Novel Foxp3− IL-10− Regulatory T-cells Induced by B-Cells Alleviate Intestinal Inflammation in Vivo

Recent studies have revealed various Foxp3− regulatory T (Treg) cell subsets effectively protect mice from colitis. In the present study, we demonstrated that B cells induced a particular subset of regulatory T (Treg-of-B) cells, expressing programmed cell death 1 (PD-1), inducible costimulator (ICOS), lymphocyte-activation gene 3 (LAG3), glucocorticoid-induced tumor necrosis factor receptor (GITR), and OX-40, did not express Foxp3. Treg-of-B cells produced abundant levels of IL-10 and low levels of IL-4 and TGF-β. Adoptive transfer of Treg-of-B cells protected mice from CD4+CD45RBhi T-cell-induced colitis, including infiltration of leukocytes, depletion of goblet cells, epithelial hyperplasia, and inhibition of Th1 and Th17 cytokines. These features were similar to IL-10-producing type 1 regulatory T (Tr1) cells; however, IL-10-deficient Treg-of-B cells maintained their suppressive function in vitro as well as in vivo, while the regulation of Tr1 cells depended on IL-10. In conclusion, Treg-of-B cells protected against experimental colitis through an IL-10-independent mechanism. We reported a novel subpopulation of regulatory T cells was different from conventional Foxp3+ Treg and IL-10-producing Tr1 cells.

In this study, we investigated the role of IL-10 in the regulatory mechanism of Treg-of-B cells in a murine model of IBD. Treg-of-B cells, induced by anti-CD3 and anti-CD28 antibodies in the presence of B cells, suppressed the proliferation of T cells in an IL-10-independent manner. Adoptive transfer of Treg-of-B cells protected mice from T cell-mediated experimental colitis. Although Treg-of-B cells expressed LAG3 and IL-10 as IL-10-producing Tr1 cells did, we showed that IL-10-deficient Treg-of-B cells exerted suppressive ability in vitro as well as in vivo. The results indicated that Treg-of-B cells possessed IL-10-independent regulatory function and differed from Foxp3 + nTreg and IL-10-producing Tr1 cells.

Result
The characteristics of Treg-of-B cells. To analyze the characteristics, Treg-of-B cells were compared to both naïve T cells and natural regulatory T (nTreg) cells. We found that Treg-of-B cells expressed high levels of Treg-related molecules, such as CD44, GITR, ICOS, LAG3, and OX40. Treg-of-B cells also expressed low levels of PD-1 and CTLA4. Unlike nTreg cells, Treg-of-B cells did not express Foxp3 (Fig. 1A). We next examined the cytokine profile of Treg-of-B cells. Treg-of-B cells produced higher levels of IL-10, IL-4, and transforming growth factor (TGF)-β compared to that of naïve T cells. Treg-of-B cells also produced higher level of interferon (IFN)-γ than that of nTreg cells (Fig. 1B).
To examine the suppressive ability of Treg-of-B cells, a suppression assay was performed. Treg-of-B cells successfully inhibited effector T cell proliferation (Fig. 1C). Despite the absence of Foxp3 expression, the suppressive (C) The suppressive function of Treg-of-B cells was analyzed with nTreg cells as a suppressor control. *P < 0.05; **P < 0.01; ***P < 0.001; not significant (NS) indicates P > 0.05. The results are represented as the mean ± SEM. Significant differences were calculated using Student's t-test. The data are representative of three independent experiments.
Scientific RepoRts | 6:32415 | DOI: 10.1038/srep32415 capacity of Treg-of-B cells was similar to nTreg cells. Together, these data suggested that B cells induced functional Foxp3 − regulatory T cells that produce IL-10.

Adoptive transfer of Treg-of-B cells protects against colitis.
To determine the ability of Treg-of-B cells to protect against colitis induced by colitogenic T cells, CD4 + CD45RB hi colitogenic T cells were adoptively co-transferred into SCID mice, with either PBS (control), Treg-of-B cells or CD4 + CD45RB lo cells. A decrease in body weight was used as a measurement of disease progression. After eight weeks, the mice that were given Treg-of-B or CD4 + CD45RB lo cells had similar body weight. However, the body weight of mice that were given PBS as a control had decreased significantly ( Fig. 2A). Colitis symptoms were also presented in mice given colitogenic T cells alone, but not in mice that were also co-injected with Treg-of-B cells or CD4 + CD45RB lo cells (Fig. 2D). The histological analysis of the colonic tissue showed the presence of disease, characterized by an infiltration of leukocytes, depletion of goblet cells, and epithelial hyperplasia. In contrast, mice that were given Treg-of-B cells had mild colitis features, and mice treated with CD4 + CD45RB lo cells had sporadic colitis features (Fig. 2B,C).

Treg-of-B cells inhibited Th1/Th17 cytokine protected in vivo.
To address whether the addition of Treg-of-B cells can suppress inflammatory cytokine production in the colons of mice adoptively transferred with colitogenic T cells, the levels of cytokines in cultured supernatant from colon explants and mesenteric lymph nodes (MLNs) were assessed. The levels of both IFN-γ and IL-1β in colon cultures significantly decreased in mice that had received either CD4 + CD45RB lo or Treg-of-B cells, and there were no significant differences between these groups (Fig. 3A). Co-transfer of either CD4 + CD45RB lo or Treg-of-B cells reduced the levels of IFN-γ , tumor necrosis factor (TNF)-α , IL-6, and IL-17 in the MLNs of mice with colitis. Therefore, Treg-of-B cells reduced Th1 and Th17 cytokine production to a level comparable to CD4 + CD45RB lo cells (Fig. 3B).  Treg-of-B cells expressed additional suppressive molecules to compensate for the lack of IL-10, we compared the signatures of WT and IL-10 KO Treg-of-B cells. The majority of the Treg-associated markers expressed by WT and IL-10 KO Treg-of-B cells were the same. However, the expression of ICOS was lower in IL-10 KO Treg-of-B cells compared to that of WT Treg-of-B cells (Fig. 4B). Next, we examined the cytokine profiles of WT and IL-10 KO Treg-of-B cells. We found that that the IL-10 KO Treg-of-B cells produced slightly higher level of IFN-γ , yet similar amounts of IL-4 and TGF-β . As expected, IL-10 KO Treg-of-B cells did not produce IL-10 ( Fig. 4C).

IL-10 KO Treg-of-B cells protected against colitis.
To assess whether IL-10 is necessary for Treg-of-B cells to protect against colitis, CD4 + CD45RB hi effector T cells were adoptively transferred into SCID mice with either WT or IL-10 KO Treg-of-B cells. After six weeks, mice that had received either WT or IL-10 KO Treg-of-B cells exhibited less weight loss ( We next examined the levels of inflammatory cytokines in culture supernatants from colon explants and MLNs. Compared with mice that received colitogenic T cells alone, mice that were co-transferred with either WT or IL-10 KO Treg-of-B cells had reduced levels of IFN-γ and IL-1β in the colon (Fig. 6A), and reduced amounts of IFN-γ , TNF-α , IL-17, IL-6, and IL-1β in the MLNs (Fig. 6B). Importantly, there was no difference between the mice that had received WT Treg-of-B cells and those that received IL-10 KO Treg-of-B cells.

Discussion
In this study, we investigated the characteristics and regulatory function of Treg-of-B cells and examined whether IL-10 is necessary for the protective effects. Using an adoptive transfer colitis model in SCID mice, we found that Treg-of-B cells were able to successfully inhibit disease progression. We also examined whether IL-10 KO Treg-of-B cells retained the capability to suppress T cell proliferation and prevent colitis, both in vivo and in vitro. Overall, we investigated the function of Treg-of-B cells and found that these cells were different from those of IL-10-producing Tr1 cells.
B cells are important to maintain intestinal homeostasis and exert their function by secreting immunoglobulin and cytokines. In addition to their effector functions, regulatory B (Breg) cells modulate Treg cell development, proliferation, and survival 19 . The absence of B cells result in the severe and rapid onset of experimental autoimmune encephalomyelitis and coliti suggesting that there is a link between B cells and Treg cells 20 . Moreover, recent studies suggest that B cells can actively induce Treg cells. Orally administration of cholera toxin B conjugated antigen increase the frequency of antigen-specific Foxp3 + T but not in B cell-deficient mice 21 . IL-10-secreting B cells can induce Tr1 cells and thereby ameliorate T cell-mediated colitis 22 . In addition, the cell-cell interactions between antigen-specific B cells and naïve T cells creates a mature immunologic synapse and can induce the de novo generation of Treg cells 6,23 . Our present results also suggested that antigen-nonspecific Treg-of-B cells induced by conventional B-2 cells in the presence of anti-CD3 and anti-CD28 antibodies 9 (Fig. 1). The linage factor Foxp3 has long been known as an essential regulator of Treg cells for their function and development 24 . However, recent studies reported the existence of the Foxp3 − Treg subsets, including Tr1 and CD4 + CD25 − LAG3 + T cells 2 . Here, we found that Treg-of-B cells, induced by anti-CD3 and anti-CD28 antibodies, were also Foxp3 − suppressive T cells (Fig. 1A). We further analyzed the characteristics of Treg-of-B cells, the phenotype and the production of IL-10 were similar to those of nTreg and Tr1 cells 25,26 (Fig. 1A,B).
The immunosuppressive function of Treg-of-B cells has been shown in many animal disease models. Reichardt et al. demonstrated that Treg-of-B cells potently suppress cutaneous hypersensitivity and ectopic allogeneic heart transplant rejection 6 . Our group has also demonstrated that Treg-of-B cells induced by splenic B-2 and Peyer's patch B cells modulated airway inflammation in the murine model of allergic asthma 7,27 . The present study examined the function of Treg-of-B cells in a T cell-mediated colitis model. The co-transfer of Treg-of-B cells with CD4 + CD45RB hi effector T cells into SCID mice effectively inhibited the symptoms of colitis and reduced the production of Th1 and Th17 cytokines in the MLNs and colons (Figs 2 and 3). These protective effects are similar to those seen with the co-transfer of CD4 + CD45RB lo T cells, which is a Treg-enriched T cell population that has been previously shown to inhibit the onset of colitis 28 .
Many studies have shown that the mucosal production of IL-10 promote intestinal tolerance. IL-10 has been shown to suppress Th1 and Th17 responses in colitis [29][30][31] , and Tr1 cells induced by immature dendritic cells also require IL-10 32 . CD4 + CD25 − LAG3 + T cells protect mice form colitis in an IL-10-dependent manner 4 . We speculated that IL-10, which is highly secreted by Treg-of-B cells, might be responsible for the protective effects of Treg-of-B cells in colitis. However, our in vitro and in vivo data show that IL-10 deficiency did not affect the immune regulatory functions of Treg-of-B cells. IL-10 KO Treg-of-B cells successfully suppressed responder T cell proliferation (Fig. 4A); moreover, IL-10 KO Treg-of-B cells were able to attenuate chronic colitis induced by the colitogenic T cells (Fig. 5) and suppress the Th1 and Th17 response (Fig. 6). We hypothesized that IL-10 KO Treg-of-B cells might upregulate other immune modulator molecules to compensate for the loss of IL-10, but we did not find increased expression of any other regulatory molecules in this study (Fig. 4B,C).
It is unclear that IL-10 production was not necessary for Treg-of-B cells to protect against colitis in our study. Although IL-10 KO CD4 + CD25 + Treg cells are less effective than WT cells, they can still prevent T cell-mediated colitis 33 . These suggest that Treg cells can inhibit colonic inflammation through other mechanisms other than the secretion of IL-10. On the other hand, our previous data showed that the suppressive capability of Treg-of-B2 cells reduced in the presence of a transwell insertion, suggested that Treg-of-B2 cell-mediated suppression required cell-cell contact 9 . These results suggest that surface molecules expressed on IL-10 KO Treg-of-B cells may play a role in the suppressive function.
Treg-of-B cells expressed several regulation-associated molecules, including CTLA-4, GITR, OX40, LAG3, and PD-1. These molecules in Treg cells can control the activation of antigen presenting cells and lead to the accumulation of Treg cells in the colon [34][35][36][37][38] . Our group also found that LAG3 + Treg-of-B cells induced by Peyer's patch B cells could alleviate airway hypersensitivity 8 . Taken together, these data provide hints about how IL-10 KO Treg-of-B cells utilize other regulatory pathways to attenuate the severity of colitis.
We also found that Treg-of-B cells share a similar phenotype with Tr1 cells. Currently, there is no lineage-defining transcription factor or signature cellular surface markers for Tr1 cells. Their characterization is based on cytokine profile (IL-10 hi IL-4 − IFN-γ lo ) and IL-10-dependent suppression mechanisms 26,39 . In vitro cultued, OVA-specific Tr1 cells prevent colitis through the IL-10 production 16 . Therefore, IL-10-independent regulatory mechanisms might provide a unique feature to distinguish Treg-of-B cells from Tr1 cells. Our group found that Treg-of-B cells did not express CD49b or CD103, both of which expressed in Tr1 cells 8 . In addition, IL-10 KO B2 cells still induced Treg-of-B cells 9 , whereas the induction of Tr1 cells requires IL-10 16 .
In conclusion, these findings shed new light on Treg-based therapies for experimental colitis. Treg-of-B cells inhibited colitis and suppressed Th1 and Th17 responses in an IL-10-independent manner. In addition, unlike the IL-10-dependent regulatory mechanisms of Tr1 cells, IL-10 is not necessary for Treg-of-B cell-mediated suppression (Fig. 7). Our study here is the first one to demonstrate the effectiveness of IL-10 deficient Treg-of-B cells in vivo. The generation of Treg-of-B cells in vitro might potentially be utilized as a new approach for IBD therapy. However, further studies are needed to understand the detailed immune modulatory mechanisms of Treg-of-B cells, distinguish them from other Treg subtypes and utilize Treg-of-B cells therapeutically in human IBD.

Methods
Mice. Female C.B17/Icr-Prkdc SCID /Crl (SCID) and BALB/c mice were from the National Laboratory Animal Center aged 6-8 weeks. BALB/c Il10 −/− (KO) mice were purchased from Jackson Laboratory. All mice were maintained in Laboratory Animal Center of the College of Medicine at National Taiwan University. All animal To analyze the production of cytokines, an ELISA was performed using the DuoSet ELISA Development kit (R&D, Minneapolis, MN, USA) according to the manufacturer's instructions. The following cytokines were assayed: IL-1β , IL-4, IL-6, IL-10, IL-17, TNF-α , IFN-γ , and TGF-β . injected with CD4 + CD45RB hi effector T cells (5 × 10 5 per mouse) from BALB/c mice. Subsequently, mice were i.p. injected with CD4 + CD45RB lo T cells (2.5 × 10 5 per mouse), Treg-of-B cells (3 × 10 6 per mouse) or PBS. The body weight of each mouse was recorded weekly to assess disease severity.
Colitis scoring and histological assessment of tissue sections. The criteria for the colitis disease score has been described elsewhere 40 . The mice were clinically evaluated prior to their sacrifice. Colitis disease score was scored as follows: 0, normal; 1, soft or loose stools; 2, diarrhea; 3, blood in stools or prolapsed rectum.
Statistical analysis. The statistical analyses were performed using GraphPad Prism 5 software.
Significant differences were determined using the unpaired Student's t-test or one-way Analysis of Variance (ANOVA). A value of P < 0.05 was considered to be statistically significant. All of the data were expressed as the mean ± SEM.