A Comprehensive Structural Analysis of Hemoglobin Adducts Formed

A widely used method for assessing occupational and environmental exposure to 1,3-butadiene involves the detection of hemoglobin adducts formed by the...
0 downloads 0 Views 178KB Size
Chem. Res. Toxicol. 2000, 13, 1103-1113

1103

A Comprehensive Structural Analysis of Hemoglobin Adducts Formed after in Vitro Exposure of Erythrocytes to Butadiene Monoxide Thomas S. Moll,† Amy C. Harms,‡ and Adnan A. Elfarra*,† Department of Comparative Biosciences and Center for Environmental Toxicology and Mass Spectrometry Facility, University of Wisconsin Biotechnology Center, University of Wisconsin, Madison, Wisconsin 53706-1102 Received July 13, 2000

A widely used method for assessing occupational and environmental exposure to 1,3-butadiene involves the detection of hemoglobin adducts formed by the reactive metabolite butadiene monoxide (BMO). This assay employs the N-alkyl Edman method, which was developed to determine adducts formed at the amine group of the N-terminal valine of hemoglobin. Disadvantages of this procedure include its limitation to detecting only one adduct per globin chain, despite the presence of numerous other, and potentially more reactive, nucleophilic amino acids in hemoglobin. The method is also not suitable for determining whether the reaction of BMO occurs at the N-terminal valine of R- or β-globin. The primary goals of the current research are to determine the degree of modification of R- and β-globin chains by BMO and to localize the reactive residues to specific regions of the globin polypeptides. The reaction products after in vitro incubation of C57Bl/6 mouse erythrocytes with BMO were isolated by acid extraction of heme and microprecipitation of globin, followed by the determination of the number and location of adducts by mass spectrometry. The modification degree was monitored by electrospray mass spectrometry, which was used to measure the time- and concentrationdependent formation of BMO-hemoglobin adducts (e10 adducts per globin). The results indicate that BMO reacts faster and to a higher degree with R-globin than with β-globin. Structural analysis was performed by peptide mapping of globin peptides after trypsin digestion using liquid chromatography/mass spectrometry. These experiments allowed the localization of BMO-hemoglobin adducts to specific regions within R- and β-globin, and also provided information about their relative reactivity. Interestingly, the initial site of adduct formation on R-globin is located near the N-terminal peptide, whereas the initial site on β-globin is located at the C-terminal region. Collectively, the results establish differences in the reactivities of Rand β-globin toward BMO, demonstrate the formation of multiple adducts at several R- and β-globin sites, and show that the N-terminal valine residues are not the first to be modified by BMO.

Introduction 1,3-Butadiene (BD)1 has recently been upgraded to “Known to be a Human Carcinogen” by the U.S. Department of Health and Human Services-National Toxicology Program on the basis of epidemiological studies and substantial mechanistic information, indicating a causal relationship between occupational exposure to BD and excess mortality from lymphatic and/or hematopoietic cancers (1). BD is used primarily as a chemical intermediate and polymer component in the manufacture of * To whom correspondence should be addressed: Department of Comparative Biosciences, School of Veterinary Medicine, University of Wisconsin, 2015 Linden Drive West, Madison, WI 53706-1102. Phone: (608) 262-6518. Fax: (608) 263-3926. E-mail: elfarraa@svm. vetmed.wisc.edu. † Department of Comparative Biosciences and Center for Environmental Toxicology. ‡ Mass Spectrometry Facility, University of Wisconsin Biotechnology Center. 1 Abbreviations: BD, 1,3-butadiene; BMO, butadiene monoxide; ES/ MS, electrospray mass spectrometry; LC/MS, liquid chromatography/ mass spectrometry; GC/MS, gas chromatography/mass spectrometry; Hb, hemoglobin; DPG, 2,3-diphosphoglycerate; TFA, trifluoroacetic acid; MTBSTFA, N-methyl-N-(tert-butyldimethylsilyl)trifluoroacetamide; TBDMS, tert-butyldimethylsilyl.

synthetic rubber and plastics, and is also listed by the EPA as a hazardous component of air pollution; environmental sources include cigarette smoke and automobile exhaust. BD is bioactivated by cytochrome P450- and myeloperoxidase-mediated oxidation to produce butadiene monoxide (BMO) (2). This primary reactive metabolite has been shown to undergo nucleophilic substitution reactions with numerous cellular macromolecules, including DNA (3) and proteins (4) both in vivo and in vitro. To date, the most commonly used method for determining the extent of modification of hemoglobin (Hb) by xenobiotics or their reactive intermediates, including BMO, is the modified Edman degradation procedure (5). Although very specific for detecting adducts formed at the free R-amine of the N-terminal valine, present in both R- and β-globin, this procedure has many disadvantages. First, only one reactive site per globin can be monitored, despite the presence of numerous other, and potentially more reactive, nucleophilic amino acids within the globin sequence. Second, the N-terminal valine is subject to modification by other compounds normally present in erythrocytes, which could decrease the sensitivity and

10.1021/tx000151f CCC: $19.00 © 2000 American Chemical Society Published on Web 09/28/2000

1104

Chem. Res. Toxicol., Vol. 13, No. 11, 2000

accuracy of detecting BMO-Hb adducts. For example, the R-amino groups of the N-terminal valines can react with CO2 to form carbamylated amines, which form salt bridges with other residues within globin that could affect modification by BMO (6). Also, the nonenzymatic condensation of glucose with N-terminal valines to form glycated amines can occur in Hb (7). Indeed, the measurement of the level of glycohemoglobin is widely used as an indicator for long-term diabetic control (8). In addition, the cofactor 2,3-diphosphoglycerate (DPG), which is present in human red blood cells at about the same molar concentration as Hb (9), binds electrostatically to the N-terminal amine group of β-globin (10) and may affect the reaction with BMO. Third, with regard to the modified Edman degradation procedure, there is an inherent racemization of amino acids upon reaction with phenyl isothiocyanate to form phenylthiohydantoins which prevents the quantitation of individual diastereomers of the N-terminal BMO-valine adducts (11, 12). Combined, these limitations of measuring the levels of N-terminal valine adducts as a marker of chemical exposure led us to consider alternative methodologies. Recent advances in mass spectrometry serve as the logical starting point for the development of methods that will enable a comprehensive and global structural analysis of Hb adducts formed after in vitro exposure of erythrocytes to BMO. Specifically, electrospray ionization mass spectrometry (ES/MS) is a very mild process with little thermal input. Electrospray ionization is capable of generating multiply charged ions with low mass/charge ratios, which can be easily analyzed by quadrupole mass spectrometry. Computer-aided algorithms are available for deriving molecular mass information from these multiply charged ion series. Averaging the results of multiply charged peaks leads to mass measurement with an accuracy of (0.01%, and can be used to measure the molecular masses of proteins and polypeptides with molecular masses of up to ∼60 kDa (13). In this study, ES/MS was used to determine the degree of modification of Hb after in vitro exposure of red blood cells to BMO. Because the ES/MS apparatus can be equipped with an HPLC system with reversed phase (RP) microbore/capillary columns for liquid chromatography/ mass spectrometry (LC/MS) experiments, LC/MS was used to determine the localization of reactive amino acids within the globin sequences after digestion of the globin chains by trypsin. To determine the amino acid functional groups involved in BMO adduct formation, globin peptide bonds were also acid hydrolyzed and the modified and unmodified amino acids were analyzed using gas chromatography/mass spectrometry (GC/MS) (14). The results obtained from these experiments enable the direct comparison of the reactivities of R- and β-globin with BMO, and identify the specific regions prone to modification. In addition, several BMO-amino acid adducts have been detected, and the possible role of these modifications in the structural and functional properties of Hb is discussed.

Materials and Methods Caution: BMO is a known mutagen and carcinogen in laboratory animals and must be handled using proper safety measures. Materials. Racemic BMO, anhydrous methanol, dimethylformamide, and trifluoroacetic acid (TFA) were obtained from

Moll et al. Aldrich Chemical Co. (Milwaukee, WI). Acetone was obtained from Sigma Chemical Co. (St. Louis, MO). N-Methyl-N-(tertbutyldimethylsilyl)trifluoroacetamide (MTBSTFA) was obtained from Pierce Chemical Co. (Rockford, IL). Trypsin (reductively alkylated) was obtained from Promega (Madison, WI). ZipTip pipet tips were obtained from Millipore (Bedford, MA). Whole blood (heparin anticoagulant) from C57Bl/6 mice was obtained from Harlan Bioproducts (Madison, WI). The C57Bl/6 mouse strain was chosen due to its close relationship to the B6C3F1 mouse, the primary strain used in long-term toxicology studies of BD. B6C3F1 is an F1 hybrid between C57Bl/6 and C3H. In addition, the Hb sequences of R- and β-globin from C57Bl/6 mice are known (15, 16), which are required for the peptide mapping analysis. Methods. (1) In Vitro Reaction of Mouse Erythrocytes with BMO. Whole blood with heparin as the anticoagulant was obtained from freshly killed mice and was processed within 24 h. Intact erythrocytes were initially separated from plasma proteins by centrifugation of whole mouse blood for 5 min at 1500g. After removal of the plasma fraction, the red blood cells were washed three times with an equal volume of isotonic saline (0.9% w/v). The red blood cells were resuspended in an equal volume of phosphate-buffered saline (0.1 M, pH 7.4). Solutions of BMO for in vitro incubation were prepared by dilution to the appropriate concentration in anhydrous methanol. After addition of BMO (25 µL/mL; final concentration of 1-150 mM), the mixture was placed in a shaking water bath at 37 °C and the reaction stopped at designated time points. After extraction of the samples with toluene to remove excess BMO, the red blood cells were lysed in an equal volume of cold distilled water. Aliquots were added dropwise with vigorous vortexing to cold acidified acetone (1% HCl) to precipitate globin. The sample was centrifuged for 15 min at 3000g, and the pellets were washed once with cold acidified acetone, once with cold acetone, and once with cold diethyl ether. The globin was dried under a gentle stream of nitrogen and stored at -20 °C. (2) Trypsin Digestion of Globin. Trypsin, stored as a lyophilized powder at -20 °C, was reconstituted in 50 mM acetic acid immediately prior to use. Proteolytic digestion of precipitated globin samples was carried out in 100 mM ammonium bicarbonate (pH 8.0) at 37 °C for 12 h. The enzyme/substrate ratio was 1/20 (w/w). The reaction was stopped by the addition of TFA to a final concentration of 0.05% (v/v). (3) ES/MS of Globin and BMO-Globin Proteins. ES/MS analysis of modified and unmodified R- and β-globin chains was performed with a Perkin-Elmer Sciex API 365 triple-quadrupole electrospray ionization mass spectrometer. Prior to injection, the globin samples were purified and concentrated using C18 ZipTip pipet tips. The spectra were scanned with a mass range between 0.6 and 1.6 kDa. Molecular masses of the globins were obtained by deconvolution of each multiply charged ion spectrum (i.e., hypermass analysis). The application software used for the reconstruction, supplied by the manufacturer, was Biospec Reconstruct (BioMultiView, version 1.3.1, Perkin-Elmer). (4) LC/MS of Trypsin Digests of Globin and BMOGlobin Adducts. LC/MS was performed on an API 300 PerkinElmer Sciex triple-quadrupole LC/MS/MS mass spectrometer coupled to a high-performance liquid chromatography (HPLC) system using a Vydac microbore C18 column (15 cm × 1.0 mm, 5 µm particle size, 300 Å pore size silica). Solvent A consisted of 0.05% TFA in water, and solvent B was 0.05% TFA in 70% acetonitrile. The column was equilibrated at 100% solvent A, and a linear gradient to 25% solvent A was programmed with a duration of 125 min. The solvent system was then changed to 1% solvent A in 10 min and held for 1 min. The flow rate was 0.02 mL/min. Masses between 0.3 and 2.3 kDa were measured using a step of 0.2 Da and a dwell time of 0.35 s. (5) Amino Acid Analysis of Globin and BMO-Globin Adducts by GC/MS. Globin and BMO-globin samples were reconstituted in 6 N HCl and placed in glass reaction vials capped with Teflon septa. The samples were flushed with N2 prior to capping and placed in a Reacti-Therm III (Pierce)

Hemoglobin Adducts of Butadiene Monoxide

Figure 1. Electrospray mass spectrometry of mouse R- and β-globin. (A) Mass spectrum of mouse R- and β-globin. (B) Deconvoluted mass spectrum of mouse R- and β-globin. heating/stirring module set at 110 °C for 24 h. The solvent was evaporated using the Reacti-Vap III (Pierce) apparatus equipped with N2. Dimethylformamide (75 µL) and MTBSTFA (75 µL) were added, and the samples were vortexed. The solution was flushed with N2 and then heated for 1 h at 75 °C. The samples were analyzed using a HP-6890 series mass selective detector interfaced with an HP 6890 series gas chromatograph (HewlettPackard, Palo Alto, CA). Separation of components was accomplished using a 30 m × 0.25 mm i.d. HP-5MS (0.25 µm) capillary column. The GC injection port temperature was maintained at 300 °C, and the inlet pressure of the ultrapure helium gas was maintained at 20 kPa during the run. All injections were made in the splitless mode. The initial oven setpoint was 120 °C, and an initial temperature gradient of 2 °C/min was used to reach 155 °C, followed by an increase at a rate of 10 °C/min to 185 °C. A gradient of 3 °C/min was used to reach 280 °C followed by an increase at a rate of 10 °C/min to 300 °C, at which the sample was held for 5 min. The mass spectrometer was programmed for scanning mode with a scan range of 50-700 Da.

Results The covalent modification of Hb with BMO at physiological pH (7.4) and temperature (37 °C) was studied by incubating mouse erythrocytes with varying concentrations (5-100 mM) of the epoxide. At the specified time points, the red blood cells were lysed and the globin was microprecipitated using acidic acetone at low temperatures. Prior to mass spectrometry, the globin samples were further purified and concentrated using C18 ZipTip pipet tips as described in Materials and Methods. Figure 1A shows the ES/MS spectra for R- and β-globin extracted from control C57Bl/6 mouse erythrocytes. Mass analyses were carried out under the positive ion mode

Chem. Res. Toxicol., Vol. 13, No. 11, 2000 1105

Figure 2. Electrospray mass spectrometry of mouse R- and β-globin after incubation with BMO. Mouse erythrocytes were incubated with BMO at pH 7.4 for 4 h at 37 °C. Globin was microprecipitated by dilution in cold acid acetone, and the samples were analyzed by ES/MS.

which resulted in spectra for the globins recorded as a series of protonated species. Control globin typically produced spectra having 10-21 positive charges; modification of the globin by BMO adduct formation slightly decreased the number of positive charges. The ES/MS instrumentation that was utilized has a mass range between 0.03 and 3 kDa with a mass accuracy of (0.1 Da. Molecular masses of the globins were obtained by deconvolution of each multiply charged ion spectrum (i.e., hypermass analysis). The application software used for the reconstruction (Biospec Reconstruct, BioMultiView, version 1.3.1) employs an algorithm that relies on Bayesian statistics to predict the initial m/z spectrum allowing for noise and instrumental peak widths. To maximally eliminate artifacts, the algorithm is iterated 20 times to produce the final solution. In essence, the reconstruction maps the spectra to the zero-charge form, which represents the correct molecular mass. Upon deconvolution of the spectra (Figure 1B), the two main peaks observed correspond to R-globin (molecular mass of 14 982 Da) and β-globin (molecular mass of 15 618 Da). The formation of BMO adducts was then monitored as a function of epoxide concentration (5-100 mM) and time (0-12 h). The nucleophilic addition of the electrophile BMO to reactive amino acids on Hb can occur at either the primary (C-1) or secondary (C-2) carbon of the epoxide; either regioisomer adduct results in a molecular mass increase of 70 Da for Hb. Typical spectra are shown in Figure 2, which shows the formation of BMO adducts at 0, 25, and 50 mM BMO after a 4 h incubation under physiological conditions. As shown, the spectra can easily distinguish mass increases of 70 Da. ES/MS has been used previously to measure the number of methylated residues (i.e., increase of 14 Da) in hemoglobin exposed

1106

Chem. Res. Toxicol., Vol. 13, No. 11, 2000

Figure 3. Reactivities of mouse R- and β-globin with BMO. The relative reactivities of the individual R- and β-globin chains were determined by monitoring the loss of unmodified globin as a function of time and epoxide concentration. The slopes of the linear regression were used to calculate second-order rate constants for the reaction of BMO with globin.

to methyl bromide (17), and to measure the number of heterozygous Hb variants with mass differences as small as 12 Da (18). The observed molecular mass increases of modified globins are within (3 Da of the calculated increases. All spectra were collected using multiscan averaging (20 scans). The in vitro kinetics of BMO-Hb adduct formation were studied by determining the overall covalent bondforming rate constants for the reaction of BMO with Rand β-globin. The rate constants were obtained using the slope of the initial linear portion of the loss of unmodified globin as a function of time and BMO concentration (Figure 3). The second-order rate constants were estimated using the following equation:

k)

[BMO-globin](k′) [globin]0[BMO]0(1 - e-k′t)

where [globin]0 is the molar concentration of R- and β-globin [4.2 mM in mouse (19)], k′ is the first-order rate constant for aqueous BMO hydrolysis [1.292 × 10-5 s-1 (20)], and [BMO-globin]/[BMO]0 is the dependence of the slope of the linear regression of adduct formation (RR ) 0.999; Rβ ) 0.973) at 4 h on the initial concentration of BMO. On the basis of the kinetic calculations, R-globin has a higher rate constant for adduct formation (1.65 × 10-3 M-1 s-1; SE ) 0.05) than β-globin (1.15 × 10-3 M-1 s-1; SE ) 0.09). The differences in the reactivity of Rand β-globin toward BMO could be attributed to sequence variation, manifested by different levels of reactive amino acid residues (e.g., C57Bl/6 mouse R-globin contains 11 histidines compared to 9 in β-globin). The variable rate

Moll et al.

constants for adduct formation could also be due to the greater availability of the reactive residues to the reagent (e.g., surface vs buried). Another possibility is that the β-globin N-terminal amine is bound to DPG, thereby inhibiting the formation of N-terminal valine-BMO adducts. The complete profile of R-globin modification as a function of time and concentration of BMO is displayed in Figure 4. The bar chart shown in Figure 4A shows the reactivity of R-globin at a low BMO concentration (5 mM). Under these conditions, the majority of modified R-globin (19%) has only one BMO bound even after a 12 h incubation, with a small portion (97% modified R-globin, with up to four bound BMO adducts detected. In addition, the quantity of each level of modification is still increasing with time. In contrast, the lower levels of modification (e.g., one or two BMO molecules per globin) at 50 mM BMO (Figure 4C) begin to decrease. Although the majority of modified R-globin at 5 and 25 mM BMO only contained one BMO molecule per globin, the majority of the modified R-globin at 50 mM BMO contained three BMO molecules per globin after 12 h. Also, up to seven BMO molecules per globin could be detected and the quantity of adducts at each time point can be represented by a normal distribution. At the high concentration of BMO (100 mM), there is a very interesting shift in the distribution patterns at any given time point (Figure 4D). For example, there appears to be a buildup of R-globin proteins with six BMO molecules bound after a 4 h incubation, resulting in a skewed distribution pattern. Interestingly, this is the major adducted globin at all time points. After 12 h, the distribution is highly skewed, with ∼50% of all modified globins having six bound BMO molecules; up to ten BMO molecules per globin can be detected. These observations strongly suggest the presence of six residues on R-globin that are more readily modified than the additional residues that eventually form BMO adducts. This could be due to either variable specific reactivities or optimal positioning of reactive groups on the three-dimensional structure of Hb. The profile of β-globin modification by BMO adduct formation is shown in Figure 5. Similar patterns and distributions can be seen for the 5, 25, and 50 mM incubations. For example, at the lowest concentration (Figure 5A), most of the modified β-globin has one BMO molecule per globin, with up to two BMO molecules per globin detected. However, only 13% of β-globin is modified after 12 h compared to 21% of R-globin. Similarly, at 25 mM BMO (Figure 5B), up to three BMO per globin was detected and the total modification of β-globin increased to 61%. However, this is far below the 97% modified found with R-globin. At 50 mM BMO (Figure 5C), the normal distribution of modified polypeptides seen in R-globin is also apparent with β-globin modifications at the 8 and 12 h incubations. Besides the overall lower reactivity of β-globin, the major difference between modification patterns can be seen at the highest concentration of BMO (Figure 5D). Again, the distribution is skewed but with different levels of bias. After 4 h, there is a buildup of the two BMO per globin species, but the

Hemoglobin Adducts of Butadiene Monoxide

Chem. Res. Toxicol., Vol. 13, No. 11, 2000 1107

Figure 4. Complete profile for the reaction of BMO with R-globin. The amount of adducts arising from the reaction of BMO with nucleophilic groups on R-globin was determined as a function of time and epoxide concentration using ES/MS: (A) 5 mM BMO, (B) 25 mM BMO, (C) 50 mM BMO, and (D) 100 mM BMO.

Figure 5. Complete profile for the reaction of BMO with β-globin. The amount of adducts arising from the reaction of BMO with nucleophilic groups on β-globin was measured as a function of time and epoxide concentration using ES/MS: (A) 5 mM BMO, (B) 25 mM BMO, (C) 50 mM BMO, and (D) 100 mM BMO.

levels of modification increase to eight BMO per globin. Similar adduct buildups of the two BMO per globin species is seen after 8 and 12 h, but an additional buildup of eight BMO per globin proteins is also seen after 12 h.

Thus, in β-globin, there are apparently two highly reactive residues followed by six less reactive groups. With this methodology, up to ten BMO adducts per β-globin were detected, similar to R-globin.

1108

Chem. Res. Toxicol., Vol. 13, No. 11, 2000

Figure 6. Liquid chromatography/mass spectrometry of trypsin digests of hemoglobin. (A) Peptide mapping of the tryptic digest of control mouse Hb. (B) Chromatogram of the tryptic digest of mouse Hb modified with BMO (150 mM BMO for 4 h at 37 °C). The appearance of modified peptides is denoted (*).

To localize the sites of BMO adduct formation on Rand β-globin, the modified globin chains were subjected to trypsin digestion and analyzed by LC/MS. Hemoglobin (unmodified control) was isolated from mouse erythrocytes and digested with trypsin. Since native trypsin is highly subject to autolysis, which would generate fragments that could interfere with peptide mapping, only trypsin which had been modified by reductive alkylation was used. This modification has been shown to be extremely resistant to autolytic digestion (21). The peptide fragments were separated on a Vydac C18 column (Figure 6A) and analyzed by mass spectrometry. Peptide mapping was performed using the software supplied by the LC/MS manufacturer (PeptideMap, BioMultiView). Briefly, amino acid sequences of C57Bl/6 mouse globins (20, 21) were imported into the sequence pane. Peptide fragments produced by trypsin digestion were predicted, and the spectra were automatically processed to determine peptide molecular masses using multiply charged ion series. Finally, coverage maps were created that identified the tryptic peptides as R- or β-globin sequences (Figure 6A). The complete peptide map of control unmodified globins is listed in Table 1. Incubations of C57Bl/6 mouse erythrocytes with BMO as a function of time and epoxide concentration were analyzed similarly by trypsin digestion and LC/MS to localize peptides containing BMO adducts. A typical spectrum of a modified globin digest is shown in Figure 6B, which shows a high degree of modification as evidenced by the appearance of numerous new peaks. To simplify the process of identifying the peptides containing BMO adducts, 5 min spectral windows were used to generate extracted ion chromatograms for both the control peptides and the control and modified peptides. The extracted ion chromatogram derived from the control

Moll et al.

Hb was then subtracted from the modified Hb chromatogram, and the remaining mass values were manually fit to calculated peptide masses having varying degrees of BMO modification. Surprisingly, with increasing levels of modification, the appearance of much higher molecular mass peptides was observed. It was hypothesized that BMO adduct formation could be inhibiting trypsin proteolytic activity, resulting in the incomplete digestion of globin. To test this hypothesis, the molecular masses of nearest-neighbor nondigested peptides were calculated, along with multiple additions of BMO, and fit to the series of multiply charged ions found in the subtracted spectra. The results were remarkably accurate in detecting several modified peptides showing varying degrees of incomplete digestion, which are listed in Table 2. The degree of inhibition of trypsin activity as evidenced by partial digestion is directly proportional to the degree of BMO adduct formation, strongly suggesting the covalent modification of lysine by BMO. Ten BMO peptides were assigned to R-globin compared to thirteen in β-globin. To identify the most reactive peptides, Hb was reacted with varying concentrations of BMO (1-150 mM) for 4 h and the trypsin digests were subsequently peptide mapped. The results from R-globin are shown in Figure 7A. At the lowest BMO concentration (1 mM), only one peptide is modified [R(8-11)]. Potential nucleophilic sites in this peptide are Ser-8 and Lys-11. At 10 mM BMO, both R(1-7) and R(8-11) are modified, in addition to a nondigested peptide consisting of R(8-11) and R(12-16). In contrast to the preferential formation of BMO adducts near the N-terminus of R-globin, the initial peptides modified in β-globin (Figure 7B) are located near the C-terminus; peptide β(133-144) and the nondigested peptide β(133-144)+β(145-146) are the initial products, suggesting reaction at Lys-144 as well as His-143 or His-146. At higher concentrations, the globins appear to be modified in distinct regions of the polypeptides. In R-globin, adducts are formed preferentially within the R(1-31) and R(61-141) regions. In contrast, the reactive areas in β-globin are the highly modified β(60-104) region followed by the N- and Cterminal peptides. Lysine residues in both R- and β-globin are readily modified, as evidenced by the appearance of partially digested peptides; five lysines in R-globin and five lysines in β-globin are modified by BMO. Although trypsin cleaves proteins at both lysine and arginine residues, BMO does not appear to react with arginine. No inhibition of trypsin activity is found at five of the six arginine residues; the only arginine not cleaved could be a result of a lysine modification only two residues upstream. To identify the reactive amino acids in Hb, acid hydrolysis was performed on globin samples followed by amino acid analysis using gas chromatography/mass spectrometry (GC/MS). To facilitate the separation and identification of amino acids, the samples were derivatized with MTBSTFA prior to analysis by GC/MS. MTBSTFA is a silylation reagent which replaces reactive hydrogens of both the amine and carboxylic acid moieties of free amino acids with the tert-butyldimethylsilyl (TBDMS) group (22). In addition, the functional groups of serine, threonine, aspartic acid, glutamic acid, lysine, histidine, and tyrosine react with MTBSTFA to produce the tri-TBDMS derivatives. Figure 8A shows the separation and identification of 16 out of the 20 amino acids; tryptophan and cysteine are destroyed under the harsh

Hemoglobin Adducts of Butadiene Monoxide

Chem. Res. Toxicol., Vol. 13, No. 11, 2000 1109

Table 1. Peptide Mapping of Hb after Trypsin Digestiona peptide R-globin 1 2 3 4 5 6 7 8 9 10 11 12 13 14 β-globin 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

residues

mass (Da)

time

sequence

1-7 8-11 12-16 17-31 32-40 41-56 57-60 61 62-90 91 and 92 93-99 100-127 128-139 140 and 141

746.38 460.26 531.28 1529.63 1028.5 1819.99 397.21 146.11 2837.14 287.2 817.43 2994.47 1251.71 337.17

25.82 17.29 30.03 39.49 45.80 47.15 nd nd 58.19 nd 39.32 66.25 53.40 16.04

VLSGEDK SNIK AAWGK IGGHAGEYGAEALER MFASFPTTK TYFPHFDVSHGSAQVK GHGK K VADALASAAGHLDDLPGALSALSDLHAHK LR VDPVNFK LLSHCLLVTLASHHPADFTPAVHASLDK FLASVSTVLTSK YR

1-8 9-17 18-30 31-40 41-59 60-61 62-65 66 67-82 83-95 96-104 105-120 121-132 133-144 145-146

911.47 887.49 1285.63 1273.72 1981.17 245.17 411.22 146.11 1756.98 1406.65 1125.56 1714.15 1293.64 1105.66 318.13

31.96 44.58 38.74 58.54 59.71 nd nd nd 55.33 48.32 42.30 71.51 71.74 43.35 12.27

VHLTDAEK AAVSGLWGK VNADEVGGEALGR LLVVYPWTQR YFDSFGDLSSASAIMGNAK VK AHGK K VITAFNDGLNHLDSLK GTFASLSELHCDK LHVDPENFR LLGNMIVIVLGHHLGK DFTPAAQAAFQK VVAGVAAALAHK YH

a Globin from C57Bl/6 mice was used in these experiments. Peptide mapping was performed by LC/MS as described in Materials and Methods.

Table 2. Peptide Mapping of Hb-BMO Adducts after Trypsin Digestiona elution (min)

sequence

charge

no. of BMO

mass(obs) (Da)

mass(calcd) (Da)

20-25 25-30

β(145-146) β(133-144)+β(145-146) R(8-11) R(8-11)+R(12-16) β(1-8) R(1-7) R(17-31) β(96-104) β(83-95) R(41-56) β(133-144) β(83-95)+β(96-104) β(105-120) β(67-82) R(100-127) R(100-127) β(105-120) R(128-139)+R(140-141) R(61-61)+R(62-90)+R(91-92)+R(93-99) R(91-92)+R(93-99)+R(100-127) β(66-66)+β(67-82)+β(83-95) β(60-61)+β(62-65) β(66-66)+β(67-82)

1 1-2 1 1-2 1-2 1 1-2 1-2 1-2 1-3 1-2 1-3 1-2 1-2 1-5 1-4 1-3 1-5 1-5 1-3 1-3 1 1-2

1 1 1 1 1 1 1 1 1 1 1 1 2 1 1 2 1 1 3 2 1 1 1

388.4 1493.2 531.4 1061.8 981.8 816.4 1600.8 1195.4 1476.6 1890.2 1175.0 2601.4 1852.0 1826.6 3064.5 3134.7 1784.2 1660.2 4298.5 4234.6 3381.4 726.2 1973.8

388.2 1493.8 531.4 1061.6 981.6 816.5 1599.7 1195.7 1476.7 1890.1 1175.8 2602.3 1854.3 1827.1 3064.6 3134.7 1784.2 1659.0 4298.2 4238.3 3379.8 726.5 1973.2

30-35 40-45 45-50 50-55 55-60 60-65 70-75 75-80 80-85 90-95 a

Identification of the BMO-peptide adducts formed after a 4 h incubation of mouse Hb with 1-150 mM BMO at 37 °C.

conditions, and asparagine and glutamine are deamidated to aspartate and glutamate, respectively. Figure 8B shows the chromatogram of Hb modified with 150 mM BMO for 4 h. Several new peaks can be distinguished, although the levels are quite lower than expected on the basis of the data obtained using ES/MS and LC/MS. Although the BMO-Hb adducts are very stable at neutral pH, it is possible that they are unstable in the harsh conditions used to digest the protein (6 N HCl for

24 h at 110 °C). To specifically identify BMO adducts, selected ion monitoring was employed using the major fragment ion common with tert-butyldimethylsilyl (TBDMS) derivatives of amino acids, the M+ - 57 ion (22). Two BMO-histidine peaks, eluting at 45.2 and 45.6 min, were detected by selected ion monitoring of the M+ - 57 ()510) ion (Figure 9A). The presence of two BMO adducts could be due to the reaction of the epoxide at the N-1 and N-3 positions of the imidazole group, the reaction of

1110

Chem. Res. Toxicol., Vol. 13, No. 11, 2000

Figure 7. Mass spectral fingerprinting of BMO-globin adducts. Peptide mapping of BMO-globin adducts as a function of epoxide concentration was performed after trypsin digestion to create a mass spectral fingerprint of the sequential modification of the individual globins: (A) R-globin and (B) β-globin.

Figure 8. Amino acid analysis by gas chromatography/mass spectrometry of acid-hydrolyzed globin. Globin samples were acid hydrolyzed and then derivatized by MTBSTFA to create the (N,O)-TBDMS derivatives. The amino acids (with or without BMO modification) were separated by GC/MS. (A) Chromatogram of control Hb. (B) Chromatogram of Hb and BMO. BMOamino acid adducts are labeled (*).

C-1 and C-2 of the epoxide with the imidazole, or the reaction of either C-1 or C-2 of the epoxide and its diastereomer. The total ion chromatograms of the BMOHis adducts (Figure 9B,C) are very similar, with the major difference being manifested in the relative abundance of the 266 and 302 ions. Using this methodology, five BMO adducts of amino acids were detected (Figure 10). BMO adducts of histidine, lysine, serine, methionine,

Moll et al.

Figure 9. Mass spectra of BMO-histidine adducts. (A) Total ion chromatogram of the two BMO-histidine adducts formed after the reaction of mouse Hb with 150 mM BMO for 4 h at 37 °C. (B and C) Mass spectra of the individual BMO-histidine adducts.

Figure 10. Mass spectra of BMO-amino acid adducts. Mass spectra of the individual BMO-amino acid adducts formed after the reaction of mouse Hb with 150 mM BMO for 4 h at 37 °C. The selected ions that are shown represent the major M+ - 57 ion of the TBDMS-BMO-amino acid adducts. The chromatograms on the left are from control (untreated) Hb.

and the N-terminal valine have been identified. It should be noted that ester-BMO adducts of aspartic and glutamic acid, if formed, would be readily hydrolyzed under the conditions used in these experiments. No cysteine ad-

Hemoglobin Adducts of Butadiene Monoxide

ducts were detected, although the presence of only one cysteine in R- and β-globin could preclude identification under these conditions. In contrast, there are 40 histidine, 44 lysine, and 44 serine residues per Hb. Also of interest is the fact that two N-terminal valine adducts of BMO are detected. As stated previously, the nucleophilic addition of BMO to reactive amino acids can occur at either C-1 or C-2 of the epoxide. Previous studies examining the N-terminal valine adducts of BMO (11) have verified that both regioisomers and their diastereomers are formed in vitro. Whether both regioisomers or one regioisomer and its diastereomer are formed in these studies cannot be determined using the current methodology.

Discussion Recent advances in mass spectrometry provided the incentive to examine the feasibility of using these procedures to examine the global modification of R- and β-globin by BMO adduct formation. The first goal was to determine the number of adducts formed on each globin chain after incubation of mouse erythrocytes with BMO. The ES/MS method developed for these studies is extremely rapid (intersample analysis time of