MS of Peptides

Journal of Proteome Research , 2003, 2 (3), pp 265–272 ... minutes at neutral pH and the PIC-labeled peptides undergo informative MS/MS fragmentatio...
0 downloads 0 Views 127KB Size
Download PDF
Quantitative Analysis of Modified Proteins by LC-MS/MS of Peptides Labeled with Phenyl Isocyanate Daniel E. Mason and Daniel C. Liebler* Southwest Environmental Health Sciences Center, College of Pharmacy, The University of Arizona, Tucson, Arizona 85721-0207 Received November 16, 2002

Stable isotope tagging methods have enabled relative quantitation of proteins between samples in LC-MS/MS analyses. However, most such methods are not applicable to the differential quantitation of modified proteins because the isotope tagging reagents only react with certain peptides or because the reagents incorporate a mass increment that is too small to allow reliable quantitation on low resolution ion trap MS instruments. Here, we describe the use of d0- and d5-phenyl isocyanate (PIC) as N-terminal reactive tags for essentially all peptides in proteolytic digests. PIC reacts quantitatively with peptide N-terminal amines within minutes at neutral pH and the PIC-labeled peptides undergo informative MS/MS fragmentation. Ratios of d0- and d5-PIC-labeled derivatives of several model peptides were linear across a 10 000-fold range of peptide concentration ratios, thus indicating a wide dynamic range for quantitation. Application of PIC labeling enabled relative quantitation of several styrene oxide adducts of human hemoglobin in LC-MS/MS analyses. PIC labeling offers a versatile means of quantifying changes in modified or variant protein forms in paired samples. Keywords: adducts • isotope tag • phenyl isocyanate • quantitation • tandem mass spectrometry

Introduction Covalent modifications are essential determinants of the location, function, and turnover of proteins.1 Diverse endogenous modifications range in size from methylation and phosphorylation to ubiquitination and complex glycosylation. Reactive intermediates generated from toxic chemicals or endogenous oxidative damage also modify proteins and these electrophile adducts play a causative role in tissue damage and cellular toxicity.2-4 LC-MS/MS1 offers a powerful approach to characterize protein modifications.5 MS/MS spectra of peptides not only allow protein identification with the aid of database searching algorithms, but also sequence specific mapping of modifications.6 We have developed the SALSA algorithm to facilitate detection of modified and variant protein forms by detecting modification-specific features or sequence motifrelated ion series in MS/MS data.7-9 An essential element of understanding protein modifications is the ability to quantify modifications under different experimental conditions or under different states of a biological system. For example, the toxic effects of xenobiotic adducts may depend on the rates of their formation and elimination from cells and tissues.3,10 Quantitation of modified proteins is a difficult task due to the large number of different proteins in real biological systems and the probability that, in many cases * To whom correspondence should be addressed at College of Pharmacy, P.O. Box 210207, The University of Arizona, Tucson, AZ 85721-0207. Phone: 520 626 4488; FAX: 520 626 6944. E-mail: [email protected]. 10.1021/pr0255856 CCC: $25.00

 2003 American Chemical Society

protein modification is nonstoichiometric (i.e., only a fraction of any particular protein may exist in any particular modified form).3 Recent work by several groups has applied stable isotope tags to differentially label proteins or peptides. This enables relative quantitation of proteins between two samples by measuring the relative abundance of labeled peptides in LCMS/MS analyses. The advantage of this LC-MS/MS approach is that peptides are unambiguously identified from MS/MS data and quantified in the same LC-MS/MS analysis. Incorporation of isotopic tags can be accomplished metabolically in cultured cells supplemented with labeled nutrients11,12 or by enzymatic digestion in H218O water.13 Metabolic labeling is an effective tool for certain in vitro systems, but some labels are costly. Technical difficulty, cost of highly pure H218O water and the efficiency of enzyme-assisted labeling limits the use of 18O labeling. Covalent modification of proteins or peptides with an isotope-coded label14-20 has become the most widely used approach. Perhaps the best-known isotope tags are the ICAT reagents described by Gygi et al.14 The ICAT reagents react with protein thiols, which are present in many proteins. Intact proteins are labeled on cysteine thiols and then subjected to proteolytic digestion. The incorporation of a biotin affinity tag in the ICAT reagent permits isolation of the ICAT-labeled cysteinyl peptides, thus simplifying subsequent LC-MS/MS analysis. However, the ICAT reagents are generally unsuitable for quantifying protein modifications because modifications may not appear on cysteine-containing peptides and because Journal of Proteome Research 2003, 2, 265-272

265

Published on Web 01/22/2003

research articles many electrophiles preferentially modify cysteines, thus precluding their reaction with the ICAT reagents. Other isotopic tagging approaches target functional groups post-digestion to prevent interferences by the label with proteolysis and to exploit newly formed amine15-17 and carboxyl groups18on each peptide. These approaches allow more extensive labeling of peptides and approaches that target carboxyls or N-terminal amines potentially enable quantitation of any peptide of a protein, including modified peptides. Nevertheless, these approaches have drawbacks that limit their application to protein adduct quantitation. The carboxyl methyl esterification strategy described by Goodlett et al.18 and the amine acetylation described by Ji et al.16 incorporate labels that differ by only 3 amu, thus complicating resolution of the M+3 isotope signal from the all-12C signal for the deuterated peptide.21 Doubly and triply charged peptide ions labeled with the d3 and d0 reagents differ in m/z by 1.5 and 1, respectively, which complicates quantitation on commonly used ion trap MS instruments, which typically are operated at unit resolution for LC-MS/MS work. Although a greater labeling mass increment can be achieved with labels containing more deuterium atoms, peptides labeled with these larger deuterated tags are often resolved chromatographically from the nondeuterated forms. This phenomenon complicates the use of isotope ratios for relative quantitation by LC-MS.21,22 Another post-digestion tagging method was recently introduced by Cagney and Emili, who used O-methylisourea to convert peptide lysine residues to homoarginine residues, thus introducing a 42-dalton mass increment.23 Paired analyses of derivatized and underivatized samples allows quantitative comparisons, but the approach is applicable only to the quantitation of lysine-containing peptides. Here, we describe a stable isotope labeling method to quantify peptide adducts by labeling peptides at the N-terminus with PIC. The deuterated (d5) reagent provides a sufficiently high mass increment to allow relative quantitation by low resolution ion trap MS instruments and is available commercially at modest cost. PIC quantitatively labels the N-termini of peptides under mild conditions and permits relative quantitation of any peptide in paired samples. We also describe application of PIC for differential quantitation of hemoglobin adducts of styrene oxide.

Experimental Procedures Peptides and Reagents. PIC (98%) was purchased from Aldrich and used without further purification. Deuterated (d5-) PIC (98%) was purchased from Isotech, Inc. TpepC (sequence: AVAGCAGAR) was purchased from Sigma-Genosys. Atrial natriuretic factor, frog (Frog24) (sequence: SSDCFGSRIDRIGAQSGMGCGRRF), human fibrinogen peptide 6A (sequence: ARPAK), and anaphylatoxin C3a peptide (sequence: ASHLGLAR) were purchased from Bachem Bioscience, Inc. Laminin B1 chain fragment (925-933; sequence CDPGYIGSR) was obtained from Synpep Corp. BSA was purchased from Sigma Chemical Co. PIC Labeling Protocol. Peptides were dissolved to 0.01-0.1 mg mL-1 in 10 mM HEPES, pH 8.0 in a final volume of 200 µL. Cysteine residues were reduced and alkylated with 5 mM TCEP and 5 mM iodoacetamide and incubation in the dark for 15 min at room temperature. PIC (6 µL of 0.1 M in acetonitrile) was added to the solution and incubated for 15 min at 37 °C, after which 1 µL of concentrated formic acid was added to 266

Journal of Proteome Research • Vol. 2, No. 3, 2003

Mason and Liebler

terminate the reaction. The solution was diluted 1:10 in 0.1% formic acid and analyzed by LC-MS. Differential Labeling of TpepC. Two samples of TpepC in separate tubes were diluted to 0.0775 mg mL-1 with 10 mM HEPES, pH 8.0 and a final volume of 200 µL. Iodoacetamide was added to a final concentration of 5 mM and the solutions were incubated in the dark for 15 min at room temperature. Unlabeled (d0-) or d5-PIC diluted in acetonitrile to 0.1 M was added to each sample to a final concentration of 3 mM. After incubation at 37 °C for 15 min, both samples then were acidified with 1 µL of concentrated formic acid, combined at ratios from 100:1 to 1:100, and diluted with 0.1% formic acid to a final volume of 100 µL prior to LC-MS analysis with a “fast” gradient (see below). For labeling of TpepC in a BSA mixture, two sets of 13 samples were prepared containing equal amounts of BSA (69 µg mL-1), 5 mM TCEP, and 50 mM HEPES pH 8.0. TpepC was added to each of the sample tubes such that when the sets were later differentially labeled with either d0-PIC or d5-PIC and combined at 1:1, the ratio of TpepC would range from 0.1 to 10. These samples also contained Frog24, peptide 6A, C3a, and B1 chain fragment peptides, which were prepared in two standard solutions in water. The concentrations of the peptides in solution 1 were 48 pmol µL-1 (frog24), 24 pmol µL-1 (peptide 6A), 120 pmol µL-1 (C3a), and 84 pmol µL-1 (B1 chain fragment). The concentration of all peptides in solution 2 was 12 pmol µL-1. Samples to be labeled with d0-PIC were spiked with 5 µL of the Solution 1 standard peptides mix and all others were spiked with the same volume of Solution 2. Iodoacetamide was added to a concentration of 10 mM and samples were incubated at room temperature in the dark for 15 min. Trypsin was then added to each tube individually at a trypsin:protein ratio of 1:25 (w/w) and allowed to digest at 37 °C for 20 h. After digestion, samples were labeled with either d0-PIC or d5-PIC in the same manner as the TpepC differential labeling experiment and combined in equal amounts. Analysis was performed by LC-MS with inline desalting (see below). LC-MS/MS Analyses. LC-MS/MS analyses were performed on a ThermoFinnigan LCQ Deca ion trap MS instrument equipped with a ThermoFinnigan Surveyor HPLC pump and microelectrospray source and operated with ThermoFinnigan Xcalibur version 1.2 system control and data analysis software. Analysis of samples was performed with an acetonitrile gradient and a Monitor C18 (Column Engineering) packed tip with 100 µm ID, 360 µm OD, and 5-15 µm tip opening. The flow from the HPLC pump was split to achieve 500 nL to 1 µL flow rate from the packed tip. Generally, two gradients were used, “fast” and “normal”, depending on the complexity of the sample being analyzed. The fast gradient began at 98:2:0.1 water: acetonitrile:formic acid, and at 2 min, a 10 min linear gradient increased acetonitrile to 80%. Total analysis time for the fast method was 30 min. The normal gradient was held at similar starting conditions for 5 min. Linear gradients increased acetonitrile to 60% by 45 min then 80% acetonitrile by 47 min. The entire analysis time for the normal method was 65 min. For analyses of more complex mixtures containing BSA peptides, LC-MS/MS analyses were done with inline desalting prior to injection on the analytical column. Captrap peptide trapping cartridges and the holder were obtained from Michrom BioResources, Inc. A KD Scientific Model 100 syringe pump was used to load the sample onto the captrap using 0.1% formic acid in water at a flow rate of 5 µL min-1. Loading and rinsing of the sample was completed in 15 min. After loading, the LCQ

Quantitative Analysis of Modified Proteins

valve was switched so that the Captrap was in line with the HPLC and the packed tip. A gradient similar to the one described in the previous section eluted peptides from the trap for separation on the packed tip without interference from salts. Relative Quantitation of Differentially Labeled Peptides. The ratio of differentially labeled peptides was calculated by integrating the area under the curve for the monoisotopic mass of the peptide with the d0- and d5-PIC. A mass window of (0.3 amu was allowed, and the ICIS peak detection algorithm feature of Xcalibur was used for automated detection of peaks. For lower abundance peaks, or for peaks severely interrupted by MS/MS scans, ICIS failed, and the peak was integrated manually. Unless otherwise noted, ratios are reported as area under the curve for d0-PIC peptide divided by area under the curve for d5-PIC peptide. Resolution of d0- and d5-PIC-labeled isotopomers was calculated as the difference in peak elution time divided by the peak width at half-maximum and was based on the assumption of Gaussian peak shapes with a 7 point smoothing algorithm in Xcalibur. Preparation of Hemoglobin Adducts of Styrene Oxide. Human hemoglobin was obtained from Sigma and diluted in isotonic saline to 4 mg mL-1 in 50 mL final volume. Concentrated styrene oxide or styrene oxide diluted in methanol was added for final concentrations of 0.0, 0.04, 0.4, 4, 20, and 40 mM and the solutions were incubated with shaking at 37 °C for 6 h. 2-Propanol (35 mL) with 50 mM HCl were added and the samples were centrifuged at 2000 rpm to precipitate the heme. Protein was precipitated from the supernatant by addition of 50 mL of ethyl acetate and brief centrifugation. Protein pellets were washed with 10 mL of ethyl acetate, vacuum filtered, and rinsed again with hexane. The air-dried protein residue was stored at -20 °C until analyzed. Analysis and Quantitation of Styrene Oxide Adducts. Protein was weighed out from each treatment sample and diluted in 6 M guanidine HCl, 50 mM HEPES, pH 8.0 to 10 mg mL-1. The protein content of each sample was measured using the BCA assay (Pierce Chemical Co.) and subsequent dilutions were adjusted to correct for variations that may have arisen from weighing error. Samples were diluted in 2 M urea, 50 mM HEPES pH 8.0 to 5 mg mL-1 protein and 5 mM TCEP prior to addition of iodoacetamide to a concentration of 10 mM and incubation in the dark at room temperature for 15 min. Samples then were diluted with the same buffer to 0.1 mg mL-1 protein and digested with trypsin as previously described. Seven separate dilutions of the 40 mM treatment were prepared and digested. These samples were labeled with d5-PIC, whereas all others were labeled with d0-PIC as previously described. Each dose was then combined 1:1 with the d5-PIC labeled 40 mM samples. Samples were analyzed by infusing 10 µL of sample into a Monitor C18 packed tip at a flow rate of 1 µL min-1. The tip was then rinsed with 98:2:0.1 water:acetonitrile:formic acid for 10 min prior to LC-MS analysis with the normal gradient as described above. Hemoglobin adducts of styrene oxide have previously been described using this system by Badghisi et al. and were located in the chromatogram/datafile by searching for the protonated monoisotopic mass and verifying the previously reported sequence from the accompanying MS/MS spectra.9 Relative quantitation of peptides was performed by integrating peak areas as described above.

Results Labeling of TpepC. Phenyl isocyanate labels the N-termini of peptides by reaction to form a phenylcarbamoyl derivative

research articles

Figure 1. Reaction of PIC with a peptide N-terminus. Hydrogen atoms of the phenyl ring are replaced with deuterium atoms for d5-PIC.

(Figure 1). The initial objective of this work was to establish conditions for N-terminal labeling of peptides with PIC. The labeling technique was optimized using the peptide TpepC (sequence AVAGCAGAR). d5-PIC was reacted with TpepC to determine the extent of reaction with the peptide and purity of d5-PIC as analyzed by LC-MS (Figure 2). A single peptide derived product was obtained at retention time 27.34 min and no residual TpepC was observed (Figure 2A). Other peaks in the chromatogram were identified as PIC or side products of PIC and were present in incubations of PIC alone (data not shown). A typical full scan spectrum of d5-PIC labeled TpepC is shown in Figure 2B with the region around the singly charged ion, m/z 956.4 is amplified 5-fold. A minor d4 isotope peak is observed at m/z 955.5 and is calculated to be present at less than 2% of the ion current for the d5-PIC derivatives. Differential Labeling of TpepC. The next objective was to determine the linear dynamic range for peptide quantitation with d0- and d5-PIC labels and the variability of the results obtained. TpepC was differentially labeled with d5- or d0-PIC, combined in d0:d5 ratios from 0.01 to 100, and then analyzed by LC-MS. This experiment was repeated three times to determine interexperiment variability and plots of expected versus observed ratios for the entire range (0.01 to 100.0) (Figure 3A) and a limited range (0.1 to 10.0) (Figure 3B). R2 values greater than 0.99 were calculated for both plots and the slope of the trend lines were 1.43 and 1.57 for the limited and entire ranges, respectively. The slope of the trend line was slightly greater than 1.0 apparently due to carryover between injections. This can be remedied by flushing the system between injections (data not shown). A set of samples from one differentially labeled TpepC experiment was analyzed three different times over several days to assess any variation imposed by length of time prior to LCMS analysis. Plots of the entire range of ratios and the more limited range were nearly identical to the previous experiments with similar linearity, slopes, and standard deviations (data not shown). Differential Labeling of TpepC in a BSA Digest. TpepC was spiked into a BSA solution along with standard peptides prior to digestion and differential PIC labeling in order to assess this labeling technique in a complex mixture. Standard peptides, peptide 6A, frog24, B1 chain fragment, and C3a, were added to each sample tube such that when the samples were combined at 1:1 after digestion and labeling the ratio of these peptides in each tube would be 2, 4, 7, and 10, d0:d5, respectively. In essence, these peptides establish a standard curve to verify labeling in general and evaluate quantitation of peptides present at different abundances. The frog24 peptide has tryptic cleavage sites and requires digestion for detection. BSA peptides should be present at a ratio of 1:1, and three were chosen at random for use in the plot of standard peptides, AEFVEVTK, LCVLHEK, and IETMR. A plot of the average and standard deviation of the observed ratios of each standard peptide and the three selected BSA peptides versus their expected ratio is shown in Figure 4. A Journal of Proteome Research • Vol. 2, No. 3, 2003 267

research articles

Mason and Liebler

Figure 2. (A) Base peak chromatogram of d5-PIC labeled TpepC. (B) Averaged full scan spectrum corresponding to d5-PIC labeled TpepC.

linear relationship with an R2 value of 0.9932 and slope slightly greater than 1.0 exists between the standard peptides. With the exception of the C3a peptide, the standard deviation of the standard peptides is small. Digestion with trypsin is potentially a problematic step in this method because the samples are digested separately and differences in digestion would result in variable and unreliable relative quantitation. The three BSA peptides used for the standard peptides curve had average ratios of 0.99 ( 0.12, 1.02 ( 0.09, and 0.97 ( 0.12, and the frog24 peptides in each sample with an expected ratio of 4 had observed ratios of 4.51 ( 0.55 and 4.34 ( 0.67. The relative ratio of TpepC was varied between samples from 0.1 to 10.0, and the graph of observed versus expected ratios was linear 268

Journal of Proteome Research • Vol. 2, No. 3, 2003

with a slope slightly greater than one (Figure 4). In this experiment, a lower concentration of TpepC was used than in the analyses depicted in Figure 3, which reduced carryover between analyses. The plot of TpepC and its trend line nearly overlap that of the standard peptides. Resolution of PIC-Labeled Isotopomers by Reverse Phase LC. An common feature of peptides labeled with deuterated isotope tags is that the isotopomers are frequently resolved in typical reverse phase LC analyses.21,22 We examined the resolution of 13 d0-/d5-PIC-labeled BSA peptide pairs in the analysis described above (Figure 5). Resolution ranged from approximately-0.2 to 0.6 and was independent of peptide elution time (Figure 5A) and peptide mass (Figure 5B). These values are in

Quantitative Analysis of Modified Proteins

Figure 3. Linearity of TpepC relative quantitation by PIC labeling and LC-MS/MS. Samples containing different amounts of TpepC were labeled with d0- or d5-PIC and then combined and analyzed by LC-MS/MS. (A) A plot of observed versus expected d0/d5PIC labeled TpepC ratios ((SD) for three experiments for the entire range of relative concentrations (0.01 to 100). (B) A plot of observed versus expected ratios for a limited range of TpepC concentration ratios (0.1 to 10) from A.

Figure 4. Linearity of TpepC relative quantitation in a BSA digest by PIC labeling and LC-MS/MS. A standard set of peptides, peptide 6A, frog24, -B1 chain fragment, and C3a, were spiked into each sample at ratios of 2, 4, 7, and 10, respectively. The average and standard deviation for each standard peptide for all samples is plotted against its expected ratio (square markers with dashed trend line, R2 ) 0.9932; slope ) 1.064). TpepC concentration ratio differed across samples from 0.1 to 10 (round marker with solid trend line, R2 ) 0.9803; slope ) 1.1041).

the range reported previously for deuterated versus nondeuterated isotope tags.22 Differential Quantitation of Styrene Oxide Adducts of Hemoglobin. Previous work in our laboratory had mapped styrene oxide adducts to individual residues in human hemoglobin, including cysteine-93 of the beta chain and multiple histidines of both the alpha and beta chain, as well as the N-terminal valine residues of both chains.9 We used this model

research articles

Figure 5. Resolution of the d0- and d5-PIC-labeled isotopomers of 13 BSA peptides by reverse phase LC. Data were taken from the analyses described in Figure 4. Panel A shows the dependence of resolution on peptide elution time. Panel B shows the dependence of resolution on peptide mass.

system to test the utility of this method to quantify relationships between adduct levels and styrene oxide concentration. Hemoglobin was treated with increasing doses of styrene oxide (0-40 mM) in isotonic saline for 6 h, then the protein was purified and digested with trypsin with some modification of previous protocols.9 Samples of hemoglobin treated with styrene oxide at different concentrations were digested and the peptides then were labeled with d0-PIC. These were mixed with an equal amount of a tryptic digest of a hemoglobin sample treated with 40 mM styrene oxide and then labeled with d5PIC. Thus, peptide adducts at each styrene oxide concentration were analyzed as a ratio of d0/d5 isotopomers, which represent the ratio of adduct at a specific exposure concentration to that at the 40 mM concentration. A concentration/adduct ratio curve was plotted for peptide adducts identified in the styrene oxide-treated hemoglobin and is provided as a semilog plot in Figure 6. Eight unique adducts were identified and verified by tandem-MS and annotated spectra of each are available as Supporting Information. Adduction at histidine of the peptides GHGK and AHGK each produced two chromatographically resolved adducts with identical tandem-MS spectra which are designated 1 and 2, based on order of elution. This may be attributed to chromatographic resolution of diastereomeric adducts of these short peptides formed by their reaction with racemic styrene oxide. Although these were quantified separately, the concentration dependence of adduction for the resolved forms was similar. The styrene oxide adduct of LHVDENFR is assumed to be present at the histidine, but tandem-MS data can only localize the adduct to the sequence HVD. Each of the peptide adducts demonstrates a concentration-dependent formation, as indicated by the increase in the relative ratio with increasing concentration. Because peptide adducts were quantified in relation to the 40 Journal of Proteome Research • Vol. 2, No. 3, 2003 269

research articles

Mason and Liebler

15 min as opposed to hours for other methods14-17,23 and yields a single peptide-derived product.

Figure 6. Concentration dependence of styrene oxide Hb adduct formation as measured by PIC labeling and LC-MS/MS. Hb was incubated for 6 h with styrene oxide at concentrations from 0 to 40 mM and the proteins then were digested and analyzed by LCMS/MS as described under “Experimental Procedures”. Ratios of adducts are for each styrene oxide concentration (d0) versus the 40 mM styrene oxide concentration (d5). The legend lists the sequence of each adducted peptide and * following an amino acid in the peptide sequence indicates the location of the adduct.

mM concentration, the ratios generally approach unity at 40 mM. Values for ratios of the peptides LLGNVLVC*VLAHHFGK and AH*GK exceed unity at the 20 mM exposure concentration. The reason for this is not clear, but may reflect differing solubilities of the adducted Hb proteins prior to workup.

Discussion Stable isotopic labeling of peptides enables differential quantitation of proteins and peptides from complex mixtures by MS. We have applied PIC as a stable isotope label for N-terminal modification of peptides and validated its application to differential quantitation of peptide adducts. The d5-PIC reagent is commercially available in high purity and at relatively low cost. The active portion of the label is an isocyanate and has reactivity toward amino,24 sulfhydryl,25 carboxyl,24 and hydroxyl groups26 depending on the solution pH. Reactions with carboxyls and hydroxyls proceed only at low pH. Although isocyanates are more reactive toward sulfhydryls than toward other functional groups, cysteines are typically reduced and alkylated as part of our digestion scheme, thus rendering them unavailable for reaction. Stark observed that reactivity toward amines was a function of pKa, and reactions with -amino groups such as lysine proceeded 100 times slower than R-amines at near neutral pH,24 thus indicating that the solution pH could be controlled to selectively direct modification of R-amines. Given the specificity of PIC- for N-terminal peptide labeling under these reaction conditions, PIC would label essentially all peptides in a mixture except for those that bore stable N-terminal modifications in proteins prior to digestion. The reaction conditions for PIC labeling described here are well-suited for differentially labeling protein digests. PIC labeling was performed in the same buffer and at the same pH as trypsin digestion, thus reducing sample handling, and excess PIC is quickly destroyed by addition of acid prior to LC-MS analysis. The high reactivity of isocyanates toward amines is demonstrated by our initial experiment where d5-PIC was incubated with TpepC and no residual unmodified peptide was detected after a 15-min incubation, thus indicating quantitative labeling of the peptide. The labeling reaction is complete by 270

Journal of Proteome Research • Vol. 2, No. 3, 2003

Our intention was to use PIC to differentially label peptides for relative quantitation. TpepC was differentially labeled in three separate experiments in order to evaluate reproducibility, dynamic range, and linearity of PIC labeling. Observed ratios of labeled TpepC were linear across a 10 000-fold range of concentration ratios from 0.01 to 100 with an R2 value of >0.99 (Figure 3), thus indicating a wide dynamic range for quantitation. Variability between experiments was less than 10% for the range of 0.1 to 10 and less than 20% for the range from 10 to 100, which is similar to or less than variability reported for other labeling techniques.16,27 Variation between analyses was similar over the course of several days, thus indicating no significant impact of day-to-day instrumental variation on quantitation. Differential quantitation with PIC was also evaluated in a tryptic digest of BSA with multiple standard peptides whose concentration ratios varied within each sample and TpepC whose concentration ratio varied between samples (Figure 4). Similar to the previous experiments (Figure 3), quantitation of TpepC was linear across the 100-fold range of concentration ratios, thus indicating that quantitation of peptides can be compared between multiple samples. Because of this, PIC labeling is applicable for determination of dose- or concentration-response and kinetics of formation of protein modifications. The observed ratios of standard peptides aligned linearly with the slope (1.06) of the trend line, also indicating that observed ratios of these peptides directly reflect actual ratios in the samples. Proteolysis was performed prior to PIC labeling, and cleavage of the frog24 and BSA was necessary for detection and quantitation of peptides that comprise the standard curve. Variation in proteolysis between samples would decrease the reliability of quantitation and increase the standard deviation between samples. Two frog24 peptides were released by proteolysis and detected for the standard curve with a standard deviation of less than 15%, thus indicating that proteolysis did not significantly affect relative quantitation of these peptides by PIC labeling. The variation of other standard peptides was less than 15% except C3a, which was 21%. Three BSA peptides were randomly selected for the standard curve and each varied by less than 10% from the expected ratio of 1. Trend lines for TpepC and the standard peptides were nearly superimposable, thus indicating that quantitation of peptides by PIC labeling is as reproducible across multiple samples as it is within single samples. Taken together, these data demonstrate that differential PIC labeling can be applied to quantify multiple peptides present at different concentration ratios within the same sample and with low variability between samples. An important consideration with any chemical tagging strategy is the effect of the label on peptide chromatography. An interesting benefit of the PIC label is that it increased detection of short (e.g., 4-5 amino acids) tryptic peptides and peptide adducts. N-terminal PIC derivatization adds a moderately hydrophobic group that increases retention on reverse phase columns. Peptides that might otherwise elute near the void volume and escape detection thus may be detected. Moreover, peptides elute in mobile phase at somewhat higher acetonitrile concentration, which increases spray stability and signal/noise in electrospray analyses. These advantages allowed us to detect the styrene oxide-histidine adducts of the peptides

research articles

Quantitative Analysis of Modified Proteins

GHGK and AHGK, which escaped detection in our earlier work with underivatized Hb peptides and peptide adducts.9 Modification of peptides with other labels has been shown to affect fragmentation in CID.17,19 Fragmentation of multiply charged PIC-modified peptides was not significantly affected by labeling. A neutral loss of PIC typically dominated singly charged spectra, but informative peptide sequence information was still obtained (data not shown). Sequest analyses28,29 of MS/ MS data generally yielded similar overall protein sequence coverage for unlabeled and PIC labeled protein digests when Sequest search parameters included possible N-terminal PIC modification. The two chromatographically resolved adducts of GHGK and AHGK may either be positional isomers that differ in position of attachment of the peptide to styrene oxide (alpha versus beta to the phenyl ring) or diastereomers formed by reaction with the racemic epoxide. Isomeric products are possible for all peptide adducts, but were only observed for the GHGK and AHGK adducts. This may reflect the relative sizes of the adducts and the modified peptides. The interaction of relatively small peptide adducts with the LC stationary phase is probably more influenced by the adduct structure than is that of larger peptide adducts. We applied the PIC labeling technique to quantify adduction of hemoglobin by styrene oxide. Peptide adducts at all concentrations along the dose response curve were quantified relative to the same adducts in the 40 mM sample, thus all adducts approach 1.0 at the 40 mM dose. All adducts demonstrated a dose response, approaching zero at the lowest dose and increasing in abundance at the 20 and 40 mM doses. This prototype experiment illustrates the value of the PIC labeling strategy in quantifying changes in protein modifications with experimental conditions. The only disadvantage of the PIC label compared to solid phase or affinity tag isotope labels (e.g., ICAT) is the lack of simplification of the mixture by selective capture of a subpopulation of peptides.14,20 Others have described selective purification of cysteine- or histidine-containing peptides prior to LC-MS analysis to simplify the mixture and facilitate protein quantitation.15,16 Thus, optimal application of PIC labels would be to samples containing a limited number of proteins or in analyses involving more extensive protein or peptide fractionation prior to MS analysis. Finally, the accuracy of post-digestion labeling strategies, such as PIC labeling assumes equal digestion efficiency in each pair of samples to be compared. We did not observe significant differences in tryptic digestion in paired samples in our studies. However, in analyses where confirmation of digestion efficiency was needed, samples could be spiked with equal amounts of some exogenous protein, whose peptides could be detected and compared to normalize comparisons between samples. Abbreviations. BSA, bovine serum albumin; Hb, hemoglobin; HEPES, N-(2-hydroxyethyl)piperazine-N′-(2-ethanesulfonic acid); HPLC, high performance liquid chromatography; ICAT, isotope-coded affinity tag; LC-MS, liquid chromatographymass spectrometry; LC-MS/MS, liquid chromatographytandem mass spectrometry; MS/MS, tandem mass spectrometry; PIC, phenyl isocyanate; TCEP, tris(carboxyethyl)phosphine.

Acknowledgment. This work was supported by NIH Grant Nos. ES10056, ES06694, and ES07091. Supporting Information Available: Annotated MSMS spectra of styrene oxide-adducted hemoglobin peptides.

This material is available free of charge via the Internet at http://pubs.acs.org.

References (1) Krishna, R. G.; Wold, F. Post-translational Modification of Proteins. Adv. Enzymol. Relat. Areas Mol. Biol. 1993, 67, 265-298. (2) Guengerich, F. P.; Liebler, D. C. Enzymatic Activation of Chemicals to Toxic Metabolites. Crit. Rev. Toxicol. 1985, 14, 259-307. (3) Cohen, S. D.; Pumford, N. R.; Khairallah, E. A.; Boekelheide, K.; Pohl, L. R.; Amouzadeh, H. R.; Hinson, J. A. Selective Protein Covalent Binding and Target Organ Toxicity. Toxicol. Appl. Pharmacol. 1997, 143, 1-12. (4) Pumford, N. R.; Halmes, N. C. Protein Targets of Xenobiotic Reactive Intermediates. Annu. Rev. Pharmacol. Toxicol. 1997, 37, 91-117. (5) Yates, J. R. Mass Spectrometry and the Age of the Proteome. J. Mass. Spectrom. 1998, 33, 1-19. (6) Liebler, D. C. Proteomic Approaches to Characterize Protein Modifications: New Tools to Study the Effects of Environmental Exposures. Environ. Health Perspect. 2002, 110 Suppl 1, 3-9. (7) Hansen, B. T.; Jones, J. A.; Mason, D. E.; Liebler, D. C. SALSA: A Pattern Recognition Algorithm to Detect Electrophile-Adducted Peptides by Automated Evaluation of CID Spectra in LC-MS/ MS Analyses. Anal. Chem. 2001, 73, 1676-1683. (8) Liebler, D. C.; Hansen, B. T.; Davey, S. W.; Tiscareno, L.; Mason, D. E. Peptide Sequence Motif Analysis of Tandem MS Data With the SALSA Algorithm. Anal. Chem. 2002, 74, 203-210. (9) Badghisi, H.; Liebler, D. C. Sequence Mapping of Epoxide Adducts in Human Hemoglobin With LC-Tandem MS and the SALSA Algorithm. Chem. Res. Toxicol. 2002, 15, 799-805. (10) Nelson, S. D. Mechanisms of the Formation and Disposition of Reactive Metabolites That Can Cause Acute Liver Injury. Drug Metab. Rev. 1995, 27, 147-177. (11) Oda, Y.; Nagasu, T.; Chait, B. T. Enrichment Analysis of Phosphorylated Proteins As a Tool for Probing the Phosphoproteome. Nat. Biotechnol. 2001, 19, 379-382. (12) Ong, S. E.; Blagoev, B.; Kratchmarova, I.; Kristensen, D. B.; Steen, H.; Pandey, A.; Mann, M. Stable Isotope Labeling by Amino Acids in Cell Culture, SILAC, As a Simple and Accurate Approach to Expression Proteomics. Mol. Cell Proteom. 2002, 1, 376-386. (13) Yao, X.; Freas, A.; Ramirez, J.; Demirev, P. A.; Fenselau, C. Proteolytic 18O Labeling for Comparative Proteomics: Model Studies With Two Serotypes of Adenovirus. Anal. Chem. 2001, 73, 2836-2842. (14) Gygi, S. P.; Rist, B.; Gerber, S. A.; Turecek, F.; Gelb, M. H.; Aebersold, R. Quantitative Analysis of Complex Protein Mixtures Using Isotope-Coded Affinity Tags. Nat. Biotechnol. 1999, 17, 994-999. (15) Wang, S.; Zhang, X.; Regnier, F. E. Quantitative Proteomics Strategy Involving the Selection of Peptides Containing Both Cysteine and Histidine From Tryptic Digests of Cell Lysates. J. Chromatogr. A 2002, 949, 153-162. (16) Ji, J.; Chakraborty, A.; Geng, M.; Zhang, X.; Amini, A.; Bina, M.; Regnier, F. Strategy for Qualitative and Quantitative Analysis in Proteomics Based on Signature Peptides. J. Chromatogr. B, Biomed. Sci. Appl. 2000, 745, 197-210. (17) Munchbach, M.; Quadroni, M.; Miotto, G.; James, P. Quantitation and Facilitated De Novo Sequencing of Proteins by Isotopic N-Terminal Labeling of Peptides With a Fragmentation-Directing Moiety. Anal. Chem. 2000, 72, 4047-4057. (18) Goodlett, D. R.; Keller, A.; Watts, J. D.; Newitt, R.; Yi, E. C.; Purvine, S.; Eng, J. K.; von Haller, P.; Aebersold, R.; Kolker, E. Differential Stable Isotope Labeling of Peptides for Quantitation and De Novo Sequence Derivation. Rapid Commun. Mass Spectrom. 2001, 15, 1214-1221. (19) Peters, E. C.; Horn, D. M.; Tully, D. C.; Brock, A. A Novel Multifunctional Labeling Reagent for Enhanced Protein Characterization With Mass Spectrometry. Rapid Commun. Mass Spectrom. 2001, 15, 2387-2392. (20) Zhou, H.; Ranish, J. A.; Watts, J. D.; Aebersold, R. Quantitative Proteome Analysis by Solid-Phase Isotope Tagging and Mass Spectrometry. Nat. Biotechnol. 2002, 20, 512-515. (21) Regnier, F. E.; Riggs, L.; Zhang, R.; Xiong, L.; Liu, P.; Chakraborty, A.; Seeley, E.; Sioma, C.; Thompson, R. A. Comparative Proteomics Based on Stable Isotope Labeling and Affinity Selection. J. Mass Spectrom. 2002, 37, 133-145. (22) Zhang, R.; Sioma, C. S.; Thompson, R. A.; Xiong, L.; Regnier, F. E. Controlling Deuterium Isotope Effects in Comparative Proteomics. Anal. Chem. 2002, 74, 3662-3669.

Journal of Proteome Research • Vol. 2, No. 3, 2003 271

research articles (23) Cagney, G.; Emili, A. De Novo Peptide Sequencing and Quantitative Profiling of Complex Protein Mixtures Using Mass-Coded Abundance Tagging. Nat. Biotechnol. 2002, 20, 163-170. (24) Stark, G. R. Reactions of Cyanate With Functional Groups of Proteins. III. Reactions With Amino and Carboxyl Groups. Biochem. 1965, 4, 1030-1036. (25) Stark, G. R. On the Reversible Reaction of Cyanate With Sulfhydryl Groups and the Determination of NH2-Terminal Cysteine and Cystine in Proteins. J. Biol. Chem. 1964, 239, 1411-1414. (26) Stark, G. R. Reactions of Cyanate With Functional Groups of Proteins. IV. Inertness of Aliphatic Hydroxyl Groups. Formation of Carbamyl- and Acylhydantoins. Biochemistry 1965, 4, 23632367.

272

Journal of Proteome Research • Vol. 2, No. 3, 2003

Mason and Liebler (27) Oda, Y.; Huang, K.; Cross, F. R.; Cowburn, D.; Chait, B. T. Accurate Quantitation of Protein Expression and Site-Specific Phosphorylation. Proc. Natl. Acad. Sci., U. S. A 1999, 96, 6591-6596. (28) Eng, J. K.; McCormack, A. L.; Yates, J. R. An Approach to Correlate Tandem Mass-Spectral Data of Peptides With Amino-AcidSequences in A Protein Database. J. Am. Soc. Mass Spectrom. 1994, 5, 976-989. (29) Yates, J. R.; Eng, J. K.; McCormack, A. L.; Schieltz, D. Method to Correlate Tandem Mass Spectra of Modified Peptides to Amino Acid Sequences in the Protein Database. Anal. Chem. 1995, 67, 1426-1436.

PR0255856