Dynamic modeling of in vitro lipid digestion: Individual fatty acid release and bioaccessibility kinetics

The aim of this study was to gain knowledge about the role of triacylglycerol (TAG) 16 composition in fatty acids (FA) of o/w emulsions on both the pancreatic lipolysis kinetics and 17 the bioaccessibility of released products (i.e. contained within the bile salt micellar phase). A 18 mathematical model was developed and its predictions were compared to a set of 19 experimental data obtained during an in vitro digestion of a whey protein stabilized emulsion. 20 Modeling results show that FA residues of TAG were hydrolyzed at specific rates, inducing 21 different bioaccessibility kinetics. The estimated lipolysis rate constants of the studied FA 22 (C8:0, C10:0 >> C18:1 n-9 >> C12:0 > C14:0 > C16:0 ≈ C16:1 n-7 > C22:6 n-3) were in 23 close agreement with the available literature on the substrate specificity of pancreatic lipase. 24 Results also suggest that lipolysis products are very rapidly solubilized in the bile salt mixed 25 micelles with no fractionation according to the FA carbon chain.

. 58 Moreover, the extent of FFA and MAG that can be solubilized into the bile salt micelles also 59 seems to be product dependent (Freeman, 1969;Hofmann, 1963). For instance, the molar  with the available literature and discussed according to their biochemical meanings.

202
These 8 FA showed coefficients of variation of less than 10.26% over the three replicate 203 measurements. Since the emulsion was prepared from a mixture of two oils, the last 2 204 columns of Table 1 where is the mass of in the oil droplets (mg), is the lipolysis rate constant of 224 (mg m -2 min -1 ), and ( ) is the interfacial area occupied by (m 2 ) at time t (min).

225
A3: the interfacial area occupied by at time t was assumed to be proportional to its 226 molar fraction within the oil droplets.
where ( ) is the number of moles of in the oil droplets (mol) at time t and (t) is 228 the interfacial area of oil droplets (m 2 ) at time t. This equation can be rewritten as where is the molar mass of (g mol -1 ).

230
A4: oil droplets in the emulsion were considered as spheres. The droplet interfacial area 231 is thus: where ( ) is the total mass (mg) of TAG at time t, is the mean mass density of the 233 TAG (taken as 0.92 mg mm -3 ) and 32 ( ) is the surface weighted mean droplet diameter 234 measured by laser diffraction (nm).
(3) can be rewritten: A5: our experimental results show that only a fraction of hydrolyzed products were 238 transferred into the bile salt micellar phase. To calculate the masses of within this phase, 239 a "micellar fraction", , was therefore introduced with the assumption that the transfer of 240 lipolysis products into the micellar phase is instantaneous: where is the mass of in the micellar phase (mg).  Table   249 1 and by adding the adequate mass of glycerol moiety:  (Table 1). One may also notice that the long chain FA showed two-stage shapes with a slower

307
It appeared that the fatty acids coming from MCT, C8:0 and C10:0, were fully hydrolyzed in 308 less than 15 min of intestinal digestion (Fig. 1A and Fig. 2), i.e. before the first sampling time.

309
The lipolysis rate constants, 8 and 10 , could therefore not be estimated due to lack of Estimated values for the micellar fraction parameters (Table 2) can be interpreted as follows.

327
About 70% (w/w) of the C8:0 lipolysis products were released in the bile salt micellar phase, 328 and about 56% for the other FA. The remaining amounts of products were retained in the oil 329 and pellet phases during the centrifugation step. The order of magnitude of these values is 330 consistent with studies in which centrifugation was also employed to measure the micellar 331 content in lipolysis products (Sek et al., 2002). Note also that the model assumes an 332 instantaneous transfer of the products into the micellar phase. The very good fits of the 333 bioaccessiblility kinetics (Fig. 3) obtained under this assumption therefore strongly suggest 334 that the solubilization of the lipolysis products in the bile salt micelles was not rate limiting. The lipolysis rate constants estimated with the model are given in Table 2. The same data, but  We may also stress that, according to our results, C8:0 and C10:0 presented rate constants that 371 were at least 3 times the one estimated for oleic acid (Table 2)  remaining oil (Fig 5), which leads to a progressive decrease of the average reaction rate and a 407 better agreement with experimental measurements of the remaining LCT at 300 min (Fig 3).