ABSTRACT
Recently, many Gram-positive bacteria as well as Gram-negative bacteria have been reported to produce membrane vesicles (MVs), but little is known regarding the regulators involved in MV formation. We found that a Gram-positive anaerobic pathogen, Clostridium perfringens, produces MVs predominantly containing membrane proteins and cell wall components. These MVs stimulated proinflammatory cytokine production in mouse macrophage-like cells. We suggested that MVs induced interleukin-6 production through the Toll-like receptor 2 (TLR2) signaling pathway. Thus, the MV could have a role in the bacterium-host interaction and bacterial infection pathogenesis. Moreover, we found that the sporulation master regulator gene spo0A was required for vesiculogenesis. A conserved, phosphorylated aspartate residue of Spo0A was indispensable for MV production, suggesting that the phosphorylation of Spo0A triggers MV production. Multiple orphan sensor kinases necessary for sporulation were also required to maximize MV production. These findings imply that C. perfringens actively produces immunoactive MVs in response to the environment changing, as recognized by membrane-spanning sensor kinases and by modulating the phosphorylation level of Spo0A.
INTRODUCTION
Many microorganisms produce 20- to 500-nm spherical structures called membrane vesicles (MVs) (1). MVs contain nucleic acids, proteins, membrane components, and quorum-sensing signals and thus function as transporters of multiple components (2, 3). MVs are abundantly found in the natural environment, facilitating interaction between cells and their growing environment (4). In many pathogens, MVs act as vehicles of bacterial antigens, such as lipopolysaccharide (LPS), peptidoglycan (PG), lipoproteins, and virulence-associated factors (such as toxins). Therefore, MVs have an important role in bacterial cell-to-cell communication and host-microbe interactions (5–8).
MVs contain microorganism-associated molecular patterns (MAMPs), such as bacterial DNA, RNA, LPS, and PG, and can manipulate the host immune response (9). MVs induce inflammation and pathology in a range of host tissues. Recently, these responses were shown to involve host pattern recognition receptors (PRRs). Salmonella enterica serovar Typhimurium and Bacteroides fragilis MVs are sensed by Toll-like receptors (TLRs) and modulate the immune response (10, 11). MVs from Pseudomonas aeruginosa, Neisseria gonorrhoeae, and Helicobacter pylori induce proinflammatory cytokine and chemokine production via nucleotide-binding oligomerization domain-containing protein 1 (NOD1) (12). Therefore, MVs have the capability of stimulating PRRs and modulating the host immune response, suggesting that MVs are closely involved in chronic and acute infection by pathogenic bacteria and colonization of human indigenous bacteria. Moreover, MVs have been a potential vaccine candidate, and a mechanistic understanding of MV immunogenicity and formation in various organisms is needed (9, 13).
Functional analysis of MVs has mainly proceeded with Gram-negative bacteria, in which the MVs are derived from the outer membrane of the bacteria. Previous studies have revealed that MV release is conserved in Gram-negative cells and that MVs are associated with virulence factor delivery, protein secretion, cell-cell communication, protection from phage infection, and other biological processes (2, 8). Although Gram-positive bacteria have no outer membrane and are covered by a rigid, thick cell wall, MV production was observed in many Gram-positive bacteria, such as Bacillus subtilis, Staphylococcus aureus, Listeria monocytogenes, Streptococcus mutans, S. pneumoniae, Mycobacteria spp., and Lactobacillus rhamnosus (14–21). In addition, Gram-positive bacterial MVs have been reported to deliver virulence factors to host cells to stimulate an immune response in recipient cells and to facilitate biofilm formation by extracellular DNA release via the MVs (17, 22, 23), suggesting that functional MV release is also conserved in Gram-positive bacteria. In Gram-negative bacteria, MV production is stimulated by accumulation of the unnecessary proteins in the periplasm, charge repulsion, and the SOS response (8, 24). Only a few factors involved in MV production have been reported in Gram-positive bacteria. The biosurfactant surfactin and serum albumin disrupt Gram-positive bacterial MVs (14, 25), suggesting that the amount and the local concentration of MVs are controlled by the bacterial and host environments. In Mycobacterium tuberculosis, virR mutant strains overproduce MVs; thus, VirR proteins inhibit vesiculogenesis (26). Mutation of srtA, which encodes the enzyme involved in anchoring cell surface proteins, affects protein composition in MVs of S. mutans but not the amount of MVs (17). However, the regulatory factor activating MV production and release is not known in Gram-positive bacteria.
Clostridium perfringens is a Gram-positive, anaerobic, spore-forming bacterium, and it causes food poisoning, gas gangrene, and antibiotic-associated diarrhea (27, 28). This bacterium adapts and survives under various stress conditions by sporulation and biofilm formation and is widely spread in the environment, including in the soil and in animal intestines (29–32). Recently, it has been found that type A strains of C. perfringens produce MVs (33). We also have succeeded in the purification of MVs produced by C. perfringens type A strain, and we have confirmed that the MVs contain proteins and nucleic acids (34). However, the role of the MVs in pathogenesis of C. perfringens remains unclear, and factors involved in MV production of C. perfringens are completely unknown. In this study, we report that C. perfringens MVs activate the host innate immune response. It is suggested that MVs induce release of the proinflammatory cytokine interleukin-6 (IL-6) via the Toll-like receptor 2 (TLR2) signaling pathway. Moreover, a sporulation master regulator, Spo0A, is required for MV production, although the sporulation-specific sigma factor is not involved in vesiculogenesis. Moreover, multiple sensor kinases are involved in MV production. spo0A mutation reduced the production of MVs, indicating that C. perfringens actively produces immunoactive MVs in response to the environment.
RESULTS
C. perfringens produces MVs at the late growth phase.We have observed MV production in C. perfringens strain 13 (34). C. perfringens was cultured in BHI medium because the cells produced the most MVs in this medium, although MV production was detected in the other media, PGY and TY-G1 (data not shown). From transmission electron microscopy (TEM) observation, C. perfringens started to release MVs into the culture supernatant at 6 h of incubation (the early stationary phase) (Fig. 1A). However, the MV production was at very low levels, and the proteins associated with the MV were below the detection limit under these conditions (Fig. 1B and C). MVs were produced more as the culture times were prolonged (Fig. 1A, B, and C). These MVs contained proteins because the BCA protein assay and SDS-PAGE detected proteins in the purified MVs (Fig. 1B and C). Moreover, purification of MVs by density gradient ultracentrifugation and TEM observation also suggested that MVs associated with these proteins (see Fig. S1 in the supplemental material). Because MVs produced by pathogens often contain virulence factors, we focused on a primary toxin of C. perfringens, phospholipase C (PLC), by Western blotting. We detected PLC in the culture supernatant but not in the MV fraction (Fig. S2). It was consistent with the previous study by Jiang et al., which also demonstrated that MVs isolated from C. perfringens type A strains were not associated with PLC (33). In addition, we found that the major proteins in the MV fraction were predicted to be membrane-associated proteins or lipoproteins (Fig. S2A). The MV fractions at all time points contained more peptidoglycans, a major cell wall component, than the membrane fractions, indicating that the C. perfringens MVs predominantly contain cell envelope and cell wall (Fig. 1D).
C. perfringens produces membrane vesicles containing proteins and peptidoglycans. (A) TEM observation of MVs produced by C. perfringens. MVs were purified from the culture supernatant of C. perfringens strain 13 grown in BHI medium at 37°C for 6 or 24 h. Bars indicate 500 nm. (B) Protein profiles in MVs. Proteins in MVs purified from the cultures grown for 6, 12, 18, or 24 h were separated by 12.5% SDS-PAGE and visualized by CBB staining. Equivalent amounts of culture (10 ml) were loaded in each lane. (C) Quantification of proteins in MVs. The proteins were quantified by BCA protein assay. (D) Quantification of peptidoglycan in MVs. Equivalent MV and membrane fractions of protein (300 ng) were used for peptidoglycan quantification using the SLP reagent kit. The values represent the means and standard deviations from an average of at least three independent experiments. Asterisks indicate statistical significance (P < 0.05), as evaluated by Student's t test.
C. perfringens MVs induce an inflammatory cytokine response in mouse macrophage-like cells.It was reported that MVs produced by various bacteria manipulate the host immune response and induce the production of proinflammatory cytokines. We added C. perfringens MVs to J774.1 mouse macrophage-like cells to assess the MV effect on mammalian cells. Morphology of the cells was changed in the presence of MVs but not in the presence of an equal amount of bovine serum albumin (BSA), suggesting that MVs interact with and influence mammalian cells (Fig. S3). To evaluate the ability of MVs to induce inflammatory cytokine expression, we measured cytokine expression in J774.1 culture medium containing MVs. Representative inflammatory cytokines IL-6 and tumor necrosis factor alpha (TNF-α) accumulated in the cultures treated with MVs in a dose-dependent manner but not in cultures treated with BSA (Fig. 2A and B). We also observed increased levels of IL-6 and TNF-α mRNAs by quantitative PCR (qPCR) (Fig. S4). We showed that the C. perfringens MVs predominantly contained cell envelope and cell wall (Fig. 1B and C). Thus, components of the cell membrane and/or the cell wall in MVs might trigger the proinflammatory cytokine response. When the membrane protein fraction was also used for cytokine enzyme-linked immunosorbent assay (ELISA), induction of IL-6 production by the membrane fraction was less than that in the MV fraction (Fig. 2C). On the other hand, TNF-α expression was induced at comparable levels in both the MV and membrane fractions (Fig. 2D). Thus, it was suggested that IL-6 induction is an MV-specific property and that the components of the membrane fraction, which the MVs also contain, were responsible for TNF-α induction.
C. perfringens MVs induce proinflammatory cytokine IL-6 and TNF-α production in J774.1 macrophage cells. J774.1 mouse macrophage-like cells were incubated with 3 to 300 ng/ml of MVs or BSA (A and B), 300 ng/ml of MVs or the membrane fraction (C and D), or 3 to 300 ng/ml of MVs or 3 to 300 ng/ml of peptidoglycans purified from C. perfringens cells for 12 h. The amount of IL-6 (A, C, and E) and TNF-α (B, D, and F) in the supernatants of the cell culture was quantified by cytokine ELISA. Means and standard deviations from three replicates are shown. Asterisks indicate statistical significance (*, P < 0.05; **, P < 0.01; ***, P < 0.001), as evaluated by one-way analysis of variance (ANOVA) with the Dunnett's post hoc test.
TLR2 signaling is involved in the effect of MV on cytokine response.The C. perfringens MVs contained more peptidoglycan, a major cell wall component, than the membrane fraction (Fig. 1D). Cell wall components such as peptidoglycan, lipoteichoic acid, and lipoprotein are known to be TLR2 ligands and stimulate the host immune response through a TLR2-dependent signaling pathway. Western blotting using an anti-TLR2 antibody and qPCR showed that the expression of TLR2 increased after incubation with MVs compared to expression in the presence of BSA (Fig. 3A and B; Fig. S4). MVs increase TLR2 mRNA levels but not TLR4 (Fig. S4). To evaluate the involvement of TLR2 in the inflammatory cytokine response, the induction of cytokine production by MVs was measured in the presence of the TLR2/4 inhibitor OxPAPC (oxidation of 1-palmitoyl-2-arachidonyl-sn-glycero-3-phosphorylcholine), and the TLR1/2 ligand Pam3CSK4 was used as a positive control. IL-6 induction in the presence of MVs or Pam3CSK4 was inhibited by the addition of OxPAPC, suggesting that IL-6 release is mediated by TLR2 (Fig. 3C). However, TNF-α expression in the presence of MVs was not affected by OxPAPC, although OxPAPC partially inhibited expression induction by Pam3CSK4 (Fig. 3D). These results suggested that MVs stimulate IL-6 production in macrophage-like cells via a TLR2 signaling pathway.
TLR2 signaling pathway is involved in the induction of the proinflammatory cytokine response by MVs. (A) Western blotting for TLR2 in J774.1 mouse macrophage cells treated with MVs. J774.1 mouse macrophage-like cells were incubated with 3 to 300 ng/ml of MV or BSA for 12 h. An equivalent amount of proteins extracted from 6.25 × 103 cells was loaded in each lane. TLR2 proteins were detected using an anti-TLR2 antibody, and GAPDH proteins were detected using an anti-GAPDH antibody that served as a loading control. (B) qPCR for tlr2 mRNA in J774.1 mouse macrophage cells treated with MVs. J774.1 mouse macrophage-like cells were incubated with 30 or 300 ng/ml of MV or BSA for 2 h. The tlr2 mRNA levels relative to nontreated cells (0 ng/ml) were shown. (C and D) Inhibitory effects of the TLR2/4 inhibitor on MV-induced cytokine production. IL-6 (B) and TNF-α (C) production in the presence of TLR2/4 inhibitors was quantified by cytokine ELISA. Mouse macrophage-like cells were incubated with 300 ng/ml MVs in the presence of the TLR2/4 inhibitor OxPAPC. The TLR1/2 ligand Pam3CSK4 was used as a positive control. Means and standard deviations from three replicates are shown. Asterisks indicate statistical significance (*, P < 0.05; ***, P < 0.001), as evaluated by one-way ANOVA with the Tukey-Kramer post hoc test.
We investigated whether the peptidoglycan that MVs predominantly contain is a trigger of IL-6 production via TLR2 signaling. We used peptidoglycans purified from C. perfringens cells for cytokine ELISA (Fig. 2E and F). These peptidoglycans never induced cytokine production. Lipoprotein and lipoteichoic acid, as well as peptidoglycan, differentially activate TLR2 (35). Therefore, cell wall components other than peptidoglycan associated with MVs might cause the induction of cytokine production.
The sporulation factor Spo0A is required for MV production. C. perfringens is a spore-forming bacterium, and sporulation is highly regulated by various regulatory factors and proteins (36, 37). The Spo0A protein is a primary regulatory protein of sporulation. We have shown that Spo0A has a pivotal role in the life cycle of the sporulating bacterium and in biofilm formation of this organism (29). We constructed an spo0A mutant by antibiotic resistance gene replacement and in-frame deletion in C. perfringens strains 13 and HN13, respectively. The HN13 strain is the galKT in-frame deletion mutant of strain 13 and is used for construction of in-frame deletion mutants with galactose counterselection (38). Because MV production in both strains is not significantly different, these strains were used as the wild type (data not shown). MV production was significantly reduced in the spo0A in-frame deletion mutant of HN13 (Fig. 4A). In the spo0A mutant of C. perfringens strain 13, the MV protein amount was also reduced, and complementation of the spo0A gene restored MV production (Fig. 4B and C). MV proteins were not detected by SDS-PAGE when culture of the spo0A mutant was prolonged to 48 h (Fig. 4C). Transmission electron microscopy (TEM) observation also showed that there were no MVs in the culture supernatant of the spo0A mutant (data not shown). C. perfringens strain 13 and its derivative, HN13, that we used in this study are poorly sporulating strains, so we also tested MV production of the sporulation-proficient strain SM101. The spo0A mutant of SM101 also showed decreased MV production (Fig. S5). Thus, spo0A could have a central role in the regulation of vesiculogenesis in C. perfringens.
Sporulation factor is involved in vesiculogenesis in C. perfringens. (A) MV production in C. perfringens HN13 and the Δspo0A and ΔsigF isogenic mutant strains. Each strain was cultured at 37°C for 24 h in BHI medium. The amount of MV production was quantified by BCA protein measurement. The values represent the means and standard deviations from an average of at least three independent experiments. Asterisks indicate statistical significance (P < 0.01), as evaluated by one-way ANOVA with the Bennett post hoc test. (B) MV protein profiles in strain 13 and an spo0A mutant harboring an empty vector or an spo0A complement vector were analyzed by SDS-PAGE followed by CBB staining. Equivalent amounts of culture (10 ml) were loaded in each lane. (C) MV production in strain 13 and an spo0A mutant harboring an empty vector or an spo0A complement vector. Each strain was cultured at 37°C for 24 h in BHI medium. The amount of MV production was quantified by BCA protein measurement.
Spo0A is a master regulator of sporulation, and the sporulation cascade might be involved in MV production by C. perfringens. To test this hypothesis, we constructed a sigF mutant of HN13 and analyzed MV production in this mutant. sigF codes for σF, which is a sporulation-specific sigma factor required for sporulation. The sigF mutant did not form the endospore (Fig. 5A) (39). However, the sigF mutant produced a comparable level of MVs (Fig. 4A). These results indicated that the spo0A activity, not the sporulation process, is important for MV production.
Deficiency of multiple orphan sensor kinases is involved in sporulation and vesiculogenesis. (A) Sporulation in orphan sensor kinase mutants. Endospore formation in each strain was observed by phase-contrast microscopy. Cells precultured in FTG medium were cultured in DS medium at 37°C for 24 h. Bright signals represent endospores. Bars, 2 μm. (B) Measurement of viable cells (gray bars) and heat-resistant spores (white bars) formed by the orphan sensor kinase mutants. The sporulation frequencies are indicated at the top of bars. Cells precultured in FTG medium were cultured in DS medium at 37°C for 24 h and heated at 75°C for 15 min. Survival of heat-resistant cells was measured by CFU counting. The values represent the means and standard deviations from an average of at least three independent experiments. Values below the detection limit of the CFU counting (10 spores ml−1) are indicated by N.D. Asterisks indicate statistical significance compared to HN13 (P < 0.05), as evaluated by Student's t test. (C) MV production in HN13 and isogenic mutants of orphan sensor kinases. Each strain was cultured at 37°C for 24 h in BHI medium. The amount of MV production was quantified by BCA protein measurement. The values represent the means and standard deviations from an average of at least three independent experiments. Asterisks indicate statistical significance (**, P < 0.01; ***, P < 0.001), as evaluated by one-way ANOVA with the Bennett post hoc test.
Conserved phosphorylated residue in Spo0A is required for MV production.Spo0A is a response regulator and is activated by phosphorylation of the aspartate residue (40). The aspartate residue D55 of B. subtilis Spo0A is conserved in C. perfringens Spo0A (D58). To determine whether the conserved aspartate residue is required for Spo0A activity and MV production, we introduced a point mutation (D58A) in the spo0A sequence of the complementation vector. We confirmed comparable expression levels of spo0A and spo0A D58A mRNA by Northern blot analysis (Fig. S6A). Introduction of the Spo0A D58A expression vector did not restore the MV production of the spo0A mutant (Fig. 4C). Thus, the conserved aspartate residue in spo0A is required for MV production, suggesting that MV production is regulated by phosphorylation of the Spo0A protein in C. perfringens. LIVE/DEAD staining and CFU numbers were not significantly different between the wild type and the spo0A mutant (Fig. S6B and C). These results suggested that the MVs are not the resultant cell debris of passive cell death and C. perfringens actively produces MVs.
Deficiency in orphan sensor kinases affects MV production.The phosphorylation of Spo0A is conducted by multiple orphan sensor kinases in some Clostridium spp. (41–43). We hypothesized that multiple orphan sensor kinases phosphorylate the Spo0A protein in response to the environment and subsequently are involved in MV production in C. perfringens. We searched the orphan sensor kinase genes in the C. perfringens strain 13 genome and constructed an in-frame deletion mutant of the putative orphan sensor kinase genes CPE0207, CPE0986, CPE1316, and reeS. The CPE0207, CPE0986, and CPE1316 genes have unknown functions, and reeS is reported to impact extracellular enzyme production (44). To evaluate whether these orphan sensor kinases are involved in sporulation, the mutants were cultured in DS medium. It was noted that C. perfringens strain 13, used in this study, is known to be a poorly sporulating strain and that the sporulation frequency of its derivative, HN13, serving as the wild type in this test, was 4.1% ± 1.3% under our experimental conditions (Fig. 5A). Phase-contrast microscopy showed that no endospores were observed in the culture of CPE1316 or reeS deletion mutants (Fig. 5A). In addition, the number of heat-resistant spores was reduced in the CPE1316 and reeS mutants (Fig. 5B). In contrast, the CPE0207 mutant sporulated at about the same frequency as the wild type. In addition, the sporulation frequency of the CPE0986 mutant was slightly but significantly increased compared to that of the wild type (4.1% ± 1.3% versus 6.1% ± 0.1% for the CPE0986 mutant; P < 0.05) (Fig. 5B). Therefore, CPE1316 and ReeS were required for efficient sporulation and CPE0986 inhibited sporulation, suggesting that these orphan sensor kinases are involved in Spo0A activity. The deletion of these orphan sensor kinase genes also impacted MV production (Fig. 5C). Mutants of CPE1316 or reeS and a double mutant of CPE1316 and reeS significantly decreased MV production (Fig. 5C). Thus, multiple orphan sensor kinases are involved in MV production in C. perfringens, suggesting that environmental signals stimulate the release of MVs via an orphan kinase-Spo0A signaling pathway.
DISCUSSION
We found that MVs produced by a Gram-positive pathogen, C. perfringens, induced proinflammatory cytokine production in host cells and that MV production was regulated by a response regulator and multiple sensor kinases. C. perfringens MVs accumulated at 20 h after inoculation in culture supernatant (Fig. 1A and B). The observed microspherical structures seem to be products from lysed bacteria. However, LIVE/DEAD staining and CFU numbers at 24 h of culture were not significantly different between the wild type and the Δspo0A vesiculation mutant (see Fig. S6B and C in the supplemental material). We therefore suggest that passive cell death in the late growth phase did not result in MV production. Moreover, deficiency in the sensor orphan kinases CPE1316 and ReeS significantly decreased the amount of MVs in culture supernatants (Fig. 5C). These results suggest that genetic regulation and environmental signals affect MV production and that C. perfringens actively produces MVs. Nutrient depletion has been implicated as the signal triggering the initiation of the sporulation process, which is regulated by Spo0A activity (36); thus, it might be a trigger for vesiculogenesis. The finding that MV production is induced at the stationary phase supports this notion. MVs serve as nutrients that may inhibit sporulation and affect cell fate decisions at the late growth phase (4).
In Gram-negative bacteria, endolysin and the ABC transporter system positively affect MV production via different molecular mechanisms (45, 46). How single-membrane, Gram-positive bacteria that have a thick peptidoglycan cell wall produce MVs is still unknown. Several hypothetical mechanisms have been proposed, because cell wall-modifying and/or penetrating proteins may enable MV formation through a thick cell wall (47). The sporulation factor and sensor kinases could regulate genes related to cell wall modification.
In Gram-positive bacteria, regulatory elements of vesiculogenesis are poorly understood. We found that Spo0A positively affects vesiculogenesis in C. perfringens type A strains. Spo0A is known as a master regulator for endospore formation in C. perfringens (48), suggesting that the sporulation pathway is involved in MV production. However, the MV formation of the sigF mutant, which does not sporulate, was comparable to that of the wild type (WT) (Fig. 4A). A point mutation on the phosphorylating aspartate residue in the Spo0A amino acid sequence resulted in deficiency of MV production. These results indicated that the function as a transcriptional factor of Spo0A was required for MV production. The spo0A gene is conserved in sporulating Firmicutes. Thus, conserved Spo0A may be important for the regulation of vesiculogenesis in sporulating Gram-positive bacteria. Recently, it was reported that in the nonsporulating Gram-positive bacterium Streptococcus pyogenes, the two-component system CovRS downregulates MV formation (49). Given the results of our study, which indicates the phosphorylated/activated response regulator is required for MV production, signal transduction from sensor kinase to response regulator or two-component system could be a key regulatory system of vesiculogenesis in both sporulating and nonsporulating Gram-positive bacteria.
The phosphorylation of the conserved aspartate residue is necessary for the activity of the response regulators. The so-called phosphorelay mediated by several orphan kinases and phosphotransfer proteins results in the activation of Spo0A in B. subtilis (40). Environmental factors such as nutrient limitation trigger autophosphorylation at the aspartate residue in the orphan kinases. These orphan kinases interact with Spo0F, transferring the phosphoryl group. Phosphorylated Spo0F interacts with Spo0B. Spo0B transfers the phosphate to Spo0A. This is the so-called phosphorelay, which is an initial step of sporulation. In contrast, orphan histidine kinases would directly phosphorylate Spo0A in Clostridia (50). Indeed, we could not find the homologues of phosphorelay proteins, such as Spo0B and Spo0F, in the C. perfringens strain 13 genome sequence. CPE0986, CPE1316, and ReeS show high sequence homology to Cac0903, Cac0323, and Cac0437, which are orphan sensor kinases involved in the phosphorylation of Spo0A and sporulation in Clostridium acetobutylicum (e values in BLASTP are less than 10−54) (41). Our results indicated that several orphan kinases, such as CPE0986, CPE1316, and ReeS, affect not only vesiculogenesis but also sporulation frequency. This implies that CPE0986, CPE1316, and ReeS directly interact with Spo0A and control the phosphorylation.
C. perfringens MVs stimulated TNF-α and IL-6 production by J774.1 and RAW264 mouse macrophage-like cells (Fig. 2 and data not shown). The induction of IL-6 cytokine release by the MV could be mediated by TLR2 signaling. TLR2 is believed to recognize peptidoglycans and subsequently to induce IL-6 expression (51, 52), and TLR2 expression increases in response to peptidoglycan (53). MVs contain much higher peptidoglycan levels than membrane fractions and are predominantly associated with cell wall components (Fig. 1C). We showed MVs induced more IL-6 expression in J774.1 cells than the membrane fraction (Fig. 2C). However, peptidoglycans purified from C. perfringens cells did not induce IL-6 production. This indicates PG is not responsible for immunoactivity of the MVs. Mycobacterium tuberculosis MVs predominantly contain lipoproteins that have immunogenic activity (19). Jiang et al. reported that MVs produced by C. perfringens type A strain CP4 contains 12 putative lipoproteins (33). We also identified 2 putative lipoproteins (CPE1030 and MalE) in the MV fraction, and these proteins accumulated more in MV than in the membrane fraction (Fig. S2). Collectively, we suggest that cell wall components other than peptidoglycans, such as lipoprotein, in MVs induce proinflammatory cytokine expression via the TLR2 signaling pathway. However, OxPAPC inhibited the induction of TNF-α by Pam3CSK4 but not by MVs, raising the possibility that C. perfringens MVs contain other components that stimulate alternative proinflammatory cytokine response pathways.
C. perfringens MVs produced under our experimental conditions did not contain PLC toxin (Fig. S2) and did not show cytotoxicity in mouse macrophage-like cells and epithelial cells (data not shown). Thus, immunomodulatory activity is suggested to be a primary function of the MVs on pathogenesis and environmental adaptation of C. perfringens. MVs might directly interact with macrophages in vivo and facilitate systemic toxicity by inducing the cytokines in C. perfringens myonecrosis (gas gangrene). PLC toxin-induced occlusion of vessels by platelets and neutrophils creates the anaerobic condition, causing rapid progression and tissue destruction (54). During the infection, TNF-α and IL-6 induced by MVs would attract macrophages, platelets, and neutrophils. Furthermore, these cytokines contribute to the development of shock and multiorgan failure (55). MVs are highly stable under the different conditions and environments (56, 57). While C. perfringens soluble toxins like PLC also have immunogenicity, stable immunoactive MV could support the infection process (4, 55). Alternatively, a prolonged stimulation of the TLR2 signaling pathway by MVs could inhibit antigen presentation and facilitate C. perfringens host colonization (47).
In the present study, we showed that MVs impact the host immune response through a specific signaling pathway. Therefore, immunoactive MVs would have an important role in the pathogenesis of the interaction between C. perfringens and host cells. MV release is a conserved phenomenon not only in Gram-negative bacteria but also in Gram-positive bacteria, because our present study and recent studies found that functional MV production occurs in many Gram-positive bacteria (14–22, 33, 34). We found that immunoactive MV production in C. perfringens is genetically regulated by a sporulation master regulator, Spo0A, and multiple orphan sensor kinases. Thus, C. perfringens could actively produce MVs in response to external environments. Spo0A controls not only sporulation, which is an important survival strategy for obligate anaerobe C. perfringens, but also vesiculogenesis, involving inflammatory responses of host cells in the infectious process. Complementation of the spo0A gene by transformation of spo0A-encoding multicopy plasmids increased MV production (Fig. 4B and C). Excess amounts and/or activity of the Spo0A protein might lead to overproduction of MVs, demonstrating that the spo0A gene is a useful target for controlling production of MVs. However, the detailed physiological functions of MVs and the mechanisms of vesiculogenesis remain uncertain, and further investigations are warranted to understand MVs produced by the thick-cell-walled, single-membrane, Gram-positive bacteria.
MATERIALS AND METHODS
Bacterial strains and culture conditions.Bacterial strains and plasmids used in this study are listed in Table S1 in the supplemental material. C. perfringens was cultured in Gifu anaerobic medium (GAM) (Nissui Co., Japan) or brain heart infusion medium supplemented with 1 mg/ml sodium thioglycolate (BHI) under anaerobic conditions using an Anaeropack (Mitsubishi Gas Chemical Co., Inc., Tokyo, Japan). Strains harboring vectors were cultured in the presence of 20 μg/ml chloramphenicol. Escherichia coli strain DH5α was grown in Luria-Bertani medium supplemented with 20 μg/ml chloramphenicol.
MV production and purification. C. perfringens strains were cultured in 45 or 400 ml of BHI medium at 37°C for 24 h. MVs were isolated and purified as previously described (58). Briefly, culture supernatants were separated by centrifugation (6,000 × g, 20 min) and filtered through 0.45-μm-pore-sized polyvinylidene difluoride (PVDF) filter membranes (Millipore, USA). The filtered supernatants were ultracentrifuged (100,000 × g, 1 h), and pellets were washed with 10 mM HEPES (pH 6.8). MVs were quantified using the bicinchoninic acid (BCA) protein assay kit (Thermo Fisher Scientific). MVs were purified as follows. Crude MVs were suspended in 400 μl of 45% (wt/vol) OptiPrep in 10 mM HEPES containing 0.85% NaCl (OptiPrep-HEPES-NaCl) and layered with 40%, 35%, 30%, 25%, 20%, 15%, or 10% OptiPrep-HEPES-NaCl (40 to 15%, 400 μl each; 10%, 200 μl). After ultracentrifugation (100,000 × g, 3 h), 250-μl aliquots of the fractions were collected from the top of each ultracentrifuge tube. For addition to cultured cells, fractions containing MVs were ultracentrifuged again, and pellets containing purified MVs were washed with 10 mM HEPES (pH 6.8) and resuspended in 10 mM HEPES (pH 6.8).
Transmission electron microscopy (TEM).Fractionated or purified MVs were analyzed using an H-7650 transmission electron microscope (Hitachi). The MVs were allowed to adhere to carbon-coated grids for 1 min and were negatively stained with 2% uranyl acetate.
Peptidoglycan quantification.Peptidoglycans in MVs and membrane fractions were measured using an SLP-HS single-reagent set (Wako Chemicals, USA) according to the manufacturer's instructions, with some modifications. Briefly, 100-μl samples containing 300 ng protein were mixed with an equal volume of SLP-diluent-reconstituted SLP reagent in a 96-well plate. The plate was then incubated at 37°C, and the optical density was measured at 650 nm at 5-min intervals for 3 h. The concentration of peptidoglycans was calculated using a standard curve.
Cell culture.J774.1 mouse macrophage-like cells were grown in RPMI 1640 medium supplemented with 10% fetal calf serum, 100 U/ml of penicillin, and 0.1 mg/ml streptomycin at 37°C in a humidified atmosphere containing 5% CO2. The cells were plated in 24-well plates at a concentration of 2.5 × 105 cells per well and incubated with 3, 30, or 300 ng/ml MVs or bovine serum albumin (BSA) for 12 h. The TLR1/2 agonist Pam3CSK4 (InvivoGen; San Diego, CA) and the TLR2/4 inhibitor OxPAPC (InvivoGen; San Diego, CA) were added at a final concentration of 300 ng/ml or 60 μg/ml, respectively, if necessary.
Cytokine ELISA.After incubation with MVs, supernatant from the cell cultures was harvested and used for detection of IL-6 and TNF-α by enzyme-linked immunosorbent assay (ELISA). A 96-well, half-area polystyrene immunoplate (Costar) was coated with anti-mouse IL-6 (2 μg/ml) or TNF-α (1 μg/ml) capture antibody (PeproTech, Rocky Hill, NJ) suspended in Na2CO3-NaHCO3 buffer (pH 9.6) containing 0.2% NaN3 and was incubated overnight at 4°C. After washing with phosphate-buffered saline (PBS) containing 0.05% Tween 20 (PBS-T), the plates were blocked with 75 μl of 1% skim milk in PBS-T for 2 h at 37°C. A 50-μl sample supernatant diluted 1:20 in 0.5% skim milk-containing PBS-T was added. After washing, the plates were incubated with anti-mouse IL-6 (0.5 μg/ml) or TNF-α (0.25 μg/ml) detection antibody (PeproTech) suspended in PBS-T containing 0.5% skim milk for 2 h at 37°C. The plate was incubated with avidin-horseradish peroxidase (HRP) conjugates (PeproTech) diluted 1:2,000 in PBS-T containing 0.5% skim milk for 30 min at 37°C. Bound antibodies were detected using 2,2′-azinobis(3-ethylbenzthiazolinesulfonic acid) (ABTS) peroxidase substrate (KPL, Gaithersburg, MD). The absorbance was measured at 405 nm with a reference wavelength of 620 nm.
Western blotting.The cells in each well were lysed by adding 200 μl of 1× SDS sample buffer (2% SDS, 50 mM Tris-HCl [pH 6.8], 10% glycerol, 2.5% 2-mercaptoethanol, 0.1% bromophenol blue) and were boiled for 5 min at 95°C. Equal amounts of each protein sample then were separated by 10% acrylamide SDS-PAGE. Subsequently, proteins were transferred onto Immobilon PVDF membranes (Millipore, Bedford, MA) and blocked with 2.5% skim milk in PBS-T for 1 h at room temperature. The membranes were probed with anti-TLR2 antibody (Wako Chemical) diluted 1:400 in Can Get Signal (Toyobo, Japan) overnight at 4°C or with anti-glyceraldehyde-3-phosphate dehydrogenase (GAPDH) antibody (Wako Chemical) diluted 1:15,000 in PBS-T for 1 h at room temperature. HRP-labeled anti-goat secondary antibody (Sigma) and ImmunoStar LD (Wako Chemical) were used to detect the antibodies.
RNA isolation and qPCR.Total RNA was isolated from J774.1 cells incubated with MVs for 2 h by using RNeasy columns (Qiagen, Valencia, CA). For reverse transcription, we used ReverTra Ace qPCR RT master mix with a genomic DNA remover kit (Toyobo, Osaka, Japan). We quantified il-6, tnfa, il-12a, tlr2, tlr4, actb, and gapdh mRNA by using a TaqMan universal PCR master mix (AmpErase UNG) (Thermo Fisher Scientific Inc., Waltham, MA, USA) in accordance with the manufacturer's instructions. All PCRs were performed and analyzed on a StepOnePlus real-time PCR system (Thermo Fisher Scientific Inc.) using Mm00446190_m1, Mm00443258_m1, Mm00434169_m1, Mm00442346_m1, Mm00445273_m1, Mm00607939_s1, and Mm99999915_g1 TaqMan gene expression assays (6-carboxyfluorescein dye-labeled MGB probes) for il-6, tnfa, il-12a, tlr2, tlr4, actb, and gapdh genes, respectively. Relative quantification (RQ) was calculated using the comparative threshold cycle method (2−ΔΔCT). The thermal profile for PCR was 2 min at 50°C and 10 min at 95°C, followed by 40 cycles of 95°C for 15 s and 60°C for 1 min. All reactions were performed in triplicate.
Peptidoglycan isolation.Peptidoglycans were isolated as described previously (59). Cells grown in 45 ml of BHI medium were collected by centrifugation, washed with 0.8% NaCl, boiled in hot 4% SDS for 30 min, and incubated at room temperature overnight. The suspension was then boiled for 10 min. The SDS-insoluble cell wall materials were collected by centrifugation. The pellet was washed four times with water and finally resuspended in 1 ml sterile water. Remaining SDS was removed by dialysis in a 1-kDa-molecular-size-cutoff FlexTube (IBI Scientific, Peosta, IA) against 2 liters of 10 mM HEPES (pH 6.8) at 4°C overnight.
RNA isolation and Northern blotting.Total RNA was isolated from C. perfringens and used for Northern blotting as previously described (60). Template DNAs for generation of digoxigenin (DIG)-labeled DNA probes were amplified using the primers listed in Table S2.
Construction of mutant strains.In-frame deletion mutants were constructed as previously described (29, 38). Upstream and downstream regions of target genes were amplified using the primers listed in Table S2 and cloned into the SalI/BamHI site of pCM-GALK. The resulting plasmids were introduced into C. perfringens HN13 by electroporation. Constructed plasmids and mutants were confirmed by DNA sequencing and PCR.
Sporulation culture.Cells were precultured in FTG medium (Difco) for 16 h at 37°C. The precultures were diluted 1:100 in DS sporulation medium (39) and incubated for 48 h at 37°C. The cultures containing mature spores were heated at 75°C for 15 min to kill the vegetative cells. To count heat-resistant spores, the cultures then were plated on GAM plates.
ACKNOWLEDGMENTS
We are grateful to H. Nariya for kindly providing strain HN13 and pCM-GALK. We thank Michiyo Kataoka and Noriko Saito for their technical assistance with electron microscopy analysis. The matrix-assisted laser desorption ionization–time of flight mass spectrometry analysis was conducted using a TOF/TOF 5800 system at the Chemical Analysis Division and Open Facility, Research Facility Center for Science and Technology, University of Tsukuba.
This work was financially supported through the Exploratory Research for Advanced Technology (ERATO) of Japan Science and Technology Agency (JST). This work was also supported by the Japan Agency for Medical Research and Development (AMED) (40105500).
FOOTNOTES
- Received 14 February 2017.
- Accepted 16 February 2017.
- Accepted manuscript posted online 21 February 2017.
Supplemental material for this article may be found at https://doi.org/10.1128/IAI.00096-17 .
- Copyright © 2017 American Society for Microbiology.
