Inflammatory bowel disease (IBD) is a chronic intestinal inflammation, but the accurate etiology remains to be elucidated. Increasing evidence has shown that macrophages polarize to different phenotypes depending on the intestinal microenvironment and are associated with the progression of IBD. In the present study, we investigated the effect of oxytocin, a neuroendocrinal, and pro-health peptide, on the modulation of macrophages polarization and the progression of experimental colitis. Our data demonstrated that oxytocin decreased the sensitivity of macrophages to lipopolysaccharide stimulation with lower expression of inflammatory cytokines, like IL-1β, IL-6, and TNF-α, but increased the sensitivity to IL-4 stimulation with enhanced expression of M2-type genes, arginase I (Arg1), CD206, and chitinase-like 3 (Chil3). This bidirectional modulation was partly due to the up-regulation of β-arrestin2 and resulted in the inhibition of NF-κB signaling and reinforcement of Signal transducer and activator of transcription (STAT) 6 phosphorylation. Moreover, oxytocin receptor (OXTR) myeloid deficiency mice were more susceptible to dextran sulfate sodium (DSS) intervention compared with the wild mice. For the first time, we reveal that oxytocin–oxytocin receptor system participates in modulating the polarization of macrophages to an anti-inflammatory phenotype and alleviates experimental colitis. These findings provide new potential insights into the pathogenesis and therapy of IBD.
Inflammatory bowel disease (IBD) is a world-wide chronic intestinal inflammation, including two major subtypes, Crohn’s disease, and ulcerative colitis . The global incidence of IBD has increased in the past few decades, especially in developing countries . It has become an enormous financial burden on the patients and governments because of its prolonged healing and repeated recurrence. It may be due to the uncontrolled immune-medicated inflammatory response combined with unknown environmental triggers, intestinal flora imbalance and specific genetic factors .
Innate immunity especially the intestinal monocyte-macrophage has been reported to play a critical role in IBD. There are a large number of intestinal macrophages in lamina propria, acting as the most essential innate immune cells in removing invasive bacteria and maintaining intestinal homeostasis . The intestinal macrophages are highly plastic and continuously replaced by incoming monocytes which acquire a tissue-protective or pro-inflammatory signature depending on the local microenvironment [5,6]. At steady state, intestinal resident macrophages prefer an anti-inflammatory phenotype, also known as alternatively activated (M2-like) macrophages, under the effect of TGF-β, and IL-10 secreted by epithelial cells, mast cells, and stromal cells. M2-like macrophages exhibit tolerogenic properties, producing less pro-inflammatory cytokines but more IL-10, mannose receptor (CD206), CD163 and arginase I (Arg1) antigens, participating in tissue restitution after injury . When pathogenic microorganisms invade the mucosa, monocytes are recruited into the intestine and tend to polarize into pro-inflammatory macrophages, also called classically activated (M1-like) macrophages, induced by IFN-γ, TNF-α, and lipopolysaccharide (LPS). The M1-like macrophages produce significant numbers of proinflammatory cytokines and chemokines, like IL-1β, IL-6, TNF-α, inducible nitric oxide synthase (iNOS) and reactive oxygen species, exacerbating the inflammatory response and tissue injury . Regulation of the balance between M1 and M2 macrophage subsets becomes an intriguing target for the treatment of IBD.
Oxytocin is traditionally recognized as a hormone produced by the supraoptic nucleus and paraventricular nucleus of the hypothalamus and released from the posterior pituitary into circulation . In mammals, it is involved in reproductive functions like uterine contraction and milk ejection, as well as the regulation of social behaviors [10,11]. Recently, some studies indicated that oxytocin receptors (OXTRs) were expressed in immune cells and participated in maintaining immune homeostasis, although the mechanisms were still unclear [12,13]. In the intestine, oxytocin is produced by the enteric neurons . Our previous studies indicated that oxytocin administration protected the intestine from dextran sulfate sodium (DSS) colitis in mice . It was reported that oxytocin decreased IL-6 mRNA expression in THP-1 cells and TNF-α mRNA expression in bone marrow-derived macrophages induced by LPS [16,17], but there were few reports about the effects of oxytocin on M2 macrophages polarization. In our present study, we first investigated that, both in vivo and in vitro experiments, oxytocin inhibited the activity of pro-inflammatory macrophages and enhanced the anti-inflammatory polarization. Considering the critical role of intestinal macrophages in the colitis, we inferred endogenous oxytocin ameliorates intestinal inflammation by modulating the polarization of macrophages.
Materials and methods
Lipopolysaccharide (100 ng ml−1), phorbol 12-myristate 13-acetate (PMA, 100 nM), oxytocin (10−8 M), and atosiban (10−6 M) were purchased from Sigma–Aldrich (St. Louis, CA). Recombinant mouse IL-4 (10 ng ml−1) and macrophage colony stimulating factor (MCSF, 20 ng ml−1) were purchased from R&D Systems (Minneapolis, MN). Recombinant human interferon-γ (IFN-γ, 20 ng ml−1) and IL-4 (20 ng ml−1) were purchased from Peprotech (Rocky Hill, NJ). Antibodies for CD206, F4/80 and GAPDH, β-arrestin2, Signal transducer and activator of transcription (STAT) 6, phospho-STAT6, p65, and phospho-p65 were from Abcam (Cambridge, U.K.) and Cell Signaling Technology (Danvers, MA). The secondary antibodies were purchased from Invitrogen Life Technology (Foster City, CA). Mouse ELISA kits were obtained from R&D Systems and Dakewe Biotech (Shenzhen, China). All reagents were analytical grade.
Cell isolation and culture
RAW264.7 macrophage-like cell line was purchased from the Cell Bank of Chinese Academy of Science (Shanghai, China). Mouse peritoneal macrophages were obtained from C57BL6/J mice as previously described . Briefly, the mouse was euthanized with no pain, spayed with 75% ethanol, and cut the outer skin of the peritoneum. 5 ml ice-cold PBS with 3% fetal bovine serum (FBS) was injected into the peritoneum cavity using a 27 g needle. The mouse was putted on ice and massaged the peritoneum for 30 min. Peritoneal fluid was gently collected with a 25 g needle, and spun at 1500 RPM for 8 min to get the cells. Cells were cultured in Dulbecco’s modified Eagle’s medium (Hyclone, Logan, Utah) supplemented 10% heat-inactivated FBS (Gibco, Foster City, CA) and 1% penicillin–streptomycin solutions (Gibco) in a humidified incubator with 5% CO2 at 37°C. Cells were treated with oxytocin (10−8 M) for 50 min before LPS (100 ng ml−1) or IL-4 (10 ng ml−1) was added. Human monocyte cell line THP-1 was purchased from the Cell Bank of Chinese Academy of Science and maintained in RMPI 1640 medium with 10% heat-inactivated fetal bovine serum and 1% penicillin–streptomycin solutions in a humidified incubator with 5% CO2 at 37°C. THP-1 cells were stimulated with 100 nM of PMA for 48 h to obtain THP-1 derived macrophages. Then cells were treated with oxytocin (10−8 M) for 50 min before LPS (100 ng ml−1) and human IFN-γ (20 ng ml−1) or IL-4 (20 ng ml−1) was added.
Wild-type C57BL6/J mice were purchased from the Animal Center of Shandong University. The OXTR floxed (OXTRFL/FL) mice were purchased from the Jackson Laboratory (stock No. 008471). The LysM-Cre transgenic mice were kindly provided by Dr. Liu Shangming (Shandong University). OXTR floxed (OXTRFL/FL) mice were crossed with LysM-Cre transgenic mice to generate myeloid-specific OXTR knockout mice (OXTRmyel-KO), and their littermates OXTRFL/FL mice were used as myeloid-specific OXTR wild-type control (OXTRmyel-WT). All mice used were male and 8–10 weeks old. Experimental mice were randomly grouped and maintained under pathogen-free conditions in the animal care facilities. All animal experiments were carried out at the Animal Center of Shandong University Cheeloo Medical College and approved by the Medical Ethics Committee for Experimental Animals, Shandong University School of Basic Medicine Sciences (ECAESDUSM 2014056). The suffering of mice was kept to the minimum.
The cell slides were fixed with 4% paraformaldehyde at room temperature for 30 min and then washed with PBS for three-times. After blocked with 10% donkey serum for 50 min at 25°C, the cells were incubated with rabbit anti-OXTR (1:100, Abcam) and rat anti-F4/80 (1:100, Abcam) primary antibodies dissolved in blocking solution overnight at 4°C. The paraffin sections of mice colon were dewaxed and then repaired by boiling slices in sodium citrate buffer (10 mM, PH = 6, Beyotime) for 30 min. After cooling down, tissues were blocked with 10% donkey serum for 50 min at 25°C and then incubated with rabbit anti-CD206 (1:500, Abcam) primary antibodies dissolved by blocking solution overnight at 4°C. After multiple washes, the slides were incubated with Alexa Fluor 568 (or 488) donkey antirabbit (1:2000, Invitrogen) or Alexa Fluor 488 donkey antirat (1:200, Abcam) secondary antibodies for 1 h and then counterstained with DAPI (1:1000, Beyotime) for 5 min. The images were observed by a fluorescent microscope (Olympus IX71).
RNA extraction and quantitative real-time PCR
The total RNA was prepared from cells and colonic tissue using a tissue/cell rapid extraction kit (Aidlab Biotechnologies, Beijing, China) in accordance with the operating instruction. RNA was reverse transcribed using a Takara PCR Thermal Cycler SP (Takara Bio, Shiga, Japan). Real-time quantitative PCR was carried out using the SYBR Premix Dimer Eraser (Takara Bio, Shiga, Japan). The level of mRNA expression was presented as ratio relative to that of the control housekeeping gene. The primers were composed by The Beijing Genomics Institute (Shenzhen, China) and the sequences of the specific primers used were as showed in Table 1.
|Gene||Species||Forward primer||Reverse primer|
|Gene||Species||Forward primer||Reverse primer|
Abbreviations: CCL2, C–C motif chemokine ligand; Chil3, chitinase-like 3; CXCL, C-X-C motif chemokine.
ELISA and NO assay
The cell supernatant was collected and centrifugalized after stimulated by drugs. The concentration of specific protein was detected with a precoated ELISA kit according to the manual instruction. The NO production in cell culture media was detected using the classic Griess method as previous described .
DSS colitis conduction
Acute colitis was induced by administration of 2.5% DSS salt (reagent-grade, MW 36–50 kDa, MP Biomedicals, Canada) in the drinking water for 7 days. Neutrophils were depleted as described by using anti-Ly6G MAb 1A8 (1 mg per mouse, i.p., Bio X Cell) 1 day before DSS administration . The body weight was measured daily. The disease activity index was measured based on weight loss, stool consistence, and occult bleeding, as previously described . At day 7, mice were fed with FITC-dextran (400 mg kg−1, Sigma) by oral gavage. Serum samples (100 ul) were accessed from the submandibular vein 4 h later, and the fluorescence intensity was measured at 520 nm. Mice were euthanized by intraperitoneal administration of sodium pentobarbital (200 mg kg−1) on the day 7, and the colons were removed following PBS infusion. To induce chronic colitis, the mice underwent three cycles of DSS treatment. One cycle consisted of 5 days of 2% DSS in drinking water followed by 9 days of normal drinking water. This treatment is suited to induce a consistent and reproducible state of chronic colitis [22–25]. Mice were euthanized by intraperitoneal administration of sodium pentobarbital on the day 41. Distal colon was embedded by 4% paraformaldehyde solution, and the transverse sections were stained with Hematoxylin and Eosin. Two analysts were needed to measure the histological scores, blinded to each animal groups. The assessment criteria included epithelial surface damage, the loss of crypts, and inflammatory infiltrate described by Zaki , ranging from 0 to 6 (combining inflammatory cells infiltration score and tissue damage score). Picrosirius red were used to evaluate the colitis-associated collagen deposition. Under polarized light, collagen I appears red or yellow and has strong birefringence. Collagen III is green and weak birefringence. The images were observed by a polarization microscope (Nikon Eclipse Ci). The results were quantified by Image-Pro plus 6.0 and expressed as mean optical density. For 7-day DSS intervention, data were expressed as mean ± SEM from three independent experiments. For chronic DSS-induced colitis, two independent experiments were performed.
Isolation of colonic lamina propria cells
Lamina propria mononuclear cells (LPMCs) were isolated using a modified technique described previously . In brief, the colons of mice were removed and washed in ice-cold PBS to remove the intestinal contents. The intestines were opened longitudinally and cut into 0.5 cm sections. Segments were twice vigorously shaken in PBS containing 5% fetal calf serum, 1% penicillin–streptomycin, 1 mM DTT (Sigma–Aldrich) and 1 nM EDTA (Sigma–Aldrich) at 37°C for 20 min and the supernatants were discarded. The remain tissue was further minced and digested in shacking RPMI 1640 with 1 mg ml−1 collagenase IV (Roche, Germany) at 37°C for 30 min. The tube was shaken vigorously every 15 min during incubation. The supernatant was passed through a 40 μm cell strainer. To get the LPMCs, cells were isolated over a 40–75% Percoll (GE Healthcare, Sweden) gradient and spun at 2000 RPM for 20 min without the breaks. Isolated cells at the interface were gently collected for further analysis.
The mice were peritoneally injected with anti-Ly6G MAb 1A8 (1 mg per mouse) to deplete neutrophils. On the day 8, the cells in the lamina propria were collected as described above to assess the number of granulocytes by standard flow cytometry. Cells from lamina propria were stained with Percp antimouse Ly6G (Biolegend, San Diego, CA) and data were acquired using a FACS flow cytometer C6 (BD Biosciences, England).
Human colonic tissue samples
The colonic tissue samples were obtained from patients undergoing abdominal surgery at Qilu Hospital of Shandong University and Jinan Central Hospital. The samples were categorized as discard tissue and macroscopically normal in appearance. Surgical specimens were taken from seven patients (five males and two females) with a median age of 50.8 years (range 35–75 years) for colonic neoplasms (n=6) and structural anomalies (n=1). After multiple washing, specimens were trimmed into 1.0 cm pieces and cultured in RPMI 1640 supplemented 5% heat-inactivated FBS and 1% penicillin–streptomycin solutions in a humidified incubator with 5% CO2 at 37°C. The specimens were treated with oxytocin (10−8 M) for 40 min before LPS (100 ng ml−1) and human IFN-γ (20 ng ml−1) added. LPMCs were collected by using previously described methods with slightly modifications after the 5-h stimulation [28,29]. Briefly, the mucosal tissue was gently dissected and twice vigorously shaken in Hanks’ balanced salt solution (HBSS, Hyclone) containing 5% fetal calf serum, 1% penicillin-streptomycin, 1 mM DTT, and 1 nM EDTA at 37°C for 40 min and the supernatants were discarded. After rinsed with HBSS, the remaining tissue was minced and digested in RPMI 1640 with 1 mg ml−1 collagenase IV, 2% fetal calf serum, 1% penicillin-streptomycin and 20 ug ml−1 DNase (Sigma–Aldrich) at 37°C for 90 min. The tube was shaken vigorously every 15 min during incubation. The supernatant was passed through a 40 μm cell strainer. To get the LPMCs, cells were isolated by Ficoll-Hypaque (GE Healthcare) density centrifugation. Isolated cells at the interface were gently collected for further analysis.
All the patients were informed and consented to the unrestricted use of discarded tissue. The research was reviewed and approved by the Medical Ethics Committee, Shandong University School of Basic Medicine Sciences (ECSBMSSDU2018-1-039).
Western blotting analysis
After incubating by specific drugs, cells were harvested in RIPA-lysis buffer (Bioster Bio, Pleasanton, CA), and protein samples were electrophoresed and transferred to PVDF membrane. The membranes were blocked with 5% non-fat dry milk dissolved by tween/tris-buffered salt solution for 1 h at 25°C and then incubated with rabbit anti-CD206 (1:1000, Abcam), anti-OXTR (1:1000, Abcam), anti-p65 (1:1000, CST), anti-phospho-p65 (1:800, CST), anti-STAT6 (1:1000, CST), antiphospho-STAT6 (1:1000, CST), anti-β-arrestin2 (1:1000, CST) antibodies, and mouse anti-GAPDH (1:2000, CST) antibody overnight at 4°C. After multiple washes, membranes were incubated with goat antirabbit or mouse IgG secondary antibodies (1:2000, Beyotime) conjugated with HRP at 25°C for 1 h. Chemiluminescence was detected by using ECL plus (Beyotime, China). The signal intensities were analyzed by ImageJ software.
The genetic sequence targeting mouse β-arrestin2 was already confirmed as follows 5′-AAGGACCGGAAAGUGUUCGUG-3 . β-arrestin2 shRNA (shβ-arr2) plasmids were constructed by Genechem (Shanghai, China), and they were cloned into the lentiviral vectors for interference. RAW264.7 cells were infected with shβ-arr2 or negative control lentivirus separately for 24 h by using lipofectamine 2000 (Invitrogen, Foster City, CA). Cells were selected by puromycin (3μg ml−1, Sigma) for 24 h, after 4-day transfection. The expression of β-arrestin2 was detected by Western blot and qRT-PCR.
Data were expressed as mean ± SEM from at least two independent experiments, and n presented the number of samples in the specific experiments. One-way ANOVA or two-tailed Student’s t-test were used to compare between groups. GraphPad Prism version 5 (La Jolla, CA) was used for statistical analysis. P<0.05 was considered statistically significant.
Oxytocin depresses proinflammatory mediators release in LPS-activated RAW264.7 and THP-1 derived macrophages
In our study, we revealed that OXTRs were expressed on the membrane of primary and subcultured macrophages (Supplementary Figure S1). First, we used LPS to promote an M1 polarization and investigated the role of oxytocin in it. We found that LPS-induced macrophages to M1 polarization and oxytocin suppressed the transcription of inflammatory cytokines in LPS-stimulated RAW264.7 cells, like IL-1β, TNF-α, iNOS, chemokines C-C motif chemokine ligand (CCL) 2, and C-X-C motif chemokine (CXCL) 10 (Figure 1A). Oxytocin (10−6 M to 10−9 M) lessened IL-6 transcription induced by LPS in a concentration-independent manner (Figure 1B). Consistently, LPS-induced release of NO, TNF-α, IL-1β, and IL-6 were also significantly restrained by pretreatment of oxytocin (Figure 1C,D). These effects were reversed by atosiban, the OXTR antagonist. To obtain THP-1 derived macrophages, THP-1 cells were stimulated with of PMA (100 nM) for 48 h. Then we used LPS (100 ng ml−1) and human IFN-γ (20 ng ml−1) to promote an M1 polarization with up-regulated transcription of inflammatory markers, like CCL2, TNF-α, and CD86. Consistently, this effect was restrained by the pretreatment of oxytocin (Figure 1E). These findings indicated that oxytocin could restrain the polarization of macrophages to M1 phenotype and exert an anti-inflammatory function in vitro.
Oxytocin inhibits proinflammatory mediators release in LPS-activated RAW264.7 and THP-1 derived macrophages
Oxytocin attenuates the release of inflammatory mediators by LPS-stimulated macrophages via its receptors
Atosiban is not entirely selective to OXTR, so we established myeloid cell-specific deletion (OXTRmyel-KO) mice to investigate the function of OXT-OXTR system specifically. First, we confirmed that the transcription and expression of OXTR on OXTRmyel-KO macrophages were depleted successfully (Supplementary Figure S2). Then we isolated peritoneal macrophages from OXTRmyel-KO mice and the wild litter mates (OXTRmyel-WT). In OXTRmyel-WT peritoneal macrophages, LPS raised the expressions of inflammatory cytokines, like IL-1β, IL-6, and TNF-α. Preincubation of oxytocin inhibited the elevation of inflammatory cytokines induced by LPS. However, in OXTRmyel-KO peritoneal macrophages, oxytocin preincubation did not affect the increased expressions of IL-1β, IL-6, and TNF-α induced by LPS (Figure 2A–C). Therefore, we believe that oxytocin requires its receptors to inhibit the release of proinflammatory cytokines induced by LPS in macrophages.
Oxytocin attenuates the release of inflammatory mediators by LPS-stimulated macrophages via its receptors
Oxytocin promotes the M2 macrophages polarization in coordination with IL-4
The results mentioned above indicated that oxytocin inhibited macrophages polarizing toward the M1 phenotype induced by LPS. Then we investigated the effect of oxytocin on M2 polarization. We used IL-4 to promote an M2 polarization and found in cell line RAW264.7, oxytocin preincubation increased the sensitivity of macrophages to IL-4 stimulation with raised M2 markers, including CD206 and Arg1 (Figure 3A). Similarly, in THP-1 derived macrophages, oxytocin preincubation enhanced the expression of CD206, F13A1 and PTGS2 induced by IL-4 (Figure 3B).
Similarly to subcultured cells, in OXTRmyel-WT peritoneal macrophages, oxytocin preincubation increased the expression of CD206, Arg1, and Chil3 induced by IL-4. However, in OXTRmyel-KO peritoneal macrophages, oxytocin did not change the sensitivity to IL-4 stimulation (Figure 3C–E). To the best of our knowledge, this is the first time that the oxytocin system has been shown to promote M2-type polarization directly.
Oxytocin promotes the M2 macrophages polarization in coordination with IL-4
OXTR myeloid lineage conditioned KO mice have the increased inflammatory response and severity in DSS-induced colitis
Our previous studies indicated that exogenous oxytocin treatment significantly ameliorated DSS colitis in mice . Welch et al.  also reported that OXTR deficient mice had more severe DSS colitis. Innate immune cells especially macrophages play a crucial role in the progression and recovery of IBD. Therefore, we compared the severity of DSS-induced colitis in OXTRmyel-WT and OXTRmyel-KO mice to investigate the possible role of macrophagic oxytocin system in IBD course. To induce acute colitis, male OXTRmyel-KO mice and their wild OXTRmyel-WT littermates aged 8–10 weeks were treated with or without 2.5% DSS dissolved in drinking water for 7 days (n=8 in each group). The control mice had no significant inflammation in the colon and mice treated with 2.5% DSS, in both OXTRmyel-WT and OXTRmyel-KO groups, developed severe colitis. Compared with OXTRmyel-WT mice, the OXTRmyel-KO mice had more weight loss, shorter colon lengths, and higher disease activity index after DSS administration (Figure 4A–C). On the day 7, histological assessment of the OXTRmyel-KO mice showed more severe epithelial destruction, crypt loss, submucosal edema, and inflammatory cell infiltration in the colon than those of their wild littermates (Figure 4D). The histological index of colitis in OXTRmyel-KO mice was also significantly higher (Figure 4E). FITC-dextran was administered orally on the day 7, and its fluorescence intensity in serum was measured after 4 h to reflect the mucosal permeability. FITC fluorescence intensity in serum was significantly greater in OXTRmyel-KO mice than that in OXTRmyel-WT mice (Figure 4F), suggesting that OXTRmyel-KO mice had more severe mucosal damage and greater macromolecular permeability. To further evaluate whether OXTR deletion in macrophages had the same effect in chronic colitis, the male OXTRmyel-WT and OXTRmyel-KO mice underwent three cycles of 2% DSS treatment. During the experiment, OXTRmyel-KO mice were more sensitive to DSS treatment. The OXTRmyel-KO mice had more weight loss during DSS intervention compared with OXTRmyel-WT mice (Figure 4G). On the day 41, more severe shortening of colon was observed in OXTRmyel-KO mice (Figure 4H). On the day 5 and day 34, the OXTRmyel-KO mice had higher disease activity index compared with OXTRmyel-WT mice (Figure 4I). The LysM-Cre transgene is expressed in granulocytic cells in addition to macrophages . According to the method of Li et al. , we depleted granulocytes with anti-Ly6G antibody prior to DSS administration to exclude a role for these cells in enhanced colitis in OXTRmyel-KO mice, and granulocytes were effectively depleted for 8 days (Supplementary Figure S3). The anti-Ly6G intervention did not alter the disease course or severity in either OXTRmyel-WT or OXTRmyel-KO mice (Figure 4J). This indicated that macrophages but not granulocytes contributed to the protective effects of OXTR in colitis. All evidence above indicates that down-regulation of the OXTRs on macrophages would aggravate the inflammatory response induced by DSS in mice.
OXTR myeloid lineage conditioned KO mice become more sensitive to DSS-induced colitis
OXTR myeloid-lineage defect causes macrophages to polarize into a pro-inflammatory phenotype in DSS colitis
Macrophages take a critical role in the pathogenesis of intestinal inflammation. Previous data showed that oxytocin could up-regulate the sensitivity of macrophages to IL-4 stimulation, but down-regulate the sensitivity to LPS stimulation. In order to confirm the hypothesis that oxytocin performed the same function in vivo, we investigated macrophage subpopulations in the LPMCs from both OXTRmyel-WT and OXTRmyel-KO mice after DSS treatment (n=8 in each group). The expression of IL-1β, IL-6 and iNOS mRNA was distinctly higher in OXTRmyel-KO than those in OXTRmyel-WT mice after the 7-day DSS administration (Figure 5A). We also investigated the M2-type macrophages infiltrating into colonic submucosal tissue during colitis by immunofluorescence. There were less CD206-positive cells recruited in the OXTRmyel-KO mice than those in OXTRmyel-WT mice after 7-day DSS treatment (Figure 5B). Consistently, the levels of CD206 and Arg1, hallmarks of M2 macrophages, were apparently lower in the LPMCs from OXTRmyel-KO mice (Figure 5A). Similarly, in chronic DSS-induced colitis, the LPMCs from OXTRmyel-KO mice expressed more IL-1β, IL-6, TNF-α and less CD206 (Figure 5C). Besides we used picrosirius red staining to evaluate the colitis-associated fibrosis in each group. There was no significant difference on the collagen deposition between OXTRmyel-WT and OXTRmyel-KO mice on the day 41 (Figure 5D,E). These results indicate that macrophages are more prone to M1-like polarization when the OXTRs on macrophages were knocked out during colitis. Human colonic specimens from patients with colonic neoplasms (n=6) and structural anomalies (n=1) were stimulated by LPS (100 ng ml−1) and human IFN-γ (20 ng ml−1) for 5 h with or without 40-min oxytocin preincubation and then LPMCs were isolated from each group. Oxytocin effectively inhibited the expression of inflammatory cytokine induced by LPS and IFN-γ in LPMC, such as IL-6, iNOS, and TNF-α. Besides, oxytocin also enhanced the transcription of CD206 (Figure 5F). In other words, oxytocin may ameliorate the symptoms of IBD by inducing macrophages to a more anti-inflammatory phenotype.
OXTR myeloid defect causes macrophages to polarize into a pro-inflammatory phenotype in DSS mice
OXT-OXTR signaling modulates macrophage polarization by inhibiting NF-κB and promoting STAT6 phosphorylation
The nuclear transcription factor NF-κB is indispensable for intestinal immune homeostasis. It is known that the release of proinflammatory mediators induced by LPS stimulation is predominantly regulated by NF-κB at transcriptional level . It has been reported that the increase of phosphorylated NF-κB could cause macrophages to polarize toward M1 phenotype with enhanced expression of inflammatory cytokines and chemokines . STAT 6 is mainly responsible for the polarization to M2 macrophages induced by IL-4 . In our study, we investigated the effect of oxytocin on the phosphorylation of NF-κB and STAT6 during macrophage polarization. After stimulated by LPS for 1 h, the degree of p65 phosphorylation in peritoneal macrophages was higher than that of the control group. Oxytocin preincubation blocked the activation of p65 induced by LPS partly (Figure 6A). We used IL-4 to promote M2 polarization. The phosphorylation of STAT6 was enhanced when peritoneal macrophages were treated with IL-4 and oxytocin potentiated this effect (Figure 6B). Oxytocin alone had no effect on the phosphorylation of p65 and STAT6 (Supplementary Figure S4). We also extracted the colonic proteins from both OXTRmyel-KO and OXTRmyel-WT mice treated by 7-day DSS and found the expression of phosphorylated p65 in OXTRmyel-KO mice was higher than that in OXTRmyel-WT mice (Figure 6D). Therefore, our results demonstrate that oxytocin modulates the polarization of macrophages to a more anti-inflammatory phenotype primarily by inhibiting the LPS/NF-κB p65 signaling as well as promoting the IL-4/STAT6 signaling. However, the exactly underlying mechanisms are still unfathomed.
OXT-OXTR signaling modulates macrophage polarization by inhibiting NF-κB and promoting STAT6 phosphorylation
Oxytocin modulates the polarization of macrophages through OXTR-β-arrestin2 pathway
β-arrestin2, as a multi-functional scaffold that regulates G protein-coupled receptors (GPCR), is reported to be an endogenous regulator of NF-κB signaling in a G protein independent manner . To explore the role of β-arrestin2 in the connection between OXT-OXTR system and NF-κB signaling, we detected the expression of β-arrestin2 after LPS or IL-4 stimulation in mouse RAW264.7 macrophages. The transcription and expression of β-arrestin2 decreased after LPS treatment. Oxytocin preincubation abrogated the down-regulation of β-arrestin2 induced by LPS (Figure 7A). On the contrary, IL-4 enhanced the expression of β-arrestin2, and oxytocin potentiated the effect of IL-4 (Figure 7B). These data indicate that β-arrestin2 might be involved in regulating the effect of OXTR on TLR4- or IL-4R-signaling. Then we used RNA interference to knock down the expression of β-arrestin2 in RAW264.7 cells to prove this hypothesis. After infected with shβ-arr2 lentivirus, the expression of β-arrestin2 in RAW264.7 cells significantly decreased compared with those infected with negative control lentivirus (Figure 7C). After shβ-arr2 lentivirus interference, LPS could still induce the expression of inflammatory cytokines, like IL-1β, IL-6, and TNF-α mRNA in RAW264.7 cells, but oxytocin pretreatment had no effect on LPS (Figure 7D). Similarly, shβ-arr2 lentivirus infection did not change the up-regulation of Arg1 and CD206 induced by IL-4 but abrogated the potentiation of oxytocin (Figure 7E). In shβ-arr2 lentivirus infected cells, LPS could still increase the phosphorylation of p65, but oxytocin lost the inhibitory effect (Figure 7F). Similarly, when β-arrestin2 was knocked down, oxytocin could no longer enhance the phosphorylation of STAT6 induced by IL-4 (Figure 7F). Consistently, the expression of β-arrestin2 in OXTRmyel-KO mice was lower than that in OXTRmyel-WT mice administrated with DSS for 7 days (Figure 6D). In conclusion, oxytocin regulates the macrophage polarization presumably through β-arrestin2 in the G protein non-dependent manner.
Oxytocin modulates the polarization of macrophages through OXTR-β-arrestin2 pathway
Our current study demonstrated that in vitro, oxytocin reduced LPS-induced M1 polarization with impaired expression of inflammatory cytokines and chemokines, but promoted IL-4-induced M2 polarization with a significant increase of expressions of M2 hallmarks, including CD206, Arg1, and Chil3. The OXTR-deficient peritoneal macrophages exerted diminished anti-inflammatory characteristic and enhanced proinflammatory phenotype. For the first time, we demonstrated that oxytocin was directly involved in the modulation of M1/M2 polarization status. Consistent with findings in vitro, OXTR macrophage deficient mice were more susceptible to DSS colitis compared with wild-type mice. The colonic macrophages in OXTR conditioned knockout mice stimulated by DSS exhibited enhanced proinflammatory properties with higher levels of IL-1β, IL-6, and iNOS, as well as lower levels of CD206 and Arg1. The anti-inflammatory effect of oxytocin functioned mainly by strengthening the expression of β-arrestin2, leading to the down-regulation of NF-κB and up-regulation of STAT6 signaling in a GPCR independent manner (Figure 8). β-arrestin2 might become a potential target for the intervention of IBD.
Proposed mechanisms for oxytocin-modulated macrophage polarization
The gut is one of the largest immune organs in the body, which consists of innate and adaptive immune cells and macrophages are one of the most abundant innate immune cells . Intestinal macrophages are continuously replaced by bone marrow-derived monocytes and embryonic precursors might also contribute to the maintaining of resident macrophage populations [6,38]. Challenges such as bacterial translocation or tissue damage will result in the recruitment of highly plastic monocytes from peripheral blood circulation and differentiate to M1-like macrophages in response to the inflammation in gut . The recruited M1-like macrophages produce large amounts of pro-inflammatory cytokines, such as IL-1β, IL-6, TNF-α, and reactive oxygen species, resulting in inflammation amplification, tissue damage, killing of intracellular microbes and promoting Th1/Th17 immune response . The unbalanced production of proinflammatory mediators contributes to the pathogenesis of IBD. Though the population of M1-like macrophages predominates during colitis, there are still M2-like macrophages maintaining the immune homeostasis, promoting the integrity of the epithelium, and producing IL-10 to facilitate the expansion of the regulatory T cells . The specific factors that lead colonic macrophages to polarize to an anti-inflammatory phenotype are not currently understood, but it may be a result from the interaction with IL-10 signaling, TLR-4 signaling, the gut microflora and the intestinal epithelial cells . Regulation of the balance between M1- and M2-like macrophages is critical to the progression of IBD. Several studies suggested that increasing the proportion of colonic M2 macrophages could reduce intestinal inflammation [42,43]. Oxytocin, as a crucial neuroimmunomodulator, could inhibit the transcription of TNF-α in macrophages induced by LPS . But the connection between oxytocin and the M2 polarization was still ambiguous. To the best of our knowledge, our study is the first to report oxytocin promoted polarization of macrophages to a more anti-inflammatory phenotype with enhanced expression of CD206, Arg1, and Chil3.
It is worth mentioning that we found the anti-inflammatory effect of oxytocin on macrophages functioned mainly by enhancing the expression of β-arrestin2, leading to the down-regulation of NF-κB signaling in the absence of G protein activation. As is known, NF-κB family is the most popular player in innate immunity and involved in the polarization of macrophages closely. When LPS binds to TLR4, TRIF- and MyD88-dependent pathway are activated and many other bridging inducers are recruited to induce the degradation of the inhibitor of NF-κB (IκBα) and activation of NF-κB. The activated NF-κB translocates into the nucleus and binds to the promotors of large amounts of inflammatory genes, such as IL-1β, IL-6, iNOS, and ROS, which lead macrophages to M1 polarization . Although molecular mechanisms involved in the modulation of M2 polarization are poorly understood, IL-4-induced STAT6 activation is playing a critical role in the expression of large amounts of genes during M2 polarization and STAT6 signaling shows extensive overlap with NF-κB pathway in modulating the polarization of macrophages .In this article, we revealed that oxytocin could regulate the anti-inflammatory polarization of macrophages through inhibiting the LPS/NF-κB as well as reinforcing IL-4/STAT6 signaling.
Based on the published data, β-arrestins are ubiquitously expressed, and canonically act on the GPCR desensitization and internalization. Recently, the uncanonical function of β-arrestins, serving as signal transduction scaffolds for numerous pathways, has come to light . Jiang et al.  have found that β-arrestin2 attenuated LPS-induced liver injury by inhibiting TLR4/NF-κB pathway. β-arrestin2 was also involved in macrophages to modulate the enhanced inflammatory response during myocardial infarction, indicating β-arrestin2 might have a role in the M2 transition . Gao et al.  also observed that β-arrestin2 directly interacted with IκBα to modulate the activation of NF-κB. β-arrestin2, as an important negative regulator of inflammatory cascade, might become an indispensable intermediate between OXT-OXTR system and NF-κB signaling in macrophages. In our research, we observed that β-arrestin2 participated in the modulation of p65 or STAT6 phosphorylation induced by LPS or IL-4 after the activation of OXTR. The discovery that β-arrestin2 is involved in regulating macrophage polarization is extremely innovative and attractive. The last but not the least, further researches are urgently desired to explore the underlying mechanisms accurately.
Consistent with the findings in vitro, we first established OXTR macrophage-deficient mice and found those conditionally deficient mice were more susceptible to DSS intervention and had a lower percentage of M2-like macrophages in the colon compared with the wild-type mice. When OXTRs were conditionally deleted on macrophages, the colonic macrophages stimulated by DSS showed enhanced proinflammatory properties with higher levels of IL-1β, IL-6 and iNOS, and lower levels of CD206 and Arg1. Therefore, regulation of the intestinal macrophages balance is a potential target for IBD intervention. Numerous researches revealed that oxytocin exerted anti-inflammatory effect in different organs [17,50]. In the intestine, oxytocin is produced by the enteric nervous system (ENS) and the concentration of oxytocin increases in the serum and colon during colitis (Supplementary Figure S5). Szeto et al.  found that under inflammatory stimulation, the expression of OXTRs increased in cultured human and mouse macrophages, responding as an acute phase protein. In the early phase of sepsis, the level of oxytocin in plasma is up-regulated, which diminishes the release of TNF-α, IL-1β, and nitrite in macrophages. Our previous studies demonstrated that oxytocin administrated by intraperitoneal injection alleviated DSS-induced colitis in mice by triggering prostaglandin E2 (PGE2) release of intestine epithelial cells . Welch et al.  also reported that the OXTR-deficient mice were more susceptible to TNBS- and DSS-associated colitis compared with their wild littermates according to clinical and histological damage scores. This was mainly caused by the increased macromolecular permeability of the GI mucosa, although the underlying mechanism was unknown. All the findings above indicated that oxytocin contributed to the elimination of inflammation and restoration of the intestinal homeostasis. As illustrated above, we hypothesize there exists a bidirectional interaction between macrophages and the oxytocinergic neurons in the intestine. Oxytocin, as the interplay between the ENS and the intestinal immune system, especially the macrophages, participates in monitoring the intestinal homeostasis and might become an intriguing therapeutic target for relieving immune diseases and injuries.
In conclusion, oxytocin and its receptors participate in the polarization of macrophages to a more anti-inflammatory phenotype during inflammation, and could have implications for the treatment of IBD.
The phenotypes of macrophages differ depending on the microenvironment and make a difference in the progression of IBD. Oxytocin, as an intestinal neuropeptide, is reported to have a role in alleviating colitis. But the exactly mechanisms are still unclear. Here, we investigated if oxytocin and OXTR were involved in modulating the polarization of intestinal macrophages to cure IBD and the underlying mechanisms.
Our ex vivo and in vivo experiments demonstrated that oxytocin was directly involved in the modulation of M1/M2 polarization status. The anti-inflammatory effect of oxytocin functioned mainly by strengthening the expression of β-arrestin2, leading to the down-regulation of NF-κB and up-regulation of STAT6 signaling in a GPCR independent manner.
Oxytocin might become an important neuroimmunomodulator for the bidirectional regulation between ENS and gut immune system and have implications for the treatment of IBD.
We thank Dr. Chunhong Ma for critical reading of the manuscript, and Dr. Shangming Liu for kindly donating the LysM-Cre transgenic mice.
Study concept: CL, YT, and JL. Experiment design and data acquisition: YT, YS, YG, and XX. YT, TH, XX, CL, and JL contributed to the analysis, interpretation of data and drafting the manuscript. Supervising the study and obtaining funding: CL.
This work was supported by grants from the National Scientific Foundation of China [grant numbers NSFC31471098 and 31671191].
The authors declare that there are no competing interests associated with the manuscript.
C–C motif chemokine ligand
cluster of differentiation
C-X-C motif chemokine
dextran sulfate sodium
enteric neuron system
fetal bovine serum
G protein-coupled receptor
Hanks’ balanced salt solution
inflammatory bowel disease
inhibitor of NF-κB
inducible nitric oxide synthase
lamina propria mononuclear cell
myeloid-specific OXTR knockout mice
myeloid-specific OXTR wild-type control
phorbol 12-myristate 13-acetate
signal transducer and activator of transcription 6
transforming growth factor β
tumor necrosis factor α