An analytical method combining high-throughput automatic solid-phase extraction with ultra-performance liquid chromatography-tandem mass spectrometry (UPLC-MS/MS) was developed to determine 16 antibiotics (macrolides, tetracyclines, quinolones, and sulfonamides) and 4 β-agonists (terbutaline, salbutamol, ractopamine, and clenbuterol) in human urine samples. After thawing at room temperature, 1 mL of urine was sampled and the internal standard was added, followed by the addition of 200 μL ammonium acetate buffer and 20 μL β-glucuronidase, and the mixture was incubated at 37 ℃ overnight. Automatic solid-phase extraction was used to extract the target compounds from the urine samples, and the recoveries were compared using different solid-phase extraction 96-well plates (PRiME MCX, Sep-Pak C18, PRiME HLB), types and volumes of rinse solutions and eluents. Satisfactory recoveries of the 20 target compounds were obtained using the Oasis PRiME HLB 96-well plate, with 1.5 mL 10% (v/v) methanol aqueous solution and 2.0 mL methanol as the rinse solution and eluent, respectively. The eluent was concentrated under nitrogen gas at 45 ℃, and the recoveries of the target compounds were compared under different conditions (completely or almost dry, drying to 1 mL, and adding water as a protective agent), and the recovery rate was optimal when water was added as a protective agent.
In this study, two types of analytical columns (ACQUITY BEH C18 and ACQUITY HSS T3) and different gradient elution procedures and mobile phases were compared. The optimal chromatographic effect was realized using an HSS T3 column (100 mm×3.0 mm, 1.8 μm) and 0.1% (v/v) formic acid aqueous solution-0.1% (v/v) formic acid in acetonitrile as the mobile phase in gradient elution at a flow rate of 0.3 mL/min. Comparing the peaks observed using different proportions of methanol aqueous solution and the initial mobile phase as the injection solvent revealed that 30% (v/v) methanol aqueous solution was the optimal solution in terms of peak shape and signal-to-noise ratio. MS was conducted using positive electrospray ionization (ESI+) in multiple reaction monitoring (MRM) mode, and the MS parameters were optimized, including the curtain (CUR) and collision gases (CAD). The standard curve obtained using this method exhibited a good linearity (correlation coefficient>0.997), and the respective limits of detection and quantification were 0.02-0.12 ng/mL and 0.06-0.41 ng/mL. At spiked levels of 0.25, 2.5, and 12.5 ng/mL, the recoveries were in the range of 81.7%-120.0% (except that of tetracycline), the intra- and inter-day RSDs (n=6) were 1.1%-11.0% and 1.2%-13.0%, respectively. Azithromycin, trimethoprim, terbutaline, salbutamol, ractopamine, and clenbuterol displayed moderate matrix effects, but all targets exhibited weak matrix effects after correction using the isotope internal standard. To evaluate the accuracy of this method, BCR-503 (containing salbutamol and clenbuterol) and internal quality control samples were used and the concentrations of salbutamol and clenbuterol were within the reference ranges. Additionally, the mean concentrations of the 20 target compounds of two different internal quality control samples after 7 measurements were in the ranges of 0.44-0.59 ng/mL (0.5 ng/mL) and 1.72-2.16 ng/mL (2.0 ng/mL), respectively, which were satisfactory. In this study, the analytical method employed automatic sample pretreatment with a 96-well solid-phase extraction plate, and the detection efficiency was considerably improved. This method displays the advantages of simple operation, ideal recovery, a high sensitivity and weak matrix effect, which satisfies the requirements for the simultaneous determination of 16 antibiotics and 4 β-agonists in human urine samples. This study provides a crucial method for use in monitoring antibiotics and β-agonists in human urine and studying their exposure characteristics and health risks.
