Lactobacillus gasseri OLL2809 and its RNA suppress proliferation of CD4⁺ T cells through a MyD88-dependent signalling pathway

Abstract

Recent studies have shown that probiotics are beneficial in prevention and improvement of inflammatory diseases. Accumulating evidence indicates that probiotics can modulate immune cell responses, although the specific molecular mechanism by which probiotics work remains elusive. Because T cells express receptors for microbial components, we examined whether the probiotic strain Lactobacillus gasseri OLL2809 (LG2809) and its components regulate murine CD4⁺ T-cell activation. LG2809, as well as two other Lactobacillus strains, inhibited proliferation of CD4⁺ T cells; LG2809 had the strongest suppressive activity among them. RNA isolated from LG2809 was also shown to have suppressive activity. We observed this suppressive effect in the culture of CD4⁺ T cells stimulated with anti-CD3/CD28 treatment, suggesting a direct effect on CD4⁺ T cells. In contrast, the suppressive effect was not observed for CD4⁺ T cells from myeloid differentiation primary response gene 88 (MyD88) protein-deficient mice, and was abrogated in the presence of an anti-oxidant reagent, N-acetyl-cysteine (NAC). These results demonstrate that the suppressive effect of LG2809 and its RNA occurred through a MyD88-dependent signalling pathway and suggest involvement of a reactive oxygen species-dependent mechanism. LG2809 RNA injected subcutaneously suppressed delayed-type-hypersensitivity response in DO11.10 mice, and the suppression was abrogated by treatment with NAC. Collectively, these results suggest that suppression of T-cell proliferation by RNA may be one of the mechanisms when a probiotic bacterial strain exerts suppressive effects on inflammatory responses.

Introduction

It is well known that effector CD4⁺ T cells, such as T helper type 1 (Th1), Th2 and Th17 cells, each of which secretes different cytokines, are involved in the exacerbation of various inflammatory diseases. Specifically, Th1 cells that secrete interferon-γ (IFN-γ) recognize cytosolic pathogens and principally mediate the cell-mediated immune responses.¹ The Th17 cells that secrete interleukin-17 (IL-17) attack extracellular pathogens.² Both Th1 and Th17 cells also trigger the pathogenesis of inflammatory and autoimmune diseases, such as type I diabetes and inflammatory bowel disease.^3,4 The Th2 cells that secrete IL-4 and IL-5 play a key role in the differentiation of B cells into IgE-producing cells, as well as in the recruitment of mast cells and eosinophils. Accordingly, Th2 cells trigger the development of allergic diseases.⁴ Therefore, it is thought that controlling these effecter T cells would make it possible to control the corresponding inflammatory diseases.

As a means of reducing allergic symptoms, recent studies have focused on activating counteracting Th1 cells using specific bacteria to regulate the effector Th2 cell-mediated immune responses.⁵ Other studies have shown that specific bacteria prevent Th1- or Th17-mediated intestinal inflammation by activating regulatory T cells.^6,7 In other words, each of these studies involved using a bacterial strategy for regulating pathogenic T cells by activating other T cells. Other studies have used an alternative strategy to inhibit the proliferation of pathogenic T cells using bacteria. For instance, Kanzato et al.⁸ showed that heat-killed Lactobacillus acidophilus L-92 induced apoptosis of antigen-stimulated murine T cells by modulating the function of dendritic cells in vitro and in vivo. Furthermore, Peluso et al.⁹ showed that live Lactobacillus paracasei subsp. paracasei B21060 directly inhibited human T-cell proliferation by lactic acid produced by the strain.

Pathogen-associated molecular patterns, including bacterial cellular components, are recognized by a variety of pattern recognition receptors, including Toll-like receptors (TLR), Nod-like receptors, RIG-I-like receptors and C-type lectin receptors. The TLRs are most frequently studied in the context of inflammatory responses and in the modulation of the functions of antigen-presenting cells (APCs). It is established that TLRs are also expressed by T cells and that stimulation of TLRs can modulate the function of T cells directly.^10–14 Myeloid differentiation primary response gene 88 protein (MyD88) is a requisite adaptor protein that functions downstream of several TLRs.¹⁵ It is a key factor that determines whether various bacterial components can influence T cells via TLRs.

Lactobacillus gasseri OLL2809 (LG2809) is a probiotic bacterial strain that has been isolated from human faeces.¹⁶ It was previously reported that oral administration of heat-killed LG2809 both effectively reduced serum antigen-specific IgE levels in mice, and reduced the symptoms of Japanese cedar pollinosis.^17,18 Hence, it is thought that heat-killed LG2809 exhibits suppressive effects on allergic responses.

In this study, we examined the effect of LG2809, as well as two other Lactobacillus strains, on CD4⁺ T-cell activation. We demonstrate that heat-killed LG2809 suppressed murine CD4⁺ T-cell proliferation in vitro. RNA was found to be an active component of LG2809. The active component of LG2809 was determined to be its RNA. Furthermore, we found that MyD88-mediated signalling was required for the suppressive effect of LG2809. In addition, our results demonstrate that LG2809 RNA suppressed delayed-type hypersensitivity (DTH) response in vivo.

Materials and methods

Mice

Female BALB/c mice (8–10 week old) were purchased from CLEA Japan (Tokyo, Japan). Female DO11.10 T-cell receptor transgenic mice (8–10 weeks old) were transgenic for ovalbumin (OVA) 323–339-specific and I-A^d-restricted T-cell receptor αβ, with a BALB/c genetic background.¹⁹ MyD88-deficient BALB/c mice were kindly provided by Prof. Shizuo Akira (Osaka University, Osaka, Japan).²⁰ All animal experiments were performed in accordance with the guidelines of the University of Tokyo.

Bacterial strains and preparation

LG2809, L. gasseri MEP225101 (MEP225101) and Lactobacillus casei MEP225102 (MEP225102) were isolated by Meiji Dairies Corporation, Japan. Lactobacillus strains were cultured in Lactobacilli MRS broth (DIFCO, Detroit, MI) at 37° for 18 hr. Bacteria were harvested by centrifugation at 1800 g at 4° for 15 min, and then washed twice with 0·85% NaCl (saline). The cells were collected and suspended in saline to achieve 5 mg dry weight/ml. Where indicated, bacteria were killed by heating at 75° for 60 min.

Preparation of heat-killed bacteria homogenate and its fractions

Heat-killed LG2809 were collected by centrifugation and suspended in saline at 5 mg dry weight/ml. The bacterial suspension was beaten with 0·5 g of glass beads (diameter, 0·1 mm; Sigma-Aldrich, St Louis, MO) by a multi-beads shocker (Yasui Kikai, Osaka, Japan), and to obtain the homogenate. After this, the homogenate was centrifuged at 300 g for 10 min at 4° and the supernatant and precipitate were obtained.

Enzymatic treatment of the bacterial homogenate supernatant

The supernatant of the LG2809 homogenate was treated as follows: (i) incubated at 37° for 4 hr; (ii) incubated at 37° for 4 hr and subsequently at 96° for 10 min; (iii) treated with 100 μg/ml DNAse I (Sigma-Aldrich) at 37° for 4 hr and subsequently at 96° for 10 min to inactivate DNAse I; (iv) treated with 100 μg/ml Proteinase K (Sigma-Aldrich) at 37° for 4 hr and subsequently 96° for 10 min to inactivate Proteinase K; or (v) treated with 100 μg/ml RNAse A (Sigma-Aldrich) at 37° for 4 hr, subsequently treated with 100 μg/ml Proteinase K at 37° for 4 hr to inactivate RNAse A, and further treated at 96° for 10 min to inactivate Proteinase K. Degradation of DNA or RNA from the supernatant of the LG2809 homogenate by DNAse I or RNAse A was confirmed by agarose gel electrophoresis and degradation of the protein by Proteinase K was confirmed by SDS–PAGE.

Bacterial RNA extraction

Bacterial RNA from LG2809 was extracted as previously described.²¹ After the live bacteria (3 mg dry weight) were collected by centrifugation, the pellet was suspended in saline. Subsequently, the pellet was beaten with 0·5 g of glass beads (0·1 mm diameter) using a multi-beads shocker. The extracted RNA was purified with an RNeasy Mini kit (Qiagen, Hilden, Germany) and the resultant RNA was stored at −80° until use.

Mammalian cell culture

Cells from mice were cultured in RPMI-1640 (Nissui Pharmaceutical, Tokyo, Japan) containing 10% fetal calf serum (Gibco, Grand Island, NY), 2 g/l NaHCO₃, 100 U/ml penicillin, 100 μg/ml streptomycin, 50 μm 2-mercaptoethanol and 300 mg/l l-glutamine at 37° in 5% CO₂ in air.

CD4⁺ T-cell proliferation assay

Splenocytes were prepared from BALB/c mice as previously described.²² We isolated T-cell-depleted splenocytes from the splenocytes by magnetic antibody cell sorting (MACS) negative selection using Thy-1.2 microbeads (Miltenyi Biotec, Bergish Gladbach, Germany). The isolated cells used as APCs (1 × 10⁵ cells/well) were cultured with each of the bacterial suspensions (three strains of Lactobacillus), the homogenate of LG2809, RNA from LG2809, or murine splenic RNA (Ambion, Austin, TX) in 96-well flat-bottomed plates at 37° for 4 hr in 5% CO₂ in air. DO11.10 CD4⁺ T cells were isolated from splenocytes by MACS positive selection using CD4 microbeads as described previously (Miltenyi Biotec).²² Isolated CD4⁺ T cells (5 × 10⁴ cells/well) and OVA 323–339 (0·1, 0·3, or 1 μm) were added to the culture of APCs plus the prepared samples. After incubation for 3 days, 37 kBq of [³H]thymidine was added per well, and the plates were cultured for an additional 24 hr. Subsequently, the cells were harvested on a March III harvester (Tomtec, Hamden, CT), and incorporated [³H]thymidine was counted on a Trilux1450 Microbeta counter (Wallac, Gaithersburg, MD) using Microbeta 270.004 software (Wallac).

For activation of T cells without APCs, 96-well flat-bottomed plates were coated with 10 μg/ml anti-CD3ε monoclonal antibody (mAb; BD Biosciences, San Jose, CA) for 2 hr at 37°. CD4⁺ T cells (> 95% purity confirmed by flow cytometric analysis) isolated from splenocytes of BALB/c or MyD88-deficient mice (BALB/c background; 5 × 10⁴ cells/well) were cultured in the wells coated with anti-CD3ε mAb with 2 μg/ml anti-CD28 mAb (BD Biosciences) in the presence of LG2809 or RNAs (LG2809 or mouse spleen RNA) for 72 hr. We measured the proliferation of the cells as described in the above section.

Cytokine analysis by ELISA

CD4⁺ T cells from BALB/c mice were cultured in the wells coated with anti-CD3ε mAb with 2 μg/ml anti-CD28 mAb in the presence of LG2809 RNA. After 48 hr of incubation, culture supernatants were collected. Determination of transforming growth factor-β (TGF-β) and IL-10 levels in culture supernatants was performed by sandwich ELISA using Mouse Opt EIA kits (BD Biosciences) as described previously.²³

Flow cytometric analysis

The flow cytometric analysis was performed using a FACS LSR with CellQuest software (BD Biosciences). For apoptotic analysis, the cells were washed in FACS buffer (1% fetal calf serum, 0·1% NaN₃ in PBS) and incubated with anti-CD16/CD32 mAb (clone 2.4G2; BD PharMingen, San Diego, CA) on ice to block non-specific binding to Fc receptors. After the cells were washed in FACS buffer, the cells were stained with Annexin V-phycoerythrin (PE) and 7-amino-actinomycin D (7-AAD) using Annexin V-PE Apoptosis Detection kit I (BD Biosciences), and then the cells were subjected to the flow cytometric analysis.

For intracellular analysis of the expression of Foxp3, the cells were fixed with Fix/Perm Concentrate (eBioscience, San Diego, CA) and Fix/Perm Diluent (eBioscience) overnight at 4°. The cells were washed with permeabilization buffer (eBioscience) and incubated with anti-CD16/CD32 mAb (BD PharMingen) on ice. Then the cells were stained with anti-Foxp3-PE mAb (clone FJK-16s; BD Biosciences) for 30 min on ice, and the cells were subjected to the flow cytometric analysis.

Reactive oxygen species detection

For detection of reactive oxygen species (ROS), 25 mm 2′,7′-dichlorodihydrofluorescein diacetate (DCF-DA; Invitrogen Carlsbad, CA), an oxidation-sensitive dye, was added to the plates 30 min before harvest. Incubation was terminated by washing with PBS. Generation of ROS was measured as the increase in mean fluorescence intensity by the flow cytometric analysis.

Measurement of DTH responses

Each DO11.10 mouse was immunized at the base of its tail with 100 μg OVA (Wako) in complete Freund's adjuvant. After 14 days, 50 μl saline was injected into the left footpad as a control, and 50 μl saline containing OVA (5 μg) or OVA plus each type of sample was injected into the right footpad. The experimental groups (n = 4 for each group) were as follows: (i) injected with saline containing OVA; (ii) injected with saline containing OVA and LG2809 RNA (0·5 μg); (iii) injected with saline containing OVA, LG2809 RNA and NAC (1·5 μmol); and (iv) injected with saline containing OVA and NAC. We calculated the DTH response using the following formula: the thickness of right footpad (mm) – the thickness of left footpad (mm).

Statistical analysis

All experimental data were expressed as the mean ± standard deviation (SD). Statistical differences were analysed by Tukey's multiple comparison tests or Student's t-tests.

Results

LG2809 strongly suppressed proliferation of splenic CD4⁺ T cells from DO11·10 mice

To study the effect of LG2809 and two other Lactobacillus strains on the activation of CD4⁺ T cells, we examined the proliferative response of antigen-stimulated CD4⁺ T cells from DO11·10 mice in the presence of the heat-killed strains (Fig. 1a). We found that all of the strains we examined suppressed CD4⁺ T cell proliferation, and LG2809 had the strongest suppressive activity among them (Fig. 1a). Furthermore, LG2809 suppressed CD4⁺ T cell proliferation in a dose-dependent manner (Fig. 1b). Thus, for all additional analyses, we focused on examination of LG2809.

Figure 1.

LG2809 suppressed proliferation of DO11.10 murine splenic CD4⁺ T cells activated with antigenic stimulation. T-cell-depleted splenocytes from BALB/c mice used as antigen-presenting cells (1 × 10⁵ cells/well) were cultured in the presence or absence of heat-killed Lactobacillus strains, LG2809, MEP225101 and MEP225102 (10 μg dry weight/ml) (a) or indicated doses of heat-killed LG2809 (b) for 4 hr. Subsequently, splenic CD4⁺ T cells from DO11.10 mice (5 × 10⁴ cells/well) and OVA 323–339 (1 μm) were added to the culture. After 72 hr, the cells were pulsed with [³H]thymidine for 24 hr, and [³H]thymidine incorporation was measured. Data are shown as the mean ± SD. Data are representative of three independent experiments. Statistical differences were analysed by Tukey's tests. Data are statistically different (P < 0.05) among those columns with different symbols.

LG2809 RNA suppressed proliferation of CD4⁺ T cells from DO11·10 mice

To identify the components of LG2809 that were responsible for the anti-proliferative effect on CD4⁺ T cells, we examined different preparations of LG2809 homogenate. We determined that supernatants, but not precipitates, of LG2809 homogenate suppressed CD4⁺ T cell proliferation (Fig. 2a). Next, we examined the effect of various enzymatic treatments of the supernatant fraction of LG2809 homogenate on the suppressive effect. As shown in Fig. 2b, the suppressive effect was not changed significantly by treatment with heat (ii), DNase I (iii), or Proteinase K (iv), compared with the original supernatant fraction (i). In contrast, the suppressive effect was significantly decreased by the treatment with RNase A (v). Furthermore, RNA isolated from LG2809 (the A₂₆₀/A₂₈₀ value was 1·86 on average for LG2809 RNA) suppressed proliferation in a dose-dependent manner (Fig. 2c). On the other hand, RNA isolated from murine splenic cells failed to inhibit the proliferative response (Fig. 2d).

Figure 2.

LG2809 RNA suppressed CD4⁺ T-cell proliferation from DO11.10 mice. Suppressive effects of bacterial samples were examined as described in the legend to Fig. 1. (a) Heat-killed LG2809 (heat-killed), the homogenate of heat-killed LG2809 (homogenate), the supernatant of the homogenate (sup) and the precipitate of the homogenate (ppt) were added at doses corresponding to 0.1, 1, 10 and 100 μg dry weight LG2809/ml. (b) Five different preparations of the LG2809 homogenate supernatant at a dose corresponding to 100 μg dry weight LG2809/ml were added to the culture. The homogenate supernatant was treated with DNAse I (iii), proteinase K (iv) and RNAse A (v). DNAse I and proteinase K were inactivated by heating at 96° for 10 min, and RNAse A was inactivated with proteinase K treatment. The homogenate supernatants incubated at 37° for 4 hr (i) or at 37° for 4 hr followed by heating at 96° for 10 min (ii) were used as controls. (c) LG2809 RNA (at 0.1, 1 and 10 μg/ml) or (d) RNA from murine splenic cells (at 1 and 10 μg/ml) were added to the culture. Data are shown as the mean ± SD. Data are representative of three independent experiments. Statistical differences were analysed by Tukey's tests. Data are statistically different (P < 0.05) among those with different symbols.

LG2809 and its RNA suppressed proliferation of CD4⁺ T cells stimulated with anti-CD3/CD28 treatment in the absence of antigen presenting cells

Because we determined that LG2809 and its RNA suppressed the proliferation of CD4⁺ T cells activated by antigenic stimulation with APCs, we also examined whether LG2809 and its RNA suppress the proliferation of CD4⁺ T cells in the absence of APCs. CD4⁺ T cells from BALB/c mice were stimulated with anti-CD3ε/CD28 mAb in the presence or absence of LG2809 or its RNA. We found that both LG2809 (Fig. 3a) and its RNA (Fig. 3b) suppressed CD4⁺ T cell proliferation in a dose-dependent manner, whereas the RNA isolated from murine splenic cells did not (Fig. 3c). These results suggest that LG2809 and its RNA affect CD4⁺ T cells directly to exhibit suppression.

Figure 3.

LG2809 and its RNA suppressed proliferation of CD4⁺ T cells stimulated with anti-CD3/CD28 treatment in the absence of antigen-presenting cells. Plates were first coated with 10 μg/ml anti-CD3ε monoclonal antibody (mAb) for 2 hr at 37°. Subsequently, splenic CD4⁺ T cells from BALB/c mice (5 × 10⁴ cells/well) and 2 μg/ml anti-CD28 mAb were added with indicated doses of heat-killed LG2809 (a), LG2809 RNA (b) or RNA from murine splenic cells (c) and were cultured for 72 hr. The cells were pulsed with [³H]thymidine for 24 hr, and [³H]thymidine incorporation was measured. Data are shown as the mean ± SD. Data are representative of three independent experiments. Statistical differences were analysed using Tukey's tests. Data are statistically different (P < 0.05) among those with different symbols.

LG2809 and its RNA suppress proliferation of CD4⁺ T cells through a MyD88-dependent signalling pathway

Because most bacterial components are recognized by TLRs that transmit signals through a MyD88-dependent pathway, we examined the involvement of a MyD88-dependent signalling pathway in this study. We determined that LG2809 and its RNA failed to suppress the proliferation of CD4⁺ T cells form MyD88-deficient mice (Fig. 4). This result indicates that the suppressive effect of LG2809 and its RNA was mediated through a MyD88-dependent signalling pathway.

Figure 4.

Suppressive effect of LG2809 on proliferation of CD4⁺ T cells was mediated through a MyD88-dependent signalling pathway. Splenic CD4⁺ T cells (5 × 10⁴ cells/well) from wild-type (a) or MyD88-deficient BALB/c mice (b) were stimulated with plate-bound anti-CD3ε and soluble anti-CD28 monoclonal antibodies, as described in the legend to Fig. 3, in the presence of heat-killed LG2809 (10 μg/ml) or LG2809 RNA (10 μg/ml). Proliferation of the cells was measured as described. Data are shown as the mean ± SD. Data are representative of three independent experiments. Statistical differences were analysed using Tukey's tests. Data are statistically different (P < 0.05) among those with different symbols.

Regulatory T cells are not induced by treatment with LG2809 RNA

We examined whether regulatory T cells are involved in the mechanisms of suppressive effect of LG2809 RNA on CD4⁺ T cell proliferation. To address this issue, secretion of TGF-β and IL-10 and the rate of Foxp3⁺ cells were measured in CD4⁺ T cells stimulated with anti-CD3ε/CD28 mAbs in the presence of LG2809 RNA after 48 hr incubation. Accordingly, no significant difference was observed in TGF-β and IL-10 secretion and the rate of Foxp3⁺ cells in CD4⁺ T cells (data not shown). These results suggest that the regulatory T cells are not involved in the suppressive effect of LG2809 RNA on CD4⁺ T cell proliferation.

Apoptosis of CD4⁺ T cells is not induced by treatment with LG2809 RNA

To examine whether induction of apoptosis is involved in the suppressive effect of LG2809 RNA, the rate of apoptotic cells were measured in CD4⁺ T cells stimulated with anti-CD3ε/CD28 mAbs in the presence of LG2809 RNA for 48 hr. There was no difference in the rate of Annexin V⁺7-AAD⁺ cells and Annexin V⁻ 7-AAD⁻ cells in the CD4⁺ T cells stimulated in the presence and absence of LG2809 RNA (data not shown), which suggests that induction of apoptosis or mortality is not involved in the suppressive effect of LG2809 RNA on CD4⁺ T cell proliferation.

Reactive oxygen species are generated in T cells treated with LG2809 and its RNA and their suppressive effect is abrogated by anti-oxidant treatment

To assess the role of ROS in the suppressive effect of LG2809 and its RNA, we performed experiments that included treatment with NAC as an anti-oxidant. These results showed that NAC abrogated the suppressive effect of both LG2809 and its RNA, whereas NAC had no effect on the control (Fig. 5a). We next examined the ROS generation in the CD4⁺ T cells by treatment with LG2809 and its RNA using a ROS-sensitive fluorescence indicator DCF-DA. As expected, LG2809 and its RNA induced ROS in the T cells after 1 hr of incubation (Fig. 5b,c) and NAC treatment abrogated the ROS generation (Fig. 5c). These results suggest that the suppressive effect of LG2809 and its RNA on the proliferative response requires an oxidative signal.

Figure 5.

Reactive oxygen species are generated in CD4⁺ T cells treated with LG2809 and its RNA and their suppressive effect is abrogated by antioxidant treatment. (a) Splenic CD4⁺ T cells from BALB/c mice (1 × 10⁵ cells/well) were stimulated with plate-bound anti-CD3ε and soluble anti-CD28 monoclonal antibodies, as described in the legend to Fig. 3, in the presence of heat-killed LG2809 (10 μg/ml) or LG2809 RNA (10 μg/ml) with or without N-acetyl-cysteine (NAC) at 25 mm. Proliferation of the cells was measured as described. Data are shown as the mean ± SD and are representative of three independent experiments. (b, c) Splenic CD4⁺ T cells from BALB/c mice (1 × 10⁵ cells/well) were cultured in the presence of heat-killed LG2809 (10 μg/ml) or LG2809 RNA (10 μg/ml) for 1 hr with or without NAC at 25 mm. After the incubation, all groups of the cells were stained with 25 mm 2′,7′-dichlorodihydrofluorescein diacetate (DCF-DA) for 30 min and analysed by flow cytometry. The open histograms indicate the cells incubated with LG2809 or LG2809 RNA and the shaded histograms indicate the cells cultured with the medium alone. (c) Data are shown as the mean ± SD of mean fluorescence intensities of triplicate culture and are representative of three independent experiments. Statistical differences were analysed by Student's t-tests. NS = not significant, *P < 0.05, and **P < 0.01.

LG2809 RNA suppresses the DTH response

Finally, we examined the suppressive effect of LG2809 RNA on antigen-activated CD4⁺ T-cell proliferation in vivo using the DTH response (Fig. 6). We found that saline/OVA-injection of mice induced swelling of the footpad, whereas LG2809 RNA administered together with saline/OVA significantly suppressed the DTH response. Moreover, NAC treatment almost fully abrogated the suppression of the DTH response by LG2809 RNA. This observation indicates that LG2809 RNA may reduce the DTH response in vivo via a ROS oxidative signal.

Figure 6.

LG2809 RNA suppressed the delayed-type hypersensitivity response. DO11.10 mice were immunized with ovalbumin (OVA) and complete Freund's adjuvant (CFA) at the base of the tail. After 14 days, saline was injected into the left footpad, and saline containing OVA with or without the sample was injected into the right footpad. After 24 hr, the footpad swelling was measured and calculated using the following formula: the thickness of right footpad (mm) – the thickness of left footpad (mm). Data are shown as the mean ± SD. Data are representative of three independent experiments. Statistical differences were analysed using a Tukey's test. Data are statistically different (P < 0.05) among those with different symbols.

Discussion

In the present study, we found that LG2809 and its RNA had suppressive effects on murine CD4⁺ T-cell proliferation. These effects were apparently caused through a MyD88-dependent signalling pathway, and were abrogated by treatment with an anti-oxidant. We also showed that LG2809 RNA suppressed the DTH response in vivo. Our results suggest that suppression of T-cell proliferation by RNA may be one of the mechanisms when a probiotic bacterial strain exerts suppressive effects on inflammatory responses.

LG2809 had the strongest suppressive activity among the three Lactobacillus strains we examined in this study. It was reported that L. acidophilus strain L-92 inhibited antigen-stimulated CD4⁺ T-cell proliferation by inducing apoptosis in an APC co-culture system.⁸ Kanzato et al. suggested that L-92 may inhibit the T-cell response by modulating the expression of costimulatory molecules on dendritic cells. In the present study, LG2809 inhibited not only CD4⁺ T-cell proliferation induced by antigen stimulation with APCs, but also proliferation induced by anti-CD3ε/CD28 mAb stimulation. We also found that LG2809 and its RNA did not induce apoptosis. Therefore, LG2809 probably inhibits proliferation by a mechanism that does not involve dendritic cells and does not induce apoptosis, which is different from the case with L-92. Peluso et al.⁹ showed that live L. paracasei subsp. paracasei B21060 directly inhibited human CD4⁺ T-cell proliferation. This report showed that lactic acid from live bacteria suppressed CD4⁺ T-cell proliferation. However, in the present study we used heat-killed LG2809 so lactic acid was not produced in our assay system. Accordingly, LG2809 inhibited proliferation by a different mechanism.

The inhibitory effect of heat-killed bacteria on T-cell proliferation has also been reported in Salmonella.²⁴ Velden et al. showed that the inhibitory effect did not require virulence genes; however, this study did not identify the bacterial component responsible for the suppressive effect and the inhibitory mechanism.

In the present study, we revealed that LG2809 RNA was responsible for the suppressive effect on CD4⁺ T-cell proliferation. To our knowledge, this is the first report that has identified bacterial RNA as the active component for the T-cell suppressive effect. To elucidate whether the inhibitory activity of RNA was unique to that of the Lactobacillus species (or the LG2809 strain), we also examined the inhibitory effect of RNA derived from mammalian cells and Escherichia coli on CD4⁺ T-cell proliferation. Interestingly, E. coli Total RNA (Ambion) exhibited a suppressive effect on CD4⁺ T-cell proliferation, whereas RNA from mouse spleen Total RNA (Clontech, Palo Alto, CA) did not. To exclude the possibility that the difference of the preparation methods led to the different activities, we confirmed that RNA extracted from splenocytes of BALB/c mice by the same method used for bacterial RNA did not suppress CD4⁺ T-cell proliferation (data not shown). These data suggest that bacterial RNA in general may have the inhibitory effect on CD4⁺ T-cell proliferation. When a probiotic bacterial strain, which contains different kinds of immune-modulating components, causes the action toward suppression of inflammatory responses, RNA from the strain would work as one of the suppressing reagents for the T-cell responses.

There are two types of receptors for bacterial RNA in mammalian cells: membrane-bound receptors, including TLR3 and TLR7, and the cytosolic receptor NALP3.^25,26 The TLR3 recognizes double-stranded RNA, whereas TLR7 and NALP3 recognize single-stranded RNA.^25,26 It has been shown that only TLR7 of these three uses MyD88 to activate the distinct signalling pathway.²⁷ In this study, LG2809 and its RNA failed to inhibit the proliferative response of CD4⁺ T cells from MyD88-deficient mice. Expression of TLR7 has been observed on mouse CD4⁺ T cells,²⁸ and we found that a TLR7 ligand, Imiquimod, but not the TLR3 ligand poly I:C had a strong suppressive effect on CD4⁺ T-cell proliferation and the DTH response (Yoshida et al., unpublished data). In addition, our data indicate that the suppressive activity of supernatant of LG2809 homogenate was lost by treatment with RNAse A that degrades single-strand RNA but not double-strand RNA. Collectively, these findings suggest that LG2809 RNA containing single-stranded RNA as the major active component may exert its suppressive effect through TLR7.

Caron et al.²⁹ found that Resiquimod, a TLR7 and TLR8 ligand, induced IFN-γ production and proliferation of human CD4⁺ T cells synergistically with anti-CD3ε mAb or anti-CD2 mAb stimulation. On the other hand, our preliminary results showed that Imiquimod, another TLR7 ligand³⁰, suppressed CD4⁺ T cell proliferation of murine CD4⁺ T cells stimulated with anti-CD3ε/CD28 mAb. Although these two results seem to be conflicting, we think that this difference would be caused by the differences in the experimental settings, such as using different TLR ligands, Resiquimod and Imiquimod, and using human or murine T cells. It is known that TLR8 works in humans but not in mice^31,32 and that TLR expression on CD4⁺ T cells is different between humans and mice.³⁰

The anti-oxidant NAC is commonly used for its ability to minimize intracellular oxidative stress. In this study, NAC abrogated the suppressive effect of both LG2809 and its RNA, which suggests that ROS are involved in this effect. We also detected ROS generation in CD4⁺ T cells stimulated with LG2809 and its RNA. There are some reports on the role of ROS in T cells. Activation-induced cell death is known as a phenomenon where activated T cells undergo apoptosis, and plays an important role in maintaining T-cell homeostasis.³³ Previous studies have suggested that T-cell receptor signalling may involve ROS generation and that inhibition of oxidative signals interferes with induction of CD95 expression and activation-induced cell death.³⁴ Moreover, mitochondrial ROS control T-cell activation by regulating IL-2 and IL-4 expression.³⁵ Our study raises the possibility that ROS is engaged in the suppression of CD4⁺ T-cell proliferation by LG2809 and its RNA. However, further studies are necessary to elucidate precisely how ROS is involved in the suppressive function of LG2809 and its RNA on CD4⁺ T-cell proliferation.

The DTH reaction is the typical in vivo manifestation of cell-mediated immunity, and the reaction is characterized by activation and recruitment predominantly of T cells. It is well known that Th1 cells mediate the DTH reaction. LG2809 RNA would inhibit Th1 proliferation to a level similar to that by which it inhibited splenic CD4⁺ T-cell proliferation in vitro, and suppress the DTH response. Because we also observed the suppressive effect in vivo in this study, LG2809 RNA may be effective as an immunosuppressive agent for treating patients with immune diseases. In addition, we found that the suppressive effect of LG2809 RNA was lost by NAC treatment, which suggests that ROS may affect the suppressive effect of LG2809 RNA on T-cell-mediated immune responses in vivo.

Although the mechanism by which lactobacilli and their components gain access to immune cells has not been fully elucidated, previous reports indicated that lactobacillus cells administered perorally can reach Peyer's patches³⁶ and a mesenteric lymph node.³⁷ The bacteria were found in the intracellular spaces of lymphocytes and macrophages in Peyer's patches.³⁶ Cukrowska et al. reported that lactobacillus cells administered intragastrically did not translocate through the intestinal barrier into blood, liver and spleen, but were found alive in a mesenteric lymph node, although the localization of the bacteria in a mesenteric lymph node was not shown. These results suggest the possibility that lactobacilli and their components can gain access to T cells at least in gut-associated lymphoid tissues.

In conclusion, LG2809 and its RNA inhibited CD4⁺ T-cell proliferation through a MyD88-dependent signalling pathway, and this effect was decreased by anti-oxidant treatment. Our results suggest that suppression of CD4⁺ T-cell proliferation by RNA may be one of the mechanisms when a probiotic bacterial strain exerts suppressive effects on inflammatory responses. LG2809 and its RNA may be beneficial for treating patients with diseases resulting from a hyper-response of CD4⁺ T cells.

Glossary

Abbreviations

7-AAD

7-amino-actinomycinD

APCs

antigen-presenting cells

DCF-DA

2′, 7′-dichlorodihydrofluorescein diacetate

DTH

delayed-type hypersensitivity

IFN

interferon

IL

interleukin

mAb

monoclonal antibody

MACS

magnetic antibody cell sorting

MyD88

myeloid differentiation primary response gene 88 protein

OVA

ovalbumin

PE

phycoerythrin

ROS

reactive oxygen species

TGF-β

transforming growth factor-β

Th1

T helper type 1

TLR

Toll-like receptor