IL-6 is a pleiotropic cytokine implicated in the pathogenesis of disorders such as sepsis and cancer. We noted that human monocytes are excellent producers of IL-6 as compared with monocyte-derived macrophages. Because macrophages from molecule containing ankyrin repeats induced by LPS (MAIL) knockout animals have suppressed IL-6 production, we hypothesized that regulation of MAIL is key to IL-6 production in humans and may explain the differences between human monocytes and macrophages. To test this hypothesis fresh human monocytes and monocyte-derived macrophages were compared for MAIL expression in response to LPS. LPS-induced monocyte MAIL expression was highly inducible and transient. Importantly for our hypothesis MAIL protein expression was suppressed during differentiation of monocytes to macrophages. Of note, the human MAIL protein detected was the 80 kDa MAIL-L form and human MAIL showed nuclear localization. Human MAIL-L bound to p50 subunit of the NF-κB and increased IL-6 luciferase promoter activity in a cEBPβ, NF-κB, and AP-1-dependent fashion. Like the differences in MAIL induction, monocytes produced 6-fold more IL-6 compared with macrophages (81.7 ± 29.7 vs 12.6 ± 6.8 ng/ml). Furthermore, suppression of MAIL by small interfering RNA decreased the production of IL-6 significantly in both THP-1 cells and in primary monocytes. Costimulation of monocytes with LPS and muramyl dipeptide induced an enhanced IL-6 response that was suppressed by siMAIL. Our data suggests that MAIL is a key regulator of IL-6 production in human monocytes and plays an important role in both TLR and NOD-like receptor ligand induced inflammation.
Molecule containing ankyrin repeats induced by LPS (MAIL),2 also called IL-1-inducible nuclear ankyrin repeat protein (INAP) or IκBζ, is a newly described homolog of IκB proteins (1, 2, 3). It is induced by TLR agonists and other proinflammatory cytokines but not by TNF-α (2, 4, 5). Unlike the other IκB proteins, MAIL has a long N-terminal region that anchors a nuclear localization sequence and a transactivation domain. The transactivation domain of MAIL actively regulates the transcriptional activation of many genes unlike the other IκB proteins that are known for their role in gene suppression (6, 7, 8, 9). In contrast, the C-terminal region of MAIL harbors six ankyrin repeats that are thought to suppress NF-κB activity (2, 10, 11). It has been demonstrated that mouse MAIL binds to the p50 subunit of NF-κB and regulates transcriptional activation (12). Extrapolating from the MAIL knockout gene expression profiles, MAIL appears to regulate the production of many proinflammatory cytokines and innate defense proteins. For example, macrophages from MAIL knockout mice are deficient in the production IL-6 upon LPS stimulation (12). Furthermore, MAIL regulates the expression of the antimicrobial defense proteins human β defensin 2 (hBD2) and neutrophil gelatinase associated lipocalin (NGAL) in epithelial cells upon IL-1β or IL-17 stimulation (13, 14).
Although the role of MAIL in regulating the expression of IL-6 is clearly demonstrated in mouse, the role of MAIL in human monocyte IL-6 production has not been reported. It has been hypothesized that differences in the N-terminal region between human and mouse MAIL may lead to the expression of distinct genes (11). Also, it has been reported that mouse MAIL and human MAIL differ in their ability to bind NF-κB proteins. Mouse MAIL binds only to p50 whereas the human MAIL binds both p65 and p50 (11, 12, 15). Differences also exist in TNF-α-mediated MAIL induction in humans and mice. In contrast to human MAIL, which has been reported to be inducible by TNF-α, mouse MAIL is not (11). Since IL-6 plays an important role in many inflammatory diseases such as sepsis, heart attacks, stroke, and in many human cancers including hepatocarcinoma, multiple carcinoma, and ovarian cancer, we wanted to elucidate the function of human MAIL in IL-6 production.
In this study, we show that human MAIL is comparable in function to mouse MAIL for the production of LPS-induced IL-6. We also show that macrophages, unlike their precursor monocytes, have less MAIL and therefore have a reduced ability to produce IL-6 in response to LPS. Furthermore, we demonstrate that intracellular NOD2 ligand (muramyl dipeptide) synergizes with LPS for the production of IL-6, and this response is dependent on the expression of MAIL.
Materials and Methods
Reagents and antibodies
LPS (Escherichia coli strain 0127:B8, Westphal preparation) and E. coli strain 0111:B4 was purchased from Difco and Alexis, respectively. Ala-γ-d-Glu-diaminopimelic acid (γ-iE-DAP) and Ala-α-dE. coli.
Cell culture and transfection
Human PBMCs were isolated by Histopaque density gradients from fresh source leukocytes from the American Red Cross. Monocytes were isolated from PBMC by CD14 positive selection (Miltenyi Biotec). In brief, blood was layered on Histopaque-1077 (Sigma-Aldrich) and spun at 600 × g for 20 min at room temperature. The mononuclear layer was collected and washed three times with RPMI 1640. Monocytes were purified from PBMCs using positive selection with anti-CD14-coated magnetic beads following the manufacturer’s recommendations (Miltenyi Biotec). This method of purification yields greater than 98% pure monocytes based on flow cytometry analysis (16). Monocytes (2 × 106/ml) were plated in RPMI 1640 supplemented with 5% FBS, 100 U/ml penicillin, and 100 μg/ml streptomycin. For culturing monocytes derived macrophages, PBMCs (2 × 106/ml) were plated in Teflon containers and cultured in RPMI 1640 supplemented with 20% human AB serum for 5 days. Five-day-old macrophages were purified by CD14 positive selection. Purified macrophages (2 × 106 6) were nucleofected with 5 μg of EGFP, EGFP-MAIL-S, or EGFP-MAIL-L plasmids. Cells were stimulated with medium alone, LPS (1.0 μg/ml) MDP (5.0 μg/ml), α-iE-DAP (5.0 μg/ml) or LPS plus MDP. Cells culture supernatants and cell lysates were harvested 4 h after stimulation.
17). The sequences of primers are available upon request.
Generation of MAIL Ab
Human MAIL-S was cloned in pET28b with both N- and C-terminal histidine tags. This construct was overexpressed in BL21-DE3 RIL strain of bacteria. The overexpressed protein was purified by using Ni-NTA column (Qiagen). Purified human MAIL-S was confirmed by liquid chromatography tandem mass spectrometry analysis (Campus Chemical Instrumentation Centre, Ohio State University). This recombinant protein was used to immunize rabbits. To confirm the specificity of the Ab, recombinant protein at two different concentrations and HEK293 cells transfected with EGFP alone or with EGFP-MAIL were used to probe with MAIL Ab (data not shown).
Preparation of cell lysates and Western blotting
Cells were lysed using lysis buffer (50 mM Tris, 125 mM NaCl, 1% Triton X100, 10 mM EDTA, 10 mM sodium fluoride, and 10 mM sodium pyrophosphate) with 0.3 mM sodium orthovanadate, protease inhibitor mixture (Sigma-Aldrich), 0.2 mM methoxysuccinyl-Ala-Ala-Pro-Val-chloromethyl ketone, and 2 mM PMSF (Sigma-Aldrich). Cell lysates were incubated on ice for 15 min. Cell lysates were centrifuged at 14,000 × g for 10 min and the supernatants was transferred to a new tube. Protein was quantified by Lowry’s assay (Bio-Rad) and equal protein amounts (10–50 μg) per lane loaded onto a SDS-PAGE. The separated proteins were transferred to nitrocellulose, and the membranes were blocked in 5% nonfat milk. The membranes were blotted with appropriate Ab overnight, and then blotted with appropriate secondary Ab followed by ECL substrate (GE Healthcare).
g for 5 min and washed 3 times (1 ml of RIPA buffer). The proteins were eluted with 2× SDS-PAGE sample buffer at 95°C for 10 min. Eluted proteins were separated by SDS-PAGE and blotted with respective Abs.
Preparation of nuclear and cytoplasmic fractions
For fractionation of THP-1 cells into nuclear and cytoplasmic fractions, THP-1 cells (10 × 106) were washed with 10 ml of PBS. Cells were centrifuged at 1500 × g for 5 min at 4°C. Cell pellet was resuspended in 1 ml of PBS and centrifuged again for 15 s at 4°C at maximum speed. Cell pellet was resuspended in buffer A (10 mM HEPES (pH 7.9), 10 mM KCl, 0.1 mM EDTA, 0.1 mM EGTA supplemented with 1 mM DTT, 0.5 mM PMSF, and 10 μg/ml each aprotinin, leupeptin, and pepstatin) and allowed to swell on ice for 15 min; 200 μl of 10% Nonidet P-40 was added to the swollen cells, and the samples were vortexed vigorously for 10 s. Cells were centrifuged at 14,000 × g. The supernatant resulting from this centrifugation was the cytosolic fraction, and the pellet was the nuclear fraction. The cytosolic fraction was diluted in buffer B (10 mM Tris (pH 7.5), 7 M urea, 1% SDS, 0.3 M sodium acetate, 20 mM EDTA). The nuclear pellet was resuspended in buffer C (20 mM HEPES (pH 7.9), 0.4 M NaCl, 1 mM EDTA, 1 mM EGTA supplemented with protease inhibitors as in buffer A) and shaken vigorously at 4°C for 2 h. The samples were centrifuged at 14,000 × g to get the nuclear fraction.
IL-6 promoter assay
The 651-bp region containing different transcription factor binding sites located immediately upstream of IL-6 start site and the mutant plasmids were a gift from Dr. Oliver Eickelberg (Giessen University School of Medicine, Giessen, Germany). The 651-bp region was subcloned into the (5′ KpnI and 3′ XhoI) pGL3 basic luciferase reporter gene vector. Mutations were made by site-directed mutagenesis of the sites AP-1 (positions −283 to −276 5′-TGAGTCAC-3′ was changed to 5′-TGCAGCAC-3′), C/EBPβ (−154 to −146 5′-TTGCACAAT-3′ was changed to 5′-CCGTTCAAT-3′) and NF-κB (−72 to −63 5′-GGGATTTTCC-3′ was changed to 5′-CTCATTTTCC-3′). These mutations have been shown to have decreased promoter activity compared with the wild-type promoter in various cells when stimulated with different ligands (18, 19, 20). HEK293 TLR4/IL-1R1/MD-2 cells were cotransfected with pGL3 basic (Promega), pIL-6 luc 651, mut NF-κB pIL-6 luc 651, mut cEBPβ pIL-6 luc 651, or mut AP-1 pIL-6 luc 651 along with MAIL-L pCDNA, MAIL-S-pCMV2A, or pCDNA 3.1. pEGFP-C2 was also used as a transfection control. Cells were harvested 48 h after transfection cells and lysed in passive lysis buffer (Promega). Lysates were used to analyze luciferase activity using a luminometer. We observed that overexpression of either MAIL-S or MAIL-L increased the activity of renilla luciferase and β-galactosidase and were thus unable to normalize our data based on these. Therefore, the transfection efficiency was normalized by Western blotting the lysates with EGFP Ab.
Small interfering RNA (siRNA)
Monocytes (10 × 106) were nucleofected with 80 μM control or MAIL siRNA (Dharmacon) using the Amaxa Nucleofector kit. The sequences for MAIL siRNA are GCACUUCACAUGCUGGAUA, GAGAACAGAUCCGACGUAU, GUAAUCAGUUUGUGGAUCU, and GCACAUCCGAAGUCAUAAA. After nucleofection, the cells were suspended in the Amaxa kit medium, supplemented with l-glutamine and 25 ng/ml M-CSF. Nucleofected cells were plated in 12-well plates at a concentration of 2.5 × 106 cells per ml for 18 h. After 18 h, cells were stimulated with LPS (1.0 μg/ml) or other NOD ligands for an additional 6 h. For THP-1 cells, 80 μM siRNA was used to transfect 1.5 × 106
Values are expressed as mean ± SEM. For simple comparisons, a standard paired t test was used, and for the multiple comparisons in the luciferase experiment, one-way ANOVA with Tukey’s post hoc analysis was used. Significance was defined as a value of p < 0.05.
MAIL is transiently induced in monocytes upon LPS stimulation
To determine the kinetic pattern of MAIL mRNA expression, monocytes were treated with LPS (1.0 μg/ml) for various time points for a period of 24 h. Cells were lysed for mRNA analysis (by real-time PCR) and protein analysis (by immunoblotting). Freshly isolated monocytes had low levels of MAIL mRNA. However, treatment with LPS induced MAIL mRNA by 15 min and the expression peaked at 2 h (Fig. 1⇓A). After 2 h, there was a gradual decrease in MAIL mRNA, which was almost undetectable by 24 h. To analyze the expression kinetics of MAIL protein, we generated a rabbit polyclonal Ab against MAIL-S. This Ab detects both EGFP-tagged MAIL-L and MAIL-S and also Flag-tagged MAIL-S and Myc/His-tagged MAIL-L (supplemental Fig. 1, A and B, respectively).3 To analyze whether MAIL mRNA data matched MAIL protein data, we analyzed the cell lysates by immunoblotting with the Ab for MAIL and actin. MAIL protein was absent in fresh monocytes, but the protein expression rapidly increased upon LPS stimulation, peaking at 4 h (Fig. 1⇓B). There was only minimally detectable protein at 24 h after LPS stimulation. Since MAIL exists in two spliced forms, MAIL-L and MAIL-S, we assayed the ∼80-kDa form with a MAIL-L-specific Ab (supplementary Fig. 1C). Its recognition indicated that MAIL-L is the predominant form seen in human monocytes as previously reported in murine cells (2, 10, 24).
As it has been previously shown that MAIL regulates IL-6 production in mouse (1, 10, 12), we hypothesized that human MAIL might also regulate IL-6 production in monocytes. So, we analyzed the mRNA expression of MAIL and IL-6 in monocytes at time points indicated upon LPS stimulation. As shown in Fig. 1⇑A, expression of MAIL mRNA was very early, peaking at 2 h after LPS stimulation and then gradually decreasing after 2 h. In contrast to MAIL, expression of IL-6 mRNA was much later, peaking at 8 h after LPS stimulation (Fig. 1⇑A). Thus MAIL, as an early inducible gene, has the appropriate kinetics to serve as a regulator of IL-6 in human monocytes upon LPS stimulation.
Monocytes express more MAIL and secrete more IL-6 compared with macrophages
Monocytes and macrophages are important components of the host innate response. As we have previously shown that monocytes and macrophages differ in their ability to process IL-1β (25, 26, 27), we wanted to analyze whether monocytes and macrophages differed in the expression of MAIL and therefore in the production of IL-6. We isolated monocytes from blood and compared the LPS induced production of MAIL and IL-6 to in vitro matured macrophages. Consistent with the IL-1β differences, macrophages expressed less MAIL (Fig. 2⇓A) and also produced less IL-6. Monocytes produced 7-fold more IL-6 compared with macrophages (81.7 ± 29.7 vs 12.6 ± 6.8 ng/ml) (Fig. 2⇓B). In contrast, TNF-α production was not different between monocytes and macrophages (Fig. 2⇓C). Furthermore, the expression of IL-8 did not differ between monocytes and macrophages (data not shown). This indicates that in vitro matured macrophages have less MAIL and less IL-6 compared with monocytes upon LPS stimulation.
MAIL is induced by IL-1β and LPS and not by TNF-α in monocytes
It is generally agreed that MAIL expression is regulated by proinflammatory stimuli such as TLR ligands and IL-1β; however, TNF-α induced expression of MAIL is controversial. Muta and colleagues have documented that murine MAIL is not induced by TNF-α (24, 28). However, recent work by Totzke et al. (11) indicated that MAIL was induced by TNF-α in HeLaD98 cells. To test for TNF-α induced MAIL expression, we examined MAIL expression in TNF-α stimulated monocytes (Fig. 3⇓). In accordance with the previously published data in mice (2, 12, 24, 28, 29), MAIL expression in human monocytes was readily induced by LPS and only minimally so by IL-1β. However, in contrast to LPS or IL-1β, TNF-α had no detectable influence on MAIL induction. Furthermore, IL-1β in combination with TNF-α added little to the MAIL expression from IL-1β alone. Pro-IL-1β expression was analyzed in the same cells and paralleled the MAIL expression pattern. These data are consistent with the prior observations that monocytes are very sensitive to LPS but less so to IL-1β since they have low expression of IL-1 receptors (30).
Of note, although TNF-α was a poor inducer of both MAIL and proIL-1β, TNF-α did induce the production of IL-8, indicating that monocytes were responsive to TNF-α. However, TNF-α-treated monocytes did not produce IL-6. Furthermore, monocytes response to IL-1β and to IL-1β/TNF-α combination was diminished when compared with LPS (Fig. 3⇑). From this experiment, we can conclude that in human monocytes MAIL is induced by TLR and IL-1 receptor ligands but not by TNF-α.
MAIL translocates to nucleus and forms a distinct speckled pattern
Because mouse MAIL has been shown to localize to the nucleus, we hypothesized that human MAIL should also be a nuclear protein. To determine the cellular localization of MAIL, HEK293/TLR4/IL-1R1/MD-2 cells were transfected with EGFP-MAIL-L, EGFP-MAIL-S, or EGFP plasmids, and the cells were examined by fluorescent microscopy. In the samples transfected with EGFP alone, GFP was localized throughout the cell, whereas in the samples transfected with EGFP-MAIL-L, the GFP was seen only in the nucleus (Fig. 4⇓). We also observed that MAIL-S was a nuclear protein and had a characteristic speckled pattern of staining as previously described (supplemental Fig. 2) (3, 10, 11). To further prove that MAIL-L is a nuclear protein, we fractionated THP-1 cells into nuclear and cytosolic fractions. The nuclear and cytosolic fractions were blotted for MAIL, pro-IL-1β, and lamin-B1. We used pro-IL-1β as a cytosolic marker and lamin B1 as a nuclear marker. In agreement with the GFP localization data, both MAIL-L and lamin B1 were detected in the nuclear fraction (Fig. 4⇓B). As expected, the cytosolic protein pro-IL-1β was detected in the cytoplasmic fraction. These experiments indicate that MAIL is a nuclear protein.
MAIL binds to p50 subunit of NF-κB
It has been reported by many groups that mouse MAIL binds to the p50 subunit of the NF-κB and is involved in transcriptional regulation of IL-6 gene. However, recent observation by Totzke et al. indicated that there may be a difference in function between human and mouse form of MAIL. They showed that human MAIL can bind both p65 and p50 subunit of NF-κB. Based on this observation they speculated that human MAIL may differ in certain functions and may induce different set of genes upon binding to both p65 and p50 compared with mouse MAIL which binds only p50. Based on the finding reported by Totzke et al., we were interested in analyzing the binding of MAIL to NF-κB subunit. We cotransfected HEK293 TLR4/IL-1R1/MD-2 cells with EGFP-MAIL-L, EGFP MAIL-S or EGFP-ASC (a control protein for EGFP expression) in combination with p50-FLAG or ASC FLAG (a control protein for FLAG expression). We immunoprecipitated with both Flag (p50) and MAIL Abs and immunoblotted with p50 and MAIL Abs. As shown in Fig. 5⇓A, both forms of MAIL (MAIL-L and MAIL-S) bind to the p50 subunit of NF-κB. This binding of MAIL to NF-κB is seen only during stimulation with IL-1β. To confirm the findings by Totzke et al., that human MAIL binds to both p50 and p65 subunit of NF-κB, we immunoprecipitated human MAIL with p65 and vice versa. However, we could not confirm the binding of p65 to MAIL in HEK293 cells expressing TLR4/IL-1R1/MD-2 even upon stimulation with IL-1β (data not shown). These differences may be attributed to differences in cell line used and other steps involved in immunoprecipitation. Nevertheless, our study confirms earlier studies done in mice and humans indicating that MAIL binds to the p50 subunit of NF-κB (10, 12, 15).
MAIL increases IL-6 promoter activity
To further analyze the mechanism of MAIL induced IL-6 production we used an IL-6 promoter luciferase reporter assay. HEK293 TLR4/IL-1R1/MD-2 cells were transfected with the pGL3 IL-6 reporter construct along with pEGFP-C2 (as transfection control), and either MAIL-L pCDNA, MAIL-S pCMV2A, or pCDNA alone. Cells were subjected to luciferase analysis after IL-1β or control stimulation. Results were referenced to the pIL-6 reporter defining the pCDNA control vector as 100 U. Cells transfected with pIL-6 alone showed a 2-fold increase in luciferase activity after IL-1β stimulation (100 ± 0.02 vs 195 ± 6.4, p < 0.05). This activity was further enhanced by transfection of MAIL-L or MAIL-S. Both MAIL-L and MAIL-S transfected cells had ∼3-fold increased luciferase activity compared with vector control transfected cells. Without IL-1β stimulation, MAIL-L and MAIL-S induced increased luciferase compared with control vector (p = 0.045 and 0.05, respectively). When cells were stimulated with IL-1β, MAIL-L and MAIL-S also enhanced reporter activity, p = 0.014 and 0.016 respectively.
As MAIL binds to p50 NF-κB subunit and this complex is thought to regulate gene expression by binding to the NF-κB DNA binding site, we hypothesized that a functional NF-κB binding site is necessary for MAIL action. As shown in Fig. 5⇑C, MAIL-L and MAIL-S induced luciferase activity is dramatically reduced in IL-6 reporter containing a mutated NF-κB site (p = 0.05), suggesting that a functional NF-κB site is required for MAIL activity.
It has been shown that certain MAIL regulated genes like NGAL/lipocalin and hBD2 require both NF-κB and cEBPβ for their expression (31). To determine whether NF-κB and cEBPβ are both required for MAIL induced IL-6 production, we used an IL-6 reporter lacking a functional cEBPβ binding site. Mutation of cEBPβ binding site in IL-6 promoter decreased IL-6 promoter activity after both MAIL-L and MAIL-S over expression when compared with the wild type cEBPβ containing IL-6 reporter construct (p < 0.05). (Fig. 5⇑C). This indicates that similar to other MAIL induced genes, i.e., NGAL/lipocalin and hBD2, both cEBPβ and the NF-κB binding site are required for full induction of IL-6 by MAIL.
Since AP-1 has been shown to regulate IL-6 expression in response to variety of stimuli (19, 20), we also tested the requirement of AP-1 in MAIL induced IL-6 production. We used the AP-1 mutated IL-6 promoter luciferase reporter construct transfected with vector alone, MAIL-L, or MAIL-S. Surprisingly, we found that AP-1 binding site is also required for MAIL induced IL-6 production. We found that mutations in AP-1 binding site resulted in significant down-regulation of luciferase activity in both MAIL-S and MAIL-L over expressed samples (p < 0.05). The role of AP-1 in MAIL induced IL-6 production has not been reported earlier and needs to be tested further. Taken together, these experiments indicate that MAIL binds to p50 subunit of NF-κB and thus increases the IL-6 reporter activity which requires functional NF-κB, cEBPβ, and AP-1 binding sites in the IL-6 promoter.
Knockdown of MAIL decreases LPS induced IL-6 production in THP-1 cells
As the function of human MAIL in regulating IL-6 production has not yet been studied, we wanted to determine the function of MAIL in monocytes. To do this, the expression of MAIL in THP-1 cells (monocytic cell line) was knocked down using siRNA technology (Fig. 6⇓). THP-1 cells were transfected with MAIL or control siRNA and were allowed to rest for 18 h, after which the cells were stimulated with LPS (1.0 μg/ml) for an additional 3 and 6h. Cell supernatants were harvested and analyzed for IL-6, IL-8 and TNF-α. Cell lysates were blotted for MAIL, IκBα and actin protein (Fig. 6⇓). We also analyzed the expression of MAIL protein by immunoblotting (Fig. 6⇓). MAIL siRNA-treated samples had decreased expression of both MAIL mRNA and protein compared with control siRNA-treated samples. As MAIL is homologous to IκBα, we wanted to check whether the MAIL siRNA affected the expression of IκBα. We found no difference in expression of IκBα mRNA (data not shown) or protein.
Suppression of MAIL was associated with suppression of IL-6 mRNA and protein. IL-6 production was decreased by ∼50% in MAIL siRNA-treated samples compared with control siRNA-treated samples (p = 0.003). To control for the specificity of the siMAIL effect we measured TNF-α and IL-8. Although the MAIL siRNA did slightly suppress the production of both IL-8 mRNA and protein, there was no appreciable change in the production of TNF-α in MAIL siRNA-treated samples compared with the control treated samples. These findings suggest that decreasing MAIL expression suppresses the LPS-induced production of IL-6 in THP-1 cells.
Knockdown of MAIL suppresses LPS induced IL-6 production in monocytes
To verify the data generated in THP-1 cells, we also studied knocking down MAIL expression in human primary monocytes. Human monocytes were nucleofected with control or MAIL siRNA for 18 h. Nucleofected cells were stimulated with LPS for additional 6 h. Cell lysates and cell culture supernatants were harvested and analyzed for expression of MAIL and for the production of IL-6. After treatment of monocytes with siMAIL, MAIL mRNA levels trended lower but did not reach statistical difference. However, there was about a 70–90% suppression of MAIL protein expression in MAIL siRNA-treated samples compared with the control siRNA-treated samples. This may be due to the fact that the 6 h time point was well after peak MAIL mRNA or that the siMAIL effect is predominantly on translation. Upon analyzing the RNA samples we saw a significant decrease in the production of IL-6 mRNA in MAIL siRNA-treated samples compared with the control siRNA-treated samples (p = 0.03) (Fig. 7⇓A). In contrast to the difference in expression of IL-6 mRNA between MAIL siRNA-treated and control siRNA-treated monocytes, we did not see any significant change in IL-8 and TNF-α mRNA expression (Fig. 7⇓A and data not shown). Importantly, suppression of MAIL protein was associated with a decrease in the IL-6 production. MAIL siRNA samples had a 5-fold decrease in IL-6 production compared with control siRNA-treated samples (Fig. 7⇓B). Although MAIL suppression was also linked to a decrease in IL-8 and TNF-α (data not shown), it did not affect the expression of IκBα protein, a homolog of MAIL, supporting that the siRNA to MAIL is target specific.
Knockdown of MAIL suppresses NOD ligand induced IL-6 production
Stimulation of the intracellular NOD like receptors activates the NF-κB pathway which leads to the production of many proinflammatory genes including IL-1β, IL-6 and IL-8 (32, 33, 34, 35). It has also been demonstrated by many different groups that TLR ligands synergize with NOD like receptor ligand for the production of many cytokines like IL-1β, IL-8 and IL-6 (36, 37, 38, 39, 40, 41, 42). Recently a microarray analysis revealed MAIL as one of the genes which was differentially regulated upon stimulation with NOD1 ligand. The authors speculated that MAIL may have a role in the enhancement of TLR signaling by the NLR ligands (32). Since MAIL is a NF-κB dependent protein and is required for the production of IL-6, we hypothesized that suppression of MAIL would also suppress NOD ligand induced IL-6 production. We made use of two NOD ligands namely γ-iE-DAP, a NOD1 agonist and MDP, a NOD2 agonist. Monocytes, treated with MAIL siRNA or the control siRNA, were stimulated with LPS, α-iE-DAP, γ-iE-DAP and MDP individually, or in combination for 6 h (Fig. 8⇓).
As we have seen before, treatment of monocytes with MAIL siRNA decreased the expression of MAIL protein. In agreement with the data shown in Figs. 5⇑ and 6⇑, LPS induced IL-6 production was decreased in MAIL siRNA-treated samples compared with control siRNA-treated samples. In line with previously published data, neither of the NOD ligands stimulated MAIL expression or IL-6 production when they were treated alone or in combination, indicating that they are not potent in stimulating the monocytes (32, 43). Monocytes treated with a combination of α-iE-DAP (a control for γ-iE-DAP which does not activate NOD receptors) and LPS, did not have a synergistic effect on IL-6 production. Interestingly, when monocytes were treated with NOD γ-iE-DAP in combination with LPS, they did not show any synergistic effect on IL-6 production, indicating that monocytes do not respond to NOD1 stimulus. However, LPS in combination with the NOD2 agonist, MDP, did have an enhanced effect on the production of MAIL and IL-6 (p = 0.03). In agreement with the previous data, MAIL siRNA-treated samples had dramatically less IL-6 production compared with the control siRNA-treated samples when they were stimulated with LPS or LPS in combination with MDP (Fig. 7⇑). As shown above, suppression of MAIL did not affect production of IL-8. Furthermore, IL-1β production was similar to IL-8 production and there was a synergistic effect of MDP and LPS (data not shown). This indicates that NOD ligands (NOD2) by themselves are not potent inducers of IL-6 production in monocytes, but they synergize with LPS for the production of IL-6. This synergistic effect on IL-6 production is dependent on MAIL expression. From this we can conclude that MAIL is an important regulator of IL-6 production in response to both extracellular (TLR) and intracellular (NLR) ligands.
MAIL is a newly described homolog of IκB protein family having six ankyrin repeats at the C-terminal region (1, 2, 3). Unlike the other IκB proteins, which are constitutively expressed, MAIL is an inducible gene. From our findings, we show that MAIL is an early gene and is transiently expressed in monocytes upon LPS stimulation. Based on this, we speculate that like IκBα, MAIL may undergo phosphorylation and ubiquitination leading to its degradation by the proteasome. In support of our findings, it has been reported by other groups that MAIL is an inducible gene and the expression of MAIL depends on the NF-κB pathway (5). So far, there are two major alternatively spliced forms of MAIL reported. In monocytes, we find that both MAIL-L and MAIL-S mRNA is expressed (data not shown). However, the predominant protein product detected was MAIL-L. In support of our findings in human monocytes, Kitamura et al. showed by histology and immunoblotting that MAIL-L is the predominant form of MAIL in murine macrophages and B lymphocytes (44).
The presence of ankyrin repeats at the C terminus end of MAIL suggests that, like other IκB proteins, it may be involved in suppression of NF-κB. It has been demonstrated by other groups that the over expression of MAIL can suppress NF-κB activity (2). However there is no direct evidence for the suppression of NF-κB activity by MAIL. In contrast to the suppressive ability, mouse MAIL has been shown to transcriptionally up-regulate IL-6 production. IL-6 is an pluripotent cytokine thought to have both pro- and anti-inflammatory effects (45). Its role in sepsis has been well documented (46). Recently the role of IL-6 in cancer progression and metastasis has been reported by two different groups (47, 48). Many groups have implicated that MAIL might be involved in the production of IL-6 based on the over expression model system. Yamamoto et al. 2004 reported that macrophages from MAIL knockout animal are deficient in LPS induced IL-6 production (12). To our knowledge, we are the first to describe the role of human MAIL in the production of IL-6. We show that MAIL is expressed by monocytes very early upon LPS stimulation, and the production of IL-6 mRNA follows MAIL protein expression. We also show that monocytes express more MAIL and produce more IL-6 upon LPS stimulation than monocyte-derived macrophages. This observation supports the data from two other groups Bauer et al. 1988 and Kotloff et al. 1990 (49, 50). Our data also extends their finding by indicating that the deficiency observed in monocyte derived macrophages for IL-6 production is due to decreased expression of MAIL. Monocyte-derived macrophages express less MAIL upon LPS stimulation and therefore are deficient in the production of IL-6. This may be due to differences in the expression of transcription factors which may regulate the expression of MAIL that together with MAIL induce IL-6. Unlike human macrophages, mouse macrophages have been shown to express MAIL (44). This difference in expression of MAIL in macrophages may be species specific. For example it has been reported by many different groups that human macrophages are different from mouse macrophages in NO production (51, 52, 53, 54, 55).
We also tried maturing THP-1 cells to macrophages using PMA. Unexpectedly, we found that PMA treated in vitro matured THP-1 cells have no difference in MAIL expression compared with THP-1 cells that are not matured (data not shown). Interestingly, in vitro matured THP-1 derived macrophages release more IL-6 and IL-8 compared with naive THP-1 cells. In support of this finding, Takashiba et al. have reported that PMA stimulation primes THP-1 cells for LPS stimulation by increasing the accumulation of cytosolic NF-κB and thereby increasing the secretion of TNF-α (56). It is our conception that at baseline, THP-1’s are premonocytic cells that gain functions with maturation. For example, it has long been recognized that PMA matured THP-1 cells acquire the ability to process and release IL-1β (57, 58, 59), which is a functional characteristic of monocytes but not differentiated macrophages. In this context, we believe that the IL-6 responsiveness of PMA matured THP-1 cells is not strictly equivalent to monocyte derived macrophages. Thus, the IL-6 responsiveness of the PMA treated THP-1 cells is likely more complex than a simple unifactorial response to MAIL up-regulation.
Since we have reported earlier that macrophages are deficient in production of IL-1β and now we show that macrophages also produce less IL-6, it is tempting to speculate that macrophages are quite stable in their reaction to various stimuli when compared with their precursor monocytes, which are exquisitely reactive (21, 25, 26). This idea is supported by the fact that alveolar macrophages (similar to in vitro derived macrophages) may be designed to provide a clearance function without inducing excess inflammation (27). To prove that MAIL maybe one of the factors essential for macrophage IL-6 production, we over expressed MAIL in monocyte derived macrophages to induce macrophage IL-6 production. However, due to the recognized difficulties in transfecting primary macrophages, we were unable to successfully express the gene in macrophages. Therefore, we expressed MAIL in THP-1 cells. Although we were able to induce the artificial expression of both MAIL-S and MAIL-L genes in THP-1 cells by nucleofection, we failed to detect any change in IL-6 production after recombinant MAIL-L or MAIL-S expression (supplemental data Fig. 3). The reason for lack of effect of over expression of MAIL in THP-1 cells is unexplained. It might be due to the fact that sufficient amounts of endogenous MAIL were also induced by the exogenous stimulation thus masking the over expression effect. Furthermore, MAIL may require additional cofactors or posttranslational processing to fully promote IL-6 regulation.
Unlike other IκB proteins, MAIL translocates to nucleus with a speckled pattern (1, 2, 3, 11). We show that like the mouse MAIL, human MAIL-L and MAIL-S are also nuclear proteins. Unlike MAIL-S, MAIL-L had fewer speckles upon transfection. We do not know the cause for this difference in speckling pattern. As reported earlier, this speckled pattern may be due to the interaction of MAIL with nuclear de-acetylases or with transcriptional machinery (11, 60, 61). Our immunoprecipitation data and IL-6 promoter luciferase reporter studies suggest that MAIL binds to p50 subunit of NF-κB and thus increases the IL-6 promoter activity. This MAIL induced IL-6 luciferase activity is dependent on NF-κB, cEBPβ and AP-1. Our siRNA studies from THP-1 cells and primary human monocytes suggest that human MAIL, like its mouse counterpart, is involved in production of IL-6. Additionally, we demonstrate for the first time that MAIL is involved in the production of both TLR and NLR induced IL-6.
Totzke et al. also reported that human MAIL, unlike the mouse MAIL, bound both p50 and p65 subunits of the NF-κB molecules (11). They hypothesized that there may be differences in the regulation of IL-6 between mouse and human MAIL. Based on their finding, we wanted to test whether there were any difference in regulation of IL-6 production between mouse and human MAIL. We observed that MAIL was not induced by TNF-α in primary human monocytes. This observation is not in line with the report by Totzke et al. (11). They reported that human MAIL was induced by TNF-α in HeLaD98 cells and MCF-7 cells. The differences observed between our studies may be the differences in the cells used. We have used primary monocytes, while they have used human cervix and breast carcinoma cells namely HeLaD98 and MCF-7 cells. Although we did not find any discrete differences between human and mouse form of MAIL in IL-6 regulation, we did note two differences between human and mouse MAIL. First, human MAIL binds to p50 subunit of the NF-κB but only upon stimulation. Second, we observed that human MAIL-L did not readily form speckles spontaneously upon over expression unlike mouse MAIL or human MAIL-S. The significance of this observation is for further investigation.
The role of MAIL in the production of innate defense proteins has been well documented. MAIL acts as a transcriptional regulator by binding to p50 of NF-κB for the induction of the antimicrobial peptides, lipocalin and hBD-2 (13, 14). The fact that MAIL knockout animals have atopic like dermatitis and eye inflammation supports a role for MAIL in innate host defense (12, 62, 63). How MAIL deficiency causes inflammation is however not so clear. It will be important to test whether the skin and eye inflammation in MAIL knockout mice is due to the lack of host defense molecules, e.g., lipocalin and/or to lack of essential cytokines, e.g., IL-6. However, the other function of MAIL i.e., suppression of NF-κB activity cannot be excluded. The atopic dermatitis and ocular inflammation may also be due to the over expression of cytokines that MAIL might inhibit via the C-terminal ankyrin region.
In summary, our data indicate that MAIL is a key positive regulator of IL-6 production in human monocytes in response to both extracellular and intracellular pathogens. Furthermore, there are major differences between macrophages and monocytes that may indicate tissue specific differences in MAIL induced IL-6 production.
The authors have no financial conflict of interest.
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
↵1 Address correspondence and reprint requests to Dr. Mark D. Wewers, The Ohio State University, 201 Davis Heart and Lung Research Institute, 473 12th Avenue, Columbus, OH 43210. E-mail address:
↵2 Abbreviations used in this paper: MAIL, molecule containing ankyrin repeats induced by LPS; INAP, IL-1-inducible nuclear ankyrin repeat protein; EGFP, enhanced GFP; MDP, muramyl dipeptide.
↵3 The online version of this article contains supplemental material.
- Received August 19, 2008.
- Accepted August 13, 2009.
- Copyright © 2009 by The American Association of Immunologists, Inc.