RESUMEN
It is well established that different factors affect milk composition in cows and that milk composition, in turn, affect both technological and nutritional qualities. In this respect the comprehension of the metabolic variability of milk composition in relation to the lactation time as well as to the genetic background may be of paramount importance for the agri-food industries. In the present study we investigated the variations of the metabolic profiles during lactation in milks obtained from Friesian and autochthonous races from Northern Italy by 1H NMR metabolomics. Furthermore, the external factors influencing the milk composition were minimized: the cows were breeded in the same farm, were fed with the same diet and were paired for the lactation interval and lactation stage. Our results showed a difference in milk composition between races and in relation to late lactation. The PLS-DA analysis permitted to distinguish the Friesian and autochthonous cow milks at the investigated different lactation times. Interestingly, the metabolites significantly involved into the discrimination between races appeared to be also technological property parameters, highlighting the importance of maintaining the biodiversity of cow breeds. Therefore, NMR-based metabolomics of milk could represent an informative tool to identify metabolites involved in milk quality both from a nutritional and industrial perspective.