cIAP1/2–TRAF2–SHP-1–Src–MyD88 Complex Regulates Lipopolysaccharide-Induced IL-27 Production through NF-κB Activation in Human Macrophages Free

Interleukin-27, a heterodimeric IL-12 family cytokine, is composed of EBV-induced gene 3 (EBI3) and p28, similar to the IL-12 p40 and p35 subunits, respectively. It is secreted mainly by APCs such as macrophages and dendritic cells (DCs). Initially reported as promoting Th1 differentiation, it exhibits both pro- and anti-inflammatory properties. It augments proinflammatory cytokine production in response to TLR2 and TLR4 stimulation and attenuates IL-10 secretion by human monocytes (1). It inhibits Th1, Th2, and Th17 immune responses (1–3), prevents the generation of inducible regulatory T cells (4), and promotes IL-10 synthesis by T cells (5). It displays anti-inflammatory properties by suppressing TNF-α and IL-1β responses in human macrophages (6). It also exhibits antitumor activity (7), inhibits HIV replication in CD4⁺ T cells and macrophages (8, 9), and has been linked to autoimmune diseases including Crohn’s disease and experimental autoimmune encephalitis (5, 10, 11). High levels of IL-27 have been detected at inflammatory sites during infection with Mycobacterium tuberculosis (12), Trichuris muris (13), and Toxoplasma gondii (14).

Recently, transcriptional regulation of p28 and EBI3 subunits has been studied in murine macrophages and DCs (15, 16). The adaptive proteins, MyD88 and TIR (Toll/IL-1R) domain–containing adaptor-inducing IFN-β (TRIF), regulate IL-27 p28 and EBI3 transcription following LPS stimulation of murine bone marrow–derived DCs and macrophages (15–17). TLR-induced IL-27 production was also found to be dependent on type I IFN responses including IRF-1 and IRF-3 in mouse DCs and human macrophages (15, 18–20). In human DCs, the c-Jun N-terminal kinase (JNK) MAPK regulated IL-27 production following stimulation with TLR4 ligand (21). The c-Rel, p50, and p65 subunits of the NF-κB pathway were implicated in IL-27 p28 and EBI3 expression in LPS-stimulated cells (16).

Interestingly, the role of upstream signaling molecules such as protein tyrosine kinases (PTKs), protein tyrosine phosphatases (PTPs), and inhibitors of apoptosis (IAPs) in IL-27 regulation remains unknown. PTKs such as Src family kinases were implicated in TNF-α and IL-6 production (22). The activity of the Src kinases is tightly regulated by reversible tyrosine phosphorylation (23). Sequestration of the Src tyrosine phosphorylated tail by its SH2 domain maintains an inactive state of the kinase. Therefore, it requires a PTP, such as SHP-1, to activate it and promote downstream phosphorylation (23, 24). SHP-1 regulates several tyrosine phosphorylated molecules in hematopoietic cells and plays a role in LPS-induced IL-12 p40 (25) and IL-6 production in LPS-stimulated bone marrow–derived murine macrophages (26).

Initially described in the context of apoptosis regulation as promoting cell survival, the IAP proteins have recently emerged as key regulators of innate immune signaling that promote inflammation (27). Macrophages of cellular IAP2 (cIAP2) null mice exhibited an impaired inflammatory response to LPS, and were resistant to endotoxic shock (28). IAPs also play a role in innate immune responses against infections with Chlamydia pneumoniae (29) and Listeria monocytogenes (30). IAPs associate with TLR4 signaling complexes to serve as ubiquitin ligases that promote degradation of adaptor proteins such as TNFR-associated factor 3 (TRAF3) (31). cIAP1 and cIAP2 associate with the TNF-α receptor complex I to mediate TNF-α–induced activation of the classical NF-κB pathway (32) and at the same time inhibit the alternative NF-κB pathway by promoting constant degradation of NF-κB–inducing kinase (NIK), the regulatory kinase of noncanonical NF-κB signaling (33). However, the involvement of IAPs in cytokine production by primary cells and particularly human macrophages is poorly understood. Herein, we show for the first time, to our knowledge, that IAPs associate with Src, SHP-1, MyD88, and TRAF1/2 to form a novel upstream signaling complex that regulates LPS-induced IL-27 production through the activation of downstream p38 and JNK MAPKs, PI3K, and NF-κB in human monocyte-derived macrophages (MDMs).

MDMs were obtained from primary monocytes isolated by the adherence method as described earlier (34–36). Blood was obtained from healthy volunteers according to a protocol approved by the Ethics Review Committee of The Ottawa General Hospital and the Children’s Hospital of Eastern Ontario, Ottawa. Briefly, human PBMCs isolated by Ficoll Paque density centrifugation (GE Healthcare Life Sciences, Buckinghamshire, U.K.) were resuspended in serum-free media (5 × 10⁶/ml) and plated in 12-well polystyrene plates (Becton Dickinson, Franklin Lakes, NJ). After being allowed to adhere for 3 h, nonadherent cells were washed off and adherent cells were cultured for 6 d in IMDM supplemented with 10% FBS and 10 ng/ml M-CSF (R&D Systems, Minneapolis, MN). Media containing M-CSF was replenished every 2 d. MDMs thus obtained were 98% CD14⁺. Primary monocytes were obtained from PBMCs by using Miltenyi monocyte negative selection kit II (Miltenyi Biotech, Auburn, CA) as per the manufacturer’s instructions. Negatively selected monocytes were purified using Automacs sorting.

LPS was obtained from Escherichia coli O111:B4 (Sigma, St. Louis, MO), dissolved in complete medium, aliquoted, and stored at −80°C before use. Second mitochondria-derived activator of caspase (Smac) mimetics (SMM) LN730 and SM164 were a generous gift from Dr. Korneluck (Children’s Hospital of Eastern Ontario Research Institute, Apoptosis Research Centre). Chemical inhibitors LY294002 (PI3K/Akt inhibitor), sodium stibogluconate (SHP-1 inhibitor), SU6656 (Src inhibitor), SP600125 (JNK inhibitor), SB203580 (p38 inhibitor), and CAPE (NF-κB inhibitor) were purchased from Calbiochem and used to treat MDMs for 2 h before LPS stimulation.

IL-27 measurement in culture supernatants was performed using IL-27 ELISA DuoSet kit purchased from R&D Systems (catalog number DY2526) following the manufacturer’s instructions and as described earlier (37). Briefly, microplates were coated overnight at 4°C in ELISA coating buffer (eBioscience, San Diego, CA) with 0.4 μg/ml IL-27 capture Ab (R&D Systems). For measurement of TNF-α secretion, 0.2 μg/ml TNF-α capture Ab (Invitrogen, Burlington, ON, Canada) was used with a similar protocol. The next day plates were washed, blocked for 2 h with 10% FBS, followed by the addition of standards and sample supernatants for another 24 h. On the third day biotinylated secondary mouse monoclonal anti-human Abs were added, followed by streptavidin-HRP. Visualization was carried out using 3,3′,5,5′-tetramethylbenzibidine one component HRP microwell substrate solution and 450 nm liquid stop solution for 3,3′,5,5′-tetramethylbenzibidine microwell substrates (BioFX Laboratories, Owings Mills, MD). Absorbance was read using Bio-Rad iMark microplate reader and data were processed using Microplate Manager 6 software.

Monocytes were initially adhered and differentiated into MDMs as described above on round glass coverslips in 12-well plates. After LPS treatment cells were fixed in 4% paraformaldehyde for 15 min, rinsed in PBS, and then permeabilized with 0.1% Triton X-100 for another 10 min. The coverslips were incubated with anti–SHP-1, anti-Src, anti-MyD88, or anti-cIAP2 rabbit primary Abs (Cell Signaling Technology, Danvers, MA) at 4°C overnight. On the second day coverslips were rinsed with 5% BSA in PBS and incubated in the same buffer with secondary Abs for 1 h at room temperature: Alexa Fluor 680 donkey anti-rabbit IgG or Alexa Fluor 350 donkey anti-rabbit IgG (Molecular Probes, Burlington, ON, Canada). Coverslips were mounted on microscopy slides using ProLong Gold antifade mounting media with DAPI (Invitrogen) and examined with a Zeiss LSM 510 Meta confocal microscope using a 488 nm (green), a 633 nm HeNe (red), or a 405 nm (blue) laser. Fluorescent images were acquired with ZEN 2009 software and analyzed with ImageJ software.

Total cell proteins obtained after lysis of cell pellets were subjected to SDS-PAGE and transferred onto a polyvinylidene difluoride membrane (Bio-Rad Laboratories, Hercules, CA) as described earlier (34, 35). Membranes were probed with Abs against cIAP1, cIAP2, XIAP, TRAF1, TRAF2, TRAF6, MyD88, and RIP-1 (all from Cell Signaling Technology). Anti-p100/p52 Ab was from Sigma-Aldrich. For SHP-1 we used a mouse anti–SHP-1 Ab generated in the laboratory. For experiments involving signaling pathways, Abs against IκBα, p-Akt, Akt, p-p38, p38 (all from Cell Signaling), p-JNK, and JNK (both from Santa Cruz Biotechnology, Santa Cruz, CA) were used. Membranes were probed with primary Abs at 4°C overnight, followed by goat secondary Abs conjugated to HRP (Bio-Rad). To control for total protein loading, membranes were stripped of the primary Abs and reprobed with anti-GAPDH (Sigma-Aldrich) or anti–β-actin (Cell Signaling) Ab. Immunoblots were visualized using the Amersham ECL Western blotting detection system. The images were obtained with the ChemiGenius Bio-imaging system and the GeneSnap software (both from Syngene).

All specific small interfering RNAs (siRNAs) were purchased from Santa Cruz Biotechnology. MDMs were transfected with RIP-1 siRNA, TRAF1/2 siRNA, cIAP1/2 siRNA, SHP-1 siRNA, Src siRNA, Akt siRNA, p38 siRNA, JNK1 and 2 siRNA (Santa Cruz Biotechnology) or control siRNA (Qiagen, Venlo, the Netherlands) using TransIT-TKO transfection reagent (Mirus Bio, Madison, WI), as described earlier (34–36). Briefly, 40 pmol siRNA was mixed with 4 μl transfection reagent in 100 μl of serum-free media to allow formation of RNA complexes. After 30 min, the siRNA mixtures were added to cells. After 5 h, the media of transfected cells was changed to complete media. The following day after transfection, cells were treated with LPS for another 24 h, then supernatants and cell pellets were collected for cytokine evaluation and protein extraction respectively.

MDMs (1 × 10⁶ cells per well) were stimulated with LPS for 4 h. RT-PCR for IL-27 subunits was performed as described earlier (37). RNA was extracted using RNeasy kit (Qiagen, Mississauga, ON, Canada) according to the manufacturer’s protocol and as described earlier (34–36). For RT-PCR, a master mix was prepared with 7.5 μl 10× reverse transcriptase buffer, 3 μl dNTP, 7.5 μl random primers, 3.75 μl reverse transcriptase, and 53.25 μl double distilled H₂O. Twenty-five microliters of master mix was added to each sample to 50 ng/μl purified RNA and RT-PCR was carried out for 2 h in a GeneAmp PCR System 2700 amplifier (Applied Biosystems, Carlsbad, CA) to yield cDNA. The PCR program was as follows: initial incubation for 2 min at 50°C and 10 min at 95°C, and cycles of denaturation at 95°C for 15 s were followed by annealing and elongation for 2 min at 60°C. Forty to fifty cycles were applied. cDNA was then used in real-time PCR reactions. For this, samples were prepared with 12.5 μl TaqMan DNA polymerases, 1.25 μl primer pairs (for each of the cytokine subunits), and 8.75 μl double distilled H₂O. Real-time PCR was carried out for 2 h in a 7500 Real-Time PCR System. The expression levels of the transcripts were shown as the ratio compared with β-actin by calculation of cycle threshold values in amplification plots (Applied Biosystems). Primer pairs (Applied Biosystems, U.K.) were as follows: IL-27 EBI3 primer: 5′-AGCAGCAGCCTCCTAGCCT-3′, 5′-ACGCCTTCCGGAGGGTC-3′; IL-27 p28 primer: 5′-GGCCAGGYGACAGGAGACC-3′, 5′-CAGCTTGTACCAGAAGCAAGGG-3′.

MDMs were treated with SMM LN730 for 24 h, followed by LPS (100 ng/ml) for another 24 h. Cell lysates were collected by treating cell pellets with lysis buffer (50 mM HEPES pH 7.5, 150 mM NaCl, 10% glycerol, 1% Triton X-100, 1.5 mM MgCl₂, 100 mM NaF, 100 mM sodium orthovanadate, and 1 mM EGTA pH 7.7) for 1 h on ice. Two hundred micrograms of protein for each sample was mixed with 20 μl of Protein A Sepharose beads (Invitrogen) and specific coimmunoprecipitation Abs (either anti-MyD88, anti-TRAF1, or anti-TRAF2 Abs, all from Cell Signaling) and the mixture was incubated overnight at 4°C. Following overnight incubation, samples were centrifuged and the beads with attached Abs and coimmunoprecipitated proteins were collected. The beads were resuspended in loading dye, boiled for 5 min to release the coimmunoprecipitated protein–Ab complexes, then separated by SDS-PAGE and evaluated by Western blotting as described above.

Data was plotted using Windows Excel 2010 (Microsoft) and GraphPad Prism 5. To minimize variability between donors in their responsiveness to LPS, data were calculated relative to LPS response, which was considered 100%. Statistical analysis of multiple groups was performed using ANOVA, followed by Tukey post hoc test. Two groups were compared using Student t test. A p value <0.05 was considered significant. Unless otherwise specified, plotted data represent the mean ± SD of at least three experiments.

IAP antagonists (SMM) have been evaluated in clinical trials for cancer treatment for their ability to promote autoubiquitination and proteosomal degradation of IAPs leading to TNF-α–induced apoptosis in cancer cells (38–41). To understand the role of IAPs on cytokine production by primary human macrophages, we investigated the impact of SMM (LN730 and SM164) on LPS-induced IL-27 production in TLR4-stimulated human MDMs. LN730 significantly decreased basal levels of cIAP1, cIAP2, and XIAP, whereas SM164 affected only cIAP1 and cIAP2 (Fig. 1A). LN730 significantly inhibited IL-27 secretion in response to LPS in MDMs (Fig. 1B) and primary monocytes (Fig. 1C). Similar results showing inhibition of LPS-induced IL-27 production following SM164 treatment of MDMs (Fig. 1D) and primary monocytes (Fig. 1E) were observed suggesting that this effect is not limited to macrophages or to one SMM. The involvement of cIAP1/2 was confirmed using specific siRNAs for cIAP1 and cIAP2, which effectively inhibited cIAP1/2 expression without affecting XIAP expression (Fig. 1F). There was a significant reduction of LPS-induced IL-27 production in MDMs transfected with cIAP1 and cIAP2 siRNAs compared with those transfected with control siRNA (Fig. 1G).

The receptor-interacting protein 1 (RIP-1), a death domain containing kinase, is involved in TNF-R1 signaling and IAP-mediated apoptosis in cancer cells, p38 MAPK activation in response to TNF-α, and NF-κB activation via TLR3 and TLR4 (42–45). To understand the involvement of IAPs, we investigated if RIP-1 was involved in SMM-mediated inhibition of LPS-induced IL-27 production. RIP-1 levels were increased following LPS stimulation of MDMs and this effect was inhibited following SMM pretreatment (Fig. 2A). The role of RIP-1 in SMM-mediated inhibition of LPS-induced IL-27 production was determined by using RIP-1 siRNA. Despite successful knockdown of RIP-1 (Fig. 2B), LPS-induced IL-27 production remained unchanged (Fig. 2C). These results suggest that SMM-mediated inhibition of LPS-induced IL-27 production does not involve RIP-1 induction.

To further dissect the mechanism of SMM inhibition of LPS signaling, we investigated the effect of SMM on proteins involved in TLR4 signaling. LPS binding to TLR4 triggers two distinct signaling cascades according to different adaptor molecules recruited to the receptor, MyD88 and TRIF, that culminate with cytokine production and IFN responses, respectively (46). TRAFs are signal transduction molecules involved in multiple transduction pathways, such as TLR and TNFR signaling. Of the six TRAF molecules, TRAF6 is implicated in TLR4 signaling downstream of MyD88, where it serves as ubiquitin ligase that activates IκB kinase and classical NF-κB signaling (47). TRAF2 is mainly involved in TNF signaling, whereas the role of TRAF1 is to recruit cIAP1 and 2 to TRAF2, thereby forming a cIAP1/2-TRAF1/2 heterocomplex (48).

Initially, we evaluated if SMM affect TRAF1, 2, 6, and MyD88 expression levels. LN730 treatment inhibited LPS-induced TRAF1, TRAF2, and MyD88, with no effect on TRAF6 (Fig. 2D) suggesting that TRAF1, TRAF2, and MyD88 may be involved in SMM-mediated inhibition of LPS-induced IL-27 secretion. To determine their specific involvement, we used siRNAs for TRAF1, TRAF2, and MyD88, which induced specific knockdown of their targets (Fig. 2E, 2F, left panels). IL-27 production was significantly inhibited in response to LPS in cells transfected with TRAF2 and MyD88 siRNA compared with cells transfected with control siRNA (Fig. 2E, 2F, right panels), suggesting that MyD88 and TRAF2 are the main adaptor molecules in LPS signaling that are responsible for IL-27 production.

The PTK Src has been shown to regulate the production of proinflammatory cytokines such as IL-6 (22, 26). Src requires a protein phosphatase SHP-1 to form a signaling tandem that dephosphorylates Src at the inhibitory tyrosine site (49). Therefore, we determined the involvement of Src and SHP-1 in the regulation of LPS-induced IL-27 production. Treatment of MDMs with either sodium stibogluconate (SS, SHP-1 inhibitor) (Fig. 3A) or SU6656 (SU, Src inhibitor) (Fig. 4A) resulted in significant suppression of IL-27 protein. Similar results were obtained with IL-27 subunits at the mRNA level, with both IL-27 p28 mRNA (Figs. 3B, 4B) and EBI3 mRNA (Figs. 3C, 4C) showing suppression following treatment with SHP-1 or Src inhibitor respectively. These results were confirmed using siRNAs specific to SHP-1 and Src, which were efficient in downregulating the basal level of their respective target proteins (Figs. 3G, 4G). There was a significantly decreased production of IL-27 protein and IL-27 p28 and EBI3 mRNAs in cells transfected with SHP-1 (Fig. 3D–F) and Src (Fig. 4D–F) siRNAs as compared with those transfected with control siRNAs.

The above results show that cIAP1/2, PTP SHP-1, PTK Src, TRAF2, and MyD88 regulate LPS-induced IL-27 production in human macrophages. Therefore, it was of interest to determine if SHP-1/Src signaling and TLR4/MyD88 signaling are interconnected or function as separate, parallel pathways. We show by immunofluorescence that SHP-1 and Src colocalize following LPS stimulation in human macrophages (Fig. 5A). Similarly, SHP-1 (Fig. 5B) and Src (Fig. 5C) were shown to coimmunofluoresce with MyD88 after LPS stimulation. Furthermore, cIAP2 and SHP-1 colocalized after LPS stimulation (Fig. 5D) suggesting that cIAP2 is also a part of the SHP-1–Src–MyD88 signaling complex.

To determine that colocalization observed with immunofluorescence is due to physical interaction of these molecules, we performed coimmunoprecipitation experiments. Using anti-MyD88 capture Ab, results show that cIAP1/2, SHP-1, and Src interacted with MyD88 in media and LPS-treated cells (Fig. 6A, lanes 3 and 4). As expected, cIAP1/2 bands were not visible in cells pretreated with SMM before LPS stimulation (Fig. 6A, lanes 5 and 6). However, SMM did not affect SHP-1 or Src binding to MyD88 (Fig. 6A, lanes 5 and 6). These results confirmed that cIAP1/2, SHP-1, Src, and MyD88 interact to form a signaling complex.

We next determined if TRAF1/2 were also a part of the signaling complex that includes cIAP1/2. TRAF1/2, the binding partners of cIAP1/2, could not be evaluated on these Western blot membranes, as TRAFs 50 kDa bands were masked by the capture Ab H chain band. Therefore, we used either TRAF1 or TRAF2 as capture Abs in subsequent coimmunoprecipitation experiments. Results show that cIAP1/2, SHP-1, Src, and MyD88 bind to TRAF1 (Fig. 6B, lanes 3 and 4) and TRAF2 (Fig. 6C, lanes 3 and 4) in media and LPS-treated cells. Similar to the MyD88 results, there was decreased binding of cIAP1/2 to TRAF1 (Fig. 6B, lanes 5 and 6) and TRAF2 (Fig. 6C, lanes 5 and 6) when cells were treated with SMM. Moreover, SMM pretreatment did not influence TRAF1/2 binding to SHP-1, Src, and MyD88. These results were confirmed using Src as the capture Ab, which showed cIAP2 interactions with the Src-SHP-1 complex upon LPS induction (Fig. 6D, lanes 3 and 4). This interaction was abolished following SMM pretreatment (Fig. 6D, lanes 5 and 6). Overall, these results suggest that an upstream signaling complex consisting of cIAP1/2–TRAF1/2–SHP-1–Src–MyD88 regulates LPS-induced IL-27 production in MDMs.

TLR4 signaling in response to LPS leads to MAPKs and NF-κB activation to promote cytokine synthesis (50). cIAPs are also required for MAPKs activation in response to LPS stimulation (51). Moreover, MAPKs can also be activated downstream of Src/SHP-1 kinases (52). To determine if IAPs influence MAPKs and PI3K/Akt activation leading to LPS-induced IL-27 production, we first determined the requirement for MAPKs and Akt in the regulation of IL-27 production by using inhibitors specific for JNK (SP600125), p38 (SB203580), and Akt (LY294002). Blocking of JNK, p38, and Akt significantly inhibited LPS-induced IL-27 production (Fig. 7A–C, left panels). Interestingly, JNK MAPK had a divergent effect on the two IL-27 subunits at mRNA levels, as it positively regulated LPS-induced IL-27 p28 mRNA without any effect on EBI3 mRNA (Fig. 7A). Similarly, p38 MAPK positively regulated LPS-induced IL-27 EBI3 mRNA, with no effect on p28 mRNA (Fig. 7B). These results suggest that one of the subunits represents the limiting factor in protein secretion. PI3K regulated both p28 and EBI3 mRNA (Fig. 7C). The requirement for JNK, p38, and Akt signaling was confirmed by using their respective specific siRNAs, which significantly inhibited LPS-induced IL-27 production compared with control siRNA treated cells (Fig. 7D). Moreover, SMM inhibited LPS-induced phosphorylation of JNK, p38, and Akt (Fig. 8A, 8B), suggesting that SMM inhibit LPS-induced IL-27 production through the inhibition of p38 and JNK MAPKs and PI3K/Akt signaling.

Because TLR4 signaling culminates with NF-κB activation, we evaluated if cIAP1/2 regulate LPS-induced IL-27 production through NF-κB signaling in MDMs. cIAP ablation by SMM as a single agent has been shown to promote classical pathway of NF-κB activation in cancer cells because of increased TNF-α secretion, which promotes degradation of IκBα protein and subsequent release of p65 and p50 subunits of NF-κB (38, 39). To determine if SMM inhibit classical NF-κB activation in human macrophages, we measured the level of IκBα protein as an indirect indicator of the classical NF-κB activity. In contrast to previous results obtained with cancer cells (38, 39), SMM alone did not affect IκBα expression in human macrophages. As expected, LPS stimulation decreased IκBα levels, indicative of NF-κB activation. However, LPS-induced degradation of IκBα was attenuated in MDMs pretreated with SMM (Fig. 9A), suggesting that LPS-induced NF-κB activation is dependent on cIAPs in human macrophages. In cancer cells SMM induce cell death by promoting the production of TNF-α, a downstream target of classical NF-κB pathway with positive feedback effect on NF-κB activation (38, 39, 45, 51, 53). However, in human macrophages, we show that SMM do not alter LPS-induced production of TNF-α (Fig. 9B), suggesting that this mechanism is not involved in regulating classical NF-κB pathway.

Noncanonical regulation of NF-κB depends on NIK, which promotes partial processing of the p100 subunit to its p52 mature form (54). In resting cells, this pathway is subdued via continuous NIK degradation performed by cIAP1/2 (33). SMM treatment as a single agent leads to IAP inhibition resulting in NIK accumulation and subsequent activation of noncanonical NF-κB pathway in cancer cells (38, 39). The effect of SMM on alternative NF-κB signaling in human macrophages is not known. As a single agent, in contrast to cancer cells (38, 39), SMM did not impact p100 processing in macrophages. However, SMM augmented LPS-induced p100 processing to p52 (Fig. 9B), indicating that SMM promote alternative NF-κB signaling, in contrast to the inhibitory effect on the classical pathway following LPS stimulation.

To determine if the SMM effect on secretion of IL-27 could be attributed to altered NF-κB signaling, we employed the chemical inhibitor CAPE to block NF-κB pathway. Inhibition of NF-κB by CAPE significantly reduced LPS-induced IL-27 production (Fig. 9C). Given the requirement for NF-κB in regulating IL-27 and the inhibitory effect of SMM on the classical pathway, these results suggest that IAPs regulate IL-27 induction following TLR4 stimulation through the classical NF-κB pathway. Moreover, activation of a noncanonical alternate pathway by SMM in LPS-activated macrophages does not impact IL-27 production (Fig. 10).

SMM compounds have been actively studied for their apoptosis-inducing abilities in cancer cells (38, 39). We and others have shown that cIAP ablation on its own does not influence survival of primary human macrophages (34, 35, 55) and murine B cells (33). However, the role of IAPs in regulating immune responses in primary cells is poorly understood. One report from mouse studies indicates that cIAPs mediate the production of proinflammatory cytokines in response to LPS, without affecting IFN responses. As a result, cIAP ablation using SMM resulted in decreased IL-6, TNF-α, and IL-12 production, with no effect on IL-10 (31). However, the effect of SMM on cytokine production in primary human cells remains unknown. SMM have also been used in phase 1 and 2 clinical trials as therapeutic options in various cancers (40, 41). Therefore, it is imperative to understand the impact of SMM on immune signaling in primary human cells. Herein, we show that cIAPs positively regulate LPS-induced IL-27 production in human macrophages by a novel signaling complex comprising cIAP1/2, TRAF2, SHP-1, Src, and MyD88 leading to p38, JNK, and Akt activation and NF-κB signaling.

LPS binding to TLR4 recruits two adaptor molecules with Toll/IL-1 receptor (TIR) domains, MyD88 and TRIF. MyD88 recruits IRAK1/4, followed by TRAF6, which ultimately mediates IκB kinase complex activation leading to secretion of inflammatory cytokines (56). SMM bind specifically to cIAP1 and 2 (45) and cause their degradation without affecting IAP-associated proteins such as TRAFs (39). TRAF1/2 bind cIAP1/2 to form a heterocomplex with key functions in TNFR signaling (57), whereas TRAF6 contributes to MyD88-dependent TLR4 signaling (58). Our results show that SMM treatment inhibited LPS-induced TRAF1 and 2, with no effect on TRAF6. However, only TRAF2 was required for IL-27 production. The mechanism responsible for IL-27 downregulation by TRAF2 and not by TRAF1 is not clear. There is a redundancy in the function and structure of TRAF1 and TRAF2. Although both TRAF1 and TRAF2 can bind cIAP1/2, TRAF1 and TRAF2 preferentially form 1:2 trimers that bind with higher affinity to cIAP2 compared with TRAF2 alone, indicating the importance of oligomerization for TRAF1/2 function (59). It is possible that when TRAF1 levels are reduced, TRAF2 is still able to bind the signaling complex and promote IL-27 production. Secondly, the RING domain and the cIAP1/2-interacting motif within TRAF2 are required to regulate downstream (NIK) stability, mediate activation of the canonical NF-κB pathway, and suppress constitutive noncanonical NF-κB activation (57, 60). Interestingly, RING domain is absent in TRAF1 (61–63). This may explain the inhibitory effect on LPS-induced IL-27 production following silencing of TRAF2, keeping in view that LPS-induced IL-27 production is regulated by the canonical pathway.

The mechanism of TRAF1 and 2 inhibitions may be an indirect effect of SMM on NF-κB signaling. Similar to cIAP1/2, TRAF1/2 are also targets of SMM on NF-κB activity (64). Therefore, by inhibiting LPS-induced classical NF-κB pathway, SMM may indirectly prevent TRAF1/2 induction. TRAF2 is a dual regulator of NF-κB that promotes activation of the classical pathway, but inhibits the noncanonical pathway (60) by inducing degradation of NIK (65). Therefore, SMM-induced TRAF2 downregulation may also potentiate its effects on NF-κB pathways by promoting inhibition of the classical pathway.

The PTP SHP-1 plays a critical role in protection against autoimmunity and inflammation and in the regulation of innate and adaptive immune responses in murine models (26, 66). By employing the me/me mice, we and others have shown SHP-1 as a key positive regulator of IL-6 and IL-10 in murine bone marrow–derived macrophages (26) as well as in IL-21 production by CD4⁺ T cells resulting in a dampened Th17 response (67). In this study, we also show that SHP-1 regulates IL-27 production downstream of TLR4 in human macrophages.

PTKs such as Src represent targets for SHP-1 activity (68). Under basal conditions, Src is present in a restrictive state, where Tyr527 in the C terminus is phosphorylated and binds with the Src SH2 domain (24). This molecular interaction sequesters the catalytic domain and holds the kinase in an inactive state and a PTP, such as SHP-1, can activate it to promote downstream phosphorylation events (24). In conformity with these observations, our results show that macrophages employ the SHP-1/Src complex downstream of TLR4 signaling to positively regulate IL-27 secretion through transcriptional regulation of its subunits.

Inflammasomes, multiprotein complexes of the innate immune system, play a key role in initiating inflammatory responses during infection (69). Their nature and exact composition depend on the microbial activator (70). This study has provided evidence for the formation of a multiprotein complex encompassing SHP-1, Src, and the TLR4-associated proteins including cIAP1/2, TRAF2, and MyD88 leading to IL-27 production in human macrophages. We and others have shown that MAPKs and PI3K regulate IL-27 production to other stimuli (37, 71). Therefore, we investigated whether the cIAP1/2–TRAF2–SHP-1–Src–MyD88 complex regulates LPS-induced IL-27 production through the activation of MAPK/PI3K pathway. Our results show that IAP abolition with SMM reduced LPS signaling through p38 and JNK MAPK and PI3K pathways, suggesting that the signaling complex regulates LPS-induced IL-27 production by inhibition of MAPKs and Akt/PI3K signaling. This study and our previous observations (37, 72) show that PI3K pathway positively regulates IL-27 production in human macrophages. However, PI3K pathway did not affect IL-27 production in murine DCs (21).Thus, the role of PI3K pathway in IL-27 production may vary from cell to cell type and the stimulus involved.

In tumor cell lines SMM alone activate both classical and alternative NF-κB pathways by promoting TNF-α secretion and stabilizing NIK, respectively (38, 39). cIAP ablation also reduces activation of the classical NF-κB pathway in response to members of the TNFR family, such as TNF-α (73, 74). However, the effect of SMM on NF-κB activity in primary macrophages, either alone or in response to other immune activators, remains unknown. We show that SMM alone did not activate either classical or alternate NF-κB pathway in human macrophages. In contrast, SMM exhibited divergent effects on LPS-induced NF-κB signaling, causing inhibition of the classical NF-κB pathway and activation of the alternative noncanonical NF-κB pathway, manifested by reduced IκB degradation and increased p100 processing, respectively. The different results obtained in cell lines and primary cells on NF-κB activation may be due to different TNF-α secretion patterns in response to SMM. SMM-sensitive cell lines secrete TNF-α in response to cIAP ablation and consequently become susceptible to TNF-α–induced apoptosis (38, 45). However, we did not detect enhanced TNF-α secretion following SMM treatment either alone or in response to LPS (Fig. 9B). This result is in keeping with the fact that SMM treatment causes apoptosis of sensitive cancer cells (38, 39) but not of primary monocytic cells (34, 35, 55). The effect on alternative NF-κB pathway can be explained by cIAP ability to bind and degrade NIK (33). In the absence of cIAP1/2 as a result of SMM treatment, NIK regulatory complex formation is impaired resulting in NIK accumulation and increased p100 to p52 processing.

In summary, our results clearly show that SMM inhibit LPS-induced IL-27 production in human macrophages without impacting the cytokine production in the absence of TLR4 stimulation. Moreover, we have elucidated the signaling pathways that regulate IL-27 production through the formation of a large signaling complex that brings together major signaling pathways, such as SHP-1, Src, cIAPs, TRAF2, and MyD88. This signaling complex impacts on p38 and JNK MAPK and PI3K and ultimately the classical NF-κB pathway to regulate LPS-induced IL-27 production in human macrophages (Fig. 10). These results assume significance when evaluating the clinical implications of SMM administration, given the constant interest in SMM as cancer treatment.

We thank Dr. M. Saxena for critically reading the manuscript. We also thank Dr. Bruno Fonseca for help in immunoprecipitation experiments.

The authors have no financial conflicts of interest.