Lowering the lactose content of milk in vivo has two potential benefits: to develop milk products for lactose intolerant people (Delmont, 1983) and to produce a more concentrated milk, reducing milk storage volume and milking frequency without affecting the overall fat and protein yields. Lactose synthesis in the lactating mammary gland results from the interaction of α-lactalbumin (α-la) with a Golgian UDP-galactosyltransferase (EC 2.4.1.22), and it has been suggested that lactose and α-la contents are directly related (Fitzgerald et al. 1970). Evidence that α-la is the only protein responsible for the in vivo induction of lactose synthesis came recently from knock-out experiments of the relevant gene (Stinnakre et al. 1994; Stacey et al. 1995), which disrupt lactation, as lactose-deprived milk of α-la-deficient mice is highly viscous. However, although the lactose, fat and protein contents of the milk were found to be directly related to the quantity of α-la present (Stinnakre et al. 1994), only in mice homozygous for the α-la null allele was there a significant alteration of milk concentration (Stacey et al. 1995). Stacey et al. (1995) also demonstrated that human α-la could substitute for its mouse counterpart via homologous recombination.

In the present study, we report the generation of transgenic mice producing only bovine α-la by back-crossing animals from four independent bovine α-la transgenic lines with α-la-deficient mice. Our work confirms that foreign α-la can substitute for the mouse protein and suggests a negative correlation between the α-la content and protein and fat concentrations in milk.