D-mannose inhibited BMDCs activation and BMDC-induced antigen-specific T-cell responsesin vitro
An in vitro study was carried out to evaluate the effect of D-mannose relative to glucose on the function of BMDCs. The viability of BMDCs greatly decreased after a 24 h culture in a glucose-free culture medium with 10 mM D-mannose alone (Fig. 1A). The addition of D-mannose to glucose did not affect DC viability (Fig. 1A).Since a glucose-free condition is not physiologically relevant, we compared 10 mM glucose (G10) to the combination of 10 mM glucose and D-mannose (G10M10) in culture media. D-Mannose enhanced the expression of the mannose receptor CD206 (Fig. 1B) whether or not the BMDCs were activated with LPS. The expression of two other surface receptors that regulate DC functions, CD32b (FcγRIIb), and inhibitory receptor, and CD64 (FcγRI), an endocytic receptor involved in antigen uptake [15], were also increased by the addition of D-mannose, even in the presence of LPS stimulation (Fig. 1C-D). Notably, this effect of D-mannose on BMDCs was observed without altering the expression of activation markers (CD40, CD80 and CD86, data not shown). Finally, the production of pro-inflammatory cytokines, IL-6, TNF-α and type 1 IFN (detected through the expression of the interferon stimulated gene Mx1), was decreased at the mRNA level after D-mannose treatment in LPS-activated BMDCs, but no difference was observed in unstimulated BMDCs (Fig. 1E-G). At the protein level, unstimulated BMDCs and BMDCs pulsed with OVA produced lower amounts of TNF-α in the presence of D-mannose (Fig. 1E). Furthermore, OVA-pulsed and LPS-stimulated BMDCs produced lower amounts of IL-6 (Fig. 1F).
In light of the critical role of DCs in priming T cells, we next investigated the effect of D-mannose on the antigen-presenting function of OVA-loaded BMDCs co-cultured with OVA-specific OT-II CD4+ T cells (Fig. 1H). The proliferation of T cells, assessed by CTV dilution, was suppressed when both BMDCs and co-cultures were exposed to D-mannose, but not when the BMDCs alone were loaded with OVA in the presence of D-mannose (Fig. 1I - J). The same result was obtained for the expression of CD69, an early T cell activation marker (Fig. 1K). Although D-mannose may also have an effect on T cells themselves, our findings suggest that D-mannose induces an immature, anti-inflammatory phenotype in DCs that inhibits antigen-specific T cell proliferation and activation during antigen presentation.
D-mannose expanded the steady-state Treg population in non-autoimmune B6 mice
In vivo, we first tested the effect of D-mannose in unmanipulated B6 mice treated for up to 9 weeks. At week 3, the frequency of circulating CD4+ Foxp3+ Treg cells was higher in D-mannose-treated mice (Fig. 2A and B), while the frequency of total CD4+ T cells remained unchanged (Fig. 2C). After 9 weeks of treatment, the frequency of Treg cells was increased in the spleen of D-mannose treated-mice (Fig. 2D). Moreover, the expression of Treg-associated receptors, including CD122, CD132 (the IL-2Rβ and γ chains, respectively), and GITR, was increased in D-mannose treated mice (Fig. 1 E - G). No difference was, however, observed for the expression of CD25, the IL-2Rα high affinity chain (data not shown). There was also a trend for a higher production of IL-10 in the Treg cells from D-mannose treated mice (data not shown). These results showed that D-Mannose expanded the Treg population and the expression of some of its markers under homeostatic conditions.
D-mannose treatment decreased autoimmune activation in the cGVHD model
We next investigated the effect of D-mannose in the cGVHD model, a well-established model of induced systemic autoimmunity [12]. After three weeks of pre-treatment with or without D-mannose, B6 recipients received CD4+ T cells from the semi-allogenic bm12 donor, and continued their treatment with or without D-mannose for another 3 weeks (Fig. 3A). Mice were sacrificed three weeks after induction, which corresponds to the peak of anti-dsDNA IgG production in this model [16]. D-mannose reduced anti-dsDNA IgG production starting 2 weeks after induction (Fig. 3B). This suppressive effect was associated with a lower number of cells in the mLN, and a similar trend was observed in the spleen (Fig. 3C). The frequency and number of splenic B cells remained unchanged (Fig. 3D - F), but their activation measured as the expression of MHC-II was decreased in D-mannose treated mice (Fig. 3G). D-mannose also decreased the frequency of germinal center (GC) B cells in both spleen and mLN, but cell numbers decreased only in mLN (Fig. 3H - I). There was also a reduction in the number of plasma cells of mLN and a similar trend in the spleen (Fig. 3J-K).
CD4+ T cell activation has been a major focus of investigation in the SLE-cGVHD model [17]. D-mannose treatment did not change the frequency or number of total CD4+ T cells (data not shown), but alleviated CD4+ T cell activation measured as the frequency of CD44+CD62L− effector memory (TEM) cells (Fig. 4A - B), as well as the TEM over naive CD44- CD62L+ (TN) ratio in spleen (Fig. 4C). In addition, the frequency of follicular helper (TFH) T cells and the ratio of TFH/follicular regulatory (TFR) in spleen were also reduced after D-mannose treatment (Fig. 4D - E). Contrary to the results observed in non-inflammatory conditions (Fig. 2), the frequency of Treg cells was not altered by D-mannose in cGVHD-induced mice (Fig. 4F). However, cGVHD expanded the frequency of Treg cells in control mice, most likely as a response to inflammation, but not in D-mannose treated mice (Fig. 3G). In addition, a higher expression of the IL-2 receptor gamma chain (CD132) was observed in the splenic Treg cells from the D-mannose treated group in both cGVHD-induced mice (Fig. 3G) and in unmanipulated mice (Fig. 2F).
Finally, we determined the effect of D-mannose on DCs by assessing their frequency and the expression of activation markers. The frequency of plasmacytoid DCs (pDCs) was comparable in D-mannose treated and non-treated mice, but there was a small expansion in the frequency of conventional dendritic cells (cDCs) in D-mannose-treated cGVHD-induced mice (Fig. 5A). In addition, the splenic cDCs from D-mannose treated mice displayed an increased expression of the mannose receptor (Fig. 5B) and a decreased expression of the activation markers CD80 and CD40 (Fig. 5C - D). Overall, these results suggest that D-mannose limits the development of autoimmune activation in cGVHD model.
D-mannose decreased the number of TEM and TFH cells and expanded the frequency of Treg cells in cGVHD recipients
To differentiate the effect of D-mannose on donor or recipient CD4+ T cells in cGVHD, we used B6.SJL as recipients, in which we can distinguish recipient CD4+ T cells (CD45.1-B6.SJL) from the donor CD4+ T cells (CD45.2+ -bm12). D-mannose affected CD4+ T cells in recipient mice, first by decreasing the frequency and number of recipient-derived Tem (Fig. 6A – B) and number of recipient Tfh cells (Fig. 6C – D) in D-mannose treated mice. Moreover, D-mannose increased the frequency of recipient Treg cells (Fig. 6E – F). It should be noted that less than 1% of the donor CD4+ T cells recovered after cGVHD expressed Foxp3 while over 80% of these donor cells expressed a TEM phenotype (data not shown). These results confirm the immunoregulatory effect of D-mannose in the cGVHD model and showed that it does not change the inflammatory phenotypes of the donor cells, but their response in recipient cells.
D-mannose treatment decreased CD4+ T cell activation in a B6.lpr mice
To assess the efficacy of D-mannose in a spontaneous model of lupus, we selected B6.lpr mice, a simplified model driven by FAS deficiency. Treatment was started at 4 months of age, when B6.lpr mice start to produce autoantibodies, and lasted for 10 weeks. There was an initial reduction of anti-dsDNA IgG production, which reached significance at week 4 after treatment, but it was not sustained thereafter (Fig. 7A). There was a trend for an increase of the frequency of TN cells (Fig. 7B) and Treg cells (Fig. 7C) by the D-mannose treatment. CD4+ T cell activation, measured as CD44 expression (Fig. 7D) and the TFH/TFR ratio, (Fig. 7E) was reduced by the D-mannose treatment. Finally, splenic cDCs from D-mannose treated mice expressed higher levels of CD206 (Fig. 7F), which was in line with the result obtained in the cGVHD model. The D-mannose treatment did not affect the number and distribution of cDC1 and cDC2 subsets (data not shown). These results showed that the immunoregulatory effect of D-mannose is not restricted to the cGVHD induced model of lupus, but also applies to the B6.lpr model, in which autoimmunity arises from a completely different pathway.