Corynebacterium pyruviciproducens-peptidoglycan: A novel bacterial peptidoglycan inhibiting overexpression of MyD88 in macrophages

Abstract

Objectives:

Bacterial peptidoglycan (PGN) is an essential ligand of TLR2 inducing inflammatory damage by boosting MyD88 overexpression in pathogen invasion, such as Methicillin-resistant Staphylococcus aureus (MRSA) infection. CP-PGN is a novel PGN from an adjuvant bacterium, displaying anti-infection immune regulation. This study aimed to clarify the unique moderation of MyD88 expression by CP-PGN.

Methods:

Compared with other ligands of TLR2, high expression of MyD88 in macrophages was established by MRSA and virus to investigate the immunomodulation of CP-PGN.

Results:

Compared with PGN derived from MRSA (M-PGN) and chemosynthetic Pam3CSK4 of model agonists of TLR2, CP-PGN could inhibit overexpression of MyD88 in a time- and dose-dependent way in infected macrophages by MRSA or Abelson leukemia virus. CP-PGN also promoted more anti-inflammatory IL-10 and less pro-inflammatory TNF-α in immature primary macrophages. Furthermore, IL-10 secretion induced by CP-PGN was reduced most significantly by blocking the dimer formation of MyD88 with ST2825 and lowering down expression by si-MyD88.

Conclusion:

CP-PGN could inhibit MyD88 overexpression by infection to moderate inflammatory cytokines. Therefore, CP-PGN is a novel potential ligand of TLR2 to induce inflammatory balance in the process of host defense against invading pathogens.

Introduction

The innate immune system plays a vital role in the fight to infectious diseases, since it can recognize pathogen-associated molecular patterns (PAMPs) from invading microorganisms through the pattern recognition receptors (PRRs) expressed on professional antigen-presenting cells (APCs), such as macrophages or dendritic cells (DCs).^1,2 Toll-like receptors (TLRs) are the most important family of PRRs, whose discovery has had a great impact in the field of innate immunity and led to two Nobel Prizes in Physiology or Medicine. Until now, 10 human and 13 murine subtypes of TLRs have been identified.^3,4 TLR2 is the only TLR that forms functional heterodimers with other TLRs (TLR1, TLR6, and TLR4).^5,6 TLR2 also interacts with different non-TLR molecules. ⁷ All of these allow TLR2 to recognize a great variety of PAMPs from all microbial phyla including viruses, fungi, bacteria, and parasites. Independent of different dimer patterns, the downstream signaling cascade proceeds through the myeloid differentiation primary response protein 88 (MyD88)-dependent pathway.^8,9

However, different ligands of TLR2 lead to adverse cytokines production, because MyD88 is a core signaling molecule to play an essential role in multiple signal transduction pathways. In a whole, these cytokines can be divided into two patterns through specific signal transduction from different agonists. For example, the pro-inflammatory cytokine of tumor necrosis factor alpha (TNF-α) leads inflammatory damage when activating the immune defense.^9,10 The anti-inflammatory cytokine of interleukin-10(IL-10) alleviates the inflammatory response caused by infection.^11,12 Therefore, TLR2-MyD88 signaling pathway may have a dual role and be a central element in the anti-infection immune response or in the pro-inflammatory processes. In general, high expression of MyD88 is accompanied with pro-inflammatory response. In-depth understanding of these two mechanisms induced by different ligands is essential for the development of new approaches to design anti-inflammatory immunotherapy.

Peptidoglycan (PGN) is the main component of bacterial cell wall and an essential ligand of TLR2, playing a vital and diverse role of regulating inflammatory response in different infections. ¹³ Muramyl dipeptide (MDP) derived from the conserved peptidoglycan is well known as a common functional unit of TLR2 agonist. ¹⁴ In fact, it has gone far beyond our cognition of the immunomodulatory mechanism of TLR2 pathway by the whole natural peptidoglycan. In recent years, it has been found that peptidoglycan of different bacteria can induce different cytokines and mediate different immune responses. For the most well-known peptidoglycan of Staphylococcus aureus (S. aureus, SAU), it has been found to promote pro-inflammatory cytokine (TNF-α) in TLR2-dependent manner.^15,16 In addition, two mutants of Listeria monocytogenes undergo a genetic variation encoding two different peptidoglycans, one of which induces significantly reduced IL-10 secretion, the other boosting increased IL-10 secretion. ¹⁷

Our previous studies indicated that the newly discovered Corynebacterium pyruviciproducens (C. pyruviciproducens, CP) and its peptidoglycan (CP-PGN) displayed significant immune adjuvant function and attenuated inflammatory damage caused by methicillin-resistant Staphylococcus aureus (MRSA) infection by TLR2-dependent pathway.^18-20 Here, in comparison with other peptidoglycan and model ligands of TLR2, we investigated the regulation of CP-PGN on MyD88 expression to understand its immunomodulatory effect against infection. This may illustrate the diversity of TLR2-MyD88 signaling in different bacterial peptidoglycans during microbial infection.

Materials and Methods

Bacterial strains and PGN extraction

C. pyruviciproducens (CCUG 57046, ATCC BAA-1742 obtained from Olive View Medicine Center, University of California, Los Angeles, CA, USA) and methicillin-resistant Staphylococcus aureus (MRSA) were cultured and collected to be inactivated by heat. CP-PGN and M-PGN were extracted according to trichloroacetic acid (TCA) method, as described before.^18-20 The extracted product was identified as a conserved bacterial peptidoglycan by Fourier spectrum analysis (Kebiao Testing Institute, Qingdao, China), presenting conserved structures of polysaccharide skeleton and polypeptide chain. In addition, this extract could be completely degraded by lysozyme. These data confirmed that this extract well maintained its nature and purity of PGN. ²¹

RAW264.7 cell line and treatments

RAW264.7 cell line was maintained in RPMI1640 medium (HyClone, USA) containing 10% fetal bovine serum (FBS) (GIBCO, USA), 1% L-glutamine (2 mM), 1% sodium pyruvate (1 mM) cultured at 37°C, and 5% CO₂ atmosphere. ²⁰ RAW264.7 cells were incubated with different concentrations of CP-PGN (10, 50, and 100 μg/mL), for different times (8, 24, and 36 h). The 50 μg/mL and 24 h of stimulation were chosen in the following experiments, due to cause none significant cell death measured by CCK-8 (Dojindo, Japan) assay. The working concentration of other stimulations as control was chosen according to the common practice and manufacturer’s instructions in our lab. RAW264.7 cells were collected to detect MyD88 protein after stimulations with CP-PGN (50 μg/mL), MRSA (50 μg/mL), or Pam3CSK4 (10 μg/mL) for 24h. To determine the regulations of different bacterial PGN, CP-PGN and M-PGN stimulated RAW264.7 at the same concentration of 50 μg/mL for 24h or 48h. To simulate the process of pathogen invasion, RAW264.7 cells were pretreated by the whole MRSA killed by heat (100 μg/mL) alone for 24h. After discarding the former medium and washing with sterile phosphate buffer saline (PBS), the infected cells were stimulated with 50 μg/mL CP-PGN and M-PGN, respectively, for the next 24 h, to detect the MyD88 protein changes. Each test was repeated 3 times independently and at least 3 wells were set in one group.

Primary peritoneal macrophages and treatment

Primary peritoneal macrophages were derived from BALB/c mice (Yangzhou University, China), referring to the previous reports.^18,20 Peritoneal macrophages were harvested immediately by lavaging with PBS repeatedly. Red blood cell (RBC) lysis buffer was used to destroy RBCs. The remaining cells were washed twice and plated in RPMI 1640 medium containing 10% FBS at 37°C, 5% CO₂ atmosphere. After 12h incubation, non-adherent cells were removed by gently washing with PBS and freshly prepared medium was added. All the cells were resuspended and adjusted to a concentration of 10⁶cells/mL before stimulations. In most experiments, macrophages above were cultured with CP-PGN (50 μg/mL), MRSA (100 μg/mL), and Pam3CSK4 (InvivoGen, USA, 10 μg/mL) in the following experiments. To determine the regulations of different TLR2 ligands, CP-PGN and Pam3CSK4 stimulated primary macrophages at the same concentration for 24h. To simulate the process of pathogen invasion, primary macrophages were pretreated by the whole MRSA (100 μg/mL) for 24h. After discarding the medium and washing with PBS, the pretreated macrophages were incubated with CP-PGN and Pam3CSK4, respectively, for the next 24 h, to detect the MyD88 protein changes. Each test was repeated 3 times independently and at least 3 wells were set in one group. All the measures taken for the mice were in accordance with approved guidelines (Guideline for the Care and Use of Laboratory Animals) established by the Chinese Council on Animals Care. In addition, all the animal experimental protocols were approved by the Ethics Committee of Soochow University (81501425).

Detecting MyD88 by Western blot

Samples from RAW264.7 and peritoneal macrophages were lysed in RIPA buffer (Beyotime, China). Protease inhibitors were added. Samples were separated on SDS-PAGE and then were transferred onto nitrocellulose membranes (Pall Corporation, USA) and blocked with 5% nonfat milk at room temperature for 1 h. Then membranes were incubated overnight at 4°C with primary antibodies. MyD88 was targeted using rabbit monoclonal anti-MyD88 antibody (Cell Signaling Technology, USA) at 1:500 dilution. Anti-β-actin mouse monoclonal primary antibody (Beyotime, USA) was used as a control. Membranes were washed with phosphate buffered saline tween-20 and incubated with secondary goat anti-rabbit (1:10,000) or goat anti-mouse (1:10,000) at room temperature for 1 h. Bands were visualized using the Odyssey Infrared Imaging System to evaluate the differences. Representative western blot images of MyD88 and actin were shown.

ELISA assay for cytokines

Culture supernatants from those stimulated cells above were collected at the specified time to measure TNF-α and IL-10. Enzyme-linked immunosorbent assay (ELISA) was performed according to the manufacturer’s instructions (PeproTech, USA).

Inhibiting MyD88 activation by ST2825

RAW264.7 cells were treated with CP-PGN (50 μg/mL), M-PGN (50 μg/mL), MRSA (100 μg/mL), and Pam3CSK4 (InvivoGen, USA, 10 μg/mL) alone for 3 h. For determining MyD88 dependence, after pretreatment of RAW264.7 cells with ST2825 (MCE, USA, 20 μg/mL) of the MyD88 inhibitor for 3h, different stimulations listed above were added for another 3h. The stimulated cells were collected to detect the TNF-α and IL-10 secretions. Three independent experiments were repeated and there were at least 3 wells in each group. For relative gene-expression analysis of TNF-α and IL-10, total RNAs were isolated with Trizol (Invitrogen, USA), and reverse transcription was performed with the Takara RNA polymerase chain reaction (PCR) kit (Takara, Japan) according to the manufacturer’s protocol. qRT-PCR was performed using a SYBR Green qRT-PCR kit (Takara, Japan) on LightCycler480 (Roche Diagnostics, USA). Relative expression was calculated and normalized to the expression of the β-actin using the 2^−∆∆CT method. The primer sets used for this study were TNF-α: forward 5’- AACTAGTGGTGCCAGCCGAT-3’ and reverse 5′-CTTCACAGAGCAATGACTCC-3′; IL-10: forward 5’-ATCGATTTCTCCCCTGTG-3’ and reverse 5′- C AATGGGAACTGAGGTATCAG -3′; β-Actin: forward 5’-TGGAATCCTGTGGCATCC ATGAAAC-3’ and reverse 5’-TAAAACGCAGC TCAGTAACAGTCCG-3’. Three independent experiments were repeated and there were at least 3 wells in each group.

Knockdown MyD88 expression by RNA interference

5 × 10⁶/well RAW264.7 were suspended with RPMI1640 medium without FBS; 8 μL si-MyD88 or si-negative control plasmid (si-N.C, GenePharma, China) was added into each well. Cooled RAW264.7 cells were treated at 340V and 20 ms by electroporation using Bio-Rad Gene Pulser electroporation apparatus. After being cooled down for 10 min on the ice, the cells were transferred to the Petri dish and cultured with complete medium including si-RNA. About 24 h after transfection, a new medium with non-interference RNA was added for subsequent experiments. Cells were transfected with si-N.C as wild type (WT). CP-PGN (50 μg/mL) stimulated the transfected cells for 24 h to detect MyD88 protein by western blot and cytokines by ELISA. Three independent experiments were repeated and there were at least 3 wells in each group.

Statistical Analysis

All the data involved in the present study were from three independent experiments (n = 3–5 wells/group) at least. Data were presented as means ± SD. The two-tailed Student’s t-test or one-way ANOVA was used for statistical analysis. Differences were considered statistically significant when p < 0.05. *p < 0.05, **p < 0.01, ***p < 0.001.

Results

CP-PGN inhibited MyD88 overexpression boosted by MRSA

Primary peritoneal macrophages, derived from BALB/c mice, are in an immature state with low level of MyD88 before stimulation. CP-PGN alone promoted a little MyD88 expression of immature primary peritoneal macrophages, but more MyD88 protein was presented by MRSA-stimulated cells (Figure 1(a)). Surprisingly, even if MyD88 in immature macrophages has been increased greatly by MRSA, the following CP-PGN stimulation could significantly revers this high expression of MyD88 to a moderate level (Figure 1(b)). In addition, CP-PGN–treated primary macrophages presented more IL-10 and less TNF-α secretions (Figures 1(c) and (d)). Pam3CSK4, a simple chemically synthesized ligand, is well known as a positive agonist of TLR2. Only CP-PGN could moderate the increased MyD88 by MRSA (Figure 2). All these results demonstrate that CP-PGN could inhibit MyD88 overexpression boosted by MRSA with more IL-10 in primary peritoneal macrophages.

Figure 1.

Corynebacterium pyruviciproducens-peptidoglycan moderates the increased MyD88 expression triggered by MRSA in primary peritoneal macrophages. Primary peritoneal macrophages obtained from BALB/c mice were stimulated with CP-PGN alone (50 μg/mL, 24h), inactivated MRSA alone (100 μg/mL, 24h), or MRSA followed by CP-PGN. MyD88 protein was detected by western blot analysis (a–b). TNF-α(c) and IL-10(d) in the supernatant were detected by ELISA. All data are expressed as mean ± s.d. in triplicate determinations. Statistical evaluation was performed by one-sided, two-sample with equal variance t-tests.*p < 0.05, **p < 0.01, ***p < 0.001.

Figure 2.

Corynebacterium pyruviciproducens-peptidoglycan inhibits the increased MyD88 expression by MRSA in primary peritoneal macrophages, instead of Pam3CSK4 of a common TLR2 agonist ligand. (a) Pam3CSK4 (10 μg/mL) or CP-PGN (50 μg/mL) alone treated primary peritoneal macrophages for 24h, detecting MyD88 by western blot. (b) Primary peritoneal macrophages were stimulated by 100 μg/mL whole MRSA cells for 24 h, followed by Pam3CSK4 (10 μg/mL) or CP-PGN (50 μg/mL) for another 24h, then MyD88 detected by western blot. Western blot shown is a representative image of triplicate experiment.

CP-PGN inhibited MyD88 expression in a time- and dose-dependent manner in RAW264.7 Cells

In primary peritoneal macrophages, hyper-expression of MyD88 was boosted by MRSA in the experiments above. Thus, we assessed the role of CP-PGN if the cell line RAW264.7 established by Abelson leukemia virus infection carried more MyD88 themselves. ²² As shown in Figure 3(a) and (b), the MyD88 protein in CP-PGN–treated RAW264.7 cells was significantly lower than that of untreated control, inhibited by CP-PGN in a time- and dose-dependent manner. Furthermore, only CP-PGN significantly suppressed the MyD88 expression of RAW264.7 cells compared with MRSA and Pam3CSK4 (Figure 3(c)). These results suggest that CP-PGN promotes a robust low expression of MyD88.

Figure 3.

Corynebacterium pyruviciproducens-peptidoglycan inhibits the high expression of MyD88 in RAW264.7 cell line in a dose- and time-dependent manner. MyD88 protein of RAW264.7 cells was detected by western blot analysis. (a) 50 μg/mL CP-PGN stimulated RAW264.7 cells for a prolonged time (8, 24, and 36 h). (b) Different doses of CP-PGN (10, 50, and 100 μg/mL) stimulated RAW264.7 cells for 24 h. (c) Only CP-PGN (50 μg/mL for 24h) significantly suppressed the MyD88 protein, compared with MRSA and Pam3CSK4. Western blot shown is a representative image of triplicate experiments.

Compared with M-PGN, CP-PGN decreased MyD88 expression in RAW264.7

Peptidoglycan of MRSA (M-PGN) is a common bacterial peptidoglycan as a contrast here to determine CP-PGN regulation of MyD88. As shown in Figure 4(a), MyD88 expression in CP-PGN–treated RAW264.7 cells was significantly lower than that of M-PGN treatment. In addition, increased MyD88 expression by MRSA could be reversed by CP-PGN, instead of M-PGN (Figure 4(b)). These results suggest that it is different from other bacterial peptidoglycan for CP-PGN to induce a low expression of MyD88.

Figure 4.

Corynebacterium pyruviciproducens-peptidoglycan, not the peptidoglycan of MRSA (M-PGN), reverses the increased MyD88 expression by MRSA in RAW264.7. MyD88 protein of the stimulated RAW264.7 cells was detected by western blot analysis. (a) CP-PGN (50 μg/mL) or M-PGN (50 μg/mL) alone stimulated RAW264.7 cells for 24h or 48h. (b) After MRSA (100 μg/mL) pretreated RAW264.7 cells for 24h, the following CP-PGN (50 μg/mL) or M-PGN (50 μg/mL), respectively, stimulated cells for another 24h. Western blot shown is a representative image of triplicate experiments.

CP-PGN promoted IL-10 Secretion in a MyD88-dependent way

All the ligands of TLR2 regulate immune responses through the MyD88, because it has been demonstrated as an essential signaling molecule of TLR2. ²³ We hypothesized that CP-PGN could not work without MyD88, although CP-PGN inhibited hyper-expression of MyD88. To test this possibility, we incubated RAW264.7 macrophages with an inhibitor ST2825 or transfected by small interfering RNA to disturb the action of MyD88. As expected, blocking the dimer formation of MyD88 with ST2825 almost completely inhibited the secretions of IL-10 boosted by CP-PGN (Figure 5). Consistently, knockdown of MyD88 with siRNA significantly decreased much more IL-10 than TNF-α by CP-PGN (Figure 6). These results indicate that like other ligands of TLR2, CP-PGN fails to boost IL-10 secretion in macrophages without MyD88.

Figure 5.

The secretions of cytokines boosted by CP-PGN can be suppressed by the MyD88 inhibitor ST2825 in RAW264.7 cells. RAW264.7 cells were treated for 3 h by CP-PGN (50 μg/mL), M-PGN (50 μg/mL), Pam3CSK4 (10 μg/mL), and ST2825 of a specific MyD88 inhibitor (20 μmol/L), with untreated cells as a control. Messenger RNA (mRNA) of cytokines was detected by qRT-PCR. M-PGN induced most TNF-α (A); however, CP-PGN boosted most IL-10 (B). After pretreatment of RAW264.7 cells with ST2825 for 3h, different stimulations listed above were added for another 3h. mRNA of TNF-α(C) and IL-10(D) was significantly inhibited by ST2825, especially for IL-10. All data are expressed as mean ± s.d. in triplicate determinations. Statistical evaluation was performed by one-sided, two-sample with equal variance t-tests. *p < 0.05, **p < 0.01, ***p < 0.001.

Figure 6.

Small interfering RNA-mediated silencing of MyD88 significantly decreases the CP-PGN-triggered secretion of cytokines in RAW264.7. RAW264.7 cells were transfected with si-MyD88 for knockdown of MyD88 expression and done with si-N.C as WT control. CP-PGN (50 μg/mL) stimulated the transfected cells for 24 h to detect MyD88 protein by western blot and cytokines by ELISA. (a) CP-PGN induced no significant changes of MyD88 protein compared with unstimulated cells when it was already at a very low level by si-MyD88. (b) Knockdown of MyD88 protein significantly decreased the secretion of IL-10 (b) and TNF-α (c) induced by CP-PGN, especially for IL-10 depression. Western blot shown is a representative image of triplicate experiments. All data are expressed as mean ± s.d. in triplicate determinations. Statistical evaluation was performed by one-sided, two-sample with equal variance t-tests. *p < 0.05, **p < 0.01, ***p < 0.001.0.01, ***p < 0.001.

Discussion

The MyD88 adaptor protein is essential and indispensable for activating TLR2 signal pathway in response to microbial ligands.^23,24 In the present study, we also demonstrated that lack of MyD88 could significantly block the cytokines secretions of macrophages by CP-PGN, especially for the anti-inflammatory IL-10 (Figure 5 and 6). These results demonstrate that CP-PGN alleviated inflammatory damage by MRSA infection through TLR2 pathway in vivo in the previous study. However, peptidoglycan of S. aureus, as a natural ligand of TLR2, has been playing the role of pro-inflammatory signal to promote some inflammatory cytokines, aggravating the damage caused by infection. In addition, it was found that pretreating BALB/c mice by peptidoglycan of MRSA (M-PGN) could aggravate MRSA-induced infection to cause more deaths than none-pretreatment control. In contrast, pretreatment of CP-PGN improved the survival of MRSA-infected mice (see Figure S1).

CP-PGN and M-PGN are extracted from two Gram-positive (G+) bacteria, an adjuvant bacterium, and a pathogenic bacterium, respectively, which display opposite regulations in process of infection by TLR2-dependent way. Thus, it is essential to investigate the underlying regulation of MyD88 expression in TLR2 downstream by these two peptidoglycans. In the present study, compared with M-PGN and chemosynthetic Pam3CSK4 of two model ligands of TLR2, only CP-PGN could reverse the overexpression of MyD88 in macrophages infected by MRSA pathogen or Abelson leukemia virus. Such moderation of MyD88 expression by CP-PGN was a typical time- and dose-dependent manner when MyD88 has been at a high level as a result of infection.

For cytokines, CP-PGN induced more anti-inflammatory IL-10²⁵ and less pro-inflammatory TNF-α ²⁶ than M-PGN, which favored its anti-infection results in vivo. Blocking activity and interfering expression of MyD88 extremely reduced IL-10 secretion in macrophages by CP-PGN, which was consistent with present cognition and knowledge.^27,28 For present bacterial peptidoglycans, it is rare for CP-PGN to promote more IL-10 by moderating MyD88 expression.^29,30

Till now, the whole bacterial PGN has always been isolated by extraction, including peptidoglycan preparations that are commercially available. Maybe, these extracted PGN contain some contaminants from cell wall components, which also function as PAMPs recognized by PRRs, like lipoteichoic acid (LTA) of a TLR2 agonist in G+ bacteria.^31,32 In the present study, residual LTA and other known PAMPs contaminants in CP-PGN could not be detected by Fourier spectrum analysis, because it had been ruled out as soon as possible by heating bacteria in TCA and washing repeatedly using PBS. For CP-PGN, the present data cannot identify which reactive fragment of peptidoglycan determines the diversity of immune regulation of TLR2-MyD88 signaling. Thus, it is essential to establish the precise spatial structure model of CP-GPN in the future, which can reveal its underlying mechanism of moderating MyD88.

Conclusion

In the present study, available data suggest that CP-GPN could inhibit overexpression of MyD88 by infection with significant promotion of anti-inflammatory IL-10. These immune regulations of CP-PGN are not very common among TLR2 agonists. More underlying mechanism research is needed to expand the knowledge of TLRs network. This specific immunoregulation of CP-PGN is important for us to understand the essential roles of bacterial peptidoglycans and MyD88 for maintaining balance between pro- and anti-inflammatory due to infection.

Supplemental Material

sj-docx-1-eji-10.1177_1721727X221095378 – Supplemental material for Corynebacterium pyruviciproducens-peptidoglycan: A novel bacterial peptidoglycan inhibiting overexpression of MyD88 in macrophages

Supplemental material, sj-docx-1-eji-10.1177_1721727X221095378 for Corynebacterium pyruviciproducens-peptidoglycan: A novel bacterial peptidoglycan inhibiting overexpression of MyD88 in macrophages by Lin Wang, Yuan Ma, Jinfang Shi, Yang Zhang, Jia Tong and Qingzhen Han in European Journal of Inflammation