Identification of a novel negative role of flagellin in regulating IL-10 production

Flagellin-stimulated peritoneal elicited macrophages produce IL-12 but not IL-10

Peritoneal elicited macrophages (PEM) were first treated in vitro with increasing concentrations of purified flagellin or with LPS for 24 h. Supernatants were collected and production of IL-12p40/p70 and IL-10 were determined by ELISA. Reminiscent of previous studies 5, LPS-treated PEM produce both the pro-inflammatory cytokine IL-12 as well as the inhibitory cytokine IL-10 in a dose-dependent manner (Fig. 1A). Treatment of PEM with flagellin also resulted in a dose-dependent production of IL-12p40/p70 (Fig. 1B). The magnitude of this response was, however, not as potent as the one induced by LPS (Fig. 1A versus 1B, black squares).

IL-10 is commonly produced by macrophages upon stimulation with different TLR ligands, but to our surprise we did not detect any IL-10 production in flagellin-stimulated PEM (Fig. 1B, open squares). The lack of IL-10 production was not the result of delayed production of this cytokine, since kinetic studies showed that even after 72 h of treatment IL-10 was still not detected in the supernatants of flagellin-treated PEM (Fig. 1C, open squares). In contrast, when PEM were stimulated with LPS, IL-10 was detected at early time points and its levels remained elevated throughout the duration of the incubation period (Fig. 1C, open diamonds).

One concern in the interpretation of the results described above is the possibility that flagellin preparations could be contaminated with LPS. However several controls have ruled out this possibility. First, using the Limulus amebocyte lysate test, no LPS contamination was detected in our flagellin preparation. Second, digestion of the flagellin preparation with proteinase K resulted in loss of its ability to induce IL-12 production in PEM (Fig. 1B, FD; Fig. 2A), indicating that this effect is induced by a protein and not by other non-proteinaceous potential contaminants in the preparation.

Finally, given the previous demonstration that elicitation of inflammatory effects by flagellin requires binding to TLR5 8, 18, we treated the macrophage cell line RAW 264.7, which lacks TLR5 20 (Fig. 2B, RAW) but still expresses TLR4, with flagellin or LPS. While treatment with LPS resulted in IL-12 production by these cells, treatment with flagellin failed to induce such a response (Fig. 2C), indicating that induction of IL-12 by flagellin requires expression of TLR5 in the target cell (Fig. 2B). The absence of IL-12 production in RAW 264.7 cells also provides additional support to our claim that the flagellin preparation used in our experiments was LPS-free, since some LPS contamination might have otherwise resulted in IL-12 production by RAW 264.7 cells. To conclude, these data show that the effects induced by flagellin upon macrophages are not due to LPS contamination and require expression of TLR5 in the responding cells.

IL-12 but not IL-10 is detected in the serum of flagellin-treated mice

To confirm the above results in vitro, we assessed the production of IL-10 and IL-12 in mice that received a single intravenous injection of purified flagellin or LPS. As shown in Fig. 3A, similar levels of IL-12p40/p70 were detected in the serum of LPS- or flagellin-treated mice. However, a completely divergent outcome in IL-10 production was observed following treatment with either TLR ligand. While significant levels of IL-10 were detected in the serum of LPS-treated mice, this cytokine could not be detected in the serum of flagellin-treated mice (Fig. 3B).

IL-10 mRNA dynamics in macrophages treated with flagellin or LPS

To better understand the mechanisms(s) by which flagellin-stimulated PEM, but not LPS-stimulated PEM, are unable to produce IL-10 protein, we next determined the kinetics of IL-10 mRNA in cells treated with either TLR ligand. In response to LPS stimulation, a rapid increase in IL-10 mRNA was observed by 3 h, followed by a peak response at 6 h. This induction is followed by a progressive decline and, after 12 h of exposure to LPS, the IL-10 mRNA levels were back to baseline (Fig. 4A, open squares). In sharp contrast, the magnitude and kinetics of IL-10 mRNA were significantly different in flagellin-treated PEM. After 3 h of incubation, induction of IL-10 mRNA was detected by RT-PCR. This initial response was followed by a rapid decline and by 6 h, IL-10 mRNA levels in flagellin-treated PEM were equivalent to the levels in untreated cells (Fig. 4A, black squares).

In order to expand the above observations further, we generated bone marrow (BM)-derived macrophages and exposed them to either LPS or flagellin. As shown in Fig. 4B, a similar pattern of IL-10 mRNA expression was observed in flagellin-treated BM-derived macrophages. Furthermore, reminiscent of our findings with PEM (Fig. 1B), IL-12p40/p70 but not IL-10 was detected in the supernatants of BM-derived macrophages that were stimulated with flagellin (Fig. 4C, D). In summary, unlike macrophages treated with LPS, flagellin-treated macrophages display a short-lived expression of IL-10 mRNA that seems to be not sufficient enough to result in IL-10 protein production.

Flagellin inhibits IL-10 production induced by specific TLR ligands

During an ongoing infection, flagellin and other TLR ligands expressed in flagellated bacteria are likely to be recognized either simultaneously or sequentially by cells of the immune system displaying TLR. In order to find whether flagellin could influence the production of IL-10 in response to other TLR ligands, we treated PEM with flagellin and the TLR4 ligand (LPS) either simultaneously (Fig. 5A–C) or sequentially (Fig. 5D–F).

As shown in Fig. 5A, no differences in IL-10 mRNA levels were observed in peritoneal macrophages after 2 h of exposure to either LPS alone (white squares) or LPS in the presence of flagellin (F/LPS: black squares). However, unlike stimulation with LPS alone, IL-10 mRNA levels declined rapidly by 4 and 6 h of simultaneous engagement of TLR4 and TLR5. Flagellin-induced IL-10 inhibition is also reflected at the protein level. As displayed in Fig. 5B, a dose-dependent inhibition of IL-10 production was observed when PEM were treated with LPS (1 μg/mL) in the presence of increasing concentrations of flagellin. Inhibition of IL-10 was not observed when the TLR5-deficient cell line RAW 264.7 was treated with LPS and flagellin, indicating that flagellin-induced IL-10 inhibition in response to LPS requires TLR5 expression (Fig. 5C).

A more profound inhibition of IL-10 mRNA expression was observed when PEM were exposed to flagellin (5 μg/mL) prior to LPS stimulation (Fig. 5D). In addition, decreased levels of IL-10 protein were observed in PEM pre-treated with flagellin (5 μg/mL) and then stimulated with increasing concentrations of LPS (Fig. 5E). In the pre-treatment setting, flagellin at a dose of 5 μg/mL, which has minimal inhibitory effect when used simultaneously with LPS (Fig. 5B), displayed stronger IL-10-inhibitory properties. Conversely, this inhibitory effect of flagellin on IL-10 production was not observed when TLR5-deficient RAW 264.7 cells were pre-treated with flagellin (Fig. 5F).

To determine whether a similar inhibitory effect on IL-10 production occurs when PEM are exposed to TLR ligands other than LPS, we treated PEM with flagellin followed by stimulation with either zymosan (TLR2/6 ligand) or CpG (TLR9 ligand). As shown in Fig. 6A, treatment of PEM with increasing concentrations of zymosan results in IL-10 production in a dose-dependent manner (open bars). However, when these cells were treated with flagellin and then zymosan, a dramatic decrease in IL-10 production was observed (Fig. 6A, black bars). In contrast, such an inhibitory effect on IL-10 production was not observed when PEM were treated with flagellin and CpG (Fig. 6B). To summarize, our results have revealed a previously unknown effect of flagellin in the regulation of IL-10 production in response to specific TLR ligands.

Flagellin treatment does not affect the stability of LPS-induced IL-10 mRNA

In order to determine whether suppression of LPS-induced IL-10 by flagellin was caused by an increase in IL-10 mRNA degradation, we treated macrophages with actinomycin D, an inhibitor of mRNA synthesis, and followed the changes induced by flagellin in the stability of LPS-induced IL-10 mRNA. Fig. 7A shows that actinomycin D suppresses mRNA synthesis. For this control experiment, cells were treated with LPS plus actinomycin D or LPS plus media for 2 h. Cells were then harvested and the levels of IL-10 mRNA were measured by RT-PCR. Actinomycin D (5 μg/mL) completely blocked the synthesis of new mRNA (LPS+ActD).

In our next set of experiments, PEM were cultured for 3 h with either flagellin or media before LPS treatment. Actinomycin D was added to the cultures 2 h after LPS treatment and cells were harvested at the indicated time points after actinomycin D treatment. As shown in Fig. 7B, an almost identical rate of IL-10 disappearance was observed, suggesting that pre-treatment with flagellin does not affect IL-10 mRNA stability in response to LPS stimulation.

Flagellin stimulation decreases ERK activation by LPS in macrophages

A general outcome of TLR engagement is the activation of NF-κB and mitogen-activated protein kinase (MAPK) signaling pathways 21. Among the MAPK, several studies have linked the extracellular signal-regulated kinases (ERK) with IL-10 regulation by TLR 22–26. In order to determine whether flagellin stimulation would affect LPS-induced ERK phosphorylation, we stimulated flagellin- or media-treated PEM with LPS. Since significant changes in the LPS-induced IL-10 mRNA levels are detected in flagellin-treated macrophages as early as at 2 h (Fig. 5D), we decided to quantify ERK phosphorylation at earlier time points (30 and 60 min). Interestingly, LPS-induced phosphorylated ERK1/2 levels were considerably reduced in flagellin-treated macrophages as compared to media-treated ones (Fig. 8). These results suggest that decreased ERK activation could play a role in flagellin-mediated IL-10 inhibition, since TLR-induced production of this cytokine has been reported to be dependent on ERK signaling.

The transcription factor NF-κB plays a central role in the induction of cytokine production by TLR. NF-κB is sequestered in the cytoplasm by molecules known as inhibitor of NF-κB (IκB). Phosphorylation of IκB results in its degradation and in the release of NF-κB that now can translocate to the nucleus and activate its target genes 2. Macrophages treated with flagellin or LPS alone show increased IκB phosphorylation, suggesting that treatment with these TLR ligands results in NF-κB activation. As seen in Fig. 8, flagellin treatment in combination with LPS did not result in decreased levels of IκB phosphorylation, suggesting that inhibition of LPS-induced IL-10 by flagellin is not caused through reducing NF-κB activation by LPS.

Flagellin enhances IL-12 production in response to specific TLR ligands

In order to determine whether the inhibitory effect of TLR5 ligation on PEM might affect cytokines other than IL-10, we determined the IL-12p40/IL-12p35 mRNA levels in macrophages treated with flagellin and LPS. First, we observed that when PEM were treated with LPS alone for 2 h (Fig. 9A, –/LPS), they display much higher levels of IL-12p40mRNA relative to PEM stimulated with flagellin alone (Fig. 9A, F/–). Surprisingly, pre-treatment with flagellin increased the level of IL-12p40 in response to LPS stimulation at 2 h treatment (Fig. 9A, black bar). This increase is short-lived since at 4 h, IL-12p40 mRNA levels were back to those observed in PEM treated with LPS alone (Fig. 9A, 4 hours). In addition, an increase in IL-12p35 mRNA levels was observed in PEM treated with flagellin and subsequently stimulated with LPS (Fig. 9B, F/LPS: black bars). However, the response duration was longer relative to IL-12p40 mRNA induction, since enhanced levels of IL-12p35 mRNA were still observed after 4 h of exposure to LPS (Fig. 9B, 4 hours).

The biologically active form of the cytokine IL-12 is a heterodimer composed of IL-12p40 and IL-12p35 proteins (IL-12p70) 4. IL-12p40 has been reported to be produced in excess over IL-12p35 which acts as a limiting factor for IL-12p70 production 27. Therefore, an increase in IL-12p35 might result in elevated IL-12p70 protein production that would not be detected using standard ELISA methods that do not discriminate between IL-12p40 and IL-12p70 production. For this reason, we specifically analyzed IL-12p70 production by PEM pre-treated with flagellin in which we detected an increase in IL-12p35 mRNA levels in response to LPS. As shown in Fig. 9C, when PEM are pre-treated with flagellin and then exposed to increased concentrations of LPS (black bars), they produced higher levels of IL-12p70 relative to PEM treated with LPS alone (open bars). In summary, PEM treated with flagellin and then stimulated with LPS display enhanced expression of both IL-12p40 mRNA (Fig. 9A) and IL-12p35 mRNA (Fig. 9B) as well as increased production of IL-12p70 protein (Fig. 9C).

In order to determine whether flagellin-mediated enhancement of IL-12p70 production occurs only in response to LPS stimulation or could also be observed in response to other TLR ligands, PEM were treated with flagellin in combination with either zymosan (TLR2/6 ligand) or CpG (TLR9 ligand). While treatment with flagellin did not enhance IL-12p70 production in response to zymosan, which remained undetectable (data not shown), a significant increase in IL-12p70 production was observed when PEM were treated with flagellin and CpG (Fig. 9D). These results indicate that ligation of TLR5 in macrophages influences the inflammatory response of these cells to TLR4 and TLR9 ligands (LPS and CpG) resulting in enhanced production of the pro-inflammatory mediator IL-12p70.

IL-10 has been shown to act as a powerful inhibitor of IL-12 production 28, 29. Therefore we analyzed whether blocking IL-10 signaling would enhance IL-12p70 production by LPS-treated PEM. As shown in Fig. 9E, culturing PEM with LPS plus increasing concentrations of IL-10-blocking antibody increased IL-12p70 production in a dose-dependent manner. These data indicate that IL-10 repression by flagellin could be, at least in part, responsible for the increase in IL-12p70 production observed in macrophages treated with flagellin in combination with LPS. Nonetheless, other mechanisms must be also playing a role, since IL-12p70 production was also enhanced in macrophages treated with a combination of flagellin and CpG, even when on those cells IL-10 production was not altered by flagellin stimulation.

Mice

Male BALB/c mice (6–8 wk old) were obtained from the National Institutes of Health (Frederick, MD). All experiments involving the use of mice were performed in accordance with protocols approved by the Animal Care and Use Committee of the University of South Florida College of Medicine.

Isolation of peritoneal elicited macrophages and generation of bone marrow-derived macrophages

BALB/c mice were injected intraperitoneally with 1 mL of thioglycollate (DIFCO Laboratories, Detroit, MI). Four days later, PEM were obtained by peritoneal lavage as previously described 42. BM-derived macrophages were differentiated from BM cells by harvesting the hind legs and flushing the BM with RPMI. Red blood cells were lysed using ACK lysis buffer and the remaining cells extensively washed. Differentiation of BM cells into macrophages was achieved by 5-6 days incubation of the cells in complete medium (20% FBS, 100 μM β-mercaptoethanol in DMEM) supplemented with L-929 cell supernatants as a source of M-CSF 43.

TLR ligands

S. typhimurium flagellin was isolated as previously described 44. Protein purity was verified by Coomassie staining and its identity confirmed by Western blot with an anti-flagellin monoclonal antibody (Igen International, Gaithersburg, MA). Endotoxin removal was accomplished by combination of Detoxi-Gel AffinityPak columns (Pierce, Rockford, IL) and filtration through 100-kD pore size centricon columns (Millipore, Billerica, MA). Endotoxin removal was confirmed using the Limulus amebocyte lysate test (Cambrex, Rutherford, NJ). In those experiments in which flagellin was digested with proteinase K (Sigma-Aldrich, St. Louis, MO) 100 μg/mL proteinase K was used, digestion carried for 4 h at 37°C, and the protease inactivated at 100°C for 1 h. Experiments were also performed with commercially available S. typhimurium flagellin (InVivoGen, San Diego, CA), which yielded identical results. LPS was obtained from Sigma-Aldrich). CpG and zymosan were from InVivoGen and used as indicated in the text.

Real-time RT-PCR analysis

Two million PEM or BM-derived macrophages were plated per well in a 24-well plate. After 2 h, non-adherent cells were washed off with media and attached cells were then treated as indicated. Total RNA was extracted using TriZol reagent (Qiagen, Valencia, CA), and cDNA obtained with the iScript cDNA synthesis kit (Bio-Rad, Hercules, CA). Target mRNA was quantified using MyIQ single-color real-time PCR detection system (Bio-Rad) and iQ SYBR green Supermix (Bio-Rad).

IL-12p35 primers (left oligo ATGGTGAAGACGGCCAGAG, right oligo CAGGTCTTCAATGTGCTGGTT), IL-12p40 primers (left oligo GCAACGTTGGAAAGGAAAGA, right oligo AAAGCCAACCAAGCAGAAGA), IL-10 primers (left oligo CAGGGATCTTAGCTAACGGAAA, right oligo GCTCAGTGAATAAATAGAATGGGAAC), TLR5 primers (left oligo GCATAGCCTGAGCCTGTTTC, right oligo AAGTTCCGGGGAATCTGTTT) and GAPDH primers (left oligo ATGGCCTTCCGTGTTCCTAC, right oligo CAGATGCCTGCTTCACCAC) were used for PCR amplification. Cycling parameters were 3 min 95°C, 15 s 95°C, 30 s 60°C, 40 repetitions, 1 min 95°C. Single-product amplification was confirmed by melting curve analysis and primer efficiency was near to 100% in all the experiments performed. Quantification is expressed in arbitrary units and target mRNA levels were normalized to GAPDH expression. Actinomycin D was purchased from Sigma-Aldrich.

Quantification of cytokines

Mice were treated with flagellin (10 μg/animal), LPS (10 μg/animal) or HBSS (0.1 mL) via tail vein injection. Ninety minutes later animals were sacrificed and blood obtained by heart puncture. Serum was obtained and the levels of IL-12 and IL-10 were determined by ELISA. For in vitro determination of cytokine production, 1×10⁵ PEM were plated by triplicate in 96-well plates and treated for 24 h unless otherwise specified. Supernatants were harvested and kept at –70°C until ELISA for IL-10, IL-12p40/p70 (BD Pharmingen, San Diego, CA) or IL-12p70 (eBioscience, San Diego, CA) was performed following manufacturers’ instructions. Data represent mean ± SD of triplicate cultures. In the experiments in which IL-10 signaling was blocked we used clone JES5-16E3 rat monoclonal antibody (BD Pharmingen). Statistic analysis was performed using Stat view 4.0.

Cell lysates and Western blot

Total cell lysates were prepared in RIPA buffer (150 mM NaCl, 10 mM Tris-HCl pH 7.4, 5 mM EDTA pH 8.0, 0.1% SDS, 1% deoxycholate, 1% Triton X-100) containing protease inhibitor cocktail (Sigma Aldrich) and phosphatase inhibitor cocktail I and II (Sigma Aldrich) and subjected to SDS-PAGE-immunoblot analysis with anti-phospho-p44/42 MAPK (Thr202/Tyr204) polyclonal antibody (Cell Signaling), phospho-IkappaBα (Ser32) (14D4) rabbit monoclonal antibody (Cell Signaling) and anti-MAPK (ERK1+ERK2) monoclonal antibody (Invitrogen).