Visual Abstract

Lactoferrin (LF) is known to possess anti-inflammatory activity, although its mechanisms of action are not well-understood. The present study asked whether LF affects the commitment of inducible regulatory T cells (Tregs). LF substantially promoted Foxp3 expression by mouse activated CD4+T cells, and this activity was further enhanced by TGF-β1. Interestingly, blocking TGF-β with anti–TGF-β Ab completely abolished LF-induced Foxp3 expression. However, no significant amount of soluble TGF-β was released by LF-stimulated T cells, suggesting that membrane TGF-β (mTGF-β) is associated. Subsequently, it was found that LF binds to TGF-β receptor III, which induces reactive oxygen species production and diminishes the expression of mTGF-β–bound latency-associated peptide, leading to the activation of mTGF-β. It was followed by phosphorylation of Smad3 and enhanced Foxp3 expression. These results suggest that LF induces Foxp3+ Tregs through TGF-β receptor III/reactive oxygen species–mediated mTGF-β activation, triggering canonical Smad3-dependent signaling. Finally, we found that the suppressive activity of LF-induced Tregs is facilitated mainly by CD39/CD73-induced adenosine generation and that this suppressor activity alleviates inflammatory bowel disease.

Regulatory T cells (Tregs) are a unique population of suppressor T cells that maintain peripheral immune tolerance and homeostasis. In vivo, Tregs are generated in the thymus and extrathymically in peripheral tissues. In vitro, inducible Tregs (iTregs) can be differentiated from naive CD4+CD25 T cells upon TCR stimulation in the presence of TGF-β and IL-2 (1, 2). TGF-β/Smad–dependent signaling strongly induces Treg differentiation, which is characterized by the upregulation of the master transcription factor Foxp3 (3, 4). TGF-β–mediated Foxp3 expression can be enhanced with retinoic acid (RA) by increasing the expression of Smad3 and p-Smad3 (5) or by enhancing p-Smad3 binding to the conserved enhancer region of the Foxp3 gene (6). In addition, several factors, including vitamin D3, contribute to the upregulation and stabilization of Foxp3 expression (7).

Lactoferrin (LF) is an 80 kDa multifunctional iron-binding glycoprotein of the transferrin family. It is found in most mammalian exocrine secretions as well as in secondary granules of neutrophils. In addition to its well-known antimicrobial activity, LF has been shown to have potent anti-inflammatory effects (8). For example, LF suppresses IFN-γ production from human CD4+T cells (9), inhibits the proliferation of Th1 cells (10), and mitigates autoimmune diseases such as experimental autoimmune encephalomyelitis (11) and autoimmune hepatitis (12) by suppressing proinflammatory cytokine production. Peroral administration of LF alleviates colitis in rats by enhancing the secretion of the anti-inflammatory cytokine IL-10 and decreasing IFN-γ production (13); it also has been shown to reduce the severity of nosocomial sepsis by increasing Treg levels (14). Finally, a recent study reported that recombinant human LF resolves inflammatory bowel disease (IBD) by increasing Tregs (15). These observations suggest that LF may suppress effector T cells through the induction of Tregs, although the molecular mechanisms are still unclear. In a previous study, we demonstrated that LF was as potent as TGF-β1 in causing IgA isotype switching in mice (16). Thus, we explored whether LF could stimulate naive T cells to differentiate into Foxp3+ Tregs and its underlying mechanisms. In the current study, we found that LF, particularly together with TGF-β1, induced the Foxp3+CD4+T cells through binding to TGF-β receptor III (TβRIII) and in turn activation of membrane TGF-β (mTGF-β), which was mediated by the induction of reactive oxygen species (ROS) in T cells. Moreover, we found that LF/TGF-β–induced Foxp3+T cells alleviated the pathological severity in a mouse colitis model through CD39/CD73-induced adenosine generation.

BALB/c mice were obtained from Orient Bio (Seongnam, Korea). C57BL/6 wild-type mice, OVA-specific TCR-transgenic OT-II mice, and Rag1−/− mice were obtained from The Jackson Laboratory (Bar Harbor, ME). The animals were fed Purina Laboratory Rodent Chow 5001 ad libitum. Mice that were 8–12 wk of age were used in this study. Animal care was performed in accordance with the institutional guidelines set forth by Kangwon National University (approval no. KW-150619-2 and KW-190515-1).

Naive CD4+CD25 T cells from the spleens of 8–12-wk-old mice were purified by selection using naive CD4+ T cell isolation kits and magnetic cell sorting (Miltenyi Biotec, Auburn, CA), according to the manufacturer’s instructions (Fig. 1A-1). For differentiation of naive CD4+CD25 T cells into Treg cells, cells were activated by plate-bound anti-CD3e Ab (2 μg/ml), soluble anti-CD28 Ab (2 μg/ml, BD Pharmingen, San Diego, CA), and IL-2 (100 IU/ml, PeproTech, Rocky Hill, NJ) and treating cells with TGF-β1 (R&D Systems, Minneapolis, MN), RA (Sigma-Aldrich, St. Louis, MO), and bovine LF (Morinaga Milk, Zama, Japan), which contained fewer than 5.0 pg mg−g of LPS (endotoxin). In some cases, cells were treated with porcine TGF-β2, anti-pan TGF-β Ab, soluble TβRIII (sTβRIII), anti–low–density lipoprotein receptor–related protein (LRP) Ab, soluble LRP (R&D Systems), anti-LF Ab (AbFrontier, Seoul, Korea), N-acetyl-L-cysteine (NAC) (ROS scavenger) (Calbiochem, San Diego, CA), Sis3 (Smad3-specific inhibitor), and SB431542 (TβRI inhibitor) (Sigma-Aldrich). For in vitro Th type 1 cell differentiation, activated cells were cultured with IL-12 (2 ng/ml) (PeproTech) and anti–IL-4 Ab (10 μg/ml) (BD Pharmingen). For Th17 differentiation, cells were cultured with TGF-β1 (2 ng/ml), IL-6 (20 ng/ml), IL-23 (10 ng/ml), IL-21 (10 ng/ml), and IL-1β (10 ng/ml) (PeproTech). For Th2 cell differentiation, cells were treated with IL-4 (12.5 ng/ml) (PeproTech), anti–IFN-γ Ab (5 μg/ml), and anti–IL-10 Ab (5 μg/ml) (BD Pharmingen). For the induction of Ag-specific Tregs, responder CD4+ T cells (1 × 105) from OT-II transgenic mice were cocultured with OVA323–339–pulsed, irradiated splenoblasts (2 × 105) in the presence of TGF-β1 (1 ng/ml), LF (100 μg/ml), or both for 3 d. Radiation (2000 rad) was performed using a Gammacell 40 Exactor (Best Theratronics, Ottawa, Ontario, Canada).

Cell surface staining was performed in PBS containing 2% FBS using the following Abs: anti-mouse CD4-FITC, CD25-PE, CD44-PE, CD62L–allophycocyanin, CD39–allophycocyanin, CD73-PE, ICOS-PE, CTLA-4-PE, biotinylated anti–TGF-β, anti–latency-associated peptide (LAP) (R&D Systems), anti-mouse IgG–allophycocyanin, streptavidin–allophycocyanin (eBioscience, San Diego, CA), and streptavidin-FITC (SouthernBiotech). Intracellular staining was performed using Fixation and Permeabilization Buffer Sets (eBioscience) with anti–Foxp3-PE/Cy7 (Invitrogen Life Technologies, Carlsbad, CA), anti–IFN-γ–allophycocyanin (BioLegend, San Diego, CA), anti–IL-17A–allophycocyanin, and anti–IL-4–allophycocyanin Abs (eBioscience), according to the manufacturer′s instructions. For the detection of mitochondrial ROS, cells were stained with 2 μM MitoSOX Red (Invitrogen) for 10 min at 37°C in a cell culture incubator. For flow cytometry, data were collected on a FACSVerse flow cytometer (BD Biosciences, San Diego, CA) and analyzed with FlowJo software (Tree Star, Ashland, MA) using unstained controls to determine gating.

RNA preparation and reverse transcription were performed as described previously (17). All reagents for RT-PCR were purchased from Promega (Madison, WI). PCR primers were synthesized by BIONEER (Seoul, Korea) and are as follows: Foxp3 forward primer (F): 5′-CCCATCCCCAGGAGTCTTG-3′ and reverse primer (R): 5′-CCATGACTAGGGGCACT GTA-3′; RORgt F: 5′-CCGCTGAGAGGGCTTCAC-3′ and R: 5′-TGCAGGAGTAGGCCAC ATTACA-3′; IL-17 F: 5′-TCTCTGATGCTGTTGCTGCT-3′ and R: 5′-CGTGGAACGGTTG AGGTAGT-3′; T-bet F: 5′-ACCAACAACAAGGGGGCTTC-3′ and R: 5′-CTCTGGCTCTC CATCATTCACC-3′; IFN-γ F: 5′-ACACTGCATCTTGGCTTTGC-3′ and R: 5′-TGGACCT GTGGGTTGTTGAC-3′; GATA3 F: 5′-ACAGAAGGCAGGGAGTGTGTGAAC-3′ and R: 5′-TTTTATGGTAGAGTCCGCAGGC-3′; IL-4 F: 5′-ATATCCACGGATGCGAC AAA-3′ and R: 5′-AAGCCCGAAAGAGTCTCTGC-3′; Bcl-6 F: 5′-CCTGTGAAATCTGTGGCAC TCG-3′ and R: 5′-CGCAGTTGGCTTTTGTGACG-3′; and CXCR5 F: 5′-GACCTTCAAC CGTGCCTTTCTC-3′ and R: 5′-GAACTTGCCCTCAGTCTGTAATCC-3′. Relative mRNA expression levels were determined by quantitative real-time PCR using the THUNDERBIRD SYBR qPCR Mix (TOYOBO, Osaka, Japan) and the CFX 96 Touch RT-PCR Thermal Cycler (Bio-Rad Laboratories, Hercules, CA). PCR reactions for β-actin were performed in parallel for normalization within each set of samples. Data were analyzed with CFX Manager 3.1 software (Bio-Rad Laboratories).

Levels of TGF-β1 and TGF-β2 in supernatants were measured by using commercial ELISA kits, according to the manufacturer’s instructions (R&D Systems). Plates were read at 450 nm with a Versamax plate reader (Molecular Devices, Sunnyvale, CA).

The TGF-β bioassay was performed using the Mv1Lu cell line. Cells were plated at 5 × 103 cells/well in 96-well plates with 200 μl of DMEM containing 10% FBS and incubated for 16 h to ensure complete adherence. Cells were cocultured with LF/TGF-β–stimulated CD4+ T cells in the presence of anti-pan TGF-β Ab (10 μg/ml). Cell number was determined using a colorimetric MTT (Sigma-Aldrich) assay as previously described (18).

For the preparation of iTregs, freshly isolated naive CD4+CD25-T cells were activated with plate-bound anti-CD3 mAb, soluble anti-CD28 mAb (2 μg/ml), and IL-2 (100 IU/ml) in the presence of TGF-β1 (0.5 ng/ml), LF (100 μg/ml), or both for 3 d. Cells were violet labeled and cocultured with CFSE-labeled responder CD4+CD25 T cells (1 × 105) under the activation of anti-CD3/CD28 Ab for 3 d.

Suppression of allogeneic T cell proliferation was assessed using an MLR. CFSE-labeled BALB/c responder CD4+ T cells (5 × 104), irradiated C57BL/6 splenoblasts (allogenic stimulators), and LF/TGF-β–induced BALB/c CD4+Foxp3+T cells (5 × 104) were cocultured for 3 d. Dilution of CFSE was measured by flow cytometry, gating on 10,000 viable cells.

Adenosine levels in conditioned media were measured by an adenosine assay kit (BioVision, Milpitas, CA), according to the manufacturer’s protocol. Fluorescence intensity was measured at an excitation/emission of 535/587 nm by a Synergy H1 Hybrid Multi-Mode Microplate Reader (BioTek Instruments, Winooski, VT).

To induce colitis, C57BL/6 Rag1−/− mice were coinjected i.p. with syngeneic naive CD4+ CD25 T cells (5 × 105 cells/mouse) and LF/TGF-β1–induced CD4+Foxp3+ T cells (2.5 × 105 cells/mouse). Mice were observed daily and weighed weekly, and any mouse that lost >20% of its starting body weight or showed severe signs of disease was euthanized. Twelve weeks after transfer, mice were sacrificed, and their intestines were removed and transferred into cold PBS. To remove mucus and intraepithelial lymphocytes, fat-free intestines were incubated at room temperature with 1 mM DTT and 30 mM EDTA (Bio-Rad Laboratories) in PBS. Next, the intestines were digested with 0.05% collagenase solution (Worthington Biochemical, Lakewood, NJ) at 37°C for 90 min. Isolated cell suspensions were then passed through 40-μm cell strainers. The number of IFN-γ+ and IL-17+ T cells were analyzed by flow cytometry.

For dextran sulfate sodium (DSS)–induced colitis, mice were given 3% DSS (MP Biomedicals, Santa Ana, CA) in their drinking water for 9 d followed by regular drinking water. LF (5 mg/kg) was administered i.p. on days 0 and 3. Daily clinical evaluations, including the assessment of body weight, stool consistency, and detection of rectal bleeding, were conducted to generate a disease activity index score. To analyze histological changes, mice were sacrificed on day 10, and their colons were removed for assessment of colon length. The distal part of the colon was fixed in 10% formalin, embedded in paraffin, and cut into 2-mm sections. The colon sections were deparaffinized and stained with H&E. The histological score of DSS colitis was determined as the sum of individual scores for inflammatory cell infiltration and tissue damage as previously described (19). The numbers of CD4+Foxp3+ T cells in each colon sample were determined by flow cytometry.

Statistical differences between experimental groups were determined by ANOVA, and p values < 0.05 by unpaired, two-tailed Student t test were considered significant.

TGF-β converts naive CD4+CD25 T cells into Foxp3+ Treg cells in the periphery and potently synergizes with RA (20, 21). Because we previously demonstrated that LF and RA, like TGF-β, increases IgA isotype switching (16, 22), we explored the effect of LF on Foxp3+ Treg cell conversion of naive CD4+CD25 T cells. TGF-β1 and RA synergized to induce Foxp3 expression as shown before, whereas LF alone substantially induced Foxp3 expression; this was further enhanced by TGF-β1 but not by RA (Fig. 1A-2). Although Foxp3+ T cells consistently expressed CD25 regardless of stimulation conditions (Fig. 1A-2), further phenotypic analysis was needed to better understand how LF and TGF-β1 induced Foxp3 expression. The majority of LF-induced Foxp3+ T cells retained CD62L (L-selectin), whereas the majority of TGFβ1-induced Foxp3+ T cells shed it (Fig. 1A-3, dotted circles). Interestingly, LF in combination with TGF-β1 markedly increased the frequencies of both Foxp3+ T cell populations (CD62L+Foxp3+ and CD62LFoxp3+ T cells). Finally, CD44, a marker for effector-memory T cells, was upregulated by TGF-β1 but downregulated by LF (Fig. 1A-3). Thus, LF and TGF-β1 had opposite effects on CD62L and CD44 but equally enhanced Foxp3 expression. Individually, LF and TGF-β1 stimulated naive CD4+ T cells to differentiate into Foxp3+CD62L+CD44lo and Foxp3+CD62LCD44hi T cells, respectively, but together, they strongly induced both Foxp3+CD62L+CD44lo and Foxp3+CD62LCD44lo T cells.

FIGURE 1.

Effects of TGF-β1 and LF on Foxp3+ T cell differentiation. (A) Freshly isolated naive CD4+ T cells (A-1) were activated with plate-bound anti-CD3 mAb, soluble anti-CD28 mAb (2 μg/ml), and IL-2 (100 IU/ml) in the presence of TGF-β1 (0.5 ng/ml), LF (100 μg/ml), and RA (25 nM). After 3 d of culture, the expression of CD25, CD62L, CD44, and Foxp3 were analyzed by flow cytometry (A-2 and A-3). (B-1) Naive CD4+ T cells were cultured as described in (A) for 2 d. Transcriptional levels of Foxp3 and other T cell subset markers were determined by real-time PCR. (B-2) Naive CD4+ T cells were cultured as in (A) for 3 d. Intracellular expression of IL-17, IFN-γ, IL-4, and Foxp3 were determined using flow cytometry following activation with PMA and ionomycin for 4 h. (C) OT-II CD4+ T cells were cocultured with OVA323–339–pulsed splenoblasts in the presence of TGF-β1 (1 ng/ml) and LF (100 μg/ml). After 3 d of culture, cells were analyzed for surface CD4 and intracellular Foxp3 using flow cytometry. The data represent the average percentage and fluorescence intensity of triplicate samples; error bars represent SEM. *p < 0.05. MFI, mean fluorescence intensity; NS, not significant.

FIGURE 1.

Effects of TGF-β1 and LF on Foxp3+ T cell differentiation. (A) Freshly isolated naive CD4+ T cells (A-1) were activated with plate-bound anti-CD3 mAb, soluble anti-CD28 mAb (2 μg/ml), and IL-2 (100 IU/ml) in the presence of TGF-β1 (0.5 ng/ml), LF (100 μg/ml), and RA (25 nM). After 3 d of culture, the expression of CD25, CD62L, CD44, and Foxp3 were analyzed by flow cytometry (A-2 and A-3). (B-1) Naive CD4+ T cells were cultured as described in (A) for 2 d. Transcriptional levels of Foxp3 and other T cell subset markers were determined by real-time PCR. (B-2) Naive CD4+ T cells were cultured as in (A) for 3 d. Intracellular expression of IL-17, IFN-γ, IL-4, and Foxp3 were determined using flow cytometry following activation with PMA and ionomycin for 4 h. (C) OT-II CD4+ T cells were cocultured with OVA323–339–pulsed splenoblasts in the presence of TGF-β1 (1 ng/ml) and LF (100 μg/ml). After 3 d of culture, cells were analyzed for surface CD4 and intracellular Foxp3 using flow cytometry. The data represent the average percentage and fluorescence intensity of triplicate samples; error bars represent SEM. *p < 0.05. MFI, mean fluorescence intensity; NS, not significant.

Close modal

Because LF upregulated the expression of Foxp3, a canonical Treg transcription factor, we assessed its effect on other transcription factors and cytokines associated with CD4+ T cell subsets. LF substantially increased transcriptional levels of Foxp3 and RORγt but not those of other canonical transcriptional factors and cytokines (Fig. 1B-1), such as intracellular levels of IFN-γ, IL-17, and IL-4 (Fig. 1B-2). We next examined whether the effect of LF on Foxp3 T cell differentiation extended to Ag-specific Foxp3 T cell differentiation. LF alone or combined with TGF-β1 converted naive OT-II CD4+CD25 T cells into Foxp3+ Treg cells in response to their cognate Ag OVA323–339 (Fig. 1C).

We further examined structural aspects of LF to gain more insight into its activity. LF is classified into apo, holo, and native forms, depending on iron saturation. All three types of LF, but not LF hydrolysate or recombinant LF, which are structurally distinct from native LF, induced the expression of Foxp3 (Supplemental Fig. 1A). This suggests that the effect of LF on Foxp3 expression depends on protein structure but not on iron saturation. Moreover, LF directly suppressed the proliferation of activated CD4+ T cells, indicating that the effect of LF on Foxp3 induction may be related to its ability to directly inhibit CD4+ T cell expansion (Supplemental Fig. 1B).

One of the most plausible underlying mechanisms of LF-facilitated Foxp3 expression is that it stimulates naive T cells to produce TGF-β or activates latent TGF-β. Indeed, anti-pan TGF-β Ab completely abrogated LF-facilitated Foxp3 expression (Fig. 2A-1), suggesting that LF causes T cells to produce TGF-β or activate latent TGF-β. Thus, we measured the levels of two forms of active TGF-β (TGF-β1 and TGF-β2) in cell culture supernatants (Fig. 2B). Both TGF-β1 and TGF-β2 were detected at low levels (<15 pg/ml for TGF-β1 and up to 20 pg/ml for TGF-β2). We then determined whether these minute concentrations of TGF-β could induce Foxp3 expression. TGF-β1 (15 pg/ml) and TGF-β2 (30 pg/ml or fewer) did not result in Foxp3 induction (Supplemental Fig. 2C). We considered the possibility that the anti-pan TGF-β Ab was binding to LF or that the LF used in our studies was contaminated with TGF-β, accounting for our results. We found no significant binding between anti-pan TGF-β Ab and LF as determined by LF ELISA (Supplemental Fig. 2A), and neither TGF-β1 nor TGF-β2 was found in the purified LF preparation as determined by TGF-β ELISA (Supplemental Fig. 2B). Finally, anti-LF Ab abrogated Foxp3 induction and restored CD44 expression (Fig. 2A-2). Together, these data suggest that LF alone stimulates T cells to express Foxp3 and that this is mediated by TGF-β but not its soluble form. Thus, we hypothesized that mTGF-β may be involved in the activity of LF as mTGF-β is expressed by Tregs (2325). We observed that mTGF-β readily appeared once T cells were activated, regardless of stimulation by LF and TGF-β (Fig. 2C-1). mTGF-β is often found bound to LAP and becomes functional when LAP is dissociated from it (26). Interestingly, LF or TGF-β1 alone diminished LAP expression as opposed to Foxp3 expression and, together, caused an even more dramatic reduction in LAP (Fig. 2C-2). Consistent with this, LAP-associated mTGF-β was substantially diminished by the addition of LF with or without TGF-β1 (Fig. 2C-3).

FIGURE 2.

LF enhances Foxp3+ T cell differentiation through activation of mTGF-β. (A) Naive CD4+ T cells were pretreated with anti-pan TGF-β Ab (10 μg/ml) (A-1) or anti-LF Ab (2 μg/ml) (A-2) for 1 h and activated with plate-bound anti-CD3 mAb, soluble anti-CD28 mAb (2 μg/ml), and IL-2 (100 IU/ml) in the presence of LF (100 μg/ml) plus or minus TGF-β1 (0.5 ng/ml). After 3 d of culture, the expression of CD25, CD44, and Foxp3 were analyzed by flow cytometry. (B) Naive CD4+ T cells were cultured as in (Fig. 1A for the indicated time points; supernatants were then measured for TGF-β1 and TGF-β2 by ELISA. (C) Naive CD4+ T cells were cultured as in (Fig. 1A for 60 h followed by Foxp3, mTGF-β, and LAP analysis by flow cytometry (C-1–C-3). (D) Mv1Lu cells were cocultured with LF/TGF-β1–stimulated T cells plus or minus anti-pan-TGF-β Ab (10 μg/ml) for 48 h. Relative cell viabilities were measured by MTT assay. Data represent the mean of triplicate samples; error bars represent SEM. *p < 0.05, **p < 0.01.

FIGURE 2.

LF enhances Foxp3+ T cell differentiation through activation of mTGF-β. (A) Naive CD4+ T cells were pretreated with anti-pan TGF-β Ab (10 μg/ml) (A-1) or anti-LF Ab (2 μg/ml) (A-2) for 1 h and activated with plate-bound anti-CD3 mAb, soluble anti-CD28 mAb (2 μg/ml), and IL-2 (100 IU/ml) in the presence of LF (100 μg/ml) plus or minus TGF-β1 (0.5 ng/ml). After 3 d of culture, the expression of CD25, CD44, and Foxp3 were analyzed by flow cytometry. (B) Naive CD4+ T cells were cultured as in (Fig. 1A for the indicated time points; supernatants were then measured for TGF-β1 and TGF-β2 by ELISA. (C) Naive CD4+ T cells were cultured as in (Fig. 1A for 60 h followed by Foxp3, mTGF-β, and LAP analysis by flow cytometry (C-1–C-3). (D) Mv1Lu cells were cocultured with LF/TGF-β1–stimulated T cells plus or minus anti-pan-TGF-β Ab (10 μg/ml) for 48 h. Relative cell viabilities were measured by MTT assay. Data represent the mean of triplicate samples; error bars represent SEM. *p < 0.05, **p < 0.01.

Close modal

In this study, it was important to know whether mTGF-β associated with low levels of LAP (LAPlo–mTGF-β) is biologically functional. This was assessed by coculturing LF/TGF-β–stimulated T cells with Mv1Lu cells, which are known to be highly sensitive to the antiproliferative activity of TGF-β. As shown in (Fig. 2D, LF/TGF-β1–stimulated T cells substantially decreased the viability of Mv1Lu cells, which was comparable to control TGF-β1 (0.5 ng/ml). This antiproliferative activity of mTGF-β was abrogated by anti-pan TGF-β Ab, indicating that LAPlo–mTGF-β is biologically active and regulates Foxp3 expression in LF/TGF-β1–stimulated T cells.

Which receptor is responsible for mediating LF-induced Foxp3 expression? We previously showed that LF binds to TβRIII, leading to IgA isotype switching in murine B cells (16). Further, it was reported that TβRIII promotes TGFβ-dependent iTreg conversion in vitro (27). We explored the involvement of this binding interaction in LF-stimulated Foxp3 expression. LF-induced Foxp3 expression was abrogated by pretreatment of LF with sTβRIII but not by Ab against LRP, a key LF receptor, or soluble LRP (Fig. 3A-1). Further, LF-induced LAP reduction was reversed by sTβRIII (Fig. 3A-2). These results indicate that TβRIII is a key LF receptor that mediates LAPlo–mTGF-β–associated Foxp3 expression.

FIGURE 3.

LF activates a TβRIII→ROS→mTGF-β pathway to induce Foxp3+ T cell differentiation. (A) Naive CD4+ T cells were pretreated with sTβRIII, soluble LRP, or anti-LRP Ab (each 10 μg/ml) for 1 h and activated with plate-bound anti-CD3 mAb, soluble anti-CD28 mAb (2 μg/ml), and IL-2 (100 IU/ml) in the presence of LF (100 μg/ml) plus or minus TGF-β1 (0.5 ng/ml) for 3 d. CD44, Foxp3, and LAP expression were analyzed using flow cytometry (A-1 and A-2). (B) Activated naive CD4+ T cells were cultured with LF (100 μg/ml) in the presence of the ROS inhibitor NAC (2.5 nM) for 3 d (B-1) and sTβRIII (10 μg/ml) for 2 d (B-2). LAP, Foxp3, and ROS expression were analyzed using flow cytometry. (C) Naive CD4+ T cells were pretreated with Smad3 inhibitor Sis3 (2 μM) and TβRI inhibitor SB431542 (10 μM) for 1 h and cultured as in (A). The expression of CD44, LAP, and Foxp3 were assessed by flow cytometry. (D) Activated naive CD4+ T cells were cultured as in (B-2) for 30 min. Phosphorylated Smad3 (℗-Smad3) and actin were then examined by Western blot analysis. Data represent the mean of triplicate samples; error bars represent SEM. *p < 0.05, **p < 0.01. NS, not significant.

FIGURE 3.

LF activates a TβRIII→ROS→mTGF-β pathway to induce Foxp3+ T cell differentiation. (A) Naive CD4+ T cells were pretreated with sTβRIII, soluble LRP, or anti-LRP Ab (each 10 μg/ml) for 1 h and activated with plate-bound anti-CD3 mAb, soluble anti-CD28 mAb (2 μg/ml), and IL-2 (100 IU/ml) in the presence of LF (100 μg/ml) plus or minus TGF-β1 (0.5 ng/ml) for 3 d. CD44, Foxp3, and LAP expression were analyzed using flow cytometry (A-1 and A-2). (B) Activated naive CD4+ T cells were cultured with LF (100 μg/ml) in the presence of the ROS inhibitor NAC (2.5 nM) for 3 d (B-1) and sTβRIII (10 μg/ml) for 2 d (B-2). LAP, Foxp3, and ROS expression were analyzed using flow cytometry. (C) Naive CD4+ T cells were pretreated with Smad3 inhibitor Sis3 (2 μM) and TβRI inhibitor SB431542 (10 μM) for 1 h and cultured as in (A). The expression of CD44, LAP, and Foxp3 were assessed by flow cytometry. (D) Activated naive CD4+ T cells were cultured as in (B-2) for 30 min. Phosphorylated Smad3 (℗-Smad3) and actin were then examined by Western blot analysis. Data represent the mean of triplicate samples; error bars represent SEM. *p < 0.05, **p < 0.01. NS, not significant.

Close modal

We next were interested in understanding how LF leads to the dissociation of LAP from mTGF-β. It was recently reported that ROS activate mTGF-β in T cells (28, 29), and this potential mechanism was examined in the current study. Indeed, blockade of ROS activity by NAC significantly decreased LF-induced Foxp3 expression and restored LAP expression (Fig. 3B-1). Congruently, LF treatment increased ROS production in activated T cells, and this was abrogated by sTβRIII pretreatment (Fig. 3B-2). Finally, the Smad3 inhibitor Sis3 and the TGF-β receptor I (TβRI) inhibitor SB431542 completely abolished LF-mediated Foxp3 induction (Fig. 3C). Consistently, LF induced phosphorylation of Smad3, and this was also abrogated by sTβRIII pretreatment (Fig. 3D). Taken together, these results strongly suggest that LF binds to TβRIII on CD4+ T cells to increase ROS expression, which in turn reduces LAP expression and activates mTGF-β, leading to the subsequent phosphorylation of Smad3 and Foxp3 expression.

Having shown that LF and TGF-β1 markedly enhanced Foxp3 expression in CD4+T cells, we were interested in determining how Foxp3+ T cells suppress effector T cells. We first asked if LF/TGF-β1–induced Foxp3+ T cells regulated the proliferation of responder CD4+ T cells. LF-induced Foxp3+ T cells substantially suppressed responder CD4+ T cell proliferation induced by anti-CD3/CD28 Ab, and this suppressive activity was further augmented by TGF-β1 (Fig. 4A-1). The same was true for responder T cells activated with allogenic APCs (Fig. 4A-2). Further, as the number of LF/TGF-β–induced Foxp3+ T cells increased, the suppressive activity increased (Fig. 4A-3). These results clearly show that LF/TGF-β1–induced CD4+Foxp3+ T cells have a regulatory function similar to iTregs (1, 2).

FIGURE 4.

LF/TGF-β–induced Foxp3+T cells suppress responder T cell activation. (A) BALB/c Foxp3+ T cells were obtained by culturing activated CD4+ T cells with LF ± TGF-β1 as described in (Fig. 1A. Responder T cells were prepared by labeling BALB/c naive CD4+ T cells with CFSE and stimulating them with anti-CD3/CD28 Ab (A-1) or allogenic APCs (irradiated C57BL/6 splenoblasts) (A-2). Two cell populations were cultured for 3 d, and proliferation was determined by analyzing CFSE dilution using flow cytometry. A different number of LF/TGF-β1–stimulated Foxp3+ T cells (Tregs*) were cocultured with responder T cells (A-3). (B) Foxp3+ T cells and responder T cells were cocultured as in (A-1) with the addition of anti-pan TGF-β Ab (10 and 20 μg/ml). Proliferation was determined by analyzing CFSE dilution using flow cytometry. (C) Activated naive CD4+ T cells were treated with LF (50 μg/ml) and TGF-β1 (0.5 ng/ml) for 48 h, and the expression of ICOS, CD39, CD73, and Foxp3 were determined by flow cytometry. (D-1) CFSE-labeled naive CD4+ T cells were cocultured with Foxp3+ T cells, which were obtained by culturing activated CD4+ T cells with LF + TGF-β1 in the presence of the CD39/CD73 inhibitor ARL67165 (100 μM). Cell division was assessed after 48 h by analyzing CFSE dilution using flow cytometry. (D-2) Levels of adenosine in supernatants from activated T cells cultured with LF + TGF-β1 for 48 h were determined by measuring adenosine deaminase activity. Data represent the mean of triplicate samples; error bars represent SEM. *p < 0.05. NS, not significant.

FIGURE 4.

LF/TGF-β–induced Foxp3+T cells suppress responder T cell activation. (A) BALB/c Foxp3+ T cells were obtained by culturing activated CD4+ T cells with LF ± TGF-β1 as described in (Fig. 1A. Responder T cells were prepared by labeling BALB/c naive CD4+ T cells with CFSE and stimulating them with anti-CD3/CD28 Ab (A-1) or allogenic APCs (irradiated C57BL/6 splenoblasts) (A-2). Two cell populations were cultured for 3 d, and proliferation was determined by analyzing CFSE dilution using flow cytometry. A different number of LF/TGF-β1–stimulated Foxp3+ T cells (Tregs*) were cocultured with responder T cells (A-3). (B) Foxp3+ T cells and responder T cells were cocultured as in (A-1) with the addition of anti-pan TGF-β Ab (10 and 20 μg/ml). Proliferation was determined by analyzing CFSE dilution using flow cytometry. (C) Activated naive CD4+ T cells were treated with LF (50 μg/ml) and TGF-β1 (0.5 ng/ml) for 48 h, and the expression of ICOS, CD39, CD73, and Foxp3 were determined by flow cytometry. (D-1) CFSE-labeled naive CD4+ T cells were cocultured with Foxp3+ T cells, which were obtained by culturing activated CD4+ T cells with LF + TGF-β1 in the presence of the CD39/CD73 inhibitor ARL67165 (100 μM). Cell division was assessed after 48 h by analyzing CFSE dilution using flow cytometry. (D-2) Levels of adenosine in supernatants from activated T cells cultured with LF + TGF-β1 for 48 h were determined by measuring adenosine deaminase activity. Data represent the mean of triplicate samples; error bars represent SEM. *p < 0.05. NS, not significant.

Close modal

We explored the possibility that TGF-β, a well-known suppressor molecule produced by Tregs (23), was inhibiting effecter T cell activation. However, the suppressor activity of LF/TGF-β1–induced T cells was not counteracted by anti–TGF-β Ab, even at high concentrations (Fig. 4B). In addition, IL-10, another well-known suppressor cytokine produced by Tregs, was detected at very low levels in supernatants from LF/TGF-β1–stimulated T cells (data not shown), suggesting that soluble mediators were not involved in the present context. Therefore, we were interested in exploring whether putative suppressive molecules present on the surface of Tregs were responsible. As shown in (Fig. 4C, ICOS, CD39, and CD73 were upregulated in LF/TGF-β1–induced Foxp3+ T cells. We focused on CD39 and CD73 because the conversion of ATP to adenosine by CD39/CD73 in Tregs has been proposed to be inhibitory and upregulated by TGF-β (30, 31). Pretreatment with CD39/CD73 inhibitor ARL67165 completely abolished effector T cell suppression by LF/TGF-β1–induced Foxp3+ T cells, suggesting that suppression is mediated by adenosine derived from ATP conversion by CD39/CD73 (Fig. 4D-1). Indeed, adenosine levels were significantly elevated in LF/TGF-β1–stimulated culture media (Fig. 4D-2). Collectively, these results strongly suggest that CD39/CD73 potentiates the generation of adenosine, which in turn acts as an important mediator of LF/TGF-β1–induced Foxp3+ T cell suppression.

We tested the anti-inflammatory effect of LF-induced Tregs in colitis mouse models. First, we explored the effect of LF/TGF-β1–induced Foxp3+ T cells on RAG−/− mice that received colitogenic naive CD4+ T cells. Cotransfer of LF/TGF-β1–induced Foxp3+ T cells with naive T cells prevented weight loss (Fig. 5A-1), and this was associated with fewer IFN-γ/IL-17+ T cells in the small and large intestine (Fig. 5A-2). Next, we tested the effect of LF in a DSS-induced colitis model. Administration of LF significantly ameliorated weight loss, colon length shortening, and pathological severity caused by DSS treatment (Fig. 5B1–B3). We also observed that LF significantly increased the frequency of intestinal CD4+Foxp3+ T cells (Fig. 5B-4). These results reveal that LF can differentiate T cells into functional Tregs and mitigate IBD, such as colitis.

FIGURE 5.

Inducible CD4+Foxp3+ T cells alleviate disease severity in colitis mouse models. (A) Naive C57BL/6 CD4+ T cells (1 × 106) were i.p. administered into RAG1−/− mice with IL-2–stimulated CD4+ T cells (5 × 105) (media) or IL-2/TGF-β1–stimulated CD4+ T cells (5 × 105) (TGF-β) or IL-2/LF/TGF-β1–stimulated CD4+ T cells (5 × 105) (LF + TGF-β). Weight changes over time were expressed as percentages of initial body weight (A-1). After 12 wk, IFN-γ+ and IL-17+CD4+ T cells in the lamina propria of the small intestine and colon were quantitated using flow cytometry (A-2). (B) Colitis was induced in BALB/c mice using 3% DSS drinking water for 9 d. LF (5 mg/kg) was i.p. administered on days 0 and 3. Weight changes (B-1), colon length (B-2), pathology of the distal colon, including histological activity and disease activity indices (B-3), and frequencies of Foxp3+ CD4+ T cells (B-4) were measured on day 10. Data represent the means of three independent experiments; error bars represent SEM. *p < 0.05.

FIGURE 5.

Inducible CD4+Foxp3+ T cells alleviate disease severity in colitis mouse models. (A) Naive C57BL/6 CD4+ T cells (1 × 106) were i.p. administered into RAG1−/− mice with IL-2–stimulated CD4+ T cells (5 × 105) (media) or IL-2/TGF-β1–stimulated CD4+ T cells (5 × 105) (TGF-β) or IL-2/LF/TGF-β1–stimulated CD4+ T cells (5 × 105) (LF + TGF-β). Weight changes over time were expressed as percentages of initial body weight (A-1). After 12 wk, IFN-γ+ and IL-17+CD4+ T cells in the lamina propria of the small intestine and colon were quantitated using flow cytometry (A-2). (B) Colitis was induced in BALB/c mice using 3% DSS drinking water for 9 d. LF (5 mg/kg) was i.p. administered on days 0 and 3. Weight changes (B-1), colon length (B-2), pathology of the distal colon, including histological activity and disease activity indices (B-3), and frequencies of Foxp3+ CD4+ T cells (B-4) were measured on day 10. Data represent the means of three independent experiments; error bars represent SEM. *p < 0.05.

Close modal

The present study demonstrates that LF enhances both Ag-specific and Ag-nonspecific CD4+Foxp3+ Treg differentiation from naive CD4+CD25 T cells. The Foxp3-inducing property of LF was identical to that of TGF-β1. However, its effects on other phenotypes were opposite to those of TGF-β1. LF caused CD4+ T cells to differentiate into central memory-like Tregs (CD44loCD62L+Foxp3+), whereas TGF-β1 caused them to differentiate into effector-like Tregs (CD44hiCD62LFoxp3+); when combined, LF and TGF-β differentiated T cells into central memory-like Tregs (Supplemental Fig. 3A). Moreover, LF, unlike TGF-β1, substantially inhibited the proliferation of activated CD4+ T cells. Our characterization of the purified LF preparation used in the current study revealed that its biological effects were not caused by TGF-β contamination. Taken together, these results strongly suggest that the underlying mechanism(s) of LF-facilitated Treg induction is quite different from that of TGF-β. Because the inhibition of cell proliferation is known to drive cell differentiation, the antiproliferative activity of LF is likely tied to the inhibition of CD62L shedding and CD44 expression and the increase in Foxp3 expression.

One of the most striking observations made in this study was that LF and TGF-β1 synergistically increase Foxp3 expression. TGF-β1 is a well-known cytokine that induces iTreg differentiation through binding of p-Smad3 to the enhancer of the Foxp3 gene (4), and this is promoted by RA. Several underlying mechanisms have been proposed for how RA enhances TGF-β-dependent Foxp3 expression (5, 6, 32); however, unlike LF, RA does not affect Foxp3 expression on its own, clearly indicating that the underlying mechanism(s) of Foxp3 expression exerted by LF and RA are different from each other.

What could be the key mechanism by which LF induces Foxp3+CD62L+CD44lo T cells? Although LF-induced Foxp3 expression was independent of TGF-β, it was important to examine its possible link to TGF-β after observing that LF increased the phosphorylation of Smad3. Surprisingly, anti–TGF-β Ab virtually abrogated LF-facilitated Foxp3 expression. It was later determined that LF converted LAPhi–mTGF-β to biologically active LAPlo–mTGF-β as analyzed by the Mv1Lu cell inhibitory assay. This is consistent with a previous study reporting that mTGF-β/GARP complexes, rather than secreted TGF-β, are potent Treg inducers in the presence of IL-2 (33). It has been shown that LAP–mTGF-β binds to αVβ8 integrin on epithelial and neuronal cells, leading to membrane-type metalloprotease–dependent release of active mTGF-β (26). However, in LF-stimulated T cell cultures, neither metalloprotease expression nor TGF-β secretion was enhanced (data not shown). Instead, we found that LF, through TβRIII, noticeably increased the level of ROS, which in turn converted LAPhi–mTGF-β to LAPlo–mTGF-β, leading to Smad3-dependent Foxp3 expression (Supplemental Fig. 3B). Consistent with this, ROS-mediated activation of latent mTGF-β in T cells has been reported by others (28, 29). Therefore, we hypothesized that the major mechanism by which LF was inducing Foxp3 gene expression was through TβRIII/ROS. We asked whether LF stimulated the expression of NADPH oxidase 4, which is known to readily generate ROS (34), but did not find this to be the case in the current study. Therefore, it remains to be determined how LF is inducing mitochondrial ROS production in activated CD4+ T cells. As for a potential LF receptor, we have already shown that LF binds directly to TβRIII (16). It is likely that LF glycosylation is necessary for TβRIII (β-glycan, a membrane-anchored proteoglycan) binding as LF has been shown to bind to heparan sulfate chains on proteoglycans (35, 36). Thus, this may explain why unglycosylated recombinant LF had no effect on Foxp3 expression, similar to that of LF hydrolysate.

The present study clearly demonstrated that LF/TGF-β1–induced CD4+Foxp3+ T cells suppressed the proliferation of responder CD4+ T cells. Our findings suggested that this suppressor activity was mediated by adenosine converted by CD39/CD73, which are membrane receptors that are highly expressed on the surface of Tregs (30, 37). CD39 is an ectonucleotidase that cleaves ATP to form AMP, which can then be cleaved by CD73 to generate adenosine (38). Consistent with our observation, it has been demonstrated that Tregs from CD39-deficient mice have reduced suppressive capacity in vitro and that these mice fail to prevent allograft rejection in vivo (30). To date, various putative Treg suppressor molecules have been reported, including soluble cytokines such as TGF-β1 and IL-10 (23, 39). However, our in vitro experiments clearly showed that TGF-β1 did not mediate the suppressor activity of LF/TGF-β1–induced CD4+Foxp3+ Tregs, which is consistent with results from other studies (40, 41). Furthermore, IL-10 was not detected in the conditioned culture media, and anti–IL-10 Ab did not abrogate suppressor activity (data not shown). Our data suggest that the predominant suppressor molecule is adenosine generated by CD39/CD73 expressed by LF/TGF-β1–induced Foxp3+ Tregs.

LF delivery in vivo has previously been shown to increase the number of Foxp3+ T cells, alleviating DSS-induced colitis (15). Using a different colitis mouse model, we found that LF/TGF-β–induced Foxp3+ T cells reduced the number of intestinal IFN-γ/IL-17+ T cells, alleviating clinical severity. This suggests that LF is an important mediator of immune homeostasis. To support this, it is speculated that as inflammation subsides, LF derived from neutrophils (a main LF producer) initiates iTreg differentiation, leading to a subsequent tolerogenic state (42). Taken together, these findings indicate that LF is a good candidate for the treatment of autoimmune disorders.

Flow cytometry was performed at the Flow Cytometry Core Facility of the College of Pharmacy at Kangwon National University.

This work was supported by National Research Foundation of Korea grants (NRF-2016R1A2B4009646 and NRF-2019R1I1A3A01048952 to P.-H.K. and NRF-2016R1D1A3B03936377 and NRF-2019R1I1A1A01043808 to Y.-S.J.) and a 2017 research grant from Kangwon National University to P.-H.K.

Y.-S.J. and P.-H.K. conceptualized the project. Y.-S.J. designed and performed experiments and analyzed the data. H.-E.S., H.-J.J., S.P., H.-W.P., and T.-G.K. aided in experiments and contributed to data analysis. G.-Y.S. performed the histological analysis. P.-H.K. conceived, designed, and coordinated the project. Y.-S.J. and P.-H.K. wrote the manuscript. S.-G.K., S.-i.Y., H.-J.K., G.-S.L., and S.-R.P. discussed the results and commented on the manuscript.

The online version of this article contains supplemental material.

Abbreviations used in this article

DSS

dextran sulfate sodium

F

forward primer

IBD

inflammatory bowel disease

iTreg

inducible Treg

LAP

latency-associated peptide

LF

lactoferrin

LRP

low-density lipoprotein receptor–related protein

mTGF-β

membrane TGF-β

NAC

N-acetyl-L-cysteine

R

reverse primer

RA

retinoic acid

ROS

reactive oxygen species

sTβRIII

soluble TβRIII

Treg

regulatory T cell

TβRI

TGF-β receptor I

TβRIII

TGF-β receptor III

1.
Abbas
A. K.
,
C.
Benoist
,
J. A.
Bluestone
,
D. J.
Campbell
,
S.
Ghosh
,
S.
Hori
,
S.
Jiang
,
V. K.
Kuchroo
,
D.
Mathis
,
M. G.
Roncarolo
, et al
2013
.
Regulatory T cells: Recommendations to simplify the nomenclature.
Nat. Immunol.
14
:
307
308
.
2.
Chen
W.
,
W.
Jin
,
N.
Hardegen
,
K. J.
Lei
,
L.
Li
,
N.
Marinos
,
G.
McGrady
,
S. M.
Wahl
.
2003
.
Conversion of peripheral CD4+CD25- naive T cells to CD4+CD25+ regulatory T cells by TGF-beta induction of transcription factor Foxp3.
J. Exp. Med.
198
:
1875
1886
.
3.
Hori
S.
,
T.
Nomura
,
S.
Sakaguchi
.
2003
.
Control of regulatory T cell development by the transcription factor Foxp3.
Science.
299
:
1057
1061
.
4.
Tone
Y.
,
K.
Furuuchi
,
Y.
Kojima
,
M. L.
Tykocinski
,
M. I.
Greene
,
M.
Tone
.
2007
.
Smad3 and NFAT cooperate to induce Foxp3 expression through its enhancer.
Nat. Immunol.
9
:
194
202
.
5.
Xiao
S.
,
H.
Jin
,
T.
Korn
,
S. M.
Liu
,
M.
Oukka
,
B.
Lim
,
V. K.
Kuchroo
.
2008
.
Retinoic acid increases Foxp3+ regulatory T cells and inhibits development of Th17 cells by enhancing TGF-beta-driven Smad3 signaling and inhibiting IL-6 and IL-23 receptor expression.
J. Immunol.
181
:
2277
2284
.
6.
Xu
L.
,
A.
Kitani
,
C.
Stuelten
,
G.
McGrady
,
I.
Fuss
,
W.
Strober
.
2010
.
Positive and negative transcriptional regulation of the Foxp3 gene is mediated by access and binding of the Smad3 protein to enhancer I.
Immunity.
33
:
313
325
.
7.
Kanamori
M.
,
H.
Nakatsukasa
,
M.
Okada
,
Q.
Lu
,
A.
Yoshimura
.
2016
.
Induced regulatory T cells: Their development, stability, and applications.
Trends Immunol.
37
:
803
811
.
8.
Kruzel
M. L.
,
M.
Zimecki
,
J. K.
Actor
.
2017
.
Lactoferrin in a context of inflammation-induced pathology.
Front. Immunol.
8
:
1438
.
9.
Li
K. J.
,
M. C.
Lu
,
S. C.
Hsieh
,
C. H.
Wu
,
H. S.
Yu
,
C. Y.
Tsai
,
C. L.
Yu
.
2006
.
Release of surface-expressed lactoferrin from polymorphonuclear neutrophils after contact with CD4+ T cells and its modulation on Th1/Th2 cytokine production.
J. Leukoc. Biol.
80
:
350
358
.
10.
Zimecki
M.
,
J.
Mazurier
,
G.
Spik
,
J. A.
Kapp
.
1996
.
Lactoferrin inhibits proliferative response and cytokine production of TH1 but not TH2 cell lines.
Arch. Immunol. Ther. Exp. (Warsz.).
44
:
51
56
.
11.
Zimecki
M.
,
M.
Kocieba
,
G.
Chodaczek
,
M.
Houszka
,
M. L.
Kruzel
.
2007
.
Lactoferrin ameliorates symptoms of experimental encephalomyelitis in Lewis rats.
J. Neuroimmunol.
182
:
160
166
.
12.
Yin
H.
,
L.
Cheng
,
C.
Agarwal
,
R.
Agarwal
,
C.
Ju
.
2010
.
Lactoferrin protects against concanavalin A-induced liver injury in mice.
Liver Int.
30
:
623
632
.
13.
Togawa
J.
,
H.
Nagase
,
K.
Tanaka
,
M.
Inamori
,
A.
Nakajima
,
N.
Ueno
,
T.
Saito
,
H.
Sekihara
.
2002
.
Oral administration of lactoferrin reduces colitis in rats via modulation of the immune system and correction of cytokine imbalance.
J. Gastroenterol. Hepatol.
17
:
1291
1298
.
14.
Akin
I. M.
,
B.
Atasay
,
F.
Dogu
,
E.
Okulu
,
S.
Arsan
,
H. D.
Karatas
,
A.
Ikinciogullari
,
T.
Turmen
.
2014
.
Oral lactoferrin to prevent nosocomial sepsis and necrotizing enterocolitis of premature neonates and effect on T-regulatory cells.
Am. J. Perinatol.
31
:
1111
1120
.
15.
MacManus
C. F.
,
C. B.
Collins
,
T. T.
Nguyen
,
R. W.
Alfano
,
P.
Jedlicka
,
E. F.
de Zoeten
.
2017
.
VEN-120, a recombinant human lactoferrin, promotes a regulatory t cell [Treg] phenotype and drives resolution of inflammation in distinct murine models of inflammatory bowel disease.
J. Crohn’s Colitis.
11
:
1101
1112
.
16.
Jang
Y. S.
,
G. Y.
Seo
,
J. M.
Lee
,
H. Y.
Seo
,
H. J.
Han
,
S. J.
Kim
,
B. R.
Jin
,
H. J.
Kim
,
S. R.
Park
,
K. J.
Rhee
, et al
2014
.
Lactoferrin causes IgA and IgG2b isotype switching through betaglycan binding and activation of canonical TGF-β signaling.
Mucosal Immunol.
8
:
906
917
.
17.
Park
S. R.
,
J. H.
Lee
,
P. H.
Kim
.
2001
.
Smad3 and Smad4 mediate transforming growth factor-beta1-induced IgA expression in murine B lymphocytes.
Eur. J. Immunol.
31
:
1706
1715
.
18.
Denizot
F.
,
R.
Lang
.
1986
.
Rapid colorimetric assay for cell growth and survival. Modifications to the tetrazolium dye procedure giving improved sensitivity and reliability.
J. Immunol. Methods.
89
:
271
277
.
19.
Stevceva
L.
,
P.
Pavli
,
G.
Buffinton
,
A.
Wozniak
,
W. F.
Doe
.
1999
.
Dextran sodium sulphate-induced colitis activity varies with mouse strain but develops in lipopolysaccharide-unresponsive mice.
J. Gastroenterol. Hepatol.
14
:
54
60
.
20.
Benson
M. J.
,
K.
Pino-Lagos
,
M.
Rosemblatt
,
R. J.
Noelle
.
2007
.
All-trans retinoic acid mediates enhanced T reg cell growth, differentiation, and gut homing in the face of high levels of co-stimulation.
J. Exp. Med.
204
:
1765
1774
.
21.
Mucida
D.
,
K.
Pino-Lagos
,
G.
Kim
,
E.
Nowak
,
M. J.
Benson
,
M.
Kronenberg
,
R. J.
Noelle
,
H.
Cheroutre
.
2009
.
Retinoic acid can directly promote TGF-beta-mediated Foxp3(+) Treg cell conversion of naive T cells.
Immunity.
30
:
471
472
,
author reply 472–473
.
22.
Seo
G. Y.
,
Y. S.
Jang
,
H. A.
Kim
,
M. R.
Lee
,
M. H.
Park
,
S. R.
Park
,
J. M.
Lee
,
J.
Choe
,
P. H.
Kim
.
2013
.
Retinoic acid, acting as a highly specific IgA isotype switch factor, cooperates with TGF-β1 to enhance the overall IgA response.
J. Leukoc. Biol.
94
:
325
335
.
23.
Nakamura
K.
,
A.
Kitani
,
W.
Strober
.
2001
.
Cell contact-dependent immunosuppression by CD4(+)CD25(+) regulatory T cells is mediated by cell surface-bound transforming growth factor beta.
J. Exp. Med.
194
:
629
644
.
24.
Annunziato
F.
,
L.
Cosmi
,
F.
Liotta
,
E.
Lazzeri
,
R.
Manetti
,
V.
Vanini
,
P.
Romagnani
,
E.
Maggi
,
S.
Romagnani
.
2002
.
Phenotype, localization, and mechanism of suppression of CD4(+)CD25(+) human thymocytes.
J. Exp. Med.
196
:
379
387
.
25.
Oida
T.
,
X.
Zhang
,
M.
Goto
,
S.
Hachimura
,
M.
Totsuka
,
S.
Kaminogawa
,
H. L.
Weiner
.
2003
.
CD4+CD25- T cells that express latency-associated peptide on the surface suppress CD4+CD45RBhigh-induced colitis by a TGF-beta-dependent mechanism.
J. Immunol.
170
:
2516
2522
.
26.
Mu
D.
,
S.
Cambier
,
L.
Fjellbirkeland
,
J. L.
Baron
,
J. S.
Munger
,
H.
Kawakatsu
,
D.
Sheppard
,
V. C.
Broaddus
,
S. L.
Nishimura
.
2002
.
The integrin alpha(v)beta8 mediates epithelial homeostasis through MT1-MMP-dependent activation of TGF-beta1.
J. Cell Biol.
157
:
493
507
.
27.
Ortega-Francisco
S.
,
M.
de la Fuente-Granada
,
E. K.
Alvarez Salazar
,
L. A.
Bolaños-Castro
,
G.
Fonseca-Camarillo
,
R.
Olguin-Alor
,
G. R.
Alemán-Muench
,
F.
López-Casillas
,
C.
Raman
,
E. A.
García-Zepeda
,
G.
Soldevila
.
2017
.
TβRIII is induced by TCR signaling and downregulated in FoxP3+ regulatory T cells.
Biochem. Biophys. Res. Commun.
494
:
82
87
.
28.
Zhang
D.
,
C.
Chia
,
X.
Jiao
,
W.
Jin
,
S.
Kasagi
,
R.
Wu
,
J. E.
Konkel
,
H.
Nakatsukasa
,
P.
Zanvit
,
N.
Goldberg
, et al
2017
.
D-mannose induces regulatory T cells and suppresses immunopathology.
Nat. Med.
23
:
1036
1045
.
29.
Zhang
D.
,
W.
Jin
,
R.
Wu
,
J.
Li
,
S. A.
Park
,
E.
Tu
,
P.
Zanvit
,
J.
Xu
,
O.
Liu
,
A.
Cain
,
W.
Chen
.
2019
.
High Glucose intake exacerbates autoimmunity through reactive-oxygen-species-mediated TGF-β cytokine activation.
Immunity.
51
:
671
681.e5
.
30.
Deaglio
S.
,
K. M.
Dwyer
,
W.
Gao
,
D.
Friedman
,
A.
Usheva
,
A.
Erat
,
J. F.
Chen
,
K.
Enjyoji
,
J.
Linden
,
M.
Oukka
, et al
2007
.
Adenosine generation catalyzed by CD39 and CD73 expressed on regulatory T cells mediates immune suppression.
J. Exp. Med.
204
:
1257
1265
.
31.
Regateiro
F. S.
,
D.
Howie
,
K. F.
Nolan
,
E. I.
Agorogiannis
,
D. R.
Greaves
,
S. P.
Cobbold
,
H.
Waldmann
.
2011
.
Generation of anti-inflammatory adenosine by leukocytes is regulated by TGF-β.
Eur. J. Immunol.
41
:
2955
2965
.
32.
Lu
L.
,
J.
Ma
,
Z.
Li
,
Q.
Lan
,
M.
Chen
,
Y.
Liu
,
Z.
Xia
,
J.
Wang
,
Y.
Han
,
W.
Shi
, et al
2011
.
All-trans retinoic acid promotes TGF-β-induced Tregs via histone modification but not DNA demethylation on Foxp3 gene locus.
PLoS One.
6
:
e24590
.
33.
Edwards
J. P.
,
H.
Fujii
,
A. X.
Zhou
,
J.
Creemers
,
D.
Unutmaz
,
E. M.
Shevach
.
2013
.
Regulation of the expression of GARP/latent TGF-β1 complexes on mouse T cells and their role in regulatory T cell and Th17 differentiation.
J. Immunol.
190
:
5506
5515
.
34.
Shiose
A.
,
J.
Kuroda
,
K.
Tsuruya
,
M.
Hirai
,
H.
Hirakata
,
S.
Naito
,
M.
Hattori
,
Y.
Sakaki
,
H.
Sumimoto
.
2001
.
A novel superoxide-producing NAD(P)H oxidase in kidney.
J. Biol. Chem.
276
:
1417
1423
.
35.
Wu
H. F.
,
R. L.
Lundblad
,
F. C.
Church
.
1995
.
Neutralization of heparin activity by neutrophil lactoferrin.
Blood.
85
:
421
428
.
36.
Cheifetz
S.
,
J. L.
Andres
,
J.
Massagué
.
1988
.
The transforming growth factor-beta receptor type III is a membrane proteoglycan. Domain structure of the receptor.
J. Biol. Chem.
263
:
16984
16991
.
37.
Borsellino
G.
,
M.
Kleinewietfeld
,
D.
Di Mitri
,
A.
Sternjak
,
A.
Diamantini
,
R.
Giometto
,
S.
Höpner
,
D.
Centonze
,
G.
Bernardi
,
M. L.
Dell’Acqua
, et al
2007
.
Expression of ectonucleotidase CD39 by Foxp3+ Treg cells: hydrolysis of extracellular ATP and immune suppression.
Blood.
110
:
1225
1232
.
38.
Mizumoto
N.
,
T.
Kumamoto
,
S. C.
Robson
,
J.
Sévigny
,
H.
Matsue
,
K.
Enjyoji
,
A.
Takashima
.
2002
.
CD39 is the dominant Langerhans cell-associated ecto-NTPDase: modulatory roles in inflammation and immune responsiveness.
Nat. Med.
8
:
358
365
.
39.
Asseman
C.
,
S.
Mauze
,
M. W.
Leach
,
R. L.
Coffman
,
F.
Powrie
.
1999
.
An essential role for interleukin 10 in the function of regulatory T cells that inhibit intestinal inflammation.
J. Exp. Med.
190
:
995
1004
.
40.
Baecher-Allan
C.
,
V.
Viglietta
,
D. A.
Hafler
.
2002
.
Inhibition of human CD4(+)CD25(+high) regulatory T cell function.
J. Immunol.
169
:
6210
6217
.
41.
Piccirillo
C. A.
,
J. J.
Letterio
,
A. M.
Thornton
,
R. S.
McHugh
,
M.
Mamura
,
H.
Mizuhara
,
E. M.
Shevach
.
2002
.
CD4(+)CD25(+) regulatory T cells can mediate suppressor function in the absence of transforming growth factor beta1 production and responsiveness.
J. Exp. Med.
196
:
237
246
.
42.
Masson
P. L.
,
J. F.
Heremans
,
E.
Schonne
.
1969
.
Lactoferrin, an iron-binding protein in neutrophilic leukocytes.
J. Exp. Med.
130
:
643
658
.

The authors have no financial conflicts of interest.

Supplementary data