Secondary Logo

Journal Logo

Basic Science Aspects

The Hydroxylase Inhibitor Dimethyloxallyl Glycine Attenuates Endotoxic Shock Via Alternative Activation of Macrophages and IL-10 Production by B1 Cells

Hams, Emily*; Saunders, Sean P.*; Cummins, Eoin P.; O'Connor, Aisling; Tambuwala, Murtaza T.; Gallagher, William M.; Byrne, Annette; Campos-Torres, Antonio; Moynagh, Paul M.; Jobin, Christian§; Taylor, Cormac T.; Fallon, Padraic G.*

Author Information
doi: 10.1097/SHK.0b013e318225ad7e

Abstract

INTRODUCTION

Sepsis is a complex condition characterized by a systemic bacterial infection, which can result in septic shock and multiorgan failure. Despite the use of broad-spectrum antibiotics and improvements in care, sepsis is still associated with a mortality rate of between 30% and 50% in severe cases (1, 2). The immunopathogenesis of sepsis is associated with overwhelming systemic inflammation and suppression of the adaptive immune response. Recent work has identified a high degree of crosstalk between oxygen-sensing pathways, responsible for adaptation to hypoxia, and regulators of innate immunity and inflammation (3). Central to the interplay between hypoxia and the innate response are the hypoxia-inducible factors (HIFs), the oxygen sensitivity of which is conferred by a family of O2-dependent hydroxylases (4). Hypoxia also activates the master regulator of innate immunity and inflammation, nuclear factor κB (NF-κB) through decreased hydroxylase activity (5). A high degree of crosstalk between NF-κB and HIF signaling exists. For example, proinflammatory mediators such as LPS increase HIF mRNA expression in macrophages (MØs) through activation of NF-κB pathways (6, 7). Conversely, NF-κB activation via HIF has been reported in neutrophils (8).

In the context of sepsis, important roles for both HIF and NF-κB have been described in the development of LPS-induced endotoxemia, with a hif1α deletion in the myeloid lineage protective against LPS-induced mortality in vivo (9, 10). In addition, hypoxia is a feature in chronic inflammatory conditions including rheumatoid arthritis, psoriasis, and inflammatory bowel disease (11-13). In various inflammatory conditions, the use and development of prolyl hydroxylase inhibitors, such as dimethyloxallyl glycine (DMOG), in the treatment of disease are showing some benefits, particularly in models of inflammatory bowel disease (13, 14).

Data presented herein will further address the role of inhibiting hydroxylases in the progression of the innate response to LPS challenge. We identify the ability of the hydroxylase inhibitor DMOG to attenuate LPS-induced NF-κB signaling in vivo and show DMOG induced tolerance to subsequent LPS challenge and promotes M2 MØ polarization. We also clearly show the ability of DMOG to upregulate IL-10 expression in models of sepsis and propose that both mechanisms may play important roles in the protective in vivo effects of hydroxylase inhibition in LPS-driven endotoxemia.

MATERIALS AND METHODS

Animals

BALB/c and C57BL/6 mice were purchased from Harlan (Bicester, UK) and Charles River (Wilmington, Mass) respectively. Nuclear factor κB-RE-luc (Oslo) mice, which carry a transgene containing three NF-κB-responsive element sites from the Igκ light chain promoter and a modified firefly luciferase cDNA (Promega pGL-3, Promega, Southampton, UK) on a BALB/c background, were obtained from Caliper LifeSciences (Hopkinton Mass). IL-10-eGFP reporter mice (15) on a C57BL/6 background were obtained from Jackson Laboratories (Bar Harbor, Me) and subsequently backcrossed to BALB/c strain in-house. IL-10−/− and RAG-1−/− mice were from Jackson Laboratories (Bar Harbor) and bred in-house. Nuclear factor κBEGFP reporter mice (16) were bred in-house. Nuclear factor κB p50−/− were purchased from Jackson Laboratories (Bar Harbor). Animals were housed in a specific pathogen-free facility in individually ventilated and filtered cages under positive pressure. All animal experiments were performed in compliance with Irish Department of Health and Children regulations and approved by the Trinity College Dublin's BioResources ethical review board.

Induction of LPS-induced shock

Endotoxin shock was induced in mice via i.p. injection of ultrapure LPS extracted from Escherichia coli (Invivogen, Toulouse, France) at a dose of 10 mg/kg (17). Dimethyloxallyl glycine (Cayman Chemical, Ann Arbor, Mich) was prepared with sterile water and screened for endotoxin contamination (chromogenic LAL; BioWhittaker, Lonza, Switzerland) with less than 0.01 endotoxin units detected per milligram. Mice were injected with DMOG (8 mg/mouse) i.p (13), 2 h before LPS treatment. Animals were assessed as described (17), and scored at hourly intervals using the following criteria: score 0 = no symptoms; score 1 = piloerection, huddling; score 2 = piloerection, huddling, diarrhea; score 3 = lack of interest in surroundings, severe diarrhea; score 4 = decreased movement, listless appearance; and score 5 = loss of self-righting reflex. When mice reach a score of 5, they were humanely killed. To assess the potential role of IL-10, anti-IL-10R monoclonal antibody (mAb) (clone 1B1.3a; ATCC, Manassas, Va), prepared as described (18), was administered i.p. (500 μg) to mice 3 h prior or 3 h after LPS treatment.

Induction of CLP-mediated sepsis

Surgically induced polymicrobial sepsis was introduced in mice using the cecal ligation and puncture (CLP) method. Briefly, the mice were anesthetized, and a small section of the cecum ligated and punctured using a 21-gauge needle to allow leakage of the cecal contents into the peritoneum. After surgery, 1 mL of sterile NaCl was administered subcutaneously. Groups of mice were given DMOG (8 mg/mouse) i.p. at 2 h before surgery and every 24 h subsequently. A sham group of mice received surgery without CLP. The mice were allowed to recover and scored every 12 h using the criteria for LPS-induced endotoxemia, as above. Blood was collected for serum cytokine analysis 3 days after surgery. Pain relief was provided throughout using buprenorphine.

In vivo M2 macrophage activation

Chitin (Sigma-Aldrich, Arklow, Ireland) was washed three times and prepared in phosphate-buffered saline (PBS) by sonication three times for 15 seconds on ice. This was filtered using 70-μM filters and the filtrate used for subsequent immunization. Approximately 800 ng was administered i.p. both alone and in combination with DMOG (8 mg/mouse). Peritoneal cells were collected by lavage after 48 h and cDNA prepared for real-time polymerase chain reaction (PCR) analysis, as outlined below.

In vivo imaging

Mice ubiquitously expressing a transgene encoding firefly luciferase under the control of a concatamer of NF-κB response elements (Caliper Life Sciences, Runcorn, UK) were given LPS and DMOG, as outlined above. Following treatment, the mice were anesthetized and injected i.p. with luciferin (150 mg/kg). In vivo luciferase activity (indicative of NF-κB activation), measured as photons per second per centimeter squared per steradien, was quantified using Living Image Software (V. 3.0.2; Caliper LS).

Cytokine enzyme-linked immunosorbent assays and multiplex cytokine analysis

Supernatants from in vitro cultures and mouse serum were analyzed by sandwich enzyme-linked immunosorbent assays (ELISAs) to quantify the levels of specific cytokines. All cytokines (IL-1β, IL-6, IL-10, and TNF-α) were measured with the DuoSet ELISA development system from R&D Systems (Abingdon, UK) following the manufacturer's protocol. Multiplex analysis was performed on serum samples from affected mice using the proinflammatory 7-plex plate from Meso Scale Discovery ([MSD] Gaithersburg, Md), which measures IL-1β, IL-6, IL-8, IL-10, IL-12p70, IFN-γ, and TNF-α. The protocol was followed as described in the manufacturer's instructions.

Flow cytometry

Surface marker expression and detection of IL-10-eGFP+ and NF-κB-eGFP+ cells were assessed by flow cytometry with data collection on a CyAn (Beckman Coulter, High Wycombe, UK). Data were analyzed using FlowJo software (Tree Star, Ashland, Ore). Cells were stained with eBiosciences mAbs (San Diego, Calif): PerCP anti-CD4 (RM4-5), Pe-Cy7 anti-CD11c (N418), APC anti-F4/80 (BM8), APC anti-B220 (RA3-6B2), and Pacific Blue anti-CD19 (eBio1D3). BD Biosciences mAbs (Oxford, UK): PE anti-CD5 (53-7.3) and PE anti-CD11b (M1/70). Flow buffers used contained 2mM EDTA to exclude doublets. Using appropriate isotype controls, quadrants were drawn, and data plotted on logarithmic scale density or dot plots.

Isolation and culture of peritoneal macrophages and B1 cells

Peritoneal MØs were isolated as described (17). In brief, the peritoneal cavity was lavaged with 5-mL ice-cold PBS. The resulting cells were washed and resuspended in RPMI 1640 (Invitrogen Life Technologies, Paisley, UK) supplemented with 2 mM l-glutamine, 100 U/mL penicillin, and 100 μg/mL streptomycin at a density of 2 × 106 cells/mL. Cells were incubated at 37°C and 5% CO2 for 2 h, and any nonadherent cells removed. The remaining adherent cells were stimulated with DMOG (1 mM) 2 h before stimulation with LPS (10 μg/mL) for between 5 and 30 min, after which cells were collected for protein extraction. In addition, adherent peritoneal cells were cultured for 2 h with DMOG (1 mM) followed by subsequent stimulation with LPS (10 μg/mL) and IFN-γ (10 ng/mL), granulocyte macrophage colony-stimulating factor (GM-CSF) (10 ng/mL), IL-4 (10 ng/mL) and IL-13 (10 ng/mL), or IL-10 (10 ng/mL). After an 18-h incubation, cells were collected, and cDNA prepared for subsequent real-time PCR.

B1 cells were FACS sorted from total peritoneal cells using positive selection of Pacific blue anti-CD19+ (eBio1D3) and PE anti-CD5+ (53-7.3) cells. Cells were cultured at 5 × 106 cells/mL in the presence of DMOG (1 mM) for 2 h followed by stimulation with LPS (10 μg/mL) for 15 and 30 min, after which cells were collected for protein extraction.

RNA isolation and real-time PCR

RNA was isolated from peritoneal exudate cells using the RNeasy kit and reverse transcribed using the Quantitect reverse transcription kit incorporating a genomic DNA elimination step (QIAGEN, Crawley, UK). Real-time quantitative PCR was performed on an ABI Prism 7900HT sequence detection system (Applied Biosystems, Paisley, UK) using predesigned TaqMan gene expression assays specific for murine iNOS (Mm00440485_m1), arginase 1 (Mm00475988_m1), Relm-α (Mm00445109_m1), and YM1 (Mm00657889_m1). Specific gene expression was normalized to murine glyceraldehyde 3-phosphate dehydrogenase (GAPDH). Fold expression was calculated using the ΔΔCT method of analysis (Applied Biosystems).

Protein extraction and immunoblotting

Whole cell extracts were prepared from peritoneal MØs following cell lysis in RIPA lysis buffer. Protein concentrations of the samples were quantified using the commercially available BCA kit (Thermo Scientific, Loughborough, UK). Cell extracts (10 μg) were separated by sodium dodecyl sulfate polyacrylamide gel electrophoresis on 12% polyacrylamide gels and transferred to PVDF membrane (Millipore, Billarica, Mass). Membranes were probed with rabbit polyclonal antibodies specific for total (H-119) and phosphorylated p50 (Ser 337), total (C-20) and phosphorylated p65 (Ser 311), total (48H2) and phosphorylated cAMP response element-binding (CREB) (Ser 133), and GAPDH (D-6), followed by incubation with a horseradish peroxidase-conjugated goat anti-rabbit secondary antibody (Dako, Stockport, UK). Immunolabeled proteins were visualized by chemiluminescence using the ECL detection system (Millipore). Quantification of protein expression from three blots representative of three individual experiments was performed by densitometry using ImageJ software (NIH, Bethesda, Md).

Statistical analysis

Statistical analysis was performed using GraphPad Instat (San Diego, Calif). Results are presented as mean ± SEM. Differences, indicated as two-tailed P values, were considered significant when P > 0.05 as assessed by unpaired Student's t test with Welch correction applied as necessary or log-rank correlation for survival statistics.

RESULTS

DMOG preconditioning promotes basal but inhibits LPS-induced NF-κB activity

We have previously demonstrated using cell lines that pharmacologic or genetic inhibition of HIF hydroxylases leads to activation of NF-κB signaling (5). Although the regulation of NF-κB by hydroxylases has been confirmed both in vitro and in vivo (6, 19-21), specific mechanisms and targets remain unknown. In addition, there are few existing data on the regulatory role of hydroxylases in controlling NF-κB activation by innate immune receptors such as Toll-like receptors (TLRs). With this in mind, we evaluated the interplay between hydroxylases and the TLR4 signaling pathway as activated by LPS. Following in vivo treatment of NF-κBEGFP mice with the hydroxylase inhibitor DMOG, there is activation of basal NF-κB in peritoneal MØs, neutrophils, and B cells (Fig. 1; Figure, Supplemental Digital Content 1, https://links.lww.com/SHK/A81). Paradoxically, although LPS treatment evoked NF-κB activation of peritoneal cell populations, cells from mice pretreated with DMOG before LPS exposure demonstrated significantly (P < 0.05-0.01) lower NF-κB activity (Fig. 1). This led us to hypothesize that the low degree of NF-κB activation induced by DMOG leads to a preconditioning effect that renders these cells tolerant to subsequent LPS exposure, similar to reported mechanisms of LPS tolerance (22). Because of the ability of DMOG alone to activate NF-κB, we tested for endotoxin contamination (detected <0.01 endotoxin units per milligram) and confirmed that DMOG activation of NF-κB is independent of TLR4 using cells from TLR4−/− mice (data not shown). To test this hypothesis in vivo, LPS-induced endotoxic shock was evoked in mice expressing luciferase under the control of a concatamer of NF-κB response elements, both with and without prior DMOG treatment, and activation of NF-κB was monitored by in vivo imaging. Consistent with the peritoneal cell data described above (Fig. 1), LPS treatment induced systemic NF-κB activity, which was significantly reduced in mice pretreated with DMOG (Fig. 2). These data were quantified by in vivo luminometry and demonstrate a significant (P < 0.01) reduction in systemic NF-κB activity in DMOG-pretreated mice compared with mice treated with LPS-alone (Fig. 2). Previous work by the group has assessed the activation of NF-κB in response to i.p. administration of LPS and found systemic activation affecting all organs (5, 13, 23). These data promoted the hypothesis that DMOG pretreatment induces a mild NF-κB activation in vivo, which tolerizes mice to subsequent LPS challenge, with decreased systemic NF-κB activation.

F1-13
Fig. 1:
Dimethyloxallyl glycine regulates LPS-induced NF-κB induction in peritoneal cells. Nuclear factor κB activation in peritoneal MØs (F4/80hiCD11bhi), neutrophils (L7-6GhiCD11bhi), B1 cells (CD19+veCD5+ve), and B2 cells (CD19+veCD5−ve), determined by NF-κB eGFP expression. Mice were treated with DMOG (8 mg/mouse), LPS (10 mg/kg), and DMOG (8 mg/mouse) 2 h before LPS (10 mg/kg); peritoneal cells were collected 1 h after LPS treatment. Data are representative of three separate experiments with five mice per group (mean ± SEM, *P > 0.05).
F2-13
Fig. 2:
Dimethyloxallyl glycine attenuates LPS-induced NF-κB activation in vivo. Nuclear factor κB activation visualized by luciferase activity, with the color indicative of the level of NF-κB activation, was assessed in whole adult mice in response to LPS stimulation with and without DMOG treatment with quantification of luciferase activity in response to i.p. injection of luciferin (150 mg/kg). Data are representative of three separate experiments with five mice per group (mean ± SEM, *P > 0.05, **P < 0.001).

DMOG treatment ameliorates LPS-mediated endotoxin shock

We next investigated the in vivo consequences of DMOG treatment of mice before LPS-induced endotoxin shock. Mice were pretreated with DMOG for 2 h before LPS administration to induce endotoxin shock. In mice treated with LPS alone, signs of disease peaked at 12 h (data not shown), and all mice were dead by 72 h (Fig. 3A). In mice pretreated with DMOG, there was marked delayed disease onset with a significant (P < 0.05) reduction in LPS-induced mortality (Fig. 3A). This protective effect of DMOG was associated with significantly (P < 0.05) decreased levels of the acute phase protein, serum amyloid A (Fig. 3B) and the proinflammatory cytokine TNF-α (Fig. 3C). LPS-induced IL-1β and IL-6 were not attenuated with prior DMOG treatment (Fig. 3, D and E). In contrast to TNF-α, DMOG-treated mice had significantly enhanced (P > 0.05) LPS-induced expression of the anti-inflammatory cytokine IL-10 compared with mice treated with LPS alone (Fig. 3F). It is noteworthy that when mice were treated with DMOG after LPS injection, the inhibitor had no effect on the development of the endotoxic shock (data not shown). Therefore, in vivo pretreatment of mice with DMOG protects mice from LPS-induced shock.

F3-13
Fig. 3:
Dimethyloxallyl glycine decreases mortality in LPS-induced endotoxin shock. A, Survival of mice receiving LPS-induced shock with and without DMOG. Shock was induced in groups of male BALB/c mice by i.p. injection of 10 mg/kg LPS; groups received either LPS alone or 8 mg DMOG 2 h before LPS. Mice were scored every 3 h for 24 h. B, Levels of serum amyloid A (SAA) were quantified by ELISA at 3 h after LPS treatment. C-F, Serum cytokines (TNF-α [C], IL-1β [D], IL-6 [E], and IL-10 [F]) were quantified by MSD multiplex cytokine array 3 h after LPS treatment. Data are representative of three separate experiments of five mice per group (mean ± SEM, *P > 0.05).

DMOG promotes macrophage M2 polarization

Macrophages are the predominant cell type in the peritoneal cavity and play an important role in the initial response to LPS. LPS stimulation evokes classic (M1) activation of MØs (24); however, it has previously been shown that activation of the alternative (M2) macrophage phenotype is associated with LPS tolerance (25). Because of the apparent tolerizing role for DMOG in attenuating NF-κB activation in response to LPS, we assessed the potential for DMOG to drive M1 or M2 polarization in vivo. Mice were injected i.p. with LPS or M2 phenotype-inducing chitin (26), and the relative expressions of M1 (iNOS; Fig. 4A) and M2 (arginase 1, Relm-α, YM1; Fig. 4B) markers on peritoneal MØs were analyzed by quantitative PCR (qPCR). In mice treated with chitin, YM1, arginase 1, and Relm-α were significantly elevated (P < 0.05) compared with control PBS-injected mice, consistent with generation of an M2 polarization (26). Dimethyloxallyl glycine treatment with chitin significantly (P < 0.05-0.01) enhanced the levels of M2 genes (Fig. 4B). Conversely, although peritoneal MØs isolated from mice treated with DMOG alone showed an upregulation of the M1 marker iNOS, expression is significantly decreased in mice treated with DMOG and LPS compared with mice treated with LPS alone (Fig. 4A). To further confirm that DMOG induced M2 cell expansion, we used an in vitro model and exposed peritoneal MØs to DMOG, followed by GM-CSF or a combination of LPS and IFN-γ to drive M1 polarization, and IL-10 or a combination of IL-4 and IL-13 to drive M2 polarization (Fig. 4, C and D). Consistent with the in vivo data, in an in vitro system, DMOG inhibited MØ polarization to an M1 phenotype (Fig. 4C), while promoting M2 polarization (Fig. 4D).

F4-13
Fig. 4:
Dimethyloxallyl glycine promotes M2 polarization in peritoneal macrophages. mRNA expression of M1 (A) and M2 (B) markers in peritoneal cells. Male BALB/c mice were given DMOG (8 mg/mouse) 2 h before LPS (10 mg/kg) or chitin i.p. Peritoneal cells were collected by lavage 6 h after LPS treatment and 48 h after chitin treatment and cDNA prepared. mRNA expression of cultured peritoneal macrophages polarized to M1 by either GM-CSF (10 ng/mL) or a combination of LPS (10 μg/mL) and IFN-γ (10 ng/mL) (C) and M2 by either IL-10 (10 ng/mL) or a combination of IL-4 (10 ng/mL) and IL-13 (10 ng/mL) (D). Cells were stimulated with DMOG for 2 h before polarization overnight; cDNA was prepared from the collected cells. Reverse transcriptase-PCR was run using probes for Arg-1, Relm-α, YM1, and iNOS. Data expressed as fold expression were calculated comparing to GAPDH as a housekeeping gene. Data are representative of three independent experiments (mean ± SEM, *P < 0.05).

Although these data are indicative that DMOG enhances M2 cell expansion in vivo in response to chitin, and also in vitro, which is consistent with DMOG evoking a state of LPS tolerance, it does not fully define the mechanism. It has recently been shown that activation of the p50 NF-κB subunit can inhibit M1 and drive M2 macrophage activation, a phenotype that is associated with LPS tolerance (27). Therefore, p50 activation in peritoneal MØs treated with DMOG and/or LPS was tested. There was an upregulation of phosphorylated p50 (S337) in MØs treated with DMOG alone, and levels are further increased 5 min after cells are additionally treated with LPS (Fig. 5, A and B). In contrast, there was no difference in phosphorylation of the p65 NF-κB subunit, which would be more associated with classic M1 activation (Fig. 5, A and B). To confirm these in vitro observations, we induced endotoxic shock with and without DMOG treatment in p50-deficient mice. Although DMOG protected wild-type (WT) mice from LPS-induced endotoxic shock, DMOG did not reduce mortality of LPS-treated p50−/− mice (Fig. 6A). In addition, although DMOG modulated the serum levels of TNF-α and IL-10 in WT mice, it had no effect on these cytokines in LPS-treated p50−/− mice (Fig. 6B). Furthermore, reverse transcriptase-PCR analysis of the peritoneal cells from p50−/− mice showed no upregulation of any of the typical genes associated with M2 macrophage polarization, consistent with the importance of p50 in generating an M2 phenotype (data not shown). As DMOG inhibits prolyl hydroxylases, we also looked at the expression of HIF-1α in response to DMOG and LPS in peritoneal macrophages. As expected, stimulation of the cells with DMOG resulted in an increase in HIF-1α and HIF-2, with no apparent effect of LPS (data not shown). These data are consistent with DMOG treatment inducing tolerance to LPS via upregulation of p50 NF-κB and an expansion of M2 cells.

F5-13
Fig. 5:
Dimethyloxallyl glycine upregulates p50 in peritoneal MØs in vitro. A, Whole cell extracts were prepared from peritoneal MØs cultured with DMOG (1 mM) for 2 h before LPS (10 ng/mL) for 5, 15, and 30 min. Western blots were run with antibodies against total p50 and phosphorylated (P) p50, total p65, and phosphorylated (P) p65; GAPDH was used as a loading control. Densitometry was performed comparing each probe to the corresponding GAPDH band (B). Data are representative of three to five mice per group (mean ± SEM, *P > 0.05).
F6-13
Fig. 6:
Dimethyloxallyl glycine treatment does not protect p50−/− mice from LPS-induced endotoxemia. A, Survival of p50−/− and WT mice receiving LPS-induced shock with and without DMOG. Shock was induced in groups of mice by i.p. injection of 10 mg/kg LPS; groups received either LPS alone or 8 mg DMOG 2 h before LPS. Mice were scored every 3 h for 96 h. B, Levels of serum TNF-α and IL-10 were quantified by ELISA at 3 h after LPS treatment. Data are representative of two separate experiments of five mice per group (mean ± SEM, *P < 0.05).

IL-10 receptor blockade ameliorates the protective role of DMOG in endotoxic shock

Dimethyloxallyl glycine increased serum levels of the anti-inflammatory cytokine IL-10 after LPS treatment (Figs. 3F, 6A). To determine if the elevated IL-10 levels induced by DMOG were relevant to the protective effects of DMOG in LPS-induced endotoxic shock, we used an mAb to block the IL-10 receptor (IL-10R) in vivo. Before LPS challenge, the anti-IL-10R mAb was injected to mice either alone or in combination with DMOG. Whereas DMOG alone rendered mice refractory to LPS-induced mortality, cotreatment with anti-IL-10R mAb blocked the protective effects of DMOG on disease severity, with 100% mortality observed at 12 h following induction of sepsis, compared with 0% mortality seen in mice treated with DMOG and LPS alone (Fig. 7A). In mice treated with anti-IL-10R mAb, there was comparable LPS-induced death as DMOG plus mAb-treated mice (Fig. 7A). The effects of blocking IL-10 signaling with DMOG treatment were functional only when anti-IL-10R mAb was administered before treatment. Mice treated with anti-IL-10R mAb 3 h following DMOG and LPS challenge had a comparable response as mice treated with LPS alone (data not shown). Serum cytokine levels were quantified at 3 h following LPS exposure (Fig. 7, B and C). Similar to earlier studies (Fig. 3), DMOG treatment decreased levels of TNF-α and increased IL-10 expression in response to LPS (Fig. 3, B and C). Anti-IL-10R mAb treatment increased TNF-α expression, which was reduced in response to DMOG treatment (Fig. 7B). IL-10 expression is also increased in response to anti-IL-10R blockade, consistent with accumulation of IL-10 after blocking of its receptor (Fig. 7C). However, IL-10 expression is significantly (P < 0.05) increased after additional treatment with DMOG (Fig. 7C), confirming that DMOG evokes IL-10 release. Furthermore, this essential role for IL-10 in DMOG-mediated protection from LPS-induced shock was confirmed with IL-10−/− animals (data not shown).

F7-13
Fig. 7:
Protective effects of DMOG in LPS-induced endotoxin shock are blocked in mice treated with an IL-10R-blocking antibody. A, Survival of mice receiving an IL-10 receptor-blocking antibody before LPS-induced shock with and without DMOG. Shock was induced in groups of male BALB/c mice by i.p. injection of 10 mg/kg LPS; groups received either anti-IL-10R alone, LPS alone, 8 mg DMOG 2 h before LPS, or anti-IL-10R 3 h and DMOG 2 h before LPS treatment. Mice were scored every 3 h for 24 h. B and C, Serum cytokines (TNF-α and IL-10) were quantified by MSD multiplex cytokine array 3 h after treatment with LPS. Data are representative of three to five mice per group (mean ± SEM, *P > 0.05)

To formally define the cellular source of IL-10 in DMOG-treated mice, we initially induced LPS-mediated shock in both untreated and DMOG-treated RAG-1−/− mice, which lack T and B cells. Whereas DMOG-treated WT mice were tolerant to LPS-induced shock, treatment of RAG-1−/− mice with DMOG caused no alteration in severity (data not shown). Consistent with DMOG stimulating IL-10 release from a T or B lymphocyte source, DMOG-treated RAG-1−/− mice had comparable IL-10 expression as untreated mice (Fig. 8A). To further define the cellular source of IL-10, we used IL-10-eGFP mice to determine IL-10-eGFP expression in peritoneal MØs and B1 and B2 cells (Fig. 8B; Figure, Supplemental Digital Content 2, https://links.lww.com/SHK/A82). Dimethyloxallyl glycine treatment of mice significantly (P < 0.05) enhances LPS-induced IL-10-eGFP expression by peritoneal B1 cells, a known cellular source of IL-10 in LPS challenge (28), but not by MØs or B2 cells (29, 30), indicating for the first time a cell type-specific effect of DMOG (Fig. 8B).

F8-13
Fig. 8:
The cellular source of IL-10 in response to DMOG is peritoneal B1 cells. A, IL-10 levels in RAG mice after LPS-induced shock. Shock was induced in groups of male RAG mice by i.p. injection of 10 mg/kg LPS. Groups received either PBS or DMOG (8 mg/mouse) 2 h before LPS treatment. Serum was collected 3 h after LPS treatment, and IL-10 levels were quantified by ELISA. B, IL-10 expression in peritoneal MØs (F4/80hiCD11bhi), B1 cells (CD19+veCD5+ve), and B2 cells (CD19+veCD5−ve), determined by IL-10 eGFP expression and presented as mean fluorescence intensity (MFI). Mice were treated with DMOG (8 mg/mouse), LPS (10 mg/kg), and DMOG (8 mg/mouse) 2 h before LPS (10 mg/kg); peritoneal cells were collected 2 h after LPS treatment. C, Whole cell extracts were prepared from peritoneal MØs and isolated B1 cells cultured with DMOG (1 mM) for 2 h before LPS (10 ng/mL) for 5 and 15 min. Western blots were run with antibodies against total CREB and phosphorylated (P) CREB; GAPDH was used as a loading control. Densitometry was performed for B1 cells (D) and macrophages (E), comparing each probe to the corresponding GAPDH band. Data are representative of three to five mice per group (mean ± SEM, *P > 0.05).

CREB has been implicated in regulating genes associated with inflammation, including IL-10, and also hypoxia and so could provide one potential mechanism linking the effects of DMOG in systemic inflammation. Analysis of in vitro CREB activation in the peritoneal B1 cell population shows upregulation of phosphorylated CREB in response to treatment with DMOG (Fig. 8, C and D), which may explain one potential mechanism by which DMOG is acting. This effect appears to be cell specific, as peritoneal macrophages do not show any significant activation of CREB in response to DMOG (Fig. 8, C and E).

DMOG treatment during polymicrobial sepsis is associated with increased disease severity

To assess the efficacy of DMOG in a clinical sepsis model, we used the CLP model in both untreated and mice treated with DMOG. Following CLP in untreated mice, there are signs of disease within 12 h after surgery compared with sham-operated mice, with death of animals occurring within 7 days of surgery (Fig. 9A). However, mice receiving DMOG treatment before surgery develop significant exacerbation of disease symptoms and demonstrate a significantly (P < 0.01) increased mortality rate when compared with those receiving surgery alone, with 100% mortality at 96 h after surgery (Fig. 9A). However, serum cytokine analysis also shows a decrease in TNF-α and increase in IL-10 in mice treated with DMOG (Fig. 9B), consistent with the effects of DMOG in the endotoxemia model and suggesting DMOG may be using the same mechanism in the two models.

F9-13
Fig. 9:
Dimethyloxallyl glycine exaggerates disease and morbidity in CLP-induced sepsis. A, Survival of mice receiving CLP-induced sepsis with and without DMOG. Sepsis was induced in groups of male BALB/c mice by ligation and puncture of a small section of the cecum; groups received either surgery alone or received 8 mg DMOG 2 h before surgery. Mice were scored every 12 h for 120 h. B, Serum cytokine levels in mice 72 h after surgery. Data are representative of 19 to 21 mice per group (mean ± SEM, *P > 0.05).

DISCUSSION

There are overlaps in pathways active in hypoxia and inflammation (31). Previous studies have clearly shown that inhibition of hydroxylases by DMOG not only affects the HIF proteins but can also upregulate NF-κB, an observation confirmed in this report, thereby showing a distinct link between the pathways (13). In this study, we have further assessed the potential role for response evoked by hypoxia in the regulation of innate immune activation. We have shown that the hydroxylase inhibitor DMOG appears to tolerize cells to LPS activation and promote M2 polarization and subsequent upregulation of IL-10 by peritoneal B1 cells, resulting in decreased susceptibility to endotoxemia.

Previous work has outlined the link between hypoxia and inflammatory conditions such as models of colitis, ischemia, and certain cancers, with groups showing potential benefits using DMOG in the treatment of these conditions (5, 13, 32). A variety of mechanisms of action have been demonstrated for the action of DMOG, with groups showing regulation of cytokine production and adhesion molecules, apoptosis, and NF-κB tolerance by DMOG (33, 34). Our data suggests two mechanisms through which DMOG is working in preventing endotoxemia: first, through tolerizing the peritoneal macrophages to NF-κB activation and promoting the generation of M2 MØs, thereby attenuating the proinflammatory responses to LPS, and second, direct upregulation of IL-10 production, involving CREB activation as an example, specifically from peritoneal B1 cells. We appreciate that CREB is just one possible mechanism linking hypoxia and IL-10, and to fully delineate the signaling underlying this response, it would also be necessary to assess activation of other transcription factors such as AP-1 and Erk. The upregulation of NF-κB observed in response to DMOG treatment is likely to be a direct activation of the p50 subunit. Pretreatment with DMOG has the effect of tolerizing the mice to be refractory to subsequent challenge with LPS. In addition, recent work has shown that activation of the p50 NF-κB subunit actually inhibits M1 and drives M2 macrophage activation, a phenotype also associated with LPS tolerance (27). This has been demonstrated herein by the ability of DMOG to enhance M2 polarization by chitin, a known inducer of M2 activation (26).

Peritoneal B1 cells are innate-type cells, which are classically associated with production of IgM and IL-10 in response to infection or inflammatory stimuli. Recent work has demonstrated that these cells are important in LPS-induced endotoxemia, with xid B1 cell-deficient mice showing higher mortality than WT mice (29). This suggests that the B1-mediated IL-10 expression is important in protecting the mice against endotoxemia, which would support our data showing that the upregulation of IL-10 by DMOG plays a role in attenuating disease. These mechanisms support a protective immunoregulatory role for DMOG in a model of endotoxin shock, resulting in decreased disease severity. Other studies have recently also linked these two observations demonstrating that B cells have the ability to drive macrophage polarization, specifically that IL-10-producing B1 cells are able to induce M2 macrophage activation (35).

This project addresses the innate response to LPS-induced inflammatory signaling, with DMOG modulating innate cell activation and downstream responses. As shown, DMOG upregulates expression of both NF-κB and HIF-1α; it would therefore be important in the future to define the relative roles of NF-κB and HIF-1α in LPS tolerance (6, 7). This would provide further mechanistic insights into the role of DMOG in innate activation. The experiments using polymicrobial sepsis, however, demonstrate that although DMOG may attenuate LPS-induced systemic inflammation, it exacerbates the effects of systemic infection despite appearing to use the same mechanism. Polymicrobial sepsis, achieved by ligation and puncture of a section of the cecum, provides a more accurate model relevant to clinical sepsis. In this case, sepsis is caused by gradual leakage of cecal contents into the peritoneum, giving a model similar to sepsis caused by severe abdominal trauma (36). The cecal contents comprise a variety of bacterial flora resulting in activation of a variety of TLRs and other pattern recognition receptors; therefore, we hypothesize that although DMOG appears to be using the same mechanism in both models, because of the polymicrobial nature of the condition, this fails to protects against disease symptoms.

In summary, we have shown the hydroxylase inhibitor DMOG to be protective against acute systemic inflammation induced by the aseptic introduction of LPS leading to endotoxic shock. We also provide further insight as to the mechanisms used by DMOG and as such show the mechanistic interplay between the pathways used by hypoxia and those used during innate activation. Studies have already demonstrated some benefit in using DMOG in more chronic inflammatory conditions; the interplay between cellular regulation in response to hypoxia or innate immunity presented herein highlights this as an area for further research into the potential benefits of blocking prolyl hydroxylases in more acute inflammatory conditions.

REFERENCES

1. Vincent JL, Sakr Y, Sprung CL, Ranieri VM, Reinhart K, Gerlach H, Moreno R, Carlet J, Le Gall JR, Payen D: Sepsis in European intensive care units: results of the SOAP study. Crit Care Med 34:344-353, 2006.
2. Danai PA, Sinha S, Moss M, Haber MJ, Martin GS: Seasonal variation in the epidemiology of sepsis. Crit Care Med 35:410-415, 2007.
3. Taylor CT: Interdependent roles for hypoxia inducible factor and nuclear factor-kappaB in hypoxic inflammation. J Physiol 586:4055-4059, 2008.
4. Kaelin WG Jr, Ratcliffe PJ: Oxygen sensing by metazoans: the central role of the HIF hydroxylase pathway. Mol Cell 30:393-402, 2008.
5. Cummins EP, Berra E, Comerford KM, Ginouves A, Fitzgerald KT, Seeballuck F, Godson C, Nielsen JE, Moynagh P, Pouyssegur J, et al.: Prolyl hydroxylase-1 negatively regulates IkappaB kinase-beta, giving insight into hypoxia-induced NFkappaB activity. Proc Natl Acad Sci U S A 103:18154-18159, 2006.
6. Frede S, Stockmann C, Freitag P, Fandrey J: Bacterial lipopolysaccharide induces HIF-1 activation in human monocytes via p44/42 MAPK and NF-kappaB. Biochem J 396:517-527, 2006.
7. Rius J, Guma M, Schachtrup C, Akassoglou K, Zinkernagel AS, Nizet V, Johnson RS, Haddad GG, Karin M: NF-kappaB links innate immunity to the hypoxic response through transcriptional regulation of HIF-1alpha. Nature 453:807-811, 2008.
8. Walmsley SR, Print C, Farahi N, Peyssonnaux C, Johnson RS, Cramer T, Sobolewski A, Condliffe AM, Cowburn AS, Johnson N, et al.: Hypoxia-induced neutrophil survival is mediated by HIF-1alpha-dependent NF-kappaB activity. J Exp Med 201:105-115, 2005.
9. Peyssonnaux C, Cejudo-Martin P, Doedens A, Zinkernagel AS, Johnson RS, Nizet V: Cutting edge: essential role of hypoxia inducible factor-1alpha in development of lipopolysaccharide-induced sepsis. J Immunol 178:7516-7519, 2007.
10. Thiel M, Caldwell CC, Kreth S, Kuboki S, Chen P, Smith P, Ohta A, Lentsch AB, Lukashev D, Sitkovsky MV: Targeted deletion of HIF-1alpha gene in T cells prevents their inhibition in hypoxic inflamed tissues and improves septic mice survival. PLoS One 2:e853, 2007.
11. Hitchon CA, El-Gabalawy HS: Oxidation in rheumatoid arthritis. Arthritis Res Ther 6:265-278, 2006.
12. Rosenberger C, Solovan C, Rosenberger AD, Jinping L, Treudler R, Frei U, Eckardt KU, Brown LF: Upregulation of hypoxia-inducible factors in normal and psoriatic skin. J Invest Dermatol 127:2445-2452, 2007.
13. Cummins EP, Seeballuck F, Keely SJ, Mangan NE, Callanan JJ, Fallon PG, Taylor CT: The hydroxylase inhibitor dimethyloxalylglycine is protective in a murine model of colitis. Gastroenterology 134:156-165, 2008.
14. Hirota SA, Fines K, Ng J, Traboulsi D, Lee J, Ihara E, Li Y, Willmore WG, Chung D, Scully MM, et al.: Hypoxia-inducible factor signaling provides protection in Clostridium difficile-induced intestinal injury. Gastroenterology 139:259-269 e253, 2011.
15. Kamanaka M, Kim ST, Wan YY, Sutterwala FS, Lara-Tejero M, Galan JE, Harhaj E, Flavell RA: Expression of interleukin-10 in intestinal lymphocytes detected by an interleukin-10 reporter knockin tiger mouse. Immunity 25:941-952, 2006.
16. Magness ST, Jijon H, Van Houten Fisher N, Sharpless NE, Brenner DA, Jobin C: In vivo pattern of lipopolysaccharide and anti-CD3-induced NF-kappa B activation using a novel gene-targeted enhanced GFP reporter gene mouse. J Immunol 173:1561-1570, 2004.
17. Saunders SP, Barlow JL, Walsh CM, Bellsoi A, Smith P, McKenzie AN, Fallon PG: C-type lectin SIGN-R1 has a role in experimental colitis and responsiveness to lipopolysaccharide. J Immunol 184:2627-2637, 2010.
18. Mangan NE, Fallon RE, Smith P, van Rooijen N, McKenzie AN, Fallon PG: Helminth infection protects mice from anaphylaxis via IL-10-producing B cells. J Immunol 173:6346-6356, 2004.
19. Chan G, Bivins-Smith ER, Smith MS, Yurochko AD: NF-kappaB and phosphatidylinositol 3-kinase activity mediates the HCMV-induced atypical M1/M2 polarization of monocytes. Virus Res 144:329-333, 2009.
20. Xue J, Thippegowda PB, Hu G, Bachmaier K, Christman JW, Malik AB, Tiruppathi C: NF-kappaB regulates thrombin-induced ICAM-1 gene expression in cooperation with NFAT by binding to the intronic NF-kappaB site in the ICAM-1 gene. Physiol Genomics 38:42-53, 2009.
21. Takeda K, Ichiki T, Narabayashi E, Inanaga K, Miyazaki R, Hashimoto T, Matsuura H, Ikeda J, Miyata T, Sunagawa K: Inhibition of prolyl hydroxylase domain-containing protein suppressed lipopolysaccharide-induced TNF-alpha expression. Arterioscler Thromb Vasc Biol 29:2132-2137, 2009.
22. Biswas SK, Lopez-Collazo E: Endotoxin tolerance: new mechanisms, molecules and clinical significance. Trends Immunol 30:475-487, 2009.
23. Fitzpatrick SF, Tambuwala MM, Bruning U, Schaible B, Scholz CC, Byrne A, O'Connor A, Gallagher WM, Lenihan CR, Garvey JF, et al.: An intact canonical NF-kappaB pathway is required for inflammatory gene expression in response to hypoxia. J Immunol 186:1091-1096, 2011.
24. Gordon S, Taylor PR: Monocyte and macrophage heterogeneity. Nat Rev Immunol 5:953-964, 2005.
25. Mantovani A, Locati M: Orchestration of macrophage polarization. Blood 114:3135-3136, 2009.
26. Satoh T, Takeuchi O, Vandenbon A, Yasuda K, Tanaka Y, Kumagai Y, Miyake T, Matsushita K, Okazaki T, Saitoh T, et al.: The Jmjd3-Irf4 axis regulates M2 macrophage polarization and host responses against helminth infection. Nat Immunol 11:936-944, 2011.
27. Nizet V, Johnson RS: Interdependence of hypoxic and innate immune responses. Nat Rev Immunol 9:609-617, 2009.
28. Siewe L, Bollati-Fogolin M, Wickenhauser C, Krieg T, Muller W, Roers A: Interleukin-10 derived from macrophages and/or neutrophils regulates the inflammatory response to LPS but not the response to CpG DNA. Eur J Immunol 36:3248-3255, 2006.
29. Barbeiro DF, Barbeiro HV, Faintuch J, Ariga SK, Mariano M, Popi AF, de Souza HP, Velasco IT, Soriano FG: B-1 cells temper endotoxemic inflammatory responses. Immunobiology 216:302-308, 2011.
30. Sica A, Porta C, Riboldi E, Locati M: Convergent pathways of macrophage polarization: the role of B cells. Eur J Immunol 40:2131-2133, 2011.
31. Eltzschig HK, Carmeliet P: Hypoxia and inflammation. N Engl J Med 364:656-665, 2011.
32. Zhao HX, Wang XL, Wang YH, Wu Y, Li XY, Lv XP, Zhao ZQ, Zhao RR, Liu HR: Attenuation of myocardial injury by postconditioning: role of hypoxia inducible factor-1alpha. Basic Res Cardiol 105:109-118, 2011.
33. Frede S, Stockmann C, Winning S, Freitag P, Fandrey J: Hypoxia-inducible factor (HIF) 1alpha accumulation and HIF target gene expression are impaired after induction of endotoxin tolerance. J Immunol 182:6470-6476, 2009.
34. Hindryckx P, De Vos M, Jacques P, Ferdinande L, Peeters H, Olievier K, Bogaert S, Brinkman B, Vandenabeele P, Elewaut D, et al.: Hydroxylase inhibition abrogates TNF-{alpha}-induced intestinal epithelial damage by hypoxia-inducible factor-1-dependent repression of FADD. J Immunol 185:6306-6316, 2011.
35. Alexander C, Rietschel ET: Bacterial lipopolysaccharides and innate immunity. J Endotoxin Res 7:167-202, 2001.
36. Medina E: Murine model of polymicrobial septic peritonitis using cecal ligation and puncture (CLP). Methods Mol Biol 602:411-415, 2011.
Keywords:

LPS; hypoxia; mouse; NF-κB; tolerance

Supplemental Digital Content

©2011The Shock Society