- Email: [email protected]
Fluorescence and terbium-sensitised luminescence determination of garenoxacin in human urine and serum
Talanta 63 (2004) 691–697
Fluorescence and terbium-sensitised luminescence determination of garenoxacin in human urine and serum J.A. Ocaña, F.J. Barragán∗ , M. Callejón Department of Analytical Chemistry, Faculty of Chemistry, E-41012 Seville, Spain Received 7 May 2003; received in revised form 11 November 2003; accepted 11 December 2003 Available online 3 February 2004
Abstract Fluorescence and terbium-sensitised luminescence properties of new quinolone garenoxacin have been studied. The fluorimetric method allows the determination of 0.060–0.600 g ml−1 of garenoxacin in aqueous solution containing HCl/KCl buffer (pH 1.5) with λexc = 282 nm and λem = 421 nm. Micellar-enhanced fluorescence was also studied, leading to a higher than 400% increase in analytical signal in presence of 12 mM sodium dodecyl sulphate (SDS), allowing the determination of 0.020–0.750 g ml−1 of garenoxacin. The terbium-sensitised luminescence method allows the determination of 0.100–1.500 g ml−1 of garenoxacin in 12 mM SDS solution containing 0.08 M acetic acid/sodium acetate buffer (pH 4.1) and 7.5 mM Na2 SO3 (chemical deoxygenation agent), with λexc = 281 nm and λem = 546 nm. Relative standard deviation (R.S.D.) values for the three methods were in the range 1.0–2.0%. The proposed procedures have been applied to the determination of garenoxacin in spiked human urine and serum. © 2004 Elsevier B.V. All rights reserved. Keywords: Garenoxacin; Fluorescence; Micelle-enhanced fluorescence; Luminescence; Terbium; Urine; Serum
1. Introduction Garenoxacin {1-cyclopropyl-8-(difluoromethoxy)-7-[(1R)-(1-methyl-2,3-dihydro-1H-5-isoindolyl]-4-oxo-1,4-dihydro-3-quinolinecarboxylic acid} is a new des-F(6)-quinolone antibacterial agent, formerly known as BMS-284756 and T-3811. This antibiotic differs from earlier quinolones in its lack of a fluorine atom in its C6 position, and the presence of an isoindolin-5-yl group at the 7th position [1]. Garenoxacin has broad-spectrum antibacterial activity against both Gram-positive and Gram-negative organisms, through inhibition of their DNA gyrase [2]. It also shows excellent oral bioavailability [3], with final concentrations in serum and urine of test-subjects of 2.0–10.0 and 20–100 g ml−1 , respectively, after 600 mg administration [4]. The analytical literature of this quinolone is very scarce. HPLC with fluorimetric detection has been applied to the determination of garenoxacin in mouse serum [5], and LC–MS for human urine and serum [6], but no other analytical ∗
Corresponding author. Tel.: +95-455-71-71; fax: +95-455-71-68. E-mail address: [email protected] (F.J. Barrag´an).
0039-9140/$ – see front matter © 2004 Elsevier B.V. All rights reserved. doi:10.1016/j.talanta.2003.12.016
reports have been found. This work constitutes the first stage of a research schedule focused on the study of the analytical properties of garenoxacin and the proposal of procedures for its determination in human biological media. Quinolones frequently exhibit native fluorescence properties that can be applied to their determination in aqueous solution and in biological matrix as urine and serum. This technique has been applied for the determination of an important number of quinolones, as ciprofloxacin [7], norfloxacin [8] and ofloxacin [9], among others. This fluorescence can be exalted in presence of an appropriate micellar medium, leading to better sensitivities and detection limits: this property has been applied to the study of quinolones as ciprofloxacin [10], levofloxacin [11] and moxifloxacin [12], among others. Quinolones also have suitable functional groups to form stable complexes with terbium ions. The presence of these ions in quinolone solutions leads to the formation of complexes that absorb energy at the characteristic wavelength of the organic ligand, and emit radiation at the characteristic wavelength of Tb3+ (546 nm). The mechanism of this emission is based in the quinolone being excited to its triplet state, which subsequently acts as an energy donor to the emitting energy level of Tb3+ . As the triplet state has a
692
J.A. Ocaña et al. / Talanta 63 (2004) 691–697
relatively long lifetime, the emitted radiation can be registered as a time-resolved luminescence signal, achieving an important decrement in background signal and better selectivity. This requires appropriate experimental conditions, as micellar medium and the presence of a deoxygenation agent as Na2 SO3 , in order to avoid the non-radiant deactivation process favoured by the long lifetime of the triplet state. The technique has been applied to the determination of the quinolones ciprofloxacin and enrofloxacin [13], levofloxacin [14], grepafloxacin [15] and trovafloxacin [16], among others. The present paper describes the spectrofluorimetric, micelle-enhanced spectrofluorimetric and terbium-sensitised luminescence properties of garenoxacin, and their application to its determination in spiked urine and serum.
10 g ml−1 was prepared daily by dilution of stock standard solution with high-purity water. 2.2. Apparatus Fluorescence and luminescence intensities were measured on a Perkin-Elmer (Norwalk, CT, USA) LS-5 luminescence spectrometer equipped with a xenon-lamp and an Acer Model 1030 computer working with the FLUORPACK software from Sciware (Mallorca, Spain). For luminescence measurements, the instrument was set in the phosphorescence mode and delay time (td ) of 0.03 ms and gate time (tg ) of 8 ms were used. All the measurements took place in a standard 10 mm pathlength quartz cell, thermostated at 25.0 ± 0.5 ◦ C, with 5 nm bandwidths for the emission and excitation monochromators. The pH was measured on a Crison (Barcelona, Spain) micropH 2002 pH-meter. 2.3. Sample preparation
2. Procedures 2.1. Reagents Garenoxacin (100.0% purity, data supplied by the provider) was kindly donated by Bristol Myers Squibb (USA). Terbium (III) chloride hexahydrate, sodium dodecyl sulphate (SDS), sodium dodecylbenzensulphonate (NaDDB), Triton® X-100 and hexadecyltrimethyl-ammonium bromide (HDTAB) were purchased from Fluka (Madrid, Spain). Concentrated acetic acid, sodium acetate, sodium sulphite, sodium hydroxide, acetonitrile and hydrochloric acid, were of analytical-reagent grade and purchased from Merck (Darmstadt, Germany). High-purity water was obtained from a Millipore (Milford, MA USA) Milli-Q Plus System. Urine and serum samples were obtained from several healthy volunteers. For fluorescence and micelle-enhanced fluorescence determinations, a HCl/KCl buffer of pH 1.5 was prepared by mixing 20.1 ml HCl 0.2 M and 29.9 ml KCl 0.2 M, and diluting to 100 ml high-purity water. For terbium-sensitised luminescence determination, a 0.8 M acetic acid/sodium acetate buffer of pH 4.1 was prepared by mixing 40 ml of 1.6 M acetic acid with 10 ml of 1.6 M sodium acetate, and diluting to 100 ml with high-purity water. Stock standard solution of 200 g ml−1 was prepared by dissolving garenoxacin in high-purity water, being stable for several weeks at room temperature. Working standard of
The proposed methods were applied to the determination of garenoxacin in spiked human urine and serum provided from several healthy volunteers. Urine and serum samples were spiked with convenient amounts of garenoxacin stock solutions, for final antibiotic concentrations in the range 20–100 and 2.0–10.0 g ml−1 , respectively. Spiked urine was 50-fold diluted with highpurity water. No serum pre-treatment was required for aqueousfluorimetric and micelle-enhanced fluorimetric methods. For terbium-sensitised luminescence method, a deproteinization process with acetonitrile was employed, as the untreated serum samples led to turbid solutions in presence of Tb3+ . So, 1 ml spiked serum samples were deproteinized with 2 ml of acetonitrile, by vortexing for 5 min and centrifuging for 5 min at 1500 × g. 2.4. Measurement of standard garenoxacin solutions For aqueous-medium fluorimetric determination, in 25 ml volumetric flasks, aliquots of working solutions of garenoxacin with 5 ml of 0.1 M HCl/KCl buffer (pH 1.5) were pipetted and diluted to the mark with water, for a final quinolone concentration in the range 0.060–0.600 g ml−1 . The obtained solutions were thermostated at 25±0.1 ◦ C, and the fluorescence emission at 421 nm against a blank solution was measured using an excitation wavelength of 282 nm. This procedure was also applied for the micelle-enhanced fluorimetric determination of garenoxacin, adding 2 ml of 0.15 M SDS solution before diluting to the mark, with a quinolone concentration in the range 0.020– 0.750 g ml−1 . For the terbium-sensitised luminescence determination, aliquots of working garenoxacin solution were pipetted
J.A. Ocaña et al. / Talanta 63 (2004) 691–697
into 10 ml calibrated flasks. Then, 1 ml of 0.8 M acetic acid-sodium acetate buffer (pH 4.1), 1 ml of 0.03 M Tb3+ solution, 1 ml of 0.15 M SDS solution and 1 ml of 0.075 M Na2 SO3 solution were added, diluting to the mark with high-purity water. Final garenoxacin concentration in samples was in the range 0.100–1.500 g ml−1 . The solutions were thermostated at 25 ± 0.1 ◦ C and the luminescence was measured after 5 min at 546 nm, using an excitation wavelength of 281 nm, against a blank solution. 2.5. Determination of garenoxacin in urine and serum For the determination of garenoxacin in urine, 1 ml of the diluted urine was pipetted in a 25 ml (aqueous and micelle-enhanced fluorescence methods) or a 10 ml (terbium-sensitised luminescence method) calibrated flask. For the determination of garenoxacin in serum, 1 ml of untreated serum was pipetted in a 25 ml calibrated flask for aqueous and micelle-enhanced fluorescence methods, and 0.25 ml of deproteinized serum was pipetted in a 10 ml calibrated flask for the terbium-sensitised luminescence method. Solutions were prepared and emission was measured as explained for standard garenoxacin solutions. In order to compensate the effect of the biological matrix in the measurements, standard addition method was applied to the quantification of the quinolone for three proposed methods in both urine and serum.
3. Results and discussion 3.1. Preliminary studies Garenoxacin showed a notable fluorescence emission at very acidic pH values, but showed a very weak emission
693
for higher pH. Fig. 1 shows the excitation and emission spectra obtained for a 0.250 g ml−1 garenoxacin solution at pH 1.5. The quinolone showed an excitation maximum at 282 nm, and an emission maximum at 421 nm. Thus, these wavelengths were selected as suitable for further fluorimetric studies. Garenoxacin shown no native time-resolved emission, but a notable signal was found in presence of 2.0 mM Tb3+ in a 10 mM SDS micellar medium with pH 4.0. Tb3+ solutions showed negligible time-resolved luminescence compared with Tb3+ –garenoxacin in the same conditions. Fig. 2 shows the excitation and emission luminescence spectra of the complex, with a maximum excitation and emission wavelengths of 281 nm (corresponding to the excitation of the quinolone) and 546 nm (corresponding to the characteristic emission of Tb3+ cation), respectively. The terbium-sensitised luminescence intensity was measured with 0.01–10 ms td and 1–10 ms tg . It was found that Rayleigh scattering could be eliminated using a 0.03 ms td , and maximum luminescence intensity was found with a 8 ms tg , so these values were selected for further studies. 3.2. Aqueous fluorescence properties The effect of pH value in the fluorescence emission of a 0.250 g ml−1 garenoxacin solution is shown in Fig. 3A. As can be seen, analytical signal exhibits a maximum when pH was 1.3–1.8, so pH of 1.5, was chosen for further studies. This pH was achieved by the addition of a HCl/KCl 0.1 M buffer, prepared as explained above. The influence of the buffer concentration on garenoxacin fluorescence was studied by the addition of increasing volumes of buffer, for final concentrations in the range 4–60 mM. Fluorescence intensity remains stable for lower than 40 mM buffer concentration, and it decreased for higher values. Accordingly, a 20 mM buffer concentration
Fluorescence intensity (arbritary units)
40
30 EM
EXC 20
10
0 250
300
350
400
450
500
λ (nm) Fig. 1. Fluorescence excitation (λem = 421 nm) and emission (λexc = 282 nm) spectra of 0.250 mg ml−1 garenoxacin at pH 1.5.
694
J.A. Ocaña et al. / Talanta 63 (2004) 691–697
Terbium sensitised luminescence intensity (arbritary units)
30
20
EM 10
EX
0 250
300
350
400
450
500
550
600
λ (nm) Fig. 2. Terbium-sensitised luminescence excitation (λem = 546 nm) and emission (λexc = 281 nm) spectra at of 0.250 mg ml−1 garenoxacin in 2.0 mM Tb3+ and 10 mM SDS medium. pH 4.0. td = 0.03 ms, tg = 8 ms.
(corresponding to a 5 ml aliquot of the prepared 0.1 M buffer solution in 25 ml total volume) was selected as a suitable volume for the recommended procedure. The influence of the temperature on the fluorescence intensity shows a nearly linear (negative) relationship between temperature and fluorescence intensity for garenoxacin, with a temperature coefficient of −1.15%. This value showed that internal conversion was probably the main non-radiant deactivation process for garenoxacin [17]. Thus, samples were thermostated at 25 ± 0.5.
tions of the surfactants SDS (cationic), HDTAB (cationic), NaDDB (anionic) and Triton® X-100 (non-ionic) at pH 1.50. It was observed that no significant fluorescence enhancement was achieved in presence of HDTAB or NaDDB. Triton® X-100 addition led an enhancement in analytical signal of only a 20%, but a notable increase in fluorescence was observed in presence of SDS. As can be seen in Fig. 4, fluorescence reached a constant and maximum value for higher than 10 mM SDS concentration, so a 12 mM SDS concentration was selected for further studies. Fig. 3B shows the influence of pH on the luminescence of 0.100 g ml−1 garenoxacin in 12 mM SDS medium. As can be seen, optimum pH range obtained was similar than that found for aqueous solution, so pH of 1.5, obtained with 5 ml of HCl/KCl 0.1 M buffer for a final solution volume of 25 ml, was chosen for the recommended procedure.
3.3. Micelle-enhanced fluorescence properties
Fluorescence and Luminescence Intensity (arbritary units)
The fluorimetric properties of garenoxacin were studied in a series of micellar media, by preparing 0.200 g ml−1 quinolone samples in the presence of increasing concentra110 B 90
70 C 50
A
30
10 1
2
3
4
5
6
7
8
pH
Fig. 3. Effect of pH in: (A) fluorescence of 0.250 g ml−1 garenoxacin (λexc = 282 nm, λem = 421 nm); (B) micellar-enhanced fluorescence of 0.100 g ml−1 garenoxacin (λexc = 282 nm, λem = 421 nm, SDS = 12 mM) and (C) terbium-sensitised luminescence of 0.300 g ml−1 garenoxacin (λexc = 281 nm, λem = 546 nm, SDS = 10 mM, Tb3+ = 2.0 mM).
J.A. Ocaña et al. / Talanta 63 (2004) 691–697
695
Fluorescence intensity (arbritary units)
80
60
40
20
0 0
4
8
12
16
20
[SDS] (mM)
Fig. 4. Effect of SDS concentration in fluorescence of 0.250 g ml−1 garenoxacin (λexc = 282 nm, λem = 421 nm, pH 1.50).
3.4. Terbium-sensitised luminescence properties Fig. 3C shows the influence of pH on luminescence intensity. As can be seen, the optimum pH value is in the range 3.8–4.5. So, acetic acid/sodium acetate buffer of pH 4.1 was selected for the recommended procedure. The influence of the buffer concentration on the analytical signal was studied in the range 0.04–0.20 M. The optimal performance was found in the range 0.06–0.10 M. Thus, a 0.08 M buffer concentration was selected as suitable for the optimised procedure. The effect of SDS concentration was studied measuring the luminescence emission of a series of samples with tensoactive concentrations in the range 4–20 mM. The luminescence intensity reached a maximum and constant value for higher than 10 mM SDS concentrations, so a value of 12 mM was selected for further studies. A series of samples with terbium concentration in the range 0.5–4.0 mM were prepared. The luminescence intensity measured showed a stable value for concentrations of Tb3+ above 2.0 mM, so a concentration of 3.0 mM was chosen as suitable. Lastly, it was observed that a minimum concentration of Na2 SO3 of 7.5 mM was required to effectively eliminate oxygen completely from the solutions, obtaining the maximum analytical signal 5 min after Na2 SO3 addition, stable for more than an hour. The luminescence intensity decreased notably for higher than 15 mM Na2 SO3 concentrations, so a Na2 SO3 concentration of 7.5 mM was chosen as suitable, which led to a final 50% enhancement in analytical signal with respect to measurements carried in absence of the deoxygenation agent.
concentrations between 0.001 and 1.0 g ml−1 were prepared by triplicate and measured by following the procedures described in Section 2. Calibration curves of fluorescence, micelle-enhanced fluorescence and terbium-sensitised luminescence intensities (Y) versus garenoxacin concentration expressed in micrograms per millilitre (X) were constructed; Table 1 resumes the results obtained from statistical analysis of data. As can be seen, micelle-enhanced fluorescence method achieved the highest sensitivity, with a 450% increase in slope with respect to the aqueous fluorescence method. Detection limit (DL) and quantification limit (QL) were calculated according to the recommendations of the Analytical Methods Committee [18]. The values obtained are resumed in Table 1. Again, best results were achieved with micellar-enhanced fluorescence method. In order to evaluate the precision of the proposed methods, 11 replicates (0.300 g ml−1 garenoxacin for aqueous and micelle-enhanced fluorescence; 0.600 g ml−1 garenoxacin
Table 1 Figures of merit of the aqueous-solution fluorescence, micelle-enhanced fluorescence and terbium-sensitised luminescence determination of garenoxacin Parameter
Aqueous solution method
Micelle-enhanced method
Time-sensitised luminescence method
3.5. Figures of merit of the proposed methods
LRC (g ml−1 ) a ± Sa b ± Sb r R.S.D. (%) DL (g ml−1 ) QL (g ml−1 )
0.060–0.600 0.3 ± 0.1 67.9 ± 0.3 0.9997 2.0 0.015 0.045
0.020–0.750 1.5 ± 0.2 297.8 ± 0.4 0.9999 1.2 0.005 0.018
0.100–1.500 5.7 ± 0.3 56.4 ± 0.3 0.9996 1.3 0.020 0.065
Once the three proposed method were optimised, three series of ten standard solutions of garenoxacin with
LRC, linear calibration range; a, intercept; b, slope; Sa , intercept standard deviation; Sb , slope standard deviation; r, correlation coefficient; R.S.D., relative standard deviation; DL, detection limit; QL, quantification limit.
696
J.A. Ocaña et al. / Talanta 63 (2004) 691–697
Table 2 Tolerance of different compounds in the determination of garenoxacin
3.6. Determination of garenoxacin in urine and serum
Compounds added
Garenoxacin in urine and serum with a content of 20–100 and 2–10 g ml−1 , respectively, was analysed as described under experimental by the three proposed procedures, using the standard addition method. External calibration was not applied in order to avoid matrix effects of urine or serum affecting the results (slope of the standard addition curves in these media were approximately 10–20% lower than slope found for aqueous standards for the three proposed methods). The garenoxacin concentration obtained for five replicates are resumed in Table 3. In order to compare the obtained recoveries with the spiked concentration, a Student’s t-test [19] was used. In all cases, no statistically significant differences between the spiked and the experimental concentrations were observed at 95% significance level.
Maximum tolerance weight ratio
Glucose, lactose, fructose, sacarose Amikacin Aspirin Imipenem Paracetamol Penicilin Sulbactam
Aqueous fluorescence
Micellar fluorescence
Terbium-sensitised luminescence
2000a
2000a
2000a
30a 30a 1.0 6.0 40a 30
30a 30a 1.6 6.0 40a 30
30a 30a 3.2 9.0 40a 40a
a Maximum weight ratio tested (no interference was found at this level).
for terbium-sensitised luminescence), were quantified. The relative standard deviation (R.S.D.) obtained were in the range 1.0–2.0% (Table 1) for three methods. Lastly, the interference of typical excipients and co-administer drugs (mainly analgesics or antibiotics) was studied by the addition of increasing concentrations of these compounds to a 0.300 g ml−1 garenoxacin solution (aqueous and micellar-enhanced fluorescence) or a 0.600 g ml−1 (terbium-sensitised luminescence). A substance was considered to interfere at the weight interference–quinolone ratio that led to errors greater than the three times the precision level for each method (3× R.S.D). For some of the excipients and drugs, no interference was found even at the maximum tested weight interference–quinolone ratio; for the rest, only negative errors were found. Maximum tolerable weight interference–quinolone ratio for the proposed methods are given in Table 2. As can be seen, best selectivity was found for the terbium-sensitised luminescence method.
4. Conclusions This work shows that fluorescence, micelle-enhanced fluorescence and terbium-sensitised luminescence may be simple, accurate, fast and precise procedures to determination of garenoxacin in both human urine and serum. Best sensitivity was achieved when micelle-enhanced fluorescence technique was applied, which led to better detection and quantification limits, but terbium-sensitised luminescence showed to be a more selective technique, tolerating the highest concentrations of other co-administrated drugs and excipients.
Acknowledgements The authors acknowledge Bristol Myers Squibb (USA) for supplying garenoxacin.
Table 3 Determination of garenoxacin in human urine and serum Spiked concentration (g ml−1 )
Urine 20 50 75 100 Serum 2.00 5.00 7.50 10.00
References
Experimental concentration (g ml−1 )a Aqueous fluorescence
20.3 50.8 76.2 101.5
± ± ± ±
2.14 4.90 7.43 10.10
0.8 1.2 1.4 2.1
± ± ± ±
0.14 0.10 0.19 0.17
Micelleenhanced fluorescence 19.8 51.2 74.8 102.3
± ± ± ±
1.96 4.84 7.58 10.20
1.0 1.1 1.8 1.9
± ± ± ±
0.12 0.14 0.14 0.24
Terbiumsensitised luminescence 20.2 49.6 76.1 99.2
± ± ± ±
2.13 5.07 7.39 9.86
0.8 0.9 1.3 2.0
± ± ± ±
0.15 0.17 0.21 0.21
a Mean ± S.D. (n = 5). In all cases, no statistically significant differences between experimental and spiked concentration were observed (95% significance level).
[1] M. Takahata, J. Mitsuyama, Y. Yamashiro, M. Yonezawa, H. Araki, Y. Todo, S. Minami, Y. Watanabe, H. Narita, Antimicrob. Agents Chemother. 43 (1999) 1077. [2] D.B. Hoellman, L.M. Kelly, M.R. Jacobs, P.C. Appelbaum, Antimicrob. Agents Chemother. 45 (2001) 589. [3] J.T. Kirby, A.H. Mutnick, R.N. Jones, D.J. Biedenbach, M.A. Pfaller,. The SENTRY Participants Group Diagn. Microb. Infect. Dis. 43 (2002) 303. [4] R. Wise, T. Gee, G. Marshall, J.M. Andrews, Antimicrob. Agents Chemother. 46 (2002) 31. [5] D. Xuan, C. Turley, C.H. Nightingale, D.P. Nicolau, J. Chromatogr. B 765 (2001) 37. [6] H. Fukumoto, H. Matsutani, Y. Yamamoto, H. Sakai, Iyakuhin Kenkyu 34 (2003) 153. [7] W. Jin, J. Jin, Zhongguo Kangshengsu Zazhi. 21 (1996) 195. [8] M. Stankov, D. Stankov, Z. Milicevic, D. Veselinovic, P. Djurdjevic, Spectrosc. Lett. 26 (1993) 1709.
J.A. Ocaña et al. / Talanta 63 (2004) 691–697 [9] N. Murugesan, S.C. Marthur, Y. Kumar, Y.K.S. Rathore, P.D. Sethi, East. Pharm. 35 (1992) 117. [10] Z. Liu, Z. Huang, R. Cai, Spectrochim. Acta A 56A (2000) 1787. [11] J.A. Ocaña, M. Callejón, F.J. Barragán, Talanta 52 (2000) 1149. [12] J.A. Ocaña, M. Callejón, F.J. Barragán, Analyst 125 (2000) 2322. [13] J.A. Hernandez-Arteseros, R. Compaño, M.D. Prat, Analyst 123 (1998) 2729. [14] J.A. Ocaña, M. Callejón, F.J. Barragán, Analyst 12 (2000) 1851.
697
[15] J.A. Ocaña, M. Callejón, F.J. Barragán, J. Pharm. Sci. 90 (2001) 1553. [16] J.A. Ocaña, M. Callejón, F.J. Barragán, Eur. J. Pharm. Sci. 13 (2001) 297. [17] W.R. Seitz, Luminiscence Spectrometry (Fluorimetry and Phosphorimetry), Wiley, New York, 1939. [18] Analytical Methods Committee, Analyst 112 (1987) 199. [19] J.C. Miller, J.N. Miller, Statistics for Analytical Chemistry second ed., Ellis Horwood, New York, 1988.
Fluorescence and terbium-sensitised luminescence determination of garenoxacin in human urine and serum J.A. Ocaña, F.J. Barragán∗ , M. Callejón Department of Analytical Chemistry, Faculty of Chemistry, E-41012 Seville, Spain Received 7 May 2003; received in revised form 11 November 2003; accepted 11 December 2003 Available online 3 February 2004
Abstract Fluorescence and terbium-sensitised luminescence properties of new quinolone garenoxacin have been studied. The fluorimetric method allows the determination of 0.060–0.600 g ml−1 of garenoxacin in aqueous solution containing HCl/KCl buffer (pH 1.5) with λexc = 282 nm and λem = 421 nm. Micellar-enhanced fluorescence was also studied, leading to a higher than 400% increase in analytical signal in presence of 12 mM sodium dodecyl sulphate (SDS), allowing the determination of 0.020–0.750 g ml−1 of garenoxacin. The terbium-sensitised luminescence method allows the determination of 0.100–1.500 g ml−1 of garenoxacin in 12 mM SDS solution containing 0.08 M acetic acid/sodium acetate buffer (pH 4.1) and 7.5 mM Na2 SO3 (chemical deoxygenation agent), with λexc = 281 nm and λem = 546 nm. Relative standard deviation (R.S.D.) values for the three methods were in the range 1.0–2.0%. The proposed procedures have been applied to the determination of garenoxacin in spiked human urine and serum. © 2004 Elsevier B.V. All rights reserved. Keywords: Garenoxacin; Fluorescence; Micelle-enhanced fluorescence; Luminescence; Terbium; Urine; Serum
1. Introduction Garenoxacin {1-cyclopropyl-8-(difluoromethoxy)-7-[(1R)-(1-methyl-2,3-dihydro-1H-5-isoindolyl]-4-oxo-1,4-dihydro-3-quinolinecarboxylic acid} is a new des-F(6)-quinolone antibacterial agent, formerly known as BMS-284756 and T-3811. This antibiotic differs from earlier quinolones in its lack of a fluorine atom in its C6 position, and the presence of an isoindolin-5-yl group at the 7th position [1]. Garenoxacin has broad-spectrum antibacterial activity against both Gram-positive and Gram-negative organisms, through inhibition of their DNA gyrase [2]. It also shows excellent oral bioavailability [3], with final concentrations in serum and urine of test-subjects of 2.0–10.0 and 20–100 g ml−1 , respectively, after 600 mg administration [4]. The analytical literature of this quinolone is very scarce. HPLC with fluorimetric detection has been applied to the determination of garenoxacin in mouse serum [5], and LC–MS for human urine and serum [6], but no other analytical ∗
Corresponding author. Tel.: +95-455-71-71; fax: +95-455-71-68. E-mail address: [email protected] (F.J. Barrag´an).
0039-9140/$ – see front matter © 2004 Elsevier B.V. All rights reserved. doi:10.1016/j.talanta.2003.12.016
reports have been found. This work constitutes the first stage of a research schedule focused on the study of the analytical properties of garenoxacin and the proposal of procedures for its determination in human biological media. Quinolones frequently exhibit native fluorescence properties that can be applied to their determination in aqueous solution and in biological matrix as urine and serum. This technique has been applied for the determination of an important number of quinolones, as ciprofloxacin [7], norfloxacin [8] and ofloxacin [9], among others. This fluorescence can be exalted in presence of an appropriate micellar medium, leading to better sensitivities and detection limits: this property has been applied to the study of quinolones as ciprofloxacin [10], levofloxacin [11] and moxifloxacin [12], among others. Quinolones also have suitable functional groups to form stable complexes with terbium ions. The presence of these ions in quinolone solutions leads to the formation of complexes that absorb energy at the characteristic wavelength of the organic ligand, and emit radiation at the characteristic wavelength of Tb3+ (546 nm). The mechanism of this emission is based in the quinolone being excited to its triplet state, which subsequently acts as an energy donor to the emitting energy level of Tb3+ . As the triplet state has a
692
J.A. Ocaña et al. / Talanta 63 (2004) 691–697
relatively long lifetime, the emitted radiation can be registered as a time-resolved luminescence signal, achieving an important decrement in background signal and better selectivity. This requires appropriate experimental conditions, as micellar medium and the presence of a deoxygenation agent as Na2 SO3 , in order to avoid the non-radiant deactivation process favoured by the long lifetime of the triplet state. The technique has been applied to the determination of the quinolones ciprofloxacin and enrofloxacin [13], levofloxacin [14], grepafloxacin [15] and trovafloxacin [16], among others. The present paper describes the spectrofluorimetric, micelle-enhanced spectrofluorimetric and terbium-sensitised luminescence properties of garenoxacin, and their application to its determination in spiked urine and serum.
10 g ml−1 was prepared daily by dilution of stock standard solution with high-purity water. 2.2. Apparatus Fluorescence and luminescence intensities were measured on a Perkin-Elmer (Norwalk, CT, USA) LS-5 luminescence spectrometer equipped with a xenon-lamp and an Acer Model 1030 computer working with the FLUORPACK software from Sciware (Mallorca, Spain). For luminescence measurements, the instrument was set in the phosphorescence mode and delay time (td ) of 0.03 ms and gate time (tg ) of 8 ms were used. All the measurements took place in a standard 10 mm pathlength quartz cell, thermostated at 25.0 ± 0.5 ◦ C, with 5 nm bandwidths for the emission and excitation monochromators. The pH was measured on a Crison (Barcelona, Spain) micropH 2002 pH-meter. 2.3. Sample preparation
2. Procedures 2.1. Reagents Garenoxacin (100.0% purity, data supplied by the provider) was kindly donated by Bristol Myers Squibb (USA). Terbium (III) chloride hexahydrate, sodium dodecyl sulphate (SDS), sodium dodecylbenzensulphonate (NaDDB), Triton® X-100 and hexadecyltrimethyl-ammonium bromide (HDTAB) were purchased from Fluka (Madrid, Spain). Concentrated acetic acid, sodium acetate, sodium sulphite, sodium hydroxide, acetonitrile and hydrochloric acid, were of analytical-reagent grade and purchased from Merck (Darmstadt, Germany). High-purity water was obtained from a Millipore (Milford, MA USA) Milli-Q Plus System. Urine and serum samples were obtained from several healthy volunteers. For fluorescence and micelle-enhanced fluorescence determinations, a HCl/KCl buffer of pH 1.5 was prepared by mixing 20.1 ml HCl 0.2 M and 29.9 ml KCl 0.2 M, and diluting to 100 ml high-purity water. For terbium-sensitised luminescence determination, a 0.8 M acetic acid/sodium acetate buffer of pH 4.1 was prepared by mixing 40 ml of 1.6 M acetic acid with 10 ml of 1.6 M sodium acetate, and diluting to 100 ml with high-purity water. Stock standard solution of 200 g ml−1 was prepared by dissolving garenoxacin in high-purity water, being stable for several weeks at room temperature. Working standard of
The proposed methods were applied to the determination of garenoxacin in spiked human urine and serum provided from several healthy volunteers. Urine and serum samples were spiked with convenient amounts of garenoxacin stock solutions, for final antibiotic concentrations in the range 20–100 and 2.0–10.0 g ml−1 , respectively. Spiked urine was 50-fold diluted with highpurity water. No serum pre-treatment was required for aqueousfluorimetric and micelle-enhanced fluorimetric methods. For terbium-sensitised luminescence method, a deproteinization process with acetonitrile was employed, as the untreated serum samples led to turbid solutions in presence of Tb3+ . So, 1 ml spiked serum samples were deproteinized with 2 ml of acetonitrile, by vortexing for 5 min and centrifuging for 5 min at 1500 × g. 2.4. Measurement of standard garenoxacin solutions For aqueous-medium fluorimetric determination, in 25 ml volumetric flasks, aliquots of working solutions of garenoxacin with 5 ml of 0.1 M HCl/KCl buffer (pH 1.5) were pipetted and diluted to the mark with water, for a final quinolone concentration in the range 0.060–0.600 g ml−1 . The obtained solutions were thermostated at 25±0.1 ◦ C, and the fluorescence emission at 421 nm against a blank solution was measured using an excitation wavelength of 282 nm. This procedure was also applied for the micelle-enhanced fluorimetric determination of garenoxacin, adding 2 ml of 0.15 M SDS solution before diluting to the mark, with a quinolone concentration in the range 0.020– 0.750 g ml−1 . For the terbium-sensitised luminescence determination, aliquots of working garenoxacin solution were pipetted
J.A. Ocaña et al. / Talanta 63 (2004) 691–697
into 10 ml calibrated flasks. Then, 1 ml of 0.8 M acetic acid-sodium acetate buffer (pH 4.1), 1 ml of 0.03 M Tb3+ solution, 1 ml of 0.15 M SDS solution and 1 ml of 0.075 M Na2 SO3 solution were added, diluting to the mark with high-purity water. Final garenoxacin concentration in samples was in the range 0.100–1.500 g ml−1 . The solutions were thermostated at 25 ± 0.1 ◦ C and the luminescence was measured after 5 min at 546 nm, using an excitation wavelength of 281 nm, against a blank solution. 2.5. Determination of garenoxacin in urine and serum For the determination of garenoxacin in urine, 1 ml of the diluted urine was pipetted in a 25 ml (aqueous and micelle-enhanced fluorescence methods) or a 10 ml (terbium-sensitised luminescence method) calibrated flask. For the determination of garenoxacin in serum, 1 ml of untreated serum was pipetted in a 25 ml calibrated flask for aqueous and micelle-enhanced fluorescence methods, and 0.25 ml of deproteinized serum was pipetted in a 10 ml calibrated flask for the terbium-sensitised luminescence method. Solutions were prepared and emission was measured as explained for standard garenoxacin solutions. In order to compensate the effect of the biological matrix in the measurements, standard addition method was applied to the quantification of the quinolone for three proposed methods in both urine and serum.
3. Results and discussion 3.1. Preliminary studies Garenoxacin showed a notable fluorescence emission at very acidic pH values, but showed a very weak emission
693
for higher pH. Fig. 1 shows the excitation and emission spectra obtained for a 0.250 g ml−1 garenoxacin solution at pH 1.5. The quinolone showed an excitation maximum at 282 nm, and an emission maximum at 421 nm. Thus, these wavelengths were selected as suitable for further fluorimetric studies. Garenoxacin shown no native time-resolved emission, but a notable signal was found in presence of 2.0 mM Tb3+ in a 10 mM SDS micellar medium with pH 4.0. Tb3+ solutions showed negligible time-resolved luminescence compared with Tb3+ –garenoxacin in the same conditions. Fig. 2 shows the excitation and emission luminescence spectra of the complex, with a maximum excitation and emission wavelengths of 281 nm (corresponding to the excitation of the quinolone) and 546 nm (corresponding to the characteristic emission of Tb3+ cation), respectively. The terbium-sensitised luminescence intensity was measured with 0.01–10 ms td and 1–10 ms tg . It was found that Rayleigh scattering could be eliminated using a 0.03 ms td , and maximum luminescence intensity was found with a 8 ms tg , so these values were selected for further studies. 3.2. Aqueous fluorescence properties The effect of pH value in the fluorescence emission of a 0.250 g ml−1 garenoxacin solution is shown in Fig. 3A. As can be seen, analytical signal exhibits a maximum when pH was 1.3–1.8, so pH of 1.5, was chosen for further studies. This pH was achieved by the addition of a HCl/KCl 0.1 M buffer, prepared as explained above. The influence of the buffer concentration on garenoxacin fluorescence was studied by the addition of increasing volumes of buffer, for final concentrations in the range 4–60 mM. Fluorescence intensity remains stable for lower than 40 mM buffer concentration, and it decreased for higher values. Accordingly, a 20 mM buffer concentration
Fluorescence intensity (arbritary units)
40
30 EM
EXC 20
10
0 250
300
350
400
450
500
λ (nm) Fig. 1. Fluorescence excitation (λem = 421 nm) and emission (λexc = 282 nm) spectra of 0.250 mg ml−1 garenoxacin at pH 1.5.
694
J.A. Ocaña et al. / Talanta 63 (2004) 691–697
Terbium sensitised luminescence intensity (arbritary units)
30
20
EM 10
EX
0 250
300
350
400
450
500
550
600
λ (nm) Fig. 2. Terbium-sensitised luminescence excitation (λem = 546 nm) and emission (λexc = 281 nm) spectra at of 0.250 mg ml−1 garenoxacin in 2.0 mM Tb3+ and 10 mM SDS medium. pH 4.0. td = 0.03 ms, tg = 8 ms.
(corresponding to a 5 ml aliquot of the prepared 0.1 M buffer solution in 25 ml total volume) was selected as a suitable volume for the recommended procedure. The influence of the temperature on the fluorescence intensity shows a nearly linear (negative) relationship between temperature and fluorescence intensity for garenoxacin, with a temperature coefficient of −1.15%. This value showed that internal conversion was probably the main non-radiant deactivation process for garenoxacin [17]. Thus, samples were thermostated at 25 ± 0.5.
tions of the surfactants SDS (cationic), HDTAB (cationic), NaDDB (anionic) and Triton® X-100 (non-ionic) at pH 1.50. It was observed that no significant fluorescence enhancement was achieved in presence of HDTAB or NaDDB. Triton® X-100 addition led an enhancement in analytical signal of only a 20%, but a notable increase in fluorescence was observed in presence of SDS. As can be seen in Fig. 4, fluorescence reached a constant and maximum value for higher than 10 mM SDS concentration, so a 12 mM SDS concentration was selected for further studies. Fig. 3B shows the influence of pH on the luminescence of 0.100 g ml−1 garenoxacin in 12 mM SDS medium. As can be seen, optimum pH range obtained was similar than that found for aqueous solution, so pH of 1.5, obtained with 5 ml of HCl/KCl 0.1 M buffer for a final solution volume of 25 ml, was chosen for the recommended procedure.
3.3. Micelle-enhanced fluorescence properties
Fluorescence and Luminescence Intensity (arbritary units)
The fluorimetric properties of garenoxacin were studied in a series of micellar media, by preparing 0.200 g ml−1 quinolone samples in the presence of increasing concentra110 B 90
70 C 50
A
30
10 1
2
3
4
5
6
7
8
pH
Fig. 3. Effect of pH in: (A) fluorescence of 0.250 g ml−1 garenoxacin (λexc = 282 nm, λem = 421 nm); (B) micellar-enhanced fluorescence of 0.100 g ml−1 garenoxacin (λexc = 282 nm, λem = 421 nm, SDS = 12 mM) and (C) terbium-sensitised luminescence of 0.300 g ml−1 garenoxacin (λexc = 281 nm, λem = 546 nm, SDS = 10 mM, Tb3+ = 2.0 mM).
J.A. Ocaña et al. / Talanta 63 (2004) 691–697
695
Fluorescence intensity (arbritary units)
80
60
40
20
0 0
4
8
12
16
20
[SDS] (mM)
Fig. 4. Effect of SDS concentration in fluorescence of 0.250 g ml−1 garenoxacin (λexc = 282 nm, λem = 421 nm, pH 1.50).
3.4. Terbium-sensitised luminescence properties Fig. 3C shows the influence of pH on luminescence intensity. As can be seen, the optimum pH value is in the range 3.8–4.5. So, acetic acid/sodium acetate buffer of pH 4.1 was selected for the recommended procedure. The influence of the buffer concentration on the analytical signal was studied in the range 0.04–0.20 M. The optimal performance was found in the range 0.06–0.10 M. Thus, a 0.08 M buffer concentration was selected as suitable for the optimised procedure. The effect of SDS concentration was studied measuring the luminescence emission of a series of samples with tensoactive concentrations in the range 4–20 mM. The luminescence intensity reached a maximum and constant value for higher than 10 mM SDS concentrations, so a value of 12 mM was selected for further studies. A series of samples with terbium concentration in the range 0.5–4.0 mM were prepared. The luminescence intensity measured showed a stable value for concentrations of Tb3+ above 2.0 mM, so a concentration of 3.0 mM was chosen as suitable. Lastly, it was observed that a minimum concentration of Na2 SO3 of 7.5 mM was required to effectively eliminate oxygen completely from the solutions, obtaining the maximum analytical signal 5 min after Na2 SO3 addition, stable for more than an hour. The luminescence intensity decreased notably for higher than 15 mM Na2 SO3 concentrations, so a Na2 SO3 concentration of 7.5 mM was chosen as suitable, which led to a final 50% enhancement in analytical signal with respect to measurements carried in absence of the deoxygenation agent.
concentrations between 0.001 and 1.0 g ml−1 were prepared by triplicate and measured by following the procedures described in Section 2. Calibration curves of fluorescence, micelle-enhanced fluorescence and terbium-sensitised luminescence intensities (Y) versus garenoxacin concentration expressed in micrograms per millilitre (X) were constructed; Table 1 resumes the results obtained from statistical analysis of data. As can be seen, micelle-enhanced fluorescence method achieved the highest sensitivity, with a 450% increase in slope with respect to the aqueous fluorescence method. Detection limit (DL) and quantification limit (QL) were calculated according to the recommendations of the Analytical Methods Committee [18]. The values obtained are resumed in Table 1. Again, best results were achieved with micellar-enhanced fluorescence method. In order to evaluate the precision of the proposed methods, 11 replicates (0.300 g ml−1 garenoxacin for aqueous and micelle-enhanced fluorescence; 0.600 g ml−1 garenoxacin
Table 1 Figures of merit of the aqueous-solution fluorescence, micelle-enhanced fluorescence and terbium-sensitised luminescence determination of garenoxacin Parameter
Aqueous solution method
Micelle-enhanced method
Time-sensitised luminescence method
3.5. Figures of merit of the proposed methods
LRC (g ml−1 ) a ± Sa b ± Sb r R.S.D. (%) DL (g ml−1 ) QL (g ml−1 )
0.060–0.600 0.3 ± 0.1 67.9 ± 0.3 0.9997 2.0 0.015 0.045
0.020–0.750 1.5 ± 0.2 297.8 ± 0.4 0.9999 1.2 0.005 0.018
0.100–1.500 5.7 ± 0.3 56.4 ± 0.3 0.9996 1.3 0.020 0.065
Once the three proposed method were optimised, three series of ten standard solutions of garenoxacin with
LRC, linear calibration range; a, intercept; b, slope; Sa , intercept standard deviation; Sb , slope standard deviation; r, correlation coefficient; R.S.D., relative standard deviation; DL, detection limit; QL, quantification limit.
696
J.A. Ocaña et al. / Talanta 63 (2004) 691–697
Table 2 Tolerance of different compounds in the determination of garenoxacin
3.6. Determination of garenoxacin in urine and serum
Compounds added
Garenoxacin in urine and serum with a content of 20–100 and 2–10 g ml−1 , respectively, was analysed as described under experimental by the three proposed procedures, using the standard addition method. External calibration was not applied in order to avoid matrix effects of urine or serum affecting the results (slope of the standard addition curves in these media were approximately 10–20% lower than slope found for aqueous standards for the three proposed methods). The garenoxacin concentration obtained for five replicates are resumed in Table 3. In order to compare the obtained recoveries with the spiked concentration, a Student’s t-test [19] was used. In all cases, no statistically significant differences between the spiked and the experimental concentrations were observed at 95% significance level.
Maximum tolerance weight ratio
Glucose, lactose, fructose, sacarose Amikacin Aspirin Imipenem Paracetamol Penicilin Sulbactam
Aqueous fluorescence
Micellar fluorescence
Terbium-sensitised luminescence
2000a
2000a
2000a
30a 30a 1.0 6.0 40a 30
30a 30a 1.6 6.0 40a 30
30a 30a 3.2 9.0 40a 40a
a Maximum weight ratio tested (no interference was found at this level).
for terbium-sensitised luminescence), were quantified. The relative standard deviation (R.S.D.) obtained were in the range 1.0–2.0% (Table 1) for three methods. Lastly, the interference of typical excipients and co-administer drugs (mainly analgesics or antibiotics) was studied by the addition of increasing concentrations of these compounds to a 0.300 g ml−1 garenoxacin solution (aqueous and micellar-enhanced fluorescence) or a 0.600 g ml−1 (terbium-sensitised luminescence). A substance was considered to interfere at the weight interference–quinolone ratio that led to errors greater than the three times the precision level for each method (3× R.S.D). For some of the excipients and drugs, no interference was found even at the maximum tested weight interference–quinolone ratio; for the rest, only negative errors were found. Maximum tolerable weight interference–quinolone ratio for the proposed methods are given in Table 2. As can be seen, best selectivity was found for the terbium-sensitised luminescence method.
4. Conclusions This work shows that fluorescence, micelle-enhanced fluorescence and terbium-sensitised luminescence may be simple, accurate, fast and precise procedures to determination of garenoxacin in both human urine and serum. Best sensitivity was achieved when micelle-enhanced fluorescence technique was applied, which led to better detection and quantification limits, but terbium-sensitised luminescence showed to be a more selective technique, tolerating the highest concentrations of other co-administrated drugs and excipients.
Acknowledgements The authors acknowledge Bristol Myers Squibb (USA) for supplying garenoxacin.
Table 3 Determination of garenoxacin in human urine and serum Spiked concentration (g ml−1 )
Urine 20 50 75 100 Serum 2.00 5.00 7.50 10.00
References
Experimental concentration (g ml−1 )a Aqueous fluorescence
20.3 50.8 76.2 101.5
± ± ± ±
2.14 4.90 7.43 10.10
0.8 1.2 1.4 2.1
± ± ± ±
0.14 0.10 0.19 0.17
Micelleenhanced fluorescence 19.8 51.2 74.8 102.3
± ± ± ±
1.96 4.84 7.58 10.20
1.0 1.1 1.8 1.9
± ± ± ±
0.12 0.14 0.14 0.24
Terbiumsensitised luminescence 20.2 49.6 76.1 99.2
± ± ± ±
2.13 5.07 7.39 9.86
0.8 0.9 1.3 2.0
± ± ± ±
0.15 0.17 0.21 0.21
a Mean ± S.D. (n = 5). In all cases, no statistically significant differences between experimental and spiked concentration were observed (95% significance level).
[1] M. Takahata, J. Mitsuyama, Y. Yamashiro, M. Yonezawa, H. Araki, Y. Todo, S. Minami, Y. Watanabe, H. Narita, Antimicrob. Agents Chemother. 43 (1999) 1077. [2] D.B. Hoellman, L.M. Kelly, M.R. Jacobs, P.C. Appelbaum, Antimicrob. Agents Chemother. 45 (2001) 589. [3] J.T. Kirby, A.H. Mutnick, R.N. Jones, D.J. Biedenbach, M.A. Pfaller,. The SENTRY Participants Group Diagn. Microb. Infect. Dis. 43 (2002) 303. [4] R. Wise, T. Gee, G. Marshall, J.M. Andrews, Antimicrob. Agents Chemother. 46 (2002) 31. [5] D. Xuan, C. Turley, C.H. Nightingale, D.P. Nicolau, J. Chromatogr. B 765 (2001) 37. [6] H. Fukumoto, H. Matsutani, Y. Yamamoto, H. Sakai, Iyakuhin Kenkyu 34 (2003) 153. [7] W. Jin, J. Jin, Zhongguo Kangshengsu Zazhi. 21 (1996) 195. [8] M. Stankov, D. Stankov, Z. Milicevic, D. Veselinovic, P. Djurdjevic, Spectrosc. Lett. 26 (1993) 1709.
J.A. Ocaña et al. / Talanta 63 (2004) 691–697 [9] N. Murugesan, S.C. Marthur, Y. Kumar, Y.K.S. Rathore, P.D. Sethi, East. Pharm. 35 (1992) 117. [10] Z. Liu, Z. Huang, R. Cai, Spectrochim. Acta A 56A (2000) 1787. [11] J.A. Ocaña, M. Callejón, F.J. Barragán, Talanta 52 (2000) 1149. [12] J.A. Ocaña, M. Callejón, F.J. Barragán, Analyst 125 (2000) 2322. [13] J.A. Hernandez-Arteseros, R. Compaño, M.D. Prat, Analyst 123 (1998) 2729. [14] J.A. Ocaña, M. Callejón, F.J. Barragán, Analyst 12 (2000) 1851.
697
[15] J.A. Ocaña, M. Callejón, F.J. Barragán, J. Pharm. Sci. 90 (2001) 1553. [16] J.A. Ocaña, M. Callejón, F.J. Barragán, Eur. J. Pharm. Sci. 13 (2001) 297. [17] W.R. Seitz, Luminiscence Spectrometry (Fluorimetry and Phosphorimetry), Wiley, New York, 1939. [18] Analytical Methods Committee, Analyst 112 (1987) 199. [19] J.C. Miller, J.N. Miller, Statistics for Analytical Chemistry second ed., Ellis Horwood, New York, 1988.