HORII & OKU: JOURNAL OF AOAC INTERNATIONAL VOL. 83, NO.

1, 2000 17

DRUGS, COSMETICS, FORENSIC SCIENCES

Rapid Determination of Nosiheptide in Meat and Egg by Liquid Chromatography with Fluorescence Detection
HORII & OKU: JOURNAL OF AOAC INTERNATIONAL VOL. 83, NO. 1, 2000 SHOZO HORII The Tokyo Metropolitan Laboratory of Public Health, 3-24-1 Hyakunin-cho, Shinjuku-ku, Tokyo 169-0073, Japan NAOTO OKU University of Shizuoka, School of Pharmaceutical Sciences, 52-1 Yada, Shizuoka 422-8526, Japan

A procedure was developed to determine nosiheptide residues in marketed meat and egg. Acetonitrile was used for the extraction, and the extract was partitioned with hexane to remove fat. The lower layer was reconstructed and quantitated by liquid chromatography using fluorescence detection at 357 nm excitation and 500 nm emission. The mobile phase consisted of 0.025% phosphoric acidacetonitrile (50 + 50, v/v). Recoveries of nosiheptide from fortified samples ranged from 91.3 to 95.2% for swine muscle, 88.6 to 92.7% for chicken muscle, and 86.3 to 86.8% for egg. The method was used to monitor swine and chicken muscle and egg (20 samples each) in the market. Nosiheptide was not determined in all 60 samples.

osiheptide (NH; Figure 1) is a polythiazole antibiotic produced by Streptomyces actuosus and effective against gram positive microorganisms, e.g., Staphylococcus aureus and Bacillus subtilis. The compound is widely added in feed (0.620 ppm) to promote the growth of pigs and chickens. Japanese regulatory ranges specified in 1987 were 2.510 and 2.520 ppm for chicken and swine, respectively, using only purified bacteria cultures. The regulation was expanded to include crude bacterial cultures in 1993. Today, NH can be synthesized by chemical reaction (13). Although microbiological (47) and liquid chromatography (LC; 8) procedures to determine the level of NH residue in feed and premix have been reported, there has been no report on residual NH in animal tissue. This study describes a simple and sensitive LC method for determining residual NH in meat and egg. METHOD

(c) Column oven.Shimadzu Model CTO-10A. Set at 40EC. (d) Detector.Shimadzu Model RF-10AXL. Set at excitation wavelength 357 nm, emission wavelength 500 nm. (e) Degasser.Shimadzu Model DGU-14A. (f) System controller and data system.Shimadzu Model 10A system. (g) Analytical column.Inertsil ODS 3.5 m, 4.6 (N) 150 mm (GL Science, Tokyo, Japan). Flow rate was 1 mL/min. (h) Centrifuge.TOMY Model LX-130 (Tomy Seiko, Tokyo, Japan); 3000 rpm, 10 min, 4EC. (i) Rotary evaporator.EYELA Model NE (Tokyo Rikakikai, Tokyo, Japan). Water bath was set below 40EC. (j) Pipettor.Eppendorf Model 4910 (Eppendorf, Hamburg, Germany). (k) Water purification system.Millipore Model Milli Q ELIX (Millipore, Milford, MA). (l) Homogenizer.Yamato Model OMNI GLH (Yamato Kagaku, Tokyo, Japan). High speed, 3 min. (m) Shaker.Sugiyamagen Model ELVIS (Sugiyamagen, Tokyo, Japan). Shake vigorously, 5 min. (n) Disk filter.Cosmofilter-S, 13 mm (N) (Nacalai Tesque, Kyoto, Japan).

Reagents
(a) Solvents.Analytical grade hexane and LC grade acetonitrile. (b) LC water.Tap water purified with Milli Q ELIX system. (c) Phosphoric acid solution, 0.025% (w/v).Transfer 173 L (purity, 85%; density, 1.7 g/mL) orthophosphoric acid (Wako, Osaka, Japan) into 1000 mL volumetric flask and dilute to mark with LC water. (d) LC mobile phase.Acetonitrile0.025% phosphoric acid solution (50 + 50, v/v). (e) Standard solution.(1) Stock solution.Weigh 0.1000 g NH (Tokyo Hisiryou Kensasho, Tokyo, Japan; purity unknown). Transfer to 100 mL volumetric flask and dilute to volume with N,N-dimethylformamide (1000 ppm). (2) Intermediate solutions.Transfer 10 mL stock solution into 100 mL volumetric flask and dilute to volume with methanol (100 ppm). Dilute solution 1:10 with methanol (10 ppm).

Apparatus
(a) Liquid chromatograph.Shimadzu Model 10AT pumps ( 2; Shimadzu, Kyoto, Japan). (b) Auto sampler.Shimadzu Model SIL-10AXL. Automatically injected 20 (standard)30 (sample) min intervals. Injection volume was 10 L.
Received June 14, 1999. Accepted by JM September 17, 1999.

18 HORII & OKU: JOURNAL OF AOAC INTERNATIONAL VOL. 83, NO. 1, 2000 Table 1. Effects of added DDCTa
Food Swine muscle DDCT, mg 0 5 50 500 Chicken muscle 0 5 50 500 Egg 0 5 50 500
a

Recovery, % 94.3 93.6 92.3 0 91.0 90.5 88.8 0 88.3 86.7 80.2 0

CV, % 3.8 3.8 5.0

3.5 6.7 4.0

6.5 7.3 5.9

Mean of 3 determinations per DDCT dosage (homogenized twice; 4 g nosiheptide added).

Figure 1. Structure of nosiheptide.

(3) Working solution.Transfer 2, 10, and 20 mL intermediate solutions (10 ppm) into separate 100 mL volumetric flasks and dilute to volume with methanol (0.2, 1, and 2 ppm). Stock and intermediate standard solutions are stable at least 6 months in freezer at 20EC. Prepare working solutions weekly; keep refrigerated and protect from light throughout analysis. (f) DDCT solution.Weigh 10 g, dissolve, and dilute to volume (50 mL) with water (20%, w/v). Dilute solution 1:10 with water (2%, w/v), and dilute once more with water (0.2%, w/v).

Results and Discussion Shimada (8) reported that minerals in premix interfered with NH analysis. To prevent the disturbance due to minerals, he added sodium N,N-diethyldithiocarbamate trihydrate (DDCT) to the sample before extraction. We also studied the influence of added DDCT in meat and egg by the following procedure: Weigh 10 g sample and fortify 2 mL 2 ppm standard solution and 2.5 mL water (DDCT, 0 mg) or DDCT solutions (5, 50, and 500 mg). Continue procedure as described in Sample Preparation (Table 1). In samples with 5 and 50 mg added DDCT, average recovery values of NH did not differ from those of samples with no added DDCT, whereas, 500 mg DDCT seriously disturbed the results. Thus, DDCT was not used to determine NH. To select optimum extraction conditions, the number of step extractions of meat or egg was varied: Weigh 10 g sample and fortify 2 mL 2 ppm standard solution. Add 50 mL acetonitrile, homogenize, and centrifuge, once or twice. Shake
Table 2. Comparison of number of homogenizationsa
Food Swine muscle No. of homogenizations 1 2 Recovery, % CV, % 77.8 94.3 74.4 91.0 70.1 88.3 7.4 3.8 6.9 3.5 6.8 6.5

Sample Preparation
Weigh 10 g (meat, with skin and fat removed; egg, whole) and fortify with 2 mL appropriate working standard solution (0.2 and 1 ppm; 0.4 and 2 g/10 g). Add 50 mL acetonitrile, homogenize, and centrifuge. Transfer upper layer to a 300 mL volumetric separate funnel. Add additional 50 mL acetonitrile to pellet and repeat procedure. Combine upper layers and shake with 100 mL hexane. Transfer lower acetonitrile layer to a 300 mL volumetric round bottom flask, and add 10 mL 1-propanol to avoid bumping. Evaporate to dryness on rotary evaporator under vacuum. Dissolve residue with 2 mL methanol and pass through filter disk into LC vial and cap. Inject 10 L aliquot of filtrate into LC.

LC Analysis
Monitor determinations at single point injections (same concentration). For market survey, use 10 ppb standard solution. Prepare calibration curve from 3 injections of standards from 1 to 100 ppb. Compute linear regression data based on peak area response vs concentration.

Chicken muscle

1 2

Egg

1 2

Mean of 3 determinations per extraction (4 g nosiheptide added).

HORII & OKU: JOURNAL OF AOAC INTERNATIONAL VOL. 83, NO. 1, 2000 19

Figure 2. Typical liquid chromatograms (all 10 mL injections) of drug-free swine muscle (A); spiked with 0.4 mg/10 g NH (B); drug-free chicken muscle (C); spiked with 0.4 mg/10 g NH (D); drug-free egg (E); spiked with 0.4 mg/10 g NH (F).

upper layer with equal volume of hexane. Continue procedure as written in Sample Preparation (Table 2). The extraction efficiency depended on the number of homogenizations with acetonitrile. NH recovery reached about 90% with 2 homogenizations; therefore, it was decided that twice would be sufficient. NH had a retention time of about 6.3 min on the LC system. Representative liquid chromatograms are presented in Figures 2 AB for swine muscle, 2 CD for chicken muscle, and 2 EF for egg. The chromatograms demonstrate that no interfering peaks were present in either solvents or drug-free tissue and that background noise was well below the desired 10 ng NH/g tissue level. Early eluting peaks in the tissue matrix (seen before 3 min) varied in some samples. This may have resulted from normal biological variations of endogenous compounds, or from differences occurring during one-step purification. However, early eluting peaks did not interfere with quantitation of NH at 6.3 min. A reagent blank liquid chromatogram showed only solvent peak.
Table 3. Recovery of nosiheptide (NH) from fortified meat and egga
Food Swine muscle NH added, g/10 gb 0.4 2 Chicken muscle 0.4 2 Egg 0.4 2
a b

Recoveries ranged from 91.3 to 95.2% for swine muscle, 88.6 to 92.7% for chicken muscle, and 86.3 to 86.8% for egg, with coefficients of variation of 3.13.5, 3.84.6, and 3.95.6%, respectively. The data in Table 3 suggest that recoveries are better for swine and chicken muscle than for egg. Calibration graph for NH was constructed in the range of 1100 ppb (ng/mL). Correlation coefficient was 0.993. Limit of detection of pure standard in solvent was 1 ppb. Limits of quantitation in meats and egg were <10 ppb (signal-to-noise ratio > 3). Conclusion An analytical method was developed for determination of nosiheptide in swine and chicken muscle and in egg by LC with fluorescence detection. The method has good sensitivity and simplicity. References
(1) Umemura, K., Tate, T., Yamaura, M., Yoshimura, J., Yonezawa, Y., & Shin, C.-G. (1995) Synthesis 11, 14231426 (2) Shin, C.-G., Yamada, Y., Hayashi, K., Yonezawa, Y., Umemura, K., Tanji, T., & Yoshimura, J. (1996) Heterocycles 43, 891898 (3) Umemura, K., Noda, H., Yoshimura, J., Konn, A., Yonezawa Y., & Shin, C.-G. (1998) Bull. Chem. Soc. Jpn. 71, 13911396 (4) Pascal, C., Gaillard, C., & Moreau, M.-O. (1979) J. Assoc. Off. Anal. Chem. 62, 976981 (5) Fukumoto, Y. (1992) Shiryo Kenkyu Hokoku 17, 7280 (6) Akimoto, K., & Fukumoto,Y. (1995) Shiryo Kenkyu Hokoku 20, 8193 (7) Yamatani, S., & Akimoto, K. (1997) Shiryo Kenkyu Hokoku 22, 116125 (8) Shimada, H. (1990) Shiryo Kenkyu Hokoku 15, 6780

Recovery, % 91.3 95.2 88.6 92.7 86.8 86.3

CV, % 3.5 3.1 3.8 4.6 5.6 3.9

Mean of 5 determinations per fortification level. 0.4 g/10 g = 40 ppb; 2 g/10 g = 200 ppb.