New Biomarkers for Monitoring the Levels of Isothiocyanates in

Feb 4, 2010 - Department of Environmental Health Sciences, School of Public Health ... 1440 Canal Street, Suite 2100 (SL-29), New Orleans, Louisiana 7...
0 downloads 0 Views 436KB Size
756

Chem. Res. Toxicol. 2010, 23, 756–765

New Biomarkers for Monitoring the Levels of Isothiocyanates in Humans Anoop Kumar and Gabriele Sabbioni* Department of EnVironmental Health Sciences, School of Public Health and Tropical Medicine, Tulane UniVersity, 1440 Canal Street, Suite 2100 (SL-29), New Orleans, Louisiana 70112 ReceiVed October 26, 2009

Isothiocyanates (ITCs) found in cruciferous vegetables have demonstrated cancer preventive activity in animals, and increased dietary intake of ITCs has been shown to be associated with a reduced cancer risk in humans. ITCs exert their cancer chemopreventive action by multiple mechanisms, for example, by modulating the activities of phase I and phase II drug metabolism enzymes, by inhibiting the cell cycle and histone deacetylase, and by causing apoptotic cell death. In cells, protein adducts account for most of total cellular ITC uptake at 4 h after treatment. The time course of this protein binding correlates well with the inhibition of proliferation and the induction of apoptosis. Animal studies have shown that glutathione conjugates are the major products of ITCs. The major urinary excretion products of ITCs in human are N-acetyl cysteine conjugates. Urinary metabolites might provide the exposure history of the last 24 h, if the urine of the full next day is collected. However, this is not feasible in large epidemiological studies. Furthermore, the mercapturic acids of ITC are not stable. Therefore, stable biomarkers are needed that reflect a larger time span of the ITC exposure history. We developed a method to determine stable (not cysteine adducts) reaction products of ITCs with albumin and hemoglobin in humans and mice. We reacted albumin with the ITCs: benzyl isothiocyanate (BITC), phenylethyl isothiocyanate (PEITC), sulforaphane (SFN), and allyl isothiocyanate (AITC). After enzymatic digestion, we found one major product with lysine using LC-MS/MS. The identity of the adducts was confirmed by comparing the analyses with synthetic standards: N6-[(benzylamino)carbonothioyl]lysine (BITC-Lys), N6-{[(2phenylethyl)amino]carbonothioyl}lysine (PEITC-Lys), N6-({[3-(methylsulfinyl)propyl]amino}carbonothioyl)lysine (SFN-Lys), and N6-[(allylamino]carbonothioyl]lysine (AITC-Lys). The adduct levels were quantified by isotope dilution mass spectrometry using the corresponding new ITC-[13C615N2]lysines as internal standards. The applicability of the method was tested for biological samples obtained from different experiments. In humans consuming garden cress, watercress, and broccoli and/or in mice exposed chronically to N-acetyl-S-{[(2-phenylethyl)amino]carbonothioyl}-L-cysteine, albumin and hemoglobin adducts were found. BITC-Lys, PEITC-Lys, and SFN-Lys released after enzymatic digestion of the proteins were quantified with LC-MS/MS. This new method will enable quantification of ITC adducts in blood proteins from large prospective studies about diet and cancer. Protein adducts are involved in the chemopreventive effects of ITCs. Therefore, blood protein adducts are a potential surrogate marker for the effects of ITCs at the cellular level. This new technique will improve the assessment of ITC exposure and the power of studies on the relationship between ITC intake and cancer. Introduction Isothiocyanates (ITCs)1 occur as glucosinolates (GLs) in cruciferous vegetables (1, 2). They occur in all parts of these plants but in different concentrations and profiles. ITCs are released from GLs by the enzyme myrosinase (3), which occurs in the same plants, separated cellularly from the GLs (4). Myrosinase mixes with the GLs and hydrolyzes the GLs when tissues of plants are damaged or chewed (5) (Figure 1). After * To whom correspondence should be addressed. Tel/Fax: 504-988-2771. E-mail: [email protected]. 1 Abbreviations: AITC, allyl isothiocyanate; AcLys, NR-acetyl-L-lysine; AITC-Lys, N6-[(allylamino]carbonothioyl]lysine; BITC, benzyl isothiocyanate; BITC-Cys, S-[(benzylamino)carbonothioyl]cysteine; BITC-Lys, N6[(benzylamino)carbonothioyl]lysine; BocLys, NR-(tert-butoxycarbonyl)-Llysine; BocCys, NR-(tert-butoxycarbonyl)-L-cysteine; CE, collision energy; DP, declustering potential; GL, glucosinolate; GSH, glutathione; Hb, hemoglobin; ITC, isothiocyanate; LOD, limit of detection; LOQ, limit of quantitation; PEITC, phenylethyl isothiocyanate; PEITC-AcCys, N-acetylS-{[(2-phenylethyl)amino]carbonothioyl}cysteine; PEITC-Lys, N6-{[(2phenylethyl)amino]carbonothioyl}lysine; SFN, D,L-sulforaphane; SFN-Lys, N6-({[3-(methylsulfinyl)propyl]amino}carbonothioyl)lysine.

cooking, the myrosinase is deactivated (4). The release of the ITCs from GLs in the human can be performed by the gut microflora (6). Naturally occurring ITCs found in cruciferous vegetables have demonstrated cancer preventive activity in animals, and increased dietary intake of ITCs has been shown to be associated with a reduced cancer risk in humans (6, 7) (Figure 1). ITCs exert their cancer chemopreventive action by modulating the activities of phase I and phase II drug metabolism enzymes (6, 8-11). ITCs and their thiol conjugates inhibit the cell cycle and cause apoptotic cell death, possibly by activation of vital signal transduction pathways (12, 13). ITCs can induce cellular oxidative stress by rapidly conjugating and thus depleting cells of glutathione (GSH) (14, 15). ITC conjugates with thiols, including the thiols in GSH and cellular proteins, are reversible (16). It has been shown with radiolabeled phenethyl isothiocyanate (PEITC) and SFN that the initial conjugation predominantly occurs with cellular GSH (12). With increasing time, protein binding gradually becomes the major reaction, at least in part because of dissociation of ITC from

10.1021/tx900393t  2010 American Chemical Society Published on Web 02/04/2010

New Biomarkers for Monitoring ITC LeVels in Humans

Chem. Res. Toxicol., Vol. 23, No. 4, 2010 757

Figure 1. Release of BITC, PEITC, AITC, and SFN from the corresponding GLs.

the unstable adducts of ITC with the thiol group of GSH (12). Eventually, proteins are the major binding sites of ITCs inside cells. For example, PEITC-protein adducts account for 87% of total cellular ITC uptake after 4 h of treatment. The time course of this protein binding correlated well with the inhibition of proliferation and the induction of apoptosis. This suggests that cellular protein adducts of ITC may be an early event for apoptosis induction (12). Therefore, biomarkers are needed that show the presence of stable reaction products with proteins. Animal studies have shown that following first pass metabolism, GSH conjugates are the major products of ITCs (reviewed in refs 17 and 18). The major urinary excretion products of GLs and breakdown products in human are N-acetyl cysteine conjugates, which have been used as urinary biomarkers of exposure to dietary GLs (reviewed in ref 18). Reaction products of thiols with ITCs are not stable (19-22), since an ITC molecule may be regenerated and then react with other nucleophiles. Most epidemiological studies on the relation of diet and cancer have relied on the information collected with questionnaires to monitor the food intake (7). Questionnaires suffer from the disadvantage of information bias, especially recall bias, which can lead to inaccurate exposure estimation. Urinary metabolites provide the exposure history of the last 24 h, if the urine of the full next day is collected. However, this is not very practicable in large epidemiological studies. Furthermore, the mercapturic acids of ITC in urine can react with other nucleophiles (21). Therefore, stable biomarkers are needed that reflect a larger time span of the ITC exposure history. Thus, we propose to determine stable (not cysteine adducts) reaction products of ITCs with albumin and hemoglobin (Hb) in human. Stable Hb adducts and albumin adducts have a lifetime of 120 days and a half-life of 20-25 days, respectively (23). This new method would enable one, for example, to quantify ITC adducts in blood proteins from large prospective studies about diet and cancer. Protein adducts have been discovered to be involved in the chemopreventive effects of ITCs. Therefore, blood protein adducts are a potential surrogate marker for the effects of ITCs at the cellular level. This new technique will improve the assessment of ITC exposure and of the power of studies on the relationship between ITC intake and cancer.

Materials and Methods Chemicals. Methanol (A454-4) for sample preparation, methanol (Optima, A456-4) for LC-MS/MS, and Amicon Ultra Centrifuge filters with 10 and 30 kDa (UFC803096) cutoffs were obtained from Fisher Scientific (New Jersey). Pronase E from Streptomyces griseus (#81748), formic acid (MS grade, #94318), ammonium formate (#17843), NR-(tert-butoxycarbonyl)-L-lysine (BocLys) (#15456), N-(tert-butoxycarbonyl)-L-cysteine (BocCys) (#15411), sodium sulfide hydrate puriss p.a. (32-38%) (#71975), allyl isothiocyanate (AITC) (#36682), and sodium hydroxide (#71687) were purchased from Fluka (Buchs, Switzerland). The reagent grade

tris(hydroxymethyl)amino-methane (#252859), sodium phosphate monobasic monohydrate (#P0662), sodium thiocyanate (#251410), [2H6]DMSO, dry 1,4-dioxane (#296309), L-Lys-[13C615N2]hydrochloride (#608041), NR-acetyl-L-lysine (Ac-Lys) (A2010), L-valine, L-aspartic acid, copper(II) carbonate basic (#207896), benzyl isothiocyanate (BITC) (#252492), S-[(benzylamino)carbonothioyl]cysteine (BITC-Cys) (#B7031), phenethyl isothiocyanate (PEITC) (#53731), trifluoroacetic acid (TFA) (#T62200), human serum albumin (#A1653), acylase I from porcine kidney (#A3010), and water for LC-MS/MS (Chromasolv; #39253) were purchased from Sigma-Aldrich (St. Louis, MO). A Chromabond C18ec cartridge (500 mg/3 mL) was purchased from Macherey-Nagel. D,LSulforaphane (S699115) (SFN) was purchased from Toronto Research Chemicals, Inc. (North York, ON, Canada). Instrumentation. An API 4000Q Trap (Applied Biosystems, Foster City, CA) mass spectrometer interfaced to a HPLC (Shimadzu Prominance 20AD) was used for the LC-MS/MS analyses. Chromatographic separation for the LC-MS/MS analyses was achieved with a Luna 3 µ C18(2) (100 Å, 150 mm × 2 mm, 3 µm) (Phenomenex Inc., Torrance, CA) attached to a C18 guard column (AJO-4287; 4 mm L × 3 mm ID; Phenomenex Inc.) and a gradient system with solvents A (10 mM ammonium formate) and B (methanol) at a flow rate of 0.2 mL/min. HPLC analyses to determine the purity of the compounds were performed on a Lichrosphere RP18 (125 mm × 4.6 mm, 5 µm) column with a 20 min 30-80% methanol gradient in ammonium acetate (10 mM), a flow rate of 1.0 mL/min, and λ ) 250 nm using a Hewlett-Packard 1100 system with a quaternary HPLC pump and a photodiode array detector. A Beckman Coulter DU 800 spectrophotometer was used for protein determination. Centrifugations were performed on a Beckman Coulter Allegra X-22R centrifuge equipped with a SX4250 swing out bucket rotor. NMR spectra were recorded on a Bruker AC 500 instrument with [2H6]DMSO as the solvent and as the internal standard (IS). The raw NMR data were processed with the program MestRe-C (Magnetic Resonance Companion, J. C. Cobas, J. Cruces, and F. J. Sordina, Departamente de Quimica Organica, Universidad de Santiago de Compostela, 15706 Santjago de Compostela, Spain). Animals. Blood samples were obtained from Dr. Stephen Hecht (University of Minnesota). The animals were part of a large experiment. Five to six week old A/J mice received eight treatments of a mixture of 4-(methylnitrosamino)-1-(3-pyridyl)-1-butanone (NNK) plus benzo(a)pyrene (BaP) (2 µmol of each) twice a week. The mice of group 6 received N-acetyl-S-(N-2-phenethylthiocarbamoyl)-L-cysteine (PEITC-AcCys) (5 µmol/g diet) and indole-3carbinol (I3C) (5 µmol/g diet) 1 week after the last carcinogen dose. The mice of group 11 were treated the same way, but myoinositol (MI) (56 µmol/g diet) was given in addition to PEITC-AcCys and I3C 1 week after the last carcinogen dose. The animals were sacrificed after 19 weeks. Pooled frozen blood samples of group 6 and group 11 were sent to Tulane University for work-up and chemical analyses. Fresh blood from control mice was obtained from Dr. C. Miller (Tulane University, New Orleans). Isolation of Albumin and Hb from Frozen Whole Blood Samples of Mice. The isolation procedure was used with a modification of the method reported by Funk et al (24). Blood (200 µL) was diluted with water (370 µL) followed by dropwise addition of cold ethanol (430 µL) while stirring for 2-3 min to precipitate

758

Chem. Res. Toxicol., Vol. 23, No. 4, 2010

the Hb. The supernatant was separated from Hb by centrifugation at 4000 rpm for 5 min. The Hb pellet was washed twice with 10 mM Tris-HCl (pH 7.5, 2 mL), followed by sequential washing with cold ethanol (2 mL), ethanol/ethyl ether (3:1, v/v, 1.5 mL), and ethyl ether (2 mL). Hb was stored at -20 °C after it was dried at room temperature in a desiccator. The purity of the Hb was checked with HPLC. An HPLC-UV method was developed to check the purity for Hb. Hb was dissolved in 50 mM ammonium bicarbonate buffer (pH 9.0). HPLC-UV analysis was performed on a Phenomenex Jupitor (250 mm × 4.6 mm, 5 µm, 300 Å) column with a 70 min gradient run: 35 (0 min), 35 (10 min), 50 (60 min), 35 (65 min), and then 35% (70 min) acetonitrile in 0.1% TFA (B) and 0.1% TFA in water (A) [flow rate of 1 mL/min, tR (heme) ) 3.2 min, tR (βa chain) ) 33.9 min, and tR (Ra chain) ) 39.5 min] (25). The purity of Hb was measured at 210 nm. In the supernatant, 600 µL of cold ethanol was added dropwise while it was constantly stirred to precipitate albumin. Precipitated albumin was dissolved in 10 mM Tris-HCl (pH 7.5, 2 mL) and centrifuged at 4000 rpm for 5 min to remove further Hb. The solution was loaded on a prewashed Amicon ultracentrifugal tube (10 kDa molecular mass cutoff) and centrifuged with 4000 rpm for 40 min at 4 °C. The procedure was repeated after 4 mL of deionized water was added. The residue was dissolved in 10 mM sodium phosphate buffer (1 mL, pH 7.0) for albumin concentration determination. The purity of the isolated albumin was controlled by HPLC on a Phenomenex Jupitor (250 mm × 4.6 mm, 5 µm, 300 Å) column with 0-80% methanol gradient in 0.1% TFA over 20 min and 5 min with 80% methanol in 0.1% TFA. Albumin eluted at 21 min using a flow rate of 1 mL/min. The purity of albumin was >99.5% at λ ) 220 nm. The retention time corresponded to commercially available albumin (A3139, Sigma Aldrich) from mouse serum. Human Sample. In a preliminary experiment, blood samples were analyzed after cruciferous vegetable intake from one subject. Garden cress (60 g) and water cress (100 g) were eaten as salad. Blood samples were collected 1 day after the meal. Garden cress and watercress were not part of the normal diet of the studied subject. Broccoli (ca. 40 g in salads) was part of the regular diet of the subject. The day before blood donation, 300 g of raw broccoli was eaten. Plasma and erythrocytes were separated with centrifugation. The erythrocytes were washed three times with saline water (0.9% NaCl). Isolation of Albumin and Hb from Human Blood. Blood was centrifuged to separate plasma from the erythrocytes. The erythrocytes were washed three times with an equivalent volume of 0.9% NaCl in water. Albumin was isolated from human plasma as described earlier (26) using HiTrap Blue HP affinity column (1 mL volume, GE Life Sciences, Piscataway, NJ) (#17-0412-01). The column was first equilibrated with 6.0 mL of binding buffer (20 mM NaH2PO4, pH 7.0). The plasma sample (0.5 mL) was diluted with binding buffer (0.5 mL), applied on a pre-equilibrated column, and washed with binding buffer (6 mL). The adsorbed albumin was eluted with 6 mL of elution buffer (20 mM NaH2PO4 + 2 M NaCl, pH 7.0). The HiTrap columns were regenerated and equilibrated by washing with 10 mL of 50 mM Tris-HCl, pH 7.4, + 0.2 M NaSCN followed by 6.0 mL of binding buffer. Purified fractions were concentrated in Amicon ultracentrifugal filter tube (30 kDa molecular mass cutoff; 4 mL) by centrifuging with 4000 rpm at 4 °C for 10-15 min, followed by desalting with water (3 × 4 mL). Samples were redissolved in 10 mM sodium phosphate buffer (pH 7.0). Hb was isolated from erythrocyte (200 µL) as described earlier (27). Erythrocytes were lysed with 4 volumes of water on ice for 30 min. The cell debris was separated with centrifugation. Cold ethanol was added (2 mL) dropwise to the supernatant to precipitate the Hb. Precipitated Hb was sequentially washed with cold ethanol (2 × 3 mL), 1.5 mL (3:1, ethanol/ethyl ether v/v), and finally with 2 mL of ethyl ether. Hb was stored at -20 °C after it was dried at room temperature for 1 h. N6-[(Benzylamino)carbonothioyl]lysine (BITC-Lys). BITC (149 µL, 1 mmol) in 1,4-dioxane was added dropwise to BocLys

Kumar and Sabbioni (246 mg, 1 mmol) in sodium bicarbonate buffer (3 mL, 0.25 M, pH 8.4) at room temperature. After 1 h at 80 °C, the reaction mixture was collected and extracted with ethyl acetate (3 × 5 mL). The organic phase was discarded, and the pH of the water phase was adjusted to pH 4. The acidified water phase was extracted with ethyl acetate (3 × 5 mL). The combined organic extracts were dried over anhydrous sodium sulfate, filtered, and evaporated under reduced pressure. The residue was covered with trifluoroacetic acid (TFA). After 1 h, the TFA was evaporated under reduced pressure, and the residue was dried using a high vacuum pump. After recrystallization in ethyl acetate, a white powder was yielded (210.7 mg, 71.3%). The purity was 99% according to HPLC analysis (254 and 244 nm). 1 H NMR ([2H6]DMSO) δ ppm: 8.02 (br s, 1H, NH), 7.78 (br s, 1H, NH), 7.29 (m, 4H, aromatic H), 7.24 (m, 1H, aromatic H in para position), 4.69 (s, 2H, C6H5CH2), 3.42 [m, 5H, CH(R), CH2(ε), NH2], 1.77 + 1.70 [m, 2H, CH2(δ)], 1.48 [m, 2H, CH2(β)], 1.39 [m, 2H, CH2(γ)]. ESI-MS (+) m/z: 613.1 [2 M + Na]+, 591.1 [2 M + H]+, 318.0 [M + Na]+, 296.0 [M + H]+. MS/MS of m/z 296.0: 189.0 (100%) [M - C7H9N]+, 130.0 (75%) [M C8H10N2S]+, 116.2 (75%) [M - C7H12N2O2S]+, 91.0 (57%) [M C7H13N3O2S]+. ESI-MS (-) m/z: 611.0 [2 M + Na]-, 589.1 [2 M - H]-, 316.0 [M + Na]-, 294.0 [M - H]-. MS/MS of m/z 294.0: 145.0 (100%) [M - C8HNS]-, 130.8 (20%) [M - C8H7N2S]- . N6-[(Benzylamino)carbonothioyl]-[13C615N2]lysine (BITC13 [ C615N2]Lys). One molequivalent of sodium hydroxide and then one molequivalent of copper(II) carbonate were added to L-lysine[13C615N2] hydrochloride (10 mg, 0.06 mmol) dissolved in water (3 mL). After 20 min at 100 °C, the reaction mixture was cooled on an ice bath. The formed black precipitate was separated from the blue supernatant by centrifugation. BITC (one molequivalent) in 1,4-dioxane (0.5 mL) was added dropwise to the blue solution. After 3 h at room temperature, sodium sulfide (16 mg, 0.2 mmol) was added to the reaction mixture. Stirring for 10 min was followed by centrifugation. The supernatant was washed with dichloromethane (3 × 5 mL) and then acidified to pH 4. The acidic solution was loaded in a preconditioned Chromabond C18ec cartridge. After it was washed with water, the compound was eluted with 0.1% formic acid/methanol (50:50 v/v, 6 mL). The eluate was collected and evaporated under reduced pressure. The residue was washed with n-hexane and dichloromethane and dried using a high vacuum pump. A yellowish powder of BITC-[13C615N2]Lys (12.2 mg, 67.6%) was obtained. ESI-MS (+) m/z: 326.1 [M + Na]+, 304.2 [M + H]+. MS/MS of 304.2: 196.9 (100%) [M - C7H9N]+ 155.2 (30%) [M C8H7NS]+, 149.9 (35%) [M - C6H14N2S2]+, 137.0 (75%) [M C8H10N2S]+, 108.1 (75%) [M - C7H12N2O2S]+, 91.1 (57%) [M C7H13N3O2S]+. ESI-MS (-) m/z: 302.2 [M - H]+. MS/MS of 302.2: 194.9 (10%) [M - C7H9N], 153.0 (100%) [M - C8H7NS]-, 134.9 (14%) [M - C8H10N2S]-. N6-{[(2-Phenylethyl)amino]carbonothioyl}lysine (PEITC-Lys). The same procedure as described for the synthesis of BITC-Lys was followed. Instead of BITC, PEITC was taken. A white powder with 75% yield was obtained. 1 H NMR ([2H6]DMSO) δ ppm: 7.55 (br s, NH), 7.51 (br s, NH), 7.29 (“t”, J ) 7.4, 7.3 Hz 2H, aromatic H in meta position), 7.23 (“d”, J ) 7.3 Hz, 2H, aromatic H in ortho position), 7.19 (“t”, J ) 7.4 Hz 1H, aromatic H in para position), 3.59 (m, 2H, C6H5CH2CH2), 3.53 (m, 1H, CHCOO), 3.34 (m, 2H, CH2NH), 2.80 (t, J ) 7.5 Hz, 2H, C6H5CH2), 1.77 + 1.70 [m, 2H, CH2(δ)], 1.47 [m, 2H, CH2(β)], 1.38 [m, 2H, CH2(γ)]. ESI-MS (+) m/z: 641.2 [2 M + Na]+, 619.2 [2 M + H]+, 332.1 [M + Na]+, 309.9 [M + H]+. MS/MS of m/z 309.9: 189.1 (85%) [M - C8H11 N]+, 147.0 (35%) [M - C9H9NS]+, 130.1 (55%) [M - C9H12N2S]+, 122.1 (100%) [M - C7H14N2O2S]+, 105.1 (30%) [M - C7H15N3O2S]+. ESI-MS (-) m/z: 639.2 [2 M + Na]-, 617.0 [2 M - H]-, 329.9 [M + Na]-, 308.1 [M - H]-. MS/MS of 308.1: 187.0 (9%) [M C8H11N]-, 145.0 (100%) [M - C9H9NS]-. PEITC-[13C615N2]Lys. The same procedure used to synthesize BITC-[13C615N2]Lys was followed and yielded PEITC-[13C615N2]Lys (white solid, 9.4 mg, 45.6% yield, purity 99% at 250 nm). ESI-

New Biomarkers for Monitoring ITC LeVels in Humans MS (+) m/z: 340.2 [M + Na]+, 318.2 [M + H]+. MS/MS of m/z 318.2: 197.1 (82%) [M - C8H11N]-, 137.0 (55%) [M C9H12N2S]+, 122.1 (100%) [M - C7H14N2O2S]+, 104.8 (30%) [M - C7H15N3O2S]+. ESI-MS (-) m/z: 632.8 [2 M - H]-, 316.1 [M - H]-. MS/MS of 316.1: 185.0 (9%) [M - C8H11N]-, 153.0 (100%) [M - C9H9NS]-, 135.0 (8%) [M - C9H12N2S]-. N6-[(Allylamino)carbonothioyl]lysine (AITC-Lys). AITC (1 mmol) in 1,4-dioxane (0.5 mL) was added dropwise to AcLys (188 mg, 1 mmol) dissolved in sodium bicarbonate (0.25 M, pH 8.4, 5 mL) at 80 °C. After 1 h, the reaction mixture was cooled down on a ice bath, extracted with ethyl acetate (3 × 5 mL), and then acidified with 2 M HCl to pH 4. AITC-AcLys was purified using a Chromabond C18ec cartridge. Before the sample was loaded, the cartridge was conditioned with methanol (3 mL) and equilibrated with 0.1% formic acid (3 mL, pH 4). AITC-AcLys was eluted with methanol (6 mL) and concentrated to dryness. The residue was dissolved in 50 mM sodium phosphate buffer (pH 7.4, 3 mL) and incubated (37 °C, 24 h) with 25 mg of acylase I from porcine kidney (#A3010 Sigma). The reaction mixture was centrifuged and acidified with 2 M HCl to pH 4.0 before solid-phase purification as described above. AITC-Lys was eluted with methanol. Precipitated acylase I was removed by centrifugation, and the supernatant was evaporated at reduced pressure and yielded AITC-Lys as a white solid product (112.7 mg, 46.0%). HPLC-UV analysis was performed on a Lichrosphere RP18 (125 mm × 4 mm, 5 µm) column with a 15 min, isocratic run: 20% methanol in 0.1% formic acid [flow rate of 1 mL/min, tR (AITC-AcLys) ) 5.9 min, and tR (AITC-Lys) ) 3.1 min]. The purity for AITC-Lys was found to be 99% at λ ) 220 nm. 1 H NMR ([2H6]DMSO) δ ppm: 7.93 (br s, 3H, NH, NH, COOH), 5.84 (m, 1H, CH2CHCH2N), 5.14 (d, J ) 17.2 Hz, 1H, trans CH2dCHCH2N), 5.04 (dd, J ) 10.4 Hz, 1.5, 1H, cis CH2dCHCH2N), 4.04 (s, 2H, CHCH2N), 3.42 [m, 2H, CH2(ε)], 3.35 (m, 1H, CHCOO), 1.77 + 1.70 [m, 2H, CH2(δ)], 1.46 [m, 2H, CH2(β)], 1.37 [m, 2H, CH2(γ)]. ESI-MS (+) m/z: 736.3 [3 M + H]+, 513.3 [2 M + Na]+, 491.1 [2 M + H]+, 268.1 [M + Na]+, 246.0 [M + H]+. MS/MS of m/z 246.0: 201.1 (100%) [M CHO2]+, 189.3 (10%) [M - C3H7N]+, 147.2 (5%) [M - C4H5NS]+, 143.2 (22%) [M - C4H9NO2]+, 129.9 (32%) [M - C4H8N2S]+, 116.8 (30%) [M - C6H11NO2]+, 58.2 (17%) [M - C7H12N2O2S]+. ESI-MS (-) m/z: 511.0 [2 M + Na]-, 489.0 [3 M - H]-, 489.0 [2 M - H]-, 266.1 [M + Na]-, 244.0 [M - H]-. MS/MS of 244.0: 186.6 (5%) [M - C3H7N]+, 145.0 (100%) [M - C4H5NS]-, 56.0 (7%) [M - C7H12N2O2S]-. AITC-[13C615N2]Lys. L-Lysine [13C615N2]hydrochloride (10 mg, 0.06 mmol) was dissolved in water (3 mL), and one molequivalent of sodium hydroxide was added followed by one molequivalent of cupric(II) carbonate and heated to 100 °C. After 20 min, the reaction was cooled down on an ice bath, and a black precipitate was eliminated by centrifugation. AITC (one molequivalent in 1,4dioxane) was added dropwise to the blue supernatant. After 4 h at room temperature, sodium sulfide (16 mg, 0.2 mmol) was added to the reaction mixture and stirred for 10 min followed by the centrifugation. The supernatant was washed with dichloromethane (3 × 3 mL), and then, the aqueous phase was acidified with 2 M HCl up to pH 4. The acidified aqueous phase was loaded on a preconditioned Chromabond C18ec cartridge, and the retained compound was eluted with 0.1% formic acid/methanol (50:50 v/v, 6 mL). Fractions were collected and evaporated under reduced pressure. The final compound was washed with n-hexane and dichloromethane and yielded AITC-[13C615N2]Lys (18.9 mg, 93.2%). HPLC analysis showed a purity of 99% at λ ) 244 nm. ESI-MS (+) m/z: 529.5 [2 M + Na]+, 507.4 [2 M + H]+, 254.0 [M + H]+. MS/MS of m/z 254.0: 58.1 (66%) [M - C7H12N2O2S]+, 89.9 (90%) [M - C7H12N2S]+, 137.0 (100%) [M - C4H8N2S]+, 149.9 (75%) [M - C4H7NO2]+, 197.1 (90%) [M - C3H7N]+. ESIMS (-) m/z: 527.3 [2 M + Na]-, 504.8 [2 M - H]-, 252.1 [M H]-. MS/MS of 252.1; 59.0 (17%) [M - C7H12N2O2S]-, 134.7 (7%) [M - C4H8N2S]-, 153.0 (100%) [M - C4H5NS]- .

Chem. Res. Toxicol., Vol. 23, No. 4, 2010 759 N6-({[3-(Methylsulfinyl)propyl]amino}carbonothioyl)lysine (SFNLys). AcLys (60 mg, 0.32 mmol) was dissolved in sodium bicarbonate (0.25 M, pH 8.4, 5 mL) buffer followed by dropwise addition of SFN (35 µl, 0.20 mM) in 1,4-dioxane (0.5 mL) at 80 °C with continuous stirring for 1 h. After 1 h, the reaction mixture was cooled down on a ice bath, extracted with ethyl acetate (3 × 5 mL), and then acidified with 2 M HCl to pH 4. SFN-AcLys was purified using a Chromabond C18ec cartridge. Before the sample was loaded, the cartridge was conditioned and equilibrated with 3 mL of methanol and 0.1% formic acid (pH 4), respectively. SFNAcLys was eluted with 6 mL of methanol and concentrated to dryness. The residue was dissolved in 50 mM sodium phosphate buffer (pH 7.4, 3 mL) and incubated (37 °C, 60 h) with 50 mg of acylase I from porcine kidney (#A3010 Sigma). The reaction mixture was centrifuged and acidified with 2 M HCl to pH 4.0. AITC-Lys was isolated using solid-phase extraction as described above. Precipitated acylase I was removed by centrifugation. The supernatant was dried using at reduced pressure and yielded SFNLys (72 mg, 69.9%) as a white solid. HPLC-UV analysis was performed on a LiChrosphere RP18 (125 mm × 4 mm, 5 µm) column with a 14 min: gradient run 5% (1 min), 10% (7 min), then hold 10% (11 min), 5% (12 min) methanol in 10 mM ammonium acetate [flow rate of 1 mL/min, tR (SFN-NR-AcLys) ) 9.2 min, tR (SFN-Lys) ) 6.3 min]. The purity for SFN-Lys was 99% at λ ) 220 nm. 1 H NMR (COSY) ([2H6]DMSO) δ ppm: 8.3 (br s, 3H, NH, NH, COOH), 3.39 (m, CH2N of SFN), 3.31 (m, CH2N of Lys, and CHCOO of Lys), 2.76 (m, 1H, CH2SO), 2.66 (m, 1H, CH2SO), 2.6 (s, 3H, CH3SO), 1.72 [m, 1H, CH2(δ)], 1.60 [m, 1H of CH2(δ) + 4H of SOCH2CH2CH2CH2], 1.44 [m, 2H, CH2(β)], 1.35 [m, 2H, CH2(γ)]. 13C NMR ([2H6]DMSO) δ ppm: 173.7 and 171.2 (COOH, CS), 54.06 (CHCOOH), 52.90 (CH2SO), 43.2 and 42.8 (CH2NHCSNHCH2), 37.95 (CH3SO), 30.84 [CH2(β)], 28.29 and 28.05 [CH2(δ) and CH2CH2NHSO], 22.42 [CH2(γ)], 19.51 (CH2CH2SO). ESI-MS (+) m/z: 669.2 [2 M + Na]+, 646.0 [2 M - H]+, 346.3 [M + Na]+, 324.1 [M + H]+. MS/MS of 324.1: 136.0 (100%) [M - C7H12N2O2S]+, 147.2 (53%) [M C6H11NOS2]+, 178.2 (13%) [M - C6H14N2O2]+, 260.3 (14%) [M - CH3OS]+. ESI-MS (-) m/z: 645.1 [2 M - H]-, 344.1 [M + Na]-, 322.0 [M - H]-. MS/MS of 322.2; 145.0 (100%) [M C6H11NOS2]-, 186.9 (13%) [M - C5H13NOS]- . SFN-[13C615N2]Lys. L-Lys [13C615N2]hydrochloride (10 mg, 0.06 mmol) was dissolved in water (3 mL), and one molequivalent of sodium hydroxide was added followed by one molequivalent of cupric(II) carbonate. After 20 min at 100 °C, the reaction was cooled down on an ice bath, and the black precipitate was eliminated by centrifugation. SFN (one molequivalent in 1,4-dioxane) was added dropwise to the blue supernatant. The reaction was carried out at room temperature for 4 h. Sodium sulfide (16 mg, 0.2 mmol) was added to the reaction mixture and stirred for 10 min at room temperature. After centrifugation, the supernatant was washed with dichloromethane (3 × 3 mL), acidified with 2 M HCl to pH 4, and loaded on a preconditioned Chromabond C18ec cartridge (500 mg/3 mL). The compound was eluted with 0.1% formic acid/methanol (50:50 v/v, 6 mL). After evaporation under reduced pressure, the residue was washed with n-hexane and dichloromethane. After this was dried, SFN-[13C615N2]Lys (16.1 mg, 81%) was obtained. ESI-MS (+) m/z: 354.2 [M + Na]+, 332.0 [M + H]+. MS/MS of 332.0: 136.0 (100%) [M - C7H12N2O2S]+, 155.0 (46%) [M C6H11NOS2]+, 178.3 (13%) [M - C6H14N2O2]+, 268.1 (14%) [M - CH3OS]+. ESI-MS (-) m/z: 330.2 [M - H]-. MS/MS of 330.2: 153.0 (100%) [M - C6H11NOS2]-, 194.9 (10%) [M - C5H13NOS]-. N-[(Benzylamino)carbonothioyl]valine (BITC-Val). BITC (149 µL, 1 mmol) in 1,4-dioxane (0.5 mL) was added dropwise to L-valine (116 mg, 1 mmol) in sodium bicarbonate (3 mL, 0.25 M, pH 8.4). After 30 min at 80 °C, the cooled reaction mixture was extracted with ethyl acetate (three times). The organic phase was discarded, and the pH of the aqueous phase was adjusted to pH 4. The acidified water phase was extracted with ethyl acetate. The organic extract was dried over anhydrous sodium sulfate, filtered, and evaporated under reduced pressure. The residue was dried using

760

Chem. Res. Toxicol., Vol. 23, No. 4, 2010

a high vacuum pump. A yellowish oily product of BITC-Val (43 mg, 16%) was obtained after recrystallization with ethyl acetate. The purity was 99% according to HPLC analysis (200 and 254 nm). The compound was not stable at -20 °C. Gradually, the benzylthiohydantoin-valine (BITC-Hyd-Val) formed. 1 H NMR ([2H6]DMSO) δ ppm: 12.5 (br s, COOH), 8.02 (t, J ) 5.1 Hz, CH2NH), 7.51 (“d”, J ) 5.7 Hz, CHNH), 7.32 (m, 4H, aromatic H), 7.25 (m, 1H, aromatic H in para position), 4.86 (m, 1H, CHNH), 4.68 (dd, J ) 2.2, 5.1 Hz, 2H, CH2NH), 2.13 [m, 1H, CH(CH3)2], 0.91 (d, J ) 6.8 Hz, 3H, CH3), 0.90 (d, J ) 6.9 Hz, 3H, CH3), 15% BITC-Hyd-Val was present. ESI-MS (+) m/z: 598.9 [2 M + 3Na]+, 576.9 [2 M + 2Na]+, 555.1 [2 M + Na]+, 289.1 [M + Na]+, 267.1 [M + H]+. MS/MS of 267.1; m/z: 71.8 [M C13H16N2O2S]+, 90.8 [M - C6H12N2O2S]+, 118.2 [M - C8H7NS]+. ESI-MS (-) m/z: 530.8 [2 M + H]+, 265.1 [M + H]+. MS/MS of 265.1; m/z: 115.8 (100%) [M - C8H7NS]- . BITC-Cys. BITC-Cys can be obtained commercially or from the reaction of BITC with BocCys for 1 h at room temperature in sodium bicarbonate (3 mL, 0.25 M, pH 8.4). LC-MS/MS chromatographic separation of BITC-Cys was achieved on a Luna 3 µ C18(2) with gradient system using solvents A (10 mM ammonium formate) and B (methanol) at a flow rate of 0.2 mL/min. The gradient program was as follows: 0.1 (B 20%), 3 (B 20%), 16 (B 90%), 20 (B 90%), and 21-26 min (B 20%). The retention time (tR) of BITC-Cyst was 14.1 min. The column flow was diverted away from the ESI ion source except for the time period from 10 to 17 min. The BITC-Cys was detected with MRM transition of 268.9/119.9 [M - H]- . ESI-MS (+) m/z: 293.1 [M + Na]+, 271.1 [M + H]+. MS/MS of m/z 271.1; 91.0 (100%) [M - C4H6N2O2S2]+, 122.0 (25%) [M - C8H8NS]+, 245.0 (15%) [M - OH]+. ESI-MS (-) m/z: 268.9 [M - H]-. MS/MS of m/z 268.9; 119.9 (100%) [M - C8H8NS]-. N-[(Benzylamino)carbonothioyl]aspartic Acid (BITC-Asp). BITC (149 µL, 1 mmol) in 1,4-dioxane (0.5 mL) was added dropwise to L-aspartic acid (133 mg, 1 mmol) in sodium bicarbonate (3 mL, 0.25 M, pH 8.4). After 15 min at 80 °C, the cooled reaction mixture was extracted with ethyl acetate (3 × 5 mL). The organic phase was discarded, and the pH of the aqueous phase was adjusted to pH 5. The acidified water phase was loaded on a preconditioned Chromabond C18ec cartridge and eluted with 0.1% formic acid/ methanol (50:50 v/v, 6 mL). The eluate was collected and evaporated under reduced pressure. The residue was washed with n-hexane:methyl t-butyl ether (1:5) and dried using a high vacuum pump. A whitish thick oil of BITC-Asp (210 mg, 74.4%) was obtained. The purity was 53% according to HPLC analysis (244 nm). HPLC-UV analysis was performed on a Lichrosphere RP18 (125 mm × 4 mm, 5 µm) column, with a linear methanol gradient in 0.1% formic acid: 30% (0-2 min), 30-90% (15 min), and BITCAsp and benzylthiohydantoin-aspartic acid [)1-benzyl-5-oxo-2thioxoimidazolidin-4-yl)acetic acid] (BITC-Asp-Hyd) were found to be 53 and 47% at λ ) 244 nm, respectively. LC-MS/MS chromatographic separation of BITC-Asp was achieved with a Luna 3 µ C18(2) with a gradient system using solvent A (10 mM ammonium formate) and B (methanol) at a flow rate of 0.2 mL/ min. The gradient program was as follows: 0.1 (B 20%), 3 (B 20%), 16 (B 90%), and 23 min (B 90%). The retention times (tR) of BITCAsp and BITC-Asp-Hyd were 11.9 and 15.8 min, respectively. ESIMS for BITC-Asp and BITC-Asp-Hyd showed molecular ions of m/z 281.0 and 263.0, respectively. ESI-MS (+) m/z: 565.2 [2 M + H]+, 305.0 [M + Na]+, 283.0 [M + H]+. MS/MS of m/z 283.0; 116.0 (55%) [M - C8H8N2S]+, 133.9 (100%) [M - C8H7NS]+. ESI-MS (-) m/z: 563.2 [2 M H]-, 281.0 [M - H]-, MS/MS of m/z 281.0; 132.0 (100%) [M C8H7NS]-. LC-MS/MS Method and Calibration Line for BITC-Lys, PEITC-Lys, AITC-Lys, and SFN-Lys. A Shimadzu Prominance 20AD interfaced to a API 4000Q Trap LC-MS/MS (Applied Biosystems) mass spectrometer system was used for all of the quantitative analyses. The MS parameters were optimized in the electrospray ionization mode (ESI). Parameter optimization was

Kumar and Sabbioni carried out with 100 pg/µL solution of analyte with a flow rate of 10 µL/min in negative ionization mode for BITC-Lys, PEITC-Lys, and AITC-Lys and positive ionization mode for SFN-Lys, showing corresponding peaks at m/z 294.0, 308.1, 244.1 [M - H]-, and 324.1 [M + H]+. Quantitative optimization mode was used to maximize the signal and set the maximum suitable MS parameters for the compounds. For better resolution and sensitivity of the analyte, quadrupole mass analyzers (Q1 and Q3) were set on 0.7 ( 0.1 amu resolution window. The mass spectrometer was operated in negative ionization mode with a electrospray voltage at -4500 V for negative ionization and +5500 V for positive ionization mode and a source temperature of 500 °C. Nitrogen was used as ion spray (GS1), drying (GS2), and curtain gas at 40, 45, and 10 arbitrary units, respectively. The declustering potential (DP) and collision energy (CE) for BITCLys, PEITC-Lys, AITC-Lys, and SFN-Lys were -45, -50, -40, +61 and -24, -24, -25, +23 V, respectively. The entrance potential (EP) for all compounds was -10 V. All data were processed using Analyst software 1.4.2 (Applied Biosystems/MDS Sciex). The BITC-Lys, PEITC-Lys, AITC-Lys, and SFN-Lys were detected with MRM transition of 294.0/145.0, 308.1/145.0, 244.1/ 145.0 [M - H]- and 324.1/136.0 [M + H]+, respectively, along with their corresponding stable isotope-labeled compounds as ISs PEITC-[13C615N2]Lys {316.1/153.0 [M - H]-}, BITC-[13C615N2]Lys {302.1/153.0 [M - H]-}, AITC-[13C615N2]Lys {252.1/153.0 [M - H]-}, and SFN-[13C615N2]Lys {332.1/136.0 [M + H]+}. Chromatographic separation of BITC-Lys and PEITC-Lys was achieved on a Luna 3 µ C18(2) column with a gradient system using solvent A (10 mM ammonium formate) and B (methanol) at a flow rate of 0.2 mL/min. The gradient program was as follows: 0.10 (B 20%), 3 (B 20%), 16 (B 90%), 20 (B 90%), and 21-26 min (B 20%). The retention times (tR) of BITC-Lys and PEITCLys were 13.7 and 15.2 min, respectively. The column flow was diverted away from the ESI ion source except for the time period from 8 to 18 min. The limits of detection (LOD) for BITC-Lys and PEITC-Lys were 1.69 and 1.61 fmol (on column), while the limits of quantitation (LOQ) were 18.8 and 17.9 fmol/mg albumin, respectively. SFN-Lys and AITC-Lys separation was achieved on a Luna 3 µ C18(2) with a gradient system using solvents A (10 mM ammonium formate) and B (methanol) at a flow rate of 0.2 mL/min. The gradient program was as follows: 0.10 (B 5%), 1 (B 5%), 8 (B 80%), 9 (B 80%), and 11-15 min (B 5%). The retention times (tR) of SFN-Lys and AITC-Lys were 7.71 and 8.30 min, respectively. The column flow was diverted away from the ESI ion source except for the time period from 4 to 10 min. The LODs for SFNLys and AITC-Lys were 1.54 and 4.07 fmol (on column), while the LOQs were 34.3 and 113.7 fmol/mg albumin, respectively. Calibration Line. To generate the calibration line, 9 mg of human serum albumin (Sigma) was spiked with different amounts of BITC-Lys (0.00, 0.17, 0.34, 3.34, 33.88, and 169.40 pmol), PEITC-Lys (0.00, 0.16, 0.32, 3.23, 32.35, and 161.75 pmol), SFNLys (0.00, 0.31, 0.77, 1.54, 3.09, 15.47, and 30.95 pmol), and AITCLys (0.00, 0.40, 1.01, 2.03, 4.07, 20.39, and 40.79 pmol) along with the corresponding BITC-[13C615N2]Lys (1.61 pmol), PEITC[13C615N2]Lys (1.57 pmol), SFN-[13C615N2]Lys (15.1 pmol), and AITC- [13C615N2]Lys (19.7 pmol). Samples were incubated with 3 mg of Pronase E in 50 mM sodium bicarbonate (2 mL, pH 8.9) at 37 °C overnight. Digested samples were acidified (up to pH, 4.0) using 2 M HCl. Chromabond C18ec cartridges were used for the solid-phase extraction cleanup. The column was first activated with 3 mL of methanol and then equilibrated with 3 mL of 0.1% formic acid (pH 4.0). The sample was applied on the column and subsequently washed with 0.1% formic acid (3 mL, pH 4.0). BITCLys, PEITC-Lys, SFN-Lys, and AITC-Lys were eluted with a 6 mL fraction of 80% methanol in 0.1% formic acid (pH 4.0). The fraction was evaporated up to 1 mL and injected in the LC-MS/ MS system for further analysis. The calibration lines for BITC-Lys, PEITC-Lys, SFN-Lys, and AITC-Lys were generated over the ranges of 0.00-169.40,

New Biomarkers for Monitoring ITC LeVels in Humans

Chem. Res. Toxicol., Vol. 23, No. 4, 2010 761

Figure 2. Synthesized adducts of BITC, PEITC, AITC, and SFN with lysine.

0.00-161.75, 0.00-30.95, and 0.00-40.79 pmol/9 mg albumin, respectively. The concentration levels were plotted against the peak area ratio of the analyte against the peak area of the IS (e.g., peak area ratio BITC-Lys/BITC-[13C615N2]Lys). The regression coefficients r2 ) 0.999 (BITC-Lys), 0.998 (PEITC-Lys), 0.999 (SFNLys), and 0.998 (AITC-Lys) were found using linear regression and a weighting factor of 1/x. Reaction of ITC with Albumin. The following procedure was applied for the in vitro modification of albumin with ITCs (BITC, PEITC, AITC, and SFN). Below the method is described in detail for BITC. Human serum albumin (20 mg) was dissolved in 50 mM sodium hydrogen phosphate buffer (3 mL, pH 7.4). A different molar ration of BITC (in 1,4-dioxane) (1:10, 1:1, 1:0.1, and 1:0.01) was added in the dissolved albumin solution. The reaction was kept at 37 °C overnight with continuous stirring. After it was stirred overnight at 37 °C, the reaction mixture was centrifuged to eliminate the formed precipitate. The supernatant was washed with ethyl acetate (three times), and the organic phase was discarded. The aqueous solution was dialyzed against 4 L for 4 h at 4 °C. Albumin was precipitated after cold acetone (4:1) and 25 µL of acetic acid were added. After filtration, the precipitate was reconstituted in 2 mL of water for protein concentration determination. The protein determination was done using a Coomassie protein Assay kit (23236) with the standards protocol. The protein recovery was 50-70%. The samples were digested with Pronase E and worked as described above. Human serum albumin (0.5 mg) was spiked with BITC-[13C615N2]Lys (82.4 pmol). Samples were digested and incubated with 166.6 µL of fresh Pronase E solution (1 mg/mL; 50 mM ammonium bicarbonate, pH 8.9) at 37 °C overnight. Digested samples were acidified (up to pH 4.0) with 2 M HCl, and Chromabond ec C18 was used for the solid-phase extraction procedure. The column was first activated with 3 mL of methanol and then equilibrated with 3 mL of 0.1% formic acid (pH 4.0). The sample was applied on the column and subsequently washed with a 3 mL fraction of 0, 10% methanol in 0.1% formic acid. BITC-Lys eluted with a 6 mL fraction of 80% methanol in 0.1% formic acid. After evaporation at the speed-vac to approximately 1 mL, 2 µL was taken for LC-MS/MS analysis. BITC-Cys Incubation with Albumin. Albumin (20 mg) was incubated with BITC-Cys (74.07 pmol) in 50 mM sodium phosphate buffer (1 mL, pH 7.4) and 50 mM sodium bicarbonate buffer (1 mL, pH 9) at 37 °C overnight. Pronase E (2 mg) was added and incubated overnight in a shaking bath at 37 °C. The samples were purified with Chromabond C18ec and analyzed with LC-MS/MS.

Results The potential reaction products of BITC with blood proteins present in vivo were synthesized from the single amino acids. Cysteine, the N-terminal amino acids, and lysine (Figure 2) are the most reactive amino acids. The N-terminal amino acids in human serum albumin and Hb are aspartic acid and valine, respectively. For the synthesis of adducts with the ε amino group of lysine, NR-Boc-protected lysine was used as reagent. The

Figure 3. Work-up scheme for the analysis of albumin adducts of BEITC, PEITC, AITC, and SFN. The ISssBITC-[13C615N2]Lys, PEITC[13C615N2]Lys, AITC-[13C615N2]Lys, and SFN-[13C615N2]Lysswere added prior to albumin digestion.

reaction was performed in sodium carbonate at 80 °C. The Boc group was cleaved in neat TFA. BITC-Cys can be obtained commercially or from the reaction of BITC with BocCys at room temperature in sodium bicarbonate. Valine and aspartic acid were reacted at room temperature with BITC. The loss of water and subsequent ring closure to the hydantoin could not be avoided. For PEITC, AITC, and SFN, only the adducts with lysine were synthesized (Figure 2). The same procedure used for BITC-Lys was applied for PEITC-Lys. In the case of SFN and AITC, AcLys was taken since the cleavage procedure of Boc with TFA yielded side products. The acetyl group from AcLys was cleaved with acylase I as shown previously for aflatoxin adducts of AcLys (28). AITC-Lys was synthesized recently from BocLys and AITC (21). No NMRs were reported. The corresponding isotope-labeled standards of BITC-Lys, PEITC-Lys, AITC-Lys, and SFN-Lys were synthesized with [13C615N2]Lys. The R amino group was protected as cupper complex (29). After reaction with the corresponding isothiocyanates, the copper complex was released using sodium sulfide as the reagent (30). The enantiomeric purity of all final products was not investigated. BITC was reacted in vitro with albumin overnight at physiological conditions. The proteins were digested with Pronase E following an optimized procedure developed for 4,4′methylenediphenyl diisocyanate adducts with albumin found in vivo (31). The digests were prepurified with solid-phase extraction (Figure 3). The extracts were analyzed with LC-MS/ MS (Figure 4). The major reaction product appears to be with lysine. No cysteine adducts were found. Other potential amino acid adducts with tryptophan, tyrosine, and serine were monitored, although no synthetic standard was available. The N-terminal adduct with aspartic acid appears not to be released after hydrolysis with Pronase E.

762

Chem. Res. Toxicol., Vol. 23, No. 4, 2010

Kumar and Sabbioni

Figure 4. LC-MS/MS analysis of albumin digested with Pronase E. The retention times of synthesized standards BITC-Lys, BITC-Asp, and BITCCys synthesized from BITC and the single amino acids were 13.1, 11.9, and 14.1, respectively. BITC-Ser, BITC-His, BITC-Trp, and BITC-Tyr standards were not available. The chosen mass fragmentations were deduced from the MS properties of the ITC adducts with other amino acids. MS/MS of the molecular ion of amino acid-ITC adducts yields the corresponding amino acid as a daughter ion.

For the analysis of biological samples, a LC-MS/MS method was developed to analyze the major adducts. First, in vitro experiments with albumin were performed to develop a method to quantify ITC adducts with amino acids. To detect single amino acid adducts, Pronase E was used to digest the proteins. The digests were purified using solid-phase extraction (Figure 3). The eluates were analyzed with LC-MS/MS. MS/MS of the molecular ion of amino acid-ITC adducts yields the corresponding amino acid as daughter ion. The sensitivity of the assay was tested by spiking albumin samples with synthesized standard compounds. For BITC-Lys, PEITC-Lys, and AITC-Lys, the best results were obtained with negative ESI. For SFN-Lys, the best results were obtained with positive ESI. Starting from 9 mg, the LOQs were established as follows: BITC-Lys, PEITC-Lys, AITC-Lys, and SFN-Lys as 18.8, 17.9, 113.7, and 34.3 fmol/ mg albumin. Albumin was modified in vitro with four different doses of BITC. The adduct levels of the formed BITC-Lys increased linearly with the dose (Figure 5). Albumin was modified in vitro with BITC, PEITC, AITC, and SFN. The reactions were performed with a molar ratio 1:1 of ITC to albumin. Albumin was digested with Pronase E and analyzed with LC-MS/MS. The adduct levels decreased in the following way: BITC-Lys > PEITC-Lys > AITC-Lys > SFNLys. The stability of cysteine adducts was tested. BITC-Cys was incubated at pH 7.4 with albumin overnight. After the digestion of albumin, no BITC-Cys could be found but newly formed BITC-Lys. The adduct formation in vivo and the applicability of the method was investigated from biological material obtained with different precursors of ITC: mercapturic acid adduct of PEITC and/or diet. Mice were chronically exposed to PEITC-AcCys. Hb and albumin were isolated from frozen whole blood samples. Hb could be not isolated pure from whole frozen blood. Impurities were present according to the HPLC analyses of Hb. Hb of the control mice obtained from washed erythrocytes showed the typical signals for the heme group and for the R and β chain of Hb (25). Therefore, Hb of the exposed mice was not analyzed. Albumin from exposed mice was obtained from ethanol precipitation. Albumin adducts of PEITC were

Figure 5. Albumin was modified in vitro with four different ratios of BITC/albumin ) 0.01/1, 0.1/1, 1/1, and 10/1. Albumin was digested and analyzed with LC-MS/MS.

found in the mice (Figure 6). In the pooled blood samples of the mice, 6ABCD and 11CD, 1.94 and 1.09 pmol PEITC-Lys per mg albumin were found. In the control mice, no such adducts were found. Human blood samples were obtained from one subject on three different occasions: (1) 1 day after eating 60 g of garden cress and 100 g of watercress, (2) 30 days after eating garden cress and watercress and 1 day after eating 300 g of broccoli, and (3) 4 months after eating garden cress and watercress. Watercress and garden cress were never eaten before and after the experiment. Broccoli was part of the daily diet. Albumin was purified with affinity chromatography, digested with Pronase E, and analyzed with LC-MS/MS (Figures 6 and 7). The results are summarized in Table 1. PEITC-Lys and BITC-Lys were present in large amounts in albumin 1 day after the meal. After 4 months, the PEITC-Lys and BITC-Lys levels were close to

New Biomarkers for Monitoring ITC LeVels in Humans

Chem. Res. Toxicol., Vol. 23, No. 4, 2010 763

Figure 6. LC-MS/MS analyses of albumin digests obtained from (A) mouse control, (B) mouse exposed to PEITC-AcCys, (C) human exposed to watercress, and (D) PEITC-Lys standard compound. The right panels show the chromatogram of the IS, PEITC-[13C615N2]Lys, which was added at the beginning of digestion procedure (Figure 3). PEITC-Lys and PEITC-[13C615N2]Lys were detected with the MRM transitions m/z 308.1/145.0 and 316.1/153.0, respectively.

Figure 7. LC-MS/MS analyses of albumin digests of a human subject regularly eating broccoli. The panels on the left represent the chromatograms of albumin digests from (A) commercially available human serum albumin, (B) human sample collected at day 1, (C) human sample collected at day 30, and (D) SFN-Lys standard compound. The right panels show the chromatogram of the IS, SFN-[13C615N2]Lys, which was added at the beginning of digestion procedure (Figure 3). SFN-Lys and the IS SFN-[13C615N2]Lys were detected with the MRM transitions m/z 324.1/136.0 and 332.1/136.0, respectively.

764

Chem. Res. Toxicol., Vol. 23, No. 4, 2010

Kumar and Sabbioni

Table 1. Albumin (Alb) and Hb in One Human Subject Eating Garden Cress and Watercress on Day 0 and Regularly Eating Broccoli (ca. 40 g Per Day), except for a Large Meal on Day 29 (300 g) pmol/mg Alb

pmol/mg Hb

days PEITC-Lys BITC-Lys SFN-Lys PEITC-Lys BITC-Lys SFN-Lys 1 30 121 a

2.365 0.823 0.046

1.011 0.480 0.029

2.203 3.306 1.372

0.050 0.020 a

0.023 0.021 a

0.140 0.203 a

Below the LOQ.

Figure 8. Kinetics of albumin adducts of a human subject eating 100 g of watercress and 60 g of garden cress on day 0.

the quantitation limit of the assay. SFN-Lys was present at all three time points since the subject was consuming broccoli regularly. The highest levels were found 1 day after the consumption of a larger amount of broccoli (300 g). Adducts with lysine were found in human Hb. The adduct with the N-terminal valine, BITC-Val, was not found. Probably BITCVal was not released with Pronase E. The levels of BITC-Lys and PEITC-Lys were approximately 50 times lower in Hb than in albumin (Table 1). Therefore, albumin adducts are a more sensitive marker to monitor ITC exposure. Plotting the adduct levels of PITC-Lys and or BITC-Lys against the days postexposure on a half-logarithmic scale (Figure 8) yielded a half-life of the albumin adducts of 21.3 and 23.2 days, which was close to the values of the half-life of albumin, 20-25 days (32). The adduct levels in albumin-depleted plasma were lower than in albumin. In the samples 1 day after watercress and garden cress, the adduct levels for albumin-depleted plasma proteins were 1.17 pmol/mg PEITC-Lys and 0.60 pmol/mg BITC-Lys. This is a factor two lower than the binding levels in albumin (Table 1).

Discussion This is the first report of albumin and Hb adducts of ITCs present in vivo. Albumin adducts were found in mice after chronic exposure to PEITC-AcCys, and albumin adducts were found in a human subject after eating broccoli, garden cress, and watercress. Biomarkers of ITC exposure have been

measured in the past in urine and in serum (see Introduction). For example, recently, mercapturic acids of ITCs have been quantified in urine (33) after the consumption of certain vegetables. Biologically free available ITCs have been found in urine and serum of humans (34). Protein adducts after in vitro experiments with cells have been published recently (35). In this case, cysteine adducts were postulated (13). Our experiments and the data from the literature (21) indicate that such adducts will not be stable. For this study, human albumin was isolated with Cibacron blue affinity columns (HiTrap Blue), which yields purer albumin than ammonium sulfate precipitation (36). In larger studies, we suggest using the precipitation method since many more samples could be processed per day (8 vs 30). The developed method was sensitive enough to analyze small samples of albumin (9 mg). Increasing the amount of albumin also increases the ion suppression effects for the detection of the adducts in LC-MS/MS. Therefore, to increase the sensitivity of the assay, the digests would have to be purified with specific ITC immunoaffinity columns. In a 75 kg man, approximately 100 g of albumin is present. It has been reported that from 30 g of watercress 46.6 µmol of PEITC can be released (37). Therefore, consuming 100 g of watercress could yield approximately 155 µmol of PEITC. We found 2.365 nmol PEITC-Lys per g of albumin. Thus, approximately 0.15% of the PEITC dose was bound to albumin. In humans eating 30 g of watercress, 30-67% of the PEITC conjugate (PEITC-AcCys) was found in 24 h urine (37). In humans eating 100 g of watercress, up to 1 µmol of free PEITC per L plasma could be found (34). Comparing these values from different laboratories, the levels of albumin adducts (present data) were 10 and 300 times lower than PEITC in plasma (34) and PEITC metabolites in urine (37), respectively. After 48 h, usually no urinary metabolites were found (37). The half-life of PEITC in plasma was 4.9 ( 1.1 h (34). After 48 h, approximately 10 half-lives, only 0.09% (0.9 nmol/L plasma) of the original PEITC levels (1 µmol/L plasma), will be available in plasma. Therefore, only blood protein adducts will be measurable 48 h after the last exposure to ITCs, since the half-life of the adducts is approximately 21-23 days. For chronically exposed people, we expect a steady state albumin adduct level that is 29 times higher adduct level after chronic dosing in comparison to a single dose (28). Therefore, in chronically exposed people, higher adduct levels will be found than after a single dose. Consequently, the concentration difference between the shortterm markers (urine and plasma metabolites) and the longterm markers (albumin adducts) will be smaller under chronic exposure.

Conclusions ITCs (PEITC, BITC, SFN, and AITC) react with lysine present in proteins such as albumin. Using LC-MS/MS, ITCLys can be determined as low as 18 fmol/mg albumin. Because cell protein adducts are involved in the chemopreventive effects of ITCs, blood protein adducts are probably not only a biomarker of exposure but also a potential surrogate marker for the effects of ITCs at the cellular level. This new method will enable one to quantify ITC adducts in blood proteins from large prospective studies about diet and cancer. This new technique will improve the assessment of ITC exposure and of the power of studies on the relationship between ITC intake and cancer. Acknowledgment. This research was supported by the Tulane Cancer Center. The NMR spectra were run by Dr. Qi Zhao at

New Biomarkers for Monitoring ITC LeVels in Humans

the Coordinated Instrumentation Facility of Tulane University. We are most grateful for the mice blood samples obtained from Dr. Stephen S. Hecht and Dr. Kassie Fekadu (both Masonic Cancer Center, University of Minnesota, Minneapolis, MN).

References (1) Fahey, J. W., Zalcmann, A. T., and Talalay, P. (2001) The chemical diversity and distribution of glucosinolates and isothiocyanates among plants. Phytochemistry 56, 5–51. (2) Rosa, E. A. S., Heaney, R. K., Fenwick, G. R., and Portas, C. A. M. (1997) Glucosinolates in crop plants. Hortic. ReV. 19, 99–215. (3) Shapiro, T. A., Fahey, J. W., Wade, K. L., Stephenson, K. K., and Talalay, P. (1998) Human metabolism and excretion of cancer chemoprotective glucosinolates and isothiocyanates of cruciferous vegetables. Cancer Epidemiol. Biomarkers PreV. 7, 1091–1100. (4) Bones, A. M., and Rossiter, J. T. (2006) The enzymic and chemically induced decomposition of glucosinolates. Phytochemistry 67, 1053– 1067. (5) Holst, B., and Williamson, G. (2004) A critical review of the bioavailability of glucosinolates and related compounds. Nat. Prod. Rep. 21, 425–447. (6) Hecht, S. S. (2000) Inhibition of carcinogenesis by isothiocyanates. Drug Metab. ReV. 32, 395–411. (7) Higdon, J. V., Delage, B., Williams, D. E., and Dashwood, R. H. (2007) Cruciferous vegetables and human cancer risk: Epidemiologic evidence and mechanistic basis. Pharmacol. Res. 55, 224–236. (8) Sparnins, V. L., Venegas, P. L., and Wattenberg, L. W. (1982) Glutathione S-transferase activity: Enhancement by compounds inhibiting chemical carcinogenesis and by dietary constituents. J. Natl. Cancer Inst. 68, 493–496. (9) Maheo, K., Morel, F., Langouet, S., Kramer, H., Le Ferrec, E., Ketterer, B., and Guillouzo, A. (1997) Inhibition of cytochromes P-450 and induction of glutathione S-transferases by sulforaphane in primary human and rat hepatocytes. Cancer Res. 57, 3649–3652. (10) Zhang, Y., Talalay, P., Cho, C. G., and Posner, G. H. (1992) A major inducer of anticarcinogenic protective enzymes from broccoli: isolation and elucidation of structure. Proc. Natl. Acad. Sci. U.S.A. 89, 2399– 2403. (11) von Weymarn, L. B., Chun, J. A., and Hollenberg, P. F. (2006) Effects of benzyl and phenethyl isothiocyanate on P450s 2A6 and 2A13: Potential for chemoprevention in smokers. Carcinogenesis 27, 782– 790. (12) Mi, L., Wang, X., Govind, S., Hood, B. L., Veenstra, T. D., Conrads, T. P., Saha, D. T., Goldman, R., and Chung, F. L. (2007) The role of protein binding in induction of apoptosis by phenethyl isothiocyanate and sulforaphane in human non-small lung cancer cells. Cancer Res. 67, 6409–6416. (13) Mi, L., Xiao, Z., Hood, B. L., Dakshanamurthy, S., Wang, X., Govind, S., Conrads, T. P., Veenstra, T. D., and Chung, F. L. (2008) Covalent binding to tubulin by isothiocyanates: a mechanism of cell growth arrest and apoptosis. J. Biol. Chem. 283, 22136–22146. (14) Zhang, Y., and Talalay, P. (1998) Mechanism of differential potencies of isothiocyanates as inducers of anticarcinogenic phase 2 enzymes. Cancer Res. 58, 4632–4639. (15) Conaway, C. C., Krzeminski, J., Amin, S., and Chung, F. L. (2001) Decomposition rates of isothiocyanate conjugates determine their activity as inhibitors of cytochrome p450 enzymes. Chem. Res. Toxicol. 14, 1170–1176. (16) Jiao, D., Conaway, C. C., Wang, M. H., Yang, C. S., Koehl, W., and Chung, F. L. (1996) Inhibition of N-nitrosodimethylamine demethylase in rat and human liver microsomes by isothiocyanates and their glutathione, L-cysteine, and N-acetyl-L-cysteine conjugates. Chem. Res. Toxicol. 9, 932–938. (17) Conaway, C. C., Yang, Y. M., and Chung, F. L. (2002) Isothiocyanates as cancer chemopreventive agents: Their biological activities and metabolism in rodents and humans. Curr. Drug Metab. 3, 233–255. (18) Zhang, Y. (2004) Cancer-preventive isothiocyanates: measurement of human exposure and mechanism of action. Mutat. Res. 555, 173– 190.

Chem. Res. Toxicol., Vol. 23, No. 4, 2010 765 (19) Brusewitz, G., Cameron, B. D., Chasseaud, L. F., Gorler, K., Hawkins, D. R., Koch, H., and Mennicke, W. H. (1977) The metabolism of benzyl isothiocyanate and its cysteine conjugate. Biochem. J. 162, 99– 107. (20) Hinman, A., Chuang, H. H., Bautista, D. M., and Julius, D. (2006) TRP channel activation by reversible covalent modification. Proc. Natl. Acad. Sci. U.S.A. 103, 19564–19568. (21) Nakamura, T., Kawai, Y., Kitamoto, N., Osawa, T., and Kato, Y. (2009) Covalent modification of lysine residues by allyl isothiocyanate in physiological conditions: Plausible transformation of isothiocyanate from thiol to amine. Chem. Res. Toxicol. 22, 536–542. (22) Xu, K., and Thornalley, P. J. (2000) Studies on the mechanism of the inhibition of human leukaemia cell growth by dietary isothiocyanates and their cysteine adducts in vitro. Biochem. Pharmacol. 60, 221– 231. (23) Skipper, P. L., and Tannenbaum, S. R. (1990) Protein adducts in the molecular dosimetry of chemical carcinogens. Carcinogenesis 11, 507– 518. (24) Funk, W. E., Waidyanatha, S., Chaing, S. H., and Rappaport, S. M. (2008) Hemoglobin adducts of benzene oxide in neonatal and adult dried blood spots. Cancer Epidemiol. Biomarkers PreV. 17, 1896– 1901. (25) Manjula, B. N., and Acharya, S. A. (2003) Hemoglobin disorders: Molecular methods and protocols. In Purification and Molecular Analysis of Hemoglobin by High-Performance Liquid Chromatography (Nagel, R. L., Ed.) pp 31-46, Humana Press Inc., Totowa, NJ. (26) Sepai, O., Henschler, D., and Sabbioni, G. (1995) Albumin adducts, hemoglobin adducts and urinary metabolites in workers exposed to 4,4′-methylenediphenyl diisocyanate. Carcinogenesis 16, 2583–2587. (27) Sabbioni, G., and Beyerbach, A. (1995) Determination of hemoglobin adducts of arylamines in humans. J. Chromatogr. B: Biomed. Appl. 667, 75–83. (28) Sabbioni, G., Skipper, P. L., Buchi, G., and Tannenbaum, S. R. (1987) Isolation and characterization of the major serum albumin adduct formed by aflatoxin B1 in vivo in rats. Carcinogenesis 8, 819–824. (29) Caldwell, J. B., Holt, L. A., and Milligan, B. (1971) E-amino derivatives of lysine. Aust. J. Chem. 24, 435–437. (30) Nowshuddin, S., and Reddy, A. R. (2006) Decomposition of copperamino acid complexes by sodium sulfide. Tetrahedron Lett. 47, 5159– 5161. (31) Kumar, A., Dongari, N., and Sabbioni, G. (2009) New isocyanatespecific albumin adducts of 4,4′-methylenediphenyl diisocyanate (MDI) in rats. Chem. Res. Toxicol. 22, 1975–1983. (32) Rappaport, S. M., Waidyanatha, S., Qu, Q., Shore, R., Jin, X., Cohen, B., Chen, L.-C., Melikian, A. A., Li, G., Yin, S., Yan, H., Xu, B., Mu, R., Li, Y., Zhang, X., and Li, K. (2002) Albumin adducts of benzene oxide and 1,4-benzoquinone as measures of human benzene metabolism. Cancer Res. 62, 1330–1337. (33) Vermeulen, M., van den Berg, R., Freidig, A. P., van Bladeren, P. J., and Vaes, W. H. (2006) Association between consumption of cruciferous vegetables and condiments and excretion in urine of isothiocyanate mercapturic acids. J. Agric. Food. Chem. 54, 5350– 5358. (34) Ji, Y., and Morris, M. E. (2003) Determination of phenethyl isothiocyanate in human plasma and urine by ammonia derivatization and liquid chromatography-tandem mass spectrometry. Anal. Biochem. 323, 39–47. (35) Mi, L., Gan, N., Cheema, A., Dakshanamurthy, S., Wang, X., Yang, D. C., and Chung, F. L. (2009) Cancer preventive isothiocyanates induce selective degradation of cellular R- and β-tubulins by proteasomes. J. Biol. Chem. 284, 17039–17051. (36) Skipper, P. L., Obiedzinski, M. W., Tannenbaum, S. R., Miller, D. W., Mitchum, R. K., and Kadlubar, F. F. (1985) Identification of the major serum albumin adduct formed by 4-aminobiphenyl in vivo in rats. Cancer Res. 45, 5122–5127. (37) Chung, F. L., Morse, M. A., Eklind, K. I., and Lewis, J. (1992) Quantitation of human uptake of the anticarcinogen phenethyl isothiocyanate after a watercress meal. Cancer Epidemiol. Biomarkers PreV. 1, 383–388.

TX900393T