Multi-class Antibiotic Screening of Honey Using Online Extraction with LC-MS/MS

Importance of the Topic

Honey production frequently employs antibiotics to prevent bacterial diseases in bee colonies. Traces of these drugs can persist in commercial honey, posing risks of allergic reactions in sensitive individuals and promoting bacterial resistance. Regulatory agencies worldwide have therefore implemented residue monitoring programs. A rapid, reliable, multi-class screening method is essential for ensuring honey safety and compliance.

Objectives and Study Overview

The primary goal was to develop an automated LC-MS/MS workflow capable of simultaneously detecting representatives of four antibiotic classes—macrolides, sulfonamides, aminoglycosides, and tetracyclines—in honey. The study presents a single, generic method with minimal sample preparation and a total run time of under 18 minutes per injection.

Methodology and Used Instrumentation

Sample Preparation:

• Wildflower honey was diluted 1:1 (w/v) in McIlvaine buffer containing 0.1 M EDTA to reduce viscosity and chelate metal ions.
• Calibration and QC stocks for ten analytes were prepared in methanol:water or water, then spiked into the honey:buffer matrix.

Online Extraction (TurboFlow):

• Dual TurboFlow columns (Cyclone MAX and Cyclone-P, 0.5 × 50 mm each) connected in tandem removed sugars and other interferences during loading.
• Analytes were eluted onto a BETASIL Phenyl/Hexyl analytical column (100 × 3 mm, 3 µm) via gradient elution.

Liquid Chromatography and Mass Spectrometry:

• Aria TLX-1 system with TurboFlow technology and Aria operating software version 1.6.2.
• Mobile phases included acidified water, acetonitrile, ammonium acetate, nonafluoropentanoic acid (NFPA), and trifluoroacetic acid (TFA) mixtures.
• Detection employed a Thermo Scientific TSQ Quantum Ultra triple quadrupole with heated electrospray ionization in positive SRM mode.
• MS parameters: spray voltage 4,000 V, vaporizer 400 °C, capillary 370 °C, collision pressure 1.2 mTorr, scan time 0.02 s.

Main Results and Discussion

The method achieved excellent linearity over two orders of magnitude (r2 > 0.99) for all ten antibiotics. Limits of quantification ranged from 0.5 to 10 ng/mL depending on the analyte. Carryover was maintained below 20% at the upper calibration limit. Precision expressed as %CV was under 8% above the lower limit and below 19% at LLOQ. Quality control samples at four concentration levels yielded RSDs below 7% (erythromycin < 15%). A representative chromatogram demonstrated clear separation of all analytes at 100 ng/mL in the matrix.

Benefits and Practical Applications

This online extraction LC-MS/MS approach dramatically simplifies sample handling by eliminating offline cleanup steps such as pH adjustment, SPE, and evaporation. It reduces analysis time and labor while covering multiple antibiotic classes in a single run. The high throughput and robust performance make it suited for routine quality control laboratories and regulatory screening of honey.

Future Trends and Potential Uses

Further enhancements may include:

• Multiplexing on Aria TLX-4 to achieve four-fold throughput increases (up to 12 samples/hour).
• Expanding the analyte panel to other veterinary drugs and environmental contaminants.
• Lowering detection limits through optimized MS acquisition and sample loading.
• Adapting the workflow for other complex food or environmental matrices.

Conclusion

The described method leverages dual online TurboFlow extraction and rapid LC-MS/MS analysis to provide a broad, sensitive, and automated screening tool for antibiotic residues in honey. It offers minimal sample preparation, robust quantification, and the potential for high throughput, addressing the analytical challenges of multi-class residue monitoring.

Reference

1. Qualitative Identification of Tetracyclines in Tissues, CLG SOP No: CLG-TET2.01, Rev 01, p. 6, United States Department of Agriculture, Food Safety and Inspection Service, Office of Public Health Science, 9/25/03.