Antibody Enrichment and Mass Spectrometry of ... - ACS Publications

Article pubs.acs.org/crt

Antibody Enrichment and Mass Spectrometry of Albumin-Cys34 Adducts Ming-Kei Chung,† Hasmik Grigoryan,† Anthony T. Iavarone,‡ and Stephen M. Rappaport*,† †

Center for Exposure Biology, School of Public Health, University of California, Berkeley, California 94720, United States Department of Chemistry and QB3 Institute, University of California, Berkeley, California 94720, United States

‡

S Supporting Information *

ABSTRACT: Untargeted analyses of tryptic peptides of human serum albumin (HSA) have been used to investigate unknown exposures to reactive electrophiles (adductomics). To reduce the complexity of the analytical matrix and thereby enhance identification of adducts by liquid chromatography-high-resolution mass spectrometry (LC-HRMS), a polyclonal anti-T3 antibody was designed to capture Cys34 adducts in tryptic digests of HSA (T3 is the third largest tryptic peptide). Epitopes were selected from sequences at both C- and N-termini based on the threedimensional structure of the T3 peptide to minimize the influence of modified Cys34 residues. The assay was simplified by attaching magnetic beads to the anti-T3 antibody. When applied to commercial HSA and to plasma samples from healthy humans and analyzed by LC-HRMS, antibody treatment greatly reduced the background of non-T3 peptides in the sample matrix. Although other lipophilic HSA peptides were still present, presumably due to nonspecific binding to the antibody-magnetic-bead surfaces, their concentrations in antibody-treated samples were reduced about 6-fold compared to the same samples that had not been treated with the antibody. Analysis of antibody-enriched HSA digests from human plasma samples revealed 10 modified T3 peptides of which 8 were identified from accurate masses. Identified peptides included Cys34 oxidation and cysteinylation products and modifications representing losses of water and Lys and transpeptidation of Arg.

■

INTRODUCTION Because they react with functional nucleophiles in DNA and proteins, electrophilic chemicals initiate and progress tumor development,1 damage organs,2 and contribute to chronic diseases.3 Although electrophiles are generally too unstable to measure directly in biological specimens, investigators have inferred their dispositions by measuring products of reactions with nucleophiles in blood, cells or tissues.4−7 One approach to untargeted analyses of such adducts, called adductomics, has focused upon modifications at the Cys34 locus of HSA (the only free thiol in HSA), which is the most abundant nucleophile in serum.8−10 The adductomics method for Cys34 adducts8 involves isolation of HSA from serum or plasma, removal of mercaptalbumin (i.e., HSA with unmodified Cys34) by thiolaffinity chromatography, and digestion of the adduct-enriched protein with trypsin. Cys34 adducts are present in the thirdlargest peptide (T3, with sequence ALVLIAFAQYLQQC34PFEDHVK). Following high performance liquid chromatography (HPLC) of tryptic digests, the fraction containing T3 peptides is collected and analyzed en masse by triple-quadrupole mass spectrometry (MS). Adducted T3 peptides are visualized in a heat map showing hits in added-mass bins from 9 to 350 Da for different HSA samples. The identities of particular T3 adducts have been explored in tryptic digests via LC-high resolution MS (HRMS), adductdatabase searches, and comparisons with synthetic T3 adducts.10 This has been challenging because digests contain © 2013 American Chemical Society

many coeluting peptides that increase background noise and suppress ionization in the electrospray. As a possible strategy for purifying the peptide matrix, we were impressed by Anderson et al.’s use of antipeptide antibodies to enrich peptides in protein digests.11−14 However, Anderson et al. focused upon unmodified peptides rather than peptides containing adduct moieties that might influence the immunogenic properties of the capture antibody. Here, we describe a method employing a polyclonal anti-T3 antibody to enrich T3 and the associated Cys34 adducts from tryptic digests of adduct-enriched HSA. By visualizing the 3D structure of the T3 peptide, we selected epitopes from the flank sequences of both the C- and N-termini while minimizing the influence of the modified Cys34 side chain. The desired epitopes motivated the development of an immunogen and, ultimately, of the anti-T3 antibody. Antibody enrichment of Cys34 adducts was further simplified by attaching magnetic beads to the antibody, again motivated by Anderson et al.12 The cross-reactivity of the anti-T3 antibody was evaluated with four synthetic T3 adducts, covering a range of added masses and sizes, and the performance characteristics of several types of magnetic beads were compared. The antibody-enrichment scheme was investigated with digests of commercial HSA and Special Issue: Systems Toxicology Received: September 16, 2013 Published: December 12, 2013 400

dx.doi.org/10.1021/tx400337k | Chem. Res. Toxicol. 2014, 27, 400−407

Chemical Research in Toxicology

Article

Figure 1. Predicted structures of the T3 peptide (ALVLIAFAQYLQQCPFEDHVK). (A) Model predicted by I-TASSER. (B) Model predicted by PEP-FOLD. Cys34 is highlighted in yellow in both structures. Both models (A and B) include a bend of the backbone around Cys34 and an α helix between Cys34 and the N-terminus. (C) Elaborated I-TASSER model. Cys34 is represented in “licorice style” with the sulfur atom shown in yellow. (D) Side view of the elaborated I-TASSER mode. NQ and 1,2-PQ, EQ and iT3-IAA were also synthesized as reported by Li et al. and Grigoryan et al.8,10 Final stock solutions of synthetic adducts were prepared in 30−45% acetonitrile and 0.1% formic acid at concentrations between 20 μM (iT3-IAA) and 100 μM (T3-quinone adducts, representing the sum of diol and quinone forms) and were stored at −80 °C prior to use. Design of the Immunogen. A three-dimensional (3D) model of the T3 peptide in aqueous solution was constructed to design an immunogen for producing anti-T3. To begin, the amino acid sequence of T3 (ALVLIAFAQYLQQCPFEDHVK, from Ala21 to Lys41 of HSA) was obtained from in silico tryptic digestion of HSA (UniProt identifier: P02768) via ProteinProspector (prospector.ucsf.edu). Next, the T3 sequence was inputted to I-TASSER15 and PEP-FOLD,16 using default settings with no additional restraints or exclusions. While ITASSER employs an iterative process to align multiple threading alignments from a library, PEP-FOLD uses a de novo solution based upon structural elements (e.g., helices and strands) following iterative Markov simulations. Visual Molecular Dynamics (v.1.9.1)17 was used to create molecular representations of predicted 3D structures, and the images were rendered with Tachyon. The final models, shown in Figure 1, had the highest c-score (I-TASSER, Figure 1A) and lowest energy (PEP-FOLD, Figure 1B). Because the models derived from such disparate methods were strikingly similar, the predicted structure is probably a good representation of the T3 peptide in an aqueous environment. The rich structural conformation of the peptide surprised us, given only 21 amino acids and no elements for a disulfide bridge. The T3 peptide displays an α helix (ALVLIAFAQYLQQ), two 90° bends around Cys34 (the free sulfhydryl group is shown in yellow) and Pro35, and a random coil in the C-terminal tail (FEDHVK). We suspect that the tail coil is influenced by Pro35 because proline is the only amino acid that can adopt the necessary cis-

HSA isolated from human plasma. Several Cys34 adducts were detected in enriched T3 peptides by LC-HRMS.

■

EXPERIMENTAL PROCEDURES

Chemicals and Reagents. The following chemicals and reagents were obtained from the listed vendors. Fisher Scientific (Pittsburgh, PA): ThermalSeal adhesive films, Tris-buffered saline (TBS, 10×), acetonitrile (Optima grade, 99.9%), formic acid (Optima, LCMS grade), iodoacetamide (IAA), dimethylsulfoxide (DMSO), ammonium sulfate, methanol (LCMS grade), tris(2-carboxyethyl)-phosphine (TCEP), acetic acid (LCMS grade), and Tris Base (≥99.9%). Sigma-Aldrich (St. Louis, MO): carbonate-bicarbonate buffer, Tween 20, 1,4-benzoquinone (1,4-BQ), 1,4-naphthoquinone (1,4-NQ), 4hydroxyestradiol (4-OH-E2), silver(I) oxide, porcine trypsin, ammonium bicarbonate, HSA (lyophilized powder, 97−99%), soybean trypsin inhibitor (Type I-S, lyophilized powder). Bethyl Lab, Inc. (Montgomery, TX): TMB peroxidase substrate. Thermo Scientific (Rockford, IL): SuperBlock, ABC Peroxidase Staining kits. The following chemicals were custom synthesized: Midwest Research Institute (Camp LeJeune, NC), 1,2-phenanthrenequinone (1,2-PQ); Biomatik (Wilmington, DE), T3 peptide with sequence ALVLIAFAQYLQQCPFEDHVK (>97%); BioMer Technology (Pleasanton, CA), isotopically modified T3 (iT3) with sequence AL-[15N, 13C-Val]LIAFAQYLQQCPFEDH-[15N, 13C-Val]-K (>95%). Synthesis of T3 Adducts and the Internal Standard. Stock solutions of T3 and iT3 were prepared in DMSO at 5 mM and then diluted with water or methanol to a final concentration of 100 μM. A carbamidomethylated Cys34 adduct of iT3 (IAA-iT3) was synthesized for use as the internal standard, and individual T3 adducts (modifications to Cys34) of diol and quinone forms of 1,4-BQ, 1,4401



Article

Figure 2. Flow scheme for enrichment and detection of T3 adducts. overnight incubation at 4 °C, the plate was blocked with SuperBlock at 200 μL/well, and the (biotinylated) anti-T3 antibody (0.5 μL/ml) was transferred to the wells. Then, ABC was added, and the plate was incubated at room temperature for 30 min. Finally, TMB was loaded, and the plate was incubated for another 30 min, in the dark. Color intensity was measured at 652 nm with a microplate spectrophotometer (ELx800, Bio-Tek, Winooski, VT). Results from direct ELISA were plotted by Qtiplot (v. 0.9.8.3, ProIndep Serv SRL).27 Preparation of Plasma Samples and Commercial HSA. Blood from three male, nonsmoking laboratory volunteers was obtained by venipuncture. Following centrifugation to remove red blood cells, the serum was immediately frozen and stored at −80 °C for about one month prior to analysis (one freeze−thaw cycle). The flow scheme for analysis is illustrated in Figure 2, based largely on the method described by Grigoryan et al.10 In brief, after treating plasma with saturated ammonium sulfate to precipitate immunoglobulins and other proteins, the supernatant (containing mostly HSA) was buffer exchanged and incubated overnight at room temperature with thiol affinity resin (Sepharose 4B) to remove mercaptalbumin. Mercaptalbumin was similarly removed from samples of commercial HSA. The flow-through fraction was collected and buffer exchanged with digestion buffer (50 mM ammonium bicarbonate and 1 mM EDTA, pH 8), and the protein content was measured with a nanodrop spectrophotometer (ND-1000, Thermo Scientific). Aliquots of 200 μL of enriched HSA were diluted to 1 mg/mL in digestion buffer containing 10% methanol and 5 mM TCEP and incubated for 15 min at 37 °C to reduce disulfide bonds. Trypsin was added at a ratio of 1:10 (trypsin to peptide, w/w), and the mixture was digested in a pressurized system (Barocycler NEP2320, Pressure Biosciences Inc.) cycling between ambient pressure (15 s) and 138 MPa (45 s) at 37 °C for 30 min. The digest was transferred to a glass vial with the addition of a soybean trypsin inhibitor (1:3, trypsin to inhibitor, w:w) and stored at −80 °C. After thawing to room temperature, the samples were carried through the procedures for antibody enrichment and LCMS/MS as described below. Antibody Enrichment of T3 Adducts. To optimize the method, the amount of anti-T3 antibody and type of magnetic bead were investigated using commercial HSA. Twenty microliters of HSA digest, containing 20 μg of HSA peptides, was diluted 10× with 10 mM Tris buffer (pH 7) to a final volume of 200 μL in a 1.5-mL tube. For each treatment, a selected volume of stock solution, containing a predetermined amount of anti-T3 antibody (3 or 24 μg), was added and incubated for 1 h at 37 °C. Magnetic beads (Dynabeads, types M270, M280, C1, or T1; Invitrogen Life Technologies, Grand Island, NY) were washed three times with TBS before mixing with the incubated digest at a ratio suggested by the manufacturer. The tube was mounted on a rotor (Labquake Rotisserie, Thermo Scientific) and incubated for 30 min at room temperature. After the bead−antibody− adduct complexes were formed, the tube was mounted to the DynaMag-2 magnet for 30 s to separate beads from the solution. Beads were washed twice with 10 mM Tris buffer (pH 7) and then twice with deionized water. The T3 adducts were released from the complexes by eluting three times with 21 μL of 1% acetic acid and then diluting to 80 μL with acetonitrile. Finally, all samples were spiked with IAA-iT3 to a concentration of 0.5 nM for commercial HSA and 0.2 nM for plasma HSA.

amide bond formation for inducing a turn. When we reran the T3 sequence without Pro35 (ALVLIAFAQYLQQCFEDHVK), the best models were simple linear helices (results not shown). We sought a polyclonal antibody that would cross-react to a wide range of unknown Cys34 adducts and, therefore, focused upon potential epitopes common to all T3 peptides. Since Cys34 is the 14th amino acid in the T3 peptide, the logical epitope would have to reside in either the N-terminal sequence (ALVLIAFAQYLQQ) that precedes Cys34, the following C-terminal sequence (PFEDHVK), or some combination of the two. The longer N-terminal sequence would maximize the affinity of the antibody, while the shorter C-terminal sequence would encourage greater specificity. To explore these options, in silico antibody specificity was investigated by aligning either the N- or C-terminal sequence to the human proteome through the BLAST server (blast.ncbi.nlm.nih.gov). Human proteins with 100% identity matches to the N-terminal sequence (29 hits) and the C-terminal sequence (32 hits) are shown in Supporting Information (Tables S1 and S2). Aside from HSA, the only match that could be classified as a blood protein was the angiomotin-like protein (Table S1, Supporting Information). Thus, we reasoned that selectivity was unlikely to be a problem and chose the full T3 sequence for the immunogen. Furthermore, most of the overlapping sequences for Nand C-terminal sequences with other human proteins were directed away from the termini (Tables S1 and S2, Supporting Information). Thus, even though acetylation or amidation is commonly employed to remove terminal charges from an immunogen, such modifications would probably not be beneficial here because the two terminal sequences are in close proximity and could be part of a discontinuous epitope (i.e., one that includes sequences on both sides of Cys34; see Figure 1C and D). Finally, we decided to cross-link the carrier protein, keyhole limpet hemocyanin (KLH), directly to the T3-peptide via the free sulfhydryl group of Cys34, thereby conserving the greatest degree of mimicry to T3 adducts. Production of the Anti-T3 Antibody. The polyclonal anti-T3 antibody was prepared to our specifications under contract by Genscript (Piscataway, NJ). In brief, the T3 peptide (2432 Da) was synthesized by Genscript to 92% purity. Then, Genscript prepared the immunogen by coupling the sulfhydryl group from the Cys34 residue of T3 to KLH with a maleimide heterobifunctional cross-linker. The immunogen was mixed with adjuvant and injected into three rabbits with immunizations repeated twice over 8 weeks. After immunization, Genscript screened the antisera for titer quality by ELISA. Because the immunogen was a KLH conjugate, antibody clones responded to KLH, and the cross-linker was removed from antisera using an affinity resin conjugated to the Cys34 residue of the T3 peptide. Finally, Genscript biotinylated the anti-T3 antibody and characterized the product by ELISA. The final product was lyophilized and stored at −80 °C prior to shipment and subsequent use in our laboratory. Prior to use, the anti-T3 antibody was dissolved in water at 1 mg/mL. Cross-Reactivity of the Anti-T3 Antibody. Direct ELISA was used to characterize the cross-reactivity of the anti-T3 antibody. Unless otherwise specified, the loading volume was 100 μL/well; 96-well MaxiSorp plates were rinsed with TBS-T (0.05% Tween) 3 times at 200 μL/well between incubation steps; and incubations were 45 min at 37 °C. In brief, solutions of T3, T3-BQ, T3-NQ, T3-PQ, T3-EQ, and solvent controls were serially diluted 2× and coated onto a plate (with 0.1 M carbonate-bicarbonate buffer, starting from 60 nM/well). After 402



Article

After reviewing results for peptide capture with commercial HSA, the above experiment was repeated with HSA isolated from three fresh plasma samples using 12 μg of anti-T3 antibody bound to T1-type magnetic beads. Mass Spectrometry. Samples were analyzed with an Agilent 1200 HPLC (Santa Clara, CA) connected in-line with an LTQ Orbitrap XL hybrid MS equipped with an Ion Max electrospray ionization source (ESI; Thermo Fisher Scientific, Waltham, MA). Instrumental settings were the same as those reported by Grigoryan et al.10 with minor modifications. The HPLC employed a reversed-phase, C8 analytical column (300SB Zorbax, 1.0 × 150 mm, 3.5-μm particles, 300 Å pore size, Agilent) with a flow rate of 80 μL/min. Solvent A was H2O/0.1% formic acid (v/v) and solvent B was acetonitrile/0.1% FA (v/v). An injection volume of 40 μL (partial loop injection mode) was used. The elution program consisted of a linear gradient from 2% to 25% B over 5 min, an isocratic hold for 5 min, a linear gradient to 60% B over 40 min, and an increase to 98% B over 0.5 min followed by an isocratic hold for 15 min, and re-equilibration of the system at 2% B for 25 min. Mass spectra were acquired in the positive ion mode over the range m/z = 750 to 1000 using the Orbitrap mass analyzer, in profile format, with a resolution setting of 6 × 104 (at m/z = 400, measured at fullwidth-at-half-maximum peak height). In the data-dependent mode, the six most intense precursor ions exceeding an intensity threshold of 10,000 counts were selected from each full-mass spectrum for further MS/MS analysis using collision-induced dissociation (CID). MS/MS spectra were acquired in the positive ion mode using the linear ion trap, in profile format, with the following parameters: precursor ion mass range for data-dependent MS/MS m/z = 750 to 1000, isolation width 3 m/z units, normalized collision energy 28%, activation time 30 ms, activation Q 0.25, and default charge state 3+. Only triply charged precursor ions were selected for CID fragmentation of precursor ions from MS1. To avoid the occurrence of redundant MS/MS measurements, real-time dynamic exclusion was enabled to preclude reselection of previously analyzed precursor ions, using the following parameters: repeat count 2, repeat duration 10 s, exclusion list size 500, exclusion duration 50 s, and exclusion width 20 ppm. LC-HRMS data were analyzed using Xcalibur software (version 2.0.7, Thermo Fisher Scientific). Quantitation of T3 and other HSA peptides was performed based on ratios of peak areas for the triply charged precursor ions of peptide analytes to peak areas for the corresponding precursor ion of the internal standard (IAA-iT3), and are reported as micromolar or nanomolar concentrations in a final volume of 80 μL. Putative T3 Adducts. The list of precursor ions related to putative T3 adducts was generated by screening the MS/MS spectra acquired during data-dependent analysis. A set of seven singly charged b-series fragment ions, b3+ (m/z 284.19), b4+ (m/z 397.28), b5+ (m/z 510.36), b6+ (m/z 581.40), b11+ (m/z 1203.42), b12+ (m/z 1331.77), and b13+ (m/z 1459.83), was selected with a mass tolerance set to ±0.5 m/z. The list of all potential precursor ions was exported to an Excel spreadsheet and inspected to identify arrays containing at least five singly charged b-series fragment ions out of a possible seven. These arrays were then filtered and arranged by retention times (RT). The monoisotopic masses were determined from the isotopic distributions of the corresponding triply charged precursor ions (within a 5-ppm mass tolerance), and the corresponding selected-ion chromatograms (SICs) were extracted from the total-ion chromatograms (TICs). The MS/MS spectrum for each detected precursor ion was then manually screened to confirm or reject the presence of a putative T3 adduct. Since only triply charged precursor ions were screened, masses of putative T3 adducts were calculated from the difference in monoisotopic masses of each putative T3 adduct and the unadducted T3 peptide as 3 × (T3-adduct mass − 811.75933). Elemental compositions of unknown adducts were inferred from accurate masses via searches of the UNIMOD protein-modifications database (www. unimod.org), the Elemental Composition search algorithm built into the Xcalibur software (version 2.0.7, Thermo Fisher Scientific), and the algorithm in Molecular Weight Calculator (v. 6.48, created by Matthew Monroe in 2010). Isotopic distributions of putative T3 adducts were simulated using Xcalibur, and mass errors were

calculated as the differences between measured monoisotopic masses and theoretical monoisotopic masses obtained from the simulated isotopic distributions.

■

RESULTS Affinity of Anti-T3 to Synthetic Adducts. To gauge the affinity of the anti-T3 antibody for T3 peptides representing a range of adduct sizes, we performed direct ELISA of the antibody with four synthetic T3 adducts, representing quinones with increasing molecular weights and numbers of fused rings (BQ = 108.09 Da, one ring; NQ = 158.15 Da, two rings; PQ = 208.21 Da, three rings; EQ = 286.15 Da, four rings). As shown in Figure 3, the order of binding strength was as follows: T3 >

Figure 3. Plots showing the relative strength of anti-T3 binding to the unmodified T3 peptide and synthetic T3 adducts (BQ, benzoquinone; NQ, naphthoquinone; PQ, phenanthrene quinone; and EQ, estrogen quinone).

T3-BQ > T3-NQ > T3-EQ > T3-PQ, suggesting that the affinity of anti-T3 decreased with increasing molecular weight (or ring number) of the modifying species. The only exception was T3-EQ, which showed a higher affinity for anti-T3 than T3PQ despite having a larger molecular weight (286 vs 208 Da) and four rather than three rings. Because binding occurs in an aqueous medium, we suspect that this disparity reflects the higher water solubility of EQ than PQ (0.57 g/L vs 0.054 g/L). That is, hydration would logically encourage the EQ adduct to move away from the peptide backbone while encouraging the PQ adduct to compress toward the peptide backbone, potentially blocking its access to epitopes. Enrichment with Commercial HSA. To assess the effects of various magnetic beads on antibody enrichment and optimize the amount of anti-T3 antibody, commercial HSA was treated with either 3 or 24 μg of anti-T3 bound to one of four types of magnetic beads. In each case, a concurrent control sample was included where HSA was incubated with an equivalent amount of magnetic beads without the antibody attached. Upon analysis by LC-HRMS, we always detected coeluting HSA peptides in the digest, notably T4 (MPCAEDYLSVVLNQLCVLHEK) and a miscleaved T4 ( m T4, RMPCAEDYLSVVLNQLCVLHEK), both of which have physiochemical properties similar to those of T3. We suspect that these other hydrophobic peptides had been nonspecifically bound to the magnetic beads and other surfaces. Figure 4 shows TICs of digests of commercial HSA without anti-T3 enrichment (Figure 3A) and after incubation with anti-T3 (Figure 4B). Note the relative absence of hydrophilic peptides, 403



Article

Figure 4. Total ion chromatograms (TICs) of commercial HSA before and after antibody enrichment. (A) HSA without anti-T3 enrichment. (B) HSA treated with 12 μg of anti-T3 antibody. The insert shows the selected ion chromatograms (SICs) at m/z = 811.7593, 802.0618, and 854.0955, for T3, T4, and the mT4 peptide, respectively.

eluting before T3 in the reversed-phase chromatogram. Following antibody enrichment, the most intense peak was that of T3, while those of coeluting T4 and mT4 were substantially reduced. Table 1 shows results from all experiments employing combinations of magnetic beads with and without the addition of anti-T3. Since the concentration of unadducted T3 was found to be about 100 times greater than the sum of all adducts, the level of the unadducted T3 peptide can be used as

a quantitative surrogate for total antibody capture of T3-related peptides. Likewise, since T4 and mT4 are the most prominent contaminating peptides in anti-T3-enriched digests of HSA, their sum (T4 + mT4) can be used as a measure of nonspecifically bound peptides in the digest. Using the ratio of these two metrics [i.e., T3/(T4 + mT4)] to gauge the efficiency of antibody enrichment, the T1 bead was roughly twice as efficient as the other bead types (Table 1). In absolute terms, the M280 beads produced the highest T3 concentrations, but this was offset by higher concentrations of T4 + mT4 signals in the samples incubated with magnetic beads only. As noted above, we suspect that the presence of T4 and mT4 in enriched T3 peptides reflects nonspecific binding of these hydrophobic peptides on the bead surface. In fact, the highest amount of T4 + mT4 was observed following incubation with M270 beads, which are the only tested beads that had not been blocked (i.e., not stored in a buffer containing bovine serum albumin). On the basis of these results, the T1 magnetic bead was chosen for immobilizing the anti-T3 antibody in subsequent experiments. Since the concentration of captured T3 peptide increased only marginally (from 0.64 μM to 1.0 μM) when the amount of anti-T3 antibody was increased from 3 μg to 24 μg with the T1 bead, an intermediate value of 12 μg of antibody was selected for experiments with plasma samples. Antibody Enrichment of HSA from Plasma Samples. Both antibody-enriched and control samples were analyzed by LC-HRMS. Data-dependent MS/MS spectra were screened to measure T3, T4, and mT4, as summarized in Table 2. Whereas the T3 concentrations were only marginally greater with antibody enrichment (mean ± SD = 0.853 ± 0.191 μM) than without (0.706 ± 0.072 μM), the corresponding concentrations of T4 + mT4 were much smaller with antibody enrichment (0.309 ± 0.093 μM) than without (1.37 ± 0.138 μM). Thus, the mean ratio of T3/(T4 + mT4) with enrichment (2.85 ±

Table 1. Peptide Concentrations Measured in Digests of Commercial HSA after Anti-T3 Enrichment or Following Incubation with Magnetic Beads Only bead type

treatment

M270

beads only enrichment

M280


C1


T1


a

anti-T3 (μg)

T3a (μM)

T4 + mT4a (μM)

T3/(T4 + mT4)

3 24 3 24 3 24 3 24 3 24 3 24 3 24 3 24

0.01 0.01 0.45 0.25 0.01 0.04 0.89 1.5 0.01 0.01 0.48 0.16 0.01 0.01 0.64 1.0

0.26 0.14 0.04 0.02 0.11 0.11 0.06 0.14 0.11 0.11 0.03 0.02 0.08 0.04 0.02 0.02

0.06 0.05 11 13 0.09 0.32 16 11 0.08 0.06 15 6.6 0.08 0.30 30 43

T3 and T4 are the 3rd and 4th largest tryptic peptides of HSA, and is miscleaved T4.

mT4

404



Article

Another modification was identified as cysteinylated Cys34 (m/ z 851.42705). Since the HSA samples had been treated with TCEP to reduce disulfide bonds prior to digestion, these cysteinylated products could represent incomplete reduction of disulfide bonds or disulfide scrambling during sample preparation.18,19 The loss of Lys from the C-terminus of T3 is a form of peptide truncation that may have resulted from sample preparation as well as in vivo proteolytic activity.20−22 The peptide with m/z 863.79285 likely represents transpeptidation of Arg at the C-terminus, which also could have occurred either in vivo23 or during tryptic digestion.24,25 We were unable to identify two putative T3 adducts with m/z 814.71186 and m/z 848.09460, respectively. In addition to the above T3 modifications, several ESIgenerated T3 adducts of single metal ions (Na+, 21.97987 Da; K+, 37.94070 Da; and Fe3+, 52.90428 Da) or a combination of metal ions [2Na+, 43.95616 Da; (Na+ + K+), 59.92251 Da] were detected as were monosodiated adducts of sulfinic acid (m/z 829.74991) and cysteinylated T3 (m/z 858.75447). Since adducts of iron are rarely observed with cysteine thiols, the Fe3+ species (52.90428 Da) could reflect the influence of iron from the magnetic beads. Subsequent experiments with nanoelectrospray ionization did not reveal the presence of these metal-generated adducts.

Table 2. Peptide Concentrations Measured in Digests of HSA Isolated from Three Human Plasma Samples with and without Antibody Enrichmenta plasma sample

antibody enrichment

T3b (μM)

T4 + mT4b (μM)

T3/(T4 + mT4)

1

yes no yes no yes no

0.88 0.79 1.0 0.65 0.65 0.68

0.40 1.5 0.31 1.3 0.22 1.2

2.2 0.52 3.4 0.49 3.0 0.55

2 3 a

HSA was isolated from plasma by treatment with ammonium sulfate, and mercaptalbumin was removed with thiol-affinity resins prior to treatment with 12 μg of anti-T3 antibody bound to T1-type magnetic beads. bT3 and T4 are the 3rd and 4th largest tryptic peptides of HSA, and mT4 is miscleaved T4.

0.616) was 5.52-fold greater than the ratio without enrichment (0.517 ± 0.032). This result confirms the earlier finding with commercial HSA that anti-T3 treatment substantially reduced concentrations of non-T3 peptides in tryptic digests of HSA. Identification of Putative T3 Modifications in Human Plasma. All precursor ions having the requisite b-series fragment ions were used to confirm the presence of modified T3 peptides in HSA isolated from plasma and treated with the anti-T3 antibody. Table 3 summarizes the 10 putative T3related precursor ions that were detected in addition to the unmodified T3 peptide. The observed monoisotopic masses for the triply charged precursor ions were estimated from SICs as mean (SD) values in the three plasma samples. All but two of the modified T3 peptides were unambiguously identified within 3 ppm. Three oxidation products of Cys34 were detected as expected,10 namely, a sulfinamide formed by an intramolecular reaction (+O, −2H, m/z 816.41911), the sulfinic acid (+2O, m/z 822.42263), and the sulfonic acid (+3O, m/z 827.75427).

■

DISCUSSION We used a 3D model of the T3 peptide in aqueous solution (Figure 1) to design an immunogen for producing a polyclonal antibody that would capture the tryptic T3 peptide of HSA and its modifications at Cys34. After preparing the antibody, we attached it to magnetic beads and used it to successfully enrich Cys34 adducts (and related modifications of the T3 peptide) in tryptic digests of HSA and released the modified peptides for subsequent LC-HRMS. To our knowledge, this is the first demonstration that a semitargeting antibody can capture a

Table 3. Putative T3 Adducts Detected in Human Plasma Samples after Digestion and Enrichment with the Anti-T3 Antibody adduct concentration (nM) observed m/za 769.06278 (9.92 × 10−04) 805.75773 (1.10 × 10−04) 811.76140 (3.53 × 10−04) 814.71187 (2.01 × 10−04) 816.41934 (2.76 × 10−04) 822.42293 (2.24 × 10−04) 827.75459 (4.30 × 10−04) 848.09444 (4.53 × 10−04) 851.42715 (8.33 × 10−05) 863.79285 (1.95 × 10−04)

theoretical m/zb

mass accuracy (ppm)c

769.06102

2.28

805.75583

2.35

811.75933

2.55

N/A

N/A

816.41911

0.282

822.42263

0.361

827.75427

0.383

N/A

N/A

851.42739

−0.286

863.79306

−0.247

ret. time (min)d

adduct mass (Da) −128.09586 (2.58 × 10−03) −18.01101 (7.37 × 10−04 0.00000 8.85142 (7.11 × 10−04) 13.97383 (7.04 × 10−04) 31.98459 (6.46 × 10−04) 47.97957 (1.22 × 10−03) 108.99912 (1.43 × 10−03) 118.99725 (9.99 × 10−04) 156.09435 (7.95 × 10−04)

sample 1 sample 2 sample 3

putative adduct

38.48 (0.010)

3.72

3.89

2.19

−Lys (from C-terminus)

36.37 (0.026)

2.84

2.73

2.43

−H2O

36.23 (0.010)

874

963

651

unadducted T3

36.82 (0.036)

6.18

3.88

2.91

unknown

34.15 (0.046)

5.32

7.29

4.25

+O, −H2

34.18 (0.017)

11.6

16.6

9.58

+2O

34.38 (0.017)

2.84

3.93

1.85

+3O

38.20 (0.023)

6.13

2.53

0.97

unknown

33.19 (0.025)

22.1

8.34

4.84

+C3H5NO2S (cysteinylation)

34.79 (0.015)

16.4

28.1

12.5

Arg transpeptidation (from Cterminus)

a

Mean (SD) monoisotopic mass-to-charge ratio (m/z) of 3+ precursor ion for three plasma samples. bTheoretical monoisotopic mass-to-charge ratio (m/z) of the 3+ precursor ion calculated from simulation of the isotopic distribution from the empirical formula. cMass accuracy of the 3+ precursor ion calculated from simulation of the isotopic distributions from the empirical formula. dMean (SD) retention time for three plasma samples. 405



Article

ionization; HSA, human serum albumin; HPLC, high performance liquid chromatography; HRMS, high resolution MS; IAAiT3, carboxyamidomethylated Cys34 adduct of iT3; iT3, isotopically modified T3; KLH, keyhole limpet hemocyanin; LC-HRMS, liquid chromatography-high-resolution mass spectrometry; MS, mass spectrometry; NQ, naphthoquinone; PQ, phenanthrene quinone; RT, retention times; SIC, selected-ion chromatogram; TBS, Tris-buffered saline; TCEP, tris(2carboxyethyl)-phosphine; TMB, 3,3′,5,5′-tetramethylbenzidine; TIC, total-ion chromatogram

diverse class of protein modifications for subsequent isolation and characterization. The capture efficiency of the antibody was tested with a series of synthetic T3-quinone adducts having between one and four rings. Although all of these quinone adducts were captured with the anti-T3 antibody, capture efficiency diminished with the increasing size of the adducting species (Figure 3). Reduced binding affinities with bulky Cys34 adducts is unlikely to cause problems because the thiol resides inside a 10 Å crevice in the HSA molecule,26,27 and previous work has shown that most Cys34 adducts had masses below 100 Da.8 Using digests of commercial HSA and HSA isolated from fresh plasma, we showed that treatment with anti-T3 increased the concentration of T3 while removing unwanted peptides (Figure 4 and Tables 1 and 2). Antibody enrichment greatly reduced but did not eliminate background effects from other hydrophobic HSA peptides (T4 and mT4) that coelute with T3 and are probably nonspecifically bound to magnetic beads. Nonetheless, antibody treatment reduced the amounts of T4 and mT4 in digests of fresh HSA by approximately 6-fold compared to that of control samples, thereby improving the quality of the matrix for mass spectrometry. Analysis of antibody-enriched HSA digests from human plasma samples revealed 10 modified T3 peptides of which 8 were identified from accurate masses. Identified peptides included Cys34 oxidation products, modifications representing losses of water and Lys, cysteinylation, and transpeptidation of Arg. Modifications of the T3 peptide at the C-terminus (loss of Lys and transpeptidation of Arg) and cysteinylation of Cys34 could have resulted from tryptic digestion and subsequent processing. Further work is needed to optimize anti-T3 enrichment and to characterize populations of adducts in a large number of human plasma samples.

■

■

(1) Miller, E. C., and Miller, J. A. (1966) Mechanisms of chemical carcinogenesis: nature of proximate carcinogens and interactions with macromolecules. Pharmacol. Rev. 18, 805−838. (2) Brodie, B. B., Reid, W. D., Cho, A. K., Sipes, G., Krishna, G., and Gillette, J. R. (1971) Possible mechanism of liver necrosis caused by aromatic organic compounds. Proc. Natl. Acad. Sci. U.S.A. 68, 160−164. (3) Liebler, D. C. (2008) Protein damage by reactive electrophiles: targets and consequences. Chem. Res. Toxicol. 21, 117−128. (4) Nair, U., Bartsch, H., and Nair, J. (2007) Lipid peroxidationinduced DNA damage in cancer-prone inflammatory diseases: A review of published adduct types and levels in humans. Free Radical Biol. Med. 43, 1109−1120. (5) Xue, W., and Warshawsky, D. (2005) Metabolic activation of polycyclic and heterocyclic aromatic hydrocarbons and DNA damage: A review. Toxicol. Appl. Pharmacol. 206, 73−93. (6) Aldini, G., Vistoli, G., Regazzoni, L., Gamberoni, L., Facino, R. M., Yamaguchi, S., Uchida, K., and Carini, M. (2008) Albumin is the main nucleophilic target of human plasma: a protective role against pro-atherogenic electrophilic reactive carbonyl species? Chem. Res. Toxicol. 21, 824−835. (7) Rubino, F. M., Pitton, M., Di Fabio, D., and Colombi, A. (2009) Toward an “omic” physiopathology of reactive chemicals: thirty years of mass spectrometric study of the protein adducts with endogenous and xenobiotic compounds. Mass Spectrom. Rev. 28, 725−784. (8) Li, H., Grigoryan, H., Funk, W. E., Lu, S. S., Rose, S., Williams, E. R., and Rappaport, S. M. (2011) Profiling Cys34 adducts of human serum albumin by fixed-step selected reaction monitoring. Mol. Cell. Proteomics 10, M110 004606. (9) Rappaport, S. M., Li, H., Grigoryan, H., Funk, W. E., and Williams, E. R. (2012) Adductomics: Characterizing exposures to reactive electrophiles. Toxicol. Lett. 213, 83−90. (10) Grigoryan, H., Li, H., Iavarone, A. T., Williams, E. R., and Rappaport, S. M. (2012) Cys34 adducts of reactive oxygen species in human serum albumin. Chem. Res. Toxicol. 25, 1633−1642. (11) Anderson, N. L., Anderson, N. G., Haines, L. R., Hardie, D. B., Olafson, R. W., and Pearson, T. W. (2004) Mass spectrometric quantitation of peptides and proteins using Stable Isotope Standards and Capture by Anti-Peptide Antibodies (SISCAPA). J. Proteome Res. 3, 235−244. (12) Anderson, N. L., Jackson, A., Smith, D., Hardie, D., Borchers, C., and Pearson, T. W. (2009) SISCAPA peptide enrichment on magnetic beads using an in-line bead trap device. Mol. Cell. Proteomics 8, 995− 1005. (13) Whiteaker, J. R., Zhao, L., Zhang, H. Y., Feng, L. C., Piening, B. D., Anderson, L., and Paulovich, A. G. (2007) Antibody-based enrichment of peptides on magnetic beads for mass-spectrometrybased quantification of serum biomarkers. Anal. Biochem. 362, 44−54. (14) Whiteaker, J. R., Zhao, L., Anderson, L., and Paulovich, A. G. (2010) An automated and multiplexed method for high throughput peptide immunoaffinity enrichment and multiple reaction monitoring mass spectrometry-based quantification of protein biomarkers. Mol. Cell. Proteomics 9, 184−196. (15) Zhang, Y. (2008) I-TASSER server for protein 3D structure prediction. BMC Bioinformatics 9, 40.

ASSOCIATED CONTENT

S Supporting Information *

Summary of BLAST results for searches with input sequences ALVLIAFAQYLQQ (Table S1) and PFEDHVK (Table S2) against the human reference protein database. This material is available free of charge via the Internet at http://pubs.acs.org.

■

REFERENCES

AUTHOR INFORMATION

Corresponding Author

*School of Public Health, University of California, Berkeley, CA 94720-7356, USA. Tel: 1-510-642-4355. Fax: 1-510-6420427. E-mail: [email protected]. Funding

This work was supported by a contract from the American Chemistry Council Long-Range Research Initiative. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS We are grateful to Jacques Riby for advice regarding the selection of an immunogen and Luca Regazzoni for assistance with mass spectrometry and bioinformatics.

■

ABBREVIATIONS 3D, three-dimensional; 4-OH-E2, 4-hydroxyestradiol; BQ, benzoquinone; CID, collision-induced dissociation; Da, Daltons; DMSO, dimethylsulfoxide; ELISA, enzyme-linked immunosorbent assay; EQ, estrogen quinone; ESI, electrospray 406



Article

(16) Maupetit, J., Derreumaux, P., and Tuffery, P. (2009) PEPFOLD: an online resource for de novo peptide structure prediction. Nucleic Acids Res. 37, W498−503. (17) Humphrey, W., Dalke, A., and Schulten, K. (1996) VMD: visual molecular dynamics. J. Mol. Graphics Modell. 14 (33−38), 27−38. (18) Wang, Z., Rejtar, T., Zhou, Z. S., and Karger, B. L. (2010) Desulfurization of cysteine-containing peptides resulting from sample preparation for protein characterization by mass spectrometry. Rapid Commun. Mass Spectrom. 24, 267−275. (19) Liu, H., Gaza-Bulseco, G., Chumsae, C., and Newby-Kew, A. (2007) Characterization of lower molecular weight artifact bands of recombinant monoclonal IgG1 antibodies on non-reducing SDSPAGE. Biotechnol. Lett. 29, 1611−1622. (20) Alves, P., Arnold, R. J., Clemmer, D. E., Li, Y., Reilly, J. P., Sheng, Q., Tang, H., Xun, Z., Zeng, R., and Radivojac, P. (2008) Fast and accurate identification of semi-tryptic peptides in shotgun proteomics. Bioinformatics 24, 102−109. (21) Olsen, J. V., Ong, S. E., and Mann, M. (2004) Trypsin cleaves exclusively C-terminal to arginine and lysine residues. Mol. Cell. Proteomics 3, 608−614. (22) Strader, M. B., Tabb, D. L., Hervey, W. J., Pan, C., and Hurst, G. B. (2006) Efficient and specific trypsin digestion of microgram to nanogram quantities of proteins in organic-aqueous solvent systems. Anal. Chem. 78, 125−134. (23) Berkers, C. R., de Jong, A., Ovaa, H., and Rodenko, B. (2009) Transpeptidation and reverse proteolysis and their consequences for immunity. Int. J. Biochem. Cell Biol. 41, 66−71. (24) Fodor, S., and Zhang, Z. (2006) Rearrangement of terminal amino acid residues in peptides by protease-catalyzed intramolecular transpeptidation. Anal. Biochem. 356, 282−290. (25) Schaefer, H., Chamrad, D. C., Marcus, K., Reidegeld, K. A., Bluggel, M., and Meyer, H. E. (2005) Tryptic transpeptidation products observed in proteome analysis by liquid chromatographytandem mass spectrometry. Proteomics 5, 846−852. (26) He, X. M., and Carter, D. C. (1992) Atomic structure and chemistry of human serum albumin. Nature 358, 209−215. (27) Christodoulou, J., Sadler, P. J., and Tucker, A. (1995) 1H NMR of albumin in human blood plasma: drug binding and redox reactions at Cys34. FEBS Lett. 376, 1−5.

407


Antibody Enrichment and Mass Spectrometry of ... - ACS Publications

Recommend Documents