l-cysteine - American Chemical Society

The stability of S-(purin-6-yl)-L-cysteine (SPC), a kidney-selective prodrug of 6-mercaptopu- rine and a putative metabolite of 6-chloropurine, was in...
0 downloads 0 Views 136KB Size
654

Chem. Res. Toxicol. 1996, 9, 654-658

Effects of pH, Temperature, and Chemical Structure on the Stability of S-(Purin-6-yl)-L-cysteine: Evidence for a Novel Molecular Rearrangement Mechanism To Yield N-(Purin-6-yl)-L-cysteine Adnan A. Elfarra* and In Young Hwang† Department of Comparative Biosciences and Environmental Toxicology Center, University of Wisconsin, Madison, Wisconsin 53706 Received November 14, 1995X

The stability of S-(purin-6-yl)-L-cysteine (SPC), a kidney-selective prodrug of 6-mercaptopurine and a putative metabolite of 6-chloropurine, was investigated under various pH and temperature conditions. At room temperature, the half-life (t1/2) of SPC at either highly acidic (pH 3.6) or basic conditions (pH 9.6) was longer than at neutral or slightly acidic or basic conditions (pH 5.7-8.75). The primary degradation product, N-(purin-6-yl)-L-cysteine (NPC), was isolated using Sephadex LH-20 chromatography and characterized by 1H NMR and FAB/ MS after derivatization with 2-iodoacetic acid. These results reveal novel stability requirements and implicate the cysteinyl amino group and the purinyl N-1 nitrogen in the mechanism of SPC rearrangement to NPC. Further evidence for this hypothesis was provided by the findings that the stability of SPC in phosphate buffer (pH 7.4) at 37 °C was similar to that of S-(guanin6-yl)-L-cysteine, whereas S-(purin-6-yl)-N-acetyl-L-cysteine and S-(purin-6-yl)glutathione which have their cysteine amino groups blocked were much more stable than SPC. S-(Purin-6-yl)L-homocysteine (SPHC) was also more stable than SPC, possibly because the formation of a 6-membered ring transition state as would be expected with SPHC is kinetically less favored than the formation of a 5-membered ring transition state as would be expected with SPC. These results may explain previous in vivo metabolism results of SPC and its analogs and may contribute to a better understanding of stability of structurally related cysteine S-conjugates.

Introduction Cysteine S-conjugates are in vivo metabolites of endogenous chemicals, drugs, herbicides, and other organic chemicals (1-3). Several cysteine S-conjugates exhibit biological activity; for example, cysteinyl leukotrienes play a role in the pathogenesis of asthma (4), and cysteine S-conjugates of halogenated hydrocarbons are implicated in nephrotoxicity, hepatotoxicity, and carcinogenicity (2, 3, 5-7). Some cysteine S-conjugates are relatively unstable; they can be degraded by nonenzymatic oxidation, hydrolysis, and cyclization reactions (3, 5-8). S,NIntramolecular rearrangements of S-(2,4-dinitrophenyl)L-cysteine and S-(pyrimidin-2-yl)-L-cysteine have been reported (9, 10), but the mechanisms of these degradation reactions have not been determined. S-(Purin-6-yl)-L-cysteine (SPC),1 a cysteine conjugate β-lyase-activated prodrug of the antitumor and immunosuppressant drug, 6-mercaptopurine, has been implicated in the antitumor properties of 6-chloropurine (1114). Because the stability of SPC is likely to be an important determinant of its biological activity, SPC degradation rates under different conditions of pH and * Corresponding author at the Department of Comparative Biosciences, 2015 Linden Dr. W., Madison, WI 53706. Phone: (608) 2626518; Fax: (608) 263-3926; E-mail: [email protected]. † Present address: Department of Environmental Science, Inje University, Kimhae, Kyungnam, S. Korea. X Abstract published in Advance ACS Abstracts, March 15, 1996. 1 Abbreviations: SPC, S-(purin-6-yl)-L-cysteine; NPC, N-(purin-6yl)-L-cysteine; CMNPC, S-(carboxymethyl)-N-(purin-6-yl)-L-cysteine; NAPC, S-(purin-6-yl)-N-acetyl-L-cysteine; SPHC, S-(purin-6-yl)-L-homocysteine; SPG, S-(purin-6-yl)glutathione; SGC, S-(guanin-6-yl)-Lcysteine; FAB/MS, fast atom bombardment mass spectrometry.

0893-228x/96/2709-0654$12.00/0

Figure 1. Chemical degradation of SPC to NPC and reaction of NPC with 2-iodoacetic acid to yield CMNPC.

temperature were determined in this study. The chemical structure of the major degradation product, N-(purin6-yl)-L-cysteine (NPC), was determined by NMR and FAB/MS after its derivatization with iodoacetic acid to yield S-(carboxymethyl)-N-(purin-6-yl)-L-cysteine (CMNPC; Figure 1). The decomposition rates of several SPC analogs (Figure 2) were also determined. The results provided strong evidence for a novel mechanism of intramolecular rearrangement of SPC to NPC.

Experimental Procedures Chemicals. SPC, S-(purin-6-yl)-N-acetyl-L-cysteine (NAPC), S-(purin-6-yl)-L-homocysteine (SPHC), S-(purin-6-yl)glutathione (SPG), and S-(guanin-6-yl)-L-cysteine (SGC) were prepared as previously described (12-15). 2-Iodoacetic acid and deuterium oxide (99.9%) were purchased from Aldrich Chemical Co. (Milwaukee, WI). Sephadex LH-20 was obtained from Pharmacia Inc. (Piscataway, NJ). All other reagents were of the highest grade available commercially. Effects of pH and Temperature on SPC Degradation Rate. The time dependent SPC degradation under different pH and temperature conditions was monitored by HPLC. SPC (0.42 mM) was dissolved in either 0.02 M sodium acetate buffer

© 1996 American Chemical Society

S,N-Rearrangement of S-(Purin-6-yl)-L-cysteine

Chem. Res. Toxicol., Vol. 9, No. 3, 1996 655

Figure 2. Chemical structures of SPC and its analogs. adjusted to pH 3.6 or 5.7 with trifluoroacetic acid, 0.02 M potassium phosphate buffer (pH 6.10 or 7.23), or 0.02 M sodium borate buffer (pH 8.25, 8.75, 9.55, or 10.50). SPC solutions were then incubated for specified time intervals in a water bath at 23 °C. In some experiments, SPC solutions in borate buffer (pH 8.6) were incubated for various time points at either 0, 20, 36, or 48 °C. For PC half-life determinations, 5-100 µL of trifluoroacetic acid (20%, v/v) was added to aliquots (0.5 mL) taken from the incubated SPC solutions to adjust the pH to 1.8-2.0 before analysis for SPC remaining in solution by HPLC as described below. Rate constants for the stability of SPC were calculated using the following first-order equation: ln (C/C0) ) -kt, where C0 was the SPC concentration at time zero, C was the SPC concentration at a specific time interval, and k was the rate constant (min-1). Incubations were carried out until the remaining SPC concentration was approximately 20% of C0 or after a few days when the degradation rate was very slow. After determining the k value for each condition, the half-life (t1/2) of SPC was calculated using the equation: t1/2 ) 0.693/k. Decomposition Rates of SPC Analogs. Effects of structural modifications on SPC degradation rates were determined by investigating the half-lives of SPC and its analogs SGC, SPHC, NAPC, or SPG (0.1-0.4 mM), after incubations at 37 °C in 0.02 M phosphate buffer (pH 7.4). The half-lives of SPC analogs were determined as described above; because of their low degradation rates, SPG, SPHC, and NAPC incubations were carried out for 5 days. Identification of NPC as the Major SPC Degradation Product. SPC (200 mg) was dissolved in 15 mL of water, and the pH was adjusted to pH 7.0 with 0.1 N HCl. The SPC solution was incubated at room temperature for 2 h to allow for SPC rearrangement to NPC, and the solution was then kept in an ice bath until an aliquot (4 mL) was applied to a Sephadex LH-20 column (95 × 2.5 cm) at 4 °C. NPC was eluted with water, adjusted to pH 10.7 with NaOH; the flow rate was maintained at 4 mL/min by a peristaltic pump (LKB 2115 multiperpex pump, Pharmacia LKB Biotechnology AB, Bromma, Sweden); the detection wavelength (Isco UA-5, ISCO Inc., Lincoln, NE) was at 254 nm. The eluting fractions containing NPC (fractions eluting from 390 to 480 mL) were pooled, and 2-iodoacetic acid (110 mg) was added with constant magnetic stirring to trap the free thiol group of NPC to form a stable product. NaOH (1 N) was also added to the reaction mixture to adjust the pH to 10. The reaction was allowed to proceed at room temperature for 30 min. This mixture was then applied to a Sephadex LH-20 column, and the expected product CMNPC, was eluted as described for NPC. The fractions containing CMNPC (eluted between 300 and 340 mL) were pooled, lyophilized, and stored in a desiccator at room temperature. The purity of CMNPC was >99% as determined by HPLC (described below). HPLC Analyses. Samples (20 µL) were injected onto a 250 × 4.6 mm Beckman ODS 5 µm column attached to an HPLC

Figure 3. HPLC chromatograms of SPC in the absence (A) and presence (B) of 2-iodoacetic acid. The SPC peak (I) was decreased over time and a new peak (II) appeared. In the presence of 2-iodoacetic acid, peak II was converted to peak III. All kinetic experiments were carried out in the absence of 2-iodoacetic acid. system consisting of a Beckman 110B and 114M HPLC pump, a 3-cm C18 guard column, and a Spectroflow 757 variable wavelength detector (Kratos Analytical, Ramsey, NJ). The eluent was run at a flow rate of 1 mL/min. For SPC, NAPC, and SPHC, isocratic analyses were carried out with 7.5% (v/v) acetonitrile in water adjusted to pH 2.5 with trifluoroacetic acid. For SPG analysis, the acetonitrile gradient was changed from 15% to 40% over 8 min, decreased to 15% over 6 min, and kept constant at 15% for a total run time of 20 min. The detection wavelengths and retention times were 266 nm and 4.3 min for SPC, 292 nm and 5.0 min for SPHC, 292 nm and 8.9 min for NAPC, and 288 nm and 5.3 min for SPG, respectively. SGC analyses were carried out as described previously (15). NMR and Mass Spectral Analysis of CMNPC. To obtain a 1H NMR spectrum of CMNPC, CMNPC was dissolved in D2O and the spectrum was obtained on a Bruker spectrometer at 500 MHz. Chemical shifts are reported in parts per million from sodium 3-(trimethylsilyl)tetradeuteriopropionate. The residual HOD signal at 4.7-4.9 parts per million was attenuated by continuous irradiation with the decoupler. The assignment of structure of CMNPC was achieved by comparing the chemical shifts, coupling patterns, and integration ratios of the various protons in the spectrum with those of published spectra of other S-substituted derivatives of cysteine, 6-thiopurine and 2-thioacetic acid (12-15). Positive ion FAB/MS analysis of CMNPC was performed with a Kratos MS-50TC spectrometer (Manchester, U.K.) equipped with a saddle field fast atom bombardment gun. Approximately 5 µg of sample was mixed with 3-nitrobenzyl alcohol matrix before placing the sample on the target. Spectra were recorded over the range of m/z 100-400.

Results Stability of SPC and Its Analogs. When SPC solutions were adjusted to pH 7.3, the SPC HPLC peak at 4.2 min decreased rapidly and a new peak, which was

656 Chem. Res. Toxicol., Vol. 9, No. 3, 1996

Elfarra and Hwang

Figure 4. SPC half-life under different pH conditions.

Figure 5. 1H NMR spectrum of CMNPC in NaOD/D2O. The chemical shifts in parts per million are relative to the internal standard, sodium 3-(trimethylsilyl)tetradeuteriopropionate. The resonance at 4.8 ppm is for residual HOD.

Table 1. The Half-Lives (t1/2) of SPC and Its Analogs at pH 7.4 and 37 °C Chemical

t1/2

Chemical

t1/2

SPC SGC SPHC

35 min 32 min >80 days

NAPC SPG

>125 days >180 days

later identified as NPC, appeared at 6.4 min (Figure 3A). The increase in the NPC peak at various pH and temperature incubation conditions inversely correlated with the decrease of the SPC peak; log plots of the SPC rate of decomposition were linear to 2-3 half-lives (data not shown). The SPC degradation rate under neutral or mildly acidic or basic conditions (pH 5.7-8.75) was higher than that under more highly acidic (pH 3.6) or basic conditions (pH 9.5 or 10.5). The SPC half-life at pH 7.3 at room temperature was approximately 76 min, whereas the SPC half-life at pH 3.6 or pH 10.5 was estimated to be higher than 2000 min (Figure 4). The SPC degradation rate at pH 8.6 was dependent upon the temperature of the incubation; the SPC half-lives at 20, 36, and 48 °C were approximately 250, 40, and 25 min, respectively. When the stability of several SPC analogs was investigated at pH 7.4 at 37 °C, the results showed that SPHC, NAPC, and SPG were much more stable than SPC (Table 1). However, SGC stability under these conditions was similar to that of SPC. Intramolecular Rearrangement of SPC to NPC. Attempts to characterize purified peak II were not successful, possibly because of rapid oxidation of the free thiol group of NPC to form unknown products. However, because peak II was successfully converted to a stable peak (peak III, retention time: 5.3 min; Figure 3B) after the addition of 2-iodoacetic acid, isolated peak II was converted to peak III, which was purified by Sephadex LH-20 chromatography and identified by 1H NMR and FAB/MS as CMNPC (Figures 5 and 6). Each proton of CMNPC was assigned by its chemical shift, integration ratio, and coupling pattern; purinyl Hs, 8.20 (1H, s) and 8.07 ppm (1H, s); cysteinyl HR, 3.26 ppm (1H, merged s); cysteinyl Hβ, 3.25 (1H, dd) and 3.05 ppm (1H, dd); carboxymethyl Hs, 1.93 ppm (2H, s), respectively. The positive ion FAB/MS of CMNPC yielded a pseudomolecular ion of m/z 298, which is consistent with the expected molecular weight of 297 for CMNPC (Figure 6). The finding that the CMNPC precursor, NPC, did not react

Figure 6. Positive ion FAB/MS of CMNPC obtained using 3-nitrobenzyl alcohol as the matrix.

with ninhydrin is consistent with the structural assignment of CMNPC.

Discussion The results described in this paper show that SPC stability was highly dependent on pH and temperature. Intramolecular rearrangement of SPC to NPC (Figure 1) may be the most important pathway for SPC decomposition, since the decomposition product did not react with ninhydrin and no other peaks were detected by the HPLC and Sephadex LH-20 methods used to monitor SPC degradation. The identity of the decomposition product was confirmed by NMR and FAB/MS analyses after allowing the generated free thiol group to react with 2-iodoacetic acid (Figure 1). Johnston et al. (16) reported the hydrolysis of 6-[(2-hydroxyethyl)thio]purine to hypoxanthine during refluxing in either 0.1 N HCl or 0.1 N NaOH. However, SPC decomposition by hydrolysis was ruled out since hypoxanthine was not detected in any of our experiments; the retention time of reference hypoxanthine under the HPLC conditions described for Figure 3 was 3.0 min. The finding that the half-life of SPC under highly acidic (pH 3.6) or basic conditions (pH 9.55 or 10.5) was higher

S,N-Rearrangement of S-(Purin-6-yl)-L-cysteine

Figure 7. Mechanism of SPC (1) rearrangement to NPC (3) involving the formation of a 5-membered ring transition state (2).

than under neutral or mildly acidic or basic conditions (Figure 4) suggests the involvement of ionic groups on the cysteinyl and purinyl moieties of the SPC molecule in its rearrangement to NPC. Further evidence for this hypothesis was provided by the finding that SPC analogs which do not have a free cysteinyl amino group (NAPC and SPG) were much more stable than SPC (Table 1). The finding that SPHC was more stable than SPC provides evidence for the carbon length of the amino acid residue being important for this intramolecular rearrangement. A mechanism for the intramolecular rearrangement of SPC to NPC which is consistent with our results is presented (Figure 7). At neutral or slighltly acidic or basic conditions, the N-1 nitrogen of the purinyl moiety is expected to be protonated; pKa values (8.8-10.0) for N-1 hydrogen of several uncharged guanine derivatives are similar to that of the protonated amino group of cysteine (17, 18). This will increase the electrophilicity of the purinyl C-6 carbon. Intramolecular attack of the amino group of the cysteinyl moiety toward the purinyl C-6 carbon would result in the breakage of the C-S bond to form NPC. As the pH is increased, the N-1 nitrogen would be deprotonated and the purine moiety would be ionized to an anion, resulting in a decrease in the electrophilicity of the C-6 carbon; as the pH is decreased, the cysteinyl amino group would become more effectively protonated and lose its nucleophilicity. Thus, at either high basic or acidic conditions, the intramolecular nucleophilic displacement reaction would be difficult, resulting in the longer t1/2 of SPC we observed compared to neutral or slightly acidic or basic conditions. Compared to SPC, the additional carbon of SPHC decreases the rate of the reaction by possibly rendering the transition state more difficult to form. With SPHC, a 6-membered ring transition state is expected, whereas with SPC, a 5-membered ring transition state may be formed. The formation of a 6-membered ring transition state is kinetically less favored than the formation of a 5-membered ring (19, 20). NAPC and SPG were more stable than SPC, possibly because the secondary amino group of NAPC or SPG is much less nucleophilic than the primary amino group of SPC. The similar stabilities of SPC and SGC suggest that the exocyclic amino group of the guaninyl moiety does not significantly affect the electrophilicity of the C-6 carbon of the purinyl moiety. In summary, the results presented in this paper provide strong evidence for intramolecular rearrangement of SPC to NPC, the mechanism of this rearrangement, and the conditions under which rearrangement can be minimized. While formation of a 5-membered ring intermediate has previously been suggested for intramolecular rearrangements of O6-substituted guanine derivatives of styrene oxide (21), these rearrangements occurred at much faster rates under alkaline conditions than under neutral conditions. To our knowledge, SPC is the

Chem. Res. Toxicol., Vol. 9, No. 3, 1996 657

first cysteine S-conjugate which is shown to be stable under highly acidic or basic conditions but degrades readily under neutral or slightly acidic or basic conditions. The SPC results are significant because they provide valuable information as to how to minimize SPC degradation during storage and provide an explanation as to why rats given SPC exhibited much higher levels of SPC metabolites in their kidneys at 30 min compared to 60 or 90 min after SPC treatment (11-14). The SPC analog data may also explain why rats given SPHC excreted into their urine higher levels of 6-mercaptopurine than rats given an equimolar dose of SPC. Furthermore, the results presented in this paper may contribute to a better understanding of the stability of other cysteine S-conjugates, such as the cysteine S-conjugates of triazine herbicides and other related chemicals.

Acknowledgment. This research was supported in part by Grant DK-44295 from NIDDK, National Institutes of Health.

References (1) Elfarra, A. A. (1995) Potential role of the flavin-containing monooxygenases in the metabolism of endogenous compounds. Chem.-Biol. Interact. 96, 47-55. (2) Elfarra, A. A. (1993) Aliphatic halogenated hydrocarbons. In Toxicology of the Kidney (Hook, J. B., and Goldstein, R. S., Eds.) 2nd ed., pp 387-413, Raven Press, New York. (3) Monks, T. J., Anders, M. W., Dekant, W., Stevens, J. L., Lau, S. S., and van Bladeren, P. J. (1990) Glutathione conjugate-mediated toxicities. Toxicol. Appl. Pharmacol. 106, 1-19. (4) Hay, D. W. P., Torphy, T. J., and Undem, B. J. (1995) Cysteinyl leukotrienes in asthma: old mediators up to new tricks. Trends Pharmacol. Sci. 16, 304-309. (5) Peterson, L. A., Harris, T. M., and Guengerich, F. P. (1988) Evidence for an episulfonium ion in the formation of S-[2-N7guanyl)ethyl]glutathione in DNA. J. Am. Chem. Soc. 110, 32843291. (6) Schasteen, C. S., and Reed, D. J. (1983) The hydrolysis and alkylation activities of S-(2-haloethyl)-L-cysteine analogs. Evidence for extended half-life. Toxicol. Appl. Pharmacol. 70, 423432. (7) Webb, W. W., Elfarra, A. A., Webster, K. D., Thom, R. E., and Anders, M. W. (1987) Role for an episulfonium ion in S-(2chloroethyl)-DL-cysteine-induced cytotoxicity and its reaction with glutathione. Biochemistry 26, 3017-3023. (8) Ito, S., Inoue, S., Yamamoto, Y., and Fujita, K. (1981) Synthesis and antitumor activity of cysteinyl-3,4-dihydroxyphenylalanines and related compounds. J. Med. Chem. 24, 673-677. (9) Burchfield, H. P. (1958) Molecular rearrangement in the reaction of cysteine with 1-fluoro-2,4-dinitrobenzene. Nature 181, 49-50. (10) Holy, A., Votruba, I., and Jost, K. (1974) S-(Pyrimidin-2-yl)-Lcysteine: Chemical synthesis and biosynthesis in Escherichia coli. Collect. Czech. Chem. Commun. 39, 634-646. (11) Elfarra, A. A., and Hwang, I. Y. (1993) Targeting of 6-mercaptopurine to the kidneys: Metabolism and kidney-selectivity of S-(6-purinyl)-L-cysteine analogs in rats. Drug Metab. Dispos. 21, 841-845. (12) Hwang, I. Y., and Elfarra, A. A. (1993) Detection and mechanisms of formation of S-(6-purinyl)glutathione and 6-mercaptopurine in rats given 6-chloropurine. J. Pharmacol Exp. Ther. 264, 41-46. (13) Hwang, I. Y., and Elfarra, A. A. (1991) Kidney-selective prodrugs of 6-mercaptopurine: Biochemical basis of the kidney selectivity of S-(6-purinyl)-L-cysteine and metabolism of new analogs in rats. J. Pharmacol Exp. Ther. 258, 171-177. (14) Hwang, I. Y., and Elfarra, A. A. (1989) Cysteine S-conjugates may act as kidney-selective prodrugs: Formation of 6-mercaptopurine by the renal metabolism of S-(6-purinyl)-L-cysteine. J. Pharmacol. Exp. Ther. 251, 448-454. (15) Elfarra, A. A., Duescher, R. J., Hwang, I. Y., Sicuri, A. R., and Nelson, J. A. (1995) Targeting of 6-thioguanine to the kidney with S-(guanin-6-yl)-L-cysteine. J. Pharmacol. Exp. Ther. 274, 12981304. (16) Johnston, T. P., Holum, L. B., and Montgomery, J. A. (1958) Synthesis of potential anticancer agents. XVI. S-Substituted derivatives of 6-mercaptopurine. J. Am. Chem. Soc. 80, 62656271.

658 Chem. Res. Toxicol., Vol. 9, No. 3, 1996 (17) Shapiro, R. (1968) Chemistry of guanine and its biologically significant derivatives. In Progress in Nucleic Acid Research and Molecular Biology (Davidson, J. N., and Cohn, W. E., Eds.) Vol. 8, pp 73-112, Academic Press, New York. (18) Walsh, C. (1979) Chapter 2: Introductory Remarks About Enzymes and Enzymatic Catalysis. In Enzymatic Reaction Mechanisms, pp 24-48, Freeman, New York. (19) Baldwin, J. E. (1976) Rules for ring closure. J. Chem. Soc., Chem. Commun., 734-736.

Elfarra and Hwang (20) March, J. (1985) Chapter 6: Mechanisms and Methods of Determining Them. In Advanced Organic Chemistry, 3rd ed., pp 179-201, Wiley, New York. (21) Moschel, R. C., Hemminki, K., and Dipple, A. (1986) Hydrolysis and rearrangement of O6-substituted guanosine products resulting from reaction of guanosine with styrene oxide. J. Org. Chem. 51, 2952-2955.

TX9501905