Interleukin 1α (IL-1α) Promotes Pathogenic Immature Myeloid Cells and IL-1β Favors Protective Mature Myeloid Cells During Acute Lung Infection

Abstract

Bacterial pneumonia is a common risk factor for acute lung injury and sepsis-mediated death, but the mechanisms underlying the overt inflammation and accompanying pathology are unclear. Infiltration of immature myeloid cells and necrotizing inflammation mediate severe pathology and death during pulmonary infection with Francisella tularensis. However, eliciting mature myeloid cells provides protection. Yet, the host factors responsible for this pathologic immature myeloid cell response are unknown. Here, we report that while the influx of both mature and immature myeloid cells is strictly MyD88 dependent, the interleukin 1 (IL-1) receptor mediates an important dual function via its ligands IL-1α and IL-1β. Although IL-1β favors the appearance of bacteria-clearing mature myeloid cells, IL-1α contributes to lung infiltration by ineffective and pathologic immature myeloid cells. Finally, IL-1α and IL-1β are not the sole factors involved, but myeloid cell responses during acute pneumonia were largely unaffected by lung levels of interleukin 10, interleukin 17, CXCL1, granulocyte colony-stimulating factor, and granulocyte-macrophage colony-stimulating factor.

Infiltration of the lung by immature myeloid cells and accompanying necrotizing inflammation mediate pathology and death during acute bacterial pneumonia. During lethal pulmonary tularemia, IL-1α contributes to the appearance of immature myeloid cells while IL-1β favors bacteria-clearing mature cells.

Bacterial pneumonia is a serious health problem worldwide and a primary risk factor for acute lung injury and sepsis [1, 2]. Neutrophils (PMNs), macrophages, and natural killer cells are major frontline defenses in acute pneumonia [2, 3]. Paradoxically, these cells also cause significant lung damage, leading to compromised pulmonary physiology and death [4–6]. However, the cellular mechanisms leading to inflammation and pathology remain unclear.

We and others have reported that accumulation of immature myeloid cells (IMCs), such as immature PMNs with “band” nuclei and immature macrophages, and necrotizing inflammation mediate death during acute lung infections [7–14]. Lethal Francisella tularensis infection activates emergency myelopoiesis, producing large numbers of IMCs infiltrating the lungs and outnumbering protective myeloid cells, including mature PMNs and mature macrophages. Infiltrating IMCs display myeloid-derived suppressor cell (MDSC) phenotypes and inhibit both T-cell proliferation and macrophage phagocytosis [7]. Depleting Ly6G⁺ IMCs (immature PMNs) with anti-Ly6G enhances survival of mice, but adoptive transfer of IMCs to naive mice causes lung damage and death [7], indicating their pathogenic function. In contrast, eliciting mature PMNs and mature macrophages protects mice from lethal F. tularensis infection [7, 8].

Several proinflammatory cytokines (eg, interleukin 1 [IL-1], interleukin 6 [IL-6], and interleukin 8) recruit innate and adaptive immune cells, while others (eg, interleukin 10, transforming growth factor β, and prostaglandin E₂) inhibit their maturation or function [11–16]. More specifically, in sterile inflammation, IL-1α and IL-1β play distinct roles by recruiting different myeloid cells and promoting inflammation at various stages [15]. However, during infection, both IL-1α and IL-1β recruit and activate PMNs and macrophages, while IL-1β mobilizes myeloid-suppressor cells in some tumor models [16–21]. IL-6, although critical for inflammatory responses, can also be antiinflammatory by controlling tumor necrosis factor (TNF) and macrophage inflammatory protein 2 levels during acute lung inflammation [22]. TNF is a key cytokine for innate myeloid cell responses during infections [23]. CXCL1 attracts neutrophils during infections and other inflammatory conditions [24]. However, the host factors driving the IMC response during lung infections are not well understood.

In the present study, we found that MyD88-dependent IL-1R signaling is important for the appearance of both immature and mature myeloid cells in the lung, with IL-1α and IL-1β mediating divergent outcomes. IL-1α contributes to lung infiltration by pathology-associated IMC, while IL-1β favors protective mature myeloid cells. Our data also suggest that other cytokines and chemokines important for PMN and macrophage responses, including interleukin 17 (IL-17) and the CXCR2 ligand CXCL1, may be less important in this context.

MATERIALS AND METHODS

Mice

C57BL/6J wild-type (Wt), Il1r1^−/− (strain B6.129S7-Il1r1^tm1Imx/J), Rag1^−/− (strain B6.129S7-Rag1^tm1Mom/J), Myd88^−/− (strain B6.129P2SJL-Myd88^tm1.1Defr/J), Il10^−/− (strain B6.129P2-Il10^tm1Cgn/J), Aim2^−/− (strain B6.129P2-Aim2^{Gt(CSG445)Byg}/J), and Casp1/11^−/− (strain B6N.129S2-Casp1^tm1Flv/J) mice were purchased from Jackson Laboratories. All mice were housed and bred in the Animal Resources Facility at Albany Medical College. Animal experiments were performed in accordance with the Guide for the Care and Use of Laboratory Animals and approved by the Albany Medical College Institutional Animal Care and Use Committee. Experiments were conducted using equal numbers of male and female mice (age, 8–10 weeks).

Bacteria

F. tularensis live vaccine strain (LVS) was cultured as described elsewhere [7]. Briefly, a single colony of F. tularensis was cultured in 5 mL of modified Muller-Hinton (MH) broth for 12 hours and subcultured (1:200) in 100 mL for another 12–16 hours. At an OD₂₆₀ of 0.2, the culture was aliquoted and stored frozen at −80°C. Colony-forming units (CFU) per milliliter were determined by plating on MH chocolate agar.

Intranasal F. tularensis Infection

Prior to infection, mice were anesthetized with a ketamine (20 mg/mL) and xylazine (1 mg/mL) mixture (80–100 μL/mouse), and 40 μL of inoculum containing 1000 CFU (the absolute lethal dose) or 500 CFU (the median lethal dose) in sterile phosphate-buffered saline (PBS) was instilled in 1 nare. Sham-inoculated control mice received an equal volume of PBS or appropriate vehicle medium. Infected mice were monitored for survival and mortality.

Necropsy, Tissue Collection, and Histologic Analysis

Mice were euthanized with ketamine (20 mg/mL) and xylazine (1 mg/mL), followed by cervical dislocation. Necropsy was performed, gross lesions were noted, and organs (lungs, liver, and spleen) were collected aseptically for preparation of homogenates, single-cell suspension, or histologic analysis [7]. Formalin-fixed tissues were processed by standard histologic procedures, and 4-μm-thick sections were stained with hematoxylin and eosin.

Bacterial Burden, Cytokine, and Chemokine Assays

Tissue homogenates were centrifuged at 14000g for 10 minutes to remove cellular debris and particulates. For bacterial counts, 100 μL was serially diluted (10-fold), and 3 dilutions were plated onto MH chocolate agar to count CFUs (expressed as log₁₀ CFUs/milliliter/organ). For cytokine and chemokine assays, cleared homogenates were stored frozen or analyzed immediately, using Mouse Group I or II Luminex assays (BioRad).

Multiparameter Flow Cytometry

Myeloid cells were identified by multiparameter flow cytometry as described elsewhere [7]. Briefly, isolated lung cells were surface stained with biotinylated antibodies targeting lymphoid markers (CD3, CD4, CD8, NK1.1, B220, CD19, and Ter-119), myeloid markers (CD11b, CD11c, F4/80, Gr-1, Ly6C, and Ly6G), and other markers (MHCII, CD80, and CD86) for 30 minutes, fixed in 1% paraformaldehyde, and analyzed using an LSRII (Becton Dickinson) with FlowJo software (version 10.0.1). To identify mature and immature myeloid cells, lymphoid cells were excluded, remaining cells were gated for CD11c^hi, CD11b⁺CD11c⁺, or CD11b^hi, with CD11b^hi cells gated further into F4/80⁻ and F4/80^hi populations. F4/80⁺ cells represent mature macrophages, based on higher expression of MHCII and Ly6C. F4/80⁻ cells were gated to identify mature PMNs (CD11b^hi Ly6G^hi Ly6C^int), immature PMNs (CD11b^hi Ly6G^hi Ly6C^low), and immature macrophages (CD11b^hi Ly6G^low Ly6C^hi). Total numbers for each population were calculated by flow cytometry and by cytospin smears of Giemsa-stained bronchoalveolar lavage fluid as described previously [7].

Cytokine Neutralization

F. tularensis–infected mice were injected intraperitoneally with 200 µg/mouse of anti–IL-1β (clone B122), anti–IL-1α (clone ALF-161), or isotype control antibody (BioXcell, West Lebanon, NH) at −1, 1, 3, 4, and 5 days after infection. Mice receiving anti-IL-1α were euthanized 6 days after infection for cellular/cytokine analysis and pathology. As anti-IL-1β-treated mice died by 6 days after infection in preliminary experiments, mice receiving anti-IL-1β were euthanized at 5 days after infection.

Isolation, Sorting and Culture of Bone Marrow Cells and Blood Cells

Bone marrow cells were isolated as described [8]. Peripheral blood was collected in ethylenediaminetetraacetic acid, layered over Ficoll-Paque Plus (GE Healthcare, Uppsala). After centrifugation, peripheral blood mononuclear cells were collected, washed in PBS, and stained with CD3 (PE), CD11b (PE/Cy7), Ly6G (AF647), and Ly6C (FITC) for sorting Ly6C+ and Ly6G+ cells (BDFacsAria). Sorted cells were cultured in Dulbecco’s modified Eagle’s medium (10% fetal bovine serum) with and without recombinant IL-1α (rIL-1α) and rIL-1β for 24 hours, and analyzed for viability and maturation as described [7]. For phagocytic assays, cells were cultured with Alexa Fluor 488–labeled Escherichia coli (K-12 Strain) Bioparticles (catalog number E-13231; Molecular Probes, Eugene, OR) for 4 hours, and the percentage of cells containing bioparticles was determined by flow cytometry. For the bacterial killing assay, cells were infected with F. tularensis LVS (multiplicity of infection, 100), and the intracellular bacterial burden was calculated as described previously [7, 8].

Statistical Analysis

Statistical analyses were performed with GraphPad Prism (version 6), using an unpaired Student t test or nonparametric Mann-Whitney test. For survival analysis, the log-rank (Mantel-Cox) test was used. A P value of < .05 was considered statistically significant.

RESULTS

Innate Myeloid Cells Mediate Acute Inflammation and Pathology in Lung Infection

Our previous work suggests that lung infiltration by myeloid cells mediates overt inflammation, tissue pathology, and death during acute pulmonary tularemia [7–10]. However, whether T and B cells are involved in these processes remains unclear. To evaluate this possibility, Rag1^−/− mice were infected with F. tularensis. Susceptibility and tissue bacterial burden were comparable between Rag1^−/− and Wt mice following lethal (Figure 1A) and sublethal infection (survival rate, 30% and 37%, respectively). Furthermore, cellular inflammatory foci in the lung, associated necrotic changes, and lung pathology scores 6 days after infection (14.6 for Wt mice vs 14.3 for Rag1^−/− mice) were comparable. Numbers of infiltrating myeloid cells (PMN, immature PMNs, mature macrophages, and immature macrophages) in Rag1^−/− mice were equivalent to those in Wt mice (Figure 1B). The slight difference in mature PMNs was not statistically significant. The T-cell–dependent cytokines IFN-γ and IL-17 were significantly lower in Rag1^−/− mice, but IL-10 levels were increased (Figure 1C). However, the levels of IL-1α, IL-1β, IL-18, and CXCL1 (Figure 1C), as well as levels of IL-6, TNF-α, IL-12, granulocyte colony-stimulating factor (G-CSF), granulocyte-macrophage (GM)–CSF, and monocyte chemoattractant protein 1 (data not shown), were not significantly different. Thus, T and B cells are not essential for overt cellular inflammation or associated pathology during acute pulmonary tularemia. Further, decreased IL-17 levels did not influence the numbers of mature or immature PMNs in the lungs of infected Rag1^−/− mice.

Figure 1.

Innate myeloid cells mediate acute inflammation and pathology in lung infection. Rag1^−/− mice (lacking T and B cells) or wild-type (Wt) mice were infected with Francisella tularensis. A, Survival of Rag1^−/− and Wt mice following lethal F. tularensis live vaccine strain infection (with 1000 colony-forming units [CFU]) and tissue bacterial burden at day 6. Data are percentage survival from 2 independent experiments (n = 10 mice; comparisons were made by the log-rank test) or median values from 2 independent experiments (comparisons were made by the unpaired Student t test). B, Lung myeloid cells analyzed by flow cytometry and identified as mature neutrophils (mPMNs), immature neutrophils (iPMNs), mature macrophages (mMΦ), and immature macrophages (iMΦ) at days 0 and 6. Data are median values from 2 independent experiments. Comparisons were made by the unpaired Student t test. C, Levels of cytokines and chemokines measured in lung homogenates at day 6, using the Luminex assay. Data are median values from 2 independent experiments. *P < .05, by the nonparametric Mann-Whitney test. IFN-γ, interferon γ; IL-1, interleukin 1; IL-10, interleukin 10; IL-18, interleukin 18.

The IL-10/IL-17 Axis Is Dispensable for the IMC Response

IL-17 attracts neutrophils important for protection against sublethal F. tularensis pneumonia [25, 26]. Il10^−/− mice infected with a sublethal dose have increased IL-17 levels and pulmonary neutrophil accumulation, but despite elevated IFN-γ levels, lung pathology and susceptibility are enhanced [20, 27]. Whether IL-10 influences IMC infiltration of the lung is unknown and was evaluated by examining Il10^−/− mice infected with a lethal dose of F. tularensis LVS. While mature and immature PMN numbers were slightly increased and decreased, respectively, cell numbers were statistically indistinguishable from those in Wt mice (Supplementary Figure 1A). Unsurprisingly, IL-17 and IFN-γ levels were increased in Il10^−/− mice (Supplementary Figure 1B). However, levels of other cytokines and chemokines, including IL-1α and IL-1β, were unchanged 6 days after infection, compared with those in Wt mice. Thus, IL-10, either alone or through regulation of IL-17 and IFN-γ, has no significant impact on myeloid cell numbers during acute lethal pneumonia.

MyD88 Is Required for Lung Infiltration by Myeloid Cells

As T cells and IL-17 appear dispensable for lung infiltrating myeloid cells, other host factors are likely responsible. MyD88 signals are required for both inflammation and clearance of lung infections [28–31]. Curiously, Myd88^−/− mice are more susceptible to highly virulent F. tularensis (SchuS4) but lack the robust cellular lung infiltrates seen in Wt mice [28]. Myd88^−/− mice infected with a lethal dose of F. tularensis LVS were more susceptible than Wt mice, with higher bacterial burdens (Figure 2A) and significantly fewer lung mature PMNs, immature PMNs, mature macrophages, and immature macrophages 6 days after infection (Figure 2B). Thus, MyD88 is required for the marked infiltration of both mature and immature myeloid cells. Lung IL-1α and IL-1β levels were dramatically lower, and IL-10 and IFN-γ levels were reduced (Figure 2C). Other cytokines, including IL-17, IL-18, CXCL1, G-CSF, and GM-CSF, were largely unchanged from Wt mice, indicating that their elaboration in the lungs of F. tularensis–infected mice is MyD88 independent. More importantly, these cytokines appear insufficient for either the IMC or mature myeloid cell response in the absence of MyD88. Together, these data suggest that the lung cellular inflammatory response to F. tularensis requires a MyD88-dependent cytokine(s) and/or chemokine(s), such as IL-1α and/or IL-1β.

Figure 2.

Myd88-dependent interleukin 1β (IL-1β) is required for a protective response during lung infection. Myd88^−/− or wild-type (Wt) mice were infected with Francisella tularensis. A, Survival of Myd88^−/− and Wt mice following lethal F. tularensis live vaccine strain infection (with 1000 colony-forming units) and tissue bacterial burden at day 6. Data are percentage survival from 2 independent experiments (n = 10 mice; **P < .01, by the log-rank or Mantel-Cox tests) or median values from 2 independent experiments (*P < .05 and **P < .01, by the unpaired Student t test). B, Lung cells analyzed by flow cytometry were identified as mature neutrophils (mPMNs), immature neutrophils (iPMNs), mature macrophages (mMΦ), and immature macrophages (iMΦ) at day 6. Data are median values from 2 independent experiments. *P < .05 and **P < .01, by the unpaired Student t test. C, Levels of cytokines and chemokines measured by Luminex assay in lung homogenates at day 6. Data are median values from 2 independent experiments. *P < .05, by the nonparametric Mann-Whitney test. IFN-γ, interferon γ; IL-10, interleukin 10; IL-18, interleukin 18.

IL-1R Is Important for Lung Infiltration by Mature Myeloid Cells and IMCs

Protection against pulmonary pathogens, including F. tularensis, requires IL-1β and the IL-1R [15–17, 32]. Blocking MyD88-dependent and MyD88-independent IL-1R signaling reduces inflammation-related pathology [33–35]. However, whether pulmonary IMC responses require IL-1R is unclear. The bacterial burden was higher in lungs of Il1r^−/− versus Wt mice 3 days after infection (2.7 × 10⁷ vs 1.3 × 10⁷ CFU) and significantly increased in lung, spleen, and liver 6 days after infection (Figure 3A). Consistent with bacterial loads, levels of mature PMNs and mature macrophages were significantly lower in Il1r^−/− mice 3 and 6 days after infection (Figure 3B). However, without IL-1R, immature PMN numbers were decreased at both time points, while immature macrophage counts were only reduced 6 days after infection. Lung IL-1α and IL-1β levels (Figure 3C and 3D), as well as IL-2, IL-4, IL-6, IL-12, IL-17, IL-18, IFN-γ, and TNF-α levels (data not shown), were unchanged in Il1r^−/− versus Wt mice. The CXCL1 level was also significantly reduced 3 and 6 days after infection (Figure 3C and 3D), suggesting that IMC and mature myeloid cell recruitment is independent of local CXCL1. IL-10 and GM-CSF levels were modestly decreased 3 days after infection but were strikingly lower 6 days after infection (Figure 3D). Thus, although other MyD88-dependent pathways are involved, the IL-1/IL-1R axis influences the magnitude of both lung IMC and mature myeloid cell responses during F. tularensis pneumonia. Further, IL-1R has competing functions contributing to both protective and pathogenic responses.

Figure 3.

Interleukin 1 (IL-1) and IL-1 receptor (IL-1R) signaling regulates myeloid cell recruitment in lung infection. Il1r1^−/− or wild-type (Wt) mice were infected with Francisella tularensis. A, Tissue bacterial burden at day 6. Data are median values from 2 independent experiments. *P < .05, by the unpaired Student t test. B, Lung cells analyzed by flow cytometry were identified as mature neutrophils (mPMNs), immature neutrophils (iPMNs), mature macrophages (mMΦ), and immature macrophages (iMΦ) at days 0, 3, and 6. Data are median values from 2 independent experiments. *P < .05 and **P < .01, by the unpaired Student t test. C and D, Levels of cytokines/chemokines measured by the Luminex assay in lung homogenates at day 3 (C) and day 6 (D). Data are median values from 2 independent experiments. *P < .05 and **P < .01, by the nonparametric Mann-Whitney test. GM-CSF, granulocyte-macrophage colony-stimulating factor; IL-10, interleukin 10.

IL-1α Promotes IMCs, While IL-1β Favors Mature Myeloid Cells

IL-1R elicits inflammatory responses via 2 ligands, IL-1α and IL-1β [35, 36]. IL-1β recruits and activates various leukocytes, including neutrophils and macrophages, controlling bacteria [14, 15]. Indeed, IL-1β is protective during many infectious diseases [15–17], and IL-1β–deficient mice are highly susceptible to F. tularensis infection [32]. IL-1α promotes neutrophilic inflammation following its release from dead and dying cells but also drives inflammatory tissue damage [37–43]. Intriguingly, mice lacking IL-1α are more resistant to F. tularensis infection [32], suggesting that IL-1α might elicit the pathogenic IMC response. Following IL-1α neutralization, lung and spleen bacterial burdens 6 days after infection were reduced (Figure 4A). While mature PMN numbers were, as anticipated, lower, immature PMN and immature macrophage numbers were also reduced (Figure 4A). Thus, IL-1α promotes lung infiltration by both types of IMCs. In contrast, IL-1β neutralization resulted in a higher bacterial burden in lung, spleen, and liver, while numbers of protective lung mature PMNs and mature macrophages were diminished (Figure 4B). Immature PMN numbers were unchanged, but immature macrophages numbers increased approximately 2-fold, suggesting that IL-1β promotes both mature PMNs and mature macrophages required for bacterial clearance, while suppressing numbers of pathology-associated immature macrophages. To evaluate how IL-1α and IL-1β might regulate IMC numbers, lung cytokines and chemokines were examined. With IL-1α neutralization, lung IL-1α levels were, as expected, diminished 6 days after infection, but IL-1β levels increased (Figure 4C). IL-17, CXCL1, G-CSF, and GM-CSF levels were indistinguishable from those in controls (Figure 4C), as were IL-6, IL-10, IL-12p70, IL-18, IFN-γ, TNF-α, and MIP-1α levels (data not shown). Thus, IL-1α promotes the appearance of IMCs independent of other key cytokines and chemokines. After IL-1β neutralization, lung IL-1β levels were very low and IL-17, CXCL1, G-CSF, and GM-CSF levels were also reduced (Figure 4D). Thus, changes in lung PMN, macrophage, and immature macrophage populations during F. tularensis infection could be direct mediated by IL-1β or achieved in conjunction with IL-17, CXCL1, G-CSF, and/or GM-CSF.

Figure 4.

Interleukin 1α (IL-1α) elicits an immature myeloid cell response, but IL-1β is required for mature myeloid cells. Wild-type (Wt) mice were treated with anti–IL-1α or anti–IL-1β and infected with Francisella tularensis. A, Tissue bacterial burden and total numbers of mature neutrophils (mPMNs), immature neutrophils (iPMNs), mature macrophages (mMΦ), and immature macrophages (iMΦ) in the lungs of IL-1α–neutralized mice at day 6. Data are median values from 2 independent experiments. *P < .05 and **P < .01, by the unpaired Student t test. B, Tissue bacterial burden and total numbers of mPMNs, iPMNs, mMΦ, and iMΦ in lungs of IL-1β–neutralized mice at day 5. Data are median values from 2 independent experiments. *P < .05 and **P < .01, by the unpaired Student t test. C, Levels of cytokines and chemokines measured by the Luminex assay in lung homogenates of IL-1α–neutralized mice at day 6. Data are median values from 2 independent experiments. *P < .05, by the nonparametric Mann-Whitney test. D, Levels of cytokines and chemokines measured by the Luminex assay in lung homogenates of IL-1β–neutralized mice at day 5. Data are median values from 2 independent experiments. *P < .05 and **P < .01, by the nonparametric Mann-Whitney test. Ab, antibody; G-CSF, granulocyte colony-stimulating factor; GM-CSF, granulocyte-macrophage colony-stimulating factor; IL-17, interleukin 17; IsoAb, isotype control antibody.

In vivo, IL-1α and IL-1β recruit neutrophils and macrophages, respectively [15], consistent with our cellular data above. As IL-1α neutralization reduced IMC numbers, IL-1α and IL-1β might also influence myeloid cell maturation. Although maturation and differentiation of Ly6C+ cells treated with rIL-1α and rIL-1β was increased as compared to control cells or cell treated with L-cell medium (positive control), maturation scores were comparable (data not shown). However, phagocytosis and bacterial killing was significantly greater for Ly6C+ cells treated with rIL-1β versus those treated with rIL-1α (Figure 5A and 5B). In contrast, many LyG+ cells were dead, and maturation was not enhanced after culture with rIL-1. Statistically, bone marrow Ly6G+ cells were more phagocytic following IL-1β treatment, but the frequency of phagocytic cells remained quite low (Figure 5C and 5D). Thus, while IL-1α and IL-1β both support in vitro maturation and differentiation of Ly6C+ myeloid cells, IL-1β may more effectively improve their phagocytic capacity than IL-1α, thus favoring bacterial clearance and survival.

Figure 5.

Interleukin 1β (IL-1β) favors myeloid cell maturation, differentiation, and phagocytosis. A and B, Ly6C+ cells from bone marrow (A) or blood (B) were or were not treated with recombinant IL-1α (rIL-1α) or rIL-1β. Left, Percentage of cells that phagocytosed the bioparticles. Data are median values with ranges, from 2 experiments. *P < .05, by the unpaired Student t test. Right, Intracellular bacterial burden. Data are median values from 2 experiments. *P < .05, by the nonparametric Mann-Whitney test. C and D, Ly6G+ cells from bone marrow (C) or blood (D) show the percentage of cells that phagocytosed the bioparticles. Data are median values with ranges, from 2 experiments. *P < .05, by the unpaired Student t test.

DISCUSSION

Acute pulmonary infections are a major cause of illness and mortality worldwide [1, 2]. Mortality has been largely attributed to overt myeloid cell infiltration of the lung and associated necrotic lung damage [7–10]. Although well-established systems to confirm the deleterious effects of excessive myeloid cell–mediated inflammation are lacking, we demonstrated the pathogenicity of IMCs by cell depletion and adoptive transfer of IMCs [7, 8]. In the present study, we determined the importance of IL-1β and IL-1α in controlling whether the lung myeloid cell response favors protection or pathology.

B and T cells were not essential for myeloid cell responses, lung pathology, or host mortality during F. tularensis infection, further demonstrating that myeloid cells are likely sufficient, as reported previously, to cause pathology accompanying acute lung infections [7–10]. However, B and T cells are essential for the development of protective, adaptive immunity [44] and thus might be important for resistance. Yet, Rag1^−/− and Wt mice were equally susceptible to sublethal or lethal F. tularensis infections. As the mature myeloid cell response is protective during sublethal infection [7–10], maintained resistance to sublethal challenge in Rag1^−/− mice supports the thinking that early adaptive responses are not essential. In contrast, early T-helper type 17 cell responses and IL-17 are reported to be important for resistance to F. tularensis pneumonia [25–27]. Our data suggest that T-cell–derived IL-17 has little impact on the lung myeloid response to F. tularensis. Contrary to studies of Il17^−/− mice [26, 27], we observed no increase in bacterial burden or a significant decrease in mature PMNs in the relative absence of IL-17 in Rag1^−/− mice. Furthermore, IL-17 levels were elevated and associated with excessive neutrophilic inflammation and severe pathology in F. tularensis–infected Il10^−/− mice, consistent with detrimental lung inflammation [8, 9, 27]. However, the elevated level of IL-17 in Il10^−/− mice did not correlate with numbers of pathology-associated IMCs. Thus, the pathologic role of IMCs during lethal bacterial pneumonia may not be enhanced or countered by IL-17 or IL-10. Alternatively, IL-17 may be required, but IL-17 levels might not reflect its function in this context, although this seems unlikely.

MyD88 was essential for substantial myeloid cell infiltration and lung damage, control of bacteria, and survival following infection with F. tularensis LVS (this study) or Schu S4 [28]. While contrasting with the hypothesis that IMCs are responsible for pathology and host mortality during F. tularensis infection, mature phagocytes are required to control infection and prevent host damage. Bacterial control is largely absent in Myd88^−/− mice, as liver and spleen bacterial burdens were similar to those in mice receiving 10 times the lethal dose of F. tularensis [7]. Although Myd88^−/− mice die from F. tularensis infection, they do not lose weight (data not shown) [45], and levels of the cytokines normally associated with septic shock are generally reduced. It is possible that burdens achieved in these mice are sufficient to interfere with normal organ function in the absence of other pathologic processes. Levels of IL-1α and IL-1β were reduced in infected Myd88^−/− mice, but levels of many other proinflammatory cytokines and chemokines were unchanged, including the CXCR2 ligand CXCL1. As CXCR2 signaling is important for myeloid cell recruitment, low numbers of these cells may reflect insufficient CXCR2 ligand levels. Further, the cytokines with levels that remain elevated in Myd88^−/− mice appear insufficient to promote either IMCs or mature myeloid cells. Although IL-1R is MyD88 dependent, the myeloid cell response of infected Il1r^−/− mice was not abrogated as in Myd88^−/− mice. Accordingly, other MyD88-coupled receptors are likely important. For example, Toll-like receptor 2 (TLR2), the major receptor for F. tularensis–derived lipoproteins may contribute. However, myeloid cell lung infiltration in F. tularensis–infected Tlr2^−/− mice exceeds that of Wt mice [46], suggesting that TLR2 may limit the magnitude or nature of the IMC response.

As reported previously, Asc^−/−, Casp1/11^−/−, Aim2^−/− [8], and Il11β^−/− [32] mice are susceptible to sublethal doses of F. tularensis, consistent with IL-1β being essential for resistance. Further, rIL-1β treatment protects Il1β^−/− mice [32], emphasizing its protective role. Indeed, here, anti–IL-1β treatment diminished the mature myeloid cell response associated with protection. In our previous study, numbers of IMCs and lung pathology were similar between Wt and Casp1/11^−/− mice [8]. An insufficient mature myeloid cell response in these mice could explain their higher susceptibility to infection. Unlike other IL-1β–deficient lines, Nlrp3^−/− mice resist F. tularensis infection, but these mice exhibit an early mature PMN response and reduced myeloid cell death independent of reduced IL-1β levels [8]. IL-1α is implicated in pathologic inflammation of the lungs and other organs [38–43], and Il1α^−/− mice are resistant to F. tularensis infection [32]. Moreover, in this study we found that neutralizing IL-1α reduced the number of lung IMCs and increased the number of mature myeloid cells. Collectively, our data support the view that IL-1β favors a protective mature myeloid cell response in the lung, while IL-1α promotes a pathogenic IMC response. IL-β and IL-1α are major cytokines with pleiotropic functions [35, 36], and while it is appreciated that these cytokines mediate both protective and pathogenic responses during pulmonary infections [39, 41], how IL-1R mediates both is unclear. While Il1r1^−/− mice had significantly reduced numbers of mature PMNs and macrophages early in infection (day 3), they also had fewer IMCs, consistent with a dual functional role for IL-1R. However, although infected Il1r1^−/− mice had lower IMC and mature myeloid cell counts, numbers of these cells were still substantial, indicating that factors other than IL-1α and IL-1β likely help mediate both pathologic and protective myeloid cell responses.

That IL-1α and IL-1β mediate opposing roles but use the same receptor is perplexing. However, as the IL-1 monoclonal antibodies used distinguish between forms of IL-1 (the most likely cross-reacting targets) and because IL-1 neutralization has little impact on the other cytokines assayed, our results are unlikely due to off-target effects. Further, the literature does not support more than one signaling receptor for IL-1α or IL-1β. Importantly, while it is generally thought that functions of IL-1α and IL-1β are identical in vitro [37], various in vitro and in vivo studies highlight differential roles [15, 32, 36, 47]. In contrast, membrane-bound IL-1α and differential expression of IL-1α and IL-1β have been suggested to explain differential in vivo functions despite a shared receptor [37].

In summary, while the influx of myeloid cells requires MyD88, the pathologic IMC response during lethal F. tularensis pneumonia involves IL-1α but is opposed by IL-1β, which promotes protective phagocytes. Although differential recruitment of neutrophils and macrophages by IL-1α and IL-1β has been reported [15], whether a similar mechanism influences immature myeloid cells remains to be tested. However, our data do suggest that IL-1β favors maturation and differentiation of more-effective phagocytic cells over IL-1α. Further, multiple MyD88-dependent receptors and cytokines and chemokines beyond IL-1α and IL-1β are required for the pathologic IMC response, but lung levels of IL-10, IL-17, CXCL1, G-CSF, and GM-CSF, which were expected to have more obvious contributions, did not correlate with the magnitude of the mature or immature myeloid cell response. Future mechanistic studies will be needed to detail the genesis of IMCs during bacterial pneumonia, the opposing roles of IL-1α and IL-1β via IL-1R, other MyD88-dependent pathways contributing to the IMC response, and how IMCs promote pathology.

Supplementary Data

Supplementary materials are available at The Journal of Infectious Diseases online. Consisting of data provided by the authors to benefit the reader, the posted materials are not copyedited and are the sole responsibility of the authors, so questions or comments should be addressed to the corresponding author.

Notes

Acknowledgment. We thank Heather Chin for technical assistance

Financial support. This work was supported by the National Institute of Allergy and Infectious Diseases, National Institutes of Health (grant P01AI056320 subproject 2 to J. A. H.).

Potential conflicts of interest. All authors: No reported conflicts of interest. All authors have submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Conflicts that the editors consider relevant to the content of the manuscript have been disclosed.