International Immunopharmacology 25 (2015) 321–331

Contents lists available at ScienceDirect

International Immunopharmacology journal homepage: www.elsevier.com/locate/intimp

Scavenger receptor for lipoteichoic acid is involved in the potent ability of Lactobacillus plantarum strain L-137 to stimulate production of interleukin-12p40 Shinya Hatano a, Yoshitaka Hirose b, Yoshihiro Yamamoto b, Shinji Murosaki b, Yasunobu Yoshikai a,⁎ a b

Division of Host Defense, Medical Institute of Bioregulation, Kyushu University, Fukuoka 812-8582, Japan Research and Development Institute, House Wellness Foods Corporation, 3-20 Imoji, Itami, Hyogo 664-0011, Japan

a r t i c l e

i n f o

Article history: Received 12 November 2014 Received in revised form 7 February 2015 Accepted 9 February 2015 Available online 16 February 2015 Keywords: Bacteria Interleukin-12 Phagocytosis Scavenger receptor Immunotherapy

a b s t r a c t Heat-killed Lactobacillus plantarum strain L-137 (HK L-137) is a more potent inducer of interleukin (IL)-12 than other heat-killed Lactobacillus strains. To elucidate the mechanism involved in this IL-12p40 induction, we compared HK L-137 with heat-killed L. plantarum strain JCM1149 (HK JCM1149) by nuclear magnetic resonance and mass spectrometry. Results showed that HK L-137 contained lipoteichoic acid (LTA) with a chemical structure similar to that of JCM1149, except for a lower degree of glucosyl substitution in the poly(glycerol phosphate) backbone. Lysozyme sensitivity and electrophoretic moiety analysis revealed that HK L-137 exposed more LTA on its cell surface than HK JCM1149. Phagocytosis of HK L-137 by splenic adherent cells was significantly greater than that of HK JCM1149. Anti-LTA antibody and anti-scavenger receptor-A (SR-A) antibody selectively inhibited phagocytosis of HK L-137, as well as IL-12p40 production, by splenic adherent cells. Thus, a higher efficiency of phagocytosis of HK L-137 via SR-A for LTA is responsible for the potent IL-12p40 induction. © 2015 Elsevier B.V. All rights reserved.

1. Introduction Many lactic acid bacteria are used as probiotics because of their health-promoting effects, including enhancement of the immune system. Lactobacillus plantarum L-137, a strain isolated from a fermented fish and rice dish from the Philippines called “burong isda” [1] is a potent inducer of interleukin (IL)-12 in vitro as well as in mice in vivo [2]. Administration of heat-killed (HK) L-137 suppressed not only IgE production against a natural antigen in a mouse model of food allergy, but also inhibited tumor growth in mice transplanted with syngeneic tumor cells [2,3]. Furthermore, it was demonstrated that, in healthy subjects, daily intake of HK L-137 enhanced acquired immunity, particularly Th1-related immune functions [4]. Lipoteichoic acid (LTA) of lactobacilli [5] is potent inducers of proinflammatory cytokines such as tumor necrosis factor (TNF)-α and IL-6. Lactobacillus acidophilus deficient in LTA [6] exhibited an impaired ability to induce production of proinflammatory cytokines such as TNF-α and IL-12 by human peripheral blood mononuclear cells and murine bone marrow-derived dendritic cells. The amount of fatty acid substitution and the content of unsaturated fatty acid in the glycolipid anchor moiety of LTA reportedly influence the cytokine-inducing capacity of LTA [7,8]. The level of D-Ala substitution in the poly(glycerol phosphate) (GroP) backbone was also reported to affect cytokine induction by LTA ⁎ Corresponding author at: Division of Host Defense, Medical Institute of Bioregulation, Kyushu University, 3-1-1 Maidashi, Higashi-ku, Fukuoka 812-8582, Japan. Tel.: +81 92 642 6962; fax: +81 92 642 6973. E-mail address: [email protected] (Y. Yoshikai).

http://dx.doi.org/10.1016/j.intimp.2015.02.011 1567-5769/© 2015 Elsevier B.V. All rights reserved.

[8]. The ability of L. plantarum, which contains much less D-Ala in its LTA than other lactobacilli [9], to induce proinflammatory cytokines is also impaired. Thus, structural differences in LTA in the cell walls of lactobacilli may affect their immunomodulatory properties. Several receptors related to pattern recognition in innate immunity have been identified as receptors for LTA. Toll-like receptor (TLR)-2 recognizes various fungal and bacterial cell wall components, including the LTA of Gram-positive bacteria, and activates a myeloid differentiation factor 88 (MyD88) nuclear factor (NF)-κB pathway, leading to cytokine induction [10–12]. CD36 (aka scavenger receptor B3) are co-receptors of TLR2 for LTA in proinflammatory cytokine responses [13]. CD14 is primarily known as a co-receptor of TLR4, and is part of the lipopolysaccharide (LPS) receptor complex [14]. CD14 also assists in LTA signaling, acting as a coreceptor for TLR2 [13]. CD36 acts as a sensor for LTA and diacylated lipopeptides [15], and also acts as a co-receptor for TLR2 in response to microbial diacylglycerides. CD36 also mediates the phagocytosis of various Gram-positive and Gram-negative bacteria, including Escherichia coli, Klebsiella pneumoniae, Salmonella typhimurium, Staphylococcus aureus, and Enterococcus faecalis [16]. Scavenger receptor A (SR-A aka CD204) binds to a broad range of polyanionic ligands, including modified lipoproteins, the LPS of Gram-negative bacteria, the LTA of Gram-positive bacteria, and bacterial CpG DNA and double-stranded RNA [17,18]. SR-A can also recognize bacterial surface LTA and lipoproteins, and mediate non-opsonic phagocytosis of Gram-positive bacteria such as Streptococcus pyogenes and S. aureus [17,19,20]. We previously reported that HK L-137 exhibited a stronger ability to produce IL-12 than HK L. plantarum JCM1149T (HK JCM1149) [21]. The

322

S. Hatano et al. / International Immunopharmacology 25 (2015) 321–331 Table 1 1 H and 13C chemical shift data of LTAs from L-137 and JCM1149. Strain

Residue

Chemical shifts of proton and carbon in D2O (δ ppm) 1

2

3

4

5

6

3.87/3.92 66.3

4.02 69.4

3.87/3.92 66.3

– –

– –

– –

4.08 63.7

5.36 74.1

4.08 63.7

– –

– –

– –

3.87/3.92 66.3

4.09 75.3

3.87/3.92 66.3

– –

– –

– –

5.15 97.4

3.52 71.5

3.74 72.9

3.40 69.7

3.89 71.9

3.75/3.86 60.5

– 170.0

4.26 48.9

1.61 15.4

– –

– –

– –

3.87/3.93 66.2

4.02 69.5

3.87/3.93 66.2

– –

– –

– –

4.08 63.7

5.37 74.2

4.08 63.7

– –

– –

– –

3.87/3.93 66.2

4.09 75.2

3.87/3.93 66.2

– –

– –

– –

5.15 97.5

3.51 71.5

3.74 73.0

3.39 69.6

3.91 71.9

3.75/3.86 60.5

− 170.0

4.27 48.9

1.61 15.4

− −

− −

− −

L-137 GroP-H 1 H 13 C GroP-d-Ala 1 H 13 C GroP-Glc 1 H 13 C Glc 1 H 13 C d-Ala 1 H 13 C JCM1149 Fig. 1. Western blot analysis of LTA. LTA from L. plantarum strains was analyzed by immunoblotting with a mouse anti-LTA antibody (clone 55). Lane 1, LTA obtained from L-137; lane 2, LTA obtained from JCM1149; lane 3, S. aureus LTA positive control.

objective of this study was to investigate the mechanisms by which HK L-137 induces potent IL-12p40 production, assessed by comparison of the chemical structure of LTA and recognition receptors between HK L-137 and HK JCM1149. 2. Materials and methods 2.1. Preparation of heat-killed L. plantarum cells

GroP-H 1 H 13 C GroP-d-Ala 1 H 13 C GroP-Glc 1 H 13 C Glc 1 H 13 C d-Ala 1 H 13 C

HK L-137 and HK JCM1149 were prepared according to previously described methods [22]. Briefly, L. plantarum L-137 and JCM1149 were

Fig. 2. NMR spectra and specific features of LTA and the glycolipid anchors obtained following HF hydrolysis. 1H NMR spectra (500 MHz, 297 K) of LTA purified from L. plantarum L-137 (A) and JCM1149 (C) suspended in D2O, and those of glycolipid anchors from L-137 (B) and JCM1149 (D) suspended in CD3OD/CDCI3 (1:1 v/v).

S. Hatano et al. / International Immunopharmacology 25 (2015) 321–331 Table 2 1 H and 13C chemical shift data of the lipid anchor of LTA from L-137 and JCM1149. Residue

Chemical shifts of proton and carbon in CD3OD/CD3Cl (δ ppm) 1

2

3

4–6

αHex 1 H 13 C

4.91 98.4

3.74 70.0

n.d n.d

n.d n.d

αHex′ 1 H 13 C

4.93 97.0

3.53 72.0

3.74 72.2

n.d n.d

βHex 1 H 13 C

4.32 103.3

3.22 73.7

3.25 76.3

n.d n.d

Gro 1 H 13 C

4.16/4.40 62.6

5.19 70.0

3.62/3.79 65.9

–

n.d = not determined because of signal overlapping.

incubated in modified maltose, yeast, and peptone-based medium (modified MYP medium) at 32 °C for 18 h, harvested by centrifugation at 4200 ×g at 4 °C for 10 min, washed three times with saline, and resuspended in distilled water. The bacteria were then heated at 80 °C for 20 min, and resultant heat-killed cells were harvested by centrifugation at 4200 ×g for 10 min at 4 °C. Cells were resuspended in distilled water and then lyophilized. The preparations were resuspended in saline and sonicated in an ultrasonic bath (AU-80C, Aiwa, Tokyo, Japan) to disperse aggregated bacterial cells prior to use in experiments.

323

the aqueous phase was evaporated and lyophilized. The extracts, which were dissolved in equilibration buffer (0.1 M ammonium acetate buffer containing 15% n-propanol, pH 4.7), were filtered through a 0.22-μm syringe filter (Millipore, Billerica, MA, USA) and then loaded onto a 1.6 × 15 cm octyl Sepharose CL-4B column (GE Healthcare, Uppsala, Sweden). LTA was eluted using a linear 15–60% 1-propanol gradient in 0.1 M ammonium acetate buffer (pH 4.7). The fractions containing LTA were pooled following phosphate determination (described in 2.2.), and the pooled sample was lyophilized after ultrafiltration with a Macrosep 1Κ Omega centrifugal device (Pall Filtron, Saint Germain en Laye, France) and desalination on a PD10 desalting column (GE Healthcare). The purity of the LTA was determined by measuring endotoxin contents using the Limulus amebocyte lysate assay (b 0.02% w/w) (Seikagaku, Tokyo, Japan), and nucleic acid and protein contamination was checked by measuring UV absorption at 260 and 280 nm using HIC. 2.3. Phosphate determination Phosphate was determined as described by Gründling and Schneewind [23]. Briefly, 140 μL of each HIC fraction was dried for 3 h at 98 °C and then hydrolyzed at 160 °C in 400 μL of acid solution (139 mL of concentrated H2SO4 and 37.5 mL of 70% (v/v) HClO4 per liter), and subsequently cooled to room temperature. Samples were further cooled on ice, then distilled water was added to achieve a final volume of 400 μL. Twenty microliters of each sample were then transferred to a 96-well assay plate, and 100 μL of a freshly prepared reduction solution (3.75 g of ammonium molybdate, 20.4 g of sodium acetate, and 10 g of ascorbic acid per liter) were added to each well. Following 2 h incubation at 37 °C, absorbance at OD710 was determined.

2.2. Preparation of LTA 2.4. Preparation of LTA glycolipid anchor n-Butanol extraction and subsequent purification of LTA by hydrophobic interaction chromatography (HIC) was performed as described by Gründling and Schneewind [23] with some modifications. Briefly, bacteria were cultured in modified MYP medium for 18 h at 32 °C. The cells were harvested, suspended in 0.1 M sodium citrate buffer (pH 4.7), and lysed in a bead-beater (BioSpec Products, Bartlesville, OK, USA) by shearing six times for 30 s on ice with 0.1 mm glass beads. The suspensions were then mixed with an equal volume of n-butanol and stirred for 2 h at room temperature. Following centrifugation at 13,000 ×g for 20 min,

The glycolipid anchor of LTA was purified as described by Gründling and Schneewind [24], with some modifications. Purified LTA (~10 mg) was dissolved in 0.5 mL of 47% hydrofluoric acid (HF) for 48 h at 4 °C, before 7 mL of saturated Na2CO3 solution was added to neutralize the reaction mixtures. Glycolipids were extracted for 1 h at room temperature by adding chloroform-methanol to obtain a final chloroform/ methanol/aqueous phase ratio of 1:1:0.9 (v/v). Samples were centrifuged for 10 min at 2600 ×g, and the chloroform phase containing the

Fig. 3. ESI MS spectra of the glycolipid anchors of L. plantarum LTA. MS spectra of glycolipids isolated from L-137 LTA (A) and JCM1149 LTA (B) following hydrofluoric acid hydrolysis.

324

S. Hatano et al. / International Immunopharmacology 25 (2015) 321–331

Table 3 The characteristics of poly (glycerol phosphate) chain in intact LTAs from L-137 and JCM1149.

Poly (GroP) chain length d-Ala substitution (%) Glucose substitution (%)

LTA L-137

LTA JCM1149

96 50 2

110 42 10

LTA glycolipid anchor was transferred to a new tube. The glycolipid solution was evaporated and dried under a stream of nitrogen. 2.5. Western blot analysis Sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDSPAGE) and western blot analyses of LTA were performed as described previously [21]. Briefly, LTA samples were dissolved in SDS-PAGE sample buffer at a final concentration of 25 ng/μL, then 10 μL of each sample were subjected to 15% polyacrylamide gel electrophoresis, followed by electrotransfer to polyvinylidene fluoride membranes. LTA derived from S. aureus (pLTA-SA; InvivoGen, San Diego, CA, USA) was used as the positive control. Pre-stained protein marker (broad range) (Nacalai-Tesque, Kyoto, Japan) was used as a rough standard to assess molecular weights. Mouse anti-LTA primary antibody (clone 55; HyCult Biotechnology, Uden, The Netherlands) and horseradish peroxidase-linked goat anti-mouse IgG antibody (Anaspec, San Jose, CA, USA) were used at 1:2500 and 1:5000 dilutions, respectively. Immunoreactive LTA was detected by chemiluminescence with Ez-Capture MG (ATTO, Tokyo, Japan). 2.6. Nuclear magnetic resonance (NMR) spectroscopy NMR experiments were performed on a JEOL ECA 500 instrument (Tokyo, Japan). 1H NMR and 13C NMR spectra of LTA (~ 10 mg of LTA was dissolved in 0.5 mL of D2O) or glycolipid anchor (the glycolipid anchor obtained from ~10 mg of LTA was dissolved in 0.5 mL of CD3OD/ CDCI3 (1:1 v/v)) samples were obtained at 500 and 125 MHz, respectively, at 297 Κ. Assignments were taken from 1H-1H correlation spectroscopy (COSY), totally correlated spectroscopy (TOCSY) with a mixing time of 50 ms, rotating-frame nuclear Overhauser effect correlation spectroscopy (ROESY) with a mixing time of 0.5 s, hetero-nuclear single quantum coherence (HSQC), and hetero-nuclear multiple-bond connectivity (HMBC) analyses. Data acquisition and processing were performed using Delta v4.3 (JEOL). The average number of repeating units in the poly(GroP) backbone, the percentage of substitutions, and the chain length of the fatty acids in the membrane anchor

were calculated from the integrals of the 1H NMR spectra of LTA and glycolipid anchor. 2.7. Mass spectrometry The glycolipid anchor of LTA was analyzed by electrospray ionization mass spectrometry (MS) and subsequent MSn analysis on a LCQ Deca instrument (Thermo Quest, Austin, TX, USA). Mass spectra were recorded in the positive and negative modes under standard instrumental parameters. The glycolipid anchor obtained from ~10 mg of LTA was dissolved in 0.5 mL of chloroform/methanol (1:1 v/v; sample solution). Ten microliters of the sample solution were dissolved in 500 μL of methanol and introduced via a syringe pump at a flow rate of 3 μL/min. The ion gauge pressure was 1.0 × 10−5 Torr. Nitrogen was used as the sheath gas. For MSn experiments, the mass window for the parent ion selection was 3.0 Th. Helium was used as the collision gas, and the normalized collision energy was adjusted to 25–40%. The electrospray voltage was set at 5.0 kV for the positive ion mode and − 5.0 kV for the negative ion mode. The capillary temperature was set at 250 °C. 2.8. Determination of lysozyme sensitivity Heat-killed lactobacilli suspended at 1 mg/mL in 50 mM Tris–HCl buffer (pH 7.5) were treated with 500 μg/mL of lysozyme (Sigma, St. Louis, MO, USA) at 37 °C for 120 min. Following incubation, 10% SDS solution was added to the reaction mixtures at a final concentration of 2%, and the mixture was stirred vigorously to dissolve protoplasts resulting from digestion of the cell walls. Optical density at 600 nm (OD600) was immediately measured. The sensitivity of bacteria to lysozyme was calculated by the formula: sensitivity (%) = ((OD600 untreated − OD600 treated) / OD600 untreated × 100) [25]. 2.9. Determination of ζ potential The electrophoretic mobility of heat-killed lactobacilli was measured using a Laser Zeta Potential Analyzer (ELS-6000; Otsuka Electronics, Osaka, Japan). Bacterial cells were suspended in 10 mM potassium citrate buffer (pH 4.0), 10 mM potassium phosphate buffer (pH 7.0), or 10 mM potassium borate buffer (pH 9.0) and injected into a quartz cell at 25 °C. Electrophoretic mobilities were converted to ζ potential using the Helmholtz–Smoluchowski equation [26]. 2.10. Mice Specific pathogen-free female BALB/c or C57BL/6 mice were purchased from Charles River Japan Inc. (Hino, Japan). Fc receptor γ-chain

Fig. 4. Differences in lysozyme sensitivity of HK L-137 and HK JCM1149. HK L-137 and JCM1149 were treated with lysozyme for 2 h. Each cell suspension was treated with SDS to dissolve protoplasts resulting from digestion of the cell walls, and OD600 was measured every 30 min (A). Lysozyme sensitivity was calculated from the OD600 data at 2 h (B). Data obtained from three individual suspensions are expressed as means ± standard deviations. Asterisks represent a significant difference between groups (**, P b 0.01).

S. Hatano et al. / International Immunopharmacology 25 (2015) 321–331 Table 4 ζ potential of HK L-137 and HK JCM1149 at pH 4–9. Strain

L-137 JCM1149

ζ potential (mV) pH 4

pH 7

pH 9

−22 ± 0.6 −13 ± 1.5

−20 ± 0.6 −16 ± 1.2

−27 ± 1.2 −21 ± 1.5

(FcRγ) knockout (KO) mice were purchased from Taconic (Germantown, NY). Mice were maintained in our facility and fed a commercial diet (CE-2; CLEA Japan, Tokyo, Japan). All experiments were performed with 6–12-week-old mice in accordance with the guidelines of the Animal Care and Use Committee of the House Wellness Foods Corporation. 2.11. Cytokine assay Spleen cells were suspended at 2.5 × 106 cells/mL in RPMI 1640 medium containing 10% fetal bovine serum (FBS), 100 μg/mL of streptomycin, 100 U/mL of penicillin and 5 x 10−5 M of 2-mercaptoethanol, and cultured for 24 or 48 h with heat-killed bacteria, LTA from L. plantarum strains, Pam3CSK4 (InvivoGen) and pLTA-SA (InvivoGen) as TLR2-positive ligands, or ultrapure LPS from E. coli 0111:B4 (upLPSEB; InvivoGen) as a TLR4 ligand. In some experiments, spleen cells were treated with anti-mouse TLR2 IgG2a (clone C9A12; InvivoGen), anti-mouse SR-A IgG2b (clone 2F8; HyCult Biotechnology), antimouse CD36 IgA (clone JC63.1; Abcam, Cambridge, MA, USA), or antimouse CD14 IgG2a (clone Sa14-2; HyCult Biotechnology) antibodies or with cytochalasin D (Biomol International, Plymouth Meeting, PA, USA) for 1 h at 37 °C before adding heat-killed bacteria. Equal concentrations of mouse IgG2a κ (clone MOPC-173; Abcam), rat IgG2b κ (clone RTK4530; Abcam), mouse IgA κ (clone S107; Abcam), and rat IgG2a κ (clone RTK2758; Abcam) were used, respectively, as isotype controls for TLR2, SR-A, CD36, and CD14 blocking experiments. In other experiments, HK L-137 cells were incubated with mouse anti-

325

LTA monoclonal antibody (mAb, clone 55) or an equal concentration of mouse IgG3 isotype control (clone 6A3; Medical & Biological Laboratories, Sakae, Japan) at 37 °C for 1 h prior to the addition of spleen cells treated with anti-CD16, anti-CD32, and anti-CD64 antibodies, or antimouse SR-A IgG2b (clone 2F8) [21]. At the end of the culture period, each supernatant was harvested to determine cytokine production. Cytokine concentration was determined by sandwich enzyme-linked immunosorbent assay (ELISA). Rat anti-mouse IL-12p40 mAb (clone 15.6; Biolegend, San Diego, CA, USA), anti-mouse TNF-α polyclonal antibody (R&D systems, Minneapolis, MN, USA), anti-mouse IL-6 mAb (clone MP520F3; R&D systems), and anti-mouse IL-10 mAb (clone JES02A5; R&D systems) were used as capture antibodies. Goat antimouse IL-12 polyclonal antibody (R&D systems), goat anti-mouse TNF-α polyclonal antibody (R&D systems), goat anti-mouse IL-6 polyclonal antibody (R&D systems), and goat anti-mouse IL-10 polyclonal antibody (R&D systems) were used as detection antibodies. 2.12. Flow cytometric analysis of phagocytosis Heat-killed L. plantarum cells were labeled with fluorescein isothiocyanate (FITC) isomer-I (Dojindo Lab., Kumamoto, Japan) as described by Shida et al. [25]. Briefly, heat-killed cells were suspended at a concentration of 5 mg/mL in 50 mM carbonate buffer (pH 9.6), reacted with FITC isomer-I (5 μg/mL) at 37 °C for 1 h, and then washed three times with sterile PBS. Splenic adherent cells were prepared by incubating splenocytes (1 × 107 cells/well) in 24-well culture plates at 37 °C for 90 min. Following incubation, the supernatant was discarded and the nonadherent cells were removed by a gentle wash with RPMI 1640 medium. Splenic adherent cells (approximately 5 × 105 cells/mL) were cultured with fluorescently labeled heat-killed lactobacilli (1 μg/mL) in 1 mL of RPMI 1640 medium in a 24-well culture plate for 4 or 24 h. In some experiments, splenic adherent cells were treated with anti-CD16/CD32 antibody (clone 2.4G2), anti-CD64 antibody (clone X54-5/7.1) or mouse IgG1 κ isotype control (clone P3) for 1 h

Fig. 5. Cytokine induction of LTA from L. plantarum L-137 and JCM1149. Spleen cells (2 × 106 cells/mL) were cultured with various doses of LTA from L-137 or JCM1149 for 48 h. The amount of IL-12p40 (A), TNF-α (B), IL-6 (C), and IL-10 (D) in supernatants was determined by ELISA. Data obtained from three individual mice are expressed as means ± standard deviations.

326

S. Hatano et al. / International Immunopharmacology 25 (2015) 321–331

at 37 °C prior to the addition of heat-killed bacteria.. The adherent cells were also treated with anti-mouse SR-A IgG2b (clone 2F8) or rat IgG2b, κ isotype control (clone RTK453) for 1 h at 37 °C prior to the addition of heat-killed bacteria.. The adherent cells were also treated with cytochalasin D for 1 h at 37 °C prior to the addition of heat-killed bacteria. Bacterial samples were incubated with mouse anti-LTA mAb (clone 55) or mouse IgG3 isotype control (clone 6A3) at 37 °C for 1 h prior to the addition of splenic adherent cells from FcRγ KO mice. Following co-cultivation of splenic adherent cells with heat-killed L. plantarum, the nonphagocytosed bacteria were removed by vigorous pipetting, and the supernatant was discarded. The adherent splenic cells were dislodged by treatment with 10 mM EDTA-PBS for 10 min, washed with 10 mM EDTA-PBS containing 2% FBS, and then suspended in 500 μL of 10 mM EDTA-PBS containing 2% FBS. Flow cytometric analysis was performed on an EPICS XL flow cytometer with Expo32 software (Beckman Coulter, Miami, FL). The discriminator was set by the size of cells (forward scatter) so as not to detect bacterial cells that were not taken up by splenic adherent cells. 2.13. Statistical analysis Statistical analysis was performed using Statcel2 software (OMS Publishing, Tokorozawa, Japan). Statistical differences in cytokine production were analyzed by analysis of variance, followed by the Tukey–Kramer multiple comparison test for comparison between groups (Figs. 5, 6, 8, and Supplemental Figs. 3–5). Student's unpaired t-test was performed to analyze statistical differences in OD600 at each time point, along with lysozyme sensitivity (Fig. 4). P-values of b0.05 were considered significant.

3. Results 3.1. Comparison of the structures of LTA from L-137 and JCM1149 We first compared the structural characteristics of purified LTA from L-137 and JCM1149. Western blot analysis of the purified LTA from L-137 and JCM1149 revealed compounds with molecular weights of 20–30 kDa, which is higher than the molecular weight of LTA-SA (Fig. 1). SDS-PAGE analysis showed that the LTA of L-137 had a slightly higher mobility than that of JCM1149. To compare the chemical structures of LTA from L-137 and JCM1149, we performed one-dimensional 1H and 13C, and two-dimensional homonuclear and heteronuclear, magnetic resonance experiments. The 1 H NMR spectra of LTA are shown in Fig. 2A and C. Both spectral patterns were similar to the spectra of LTA from L. plantarum KCTC 10887BP [7], indicating that LTA from different L. plantarum strains consists of similar components. However, we could not identify the galactosyl substitution on the poly(GroP) chain that was reported by Jang et al. HSQC and HMBC spectra allowed residues in the poly(GroP) chain to be assigned (Table 1, Supplemental Fig. 1A, B). The assignment of glucose residues shown in HSQC spectra was based on the report of Sanchez Carballo et al. [27]. In addition, HMBC spectra demonstrated that the δC 170.0 had cross-peaks with the α-proton (δH 4.26 or 4.27) and β-proton (δH 1.61) of D-Ala (data not shown). We then analyzed the chemical structure of LTA glycolipid anchors from both strains. The 1H NMR spectra of the glycolipid anchors are shown in Fig. 2B and D. The relative areas of corresponding peaks in the two spectra matched closely. 1H-1H COSY, HSQC, and TOCSY spectra indicated that two α-anomers and one β-anomer existed in three

Fig. 6. IL-12p40 induction by HK L-137 and HK JCM1149 in splenocytes treated with blocking antibodies against phagocytosis-related receptors. Spleen cells (2 × 106 cells/mL) were pretreated with anti-mouse SR-A IgG2b (clone 2F8) (A), anti-mouse CD36 IgA (clone JC63.1) (B), anti-mouse SR-A IgG2b (clone 2F8), anti-mouse CD36 IgA (C), anti-mouse CD14 IgG2a (D), and anti-mouse TLR2 IgG2a (clone C9A12) (E) at the indicated final concentrations for 1 h, then cultured with 100 ng/mL of heat-killed L. plantarum strains for 24 h. The amount of IL-12p40 in supernatants was determined by ELISA. Data obtained from three individual mice are expressed as means ± standard deviations. Asterisks represent a significant difference between groups (*, P b 0.05; **, P b 0.01).

S. Hatano et al. / International Immunopharmacology 25 (2015) 321–331

corresponding sugar residues and a 1,2-diacylglyceride moiety (Table 2, Supplemental Fig. 1C–E). The number of protons and carbons in the sugar residues in the glycolipid anchors could not be determined because of signal overlap. The ROESY spectra indicated that the H-3 of glyceride (δH 3.62) had cross-peaks with one α-anomeric proton (δH 4.93), suggesting linkage between a sugar residue and a diacylglyceride moiety (Supplemental Fig. 1F). To confirm the chemical structure of glycolipid anchors, we further performed tandem MS analysis. The MS spectra of glycolipid anchors from both strains in the positive ion mode contained two groups of peaks (Fig. 3). A comparison of spectra obtained from the positive and negative ion modes revealed that each peak was detected as a sodium adduct ion in the positive ion mode, and as a deprotonated ion in the negative ion mode (data not shown). Subsequent MSn analysis revealed that the group of peaks between m/z 1050 and 1300 (Fig. 3) consisted of a trihexoside moiety and a diacylglyceride moiety, while the second group of peaks between m/z 1350 and 1500 consisted of a monoacyltrihexoside moiety and a diacylglyceride moiety (Supplemental Fig. 2, Supplemental Table 1). These results are consistent with a report on the glycolipid anchor of LTA from L. plantarum KCTC 10887BP [7]. In addition, LTA glycolipid anchors from L-137 and JCM1149 mainly contained C14:0, C16:0, C18:1, and C19:1 fatty acids [28,29], along with an unidentified 315-kDa fatty acid (Supplemental Table 1). We then estimated the polymer length of poly(GroP), as well as the substitution rate with D-Ala and glucose in poly(GroP), based on the peak areas of the methyl protons in the 1H NMR spectra of LTA. The area was calibrated by the ratio of methyl protons to α-anomeric protons in the 1H NMR spectra of the glycolipid anchor (Table 3). The polymer length of poly(GroP) and the substitution rate with D-Ala showed similar values. The major difference between the two types of LTA was the degree of glucose substitution; LTA from HK L-137 had lower glucose substitution than HK JCM1149.

327

3.2. Comparison of HK L-137 and HK JCM1149 cell wall characteristics We next investigated the lysozyme sensitivity of HK lactobacilli (Fig. 4). HK L-137 showed markedly higher sensitivity to lysozyme than HK JCM1149, implying that HK L-137 had a more fragile cell wall than HK JCM1149. The ζ potential at pH 4, 7, and 9, was also determined for the two strains to further investigate characteristics of their cell walls. As shown in Table 4, the ζ potential of HK L-137 at each pH was higher than that of HK JCM1149 at the corresponding pH, suggesting that HK L-137 cells are more negatively charged at pH 4–9 than HK JCM1149 cells. The isoelectric points (ζ potential = 0 mV) of both HK strains were bpH 4. According to these results, the chemical structure of L-137 LTA appeared to be very similar to that of JCM1149, except for the lower level of glucosyl substitution in the poly(GroP) backbone. However, the cell wall structure and/or architecture appeared to differ markedly between the two strains. The HK L-137 cell wall is more fragile and more negatively charged cell than that of HK JCM1149, suggesting that there is a larger amount of exposed LTA on the surface of L137 cells compared with HK JCM1149. 3.3. Comparison of cytokine-inducing activities of LTA from L-137 and JCM1149 Mejierink et al. reported that higher glucosyl substitution with teichoic acid, including LTA, in L. plantarum might enhance IL-12 induction in monocyte-derived cells in a co-culture system [30]. Because of the observed differences in the degree of glucose substitution in the LTA of HK L-137 and HK JCM1149, we next compared the cytokineinducing activities of LTA from L-137 and JCM1149. As shown in Fig. 5,

Fig. 7. Phagocytosis of HK L-137 and HK JCM1149 by splenic adherent cells. Splenic adherent cells (approximately 5 × 105 cells/mL) were cultured with 1 μg/mL of FITC-labeled HK L-137 or HK JCM1149 for 4 or 24 h (A). Splenic adherent cells were pretreated with cytochalasin D (Cyt-D; 5 μg/mL) or 4 μg/mL of anti-mouse SR-A IgG2b (clone 2F8) for 1 h, then cultured with 1 μg/mL of FITC-labeled HK L-137 for 4 h (B). Splenic adherent cells pretreated with anti-mouse SR-A IgG2b (clone 2F8) were cultured with 1 μg/mL of FITC-labeled HK L-137 or HK JCM1149 for 24 h (C). Representative figures from one of three independent experiments are shown. Vertical axis indicates relative cell number and horizontal axis indicates fluorescence intensity.

328

S. Hatano et al. / International Immunopharmacology 25 (2015) 321–331

the cytokine-inducing activity of the LTA from L-137 was almost equivalent to that of the LTA from JCM1149 in a murine splenocyte in vitro model. The levels of each of the investigated cytokines (IL-6, IL-10, TNF-α, and IL-12p40) increased in a dose-dependent manner from 10–100 μg/mL of LTA. These results suggested that the differences in the structure of LTA between the two strains might not be responsible for the difference in IL-12p40-inducing activity of HK L-137 compared with HK JCM1149. 3.4. Effects of LTA-related receptor blocking antibodies on IL-12p40 induction by splenocytes stimulated with HK L-137 or HK JCM1149 We next investigated the roles of SR-A, CD36, CD14, and TLR2, which have been reported to recognize or capture LTA [13,17], during IL-12p40 production by splenocytes stimulated with HK L-137 or HK JCM1149. The splenocytes were pretreated with antibodies against SR-A, CD36, CD14, and TLR2 for 1 h prior to the addition of HK bacteria. As shown in Fig. 6A, blocking SR-A did not affect the induction of IL-12p40 by HK JCM1149, while it selectively inhibited IL-12p40 induction by HK L-137 in a dose-dependent manner. Treatment of the splenocytes with anti-SR-A antibody at 1 μg/mL reduced IL-12p40 induction by HK L-137 to levels equivalent to those induced by HK JCM1149 (Fig. 6A). On the other hand, blocking CD36 inhibited IL-12p40 induction by both HK L-137 and HK JCM1149 at the highest dose of blocking antibody (Fig. 6B). Treatment with both anti‐SR-A and anti-CD36 antibodies exerted an additive inhibitory effect on IL-12p40 production (Fig. 6C). No inhibitory effects of anti-CD14 or TLR2 antibodies on IL-12p40 induction were observed for either of the HK lactobacilli (Fig. 6D, E).

LTA is reportedly a TLR2 ligand that can induce cytokines via a MyD88-dependent pathway [10–12]. Therefore, we checked the dependence of murine splenocytes on TLR2 for cytokine production. Murine splenocytes were stimulated with TLR2 ligands Pam3CSK4 and pLTASA and then treated with a TLR2-blocking antibody (Supplemental Fig. 3). We found that the induction of IL-6, IL-10, and TNF-α by Pam3CSK4 and pLTA-SA was significantly inhibited by the anti-TLR2 antibody; however, induction by LPS was not affected. This indicated that the anti-TLR2 antibody effectively inhibited TLR2-mediated recognition. On the other hand, IL-12p40 induction was only marginal when stimulated with Pam3CSK, and no inhibitory effects of the anti-TLR2 antibody on IL-12p40 induction by Pam3CSK or pLTA-SA were observed. These results suggested that IL-12p40 induction by HK L-137 and HK JCM1149 is independent of TLR2 ligand-binding on murine splenocytes, although a scavenger receptor, CD36, is involved in IL-12p40 induction by both heat-killed lactobacilli. Moreover, SR-A may be responsible for the potent ability of HK L-137 to induce IL-12p40. 3.5. Inhibition of phagocytosis of HK L-137 by treatment with an SR-Ablocking antibody SR-A and CD36 are receptors related to non-opsonic phagocytosis of bacteria [17]. Therefore, to compare the efficiency of bacterial phagocytosis of HK L-137 and HK JCM1149, splenic adherent cells were cocultured with fluorescently labeled heat-killed lactobacilli for 4 or 24 h, and phagocytosis was analyzed by flow cytometry (Fig. 7). HK L-137 and HK JCM1149 showed a similar labeling efficiency (Supplemental Fig. 4). Approximately 23% of splenic adherent cells

Fig. 8. Inhibition of non-opsonic phagocytosis of HK L-137 and IL-12p40 production by anti-LTA antibody. HK L-137 was pretreated with 16 μg/mL of anti-LTA antibody (clone 55) or isotype control (clone 6A3). Splenic adherent cells (approximately 5 × 105 cells/mL) were cultured with 1 μg/mL of FITC-labeled HK L-137 for 4 h in the presence of anti-FcRs antibodies (anti CD16/32 and anti-CD64 mAbs). Splenic adherent cells from FcRγ knockout mice were cultured with 1 μg/mL of the FITC-labeled HK L-137 (A). Splenic adherent cells were cultured with 1 μg/mL of the FITC-labeled HK L-137 for 24 h in the presence of anti-FcRs antibodies (anti CD16/32 and anti-CD64 mAbs). Splenic adherent cells from FcRγ knockout mice were cultured with 1 μg/mL of the FITC-labeled HK L-137 for 24 h. The amount of IL-12p40 in supernatants was determined by ELISA (B). Data obtained from three individual mice are expressed as means ± standard deviations. Vertical axis indicates relative cell number and horizontal axis indicates fluorescence intensity. Asterisks represent a significant difference between groups (*, P b 0.05; **, P b 0.01). n.s. indicates no significant difference between groups. Representative figures from one of three independent experiments are shown.

S. Hatano et al. / International Immunopharmacology 25 (2015) 321–331

phagocytosed HK L-137, while very few splenic adherent cells phagocytosed HK JCM1149 after 4 h of incubation. At 24-h postinoculation, approximately 47% of HK L-137 and approximately 23% of HK JCM1149 cells were phagocytosed (Fig. 7A). These results indicate that HK L-137 was more efficiently ingested by splenic adherent cells than HK JCM1149. To confirm that SR-A acts as a selective receptor for non-opsonic phagocytosis of HK L-137 and enhances the uptake of HK L-137, we further investigated the contribution of SR-A to the phagocytosis of HK L-137 by splenic adherent cells. The adherent cells were pretreated with an antibody against SR-A prior to the addition of fluorescently labeled heat-killed lactobacilli. Blocking SR-A almost completely inhibited the ingestion of HK L-137 by splenic adherent cells after 4 h of co-culture (Fig. 7B). This treatment was comparable to cells treated with cytochalasin D, a potent inhibitor of actin polymerization. Following 24 h of culture, the capacity of the SR-A-blocked cells to take up HK L-137 bacteria decreased to a level similar to that of HK JCM1149 uptake, whose phagocytosis was minimally affected by blocking SR-A (Fig. 7C). Thus, SR-A appears to be responsible for the higher level of phagocytosis of HK L-137 by splenic adherent cells.

3.6. Surface LTA is involved in SR-A-mediated non-opsonic phagocytosis of HK L-137 and higher IL-12p40 induction We found that HK L-137 had a larger amount of exposed LTA on its surface than HK JCM1149. Therefore, LTA may be responsible for the higher degree of phagocytosis via SR-A. To test this possibility, we conducted dual inhibition experiments using an anti-LTA antibody to examine IL-12p40 production and phagocytosis of HK L-137. As shown in Supplemental Fig. 5, the same amount of FITC-labeled HK L-137 treated with anti-LTA antibody was phagocytosed by splenic adherent cells as HK L-137 treated with the isotype control, although the fluorescent intensity of ingested FITC-labeled HK L-137 appeared to be lower following treatment with anti-LTA antibody compared with treatment with the isotype control. Because the complete anti-LTA IgG3 molecule was used, opsonic phagocytosis may mask non-opsonic phagocytosis. Therefore, to avoid IgG3-mediated opsonic phagocytosis, we prepared splenic adherent cells using FcR γ-chain knockout mice, which are defective in CD16/ CD32-mediated opsonic phagocytosis, or by adding anti-CD16/CD32 and anti-CD64 mAbs. As shown in Fig. 8A, FITC-labeled HK L-137 treated with anti-LTA antibody was phagocytosed at a lower level by splenic adherent cells from FcRγ KO mice, or in the presence of anti-FcRs mAbs, than HK L-137 treated with the isotype control. The fluorescent intensity of splenic adherent cells containing phagocytosed FITC-labeled HK L-137 treated with anti-LTA antibody was lower than those cells containing FITC-labeled HK L-137 treated with the isotype control. The decrease in fluorescent intensity might be caused by the decomposition of fluorescent dye molecules, along with enhanced phagocytosis, or by a decrease in the amount of ingested bacteria in each splenic adherent cell. Our results show a close correlation between the efficiency of phagocytosis and IL-12p40 induction ability of heat-killed lactobacilli, suggesting the importance of bacterial phagocytosis for IL-12p40 production. To confirm this possibility, we examined the role of bacterial phagocytosis in IL-12p40 production using cytochalasin D. IL-12p40 production by HK L-137-simulated splenocytes markedly decreased following cytochalasin D treatment (Supplemental Fig. 6), which was correlated with the inhibition of phagocytosis (Fig. 7B). These results suggest that IL-12p40 induction by heat-killed lactobacilli requires phagocytosis. Treatment of HK L-137 with anti-LTA antibody also inhibited IL-12p40 production by splenic adherent cells under nonopsonic conditions (Fig. 8B). Thus, these results suggest that the higher efficiency of non-opsonic phagocytosis of HK L-137 via LTA and SR-A is responsible, at least partly, for the potent IL-12p40 inducing ability of HK L-137.

329

4. Discussion To elucidate the mechanisms by which HK L-137 causes potent IL12p40 induction, we compared the chemical structure of LTA isolated from HK L-137 and another L. plantarum strain, HK JCM1149, which causes lower IL-12p40 induction. The chemical structures of LTA from L-137 and JCM1149 were similar, and were comparable to those of L. plantarum strains previously identified by Fischer and Rösel [31] and Jang et al. [7]. The only major difference between the structures was the degree of glucosyl substitution in the poly(GroP) chain of LTA. Mejierink et al. reported that higher levels of glucosyl substitution in TA (including LTA) of L. plantarum might enhance IL-12 induction in a monocyte co-culture system [30]. However, inconsistent with their findings, we found that HK L-137 LTA had a lower amount of glucosyl substitution in the poly(GroP) chain than HK JCM1149, leading us to presume that glucosyl substitution in the poly(GroP) chain of LTA is not responsible for the higher IL-12p40-inducing activity of HK L-137. We previously reported that the cell walls of L-137 contain abundant levels of alanine, according to the Ala/meso-DAP molar ratio, raising the possibility that the substitution rate of D-Ala in the LTA may be responsible for the potent IL-12p40 induction by HK L-137 [21]. However, the levels of D-Ala substitution were similar between strains L-137 and JCM1149, excluding this possibility. Shida et al. reported that the cell walls of L. casei Shirota lost their rigidity by depletion of polysaccharides, becoming susceptible to digestion by N-acetylmuramidase [25]. This suggested that the rigidity of the cell walls of some Lactobacillus strains might be ascribed to extracellular polysaccharide(s) (EPS) and/or surface layer (S-layer) protein(s). In some Lactobacillus strains, cell surface LTA can be masked by EPS or S-layer proteins [32]. Our findings showed that there is a larger amount of LTA exposed on the surface of L-137 compared with JCM1149, and that L-137 cells are more susceptible to lysozyme, suggesting that the cell walls of L. plantarum L-137 might be less modified by EPS or S-layer proteins than those of JCM1149 [32–34]. Schar-Zammaretti et al. examined the ζ potential of several strains of Lactobacillus [32,35] and found that the isoelectric points of most lactobacilli are around pH 4, except for L. johnsonii ATCC332 (isoelectric point b pH 3). In our study, the isoelectric points of both HK L. plantarum strains were b pH 4, suggesting that the cell walls of both strains contain substantial amounts of acidic groups with a low pKa, such as phosphate groups derived from TAs and/or anionic polysaccharides [32]. The higher ζ potential of HK L-137 at pH 4 suggests that the cell walls of HK L-137 contain acidic groups with lower pKa, more acidic groups, or fewer basic groups derived from S-layer proteins with a high pKa than the cell walls of HK JCM1149 [32]. Taken together, our analysis of the structure of LTA shows that HK L-137 has a more fragile and negatively charged cell wall than HK JCM1149, suggesting that HK L-137 exposes more LTA on its cell surface than HK JCM1149. A notable finding of the present study is that SR-A is responsible for the higher levels of non-opsonic phagocytosis of HK L-137 by splenic adherent cells than HK JCM1149. SR-A recognizes bacterial surface LTA and lipoprotein and mediates non-opsonic phagocytosis of Grampositive bacteria such as S. pyogenes and S. aureus [17,19,20]. SR-A can also bind a broad range of polyanionic ligands, including the LTA of Gram-positive bacteria, [17,18]. We demonstrated in this study that pretreatment with anti-LTA antibody selectively blocked the nonopsonic phagocytosis of HK L-137 by splenic adherent cells. HK L-137 is therefore likely to expose more LTA and is more negatively charged than HK JCM1149 at pH 4–9. Therefore, we hypothesize that HK L-137 may have greater affinity for SR-A than HK JCM1149, thereby being phagocytosed more efficiently. We demonstrated that blocking SR-A did not completely inhibit the phagocytosis of HK L-137 following 24 h of incubation (Fig. 7C). These results indicate that SR-A likely contributes to the rapid uptake of most HK L-137 cells while the remaining HK L-137 cells, and most JCM1149 cells, are only slowly ingested via

330

S. Hatano et al. / International Immunopharmacology 25 (2015) 321–331

alternative receptors(s). CD36 also mediates phagocytosis of various Gram-positive bacteria [16]. As can be seen in Fig. 6C, blocking CD36 inhibited phagocytosis and IL-12p40 induction by both HK L-137 and HK JCM1149, suggesting that CD36 may act as a receptor for nonopsonic phagocytosis of both strains of L. plantarum [16,17]. In this study, we demonstrated that phagocytosis is crucial for IL-12p40 induction by both HK L-137 and HK JCM1149, and that SR-A recognition of HK L-137 is selectively responsible for the higher IL-12p40 induction by murine splenocytes. SR-A has no proinflammatory signaling pathway for IL-12p40 induction, but rather, down-regulates the constitutive expression of some inflammatory genes by inhibiting the activation of TNF receptor-associated factor 6 [36]. Therefore, it is plausible that pattern recognition receptor(s) other than SR-A may be involved in IL-12p40 induction in HK L-137. In the present study, an anti-LTA antibody inhibited the non-opsonic phagocytosis of HK L-137, and reduced IL-12p40 production by HK L-137-stimulated phagocytes. It is possible that the rapid and efficient uptake via SR-A is a prerequisite for activation of IL-12p40-inducing signals. Blocking bacterial uptake might cancel the effect of factor(s) that impede subsequent steps, such as forming endosomes, digestion of bacteria, and endosomal pattern recognition. However, we do not know whether LTA or other components of HK L-137 cells can induce IL-12p40 through pattern recognition receptor(s) other than SR-A. Several receptors related to pattern recognition in innate immunity have been identified as receptors for LTA. TLR2 recognizes various fungal and bacterial cell wall components, including the LTA of Gram-positive bacteria. CD14 and CD36 are co-receptors for TLR2 during the uptake of proinflammatory cytokine responses to LTA [13]. In the present study, anti-TLR-2 or anti-CD14 antibodies did not inhibit IL-12p40 production. The specific receptor for LTA or other components of the HK L-137 surface remain unclear, and further studies are required to identify the endosomal pattern recognition receptor for HK L-137. In conclusion, the higher level of exposure of LTA on the cell surface of HK L-137 might cause it to be more readily phagocytosed and cause higher IL-12p40 induction than HKJCM1149. The affinity of SR-A for LTA might be responsible, at least in part, for the potent IL-12p40inducing ability of HK L-137. Therefore, the level of LTA exposed on the cell surface of a Lactobacillus strain may be important for determining its IL‐12p40-inducing ability. These results will be useful for the selection of lactic acid bacteria as probiotics.

Author contributions S.H., Y.H., S.M. and Y.Y. designed the research; S.H. and Y.H. performed the experiments and analyzed the data; Y.Y. and S.H. supervised the experimental work; and Y.H. and Y.Y. wrote the manuscript.

Funding This work was supported by a Grant-in-Aid from the Japan Society for the Promotion of Science, and grants from the Japanese Ministry of Education, Culture, Sports, Science and Technology (Y.Y.), and the Takeda Science Foundation (Y.Y.).

Acknowledgments We are grateful to Mr. Atsushi Kotani (Somatech Center, House Foods Corporation) for his kind instructions on the use of the jar fermenter. We also thank Mr. Hiroshi Sasako (Somatech Center, House Foods Corporation) and Dr. Katsuhiko Minoura (Department of Pharmacy, Osaka University of Pharmaceutical Sciences) for helping with NMR spectroscopy. We also thank Akiko Yano, Miki Kijima, and Mihoko Ohkubo for their technical assistance.

Appendix A. Supplementary data Supplementary data to this article can be found online at http://dx. doi.org/10.1016/j.intimp.2015.02.011. References [1] Olympia M, Ono H, Shinmyo A, Takano M. Lactic-acid bacteria in a fermented fishery product, burong bangus. J Ferment Bioeng 1992;73:193–7. [2] Murosaki S, Yamamoto Y, Ito K, Inokuchi T, Kusaka H, Ikeda H, et al. Heat-killed Lactobacillus plantarum L-137 suppresses naturally fed antigen-specific IgE production by stimulation of IL-12 production in mice. J Allergy Clin Immunol 1998;102:57–64. [3] Murosaki S, Muroyama K, Yamamoto Y, Yoshikai Y. Antitumor effect of heat-killed Lactobacillus plantarum L-137 through restoration of impaired interleukin-12 production in tumor-bearing mice. Cancer Immunol Immunother 2000;49:157–64. [4] Hirose Y, Murosaki S, Yamamoto Y, Yoshikai Y, Tsuru T. Daily intake of heat-killed Lactobacillus plantarum L-137 augments acquired immunity in healthy adults. J Nutr 2006;136:3069–73. [5] Matsuguchi T, Takagi A, Matsuzaki T, Nagaoka M, Ishikawa K, Yokokura T, et al. Lipoteichoic acids from Lactobacillus strains elicit strong tumor necrosis factor alpha-inducing activities in macrophages through Toll-like receptor 2. Clin Diagn Lab Immunol 2003;10:259–66. [6] Mohamadzadeh M, Pfeiler EA, Brown JB, Zadeh M, Gramarossa M, Managlia E, et al. Regulation of induced colonic inflammation by Lactobacillus acidophilus deficient in lipoteichoic acid. Proc Natl Acad Sci U S A 2011;108(Suppl. 1):4623–30. [7] Jang KS, Baik JE, Han SH, Chung DK, Kim BG. Multi-spectrometric analyses of lipoteichoic acids isolated from Lactobacillus plantarum. Biochem Biophys Res Commun 2011;407:823–30. [8] Morath S, Geyer A, Spreitzer I, Hermann C, Hartung T. Structural decomposition and heterogeneity of commercial lipoteichoic acid preparations. Infect Immun 2002;70: 938–44. [9] Grangette C, Nutten S, Palumbo E, Morath S, Hermann C, Dewulf J, et al. Enhanced antiinflammatory capacity of a Lactobacillus plantarum mutant synthesizing modified teichoic acids. Proc Natl Acad Sci U S A 2005;102:10321–6. [10] Draing C, Traub S, Deininger S, Mang P, Möller HM, Manso M, et al. Polypropylene glycol is a selective binding inhibitor for LTA and other structurally related TLR2 agonists. Eur J Immunol 2008;38:797–808. [11] Lotz S, Aga E, Wilde I, van Zandbergen G, Hartung T, Solbach W, et al. Highly purified lipoteichoic acid activates neutrophil granulocytes and delays their spontaneous apoptosis via CD14 and TLR2. J Leukoc Biol 2004;75:467–77. [12] Ozinsky A, Underhill DM, Fontenot JD, Hajjar AM, Smith KD, Wilson CB, et al. The repertoire for pattern recognition of pathogens by the innate immune system is defined by cooperation between toll-like receptors. Proc Natl Acad Sci U S A 2000; 97:13766–71. [13] Nilsen NJ, Deininger S, Nonstad U, Skjeldal F, Husebye H, Rodionov D, et al. Cellular trafficking of lipoteichoic acid and Toll-like receptor 2 in relation to signaling: role of CD14 and CD36. J Leukoc Biol 2008;84:280–91. [14] Chow JC, Young DW, Golenbock DT, Christ WJ, Gusovsky F. Toll-like receptor-4 mediates lipopolysaccharide-induced signal transduction. J Biol Chem 1999;274: 10689–92. [15] Hoebe K, Georgel P, Rutschmann S, Du X, Mudd S, Crozat K, et al. CD36 is a sensor of diacylglycerides. Nature 2005;433:523–7. [16] Baranova IN, Kurlander R, Bocharov AV, Vishnyakova TG, Chen Z, Remaley AT, et al. Role of human CD36 in bacterial recognition, phagocytosis, and pathogen-induced JNK-mediated signaling. J Immunol 2008;181:7147–56. [17] Areschoug T, Gordon S. Scavenger receptors: role in innate immunity and microbial pathogenesis. Cell Microbiol 2009;11:1160–9. [18] Haworth R, Platt N, Keshav S, Hughes D, Darley E, Suzuki H, et al. The macrophage scavenger receptor type A is expressed by activated macrophages and protects the host against lethal endotoxic shock. J Exp Med 1997;186:1431–9. [19] Areschoug T, Waldemarsson J, Gordon S. Evasion of macrophage scavenger receptor A-mediated recognition by pathogenic streptococci. Eur J Immunol 2008;38:3068–79. [20] Sever-Chroneos Z, Krupa A, Davis J, Hasan M, Yang CH, Szeliga J, et al. Surfactant protein A (SP-A)-mediated clearance of Staphylococcus aureus involves binding of SP-A to the staphylococcal adhesin eap and the macrophage receptors SP-A receptor 210 and scavenger receptor class A. J Biol Chem 2011;286:4854–70. [21] Hirose Y, Murosaki S, Fujiki T, Yamamoto Y, Yoshikai Y, Yamashita M. Lipoteichoic acids on Lactobacillus plantarum cell surfaces correlate with induction of interleukin12p40 production. Microbiol Immunol 2010;54:143–51. [22] Fujiki T, Hirose Y, Yamamoto Y, Murosaki S. Enhanced immunomodulatory activity and stability in simulated digestive juices of Lactobacillus plantarum L-137 by heat treatment. Biosci Biotechnol Biochem 2012;76:918–22. [23] Gründling A, Schneewind O. Synthesis of glycerol phosphate lipoteichoic acid in Staphylococcus aureus. Proc Natl Acad Sci U S A 2007;104:8478–83. [24] Gründling A, Schneewind O. Genes required for glycolipid synthesis and lipoteichoic acid anchoring in Staphylococcus aureus. J Bacteriol 2007;189:2521–30. [25] Shida K, Kiyoshima-Shibata J, Nagaoka M, Watanabe K, Nanno M. Induction of interleukin-12 by Lactobacillus strains having a rigid cell wall resistant to intracellular digestion. J Dairy Sci 2006;89:3306–17. [26] Evans D, Hakån W. The colloidal domain: where physics, chemistry, biology, and technology meet. New York: VCH Publishers; 1994. [27] Sánchez Carballo PM, Vilen H, Palva A, Holst O. Structural characterization of teichoic acids from Lactobacillus brevis. Carbohydr Res 2010;345:538–42.

S. Hatano et al. / International Immunopharmacology 25 (2015) 321–331 [28] Fischer K, Stein K, Ulmer AJ, Lindner B, Heine H, Holst O. Cytokine-inducing lipoteichoic acids of the allergy-protective bacterium Lactococcus lactis G121 do not activate via Toll-like receptor 2. Glycobiology 2011;21:1588–95. [29] Johnsson T, Nikkila P, Toivonen L, Rosenqvist H, Laakso S. Cellular fatty acid profiles of Lactobacillus and Lactococcus strains in relation to the oleic acid content of the cultivation medium. Appl Environ Microbiol 1995;61:4497–9. [30] Meijerink M, van Hemert S, Taverne N, Wels M, de Vos P, Bron PA, et al. Identification of genetic loci in Lactobacillus plantarum that modulate the immune response of dendritic cells using comparative genome hybridization. PLoS One 2010;5:e10632. [31] Fischer W, Koch HU, Rösel P, Fiedler F, Schmuck L. Structural requirements of lipoteichoic acid carrier for recognition by the poly(ribitol phosphate) polymerase from Staphylococcus aureus H. A study of various lipoteichoic acids, derivatives, and related compounds. J Biol Chem 1980;255:4550–6.

331

[32] Schär-Zammaretti P, Ubbink J. The cell wall of lactic acid bacteria: surface constituents and macromolecular conformations. Biophys J 2003;85:4076–92. [33] Andre G, Deghorain M, Bron PA, van Swam II, Kleerebezem M, Hols P, et al. Fluorescence and atomic force microscopy imaging of wall teichoic acids in Lactobacillus plantarum. ACS Chem Biol 2011;6:366–76. [34] Sánchez B, Urdaci MC. Extracellular proteins from Lactobacillus plantarum BMCM12 prevent adhesion of enteropathogens to mucin. Curr Microbiol 2012;64:592–6. [35] Schär-Zammaretti P, Dillmann ML, D'Amico N, Affolter M, Ubbink J. Influence of fermentation medium composition on physicochemical surface properties of Lactobacillus acidophilus. Appl Environ Microbiol 2005;71:8165–73. [36] Yu X, Yi H, Guo C, Zuo D, Wang Y, Kim HL, et al. Pattern recognition scavenger receptor CD204 attenuates Toll-like receptor 4-induced NF-kappaB activation by directly inhibiting ubiquitination of tumor necrosis factor (TNF) receptor-associated factor 6. J Biol Chem 2011;286:18795–806.