Evan Abrahamse

Protein is an important part of human nutrition; it delivers essential amino acids (AA) and nitrogen-containing molecules, which are required for maintenance and growth. The AA requirements are described in age-dependent reference AA composition patterns. Current protein quality definitions, consider how well the source AA composition matches the reference and its total digestibility in the gastrointestinal tract. However, the kinetics of availability, reflected in plasma AA peak concentrations after ingestion, are usually not considered, while this is an important determinant of the use of the AA building blocks for tissue protein synthesis or oxidation. Another clinical outcome related to dietary protein intake is the amount of protein reaching the colon, which can negatively influence gut heath. Chapter 1 describes in detail the key determinants of postprandial plasma (pp) AA peak concentrations and colonic protein flow; the physiological, physical, and chemical processes that occur in the stomach and small intestine, including gastric emptying rate, protein hydrolysis kinetics and total protein digestibility. In addition, the current state of knowledge on the influence of characteristics of the ingested food, such as protein composition, product processing and product matrix on the mentioned gastrointestinal processes is provided. The relationships between gastrointestinal processes and product characteristics and their interplay are not well understood. Therefore, the overall research question addressed in this thesis was: How do these food characteristics impact overall protein digestion, and how is this reflected in postprandial plasma (pp) AA peak concentration and colonic protein flow?

The specific aims of this thesis were to:
1) Gain insights in the relative importance of gastric emptying and protein hydrolysis kinetics in determining the postprandial AA peak.
2) Differentiate the individual effects of the mentioned food characteristics on gastrointestinal processes.
3) Assess and improve correlation between in vitro and in vivo studies

The work of this thesis combined in vitro and in vivo digestion studies on milk proteins in humans and piglets to advance the ability to formulate products with an improved protein quality tailored to the specific needs of vulnerable populations as for instance infants or elderly people.

The requirement of leucine and essential AA (EAA) to stimulate muscle protein synthesis increases with age. To target muscle anabolism, it is suggested that higher pp concentrations of leucine and EAA are needed in elderly people. In chapter 2 the matrix effect of increased energy density on the pp AA concentrations of whey protein and casein was investigated in healthy elderly. Furthermore, to better understand the relative importance of gastric emptying and protein hydrolysis, in vitro investigations in a newly developed semi-dynamic model of the gastrointestinal tract (SIM) simulating elderly conditions were undertaken. Four iso-nitrogenous protein (21g) supplements were studied containing leucine-enriched whey protein with 150/320 kcal (WNSO/WP2O) or casein protein with 150/320 kcal (CNSO/CP2O); all products contained carbohydrates (10 or 32g) and fat (3 or 12g). The pp leucine peak concentration was twofold higher for WNSO vs. CNSO (521±15 vs. 260±15 μmol/L), higher for WP2O vs. CP2O (406±15 vs. 228±15 μmol/L), and higher for low-caloric vs. high-caloric products. Similar effects were observed for the peak concentrations of EAA and total AA (TAA). In vitro gastric coagulation was observed only for the casein protein supplements. Intestinal digestion for 90 min resulted in higher levels of free TAA, EAA, and leucine for WNSO vs. CNSO, for WNSO vs. WP2O, and for CNSO vs. CP2O. Addition of extra carbohydrates and lipids lowered the pp AA concentrations of whey protein and casein in healthy elderly. These differences appear to be mediated by the gastrointestinal behaviour of these products.

Multiple studies have indicated that formula-fed infants show a different growth trajectory compared to breast-fed infants. The observed growth rates are suggested to be linked to higher postprandial levels of branched chain AA (BCAA) and insulin related to differences in protein quality. In chapter 3 the effect of whey protein denaturation, as well as the effect of changing native milk protein composition by incorporation of non-clotting casein, on pp AA, insulin, GLP-1 and glucose concentrations in neonatal piglets was studied. Furthermore, gastric emptying measurement and pharmacokinetic modelling of pp AA data was performed to gain insights in the role of protein hydrolysis. Neonatal piglets were bolus fed two of three milk protein solutions: native whey protein isolate (NWPI), denatured whey protein isolate (DWPI), or protein base ingredient (PBI) for infant formula (IF) comprising whey and β-casein. DWPI (91% denatured protein) vs NWPI (91% native protein) showed lower essential AA (EAA) (~10%) and BCAA (13-19%) concentrations in the first 30-60 min. However, TAA concentration per time-point and AUC, as well as EAA and BCAA AUC were not different. PBI induced a ~30% lower postprandial insulin spike than NWPI, yet pp TAA concentration at several time-points and AUC was higher in PBI than NWPI. The TAA rate constant for absorption (ka) was twofold higher in PBI than NWPI. pp EAA concentrations and AUCs in PBI and NWPI were not different. Changing native whey protein composition by inclusion of non-clotting casein increased pp AA concentrations in neonatal piglets, but whey protein heat induced denaturation did not. Both interventions did not affect gastric emptying. The differences between PBI and NWPI were partly explained by the difference in AA composition, but more likely differences in protein hydrolysis and absorption kinetics.

Knowledge about how molecular properties of proteins affect their digestion kinetics is crucial to understand protein pp AA responses. Therefore, to further elucidate mechanisms of the observations in chapter 3, in chapter 4 protein hydrolysis and absorbable product release kinetics and mechanisms as affected by milk protein composition or denaturation were investigated in SIM and a dynamic in vitro model of the gastrointestinal tract (TIM-1) simulating infant conditions. In addition, directions for improved correlations between in vitro digestion measures and in vivo pp AA concentrations was sought. In both models, the degree of hydrolysis (DH), loss of intact protein, and release of absorbable products (SIM: <0.5 kDa peptides and free AA, TIM-1: bioaccessible AA) were monitored. Additionally, in SIM, intermediate product amounts and their characteristics were determined. DWPI showed considerably faster intact protein loss, but similar DH and absorbable product release kinetics compared with NWPI in both models. Furthermore, more, relatively large, intermediate products were released from DWPI than from NWPI. PBI showed increased intact protein loss, similar DH, and absorbable product release kinetics, but more, relatively small, intermediate products than NWPI. Thus, both whey protein denaturation and β-casein inclusion increased the rate of intact protein loss without affecting absorbable product release during in vitro digestion. Our results suggest that intermediate digestion product characteristics are important in relation to pp AA responses, as it was found that the correlation between in vitro digestion measures and in vivo pp AA concentrations was greatly improved when intermediate digestion products were taken into consideration. Chapter 5 provided detailed insights in the differences in kinetics of protein hydrolysis and absorbable product release of human milk (HM) and cow’s milk based IF in SIM simulating infant conditions. It has been shown that the pp EAA peak concentrations of IF are higher than those of HM in infants. In addition, several HM proteins have been recovered intact in infant stool and appeared digestion resistant in vitro. It was therefore hypothesized that gastrointestinal protein hydrolysis of IF is faster than HM and leads to accelerated absorbable digestion product release. The time course of DH, loss of intact protein and release of free AA and peptides was evaluated. Gastric DH increase was similar for IF and HM, but the rate of intestinal DH increase was 1.6 times higher for IF than HM. Intact protein loss in IF was higher than HM from 120 min gastric phase until 60 min intestinal phase. Intestinal phase total digestion product (free AA + peptides <5 kDa) concentrations increased ~2.5 times faster in IF than HM. IF gastrointestinal protein hydrolysis and absorbable product release is faster than HM, possibly due to the presence of digestion resistant proteins in HM. This might present an opportunity to further improve IF bringing it closer to HM. In chapter 6 IF protein digestibility and colonic protein flow as affected by formula protein composition and matrix was investigated in ileal cannulated piglets. An IF containing milk fermented by the bacteria Bifidobacterium breve and Streptococcus thermophilus has been reported to alleviate functional digestive symptoms in infants. It was hypothesized that improved protein digestibility of the fermented infant formula could contribute to this effect. The aim of this study was to evaluate the protein digestibility of a specific fermented (FF), a standard (SF), and an extensively hydrolyzed protein based (HF) formula. Four-week-old piglets (n=7) were fitted with a T-cannula at the terminal ileum and received each formula in a Latin square design. Ileal digesta were collected and analyzed for AA and proteolytic activity. FF had a significantly higher apparent ileal crude protein digestibility (92.1±1.0%) compared to SF and HF (84.4±1.0% and 83.9±0.9% respectively). The ileal crude protein flow of FF was significantly lower compared to that of SF and HF. The ileal flow of FF total proteolytic activity was significantly lower than that of SF, but not significantly different from that of HF (412±163 vs. 1530±163 and 703±156 kU/8h, respectively). Fermented formula had in piglets a significantly higher apparent ileal crude protein digestibility compared to the standard and the hydrolyzed formula, and displayed lower ileal proteolytic activity compared to standard formula, possibly via a mechanistic pathway that involves a different physiological response. Both effects may contribute to the alleviation of common digestive symptoms reported in infants fed fermented infant milk formula. In chapter 7 all observations were discussed, and future directions in research and nutritional solutions for infants and elderly people are proposed. It was concluded that not only gastric emptying rate, but also protein hydrolysis kinetics determine the pp AA peak concentration. However, protein hydrolysis kinetics have been generally overlooked and deserve more emphasis in future studies that try to explain or predict pp AA peak concentrations. Regarding the individual effects of food characteristics, firstly, it was concluded that protein composition is a major determinant of both gastric emptying and protein hydrolysis kinetics and is therefore a key means to further optimize pp AA peak concentration of nutritional solutions to meet the specific requirements of infants and elderly. Secondly, heat induced protein denaturation only limitedly affects gastric emptying, protein hydrolysis kinetics, and pp AA concentrations. However, the impact of heat induced denaturation on intermediate digestion product characteristics is large, which could potentially infer a change in presented epitopes and immunological activity or other bioactivity relevant for infant health. Interesting further investigations could therefore include the analysis of the molecular composition and bioactivity of these peptide fractions. Lastly, protein product matrix’s main effects are on gastric emptying via physical and physiological feedback mechanisms, but protein hydrolysis can also be affected. Concerning the correlation of in vitro and in vivo, the main challenge of in vitro protein digestion approaches remains to appropriately mimic the in vivo release of absorbable protein digestion products. In vitro protein digestion product release that includes intermediate peptides (i.e., TP, <5 kDa) correlates with in vivo postprandial AA peak concentrations better than does absorbable digestion product (FAA + di- and tripeptides) release only. However, further studies using BBM peptidases to determine the most appropriate molecular weight cut-off for the intermediate peptides are required. Lastly, recommendations for improvement of milk products to optimize protein quality tailored to the needs of elderly and infants were made that involve specific changes in protein composition. ABBREVIATIONS AA Amino acid ACW α-lactalbumin enriched cow`s whey AGE Advanced glycation end-products Ala α-lactalbumin AUC Area under the curve BCAA Branched chain amino acid BBM Brush border membrane Blg β-lactoglobulin CNSO Casein based nutritional supplement low caloric CP2O Casein based nutritional supplement high caloric CNS Casein based nutritional supplements CS Cleavage sites CM Cow`s milk CP Crude protein CW Cow`s whey DH Degree of hydrolysis DIAAS Digestible indispensable amino acid score DMI Dry matter intake DWPI Denatured whey protein isolate EAA Essential amino acid EMM Estimated marginal mean FAA Free amino acids FF Fermented infant formula GER Gastric emptying rate GI Gastrointestinal GIT Gastrointestinal tract HF Hydrolyzed protein based infant formula HCM Humanized cow`s milk HCW Humanized cow`s whey HM Human Milk HMW High molecular weight peptides (5-13 kDa) iAUC Incremental area under the curve IF Infant formula Ig Immunoglobulins Lf Lactoferrin Lz Lysozyme LMW Low molecular weight peptides (<0.5 kDa) MMW Medium molecular weight peptides (0.5-5 kDa) MPS Milk protein solutions N Nitrogen NPN Non protein nitrogen NWPI Native whey protein isolate PBI Protein base ingredient (for infant formula) PDCAAS Protein digestibility corrected amino acid score pp Postprandial plasma SA Serum albumin SF Standard infant formula SGF Simulated gastric fluid SIF Simulated intestinal fluid SIM Semi-dynamic in vitro model of the gastrointestinal tract SSF Simulated saliva fluid TIM-1 TNO dynamic in vitro model of the gastrointestinal tract TAA Total amino acids TP Total digestion products (<5 kDa) UHT Ultra-high temperature WPI Whey protein isolate WPNS Whey protein based nutritional supplements WNSO Whey protein based nutritional supplement low caloric WP2O Whey protein based nutritional supplement high caloric