William A. Moats

Liquid chromatographic analysis of antibacterial drug residues in food products of animal origin

This paper reviews recent developments in the liquid chromatographic (LC) methods of analysis for the residues of antibiotics (aminoglycosides, chloramphenicol, sulfonamides, tetracyclines, macrolides, beta-lactams, etc.) in food products of animal origin. The review also covers clean-up procedures, such as, ultrafiltration, liquid-liquid partition, solid-phase extraction, immunoaffinity, and matrix solid-phase dispersion, for use as extraction, deproteination, and concentration steps. The LC methods offer considerable potential for rapid automated analysis, and some may be used as direct screening for residues in meat and milk.

Determination of penicillin G in milk by high-performance liquid chromatography with automated liquid chromatographic cleanup

Specific confirmatory tests are needed to identify and quantify beta-lactam antibiotic residues detected in milk at levels of less than 10 parts per 10(9) (ppb) by screening tests. A liquid chromatographic method for penicillin G was developed using the liquid chromatography system for cleanup as well as analysis. Milk was deproteinized with two volumes of acetonitrile. The acetonitrile was extracted with hexane-methylene chloride (1:1) and the remaining water layer was concentrated by evaporation. The water layer (2 ml = 5 ml milk) was injected onto a Polymer Laboratories PLRP-S column using a WISP autosampler with the solvent, 0.01 M pH 7.0 phosphate buffer (A). Penicillin G was eluted with acetonitrile (B) gradient A-B (100:0) (0-3 min)-(40:60) (25 min). Penicillin G eluted as a narrow band in less than 0.5 min. A narrow fraction containing penicillin G was collected and rechromatographed on the same type of column at low pH (1.96). This effectively separated penicillin G from interferences. Recoveries were 92 +/- 9% with a sensitivity limit near 2 ppb. The approach used is applicable to determination of other beta-lactam antibiotics but specific conditions for analysis must be determined for each one. The cleanup procedure can be automated using an autosampler, gradient controller, and fraction collector.

Multiresidue determination of β-lactam antibiotics in milk and tissues with the aid of high-performance liquid chromatographic fractionation for clean up

Screening of milk shipments for beta-lactam antibiotic residues is mandatory in the USA and is widely used in other countries. Interpretation of positive screening test results has been difficult. Only six beta-lactam antibiotics are approved for use in food-producing animals in the USA but many others are used in other countries. A multiresidue procedure was developed for identification and quantitation of unknown beta-lactam antibiotics. The residues were extracted with acetonitrile and tetraethylammonium chloride. The extract was concentrated by evaporation and filtered. The concentrated extract was then loaded onto an HPLC column in 100% 0.01 M KH2PO4 and eluted with an acetonitrile gradient. Fractions corresponding to analytes of interest were collected and tested for antibiotics using rapid milk screening tests. Fractions testing positive were analyzed by HPLC. The identity of beta-lactams was confirmed by treating a replicate with beta-lactamase.

Determination of tetracycline antibiotics in tissues and blood serum of cattle and swine by high performance liquid chromatography

Tissue homogenates and blood serum were acidified with hydrochloric acid and deproteinized with acetonitrile. Tetracyclines were partitioned into water and concentrated by solid-phase absorption on the analytical column form 0.01 M phosphoric acid-methanol (80:20). Tetracyclines were eluted with an acetonitrile gradient. An all-organic polymeric column (Polymer Labs. PLRP-S) was used. Similar results were obtained on a bonded reversed-phase column after addition of tetramethyl-ammonium chloride to the mobile phase. Recoveries were near 100% from blood serum, 83-94% from muscle, and 80-100% from liver and kidney with sensitivities of 0.1 ppm or less for muscle and blood serum.

Effect of the silica support of bonded reversed-phase columns on chromatography of some antibiotic compounds.

Chromatographic behavior of three types of antibiotics was investigated on bonded C18 and polymeric reversed-phase columns. With penicillins with carboxylic acid functions only, retention and separations on the two types of columns were similar. beta-Lactam antibiotics with basic functions did not give as sharp peaks on the C18 column unless a silanol blocking agent, tetramethyl ammonium chloride (TMA) was added. In 0.01 M orthophosphoric acid-acetonitrile, tetracyclines were separated on the polymeric reversed-phase columns, but not on the C18 columns. With addition of TMA, results on C18 and polymeric reversed-phase columns were nearly identical. Addition of an ion pair also improved separations on the C18 columns, but not as much as TMA. Interaction with the silica support of C18 columns was used to separate tylosin from interferences in extracts of biological materials. The results demonstrate the importance of interactions with the silica support in chromatography of basic antibiotics on C18 packings. These interactions can be either beneficial or detrimental to separations, depending on the conditions used.

High-performance liquid chromatographic determination of penicillin G, penicillin V and cloxacillin in beef and pork tissues

The objective was to develop confirmatory high-performance liquid chromatographic methods for penicillin residues in animal tissues with detection limits of less than or equal to 10 ng/g. A previously described procedure was modified by using a larger sample size and isocratic analysis. Tissues (15 g) were blended with 45 ml of water and 20 ml of homogenate were mixed with 40 ml acetonitrile and filtered. The filtrate (30 ml) was mixed with 10 ml of 0.2 M H3PO4 and extracted with methylene chloride. The combined methylene chloride layers were mixed with acetonitrile and hexane, washed with two 4-ml portions of water and then extracted with four 1-ml portions of 0.01 M phosphate buffer (pH 7). The combined buffer extracts were concentrated to 1 ml under reduced pressure. Analysis was isocratic during 0.01 M phosphate buffer (pH 7)-acetonitrile with proportions 85:15 (penicillin G), 82:18 (penicillin V) or 78:22 (cloxacillin). A polystyrene-divinylbenzene copolymer column, 150 x 4.6 mm I.D. (Polymer Labs. PLRP-S), was used with a flow-rate of 1 ml/min and detection at 210 nm. The presence of penicillins was confirmed by treating a duplicate sample with penicillinase. Recoveries were greater than 90% in most instances. Detection limits were 5 ng/g in muscle and higher in liver and kidney. The procedure is a simple and sensitive method for confirming the presence of penicillins in animal tissues.

Determination of penicillin G and cloxacillin residues in beef and pork tissue by high-performance liquid chromatography

Tissues were homogenized and then deproteinized with acetonitrile. The acetonitrile extract was partitioned between dichloromethane and pH 2.2 buffer and then extracted with pH 7 buffer. After addition of ammonium sulfate to the aqueous solution, it was mixed with acetonitrile. The acetonitrile extract was separated and evaporated, and the residue was taken up in water. The aqueous solution was analyzed by high-performance liquid chromatography (HPLC) on a C18, 10-microns particle size, reversed-phase column using gradient elution with 0.01 M orthophosphoric acid-acetonitrile (from 80:20 to 40:60 in 20 min) at a flow-rate of 1 ml/min and UV absorbance detection at 220 nm. Recoveries were generally greater than 90% with all tissues. Data on incurred residues from a treated cow showed recoveries of penicillin which were frequently several times higher by HPLC than by bioassay. Sensitivity limits in muscle were about 0.05 ppm for both penicillin G and cloxacillin, but higher in liver and kidney because of interferences. The method is suitable for other monobasic penicillins but not for dibasic or amphoteric penicillins.

Comparison of bonded, polymeric and silica columns for chromatography of some penicillins.

The retention of ampicillin, amoxicillin, and penicillin G was determined on ODS and cyanopropyl (CN)-bonded to silica, on an all-organic polymeric reversed-phase column (styrene-divinylbenzene) and on bare silica. Solvents tested were 0.01 M H3PO4 (pH 1.6)-acetonitrile and 0.002 M KH2PO4 (pH 4.6)-acetonitrile in the proportions 100:0 to 10:90. Retention with 0.002 M KH2PO4 was weaker than with 0.01 M H3PO4 at all proportions of acetonitrile on all packings. With 0.01 M H3PO4, retention of penicillins on the polymeric column decreased rapidly from 100 to 50% aqueous phase and remained low to 10% aqueous phase. On the ODS bonded column, retention was similar, except that retention of the aminopenicillins, amoxicillin and ampicillin increased at less than 30% aqueous phase. On the CN-bonded column, results were similar to those on the ODS column, except that the aminopenicillins were less strongly retained when using 100% 0.01 M H3PO4. On the bare silica, penicillin G was poorly retained at all proportions of acetonitrile, while retention of aminopenicillins increased rapidly at less than 50% 0.01 M H3PO4. Retention of aminopenicillins at less than 50% 0.01 M H3PO4 was parallel on bonded and bare silica but weaker on the bonded silica. A bonded organic layer weakened but did not prevent binding of amines to silica.