- Email: [email protected]
Simultaneous determination of nalidixic acid, oxolinic acid and piromidic acid in fish by high-performance liquid chromatography with fluorescence and uv detection
Journal of Chromatography, 402 (1987) 301-308 Elm&f Science Publishers B.V., Amsterdam - Printed in The Netherlands CHROM. 19 585
SIMULTANEOUS DETERMINATION OF NALIDIXIC ACID, OXOLINIC ACID AND PIROMIDIC ACID IN FISH BY HIGH-PERFORMANCE LIQUID CHROMATOGRAPHY WITH FLUORESCENCE AND UV DETECTION
MASAKAZU HORIE*, KOUICHI SAITO, YOUJI HOSHINO and NORIHIDE NOSE Saitama Prefectural Institute of Public Health, 639-1. Kamiokubo. Urawa, Saitama 338 (Japan) EMIKO MOCHIZUKI Yamanashi Institute of Public Health, l-7-31, Fuji& Kofu, Yamanashi 400 (Japan) and HIROYUKI NAKAZAWA The Institute of Public Health, 4-6-1, Shirokanedai. Minato-ku. Tokyo 108 (Japan) (Received March 17th, 1987)
SUMMARY
A simple and rapid method for the simultaneous determination of nalidixic acid (NA), oxolinic acid (OXA) and piromidic acid (PMA) in cultured fish has been developed by high-performance liquid chromatography (HPLC). The drugs were extracted with 0.1% metaphosphoric acid-methanol (6:4), followed by a Sep-Pak C1 a clean-up procedure. The HPLC separation was carried out on a Kaseisorb LC ODS 300-5 cohunn (25 cm x 4.6 mm I.D.) using 5 mM phosphate buffer-acetonitrile (6:4) as a mobile phase. A fluorescence detector was used for NA and OXA at the excitation wavelength of 325 mu and the emission wavelength of 365 nm and an ultraviolet detector at 280 nm for PMA. The calibration graphs were rectilinear from 1 to 10 ng for OXA, from 2 to 20 ng for NA and PMA. The recoveries of NA, OXA and PMA added to fish were 81.5-85.3,83.7-88.7 and 80.9-84.9%, respectively, with high accuracy. The limits of detection were 0.01 pg/g for each drug.
INTRODUCTION
With the development of farming of fish such as eel and yellowtail, various kinds of antibiotics and synthetic antibacterials have been widely used for the prevention and treatment of infectious diseases in fish. Concern has arisen as to the presence of drug residues in fish tissues, and demand for a rapid, simple and sensitive analytical method of determining them has increased. Nalidixic acid (NA), oxolinic acid (OXA) and piromidic acid (PMA) are widely applied to fish such as eel, yellowtail, rainbow trout, sweet fish and common carp to treat a variety of gram-negative organisms l. Several methods have been developed for the assay of NA, OXA and PMA using thin-layer chromatography (TLC)Z, gas chromatography3-5, fluorometry6 and high-performance liquid chromatography 0021-9673/87/$03.50
0
1987 Elsevier Science Publishers B.V.
302
M. HORIE et al.
(HPLC)‘-l l. Most methods were developed for the assay of pharmaceutical preparations, and plasma and urine samples, only few for the residue analysis of fish sampleslz-14. A few HPLC methods with ultraviolet (W) detection were recently reported for the simultaneous determination of NA, OXA and PMA in fishlS,16. However, they were tedious and insufficiently sensitive. This paper describes a simple, specific and rapid HPLC method for the simultaneous determination of NA, OXA and PMA using W and fluorometric detection, and Sep-Pak Cl8 cartridges as a clean-up step. EXPERIMENTAL
Materials and reagents
The edible muscle tissues of yellowtail, eel, rainbow trout, sweet fish and common carp served as samples. NA, OXA and PMA were obtained from Daiichi Pharmaceutical (Tokyo, Japan), Tanabe Pharmaceutical (Osaka, Japan) and Dainihon Pharmaceutical (Osaka, Japan), respectively. Each standard (10 mg) was weighed accurately into a loo-ml flask and diluted to volume in acetonitrile. Subsequent dilutions were made in the HPLC mobile phase. Sep-Pak Crs cartridges (Waters Assoc., Milford, MA, U.S.A.) were washed with 10 ml of methanol and then 20 ml of distilled water before use. Hyflo Super-Cel was obtained from Johns-Manville (Denver, CO, U.S.A.). Other chemicals were of reagent grade or of HPLC grade. Apparatus
The HPLC system consisted of a solvent-delivery system LC-SA, an UV detector SPD-2A operated at 280 nm, a fluorescence detector RF-530 operated at an excitation wavelength of 365 mu and an emission wavelength of 325 nm and a Chromatopak C-R3A data system, all from Shimadzu Seisakusho (Kyoto, Japan). The separation of NA, OXA and PMA was performed on a Kaseisorb LC ODS 300-5 column (250 mm x 4.6 mm I.D.; Tokyo Kasei Kogyo, Tokyo, Japan) with 5 mM sodium dihydrogenphosphate-acetonitrile (6:4) as a mobile phase at a flow-rate of 0.5 ml/min. A mass spectrometer, GCMS-QP 1000 (Shimadzu), was employed in electron impact (El) mode under the following conditions; ionization energy, 70 eV; ionsource temperature, 250°C. The other instruments used were a Type 330 spectrophotometer (Hitachi, Tokyo, Japan), a Type 650-40 spectrofluorometer (Hitachi) and an Hiscotolon homogenizer (Nichion Irika Kikai, Tokyo, Japan). Sample preparation
A 5-g amount of sample was homogenized with 100 ml of 0.1% metaphosphoric acid-methanol (6:4) as deproteinizing extractant at high speed for 2 min. The homogenate was filtered through 2 mm Hyflo Super-Cel coated on a suction funnel. The filtrate was evaporated under reduced pressure at a temperature of 40°C. Evaporation was interrupted when 20 ml of solution remained in the flask. The flask contents were applied to a Sep-Pak C 18 cartridge and eluted with a flow-rate of 5 ml/mm. After washing with 20 ml of distilled water and 10 ml of S%.methanol, the
HPLC OF NALIDIXIC, OXOLINIC AND PIROMIDIC ACIDS
303
cartridge was eluted with 10 ml of methanol. The eluate was evaporated to dryness under reduced pressure and the residue dissolved in 1 ml of HPLC mobile phase; 10 ~1 of the solution were injected for HPLC. Calibration curves
Standards at concentrations of 0.1, 0.2, 0.4, 0.6, 0.8 and 1.0 pg/ml of OXA, 0.2,0.4,0.8, 1.2, 1.6 and 2.0 fig/ml of NA and PMA were prepared from the standard stock solutions. A lo-p1 volume of these solutions was injected into the column. Calibration curves were obtained by the measurement of peak height. Conjirmation of OXA by mass spectrometry (MS)
The identity of OXA found in sweet fish was confirmed by MS. OXA was fractionated preparatively by HPLC under the conditions described in Apparatus. The fractionated solution containing OXA was removed under reduced pressure. The residue was dissolved in 100 ml of distilled water and adjusted to pH 3. The solution was then extracted with 30 ml of chloroform, and the chloroform phase was evaporated to dryness under reduced pressure for MS analysis. RESULTS AND DISCUSSION
Chromatographic conditions
Excitation and emission spectra of NA, OXA and PMA dissolved in the HPLC mobile phase are presented in Fig. 1 and UV absorption spectra in Fig. 2. The wavelengths of maximum excitation of NA and OXA were 326 and 334 nm, and the wavelengths of maximum emission were 357 and 382 nm, respectively. Therefore, wavelengths of 325 nm for excitation and 365 nm for emission were chosen for detection of NA and OXA. Fluorometry is desirable because of its specificity. However, fluorometric detection of PMA was abandoned due to its low intensity. Thus, UV detection at 280 nm, which is the wavelength of maximum UV absorption of PMA, was adopted.
iA)
(B) A,,326
o-
mn
A,,334
mn
1 . 300 400 400 Wavelength (nm) Wavelength (run) Fig. 1. Excitation and emission spectra of (A) nalidixic acid and (B) oxolinic acid in the mobile phase. 300
304
M. HORIE et al. (280 tnn) 0.8 0.7 0.6
9) 0.5 B 2
0.4
f 2
0.3 0.2
0.1
;: : :: :......”
.._. .......
0 250
300
Wavelength
350
400
(nm)
Fig. 2. Absorption spectra of 4 pg/ml nalidixic acid (piromidic acid (. . . . .) in the mobile phase.
), 4 pg/ml oxolinic acid (. -
. -.) and 4 pg/ml
It is known that NA, OXA and PMA give broad or tailing peaks in reversedphase chromatography. Although methylationQ*13 and paired-ion techniques 10,11~14,15have been investigated, these methods did not give suitable results in respect of separation or sensitivity for residual analysis. In this study, Nucleosil 5C1s and Kaseisorb LC ODS 300-5 were compared for the separation of these three drugs, using phosphate buffer-acetonitrile as a mobile phase. The latter is a reversed-phase column with silica gel of pore size 300 A, commonly used for protein and peptide analysis, and the former is a popular ODS column with silicas gel of pore size 100 A. Kaseisorb LC ODS 300-5 allowed more efficient separation of chromatographic peaks and thus it was used for subsequent experiments. This column has a smaller surface area than silica gel columns of pore sizes 60 and 100 A, and has a lower amount of silanol groups on its surface. Thus, the carbon number ratio is believed to be small. In addition, it appears to have less effect on the chromatography, and therefore, should be suitable for these three drugs which are retained more strongly and give tailing peaks on the popular ODS columns. The effects of the mixing ratio, sodium dihydrogenphosphate concentration and pH of the mobile phase solution (phosphate buffer-acetonitrile) on the capacity factor, k’, and separation were studied. On increasing the content of acetonitrile, the k’ value of OXA, NA and PMA decreases remarkably. Considering the separation and retention times of the three compounds, a phosphate buffer-acetonitrile ratio of 6:4 was chosen for the analysis. Under these conditions, the effect of the sodium dihydrogenphosphate concentration was studied up to 50 mM. No separation was observed in the absence of phosphate buffer, while 5 mM phosphate gave complete separation. No significant
HPLC OF NALIDIXIC, OXOLINIC AND PIROMIDIC ACIDS
-
(Cl
(B)
(A)
--T w
2
--
I
II
t
12
3
tt
1
tt
t
12
123
2
305
3
1
FL
, 0
12
3
111,
III i!
10 Timetmin)
0
10 Time(min)
0
10 Timetmin)
Fig. 3. Typical chromatograms of extracts from cultured fish. Top traces: ultraviolet absorbance (280 nm). Bottom traces: fluorescence detection at the excitation wavelength of 325 nm and the emission wavelength of 365 nm. (A) Standard mixture. Peaks: 1 = OXA (10 ng); 2 = NA (20 ng); 3 = PMA (20 ng). (B) Yellowtail extract. (C) Sweet fish extract.
differences in k’ and peak shapes were observed at concentrations higher than 5 mM. Thus, a 5 mM phosphate buffer was employed. The pH was found to have no effect in the range studied, 2.5-5.0. On the basis of these experiments, 5 mM sodium dihydrogenphosphate-acetonitrile (6:4) without pH conditioning was used as a mobile phase. Fig. 3A shows typical chromatograms of NA, OXA and PMA under the established conditions. Clean-up
Liquid-liquid partition using ethyl acetate12, dichloromethane13 and chloroform-methanol14 have been reported as means of clean up from fish. An unified extraction clean-up procedure is desirable to establish a rapid and widely used residual analysis for antibiotics and synthetic antibacterials in animal and fish samples. Attempts were made to determine NA, OXA and PMA using the previously proposed method” for simultaneous residual analysis of veterinary drugs in fish culture: extraction solvent, 0.5% metaphosphoric acid (MPA)-methanol (8:2); SepPak Cl s cartridges. Table I shows the results of recovery experiments on yellowtail spiked with 0.2 (OXA), 0.4 pg/g (NA and PMA). The recoveries of NA and OXA were > 70%, but that of PMA was as low as 65%. The effect of the methanol content on the recovery of OXA, NA and PMA was examined. Over 40% methanol provided significant recoveries, whereas larger contents resulted in an increase in interfering peaks. Thus, 40% methanol was chosen for the analysis. Table II shows the effect of the MPA concentration on the recovery of OXA, NA and PMA. The results indicated that the MPA content was of secondary im-
M. HORIE et al.
306 TABLE I
EFFECT OF METHANOL CONTENT IN THE EXTRACTING SOLVENT ON THE RECOVERY OF OXOLINIC ACID, NALIDIXIC ACID AND PIROMIDIC ACID FROM YELLOWTAIL Values are mean f S.D. (n = 5). Samples were spiked with 0.2 pg/g of oxolinic acid, 0.4 &g acid and 0.4 &g of piromidic acid.
Extracting solvent
Recovery ( %) OXA
0.5% MPA 0.5% 0.5% 0.5% 0.5% 0.5% 0.5%
MPA-methanol MPA-methanol MPA-methanol MPA-methanol MPA-methanol MPA-methanol
of nalidixic
(9: 1) (8:2) (7:3) (64) (5:5) (3:7)
61.5 71.0 77.8 80.7 86.8 85.0 89.5
NA f 2.3 f 2.4 f 4.3 f 2.2 f 2.3 f 3.7 f 3.3
65.7 69.4 75.5 78.9 84.2 84.6 86.4
PMA f 2.2 f 2.9 f 3.5 f 3.9 f 2.3 f 3.2 f 2.1
49.7 55.5 65.0 75.7 82.5 87.5 88.1
f f f f f f f
1.7 2.1 2.5 2.1 2.1 3.2 1.8
portance. Lower concentrations of MPA gave less interfering peaks. An extractant of 0.05% MPA-methanol (6:4) was not effective enough with regard to deproteinization. and successive filtration. Based on the above experiments, 0.1% MPA-methanol (6:4) was chosen as a deproteinization extractant. Fig. 3B and C show chromatograms of yellowtail and sweet fish samples. Similar chromatograms were obtained from eel, rainbow trout and common carp samples. Recovery
Linear calibration graphs were obtained from 1 to 10 ng OXA, 2 to 20 ng NA and PMA. Table III summarizes the recoveries of the drugs from commercial samples of yellowtail, eel, rainbow trout, sweet fish and common carp spiked with 0.2 ,ug/g 0x4, 0.4 pg/g NA and PMA. Greater than 80% overall mean recoveries and 5% standard deviations within were obtained from every fish sample. The detection limits of the method were 0.01 pg/g for each drug. TABLE II EFFECT OF THE CONCENTRATION OF METAPHOSPHORIC ACID IN THE EXTRACTING SOLVENT ON THE RECOVERY OF OXOLINIC ACID, NALIDIXIC ACID AND PIROMIDIC ACID FROM YELLOWTAIL Details as in Table I.
Extracting solvent
Recovery (%) OXA
0.1% 0.2% 0.5% 1.0% 2.0% 4.0%
MPA-methanol(64) MPA-methanol (6:4) MPA-methanol (64) MPA-methanol (6:4) MPA-methanol (64) MPA-methanol (6:4)
83.7 84.3 86.8 83.6 84.1 83.9
PMA
NA f 3.5 f 2.4 f 2.3 f 3.6 f 3.1 f 3.9
81.9 82.0 84.2 87.3 85.6 86.0
f 3.7 f 4.3 f 2.3 f 3.1 f 3.2 f 3.5
80.9 82.5 82.5 80.1 80.6 78.7
f f f f f f
3.3 2.8 2.1 2.4 3.2 2.7
HPLC OF NALIDIXIC,
30-I
OXOLINIC AND PIROMIDIC ACIDS
TABLE III RECOVERIES OF OXOLINIC ACID, NALIDIXIC ACID AND PIROMIDIC ACID FROM CULTURED FISH Details as in Table I. Sample
Recovery (%) OXA
Yellowtail Eel Rainbow trout Sweet fish Carp
83.1 85.1 88.7 85.7 84.6
PMA
NA f f f f f
3.5 4.6 2.5 2.4 2.1
81.9 81.5 85.3 83.3 81.7
f f f f f
3.7 3.5 2.1 2.0 2.6
80.9 83.3 84.9 81.1 83.1
f f f f f
3.3 2.0 1.5 2.7 1.7
Drug interference studies No fluorescence was observed under the analytical conditions from the antibiotics oxytetracycline, chlortetracycline, tetracycline, doxycycline, spiramycin, erythromycin, ampicillin and colistin, and the synthetic antibacterials sulphamonomethoxine, sulphisozole, sulphadimethioxine, thiamphenicol and sodium nifulstyrate. OXA, NA and PMA were apparently separated from tetracyclines and sulphonamides, which have maximum absorption at 280 nm. In another words, the present procedure allowed over 70% recoveries of drugs such as oxytetracycline, chlortetracycline, spiramycin, sulphamonomethoxine and sulphadimethoxine. The sampling procedure can also be applied to determination of these drugs under previously reported HPLC conditions’ 7~18. Analysis of commercial samples This method was applied to the analysis of 10 commercial samples of yellowtail, eel and common carp, 20 samples of rainbow trout and 85 samples of sweet fish. OXA was found in 24 samples of sweet fish (28.2%) at levels ranging from 0.01 to 1.90 pg/g. Four samples contained over 1 ,ug/g OXA.
gl~~~~ ff--jfg 150
200
250
m/z
Fig. 4. Mass spectra of (A) oxolinic acid isolated by HPLC from sweet fish, and (B) authentic drug.
308
M. HORIE et al.
MS identification was carried out to confirm OXA in fractions obtained by NPLC and purified. Fig. 4 demonstrates the mass spectra of standard OXA (C13H21N05, MW 261) and the sample. The positions of the sample peaks coincide with those of the standard OXA and with the mass spectrum of OXA reported by Dicarlo et aLIP; a molecular ion peak at m/z 261 and a parent peak at m/z 217. The compound found in sweet fish was, therefore, identified as OXA. REFERENCES 1 2 3 4 5 6
T. Endo, K. Ogishima, H. Hayasaka and S. Kaneko, Bull. Jpn. Sot. Sci. Fish., 39 (1973) 165. H. K. L. Hundt and E. C. Barlow, J. Chromutogr., 223 (1981) 165. H.-L. Wu, T. Nakagawa and T. Uno, J. Chromatogr., 157 (1978) 297. H. Roseboom, R. H. A. Sorel, H. Lingeman and R. Bouman, J. Chromatogr., 163 (1979) 92. H.-L. Wu, L.-C. Hsu and C.-Y. Hsil, J. Chromafogr., 193 (1980) 476. E. W. McChesney, E. J. Froelich, G. Y. Lesher, A. V. R. Crain and D. Rosi, Toxicol. Appl. Pharmacol.,
6 (1964) 292. 7 L. Shargel, R. F. Koss, A. V. R. Crain and V. J. Boyle, J. Pharm. Sci., 62 (1973) 1452. 8 D. L. Sondack and W. L. Koch, J. Chromatogr., 132 (1977) 352. 9 R. H. A. Sore1 and H. Roseboom, J. Chromatogr., 162 (1979) 461. 10 G. Cuisinaud, N. Ferry, M. Se&a, N. Bernard and J. Sassard, J. Chromatogr., 181 (1980) 399. 11 R. H. A. Sorel, A. HulshotT and C, Snelleman, J. Chromaiogr., 221 (1980) 129. 12 Y. Kasuga, K. Otsuka, T. Sugiya and F. Yamada, J. Food. Hyg. Sot. Jpn., 22 (1981) 479. 13 OfJicial Publication, Veterinary Sanitation Division, Environmental Health Bureau, Ministry of Health
and Welfare, Tokyo, Vol. 2, 1982, No. 5, pp. I-11. 14 Oficial Publication, Veterinary Sanitation Division, Environmental Health Bureau, Ministry of Health and Welfare, Tokyo, Vol. 2, 1984, No. 7, pp. 1l-18, 15 S. Horii, C. Yasuoka and-M. Matsumoto, J. Chromutogr., 388 (1987) 459, 16 Y. Kasuga, T. Sugiya and F. Yamada, J. Food. Hyg. Sot. Jpn., 23 (1982) 344. 17 M. Horie, Y. Hoshino, N. Nose, H. Iwasaki and H. Nakazawa, Eisei Kagaku, 31 (1985) 371. 18 M. Horie, Y. Hoshino, N. Nose, H. Iwasaki, Y. Shida, H. Nakazawa and M. Fuji@ Bunseki Kugaku, 35 (1986) 219. 19 F. J. Dicarlo, C. C. Malcolm and R. C. Greenough, Arch. Biochem. Biophys., 127 (1968) 503.
SIMULTANEOUS DETERMINATION OF NALIDIXIC ACID, OXOLINIC ACID AND PIROMIDIC ACID IN FISH BY HIGH-PERFORMANCE LIQUID CHROMATOGRAPHY WITH FLUORESCENCE AND UV DETECTION
MASAKAZU HORIE*, KOUICHI SAITO, YOUJI HOSHINO and NORIHIDE NOSE Saitama Prefectural Institute of Public Health, 639-1. Kamiokubo. Urawa, Saitama 338 (Japan) EMIKO MOCHIZUKI Yamanashi Institute of Public Health, l-7-31, Fuji& Kofu, Yamanashi 400 (Japan) and HIROYUKI NAKAZAWA The Institute of Public Health, 4-6-1, Shirokanedai. Minato-ku. Tokyo 108 (Japan) (Received March 17th, 1987)
SUMMARY
A simple and rapid method for the simultaneous determination of nalidixic acid (NA), oxolinic acid (OXA) and piromidic acid (PMA) in cultured fish has been developed by high-performance liquid chromatography (HPLC). The drugs were extracted with 0.1% metaphosphoric acid-methanol (6:4), followed by a Sep-Pak C1 a clean-up procedure. The HPLC separation was carried out on a Kaseisorb LC ODS 300-5 cohunn (25 cm x 4.6 mm I.D.) using 5 mM phosphate buffer-acetonitrile (6:4) as a mobile phase. A fluorescence detector was used for NA and OXA at the excitation wavelength of 325 mu and the emission wavelength of 365 nm and an ultraviolet detector at 280 nm for PMA. The calibration graphs were rectilinear from 1 to 10 ng for OXA, from 2 to 20 ng for NA and PMA. The recoveries of NA, OXA and PMA added to fish were 81.5-85.3,83.7-88.7 and 80.9-84.9%, respectively, with high accuracy. The limits of detection were 0.01 pg/g for each drug.
INTRODUCTION
With the development of farming of fish such as eel and yellowtail, various kinds of antibiotics and synthetic antibacterials have been widely used for the prevention and treatment of infectious diseases in fish. Concern has arisen as to the presence of drug residues in fish tissues, and demand for a rapid, simple and sensitive analytical method of determining them has increased. Nalidixic acid (NA), oxolinic acid (OXA) and piromidic acid (PMA) are widely applied to fish such as eel, yellowtail, rainbow trout, sweet fish and common carp to treat a variety of gram-negative organisms l. Several methods have been developed for the assay of NA, OXA and PMA using thin-layer chromatography (TLC)Z, gas chromatography3-5, fluorometry6 and high-performance liquid chromatography 0021-9673/87/$03.50
0
1987 Elsevier Science Publishers B.V.
302
M. HORIE et al.
(HPLC)‘-l l. Most methods were developed for the assay of pharmaceutical preparations, and plasma and urine samples, only few for the residue analysis of fish sampleslz-14. A few HPLC methods with ultraviolet (W) detection were recently reported for the simultaneous determination of NA, OXA and PMA in fishlS,16. However, they were tedious and insufficiently sensitive. This paper describes a simple, specific and rapid HPLC method for the simultaneous determination of NA, OXA and PMA using W and fluorometric detection, and Sep-Pak Cl8 cartridges as a clean-up step. EXPERIMENTAL
Materials and reagents
The edible muscle tissues of yellowtail, eel, rainbow trout, sweet fish and common carp served as samples. NA, OXA and PMA were obtained from Daiichi Pharmaceutical (Tokyo, Japan), Tanabe Pharmaceutical (Osaka, Japan) and Dainihon Pharmaceutical (Osaka, Japan), respectively. Each standard (10 mg) was weighed accurately into a loo-ml flask and diluted to volume in acetonitrile. Subsequent dilutions were made in the HPLC mobile phase. Sep-Pak Crs cartridges (Waters Assoc., Milford, MA, U.S.A.) were washed with 10 ml of methanol and then 20 ml of distilled water before use. Hyflo Super-Cel was obtained from Johns-Manville (Denver, CO, U.S.A.). Other chemicals were of reagent grade or of HPLC grade. Apparatus
The HPLC system consisted of a solvent-delivery system LC-SA, an UV detector SPD-2A operated at 280 nm, a fluorescence detector RF-530 operated at an excitation wavelength of 365 mu and an emission wavelength of 325 nm and a Chromatopak C-R3A data system, all from Shimadzu Seisakusho (Kyoto, Japan). The separation of NA, OXA and PMA was performed on a Kaseisorb LC ODS 300-5 column (250 mm x 4.6 mm I.D.; Tokyo Kasei Kogyo, Tokyo, Japan) with 5 mM sodium dihydrogenphosphate-acetonitrile (6:4) as a mobile phase at a flow-rate of 0.5 ml/min. A mass spectrometer, GCMS-QP 1000 (Shimadzu), was employed in electron impact (El) mode under the following conditions; ionization energy, 70 eV; ionsource temperature, 250°C. The other instruments used were a Type 330 spectrophotometer (Hitachi, Tokyo, Japan), a Type 650-40 spectrofluorometer (Hitachi) and an Hiscotolon homogenizer (Nichion Irika Kikai, Tokyo, Japan). Sample preparation
A 5-g amount of sample was homogenized with 100 ml of 0.1% metaphosphoric acid-methanol (6:4) as deproteinizing extractant at high speed for 2 min. The homogenate was filtered through 2 mm Hyflo Super-Cel coated on a suction funnel. The filtrate was evaporated under reduced pressure at a temperature of 40°C. Evaporation was interrupted when 20 ml of solution remained in the flask. The flask contents were applied to a Sep-Pak C 18 cartridge and eluted with a flow-rate of 5 ml/mm. After washing with 20 ml of distilled water and 10 ml of S%.methanol, the
HPLC OF NALIDIXIC, OXOLINIC AND PIROMIDIC ACIDS
303
cartridge was eluted with 10 ml of methanol. The eluate was evaporated to dryness under reduced pressure and the residue dissolved in 1 ml of HPLC mobile phase; 10 ~1 of the solution were injected for HPLC. Calibration curves
Standards at concentrations of 0.1, 0.2, 0.4, 0.6, 0.8 and 1.0 pg/ml of OXA, 0.2,0.4,0.8, 1.2, 1.6 and 2.0 fig/ml of NA and PMA were prepared from the standard stock solutions. A lo-p1 volume of these solutions was injected into the column. Calibration curves were obtained by the measurement of peak height. Conjirmation of OXA by mass spectrometry (MS)
The identity of OXA found in sweet fish was confirmed by MS. OXA was fractionated preparatively by HPLC under the conditions described in Apparatus. The fractionated solution containing OXA was removed under reduced pressure. The residue was dissolved in 100 ml of distilled water and adjusted to pH 3. The solution was then extracted with 30 ml of chloroform, and the chloroform phase was evaporated to dryness under reduced pressure for MS analysis. RESULTS AND DISCUSSION
Chromatographic conditions
Excitation and emission spectra of NA, OXA and PMA dissolved in the HPLC mobile phase are presented in Fig. 1 and UV absorption spectra in Fig. 2. The wavelengths of maximum excitation of NA and OXA were 326 and 334 nm, and the wavelengths of maximum emission were 357 and 382 nm, respectively. Therefore, wavelengths of 325 nm for excitation and 365 nm for emission were chosen for detection of NA and OXA. Fluorometry is desirable because of its specificity. However, fluorometric detection of PMA was abandoned due to its low intensity. Thus, UV detection at 280 nm, which is the wavelength of maximum UV absorption of PMA, was adopted.
iA)
(B) A,,326
o-
mn
A,,334
mn
1 . 300 400 400 Wavelength (nm) Wavelength (run) Fig. 1. Excitation and emission spectra of (A) nalidixic acid and (B) oxolinic acid in the mobile phase. 300
304
M. HORIE et al. (280 tnn) 0.8 0.7 0.6
9) 0.5 B 2
0.4
f 2
0.3 0.2
0.1
;: : :: :......”
.._. .......
0 250
300
Wavelength
350
400
(nm)
Fig. 2. Absorption spectra of 4 pg/ml nalidixic acid (piromidic acid (. . . . .) in the mobile phase.
), 4 pg/ml oxolinic acid (. -
. -.) and 4 pg/ml
It is known that NA, OXA and PMA give broad or tailing peaks in reversedphase chromatography. Although methylationQ*13 and paired-ion techniques 10,11~14,15have been investigated, these methods did not give suitable results in respect of separation or sensitivity for residual analysis. In this study, Nucleosil 5C1s and Kaseisorb LC ODS 300-5 were compared for the separation of these three drugs, using phosphate buffer-acetonitrile as a mobile phase. The latter is a reversed-phase column with silica gel of pore size 300 A, commonly used for protein and peptide analysis, and the former is a popular ODS column with silicas gel of pore size 100 A. Kaseisorb LC ODS 300-5 allowed more efficient separation of chromatographic peaks and thus it was used for subsequent experiments. This column has a smaller surface area than silica gel columns of pore sizes 60 and 100 A, and has a lower amount of silanol groups on its surface. Thus, the carbon number ratio is believed to be small. In addition, it appears to have less effect on the chromatography, and therefore, should be suitable for these three drugs which are retained more strongly and give tailing peaks on the popular ODS columns. The effects of the mixing ratio, sodium dihydrogenphosphate concentration and pH of the mobile phase solution (phosphate buffer-acetonitrile) on the capacity factor, k’, and separation were studied. On increasing the content of acetonitrile, the k’ value of OXA, NA and PMA decreases remarkably. Considering the separation and retention times of the three compounds, a phosphate buffer-acetonitrile ratio of 6:4 was chosen for the analysis. Under these conditions, the effect of the sodium dihydrogenphosphate concentration was studied up to 50 mM. No separation was observed in the absence of phosphate buffer, while 5 mM phosphate gave complete separation. No significant
HPLC OF NALIDIXIC, OXOLINIC AND PIROMIDIC ACIDS
-
(Cl
(B)
(A)
--T w
2
--
I
II
t
12
3
tt
1
tt
t
12
123
2
305
3
1
FL
, 0
12
3
111,
III i!
10 Timetmin)
0
10 Time(min)
0
10 Timetmin)
Fig. 3. Typical chromatograms of extracts from cultured fish. Top traces: ultraviolet absorbance (280 nm). Bottom traces: fluorescence detection at the excitation wavelength of 325 nm and the emission wavelength of 365 nm. (A) Standard mixture. Peaks: 1 = OXA (10 ng); 2 = NA (20 ng); 3 = PMA (20 ng). (B) Yellowtail extract. (C) Sweet fish extract.
differences in k’ and peak shapes were observed at concentrations higher than 5 mM. Thus, a 5 mM phosphate buffer was employed. The pH was found to have no effect in the range studied, 2.5-5.0. On the basis of these experiments, 5 mM sodium dihydrogenphosphate-acetonitrile (6:4) without pH conditioning was used as a mobile phase. Fig. 3A shows typical chromatograms of NA, OXA and PMA under the established conditions. Clean-up
Liquid-liquid partition using ethyl acetate12, dichloromethane13 and chloroform-methanol14 have been reported as means of clean up from fish. An unified extraction clean-up procedure is desirable to establish a rapid and widely used residual analysis for antibiotics and synthetic antibacterials in animal and fish samples. Attempts were made to determine NA, OXA and PMA using the previously proposed method” for simultaneous residual analysis of veterinary drugs in fish culture: extraction solvent, 0.5% metaphosphoric acid (MPA)-methanol (8:2); SepPak Cl s cartridges. Table I shows the results of recovery experiments on yellowtail spiked with 0.2 (OXA), 0.4 pg/g (NA and PMA). The recoveries of NA and OXA were > 70%, but that of PMA was as low as 65%. The effect of the methanol content on the recovery of OXA, NA and PMA was examined. Over 40% methanol provided significant recoveries, whereas larger contents resulted in an increase in interfering peaks. Thus, 40% methanol was chosen for the analysis. Table II shows the effect of the MPA concentration on the recovery of OXA, NA and PMA. The results indicated that the MPA content was of secondary im-
M. HORIE et al.
306 TABLE I
EFFECT OF METHANOL CONTENT IN THE EXTRACTING SOLVENT ON THE RECOVERY OF OXOLINIC ACID, NALIDIXIC ACID AND PIROMIDIC ACID FROM YELLOWTAIL Values are mean f S.D. (n = 5). Samples were spiked with 0.2 pg/g of oxolinic acid, 0.4 &g acid and 0.4 &g of piromidic acid.
Extracting solvent
Recovery ( %) OXA
0.5% MPA 0.5% 0.5% 0.5% 0.5% 0.5% 0.5%
MPA-methanol MPA-methanol MPA-methanol MPA-methanol MPA-methanol MPA-methanol
of nalidixic
(9: 1) (8:2) (7:3) (64) (5:5) (3:7)
61.5 71.0 77.8 80.7 86.8 85.0 89.5
NA f 2.3 f 2.4 f 4.3 f 2.2 f 2.3 f 3.7 f 3.3
65.7 69.4 75.5 78.9 84.2 84.6 86.4
PMA f 2.2 f 2.9 f 3.5 f 3.9 f 2.3 f 3.2 f 2.1
49.7 55.5 65.0 75.7 82.5 87.5 88.1
f f f f f f f
1.7 2.1 2.5 2.1 2.1 3.2 1.8
portance. Lower concentrations of MPA gave less interfering peaks. An extractant of 0.05% MPA-methanol (6:4) was not effective enough with regard to deproteinization. and successive filtration. Based on the above experiments, 0.1% MPA-methanol (6:4) was chosen as a deproteinization extractant. Fig. 3B and C show chromatograms of yellowtail and sweet fish samples. Similar chromatograms were obtained from eel, rainbow trout and common carp samples. Recovery
Linear calibration graphs were obtained from 1 to 10 ng OXA, 2 to 20 ng NA and PMA. Table III summarizes the recoveries of the drugs from commercial samples of yellowtail, eel, rainbow trout, sweet fish and common carp spiked with 0.2 ,ug/g 0x4, 0.4 pg/g NA and PMA. Greater than 80% overall mean recoveries and 5% standard deviations within were obtained from every fish sample. The detection limits of the method were 0.01 pg/g for each drug. TABLE II EFFECT OF THE CONCENTRATION OF METAPHOSPHORIC ACID IN THE EXTRACTING SOLVENT ON THE RECOVERY OF OXOLINIC ACID, NALIDIXIC ACID AND PIROMIDIC ACID FROM YELLOWTAIL Details as in Table I.
Extracting solvent
Recovery (%) OXA
0.1% 0.2% 0.5% 1.0% 2.0% 4.0%
MPA-methanol(64) MPA-methanol (6:4) MPA-methanol (64) MPA-methanol (6:4) MPA-methanol (64) MPA-methanol (6:4)
83.7 84.3 86.8 83.6 84.1 83.9
PMA
NA f 3.5 f 2.4 f 2.3 f 3.6 f 3.1 f 3.9
81.9 82.0 84.2 87.3 85.6 86.0
f 3.7 f 4.3 f 2.3 f 3.1 f 3.2 f 3.5
80.9 82.5 82.5 80.1 80.6 78.7
f f f f f f
3.3 2.8 2.1 2.4 3.2 2.7
HPLC OF NALIDIXIC,
30-I
OXOLINIC AND PIROMIDIC ACIDS
TABLE III RECOVERIES OF OXOLINIC ACID, NALIDIXIC ACID AND PIROMIDIC ACID FROM CULTURED FISH Details as in Table I. Sample
Recovery (%) OXA
Yellowtail Eel Rainbow trout Sweet fish Carp
83.1 85.1 88.7 85.7 84.6
PMA
NA f f f f f
3.5 4.6 2.5 2.4 2.1
81.9 81.5 85.3 83.3 81.7
f f f f f
3.7 3.5 2.1 2.0 2.6
80.9 83.3 84.9 81.1 83.1
f f f f f
3.3 2.0 1.5 2.7 1.7
Drug interference studies No fluorescence was observed under the analytical conditions from the antibiotics oxytetracycline, chlortetracycline, tetracycline, doxycycline, spiramycin, erythromycin, ampicillin and colistin, and the synthetic antibacterials sulphamonomethoxine, sulphisozole, sulphadimethioxine, thiamphenicol and sodium nifulstyrate. OXA, NA and PMA were apparently separated from tetracyclines and sulphonamides, which have maximum absorption at 280 nm. In another words, the present procedure allowed over 70% recoveries of drugs such as oxytetracycline, chlortetracycline, spiramycin, sulphamonomethoxine and sulphadimethoxine. The sampling procedure can also be applied to determination of these drugs under previously reported HPLC conditions’ 7~18. Analysis of commercial samples This method was applied to the analysis of 10 commercial samples of yellowtail, eel and common carp, 20 samples of rainbow trout and 85 samples of sweet fish. OXA was found in 24 samples of sweet fish (28.2%) at levels ranging from 0.01 to 1.90 pg/g. Four samples contained over 1 ,ug/g OXA.
gl~~~~ ff--jfg 150
200
250
m/z
Fig. 4. Mass spectra of (A) oxolinic acid isolated by HPLC from sweet fish, and (B) authentic drug.
308
M. HORIE et al.
MS identification was carried out to confirm OXA in fractions obtained by NPLC and purified. Fig. 4 demonstrates the mass spectra of standard OXA (C13H21N05, MW 261) and the sample. The positions of the sample peaks coincide with those of the standard OXA and with the mass spectrum of OXA reported by Dicarlo et aLIP; a molecular ion peak at m/z 261 and a parent peak at m/z 217. The compound found in sweet fish was, therefore, identified as OXA. REFERENCES 1 2 3 4 5 6
T. Endo, K. Ogishima, H. Hayasaka and S. Kaneko, Bull. Jpn. Sot. Sci. Fish., 39 (1973) 165. H. K. L. Hundt and E. C. Barlow, J. Chromutogr., 223 (1981) 165. H.-L. Wu, T. Nakagawa and T. Uno, J. Chromatogr., 157 (1978) 297. H. Roseboom, R. H. A. Sorel, H. Lingeman and R. Bouman, J. Chromatogr., 163 (1979) 92. H.-L. Wu, L.-C. Hsu and C.-Y. Hsil, J. Chromafogr., 193 (1980) 476. E. W. McChesney, E. J. Froelich, G. Y. Lesher, A. V. R. Crain and D. Rosi, Toxicol. Appl. Pharmacol.,
6 (1964) 292. 7 L. Shargel, R. F. Koss, A. V. R. Crain and V. J. Boyle, J. Pharm. Sci., 62 (1973) 1452. 8 D. L. Sondack and W. L. Koch, J. Chromatogr., 132 (1977) 352. 9 R. H. A. Sore1 and H. Roseboom, J. Chromatogr., 162 (1979) 461. 10 G. Cuisinaud, N. Ferry, M. Se&a, N. Bernard and J. Sassard, J. Chromatogr., 181 (1980) 399. 11 R. H. A. Sorel, A. HulshotT and C, Snelleman, J. Chromaiogr., 221 (1980) 129. 12 Y. Kasuga, K. Otsuka, T. Sugiya and F. Yamada, J. Food. Hyg. Sot. Jpn., 22 (1981) 479. 13 OfJicial Publication, Veterinary Sanitation Division, Environmental Health Bureau, Ministry of Health
and Welfare, Tokyo, Vol. 2, 1982, No. 5, pp. I-11. 14 Oficial Publication, Veterinary Sanitation Division, Environmental Health Bureau, Ministry of Health and Welfare, Tokyo, Vol. 2, 1984, No. 7, pp. 1l-18, 15 S. Horii, C. Yasuoka and-M. Matsumoto, J. Chromutogr., 388 (1987) 459, 16 Y. Kasuga, T. Sugiya and F. Yamada, J. Food. Hyg. Sot. Jpn., 23 (1982) 344. 17 M. Horie, Y. Hoshino, N. Nose, H. Iwasaki and H. Nakazawa, Eisei Kagaku, 31 (1985) 371. 18 M. Horie, Y. Hoshino, N. Nose, H. Iwasaki, Y. Shida, H. Nakazawa and M. Fuji@ Bunseki Kugaku, 35 (1986) 219. 19 F. J. Dicarlo, C. C. Malcolm and R. C. Greenough, Arch. Biochem. Biophys., 127 (1968) 503.