CXCL1 Regulates Pulmonary Host Defense to Klebsiella Infection via CXCL2, CXCL5, NF-κB, and MAPKs

Despite the development of newer generation antibiotics, the morbidity and mortality associated with bacterial pneumonia is still extensive (1–3). In addition, clinical management of patients infected with antibiotic-resistant bacterial strains, including Klebsiella pneumoniae, has been difficult. Innate host defense mediated through neutrophils is pivotal to controlling bacterial replication in the host. However, neutrophil influx acts as a double-edged sword: an insufficient neutrophil recruitment can lead to life-threatening infection, whereas an extreme accumulation of neutrophils can result in excessive lung injury, such as acute lung injury/adult respiratory distress syndrome (4–6). Therefore, the ideal therapeutic approach to target neutrophils would be to attenuate their destructive potential whereas maintaining their antibacterial defense function. However, understanding the mechanisms contributing to neutrophil influx is a prerequisite to designing improved treatment or prevention strategies, or both, to attenuate excessive neutrophil-mediated lung parenchymal damage.

Bacterial lung infection is characterized by excessive capillary leakage, activation of transcription factors, upregulation of cell adhesion molecules, expression of proinflammatory mediators, and accumulation of neutrophils in the alveolar spaces with diffuse damage of alveolar epithelial and endothelial cells (7–9). The production of chemokines is a critical step leading to the recruitment of neutrophils to organs during microbial insult (7–9). These chemokines are classified into four groups based on the arrangement of a cysteine motif positioned near the N terminus: C, CC, CXC, and CX₃C. These groups are further characterized by the position of the ELR motif (glutamic acid-leucine-arginine) before the CXC sequence. These ELR⁺ CXC chemokines are powerful neutrophil chemoattractants, and seven of this group have been identified in humans: IL-8; neutrophil-activating peptide 2; growth-related oncogene-α, -β, and -γ; epithelium-derived neutrophil-activating peptide 78; and granulocyte chemotactic protein 2. Of these, IL-8 is the most potent neutrophil chemoattractant in humans. Although a homolog of human IL-8 has not been identified, CXCL1/keratinocyte cell-derived chemokine (KC) (10, 11), CXCL2/MIP-2 (12, 13), and CXCL5/LPS-induced CXC chemokine (LIX) (14, 15) have been found to be potent neutrophil chemoattractants in mice.

KC has been shown to be essential in inducing neutrophil recruitment to the lungs because 1) inhibition of KC using a blocking Ab resulted in attenuation of neutrophil migration to the airspaces after Escherichia coli LPS challenge in a rat model (10, 11), and 2) transgenic KC mice, which constitutively expressed KC within the lungs, have more neutrophil influx and less bacterial burden in the organs after challenge with K. pneumoniae (16). However, the signaling cascades associated with neutrophil-mediated bacterial clearance in the lungs have yet to be explored. With the advent of gene deletion technology in mice, it is possible to determine the mechanisms of these chemokines in pulmonary host defense. In this study, we sought to determine the mechanisms underlying KC-mediated pulmonary defense using mice with a targeted disruption in KC. Our results reveal that KC derived from both hematopoietic and resident cells regulate host defense in the lung via expression of other neutrophil chemokines and activation of NF-κB and MAPKs. Moreover, our findings suggest that KC mediates pulmonary host defense in an autocrine and paracrine manner.

KC gene-deficient mice (KC^−/−) (17) were backcrossed 10 times with C57BL/6; therefore, C57BL/6 mice were used as control mice (KC^+/+). All animal experiments were conducted using a protocol approved by the Louisiana State University Animal Welfare Committee. Female mice 8–10 wk old, ranging from 20 to 25 g in weight, were used in our experiments. Mice were maintained under specific pathogen-free conditions and kept on a 12:12 h light/dark cycle with free access to food and water.

We used serotype 2 of K. pneumoniae strain (ATCC 43816; American Type Culture Collection, Manassas, VA) because this strain induces substantial inflammation in mice at 10³ CFU/animal (18). In brief, the bacteria were grown for 6 h to midlogarithmic phase at 37°C in tryptic soy broth while shaking at 225 rpm, harvested by centrifugation, and washed twice in sterile isotonic saline. The bacteria were resuspended in saline at a concentration of 10³ CFU/50 μl/mouse for in vivo experiments. The mice were anesthetized with ketamine (100 mg/kg)/xylazine (10 mg/kg) mixture, and a midventral incision was performed. After isolation of muscles, the trachea was exposed and inoculated with 10³ CFU K. pneumoniae. A 50-μl aliquot of a serially diluted suspension of the initial inocula was plated onto tryptic soy agar and MacConkey plates for validation of CFUs.

Mice were inoculated intratracheal (i.t.) with 10³ CFU K. pneumoniae in 50 μl of 0.9% saline, and survival was examined up to 15 d.

The lungs and spleens of mice were weighed and homogenized in 1 ml of 0.9% saline using a tissue homogenizer. The solid tissue was allowed to sediment for 10 min at room temperature, and the supernatants were serially diluted. Twenty-microliter aliquots of each sample were plated on tryptic soy agar and MacConkey agar plates, and bacterial colonies were enumerated after incubation at 37°C.

At the designated time points, the animals were euthanized and exsanguinated by cardiac puncture. The trachea was cannulated with a 20-gauge catheter, and bronchoalveolar lavage fluid (BALF) was collected as described earlier (18–21). A total of 3.0 ml BALF was retrieved from each mouse, and 140 μl BALF was centrifuged and placed on glass cytospin slides, which were then stained by Diff-Quick reagents (Fisher Scientific, Chicago, IL) to enumerate leukocyte subtypes. Two milliliters of the undiluted BALF was centrifuged, passed via a 0.22-μm filter, and used immediately or stored at −80°C for future use.

Neutrophil migration to the lung can be assessed by myeloperoxidase (MPO) activity in the lungs, which was determined as previously described (18–21). In brief, the lungs were weighed, homogenized, centrifuged, and the pellet was resuspended in 50 mM potassium phosphate buffer, pH 6.0 (supplemented with 0.5% hexadecyltrimethylammonium bromide). Samples were then sonicated, incubated at 60°C for 2 h, and assayed for MPO activity in a hydrogen peroxide/O-dianisidine buffer at 460 nm. The MPO activity was calculated between 0 and 90 s. The lung samples were used for MPO activity within 3 wk postharvest.

Cytokine and chemokine levels were measured in BALF, lung homogenates, culture supernatants obtained from bone marrow-derived macrophages (BMDMs), and alveolar macrophages (AMs) using a specific sandwich ELISA as described in earlier publications. The minimum detection limit is 8 pg/ml cytokine protein (18–21).

The harvested lungs were homogenized in 1 ml PBS buffer containing 0.1% Triton X-100, and complete protease and phosphatase inhibitor mixture (Roche, Indianapolis, IN) using a TissueLyser II (Qiagen, Valencia, CA) for 30 s in chilled conditions and centrifuged at maximum speed in a microcentrifuge at 4°C for 20 min to remove debris. The resulting supernatant fluids were used for Western blotting. To ensure equal loading of protein onto the gel, we performed a Bradford protein assay (Bio-Rad, Hercules, CA). The lung homogenates were analyzed on 8–15% Tris-glycine gels. The resolved proteins were transferred onto a polyvinylidene difluoride membrane using standard protocols. The specific Abs to phospho-p44/42 MAPK (Erk1/2) (Thr202/Tyr204), phospho-p38 MAPK (Thr180/Tyr182), phospho–stress-activated protein kinase/JNK (Thr183/Tyr185), phospho-Iκ kinase (IKK)α/β (Ser176/180), phospho–NF-κB/p65 (Ser536), phospho–Iκ-Bα (Ser32), IKKβ, NF-κB/p65, IκBα, VCAM-1, and ICAM-1 were added at a 1:1000 dilution, whereas Abs to total p38 and GAPDH were added to the concentration of 1:5000. Immunostaining was performed using an appropriate secondary Ab at a dilution of 1:2000 and developed with ECL plus Western blot detection system (Thermo-Fisher, Piscataway, NJ).

An ELISA-based NF-κB DNA binding assay was performed in the lungs according to the manufacturer’s protocol (Active Motif, Carlsbad, CA) to detect the activation of the p65 subunit of NF-κB as described earlier (18). Nuclear extract was prepared using the Nuclear Extraction Kit from Active Motif. In brief, a total of 7.5 μg nuclear extract from each lung sample post-K. pneumoniae challenge was added to the NF-κB–specific oligonucleotide coated 96-well plate and incubated for 1 h at room temperature. After washing the plate three times, a primary Ab specific for NF-κB/p65 was added and incubated for 1 h at room temperature. After three washings to remove excess primary Ab, an anti-HRP conjugate was added to the plate, and the color development was monitored at the OD of 450 nm.

Donor and recipient mice between 6–8 wk old were used to generate chimeras, as described in our earlier publication (19). Recipient mice were gamma-irradiated in two 525-rad doses between 3 h. Bone marrow cells (8 × 10⁶/mouse) were injected into the tail vein of the irradiated recipients, and mice were maintained on 0.2% neomycin sulfate for the first 3 wk. The reconstituted mice were used for experiments 6 wk after the transplantation. We found that greater than 84% of blood leukocytes were derived from donor mice at the time the mice were used for experiments (6–8 wk after transplantation). Irradiated mice that were not transplanted with donor cells died between days 18 and 22 after transplantation (data not shown).

The protocol used to deplete neutrophils (GR1⁺) in mice has been described in our earlier reports (21) with modifications. A total of 50 μg anti-GR1⁺ mAb (sodium azide-free and low LPS content; RB6-8C5 from BD Pharmingen, San Diego, CA) in 50 μl was administrated i.p. at 12 and 2 h before bacterial infection. In control experiments, 50 μg of nonspecific isotype-matched control mAb in equal volumes was administrated at the same time points before bacterial infection. To validate the efficiency of neutrophil depletion by the Ab, we obtained differential WBC counts in blood every 12 h up to 3 d, and ≤2% neutrophils were found up to 3 d after depletion.

BMDMs were differentiated using DMEM containing 10% FBS and M-CSF. BM cells were differentiated in culture dishes by supplementing DMEM with M-CSF on days 3 and 5 for 7 d. Cells were then stimulated for the indicated times with K. pneumoniae (multiplicity of infection [MOI] of 1) and processed in urea/CHAPS/Tris buffer before being separated on SDS polyacrylamide gels. Proteins were transferred onto polyvinylidene difluoride membranes and were detected using the indicated Ab specific for activated/phosphorylated NF-κB or total MAPKs, such as JNK, ERK, and p38. The Western blots were quantitated against GAPDH and expressed as fold increase in response to K. pneumoniae after background (unstimulated) subtraction. In addition, culture media obtained from macrophages at 18 h after K. pneumoniae infection were used for cytokine/chemokine determination by ELISA.

AMs were isolated as described in earlier publications (22, 23). Positive selection by CD11b MicroBeads (Stem Cell Technologies, Vancouver, British Columbia, Canada) was applied to isolate AMs from BAL cells. Pooled BAL cells were resuspended in RoboSep buffer, and CD11b PE labeling reagent was mixed with cell suspension, followed by incubation at room temperature for 15 min. EasySep PE selection mixture was thereafter added to the cell suspension, and the suspension was further incubated for 15 min at room temperature. EasySep Magnetic nanoparticles were lastly mixed with the cell suspension, followed by magnetic separation of CD11b⁺ cells. Cells were washed with PBS, centrifuged at 1200 rpm for 10 min, and resuspended in DMEM containing 5% FBS. Cells were infected with K. pneumoniae (MOI of 1) and harvested at designated time points using urea/CHAPS/Tris buffer. Samples were centrifuged and supernatants were used for Western blotting. In addition, culture media were collected at 18 h postinfection and were used for cytokine/chemokines determination. Finally, the total number of live cells and the percentage of purified AMs were determined by utilizing trypan blue exclusion and cytospin preparations stained with Wright-Giemsa, respectively. Percentage recovery of AMs was calculated as the ratio of AMs to the total macrophages isolated from the BALF. Using this technique, we obtained ∼85–90% AMs.

Data are expressed as mean ± SE. The intensity of immunoreactive bands was determined using a Gel Digitizing Software (UN-SCAN-IT gel) from Silk Scientific (Orem City, UT). Data were analyzed by one-way ANOVA, followed by Bonferroni’s post hoc analysis for multiple comparisons. All statistical calculations were performed using InStat software and GraphPad Prism 4.0. Differences were considered statistically significant at p < 0.05 when compared with control. Survival curves were compared by Wilcoxon rank sign test.

To assess the requirement of KC in host defense against K. pneumoniae, we infected KC^−/− and KC^+/+ animals with K. pneumoniae (10³ CFU/mouse) through the i.t. route, and survival was monitored for up to 15 d. We observed decreased survival in the KC^−/− group, and only 14.3% of mice survived to day 15 (Fig. 1A). In contrast, 75% of K. pneumoniae-infected KC^+/+ mice survived up to 15 d after K. pneumoniae infection (Fig. 1A).

To determine whether the enhanced mortality in KC^−/− mice to K. pneumoniae was due to an impaired ability to limit bacterial growth in the lungs and/or bacterial dissemination to the bloodstream, we infected KC^−/− and their control mice (KC^+/+) with 10³ CFU K. pneumoniae, and assessed bacterial numbers in the lung and spleen at 24 and 48 h. As compared with KC^+/+ (C57BL/6) mice, KC^−/− animals had a greater bacterial burden in the lungs at 48 h and more bacterial dissemination at 24 h (Fig. 1B, 1C).

To assess the myeloid cells recruitment to the lungs of KC^−/− and KC^+/+ mice post-K. pneumoniae infection, we i.t. infected both groups of mice with K. pneumoniae and then enumerated the recruited cells. In KC^−/− mice, pulmonary total WBC and neutrophil recruitment was attenuated at 24 and 48 h after K. pneumoniae challenge. In control (saline-challenged) groups, however, no neutrophil influx in the lungs was observed in both KC^−/− and KC^+/+ mice (Fig. 1D, 1E). We also measured MPO activity in the lung parenchyma of infected KC^−/− and KC^+/+ mice, and observed decreased activity in the lungs of KC^−/− mice as compared with their WT counterparts (Fig. 1F), demonstrating attenuated neutrophil recruitment to the lung parenchyma.

Because the impairment in K. pneumoniae-induced neutrophil accumulation observed in KC^−/− mice could reflect a decrease in the levels of cytokines/chemokines, we quantified the expression of TNF-α, MIP-2, and LIX in BALF at 24 and 48 h after K. pneumoniae challenge. The levels of MIP-2 were found to be attenuated in KC^−/− mice at both 24 and 48 h after K. pneumoniae infection, whereas the levels of LIX were found to be decreased at 24 h postinfection (Fig. 2A, 2B). However, TNF-α levels in BALF were not different between KC^−/− and KC^+/+ mice (Fig. 2C).

Previous reports have demonstrated that NF-κB is a major regulator of cytokines and chemokines, we next examined whether KC modulates NF-κB activation post-K. pneumoniae challenge. To elucidate NF-κB activation, we first determined the effect of K. pneumoniae-induced KC on the phosphorylation of IKK, the upstream kinase of NF-κB and Iκ-Bα. Our data revealed activation of IKKα/β by KC. In a similar manner, we observed an increase in NF-κB /p65 phosphorylation at Ser536 at 48 h postinfection. Although we also observed a modest decrease in the expression of IKKβ and NF-κB/p65 proteins in KC^−/− mice after K. pneumoniae infection, these changes were not significant as compared with the posttranslational modifications (phosphorylation) of NF-κB (Fig. 3A–D). In addition, we repeatedly could not detect phosphorylated IKKα/β or NF-κB/p65 in the control (saline-challenged) KC^−/− and KC^+/+ animals. Moreover, our time-course experiments revealed a decrease in the phosphorylation of IκBα (Ser32) in the lungs of KC^−/− mice starting from 24 h posttreatment and persisted up to 48 h post-K. pneumoniae infection. Concomitantly, we also observed an increased accumulation of IκBα protein in KC^−/− mice at 48 h post-K. pneumoniae infection (Fig. 3A–D). To confirm decreased activation of NF-κB /p65 in KC^−/− mice after K. pneumoniae infection, we determined p65 DNA binding using the nuclear extracts from the lungs. Our findings revealed a decrease in NF-κB/p65 DNA binding in KC^−/− mice as compared with their wild type counterparts, although this decrease was significant only at a later time point (48 h) (Fig. 3B, 3D).

Decreased neutrophil accumulation in the lungs in response to K. pneumoniae in KC^−/− mice could be attributed to reduced activation of MAPKs, which can then regulate other transcription factors besides NF-κB, such as AP-1 and STAT-1 (24, 25). Therefore, we examined the activation of JNK, ERK, and p38 kinases in the lungs of KC^−/− and KC^+/+ mice after K. pneumoniae infection. Activation of JNK was attenuated both at 24 and 48 h after K. pneumoniae infection, whereas the activation of ERK was reduced only at 48 h (Fig. 3C, 3D). In addition, activation of p38 was attenuated in the lungs of KC^−/− mice at 24 h postinfection (Fig. 3C, 3D).

Successful neutrophil recruitment to the lung in response to infection requires the efficient upregulation of cell adhesion molecules on vascular endothelium and leukocytes. We therefore examined the expression of ICAM-1 and VCAM-1 in lung homogenates after K. pneumoniae infection. In KC^−/− mice, the expression of both ICAM-1 and VCAM-1 was attenuated in the lung at 48 h after K. pneumoniae infection relative to the KC^+/+ mice (Fig. 3C, 3D).

We next studied the relative importance of bone marrow (radiosensitive) versus resident/nonhematopoietic (radioresistant) cell-derived KC in host defense using bone marrow chimeras. Lung CFUs in response to K. pneumoniae infection were increased in irradiated KC^−/− reconstituted with KC^+/+ bone marrow (Fig. 4A). K. pneumoniae-induced total leukocyte and neutrophil influx, as well as KC production, was attenuated in irradiated KC^−/− reconstituted with KC^−/− bone marrow (Fig. 4B–D). As compared with irradiated KC^+/+ mice reconstituted with KC^+/+ bone marrow, irradiated KC^−/− mice reconstituted with KC^+/+ bone marrow or KC^+/+ irradiated mice reconstituted with KC^−/− bone marrow did not show differences in lung CFU, neutrophil influx, or KC levels after K. pneumoniae infection (Fig. 3A–D) As expected, neutrophil recruitment to the lung or KC production was not observed in reconstituted mice after saline challenge (data not shown).

We observed substantial neutrophil recruitment to the lung at 48 h post-K. pneumoniae infection (Fig. 1C–E). Therefore, we depleted neutrophils from KC^+/+ and KC^−/− mice to assess their role in the production of cytokines/chemokines, as well as activation of NF-κB and MAPKs in the lungs. To deplete neutrophils, we treated mice with anti–Gr-1 (RB6-8C5) Ab at 12 and 2 h before K. pneumoniae challenge, whereas isotype-matched Ab was used as a control. Neutrophil depletion did not substantially alter the differences in MIP-2, LIX, and TNF-α levels between KC^+/+ and KC^−/− mice after K. pneumoniae infection (Fig. 5). In a similar manner, activation of NF-κB and MAPKs, and expression of VCAM-1 and ICAM-1 was not different between KC^+/+ and KC^−/− mice postinfection (Fig. 6).

We next assessed whether KC deficiency has any effect on activation of NF-κB and MAPKs in myeloid cells. In particular, we have used macrophages because these cells are known to mediate host defense. When KC^−/− macrophages were stimulated with K. pneumoniae, reduced activation of NF-κB, as well as JNK, ERK, and p38, was observed (Fig. 7A, 7B). We observed attenuated MIP-2 production in macrophages obtained from KC^−/− mice, although TNF-α levels between KC^−/− and KC^+/+ mice were not different after K. pneumoniae infection (Fig. 7C–E). To examine the basal levels of cytokines/chemokines in BMDMs, we measured the protein levels of MIP-2, KC, and TNF-α. Our results also show that BMDMs release KC, MIP-2, and TNF-α during differentiation on day 7 in the absence of K. pneumoniae infection, although these levels are much less than in the presence of K. pneumoniae infection (Fig. 8A–C).

To test the hypothesis that KC-induced production of cytokine/chemokines and activation of NF-κB and MAPKs involves autocrine and paracrine responses in BMDMs, we used a KC neutralizing Ab. As demonstrated in Figs. 8 and 9, blocking KC reduced release of KC and MIP-2, but not TNF-α (Fig. 8D, 8E), and activation of NF-κB and MAPKs (Fig. 9). We used murine AMs to confirm the findings obtained from BMDMs, and demonstrated that KC and MIP-2 production and activation of NF-κB and MAPKs were attenuated in AMs obtained from KC^−/− mice as compared with cells obtained from their littermate control mice (Fig. 10).

Community-acquired pneumonia caused by K. pneumoniae is a leading cause of death due to extensive lung damage (26). Unlike many other Gram-negative bacterial pathogens, K. pneumoniae can induce disseminated infection even with a relatively low dose. Although initial treatment of community-acquired pneumonia is empirical antibiotic coverage, increasing numbers of immunocompromised patients, a growing proportion of elderly individuals, and the emergence of antibiotic-resistant K. pneumoniae makes this treatment difficult. Therefore, modulation of the host immune response itself is an attractive option for clinical management, particularly among patients with immune system impairment or infected with antibiotic-resistant bacteria. Development of more effective agents requires a better understanding of the mechanisms by which innate defense against bacteria is regulated.

Successful host defense against bacteria involves the brisk recruitment of neutrophils because they phagocytose and eliminate microbes. Neutrophil migration to the lungs involves the expression of chemokines. Earlier reports have demonstrated that KC, MIP-2, and LIX are important for neutrophil accumulation in the lungs (11, 13–15). These investigations have primarily been performed using blocking Abs. Subsequently, several KC transgenic mouse strains have been used to demonstrate the importance of KC in neutrophil influx in the lungs, although the transgene expression was expressed in all tissues, including the lung (27, 28). Therefore, a lung-specific KC transgene-expressing mouse was later developed to demonstrate the importance of KC in K. pneumoniae-induced neutrophil recruitment and bacterial clearance. An important challenge that remains is the determination of the KC-mediated signaling mechanisms in the lungs during pulmonary bacterial infection. In this report, we have used a loss-of-function model using mice deficient in KC to delineate the mechanisms associated with K. pneumoniae-induced neutrophil-mediated bacterial clearance in the lungs. This model provided an opportunity to address the mechanisms related to KC-induced, neutrophil-mediated host defense in the lungs against bacterial pathogens, including K. pneumoniae. Using this murine model, we confirmed the previous observations (27, 28) that KC is essential for survival and the clearance of K. pneumoniae from the host.

A role for chemokines to amplify the inflammation in the lungs has been documented in an investigation using KC transgenic mice (28). That report provided evidence that KC regulates the expression of MIP-2 in the lungs in KC transgenic mice infected with K. pneumoniae, whereas KC does not regulate TNF-α, IFN-γ, IL-12, and IL-10 (28). Our data show that KC also regulates other chemokines, such as LIX, in response to K. pneumoniae infection. These observations reveal that KC can augment neutrophil-mediated lung inflammation and host defense against bacteria via the upregulation of MIP-2 and LIX. These findings support that KC not only has a direct role but also has an indirect role via upregulating other proinflammatory mediators in the lungs to augment host defense during bacterial infections.

Neutrophil infiltration into mucosal epithelia is a hallmark of bacterial pneumonia, and this involves a multistep paradigm. Neutrophils in the bloodstream adhere to the vascular endothelium and migrate across the endothelium to reach sites of inflammation in various organs, including the lungs (29). This process is complex and involves multiple adhesion molecules from different families that are expressed on the surface of endothelial cells and leukocytes in response to different cytokines. These endothelial proteins recognize specific leukocyte glycoproteins and facilitate their reversible adherence to the vessel wall, known as tethering. This is followed by firm attachment of neutrophils to the endothelium and eventual extravasation through diapedesis. In this investigation, we identified that K. pneumoniae regulates the expression of VCAM-1 and ICAM-1 after K. pneumoniae infection, and this is one of the KC-mediated defense mechanisms involved in neutrophil accumulation and bacterial clearance in the lungs.

A central feature of inflammation is the activation of transcription factors. Among these, NF-κB is well studied and becomes activated in response to stress, cytokines, free radicals, UV irradiation, oxidized low-density lipoprotein, and bacterial or viral Ags. NF-κB activation has been reported to contribute to neutrophil accumulation in the lungs by bacteria via multiple mechanisms, including upregulation of cell adhesion molecules (VCAM-1 and ICAM-1). In this regard, studies demonstrate that endogenous NF-κB protects the host by preventing excessive inflammation and injury during E. coli (30) and pneumococcal pneumonia (31). Moreover, NF-κB is shown to be essential for antibacterial host defense (30, 31). Our findings from this study support an important role for KC in late NF-κB activation (48 h) in the lungs after K. pneumoniae infection. In addition, our findings also suggest the involvement of other transcription factors such as AP-1/STAT-1 because MAPKs are known to regulate AP-1 and STAT-1 (32), and our results reveal that KC mediates activation of all three MAPKs at a late time point (48 h).

The relative contribution of myeloid cells (macrophages and neutrophils) versus resident tissue cells in neutrophil accumulation into the lung from the bloodstream has not been clear. However, the contribution of specific cell type is dependent on the stimulus, because both of these cell types are exposed to inflammatory components during an insult. In earlier studies, it has been proposed that myeloid cells in the lung produce a battery of neutrophil chemotactic substances such as KC (10, 11) and MIP-2 (12, 13), whereas resident cells, including alveolar epithelial type II cells, produce other neutrophil chemoattractants, such as LIX (14). An interesting finding of this investigation was the dependence on both hematopoietic/myeloid cell and resident/nonhematopoietic cell-derived KC for host defense during K. pneumoniae infection. The conclusions obtained from this investigation are consistent with prior reports of the role of hematopoietic and nonhematopoietic cells in pulmonary inflammation. For example, an earlier study has shown that MyD88 derived from hematopoietic cells is more important for LPS-induced expression of TNF-α and IL-12p40 (33), although both hematopoietic and resident cell-derived MyD88 are important for LPS-induced neutrophil influx (31, 34, 35). We have shown earlier that MD-2 in both cell types is important for neutrophil-mediated inflammation, and the expression of MIP-2, TNF-α, and IL-6 is mediated by both cell types in the lungs after LPS challenge (19).

A key finding of our investigation is that depletion of neutrophils resulted in no substantial change in the production of KC in the lungs. These results are consistent with the fact that several other cell types are involved in the initiation of KC production during pulmonary infectious challenge. However, Kp-induced production of TNF-α was substantially reduced in the lungs of neutrophil-depleted KC^+/+ and KC^−/− mice, whereas Kp-induced TNF-α levels were not different between KC^+/+ and KC^−/− mice. Our observations suggest that neutrophils may be the key source for TNF-α, a cytokine that is not regulated by KC-dependent signaling cascade(s). We used macrophages in our in vitro system because this cell type is a vital mediator of host defense in the lungs and found that macrophages modulate lung inflammation on K. pneumoniae infection via KC-mediated activation of NF-κB and MAPKs. Our data also show that KC regulates MIP-2 in macrophages on K. pneumoniae infection. Because KC controls the production of MIP-2 from macrophages and LIX from alveolar epithelial type II cells in the lung, it is likely that KC-mediates both autocrine and paracrine effects on K. pneumoniae infection. In this regard, we have observed expression of KC and its receptor (CXCR2) in large and small airway cells, alveolar type II cells, macrophages, and neutrophils (data not shown).

KC^−/− mice have been used in a few prior studies: 1) KC/low-density lipoprotein receptor^−/− mice exhibit protective effects against the development of atherosclerosis (36), and 2) KC^−/− mice also show protection against dextran sulfate sodium-induced colitis (17). We took advantage of the gene-deficient mice to demonstrate the host defense mechanisms induced by KC after K. pneumoniae infection. The findings from this study unveil that multiple signaling cascades involving transcription factors, kinases, and the expression of proinflammatory mediators are altered in Gram-negative pneumonia via KC (Fig. 11, proposed model). This investigation offers the first step, to our knowledge, into the biology of KC in mediating pulmonary antibacterial defense on bacterial infection.

We thank Nades Palaniyar (University of Toronto, Toronto, Ontario, Canada) for helpful discussions and Rachel Zemans and Ken Malcolm (National Jewish Health, Denver, CO), Dan Chisenhall, Pete Mottram (Louisiana State University), and Lung Biology laboratory members, and Gayathriy Balamayooran and Himanshu Raje for critical reading of the manuscript.

Disclosures The authors have no financial conflicts of interest.