Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2026 Jan 1.
Published in final edited form as: Food Chem. 2024 Aug 23;462:140973. doi: 10.1016/j.foodchem.2024.140973

Comparative proteomic analysis of donor human milk treated by high-pressure processing or Holder pasteurization on undigested proteins across dynamic simulated preterm infant digestion

Michael A Pitino 1, Deborah L O’Connor 2,3,4,5, Sharon Unger 2,3,4,6, Bum Jin Kim 1, Alain Doyen 7, Md Abdul Wazed 1, Sudarshan Kumar 8, Yves Pouliot 7, Debbie Stone 5, David C Dallas 1
PMCID: PMC11413619  NIHMSID: NIHMS2019934  PMID: 39208730

Abstract

High-pressure processing (HPP) of donor human milk (DM) minimally impacts the concentration and bioactivity of some important bioactive proteins including lactoferrin, and bile salt-stimulated lipase (BSSL) compared to Holder pasteurization (HoP), yet the impact of HPP and subsequent digestion on the full array of proteins detectable by proteomics remains unclear. We investigated how HPP impacts undigested proteins in DM post-processing and across digestion by proteomic analysis. Each pool of milk (n=3) remained raw, or was treated by HPP (500 MPa, 10 min) or HoP (62.5 °C, 30 min), and underwent dynamic in vitro digestion simulating the preterm infant. In the meal, major proteins were minimally changed post-processing. HPP-treated milk proteins better resisted proximal digestion (except for immunoglobulins, jejunum 180 min) and the extent of undigested proteins after gastric digestion of major proteins in HPP-treated milk was more similar to raw (e.g., BSSL, lactoferrin, macrophage-receptor-1, CD14, complement-c3/c4, xanthine dehydrogenase) than HoP.

Keywords: breastmilk, non-thermal processing, high hydrostatic pressure, low-temperature long-time pasteurization, dynamic in vitro digestion, bioactive proteins

1. Introduction

Maternal milk is the ideal nourishment for all newborns, offering an array of nutrients essential for supporting infant growth, neurodevelopment, and long-term health (Victora et al., 2016). Beyond nutrition, maternal milk contains numerous unique bioactive components, growth factors, and hormones, many of which are proteins, that function to enhance: (1) immune development (e.g., bacteriostasis, anti-bacterial/anti-viral activity, protection against atopic diseases (Quitadamo et al., 2021); (2) digestion (via provision of exogenous digestive enzymes) (Dallas & German, 2017); (3) gastrointestinal (GI) health (Donovan, 2006); and (4) anti-inflammatory effects (Goldman et al., 1990).

Though critical for all infants, preterm infants greatly benefit from the bioactive components provided by their parent’s milk, as they are inherently at a higher risk of health complications, including necrotizing enterocolitis, infection, sepsis, and poor growth (Underwood, 2013). Whenever parent’s milk is unavailable in sufficient volume, it is recommended that pasteurized donor human milk be used as a supplement instead of preterm infant formula, to protect against developing necrotizing enterocolitis (O’Connor et al., 2016; Quigley et al., 2019). Holder pasteurization (HoP) (62.5 °C, 30 min) is typically used to pasteurize donor milk to provide a 5-log reduction in potential pathogens known to be found in human milk (Human Milk Banking Association of North America, 2020). However, compared to raw, HoP negatively impacts heat-sensitive nutrients, and functional bioactive components including proteins inherently susceptible to denaturation (Peila et al., 2016). Functional components likely denatured during processing are predominantly proteins including enzymes (e.g., proteases, lipases), enzyme inhibitors (e.g., anti-trypsin, anti-chymotrypsin) and antimicrobial proteins (e.g., lactoferrin, lysozyme).

In a recent scoping review of the literature, we reported that HoP alters the kinetics of protein digestion during in vitro experiments, likely due in part to protein unfolding and aggregation, which increases steric accessibility of proteolytic sites for digestive proteases to function (Pitino, Beggs, et al., 2023). Moreover, both in vivo and in vitro digestion studies provide compelling evidence that HoP of human milk may result in less undigested lactoferrin and may increase the digestion of caseins. In a crossover study where infants were fed with raw or HoP maternal milk, heat treatment resulted in poorer fat absorption and linear growth (Andersson et al., 2007). Similarly, in a preterm piglet study, feeding HoP donor milk versus raw donor milk resulted in significantly less weight gain over an 8 day period (Li et al., 2017). Even when compared to formula, feeding pasteurized donor milk to preterm infants reduces growth (linear, head circumference, weight gain) (Quigley et al., 2019). Further, we and others, speculate that the impact of heat on milk constituents may be responsible for observations that, unlike maternal milk, there is no advantage of supplemental pasteurized donor milk over preterm formula for preventing late-onset sepsis (Schanler et al., 2005) or improving neurodevelopment (O’Connor et al., 2016).

To overcome the limitations of HoP, alternative pasteurization technologies are being investigated that minimize changes to donor milk proteins, while providing adequate inactivation of potential pathogens (e.g., bacteria, viruses) that could harm a vulnerable preterm infant. High-pressure processing (HPP) is one promising non-thermal technology which minimally alters the composition of human milk compared to raw milk overall, and has been shown to better preserve the concentration of heat-sensitive nutrients and bioactive components compared to HoP (e.g., lactoferrin, bile salt-stimulated lipase [BSSL], lipoprotein lipase, alkaline phosphatase, insulin, adiponectin, leptin, vitamin B6, vitamin B12, vitamin C, folate, immunoglobulin [Ig]M, IgG, volatile profile), while inactivating bacteria and viruses (Pitino et al., 2019, 2022; Wesolowska et al., 2019). Some studies also report increases in certain components following HPP treatment of human milk compared to raw, including lysozyme activity, and interleukins (IL-6, IL-8, Il-13) (Wesolowska et al., 2019). Although HPP appears to be superior to HoP in terms of retaining milk components, some discordant literature suggests HPP may still alter the composition of milk in some way compared to raw milk, including reductions in lactoferrin (3–48%), IgA (up to 42%), insulin (5–18%), vitamin C (<5%) and tocopherols (reduced); however, these results originate from very few studies and reductions have been primarily associated with pressures at or above 600 MPa (Wesolowska et al., 2019).

Beyond changes in individual components, the impact of HPP on human milk protein digestion is not well-defined. Our recently published paper which examined proteins by SDS-PAGE, as well as lysozyme activity, demonstrated that lactoferrin, milk fat globule membrane associated proteins, and lysozyme activity better resisted simulated preterm infant digestion following HPP than following HoP (Pitino, Unger, et al., 2023). Given that the proteins we could identify by SDS-PAGE only represent a very small fraction of all previously identified proteins in human milk, a thorough proteomic investigation was warranted to characterize the non-digested human milk proteins with greater granularity. Moreover, it is important to characterize the impact of HPP on protein digestion given that in bovine milk, pressurization is known to reversibly impact the size distribution of casein micelles, and that changes in the colloidal properties may affect digestion of caseins and other proteins with which they interact (Mackie & Macierzanka, 2010; Sergius-Ronot et al., 2022). To date, very little is known about the extent to which processing (HPP or HoP) impacts the hundreds of human milk proteins and whether they resist or become more susceptible to digestion. These include numerous immune-related proteins which can help protect against infection or develop the infant’s immune system (e.g., Igs, complement factors, immunomodulatory factors/cytokines), (2) biologically active enzymes (e.g., BSSL, lipoprotein lipase, fatty acid synthase) which are thought to enhance nutrient digestion, (3) micronutrient binding proteins (e.g., vitamin D-, vitamin B12-, folate- binding proteins) which can enhance nutrient absorption, and (4) specific casein species (e.g., αS1, β, κ) important for curd formation and mineral absorption (Lönnerdal, 2016).

Thus, leveraging a dynamic in vitro digestion system to simulate the preterm infant, the objective of this study was to compare the impact of HPP and HoP on the donor milk proteome (including all proteins of human origin), to determine the abundance of undigested donor milk proteins throughout digestion, and to identify proteins digested differently compared to raw donor milk. We hypothesized that milk treated by HPP would be no different than raw milk in protein composition and would be digested no differently compared to raw milk, yet different compared to HoP processed milk. We also hypothesized that HoP would result in increased proteolysis of key bioactive proteins, yielding lower abundances of undigested proteins.

2. Materials and methods

2.1. Overall study design

The study is part of a larger project investigating the impact of donor milk processing on simulated preterm infant digestion and absorption (Pitino, Unger, et al., 2023). Briefly, unprocessed (raw) and pasteurized (HPP or HoP) pools of donor milk were subjected to in vitro digestion for 180 min using a dynamic model programmed using physiological parameters approximating the conditions of the GI tract of preterm infants. The overall design of this study, and an overview of the sample preparation steps prior to proteomic analysis are summarized in Fig.1.

Fig. 1.

Fig. 1.

Open in a new tab

Overall study design and analysis approach. TIM-1, TNO gastrointestinal model.

2.2. Collection and processing of milk samples

Collection and processing of donor milk samples was conducted as previously described (Pitino, Unger, et al., 2023). Briefly, frozen raw milk from 9 donors that did not qualify for processing due to high bacterial load (>5 × 107 CFU/L) was collected from the Rogers Hixon Ontario Human Milk Bank and immediately placed at −80 °C until processing. After thawing overnight (4 °C), donor milk was combined (n=3 pools) to create pools with milk that had been frozen for a similar amount of time, and milk of comparable lactational stage. Each pool (~1L) containing milk from 3 donors, was first divided in thirds (120 g each); one of the thirds was left untreated, whereas the remaining two thirds were processed by either HPP (500 MPa, 10 min, 4 °C) using commercial-scale equipment (Hiperbaric 55, Burgos, Spain) or by HoP (62.5 °C, 30 min) using a shaker waterbath. The HPP parameters were selected based on their ability to reduce bacteria and viruses in donor milk while preserving components (Pitino et al., 2019, 2022). All milk samples were kept at 4 °C until processing (HPP and HoP), which was completed within a 24-h period. During HPP treatment, the increase in temperature due to adiabatic heating was estimated to be approximately 15 °C, the come-up time for pressurization was 2.2 min and depressurization was instantaneous (3 sec). The HoP process ramp-up time to 62.5 °C and ramp-down time to 4 °C were 12 min and 17 min, respectively (total process time = 59 min). Processed milk was frozen at −80 °C until in vitro digestion.

2.3. Dynamic in vitro digestion

The TNO Gastrointestinal Model (TIM-1) (TNO, The Hague, Netherlands) was used to simulate the preterm infant GI tract as described previously (Pitino, Unger, et al., 2023). This dynamic in vitro model contains 4 separate compartments (simulated stomach, duodenum, jejunum, and ileum) and the rate of stomach acidification, intestinal pH, and flow of simulated gastric/intestinal fluid and digestive enzymes are computer-controlled. Parameters for this study were tailored to the preterm infant using literature estimates (De Oliveira et al., 2016), with adaptation specifically for the TIM-1 system (Pitino, Unger, et al., 2023). Samples analyzed in this study were collected from raw and processed milk (meal) and across digestion in the stomach (30 min), duodenum (45 min), jejunum (60 min, 180 min) and ileum (180 min) (Fig. 1). The antiproteases pepstatin-A (1 μM) (Millipore Sigma, Oakville, Canada) and Pefabloc© (Roche, Mannheim, Germany) (5 mM) were immediately added to gastric and intestinal samples respectively, to arrest proteolysis prior to storage at −80 °C until analysis.

2.4. Analysis

2.4.1. Undigested protein extraction and preparation of tryptic peptides

Undigested proteins from raw donor milk and digesta were extracted using a previously published protocol, with some modification (Kim et al., 2023). Briefly, samples were first thawed on ice, gently mixed by inversion, and defatted by centrifugation and isolation of the infranatant (4,000 × g, 30 min, 4 °C). The undigested soluble proteins were precipitated by adding 250 μL of cold (−20 °C) high performance liquid chromatography (HPLC)-grade ethanol (111HPLC200, PharmCO, Brookfield, CT, United States) to each 50 μL sample. After mixing via vortex, suspensions were kept at −20 °C for 1 h and centrifuged (12,000 × g, 20 min, 4 °C) to precipitate proteins. After the protein pellet was resuspended in 100 μL of 50 mM ammonium bicarbonate buffer (Fisher Scientific, A643–500, Fairlawn, NJ, United States) in deionized water (ultrapure H2O), the crude protein concentration was determined by the bicinchoninic acid assay (Pierce BCA Protein Assay, Thermofisher, A53225, Rockford, IL, United States) and the value later used to determine the amount of Trypsin/Lys-C required for the preparation of tryptic peptides. Two microliters of 550 mM dithiothreitol (VWR 0281–25G, Solon, OH, United States) was added to the protein solution for 50 min at 50 °C, followed by the addition of 4 μL of 450 mM iodoacetamide (Sigma Aldrich, I11495G, St. Louis, MO, United States) incubated for 1 h in the dark (room temperature) to facilitate alkylation of thiol groups. Trypsin/Lys-C (Promega, V5072, Madison, WI, United States) (1 μg/μL) was then added to each sample to achieve a 1:50 w/w protein ratio and incubated at 37 °C for 16 h under constant stirring.

2.4.2. Purification of tryptic peptides generated from undigested proteins

Tryptic peptides from the previously undigested proteins were purified by solid-phase extraction using C18 cartridges (5 mL tube volume, 500 mg bed weight, 45 μm particle size, Sigma Aldrich Supelco Discovery SPE, 52604-U, St. Louis, MO, United States) as previously described (Kim et al., 2023) using 80% v/v acetonitrile (Sigma Aldrich, 34998-4L, Burlington, MA, United States) and 0.1 % v/v trifluoracetic acid (Millipore TX1257-1, Billerica, MA, United States) in ultrapure H2O as the elution buffer. Tryptic peptides were eluted and dried by vacuum centrifugation (Genevac, EZ 2.3 Plus, Stone Ridge, NY, United States) and stored at −80 °C until liquid chromatography-tandem mass spectrometry (LC-MS/MS) analysis.

2.4.3. LC-MS/MS and bioinformatic analyses

Immediately prior to analysis, dried tryptic peptides were reconstituted in 3% v/v acetonitrile and 0.1% v/v formic acid in ultrapure H2O (100 μL) solution, diluted 25-fold further in the same buffer, and analyzed via an Orbitrap Fusion™ Lumos™ Tribrid™ mass spectrometer (Thermo Scientific, Waltham, MA, United States) connected to a Waters Nano Acquity Ultra HPLC (Waters Corporation, Milford, MA, United States), as described with some modification (Kim et al., 2023). One microliter of sample was injected onto a nanoAcquity UPLC C18 trap column (180 μm × 20 nm, 5 μm bead, Waters Corporation) for enrichment and desalting, and separated on a 100 μm × 100 mm, 1.7 μm bead Acquity UPLC Peptide BEH C18 column (Waters Corporation) over 60 min. A flowrate of 0.5 μL/min was used to separate peptides using gradient elution of the mobile phase A (0.1% v/v formic acid [Millipore Sigma, F0507, St. Louis, MO, United States] in ultrapure H2O) and acetonitrile with 0.1% v/v formic acid (mobile phase B). The solvent profile was 3% B for the first 3 min, 3–15% B from 3–8 min, 15–30% B from 8–40 min, 30–95% B from 40–50 min, and held at 95% B for 5 min. Re-equilibration of the column was conducted using 3% B for 5 min. Tryptic peptides were ionized by electrospray (2,350 V, 300 °C) and MS spectra over an m/z range of 300–2000 were acquired in positive ionization mode with the Orbitrap of a resolution of 60,000. In the following data-dependent MS/MS scan, precursor ions were automatically selected (ion-intensity threshold of 5.0 × 104, charge state of 2–8 and exclusion duration of 60 s and mass tolerance of 10 ppm) and fragmented by higher-energy collision-induced dissociation activation with 30% of collision energy. MS/MS spectra were acquired in the positive ionization mode with the Orbitrap at a resolution of 30,000.

To determine the identity of the extracted undigested proteins from which tryptic peptides were derived, matching to a Sequest HT search of the human proteome database (Uniprot, n=20,307) using Proteome Discoverer (version 3.0) was carried out. Cleavage sites were set to C-terminal arginine and lysine (full specific digestion specificity with a maximum of two missed cleavages). Tolerance of the precursor mass and fragment mass were set to 10 ppm and 0.1 Da, respectively. Potential modifications allowed included phosphorylation of serine and threonine, and oxidation of methionine. Static modification included carbamidomethylation of cysteine. Only proteins identified with high confidence (false discovery rate, P < 0.01) were included. Integration of each eluted peak (ion intensity) was used to calculate the raw abundance values.

2.5. Data processing and statistical analysis

Protein abundance data were first adjusted for dilutions that occurred within the TIM-1 system which were 1.19, 2.71, 3.03, 4.48, and 3.75 in the simulated stomach (30 min), duodenum (45 min), jejunum (60 min), jejunum (180 min) and ileum (180 min), respectively. Individual protein row z-scores were calculated across digestion for each treatment (i.e., raw, HPP, HoP), representing the average protein abundance across all pools and scaled to the row average. A heat map was then generated to visualize the changes in each individual protein, during digestion; proteins not identified in any of the raw milk pools were excluded from this analysis. To highlight changes in the most highly abundant proteins in human milk, unstacked and stacked bar charts were used to qualitatively describe changes in protein composition of the top 10 proteins, reported as raw abundance and normalized abundance. Normalized abundance is the ratio of each detected protein abundance to the sum total of all protein abundances, expressed as a percentage.

Volcano plots were generated to assess relevant changes in the abundance of undigested proteins and comparisons were made between each treatment (HPP/HoP) and raw, as well as HoP versus HPP. The calculated log-fold change in protein abundance was plotted against the - log(p-value), calculated using a mixed models approach with pairwise comparisons (LS-MEANS) as described below. Only proteins present in all raw samples were included in this analysis. Meaningful changes in undigested proteins were defined as having a large fold change (≤ −2 or ≥ −2) and were statistically different compared to raw milk (P-value < 0.05).

In a subgroup analysis, changes in the abundance of some key undigested bioactive proteins were determined across digestion and compared among all treatment groups. These included (1) human milk proteins previously identified to better resist digestion in HPP-treated samples compared to HoP (e.g., lactoferrin, lysozyme, and milk fat globule membrane-associated proteins [macrophage mannose receptor-1, xanthine dehydrogenase] (Pitino, Unger, et al., 2023); (2) digestive enzymes (e.g., BSSL) shown to be better preserved by HPP (Pitino et al., 2019)); and (3) Ig. The abundance of Ig for the primary isotypes are shown (IgA/sIgA, IgG and IgM) and were calculated from individual Ig chains. Abundances from known Ig isotype constituents, including Ig α chain (from IgA or sIgA), Ig γ chain (from IgG), and Ig μ chain (from IgM of sIgM) were aggregated. Individual proteins belonging to each Ig isotype are summarized in Supplementary Table 1.

For all analyses, log-transformed mean protein abundances were compared among treatment groups within a given compartment using repeated measures mixed-effects models. If data were not normally distributed (PROC UNIVARIATE), a non-parametric repeated-measures rank-based analysis was conducted. Post hoc pairwise comparisons (LS-MEANS) between groups were reported for a specific digestion time/compartment separately. P values <0.05 were considered significant. All analyses were conducted using SAS Statistical Software (version 9.4; SAS Institute, Cary, NC, United States). Data visualization was conducted in R Statistical Software (R Foundation for Statistical Computing version 4.1.2) and GraphPad Prism (version 10.0.3; Dotmatics, Boston, MA, United States).

3. Results

3.1. Characterization of donor milk pools

Samples of donor milk (n=9) used in this study were frozen at −20 °C for approximately 1.3 ± 0.2 months (mean ± SD) prior to pooling and were predominantly from mature (3.3 ± 1.4 months post-partum), primiparous donors (5/9). The characteristics of individual milk donations prior to pooling are summarized in Supplementary Table 2. The macronutrient composition of the pools was reported previously (Pitino, Unger, et al., 2023), and was typical of donor milk (fat: 4.0 ± 0.3 g/dL; crude protein: 1.1 ± 0.1 g/dL; carbohydrate: 8.1 ± 0.2 g/dL) and unaffected by either HPP or HoP processing.

3.2. Counts of identified undigested proteins and changes in the top 10 most abundant proteins across digestion

On average, 241± 32 proteins were identified in all meals and in all compartments. There were no statistically significant differences in mean protein counts among treatment groups for any given simulated digestion compartment (Supplementary Fig. 1). In all pools, protein counts in the meal ranged from 226 to 295 (261 ± 23), in the stomach from 232 to 298 (270 ± 20), in the duodenum from 228 to 263 (238 ± 12), in the jejunum (60 min) from 212 to 273 (246 ± 19), in the jejunum (180 min) from 173 to 232 (198 ± 28), and in the ileum from 199 to 292 (234 ± 29). To qualitatively characterize how processing and simulated digestion affected the most abundant human milk proteins, the top 10 proteins were identified in raw milk (based on abundance) and across digestion. These included (in order of greatest abundance): (1) α-lactalbumin; (2) β-casein; (3) lactoferrin; (4) serum albumin; (5) αS1-casein; (6) BSSL; (7) osteopontin; (8) κ-casein; (9) polymeric Ig receptor; and (10) Ig heavy chain α−1 (from IgA). Changes in these top 10 proteins are shown across digestion as a function of protein abundance (Fig. 2A) and normalized abundance (Fig. 2B) and illustrate an overall decrease in abundance of all top 10 proteins. More distally in digestion (i.e., jejunum, ileum), α-lactalbumin and serum albumin together constitute approximately >75% of all proteins by abundance, resisting digestion to a certain degree among all treatments compared to other proteins including β-casein, αS1-casein and osteopontin.

Fig. 2. Protein abundance and normalized protein abundances (%) for the top 10 proteins (by abundance).

Fig. 2.

Open in a new tab

Note. (A) Mean protein abundances across digestion for each processing method adjusted for the dilution within the TIM-1 simulated digestion system (N=3 pools); Bars and error bar represent the mean (SD) of protein abundance adjusted for the dilution in the TIM. (B) Normalized protein abundances across digestion for each processing method adjusted for the dilution within the TIM-1 simulated digestion system. All identified proteins were ranked, and the top 10 proteins were selected in terms of abundance. The top 10 proteins represent approximately 93.4% of all identified proteins (on average) in all meal samples. In panel A, each calculated abundance value was adjusted for the dilution in the TIM-1 system. In panel B, each calculated abundance value was normalized by dividing a protein’s abundance by the sum of all protein abundances. The quotient expressed as a percentage represents the percent composition of each protein. HoP, Holder pasteurization; HPP, high-pressure processing.

3.3. Qualitative changes in protein abundance in all identified proteins

Changes (z-scores) in the raw abundances of all unique proteins identified in raw milk by treatment group and across digestion are summarized in a heat map as Fig. 3. Overall, even after accounting for the meal dilution, protein abundance declined throughout digestion among all processing treatments (Fig.3). Qualitatively, the protein abundances in the meal compartment for raw and HPP appeared to be most similar (Fig.3) and clustered together (Supplementary Fig.2), while some proteins in HoP treated milk appeared below average (Fig.3). HPP and HoP-treated milk clustered together in the stomach and duodenum compartments, while in the jejunum at 60 min, raw and HPP milk clustered more closely than HoP milk (Supplementary Fig.2).

Fig. 3. Heat map illustrating changes in undigested protein abundance as row z-scores across digestion among processing methods.

Fig. 3.

Open in a new tab

Note. Each row of the heat map represents an individual protein identified (N=198). Proteins not identified in any of the raw milk pools were excluded from this analysis. Row z-scores were calculated for each protein across digestion among the three processing methods (raw, high-pressure processing and Holder pasteurization) by comparing each value to the row average. The z-score value is depicted in colour (from purple [larger negative Z-score] to yellow [larger positive Z-score]) showing the direction and magnitude. HoP, Holder pasteurization; HPP, high-pressure processing.

3.4. Quantitative changes in protein abundance in all identified proteins

Volcano plots depicting the fold change in protein intensity across digestion compared to raw are illustrated in Fig. 4. A number of minor proteins were significantly lower (P <0.05) in abundance (not present in their undigested form) after HoP and HPP compared with raw milk (Fig. 4). Following HPP, the abundance of several minor proteins including calmodulin-like protein-3, sclerostin-domain containing protein-1, fibrinogen γ/α chain, biotinidase, ATP citrate synthase, were significantly lower compared with raw milk (Fig. 4A). In HoP-treated milk, proteins including tumor necrosis factor receptor, fibronectin, phosphoglucomutase-1 and biotinidase were significantly lower compared to raw (Fig. 4B). Similarly, when compared to HPP (Fig. 4C), the abundance of minor proteins was observed to either be significantly lower (e.g., protein disulfide-isomerase, complement C1q tn related protein) or higher (e.g., CD59 glycoprotein, stomatin) following HoP.

Fig. 4. Volcano plots depicting the fold change in protein intensity across digestion by processing method.

Fig. 4.

Open in a new tab

Note. Each volcano plots represents a comparison of mean protein abundances (adjusted for the TIM dilution) between (1) high-pressure processing and raw treatment groups; (2) Holder pasteurization and raw treatment groups; and (3) high-pressure processing and Holder pasteurization, respectively (N=3 pools in each group). Panels (A), (B) and (C) represent the meal or milk, panels (D), (E), and (F) represent the stomach compartment at 30 min, panels (G) (H), and (I) represent the duodenum compartment at 45 min, panels (J), (K), and (L) represent the jejunum compartment at 60 min, panels (M), (N), and (O) represent the jejunum at 180 min and panels (P), (Q), and (R) represent the ileum compartment at 180 min. The x-axis represents the log2 (fold change) of each identified protein following a processing method (high-pressure processing or Holder pasteurization) with respect to raw; the y-axis represents the negative log of the calculated p-value. P-values were calculated using a mixed models approach with follow-up pairwise comparisons (LS-MEANS). The vertical dotted lines indicate a positive and negative 2-fold change respectively and the horizontal zotted line indicates significance (p<0.05). Each data point represents an individual identified protein while red-filled points denote significance and are annotated where possible. HoP, Holder pasteurization; HPP, high-pressure processing.

Following 30 min of gastric digestion, mean differences emerged in which the abundances of undigested proteins were reduced according to processing. In HPP-treated samples, low-abundance proteins including ceruloplasmin, desmoplankin, tenascin and C4b binding protein-α were significantly reduced compared with raw milk (Fig. 4D). In contrast, major proteins from HoP-treated milk were significantly lower in abundance in the stomach compared to raw, and included lactoferrin, BSSL, macrophage mannose receptor-1, malate dehydrogenase, xanthine dehydrogenase, complement C3/C4, CD14, and mucin (Fig. 4E). Similarly, when compared to HPP in the stomach, major proteins including BSSL, lactoferrin and macrophage mannose receptor-1 as well as some minor proteins, such as malate dehydrogenase, retinol inducible serine carboxypeptidase, and zinc α−2 glycoprotein were significantly lower in HoP milk samples (Fig. 4F).

In HPP-treated milk samples, proteins including angiotensinogen, ceruloplasmin, and CD14 were significantly lower in the duodenum (45 min) compared to raw milk (Fig. 4G). In HoP-milk, proteins including lactoferrin, macrophage mannose receptor-1 and CD14 remained significantly lower compared with raw in the duodenum at 45 min, as well as other proteins such as β−2 glycoprotein, superoxide dismutase, syntenin-1, leucine-rich glycoprotein (Fig. 4H). Macrophage mannose receptor-1 remained significantly lower in HoP treated milk compared to HPP in the duodenum at 45 min, in addition to β−2 glycoprotein (Fig. 4I).

Though very few proteins were found to be significantly lower in treated milk samples in the jejunum compartment, at the end of simulated digestion (ileum, 180 min), numerous proteins were significantly reduced compared to raw milk in HPP-treated samples (Fig. 4M). These proteins include lactoferrin, glutathione peroxidase, BSSL, CD14, cathepsin D and numerous Ig chains. Similarly, in HoP-treated milk samples in the jejunum at 180 min, the abundance of undigested lactoferrin, BSSL, lactoperoxidase, and folate receptor-α was significantly lower compared to raw (Fig. 4N). There was also significant overlap in proteins identified reduced in HPP milk compared to raw and those proteins reduced in HPP milk compared to HoP (Fig. 4O). In the ileum at 180 min, less abundant proteins appeared to be reduced in HPP treated milk (Fig. 4P) (e.g., α-enolase, profilin 1, leucine-rich glycoprotein) compared with those reduced in HoP-treated milk (Fig. 4Q) (e.g., BSSL, lactoferrin, transcobalamin, vitamin D binding protein, polymeric Ig receptor, macrophage mannose receptor-1.) Lactoferrin, polymeric Ig receptor, macrophage mannose receptor-1 and BSSL were lower in HoP-treated milk in the ileum at 180 min compared to HPP (Fig. 4R).

For reference, the abundance values, and normalized abundance values of the top 40 identified proteins which constitute approximately > 99.9% of all identified undigested proteins (by abundance) are summarized in Supplementary Table 3A/B and Supplementary Table 4A/B respectively.

3.5. Abundances of select bioactive proteins across digestion characterized by processing method

Changes in the abundance of select bioactive proteins that have been previously shown to be retained following HPP and/or have been previously shown to resist digestion (Pitino, Unger, et al., 2023) are shown in Fig. 5. In the meal, no significant differences were observed in the abundance of any of these bioactive proteins including lactoferrin, BSSL, lysozyme, IgA/G/M, macrophage mannose receptor-1, and xanthine dehydrogenase following either HPP or HoP compared to raw milk. In the stomach, raw and HPP-treated samples had a significantly higher abundance of lactoferrin (Fig. 5A), BSSL (Fig. 5B), and macrophage mannose receptor-1 (Fig. 5E) compared to HoP. The abundance of undigested xanthine dehydrogenase was significantly higher in raw milk compared to HoP, but not different compared to HPP (Fig.5F). Similarly in the duodenum, undigested lactoferrin in raw milk was significantly higher than HoP, but not different from HPP (Fig.5A). Undigested macrophage mannose receptor-1 was significantly higher in raw and HPP milk compared with HoP (Fig. 5E). In the jejunum at 180 min, both HPP and HoP-treated samples had significantly lower abundances of undigested lactoferrin and BSSL than raw (Fig. 5A, Fig. 5B). Raw and HoP had significantly higher undigested xanthine dehydrogenase abundance compared with HPP, and the abundance of IgA- and IgM-related protein chains in HPP-treated milk were significantly lower than in raw or HoP-treated milk (Fig. 5D). In the ileum, raw and HPP-treated milk had significantly higher undigested lactoferrin (Fig. 5A) and BSSL (Fig. 5B) than HoP, and lysozyme abundance was significantly lower in HoP versus raw, but not HPP (Fig.5C).

Fig. 5. Changes to protein abundances of select key immunological and bioactive proteins across digestion for each treatment group.

Fig. 5.

Open in a new tab

Note. Bars and error bar represent the mean (SD) of protein abundances (N=3 pools) adjusted for the dilution in the TIM-1 system. A) Lactoferrin; (B) Bile salt-stimulated lipase; (C) Lysozyme; (D) Immunoglobulin Abundance; (E) Macrophage Mannose Receptor-1; (F) Xanthine dehydrogenase. The abundance of immunoglobulins for the primary isotypes are shown (IgA/sIgA, IgG and IgM) and were calculated from individual immunoglobulin chains. Abundances from known immunoglobulin isotype constituents, including Ig α chain (from IgA or sIgA), Ig γ chain (from IgG), and Ig μ chain (from IgM of sIgM) were compiled. Log-transformed mean protein abundances were compared among treatment groups within a given compartment using repeated measures mixed-effects models. Post hoc pairwise comparisons (LS-MEANS) between groups were reported for a specific digestion time/compartment separately. P values <0.05 were considered significant (* p<0.05; ** p<0.01; *** p<0.0001). HoP, Holder pasteurization; HPP, high-pressure processing.

4. Discussion

Improving the quality of donor human milk while ensuring safety is critical for improving health outcomes of vulnerable preterm infants who rely on its beneficial components to support growth, neurodevelopment, and GI health (Donovan, 2006; Underwood, 2013). While infants fed HoP donor milk benefit from a decreased risk of developing necrotizing enterocolitis (Quigley et al., 2019), donor milk-fed infants have difficulty growing, and do not receive the complete immunomodulatory benefits of milk afforded to infants fed their own parent’s milk (Neu, 2019). Although HPP has been identified as a candidate replacement technology that better preserves bioactive proteins than HoP; there is a paucity of research on which proteins remain present after processing and which remain undigested throughout digestion, potentially impacting the bioactivity of donor milk.

4.1. Effect of processing on identified proteins in donor human milk

Overall in the meal, none of the top 40 most abundant proteins were significantly lowered by either HPP or HoP compared with raw milk (Supplementary Table 3A). It was originally hypothesized that HoP would reduce the abundance of proteins present compared to raw or HPP milk; however, previous studies that have reported changes in milk protein concentration post HoP typically use analytical methods that are impacted by protein structure (e.g., enzyme-linked immunosorbent assay, HPLC) (Peila et al., 2016), unlike the proteomics technique which can still detect proteins if their tertiary structure is altered in some way. In spite of this, some less abundant proteins were significantly lower in abundance in HPP and HoP. Biotinidase (significantly lower in both HPP and HoP compared to raw) may play a role in the intestinal absorption of biotin during infancy (Oizumi et al., 1989). Fibronectin, which is significantly lower in HoP but not HPP milk, is essential for osteoblast mineralization. In contrast, fibrinogen, significantly lower in HPP, but not HoP milk, may be related to coagulant activity in milk and to improvements in GI outcomes (Hu et al., 2022). In a previous study, milk clotting time was used as a functional marker of human milk bioactivity, and the results indicated that milk processed by HPP (500 MPa, 5 min) better preserved this activity compared to HoP, high-temperature short-time and thermoultrasonication (Hu et al., 2022). Though clotting activity was not measured directly, losses in fibrinogen content post-HPP could be related to the additional 5 min of pressurization in this study required to inactivate pathogens. A recent study from Zhang et al. 2023, which investigated the impact of HPP on the human milk proteome, reported that protein profiles of HPP-treated milk more closely resembled raw milk compared to HoP (i.e., fewer down-regulated proteins in whey and casein fractions) (J. Zhang, Duley, Cowley, Shaw, Koorts, et al., 2023). Inconsistent with our findings, Zhang et al. 2023 noted a decrease in BSSL abundance in HoP compared to raw or HPP, though this was likely due to various factors including their protein extraction protocol (e.g., centrifugation at 25,000 × g vs. 12,000 × g, acetone vs. ethanol precipitation, a sonication step with lysis buffer solution), and the fact that our study analyzed three independent pools of previously frozen milk, whereas a single pool of fresh milk was tested in the aforementioned study (Zhang et al., 2023).

4.2. The impact of processing on undigested proteins throughout simulated preterm infant digestion

This study is the first to comprehensively characterize the fate of undigested proteins in human milk treated by HPP throughout simulated preterm infant digestion using a dynamic in vitro system. Based on preliminary protein gel analyses showing similar protein band intensities among raw and HPP milk throughout digestion, our original hypothesis was that the extent of proteins digestion from milk treated by HPP would not be significantly different from raw milk compared to HoP-treated milk (Pitino, Unger, et al., 2023). To a certain extent, this study confirms our previous findings and our hypothesis, and provides greater granularity as to specific proteins which resist digestion, and in which simulated compartment.

Interestingly, after adjustment for the meal dilution in the TIM-1, we also reported that the abundance of certain proteins including BSSL, lactoferrin, macrophage mannose receptor-1, CD14, complement C3/C4, malate dehydrogenase, xanthine dehydrogenase, and some Ig chains in HoP-treated donor milk were significantly lower upon gastric digestion compared to raw milk, yet these same proteins were not different from HPP milk. It is likely that varying degrees of heat-induced denaturation and unfolding of whey proteins during HoP, but not HPP, enhanced the accessibility of certain peptide bonds for proteases (i.e., pepsin) to act upon (Mackie & Macierzanka, 2010; Pitino, Beggs, et al., 2023). Although there were no significant differences in the abundance of caseins across digestion by treatment, the abundance of β-casein, αS1-casein, and κ-casein consistently trended lower in HoP and HPP compared to raw milk. Though caseins are not denatured by HoP, heat denatured whey proteins can aggregate with colloidal casein micelles (via κ-casein), which could impact their susceptibility to digestion by proteases (Anema, 2021). Studies have also reported a smaller micellar size and dissolution of colloidal calcium phosphate following HPP of human milk at 400–600 MPa (Sergius-Ronot et al., 2022). Though the full impact of HPP on casein micelles remains unclear, changes to their structures may similarly alter their susceptibility to digestion by proteases (Mackie & Macierzanka, 2010).

A notable observation in this study is that while differences in the milk proteome are minimal after processing in the milk itself, large differences in the abundances of major whey proteins emerge upon gastric digestion. Importantly, this trend does not appear for HPP-treated milk, and the above-mentioned proteins reduced in the stomach with HoP-treated milk, remain undigested—no different than raw milk. Taken together with the fact that retention of components in more proximal compartments reflect the composition of influx to the adjacent compartment (e.g., proteins resisting digestion measured in the stomach continually flow to the duodenum), it is likely these proteins, among others, are present within the duodenum and may be functional. The presence of functional proteins within the proximal small intestine is important given that it is the site of primary nutrient digestion and absorption. As BSSL is present in HPP-treated milk, and remains undigested in the gastric compartment, increased fat digestion in vivo with HPP-treated milk is plausible and requires further investigation. Undigested lactoferrin may still possess antibacterial and anti-infective properties, which could be beneficial in the prevention of necrotizing enterocolitis and sepsis, in addition to facilitating absorption of iron and preventing the generation of oxidation products and free radicals by binding to iron. In the one other study that assessed the impact of HPP (400 MPa, 5 min, 20 °C) compared to HoP on in vitro static digestion simulating the term infant, no differences were observed in terms of lactoferrin proteolysis among raw and HPP treatment groups (Zhang et al., 2022). Despite measurements by gel electrophoresis, the findings of Zhang et al. were consistent with our proteomic results. Lower abundances of complement proteins C3/C4, CD14, and macrophage mannose receptor-1 after gastric digestion in HoP-treated milk may play a role in poorer infection resistance, development of atopic dermatitis and impairments in host defense, respectively (Fikri et al., 2019; Ogundele, 2000). In gastric digesta from HoP milk, a lower abundance of xanthine dehydrogenase, a potent nitric oxide-producing enzyme, could potentially reduce overall antibacterial activity within the GI tract compared to raw or HPP. This reduction could potentially explain, in part, why donor milk does not protect against sepsis as well as parent’s milk.

Although the release of endogenous peptides was not directly measured in this study, the interpretation of our HoP results is consistent with previously published studies, under the assumption that decreased abundance of undigested proteins is inversely correlated with the abundance of peptides. In two separate studies investigating the impact of HoP in a dynamic in vitro digestion model simulating both term and preterm infants, pasteurization induced different kinetics of peptide release, especially for heat-denatured proteins such as BSSL and lactoferrin (Deglaire et al., 2016, 2019). Furthermore, HoP-treatment of donor milk was associated with a higher intestinal digestion of lactoferrin, and an expedited release of peptides in raw milk, regardless of the type of digestion modelled and system used (Giribaldi et al., 2022; Inglingstad et al., 2010). Our results suggest that processing impacts the timing (GI compartment) of when proteins remain undigested, which could result in potential health implications for preterm infants. This finding begs the question of which proteins should remain undigested, and in what proportion, to elicit the greatest benefit to the infant. This fact is neither known for the term infant under normal physiological conditions, nor for the preterm infant with impaired digestive capacity for proteins. It is typical for studies to measure the presence of undigested proteins in the stool, given the challenges and limitations of collecting samples from various points in the GI tract (e.g., stomach, duodenum, jejunum, ileum, colon, etc.), limiting our understanding of optimal protein digestion and its dynamics. It is likely that protein digestion should be optimally balanced between proteins resisting digestion to perform a specific function within the GI tract (e.g., immune system modulation, contributing enzyme activity, reducing inflammation etc.), proteins digested to release bioactive peptides, and protein catabolism to supply free amino acids to the infant to support protein synthesis, and growth. Several human milk-derived peptides (e.g., lactoferricin from lactoferrin) released during digestion may also contribute to additional bioactivity (e.g., preventing infection, stimulating the immune system) (Beverly et al., 2019; Lönnerdal, 2016). Currently, this balance is not well-understood and should be further investigated using innovative techniques to track protein digestion and peptide release within each segment of the GI tract. Future work should also focus on analysis of individual free amino acids to quantify metrics of milk protein digestibility (e.g., DIAAS) in the preterm infant and the subsequent impact of processing on these metrics.

4.3. Lower abundance of undigested immunoglobulins in HPP milk at the end of digestion

One notable finding was the increased digestion of Ig (IgA/sIgA and IgG) in HPP-treated milk compared to raw milk or HoP (Fig. 5D). This difference was only observed at the end of digestion in the jejunum (180 min) for IgA/sIgA and IgG, but not IgM. Ig (primarily IgA/sIgA) are commonly detected undigested in infant stool, including preterm infants fed maternal milk (Granger et al., 2022), though the proportion of Ig from milk remaining undigested through the GI is not fully understood, in particular, the proportion of Ig surviving the upper GI tract. A recently published human study comparing gastric digestion of Ig from preterm infants fed raw maternal milk demonstrated that total IgA (sIgA/IgA) concentration was reduced by 76% after 3 hours (Demers-Mathieu, Underwood, Beverly, & Dallas, 2018). In contrast, a study investigating the gastric and intestinal survival of palivizumab added to human milk fed to infants reported reductions of Ig stability (by ELISA) ranging from 26.5% to 62.2% for gastric and intestinal survival respectively (Lueangsakulthai et al., 2021). To a certain degree, Ig (all isotypes) are digested throughout simulated digestion, given that their total abundances decline from the meal though to the ileum; however, the clinical significance of increased IgA/IgG digestion in HPP-treated milk at the end of digestion is unclear, and possibly confounded by the fact that digestion of Ig may still yield functional antigen-binding fragments. Delineating this balance requires further investigation.

4.4. Strengths and limitations

This study has many strengths. First, the paired design and the use of three unique pools of milk allowed greater statistical power to detect differences among the treatments. We selected to use this pooling strategy as it best matches current practices at donor milk banks which increases the translatability of our findings. This study was also pragmatic in that the HPP parameters selected were previously shown to preserve nutrients and bioactive components while inactivating bacteria and viruses (Pitino et al., 2019, 2022). This study also leveraged a dynamic model of the GI tract which is thought to be superior to traditional static in vitro models in its ability to mimic the dynamics of digestion and gut physiology (e.g., gastric/intestinal emptying, pH, flow, and concentration of enzymes) (Dupont et al., 2019).

This study also has several limitations. The high variability in protein abundance among our three pools may have minimized our ability to detect significant differences in undigested proteins. Despite this limitation, this study had sufficient power to detect significant differences in several proteins. Also, analysis of three pools may not completely represent the typical variability that exists in pooled milk bank samples (e.g., protein abundances, protease activities). Therefore, digestion of several additional pools of milk in future studies would result in improved generalizability of findings and would likely assist in detecting differences in additional proteins. Moreover, our study design did not enable disambiguation of the relative contributions of milk proteases and gastrointestinal proteases to milk protein digestion. Future studies should investigate this question further as different processing may impact protease activity. This study also leveraged the use of milk that did not qualify for processing due to high bacterial load (>5 × 107 CFU/L). As bacteria present in milk could produce proteases that could contribute to protein digestion within milk itself and across digestion, it is possible our findings reflect higher overall protein digestion across all treatment groups than is representative of milk with lower bacterial loads. Future studies should investigate the contributions of proteases produced by bacteria in milk on protein digestion. Moreover, while the in vitro digestion model used best approximates the GI of a preterm infant, it lacks a biologically active intestinal epithelium that would further our understanding as to how components in the meal could interact with the epithelium physiologically, enhancing signalling and exerting bioactive properties (e.g., antibacterial, improving gut barrier function etc.), given that many of the identified undigested bioactive proteins in milk act locally in the gut. Though this interface of digesta and the GI epithelium cannot be replicated with this model, future investigations using in vitro gut enteroid (i.e., intestinal organoid) models could assist in elucidating the functional impact to underlying physiology.

Finally, although this study was able to demonstrate that certain proteins in HPP milk remain more abundantly undigested in proximal digestive compartments, we cannot discount the possibility these proteins may not function as efficiently as native proteins. Functionality cannot be ascertained by proteomic analyses and must be determined via follow-up functional/enzymatic assays, or via animal and human clinical studies where outcomes secondary to digestion can be measured including GI health, gut barrier function, bacterial translocation, weight gain, growth, and fat absorption. To date, no study has assessed the impact of HPP-treated milk in animal or human trials. This gap must be addressed in future research required to determine the clinical relevance of these findings in terms of health outcomes in preterm infants.

5. Conclusion

Overall, HPP is a promising alternative to HoP for processing donor milk and should be assessed clinically. Future research should consider all aspects of donor milk processing holistically, including HPP and other non-thermal processing technologies, to improve both the quality of milk and health outcomes in recipient infants.

Supplementary Material

1

Highlights.

  • Study assessing how human milk pasteurization impacts protein digestion in vitro.

  • Pasteurization minimally changes proteins (count, abundance) pre-digestion.

  • Increased proteolysis in Holder vs. high-pressure milk during gastric phase.

  • Simulated digestion of raw and high-pressure milk yields similar profiles.

  • Key proteins in high-pressure milk better resist proximal digestion vs. Holder.

Acknowledgements

This research has been financially supported by the ISRHML— Family Larsson-Rosenquist Foundation Trainee Expansion Program (TEP) to MAP, the Hospital for Sick Children Restracomp Scholarship to MAP, the Canadian Institutes of Health Research to DLO, SU, YP (CIHR FDN#: 143233) and the National Institutes of Health to DCD (R01HD106140). We acknowledge the Mass Spectrometry Center at Oregon State University supported in part by the National Institute of Health grant (NIH#: 1S10OD020111-01).

List of Abbreviations

BSSL

Bile salt-stimulated lipase

GI

Gastrointestinal

HoP

Holder pasteurization

HPLC

High-performance liquid chromatography

HPP

High-pressure processing

Ig

Immunoglobulin

Il

Interleukin

MS

Mass spectrometry

TIM-1

TNO Gastrointestinal model-1

Footnotes

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

CRediT authorship contribution statement.

Michael A. Pitino: Conceptualization, project administration, methodology, formal analysis, investigation, writing- original draft, visualization, funding acquisition. Deborah L. O’Connor: Conceptualization, methodology, writing- review and editing, funding acquisition. Sharon Unger: Conceptualization, methodology, writing- review and editing, funding acquisition. Bum Jin Kim: Methodology, investigation, writing- review and editing. Alain Doyen: Conceptualization, Methodology, Writing- review and editing. Md Abdul Wazed: Validation, methodology, investigation, writing- review and editing. Sudarshan Kumar: Validation, methodology, writing- review and editing. Yves Pouliot: Conceptualization, methodology, writing- review and editing, funding acquisition. Debbie Stone: Resources, writing- review and editing. David C. Dallas: Supervision, methodology, resources, writing- review & editing.

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

References

  1. Andersson Y, Sävman K, Bläckberg L, & Hernell O (2007). Pasteurization of mother’s own milk reduces fat absorption and growth in preterm infants. Acta Paediatrica, 96(10), 1445–1449. 10.1111/j.1651-2227.2007.00450.x [DOI] [PubMed] [Google Scholar]
  2. Anema SG (2021). Heat-induced changes in caseins and casein micelles, including interactions with denatured whey proteins. International Dairy Journal, 122, 105136. 10.1016/j.idairyj.2021.105136 [DOI] [Google Scholar]
  3. Beverly RL, Underwood MA, & Dallas DC (2019). Peptidomics Analysis of Milk Protein-Derived Peptides Released over Time in the Preterm Infant Stomach. Journal of Proteome Research, 18(3), 912–922. 10.1021/acs.jproteome.8b00604 [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Dallas DC, & German JB (2017). Enzymes in Human Milk (pp. 129–136). 10.1159/000455250 [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. De Oliveira SC, Bourlieu C, Ménard O, Bellanger A, Henry G, Rousseau F, Dirson E, Carrière F, Dupont D, & Deglaire A (2016). Impact of pasteurization of human milk on preterm newborn in vitro digestion: Gastrointestinal disintegration, lipolysis and proteolysis. Food Chemistry, 211, 171–179. 10.1016/j.foodchem.2016.05.028 [DOI] [PubMed] [Google Scholar]
  6. Deglaire A, De Oliveira SC, Jardin J, Briard-Bion V, Emily M, Ménard O, Bourlieu C, & Dupont D (2016). Impact of human milk pasteurization on the kinetics of peptide release during in vitro dynamic term newborn digestion. Electrophoresis, 37(13), 1839–1850. 10.1002/elps.201500573 [DOI] [PubMed] [Google Scholar]
  7. Deglaire A, De Oliveira S, Jardin J, Briard-Bion, Kroell F, Emily M, Ménard O, Bourlieu C, & Dupont D (2019). Impact of human milk pasteurization on the kinetics of peptide release during in vitro dynamic digestion at the preterm newborn stage. Food Chemistry, 281, 294–303. 10.1016/j.foodchem.2018.12.086 [DOI] [PubMed] [Google Scholar]
  8. Demers-Mathieu V, Underwood MA, Beverly RL, & Dallas DC (2018). Survival of Immunoglobulins from Human Milk to Preterm Infant Gastric Samples at 1, 2, and 3 h Postprandial. Neonatology, 114(3), 242–250. 10.1159/000489387 [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Donovan SM (2006). Role of human milk components in gastrointestinal development: Current knowledge and future NEEDS. The Journal of Pediatrics, 149(5), S49–S61. 10.1016/j.jpeds.2006.06.052 [DOI] [Google Scholar]
  10. Dupont D, Alric M, Blanquet-Diot S, Bornhorst G, Cueva C, Deglaire A, Denis S, Ferrua M, Havenaar R, Lelieveld J, Mackie AR, Marzorati M, Menard O, Minekus M, Miralles B, Recio I, & Van den Abbeele P (2019). Can dynamic in vitro digestion systems mimic the physiological reality? Critical Reviews in Food Science and Nutrition, 59(10), 1546–1562. 10.1080/10408398.2017.1421900 [DOI] [PubMed] [Google Scholar]
  11. Fikri B, Tani Y, Nagai K, Sahara M, Mitsuishi C, Togawa Y, Nakano T, Yamaide F, Ohno H, & Shimojo N (2019). Soluble CD14 in Breast Milk and Its Relation to Atopic Manifestations in Early Infancy. Nutrients, 11(9), 2118. 10.3390/nu11092118 [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Giribaldi M, Nebbia S, Briard-Bion V, Jardin J, Peila C, Coscia A, Dupont D, Cavallarin L, & Deglaire A (2022). Simulated dynamic digestion reveals different peptide releases from human milk processed by means of holder or high temperature-short time pasteurization. Food Chemistry, 369. 10.1016/j.foodchem.2021.130998 [DOI] [PubMed] [Google Scholar]
  13. Goldman AS, Goldblum RM, & Hanson LÅ (1990). Anti-Inflammatory Systems in Human Milk (pp. 69–76). 10.1007/978-1-4613-0553-8_6 [DOI] [PubMed] [Google Scholar]
  14. Granger CL, Lamb CA, Embleton ND, Beck LC, Masi AC, Palmer JM, Stewart CJ, & Berrington JE (2022). Secretory immunoglobulin A in preterm infants: determination of normal values in breast milk and stool. Pediatric Research, 92(4), 979–986. 10.1038/s41390-021-01930-8 [DOI] [PubMed] [Google Scholar]
  15. Hu Y, Kontopodi E, Mank E, van den Akker CHP, van Goudoever JB, Hettinga K, van Elburg RM, Thaler J, & Nieuwland R (2022). Processing methods of donor human milk evaluated by a blood plasma clotting assay. Innovative Food Science & Emerging Technologies, 76, 102938. 10.1016/j.ifset.2022.102938 [DOI] [Google Scholar]
  16. Human Milk Banking Association of North America. (2020). HMBANA Standards for Donor Human Milk Banking: An Overview. https://www.hmbana.org/file_download/inline/95a0362a-c9f4-4f15-b9ab-cf8cf7b7b866
  17. Inglingstad RA, Devold TG, Eriksen EK, Holm H, Jacobsen M, Liland KH, Rukke EO, & Vegarud GE (2010). Comparison of the digestion of caseins and whey proteins in equine, bovine, caprine and human milks by human gastrointestinal enzymes. Dairy Science and Technology, 90(5), 549–563. 10.1051/dst/2010018 [DOI] [Google Scholar]
  18. Kim BJ, Koh J, Liang N, Yang J, Ozturk G, Barile D, & Dallas DC (2023). Effect of vat pasteurization, ultra-high temperature sterilization, retort sterilization and homogenization on soluble proteins in donor human milk detected via proteomics. LWT, 182, 114842. 10.1016/j.lwt.2023.114842 [DOI] [Google Scholar]
  19. Li Y, Nguyen DN, de Waard M, Christensen L, Zhou P, Jiang P, Sun J, Bojesen AM, Lauridsen C, Lykkesfeldt J, Dalsgaard TK, Bering SB, & Sangild PT (2017). Pasteurization procedures for donor human milk affect body growth, intestinal structure, and resistance against bacterial infections in preterm pigs. Journal of Nutrition, 147(6), 1121–1130. 10.3945/jn.116.244822 [DOI] [PubMed] [Google Scholar]
  20. Lönnerdal B (2016). Human Milk: Bioactive Proteins/Peptides and Functional Properties (pp. 97–107). 10.1159/000442729 [DOI] [PubMed] [Google Scholar]
  21. Lueangsakulthai J, Kim BJ, Demers-Mathieu V, Sah BNP, Woo Y, Olyaei A, Aloia M, O’Connor A, Scottoline BP, & Dallas DC (2021). Effect of digestion on stability of palivizumab IgG1 in the infant gastrointestinal tract. Pediatric Research, 90(2), 335–340. 10.1038/s41390-020-01271-y [DOI] [PubMed] [Google Scholar]
  22. Mackie A, & Macierzanka A (2010). Colloidal aspects of protein digestion. Current Opinion in Colloid & Interface Science, 15(1–2), 102–108. 10.1016/j.cocis.2009.11.005 [DOI] [Google Scholar]
  23. Neu J (2019). Mother’s Own Milk: How Does It Differ from Donor Milk for the Baby. Breastfeeding Medicine, 14(S1), S-3–S-4. 10.1089/bfm.2019.0036 [DOI] [PubMed] [Google Scholar]
  24. O’Connor DL, Gibbins S, Kiss A, Bando N, Brennan-Donnan J, Ng E, Campbell DM, Vaz S, Fusch C, Asztalos E, Church P, Kelly E, Ly L, Daneman A, & Unger S (2016). Effect of Supplemental Donor Human Milk Compared With Preterm Formula on Neurodevelopment of Very Low-Birth-Weight Infants at 18 Months. JAMA, 316(18), 1897. 10.1001/jama.2016.16144 [DOI] [PubMed] [Google Scholar]
  25. Ogundele MO (2000). Activation and deposition of human breast-milk complement C3 opsonins on serum sensitive Escherichia coli 0111. Journal of Reproductive Immunology, 48(2), 99–105. 10.1016/S0165-0378(00)00067-X [DOI] [PubMed] [Google Scholar]
  26. Oizumi J, Hirano M, Hayakawa K, Daimatsu T, & Zaima K (1989). The Significance of Breast Milk Biotinidase. Pediatrics International, 31(4), 424–427. 10.1111/j.1442-200X.1989.tb01328.x [DOI] [PubMed] [Google Scholar]
  27. Peila C, Moro G, Bertino E, Cavallarin L, Giribaldi M, Giuliani F, Cresi F, & Coscia A (2016). The Effect of Holder Pasteurization on Nutrients and Biologically-Active Components in Donor Human Milk: A Review. Nutrients, 8(8), 477. 10.3390/nu8080477 [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Pitino MA, Beggs MR, O’Connor DL, Doyen A, Pouliot Y, Sergius-Ronot M, & Unger S (2023). Donor human milk processing and its impact on infant digestion: A systematic scoping review of in vitro and in vivo studies. Advances in Nutrition, 14(1), 173–189. 10.1016/j.advnut.2022.11.004 [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Pitino MA, Unger S, Doyen A, Pouliot Y, Aufreiter S, Stone D, Kiss A, & O’Connor DL (2019). High Hydrostatic Pressure Processing Better Preserves the Nutrient and Bioactive Compound Composition of Human Donor Milk. The Journal of Nutrition, 149(3), 497–504. 10.1093/jn/nxy302 [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Pitino MA, Unger S, Doyen A, Pouliot Y, Kothari A, Sergius-Ronot M, Bazinet RP, Stone D, & O’Connor DL (2023). Digestion of human milk processed by high pressure processing and Holder pasteurization using a dynamic in vitro model of the preterm infant. Food Chemistry, 411, 135477. 10.1016/j.foodchem.2023.135477 [DOI] [PubMed] [Google Scholar]
  31. Pitino MA, Unger S, Gill A, McGeer AJ, Doyen A, Pouliot Y, Bazinet RP, Kothari A, Mazzulli T, Stone D, & O’Connor DL (2022). High pressure processing inactivates human cytomegalovirus and hepatitis A virus while preserving macronutrients and native lactoferrin in human milk. Innovative Food Science & Emerging Technologies, 75, 102891. 10.1016/j.ifset.2021.102891 [DOI] [Google Scholar]
  32. Quigley M, Embleton ND, & McGuire W (2019). Formula versus donor breast milk for feeding preterm or low birth weight infants. The Cochrane Database of Systematic Reviews, 7, CD002971. 10.1002/14651858.CD002971.pub5 [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Quitadamo PA, Comegna L, & Cristalli P (2021). Anti-Infective, Anti-Inflammatory, and Immunomodulatory Properties of Breast Milk Factors for the Protection of Infants in the Pandemic From COVID-19. Frontiers in Public Health, 8. 10.3389/fpubh.2020.589736 [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Schanler RJ, Lau C, Hurst NM, & Smith EO (2005). Randomized Trial of Donor Human Milk Versus Preterm Formula as Substitutes for Mothers’ Own Milk in the Feeding of Extremely Premature Infants. Pediatrics, 116(2), 400–406. 10.1542/peds.2004-1974 [DOI] [PubMed] [Google Scholar]
  35. Sergius-Ronot M, Pitino MA, Suwal S, Shama S, Unger S, O’Connor DL, Pouliot Y, & Doyen A (2022). Impact of holder, high temperature short time and high hydrostatic pressure pasteurization methods on protein structure and aggregation in a human milk protein concentrate. Food Chemistry, 374, 131808. 10.1016/j.foodchem.2021.131808 [DOI] [PubMed] [Google Scholar]
  36. Underwood MA (2013). Human Milk for the Premature Infant. Pediatric Clinics of North America, 60(1), 189–207. 10.1016/j.pcl.2012.09.008 [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Victora CG, Bahl R, Barros AJD, França GVA, Horton S, Krasevec J, Murch S, Sankar MJ, Walker N, & Rollins NC (2016). Breastfeeding in the 21st century: epidemiology, mechanisms, and lifelong effect. Lancet, 387(10017), 475–490. 10.1016/s0140-6736(15)01024-7 [DOI] [PubMed] [Google Scholar]
  38. Wesolowska A, Sinkiewicz-Darol E, Barbarska O, Bernatowicz-Lojko U, Borszewska-Kornacka MK, & van Goudoever JB (2019). Innovative Techniques of Processing Human Milk to Preserve Key Components. Nutrients, 11(5). 10.3390/nu11051169 [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Zhang J, Duley JA, Cowley DM, Shaw PN, Koorts P, & Bansal N (2023). Comparative proteomic analysis of donor human milk pasteurized by hydrostatic high-pressure. Food Chemistry, 403, 134264. 10.1016/j.foodchem.2022.134264 [DOI] [PubMed] [Google Scholar]
  40. Zhang J, Lee NA, Duley JA, Cowley DM, Shaw PN, & Bansal N (2022). Comparing the effects of hydrostatic high-pressure processing vs holder pasteurisation on the microbial, biochemical and digestion properties of donor human milk. Food Chemistry, 373. 10.1016/j.foodchem.2021.131545 [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

1
NIHMS2019934-supplement-1.docx (228.1KB, docx)

RESOURCES