A Raman-spectroscopy-based approach for detection and discrimination of Streptococcus thermophilus and Lactobacillus bulgaricus phages at low titer in raw milk

In this study, a method combining Raman spectroscopy with chemometric analysis was developed for detection of phage presence in raw milk and discrimination of Streptococcus thermophilus and Lactobacillus bulgaricus phages which are among the main phages causing problems in dairy industry. For this purpose, S. thermophilus and L. bulgaricus phages were added into raw milk separately, and then some pretreatments such as fat separation, removal of casein, and filtration were applied to the raw milk samples. Raman spectra of the samples were collected and then analyzed using principal component analysis in order to discriminate these phages in raw milk. In the next step, dilutions of S. thermophilus phages in pretreated raw milk were prepared, and Raman spectra were collected. These spectra were analyzed by using partial least squares method to quantify phages in low titer. Consequently, it has been demonstrated that S. thermophilus and L. bulgaricus phages, which have titers sufficient to fail the fermentation (~?10⁷ pfu/mL) and have lower titers (10²–10³ pfu/mL), could be discriminated from antibiotic and each other. Additionally, low concentrations of S. thermophilus phages (10² pfu/mL) could be detected through Raman spectroscopy with a short analysis time (60 min) and high coefficient of determination (R²) values for both calibration (0.985) and validation (0.906) with a root mean square error of calibration of 70.54 and root mean square error of prediction of 165.47. However, a lower success was achieved with L. bulgaricus phages and the obtained coefficient of determination values were not sufficiently high (0.649).