of 5 points, prepared from a 1:4 serial dilution starting from a 1:10 dilution of a pool of all the cDNAs. To verify the specificity of the reaction, each assay also included a negative control, and melt curve analysis was performed for all reference and target gene assays. The amplification efficiency of each primer pair was estimated using the slope of the regression line, according to the equation: E = 10^(−1/slope) and relative expression was calculated as efficiency-corrected ΔΔCt [43]. The EsGAPDH and EsUBC genes were selected as reference genes, and the average Ct of untreated (control) samples was used as a calibrator. Data are presented as expression levels relative to the control.

2.5. Statistical Analysis

All the analyses were carried out at least in three technical replicates for GSL analysis and two for gene expression analysis for each biological replicate (n = 3). ANOVA and post hoc LSD tests were performed using the R environment for statistical computing, version 4.4.1 (2024) and package ‘agricolae’ [44]. Correlation analysis was performed with the R package ‘corrplot’ [45].

3. Results and Discussion

E. sativa is rich in GSLs, making it an interesting plant species for the development of bioproducts. Salinity stress is one of the abiotic stresses that showed a correlation with GSLs content in Brassicaceae [46] and, according to Lopez-Berenguer et al., 2009 and Pang et al., 2012 [47,48], AGs are the most effective osmolytes in plants salt stress activated response. Broccoli exposed to salinity stress exhibit different effects on GSL synthesis and accumulation in different cultivars, with variations in both AGs and indole GSLs, while in B. rapa salt treatments, and, in particular, Na₂SO₄, caused an increase in the concentration of aliphatic, indolic and aromatic GSLs, as well as the expression of key AG biosynthetic genes like CYP79F1 [31,32]. Furthermore, a mild salt stress (34 mM NaCl) positively impacts on glucoerucin, 4-methylthiobutyl GSL, (GER) and glucoraphanin, 4-methylsulfinylbutyl GSL, (GRA) content in E. sativa shoots [36], while no significative differences in glucosinolate hydrolysis products in Eruca sativa leaves were evidenced at higher NaCl levels (65 and 130 mM) [49].

In the present study, we investigated Na₂SO₄ elicitation as a strategy for AGs modulation in different organs of E. sativa examined at the flowering stage, and the modulation effect on the expression profile of selected putative key biosynthetic and regulatory genes was also evaluated. Starting from the roots upwards, an increasing concentration of total AGs was observed, ranging from about 15–20 μmol/g in roots to 25–50 μmol/g in apical leaves and more than 60 μmol/g in flowers, with a different composition, GER being the main one in roots, dimeric 4-mercaptobutyl GSL (dimeric 4-MBGSL) in leaves and GRA in flowers. The full characterization of desulfated dimeric 4-MBGSL is reported in Supplementary file (Figures S2 and S4; Tables S2 and S3).

In roots, AGs were 94–98% of the total GSL content in all samples: GER was the main GSL, followed by GRA (Figure 2a,d); while indolic GSLs (4-hydroxyglucobrassicin and neoglucobrassicin) were detected in traces. GER increased upon the highest dose of Na₂SO₄ in roots by 2.5 times (Figure 2a). Interestingly, at transcriptional level, the elicitation with Na₂SO₄ at the highest concentration determined also a significant upregulation of the BCAT4 gene (Figure 2d). However, in the same organ and conditions, the other biosynthetic genes examined in the current study did not show changes or were downregulated. Regarding the regulatory TFs, only MYC4 was found slightly, but significantly upregulated upon salt treatment at 25 mM and MYC3 slightly downregulated at the highest concentration (Figure 2d). GSL profiles of leaves were characterized by the presence of GRA, GER and dimeric 4-MBGSL, with dimeric 4-MBGSL being the main GSL (Figure 2b). Traces of 4-(-β-D-glucopyranosyldisulfanyl)butyl GSL were also detected by HPLC and identified in comparison to the retention times and UV spectra of the standard purified as described in “Glucosinolate extraction and analysis” section and reported in Supplementary file (Figures S1, S3 and S4; Tables S2 and S3). In apical leaves, the concentration of dimeric 4- MBGSL roughly doubled due to the treatment with Na₂SO₄ (Figure 2b). Interestingly, upon salt treatment, BCAT4, MAM1 and CYP83A1 genes were downregulated, and the GER content was also lower, less than 50% compared to the control (Figure 2b,e) suggesting a role in AG synthesis in E. sativa yet to be elucidated. A decrease in GER in basal leaves, which have the lowest content of GSLs (15–20 μmol/g) compared to the other epigeal organs (Figure S5), but a higher total biomass, could suggest a possible role in the increase in dimeric 4-MBGSL in apical leaves not of de novo synthesis, but of translocation of GSLs, which are known to move upwards to seeds during maturation [50], and of non-enzymatic dimerization [51]. Moreover, in apical leaves a downregulation of MYC3-4 was observed upon high doses of salt treatment, suggesting a direct involvement in BCAT4, MAM1 and CYP83A1 regulation, which is indeed, not affected by MYB28a-b in line with the reported independent regulatory roles in aliphatic GSL synthesis [17]. Flowers represent the organ with the highest concentration of GSLs among the samples; the main GSL in flowers is GRA, which is the relevant chemical tag of health studies on Brassicaceae. The salt elicitation boosted an increase in GRA by 1.3 times (Figure 2c) and consequently of total GSLs.

The biosynthetic machinery leading to AGs in E. sativa was proven to present unique features in terms of gene families respect to other species of Brassicales, due to gene duplication event [34]. Indeed, the major regulatory gene MYB28 is present in three paralogous sequences in E. sativa genome (MYB28a, MYB28b, MYB28c) although the ones correlated to AGs modulation are mainly MYB28a and MYB28b. Taken as a whole, in the transcriptional profiling of the MYB28a and MYB28b genes, no significant modulations were observed, suggesting that the expression of the MYB28c should also be evaluated. Moreover, in terms of experimental design, the ontological stage for elicitation treatments should also be considered, since previous studies evidence that differences in both the content and the spatial distribution of GSL together with a species—specific modulation of their profile can be observed [36].

Since in some cases the variability across biological replicates of the same organ and treatment was too high to highlight significant modulation in gene expression, with the aim of descriptively correlating biochemical quantification of AGs in E. sativa plants and transcriptional profiling of selected biosynthetic and regulatory genes across organs, a correlation analysis was performed (Figure 3). At the metabolite level, GER negatively correlates to the dimeric 4-MBGSL, as expected, being GER a putative precursor 4-MBGSL. MAM1 and SUR1, which are involved in subsequent steps of AGs synthesis (Figure 1), positively correlate with MYC3, but MAM1 negatively correlates with MYB28b, which also negatively correlates with GRA. A positive correlation was also found between MYC3 and MYC4, even though it was not statistically significant. CYP83A1 appeared scarcely in correlation with both biosynthetic and regulatory genes; thus, other mechanisms yet to be investigated could be involved. Overall, these results show some independent correlations of both MYC3 and MYB28b with biosynthetic genes and AGs, while there is no evidence of correlation for MYB28a and MYC4. Possible causal relations to AG modulation of MYC3 and MYB28b would deserve further investigation. Moreover, according to the A. thaliana aliphatic GSL model [16], the lack of clear relationship between expression levels of biosynthetic genes and metabolite levels could suggest post-transcriptional mechanisms and translocation across organs, which would also be worthy of further studies in this species.

4. Conclusions

Taken together, these results suggest that E. sativa is a reliable source for the extraction of AGs for the development of bioproducts, with different organs carrying different AG profiles, and Na₂SO₄ is a promising elicitor. Salt enrichment one week after seedling emergence seems to have an impact on GER accumulation in roots and of dimeric 4-MBGSL in apical leaves. These two metabolites exhibited a negative correlation with each other. Moreover, an increase in GER has been observed to be associated with an increase in the BCAT4 gene in roots, while the increase in dimeric 4-MBGSL in leaves does not seem to directly depend on any of the tested biosynthetic genes, which are interestingly all downregulated. This elicitation strategy, even though it requires further optimization, seems promising especially in modulating GER content in roots, which are a byproduct of E. sativa salad cultivation and, thus, interesting in a circular economy perspective. Further studies with different processing and sampling times, and a fine characterization of E. sativa gene families through genome sequencing, RNA-seq and their integration with metabolomic data, would provide insights into their organ-specific roles in the synthesis, regulation, transport and accumulation of AGs. Ultimately, this could also help define the best rocket management practices for selective enhancement of AGs and develop strategies aimed at improving the yield in specific plant organs. Such controlled modulation would be useful for downstream industrial applications, both in the agronomic sector for the development of products that exploit the antimicrobial properties of GSL hydrolysis products, and in the food sector for ingredients enriched in GSLs, which have been demonstrated to have health benefits.