Relevant infant in vitro digestive models

Milk lipids supply 50-60 % of the calories necessary for newborn growth under the form of maternal milk or infant formulas. Infant formulas have been optimized in terms of chemical composition but not with regards to the structure of the emulsion. An important difference remains between the native emulsions of milk fat globules (4 µm) as opposed to the processed submicronic emulsions (0.5 µm) of infant formulas. This structural difference may modify the kinetics of digestion and disintegration of the emulsion in the digestive tract of the newborns. Indeed the native emulsions develop an interfacial surface of 1.8 m2/g of lipid stabilized by a natural trilayered membrane based of phospholipids and proteins, whereas infant formulas present a much higher neoformed interfacial surface (30 m2/g of lipid) stabilized by dairy proteins. This higher interfacial surface is supposed to increase infant formula susceptibility to lipolysis but also to proteolysis through the higher accessibility of proteases to adsorbed proteins. To check this hypothesis, 3 matrices were formulated: M1 containing raw bovine milk fat globules dispersed in an infant formula-type serum phase, M2 similar to M1 with an additional high pressure homogenization treatment and M3 similar to M2 with the addition of a pasteurization treatment. The matrices were subjected to the semi-dynamic gastric phase of an in vitro digestion mimicking the conditions reported in newborns. The disintegration and the kinetics of lipolysis and proteolysis were monitored during the 180 minutes of digestion. The initial structure affected both the kinetics of the hydrolysis and the disintegration of the matrices. The initial rate of lipolysis of M1 was ~ 55 times lower than for M2 or M3 which was partially explained by its lower specific surface. Total proteolysis of caseins was observed at 30 minutes on M3 whereas it only occurred after 120 minutes for M1. The initial structure impacted the gastric destabilization with the formation of larger aggregates for M1 (mode at 190 µm at 30 min) than on the two other matrices. Since the free fatty acids and amino acids liberated in the gastric phase can act as second messenger and modulate the gastric emptying, this effect of structure may have important physiologic consequences. The gastric in vitro semi-dynamic model allowed rapid screening of matrices and limitations of consumption of digestive enzymes. In the future, it will be employed on maternal milk to determine the additional effect of the endogenous lipase on the digestion.