Abstract

   The goblet cells of the gastrointestinal tract (GIT) produce
   glycoproteins called mucins that form a protective barrier from
   digestive contents and external stimuli. Recent evidence suggests that
   the milk fat globule membrane (MFGM) and its milk phospholipid
   component (MPL) can benefit the GIT through improving barrier function.
   Our objective was to compare the effects of two digested MFGM
   ingredients with or without dextran sodium sulfate (DSS)-induced
   barrier stress on mucin proteins. Co-cultured Caco-2/HT29-MTX
   intestinal cells were treated with in vitro digests of 2%, 5%, and 10%
   (w/v) MFGM or MPL alone for 6 h or followed by challenge with 2.5% DSS
   (6 h). Transepithelial electrical resistance and fluorescein
   isothiocyanate (FITC)-dextran (FD4) permeability measurements were used
   to measure changes in barrier integrity. Mucin characterization was
   performed using a combination of slot blotting techniques for secreted
   (MUC5AC, MUC2) and transmembrane (MUC3A, MUC1) mucins, scanning
   electron microscopy (SEM), and periodic acid Schiff (PAS)/Alcian blue
   staining. Digested MFGM and MPL prevented a DSS-induced reduction in
   secreted mucins, which corresponded to the prevention of DSS-induced
   increases in FD4 permeability. SEM and PAS/Alcian blue staining showed
   similar visual trends for secreted mucin production. A predictive
   bioinformatic approach was also used to identify potential KEGG
   pathways involved in MFGM-mediated mucosal maintenance under colitis
   conditions. This preliminary in silico evidence, combined with our in
   vitro findings, suggests the role of MFGM in inducing repair and
   maintenance of the mucosal barrier.

   Keywords: mucin, milk fat globule membrane, phospholipids, in vitro
   digestion, cell culture, HT29-MTX, scanning electron microscopy

1. Introduction

   Mucosal barrier function of the gastrointestinal tract (GIT) is
   orchestrated by intestinal epithelial cells (IECs), immune cells, and
   the mucus layer to defend against external stimuli, absorb nutrients,
   and eliminate waste. The small intestine serves as the main site of the
   GIT to protect, sense, and respond to external stimuli by means of its
   physical and molecular components. The physical components of the
   intestinal barrier are composed of the mucus layer, which acts as the
   first barrier of defense, and IECs that tightly regulate the
   paracellular and transcellular passage of luminal contents. The
   intestinal mucus is made up of high-molecular-weight glycoproteins
   called mucins, along with lipids, secretory IgG, water, and
   antimicrobial peptides [[28]1]. Intestinal mucins include both
   secreted, gel-forming mucins (i.e., MUC2, MUC5AC), which form the
   protective layer on the epithelial surface, and transmembrane mucins
   (i.e., MUC3A, MUC1), which are involved in cell signaling
   [[29]2,[30]3]. Perturbations in the mucosal barrier through various
   stimuli such as pathogenic infection, dysbiosis of the gut microbiota,
   various medications, stress, and poor nutrition increase the host’s
   susceptibility to damage and potentiates the onset of intestinal
   barrier dysfunction-related diseases, such as inflammatory bowel
   disease (IBD) and colitis [[31]4]. Furthermore, mucosal barrier
   dysfunction predisposes individuals to the development of food
   allergies, celiac disease, and diabetes [[32]5,[33]6]. Hallmarks of gut
   barrier disruption include increased gut permeability via decreased
   tight junction protein expression, infiltration of immune cells and
   their cytokine responses, and decreased mucin through the depletion of
   goblet cells [[34]6,[35]7]. Once barrier function is compromised,
   individuals are at heightened risk for insult by further stressors,
   underscoring the need to find preventative and restorative solutions
   for adequate barrier function.

   Dextran sodium sulfate (DSS) is frequently used as a model for barrier
   disruption and a colitis-like physiology in both cell culture and in
   vivo animal models [[36]8,[37]9]. DSS is a water-soluble polysaccharide
   with a negative charge that is proposed to penetrate IECs, leading to
   gut leakiness and a subsequent immune response [[38]9]. Compared to
   other models, DSS is favored for its reproducibility and
   controllability, serving as a tunable model for severity of
   inflammation and barrier disruption based on dosage and exposure time
   [[39]9]. Typical physiological changes that occur after DSS exposure
   include reduction in tight junction protein expression (i.e.,
   occludins, Zonula occludin (ZO-1), claudins), ulceration, and
   infiltration of immune cells, mimicking many of the physiological
   changes that are observed with the pathogenesis of IBD [[40]10,[41]11].
   Further characterized by the depletion of goblet cells and the
   corresponding decrease in mucin production, DSS is associated with a
   reduction in secreted mucins, including MUC2, which is the predominant
   secreted mucin in the small intestine and colon. Through various mucin
   knockout studies and the current evidence on DSS models, the depletion
   of mucins and changes in their glycosylation patterns are associated
   with dramatic changes in the gut microbiota and long-term consequences
   to human health [[42]12,[43]13].

   The use of restorative and preventative agents of gut barrier
   dysfunction, including bioactive nutritional components, have the
   potential to impact the mucosal barrier through maintenance of its
   physical and molecular components [[44]14]. Mammalian milk has long
   been considered a rich consortium of bioactive components, including
   the milk fat globule membrane (MFGM), which has gained substantial
   attention in recent years for its copious health benefits. The MFGM
   consists of a complex tri-layer structure that emulsifies fat globules
   in milk and is composed of more than 200 different proteins and
   numerous lipids with unique structural and functional properties
   [[45]15]. The protein component of the MFGM includes eight major
   proteins, such as lactadherin (PAS6/7), butyrophilin (BTN), mucin-1
   (MUC1), fatty acid binding protein (FABP), mucin-15 (MUC15), xanthine
   dehydrogenase/oxidase (XDH/XO), adipophilin (ADPH), and hundreds more
   [[46]15,[47]16]. The predominant lipid components of the MFGM include
   milk phospholipids (MPL), which consist of phosphatidylcholine (PC;
   19–43%), phosphatidylethanolamine (PE; 26–46%), phosphatidylserine (PS;
   3–16%), phosphatidylinositol (PI; 1–14%), and sphingomyelin (SM;
   18–29%) [[48]17].

   Based on in vivo models and clinical trials of infant nutrition, the
   current evidence supports the role of the MFGM and MPL fraction in gut
   development [[49]18], brain and cognitive function [[50]19,[51]20],
   bone growth and remodeling [[52]21,[53]22], cardiometabolic health
   [[54]23,[55]24,[56]25], and more [[57]26], leading to its incorporation
   into infant formulas. With respect to its gut-benefiting properties,
   the MFGM increases the length of the villi of the developing intestine,
   downregulates pro-inflammatory cytokines, and increases tight junction
   protein expression [[58]18]. The effects of the MFGM on the gut
   microbiota are variable. However, there is a consistent finding across
   studies that it favors increased abundance of bifidobacteria, which are
   known for their beneficial properties in the gut [[59]26,[60]27]. A
   potential mechanism of these gut microbiota-related changes upon MFGM
   supplementation is through regulation of mucin proteins—a key point of
   interaction for gut microbes. A few studies have previously reported
   the effects of MFGM on goblet cells and intestinal mucin gene
   expression [[61]28,[62]29]. Early evidence in short-bowel rat models
   suggests that the MFGM can increase goblet cell count and the number of
   MUC1/MUC2-positive cells [[63]28]. Additionally, in an acute colitis
   prevention mouse model, prophylactic MFGM treatment resulted in an
   upregulation in Muc2 and Muc4 gene expression [[64]29].

   As the phospholipid component of the MFGM is often attributed to being
   the key contributor to MFGM’s health benefits [[65]26], we were
   interested in first determining whether two MFGM ingredients with
   varying phospholipid content could influence a variety of mucin
   proteins (both secreted and transmembrane) using an in vitro digestion
   approach coupled with intestinal cell co-culture (Caco-2/HT29-MTX). In
   addition, the current evidence of MFGM on mucin implies the
   pre-requisite of intestinal stress. Therefore, we also aimed to
   investigate the preventative role of the MFGM on intestinal mucins with
   or without challenge using DSS. We used three nutritionally relevant
   supplementation levels that are translatable to various food
   formulations (2, 5, 10% (w/v)) prior to digestion.

2. Materials and Methods

2.1. In Vitro Digestion of MFGM Ingredients

   Two ingredients, a commercial whey-derived MFGM ingredient (~7%
   phospholipids by weight) and a commercial beta serum powder-derived MPL
   ingredient (~70% phospholipids by weight), were subjected to simulated
   digestion following the static INFOGEST 2.0 digestion model and
   guidelines consisting of oral, gastric, and intestinal phases [[66]30].
   MFGM ingredients were resuspended in DI H[2]O at concentrations of 2,
   5, and 10% (w/v) with gentle agitation. For the oral phase, 5 mL of
   each solution were mixed with 4 mL of simulated salivary fluid, 1.5 mM
   CaCl[2], 75 U/mL salivary α-amylase, and DI water up to a final volume
   of 10 mL. The oral phase was incubated at 37 °C for 2 min with mixing
   on a rocking platform (VWR, Radnor, PA, USA) at 95 rpm. For the gastric
   phase, 8 mL of simulated gastric fluid were added to the same tube and
   the pH was adjusted to 3.0 using 5 M HCl. Then, 2000 U/mL pepsin, 0.15
   mM CaCl[2], and DI water up to a final volume of 20 mL were added and
   the tubes were incubated at 37 °C for 2 h with mixing. For the
   intestinal phase, 8 mL of simulated intestinal fluid were added to each
   tube after the oral and gastric phases and the pH was adjusted to 7.0
   using 5 M NaOH. Tubes were incubated with porcine bile salts (10 mM)
   for 30 min with rocking. Following bile salt incubation, pancreatin
   (trypsin activity 100 U/mL), 0.6 mM CaCl[2], and water up to a final
   volume of 40 mL were added and incubated for 2 h at 37 °C. Preliminary
   experiments were conducted to determine the appropriate volumes of 5 M
   HCl and 5 M NaOH for pH adjustments. At the end of the intestinal
   phase, the resulting digested fluids were immediately centrifuged
   (15,000× g, 4 °C, 45 min) and the micellar fractions, or bioaccessible
   fractions, were collected from the mid-section of the supernatant with
   a syringe and filtered through a 0.2 μm nylon filter. Samples were kept
   frozen at −80 °C until use. All digestion reagents were purchased from
   Sigma Aldrich (St. Louis, MO, USA).

2.2. Caco-2/HT29-MTX Co-Culture

   The HT29-MTX-E12 goblet-like cell line and Caco-2 enterocyte cell line
   were obtained from the European Collection of Authenticated Cell
   Cultures (Salisbury, UK) and the American Type Culture Collection
   (Manassas, VA, USA), respectively. Both cell lines were maintained
   independently in 25 mM high-glucose Dulbecco’s Modified Eagle Medium
   (DMEM) containing 10% heat-inactivated fetal bovine serum, 1%
   penicillin–streptomycin (100 units/mL penicillin and 100 units/mL
   streptomycin), 1% non-essential amino acids (100×), and 1% 200 mM
   L-glutamine, with media replacement every 48 h. At 80–90% confluency,
   the two cell lines were passaged and mixed at the physiologically
   relevant ratio of 90:10 Caco-2 to HT29-MTX cells to mimic the
   proportions of absorptive and goblet cells found in the small intestine
   [[67]31]. Co-cultures were plated in 12-well Transwell plates using a
   seeding density of 4 × 10^5 cells/insert and incubated at 37 °C in a 5%
   CO[2] humidified atmosphere for 21 days post-confluency. The media were
   replaced with serum-free media one day prior to all experiments. All
   cell culture reagents were purchased from Thermo Fisher (Gibco,
   Waltham, MA, USA).

   For cell treatment, micellar fractions were diluted to 1:100 in Hank’s
   Balanced Salt Solution (HBSS; Ca^2+- and Mg^2+-free) and then further
   diluted to 1:10 in serum-free media to maintain the osmotic balance of
   the cells based on cytotoxicity measurements. Then, 500 μL of the
   micellar fractions in media were applied to the apical chamber of the
   cells and fresh serum-free media were applied to the basolateral
   compartment. After 6 h, cells were either harvested for subsequent
   assays or further treated with dextran sodium sulfate (DSS) for an
   additional 6 h to achieve a final concentration of 2.5% DSS in the
   apical compartment. DSS was applied without removing micellar
   fraction-containing media to mitigate any effect of washing on the
   mucosal layer. A graphical representation of cell culture treatments
   and time points are described in [68]Figure 1.

Figure 1.

   [69]Figure 1
   [70]Open in a new tab

   Overview of treatments and measurable outcomes to assess the
   preventative effects of the milk fat globule membrane (MFGM) and milk
   phospholipids (MPLs) on dextran sodium sulfate (DSS)-induced mucosal
   barrier damage using the Caco-2/HT29-MTX (90:10) intestinal cell
   culture model.

2.3. Transepithelial Electrical Resistance (TEER) and FITC-Dextran
Permeability

   The transepithelial electrical resistance (TEER) of each well was
   recorded after washing twice with HBSS (with Ca^2+ and Mg^2+) at 0, 6,
   and 12 h using an EVOM Epithelial Voltohmmeter with an STX2 probe
   (Sigma). The relative change in TEER value was determined for each well
   compared to the initial value. Paracellular permeability was determined
   by measuring the transport of Fluorescein isothiocyanate (FITC)-dextran
   (FD4; MW = 4 kDa; Sigma) across the cell monolayer in phenol red- and
   serum-free media. A final concentration of 2 mg/mL was applied to the
   apical side of the inserts for 2 h at 37 °C in the dark. At the end of
   incubation, 100 μL aliquots were collected from the basolateral
   compartment and transferred to black 96-well plates. Fluorescence was
   measured using a SpectraMaxM2 fluorescent plate reader (490 nm
   excitation/520 nm emission; Molecular Devices, San Jose, CA, USA) and
   quantified based on a standard curve of FD4 concentrations ranging from
   0 to 2.56 ng/μL.

2.4. Slot Blot for Mucin Detection

   Cells were washed 2× with PBS and incubated for 30 min on ice with 200
   μL RIPA buffer containing complete protease inhibitor (Roche
   Diagnostics, Indianapolis, IN, USA). Lysates were centrifuged (11,000×
   g, 10 min, 4 °C) and the supernatants were transferred to new 1.5 mL
   tubes. Protein concentration was determined using the Pierce microBCA
   kit (Thermo Fisher). For mucin detection, a nitrocellulose membrane
   (0.45 μm) was pre-wet for 10 min in 1× tris-buffered saline (TBS), and
   the Bio-Dot SF apparatus (BioRad, Hercules, CA, USA) was assembled
   according to the manufacturer’s recommendations. After washing each
   well with 1× TBS buffer, 200 μL of sample (diluted to 17 μg/mL) were
   loaded per well and pulled onto the membrane via vacuum. The membrane
   was transferred to a blocking buffer consisting of 3% cold-soluble fish
   gelatin in TBS buffer for 1 h at 25 °C on a rotating platform (70 rpm).
   Each membrane was subsequently washed 3× in TBS buffer with 0.05%
   Tween-20 (TBST) for 5 min. For primary antibody incubation, each
   antibody was diluted in antibody buffer (AB) consisting of 1% gelatin
   in TBST and incubated overnight at 4 °C on a rotating platform (70
   rpm). The primary antibodies used included: mouse monoclonal MUC5AC
   (1:5000, Cat. No. WH0004586M7, Sigma Aldrich), mouse monoclonal MUC2
   (1:3000, Cat. No. ab11197, Abcam, Cambridge, UK), rabbit polyclonal
   MUC3A (1:2000, Cat. No. PA5100199, Invitrogen, Waltham, MA, USA), and
   mouse monoclonal MUC1 (1:3000, Invitrogen, Cat. No. MA1-06503).
   Following 3× washing with TBST (5 min each), each membrane was treated
   with 1:3000 Immun-Blot Goat Anti-Mouse IgG-AP Conjugate or Immun-Blot
   Goat Anti-Rabbit IgG-AP Conjugate (BioRad) in AB and incubated for 2 h
   at 25 °C with rocking. Membranes were washed again 3× with TBST
   followed by one wash in TBS, and the AP Color Development kit (BioRad)
   was used according to the manufacturer’s instructions. Finally, each
   membrane was washed 2× with DI H[2]O (5 min each) and the membranes
   were imaged while wet using a BioRad ChemiDoc MP Imaging System.

2.5. Scanning Electron Microscopy (SEM)

   Cells were washed 2× with PBS and fixed with 3% glutaraldehyde, 2%
   paraformaldehyde in 0.1 M potassium phosphate buffer (PB; pH 7.2)
   overnight at 25 °C. Samples were washed with PB for 10 min followed by
   a single wash with ddH[2]O for 10 min prior to post-fixation with 2%
   OsO[4] in ddH[2]O for 30 min. After transferring to a falcon tube,
   samples were dehydrated using an ethanol series (25% EtOH, 1 × 15 min;
   50% EtOH, 1 × 15 min; 70% EtOH, 1 × 15 min; 90% EtOH, 1 × 15 min; 100%
   EtOH, 3 × 15 min). Finally, after critical point trying, samples were
   sputter-coated with a 3 nm Pt coating. SEM micrographs were captured
   using a Hitachi Schottky Field Emission SU5000 Scanning Electron
   microscope (Hitachi, Tokyo, Japan) at the Molecular and Cellular
   Imaging Center (MCIC) located at The Ohio State University, Ohio
   Agricultural Research and Development Center (OARDC) in Wooster, OH,
   USA. The % coverage was determined using the Fiji software V2.15
   [[71]32].

2.6. Glycoprotein Staining

   Cells were washed 2× with PBS and fixed with 10% neutral buffered
   formalin. Samples were then sent to the Comparative Pathology & Digital
   Imaging Shared Resource (CPDISR) facility at The Ohio State University
   for processing, paraffin embedding, and combined periodic acid Schiff
   (PAS) and Alcian blue staining. Slides were imaged using an EVOS FL
   color microscope (Invitrogen) under 40× magnification.

2.7. Bioinformatic Analysis

   Prediction of MFGM lipid target action was determined by first
   obtaining the 2D structures of the digestion end products of the three
   most predominant phospholipid classes from bovine MFGM using PubChem:
   lysophosphatidylethanolamine 18:1/0:0 (LPE 18:1/0:0);
   lysophosphatidylcholine 16:0/0:0 (LPC 16:0/0:0); and
   sphingosine-1-phosphate (S1P) [[72]33,[73]34]. Using these 2D
   structures, a list of predicted human targets for each PL product was
   obtained via Swiss Target Prediction v2023 ([74]Supplementary Tables
   S1–S3). Targets for colitis were obtained from the GeneCards v5.16 and
   DisGeNet v7.0 databases ([75]Supplementary Table S4) [[76]35,[77]36].
   MFGM lipid targets were cross-referenced to colitis targets using a
   Venn diagram to discern potential targets of MPL for the prevention of
   DSS-induced colitis ([78]Supplementary Table S5). Kyoto Encyclopedia of
   Genes and Genomes (KEGG) pathway enrichment analysis of common targets
   between MFGM lipids and colitis was performed in R using
   clusterProfiler 4.0 against the Homo sapiens genome using p < 0.05 as a
   cutoff criterion ([79]Supplementary Table S6) [[80]37]. The output
   containing the top 10 pathways ranked by adjusted p-value was
   visualized using a dot plot.

2.8. Statistical Analysis

   All experiments were conducted at least in triplicate. Data presented
   are represented as the mean ± SEM where appropriate. Statistical
   analyses were conducted using one-way analysis of variance (ANOVA) with
   a post hoc Tukey test for pairwise comparisons or using a
   Kruskal–Wallis test with a post hoc Dunn’s test where appropriate.
   Differences of p < 0.05 were considered statistically significant. All
   statistical analysis was performed in GraphPad Prism V10 unless
   otherwise noted.

3. Results

3.1. Protective Effects of MFGM and MPL on Caco-2/HT29-MTX Barrier Integrity
and Permeability

   Caco-2/HT29-MTX (90:10) cells in Transwell plates were treated with
   digested 0, 2, 5, and 10% MFGM or MPL for 6 h alone (no challenge) or
   with an additional 6 h 2.5% DSS treatment to induce barrier damage
   (prevention). The potential of the digested MFGM and MPL ingredients to
   prevent DSS-induced barrier damage was observed without removing
   digesta-containing media prior to DSS treatment. The effects of these
   ingredients on TEER and FD4 permeability were measured under the
   no-challenge or prevention conditions.

   Under no-challenge conditions, the various digested MPL concentrations
   significantly increased TEER values greater than the negative control,
   whereas the various MFGM concentrations significantly decreased TEER
   below the negative control ([81]Figure 2A). Under the FD4 permeability
   measurement, the MFGM and MPL treatments did not significantly differ
   from the FD4 permeability of the control ([82]Figure 2B). Changes in
   TEER and FD4 values are associated with changes in the transport of
   ions (conductance) and barrier permeability, respectively [[83]38].
   Together, our findings indicate the intactness of the cell barrier
   after digested MFGM or MPL treatments but may suggest possible changes
   in ion transport through the regulation of tight junction protein
   expression and/or tight junction pore size, which has been previously
   reported for MFGM ingredients [[84]39,[85]40].

Figure 2.

   [86]Figure 2
   [87]Open in a new tab

   Effects of MFGM and MPL alone (A,B) or in the context of DSS challenge
   (C,D) on transepithelial electrical resistance (TEER) and paracellular
   permeability of FITC-dextran (FD4). Statistical significance (p < 0.05)
   is denoted for comparisons between negative and/or DSS + blank digesta
   controls and each treatment using a one-way ANOVA with a post hoc Tukey
   test (A,D) and the Kruskal–Wallis test with a post hoc Dunn test (B,C).

   With the preventative assessment, digested 2 and 10% MPL exhibited
   significantly higher TEER values compared to the DSS + blank digesta
   control—a positive control to induce mucosal barrier damage containing
   both digested water (background digestion elements) and 2.5% DSS
   ([88]Figure 2C). The same MPL treatments did not significantly differ
   from the negative control, indicating the prevention of a DSS-induced
   decrease in TEER by MPL. Despite MFGM showing dose-dependent trends, no
   significant differences were observed between the MFGM samples and
   either the DSS + blank digesta or negative controls. For FD4
   permeability, all MFGM and MPL concentrations inhibited DSS-induced
   permeability, suggesting fortification of the mucus layer ([89]Figure
   2D).

3.2. Effects of MFGM or MPL on DSS-Treated Cells Demonstrated Protective
Effects on Mucin Protein Abundance

   Given the differences in permeability we observed between digested
   MFGM/MPL-treated cells with and without DSS challenge, we were
   interested in observing whether these changes were related to the
   mucins, as mucin levels are associated with intestinal permeability
   [[90]41,[91]42]. In general, DSS interferes with standard qPCR
   reactions [[92]43], so we focused our efforts on characterizing the
   effects of digested MFGM and MPL on both secreted (MUC5AC, MUC2) and
   transmembrane (MUC3A, MUC1) mucin protein. MUC5AC and MUC2 were
   selected, as they represent the predominant secreted mucin for HT29-MTX
   cells and the predominant human intestinal mucin in vivo, respectively
   [[93]44,[94]45]. MUC3A and MUC1 were selected, as MUC3A is one of the
   most well-understood and predominant transmembrane mucin of HT29-MTX
   cells and MUC1 is known to have aberrant expression in colitis-like
   phenotypes [[95]44,[96]46].

   Neither digested MFGM nor MPL treatments had major effects on MUC5AC,
   MUC2, or MUC1 mucin protein expression without challenge ([97]Figure
   3A–C); however, digested 2 and 5% MFGM and 2% MPL led to significant
   reductions in MUC3A ([98]Figure 3D).

Figure 3.

   [99]Figure 3
   [100]Open in a new tab

   Effects of MFGM and MPL on secreted MUC5AC (A) and MUC2 (B) and
   transmembrane MUC1 (C) and MUC3A (D) in Caco-2/HT29-MTX cells using a
   slot blot. Prevention of DSS-induced disruption of secreted MUC5AC (E)
   and MUC2 (F) and transmembrane MUC1 (G) and MUC3A (H). Statistical
   significance (p < 0.05) is denoted for comparisons between negative
   and/or DSS + blank digesta controls and each treatment using a one-way
   ANOVA with a post hoc Tukey test.

   Upon perturbations in the mucosal barrier through DSS challenge, we
   observed a significant reduction in both secreted mucins, MUC5AC and
   MUC2 (p = 0.0068 and p = 0.0027), in response to DSS treatment
   ([101]Figure 3E,F). The significant reduction in MUC5AC was prevented
   by all digested MFGM concentrations and digested 5 and 10% MPL (similar
   to negative control). However, the reduction in MUC2 was partially
   restored through pre-treatment with 2 and 10% MPL, as these treatments
   were significantly different with neither the negative control nor the
   DSS + blank digesta control.

   Transmembrane mucin MUC1 is dysregulated in the progression of colitis
   to colon cancer [[102]47]. In our DSS-induced colitis model, DSS
   treatment yielded a slight though non-significant increase in
   transmembrane MUC1 ([103]Figure 3G). However, both MFGM and MPL digesta
   significantly reduced MUC1 protein levels compared to the DSS + blank
   digesta control and for some treatments (10% MFGM and 2% MPL) compared
   to the negative control.

   MUC3A was not significantly altered by DSS treatment or the majority of
   MFGM and MPL digesta concentrations apart from 10% MPL + DSS, which
   showed an increase in MUC3A (p = 0.0137; [104]Figure 3H).

   The mucin protein levels under both no-challenge and prevention
   conditions show similar trends to the data observed in the permeability
   assessment ([105]Figure 2D) where improvements in DSS-increased
   permeability are associated with the restoration of secreted MUC5AC
   levels and, to a lesser degree, MUC2 ([106]Figure 3E,F). Conversely,
   the absence of weakened permeability in no-challenge conditions
   ([107]Figure 2B) seems to be associated with no changes in secreted
   mucins ([108]Figure 3A,B).

3.3. DSS-Induced Impaired Brush Border Is Prevented through Digested MFGM and
MPL Treatments

   Based on the previous results, we were interested in whether the cell
   layer structure was different between treatments. To visually observe
   the effect of MFGM and MPL on cell layer structure, we employed SEM and
   PAS/AB staining. Cells were once again treated for 6 h with MFGM or MPL
   and, without removing media, treated with 2.5% DSS for an additional 6
   h to induce barrier damage. Only the 2 and 10% concentrations for each
   ingredient were selected to represent the lowest and highest
   concentrations, respectively. We observed an intact brush border with
   mucin production (light blue arrow) for the negative control by SEM
   imaging ([109]Figure 4A). Treatment with DSS led to impairment of the
   intestinal barrier through altering surface coverage ([110]Table 1) as
   well as the absence of mucin-like structures at the cell surface
   ([111]Figure 4B). Both 2% MFGM and 2% MPL with DSS treatments led to
   varying degrees of coverage and mucin abundance ([112]Figure 4C,D);
   however, the 10% concentrations of these ingredients led to nearly
   complete surface coverage and apparent mucin abundance ([113]Figure
   4E,F).

Figure 4.

   [114]Figure 4
   [115]Open in a new tab

   Scanning electron microscopy (SEM) of Caco-2/HT29-MTX cells exposed to
   digested MFGM or MPL and DSS. (A) Negative control with intact brush
   border and mucosal barrier. (B) DSS + blank digesta control with
   impaired mucosal barrier and brush border. (C,D) 2% MFGM + DSS and 2%
   MPL + DSS treatments with partially intact brush border and mucosal
   barrier after DSS challenge. (E,F) 10% MFGM + DSS and 10% MPL + DSS
   treatments with intact mucosal barrier and brush border after DSS
   challenge. Blue arrow indicates mucin production.

Table 1.

   Surface coverage of intestinal barrier from SEM images after MFGM or
   MPL treatment and subsequent DSS challenge.
   Negative Control DSS + Blank Digesta 2% MFGM + DSS 2% MPL + DSS 10%
   MFGM + DSS 10% MPL + DSS
   % Coverage 86.6 ± 2.04 71.2 ± 2.03 78.2 ± 6.74 88.9 ± 0.714 99.3 ±
   0.501 99.8 ± 0.087
   [116]Open in a new tab

   To further observe whether these effects could be related to changes in
   mucin production, a cross-section of Caco-2/HT29-MTX cells was stained
   with PAS and Alcian blue to indicate the presence of neutral mucins and
   acidic/sialylated mucins, respectively. As shown by the white arrows in
   [117]Figure 5A, the negative control exhibits both acidic/sialylated
   and neutral mucins, whereas a depletion of these mucins is observed in
   the DSS + blank digesta control ([118]Figure 5B). Both digested 10%
   MFGM and 10% MPL prevent this ablation, as noted by the presence of
   acidic and/or sialylated mucins ([119]Figure 5E,F) and the 2%
   concentrations of these ingredients appear to have an intermediary
   effect ([120]Figure 5C,D).

Figure 5.

   [121]Figure 5
   [122]Open in a new tab

   Periodic acid Schiff (PAS) and Alcian blue (AB) staining for neutral
   and sialylated mucins (pink magenta) and acidic mucins (blue) of
   Caco-2/HT29-MTX cells exposed to digested MFGM or MPL and DSS. (A)
   Negative control with acidic mucin production barrier. (B) DSS + blank
   digesta control with impaired mucin production. (C,D) 2% MFGM + DSS and
   2% MPL + DSS treatments with partial recovery of mucin. (E,F) 10% MFGM
   + DSS and 10% MPL + DSS treatments with acidic mucin production after
   DSS challenge. The white arrows represent acidic mucin production. The
   green arrow indicates nuclei.

3.4. In Silico Prediction of Pathways Involved in Prevention of DSS-Induced
Mucosal Damage by MFGM Lipids

   Regulation of mucin production in intestinal cells is not well
   understood to date. Therefore, as a beginning insight to explore
   potential metabolic pathways involved in mucosal barrier function and
   influenced by MFGM lipids, we implemented a bioinformatic approach. The
   three predominant digested phospholipid products derived from both MFGM
   ingredients include LPE (18:1/0:0), LPC (16:0/0:0), and S1P
   [[123]33,[124]34]. We obtained the potential target genes for these
   lipids using SwissTargetPrediction and cross-referenced them with
   colitis-associated genes using GeneCards and DisGeNet databases
   ([125]Supplementary Tables S1–S4). We observed 100 targets that
   intersected between at least one colitis target and at least one
   hydrolyzed phospholipid target ([126]Figure 6A; [127]Supplementary
   Table S5). Using these targets, a KEGG pathway analysis was performed
   to determine potential interactions of the MFGM lipids in preventing
   DSS-induced mucosal damage ([128]Figure 6B). Our in silico findings
   indicate that the MFGM lipids may be acting through a variety of
   metabolic pathways involved in cell survival, growth, and responses to
   changes in the cellular environment, including the PI3K-Akt signaling
   pathway, Ras signaling pathway, EGFR tyrosine kinase inhibitor
   resistance pathway, and HIF-1 signaling pathway. Lipid signaling
   pathways, including sphingolipid signaling and endocrine resistance,
   appear to also be targets of the MFGM lipids. Notably, the calcium
   signaling pathway was predicted to be a target of MFGM lipids, which
   appears to be at the epicenter of most of the predicted target
   pathways. Intracellular Ca^2+ is a potent secondary messenger that
   mediates the crosstalk between signaling pathways [[129]48]. Therefore,
   investigation into the role of calcium in directing the MFGM lipid
   response is warranted and may provide further explanation for the role
   of MFGM lipids in maintaining mucin production, as mucin secretion is
   Ca^2+-dependent [[130]49].

Figure 6.

   [131]Figure 6
   [132]Open in a new tab

   Prediction of pathways involved in prevention of DSS-induced mucosal
   damage by MFGM lipids. (A) Venn diagram of shared and exclusive targets
   for MFGM lipids (LPE 18:1/0:0, LPC 16:0/0:0, S1P) and colitis gene
   databases (GeneCards and DisGeNet). (B) Dot plot of top 10 enriched
   KEGG pathways for 100 shared genes between at least one digested MFGM
   lipid and one colitis gene database list ranked by gene ratio. Gene
   ratio is the ratio of input of shared genes to the total genes in a
   given pathway. The dot color indicates adjusted p-value.

4. Discussion

   Permeability of the mucosal barrier is associated with a reduction in
   secreted mucins and changes in IECs, among other factors
   [[133]4,[134]50]. In our study, when comparing permeability and mucin
   under no challenge conditions, we observed no significant differences
   between treatments and controls for either FD4 permeability or secreted
   mucin expression (MUC5AC and MUC2). In contrast, improvements in FD4
   permeability were associated with fully or partially restored secreted
   mucin production in DSS-challenged cells. This relationship suggests
   the relationship of MFGM or MPL in maintenance of the mucosa under
   stress conditions and minor effects under non-stressed conditions.

   The depletion of goblet cells in DSS-induced colitis and other IBDs is
   associated with varying changes in mucin proteins. In DSS-induced
   colitis, MUC1 expression is increased [[135]46] and Muc1−/− mice have
   lower colitis severity accompanied by an increase in expression of MUC2
   and MUC3 in a compensatory manner [[136]51,[137]52]. In our study, we
   observed an increase in MUC1 protein level in response to the DSS
   challenge, which was retained to levels of the negative control with
   MFGM and MPL pre-treatment. Using knockout mouse models with chemically
   induced colitis, MUC5AC expression has been shown to protect the
   intestinal barrier through preventing bacterial translocation across
   the mucosal layer [[138]53]. Similarly, MUC2 knockout mice are more
   susceptible to the development of colitis [[139]54]. In our study,
   digested MPL pre-treatment fully prevented DSS-induced reduction in
   MUC5AC and MUC2 protein levels ([140]Figure 3E,F). We observed a
   similar trend for digested MFGM, as MUC5AC was maintained after DSS
   treatment; however, MUC2 protein levels remained at levels similar to
   the DSS control ([141]Figure 3F). Taken together, digested MPL and MFGM
   protect against DSS-induced changes in mucin protein in Caco-2/HT29-MTX
   co-culture. MPL and MFGM have similar, albeit slightly different,
   effects.

   The differences in mucin protein in response to digested MFGM and MPL
   could be due to the inherent differences in ingredient composition that
   influence both the relative bioactive lipid purity and digestive
   processes. The MFGM ingredient contains MFGM proteins, phospholipids
   (~7 g/100 g powder), whey proteins, caseins, lactose, and more, whereas
   the MPL ingredient contains a mixture of phospholipids (~70 g/100 g
   powder), peptides, lactose, and other minor lipid components, and lacks
   protein entirely. The differences in purity level of phospholipids, and
   thus the final concentration of phospholipids available to the
   intestinal cells, could have contributed to the differences observed in
   mucin protein responses. Despite the differences in phospholipid
   content between ingredients, we chose to use 2, 5, and 10% (w/v)
   solutions at the start of digestion, as these concentrations represent
   supplementation levels that could be realistically incorporated into
   foods.

   Additionally, the proteins present in the ingredient may have affected
   the digestive processes and bioaccessibility of MFGM lipids.
   Glycerophospholipids are hydrolyzed in the sn-2 position by
   phospholipase A2 (PLA2) in adults and likely by pancreatic
   lipase-related protein 2 (PLRP2) in infants [[142]55]. It has been
   suggested that further hydrolysis of lyso-PC and lyso-PE can occur at
   the mucosal barrier by bile salt-dependent phospholipase B (PLB) in
   adults [[143]56]. Phospholipid hydrolysis is dependent on dietary
   components, bile salt concentration, age, and more [[144]56]. For
   example, protein-rich matrices enhance the bioaccessibility of milk
   polar lipids and reduce the bioaccessibility of neutral lipids and
   cholesterol [[145]57]. Additionally, various milk proteins, such as
   β-lactoglobulin, have been shown to bind milk polar lipids and confer
   protection from GIT conditions, potentially influencing their fate to
   mediate cellular metabolic processes [[146]58,[147]59]. These
   interactions, as well as the action of MFGM-derived peptides, may help
   explain some of the differences in mucin protein expression between
   digested MFGM and MPL samples; however, both MFGM and MPL delivered
   beneficial outcomes on mucin protein.

   Previous literature has suggested the role of MFGM and its components
   in modulation of the mucosal barrier. In an acute colitis mouse model,
   a prophylactic MFGM treatment followed by DSS increased the gene
   expression of Muc2 and Muc4, which was associated with the enrichment
   of nine genera, including Bifidobacterium [[148]60]. The MFGM was also
   found to normalize intestinal development through regulation of villus
   length and crypt depth, tight junction protein expression, and
   promotion of goblet cell number using a pup-in-a-cup model [[149]18].
   In a similar study investigating the effects of MFGM supplementation in
   infant formula (IF), MFGM-containing IF restored MUC2 expression to
   levels of breast milk in a dose-dependent manner in Sprague-Dawley rat
   pups [[150]61]. Using a piglet model, MFGM supplementation in a basal
   diet for late gestational pregnant sows resulted in increases in Muc1,
   Muc2, Muc4, and Muc13 gene expression in neonatal piglets [[151]62].
   Taken together with these studies, our findings support the role of the
   MFGM and MPL in modification of mucin protein under stress conditions.
   In general, the modifications of mucin proteins during stress offer a
   potential explanation for the role of the MFGM in colitis and on the
   gut microbiota, as changes in mucin protein expression and
   glycosylation patterns directly influence resident gut microbes
   [[152]13]. Other critical components of the mucosal barrier, including
   tight junction proteins and cytokine production, are also positively
   influenced by MFGM supplementation. The MFGM maintains tight junction
   protein expression, such as claudin-4 and zonula occludin-2, that aids
   with the maintenance of barrier integrity in a murine colitis model
   [[153]39]. The MFGM also reduces the production of several
   pro-inflammatory cytokines, such as IL-1β and IL-6, to help maintain
   immune balance [[154]29]. Ultimately, these multiple benefits of the
   MFGM strengthen its role in the prevention of gut dysfunction-related
   diseases.

   Based on our findings, it remains unclear whether the digested MFGM and
   MPL components are embedding themselves into the mucosal layer to alter
   the viscosity of the mucosa and thus inhibit DSS penetration, or
   whether these components are activating signaling pathways related to
   mucins. Previous studies have shown the PC can be arranged at the
   interface between the luminal contents and the mucus layer to confer
   protection against bacterial translocation [[155]63]. However, the MFGM
   and its components are vastly complex, which could lead to the
   activation of a multitude of signaling pathways. We used a predictive
   approach to identify potential pathway targets for the major digested
   phospholipid components in the context of colitis and mucin. The top
   enriched KEGG pathways, such as PI3K-Akt signaling, Ras signaling,
   HIF-1 signaling, and the EGFR tyrosine kinase inhibitor resistance
   pathway, are associated with cell survival, proliferation, and
   differentiation. Jiang and colleagues also identified the PI3K-Akt
   signaling pathway as a target of the MFGM using immunoblotting
   techniques [[156]39]. Furthermore, Wu and colleagues used a gene
   ontology (GO) pathway analysis of differentially expressed genes
   between DSS control mice and DSS + prophylactic MFGM-treated mice,
   which also found an upregulation in pathways involved in epithelial
   cell proliferation [[157]29]. The other enriched pathways in our
   predictive analysis, including sphingolipid signaling and endocrine
   resistance, have a direct role in lipid metabolism. At the center of
   these pathways, Ca^2+ signaling serves as a potent secondary messenger,
   coordinating the action of a multitude of pathways [[158]64].
   Intracellular Ca^2+ also regulates mucin secretion. In goblet cells,
   mucin secretion in response to external and internal stimuli is
   regulated by intracellular Ca^2+ levels by which mucin vesicle contents
   are released from the cell through fusion with the plasma membrane,
   resulting in up to a 1000-fold volumetric expansion [[159]49]. Based on
   this preliminary predictive approach and the evidence of the MFGM’s
   action on mucin protein, this study lays the foundation for future work
   to investigate the role of Ca^2+-mediated signaling by the MFGM and MPL
   to regulate mucin secretion.

5. Conclusions

   In this study, we examined the effects of two digested MFGM-based
   ingredients with varying phospholipid content on mucin protein using
   Caco-2/HT29-MTX cells with or without DSS-induced barrier disruption.
   Both in vitro digested MFGM and MPL fortified the mucosal barrier to
   protect against DSS-induced loss in mucin production; however, marginal
   effects were observed for these ingredients on mucin protein under no
   challenge conditions. These findings were further corroborated through
   measurements of FD4 permeability where the MFGM ingredients prevented
   DSS-induced increases in permeability and by SEM and PAS/Alcian blue
   staining. Using a predictive bioinformatic approach, these MFGM
   ingredients appear to influence calcium signaling pathways, which is
   directly involved in mucin secretion. The findings from this work
   warrant further exploration of the MFGM to reduce intestinal barrier
   damage using more complex models such as intestinal organoids that more
   closely mimic the heterogeneity of the intestinal epithelium and in
   vivo models, as a major limitation of this study is the lack of
   interaction with the gut microbiota.

Acknowledgments