ANALYTICAL SCIENCES JANUARY 2006, VOL. 22 25 2006 © The Japan Society for Analytical Chemistry

Sensitive Assay for Catecholamines in Pharmaceutical Samples and Blood Plasma Using Flow Injection Chemiluminescence Analysis

Chunling YU, Yuhai TANG,† Xiaonian HAN, and Xiaohui ZHENG

Department of Chemistry, College of Science, Xi’an Jiaotong University, Xi’an 710061, P. R. China

A rapid and sensitive chemiluminescence (CL) method using flow injection analysis was described for the determination of three catecholamines: dopamine, adrenaline and dobutamine, based on their greatly enhancing effects on the CL reaction of luminol–potassium periodate in basic solutions. Under the optimized conditions, the calibration graphs relating the increase of CL intensity to the concentration of the analytes were linear. The present method allows for the determination of dopamine, adrenaline, and dobutamine over the range of 1.0 × 10–10 – 1.0 × 10–7 g/ml. The relative standard deviations for measurements (n = 11) of dopamine, adrenaline and dobutamine were 2.9, 2.3 and 1.8% when the concentrations of three catecholamines were at 1.0 × 10–9 g/ml, respectively. The detection limits of the method were 2.0 × 10–11 g/ml dopamine, 1.0 × 10–11 g/ml adrenaline and 4.0 × 10–11 g/ml dobutamine. The method was successfully applied to the determination of three catecholamines in pharmaceutical samples and blood plasma.

(Received July 6, 2005; Accepted October 25, 2005)

method allowed the determination of dopamine over the range of Introduction 1.0 × 10–5 – 1.0 × 10–7 g/ml with a detection limit of 5.3 × 10–8 g/ml. Dopamine, adrenaline and dobutamine are known as In our work we found that catecholamines could strongly catecholamines, which play important roles in the central enhance the CL reaction of luminol–potassium periodate in an nervous system as neurotransmitters. These catecholamines, as alkaline solution; the increase of the CL intensity was well as metaraminol and phenylephrine with a structure of the dependent on the concentration of the studied drug. Based on phenylethyl amine group, are widely used to treat hypertension, these observations, a simple, sensitive and rapid new assay for bronchial asthma, cardiac arrest, myocardial infarction, and catecholamines has been developed using the flow-injection CL cardiac surgery. Therefore, a highly sensitive method is technique. The proposed method was successfully applied to necessary for the determination of the catecholamines, both in the determination of dopamine, adrenaline and dobutamine, not pharmaceutical samples and biological fluids. only in pharmaceutical samples, but also in blood plasma. A variety of techniques have been utilized for the determination Moreover, the possible mechanism of the CL reaction was of three catecholamines, such as chromatography,1–3 capillary briefly considered. electrophoresis,4–6 spectrophotometry,7 fluorometry,8–10 and electrochemical11,12 detection. However, these methods are complicated (they need derivatization or combination with Experimental various means of detection). Also, some of these methods suffer low sensitivity and specificity, such as the detection limit Apparatus of the method,9 which is 1.0 × 10–7 g/ml. A schematic diagram of the flow-injection chemiluminescence The CL methods for the determination of active compounds in (FI-CL) system used in this work is shown in Fig. 1. One pharmaceutical samples or biological fluids are widely used peristaltic pump was used to deliver luminol and KIO4 solutions because of their simple, rapid, and sensitive properties. In with a flow rate of 2.7 ml/min. Another peristaltic pump was recent years, there have been a few reports on the determination used to pump the sample solution at a flow rate of 0.8 ml/min. of catecholamines using the chemiluminescence (CL)13–16 PTFE tubing (0.8 mm i.d.) was used to connect all components technique. Among these methods, many of them determine the in the flow system. A six-way injection valve was used to inject content of catecholamines in pharmaceutical preparations. Only the sample solutions. The streams of luminol, KIO4 in the Li et al.16 combined on-line microdialysis sampling with a plant alkaline medium, and analytes were combined in a flow cell. tissue-based chemiluminescence flow biosensor to monitor the The CL signal produced in the flow cell was collected with a variation of the dopamine level in the blood of rabbit. In this photomultiplier tube. Data acquisition and treatment were method, dopamine was oxidized by oxygen under the catalysis handed by an IBM-compatible computer, running IFFL-DD of polyphenol oxidase in the tissue column to produce hydrogen flow-injection CL analysis system software (Xi’an Rumax peroxide, which reacted with luminol in the presence of the Electronic Equipment Corporation, Xi’an, China). peroxidase of potato tissue, and generated a CL signal. The Reagents and solutions † To whom correspondence should be addressed. All chemicals used were of analytical reagent grade; doubly E-mail: [email protected] distilled water was used through the experiment. Dopamine 26 ANALYTICAL SCIENCES JANUARY 2006, VOL. 22

Table 1 Calibration curves of catecholamines

Con. Limit of Calibration curve Correlation Catecholamine range/ detection/ ∆I = a + bC a cofficient ng ml–1 ng ml–1 (3σ )

Dopamine 0.1 – 1 ∆I =3.01 + 15.37C 0.9969 1 – 10 0.02 ∆I =3.56 + 13.01C 0.9989 10 – 100 ∆I =113.20 + 0.99C 0.9965 Adrenaline 0.1 – 1 ∆I =6.58 + 44.63C 0.9964 Fig. 1 Schematic diagram of the flow system for the determination 1 – 10 0.01 ∆I =18.97 + 11.79C 0.9987 of catecholamines. a, Sample solution; b, luminol solution; c, KIO4 10 – 100 ∆I =107.82 + 3.46C 0.9982 solution; P1 and P2, peristaltic pump 1 and 2; V, six-way injection Dobutamine 0.1 – 1 ∆I =8.29 + 72.92C 0.9985 valve; F, flow cell; PMT, photomultiplier tube; HV, high voltage 1 – 10 0.04 ∆I =30.06 + 16.26C 0.9989 (operated at –600 V); COM, computer; W, waste solution. 10 – 100 ∆I =156.30 + 5.59C 0.9975

a. ∆I = a + bC, where C is the concentration of catecholamines in ng/ml. hydrochloride, adrenaline hydrochloride and dobutamine hydrochloride were purchased from the National Institute for the Control of Pharmaceutical and Biological Products (Beijing, were prepared by diluting respectively prepared standard China). Luminol was obtained from the Institute of Analytical solutions with water. A certain amount of catecholamines Science of Shaanxi Normal University (Xi’an, China). solution was injected into the flow system by a six-way

Potassium periodate (KIO4) was obtained from Guangzhou injection valve and mixed with the mixed luminol–KIO4 Chemical Reagent Factory (Guangzhou, China). solution. Then, the mixed solution was transferred into the flow A 0.01 mol/l luminol stock solution was prepared by cell and an increased CL signal was obtained. The increased dissolving appropriate luminol in a 0.02 mol/l sodium value of the CL intensity (∆I, the difference of CL intensity hydroxide solution. A luminol working solution was prepared between the catecholamine solution and the blank reagent by directly diluting this stock solution to 1.0 × 10–4 mol/l with without the catecholamine) was proportional to the water. Potassium periodate stock solution was prepared by concentration of the catecholamine. dissolving 0.5750 g of potassium periodate in water and diluting to 250 ml with water. The working solution of potassium periodate was prepared by diluting this stock solution to 1.0 × Results and Discussion 10–4 mol/l with an approximate amount of 1.0 mol/l sodium hydroxide to provide a final concentration of 0.1 mol/l sodium CL kinetic property hydroxide in the diluted solution. In this work, it was observed that the CL reaction of –4 Standard solutions (1.0 × 10 g/ml) of dopamine, adrenaline luminol–KIO4 could be strongly enhanced by dopamine and dobutamine were prepared freshly with water. The standard hydrochloride in alkaline solution. The CL signal reached the solutions were stored in a refrigerator (4˚C) and protected from maximum intensity at 2.4 s, and was then extinguished within light before use. Before analysis, working solutions were 5.4 s after mixing the dopamine hydrochloride solution and the prepared by appropriate dilution of these standard solutions with luminol solution in a sodium hydroxide with KIO4 solution. water to ensure concentrations within the linear range of the calibration graphs (Table 1). Effect of different alkaline concentrations It is well known that the luminol CL reaction requires alkaline Preparation for the studied catecholamines injections conditions. Thus, several alkaline buffer solutions (carbonate, Pharmaceutical injections containing dopamine hydrochloride phosphate, borate and sodium hydroxide) were tested, and we (20 mg per ampoule), adrenaline hydrochloride (1 mg per found that sodium hydroxide was the most suitable medium for ampoule), and dobutamine hydrochloride (20 mg per ampoule) the present CL reaction. In addition, we also found that the CL were obtained from the First Affiliated Hospital of Xi’an Jiaotong intensity was stronger when an appropriate amount of NaOH

University (Xi’an, China). The solution was diluted appropriately was added into the KIO4 solution, rather than being added into with water to a certain concentration, respectively, before use, the luminol solution. Based on these observations, a series of and the content was calculated from the corresponding experiments were performed to select the optimum reaction calibration graph or by using the regression equation (Table 1). conditions. Also, the effect of NaOH within the range of 0.01 – 0.30 mol/l concentration on the CL reaction was further tested

Preparation for plasma when the concentrations of luminol and KIO4 were both fixed at Fresh whole blank blood was centrifugated at 3000 rpm for 10 1.0 × 10–4 mol/l (Fig. 2). The ∆I continued to change with min. The supernatant was transferred to a test tube and used as increasing the NaOH concentration. When 0.2 mol/l NaOH was blank plasma. Acetonitrile (0.2 ml) was added to an aliquot of used in the KIO4 solution, the CL reaction gave the maximum plasma (0.1 ml) in a 1.5 ml centrifuge tube, and the solutions ∆I. Considering the stability of CL signal, the best were shaken for 5 min, and homogenized. The precipitated concentration of NaOH was 0.1 mol/l. protein was removed by centrifugation for 10 min at 10000 rpm. The supernatant was blowed dry with nitrogen gas and Effect of the luminol concentration transferred into a 100 ml calibrated flask and completed to As the luminescence reagent of the CL reaction, the luminol volume with distilled water. concentration affected the CL intensity. The influence of 1.0 × 10–5 to 3.0 × 10–4 mol/l luminol on the CL reaction was Procedure examined with 1.0 × 10–9 g/ml of a dopamine hydrochloride

A series of working solutions with different concentrations solution as the analyte when the concentration of KIO4 was ANALYTICAL SCIENCES JANUARY 2006, VOL. 22 27

Table 2 Calibration curves of catecholamines in rabbit blank plasma

Con. Limit of Calibration curve Correlation Catecholamine range/ detection/ ∆I = a + bC a cofficient ng ml–1 ng ml–1 (3σ )

Dopamine 1 – 10 0.3 ∆I = 1.06 + 5.76C 0.9989 10 – 100 ∆I = 49.24 + 0.66C 0.9989 Adrenaline 1 – 10 0.4 ∆I = 5.21 + 27.46C 0.9969 10 – 100 ∆I = 21.68 + 1.61C 0.9980 Dobutamine 1 – 10 0.2 ∆I = 10.48 + 15.18C 0.9975 10 – 100 ∆I = 98.50 + 5.08C 0.9979 a. ∆I = a + bC, where C is the concentration of catecholamines in ng/ml. Fig. 2 Effect of the NaOH concentration on the CL intensity. –4 –4 (Luminol 1.0 × 10 mol/l, KIO4 1.0 × 10 mol/l, dopamine hydrochloride 1.0 × 10–9 g/ml). fixed at 1.0 × 10–4 mol/l (Fig. 3). The maximum ∆I was obtained when the concentration of luminol was 1.0 × 10–4 mol/l, and thus 1.0 × 10–4 mol/l luminol was used in a further study.

Effect of KIO4 concentration KIO4 was used as the oxidant in the reaction. The influence of the KIO4 concentration on the CL reaction was examined in the range of 1.0 × 10–5 – 3.0 × 10–4 mol/l when the concentrations of NaOH and luminol were at 0.1 mol/l and 1.0 × 10–4 mol/l (Fig. 4). The results showed that the CL intensity increased as the concentration of KIO4 was increased to 1.0 × –4 10 mol/l. When the concentration of KIO4 was above 1.0 × 10–4 mol/l, the CL intensity decreased. Thus, the concentration –4 of 1.0 × 10 mol/l KIO4 was used in the following experiments. Fig. 3 Effect of the luminol concentration on the CL intensity. × –4 Calibration curves of the studied catecholamines in (NaOH 0.1 mol/l, KIO4 1.0 10 mol/l, dopamine hydrochloride 1.0 × –9 pharmaceutical injections 10 g/ml). The working curves of the studied catecholamines were established under the optimum experimental conditions mentioned above. A 1.0 × 10–5 g/ml stock solution of the studied catecholamines was prepared in distilled water. Serial dilutions with distilled water were made to cover the working ranges (Table 1). Table 1 lists the parameters of the calibration curves, and the relative standard deviations for the detections (n = 11) of dopamine, adrenaline and dobutamine were 2.9, 2.3 and 1.8% when the concentrations of the three catecholamines were at 1.0 × 10–9 g/ml, respectively.

Calibration curves of dopamine hydrochloride in rabbit blank plasma Following the method described in the section “Preparation for plasma”, the relative standard deviations for the Fig. 4 Effect of the KIO4 concentration on the CL intensity. measurement (n = 11) of dopamine, adrenaline and dobutamine (NaOH 0.1 mol/l, luminol 1.0 × 10–4 mol/l, dopamine hydrochloride × –9 were 2.1, 2.7 and 2.3% when the concentrations of three 1.0 10 g/ml). catecholamines were at 5.0 × 10–9 g/ml, respectively. Table 2 gives the results of this experiment. Determination of the studied catecholamines injections Interference studies Following the method described in the section “Preparation The effects of various substances (excipients and common ions) for the studied catecholamines injections”, the results are listed that often accompany injections of the three catecholamines in Table 3, and agreed well with those obtained by the method were studied by analyzing a standard solution of dopamine recommended by the Pharmacopoeia of People’s Republic of hydrochloride (1.0 × 10–9 g/ml) into which increasing amounts China.17 of interfering analyte were added. It was considered not to interfere if a foreign species caused a relative error of less than Determination of catecholamines in rabbit plasma ±5% during the determination of 1.0 × 10–9 g/ml of dopamine The injection of 0.1 ml catecholamine (4.0 × 10–6 g/ml) was hydrochloride. No interference was found when including up to added to 0.1 ml of blank plasma; following the method described + + – 2– – 2– 3– 1000-fold K , Na , Cl , SO4 , NO3 , SO3 , PO4 ; 100-fold in section “Preparation for plasma”, the results are listed in Table 2+ 2+ 2+ NH4 , Ba , Mg , glucose, lactose; 50-fold sodium pyrosulfate; 4. As can be seen, the proposed method proved to be satisfactory and 1-fold Ca2+, Al3+, ascorbic acid. for determining the studied catecholamines in rabbit plasma. 28 ANALYTICAL SCIENCES JANUARY 2006, VOL. 22

Table 3 Determination of the catecholamine injections Table 4 Determination of the studied catecholamines in rabbit plasma Claimed/ Founda/ Added/ Recovered/ Rec., Official methoda/ mg ml–1 mg ml–1 ng m–1 ng ml–1 % mg ml–1 Concentration/ Founda/ Added/ Recovered/ Rec., µg ml–1 µg ml–1 ng m–1 ng ml–1 % Dopamine 10 9.87 2.00 1.95 97.5 Dopamine 3.00 3.04 101.3 9.91 4.0 3.87 2.00 2.09 104.5 4.00 3.93 98.3 3.00 2.92 97.3 Adrenaline 4.00 4.09 102.3 1 0.98 2.00 2.09 104.5 Adrenaline 3.00 2.87 95.7 0.96 4.0 4.08 2.00 1.91 95.5 4.00 3.78 94.5 3.00 3.14 104.7 Dobutamine 4.00 4.17 104.3 10 9.65 2.00 1.89 94.5 Dobutamine 3.00 2.90 96.7 9.78 4.0 4.16 2.00 1.95 97.5 4.00 4.17 104.3 3.00 2.92 97.3 4.00 4.09 102.3 a. Mean value of five measurements. a. Mean value of five measurements.

Possible CL reaction mechanism Under the experimental conditions described above, the CL spectra of the luminol–KIO4 reaction in both the absence and References presence of the studied catecholamines were obtained using a refitted RF-540 spectrofluorometer. It was observed that the CL 1. P. Kumarathasan and R. Vincent, J. Chromatogr., A, 2003, spectra of the luminol–KIO4 reaction had the maximum 987, 349. emission wavelength at 425 nm, which is the characteristic 2. G. H. Ragab, H. Nohta, and K. Zaitsu, Anal. Chim. Acta, spectrum of the luminol reaction.18 These results indicated that 2000, 403, 155. the potential CL emitter is the excited state of 3-aminophthalate 3. A. Hansen, J. Kristiansen, J. Løgsted Nielsen, K. Byrialsen, ions in the CL reaction of luminol–KIO4 in a basic solution. It and J. Christensen, Talanta, 1999, 50, 367. was also reported that polyhydroxyphenols could be oxidized to 4. Y. S. Wang, D. S. Fice, and P. K. F. Yeung, J. Pharm. produce hydroxy and/or superoxide radicals by periodate, and Biomed. Anal., 1999, 21, 519. that the dissolved oxygen could produce an enhancing effect on 5. C. Sabbioni, M. A. Saracino, R. Mandrioli, S. Pinzauti, S. the CL reaction of the luminol system in an alkaline Furlanetto, G. Gerra, and M. A. Raggi, J. Chromatogr., A, solution.19,20 Thus, it can be assumed that the studied dopamine 2004, 1032, 65. with a polyphenol group could also be oxidized to yield 6. L. Y. Zhang, S. F. Qv, Z. L. Wang, and J. K. Cheng, J. hydroxy and/or superoxide radicals, and that the dissolved Chromatogr., B, 2003, 792, 381. oxygen had a synergetic effect on the CL reaction of 7. J. J. Berzas Nevado, J. M. Lemus Gallego, and P. Buitrago luminol–KIO4. Because the oxidation of the luminol–KIO4 Laguna, J. Pharm. Biomed. Anal., 1996, 14, 571. system could be accelerated by hydroxy and/or superoxide, the 8. H. Y. Wang, Y. Sun, and B. Tang, Chin. J. Anal. Lab., CL emission was enhanced. Based on the above discussion, the 2003, 22, 45. possible mechanism of the CL reaction of luminol–KIO4–dopamine 9. H. Y. Wang, Q. S. Hui, L. X. Xu, J. G. Jiang, and Y. Sun, in a basic medium was suggested to be as follows: Anal. Chim. Acta, 2003, 497, 93. (1) Catecholamines + dissolved oxygen + NaOH → hydroxy 10. H. Y. Wang, Y. Sun, and B. Tang, Talanta, 2002, 57, 899. and/or superoxide radicals + the other products; 11. M. D. Rubianes and G. A. Rivas, Anal. Chim. Acta, 2001, (2) Luminol + hydroxy radicals and/or superoxide radicals + 440, 99.

KIO4 → excited 3-aminophthalate ions; 12. F. Xu, M. Gao, L. Wang, G. Shi, W. Zhang, L. Jin, and J. (3) Excited 3-aminophthalate ions → 3-aminophthalate ion + Jin, Talanta, 2001, 55, 329.

hv (λmax = 425 nm). 13. F. Li, H. Cui, and X. Lin, Anal. Chim. Acta, 2002, 471, 187. Conclusion 14. C. X. Zhang, J. C. Huang, Z. J. Zhang, and M. Aizawa, Anal. Chim. Acta, 1998, 374, 105. Based on the enhancement effects of catecholamines on the CL 15. E. Nalewajko, R. B. Ramyrez, and A. Kojo, J. Pharm. from the reaction of luminol with KIO4 in an alkaline solution, a Biomed. Anal., 2004, 36, 219. new flow injection CL method was developed for the 16. B. X. Li, Z. J. Zhang, and Y. Jin, Biosens. Bioelectron., determination of dopamine, adrenaline and dobutamine. The 2002, 17, 585. present method was applied to the analysis of three 17. Editorial Committee of the Pharmacopoeia of P. R. China, catecholamines in pharmaceutical samples and blood plasma “The Pharmacopoeia of People’s Republic of China”, 2nd with satisfactory results. ed., 2000, Chemical Industry Press, Beijing. 18. Z. J. Zhang, J. Lu, and X. R. Zhang, Huaxue Shiji, 1987, 3, Acknowledgements 149. 19. J. M. Lin and Y. Masaak, Anal. Chem., 1999, 71, 1760. We gratefully acknowledge financial support from Xi’an 20. Y. H. Li, Y. H. Tang, H. Yao, and J. M. Fu, Luminescence, Jiaotong University. 2003, 18, 313.