The onset of type 1 diabetes in NOD mice is delayed by oral administration of a bacterial extract (OM-85) and can be completely prevented by its intraperitoneal administration. Optimal prevention is observed when starting treatment at 3 or 6 weeks of age, and some effect is still observed with treatment at 10 weeks of age. Using genetically deficient mice and cytokine-neutralizing monoclonal antibodies, we demonstrate here that the therapeutic effect does not involve T-helper type 2 cytokines (interleukin [IL]-4 and -10) but is tightly dependent on transforming growth factor (TGF)-β. Natural killer T-cells also participate in the therapeutic effect because CD1d−/− NOD mice are partially resistant to the protective effect of OM-85. The question remains of the specificity of the protective effect of OM-85, which may include proinflammatory components. It will thus be important to further characterize the molecular components that afford protection from type 1 diabetes. Lipopolysaccharide is excluded, but other Toll-like receptor (TLR) agonists could be involved because OM-85 stimulated dendritic cells and induced TGF-β production by splenocytes in a TLR-2–, TLR-4–, and MyD88-dependent fashion.

The incidence of type 1 diabetes is steadily increasing in Western countries (1). There is a clear parallel trend toward decreased infectious load in these countries linked to improved hygiene and medical care, including extensive usage of vaccinations and antibiotics (23). A similar association is observed in animal models of the disease, such as nonobese diabetic (NOD) mice and Bio-Breeding (BB) rats. The causal relationship between the low rate of infections and high disease incidence was directly demonstrated in the NOD mouse by the increased disease incidence observed after decontamination of mice bred in conventional facilities (2) and by the disease prevention achieved by deliberate infection with a variety of bacteria, viruses, and parasites (49). The mechanisms mediating the protective effect of infections on type 1 diabetes are still ill defined.

Here, we present a systematic analysis of the potential mechanisms of diabetes protection in NOD mice afforded by OM-85, an extract manufactured from bacterial species frequently responsible for respiratory tract infections. We took advantage in this study of the availability in the laboratory of several strains of genetically deficient IL4−/− and CD1d−/− NOD mice and of antibodies that neutralize transforming growth factor (TGF)-β and block the interleukin (IL)-10 receptor. The results obtained indicate the multifactorial nature of diabetes protection by OM-85, involving TGF-β and natural killer T (NKT) cells.

Female NOD mice (Kd, I-Ag7, and Db) were bred in our animal facility. Female CD1d−/− NOD mice were originally provided by Dr. L. Van Kaer at backcross 6 and were bred to backcross 12 in our facility. Female IL-4−/− NOD mice were kindly provided by Dr. D. Mathis. MyD88−/−, TLR-2−/−, and TLR-4−/− lines were kindly provided by Dr. S. Akira and backcrossed to C57BL/6 mice in our animal facility. All mice were bred under specific pathogen-free conditions and fed a standard diet (γ-irradiated R04-10; UAR, Augy, France).

Diabetes monitoring.

Mice were monitored once a week for glycosuria, using colorimetric strips (Glukotest; Boehringer-Mannheim, Mannheim, Germany), until diabetes occurred. Diabetes was confirmed by evaluating fasting glycemia (Hemoglukotest and Reflolux F; Boehringer-Mannheim). All mice were culled when three consecutive glycosuria and hyperglycemia tests were found to be positive.

OM-85 extract.

OM-85 (Broncho-Vaxom, OM Pharma, Meyrin/Geneva, Switzerland) is an extract composed of both gram-positive and -negative bacterial species frequently responsible for respiratory tract infections: Haemophilus influenzae, Diplococcus (Streptococcus) pneumoniae, Klebsiella pneumoniae and ozaenae, Staphylococcus aureus, Streptococcus pyogenes and viridans, and Neisseria (Branhamella) catarrhalis. It mainly consists of acidic proteins, peptides, and amino acids. During the production process, bacteria are submitted to alkaline treatment. This leads to chemical modifications (such as partial hydrolysis of proteins), which shifts the isoelectric points toward low values characteristic of acidic proteins (isoelectric point between 2.8 and 5.2) and generates smears of proteins and peptides ranging from a few hundred to a few hundred thousand daltons, as detected by SDS-PAGE analysis. By gel filtration, the major peak is observed between 7 and 8 kDa. Importantly, OM-85 also includes, though in minor quantities, other components (i.e., detoxified lipopolysaccharides, lipid A molecules and lipoteichoic acids, sugars, and fatty acids). The final OM-85 lyophilizate contains 17.5% active principle (bacterial extract) and 82.5% excipients (mannitol, sodium glutamate, and propyl gallate). Doses administered are referred to as milligrams of active principle. For each experiment, a fresh solution was prepared daily.

OM-85 oral administration.

Each mouse received 1, 10, or 25 mg/day of OM-85 active principle orally, 5 days a week for 10 or 30 weeks. The bacterial extract was dissolved in drinking water. Control mice received drinking water without OM-85.

OM-85 intraperitoneal administration.

For intraperitoneal injection, fresh solutions of OM-85 were dissolved in saline (NaCl 0.9%). Each mouse received 1 mg of OM-85 active principle per injection, 5 days a week for 10 weeks.

Different compounds were used as controls. In some experiments control mice received intraperitoneal injections of saline without OM-85. In addition, two proinflammatory compounds devoid of any bacterial component, namely, incomplete Freund’s adjuvant (IFA) and N-acetylmuramyl-l-alanyl-d-isoglutamine (most commonly termed muramyldipeptide [MDP]) (1012), served as control treatments. In these experiments each NOD mouse received at 4 or 10 weeks of age a footpad injection of either MDP (25 μg dissolved in NaCl) emulsified in 50 μl IFA or saline emulsified in 50 μl IFA. As a positive control for IFA and MDP+IFA, some NOD mice also received a footpad injection of a complete Freund’s adjuvant (CFA) emulsion according to an identical experimental protocol (at 4 or 10 weeks of age), as previously described (13).

Anticytokine treatment.

When needed mice were treated with the 2G.7 monoclonal antibody to human TGF-β that cross-reacts and neutralizes in vivo murine TGF-β (14) (provided by C.J.M. Melief, Leiden University Medical Center, Leiden, the Netherlands) or with a rat monoclonal antibody specific to the mouse IL-10 receptor (15) (provided by A. O’Garra, DNAx, Palo Alto, CA). The antibodies were administered at a dose of 1 mg · mouse−1 · injection−1 i.p. three times a week for 4 weeks.

Limulus amebocyte assay.

The absence of any significant endotoxin activity in OM-85 preparation was checked by a limulus amebocyte assay (Invitrogen, Clergy Pontoise, France) following the manufacturer’s recommendations.

Preparation and culture of dendritic cells.

Primary bone marrow dendritic cells were isolated from femora of wild-type or genetically Toll-like receptor (TLR)- or MyD88-deficient mice. Briefly, bone marrow was flushed out with Hanks balanced salt solution supplemented with 100 units/ml penicillin and 100 μg/ml streptomycin. Cells were cultured at 2 × 106 cells/ml and per well in RPMI-1640 medium supplemented with 10% fetal bovine serum, 25 mmol/l HEPES, 1% minimum essential medium nonessential amino acid, 1% minimum essential medium sodium pyruvate, 0.5 mmol/l β-mercaptoethanol, 0.1 mg/ml gentamicin, and 10% supernatant of XR663 cell line–secreting granulocyte macrophage-colony stimulating factor. Fresh medium was added every 3 days. At day 9, immature dendritic cells were collected and stimulated by different concentrations of OM-85 in fresh medium for 24 h. Supernatants were then collected and assayed using a murine IL-12p40 enzyme-linked immunosorbent assay kit (R&D Systems, Lille, France).

Cell recovery and T-cell assays.

For analysis of cytokine production, total splenocytes or purified T-cell subsets were cultured in triplicate in 96-well flat-bottom microtiter plates (Costar, Cambridge, MA) at 2 × 105 cells/well in the presence of OM-85 (at different doses as indicated in figure legends) or medium alone. Cells were cultured at 37°C in a humidified atmosphere containing 5% CO2 in RPMI supplemented with 10% heat-inactivated FCS, 2 mmol/l l-glutamine, 100 units/ml penicillin, 100 μg/ml streptomycin (Life Technologies, Cergy Pontoise, France), and 0.5 mmol/l β-mercaptoethanol, and they were stimulated with immobilized anti-CD3 (145 2C11, 5 μg/ml final concentration; provided by Dr. J.A. Bluestone, University of California at San Francisco, San Francisco, CA). Supernatants were recovered at 24 h (to measure γ-interferon [IFN-γ]), 48 h (to measure IL-4 and -10), and 72 h (to measure TGF-β) and stored at −80°C until tested.

Enzyme-linked immunosorbent assay for cytokine production.

For cytokine measurement DuoSet kits quantifying TGF-β, IFN-γ, IL-4, IL-12p40, and IL-10 (R&D Systems) were used. Recombinant cytokines were used as standards. In vitro stimulation assays were repeated at least three times. Measurements were performed in duplicate or triplicate. Data are the means ± SE.

Statistical analysis.

Diabetes incidence was plotted using the Kaplan-Meier method, i.e., nonparametric cumulative survival plot. Statistical comparisons between the curves were performed using the log rank (Mantel-Cox) test, which provided the corresponding χ2 values. When needed, statistical comparison of mean values was performed using Student’s t test.

Diabetes prevention after intraperitoneal administration of OM-85.

In our colony, female NOD mice spontaneously develop overt diabetes starting at 14–16 weeks of age (usually with an incidence of ∼80% at 35 weeks of age). Female NOD mice were treated with OM-85 starting at 3, 6, 10, or 15 weeks of age. The product was administered at a dose of 1 mg/day i.p. of active principle, 5 days a week for 10 weeks.

In the group of mice injected with OM-85 from 3 weeks of age, a significant and long-term protection from diabetes was observed (Fig. 1A and Table 1). A similar protection was observed when treatment was started at 6 weeks of age (Fig. 1B) (experimental groups treated with OM-85 included six mice each; the control group included five mice). When treatment was started at 10 or 15 weeks of age, a partial though nonsignificant protection was observed.

As additional controls, in parallel to saline, we tested two proinflammatory compounds, namely, IFA and synthetic MDP. MDP is a component of the mycobacterial cell wall, initially characterized by Lederer and Chedid and colleagues (1011), which is nonimmunogenic per se while expressing proinflammatory adjuvant-like properties. In fact, it acts as an hapten when conjugated with various carriers (1012,1618). As a positive control for IFA and MDP+IFA, NOD mice were also treated with a CFA emulsion as previously described (13). Importantly, no protection was observed after parenteral administration of IFA or MDP+IFA in experimental conditions where CFA showed full protective activity (10 mice were included in each experimental group) (Fig. 2). One may also note that CFA, like OM-85, is active when treatment is started at a young age, but it loses most of its activity when given at later disease stages.

Effect of oral administration.

Mice were treated orally with different doses of OM-85 (1, 10, or 25 mg/day of active principle), 5 days a week for 10 or 30 weeks. The treatment was started at 3, 6, 10, or 15 weeks of age. Results of several experiments showed some variability. Consistent, though statistically borderline, protection (P = 0.056) was obtained after administration of 25 mg OM-85 for 10 weeks, starting treatment at 6 weeks of age (each experimental group—OM-85 and controls—included seven mice) (Fig. 1C). Some protection was also observed with the 10-mg dose, starting treatment at 6 weeks of age. No effect was observed with the 1-mg dose, even when starting treatment at 3 weeks of age (P = 0.477), a schedule fully efficacious when using the intraperitoneal route (Fig. 1A).

Diabetes protection by OM-85 does not depend on T-helper type 2 cytokines.

IL-4 dependency was studied by comparing the effect of 1 mg OM-85 administered intraperitoneally in IL-4−/− and wild-type NOD mice. A similar protection was observed in the two strains (OM-85–treated NOD IL-4−/−: n = 9; control NOD IL-4−/−: n = 6; no significant difference, P = 0.216) (Fig. 3A), thus suggesting that IL-4 is not implicated in the therapeutic effect of OM-85 .

To study the role of IL-10 on the effect of OM-85 in type 1 diabetes, OM-85 was administered intraperitoneally in NOD mice from 3 to 13 weeks of age, followed by administration of a monoclonal antibody directed to the IL-10 receptor. The antibody was given at a dose of 1 mg three times a week, from 13 to 16 weeks of age. As shown in Fig. 3B, the treatment did not abrogate the protective effect of OM-85 (OM-85–treated NOD: n = 10; OM-85 + anti–IL-10 receptor–treated NOD: n = 7; control NOD: n = 11; P = 0.600). Similar data were obtained in a second experiment (data not shown).

Contribution of TGF-β to the protective effect of OM-85.

To analyze the possible role of TGF-β in the OM-85–mediated protection of diabetes, treatment with OM-85 (1 mg/day i.p. from 3 to 13 weeks of age) was followed by the administration of a neutralizing anti–TGF-β antibody, given at the dose of 1 mg i.p. three times a week, from 13 to 16 weeks of age. Results shown in Fig. 3C clearly indicate that anti–TGF-β abrogated the protective effect of OM-85 (OM-85–treated NOD: n = 10; OM-85 + anti–TGF-β–treated NOD: n = 6; control NOD: n = 11; P = 0.026), thus suggesting that OM-85–mediated protection is a TGF-β–dependent active phenomenon. Similar data were obtained in a second experiment (data not shown).

Role of NKT cells in OM-85–mediated protection.

CD1d−/− mice are devoid of NKT cells, whose intrathymic selection depends on CD1d (19). CD1d−/− NOD mice were treated with OM-85 (1 mg/day i.p., 5 days a week for 10 weeks, starting treatment at 3 or 4 weeks of age, depending on the experiment). In both experiments shown in Fig. 4, diabetes incidence was similar in CD1d−/− and wild-type NOD mice. The following number of mice were used in each experiment presented: Fig. 4A: n = 10 for OM-85–treated NOD CD1d−/−, n = 10 for control NOD CD1d−/−, n = 10 for OM-85–treated NOD, and n = 11 for control NOD; Fig. 4B: n = 5 for OM-85–treated NOD CD1d−/−, n = 11 for control NOD CD1d−/−, n = 5 for OM-85–treated NOD, and n = 12 for control NOD. However, it appeared that CD1d−/− NOD mice were less well protected from diabetes by OM-85 than wild-type NOD mice, though not significantly (cumulative analysis P = 0.087). This suggests that the OM-85 protective effect may depend, but only in part, on NKT cells.

Search for TLR ligands in OM-85.

To test whether OM-85 contains components that might stimulate TLRs and thus potentially contribute to the immunomodulatory effect of the product, its capacity to stimulate cytokine production in a TLR-dependent fashion was examined. The experimental model consisted of incubating primary bone marrow dendritic cells from C57BL/6 or genetically TLR- or MyD88-deficient mice with different concentrations of OM-85. As shown in Fig. 5, OM-85 stimulated the production of IL-12 in vitro. This response was not observed after stimulation of dendritic cells obtained from MyD88−/− mice. When using dendritic cells from TLR-4−/− and TLR-2−/− mice, total or partial inhibition of IL-12 production was observed, respectively (Fig. 5). These results show that in dendritic cells OM-85 signals exclusively via the MyD88 pathway, at least partially by TLR-2 and -4.

OM-85 modulates TGF-β production via the TLR-MyD88 pathway.

We next tested the ability of OM-85 to stimulate the in vitro production of IL-4, IL-10, IFN-γ, and TGF-β using murine splenocytes. Importantly, OM-85 stimulation induced very high production of TGF-β (which was maximal already at 1 μg/ml in wild-type mice), high production of IL-10 (from 10 μg/ml), and moderate production of IFN-γ (from 100 μg/ml), whereas IL-4 production was undetectable (Fig. 6). Interestingly, TGF-β and IL-10 production were abolished in MyD88−/− mice (Fig. 6). TGF-β production was also completely abolished in TLR-4−/− mice and partially in TLR-2−/− mice compared with wild-type mice. The presence of TLR-4 agonists in OM-85, as suggested by these results, prompted us to examine the content of lipopolysaccharide in OM-85, using a limulus amebocyte assay. The level of lipopolysaccharide was found to be lower than that required for stimulating mononuclear phagocytes (<0.1 pg of lipopolysaccharide per milligram of OM-85). One may thus assume that such traces of lipopolysaccharide cannot account for the observed TLR-4–mediated stimulatory capacity of OM-85.

The mechanisms of the protective effect of infections on type 1 diabetes are still elusive. Limited data are available in models based on infections in autoimmunity-prone mouse strains (2). Simplified approaches using bacterial components or viral proteins are difficult because there is no indication of the nature of the molecules responsible for the protection. It was thus of particular interest to use a bacterial extract that would putatively contain the various components participating in bacteria-mediated protection, whose handling is easier than live infections. In this vein, we selected OM-85, a bacterial extract prepared from eight bacterial species frequently responsible for respiratory tract infections. It has been reported earlier that another bacterial extract, OM-89, can afford protection in animals and/or humans against autoimmune diseases such as rheumatoid arthritis (2021), lupus erythematosus (22), spondyloarthropathies (23), and diabetes (2425). The product tested in the present study (OM-85) is of particular pertinence because it has extensively and safely been used in children suffering from repeated upper respiratory tract infections (2627), thus paving the road to investigate its preventive potential in children prone to developing type 1 diabetes.

Results obtained in this study first show that OM-85 can prevent or delay type 1 diabetes onset when administered intraperitoneally at a low dosage. The product can also afford protection when given orally, but it must then be administered at a higher dosage. Importantly, the effect is optimal when the treatment is started early (3 or 6 weeks of age), but some protection is still achieved when the treatment is started at 10 weeks of age. Such kinetics are similar to those described for CFA (13). Importantly, two proinflammatory adjuvant-like yet nonimmunogenic compounds, namely, IFA and MDP (1012,1618), did not confer protection from diabetes, as opposed to OM-85 bacterial extract. However, currently, one cannot exclude that some proinflammatory compounds included in OM-85 actively participate in its protective effect in this model.

We examined the role of T-helper type 2 cytokines, namely IL-4 and -10, in the protective effect of OM-85. No responsibility of IL-4 or -10 could be evidenced by using genetically deficient mice for IL-4 and a neutralizing monoclonal antibody for IL-10 receptor. It is interesting to note that in the particular setting of newborn rats, OM-85 increased the production of IFN-γ and decreased that of IL-4 by spleen cells (28). When a neutralizing anti–TGF-β antibody was administered immediately after OM-85, a major role was evidenced for TGF-β because the protective effect of the drug was essentially lost. It is not clear, however, which cells produce TGF-β in this context and what its cellular target is. TGF-β could be produced by regulatory T-cells or induce their differentiation, as suggested in other models (2933). Our in vitro studies show that OM-85 stimulates dendritic cells and induces very high levels of TGF-β production in a TLR-dependent fashion by splenocytes (TGF-β production was not observed using splenocytes derived from TLR-4−/− or MyD88−/− mice). This TLR-mediated effect of OM-85 may also involve TLR-2, although TLR-2−/− mouse splenocytes are still at least partly sensitive to OM-85. Other cytokines are produced on OM-85 stimulation, such as IL-12 and –10, which could open other perspectives. In any event, it is interesting to mention our unpublished recent results showing that various chemically defined TLR agonists afford type 1 diabetes protection in the same manner as OM-85: Pam3Cys and P40 for TLR-2, poly(I:C) for TLR-3, and lipopolysaccharide for TLR-4. As far as OM-85 is concerned, lipopolysaccharide is probably not the moiety involved in TLR-4 stimulation because it is not found in the extract, even at a very low concentration.

We recently reported a critical role for CD1d-dependent NKT cells in exacerbated T-helper type 2–mediated airway inflammation and hyperresponsiveness (34) and in autoimmune diabetes–prone NOD mice (3536). The data presented in this article confirm that NKT cells are implicated in type 1 diabetes. One should mention, however, that in our present study, both mouse strains (CD1d−/− and wild type) showed a similar type 1 diabetes incidence, a finding in clear contrast with another report (37), but in the line of that of Van Kaer (38), who provided us with the mice used here. If NKT cells are indeed involved in the OM-85–mediated protection, it will be important to determine which component within OM-85 stimulates NKT cells and what mechanisms mediate NKT cell involvement. IL-4, as suggested by Oki et al. (39) and Mi et al. (40), and IL-10, as suggested by Mi et al. (41), could be incriminated, even though we have a priori excluded these two cytokines as major protagonists in our type 1 diabetes model. In any case, the role of NKT cells in type 1 diabetes protection is only partial because CD1d−/− NOD mice still show some sensitivity to OM-85.

An implication of effector cells is possible though unlikely in our model because treating NOD SCID mice with OM-85 after transfer of diabetogenic T-cells does not delay diabetes onset (data not shown). Note, however, that other mechanisms cannot be excluded. Thus, it has been shown that Bacille Calmette-Guerin, CFA (42), or the lymphocytic choriomeningitidis virus (43) may induce protection through IFN-γ–and TNF-α–mediated T-cell apoptosis.

To conclude, these studies bring new insight into the cellular and molecular mechanisms underlying the type 1 diabetes protection afforded by OM-85. Our data incriminate TGF-β and NKT cells, possibly through TLR stimulation. One possibility linking TGF-β and the antidiabetic effect observed in our model could be that the OM-85–induced secretion of TGF-β (as reported here) would in turn induce the production of regulatory FoxP3+ cells from CD4+CD25+ precursors, as recently suggested (44). Further studies are needed to elucidate this point. Other mechanisms are of course not excluded. One may assume that the complexity of the OM-85 molecular composition contributes to the multifactorial nature of its mode of action. It would be interesting to analyze the action of individual components of the extract. On the other hand, one must take into account the fact that OM-85 is used clinically and that the simultaneous presence of several pharmacologically active components represents an interesting correlation to the overall effect of bacterial infections.

FIG. 1.

Protection from diabetes in NOD mice treated with OM-85. A: Diabetes prevention in NOD mice treated intraperitoneally with OM-85 (1 mg/day, 5 days a week for 10 weeks). Results are from one representative experiment of three performed (Table 1). No protection was observed after oral administration of the same dose. B: Protection from diabetes is observed in the long term only when OM-85 treatment is started at an early disease stage (6 weeks of age, P = 0.007). Diabetes onset was only delayed when treatment was started at 10 or 15 weeks of age, but no significant protection was observed (P = 0.451 and P = 0.355, respectively). Experimental groups treated with OM-85 included six mice each; the control group included five mice. C: Results show consistent protection from diabetes in mice treated orally with high-dose OM-85 compared with controls. Experimental groups (OM-85 and controls) included seven mice each.

FIG. 1.

Protection from diabetes in NOD mice treated with OM-85. A: Diabetes prevention in NOD mice treated intraperitoneally with OM-85 (1 mg/day, 5 days a week for 10 weeks). Results are from one representative experiment of three performed (Table 1). No protection was observed after oral administration of the same dose. B: Protection from diabetes is observed in the long term only when OM-85 treatment is started at an early disease stage (6 weeks of age, P = 0.007). Diabetes onset was only delayed when treatment was started at 10 or 15 weeks of age, but no significant protection was observed (P = 0.451 and P = 0.355, respectively). Experimental groups treated with OM-85 included six mice each; the control group included five mice. C: Results show consistent protection from diabetes in mice treated orally with high-dose OM-85 compared with controls. Experimental groups (OM-85 and controls) included seven mice each.

FIG. 2.

Protection from diabetes in NOD mice treated with MDP, IFA, or CFA. No protection is observed after parenteral administration of IFA alone or in association with MDP in mice treated at 4 (A) or 10 (B) weeks of age. At 4 weeks of age, mice treated with CFA are protected from diabetes, compared with controls treated with IFA (P = 0.002) or IFA and MDP (P = 0.008). Protection by CFA is partial when treatment is started at 10 weeks of age (compared with IFA, P = 0.051, or with IFA+MDP, P = 0.220). Experimental groups (NaCl+CFA, MDP+IFA, and NaCl+IFA) included 10 mice each.

FIG. 2.

Protection from diabetes in NOD mice treated with MDP, IFA, or CFA. No protection is observed after parenteral administration of IFA alone or in association with MDP in mice treated at 4 (A) or 10 (B) weeks of age. At 4 weeks of age, mice treated with CFA are protected from diabetes, compared with controls treated with IFA (P = 0.002) or IFA and MDP (P = 0.008). Protection by CFA is partial when treatment is started at 10 weeks of age (compared with IFA, P = 0.051, or with IFA+MDP, P = 0.220). Experimental groups (NaCl+CFA, MDP+IFA, and NaCl+IFA) included 10 mice each.

FIG. 3.

Major role of TGF-β in the protective effect of OM-85. A: Female IL-4−/− and wild-type NOD mice were treated with OM-85 (1 mg i.p./day). A similar diabetes incidence was observed between the two groups (OM-85–treated NOD IL-4−/−: n = 9; and control NOD IL-4−/−: n = 6). B: Treatment with anti–IL-10r antibody (1 mg i.p. three times a week) did not reverse protection induced by OM-85 (1 mg/day i.p.) (OM-85–treated NOD: n = 10; OM-85 + anti–IL-10r–treated NOD: n = 7; and control NOD: n = 11). C: Treatment with a neutralizing anti–TGF-β antibody (1 mg i.p. three times a week) effectively reversed protection induced by OM-85 (1 mg/day i.p.). (OM-85–treated NOD: n = 10; OM-85 + anti–TGF-β–treated NOD: n = 6; and control NOD: n = 11).

FIG. 3.

Major role of TGF-β in the protective effect of OM-85. A: Female IL-4−/− and wild-type NOD mice were treated with OM-85 (1 mg i.p./day). A similar diabetes incidence was observed between the two groups (OM-85–treated NOD IL-4−/−: n = 9; and control NOD IL-4−/−: n = 6). B: Treatment with anti–IL-10r antibody (1 mg i.p. three times a week) did not reverse protection induced by OM-85 (1 mg/day i.p.) (OM-85–treated NOD: n = 10; OM-85 + anti–IL-10r–treated NOD: n = 7; and control NOD: n = 11). C: Treatment with a neutralizing anti–TGF-β antibody (1 mg i.p. three times a week) effectively reversed protection induced by OM-85 (1 mg/day i.p.). (OM-85–treated NOD: n = 10; OM-85 + anti–TGF-β–treated NOD: n = 6; and control NOD: n = 11).

FIG. 4.

Involvement of NKT cells in the diabetes protection afforded by OM-85 treatment. Female CD1d−/− or wild-type NOD mice were treated intraperitoneally with OM-85 (1 mg/day, 5 days a week). Results from two independent experiments are detailed in A (OM-85–treated NOD CD1d−/−: n = 10; control NOD CD1d−/−: n = 10; OM-85–treated NOD: n = 10; and control NOD: n = 11) and B (OM-85–treated NOD CD1d−/−: n = 5; control NOD CD1d−/−: n = 11; OM-85–treated NOD: n = 5; and control NOD: n = 12). A significantly lower diabetes incidence was observed in OM-85–treated CD1d−/− NOD mice compared with untreated CD1d−/− NOD mice (cumulative analysis P = 0.032). A trend toward lower diabetes incidence in OM-85–treated CD1d−/− NOD mice versus OM-85–treated wild-type NOD mice was also observed. Experimental groups included 5–12 mice.

FIG. 4.

Involvement of NKT cells in the diabetes protection afforded by OM-85 treatment. Female CD1d−/− or wild-type NOD mice were treated intraperitoneally with OM-85 (1 mg/day, 5 days a week). Results from two independent experiments are detailed in A (OM-85–treated NOD CD1d−/−: n = 10; control NOD CD1d−/−: n = 10; OM-85–treated NOD: n = 10; and control NOD: n = 11) and B (OM-85–treated NOD CD1d−/−: n = 5; control NOD CD1d−/−: n = 11; OM-85–treated NOD: n = 5; and control NOD: n = 12). A significantly lower diabetes incidence was observed in OM-85–treated CD1d−/− NOD mice compared with untreated CD1d−/− NOD mice (cumulative analysis P = 0.032). A trend toward lower diabetes incidence in OM-85–treated CD1d−/− NOD mice versus OM-85–treated wild-type NOD mice was also observed. Experimental groups included 5–12 mice.

FIG. 5.

OM-85 induces IL-12 production by dendritic cells. The response was completely abolished in dendritic cells obtained from MyD88−/− mice and partially inhibited in dendritic cells from TLR-2−/− and TLR-4−/− mice. Results are representative of one experiment of four performed. WT, wild type.

FIG. 5.

OM-85 induces IL-12 production by dendritic cells. The response was completely abolished in dendritic cells obtained from MyD88−/− mice and partially inhibited in dendritic cells from TLR-2−/− and TLR-4−/− mice. Results are representative of one experiment of four performed. WT, wild type.

FIG. 6.

OM-85 induces IL-10, IFN-γ, and TGF-β production by murine splenocytes. IL-10 and TGF-β production was completely abolished in splenocytes from MyD88−/− mice and partially decreased in splenocytes from TLR-2−/− and TLR-4−/− mice. Results are from one representative experiment of three performed.

FIG. 6.

OM-85 induces IL-10, IFN-γ, and TGF-β production by murine splenocytes. IL-10 and TGF-β production was completely abolished in splenocytes from MyD88−/− mice and partially decreased in splenocytes from TLR-2−/− and TLR-4−/− mice. Results are from one representative experiment of three performed.

TABLE 1

Protection from diabetes in NOD mice treated intraperitoneally (i.p.) with OM-85

n% Diabetes
P
Week 10Week 20Week 30Week 40
Experiment 1      0.0013 
    Control 25 75 100  
    OM-85 i.p.  
Experiment 2      0.0014 
    Control 50 75 75  
    OM-85 i.p. 20 40  
Experiment 3      0.0075 
    Control 11 64 73 73  
    OM-85 i.p. 10 10 20 40  
n% Diabetes
P
Week 10Week 20Week 30Week 40
Experiment 1      0.0013 
    Control 25 75 100  
    OM-85 i.p.  
Experiment 2      0.0014 
    Control 50 75 75  
    OM-85 i.p. 20 40  
Experiment 3      0.0075 
    Control 11 64 73 73  
    OM-85 i.p. 10 10 20 40  

The cumulative analysis of the data from the three independent experiments reported (n = 22 NOD treated mice, n = 19 controls) showed a highly significant decrease in diabetes incidence (P = 0.0008) (see Fig. 1A and research design and methods for dosage and timing schedule).

This work was supported by grants from the Juvenile Diabetes Research Foundation and the Fondation pour la Recherche Médicale and by a contribution from OM Pharma.

The authors thank L. Van Kaer and A. Herbelin for providing CD1d−/− NOD mice at backcross 6 and 12, respectively. The authors acknowledge S. Akira for providing MyD88−/−, TLR-2−/−, and TLR-4−/− mice and D. Mathis for providing IL-4−/− NOD mice. We are indebted to Anne Esling and Berta Segovia for technical help and to Martine Netter for iconography. We also thank J.A. Bluestone and A. O’Garra for providing the monoclonal antibodies used in the study.

1.
Gale EA: The rise of childhood type 1 diabetes in the 20th century.
Diabetes
51
:
3353
–3361,
2002
2.
Bach JF: The effect of infections on susceptibility to autoimmune and allergic diseases.
N Engl J Med
347
:
911
–920,
2002
3.
Feillet H, Bach JF: On the mechanisms of the protective effect of infections on type 1 diabetes.
Clin Dev Immunol
11
:
191
–194,
2004
4.
Harada M, Kishimoto Y, Makino S: Prevention of overt diabetes and insulitis in NOD mice by a single BCG vaccination.
Diabetes Res Clin Pract
8
:
85
–89,
1990
5.
Bras A, Aguas AP: Diabetes-prone NOD mice are resistant to Mycobacterium avium and the infection prevents autoimmune disease.
Immunology
89
:
20
–25,
1996
6.
Oldstone MB: Viruses as therapeutic agents. I. Treatment of nonobese insulin-dependent diabetes mice with virus prevents insulin-dependent diabetes mellitus while maintaining general immune competence.
J Exp Med
171
:
2077
–2089,
1990
7.
Takei I, Asaba Y, Kasatani T, Maruyama T, Watanabe K, Yanagawa T, Saruta T, Ishii T: Suppression of development of diabetes in NOD mice by lactate dehydrogenase virus infection.
J Autoimmun
5
:
665
–673,
1992
8.
Wilberz S, Partke HJ, Dagnaes-Hansen F, Herberg L: Persistent MHV (mouse hepatitis virus) infection reduces the incidence of diabetes mellitus in non-obese diabetic mice.
Diabetologia
34
:
2
–5,
1991
9.
Cooke A, Tonks P, Jones FM, O’Shea H, Hutchings P, Fulford AJ, Dunne DW: Infection with Schistosoma mansoni prevents insulin dependent diabetes mellitus in non-obese diabetic mice.
Parasite Immunol
21
:
169
–176,
1999
10.
Audibert F, Chedid L, Lefrancier P, Choay J, Lederer E: Relationship between chemical structure and adjuvant activity of some synthetic analogues of N-acetyl-muramyl-L-alanyl-D-isoglutamine (MDP).
Ann Immunol (Paris)
128C
:
653
–661,
1977
11.
Chedid L, Parant M, Parant F, Audibert F, Lefrancier F, Choay J, Sela M: Enhancement of certain biological activities of muramyl dipeptide derivatives after conjugation to a multi-poly(DL-alanine)–poly(L-lysine) carrier.
Proc Natl Acad Sci U S A
76
:
6557
–6561,
1979
12.
Chang YH, Pearson CM, Chedid L: Adjuvant polyarthritis. V. Induction by N-acetylmuramyl-L-alanyl-D-isoglutamine, the smallest peptide subunit of bacterial peptidoglycan.
J Exp Med
153
:
1021
–1026,
1981
13.
Shehadeh N, Calcinaro F, Bradley BJ, Bruchlim I, Vardi P, Lafferty KJ: Effect of adjuvant therapy on development of diabetes in mouse and man.
Lancet
343
:
706
–707,
1994
14.
Lucas C, Bald LN, Fendly BM, Mora-Worms M, Figari IS, Patzer EJ, Palladino MA: The autocrine production of transforming growth factor-beta 1 during lymphocyte activation: a study with a monoclonal antibody-based ELISA.
J Immunol
145
:
1415
–1422,
1990
15.
Castro AG, Neighbors M, Hurst SD, Zonin F, Silva RA, Murphy E, Liu YJ, O’Garra A: Anti-interleukin 10 receptor monoclonal antibody is an adjuvant for T helper cell type 1 responses to soluble antigen only in the presence of lipopolysaccharide.
J Exp Med
192
:
1529
–1534,
2000
16.
Bahr GM, Carelli C, Audibert F, Modabber F, Chedid L: Analysis of the antigenic relationship of various derivatives of n-acetyl-muramyl-l-ala-d-isoglutamine (MDP), using anti-MDP antibodies.
Mol Immunol
19
:
737
–745,
1982
17.
Specter S, Cimprich R, Friedman H, Chedid L: Stimulation of an enhanced in vitro immune response by a synthetic adjuvant, muramyl dipeptide.
J Immunol
120
:
487
–491,
1978
18.
Parant MA, Pouillart P, Le Contel C, Parant FJ, Chedid LA, Bahr GM: Selective modulation of lipopolysaccharide-induced death and cytokine production by various muramyl peptides.
Infect Immun
63
:
110
–115,
1995
19.
Bendelac A, Lantz O, Quimby ME, Yewdell JW, Bennink JR, Brutkiewicz RR: CD1 recognition by mouse NK1+ T lymphocytes.
Science
268
:
863
–865,
1995
20.
Bloemendal A, Van der Zee R, Rutten VP, van Kooten PJ, Farine JC, van Eden W: Experimental immunization with anti-rheumatic bacterial extract OM-89 induces T cell responses to heat shock protein (hsp)60 and hsp70; modulation of peripheral immunological tolerance as its possible mode of action in the treatment of rheumatoid arthritis (RA).
Clin Exp Immunol
110
:
72
–78,
1997
21.
Vischer TL: Subreum (OM-8980): an anti-arthritic E. coli extract.
Clin Exp Rheumatol
11 (Suppl. 8)
:
S121
–S123,
1993
22.
Dhaher YY, Khamashta MA, Hartley B, Taub N, Farine JC, Hughes GR: The effect of OM-89 (Subreum) on the murine model of systemic lupus erythematosus MRL-lpr/lpr.
Lupus
6
:
436
–440,
1997
23.
Keyser FD, Mielants H, Veys EM: Current use of biologicals for the treatment of spondyloarthropathies.
Expert Opin Pharmacother
2
:
85
–93,
2001
24.
Kolb H, Worz-Pagenstert U, Kleemann R, Rothe H, Rowsell P, Scott FW: Cytokine gene expression in the BB rat pancreas: natural course and impact of bacterial vaccines.
Diabetologia
39
:
1448
–1454,
1996
25.
Hartmann B, Bellmann K, Ghiea I, Kleemann R, Kolb H: Oral insulin for diabetes prevention in NOD mice: potentiation by enhancing Th2 cytokine expression in the gut through bacterial adjuvant.
Diabetologia
40
:
902
–909,
1997
26.
Schaad UB, Mutterlein R, Goffin H: Immunostimulation with OM-85 in children with recurrent infections of the upper respiratory tract: a double-blind, placebo-controlled multicenter study.
Chest
122
:
2042
–2049,
2002
27.
Del-Rio-Navarro BE, Luis Sienra-Monge JJ, Berber A, Torres-Alcantara S, Avila-Castanon L, Gomez-Barreto D: Use of OM-85 BV in children suffering from recurrent respiratory tract infections and subnormal IgG subclass levels.
Allergol Immunopathol (Madr)
31
:
7
–13,
2003
28.
Bowman LM, Holt PG: Selective enhancement of systemic Th1 immunity in immunologically immature rats with an orally administered bacterial extract.
Infect Immun
69
:
3719
–3727,
2001
29.
Kitani A, Fuss I, Nakamura K, Kumaki F, Usui T, Strober W: Transforming growth factor (TGF)-beta1-producing regulatory T cells induce Smad-mediated interleukin 10 secretion that facilitates coordinated immunoregulatory activity and amelioration of TGF-beta1-mediated fibrosis.
J Exp Med
198
:
1179
–1188,
2003
30.
Nakamura K, Kitani A, Strober W: 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,
2001
31.
Nakamura K, Kitani A, Fuss I, Pedersen A, Harada N, Nawata H, Strober W: TGF-beta 1 plays an important role in the mechanism of CD4+CD25+ regulatory T cell activity in both humans and mice.
J Immunol
172
:
834
–842,
2004
32.
Belghith M, Bluestone JA, Barriot S, Megret J, Bach JF, Chatenoud L: TGF-beta-dependent mechanisms mediate restoration of self-tolerance induced by antibodies to CD3 in overt autoimmune diabetes.
Nat Med
9
:
1202
–1208,
2003
33.
Bach JF, Francois Bach J: Regulatory T cells under scrutiny.
Nat Rev Immunol
3
:
189
–198,
2003
34.
Araujo LM, Lefort J, Nahori MA, Diem S, Zhu R, Dy M, Leite-de-Moraes MC, Bach JF, Vargaftig BB, Herbelin A: Exacerbated Th2-mediated airway inflammation and hyperresponsiveness in autoimmune diabetes-prone NOD mice: a critical role for CD1d-dependent NKT cells.
Eur J Immunol
34
:
327
–335,
2004
35.
Sharif S, Arreaza GA, Zucker P, Mi QS, Sondhi J, Naidenko OV, Kronenberg M, Koezuka Y, Delovitch TL, Gombert JM, Leite-De-Moraes M, Gouarin C, Zhu R, Hameg A, Nakayama T, Taniguchi M, Lepault F, Lehuen A, Bach JF, Herbelin A: Activation of natural killer T cells by alpha-galactosylceramide treatment prevents the onset and recurrence of autoimmune type 1 diabetes.
Nat Med
7
:
1057
–1062,
2001
36.
Lehuen A, Lantz O, Beaudoin L, Laloux V, Carnaud C, Bendelac A, Bach JF, Monteiro RC: Overexpression of natural killer T cells protects Valpha14- Jalpha281 transgenic nonobese diabetic mice against diabetes.
J Exp Med
188
:
1831
–1839,
1998
37.
Shi FD, Flodstrom M, Balasa B, Kim SH, Van Gunst K, Strominger JL, Wilson SB, Sarvetnick N: Germ line deletion of the CD1 locus exacerbates diabetes in the NOD mouse.
Proc Natl Acad Sci U S A
98
:
6777
–6782,
2001
38.
Van Kaer L: Natural killer T cells as targets for immunotherapy of autoimmune diseases.
Immunol Cell Biol
82
:
315
–322,
2004
39.
Oki S, Chiba A, Yamamura T, Miyake S: The clinical implication and molecular mechanism of preferential IL-4 production by modified glycolipid-stimulated NKT cells.
J Clin Invest
113
:
1631
–1640,
2004
40.
Mi QS, Ly D, Zucker P, McGarry M, Delovitch TL: Interleukin-4 but not interleukin-10 protects against spontaneous and recurrent type 1 diabetes by activated CD1d-restricted invariant natural killer T-cells.
Diabetes
53
:
1303
–1310,
2004
41.
Mi QS, Meagher C, Delovitch TL: CD1d-restricted NKT regulatory cells: functional genomic analyses provide new insights into the mechanisms of protection against type 1 diabetes.
Novartis Found Symp
252
:
146
–160 [discussion 160–144, 203–110],
2003
42.
Qin HY, Chaturvedi P, Singh B: In vivo apoptosis of diabetogenic T cells in NOD mice by IFN-gamma/TNF-alpha.
Int Immunol
16
:
1723
–1732,
2004
43.
Christen U, Benke D, Wolfe T, Rodrigo E, Rhode A, Hughes AC, Oldstone MB, von Herrath MG: Cure of prediabetic mice by viral infections involves lymphocyte recruitment along an IP-10 gradient.
J Clin Invest
113
:
74
–84,
2004
44.
Fu S, Zhang N, Yopp AC, Chen D, Mao M, Zhang H, Ding Y, Bromberg JS: TGF-beta induces Foxp3 + T-regulatory cells from CD4 + CD25 - precursors.
Am J Transplant
4
:
1614
–1627,
2004