Molecular identification, isolation and functional characterization of a glutathione S-transferase gene CsGST in saffron (Crocus sativus L.)

Abstract

Crocus sativus L., an herbaceous species belonging to the Iridaceae family is renowned for its medicinal properties. The vibrant color of its petals and leaves is attributed to anthocyanins, whereas the highly valued stigma predominantly accumulates apocarotenoids, such as crocin, as its main pigments. Glutathione S-transferases (GSTs) are multifunctional enzymes, with many members playing a crucial role in the transport of anthocyanins and other flavonoids. In this study, we isolated the cDNA sequence of a GST gene from saffron, designated CsGST, using RT-PCR and RACE techniques. The sequence revealed a 795 bp open reading frame encoding 264 amino acids. The corresponding genomic sequence was 1074 bp, containing two exons and one intron. Phylogenetic and conserved domain analysis classified CsGST as a Tau-type GST. Expression analysis detected CsGST transcripts in leaves, petals, pistils, and stamens, with its expression in petals showing a significant positive correlation with anthocyanin content. The recombinant CsGST protein was expressed in E. coli and exhibited fundamental GST enzymatic activity. Furthermore, molecular docking analysis suggested a potential interaction between CsGST and crocin, the main carotenoid pigment in the stigma. This study represents the first comprehensive identification and characterization of a GST gene in saffron, providing a foundational basis for future research into its roles in the transport of secondary metabolites.

Supplementary Information

The online version contains supplementary material available at 10.1038/s41598-026-37233-3.

Introduction

Crocus sativus L. (saffron) is an important medicinal plant, prized primarily for its stigma (pistil), which is the source of its valuable bioactive compounds¹. Saffron is one of the world’s most valuable spices by weight, with major cultivation centers in Iran, India, Greece, Morocco, and Spain². The global production of saffron faces challenges due to its labor-intensive harvesting process, requiring manual collection of stigmas, which contributes to its high economic value. Saffron contains a variety of bioactive compounds, including carotenoids, anthocyanins, flavonoids, terpenoids, and phenolic carboxylic acids, among others^2,3. The characteristic red coloration of saffron stigma is primarily attributed to the bioactive apocarotenoids, crocin and crocetin, which can constitute up to 10% of the dry weight, unlike the more common anthocyanin pigments found in many other plants⁴. This chemical composition underpins the use of dried stigmas as a culinary spice, a natural colorant, and a therapeutic agent in traditional medicine⁵. Furthermore, the unique organoleptic properties of saffron are defined by three key apocarotenoids: crocins, responsible for the color; picrocrocin, which contributes to the bitter taste; and safranal, which provides the distinct aroma³.

In contrast to the stigma, the vibrant colors of saffron petals and other floral tissues are primarily due to the accumulation of anthocyanins⁶. Anthocyanins are water-soluble flavonoids located within the vacuoles of plant cells and are known for their significant biological activities, including antioxidant and anti-inflammatory effects⁷. Additionally, they are essential in the pigmentation of plant organs^8,9. Following their synthesis in the cytoplasm, anthocyanins must be transported and sequestered into the vacuole for stable storage. This critical step is often facilitated by specific transporter proteins, among which glutathione S-transferases (GSTs) play a well-documented role^10,11.

GSTs are a multifunctional enzyme family characterized by the presence of two conserved domains: an N-terminal glutathione-binding domain (G-site) and a C-terminal substrate-binding domain (H-site), which together form the active site¹². GST genes have been identified across numerous plant species, with multiple homologous genes found in each genome, including 53 in Arabidopsis thaliana (L.) Heynh., 20 in Dracaena cambodiana Pierre ex Gagnep. 139 in Litchi chinensis Sonn., 61 in Citrus sinensis (L.) Osbeck, 90 in Solanum lycopersicum L., and 62 in Mangifera indicaL^13–18.. According to gene sequence, conservative domain analysis, substrate specificity, and immune characteristics, the plant GST family has been classified into five categories: F (phi), Z (zeta), U (tau), T (theta), and L (lamda)¹⁹. Although the majority of well-characterized anthocyanin transporters are classified within the Phi class, such as AtTT19 in Arabidopsis and PhAN9 in Petunia, the Tau class encompasses a significant exception in the form of BZ2 in Zea mays L., which is crucial for anthocyanin accumulation^20–23. This highlights the functional diversity present among GST classes.

Despite the well-established role of GSTs in anthocyanin transport in model plants and several crops, the specific GST genes responsible for this process remain entirely unknown in saffron. Therefore, the objectives of this study were to isolate and characterize a GST gene from saffron, designated CsGST, and to investigate its potential role in pigment metabolism. We report the cloning of the full-length CsGST cDNA, its phylogenetic classification as a Tau-type GST, and its expression profile across different tissues and developmental stages. Furthermore, we heterologously expressed the recombinant protein to confirm its fundamental GST activity and employed molecular docking to explore a potential interaction with crocin. These findings establish a foundational framework for elucidating the role of GSTs in saffron.

Materials and methods

Experimental materials and reagents

Saffron was cultivated under controlled conditions at the Huzhou South Tai Lake Agricultural Park, China (longitude 120.6°E, latitude 30.52°N, elevation 0 m), and the management department of the agricultural park explicitly allowed it to be used for research²⁴. In May of the following year, dormant bulbs were excavated from the field and cultured in the room without soil until they blossomed. During its blooming period in October, samples of saffron leaves, petals, pistils, and stamens were collected across four developmental stages, designated as Phases I-IV. Samples were collected every 10 days, during which time the color of the stigma transitioned from yellow to scarlet, passing through intermediate shades of orange and red. The petals, stamens, pistils, and leaves of saffron at various stages of growth and development were rapidly frozen using liquid nitrogen and subsequently stored at −80 °C for later RNA extraction.

Total RNA extraction

All tissues were ground in liquid nitrogen. Total RNA was extracted from tissues using the RNeasy Plant Mini Kit (Qiagen) according to the manufacturer’s instructions. RNA integrity was verified by 1.0% agarose electrophoresis and Agilent 2100 Bioanalyzer analysis, confirming all samples had RNA Integrity Number (RIN) values ≥ 8.5^25,26. RNA purity (OD 260/280 = 2.0–2.2.0.2, A260/A280 = 1.8–2.1) was quantified by Nanodrop 2000 (Thermo Scientific, Waltham, MA, USA), and the concentration of RNA samples with a Qubit 2.0 fluorometer (Thermo Scientific).

RACE-PCR amplification

Synthesis of RACE-ready cDNA

The synthesis of 5’ and 3’ RACE-ready cDNA was performed using the SmarterRace 5′/3′ Kit (Clontech, Takara Bio USA, Inc) with 1 µg of total RNA according to the manufacturer’s protocol²⁷. For the standard RT-PCR used to clone the coding region, the use of a no-reverse-transcriptase control is standard practice to rule out genomic DNA contamination. Detailed procedures can be found in the SmarterRace 5’/3’ Kit User Manual.

RACE-PCR amplification

Gene-specific primers (GSP) were designed based on the PB62685 sequence (Table 1). RACE-PCR amplification was carried out using the Advantage 2 PCR Kit (TaKaRa) in a 50 µL reaction volume. Sterilized nuclease-free water was used as a no-template control in all PCR and RACE amplification reactions to confirm the absence of contamination. The amplification protocol consisted of Touchdown PCR: initial 5 cycles of 94 °C for 30 s and 72 °C for 3 min; followed by 5 cycles of 94 °C for 30 s, 70 °C for 30 s, and 72 °C for 3 min; and final 25 cycles of 94 °C for 30 s, 68 °C for 30 s, and 72 °C for 3 min. The amplified products were analyzed by agarose gel electrophoresis.

Table 1.

The primer sequence used to isolate of CsGSTs.

Primer name	Prim sequence (5’−3’)	Purpose
CsGST−5’ GSP	GTCCCTCGTCATCCTCCAATACATC	5’RACE-PCR
CsGST−3’ GSP	CATCCCAGCACGATGTCGAGCAAC	3’RACE-PCR
CsGST-CDS/DNA	For: ATGGCAGATCCCTCTTCCGATTGC Rev: TTAGCTCTCGGAAGTCACAACCATAC	RT-PCR, DNA amplification
CsGST-Q	For: TGGGCAAGCTCCTACACTCAC Rev: TCCTCCGTGGAGGAGGACAG	Q-PCR

Product recovery, conversion and transformation

The amplified RACE products were purified from agarose gels using the NucleoSpin Gel and PCR Clean-up Kit (TaKaRa). The purified products were ligated into the pMD19-T vector using the Takara DNA Ligation Kit and transformed into E. coli DH5α competent cells. Positive clones were selected by blue-white screening and sent for sequencing.

Cloning of coding region and amplification of genomic DNA sequence

The complete cDNA sequence of the CsGST gene was obtained by assembling the sequenced RACE-PCR products²⁷. Primers were designed to clone the coding sequence (CDS) and amplify the corresponding genomic DNA (gDNA). Saffron tissue gDNA was extracted using the Plant Genomic DNA Kit (Tengen Organisms). The CDS and gDNA were amplified using the Advantage 2 PCR Kit (TaKaRa) with the following cycling conditions: initial denaturation at 95 °C for 1 min; 35 cycles of 95 °C for 15 s, 68 °C for 1 min (for CDS) or 3 min (for gDNA); and a final extension at 68 °C for 5 min. The resulting PCR products were cloned into the pMD19-T vector, transformed, and sequenced as described above.

Bioinformatic analysis of CsGST

Sequence alignment was performed using DNAMAN 6.0 (Lynnon Biosoft, San Ramon, CA, USA). The molecular weight and isoelectric point of the CsGST-encoded protein were predicted using ProtParam (http://au.expasy.org/tools/protparam.h.html). Conserved domains were identified using NCBI’s Conserved Domain Database (CDD)^28,29. Multiple sequence alignment of GST amino acid sequences from different species was performed using ClustalW (http://www.ch.embnet.org/software/ClustalW.html)³⁰. The phylogenetic tree was constructed using the Neighbor-Joining method in MEGA 12 software³¹.

Expression pattern analysis of target gene

The expression patterns of the GST gene across various tissues and developmental stages of saffron were examined using quantitative real-time PCR (qRT-PCR)³². Tubulin served as the internal reference gene, with the primer sequences detailed in Table 1. The reaction system and amplification protocol were developed in accordance with the SYBR PrimeScript TM RT-PCR Kit (TaKaRa). The relative expression levels of the target gene were quantified using the 2^^–ΔΔCt method, where ΔΔCt = (Ct_target - Ct_tubulin) sample - (Ct_target - Ct_tubulin) calibrator. Phase I leaf tissue was used as the calibrator (baseline) sample for all calculations³³.

Prokaryotic expression, protein purification, and activity identification of CsGST

CsGST-encoded protein is induced for expression in Escherichia coli

The recombinant plasmid PET-28 A-CsGST was engineered by YouBio and subsequently transformed into competent E. coli BL21(DE3)-RP cells. The resulting strain, designated as BL21/PET-28 A/CsGST, was initially cultured in 3 mL of Luria-Bertani (LB) broth containing 50 µg/mL kanamycin and incubated overnight at 37 °C with agitation at 220 rpm. Following this, the BL21/PET-28 A/CsGST strain was transferred into another 3 mL of LB broth with the same concentration of kanamycin and incubated under identical conditions. A 1% inoculum of the bacterial culture was then introduced into 50 mL of fresh LB medium and incubated at 37 °C until the optical density at 600 nm (OD₆₀₀) reached 0.4 to 0.6. At this point, isopropyl-β-D-thiogalactopyranoside (IPTG) was added to a final concentration of 1.0 mmol/L to induce protein expression, which was carried out at 37 °C with shaking at 220 rpm. After a 4-hour induction period, the expression of the recombinant protein was analyzed using SDS-PAGE.

Purification of the CsGST protein

Cells were harvested and centrifuged at 5,000 × g for 10 min at 4 °C. The supernatant was discarded, and the cell pellet was washed 1–2 times with ice-cold PBS, with the supernatant being carefully removed after each centrifugation step. All subsequent procedures were performed on ice to minimize protein degradation. The pellet was resuspended in lysis buffer (50 mM Tris-HCl, pH 8.0, 300 mM NaCl, 10 mM imidazole) at a ratio of 10 mL buffer per gram of wet cell mass. The cell suspension was then subjected to sonication on ice using a regimen of 20 cycles, each consisting of a 3-sec pulse followed by a 5-sec rest interval, until the suspension clarified and lost its viscosity. The lysate was centrifuged at 12,000 × g for 20 min at 4 °C to remove insoluble debris, and the supernatant, containing the soluble His-tagged recombinant CsGST protein, was collected. Finally, the recombinant CsGST protein was purified from the supernatant using His-tag Protein Purification Beads (Solarbio, M2300) according to the manufacturer’s instructions.

Activity assay

The conjugation of glutathione (GSH) with 1-chloro-2,4-dinitrobenzene (CDNB) catalyzed by GST produces a conjugate with a characteristic absorption peak at 340 nm³⁴. The enzymatic activity of CsGST was determined by monitoring the rate of absorbance increase at this wavelength using UV spectrophotometry. The enzymatic activity of the recombinant CsGST protein was determined using a commercial Glutathione S-transferase Detection Kit (Solarbio, China, Catalog No. BC0355). The assay was performed according to the manufacturer’s instructions, which is based on the conjugation of GSH to CDNB, catalyzed by GST. The reaction product exhibits a characteristic absorption peak at 340 nm, and the rate of increase in absorbance at this wavelength was monitored to calculate enzyme activity. Briefly, reaction reagents were added to cuvettes and rapidly mixed. Initial absorbance values (A1) for both test and blank samples were recorded at 10 s. The samples were then incubated at 37 °C for 5 min in a water bath, after which the final absorbance values (A2) were measured. The absorbance change (ΔA) was calculated as ΔA = (A2 test - A1 test) - (A2 blank - A1 blank). According to the liquid volume calculation:

Enzyme activity was defined as the amount of enzyme required to catalyze the conversion of 1 µmol of CDNB per minute per milliliter of reaction mixture at 37 °C. ε: molar extinction coefficient of the product, 9.6 × 10³ L/mol/cm; d: the optical diameter of the cuvette, 1 cm; Vt: the total volume of the reaction system, 1100 µL = 0.0011 L; Vs: volume of the supernatant added to the reaction system, 100 µL = 0.1 mL; T: reaction time (min), 5 min.

Total anthocyanin content determination

The total anthocyanin content was determined according to the pH differential method as described by Lee et al.³⁵. Saffron tissues were dehydrated at 45 ℃, pulverized with liquid nitrogen, and subjected to extraction with 80% acidified ethyl alcohol through ultrasonication for 30 min. The resulting crude anthocyanin extract was filtered and subsequently diluted with pH 1.0 potassium chloride buffer and pH 4.5 sodium acetate buffer at a 1:5 ratio for analysis. The samples were kept in the dark for 20 to 30 min, and then the absorption (A) was measured at λ = 520 nm and λ = 700 nm. The anthocyanin pigment concentration, expressed as cyanidin-3-glucoside equivalents, was calculated as follows:

where: A = (A520 - A700) pH 1.0- (A520 - A700) pH 4.5; MW (molecular weight) = 449.2 g mol^− 1 for cyanidin-3-glucoside; DF = dilution factor; l = cuvette path length in cm (1 cm); ε = 26,900 mol extinction coefficient, in L × mol^–1 × cm^–1 for cyanidin-3-glucoside.

Molecular docking

The Molecular docking analysis of the GST protein was done with crocin. The three-dimensional structure of CsGST was constructed by homology modeling using the SWISS-MODEL server with PDB ID: 6ep7 as a template. The resulting model was processed with PyMOL software (Version 2.5.1) to remove water molecules and co-crystallized ligands. The three-dimensional structure of crocin was retrieved from the PubChem database (https://pubchem.ncbi.nlm.nih.gov/). Protein-ligand docking was carried out using AutoDock VINA 1.1.2, which employs a semi-empirical scoring function to predict binding affinity. A binding energy of ≤ −7.0 kcal/mol was considered indicative of spontaneous binding. The docking results were visualized and analyzed using PyMOL for three-dimensional representations and LigPlot for two-dimensional interaction diagrams.

Statistical analysis

All data were expressed as mean ± standard deviation (SD) from three independent biological replicates (n = 3). Statistical analyses were performed using GraphPad Prism 9.0. The normality of data distribution was verified using the Shapiro-Wilk test. For comparisons across multiple groups involving two factors (tissue type and developmental stage), a two-way analysis of variance (ANOVA) was employed, followed by Tukey’s honestly significant difference (HSD) post-hoc test for multiple comparisons. Statistical significance was defined as p < 0.05. Additionally, Pearson correlation analysis was conducted to assess the relationship between CsGST expression levels and anthocyanin content.

Results

Cloning of CsGST gene in saffron

Through an analysis of the saffron transcriptome database, we identified a Unigene exhibiting high homology with GSTs from other species, suggesting its potential involvement in anthocyanin transport. This gene was designated as CsGST in the present study. Gene-specific primers were designed, and the full-length cDNA sequence of the CsGST gene was obtained using RACE technology (Supplementary Figure S1A). Primers targeting the CsGST coding sequence (CDS) and genomic DNA (gDNA) were designed based on the start and stop codon positions. The coding region and the corresponding gDNA sequence were cloned using cDNA and gDNA from saffron floral bud tissue as templates. Sequencing revealed a 795 bp CDS encoding a 264-amino-acid protein, and a corresponding 1074 bp gDNA sequence (Supplementary Figure S1B and C). Comparative analysis showed that the CsGST CDS contains two exons and one 279 bp intron (Fig. 1), adhering to the canonical GT-AG splicing rule—a genomic structure consistent with most anthocyanin-related GST genes.

Fig. 1.

(A) The full-length sequences and ORF sequence encoding amino acids of CsGST. (B) The DNA sequence corresponding to the coding region contains two exons and one intron.

Protein characterization and phylogenetic analysis

ProtParam analysis predicted the CsGST-encoded protein to be composed of 264 amino acids, with a molecular weight of 28.954 kDa and a theoretical isoelectric point (pI) of 5.12. BlastP analysis confirmed the presence of two conserved domains, GST_N_Tau and GST_C_Tau, classifying CsGST as a member of the Tau-type GST superfamily (Fig. 2). Typically, soluble GST functions as a dimer, featuring a conserved glutathione-binding site (G-site, ☆labeled) in the N-terminal domain and a more variable substrate-binding site (H-site, ★labeled) in the C-terminal domain (Fig. 3).

Fig. 2.

The protein encoded by CsGST was predicted by BlastP, and the conservative domain of CsGST amino acid sequence was deduced.

Fig. 3.

The full-length cDNA sequence and deduced amino acid sequence of CsGST gene, ☆indicates 7 amino acide residues of the G site and ★indicates 12 amino acide residues of the G site.

To elucidate the evolutionary relationships of CsGST, a phylogenetic tree was constructed using GST amino acid sequences from various plant species. The analysis revealed that CsGST clusters robustly within the Tau GST (GSTU) clade. It showed the closest evolutionary relationship with GSTs from other monocot species, particularly D. cambodiana, while forming a distinct branch separate from Tau GSTs of dicot species (Fig. 4). This phylogenetic placement within the monocot-specific Tau GST lineage supports its functional characterization and provides a comparative framework for future studies.

Fig. 4.

Phylogenetic analysis of CsGST. The evolutionary tree was inferred using the Neighbor-Joining method from aligned GST amino acid sequences. Bootstrap values from 1000 replicates are shown at the branches (only values > 50% are displayed). The red box indicated the GST protein of saffron is closely related to D. cambodiana. The scale bar represents the number of amino acid substitutions per site.

SDS-PAGE analysis of Recombinant CsGST

The recombinant expression plasmid was introduced into E. coli BL21 (DE3) and cultured in LB medium supplemented with kanamycin at 37 °C. The expression of the CsGST protein was induced using varying concentrations of IPTG and subsequently analyzed via SDS-PAGE (Supplementary Figure S2A). The observed molecular weight of the CsGST protein was approximately 28 kDa, aligning with the theoretically predicted molecular mass of CsGST. The recombinant CsGST protein, induced by 1 mM IPTG, was purified using His-tag Protein Purification Beads and eluted with imidazole buffers of varying concentrations. The protein was partially and predominantly eluted at imidazole concentrations of 100, 200, 300, 400, and 500 mM (Supplementary Figure S2B). The findings indicated that the optimal elution concentration was 300 mM. SDS-PAGE analysis revealed a single protein band with a molecular mass of approximately 28 kDa, confirming the successful purification of the recombinant protein. This result is consistent with the molecular weight predicted by online bioinformatics analysis.

Detection of purified protein activity

The activity of the purified protein CsGST was evaluated using CDNB and reduced GSH as substrates. Enzyme activity was quantified in units (U), where one unit is defined as the amount of enzyme required to catalyze the conversion of 1 µmol of substrate into product per minute at 37 °C, equivalent to 1 U = 1 µmol/min. The specific activity was subsequently calculated based on the enzyme’s liquid volume. The specific activity of the purified protein CsGST was 0.0514 U/mL, indicating a statistically significant increase compared to the control group (0.0023 U/mL) (Table 2). This finding substantiates that the purified CsGST protein possesses fundamental GST enzymatic activity, establishing it as a functional member of the GST family.

Table 2.

Specific activities of purified CsGST.

Sample	A1340 nm	A2340 nm	activity (U/mL)	Mean Activity ± SD (U/mL)
Blank	0.44	0.46	/	/
	0.43	0.45	/
	0.45	0.48	/
Control	0.46	0.50	0.0046	0.0023 ± 0.0023
	0.49	0.53	0.0023
	0.51	0.54	0
purified protein CsGST	0.46	0.75	0.0621	0.0514 ± 0.0092*
	0.47	0.70	0.0483
	0.46	0.68	0.0437

Note: Blank: reaction reagent without any protein sample; Control: lysate from E. coli containing empty plasmid without the CsGST insert. Activity was calculated as described in the Methods. *, significantly different from the Control group (p < 0.05, unpaired t-test).

Expression pattern analysis of CsGST gene and total anthocyanin content determination

We assessed CsGST expression and total anthocyanin content across various saffron tissues. The qRT-PCR results revealed higher expression levels of CsGST in leaves and stamens during the initial developmental stages (Phase I and Phase II) compared to later stages (Phase III and Phase IV. In contrast, the transcript level of CsGST in petals increased consistently, peaking during Phase IV, while in the pistil, expression levels rose from Phase I to Phase III before undergoing a significant reduction in Phase IV (Fig. 5A). The study revealed differential expression patterns of the CsGST gene across distinct growth stages and tissues. Total anthocyanin content was determined using the pH differential method (Fig. 5B). As anticipated, the highest anthocyanin concentration was detected in petals. In contrast, anthocyanin levels in the pistils were considerably lower, which is consistent with the fact that the characteristic pigmentation of the stigma is primarily derived from crocins, not anthocyanins. Pearson correlation analysis revealed a significant positive correlation between CsGST transcript levels and anthocyanin content exclusively in petal tissues (r = 0.9167, p < 0.05); no significant correlations were detected in pistils, leaves, or stamens (Supplementary Figure S3). These results indicate a petal-specific role for CsGST in anthocyanin accumulation and suggest divergent physiological functions in other tissues.

Fig. 5.

The changes of CsGST expression and total anthocyanin content in various tissues and different growth stages in saffron. (A) The expression changes of CsGST were detected by qRT-PCR. (B) The total anthocyanin content was detected by the pH differential method. (C) The content of crocin in the stigma of saffron was detected by HPLC. (D) Representative images showing the developmental stages (Phase I to IV) of saffron stamens, pistils, petals, and leaves. Data are presented as mean ± SD from three independent biological replicates (n = 3). Significant differences were determined by two-way ANOVA followed by Tukey’s post-hoc test for multiple comparisons. *p < 0.05, **p < 0.01, ***p < 0.001 indicate significant differences compared to Phase I within the same tissue.

Molecular docking of crocin and GST

Given the concurrent accumulation of crocin and expression of CsGST in the stigma (Fig. 5A, C), we performed molecular docking to investigate a potential interaction. Docking studies using VINA 1.1.2 yielded a binding energy of −7.0 kcal/mol, suggesting spontaneous binding between crocin and CsGST. Visualization with LigPlot and PyMOL showed that crocin binds stably within the protein cavity, primarily through hydrogen bonds (with ASN14, HIS15, ASP108, ASP109, TYR220, LEU228) and hydrophobic interactions (with LEU41, SER71, LYS57) (Fig. 6).

Fig. 6.

Molecular docking analysis of the CsGST-crocin interaction. (A) Three-dimensional representation of crocin (shown as stick model) docked into the binding pocket of CsGST (shown as cartoon representation). The protein surface is displayed in light gray. (B) Two-dimensional interaction diagram generated by LigPlot, detailing the specific amino acid residues involved in hydrogen bonding (green dashed lines) and hydrophobic interactions (red arcs) with crocin.

Discussion

This study provides a systematic identification and characterization of a GST gene in saffron. Phylogenetic analysis confirmed its classification within the Tau class (Fig. 4). Although anthocyanin transport is typically associated with Phi-class GSTs, the well-documented function of the Tau-class protein maize BZ2 sets a precedent for CsGST’s potential role in flavonoid transport^36,37. This hypothesis is strongly supported by our expression analysis, which showed a significant positive correlation between CsGST transcript levels and anthocyanin content specifically in petals, indicating a tissue-specific function.The absence of this correlation in pistils aligns with the distinct pigment chemistry of this organ, where stigma coloration is primarily due to crocins, whose vacuolar sequestration is mediated by ABCC transporters rather than GSTs⁴.

The recombinant CsGST protein exhibited significant enzymatic activity with the generic substrate CDNB, confirming its function as a glutathione S-transferase³⁸. However, it is important to emphasize that the enzymatic activity observed, although statistically significant compared to control levels, was assessed using the generic GST substrate CDNB³⁹. The specific activity of CsGST toward native anthocyanin substrates remains undetermined and necessitates further investigation through direct transport assays. However, the specific activity was relatively low, a common observation for plant GSTs with non-native substrates. This may indicate the enzyme’s specialization for certain endogenous compounds in saffron, such as specific anthocyanin derivatives or crocin⁴⁰. Future research comparing its kinetics with established anthocyanin transporters, such as AtTT19 or BZ2, using their native substrates, would provide valuable insights.

The anthocyanin transporter protein GST is intricately associated with tissue localization and developmental stages¹². Numerous studies have demonstrated a positive correlation between GST and anthocyanin accumulation across various species^19,41. This investigation focused on the expression variability of CsGST in different tissues and developmental phases of saffron. The findings of this study demonstrated that CsGST is expressed in petals, leaves, stamens, and pistils. Our investigation highlighted the variability in anthocyanin concentrations across these distinct tissues. Through statistical correlation analysis, we identified a significant positive correlation between CsGST expression and anthocyanin content specifically within petal tissues (Supplementary Figure S3), suggesting its potential involvement in petal pigmentation. The tissue-specific expression pattern of CsGST, particularly its lack of correlation with anthocyanin accumulation in non-petal tissues, suggests additional physiological roles beyond flavonoid transport. In leaves, CsGST may participate in detoxification processes and oxidative stress response, given the constant exposure to environmental stresses and the well-established role of GSTs in cellular defense mechanisms^42,43. In reproductive tissues like stamens and pistils, CsGST could be involved in the metabolism of secondary metabolites, including potential interactions with hormone conjugates or other specialized compounds, thus contributing to developmental regulation and stress adaptation. This functional diversification within the GST family underscores the complex regulatory networks governing secondary metabolism in different saffron tissues^44,45.

Crocin, the primary bioactive compound found in the stigma of saffron, is a water-soluble carotenoid that has shown inhibitory effects on tumor growth across various animal cancer models^46,47. Our research team previously investigated the impact of different cultivation methods (outdoor and indoor) on crocin content during pistil development³⁵. We compiled data on changes in crocin content in pistils from indoor cultivation at four sampling stages pertinent to this study (Fig. 5C) (This data was sourced from published articles by our team⁴⁸. The levels of crocin demonstrated an initial increase during Phases I to III, followed by a decrease in Phase IV. This pattern aligns with the expression trend of the CsGST gene in pistils observed in this study, which similarly shows an initial increase followed by a decrease. We hypothesize that crocin, as a natural pigment similar to anthocyanin, may be associated with the CsGST gene. Molecular docking revealed a binding energy of −7.0 kcal/mol for the CsGST-crocin interaction (Fig. 6), intriguingly comparable to that for its endogenous ligand, GSH (Supplementary Figure S4). While GSTs are known to bind anthocyanins, interactions with carotenoid-derived compounds like crocin are less explored. This computational prediction suggests a novel hypothesis that CsGST may be involved in crocin binding or sequestration, alongside its potential role with flavonoids. It is crucial to note that this requires experimental validation, for instance by isothermal titration calorimetry (ITC) or surface plasmon resonance (SPR).

In summary, this study provides a comprehensive molecular characterization of the CsGST gene from saffron. We have established its sequence, phylogenetic relationships, expression profile, and enzymatic activity. Our findings support the hypothesis that CsGST, a Tau-class GST, may function analogously to BZ2 in petal anthocyanin pigmentation, while also suggesting a novel avenue of research concerning its potential interaction with crocin. This work provides a critical foundation for future functional studies, including genetic transformation and direct transport assays, aimed at elucidating the precise physiological roles of CsGST in saffron.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

Yan sicheng: Writing–original draft. **Zhang Xingchang: ** Data curation. **Li Jing** : Validation, Investigation. **Jiang Fengqin: ** Data curation. **Xi Xiaoyuan: ** Formal analysis. **Zhou Guifen: ** Supervision. **Qian Xiaodong: ** Investigation. **Li Liqin: ** Writing–review and editing.

Funding

This research was supported by the National Natural Science Foundation of China (NO.32570418, NO.32170368), and the Natural Fund of Huzhou Science and Technology Bureau (NO.2023GYB23, NO.2018YZ12). This research was supported by the Zhejiang Provincial Natural Science Foundation of China under Grant No. LQ23H280018.

Data availability

The data reported in this paper have been deposited in the GenBase in National Genomics Data Center, Beijing Institute of Genomics, Chinese Academy of Sciences/China National Center for Bioinformation, under accession number C_AA125964.1 that is publicly accessible at [https://ngdc.cncb.ac.cn/genbase.].

Declarations

Competing interests

The authors declare no competing interests.