Article pubs.acs.org/jpr
Differential Digestion of Human Milk Proteins in a Simulated Stomach Model Qiang Zhang,*,† Judy K. Cundiff,† Sarah D. Maria,† Robert J. McMahon,† Martin S. J. Wickham,‡ Richard M. Faulks,‡ and Eric A. F. van Tol† †
Pediatric Nutrition Institute, Mead Johnson Nutrition, 2400 West Lloyd Expressway, Evansville, Indiana 47721, United States Institute of Food Research, Norwich, U.K.
‡
S Supporting Information *
ABSTRACT: A key element in understanding how human milk proteins support the health and development of the neonate is to understand how individual proteins are affected during digestion. In the present study, a dynamic gastric model was used to simulate infant gastric digestion of human milk, and a subsequent proteomic approach was applied to study the behavior of individual proteins. A total of 413 human milk proteins were quantified in this study. This approach demonstrated a high degree of variability in the susceptibility of human milk proteins to gastric digestion. Specifically this study reports that lipoproteins are among the class of slowly digested proteins during gastric processes. The levels of integral lysozyme C and partial lactadherin in milk whey increase over digestion. Mucins, ribonuclease 4, and macrophage mannose receptor 1 are also resistant to gastric digestion. The retention or enhancement in whey protein abundance can be ascribed to the digestive release of milk-fat-globule-membrane or immune-cell enclosed proteins that are not initially accessible in milk. Immunoglobulins are more resistant to digestion compared to total milk proteins, and within the immunoglobulin class IgA and IgM are more resistant to digestion compared to IgG. The gastric digestion of milk proteins becomes more apparent from this study. KEYWORDS: milk proteins, gastric digestion, proteomics, lipoproteins, lysozyme C, lactadherin, mucin, ribonuclease, immunoglobulins
■
influencing signaling pathways in various cellular processes.13 For instance, intestinal neonatal Fc receptors transfer IgGantigen complexes from the lumen across enterocytes and localize the antigens in the intestinal lamina propria. The sotransferred antigens stimulate pro-inflammatory effects and, in this way, help bolster the acquired immune system.14 The dynamics of digestion in the infant have been well characterized,15,16 and yet digestive studies are constrained by the implications of invasive research and in vitro models must often be used as a proxy. Many such in vitro models are overly simplified in their representation of the intricate chemical and physical forces in the digestive tract, which limit the direct applicability of findings to de facto digestive processes. Recent development of a sophisticated dynamic gastric model (DGM) has incorporated a more realistic representation of gastric anatomy and allows manipulation of gastric pH, enzymatic content, contraction, and emptying time, all which can be programmed to specifically mimic digestion in the infant.17 This model allows sampling over the course of digestion,
INTRODUCTION Human milk contains over 1500 proteins,1,2 many of which have been demonstrated to promote the health and development of infants in many ways in addition to simple nutrition.3−5 These include proteins such as lactoferrin (LTF) and lysozyme (LYZ) with purported bactericidal activity, the glycoprotein lactadherin (MFGE8) that appears to play a role in maintaining gut barrier function, and many others. It remains unclear how the gastric phase of digestion alters these proteins as they are emptied into the proximal intestine. This study uses a simulated infant stomach digestion model coupled with high resolution LC−MS/MS proteomic analysis to evaluate this issue. Gastrointestinal (GI) proteolysis in both the gastric and intestinal compartments has been demonstrated to reduce the quantities of intact proteins; nevertheless, several studies showed that specific proteins in human milk remain available after their passage through the GI tract in either primarily intact or partially digested forms.6−10 Immunoglobulin A (IgA) and LTF have been found to exert antimicrobial function in the intestinal lumen through direct interaction with pathogens, an activity that presumably requires at least partial retention of appropriate protein conformation.11,12 In addition, small quantities of milk proteins and peptides may be capable of © 2013 American Chemical Society
Received: October 21, 2013 Published: December 10, 2013 1055
dx.doi.org/10.1021/pr401051u | J. Proteome Res. 2014, 13, 1055−1064
Journal of Proteome Research
Article
Figure 1. Quantitative analysis of human milk proteins over the course of simulated gastric digestion. (A) Schematic representation of changes in gastric pH and concentration of whey proteins during dynamic gastric digestion. Half gastric emptying occurs at 48 min of digestion. (B) SDS-PAGE of the whey proteins from milk digesta.
pattern without a lag phase20−24 where 48 min corresponds to half gastric emptying of breast milk for young infants.25−27 In this study, human milk samples were donated from anonymous mothers. Milk samples obtained were collected from one breast, pooled separately into two 150 mL sets (approximately 3−5 donors for each set), frozen at −80 °C, and no institutional review board approval was required for the deidentified samples. When ready for use, each milk sample was thawed at 4 °C and 15 mL was removed for use as control at t = 0 min of gastric digestion. The remaining sample was added to the primed DGM, followed by addition of the gastric enzymes pepsin and lipase and acid at an initial rate of 0.5 mL/ min. The gastric fluid containing digestive enzymes, NaCl, CaCl2, NaH2PO4, and phosphatidyl choline was added dynamically to the DGM throughout the gastric process. Digestion continued over a 60-min period during which milk digesta was emptied and collected from the antrum of the DGM at 12-min intervals (Figure 1A). The pH and weight of the digesta at each time intervals, ranging from 12 to 60 min, were recorded (Supplementary Table 1, Supporting Information).
providing human milk digesta that can be characterized to understand protein digestion. Recently, we have expanded the human milk proteome, which provides a key foundation for the in-depth exploration of gastric digestion of milk proteins and their subsequent intestinal function.1,2 Using human milk digesta generated by the dynamic gastric model, the goal of this study is to establish the quantitative proteome of human milk digesta and the differential digestion of milk proteins during gastric processes, providing novel insights into the mechanisms that milk may provide to tailor the fate of proteins appropriately for their functionality. Quick digestion of some proteins may be advantageous to facilitate the bioavailability of nutrients, for example, providing quick absorption of active peptides or building block amino acids. This study also illuminates the mechanisms for retaining milk proteins in the gastrointestinal tract by protecting proteins from rapid gastric breakdown and even enhancing the availability of certain proteins in the aqueous phase of milk over the course of digestion. Though gastric digestibility of proteins is clearly not the only predictor of their intestinal bioavailability and function, this improved understanding may encourage further exploration of the crucial dietary benefits that milk proteins can provide for infant development.
■
In-Solution Tryptic Digestion, Isobaric TMT Labeling and ERLIC Fractionation
For each of the two sample sets, milk digesta from 0, 12, 24, 36, 48, and 60 min were thawed at 4 °C and then ultracentrifuged at 4 °C (100000g for 60 min) so that samples had a pellet on the bottom, a fat layer on the top, and delipidated whey supernatant in the middle. Gastric digestion typically reduces the quantities of intact milk proteins and leads to the formation of partial proteins (Figure 1B). In bottom-up proteomics, in vitro tryptic digestion further breaks down intact or partial proteins that had survived gastric digestion into tryptic peptides. To avoid the possible complication between protein gastric digestion and proteomic procedures, small peptides derived from gastric digestion were removed prior to proteomic tryptic digestion (Supplementary Figure 1, Supporting Information). To achieve this, the whey layer was filtered using a 10 kDa molecular-weight cutoff (MWCO) device (Millipore, Billerica, MA) and subjected to buffer-exchange with water. Protein concentration for filtered whey samples was determined with Dumas combustion methodology using an FP2000 analyzer (LECO, St. Joseph, MI). For each time point, 10 μL of filtered whey was removed for reduction and alkylation, followed by tryptic digestion, isobaric tagging, quenching of
MATERIALS AND METHODS
Chemicals
Liquid chromatography mass spectrometric- or proteomicgrade ammonium bicarbonate (NH4HCO3), acetic acid (HOAc), formic acid (FA), acetonitrile (ACN), H2O, and trypsin were purchased from Sigma-Aldrich (St. Louis, MO). Criterion Tris-HCl gradient gels, Laemmli sample buffer (4% SDS, 20% glycerol, 10% 2-mercaptoethanol, 0.004% bromophenol blue and 0.125 M TrisHCl, pH 6.8), Tris/glycine/SDS running buffer (25 mM Tris, 192 mM glycine, 0.1% SDS, pH 8.3), silver stain and Immun-Star Western C chemiluminescent reagent were purchased from Bio-Rad Laboratories (Hercules, CA). Simulated Dynamic Gastric Digestion
Gastric digestion was carried out for two sample sets using an in vitro stomach model, dynamic gastric model (DGM), with conditions mimicking pediatric gastric digestion at the age of 9−12 months.17−19 It is well established that the gastric emptying of liquids, including milk, follows an exponential 1056
dx.doi.org/10.1021/pr401051u | J. Proteome Res. 2014, 13, 1055−1064
Journal of Proteome Research
Article
normalized collision energy setting of 35, and minimum signal threshold of 10000 counts. In addition, three high-energy collision dissociations (HCD) were performed with a mass resolution set to 7500, an isolation width of m/z 3, a normalized collision energy setting of 50, and minimum signal threshold of 10000 counts. Stepped collision energy was enabled for HCD with a normalized collision energy width of 10 in two steps. For both CID and HCD, each MS/MS spectrum was averaged from three microscans. The maximum injection time is 500 ms for parent-ion analysis and 150 ms for product-ion analysis. Target ions already selected for MS/MS with a repeat count of 2 during a 30-s duration were dynamically excluded for 60 s. An AGC target value of 500 000 ions was used for full MS scans and 30 000 ions for MS/ MS scans. Duplicated LC/MS/MS analyses were carried out for each fraction.
unreacted TMT reagents, and peptide pooling according to the TMT protocol (TMT 6-plex Isobaric Label Reagent Set, Thermo-Fisher Scientific, San Jose, CA) with the following modification in protein precipitation: after protein alkylation, buffer exchange with 100 mM TEAB was used as a replacement step for overnight cold-acetone precipitation. For each of the two sample sets, milk digesta collected at 0, 12, 24, 36, 48, and 60 min of simulated gastric digestion were labeled with isobaric tags of 126, 127, 128, 129, 130, and 131 Da, respectively (Supplementary Figure 1). Instead of targeting kinetics of gastric digestion for intact proteins, this setup differentiates more rapidly digested proteins from more slowly digested ones. Though proteins released from the stomach would be subjected to further digestion during their intestinal passage, some quantities of the proteins may remain to exert functionality in the intestinal space. While the kinetics of gastric breakdown obtained couldbe used in part to assess protein bioavailability in intestinal space, the intestinal digestion of milk proteins is beyond the scope of this study and will not be discussed. Each TMT 6-plex peptide mixture was pooled, lyophilized, and resuspended in 200 μL of 90%ACN/0.1%HOAc for injection onto a 4 mm i.d. × 10 mm WAX guard column (PolyWAX LP, particle size 5 μm, pore size 1,000 Å, PolyLC Inc., Columbia, MD) connected to a 2.1 mm i.d. × 200 mm WAX column (PolyWAX LP, particle size 5 μm, pore size 300 Å, PolyLC Inc.). Electrostatic repulsion−hydrophilic interaction chromatography (ERLIC) peptide separation28 was carried out via HPLC (U3000, Dionex, Sunnyvale, CA) at a flow rate of 200 μL/min. A gradient was started with 100% A (98% ACN with 0.1% HOAc) for 10 min and ramped to 28% B (30% ACN with 0.1% FA) over 68 min and held for 20 min, followed by a gradient ramped to 100% B over 20 min and then held at 100% B for 10 min. UV absorption was monitored at 280 nm (Supplementary Figure 2, Supporting Information). Approximately 20 fractions with retention times ranging from 20 to 100 min were collected at 4-min intervals. Each fraction was dried under reduced pressure, reconstituted in 20 μL of H2O/0.1% FA, stored at −80 °C, and thawed at 4 °C when ready for LC−MS analysis.
Protein Identification and Quantification
Peaks in MS/MS spectra are in centroid format and were converted into peak profiles using Mascot Distiller (Matrix Science, London, UK; version 2.4.3). MS/MS spectra with charges +2, +3, and +4 were analyzed using Mascot search engine (Matrix Science, London, UK; version 2.3.2). Mascot was set up to search against the human UniprotKB database (20 319 entries; version 2011_08) assuming the digestion enzyme was trypsin with a maximum of one missed cleavage allowed. The searches were performed with a fragment ion mass tolerance of 0.6 Da and a parent ion tolerance of 10 ppm. Iodoacetamide derivatization of cysteine and TMT 6-plex derivatization of N-terminus and lysine were specified in Mascot as fixed modifications. Deamidation of asparagine and glutamine, oxidation of methionine, and acetylation of the Nterminus were specified in Mascot as variable modifications. Relative intensities of TMT 6-plex reporter ions for a given peptide were obtained from the Mascot program. The protein ratios were calculated using the weighted average of the individual ratios of the peptides that can be assigned to that protein. False-discovery rate (FDR) was estimated to be
