Milk osteopontin promotes intestinal development by up‐regulating the expression of integrin αvβ3 and CD44

Osteopontin (OPN) is a pleiotropic protein involved in numerous biological processes such as cell proliferation and differentiation. Since OPN is abundantly present in milk and is known to be relatively resistant to in vitro gastrointestinal digestion, the current study aimed to investigate the roles of oral intake of milk OPN in intestinal development using an established OPN knockout (KO, OPN−/−) mouse model, in which wild‐type (WT, OPN+/+) mouse pups were nursed by either WT (OPN+/+OPN+ group) or OPN KO dams (OPN+/+OPN− group; +/+ indicates genotype and − indicates milk without OPN), receiving milk with or without OPN from postnatal days 0 to 21 (P0−P21). Our results showed that milk OPN is resistant to in vivo digestion. Compared to OPN+/+OPN− pups, OPN+/+OPN+ pups at P4 and P6 had significantly longer small intestines, at P10 and P20 had larger inner jejunum surfaces, and at P30 exhibited more mature/differentiated intestines, as revealed by higher activities of alkaline phosphatase in brush border and more goblet cells, enteroendocrine cells, and Paneth cells. qRT‐PCR and immunoblotting results showed that milk OPN increased the expression of integrin αv, integrin β3, and CD44 in jejunum of mouse pups (P10, P20, and P30). Immunohistochemistry analysis showed that both integrin αvβ3 and CD44 are localized in jejunum crypts. In addition, milk OPN increased the phosphorylation/activation of the ERK, PI3K/Akt, Wnt, and FAK signaling pathways. In summary, oral intake of milk OPN in early life promotes intestinal proliferation and differentiation by upregulating the expression of integrin αvβ3 and CD44 and thus regulates OPN‐integrin αvβ3 and OPN‐CD44 mediated cellular signaling pathways.


| INTRODUCTION
Osteopontin (OPN) is a phosphorylated glycoprotein. It is an intracellular protein 1 and is also secreted by various cell types, appearing in most body fluids including milk. 2 OPN is multifunctional and involved in a wide range of bioactivities including cell proliferation and differentiation, mineralization, and immunomodulation. 3,4 OPN contains arginine-glycine-aspartate (RGD)-dependent and -independent integrin-binding sites at its N-terminus and CD44-binding domains at its C-terminus, respectively, 5 and it exerts its pleiotropic functions by binding to integrin and/ or CD44 to activate cellular signaling pathways. 6 CD44 and multiple integrins including αvβ3 have been reported to serve as OPN receptors and mediate functions of OPN. [7][8][9][10] OPN is present at a high concentration in human milk (~170 mg/L in colostrum and ~130 mg/L in transitional milk) 11 and mouse milk (~150 mg/L), 12 but its concentration is much lower in bovine milk (~18 mg/L). 13 OPN has been shown to be partly resistant to in vitro gastric digestion 14 and gastrointestinal digestion. 15 The early postnatal period is a critical time for intestinal development 16 ; proper establishment of the small intestine epithelium and crypts during early life influences the responsiveness of intestine to dietary and pathological challenges later in life. 17 Due to the appearance of abundant OPN in milk and the resistance of milk OPN to gastrointestinal digestion, milk OPN may contribute significantly to intestinal development in early infancy. In a study on infant rhesus monkeys fed regular formula or formula with supplemental bovine milk OPN (125 mg/L), the intestinal transcriptome in the bovine milk OPN supplemented infant monkeys was more similar to that of breast-fed rhesus infants. 18 In a preterm piglet study, oral supplementation of bovine milk OPN (46 mg/kg/day) from P1 to P19 increased the villus to crypt ratio across the small intestine. 19 To investigate the functions of milk OPN, we have previously established an OPN knockout (KO) mouse model, in which OPN wild-type (WT, OPN +/+ ) mouse pups are nursed by either OPN wild-type (WT, OPN +/+ ) dams (OPN +/+ OPN+ group) or OPN knockout (KO, OPN −/− ) dams (OPN +/+ OPN− group), receiving milk with or without a high concentration of OPN. 12 In the current study, effects of oral intake of milk OPN during early life on intestinal development were examined using this mouse model.

| Animals
All experiments and animal procedures were conducted according to the NIH guidelines for the care and use of laboratory animals, under a protocol approved by the Institutional Animal Care and Use Committee of the University of California, Davis. OPN KO mice [B6.129S6 (Cg)-Spp1tm1Blh/J, 8-10 weeks old, Jackson Laboratory] and WT mice (C57BL/6J, 8-10 weeks old, Jackson Laboratory) were allowed to become pregnant, deliver, and nurse. All breeders used in this study were 2-4 months old. Mice were housed in groups of 4 animals per cage and maintained in a humidity (30%-50%) and temperature (22 ± 1°C) controlled room under a 12 h/12 h light/dark cycle (lights on 7 am-7 pm). Water and food were supplied ad libitum. Pregnant mice were placed in separate cages. Following delivery, the newborn WT pups and KO pups (postnatal day 0, P0) were nursed by either OPN KO dams or WT dams. Each litter was adjusted to 8 pups. Offspring were weaned at P21. Pups from at least three different litters were used for each experiment. No significant gender differences were found in experiments in this study. Animals were gender matched within each experiment, and the data presented represent a combination of results from matched mice.

| In vivo digestion
P12 mouse pups were used for the in vivo OPN digestion experiment because mouse pups usually start eating solid food at P13. P12 KO mouse pups that were previously nursed by KO dams were fasted for 6 h then nursed by WT dams for 1 h. Thirty minutes later, the pups were killed by decapitation. Stomach and intestine contents were collected and stored at −80°C until immunoblotting analysis.

| Length of the small intestine and histological analysis (hematoxylin and eosin stain)
At P4, P6, P8, P10, P20, and P30, pups were weighed, killed by decapitation or CO 2 inhalation, and length of their small intestines was measured. Jejunum samples were collected at P10, P20, and P30 according to a method described previously, 20 fixed in paraformaldehyde [4% in phosphate-buffered saline (PBS)] overnight, embedded in paraffin, and sectioned on a microtome. Sections (6 μm) were deparaffinized and rehydrated using standard methods. Sections were stained with hematoxylin and eosin (Sigma-Aldrich) using a general protocol. 21 Villous height and crypt depth in images (20× fields) were determined using Image J. 22 Measurements of villous height and crypt depth were made under blind conditions.

| Measurement of alkaline phosphatase (ALP) activity
Differentiation of intestinal epithelial cells was determined by measuring the activity of ALP, a known intestinal differentiation marker. 23 Brush border was isolated from the duodenum and jejunum of P30 mouse pups. Activity of ALP was measured using pNPP (Sigma-Aldrich) as a substrate.

| Immunohistochemistry analysis
Mouse pups (P6, P20, and P30) were killed by decapitation or CO 2 inhalation, jejunum samples were collected, fixed in paraformaldehyde (4% in PBS) overnight, embedded in paraffin, and sectioned on a microtome. Sections (6 μm) were deparaffinized and rehydrated using standard methods. A Vectastain ABC HRP Kit (Vector Laboratories) was used according to the manufacturer's instructions. Crypt cells at P6, enteroendocrine cells at P30, and Paneth cells at P30 were identified by binding of antibodies to Ki67, synaptophysin, and lysozyme, respectively (Abcam). Goblet cells at P30 were recognized on the basis of the reactivity of their cytoplasm with Alcian blue and by their shape. Expression and localization of OPN receptors (integrin αvβ3 and CD44) in P20 mouse pups were examined by incubation with antibodies (anti-integrin αvβ3, Santa Cruz biotechnology; anti-CD44, Cell Signaling Technology). Specificity of the staining was confirmed by omission of the primary antibodies (data not shown). Positively stained cells in immunohistochemistry (IHC) images (20× or 40× fields) were evaluated with ImageJ. RNA (1 μg) was reverse transcribed to cDNA with a High-Capacity cDNA Reverse Transcription Kit (Thermo Fisher Scientific) according to the manufacturer's instructions. Gene-specific primers are listed in Table 1. Glyceraldehyde-3-phosphate dehydrogenase (GAPDH) was used as internal control. qRT-PCR was performed on the cDNA reaction mixture (2 μL) and Sybr Green (Bio-Rad) using the iCycler Real-Time PCR System (CFX 96; Bio-Rad). The cycling parameters were 95°C for 15 min, and 40 cycles including 95°C for 15 s, 60°C for 30 s, and 72°C for 30 s. For OPN +/+ OPN+ samples, values are shown as mean fold changes ± SD, relative to OPN +/+ OPN− samples (set to 1).

| Immunoblotting
Jejunum samples (P10, P20, and P30) were collected, immediately snap frozen, and stored at −80°C until analysis. They were then homogenized with a polytron homogenizer in RIPA buffer (Cell Signaling Technology) with 1× halt protease and phosphatase inhibitor cocktail (Thermo Fisher Scientific) and centrifuged at 5000 g for 10 min at 4°C to collect the supernatant. Protein concentration was measured by the Bradford assay. Proteins (20 μg/lane) were separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and transferred to nitrocellulose membranes for 1 h at 180 V using the Trans-Blot Turbo Transfer System (Bio-Rad). Membranes were blocked with blocking buffer [3% bovine serum albumin in TBST (Tris buffered saline containing 0.1% Tween-20)] for 45 min at room temperature and then incubated with primary antibodies against mouse OPN (R&D Systems), OPN receptors (integrin αv, integrin β3, and CD44, Cell Signaling Technology), components of cell signaling pathways and an internal control (anti-P-ERK, anti-T-ERK, anti-P-Akt, anti-T-Akt, anti-P-catenin, anti-T-catenin, anti-GAPDH, Cell Signaling Technology) overnight at 4°C. After three washes with TBST, primary antibodies were detected with horseradish peroxidase (HRP)conjugated donkey anti-rabbit IgG (Amersham Pharmacia Biotech, Piscataway, NJ) or HRP-conjugated goat anti-mouse Effects of OPN on expression of OPN receptors were evaluated by immunoblotting as described above.

| Statistical analysis
Each experiment was repeated 2-3 independent times, and average parameters from these sets of experiments were determined and compared by Student's t test or one-way anova using GraphPad Prism 9 (GraphPad Software Inc.). A value of p < .05 was considered statistically significant.

| Workflow diagram for the experimental design
As shown in Figure 1, two in vivo mouse models, OPN +/+ mouse pups nursed by either OPN +/+ or OPN −/− dams and OPN −/− mouse pups nursed by either OPN +/+ or OPN −/− dams, were used to determine effects of oral intake of milk OPN on intestinal development in early life. Furthermore, effects of milk OPN on expression of integrin αvβ3 and CD44 in human intestinal epithelial cells were evaluated using HIECs.
F I G U R E 1 Workflow diagram. In vivo and in vitro models were used to evaluate effects of milk OPN on intestinal development.

| Milk OPN is partly resistant to in vivo digestion
OPN KO mouse pups were used for the in vivo digestion assay to eliminate effects from endogenous OPN. The specificity of the mouse OPN antibody was evaluated using milk samples from OPN +/+ and OPN −/− dams, respectively. As shown in Figure 2A, this antibody specifically probes OPN in mouse milk from OPN +/+ dams. Both intact milk OPN and multiple OPN fragments (~20-50 kDa) appeared in contents of the stomach and small intestine, suggesting that partly digested milk OPN may contribute to intestinal development in early life ( Figure 2B).

| Milk OPN stimulates growth of the small intestine
The full length of the small intestine of mouse pups was measured from P4 to P30. As shown in Figure 3A, OPN +/+ OPN+ mouse pups showed significantly higher ratios of small intestine length to body weight than did OPN +/+ OPN− mouse pups at P4 and P6, indicating that P4 and P6 OPN +/+ OPN+ mouse pups have longer small intestines than do OPN +/+ OPN− mouse pups. Furthermore, the inner surface area of the small intestine was evaluated by the ratio of villus height to crypt depth. Compared to OPN +/+ OPN− mouse pups, P10 and P20 OPN +/+ OPN+ mouse pups had a significantly higher ratio of villus height to crypt depth, indicating that milk OPN plays a role in enhancing the inner surface area ( Figure 3B). As shown in Figure 3A,B, no differences in intestinal growth were found between P30 OPN +/+ OPN+ and P30 OPN +/+ OPN− mouse pups. Since the mouse pups were weaned at P21, direct effects of milk OPN on the intestine did not persist in P30 mouse pups, suggesting that these effects of milk OPN on proliferation occur only during the nursing period. To confirm the effects of milk OPN on intestinal proliferation, jejunum samples from P6 OPN +/+ OPN+ and OPN +/+ OPN− mouse pups were collected for immunohistochemistry analysis and stained with anti-Ki67, a proliferation marker ( Figure 4). As shown in Figure 4, significantly more intense Ki67 staining was found in P6 OPN +/+ OPN+ mouse pups than in P6 OPN +/+ OPN− mouse pups.

| Milk OPN promotes differentiation of intestinal epithelial cells
To determine the long-term (post-weaning) effect of milk OPN on the small intestine, effects of milk OPN on intestinal differentiation were determined using P30 mouse pups. ALP was used as an intestinal differentiation marker. As shown in Figure 5A, ALP activity was significantly higher in the brush border of duodenum and jejunum from P30 OPN +/+ OPN+ mouse pups, suggesting that milk OPN increases intestinal differentiation. Intestinal differentiation was further examined by identifying different cell types using histological or IHC F I G U R E 2 Milk OPN is partly resistant to in vivo digestion. (A) Specificity of the mouse OPN antibody was evaluated using mouse milk from OPN +/+ and OPN −/− dams. (B) Resistance of milk OPN to gastrointestinal digestion was assessed using contents from the stomach and small intestine. After P12 OPN KO pups were fasted for 6 h, they were nursed by WT dams for 1 h. Thirty minutes later, stomach and intestinal contents were isolated for immunoblotting analysis. SI1, SI2, and SI3 represent contents from three equal length sections of small intestine. S represents content from the stomach. SI1 is the segment adjacent to the stomach.
analysis. Goblet cells were identified by staining with Alcian blue, enteroendocrine cells were detected by a synaptophysin antibody, and Paneth cells were detected by a lysozyme antibody. Per mm 2 significantly more goblet cells ( Figure 5B), enteroendocrine cells ( Figure 5C), and Paneth cells ( Figure 5D) were found in P30 OPN +/+ OPN+ mouse pups than in P30 OPN +/+ OPN− mouse pups. Results indicate that milk OPN promotes differentiation of the small intestine.

| Milk OPN may enhance intestinal development by increasing levels of integrin αvβ3 and CD44
Subsequently, we designed experiments to investigate the mechanism by which milk OPN promotes intestinal proliferation and differentiation. Since OPN exerts its multiple functions by binding to its receptors and then activating signaling pathways, we tested effects of milk OPN on transcription of OPN receptors including integrin αv, integrin β3, and CD44 in duodenum and jejunum. As shown in Figure 6A,B, compared to OPN +/+ OPN− pups, OPN +/+ OPN+ pups showed significantly increased transcription of these OPN receptors from P4 to P20. At P30 no significant differences between OPN +/+ OPN+ and OPN +/+ OPN− pups were observed probably because mouse pups were weaned at P21, which resulted in diminishing effects of milk OPN on those OPN receptors. Immunoblotting was performed next to determine whether the protein levels of these OPN receptors were increased as well. As shown in Figure 7A, milk OPN significantly increased protein levels of all three OPN receptors in the jejunum at P10, P20, and P30. To identify whether the upregulation of the OPN receptor protein level was caused by dietary OPN directly, OPN KO pups (OPN −/− ) were nursed by either WT dams (OPN −/− OPN+) or OPN KO dams (OPN −/− OPN−) for 20 days and then expression of the three OPN receptors in jejunum was evaluated. As shown in Figure 7B, the protein levels of these three OPN receptors were remarkably elevated in OPN KO pups nursed by WT dams, which indicates that oral intake of dietary (milk) OPN directly increases the expression of the OPN receptors. Additionally, human intestinal epithelial cells, HIECs, were treated with human milk OPN and then immunoblotting was conducted. As shown in Figure 7C, human milk OPN treatment increased protein levels of the OPN receptors.
Expression and localization of integrin αvβ3 and CD44 in the intestine was assessed using IHC. As shown in  Figure 8A,B, integrin αvβ3 and CD44 were localized at jejunal crypts, and more integrin αvβ3 and CD44 stained cells appeared in the crypts of P20 OPN +/+ OPN+ mouse pups than in P20 OPN +/+ OPN− mouse pups.

| Milk OPN may promote intestinal development by activating signaling pathways
To further investigate how milk OPN regulates intestinal development in early life, immunoblotting was used to evaluate the expression of three signaling pathways related to cell proliferation and differentiation, the ERK, PI3K/Akt, and Wnt signaling pathways, as well as expression of focal adhesion kinase (FAK), which is known to regulate integrin-mediated signaling pathways. 24,25 Compared to OPN +/+ OPN− mouse pups, OPN +/+ OPN+ pups exhibited higher levels of activated components of these four cell signaling pathways (Figure 9), suggesting that milk OPN increases intestine development by upregulating the expression of its receptors and thereby increasing OPN-integrin αvβ3 and OPN-CD44-mediated cell signaling pathways.

| DISCUSSION
In the present study, milk OPN was found to be relatively resistant to in vivo digestion. Both full length OPN and OPN fragments (~20-50 kDa) were observed in the stomach and small intestine (Figure 2). In previous studies, only small OPN fragments (<48 kDa) were seen after in vitro simulated gastrointestinal digestion of human and bovine milk OPN. 15,26 It is possible that different in vivo and in vitro transit time and the ratio of enzyme to substrate resulted in somewhat different extent of proteolysis. OPN is a multifunctional protein that exerts its pleiotropic functions by binding to its receptors on the cell membrane. A recent study found that bovine milk OPN peptides that survived in vitro simulated gastrointestinal digestion were able to bind to the integrin αvβ3 receptor on the cell membrane of human MDA-MB-435 cells. 27 Therefore, milk OPN may contribute to intestinal development in infancy.
Milk OPN dynamically and variously regulates intestinal development at different stages. Milk OPN concentrations change dynamically throughout lactation in milk from both human and mouse. 11,12,28 Consistently, milk OPN variously influences intestinal proliferation F I G U R E 8 Localization of integrin αvβ3 (A) and CD44 (B) in the intestine. P20 jejunum samples from OPN +/+ OPN+ and OPN +/+ OPN− mouse pups were fixed, sectioned, and stained with anti-integrin αvβ3 (A) and anti-CD44 (B). Data are means ± SD. n = 8. Bars with * are significantly different (p < .05). Scale bars, 25 μm. and differentiation at different developmental stages. At P4 and P6, OPN +/+ OPN+ mouse pups showed longer small intestines than OPN +/+ OPN− mouse pups, and at P6 more ki67 in jejunum crypts of OPN +/+ OPN+ mouse pups indicated greater cell proliferation. Additionally, a large inner surface area was found in OPN +/+ OPN+ mouse pups at P10 and P20. Pups were weaned at P21. At P30, the jejunums of OPN +/+ OPN+ mouse pups appeared more differentiated than those of OPN +/+ OPN− mouse pups, as revealed by higher ALP activity and more differentiated cells including Paneth cells, enteroendocrine cells, and goblet cells in jejunum. Since there is a significant and immediate demand on the gastrointestinal tract to digest and absorb nutrients efficiently to maintain the high rate of growth in infancy, the positive effects of milk OPN on intestinal development may have long-term impact on the healthy development of infants.
It has previously been reported that OPN promoted transcription of OPN receptors including integrin αv, β1, β3, β5, and β8 when oligodendrocyte progenitor cells derived from human embryonic stem cells were treated with recombinant OPN. 29 Similarly, in studies primarily using cancer cell lines such as the breast cancer cell line 21NT, 30 liver carcinoma cell line HepG2, 31 melanoma cells, 32 and macrophages, 33 OPN was shown to increase the expression of CD44. Our study showed that milk OPN stimulates intestinal development by upregulation of the functional integrin receptors αvβ3 and CD44. The markedly higher levels of integrin αvβ3 and CD44 in jejunum of OPN KO pups nursed by OPN WT dams compared to OPN KO pups nursed by OPN KO dams indicate direct effects from oral intake of milk OPN. At P10 and P20, milk OPN significantly increased the transcription of αv, β3, and CD44, and at P10, P20, and P30, milk OPN increased the expression of αv, β3, and CD44. Both integrin αvβ3 and CD44 are localized in jejunal crypts (Figure 8), so it is likely that intact and partly digested milk OPN binds to integrin αvβ3 and CD44 on the surface of crypt cells and triggers cellular signaling to promote intestinal proliferation and differentiation. Integrins are large heterodimeric transmembrane receptors comprised α and β subunits that interact non-covalently to form different heterodimeric receptors that respond to various ligands. 24 Integrins permit dynamic bidirectional transmembrane signaling that is essential in cell adhesion, migration, differentiation, and survival. 34 Interestingly, integrin αvβ3 also plays roles in activation of innate responses against virus infection. 35 OPN interacts with integrin αvβ3 through functional RGD-dependent and/or RGD-independent cell binding sequences in its N-terminus. 36,37 CD44 is a cell surface adhesion receptor, involved in cellular signaling cascades which mediate growth, survival, F I G U R E 9 Effects of milk OPN on activation of ERK, PI3K/Akt, Wnt, and FAK signaling pathways. Jejunum samples from P10, P20, and P30 OPN +/+ OPN+ and OPN +/+ OPN− samples were analyzed by immunoblotting. Data are means ± SD. n = 8. Bars with * are significantly different (p < .05). proliferation, differentiation, and migration. 38 Other than regular CD44, variant CD44 isoforms have been found in numerous cancers, and they play roles in progression of cancer cells. 39 The primary domains of CD44 are the extracellular domain (or ectodomain), the transmembrane domain, and the intracellular domain/cytoplasmic domain. 40 OPN has CD44-binding domains at its C-terminus, which bind to the extracellular domain of CD44. OPN fragments generated by gastrointestinal digestion may interact with its receptors if the fragments contain OPN receptor binding sites.
Milk OPN increases phosphorylation of signaling pathways related to intestinal development and integrin functions. The ERK, PI3K/Akt, and Wnt signaling pathways play important roles in intestinal development, [41][42][43] and FAK mediates integrin-involved signaling pathways. Both OPN receptors, integrin and CD44, are involved in a variety of biological processes including cell proliferation, differentiation, and migration by activating signaling pathways. 44,45 Our results show that phosphorylation of ERK, PI3K/Akt, Wnt, and FAK signaling pathways is enhanced by milk OPN-integrin αvβ3 and OPN-CD44 interactions, implying that milk OPN promotes intestinal development by increasing the expression of integrin αvβ3 and CD44 and thus activates multiple cell signaling pathways mediated by OPNintegrin αvβ3 and OPN-CD44.
In summary, milk OPN is relatively resistant to gastrointestinal digestion and promotes intestinal development in infancy by elevating the expression of integrin αvβ3 and CD44 and then by triggering ERK, PI3K/Akt, Wnt, and FAK signaling pathways.

AUTHOR CONTRIBUTIONS
Rulan Jiang, Christine Prell, and Bo Lönnerdal conceived the study; Rulan Jiang, Jamie Lo, and Christine Prell performed the investigations; and Rulan Jiang, Jamie Lo, Christine Prell, and Bo Lönnerdal wrote the manuscript.