Introduction

Tyrosine (Tyr) nitration is becoming increasingly recognized as a prevalent, functionally significant post-translational protein modification (PTM), which can occur in cells during oxidative stress and over-production of nitric oxide [1]. This modification is involved in the control of fundamental cellular processes including cell cycle, cell adhesion and cell survival, as well as cell proliferation and differentiation [2]. The addition of NO2 group to the ortho-position of Tyr confers particular physicochemical properties to the modified amino acid and the corresponding proteins, as a consequence of pK a reduction of about three units [3]. These changes in protein conformation may have important functional consequences, such as a loss, an increase, or no effect on protein function [4–7]. Elevated levels of 3-nitrotyrosine (NO2Tyr) have been reported in a range of pathological conditions including inflammatory, neurodegenerative, and cardiovascular disorders, among others [8, 9]. Moreover, emerging data indicate a novel biological function for Tyr nitration in the regulation of immune responses [1]. Therefore, in mammals Tyr nitration is being intensively studied because it can be used as a biomarker not only of nitrosative stress but also of certain pathological and physiological conditions [10, 11]. Additionally, new studies emphasize the possible involvement of Tyr nitration in signaling pathways mediated by NO [1].

On the other hand, in plants the information available on protein nitration under normal conditions is rather limited [12]. Even though previous data indicate the existence of a basal nitration present in the plant tissues analyzed, there are published data which indicate that an increase in the number of proteins or an intensification of specific proteins resulting from Tyr nitration could be considered as an indicator of nitrosative stress in plants [7, 13–15]. Therefore, protein Tyr nitration might be a good starting point in the search of nitrosative stress markers in plant cells [13]. Nevertheless, since the actual number of nitrated Tyr residues in proteins is unknown, it is by far more preferable to use molar ratio of nitrated Tyr residues to non-nitrated Tyr residues [16]. However, the overall concentration of nitrated Tyr residues is typically low [9]. Hence, assays applied to the analysis of NO2Tyr in biological samples must offer a low limit of detection, accuracy and precision.

Detection of NO2Tyr in biological samples has been extensively reported in the literature. These methods fall into two basic categories: molecular analysis using NO2Tyr antibody-staining techniques [13] and chemical analysis using HPLC and GC [14], mainly using mass spectrometers as detectors. The source and nature of analytical problems, shortcomings, and pitfalls associated with NO2Tyr determination have been reviewed by Duncan [17] and Tsikas [16, 18]. The main drawbacks are both the low abundance of nitrated species and lack of efficient enrichment methods [2].

Mass spectrometry (MS) is a powerful analytical technique with inherent selectivity, sensitivity and precision when applied to NO2Tyr determination. Moreover, NO2Tyr immunoassays, unlike GC-MS- and LC-MS-based methods, cannot provide important information about NO2Tyr/Tyr ratio [17, 19]. In view of the complexity inherent in the determination of NO2Tyr, and the confounding results evident in the literature, MS has thus been adopted by several groups [18]. Furthermore, comparing with GC-based methods, LC-MS methods offer advantages such as that it is no longer necessary to modify the analyte to impart volatility. Because chemical manipulation can be eliminated, sample handling, the potential for side reactions, losses, and contamination are also minimized [17]. These complex matrices require, however, a careful consideration in order to evaluate and eliminate matrix effects when developing an LC-MS assay, particularly because of matrix effects/signal suppression, the Achilles’ heel of quantitative LC-electrospray (ESI)-MS [20]. In LC-ESI-MS, methods skipping sample cleanup stages lead to poor analytical performance, in particular, when complex matrices are addressed and sensitive methods are needed. In the present work, a sensitive, simple, and specific sample preparation method based on mixed-mode solid-phase extraction (SPE) was developed for the accurate quantification of trace NO2Tyr in plant tissues by liquid chromatography–electrospray time-of-flight mass spectrometry (LC-TOFMS) using Arabidopsis thaliana, as model sample.

Materials and methods

Reagents and materials

Tyrosine (Aldrich) and 3-nitro-l-tyrosine (Aldrich) standards were purchased from Sigma-Aldrich (St. Louis, MO, USA). Stock solutions of the studied compounds (1.77 mmol L−1 of nitrotyrosine and 1.10 mmol L−1 of Tyr) were prepared in water and stored at −20 °C. HPLC-grade solvents acetonitrile (Chromasolv® Gradient) and methanol (Chromasolv® for HPLC) were purchased from Sigma-Aldrich. Formic acid was obtained from Fluka (Buchs, Switzerland). A solution of 5 % (v/v) ammonium hydroxide (Sigma-Aldrich) in methanol was employed in SPE procedure. A Milli-Q-Plus ultrapure water system from Millipore (Milford, MA) was used throughout the study to obtain the HPLC-grade water. The SPE cartridges evaluated for comparing cleanup were Strata™-X-C cartridges with a capacity of 30 mg, (Phenomenex, Torrance, CA, USA); AccuBONDII SCX cartridges (200 mg, 3 mL) were acquired from Agilent Technologies (Waldbronn, Germany); Oasis MCX SPE cartridges (150 mg, 6 mL) and Oasis HLB (200 mg, 6 mL) were purchased from Waters (Milford, MA, USA). Additionally, a Supelco (Bellefonte, PA, USA) Visiprep™ SPE vacuum system was also employed.

Sample preparation and treatment

A. thaliana L. (Columbia ecotype) cell suspension culture was kindly provided by the Instituto de Recursos Naturales y Agrobiología de Salamanca (IRNASA-CSIC), Salamanca (Spain). The culture was maintained in 200 mL of liquid growth medium [21, 22] by gentle agitation at 120 rpm and 24 °C under continuous illumination (50 μE m−2 s−1) in an incubator shaker. Cells were sub-cultured with a one twentieth dilution every 7 days. The treatment of the cell culture was performed as described by Chaki et al. [23, 24]. The cell culture was treated with different concentrations of peroxynitrite by infusion for one hour in the same cell culture conditions. After an hour, cell suspension culture was grounded and homogenized in liquid nitrogen using a mortar and pestle. The resulting powder was suspended into 1/2 (w/v) digestion buffer (50 mmol L−1 sodium acetate, pH 6.5) according to Hensley et al. [25]. Homogenates were then filtered through one layer of Miracloth (Calbiochem, San Diego, CA, USA) and centrifuged at 3,000×g for 10 min. The supernatant proteins were then precipitated by the addition of 10 % trichloroacetic acid (TCA). After incubation at 4 °C for 20 min, the samples were centrifuged at 14,000×g for 10 min. Protein pellets were washed twice with acetone at −20 °C, air-dried, and re-suspended in 1 mL of digestion buffer containing 4 mg of pronase (Calbiochem), and incubated at 50 °C for 30 h with gentle stirring. The digested samples were treated with 10 % TCA at 4 °C for 20 min followed by centrifugation at 14,000×g for 10 min. The pH of the supernatant was adjusted to 3. The supernatants were passed through 0.45 μm PVDF filter.

Mixed-mode solid-phase extraction cleanup

Strata™-X-C cartridges cation-exchange cartridges with a capacity of 30 mg, with a mixed-mode stationary phase (strong cation-exchange and reverse-phase) were used to perform the SPE-based cleanup. The cartridges were placed on a vacuum SPE manifold being preconditioned with 1 mL of methanol and 1 mL of 0.1 N HCl in water at a flow rate of 2 mL min−1. Subsequently, 2.5 mL of plant extract (previous adjustment to pH 3) was loaded onto the SPE cartridge, at a flow rate of 1 mL min−1. Finally, the sample was eluted into the test tube using twice 2 mL of 5 % (v/v) ammonium hydroxide in methanol at 1 mL min−1. The eluate pH was then neutralized by vacuum evaporation of the ammonium hydroxide. Samples were evaporated until near dryness by a gentle nitrogen stream and reconstituted with 500 μL of methanol/H20 (20 %, v/v; final preconcentration factor 5:1) prior to analysis. The extract was finally filtered through a 0.45 μm PTFE filter (Millex FG, Millipore, Millford, MA, USA). For validation and quantitation purposes, matrix-matched standards were prepared by spiking the extracts with appropriate volume of NO2Tyr working standard solution before the SPE extraction procedure.

Additional experiments were also performed using cation-exchange and reverse-phase type SPE cartridges. Two cation-exchange cartridges (AccuBONDII SCX (200 mg, 3 mL) and Oasis MCX SPE cartridges (150 mg, 6 mL)) were also tested although they were not selected as the final optimized method. The cation-exchange SPE cartridges were washed with MeOH (5 ml) and 5 mL of 0.1 M HCl in water at a flow rate of 2 mL min−1. Subsequently, 10 mL of plant extract (previous adjustment to pH 3) was loaded onto the SPE cartridge, at a flow rate of 1 mL min−1. Finally, the sample was eluted into the test tube using twice 2.5 mL of 5 % (v/v) ammonium hydroxide in methanol at 1 mL min−1. The resulting extract were evaporated until near dryness by a gentle nitrogen stream and reconstituted with 2 mL of methanol/H20 (20 %, v/v) prior to LC-MS analysis.

Besides, a hydrophilic–lipophilic balanced Oasis HLB cartridge was also tested (200 mg, 6 mL). The cartridge was washed with MeOH (5 ml) and 5 mL of mQ water at a flow rate of 2 mL min−1. Subsequently, 10 mL of plant extract (previous adjustment to pH 3) was loaded onto the SPE cartridge, at a flow rate of 1 mL min−1. Finally, the sample was eluted into the test tube using twice 5 mL methanol at 1 mL min−1. The resulting extract was evaporated until near dryness by a gentle nitrogen stream and reconstituted with 2 mL of methanol/H20 (20 %, v/v) prior to analysis.

Liquid chromatography–electrospray time-of-flight mass spectrometry

The separation of the species from the whole SPE extracts was carried out using an HPLC system (consisting of vacuum degasser, auto-sampler, and a binary pump; Agilent 1290 Infinity, Agilent Technologies, Santa Clara, CA, USA). Optimization studies were carried out with standard mixtures performing chromatographic separation on an Agilent ZORBAX Eclipse XDB-C18, Rapid Resolution HT (4.6 × 100 mm, 1.8 μm). For the elution, 0.1 % (v/v) formic acid in high-purity water (mobile phase A) and acetonitrile (mobile phase B) were used as solvents at a flow rate of 500 μL min−1. The gradient program started at 5 % B and after 2 min of isocratic run solvent B was increased linearly and reached 50 % at 10 min, then 100 % at 13 min. Finally, 100 % B was kept constant for 2 min (until 15 min) and after the acquisition 10 min post-time was set for the equilibration of the initial solvent composition. The column temperature was maintained at 24 °C and an injection volume of 20 μL was used in all experiments.

The HPLC system was connected to a time-of-flight mass spectrometer Agilent 6220 TOF (Agilent Technologies, Santa Clara, CA) equipped with an ESI interface operating in positive or negative ion mode, using the following operation parameters: capillary voltage, ±4,000 V; nebulizer pressure, 40 psig; drying gas flow rate, 9 L min−1; gas temperature, 325 °C; skimmer voltage, 65 V; and fragmentor voltage (in-source CID fragmentation), 170 V in positive ion mode. LC–MS accurate mass spectra were recorded across the range of 50–1,000 m/z. Accurate mass measurements of each peak from the total ion chromatograms (TICs) were obtained using an automated calibrant delivery system to provide the correction of the masses. The instrument performed the internal mass calibration automatically, using a dual-nebulizer ESI source with an automated calibrant delivery system, which introduces the flow from the outlet of the chromatograph together with a low flow (approximately 10 μL min−1) of a calibrating solution which contains the internal reference masses purine (C5H4N4 at m/z 121.050873) and HP-0921 ([hexakis-(1H,1H,3H-tetrafluoropentoxy)-phosphazene] (C18H18O6N3P3F24) at m/z 922.009798)). The full-scan data recorded were processed with Agilent Mass Hunter software (version B.04.00). Extracted ion chromatograms (EICs) were obtained throughout the study using ±5 mDa mass window.

Results and discussion

Identification and confirmation of Tyr and nitrotyrosine by LC-ESI-TOFMS: in-source CID fragmentation and accurate mass measurements

The fragmentor voltage is the parameter that establishes the extent in which in-source CID fragmentation is carried out, which may have a strong influence on the sensitivity and relative abundance of protonated molecules [26]. Due to the low masses of both Tyr and NO2Tyr, the fragmentor voltage was set at 170 V (mild conditions), as a compromise value between sensitivity for quantitation and additional mass spectrum information for confirmation purposes. Using the selected conditions, useful fragmentation was obtained. Table 1 shows the fragmentation of Tyr and NO2Tyr and the relative abundances of the different species formed.

Table 1 Identification of tyrosine and nitrotyrosine by LC-TOFMS. Accurate mass measurements of the protonated molecules and the main fragment ions using Arabidopsis thaliana extractsa
Full size table

Primary identification of both compounds was performed basically by retention time matching and accurate mass measurements of the targeted protonated molecules and their main fragment ions. By using high-resolution mass spectrometry data with high mass accuracies, as those shown in Table 1, unambiguous identification of the targeted species was accomplished. For identification and quantitation purposes, extracted ion chromatograms were employed, using a mass-window width of 5 mDa ([M+H]+ ± 5 mDa). The protonated molecule ([M+H]+) was used for both identification and quantitation purposes for NO2Tyr and Tyr. Accurate-mass data from additional fragment ions available for NO2Tyr and Tyr were used for further confirmation. Figure 1 shows LC-TOFMS mass spectra of Tyr and NO2Tyr obtained in the positive ionization mode.

Fig. 1
figure 1

LC-electrospray TOFMS mass spectra of Tyr and NO2Tyr acquired in the positive ionization mode

Full size image

Sample treatment and recovery studies

After unsuccessful attempt of direct injection of the A. thaliana extract (Figs. 2 and 3), in order to eliminate additional interfering species from the sample extract, a SPE cleanup step was evaluated and included in the method. Although slightly more time-consuming, the improvement in chromatographic performance provided by the SPE step was significant. Additionally, the extraction method could be easily automated using a SPE-LC-TOFMS assembly, thus increasing the throughput and automation degree of the procedure. Inspection of the structure, solubility data, and acid/base properties of Tyr and NO2Tyr suggests that it can be extracted by different mechanisms. For example, ionic interactions could be increased through pH variation. Furthermore, the aromatic side chain of Tyr and NO2Tyr (Fig. 1) can be involved in stacking (non-polar) interactions with other aromatic side-chains; and the reactive hydroxyl group can be involved in polar interactions such as hydrogen bonding. Therefore, different sorbent materials with non-polar, polar, or ion-exchange properties were evaluated.

Fig. 2
figure 2

Total ion chromatograms (TICs) from the plant extracts obtained from the raw extract (a) without further treatment and (b) with the proposed SPE-based cleanup method

Full size image
Fig. 3
figure 3

Extracted ion chromatograms (EICs) for the detection of NO2Tyr in the studied A. thaliana cell culture extracts: (a) EIC obtained from the raw extract without further treatment ([NO2Tyr] = 500 nmol L−1) and (b) EIC obtained with the proposed SPE-based cleanup method ([NO2Tyr] = 100 nmol L−1)

Full size image

Among strong cation exchangers (SCX), a cartridge based on silica (AccuBONDII SCX) was tested. It is generally employed to extract positively charged basic compounds. Moreover, this benzene-sulphonic acid-based sorbent has significant non-polar secondary interactions. Different cartridges with a mixed-mode stationary phase (MCX) with reverse-phase and cation-exchange dual functionality, such as Oasis MCX and Strata™-X-C were also evaluated. Besides, HLB cartridges, which have both hydrophilic and lipophilic properties, generally employed to extract a variety of polar and non-polar compounds, were also considered. The cleanest chromatograms were obtained when cation exchange-based materials were employed. Among these materials, best recoveries were obtained when Strata™-X-C cartridges were employed (91 and 83 % recovery for Tyr and NO2Tyr for Strata versus 61 and 46 % for Tyr and NO2Tyr with Oasis MCX), and the extracts obtained were particularly clean. Therefore, in order to maximize the retentive differences between the analytes and the vegetable matrix, Strata™-X-C cartridges were employed for isolate the analytes from the matrix. For the SPE step, 2.5 mL of vegetable matrix sample were selected as the loaded volume. The preconcentration factor achieved in the final extract (500 μL) was 5:1.

The pH is a significant variable when developing a SPE method. Interactions between the matrix components and the target analytes in biological samples may be disrupted by a change in pH [27]. Thus, spiked matrix stabilized at neutral (pH 7) and acidic (pH 3) pHs were evaluated for both MCX cartridges. A significant improvement on analytes recoveries was observed when Strata cartridges at acidic pH (85–90 % recovery for Tyr and NO2Tyr) were employed, comparing with those at neutral pH (lower than 10 % for both analytes). Therefore, pH from samples was adjusted to 3 before SPE.

Figure 2 shows a comparison of TICs obtained from the raw A. thaliana extract without further treatment (Fig. 2a) and with the proposed SPE-based cleanup method (Fig. 2b). The TICs accounts for the sum of signals for all coeluting ions at each individual acquired spectrum. This can be used as an indicator of the complexity of a matrix and to evaluate the degree of efficiency of a cleanup step. Note that the TIC obtained with the SPE procedure is cleaner (lower average signal) than the raw extract, even considering the 5:1 preconcentration factor, as shown in the chromatogram where major matrix peaks are baseline-separated and the TIC current is similar to the raw extract despite the preconcentration factor (5:1). With this SPE approach, the chromatogram region where NO2Tyr is detected is free of several coeluting interfering species. In Fig. 3, EICs for the detection of NO2Tyr in the studied plant extracts obtained from (a) the raw extract without further treatment (500 nmol L−1 NO2Tyr) and (b) with the proposed SPE-based cleanup method (100 nmol L−1 NO2Tyr in the original A. thaliana extract), are shown. It can be seen that at the studied concentration level (500 nmol L−1), NO2Tyr could not be detected in the raw extract due to strong signal suppression due to matrix coeluting components. In contrast, the identification of NO2Tyr with the SPE approach was straightforward.

Even though after the sample treatment protocol a cleaner extract is obtained, the impact of the matrix on the ionization suppression/enhancement on the analytes was still significant. Therefore, a calibration with matrix-matched standards was employed throughout the study to minimize errors due to matrix effects.

Analytical performance: in vitro nitration of A. thaliana cells

To evaluate the analytical features of the proposed method, calibration curves were constructed at different concentrations, in the range 10–500 and 50–2,500 nmol L−1 of Tyr and NO2Tyr, respectively, using vegetable extracts to prepare matrix-matched standards at several concentration levels (2–100 and 10–500 nmol L−1 of Tyr and NO2Tyr, respectively), considering the SPE preconcentration factor. The results obtained are shown in Table 2 where the calibration curves are summarized together with the limits of quantitation (LOQs), matrix effects and relative standard deviation (RSD, %). The linearity of the analytical response across the studied range was excellent, taking into account that the calibration curves of the analyzed compounds showed correlation coefficients higher than 0.996. The RSD (n = 6) values for run-to-run study were 2.7 and 3.4 % for Tyr and NO2Tyr, respectively. These results demonstrate the precision of the developed method and the potential of the proposed approach for quantitative purposes. The LOQs were estimated as the minimum concentration of analyte corresponding to a signal-to-noise ratio (S/N) = 10:1. This was experimentally calculated from the injection of matrix-matched standard solutions at low concentration levels, using the more abundant ion for each compound based on the signal from high-resolution EICs with narrow mass windows (targeted mass ± 5 mDa). The LOQ obtained for NO2Tyr was 3 nmol L−1. Compared with the concentration levels that were achieved in previous reported methods for other biological matrices and considering the complexity of the studied extract, the LOQs reported here can be considered very satisfactory for the targeted application [28]. In the case of Tyr, LOQ could not be calculated because it is already present at large excess compared with NO2Tyr in the studied samples. The LOQ of neat Tyr solvent standard was 10 nmol L−1 (without preconcentration step).

Table 2 Analytical parameters for the detection of nitrotyrosine in Arabidopsis thaliana cell culture extracts by LC-TOFMS
Full size table

To evaluate the effectiveness of the extraction method, a recovery study was carried out. A. thaliana L. cell culture was incubated with two different concentrations of pure peroxynitrite (1 and 5 mmol L−1), which had been shown to mediate Tyr nitration [24]. After sample preparation (explained in “Sample preparation and treatment”), the aliquots were spiked at different concentration levels (0.5–1 μmol L−1) with the working standard solutions of Tyr and NO2Tyr. The spiked samples were extracted with the SPE method described and then analyzed with the developed LC-TOFMS method. Due to the high concentration level differences between Tyr and NO2Tyr (ca. 3 orders of magnitude) it was extremely difficult to accurately measure both Tyr and NO2Tyr in the same run. This limitation is set by mass spectrometer, which usually features 2.5–3 orders of linear dynamic range. For this reason, and also to skip matrix effects for accurate Tyr quantitation, a 1:100 dilution of the extract was also analyzed which enabled the determination of Tyr without matrix effects, just by using external solvent-based calibration. A LC-ESI(+)TOFMS identification of NO2Tyr in cell cultures exposed to peroxynitrite is shown in Fig. 4. The obtained recoveries rates for NO2Tyr were in the range 92–101 %, as shown in Table 3. These results show the feasibility of the studied extraction method for NO2Tyr determination in the studied vegetable extracts. Besides, in the samples tested, both Tyr and NO2Tyr were calculated for both experiments (1 and 5 mmol L−1 of peroxynitrite). Interestingly, a linear correlation tendency between concentration of peroxynitrite and NO2Tyr/Tyr ratio was observed. This proportional increase in the concentration of NO2Tyr when increasing the concentrations of peroxynitrite corroborates the use of NO2Tyr as a marker of nitrosative stress in plants.

Fig. 4
figure 4

LC-ESI(+)TOFMS identification of NO2Tyr in A. thaliana cell cultures exposed to peroxynitrite (nitrating compound). Left extracted ion chromatogram (m/z 227.0667); right electrospray TOFMS spectrum including ions at m/z 227.0666 and 181.0611 that provides the unambiguous identification of NO2Tyr in the studied extracts

Full size image
Table 3 Recovery studies on Arabidopsis thaliana cell culture extracts treated with peroxynitrite (1 and 5 mmol L−1)
Full size table

Conclusions

The present work described a new method based on SPE and LC-TOFMS for quantitative analyses of NO2Tyr and Tyr in A. thaliana cell culture extracts. The method is simple, since it involves a quick cleanup step with SPE employing polymer-based cartridges before measurement with LC-TOFMS. Satisfactory recoveries were obtained for both studied compounds. Moreover, the high sensitivity obtained with the proposed method compares well with previous LC-MS/MS methods described for the analyses of NO2Tyr and Tyr in biological matrices.

The potential of the proposed method was demonstrated by analyzing real samples with excellent selectivity and sensitivity, thus enabling the unambiguous identification, by means of accurate mass analysis, and quantitation of low levels of NO2Tyr in A. thaliana cell culture extracts.

The proposed LC-TOFMS method also offers the possibility of performing a posteriori (non-target) analysis of the samples, such as the search and identification of others PTM-Tyr compounds (such as sulfation, phosphorylation, or carbonylation), involved in the regulation of a wide range of biological processes [29]. All the data are saved and can be re-examined to check for compounds that previously were not expected or were not subjected to control. This is an additional attractive feature that highlights the potential application of this method based on LC-TOFMS for studies related to PTM in biochemical laboratories worldwide.