Magnetic Solid-Phase Extraction and in Situ Derivatization for Determining Phytohormones and in Oilseeds by Ultra-Performance Liquid Chromatography-tandem Mass Spectrometry

Background: Phytohormones are a group of naturally occurring signaling molecules which inuence physiological processes of oil crops. Simultaneous determination of multiple phytohormones in oilseeds is still a challenge due to their trace concentrations, species diversity, and lipid interference. Therefore, a simple and selective method for the simultaneous determination of multiple phytohormones in oilseeds is urgently needed. Results: In this study, the Fe 3 O 4 @Ti 3 C 2 @β-CD nanoparticles were successfully synthesized and used for the rst time as an adsorbent for the magnetic dispersive solid-phase extraction of phytohormones from oilseeds. The magnetic dispersive solid-phase extraction and in situ derivation by the addition of N-(3-dimethylaminopropyl)-N′-ethylcarbodiimide (EDC) were ingeniously combined. This ecient pre-treatment method integrated the extraction, purication, and derivatization processes into one single step. Several parameters affecting the eciency of extraction and derivatization were evaluated. Under the optimized analysis conditions, satisfactory methodological performance was achieved. High linearities (R 2 > 0.9928) at three spiked levels, as well as the low matrix effect (ranged from 16.63 % to 17.06 %) and limits of detection (0.89-13.62 pg/mL) were also obtained. The intra and inter-day relative standard deviations (RSDs) were less than 13.7 % and 11.6 %, respectively. The recoveries were ranged from 80.4 % to 115.1 %. This method was successfully employed for analyzing 12 phytohormones in different oilseeds samples. Conclusion: A simple and sensitive method based on the magnetic solid phase extraction integrated with in situ derivations for the proling of 12 phytohormones, including 9 gibberellins (GAs), indole-3-acetic acid (IAA), abscisic acid (ABA), and jasmonic acid (JA) in a single rapeseed seed was developed by using ultra-performance


Background
Phytohormones are naturally occurring signaling molecules attributing to the regulation of plant growth and development [1]. These molecules are present at trace amounts in plant tissues, and their regulatory mechanisms frequently rely on the complex crosstalk networks among different classes of phytohormones [2,3]. The study of multiple phytohormones such as Gibberellins (GAs), indole-3-acetic acid (IAA), abscisic acid (ABA), and jasmonic acid (JA) is closely related to agricultural production and the green revolution technology [4]. Phytohormones are usually present at extremely low concentrations, instability, and spatio-temporal distribution in certain plant organs [5,6]. To elucidate the regulatory mechanisms of phytohormones on plant growth and development, highly sensitive pro ling phytohormones in speci c organs such as seed has received persistent attention.
Oil crops contribute a total of 93% to the vegetable oils for human consumption worldwide [7]. Vegetable oils are the foremost among all oil products exhibiting speci c bene cial and functional properties towards human [8]. In oil crops, accumulating research reported that the phytohormones exhibit a signi cant effect on seed germination, seedling growth, and yield [9][10][11], thus the quanti cation of the phytohormones is critical for the underlying regulation mechanism. However, the analysis of phytohormones in oilseeds matrices is yet a challenging issue, because of the inherent complexity of the matrix. lipid matrix interferences in oilseeds make it di cult for quanti cation of phytohormones due to the high oil content. Lipids are di cult to avoid the co-extraction with the phytohormones in organic solvents during the preparing process, the detection sensitivity of the target analytes will be signi cantly reduced due to the existence of lipids [8]. The structural and chemical diversity of phytohormones make it challenge to simultaneous determination of multiple phytohormones. Therefore, a highly sensitive and reliable quanti cation method to determine multiple phytohormones in oilseeds rich in fatty acids is extremely demanding.
Nowadays, the ultra-high performance liquid chromatography-tandem mass spectrometry (UPLC-MS/MS) is being widely acknowledged for analyzing the phytohormones [12]. Chemical-derivatization assisted LC-MS methods have been demonstrated to be powerful tools for sensitive detection of trace phytohormones with poor MS response in negative mode [13,14]. Although derivatization reactions are tedious and time-consuming, various derivatization reactions have been used to enhance both the ionization e ciency and detection sensitivity of the phytohormones. Many of the reported derivatization reactions of acid phytohormones mostly occur in an organic medium [15,16]. GAs, IAA, ABA, and JA contain carboxyl groups which could derivatizate by reagents with a quaternary ammonium group such as N-(3-dimethylaminopropyl)-N′-ethylcarbodiimide hydrochloride (EDC), what's more, the EDC-derived phytohormones could be generated in an aqueous medium [17]. The implementation of in situ aqueous derivatization coupled with the sample pre-treatment technique for phytohormones might simplify the analytical methods. The combination of extraction and in situ derivatization is considered as a promising sample preparation method.
For the oilseeds, samples were often extracted and puri ed in multiple steps during the pre-treatment process to reduce the interference of lipid matrix [18]. In recent years, the matrix dispersion technology was developed for the treatment of fatty solid samples by mixing with clean-up sorbents [19]. The common dispersive sorbents used in fatty matrices include C18, primary secondary amine (PSA), and orisil [20]. Nowadays, magnetic dispersion solid-phase extraction (MD-SPE) is being applied to the plant samples for quanti cation of phytohormones [17,21]. In brief, MD-SPE is more convenient than the conventional SPE due to the advantages of simplicity, time saving, and excellent adsorption e ciency [22]. Therefore, the highly selective and functional decorated adsorbent has attracted more attention to enrich the phytohormones. Ti 3 C 2 MXene exerts a two-dimensional micro-crack structure allowing large speci c surface area, high porosity, and good stability [23]. It have demonstrated that Ti 3 C 2 exhibits a greater ability to adsorb a variety of environmental pollutants, including organic dyes, heavy metal ions, and gas molecules [24,25]. β-Cyclodextrin (β-CD) is a type of oligosaccharide having a unique appearance of hydrophilic and inner cavity hydrophobic cup structure. The β-CD could selectively bind phytohormones with the special hydrophobic cavity [21]. The magnetic Ti 3 C 2 composite conjoins the high adsorption capacity with the convenient magnetic separation. Supporting this statement, further study demonstrated that magnetic Ti 3 C 2 can be modi ed by β-CD and is possible to improve the selectivity of phytohormones in oilseeds samples. To the best of our knowledge, this is the rst study reporting the use of Fe 3 O 4 @Ti 3 C 2 @β-CD to selective enrichment of phytohormones. However, the application of functionalized Ti 3 C 2 MXene for solid extraction adsorbent to the analysis of phytohormones needs to be further explored.
The present study aimed at developing a rapid, eco-friendly, and effective method based on the in situ derivatization and MD-SPE for the extraction of phytohormones in the oilseeds samples followed by LC-MS/MS determination. In this study, oilseeds samples and sorbent were ground by the matrix dispersion to degrease the fatty acids. The Fe 3 O 4 @Ti 3 C 2 @β-CD was prepared as the sorbent material showing good extraction e ciency for phytohormones. The extraction, puri cation, and derivatization were integrated into one single step by adding an EDC derivatizing agent and performed in a micro-centrifuge tube without any sample transfer. The main parameters involved were optimized. Finally, the proposed method was successfully applied to different oilseeds.

Results And Discussion
Characterization of Fe 3 O 4 @Ti 3 C 2 @β-CD composite Fig. 1a illustrates the X-ray diffraction (XRD) patterns of the Fe 3 O 4 @Ti 3 C 2 and Fe 3 O 4 @Ti 3 C 2 @β-CD. The characteristic diffraction peaks (002) of the material were tted well with Ti 3 C 2 , which was consistent with the report [26]. The characteristic diffraction peaks of Fe 3 O 4 matched well with the standard XRD data of magnetite [22]. After Fe 3 O 4 @Ti 3 C 2 was grafted with β-CD, the pattern was unchanged in its peak positions, suggesting that the crystal structure of Lipid components are the most common coexisting matrix in the extraction phase of oilseeds. So the effects of different sorbents (C18, PSA, kieselguhr, alumina, and orisil) on extraction e ciency were investigated. As depicted in Fig. 3a, the effects of the orisil sorbent on extraction e ciency were signi cantly higher on the phytohormones compared to other clean-up sorbents. Florisil as matrix sorbent may be attributed to purify the oilseeds samples, also has advantageous effects on decrease the retention of targets. Thus, orisil sorbent was employed for the following experiments.

Selection of extraction solvent types and volumes
The selection of an appropriate extraction solvent can achieve higher extraction e ciency of phytohormones. In this study, four organic solvents (acetonitrile, methanol, 80% acetonitrile, and 80% methanol) were evaluated. As depicted in Fig. S1, methanol enhances the extraction e ciencies for multiple phytohormones. Hence, methanol was preferred as an appropriate extraction organic solvent. Further, the in uences of the extraction solutions' volume on the phytohormones were investigated. As shown in Fig. S2, the extraction solution increased from 200 μL to 400 μL exhibiting a higher distribution ratio of the analytes in the organic phase, thereby decreasing the extraction e ciencies. Therefore, 200 μL of methanol was considered as the extraction solution.
Optimization of the amount of magnetic solid-phase extraction sorbents The EDC-derived phytohormones possess a positively charged quaternary ammonium group that may attract the hydroxyl groups or oxygen groups existing on the surface of Ti 3 C 2 [27]. At the same time, the β-CD could selectively bind the phytohormones by molecular piston [28]. The effect of the amount of magnetic solid-phase extraction sorbents were investigated by adding different amounts Fe 3 O 4 @Ti 3 C 2 @β-CD. As shown in Fig 3b, the peak area of the phytohormonal derivatives increased with the change of Fe 3 O 4 @Ti 3 C 2 @β-CD from 1 mg to 5 mg. The increase in the adsorption capacity could attribute to the availability of a greater number of adsorption sites. Only a slight increase in the adsorption was observed when the sorbent amount exceeded 5 mg. Thus, 5 mg Fe 3 O 4 /Ti 3 C 2 /β-CD nancomposite was used in the following experiments.

Optimization of simultaneous derivatization and magnetic solid-phase extraction time
The extraction and in situ derivatization processes were the pivotal steps in this method. Therefore, the effect of extraction and in situ derivatization time were investigated by varying from 30 to 120 min. As shown in Fig 3c, the signals of derivatives increased with the extension of reaction time until 90 min, and then insigni cant changes were observed. Therefore, 90 min was chosen to mediate the extraction and in situ derivatization.

Optimization of desorption solution volume and desorption time
It should be noted that water is a strong desorption solvent in the hydrophilic retention systems [29]. Therefore, water was used as the desorption solution for the phytohormonal derivatives. The volume of the desorption water solution was further optimized. As shown in Fig S3, 50 μL of desorption solution was enough for the desorption of the derivatives from the sorbent. The dilution effect of the desorption was observed when the volume of the desorption solution increased from 50 μL to 100 μL. Hence, 50 μL water was used as the desorption solution. The optimization of desorption time ranged from 10 to 30 min was performed, which indicated that 20 min was enough for e cient desorption (Fig. 3d). The mass transfer between the sorbent and acidic phytohormonnal derivatives may be completed within 20 min. Hence, the desorption time of 20 min was selected.

Method validation
As depicted in Table 1, the calibration curves of all the analytes exhibited good linear determination coe cients R 2 (≥ 0.9928). The obtained calibration curves demonstrate the excellent linearity for the range studied in this work. The LOD and LOQ values were within the range of 0.89-13.62 pg/mL and 2.99-45.39 pg/mL for all target analytes, respectively. In Table 1, the recoveries were ranged from 80.4% to 117.7% for all target phytohormones. The RSD values of the intra-day precisions and inter-day precisions were in the range of 1.1-13.7 % and 0.1-11.6 %, respectively. These results indicate the acceptable repeatability and reproducibility of the proposed method. A matrix-effect is considered to be a suppression or enhancement of the analyte response due to the presence of co-eluting matrix constituents during the chromatographic run [30]. Signal suppression or enhancement may exert negative or positive ME values, respectively. A ME < -20 % indicates high ion suppression whereas an ME > 20 % indicates high signal enhancement effect and -20 % < ME < 20 % indicates no matrix effects [31]. In this study, moderate signal suppressions (-16.63 % < ME < 17.06 %) were observed for all analytes. The absolute values of the ME were signi cantly low due to the derivatization or the matrix clean-up of orisil adsorbent. For example, the EDC derivatization could increase the target molecules' sensitivity avoiding matrix interference [32].
Appling of the proposed method for the analysis of different oilseed samples The validated analytical method was employed to determine the content of phytohormones in different oilseeds plants (bean, peanut, and sesame). The total fatty acid contents of different oilseeds were shown in Fig S5. The concentrations of 12 phytohormones are listed in Table 3 ranging from ND (not detected) to 1289.5 ng/g FW. The total content of the phytohormones in bean was higher than in peanut and in sesame, especially for ABA. High content of ABA in bean is similar with the previously report [33].
The recovery experiments were carried out and the obtained results are depicted in Table 2. These recoveries and SD indicated that the present method is suitable for the determination of phytohormones in different oilseeds plants.

Comparison of the proposed method with other methods
In Table 3, the proposed method was compared with other solid-phase extraction methods for analyzing the phytohormones in the plant from previous reports. Graphene oxide [21], SiO 2 [34], and TiO 2 [17] were reported as the adsorbents for phytohormones during the solid-phase extraction. The MD-SPE is more convenient than the conventional SPE cartridges due to less consumption time and organic solvent [35]. These reported methods limited the plant sample consumption between 100 mg to 3.0 g FW for multiphytohormonal pro les, but more sample requirements could make it di cult to detect the minute plant organs [21,36,37]. In contrast, this method provides quanti cation of most major plant hormones from a single rapeseed (4-6 mg). Compared with other solid-phase extraction coupled with UPLC-MS/MS methods, the LODs of this method were lower than the LODs of previouly reported about for target phytohormones. Fe 3 O 4 @Ti 3 C 2 @β-CD nano-composite material and EDC derivatization contributed to the increased detection sensitivity of the methods. This simple procedure deduced the losses from transfers and may ensure extractive and reliable recoveries (80.4-117.7 %). Hence, the proposed method has several advantages over the other reported techniques, being simple, effective, and eco-friendly.

Conclusion
In this study, the Fe 3 O 4 /Ti 3 C 2 /β-CD was successfully synthesized and developed as the magnetic solidphase extraction adsorbents for the selective extraction of multiple gibberellins and several phytohormones. A new version of prepared method concluded that the magnetic solid-phase combined with in situ EDC derivatization is the best approach for studying the distribution of 12 phytohormones in rapeseed as it provides simpli cation, adequate selectivity, and sensitivity. This method was also successful for analyzing phytohormones and resolving the problems such as unstability, trace amount, and matrix interference in different oilseeds samples.

Chemical and reagents
Gibberellin A 1 (GA 1 ), GA 3 , GA 4 , GA 5 , GA 7 , GA 8   min, the eluent was collected and determined by UHPLC-ESI-MS/MS. The general process of sample pretreatment is illustrated in Fig.4.

Instruments and analytical conditions
The UPLC-MS/MS was equipped with an Agilent 1290 series (Agilent Technologies, USA) and Agilent 6460 triple quadruple mass spectrometer (Agilent Technologies, USA). The analytes were separated on a Waters ACQUITY UPLC HSS T3 column (100 mm × 2.1 μm). The optimized separation conditions were as follows: the column oven temperature was kept at 40 °C , and the sample injection volume was 10 μL, the ow rate of the mobile phase was 0.3 mL/min. The elution gradient program of the positive ion mode was performed, as depicted in Table S1.
The multiple reaction monitoring (MRM) was employed for the quantitative analysis of the targeted compounds. Nitrogen gas was used as the drying and collision gas. The ionization source conditions were as follows: the ow rate of the nebulizer gas was 8 L/min, the source temperature of the mass spectrometer was 300 °C , the nebulizer pressure was 50 psi, and the capillary voltage was 3500 V. The details of the EDC-derived phytohormones and their optimized MRM parameters are listed in Table 4. The MRM chromatograms of the target EDC-derived phytohormones were shown in Fig. 5. The MRM chromatograms with subsection of target phytohormone derivates in a single rapeseed were shown in Fig. 6.

Method validation
To validate the developed method, the linearity, limit of detection ( Accuracy was evaluated by the recovery of each target analyte at low, medium, and high levels, respectively. The precision was investigated by the intra-day precision (repeatability) and inter-day precision (reproducibility). The intra-day precision was evaluated by analyzing 7 replicates on the same day while the inter-day precision was carried out for 3 consecutive days (7 replicates per day). During the trace analyses with a complex matrix, the ME usually occurs and its percentage can be calculated as per the equation below: where, A extract stands for the slope of matrix-match calibration curves, A solvent stands for the slope of    Figure 1 The MRM chromatograms of target phytohormones in a single rapeseed Page 20/23