Mercury alters endogenous phosphorylation profiles of SYK in murine B cells

Murine WEHI-231 B cell lymphoma cells were obtained from the American Type Culture Collection, Rockville MD. Cells were maintained in RPMI 1640 supplemented with 10% fetal bovine serum, 2 mM glutamine, 50 μM mercaptoethanol, 100 U/ml penicillin and 100 μg/ml streptomycin in a humidified 5% CO₂ atmosphere. Cells were passaged three times a week and harvested for experiments while in logarithmic growth. On the day of the experiment, cells were washed, counted, and resuspended in serum-free medium.

7.5 × 10⁷ WEHI cells per 5 ml were mock treated or treated with activating antibody (100 μg of anti-immunoglobulin M per 10⁶ cells) for the indicated times, or the indicated concentrations of pervanadate or HgCl₂ for 10 min at 37 °C. At the endpoint, cells were cooled in an ice bath, pelleted at 4 °C, washed with ice-cold wash buffer [150 mM NaCl, 50 mM Tris (pH 7.5), 1 mM NaF, 1 mM Na₃VO₄, 2 mM β-glycerophosphate], and re-pelleted. Cells were lysed with wash buffer containing 1% NP-40, 0.5% deoxycholate, 0.1% SDS, 2 mM MgCl₂ and 100 U Benzonase (Novagen). After a 5 min digestion at 37 °C, 1X Protease Inhibitor Cocktail (Sigma) and 10 mM EDTA were added. Samples were centrifuged at high speed and supernatants transferred to new tubes, pre-cleared with Protein A/G-agarose, and then immunoprecipitated with 3 μg each of 2 anti-SYK antibodies (mouse monoclonal sc-73089 and rabbit polyclonal sc-1077, Santa Cruz Biotechnology) and 50 μl of Protein A/G-agarose overnight with end-over-end rotation at 4 °C. The following day, the beads were washed and proteins eluted by boiling with SDS sample buffer under reducing conditions. Proteins were separated by SDS-PAGE using 4-12% Bis-Tris gels (Life Technologies). For initial experiments proteins were transferred to nitrocellulose for western blotting to determine the migration of SYK. In subsequent experiments gels were stained with Coomassie Blue and gel slices were excised for mass spectrometric analysis.

WEHI cells were treated with anti-IgM (100 μg/10⁶ cells) for 0, 1, 2, 3, 5 and 15 min; or with HgCl₂ for 10 min at 0, 1, 5, 10, 20, 50 and 100 μM concentrations; or with pervanadate for 10 min at 0, 1, 3, 10, 30 and 100 μM concentrations. At the endpoint, 9 ml of cold HANKS buffer was added, and cells were cooled in an ice bath. Cells were pelleted at low speed at 4 °C, and then transferred to 1.5 ml tubes in 1 ml of HANKS. Cells were pelleted and resuspended in 300 μl of 40 mM Tris, pH 8.0, 0.5% lithium dodecyl sulfate (LiDS) and the sample heated to 90 °C for 5 min. Protein concentration was estimated using the BCA assay. Samples were reduced using 5 mM TCEP, alkylated with 15 mM iodoacetamide, and digested overnight with trypsin at 37 °C. For phosphotyrosine enrichment, 1 mg of protein digest was passed through a YM-10 cartridge (Millipore) to remove excess trypsin, and a Detergent Removal Column (Pierce) to remove LiDS. Samples were brought up to a 1X IAP buffer solution (50 mM MOPS (pH 7.2), 10 mM Na₂HPO₄, 50 mM NaCl). 4G10 anti-phosphotyrosine-agarose conjugate (Millipore) was washed with IAP buffer and 50 μg was added to each sample followed by incubation overnight at 4 °C with end-over-end rotation. The 4G10 beads were set aside after centrifugation and the supernatants transferred to new tubes containing washed PTMScan beads (20 μl slurry of P-Tyr-100 and P-Tyr-1000, Cell Signaling Technology) followed by incubation for 4 h at 4 °C with end-over-end rotation. Both sets of beads were sequentially washed several times with IAP buffer and then water. Peptides were eluted from the beads using 2% acetonitrile (ACN), 0.1% formic acid, and then dried for storage. For phosphopeptide enrichment using titanium dioxide, 1 mg of digested protein was resuspended in a saturated glutamic acid solution containing 65% ACN, 2% trifluoroacetic acid (TFA). 10 mg of titanium dioxide beads was added (GL Sciences) followed by end-over-end rotation for 20 min at RT. After a series of washes, the phosphopeptides were eluted from the beads stepwise using a gradient of NH₄OH from 300 mM to 1 M, and 60% ACN. Elutions were pooled and acidified using formic acid (1.3% final), and dried for storage.

Peptides were resuspended in 5% ACN, 0.1% formic acid prior to reverse phase chromatographic separation using an EASY nLC-1000 UHPLC system (Thermo). For SYK immunoprecipitation experiments, peptides were analyzed with an Orbitrap Fusion mass spectrometer (Thermo). MS1 scan details include 1.2 × 10⁵ resolution, 4 × 10⁴ AGC target, 350–1600 m/z scan range; MS2 scan details include CID fragmentation and ion trap detection, an AGC target of 1 × 10⁴, and 35% collision energy. Samples derived from phosphopeptide enrichment experiments were analyzed on a Q Exactive mass spectrometer (Thermo). MS1 scan details include 7 × 10⁴ resolution, 3 × 10⁶ AGC target, 350–1800 m/z scan range; MS2 scan details include HCD fragmentation and Orbitrap detection, an AGC target of 2 × 10⁵, 1.75 × 10⁴ resolution and 30% normalized collision energy.

For each sample, the area of the gel lane where SYK was pre-determined to localize, between 60 and 90 kDa, was excised. Gel slices were reduced with 5 mM TCEP, alkylated with 15 mM iodoacetamide and in-gel digested with 0.04 μg trypsin (Promega). Eluted peptides were resuspended in 5% ACN, 0.1% formic acid, and 0.005% TFA and separated by reverse phase chromatography. Samples were analyzed on a TSQ Vantage triple quadrupole mass spectrometer (Thermo). Instrument settings included a FWHM of 0.7 and cycle time of 1.75 s. Method optimization was performed to determine i) the transitions which gave the highest signal, ii) the optimal collision energy for each transition, and iii) the peptide retention times. 27 transitions were monitored for 4 min each within a 35 min gradient. Data was analyzed using Skyline software (MacCoss lab, University of Washington, version 2.5). Peak areas for each transition were integrated, averaged for 3 biological replicates, and then normalized against an unphosphorylated control SYK peptide to control for changes in SYK protein abundance. The untreated samples were then normalized to a value of 1 and the sample groups were adjusted accordingly. Statistical analyses were performed using two-tailed T-tests assuming unequal variances.

The two main goals of this work are to determine i) all the potential phosphosites of murine SYK, and ii) how the phosphorylation status of these sites are altered in response to mercury exposure. WEHI-231 is a B cell lymphoma cell line that is commonly used to study immature B lymphocytes since it readily undergoes apoptosis in response to BCR signaling [9]. Phosphosite determination was made for SYK in WEHI protein digests by utilizing LC-MS/MS on peptides after anti-SYK or -phosphotyrosine immunoprecipitations, or after phosphopeptide enrichment using titanium dioxide chromatography.

Peptides can possess multiple serine, threonine and/or tyrosine residues, therefore localization of phosphorylated residues is predicated on the internal cleavage pattern after isolation and fragmentation within the mass spectrometer. For collision-induced dissociation (CID) or higher-energy collisional dissociation (HCD) cleavage reactions, this is determined by the overlapping b- and y-ion series. For example, spectra utilized to identify phosphorylated S270 and S279 are shown in Fig. 1 for the peptide 261-IGAQMGHPGSPNAHPVTWSPGGIISR-286. Note that b ₁₂, b ₁₃ and b ₁₄ are shifted to the right when S270 (Serine²⁷⁰) is phosphorylated in A, whereas y ₁₀, y ₁₁ and y ₁₂ are shifted to the right when S279 is phosphorylated in B. These fragmentation ions, or ‘transitions’, are of prime importance for accurate quantitation of phosphosites by MRM.

Based upon published genomic data, 96.7% of SYK amino acid residues were identified and sequenced in our experiments (Fig. 2a and Additional file 1). The only potential phosphosites not sequenced were Y215 and T392. In total, when results from all experimental conditions are included, 23 of the potential phosphosites were positively determined to be phosphorylated with high localization probability (Fig. 2 and Additional file 2). This included 4 phosphoserine residues in the central portion of SYK, and the remainder as phosphotyrosines along the length of the entire sequence. Phosphorylation of threonine residues were not identified by this unbiased approach in which site specific probability was estimated to be ≥95%.

Significantly, the profile of phosphorylated SYK residues depends upon the external cell environment. When the source of the phosphopeptides are considered according to cell treatment (Fig. 2c), it is evident that the distribution is not randomly dispersed for phosphosites that are observed before and after BCR activation. For example, after BCR activation with anti-IgM, we find several sites which are phosphorylated within the Linker B region and the C-terminal portion of the kinase domain, while in control cells we find no evidence that these sites are phosphorylated. On the other hand, mercury treatment led to identification of 4 phosphosites not observed in controls or the anti-IgM activated group, In this case Y46, Y358 and Y501 identifications were only observed at the highest and very toxic mercury concentration (100 μM), and with only 1 or 2 spectral identifications for each site (Additional file 3). In contrast, Y73 had dozens of spectral identifications at either 50 or 100 μM Hg²⁺, indicating it was more susceptible to modification following mercury exposure.

Pervanadate was included in these experiments since this compound effectively inhibits protein tyrosine phosphatases, particularly PTP1B, by irreversible oxidation of the active-site cysteine residues to cysteic acid [10]. Therefore, pervanadate would unmask weakly phosphorylated tyrosine residues in control cells by shifting the dynamic equilibrium of a phosphosite to the phosphorylated form. By spectral counting, pervanadate increased the number of phosphosites identified (Fig. 2c), and accounted for about 62% of all phosphopeptides identified (Additional file 3). The majority of the phosphotyrosine sites uncovered by pervanadate treatment (9 of 10) occur within SH2 domains which are critical for protein-protein interaction.

To achieve greater sensitivity for profiling changes in SYK phosphorylation status due to mercury exposure, an approach was undertaken which we previously utilized to characterize the tyrosine-protein kinase LYN [8]. Sensitivity of mass spectrometry is greatly enhanced by reducing interfering background ions. To this end, after mercury exposures, SYK was first immunoprecipitated from cell extracts, and then the samples were further purified by excising defined areas from SDS-PAGE gels (Fig. 3). Peptides were prepared from gel slices and analyzed by MRM. The advantage of this approach over antibody-based methods is the specificity that is attained for protein phosphosites through selection of unique transition ions.

Phosphopeptides which were associated with mercury exposure were tested to determine whether they were amenable to MRM analysis. For transitions to be successfully incorporated into MRM, the parent peptide should have a low charge state and good signal intensity, and fragmentation efficiency should be great enough to yield high signal intensity for the selected transition ions. Phosphosites from 4 clusters met these criteria: S270/S279 and Y342/Y346 clusters in the Linker B region; and Y519/Y520 and Y624/Y625 clusters in the kinase domain (Table 1 and Fig. 4). A probability-based localization algorithm (Ascore) indicated the unlikelihood of phosphorylation at T277, S285, T339, S344 and Y625 within these cluster regions. However it cannot be ruled out with 100% certainty that these sites are not phosphorylated based upon their location within the peptide sequence relative to the cleavage site so they are included in the results.

Each transition was optimized for collision energy (Table 1) and chromatographic retention time and then a method was developed to monitor the signal intensity of all 27 transitions within a single LC-MS/MS run. Samples were treated in triplicate for each dosage within an experiment, and the final data represent the average of 3 separate experiments (Fig. 4). The results suggest that mercury can increase the phosphorylation of selected SYK phosphosites in a dose-dependent manner. This finding was most evident for phosphorylation of the Y342/Y346 cluster, which was induced to levels which were up to 13-fold higher with 25 μM mercury and 94-fold higher with 50 μM compared to mock-treated cell cultures. For all the sites, 50 μM mercury induced at least a 2-fold increase in phosphorylation, whereas 5 and 25 μM had no or more modest effects.

WEHI cell cultures were also exposed to either an anti-IgM antibody or pervanadate as a reference for SYK activation via BCR activation and phosphatase inhibition, respectively. Neither of the agents at the dosages and times tested activated phosphosites in the S270/S270 cluster, but both agents activated the Y342/Y346 and Y624/Y625 clusters (Fig. 4b). Antibody exposure led to increased phosphorylation of Y519 or Y520 (up to 79-fold higher relative to control), whereas pervanadate had little effect on this cluster. These results demonstrate that exogenous agents can impart unique phosphorylation profiles on SYK, and that these signatures can be monitored using sensitive and selective mass spectrometry-based techniques.