Optimized Extraction and Cleanup Protocols for LC-MS/MS Multi-Residue Determination of Veterinary Drugs in Milk

In order to insure public health and safety, a reliable screening analysis is necessary to determine veterinary drug residue levels in milk samples. The compounds of interest range from highly polar water-soluble compounds to very non-polar fat-soluble compounds. There exists very effective extraction and cleanup  procedures for individual compounds or compound classes, but these methods are not well suited for a multi-class, multi-residue screening analysis.

• Solvent extraction (with excess acetonitrile or methanol) can be effective for many veterinary drug residues in milk, but highly water-soluble drugs, such as salbutamol, are not well recovered using this approach.
  
• Aqueous buffer extraction can also be effective for many compounds, but fatsoluble compounds, such as dexamethasone, are not well recovered using this approach.
  
• Traditional solid-phase extraction (SPE) enrichment and cleanup (retention/ wash/elution) has limited utility for multi-residue analysis. Because the range of acidity/polarity/solubility among the compounds is so broad, dispersive or pass-through SPE is preferred for multi-residue methods

Benefits

• Efficient, timesaving multi-class/ multi-residue methodology
• Straightforward sample preparation  for diverse range of analytes
  
•  Fast, sensitive UPLC/MS analysis
  

Optimized sample preparation and analysis protocols were developed for tandem LC-MS/MS determination of a wide variety of veterinary drug residues in milk samples. Samples are initially precipitated and extracted with an equal volume of acetonitrile. After the resulting extract is treated with acidified acetonitrile to precipitate remaining proteins, a simple SPE cleanup is performed using a Sep-Pak C₁₈ cartridge. After evaporation and reconstitution, the sample is analyzed using tandem LC-MS. Representative compounds were chosen from major classes of veterinary drugs including tetracyclines, fluoroquinolones, sulfonamides, macrolides, beta-lactams, NSAIDS, steroids, and beta-andrenergids.

Table 1 summarizes the MRM transitions and instrument parameters used for this study. Also presented in Table 1 are matrix-matched calibration data for each compound (calculated using the primary transition). 

Figure 2 shows a typical LC-MS chromatogram obtained from analysis of a matrix matched standard of erythromycin at 6.7 ng/g. Performance of the other compounds was similar. Table 2 shows the recovery and matrix effects observed for multiresidue milk analysis.

The procedure chosen for the milk analysis was to initially extract and precipitate a milk sample with an equivalent amount of acetonitrile and then to precipitate the remaining protein from the supernatant with acidified acetonitrile. A single step procedure was also considered by which the milk sample was directly precipitated with 0.2% formic acid in 80% acetontrile. This procedure was more straightforward and produced a final extract of similar cleanliness compared with the chosen protocol. However, there was significantly lower recovery for the most polar compounds, such as sulfanilamide, and virtually no recovery of chlorotetracycline. Another approach was considered, by which two separate extractions were performed. The first extraction, for the water soluble compounds, was accomplished using aqueous succinic buffer. The second, performed on the re-suspended pellet, was with acetonitrile. This approach requires that each fraction be worked up independently before ultimately combining fractions for a single injection. Performance was marginally better than the chosen procedure but at a much greater cost of time and materials. Although the chosen procedure requires two precipitation steps, it provided the best balance of preparative time with good method performance.

• A two step extraction/protein precipitation procedure was developed and demonstrated for milk analysis

The procedure was suitable for screening for a wide  range of veterinary drug residues

Recoveries averaged 67% (22-110) with the lowest  values for tetracyclines.

• A pass-thru SPE cleanup protocol using Sep-Pak C₁₈ was utilized for effective removal of residual fats. 
  
• The sample preparation methodology for milk produced an extract for LC/MS that was free of particulates and required no subsequent filtration prior to LC-MS analysis.