Injury-induced MRP8/MRP14 stimulates IP-10/CXCL10 in monocytes/macrophages

Abstract

Trauma/hemorrhagic shock is associated with morbidity and mortality due to dysregulated inflammation, which is driven in part by monocytes/macrophages stimulated by injury-induced release of damage-associated molecular pattern (DAMP) molecules. MRP8/MRP14 is an endogenous DAMP involved in various inflammatory diseases, though its mechanism of action is unclear. Circulating MRP8/MRP14 levels in human blunt trauma nonsurvivors were significantly lower than those of survivors (P < 0.001). Human monocytic THP-1 cells stimulated with MRP8/MRP14 expressed the chemokine IFN-γ inducible protein 10 (IP-10)/CXCL10. Circulating IP-10 levels in human blunt trauma patients were correlated positively with MRP8/MRP14 levels (r = 0.396, P < 0.001), and were significantly lower in trauma nonsurvivors than in survivors (P < 0.001). We therefore sought to determine the mechanisms by which MRP8/MRP14 stimulates IP-10 in monocytes/macrophages, and found that induction of IP-10 by MRP8/MRP14 required Toll-like receptor 4 and TRIF but not MyD88. Full induction of IP-10 by MRP8/MRP14 required synergy between the transcription factors NF-κB and IFN regulatory factor 3 (IRF3). The receptor for IP-10 is CXCR3, and MRP8/MRP14-induced chemotaxis of CXCR3⁺ cells was dependent on the production of IP-10 in monocytes/macrophages. Furthermore, in vivo study with a mouse trauma/hemorrhagic shock model showed that administration of neutralizing antibody against MRP8 prevented activation of NF-κB and IRF3 as well as IP-10 production. Thus, the current study identified a novel signaling mechanism that controls IP-10 expression in monocytes/macrophages by MRP8/MRP14, which may play an important role in injury-induced inflammation.—Wang, J., Vodovotz, Y., Fan, L., Li, Y., Liu, Z., Namas, R., Barclay, D., Zamora, R., Billiar, T. R., Wilson, M. A., Fan, J., Jiang, Y. Injury-induced MRP8/MRP14 stimulates IP-10/CXCL10 in monocytes/macrophages.

Trauma remains the most common reason for productive years lost and is the primary cause of death and morbidity in people under the age of 54 y (1, 2). At an adaptive level, posttrauma inflammation is required to activate protective cell responses and to enhance antimicrobial defenses (3, 4). However, excessive and sustained escalation of immune pathways occurs in the process of immune dysregulation, which is associated with morbidity and mortality (5). Key to the induction of injury-induced inflammation are damage-associated molecular pattern (DAMP) molecules (6, 7).

Myeloid-related protein (MRP) 8 (S100A8, calgranulin A) and MRP14 (S100A9, calgranulin B) belongs to the S100 calcium-binding protein family (8, 9), which comprises more than 20 members, three of which are DAMPs that have been linked to innate immune functions, i.e., MRP8, MRP14, and MRP6 (S100A12 or calgranulin C) (9, 10). MRP8 and MRP14 are expressed at high levels by granulocytes and monocytes, accounting for up to 40% to 50% of total cytosolic proteins in neutrophils (9). MRP8/MRP14 are released from phagocytic cells either passively (from damaged or necrotic cells) or actively (by activated phagocytes); the latter is the major physiologic source for extracellular MRP8/MRP14 (9, 11). MRP8 and MRP14 form homodimers, heterodimers, and higher-order complexes. Evidence demonstrates that the MRP8/MRP14 heterodimer (also named calprotectin) is the dominant extracellular form (9, 12).

Importantly, the MRP8/MRP14 heterodimer has been identified recently as a DAMP (8), inducing a variety of inflammatory reactions by stimulating the release of proinflammatory cytokines such as TNF-α and IL-1β (8, 13). MRP8/MRP14 is an important mediator and biomarker for various infections or inflammation-associated diseases, including rheumatoid arthritis, inflammatory bowel diseases, sepsis, myocardial infarction, and allograft rejection (9). By up-regulating the expression of chemokines and integrins, MRP8/MRP14 enhances chemotaxis and migration of immune cells to the sites of infection, thus promoting the innate and adaptive immune response (9, 14, 15). Although several receptors for MRP8/MRP14, including TLR-4 (8), the receptor for advanced glycation end products (RAGE) (16), CD36 (17), special carboxylated N-glycans (18), and heparin-like glycosaminoglycans (19) have been reported, TLR-4 and RAGE are considered as 2 major receptors on macrophages for MRP8 and MRP14.

In the present study, we sought to determine if MRP8/MRP14 was produced in the setting of trauma/hemorrhage, and if so, to define the primary inflammatory mechanism or mechanisms stimulated by this DAMP. We showed that human blunt trauma survivors produced higher levels of MRP8/MRP14 than nonsurvivors, and a preliminary screen suggested that IFN-γ inducible protein 10 (IP-10; CXCL10) was significantly increased in monocytes/macrophages stimulated with MRP8/MRP14. In agreement with this finding, circulating IP-10 levels were higher in blunt trauma survivors vs. nonsurvivors. Thus, we characterized the signaling pathway involved in IP-10 induction by MRP8/MRP14 in monocytes/macrophages, as well as establishing an in vivo role for IP-10 in MRP8/MRP14-induced chemotaxis.

MATERIALS AND METHODS

Human trauma

All human sampling protocols were approved by the University of Pittsburgh institutional review board. Informed consent was obtained from each patient or next of kin as per institutional review board regulations. Blood samples from 32 human trauma victims including 16 survivors (13 male and 3 female subjects) and 16 nonsurvivors (13 male and 3 female subjects) of motor vehicle accidents (16) or falls (16) were studied. Plasma samples were obtained from day 1 to day 7 after trauma.

Reagents and mice

Male C57BL/6 wild-type (WT), TLR4^−/−, RAGE^−/−, MyD88^−/−, TRIF^−/−, and CD14^−/− mice were purchased from Jackson Laboratory (Bar Harbor, ME, USA). THP-1 cells and RAW264.7 cells were purchased from the American Type Culture Collection (ATCC, Manassas, VA, USA). The RNeasy Mini Kit was purchased from Qiagen (Hilden, Germany). An MRP8/MRP14 human ELISA kit was from Bühlmann Laboratories AG (Schönenbuch, Switzerland). NF-κB EMSA Kit was from Viagene Biotech (Ningbo, China). The human/mouse IP-10/CXCL10 immunoassay kit was from Millipore (Billerica, MA, USA). Recombinant mouse IP-10 protein and antibody against mouse NF-κB p65 were obtained from Abcam (Cambridge, United Kingdom). Recombinant human IFN-γ, anti-human IFN-γ Ab, mouse IP-10 antibody were purchased from R&D Systems (Minneapolis, MN, USA). All phospho-specific antibodies against IKKα (Ser¹⁸⁰)/IKKβ (Ser¹⁸¹), NF-κB p65 (Ser⁵³⁶), IFN regulatory factor 3 (IRF3) (Ser³⁹⁶), TBK1 (Ser¹⁷²) or IKKε (Ser¹⁷²), and Alexa-Fluor 488 conjugated anti-rabbit or goat IgG were purchased from Cell Signaling Technology (Danvers, MA, USA). Antibody against mouse IRF3 and isotype control IgG were from Santa Cruz Biotechnology (Santa Cruz, CA, USA). Actinomycin D (Act-D), cycloheximide, polymyxin B (PMB), pyrrolidine dithiocarbamate (PDTC), and Bay11-7082 were purchased from Calbiochem (San Diego, CA, USA). Specific siRNA targeting mouse IRF3 (siGENOME) and its control siRNA were purchased from Thermo Scientific (Waltham, MA, USA). Cay10576 was purchased from Cayman (Ann Arbor, MI, USA). Dual-Glo Luciferase Assay System was purchased from Promega (Madison, WI, USA). Neutralizing rabbit IgG anti-MRP8 polyclonal antibody was from Biogot Biotechnology Company (Nanjing, China).

Preparation of MRP8/MRP14 heterodimer

A prokaryotic vector expressing human MRP8 or MRP14 was constructed according to routine molecular cloning methods, and recombinant proteins were purified with a Ni²⁺-NTA-agarose column, separated by SDS-PAGE, and quantified with the Bradford method. The MRP8/MRP14 heterodimer complex was prepared as previously described, with a minor modification (20, 21). Briefly, heterodimerization was performed by mixing MRP8 and MRP14 proteins in an equimolar solution containing 10 mM Tris-HCl, pH 7.4; 0.1% sodium cholate; 1 mM EDTA; and 1 mM β-mercaptoethanol. Then CaCl₂ from a stock solution was added to a final concentration of 2 mM followed by a 10-min incubation at ambient temperature. Procedure efficiency was assessed with an ELISA kit (Bühlmann Laboratories AG) to detect MRP8/MRP14 heterodimers. Endotoxin contaminations were excluded using Limulus amebocyte lysate assay (minimum LPS sensitivity = 0.125 EU/ml).

Cell culture

THP-1 and RAW264.7 cells were respectively cultured with RPMI 1640 or DMEM medium containing 10% FBS, 100 U/ml penicillin, and 100 mg/ml streptomycin at 37°C in a 5% CO₂ incubator. Alveolar macrophages (AMs) were collected from bronchoalveolar lavage fluid of mice (22), and bone marrow–derived macrophages (BMDMs) were cultured with macrophage colony-stimulating factor (10 ng/ml) as reported previously (23). Cultured cells were stimulated with recombinant MRP8/MRP14 after various experimental pretreatments.

MRP8/MRP14 measurement

Plasma MRP8/MRP14 was measured with the MRP8/MRP14 human ELISA kit (Bühlmann Laboratories AG), according to the manufacturer’s instructions.

Cytokine measurement

Plasma IP-10 from patients or cultured cell supernatants, cytokines from cultured cell supernatants were measured with the Luminex Multiplex assay, which uses a bead-based xMAP (flexible multianalyte profiling) technology (24), in accordance with the manufacturer’s instructions.

RT-PCR and qPCR

Total RNA was extracted using an RNeasy Mini Kit according to the manufacturer’s instructions. For the RT-PCR experiment, PCR amplification was performed on the resulting cDNAs with a pair of specific primers for human IP-10 (forward, 5′-GTGGCATTCAAGGAGTACCTC-3′; reverse, 5′-GCCTTCGATTCTGGATTCAGACA-3′) or for human β-actin (forward, 5′-AGCGAGCATCCCCCAAAGTT-3′; reverse, 5′-GGGCACGAAGGCTCATCATT-3′), for mouse IP-10 (forward, 5′-GCCGTCATTTTCTGCCTCAT-3′; reverse, 5′-GCTTCCCTATGGCCCTCATT-3′) or for mouse β-actin (forward, 5′-GAGACCTTCAACACCCCAGC-3′; reverse, 5′-ATGTCACGCACGATTTCCCT-3′) at 94°C for 30 s, 57°C for 30 s, and 72°C for 60 s for 30 cycles.

PCR products were separated by 1.2% agarose gel and visualized by ethidium bromide staining. Quantitative real-time PCR (qPCR) was performed using a 7500 real-time PCR system (Applied Biosystems, Foster City, CA, USA) to verify the specific amplification. A melting point analysis was accomplished after the last cycle by cooling the samples to 55°C and then increasing the temperature to 95°C at a speed of 0.2°C/s. Each sample was run in triplicate. Relative RNA was calculated and normalized with β-actin as an internal control.

Western blot analysis

Cells were lysed with lysis buffer (Cell Signaling Technology), and cell extracts were resolved with 10% SDS-PAGE for Western blot analysis.

EMSA for detection of activity of NF-κB

Cells were collected and centrifuged at 1000 × g for 5 min. The pellets were resuspended in ice-cold homogenization buffer (10 mM HEPES, 1.5 mM MgCl₂, 10 mM KCl, 0.3 mM sucrose, 0.1 mM ethylene glycol bis [2-aminoethylether] tetra-acetic acid [EGTA], 0.5 mM dithiothreitol [DTT], 0.5 mM phenylmethylsulfonyl fluoride [PMSF], 1 mM Na₃VO₄, 1 mM sodium fluoride [NaF], proteinase inhibitor cocktail and 0.1% Nonidet P-40, pH 7.9) and mixed completely, followed by centrifugation at 3000 × g for 20 min at 4°C. The pellets were then resuspended in 10 μl nuclear extraction buffer (20 mM HEPES, 420 mM NaCl, 1.5 mM MgCl₂, 0.1 mM EGTA, 0.2 mM EDTA, 25% glycerol, 1 mM DTT, 0.5 mM PMSF, 1 mM Na₃VO₄, and proteinase inhibitor cocktail, pH 7.9), then incubated for another hour on ice and centrifuged at 14,000 × g for 20 min at 4°C. The supernatants (nuclear fractions) were then stored at −80°C for further use. Protein concentrations were measured using the Bradford method.

EMSA was performed using the NF-κB EMSA Kit according to the manufacturer's instructions. Nuclear extracts (15 μg) were incubated with biotin-labeled double-stranded oligonucleotide probe in binding buffer for 20 min. The complex was separated on a 6.5% polyacrylamide gel and transferred onto a nylon membrane, then covalently cross-linked by UV irradiation for 10 min followed by chemiluminescence with a Kodak 4000R image workstation (Eastman Kodak, Rochester, NY, USA). To verify binding specificity of NF-κB, competition analysis was performed by adding excess unlabeled competitor to the mixture and incubating on ice for 20 min prior to the binding reaction.

Cell immunostaining

AMs were seeded on plates at a density of 2 × 10⁵ cells/ml and stimulated with MRP8/MRP14. After washing twice with PBS, the cells were fixed in a mixture of 4% paraformaldehyde followed by permeabilization with PBS containing 0.25% Triton X-100 for 15 min. After 3 washings with PBS, cells were blocked with 1.0% BSA and incubated with a primary antibody against NF-κB p65 or IRF3, respectively. Cells were then washed 3 times with PBS and incubated with Alexa-Fluor 488–conjugated anti-rabbit or anti-goat IgG for 60 min and 1 μg/ml Hoechst dye for 15 min at room temperature in the dark. The plates were washed 3 times with PBS. Fluorescent analysis was conducted using a FluoView Confocal Microscope (Olympus, Japan).

Luciferase assay

RAW264.7 cells were plated in 24-well plates for transfection with reporter gene plasmids using Lipofectamine 2000 (Invitrogen, Carlsbad, CA, USA). Twenty-four hours after transfection, cells were stimulated with MRP8/MRP14 for 6 h, then collected for the Luciferase assay by using a Dual-Glo Luciferase Assay System (Promega) and normalized to Renilla luciferase.

siRNA transfection

RAW 264.7 cells (1 × 10⁵) were transfected either with SMARTpoolt mouse IRF3 siRNA or with nonspecific siRNA (Cont.siRNA) using the Neon Transfection System (Invitrogen) according to the manufacturer’s instructions with some modifications. Briefly, appropriate siRNA was added to cells cultured in an antibiotic-free medium for electroporation using optimized parameters of 1680 V and a 20 ms pulse width, with 1 pulse, followed by incubation at 37°C for 36 h. After treatment with MRP8/MRP14 for 6 h, cells were collected to measure IP-10 mRNA by RT-PCR.

Chemotaxis assay

CXCR3⁺ T lymphocyte migration was assessed in a 48-well chemotaxis microchamber (NeuroProbe, Gaithersburg, MD, USA) using polycarbonate membranes with 5 μm pores, according to the manufacturer’s instructions. CXCR3⁺ T lymphocytes were prepared as previously described (25). First, the membrane was coated with 10 μg/ml collagen IV at 37°C for 2 h. Then supernatants derived from BMDMs of different groups (including WT or genetic knockout mice) with and without MRP8/MRP14 stimulation or IP-10 recombinant protein (1 μg/ml) was added to the lower wells of the chamber, and 1 × 10⁵ CXCR3⁺ T lymphocytes suspended in RPMI 1640 were added to the top wells of the chamber. Chambers were then incubated for 180 min in a 37°C, 5% CO₂ atmosphere. After incubation, filters were removed, washed with PBS on the upper side, fixed with methanol, and stained with Giemsa. Cells that migrated to the lower surface were counted under a microscope (Olympus, Japan) in five random high-powered fields. All assays were performed in triplicate. Spontaneous migration was determined in the absence of supernatants.

In vivo air pouch animal model

Mouse air pouches were prepared as previously described (26). In brief, mice were anesthetized, followed by injection of 5 ml sterile air under the dorsal skin. On day 3, the resultant space was again injected with 3 ml sterile air. On day 5, LPS was administrated intraperitoneally to activate lymphocytes for high CXCR3 expression. Supernatants of BMDMs from WT or TLR4^−/− mice stimulated with or without MRP8/MRP14 were injected into dorsal air pouches of WT mice at specific time points after LPS administration (0, 8, 16 h). Air pouches were washed with PBS after a fixed migration time of 8 h, and the lavage fluids were collected for lymphocyte counting.

Mouse MRP8 neutralizing antibody activity assay

RAW264.7 cells were incubated with recombinant mouse MRP8 protein (1.5 μg/ml), which was preincubated with various concentrations of MRP8 neutralizing antibody or control IgG (0–100 μg/ml) for 1 h at 37°C. After 24 h of incubation, the supernatant was collected and the levels of IP-10 quantified using Luminex.

Mouse model of hemorrhagic shock and femur fracture (HS/FF)

C57BL/6 mice were anesthetized with ketamine (50 mg/kg) and xylazine (5 mg/kg) via intraperitoneal administration. Bilateral femur fractures were manually induced using hemostats as previously described (27). This resulted in an A-type femoral fracture combined with a moderate soft tissue injury. Hemorrhagic shock was initiated by blood withdrawal through the cannulated femoral artery and reduction of the mean arterial pressure to 40 mm Hg within 20 min. After a hypotensive period of 1 h, animals were resuscitated by transfusion of the shed blood and Ringer’s lactate in a volume equal to that of shed blood over a period of 20 min. Sham-treated animals underwent the same surgical procedures without HS/FF. Anesthesia was maintained during the entire study period.

Neutralizing anti-MRP8 polyclonal antibody or control IgG (4 mg/kg body weight) was administered to some animals 2 h before HS/FF model on the basis of previous publications (28, 29). Blood and AMs were collected at 4 h after resuscitation for experimental analysis.

Statistical analysis

All data were obtained from at least three independent experiments and presented as mean ± sd. Statistical significance among group means was performed by ANOVA. Student-Newman-Keuls post hoc test was performed, and correlations between MRP8/MRP14 and IP-10 or IL-8 were assessed with Spearman’s rank correlation coefficient. P values of <0.05 were considered to be statistically significant.

RESULTS

Association between MRP8/MRP14 and IP-10 in human blunt trauma patients

We hypothesized that MRP8/MRP14 was a DAMP in the setting of human blunt trauma. As seen in Fig. 1A, plasma samples from human blunt trauma patients were assessed for MRP8/MRP14. This analysis demonstrated that circulating MRP8/MRP14 levels were significantly higher in survivors of blunt trauma vs. nonsurvivors (P < 0.001). Both groups showed very close clinical and demographic features, including age, degree of infection, and Injury Severity Score values (Supplemental Table 1). We hypothesized that monocytes/macrophages were a major source of the inflammatory mediators produced in settings such as trauma/hemorrhage; accordingly, we screened human THP-1 monocytes/macrophages for expression of a panel of inflammatory mediators in response to MRP8/MRP14. As seen in Fig. 1B, the 2 predominant inflammatory mediators expressed by THP-1 cells in response to MRP8/MRP14 were the chemokines IP-10/CXCL10 and IL-8. We next sought to determine if the production of IP-10 and/or IL-8 was associated with the production of MRP8/MRP14. This analysis suggested a positive correlation between circulating MRP8/MRP14 and IP-10 (r = 0.396, P < 0.001) (Fig. 1C), but we found no association between MRP8/MRP14 and IL-8 (Supplemental Fig. 1). Importantly, blunt trauma survivors exhibited significantly higher (P < 0.001) circulating IP-10 levels vs. nonsurvivors (Fig. 1D).

Figure 1.

Correlation between MRP8/MRP14 and IP-10 in trauma patients and human monocytes/macrophages. (A) Plasma from survivors and nonsurvivors was sampled at different days (from day 1 to day 7). MRP8/MRP14 in patient plasma was measured using specific ELISA kit. (B) THP-1 cells were challenged with 3 μg/ml recombinant MRP8/MRP14. Cytokine profiles in the culture supernatants were screened using multiplex cytokine assay kit by Luminex system. (C) Correlation between plasma MRP8/MRP14 and IP-10 was analyzed by Spearman’s rank correlation coefficient. (D) Plasma IP-10 levels from survivor and nonsurvivor patients were determined by Luminex.*P < 0.001 vs. nonsurvivor group; ^ΔΔP < 0.01 vs. respective control group.

Transcriptional induction of IP-10 by MRP8/MRP14

We next sought to determine the mechanisms by which MRP8/MRP14 induced IP-10 expression in monocytes/macrophages. qPCR analysis showed that MRP8/MRP14 induced IP-10 mRNA accumulation starting at 2 h, reaching its peak around 6 h, and declining to near baseline levels at 24 h (Fig. 2A). Correspondingly, IP-10 protein in the supernatants was increased significantly at 6 h on the treatment with MRP8/MRP14 and further increased until 24 h (Fig. 2B). The MRP8/MRP14-induced IP-10 protein production presented a concentration-dependent manner (Fig. 2C). Furthermore, pretreatment of THP-1 cells with Act-D, an inhibitor of gene transcription, significantly inhibited IP-10 gene transcription induced by MRP8/MRP14 (Fig. 2D), indicating a control at transcriptional level.

Figure 2.

MRP8/MRP14 induces IP-10 expression at transcriptional level. (A) THP-1 cells were incubated with MRP8/MRP14 (1.5 μg/ml) for 0 to 24 h. qPCR was performed to measure IP-10 mRNA expression. (B) IP-10 protein in the culture supernatants of THP-1 cells treated with 1.5 μg/ml MRP8/MRP14 was quantified by Luminex. (C) THP-1 cells were incubated with 0 to 6 μg/ml MRP8/MRP14 for up to 24 h, and IP-10 in the cultured supernatants was detected by Luminex. (D) THP-1 cells were pretreated with 5 μg/ml Act-D or 5 μg/ml cycloheximide, then stimulated with or without MRP8/MRP14 (1.5 μg/ml) for 4 h; IP-10 mRNA was quantified by qPCR. (E) THP-1 cells were challenged with MRP8/MRP14 or LPS alone, or MRP8/MRP14 or LPS inactivated by 10 μg/ml PMB or heated at 80°C for 30 min (to denature). IP-10 levels were determined 24 h after stimulation. Data represent mean ± sd from 3 independent experiments. **P < 0.01 vs. untreated group; ^ΔΔP < 0.01 vs. group treated with MRP8/MRP14 or LPS.

LPS contamination of recombinant MRP8/MRP14 protein was excluded by using PMB or heat-inactivated (denature), as shown in Fig 2E. The effect of MRP8/MRP14 could not be inhibited by PMB (10 μg/ml) at concentrations that efficiently blocked the activity of 100 ng/ml LPS. In addition, denatured MRP8/MRP14 lost the ability to induce IP-10 expression. LPS is not inactivated at temperatures used to denature MRP8/MRP14 (Fig. 2E).

Induction of IP-10 by MRP8/MRP14 depends on calcium but not IFN-γ

Calcium has been suggested as a prerequisite for the biologic functions of MRP8/MRP14 (12). Our study also suggests that MRP8/MRP14-induced IP-10 expression requires calcium (Supplemental Fig. 2A).

Previous studies demonstrated that IFN-γ is an IP-10 inducer (30–34). Thus, we sought to determine if MRP8/MRP14-induced IP-10 requires IFN-γ production. As shown in Supplemental Fig. 2B, there is no significant correlation between MRP8/MRP14-induced IP-10 production and IFN-γ expression within 24 h. Furthermore, a neutralizing antibody against IFN-γ failed to show effect on the induction of IP-10 in THP-1 cells treated with MRP8/MRP14, whereas the neutralizing antibody led to suppression of IP-10 induction by IFN-γ (Supplemental Fig. 2C).

MRP8/MRP14-induced IP-10 expression is TLR4-TRIF dependent

To determine the receptor through which MRP8/MRP14 induces IP-10 expression, we isolated AMs and BMDMs from WT, TLR4^−/−, and RAGE^−/− mice, respectively. IP-10 transcripts and protein were significantly induced in response to MRP8/MRP14 in both WT and RAGE-deficient macrophages, but not in TLR4-deficient cells (Fig. 3A, B). To further determine if MyD88, TRIF (35, 36), and CD14 (37) were involved in the MRP8/MRP14-induced IP-10 production, macrophages from MyD88-, TRIF-, and CD14-deficient mice were utilized. As shown in Fig. 3C, D, the absence of MyD88 had no effect on MRP8/MRP14-induced IP-10 mRNA and protein expression; however, TRIF deficiency markedly prevented IP-10 expression induced by MRP8/MRP14, indicating that MRP8/MRP14-induced IP-10 expression is mediated by a TLR4/TRIF-dependent pathway. CD14 deficiency failed to block MRP8/MRP14-induced IP-10 mRNA expression (Fig. 3E), indicating that CD14-mediated TLR4 endocytosis (37) was not involved in the production of IP-10 induced by MRP8/MRP14.

Figure 3.

TLR4-TRIF pathway, but not RAGE, MyD88, and CD14 pathway, mediates induction of IP-10 by MRP8/MRP14. (A, C, E) BMDMs isolated from C57BL/6 (WT), TLR4^−/−, RAGE^−/−, MyD88^−/−, TRIF^−/−, or CD14^−/− mice were stimulated with or without MRP8/MRP14 (1.5 μg/ml) for 6 h. RT-PCR was used to detect IP-10 mRNA expression; β-actin was used as internal control. (B, D) AMs isolated from WT, TLR4^−/−, RAGE^−/−, MyD88^−/−, and TRIF^−/− mice were treated with MRP8/MRP14 for 12 h. IP-10 protein was detected by Luminex. Data represent mean ± sd from 3 independent experiments. **P < 0.01 vs. WT untreated group; ^ΔΔP < 0.01 vs. RAGE^−/− untreated group; ^##P < 0.01 vs. MyD88^−/− untreated group.

MRP8/MRP14 activates NF-κB pathway

Bioinformatic analysis and previous studies demonstrated that in the promoter region of the IP-10 gene there are 3 conserved cis elements, i.e., 1 IFN-stimulated response element (ISRE) and 2 NF-κB binding sites (κB1 and κB2) (32–34). We hypothesized that one or more of these cis elements might be involved in MRP8/MRP14-induced IP-10 expression. Western blot analysis performed with antibodies specific to phosphorylated IKKα/β or NF-κB p65 showed that in macrophages treated with MRP8/MRP14 for 1 h, both IKKα/β (Fig. 4A) and NF-κB p65 (Fig. 4B) were significantly phosphorylated, which led to NF-κB p65 location into the nuclei of macrophages as determined by Western blot analysis and confocal immunofluorescence microscopy (Fig. 4C, D). Electromobility shift assay consistently demonstrated that, like LPS, MRP8/MRP14 enhanced specific NF-κB binding activity in THP-1 cells (Fig. 4E). Experiments with the NF-κB-specific inhibitors PDTC and Bay11-7082 suggested a key role for NF-κB in MRP8/MRP14-induced IP-10 expression (Fig. 4F). Furthermore, the study demonstrated that MRP8/MRP14-activated NF-κB was mediated by TLR4 and TRIF, since MRP8/MRP14 failed to induce p65 phosphorylation in the TLR4^−/− and TRIF^−/− BMDMs (Fig. 4G).

Figure 4.

Functional activation of NF-κB in response to MRP8/MRP14. (A, B) THP-1 cells were stimulated with MRP8/MRP14 (1.5 μg/ml) for 0, 30, 60, 120, and 240 min. Total cellular proteins were extracted for immunoblot analysis. (A) Analysis of phosphorylated and total IKKα/β. (B) Analysis of phosphorylated and total NF-κB p65. (C) THP-1 cells were stimulated with MRP8/MRP14 for the indicated time. Nuclear proteins were extracted to detect the expression of NF-κB p65; lamin was used as the internal control. (D) AMs from WT mice were stimulated with MRP8/MRP14 for 1 h, followed by staining with antibody (Ab) against NF-κB p65 as well as nucleus Hoechst dye (1 μg/ml), and imagings were observed by confocal microscopy. (E) After treatment with MRP8/MRP14 (1.5 μg/ml) or LPS (1 μg/ml) as positive control, THP-1 cells were collected for nuclear protein extraction. EMSA was performed by incubation of nuclear extracts with streptavidin-coated probe for NF-κB p65. The specificity of the assay was ascertained by adding cold probe. Results are representative of 2 independent experiments. (F) THP-1 cells were pretreated with NF-κB-specific inhibitors PDTC (50 μM) or Bay11-7082 (10 μM) for 1 h, followed by the treatment with MRP8/MRP14. IP-10 protein was quantified using Luminex. Data represent mean ± sd from 3 independent experiments. **P < 0.01 vs. untreated group; ^ΔP < 0.05 vs. group treated with MRP8/MRP14. (G) BMDMs from WT, TLR4^−/−, or TRIF^−/− mice were stimulated with MRP8/MRP14 for 1 h; immunoblot analysis was performed to detect phosphorylated NF-κB p65.

MRP8/MRP14 activates IRF3 signaling pathway

The transcription factor IRF3 is involved in the regulation of IFN-stimulated genes by binding the consensus ISRE motif in gene promoter regions (35, 36, 38, 39). We hypothesized that IRF3 signaling is activated by MRP8/MRP14 and subsequently drives the expression of IP-10. Immunoblotting with specific phosphorylation antibodies showed that IRF3 as well as its upstream kinases, TBK1 and IKKε, were significantly phosphorylated at 30 to 60 min after stimulation with MRP8/MRP14 (Fig. 5A–C). IRF3 translocated from cytosol to nucleus by approximately 1 h after stimulation of MRP8/MRP14 (Fig. 5D).

Figure 5.

Activation of IRF3 induced by MRP8/MRP14 is involved in transactivation of IP-10. (A–C) THP-1 cells were treated with MRP8/MRP14 (1.5 μg/ml) for the indicated time. Total cellular proteins were extracted for immunoblot analysis. (A) Phosphorylated and total TBK1. (B) Phosphorylated and total IKKε. (C) Phosphorylated and total IRF3. (D) AMs from WT mice were treated with MRP8/MRP14 for 2 h, followed by double staining with IRF3 Ab and Hoechst dye to visualize IRF3 translocation using confocal microscopy. (E) RAW264.7 cells were transfected by electroporation with IRF3 siRNA or nonspecific siRNA (Cont. siRNA) for 36 h. IRF3 was detected by immunoblot analysis. (F) RAW264.7 cells with or without knocked-down IRF3 were stimulated with MRP8/MRP14 for 6 h. IP-10 mRNA levels were measured with RT-PCR. Graph depicts mean ± sd of percentage changes in IP-10 mRNA expression from three independent experiments. **P < 0.01 vs. untreated group; ^ΔP < 0.05 vs. group treated with MRP8/MRP14. (G) BMDMs were isolated from WT, TLR4^−/−, or TRIF^−/− mice and stimulated with MRP8/MRP14 for 2 h. Immunoblot analysis was performed to detect phosphorylated and total IRF3.

The requirement of IRF3 for the MRP8/MRP14-induced IP-10 expression was tested using siRNA approaches to knockdown IRF3 (approximately 78%) in RAW264.7 cells (Fig. 5E). Figure 5F shows that IRF3 knockdown significantly attenuated IP-10 mRNA expression after MRP8/MRP14 treatment vs. control (P < 0.01). We further demonstrated in Fig. 5G that TLR4 and TRIF deficiency in BMDMs prevented MRP8/MRP14-induced IRF3 phosphorylation.

Synergy between NF-κB and IRF3 is involved in IP-10 induction by MRP8/MRP14

Both κB and ISRE sites have been reported to be critical for IP-10 expression (32–34). We hypothesized that NF-κB and IRF3 synergized for the induction of IP-10 by MRP8/MRP14. Accordingly, we constructed WT or mutant IP-10 promoter-driven luciferase reporter gene constructs in order to evaluate the contribution of these cis elements on IP-10 expression induced by MRP8/MRP14. The WT IP-10 promoter-driven luciferase reporter gene was associated with an apparent response in RAW264.7 cells after stimulation with MRP8/MRP14. In contrast, inhibition of both the NF-κB and IRF3 pathways reduced IP-10 promoter activation induced by MRP8/MRP14 (Fig. 6A).

Figure 6.

Induction of IP-10 by MRP8/MRP14 is mediated by the cooperation between κB and ISRE cis elements. A) RAW264.7 cells were pretreated with or without PDTC, Bay11-7082, or Cay10576 for 1 h, then transfected with plasmid expressing Renilla luciferase (pRL-TK) combined with a firefly-harbored plasmid containing WT IP-10 promoter sequence (−435/+97) (pGL-IP10WT). After transfection for 24 h, cells were treated with MRP8/MRP14 for 6 h, followed by luciferase activity assay with the Dual-Glo Luciferase Assay System. Firefly luciferase activity normalized to Renilla is represented as fold induction over the reporter activity in unstimulated cells. B) RAW264.7 cells were transfected with pRL-TK plus either a plasmid containing IP-10 promoter sequence (−435/+97) (pGL-IP10WT) or that with mutants of κB sites and ISRE, i.e., pGL-IP10 (κB1mut), pGL-IP10 (κB2mut), pGL-IP10 (ISREmut), and pGL-IP10 (κB2/ISREmut). After transfection for 24 h, the cells were treated with MRP8/MRP14 for 6 h, followed by the luciferase activity assay as above. Data represent mean ± sd from 3 independent experiments. **P < 0.01 vs. untreated group; ^ΔΔP < 0.01, ^ΔP < 0.05 vs. the IP-10 reporter gene group treated with MRP8/MRP14.

Further experiments with WT or mutant IP-10 promoter-driven luciferase reporter gene constructs suggested that the ISRE and κB2 in the distal region were necessary for full activation of MRP8/MRP14-induced IP-10 promoter activity (Fig. 6B). Importantly, targeting NF-κB and ISRE together resulted in a much stronger inhibition than targeting either transcription factor alone: the activity of the IP-10 promoter/reporter construct induced by MRP8/MRP14 was nearly abolished in double mutants of κB2 and ISRE (Fig. 6B). These results indicate that MRP8/MRP14-induced IP-10 transcription requires synergy between NF-κB and IRF3 for full activation.

Pivotal role of TLR4-TRIF-IP-10 signaling in CXCR3⁺ T lymphocyte migration is induced by MRP8/MRP14

To address whether the chemoattractant activity of IP-10 plays a role in bridging innate and adaptive immunity after trauma, we evaluated the biologic consequence of MRP8/MRP14-induced IP-10 using an in vitro chemotaxis assay and an in vivo air pouch model.

In these two models, IP-10-induced T lymphocyte migration was assessed. The receptor for IP-10 is CXCR3 (25, 40), and this receptor is highly expressed in activated (CD69⁺) T lymphocytes (40–42). Accordingly, we examined the high expression of CXCR3 in activated T lymphocytes both in in vitro cultured lymphocytes and in in vivo lymphocytes after LPS administration (data not shown).

Via in vitro chemotaxis assay, MRP8/MRP14 dramatically induced the chemotaxis of CXCR3⁺ T lymphocytes compared with control, and IP-10 neutralizing antibody significantly reduced CXCR3⁺ T lymphocytes migration (Fig. 7A). These data suggest that IP-10 induced by MRP8/MRP14 plays an important role in the migration of CXCR3⁺ T lymphocytes. Further experiments with gene-deficient mice showed that either TLR4 or TRIF gene deficiency significantly blocked MRP8/MRP14-induced chemotaxis of CXCR3⁺ T lymphocytes, while RAGE or MyD88 gene knockout failed to show any significant effect on the chemotaxis (Fig. 7B).

Figure 7.

Migration of CXCR3⁺ T lymphocytes induced by MRP8/MRP14 is regulated by TLR4-TRIF-IP-10 signaling axis. A) Chemotactic response of CXCR3⁺ T lymphocytes was explored by the supernatants of BMDMs from WT mice stimulated with MRP8/MRP14 (1.5 μg/ml) in the absence or presence of the IP-10-neutralizing Ab or an isotype control IgG. Migrated cells were counted in five random high-powered fields. B) Chemotaxis assay was performed to detect the chemotactic activity of the supernatants of BMDMs from WT, TLR4^−/−, RAGE^−/−, MyD88^−/−, or TRIF^−/− mice, with or without stimulation of MRP8/MRP14. C) In vivo air pouches were prepared in mice for 5 d; then LPS (5 mg/kg body weight) was administered intraperitoneallySupernatants from MRP8/MRP14-stimulated BMDMs from WT or TLR4^−/− mice were utilized to induce the migration of lymphocytes systemically. An IP-10 neutralizing Ab was injected as indicated to clarify the pivotal role of IP-10 in vivo. Each assay was done in triplicate; data represent mean ± sd from 3 independent experiments. **P < 0.01 vs. untreated group; ^ΔΔP < 0.01, ^ΔP < 0.05 vs. group treated with MRP8/MRP14 alone.

In vivo, MRP8/MRP14 dramatically induced the chemotaxis of CXCR3⁺ lymphocytes compared with control at different time points (8, 16, and 24 h) after LPS administration, and that the chemotaxis occurs most obviously at 16 h. Intrapouch injection of IP-10 neutralizing antibody dramatically reduced the migration of CXCR3⁺ lymphocytes (P < 0.01), whereas the supernatants from BMDMs derived from TLR4^−/− mice failed to induce CXCR3⁺ lymphocyte migration (Fig. 7C). These data further support the hypothesis that MRP8/MRP14 induces IP-10 production through signaling involving TLR4 and TRIF, and that this signaling attracts CXCR3⁺ lymphocytes to sites of infection or injury.

Blockade of MRP8 prevented release of IP-10 in blood, activation of NF-κB and IRF3 in AMs in mouse model of HS/FF

To further determine the contribution of MRP8/MRP14-IP-10 signaling to injury-induced inflammation in vivo, mouse model of HS/FF using MRP8 neutralizing antibody or control IgG was performed to highlight the crucial role of MRP8/MRP14. First, MRP8 neutralizing antibody activity was measured, as shown in Fig. 8A, MRP8 antibody can reduce IP-10 induction stimulated by recombinant MRP8 protein in RAW264.7 cells in a concentration-dependent manner, and a concentration of 1 μg/ml antibody was able to reduce approximately 50% of IP-10 induction. Furthermore, in vivo study showed that high levels of IP-10 were detected in serum of a HS/FF mouse model after resuscitation for 4 h. However, injection of neutralizing antibody against MRP8 dramatically decreased IP-10 release (Fig. 8B) (P < 0.01). Phosphorylation of NF-κB p65 and IRF3 were also partially attenuated by administration of neutralizing antibody in AMs (Fig. 8C–D). These results preliminarily suggest the role of MRP8/MRP14-IP-10 signaling in injury-induced inflammation (Fig. 9).

Figure 8.

Neutralization of MRP8 attenuated production of IP-10, activation of NF-κB and IRF3 in HS/FF mouse model. A) Neutralizing activity assay of MRP8 antibody. IP-10 induction was analyzed by stimulation with recombinant MRP8 protein (1.5 μg/ml) combined with different concentrations of MRP8 antibody or control IgG (0–100 μg/ml). Data represent mean ± sd from three independent experiments. B–D) C57BL/6 mice were used for the HS/FF model. Neutralizing anti-MRP8 polyclonal antibody or control IgG (4 mg/kg body weight) was administered to some animals 2 h before HS/FF model. Blood and AMs were collected at 4 h after resuscitation. B) IP-10 level in the serum was detected using Luminex. Data represent mean ± sd, n = 3. **P < 0.01 vs. sham-treated group without antibody; ^ΔΔP < 0.01 vs. HS/FF group without antibody. C) Phosphorylated and total NF-κB p65 in AMs were detected by immunoblot analysis. (D) Phosphorylated and total IRF3 in AMs were detected by immunoblot analysis. Results are representative of three independent experiments, n = 3.

Figure 9.

Signaling model for IP-10 production induced by MRP8/MRP14. As an important DAMP molecule, MRP8/MRP14 activates TLR4 and its downstream TRIF, and then promotes the activation of both IKKα/β-NF-κB and TBK1-IKKε-IRF3 pathway, initiating the production of IP-10 in monocytes/macrophages. As a chemoattractant, IP-10 attracts CXCR3⁺ lymphocytes to the infection/injury sites, which establishes an in vivo role for IP-10 in MRP8/MRP14-regulated immune responses.

DISCUSSION

In the current study, we demonstrated that blunt trauma patients produce MRP8/MRP14, and that blunt trauma survivors produce more MRP8/MRP14 than nonsurvivors. Furthermore, IP-10, a CXC family chemokine (30, 31), was significantly increased in monocytes/macrophages challenged with MRP8/MRP14. Importantly, we observed a positive correlation in trauma patients between MRP8/MRP14 and IP-10. In other studies from our group, we were able to segregate the outcomes of blunt trauma survivors based on circulating IP-10; specifically, we found that highly elevated IP-10 is a biomarker of morbidity posttrauma (43). Taken together, these studies suggest that underproduction of IP-10 is associated with mortality, but overly elevated IP-10 is also detrimental in that it drives higher levels of organ dysfunction (but not death).

These dichotomous effects of IP-10 have been observed in other contexts as well. One study found that Th1-type immune response induced by IP-10-CXCR3 interaction is beneficial to the control of inflammation (44). However, other studies suggested that CXCR3 exerts deleterious functions by controlling the trafficking of lymphocytes in experimental septic shock (45).

An intriguing question, then, is how IP-10 could mediate these different effects. The current study addresses this question, at least in part. Our air pouch studies suggested that MRP8/MRP14-induced IP-10 can lead to the influx of CXCR3⁺ T cells. Previous studies have shown that CXCR3⁺ T cells can be either proinflammatory or regulatory T cells (40, 44–48). Thus, MRP8/MRP14 released in the setting of blunt trauma may lead to resolution of inflammation when it stimulates IP-10 levels to an appropriate degree. Postinjury release of MRP8/MRP14 below a certain threshold, and consequently inadequate production of IP-10, would associate with overly robust inflammation. However, overly robust production of IP-10, presumably because of overly robust release of MRP8/MRP14 [or because of synergy with other DAMPs such as HMGB1 (49)] would lead to morbidity. Although our studies do not fully elucidate all of these aspects, they do partially support these hypotheses (50, 51).

Although both TLR4 and RAGE were reported to be functional receptors for MRP8/MRP14 (8, 16), we found that MRP8/MRP14 induced IP-10 expression in a TLR4-dependent, but not RAGE-dependent, manner. These studies do not rule out a role for other receptors in MRP8/MRP14-induced IP-10. For example, TLR3 was involved in IP-10 induction in the setting of infection; however, TLR3 is mainly activated by RNA viruses or double-stranded RNA (52, 53). Furthermore, in contrast to TLR4, TLR3 is expressed at a very low level in monocytes/macrophages (54, 55). Thus, the current study supports the notion that TLR4 is the major receptor in macrophages that drives IP-10 induction by MRP8/MRP14.

Upon binding with its cognate ligands, TLR4 activates both MyD88-dependent and TRIF-dependent signaling pathways (35, 36). The MyD88-dependent pathway induces inflammatory responses such as the production of TNF-α and IL-6 via the activation of NF-κB and mitogen-activated protein kinases, whereas the TRIF-dependent pathway controls the expression of chemokines such as type I IFN and RANTES (56–58). This latter effect occurs via the activation of IRF-3 and NF-κB, which is important for the induction of adaptive immune responses (56). The MyD88-dependent pathway was reported to be involved in cytokine release induced by MRP8/MRP14 (8). In the current study, we identify a novel MyD88-independent and TRIF-dependent pathway that controls MRP8/MRP14-induced IP-10 expression in monocytes/macrophages.

NF-κB can be activated through both MyD88-dependent and MyD88-independent pathways. MyD88-dependent TRAF6 complex formation induces the early phase activation of NF-κB, while TRIF-dependent activation of TRAF6 or RIP1 mediates the later activation of NF-κB after challenge with LPS (35, 36). In the present study, we demonstrate a late activation of NF-κB induced by MRP8/MRP14, consistent with the characteristics identified in the TLR4/TRIF-dependent LPS signaling pathway, in which phosphorylation of TBK1 and IKKε are followed.

LPS-induced activation of both IRF3 and NF-κB contributes to the expression of IFN-inducible genes (59). The IP-10 promoter contains 1 ISRE and 2 NF-κB binding sites (κB1 and κB2) (33, 34), and these ISRE and κB sites could contribute to the induction of IP-10 gene either individually or cooperatively. In human astrocytoma cells, IFN-γ-induced IP-10 production appears to occur through the activation of the ISRE (33), while TNF-α-enhanced IP-10 production occurs mainly through the κB2 site (60). More importantly, a synergistic effect on the induction of IP-10 gene was observed in cells costimulated with both IFN-γ and TNF-α which involved both the ISRE and κB sites, indicating that both sites are essential for the optimal induction of IP-10 (34). In agreement with these prior studies, we found that MRP8/MRP14 induced ISRE- as well as κB-dependent induction of IP-10 gene transcription.

One possible limitation of our studies is that the MRP8/MRP14 used was expressed using a prokaryotic expression system and purified from Escherichia coli cultures. This raises the issue of potential LPS contamination, especially for the study of TLR4-mediated signaling process. However, multiple lines of evidence indicate that MRP8/MRP14, and not contaminating LPS, is responsible for the IP-10 production. First, LPS was not detectable in the MRP8/MRP14 preparations by the Limulus amebocyte assay. Second, the effect of MRP8/MRP14 could not be inhibited by polymyxin B (10 μg/ml) at concentrations that efficiently blocked the activity of 100 ng/ml LPS. Third, heat-inactivated MRP8/MRP14 lost the ability to induce IP-10 expression; LPS is not inactivated at the temperatures to denature MRP8/MRP14.

In summary, we demonstrate that blunt trauma patients have elevated circulating MRP8/MRP14, that this DAMP drives IP-10 production, and that both MRP8/MRP14 and IP-10 are drastically lower in nonsurvivors of blunt trauma compared with survivors. We show that lymphocyte trafficking in vivo is controlled by MRP8/MRP14 in a manner that depends on IP-10, and we elucidate the signaling pathways that control IP-10 expression in monocytes/macrophages stimulated by MRP8/MRP14 (Fig. 9). These findings may provide novel diagnostic or therapeutic strategies in the setting of injury.

Glossary

Act-D

actinomycin D

AM

alveolar macrophage

BMDM

bone marrow–derived macrophage

DAMP

damage-associated molecular pattern

HS/FF

hemorrhagic shock and femur fracture

IP-10

IFN-γ inducible protein 10

IRF3

IFN regulatory factor 3

ISRE

IFN-stimulated response element

MRP

myeloid-related protein

PDTC

pyrrolidine dithiocarbamate

PMB

polymyxin B

RAGE

receptor for advanced glycation end products

WT

wild-type