Anal. Chem. 2008, 80, 209-216
Electrochemical Determination of HIV Drug Abacavir Based on Its Reduction Burcu Dogan,† Bengi Uslu,† Sibel A. Ozkan,† and Petr Zuman*,‡
Faculty of Pharmacy, Department of Analytical Chemistry, Ankara University, 06100 Ankara, Turkey, and Department of Chemistry, Clarkson University, Potsdam, New York 13699-5810
Abacavir (I), a drug used in the treatment of HIV, is electrochemically reduced at the dropping mercury electrode in a four-electron process, similar to structurally related adenine (III) and adenosine triphosphate (IV). To undergo the reduction, the species is protonated in the vicinity of the electrode. The protonations take place on the 6-amino group and on one of the pyrimidine ring nitrogens. The role of covalent hydration of the pyrymidine ring has been interpreted. Best suited as supporting electrolytes for analytical purposes are solutions of 0.11.0 M sulfuric, perchloric, or hydrochloric acids. Procedures of analyses of tablets containing I were established and validated, based on peak currents obtained by linear sweep, differential pulse, or square-wave voltammetry with a hanging mercury drop electrode as indicator electrode. The procedure proved to be more sensitive and more reliable than that based on oxidation on a glassy carbon electrode, proposed previously. Abacavir (Ziagen) (I, ABA) {(1S,cis)-4-[2-amino-6-(cyclopropylamino)-9H-purin-9-yl]-2-cyclopentene-1-methanol] is a synthetic
carbocyclic nucleoside analogue, that has an inhibitory activity against HIV. In the treatment of an HIV infection, I is administered * To whom correspondence should be addressed. E-mail: clarkson.edu. Fax: 001-315-268 6610. † Ankara University. ‡ Clarkson University. 10.1021/ac0713151 CCC: $40.75 Published on Web 11/28/2007
zumanp@
© 2008 American Chemical Society
either alone or with another antiretroviral drug. Abacavir is converted intracellularly by several steps into carbovir triphosphate II. Carbovir triphosphate is an analogue of deoxyguanosine
5′-triphosphate and inhibits the activity of HIV-1 reverse transcriptase. This is achieved both by competing with the natural substrate (deoxyguanosine 5′-triphosphate) and by its incorporation into viral DNA. Abacavir shows a good oral bioavailability and significantly penetrates the cerebrospinal fluid.1,2 There have been few analytical procedures for the determination of abacavir, based on liquid chromatography with UV detection3-7 and with tandem mass spectrometry8 and based on voltammetry.9 The reported liquid chromatographic methods are affected by the interference of endogenous substances and potential loss of drugs in the re-extraction procedure. They involve lengthy, tedious, and time-consuming plasma sample preparation and extraction operations and require expensive instrumentation. The voltammetric analysis involved oxidation of abacavir on a glassy carbon electrode.9 To develop an electroanalytical procedure for determination of ABA (I), based on its reduction, its behavior was compared with the reduction of electroreduction of adenine and its phosphate (1) Sweetman, S. C., Ed. Martindale, The Complete Drug Reference, 32th ed.; Pharmaceutical Press: London, 2002; p 612. (2) Physicians Desk Reference (PDR); Medical Economics Co. Inc.: Montvale, NJ, 2003; p 1664. (3) Veldkamp, A. I.; Sparidans, R. W.; Hoetelmans, R. M. W.; Beijnen, J. H. J. Chromatogr., B 1999, 736, 123-128. (4) Aymard, G.; Legrand, M.; Trichereau, N.; Diquet, B. J. Chromatogr., B 2000, 744, 227-240. (5) Sparidans,R. W.; Hoetelmans, R. M. W.; Beijnen, J. H. J. Chromatogr., B 2001, 750, 155-161. (6) Rezk, N. L.; Tidwell, R. R.; Kashuba, A. D. M. J. Chromatogr., B 2003, 791, 137-147. (7) Ozkan, Y.; Savaser, A.; Ozkan, S. A. J. Liq. Chromatogr. Relat. Technol. 2005, 28, 423-437. (8) Fung, E. N.; Cai, Z.; Burnette, T. C.; Sinhababu, A. K. J. Chromatogr., B 2001, 754, 285-295. (9) Uslu, B.; O ¨ zkan, S. A. Electrochim. Acta 2004, 49, 4321-4329.
Analytical Chemistry, Vol. 80, No. 1, January 1, 2008 209
derivatives, the only 6-aminopurines reduced at the dropping mercury electrode (DME).10-20 Unfortunately, the conclusions about the electrode processes involved in the reduction differed considerably. There was no agreement about the number of electrons transferred in the first reduction step, the number of reduction waves, or the role of pH and the buffer composition. The situation was partly caused by the neglect of covalent hydration, partly by using the results obtained by a controlled potential electrolysis using a mercury electrode with a large surface area for interpretation of processes at DME. As the time scale, during which electrolysis is carried out, differs by several orders of magnitude for these two types of electrodes, it is now accepted,21,22 that for electroreductions, involving formation of reactive products or intermediates, the processes occurring when these two techniques were used can follow different reaction patterns. Furthermore, the products obtained, when these two techniques were used, may be different.21,22 It was hence necessary, before developing a reliable analytical procedure for the determination of ABA (I) based on electroreduction, to obtain some basic information concerning the processes involved. Such information was obtained using DME and dc polarography, which often offers most straightforward information about the sequences and numbers of proton transfers and electroreductions in protic media. To achieve this, the behavior of ABA (I) on DME was compared with that of adenine (III) and adenosine triphosphate (IV). The information obtained in such
comparison was then used for interpretation of the electrochemical behavior of I, when linear sweep and cyclic voltammetry on a hanging mercury drop electrode (HMDE) were used. Such electrode and techniques were then used in practical analytical applications. (10) Heath, J. C. Nature 1946, 158, 23. (11) McGinn, F. A.; Brown, G. B. J. Am. Chem. Soc. 1960, 82, 3193-3195. (12) Luthy, N. G.; Lamb, B. J. Pharm. Pharmacol. 1956, 8, 410-416. (13) Skulachev, V. P.; Denisovich, L. I. Biokhimiya 1966, 31, 132-136. (14) Janik, B.; Elving, P. J. Chem. Rev. 1968, 68, 295-319. (15) Smith, D. L.; Elving, P. J. J. Am. Chem. Soc. 1962, 84, 1412-1420. (16) Smith, D. L.; Elving, P. J. Anal. Chem. 1962, 34, 930-936. (17) Dryhurst, G.; Elving, P. J. Talanta 1969, 16, 855-874. (18) Janik, B.; Elving, P. J. J. Am. Chem. Soc. 1970, 92, 235-243. (19) Janik, B.; Elving, P. J. J. Electrochem. Soc. 1970, 117, 457-460. (20) Elving, P. J.; Page, S. J.; O’Reilly, J. E. J. Am. Chem. Soc. 1973, 95, 647658. (21) Zuman, P. In Organic Electrochemistry, 2nd ed.; Baizer, M. M., Lund, H., Eds.; Marcel Dekker Pub.: New York, 1983; p 151. (22) Zuman, P.; Ludvik, J. Electroanalysis 2000, 12, 879-888.
210
Analytical Chemistry, Vol. 80, No. 1, January 1, 2008
EXPERIMENTAL SECTION Instrumentation. Polarographic measurements were carried out using a Lynseis LY 1600 dc polarograph. The mercury dropping electrode had a t1 of 3.2 s, m of 2.3 mg‚s-1 at h of 70 cm. A saturated calomel electrode was used as the reference electrode. For linear sweep, cyclic, square wave, and differential pulse voltammetry a hanging mercury drop electrode was used. The current-voltage curves were recorded using a Bioanalytical Systems (BAS 100 W) electrochemical analyzer, with Pt wire (BAS) counter electrode and an Ag/AgCl (BAS; 3 M KCl) reference electrode. Operating conditions for CV were as follows: scan rate, between 10 and 100 mV/s; initial direction, negative. For square wave voltammetry (SWV) they were as follows: pulse amplitude, 25 mV; frequency,15 Hz; and potential step, 4 mV. For differnetial pulse voltammetry (DPV) they were as follows: pulse amplitude, 25 mV; pulse width, 50 ms; and scan rate, 20 mV/s. pH values were determined using a model 538 pH meter (WTW) and calibrated with standard buffers (Fixanal, Riedel-de Haen, Germany) at room temperature. A Shimadzu 1601 PC double-beam spectrophotometer equipped with 1.0-cm quartz cells with a fixed slit width (2 nm) was coupled an IBM-PC computercontrolled spectrophotometric Schimadzu UVPC software. Reagents. Abacavir and its pharmaceutical dosage form (Ziagen) were kindly provided by Glaxo Smith Kline Pharm. Ind. (Istanbul, Turkey). Adenine and adenosine were supplied by Sigma. All chemicals for preparation of buffers and supporting electrolytes were reagent grade (Merck or Sigma). All stock solutions (typically 1 mM) of electroactive species were prepared freshly every week and kept in a refrigerator. The following supporting electrolytes were used: sulfuric acid (1, 0.3, 0.2, 0.1, and 0.03 M with 0.07 M Na2SO4 and 0.01 M with 0.1 M Na2SO4), acetate buffers (0.1 M, pH 3.7-5.7), phosphate buffers (0.05 M, pH 2.0 -11.0), a borate buffer (0.2 M, pH 9.3), and 0.1 M sodium hydroxide solutions. Procedures. Solutions of chosen supporting electrolytes were placed into the electrolytic cell at room temperature, and oxygen was removed by passage of nitrogen through the solution for 2 min. A stock solution of the electroactive compound was added to a final concentration varying between 1 × 10-4 and 2.0 mM. Nitrogen was introduced for another 1 min and the currentvoltage curve recorded. In the supporting electrolytes used, the current-voltage curves remained unchanged for at least 72 h. RESULTS AND DISCUSSION Spectrophotometry. In aqueous solutions of I, characteristic absorption spectra were observed in two wavelengths regions: one between 240 and 260 nm and the other between 280 and 310 nm. The absence of acid-base equilibria between pH 7 and 13 was confirmed by the absence of changes in this pH range. Two acidbase equilibria caused variations in absorbance in more acidic media. At a pH between 7 and 3, a decrease of the absorbance at 284 nm with increasing acidity was accompanied by an increase of that at 295 nm. The dependences of these absorbances on pH correspond to an acid dissociation with pK2 ) 5.0, in agreement with the reported value23 of pK2 ) 5.1. Practically the same value of pK2 was obtained by a potentiometric titration of 0.001 M (23) http://www.gsk.ca/en/products/prescription/Trizivir_PM_20070314.pdf.
solution of I by 0.1 M NaOH. In acidic solutions containing between 0.01 and 10.0 M H2SO4, with increasing acid concentration, the absorbance at 295 nm decreased and that at 315 nm increased. At [H2SO4] > 3.0 M, the absorbance was extrapolated to t ) 0 to correct for the effect of slow hydrolysis. At a pH of pKa with a slope of ∼75 mV/pH. The pH at the intersection of the two linear segments (Figure 2B) corresponds to pKa. The value for pK1 obtained for I from the shifts of half-wave potentials with pH was ∼0.8. This is within an acceptable agreement with the value of pK1 ) 0.3 obtained by spectrophoAnalytical Chemistry, Vol. 80, No. 1, January 1, 2008
211
Figure 2. Dependence of polarographic limiting currents (A) and of half-wave potentials (B) on pH of 0.1 mM solutions of abacavir (I) (0) in solutions of sulfuric acid and in solutions of adenine (III) in sulfuric acid (2) in acetate buffers (3) and in phosphate buffers (4) and of adenosine triphosphate (IV) in buffers.(b). Limiting currents in the same buffers in buffers corrected for the increase of the molecular mass (×).
tometry, taking into account the differences in ionic strengths and the limited accuracy of the value obtained by polarography, due to the pH range limited to pH
