Nucleophilic Reactions between Thiols and a Tobacco Specific

4-(methylnitrosamino)-1-(3-pyridyl)-1-butanone (NNK) and ... is also a breakdown product of covalently bound pyridyloxobutyl adducts resulting from NN...
0 downloads 0 Views 114KB Size
Download PDF
Chem. Res. Toxicol. 2003, 16, 661-667

661

Nucleophilic Reactions between Thiols and a Tobacco Specific Nitrosamine Metabolite, 4-Hydroxy-1-(3-pyridyl)-1-butanone Lisa A. Peterson,* Daniel P. Predecki, Nicole M. Thomson, Peter W. Villalta, and Elizabeth E. Donaldson Division of Environmental and Occupational Health and Cancer Center, University of Minnesota, Minneapolis, Minnesota 55455 Received December 13, 2002

4-Hydroxy-1-(3-pyridyl)-1-butanone (HPB) is a metabolite of the tobacco specific nitrosamines, 4-(methylnitrosamino)-1-(3-pyridyl)-1-butanone (NNK) and N′-nitrosonornicotine (NNN). HPB is also a breakdown product of covalently bound pyridyloxobutyl adducts resulting from NNK and NNN exposure. HPB released from DNA or hemoglobin has been used as an important dosimeter of tobacco specific nitrosamine exposure in a variety of studies. This compound is not reactive with cellular nucleophiles under biological conditions. We have discovered that HPB reacts with nucleophiles under acidic conditions to form cyclic tetrahydrofuranyl reaction products. Dithiothreitol, 2-mercaptoethanol, and N-acetylcysteine all reacted with HPB under these reaction conditions. In addition, reactions were observed with buffer chemicals such as 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid and tris(hydroxymethyl)aminomethane. The resulting cyclic adducts were unstable at room temperature. Their half-lives were significantly longer under neutral conditions than under acidic conditions. NMR studies established that the cyclic form of HPB, 2-hydroxy-2-(3-pyridyl)-2,3,4,5-THF, is present at significant concentrations in acidic solutions. The observation of this cyclic compound suggests that the reaction with nucleophiles may occur via a cyclic oxonium ion intermediate. This reaction was significant in our biological samples; there was up to 40% conversion of [5-3H]HPB to cyclic DTT-derived compounds when acidic DNA repair reactions containing [5-3H]pyridyloxobutylated DNA were stored overnight at -20 °C. Therefore, long-term storage of acid hydrolysates of pyridyloxobutylated DNA or protein for the analysis of HPB-releasing adducts could result in an underestimation of HPB-releasing adduct in those samples. In addition, these observations provide a mild synthetic method to prepare large quantities of cyclic 2-(3-pyridyl)-2,3,4,5-THF adducts predicted to result from pyridyloxobutylation of important cellular nucleophiles as a result of NNK and/or NNN exposure.

Introduction The tobacco specific nitrosamines, NNK1 and NNN, are potent carcinogens in laboratory animals (1). These carcinogens require metabolic activation to generate reactive intermediates capable of DNA and protein alkylation (2, 3). Both compounds can be metabolized to intermediates that pyridyloxobutylate DNA and proteins such as hemoglobin. NNK is activated to a pyridyloxobutylating intermediate upon methyl hydroxylation whereas NNN is converted to a pyridyloxobutylating species upon 2′-hydroxylation (Figure 1). The intermediate reacts with water to generate HPB, a metabolite of both NNK and NNN (1). Studies with model pyridyloxobutylating agents indicate that the reactive 4-oxo-4-(3-pyridyl)-1-butanediazonium ion can cyclize to form an electrophilic cyclic oxonium ion, which then reacts with nucleophiles (4, 5). * To whom correspondence should be addressed. 1 Abbreviations: AGT, O6-alkylguanine-DNA alkyltransferase; DTT, dithiothreitol; HPB, 4-hydroxy-1-(3-pyridyl)-1-butanone; NNK, 4-(methylnitrosamino)-1-(3-pyridyl)-1-butanone; NNKOAc, 4-(acetoxymethylnitrosamino)-1-(3-pyridyl)-1-butanone; NNN, N′-nitrosonornicotine; O6-pobG, O6-[4-oxo-4-(3-pyridyl)butyl]guanine.

Both cyclic and open chain adducts were observed. Therefore, it is likely that both cyclic and open chain adducts are formed with nucleophilic sites in protein and DNA. The structure of the pyridyloxobutyl protein and DNA adducts derived from NNN or NNK is largely unknown as a result of their instability under hydrolysis conditions. A portion of the hemoglobin adducts (15-40%) can be released as HPB with mild base treatment (3, 6). One HPB-releasing adduct is thought to be a 4-(3pyridyl)-4-oxobutyl carboxylic ester (7). The majority of the pyridyloxobutyl DNA adducts decompose to HPB under acid or enzyme DNA hydrolysis conditions (2, 8). One HPB-releasing DNA adduct has been identified as O6-pobG (9, 10). HPB released from hemoglobin or DNA has been used as a dosimeter of NNK and NNN activation in numerous animal and human studies (1, 6). Multiple studies indicated that the levels of HPB released from DNA and hemoglobin were attributed to the activation of NNK or NNN to a pyridyloxobutylating intermediate (3, 6). HPB was determined to be a breakdown product of covalently bound pyridyloxobutyl adducts and was not a reactant under biological conditions.

10.1021/tx025676h CCC: $25.00 © 2003 American Chemical Society Published on Web 04/16/2003

662

Chem. Res. Toxicol., Vol. 16, No. 5, 2003

Peterson et al.

Experimental Procedures

Figure 1. Pathways leading to pyridyloxobutyl adducts in the metabolic activation of NNK and NNN.

Figure 2. Reaction products formed between HPB and various nucleophiles.

In this paper, we describe the ability of HPB to react with thiol nucleophiles under acidic conditions to form cyclic products (Figure 2). Similar reaction products were observed for alcohol nucleophiles. The stability of the product varied with the nucleophile. These observations indicate that HPB forms a reactive cyclic electrophile under acidic conditions. This reaction was significant in our biological samples; there was up to 40% conversion of [5-3H]HPB to cyclic DTT-derived compounds when acidic DNA repair reactions containing [5-3H]pyridyloxobutylated DNA were stored overnight at -20 °C. HPB was also found to react with buffer components to produce reaction products that interfered with the analysis of pyridyloxobutyl DNA adducts. Therefore, this acidic reaction of HPB with nucleophiles may confound the analysis of DNA and protein samples for the levels of pyridyloxobutyl adducts.

Materials. HPB was prepared according to a previously reported method (5). N-Acetyl Cys (NAC) was purchased from Aldrich Chemical Co. (Milwaukee, WI). β-Mercaptoethanol (2ME) was purchased from Bio-Rad Pharmaceuticals (Hercules, CA). All NMR solvents were purchased from Cambridge Isotope Labs (Andover, MA). DTT and all remaining reagents were purchased from Fisher (Fair Lawn, NJ). Apparatus. HPLC analyses were performed on a Waters 600 system equipped with a C18 column (Phenomenex Bondclone 10 µm, 300 mm × 3.9 mm, Torrance, CA) and a photodiode array spectrophotometer (Waters Division of Millipore, Milford, MA). NMR experiments were carried out on a Varian VAC-300 or a gradient equipped Varian VI-500 (Varian, Inc., Palo Alto, CA). Mass Spectral Analysis. HPLC-electrospray mass spectra were acquired on a Thermo Finnigan LCQ Deca (San Jose, CA) in the positive ion mode. The analyses were performed with a C18 column (Phenomenex Bondclone 10 µm, 300 mm × 3.9 mm). DTT-HPB reaction mixtures were separated with an isocratic solvent system of 100 mM ammonium acetate containing 30% acetonitrile. 2ME-HPB reaction mixtures were eluted with 100 mM ammonium acetate containing 20% acetonitrile. NAC-HPB reaction mixtures were eluted with 100 mM ammonium acetate containing 95% acetonitrile. Reaction with Thiol Nucleophiles. HPB (5 mg, 30 µmol) was combined with DTT (18 mg, 120 µmol), 2ME (12 mg, 150 µmol), or NAC (24 mg, 150 µmol) in 0.1 N HCl at room temperature or -20 °C (total volume: 1 mL). The mixtures were analyzed by HPLC, following neutralization with either 1 M NaOH or 1 M sodium phosphate, pH 7.5. The column was isocratically eluted with 20 mM sodium phosphate buffer (pH 7) containing 20% acetonitrile (flow ) 1 mL/min). The stability of the reaction products at room temperature was determined as follows: first, solutions of HPB (5 mg, 30 µmol) and DTT (18 mg, 120 µmol), 2ME (12 mg, 150 µmol), or NAC (24 mg, 150 µmol) in 0.1 N HCl (total volume, 1 mL) were stored at -20 °C for greater than 72 h to convert approximately 90% of HPB to reaction products. Then, the solutions were thawed and warmed to room temperature. The stability of the reaction products under the 0.1 N HCl conditions was determined by HPLC analysis at various times after reaching room temperature with monitoring at 254 nm. A portion of each mixture was also neutralized to pH 7.0 with either 1 N NaOH or 1 M sodium phosphate, pH 7.5, and the decomposition of the adducts was monitored by HPLC analysis as described above. For structural characterization, HPB (46 mg, 280 µmol) and DTT (43 mg, 280 µmol), 2ME (22 mg, 280 µmol), or NAC (24 mg, 150 µmol) were combined in 0.1 N HCl (1 mL) and stored at -20 °C for greater than 72 h. The reaction mixture was neutralized with 1 N NaOH, and the reaction products were purified by HPLC utilizing a semipreparative Phenomenex Bondclone C18 column (10 µm, 250 mm × 10 mm). 1-(2-Pyridin-3-yl-THF-2-ylsulfanyl)-4-mercapto-butane2,3-diol (1). The DTT-HPB reaction products were eluted from the semipreparative column with a linear gradient from water containing 10% acetonitrile to water containing 50% acetonitrile over 60 min (flow ) 4 mL/min). The diastereomeric products, which eluted between 30 and 40 min, were collected and concentrated under reduced pressure for structural analysis. 1H NMR (D2O): δ 8.71 (s, 1H, 2′H), 8.49 (m, 1H, 6′H), 8.01 (m, 1H, 4′H), 7.51 (m, 1H, 5′H), 4.20 (m, 2H, 4-CH2), 3.58 (m, 1H, 2′′CHOH), 3.50 (m, 1H, 3′′-CHOH), 2.54 (m, 2H, 1′′-CHaS and 2-CHa), 2.51 (m, 2H, 4′′-CH2SH), 2.42 (m, 1H, 1′′-CHbS), 2.32 (m, 1H, 2-CHb), 2.21 (m, 1H, 3-CHa), 2.02 (m, 1H, 3-CHb). 13C NMR (80% D2O/20% CD3CN): δ 148.70 and 148.65 (C6′), 147.1 (C2′), 141.1 (C3′), 135.9 and 135.8 (C4′), 124.9 (C5′), 94.6 (C1), 74.8 and 74.7 (C3′′), 72.0 and 71.8 (C2′′), 69.3 (C4), 41.7 and 41.6 (C2 and C4′′), 33.4 and 33.3 (C1′′), 25.5 and 25.4 (C3). UV λmax: 263 nm. LC/MS: m/z 302 (M + H). LC/MS/MS (rel intensity): m/z 148 (M - DTT, 100). Reanalysis of aqueous solutions of 1-(2-pyridin-3-yl-THF-2ylsulfanyl)-4-mercapto-butane-2,3-diol by HPLC indicated that

Characterization of Cyclic HPB Adducts these diastereomers were unstable; they decomposed to form HPB and three new peaks in the HPLC trace. The new reaction products were identified as diastereomeric bis-1,4-(2-pyridin3-yl-THF-2-ylsulfanyl)butane-2,3-diol (2). 1H NMR (CDCl3): δ 8.8 (2H, m, 2′H), 8.5 (2H, m, 6′H), 7.9 (2H, m, 4′H), 7.3 (2H, m, 5′H), 4.2 (4H, m, 4-CH2), 3.48 (2H, m, CHOH), 2.6-1.9 (12H, m, CH2S, 3-CH2, 2-CH2). UV λmax: 219 and 263 nm. LC/MS (rel intensity): m/z 449 (M + H, 100), 302 (M - 2-pyridin-3-yl-THF, 8), 148 (M - DTT, 8). LC/MS/MS (rel intensity): m/z 302 (M HPB, 100), 148 (M - DTT - HPB, 8). 2-(2-Pyridin-3-yl-THF-2-ylsulfanyl)ethanol (3). The 2MEHPB reaction product was eluted from the semipreparative C18 column with water containing 20% acetonitrile (flow ) 4 mL/ min). The product, which eluted at 27.3 min, was collected and concentrated under reduced pressure. 1H NMR (D2O): δ 8.49 (s, 1H, 2′H), 8.26 (m, 1H, 6′H), 7.80 (m, 1H, 4′H), 7.3 (m, 1H, 5′H), 4.00 (m, 2H, 4-CH2), 3.30 (m, 2H, CHOH), 2.36 (m, 2H, CHaS and 2-CHa), 2.25 (m, 1H, CHbS), 2.16 (m, 1H, 2-CHb), 1.99 (m, 1H, 3-CHa), 1.80 (m, 1H, 3-CHb). 13C NMR (95% D2O/5% CD3CN): δ 148.8 (C6′), 147.2 (C2′), 141.2 (C3′), 136.1 (C4′), 125.0 (C5′), 94.8 (C1), 69.6 (C4), 61.7 (CH2OH), 41.8 (C2), 32.6 (SCH2), 25.5 (C3). UV λmax: 263 nm. LC/MS (rel intensity): m/z 226 (M + H). LC/MS/MS (rel intensity): m/z 149 (73), 148 (M - 2-ME, 100). 2-Methoxycarbonylamino-3-(2-pyridin-3-yl-THF-2-ylsulfanyl)propanoic Acid (4). The NAC-HPB reaction product was eluted from the semipreparative column with a 14 min linear gradient from 100% 25 mM sodium acetate, pH 5.5, to 86% sodium acetate, pH 5.5, containing 14% methanol and held isocratic at this concentration for 60 min (4). HPB eluted at 35.1 min and the diastereomeric products at 57.9 and 64.8 min. The structural properties of these reaction products were identical to those previously published for 4. Reaction with Hydroxyl Groups in Tris and Hepes. HPB (3 mM) in 50 mM Tris, 0.1 mM EDTA, 0.1 N HCl, or in 50 mM Hepes, 1.0 mM EDTA, 0.1 N HCl was stored at -20 °C for 5 or more days. The neutralized reaction mixtures (2 µL 1 M sodium phosphate, pH 7.5, 10 µL reaction mixture) were eluted from a Phenomenex Prodigy ODS 3 column with a linear gradient from 100 mM ammonium acetate to 100 mM ammonium acetate containing 50% acetonitrile over 60 min (flow ) 1 mL/min). The retention time of HPB on this system was 29.4 min. The TrisHPB reaction product eluted at 36.2 min and was identified as 1-O-(2-pyridin-3-yl-THF-2-yl)-2-amino-2-(hydroxymethyl)-1,3propanediol (5). UV λmax: 254, 259.1, and 265 nm; m/z (rel intensity): 269 (M + H, 46), 166 [M + H - 2-amino-2(hydroxymethyl)-3-propanol, 17], 148 (M + H - Tris, 100), 122 (M + H - 2-pyridin-3-yl-THF, 43). The Hepes reaction product eluted at 44.3 min and was identified as N-[2-O-(2-pyridin-3yl-THF-2-yl)hydroxyethyl]piperazine-N′-(2-ethanesulfonic acid) (6). UV λmax: 254, 259.1, and 265 nm. LC/MS m/z (rel intensity): 444 (M + acetate, 100), 386 (M + H, 51), 239 (M + H 2-pyridin-3-yl-THF, 13). 2-Pyridin-3-yl-THF-2-ol. HPB (4.8 mg, 29 µmol) in 0.1 M DCl or 0.1 M DClO4 solution (0.6 mL) generated a mixture containing approximately 60% HPB and 40% 2-pyridin-3-ylTHF-2-ol. HPB: 1H NMR (0.1 M DCl or DClO4 in D2O): δ 9.14 (s, 1H, 2′-H), 8.93 (m, 1H, 6′-H), 8.8 (m, 1H, 4′-H), 8.1 (m, 1H, 5′-H), 3.5 (m, 2H, 4-CH2), 3.1 (m, 2H, 2-CH2), 1.8 (m, 2H, 3-CH2). 13C NMR (0.1 M DCl in D O containing 10% acetonitrile-d ): δ 2 3 198.2 (C1), 145.7 (C6′), 144 (C4′), 141.5 (C2′), 135 (C3′), 127.7 (C5′), 60.6 (C4), 35.4 (C2), 25.4 (C3). 2-Pyridin-3-yl-THF-2-ol: 1H NMR (0.1 M DCl or DClO in D O): δ 8.75 (s, 1H, 2′-H), 4 2 8.62 (m, 1H, 6′-H), 8.60 (m, 1H, 4′-H), 7.9 (m, 1H, 5′-H), 4.1 (m, 2H, 4-CH2), 2.1 (m, 1H, 2-CHa) 2.0 (m, 1H, 2-CHb) 1.9 (m, 2H, 3-CH2). 13C NMR (0.1 M DCl in D2O containing 10% acetonitriled3): δ 144.4 (C6′), 140.7 (C4′), 138.6 (C2′), 127.2 (C5′), 103.5 (C1), 69.6 (C4), 40.1 (C2), 24.3 (C3).

Results During our investigation of the repair of pyridyloxobutyl DNA adducts by AGT, we noticed that overnight

Chem. Res. Toxicol., Vol. 16, No. 5, 2003 663

Figure 3. Radiogram of 0.1 N HCl hydrolysate of [5-3H]NNKOAc-treated calf thymus DNA in the presence of AGT reaction buffer.

storage of the 0.1 N HCl hydrolysates of the repair reactions at -20 °C resulted in the formation of two unknown peaks in the HPLC traces (Figure 3). The formation of these new peaks corresponded to a reduction in the HPB peak. Subsequent studies indicated that formation of these new compounds required acid, HPB, and DTT, a reagent required for maintenance of AGT activity. The extent of conversion depended upon the length of storage time and DTT concentration (typically 1-5 mM). Chemical characterization of these reaction products indicated that they were diastereomers of 1 (Figure 2). Both products had identical UV spectra; a comparison with the UV spectrum of HPB indicated that the λmax values had shifted to slightly lower wavelengths at 219 and 264 nm (Figure 4B). The mass spectra for both products contained a protonated molecular ion at m/z 302, which corresponds to the combination of HPB and DTT minus a molecule of water. Both compounds fragmented to yield an ion at m/z 148, which is consistent with the loss of the DTT moiety. 1H NMR spectral data were consistent with the cyclic structure. The proton on carbon 2′ of the pyridine ring has a chemical shift of 8.71 ppm, which is significantly upfield from the 2′ proton of HPB. A similar upfield shift of the 2′ proton is characteristic of the loss of the carbonyl group at the 3 position of the pyridine moiety (4). The 1H NMR spectrum of the aliphatic region was consistent with the substituted tetrahydrofuranyl ring protons (4). Initial characterization of the DTT-HPB reaction products was complicated by their instability. Collection of the products from HPLC and storage in HPLC buffer (initially 20 mM sodium phosphate, pH 7, or 100 mM ammonium acetate with 30% acetonitrile) led to the formation of HPB and multiple secondary products that were isomers of one another. The UV spectra of these products were identical to the initial reaction products (data not shown). Mass spectral analysis of these secondary compounds yielded molecular ions at m/z 449. These data indicate that the secondary products are isomers containing one molecule of DTT and two molecules of HPB following the loss of two water molecules. Integration of the signals in the NMR spectrum of one of the

664

Chem. Res. Toxicol., Vol. 16, No. 5, 2003

Figure 4. UV spectra of HPB and the reaction products of HPB with various nucleophiles: (A) HPB; (B) DTT-HPB, 1; (C) 2ME-HPB, 3; (D) NAC-HPB, 4; (E) Tris-HPB, 5; and (F) HEPES-HPB, 6.

isomers indicated that there were two substituted THF rings relative to one DTT moiety. As with the primary reaction product, the 2′ proton of the pyridyl rings has a chemical shift of 8.8 ppm, consistent with the loss of the carbonyl group in the 3′ position. On the basis of all of this evidence, the secondary adducts were assigned the structure 2 (Figure 2). Because this bis-substituted DTT derivative contains four chiral centers, the diastereomers of this compound appear as multiple peaks in the HPLC trace. To avoid the complication of these secondary reaction products and to confirm our NMR assignments, we investigated the ability of 2ME to undergo similar reactions. As with DTT, reaction of 2ME with HPB in 0.1 N HCl at -20 °C led to the quantitative conversion of HPB to a new reaction product (3, Figure 2). Because 2ME lacks a chiral center, the reaction product was a mixture of enantiomers. The UV of this product was identical to that obtained with DTT and HPB (Figure 4C). The molecular ion in the mass spectrum of the adduct was m/z 226, consistent with the addition of 2ME to HPB with the loss of water. As with the DTT reaction products, the NMR spectrum indicated the presence of a substituted tetrahydrofuranyl group and the absence of a carbonyl moiety in the 3′ position of the pyridyl group. The point of attachment of the tetrahydrofuranyl moiety to the thiol of 2ME was determined by the presence of diastereotopic protons at 2.25 and 2.16 ppm, which were assigned to the methylene group adjacent to the sulfur atom. These protons were coupled to a quartet, which has a chemical shift of 3.30 ppm. This signal was assigned to the methylene protons of the unsubstituted hydroxyl group. Therefore, the 2ME reaction product was assigned the structure 3 (Figure 2). NAC was also used as a nucleophile in these reactions. The UV of the reaction products was identical to that obtained for the DTT and 2ME cyclic adducts (Figure 4D). Because we used the L isomer, two diastereomeric products were formed. These adducts (4, Figure 2) had

Peterson et al.

Figure 5. HPLC traces of HPB stored in acidic solutions. (A) In 0.1 N HCl; (B) in Tris buffer and 0.1 N HCl; (C) in Hepes buffer and 0.1 N HCl. The HPLC column was eluted with a linear gradient from 20 mM sodium phosphate buffer to 20 mM sodium phosphate buffer containing 50% methanol over 60 min.

Figure 6. 1H NMR spectra of HPB in D2O (A and B) or 0.1 N DCl (C and D).

been previously characterized as the reaction products from the reaction of model pyridyloxobutylating agents with NAC (4). For all three thiol reagents, no evidence of open chain adducts was observedsall adducts had a cyclized structure. HPB also reacted with Tris and Hepes buffer. Overnight storage of HPB in acidic Tris or Hepes at -20 °C led to the formation of a new compound that required HPB, buffer, and 0.1 N HCl (Figure 5); approximately 5% of HPB had been converted to the reaction product. Attempts to isolate these reaction products for NMR

Characterization of Cyclic HPB Adducts

Chem. Res. Toxicol., Vol. 16, No. 5, 2003 665

Figure 7. Reaction between HPB and nucleophiles under acidic conditions.

analysis were unsuccessful as a result of their instability. LC-MS analysis indicated that the compounds were addition products of HPB and Tris or Hepes following the loss of water with the molecular ions at m/z 269 and 386, respectively. HPLC analysis with diode array detection revealed that both reaction products had identical UV spectra with three λmax values at 254, 259, and 265 nm (Figure 4E,F). These UV spectra are identical to that reported for 2-methoxy-2-(3-pyridyl)-2,3,4,5-THF (4). As a result, we believe these reaction products are cyclic HPB adducts attached at the oxygen atom of the buffers (5 and 6, Figure 2). The chemical structure of these reaction products indicated that HPB may react with these nucleophiles via a cyclic intermediate. NMR studies indicate that the cyclic form of HPB (7) represents a significant structure under acidic conditions. The 1H NMR spectrum of HPB in 0.1 N DCl contains signals for two structurally related compounds. One set of signals corresponds to openchained HPB and the other set of signals corresponds to the cyclic form, 2-hydroxy-2-(3-pyridyl)-2,3,4,5-THF (Figure 6). The general downfield shift of the pyridyl protons results from the protonation of the pyridyl nitrogen in the acid solution. In the cyclic form, the pyridyl protons are shifted upfield relative to those assigned to the open chain compound. In the aliphatic region of the NMR spectrum, there is the appearance of signals consistent with the formation of a THF ring. Consistent with the loss of the carbonyl carbon and the formation of a hemiacetal, the chemical shift of the C1 carbon changes from 198.2 ppm in the open-chained HPB to 103.5 ppm in the cyclic form. We believe that the 2 position contains a hydroxyl group, not a chlorine atom, since there are no changes in either the proton or the carbon chemical shifts of the cyclic compound when DCl is substituted with DClO4. Given the novelty of these reactions, we explored the formation of the HPB-thiol reaction under a variety of conditions. HPB (30 mM) reacted with the various thiol compounds (120 or 150 mM) in 0.1 N HCl at room temperature. There was no significant reaction between HPB and DTT or 2ME in sodium phosphate, pH 7. However, a trace amount of compound 4 (∼2%) was observed when NAC and HPB were combined under these conditions. The acidic DTT (120 mM) and 2ME (150 mM) reaction mixtures reached equilibrium approximately 48 h after mixing the two reactants. Products 1 and 3 represented approximately 26 (7.8 ( 0.4 mM) and 15% (4.5 ( 0.6 mM) of the starting HPB concentration, respectively. The NAC reaction mixture had not reached equilibrium after 8 days. Freezing the acidic mixtures significantly increased the reaction yield. Almost all of the HPB (g90%) was converted to cyclic reaction products upon storage at -20 °C for up to 3 days when the nucleophiles were DTT or

Table 1. Half-Lives (h) of Cyclic Thiol-HPB Adducts at Room Temperaturea cyclic HPB-thiol adduct

neutral

acidic

1 3 4

148 ( 2 206 ( 12 271 ( 16

13 ( 5 9(3 70 ( 16

a Solutions of 30 mM HPB and 120 mM DTT, 150 mM 2ME, or 150 mM NAC in 0.1 N HCl were stored at -20 °C for at least 3 days. These reaction mixtures were neutralized with 1 M sodium phosphate buffer, pH. 7.5, to determine the half-lives at pH 7. The rates of decomposition of HPB-thiol adducts were monitored by HPLC. The values are averages ( SD from three experiments. See the Experimental Procedures for details.

2ME; the reaction with NAC required more than a week at -20 °C to convert >90% of HPB to cyclic products. The reason for this enhanced yield when the reactions were frozen is not known. Once the acidic reaction mixtures were warmed to room temperature, the reaction products decomposed to HPB and thiol. The decomposition reaction reached equilibrium for the DTT and 2ME reaction products after approximately 3 days at room temperature (equilibrium concentration for product 1, 5.3 ( 0.9 mM, and for product 3, 4.6 ( 0.6 mM). At equilibrium, the reaction products represented about 1520% of the starting HPB concentration in the reaction mixture. The NAC reaction product 4 was significantly more stable and had not reached equilibrium after 8 days at room temperature. At 8 days, the cyclic reaction product 4 represented approximately 35% of the total HPB in the solution. If the reaction mixtures were neutralized and allowed to sit at room temperature, the cyclic reaction products were significantly more stable. The half-lives of the different thiol reaction products at room temperature under neutral and acidic conditions are listed in Table 1. The adducts were significantly less stable at 37 °C. Under acidic conditions at 37 °C, compounds 1 and 3 both had half-lives of 2 h whereas the half-life of adduct 4 was 18 h. Under neutral conditions, the half-lives increased to 14, 36, and 24 h, respectively, at 37 °C.

Discussion Our results indicate that HPB can react with thiol and hydroxyl nucleophiles under acidic conditions to form cyclic adducts. Because NMR studies established that the cyclic form of HPB, 2-hydroxy-2-(3-pyridyl)-2,3,4,5-THF, is present at significant concentrations in acidic solutions, we postulate that this hemiketal reacts with nucleophiles via a cyclic oxonium ion to produce cyclic adducts (Figure 7). Alternatively, the reaction may be initiated by nucleophilic addition to the carbonyl carbon atom. In either case, the net result is the formation of cyclic adducts.

666

Chem. Res. Toxicol., Vol. 16, No. 5, 2003

Structurally related cyclic 2-(3-pyridyl)-2,3,4,5-THF adducts have been observed in reactions between activated pyridyloxobutylating nitrosamines and model nucleophiles such as methanol or NAC (4, 5). In those studies, cyclic adducts were the predominant reaction products, with significantly lower levels of open chain adducts formed. Only cyclic adducts were observed when methanol was the nucleophile (5). Cyclic and open chain adducts were observed in a ratio of 3:1 when NAC was the nucleophile (4). These studies provided evidence that pyridyloxobutylating agents preferentially react with nucleophiles via a cyclic oxonium ion intermediate. Despite the importance of these cyclic adducts in in vitro alkylation reactions by pyridyloxobutylating agents, there is scant evidence for cyclic 2-(3-pyridyl)-2,3,4,5-THF adducts in hemoglobin isolated from NNK-treated rats. The vast majority of the HPB-releasing adducts in hemoglobin appears to be derived from open chain 4-(3pyridyl)-4-oxobutyl carboxylic acid adducts (7). Given the instability of the cyclic derivatives, it is not clear that these adducts will survive the acidic hemoglobin isolation procedure (3). DNA from animals treated with pyridyloxobutylating nitrosamines contains adducts that release HPB (2, 11, 12). We have identified one open chain adduct, O6-pobG (10). This adduct is stable to 0.1 N HCl DNA hydrolysis conditions. However, the levels of this adduct are low relative to levels of HPB-releasing DNA adducts, representing only about 3-5% of the total HPB-releasing adducts in in vitro alkylation reactions (9). Recently, 7-[4oxo-4-(3-pyridyl)but-1-yl]-2′-deoxyguanosine and N2-[4oxo-4-(3-pyridyl)but-1-yl]-2′-deoxyguanosine were identified as open chain adducts in NNKOAc-treated calf thymus DNA (15). The 7-alkylguanine adduct accounted for 30-35% of the total HPB-releasing adducts in these studies. It is possible that cyclic DNA adducts account for a significant portion of the uncharacterized HPBreleasing adducts present in DNA from NNK- or NNNtreated animals. The cyclic adducts may also be reversible adducts. This is supported by the observation that the bis-DTT-HPB adduct 3 is formed in much higher concentrations during the purification of the mono-DTT-HPB product 1. Therefore, the cyclic adducts form a possible reservoir of pyridyloxobutylating agent in the cell since an oxonium ion is a likely intermediate in the decomposition process. The formation of the HPB-nucleophile adducts under acidic conditions also demonstrates that long-term storage of acid hydrolysates of pyridyloxobutylated DNA or protein for the analysis of HPB-releasing adducts is not advised. Because HPB released from DNA or hemoglobin has been used as a dosimeter of tobacco specific nitrosamine exposure in several studies (13, 14), storage of acidic hydrolysates at -20 °C should be avoided as it may lead to an underestimation of the total HPB-releasing adducts. We observed up to 40% conversion of [5-3H]HPB to cyclic DTT-derived compounds when our acidic DNA repair reactions were stored overnight at -20 °C; the HPB concentrations in these samples were approximately 1 nM. Another potential confounder is the potential reaction of HPB with buffer components under acidic conditions. The Hepes-HPB adduct 6 coeluted with O6pobG, complicating our attempts to measure repair of O6pobG by AGT. Substitution of Hepes with Tris eliminated

Peterson et al.

this problem since the Tris-HPB reaction product 5 eluted at an earlier time. Finally, our findings provide a mild synthetic method to prepare large quantities of cyclic 2-(3-pyridyl)-2,3,4,5THF adducts predicted to result from pyridyloxobutylation of important cellular nucleophiles during NNK and/ or NNN exposure. This method of generating exclusively cyclic adducts will allow development of mild isolation procedures that minimize decomposition of these unstable adducts. As a result, we will be able to better assess the importance of cyclic protein and DNA adducts to the toxicological properties of pyridyloxobutylating agents.

Acknowledgment. We thank Drs. Richard Loeppky and Shana Sturla for helpful discussions. The mass spectral analyses were performed in the Analytical Biochemistry Core Facility at the University of Minnesota Cancer Center, which is funded by the National Cancer Institute center Grant CA-77598. This research was funded by CA-59887 from the National Cancer Institute.

References (1) Hecht, S. S. (1998) Biochemistry, biology, and carcinogenicity of tobacco-specific N-nitrosamines. Chem. Res. Toxicol. 11, 560-603. (2) Hecht, S. S., Spratt, T. E., and Trushin, N. (1988) Evidence for 4-(3-pyridyl)-4-oxobutylation of DNA in F344 rats treated with the tobacco specific nitrosamines 4-(methylnitrosamino)-1-(3pyridyl)-1-butanone and N′-nitrosonornicotine. Carcinogenesis 9, 161-165. (3) Carmella, S. G., and Hecht, S. S. (1987) Formation of hemoglobin adducts upon treatment of F344 rats with the tobacco-specific nitrosamines 4-(methylnitrosamino)-1-(3-pyridyl)-1-butanone and N′-nitrosonornicotine. Cancer Res. 47, 2626-2630. (4) Carmella, S. G., Kagan, S. S., Spratt, T. E., and Hecht, S. S. (1990) Evaluation of cysteine adduct formation in rat hemoglobin by 4-(methylnitrosamino)-1-(3-pyridyl)-1-butanone and related compounds. Cancer Res. 50, 5453-5459. (5) Spratt, T. E., Peterson, L. A., Confer, W. L., and Hecht, S. S. (1990) Solvolysis of model compounds for R-hydroxylation of N′nitrosonornicotine and 4-(methylnitrosamino)-1-(3-pyridyl)-1-butanone: Evidence for a cyclic oxonium ion intermediate in the alkylation of nucleophiles. Chem. Res. Toxicol. 3, 350-356. (6) Peterson, L. A., Carmella, S. G., and Hecht, S. S. (1990) Investigations of metabolic precursors to hemoglobin and DNA adducts of 4-(methylnitrosamino)-1-(3-pyridyl)-1-butanone. Carcinogenesis 11, 1329-1333. (7) Carmella, S. G., Kagan, S. S., and Hecht, S. S. (1992) Evidence that a hemoglobin adduct of 4-(methylnitrosamino)-1-(3-pyridyl)1-butanone (NNK) is a 4-(3-pyridyl)-4-oxobutyl carboxylic acid ester. Chem. Res. Toxicol. 5, 76-80. (8) Spratt, T. E., Trushin, N., Lin, D., and Hecht, S. S. (1989) Analysis for N2-(pyridyloxobutyl)deoxyguanosine adducts in DNA of tissues exposed to tritium labeled 4-(methylnitrosamino)-1-(3-pyridyl)1-butanone and N′-nitrosonornicotine. Chem. Res. Toxicol. 2, 169173. (9) Wang, L., Spratt, T. E., Liu, X. K., Hecht, S. S., Pegg, A. E., and Peterson, L. A. (1997) Pyridyloxobutyl adduct O6-[4-oxo-4-(3pyridyl)butyl]guanine is present in 4-(acetoxymethylnitrosamino)1-(3-pyridyl)-1-butanone-treated DNA and is a substrate for O6alkylguanine-DNA alkyltransferase. Chem. Res. Toxicol. 10, 562567. (10) Thomson, N. M., Kenney, P. M., and Peterson, L. A. (2003) The pyridyloxobutyl DNA adduct, O6-[4-oxo-4-(3-pyridyl)butyl]guanine, is detected in tissues from 4-(methylnitrosamino)-1-(3pyridyl)-1-butanone-treated A/J mice. Chem. Res. Toxicol. 16, 1-6. (11) Peterson, L. A., Mathew, R., Murphy, S. E., Trushin, N., and Hecht, S. S. (1991) In vivo and in vitro persistence of pyridyloxobutyl DNA adducts from 4-(methylnitrosamino)-1-(3-pyridyl)-1butanone. Carcinogenesis 12, 2069-2072. (12) Peterson, L. A., and Hecht, S. S. (1991) O6-Methylguanine is a critical determinant of 4-(methylnitrosamino)-1-(3-pyridyl)-1butanone tumorigenesis in A/J mouse lung. Cancer Res. 51, 55575564.

Characterization of Cyclic HPB Adducts (13) Staretz, M. E., Foiles, P. G., Miglietta, L. M., and Hecht, S. S. (1997) Evidence for an important role of DNA pyridyloxobutylation in rat lung carcinogenesis by 4-(methylnitrosamino)-1-(3pyridyl)-1-butanone: effects of dose and phenethyl isothiocyanate. Cancer Res. 57, 259-266. (14) Foiles, P. G., Akerkar, S. A., Carmella, S. G., Kagan, M., Stoner, G. D., Resau, J. H., and Hecht, S. S. (1991) Mass spectrometric analysis of tobacco-specific nitrosamine DNA adducts in smokers and nonsmokers. Chem. Res. Toxicol. 4, 364-368.

Chem. Res. Toxicol., Vol. 16, No. 5, 2003 667 (15) Wang, M., Cheng, G., Sturla, S. J., Shi, Y., McIntee, E. J., Villalta, P. W., Upadhyaya, P., and Hecht, S. S. (2003) Identification of adducts formed by pyridyloxobutylation of deoxyguanosine and DNA by 4-(acetoxymethylnitrosamino)-1-(3-pyridyl)-1-butanone, a chemically activated form of tobacco-specific carcinogens. Chem. Res. Toxicol. Published ASAP 4/3/03.

TX025676H