Use chemometric techniques in the optimization of a specific bioassay for betalactams in milk

A microbiological bioassay using Geoacillus stearothermophilus was optimized to detect betalactams at concentrations near to the Maximum Residue Limits (MRLs), with low cross-specificity for tetracycline.

A factorial design (3 × 4) was used to evaluate the effects of concentration of spores (2·0 × 10(6), 4·0 × 10(6) and 8·0 × 10(6) spores ml(-1) ) and incubation time (3·0, 3·5, 4·0 and 4·5 h) on the response of the bioassay. Then, desirability function to raise the detection capabilities (CC(β) ) of tetracyclines and increase sensitivity to betalactams was implemented. Significant effects of Log[S] and incubation time [It] on the CC(β) of betalactams and tetracyclines were observed. Finally, high value of global desirability (D = 0·853), adequate betalactams CC(β) (3·8 μg l(-1) of penicillin 'G', 27 μg l(-1) of oxacillin, 8·1 μg l(-1) of ampicillin, 48 μg l(-1) of cloxacillin) and high tetracyclines CC(β) (5260 μg l(-1) chlortetracycline, 1550 μg l(-1) of oxytetracycline, 1070 μg l(-1) of tetracycline) were calculated.

The application of chemometric tools allows the optimization of a bioassay that detects betalactam residues in milk. The more robust conditions have been achieved in Log[S] = 6·30 and [It] = 4·20 h.

The logistic regression model and the desirability function are adequate chemometric techniques to improve the properties of the methods, because it is possible to increase sensitivity and decrease cross-specificity simultaneously.