INFECTION AND IMMUNITY, June 2010, p. 2620–2630 Vol. 78, No. 60019-9567/10/$12.00 doi:10.1128/IAI.01169-09Copyright © 2010, American Society for Microbiology. All Rights Reserved.

Importance of CXC Chemokine Receptor 2 in Alveolar Neutrophiland Exudate Macrophage Recruitment in Response to

Pneumococcal Lung Infection�

Wiebke Herbold,1 Regina Maus,1 Ines Hahn,1 Nadine Ding,1 Mrigank Srivastava,2John W. Christman,3 Matthias Mack,4 Jorg Reutershan,5 David E. Briles,6James C. Paton,7 Christine Winter,1 Tobias Welte,1 and Ulrich A. Maus1*

Department of Pulmonary Medicine, Laboratory for Experimental Lung Research, Hannover School of Medicine, Hannover,Germany1; Dana Farber Cancer Institute, Harvard Medical School, Boston, Massachusetts2; Center for Lung and

Vascular Biology, University of Illinois at Chicago, Chicago, Illinois3; Department of Nephrology, University ofRegensburg, Regensburg, Germany4; Department of Anaesthesiology and Intensive Care Medicine, University of

Tubingen, Tubingen, Germany5; Department of Microbiology, University of Alabama at Birmingham,Birmingham, Alabama6; and Research Centre for Infectious Diseases, School of Molecular and

Biomedical Science, University of Adelaide, Adelaide, Australia7

Received 16 October 2009/Returned for modification 17 December 2009/Accepted 26 March 2010

Sustained neutrophilic infiltration is known to contribute to organ damage, such as acute lung injury. CXCchemokine receptor 2 (CXCR2) is the major receptor regulating inflammatory neutrophil recruitment in acuteand chronic inflamed tissues. Whether or not the abundant neutrophil recruitment observed in severe pneu-monia is essential for protective immunity against Streptococcus pneumoniae infections is incompletely defined.Here we show that CXCR2 deficiency severely perturbs the recruitment of both neutrophils and exudatemacrophages associated with a massive bacterial outgrowth in distal airspaces after infection with S. pneu-moniae, resulting in 100% mortality in knockout (KO) mice within 3 days. Moreover, irradiated wild-type micereconstituted with increasing amounts of CXCR2 KO bone marrow (10, 25, 50, and 75% KO) have corre-spondingly decreased numbers of both neutrophils and exudate macrophages, which is associated with astepwise increase in bacterial burden and a reciprocal stepwise decrease in survival in S. pneumoniae-inducedpulmonary infection. Finally, application of the CXCR2 antagonist SB-225002 resulted in decreased alveolarneutrophil and exudate macrophage recruitment in mice along with increased lung bacterial loads afterinfection with S. pneumoniae. Together, these data show that CXC chemokine receptor 2 serves a previouslyunrecognized nonredundant role in the regulation of both neutrophil and exudate macrophage recruitment tothe lung in response to S. pneumoniae infection. In addition, we demonstrate that a threshold level of 10 to 25%of reduced neutrophil recruitment is sufficient to cause increased mortality in mice infected with S. pneumoniae.

Innate immune responses to acute bacterial infections in thelung are characterized by successive waves of early recruitedpolymorphonuclear neutrophils (PMN) followed by recruitedmononuclear phagocytes (24). Recruitment of neutrophils byalveolar macrophages is essential for reducing the bacterialburden in the alveolar space. Newly recruited exudate macro-phages, which mature into alveolar macrophages, may alsocontribute to bacterial containment and the resolution/repairphase of pulmonary infections (14, 22, 31). Despite their un-questionably important role in lung innate immunity, recruitedneutrophils rapidly undergo apoptosis or necrosis and subse-quently contribute to the formation of consolidated infiltrates,which disrupt alveolar gas exchange and may impair host de-fense functions of the lung (31). In addition, neutrophils, dueto their enormous capacity to release reactive oxygen speciesand proteolytic enzymes, can directly contribute to acute lunginjury as “collateral damage” (1, 13, 32). Even nonlethal chal-

lenge per se of the lung with bacteria and/or bacterial toxinsresults in the recruitment of huge numbers of neutrophils intoboth the lung interstitial and alveolar compartments, which ischaracteristic of acute respiratory distress syndrome (ARDS)(15, 17). However, whether such an uncontrolled neutrophilicburden developing in response to bacterial infections in thelungs of both humans and mice is helpful or harmful for elim-ination of pulmonary pathogens is incompletely defined.

Therapeutic strategies to accelerate the removal of consol-idated infiltrates in infected organs may either target macro-phage pool sizes or aim at inhibiting chemokine receptor sig-naling that results in the neutrophil recruitment and activationprocess. Our group has recently reported that constitutivelyincreased, CC chemokine receptor 2 (CCR2)-dependentmononuclear phagocyte subset recruitment significantly im-proves the lung resolution/repair process and outcome in re-sponse to Streptococcus pneumoniae challenge (31). On theother hand, CXC chemokine receptor 2 (CXCR2) antagonistswere found to reduce neutrophil recruitment and effector cellfunctions in lipopolysaccharide (LPS)-induced lung neutro-philia in rodent models and could have a therapeutic benefit inneutrophilic pulmonary diseases, including chronic obstructivepulmonary disease (COPD) (5, 6).

* Corresponding author. Mailing address: Department of Pulmo-nary Medicine, Hannover School of Medicine, Hannover 30625, Ger-many. Phone: 49-511-532-9617. Fax: 49-511-532-9616. E-mail: Maus.Ulrich@mh-hannover.de.

� Published ahead of print on 5 April 2010.

2620

on June 9, 2020 by guesthttp://iai.asm

.org/D

ownloaded from

The murine CXCR2 (mCXCR2) and its ligands, CXCL1(keratinocyte-derived chemokine [KC]) and CXCL2/3 (macro-phage inflammatory protein 2 [MIP-2]), functional homo-logues of human growth-related oncogenes (GRO�, -�, and-�), have been identified as the primary chemokine receptorsystem mediating neutrophil recruitment in various models ofacute and chronic inflammation, including Escherichia coli en-dotoxin-induced acute lung inflammation and injury (6, 19, 20,23). Both CXCR2 knockout (KO) mice as well as small-mol-ecule inhibitors that interfere with CXCR2 receptor functionhave shown a role for neutrophils in mediating hyperoxic orventilator-induced lung injury and in collagen deposition inbleomycin-induced fibrosis (23). Unlike in influenza viruspneumonia, where CXCR2 is required for neutrophil recruit-ment but not viral clearance, CXCR2 is important for neutro-phil recruitment as well as pathogen elimination in Pseudomo-nas aeruginosa or Nocardia asteroides pneumonia (18, 29).Also, a role for CXCR2 in mediating at least neutrophil re-cruitment and neutrophil-dependent killing in S. pneumoniae-induced lung infection was reported recently (26). However, itis currently not clear whether the abundant neutrophil recruit-ment observed in pneumococcal lung infection is really re-quired for mounting protective immunity against this pathogenand what threshold level of CXCR2 inhibition/deletion causesdecreased lung protective immune responses to challenge withS. pneumoniae. In addition, the role of CXCR2 in regulating S.pneumoniae-induced lung exudate macrophage recruitment isalso unknown. Therefore, wild-type (WT) mice with or withoutpretreatment with CXCR2 antagonist SB-225002 or CXCR2KO mice as well as mixed bone marrow chimeric mice recon-stituted with an increasingly CXCR2-deficient hematopoieticsystem were infected with S. pneumoniae, and neutrophil ver-sus monocyte-derived exudate macrophage recruitment pro-files and the respective CC and CXC chemokine responses aswell as bacterial loads and survival were determined over time.

MATERIALS AND METHODS

Mice. Wild-type BALB/c mice (weight, 20 to 22 g) were purchased fromCharles River (Sulzfeld, Germany). CXCR2 KO mice (CXCR2�/�) on theBALB/c background were purchased from Jackson Laboratories (Bar Harbor,ME). Both female wild-type mice and female CXCR2 KO mice (20 to 22 g) werekept under specific-pathogen-free conditions with free access to autoclaved foodand water and were routinely screened for major pathogens according to theFederation of European Laboratory Animal Science Associations recommenda-tions to exclude any opportunistic infections that would possibly confound theexperimental results. Mice were used in all experiments at 8 to 12 weeks of agein accordance with the guidelines of our Institutional Animal Care and UseCommittee of Hannover School of Medicine, Germany. Animal experimentswere approved by our local government authorities. Experiments were per-formed under blinded conditions.

Reagents. A Roti-Quick kit for total RNA isolation was purchased from CarlRoth (Karlsruhe, Germany). All other reagents for reverse transcription (ran-dom hexamer primers, Moloney murine leukemia virus reverse transcriptase[MMLV-RT], recombinant RNase inhibitor, deoxynucleoside triphosphates, anddithiothreitol) were purchased from Promega (Madison, WI). A SYBR green 1kit was purchased from Eurogentec (Seraing, Belgium). Selective nonpeptideCXCR2 antagonist SB-225002 was purchased from Biozol (Eching, Germany)and was dissolved at 10 mg/ml in ethanol, according to the manufacturer’sinstructions. For in vivo application, aliquots of the inhibitor were dissolved invehicle (0.9% NaCl solution containing 0.33% Tween 80) just before use (3).

Total cellular RNA isolation, cDNA synthesis, and real-time RT-PCR. Totalcellular RNA was isolated from bone marrow cells using the Roti-Quick kit(Roth) following the instructions of the manufacturer. RNA quantification andpurity were determined on an Agilent Bioanalyzer 2100 (Agilent Biosystems),

and only those RNA preparations exceeding absorbance ratios (A260/A280) of�1.90 were further processed for reverse transcription and real-time RT-PCRexperiments. Synthesis of cDNA and real-time RT-PCR experiments were per-formed as recently described (25). Primers for detection of mCXCR2 mRNAwere purchased from Operon Biotechnologies (Cologne, Germany) and had thefollowing sequences: forward primer, 5�-AGCAAACACCTCTACTACCCTCTA-3�; reverse primer, 5�-GGGCTGCATCAATTCAAATACCA-3�. Real-timeRT-PCR experiments were run on an Applied Biosystems ABI-7300 analyzer.

Culture and quantification of Streptococcus pneumoniae. We used a capsulartype 19 S. pneumoniae strain (EF3030) for infection experiments. This strain hasbeen demonstrated to primarily cause lobar pneumonia in mice without bacte-remia (4, 30). S. pneumoniae was grown in Todd-Hewitt broth (THB; Difco)supplemented with 0.1% yeast extract to mid-log phase, and aliquots were snap-frozen in liquid nitrogen and stored at �80°C until use, essentially as describedrecently (14, 27). Quantification of S. pneumoniae stocks was done by platingserial dilutions on sheep blood agar plates (BD Biosciences, Heidelberg, Ger-many), followed by incubation of the plates at 37°C with 5% CO2 for 18 h and thedetermination of CFU.

Generation of bone marrow chimeric mice. Bone marrow chimeric mice weregenerated as recently described (16, 27). Briefly, bone marrow cells were isolatedunder sterile conditions from tibias and femurs of sex-matched wild-type andCXCR2-deficient donor mice. Single-cell suspensions were carefully preparedfrom the bone marrow isolates and filtered through 70- and 40-�m nylon meshes(BD Biosciences, Heidelberg, Germany) to remove residual cell aggregates. Thecells were washed in RPMI 1640 supplemented with 10% fetal calf serum(Gibco) before transplantation. Recipient wild-type mice received a total bodyirradiation of 8 Gy at a dose rate of 1.8 Gy/min delivered by a linear accelerator(Siemens MD 2) operating in a 6-MV high-energy photon delivery mode (27).Subsequently, sedated wild-type mice received donor bone marrow cell prepa-rations collected from wild-type and CXCR2 KO mice mixed at ratios of 90:10;75:25; 50:50, and 25:75 (WT:KO ratios; 1 � 107 cells/mouse) with Leibovitz L-15medium (Gibco) via lateral tail vein injections. Resulting mixed chimeric wild-type mice were then housed under specific-pathogen-free (SPF) conditions withfree access to autoclaved food and water for at least 6 weeks prior to infectionexperiments.

Experimental groups. The following experimental groups were defined: Mock-infected mice received intratracheal instillations of vehicle (50 �l of THB/mouse)and were analyzed at 24 h posttreatment. Wild-type mice and CXCR2 KO micewere either mock infected or were infected with serotype 19 S. pneumoniae(3 � 106 CFU/mouse) and were analyzed at the indicated time points postin-fection. Moreover, groups of irradiated wild-type mice transplanted with mix-tures of CXCR2 KO and wild-type bone marrow cells (at ratios of 10:90, 25:75,50:50, 75:25 [KO:WT]; 1 � 107 cells/mouse) (Table 1) were infected with sero-type 19 S. pneumoniae (3 � 106 CFU/mouse) and subsequently analyzed forbacterial load, inflammatory leukocyte recruitment, and CC and CXC chemo-kine profiles, as indicated. In selected experiments, wild-type mice receivedintraperitoneal applications of vehicle or CXCR2 antagonist SB-225002 (1.5 or15 �g/g of body weight) twice daily prior to infection of mice with S. pneumoniae.Twenty-four hours after infection, bronchoalveolar lavage (BAL) fluid and lungtissue CFU counts and numbers of alveolar recruited neutrophils and exudatemacrophages were determined.

Infection of mice with Streptococcus pneumoniae. CXCR2 KO mice, wild-typemice, and chimeric wild-type mice were infected with serotype 19 S. pneumoniae,using freshly prepared dilutions of thawed aliquots, as recently described (14,27). For infection, mice were fixed in a vertical position followed by visualizationof the pharynx via transillumination of the neck with a cold light source (Leica,Wetzlar, Germany). Subsequently, the trachea was intubated with a 26-gaugecatheter (Abbocath) followed by careful intratracheal instillation of S. pneu-moniae into the lungs. After instillation, mice were kept in individually ventilatedcages (IVC) with free access to autoclaved food and water and were monitoredtwice daily for disease symptoms and survival during the entire observationperiod (2).

Determination of bacterial loads in bronchoalveolar lavage fluid and lungtissue homogenate. Bacterial loads in the lungs of S. pneumoniae-infected miceof the various experimental groups were determined in both BAL fluid and lungtissue homogenates. Briefly, mice were euthanized with an overdose of isoflu-rane, and tracheas of the mice were exposed and cannulated with a shortened20-gauge needle that was firmly fixed to the trachea. Subsequently, 300-�l ali-quots of ice-cold sterile phosphate-buffered saline (PBS) were instilled andcarefully aspirated until a first BAL fluid volume of 1.5 ml was collected. Sub-sequently, the collection of BAL was continued until an additional BAL fluidvolume of 4.5 ml was collected. The 1.5-ml and 4.5-ml BAL fluid samples (wholelung washes) collected from S. pneumoniae-infected mice of either treatment

VOL. 78, 2010 CXCR2 IN PNEUMOCOCCAL LUNG INFECTION 2621

on June 9, 2020 by guesthttp://iai.asm

.org/D

ownloaded from

group were immediately processed for determination of bacterial loads by plating100 �l of the respective BAL fluid aliquot in 10-fold serial dilutions on sheepblood agar plates followed by incubation of the plates at 37°C with 5% CO2 for18 h. Subsequently, CFU were counted and bacterial loads in whole lung washeswere calculated. Whole lung washes were further subjected to centrifugation at1,400 rpm (4°C, 10 min), and cell pellets were pooled to determine total numbersof BAL fluid leukocytes. In addition, BAL fluid CC (CCL2 [JE, monocytechemoattractant protein 1, or MCP-1], and CCL12 [MCP-5]) and CXC chemo-kines (CXCL1 [KC] and CXCL2 [MIP-2]) were measured in cell-free BAL fluidsupernatants of the individual 1.5-ml BAL fluid aliquots. Subsequent to thecollection of BAL fluid, lung tissue was homogenized in 2 ml of Hanks balancedsalt solution without supplements using a tissue homogenizer (IKA, Staufen,Germany), and 10-fold serial dilutions of the lung tissue homogenates wereplated on sheep blood agar plates followed by incubation of the plates at 37°Cwith 5% CO2 for 18 h for CFU determination. Since the chimeric wild-type micetransplanted with 75% CXCR2-deficient bone marrow demonstrated 100% mor-tality during day 3 after infection with S. pneumoniae, determinations of leuko-cyte counts and CFU in BAL fluids and lung tissue of the 75% chimeric micewere done immediately after reaching the 72-h time point of infection, in orderto avoid unnecessary repetition of this experiment due to increased mortality.

Immunophenotypic analysis of mononuclear phagocyte subsets in bronchoal-veolar lavage fluid. Mononuclear phagocyte subset populations (alveolar mac-rophages and alveolar exudate macrophages) contained in the bronchoalveolarlavage fluid of mock-infected or S. pneumoniae-challenged mice of the variousexperimental groups were subjected to immunophenotypic analysis of their cellsurface antigen expression profiles, using a BD FACSCanto flow cytometerequipped with an argon ion laser operating at a 488-nm excitation wavelengthand a helium neon laser operating at a 633-nm wavelength (BD Biosciences,Heidelberg, Germany), according to recently published protocols (14, 27, 30).Briefly, cells preincubated with octagam (Octapharma, Langenfeld, Germany)were stained for 15 min at 4°C with various combinations of appropriately dilutedfluorochrome-conjugated monoclonal antibodies with specificities for the follow-ing cell surface molecules: phycoerythrin (PE)-Cy7-conjugated anti-CD11b, PE-Cy5.5-conjugated anti-CD11c, PE-conjugated anti-CD86, PE-conjugated anti-major histocompatiblity complex (MHC) class II (all from BD Biosciences), andallophycocyanin (APC)-conjugated anti-F4/80 (Serotec, Dusseldorf, Germany).Subsequently, cells were washed in PBS–0.1% bovine serum albumin (BSA)–0.02% sodium azide, and cell acquisition was performed on a BD FACSCantoflow cytometer (BD Biosciences). Gating of the respective mononuclear phago-cyte subsets in BAL fluid was done according to their forward scatter area(FSC-A) versus side scatter area (SSC-A) characteristics and FSC-A versusF4/80-APC fluorescence emission characteristics to exclude debris and neutro-phils from further analysis. Resident alveolar macrophages, which were collectedby bronchoalveolar lavage, were identified according to their CD11cpos,CD11bneg/low, CD86neg, and MHC-IIneg cell surface antigen expression profile.In contrast, alveolar exudate macrophages contained in BAL fluid were identi-fied according to their cell surface expression of CD11cpos, CD11bhigh, CD86neg,and MHC-IIneg, essentially as described recently (14). Data analysis and carefulpostacquisition compensation of spectral overlaps between the various fluores-cence channels were performed using BD FACSDiva software (BD Biosciences).For quantification of neutrophils in BAL fluids, cells were counted in Pappen-heim-stained cytocentrifuge preparations according to their staining pattern andmorphological criteria, including their nuclear shape, followed by multiplicationof this value with total BAL fluid cell numbers. In selected experiments, wedetermined CXCR2 and CCR2 chemokine receptor expression profiles on pe-ripheral blood neutrophils and inflammatory monocytes or alveolar exudatemacrophages. Briefly, cells were gated according to their FSC area versus SSCarea characteristics, followed by hierarchical subgating of the Gr-1high F4/80neg

neutrophils or the Gr-1pos F4/80pos inflammatory monocytes (13) or alveolarexudate macrophages, as described above, and subsequent determination of theirCXCR2 and CCR2 receptor expression profiles. Anti-CCR2 antibody MC21 wasemployed for these experiments as described recently (13, 15). Detection ofCXCR2 expression on neutrophils and exudate macrophages was done using apreviously described protocol (16). Briefly, after incubation with Fc block (10 �l;BD Biosciences), cells were incubated with anti-CXCR2 antibody or isotypecontrol antibody (R&D Systems, Wiesbaden, Germany), washed three times,and incubated on ice with biotinylated F(ab�)2 for 30 min. Cells were thenwashed and stained with PE-conjugated streptavidin (BD Biosciences) for 15 minon ice in the dark.

ELISA. CC chemokines CCL2 and CCL12 and CXC chemokines CXCL1 andCXCL2 were determined in BAL fluids of mock-infected or S. pneumoniae-infected wild-type mice or CXCR2 KO mice or the various groups of chimericwild-type mice by using commercially available enzyme-linked immunosorbentassays (ELISAs) according to the manufacturer’s instructions (R&D Systems,Wiesbaden, Germany).

Statistics. All data are given as means standard errors of the means (SEM).Differences between treatment groups over time were analyzed by analysis ofvariance followed by a post hoc Dunnett test. Significant differences betweengroups were analyzed by Levene’s test for equality of variances followed byStudent’s t test using SPSS 15.0 for Windows software package. Survival curveswere compared by the log-rank test. Statistically significant differences betweengroups were assumed when P values were �0.05.

RESULTS

CXCR2 KO mice are highly susceptible to infection with S.pneumoniae. To determine the role of CXCR2 in lung protec-tive immunity to pneumococcal infection, wild-type mice andCXCR2 KO mice were challenged with serotype 19 S. pneu-moniae. As shown in Fig. 1, low-dose pneumococcal infectionof wild-type mice with 3 � 106 CFU/mouse did not cause anymortality during an observation period of 8 days. In sharpcontrast, CXCR2 KO mice showed a progressive mortalitystarting at day 2 postinfection, with 100% mortality in thisgroup by day 4 of infection (Fig. 1A). Bacterial loads in BALfluids and lung tissue homogenates of wild-type mice rapidlydeclined within 3 days of infection (Fig. 1B and C), while at thesame time, CXCR2 KO mice demonstrated a drastic out-growth of S. pneumoniae in BAL fluids and lung parenchymaltissue up until day 3 postinfection (Fig. 1B and C), which wasaccompanied by a sharp decline in survival just on days 3 and4 postinfection (Fig. 1A to C). Such immediate early mortalityobserved in the CXCR2 KO mice suggests impaired innateimmune responses to infection with S. pneumoniae, therebystrongly impacting survival.

Deletion of CXCR2 attenuates alveolar neutrophil andexudate macrophage recruitment in response to S. pneu-moniae infection. Alveolar macrophages and neutrophils arecentrally involved in uptake and elimination of inhaled bacte-ria (7, 24). Therefore, we next determined the impact of

TABLE 1. Treatment groups and hematopoietic CXCR2 expression profiles

Group TreatmentHematopoietic

CXCR2expression (%)

WT None 100CXCR2 KO None 0WT with WT transplant Irradiation, reconstitution with 100% WT bone marrow (BM) 100Chimeric WT, 10% Irradiation, reconstitution with 90% WT and 10% CXCR2 KO BM 90Chimeric WT, 25% Irradiation, reconstitution with 75% WT and 25% CXCR2 KO BM 75Chimeric WT, 50% Irradiation, reconstitution with 50% WT and 50% CXCR2 KO BM 50Chimeric WT, 75% Irradiation, reconstitution with 25% WT and 75% CXCR2 KO BM 25

2622 HERBOLD ET AL. INFECT. IMMUN.

on June 9, 2020 by guesthttp://iai.asm

.org/D

ownloaded from

CXCR2 deletion on profiles of resident and inflammatory leu-kocyte subsets in BAL fluids of wild-type and CXCR2 KO micechallenged with S. pneumoniae. Both mock-infected wild-typemice and CXCR2 KO mice had similar numbers of residentalveolar macrophages and few recruited exudate macrophagesin their BAL fluids, as shown in Fig. 2A and B (0-day timepoints). In response to low-dose challenge with S. pneumoniae,both wild-type and CXCR2 KO mice developed significantlyreduced numbers of resident alveolar macrophages by days 1and 2 postinfection compared to mock-infected mice on day 0,

most probably due to S. pneumoniae-induced apoptosis induc-tion in these cells, according to recent reports (7, 27) (Fig. 2Aand B). Unexpectedly, however, we observed that unlike wild-type mice, which developed a coordinated recruitment of ex-udate macrophages into their lungs in response to S. pneu-moniae, CXCR2 KO mice had significantly reduced numbersof exudate macrophages in the bronchoalveolar compartment,

FIG. 1. Survival and bacterial pathogen elimination in wild-typemice and CXCR2 KO mice infected with S. pneumoniae. (A) Wild-type mice and CXCR2 KO mice were infected with serotype 19 S.pneumoniae (3 � 106 CFU/mouse), and survival (12 mice per group)was monitored during an observation period of 8 days. (B and C)Bacterial loads were determined in bronchoalveolar lavage fluids andlung parenchymal tissue of wild-type and CXCR2 KO mice on days 1to postinfection, as indicated. The data are means SEM (n � 5 to 9mice per time point and treatment group). Asterisks indicate a signif-icant difference compared to the wild type (*, P � 0.05; **, P � 0.01;***, P � 0.001). Plus signs indicate a significant difference comparedto the respective day zero value ( , P � 0.05; , P � 0.01; ,P � 0.001).

FIG. 2. Analysis of professional phagocyte subsets in the lungs ofwild-type mice and CXCR2 KO mice infected with S. pneumoniae.Wild-type mice and CXCR2 KO mice were infected with serotype 19S. pneumoniae (3 � 106 CFU/mouse). At the indicated time points,mice of either treatment group were subjected to bronchoalveolarlavage and recruitment profiles of resident alveolar macrophages (A),inflammatory recruited alveolar exudate macrophages (B), and neu-trophils (C) were analyzed. The data shown are means SEM (n � 5to 9 mice per time point and treatment group). Asterisks indicate asignificant difference compared to the wild type (**, P � 0.01; ***,P � 0.001). Plus signs indicate a significant difference compared to therespective mock-infected mice at the day zero time point ( , P � 0.05; , P � 0.01; , P � 0.001).

VOL. 78, 2010 CXCR2 IN PNEUMOCOCCAL LUNG INFECTION 2623

on June 9, 2020 by guesthttp://iai.asm

.org/D

ownloaded from

particularly on days 2 and 3 postinfection (Fig. 2B). Further-more, repopulation of the alveolar air space with alveolar mac-rophages, as observed in S. pneumoniae-infected wild-typemice, was significantly impaired in the CXCR2 KO mice onday 3 postinfection (Fig. 2A). In addition, CXCR2 KO micehad a markedly reduced neutrophil recruitment in response toS. pneumoniae challenge on days 1 and 2 postinfection com-pared to the wild-type control mice (Fig. 2C). Together, thesedata show that deletion of CXCR2 causes severe perturbationsof both neutrophil and exudate macrophage trafficking in re-sponse to pneumococcal infections, which results in a markeddiminution of the entire pulmonary macrophage population ininfected lungs of CXCR2 KO mice.

Increasing levels of deletion of hematopoietic CXCR2 resultin an increasing loss of lung protective immunity to pneumo-coccal challenge. Although acute bacterial infections of thelung are characterized by a dramatic recruitment of neutro-phils into both the lung parenchymal tissue and bronchoalveo-lar space (24), the pathophysiological value of an abundantneutrophilic response is currently unclear. Since global dele-tion of CXCR2 in mice led to 100% mortality within 3 to 4days of infection, we next examined (i) at what thresholdlevel of hematopoietic CXCR2 ablation a reduced lung neu-trophil and/or exudate macrophage burden would be ob-served and (ii) to what degree partial CXCR2 ablation af-fected lung protective immune responses to challenge withS. pneumoniae. We therefore generated chimeric wild-typemice reconstituted with a successively CXCR2-deleted he-matopoietic system (Table 1).

Figure 3A shows a representative real-time RT-PCR ampli-fication curve of CXCR2 mRNA levels in wild-type mice com-pared to CXCR2 KO mice and chimeric wild-type mice with a50% CXCR2-negative hematopoietic system. Moreover, wealso determined baseline expression profiles of chemokine re-ceptors CXCR2 and CCR2 on the cell surface of peripheralblood neutrophils and monocytes, as well as on the cell sur-faces of recruited neutrophils and exudate macrophages inBAL fluids of S. pneumoniae-infected mice (13). As shown inFig. 3, Gr-1high peripheral blood neutrophils from wild-typemice (Fig. 3B), but not CXCR2 KO mice (Fig. 3G), demon-strated strong CXCR2 expression, while neutrophils collectedfrom chimeric wild-type mice with a 50% CXCR2-negativehematopoietic system demonstrated an 1:1 distribution ofCXCR2-positive to CXCR2-negative Gr-1high neutrophils(Fig. 3I), thus confirming the validity of the chosen experimen-tal approach. Moreover, we found that inflammatory Gr-1pos

blood monocytes from wild-type mice (Fig. 3D) or CXCR2 KOmice (Fig. 3H) demonstrated similar cell surface expression ofCCR2 but not of CXCR2 (Fig. 3C). However, in response topneumococcal challenge, alveolar recruited neutrophils main-tained their cell surface CXCR2 expression (Fig. 3E), whereasalveolar recruited exudate macrophages (Fig. 3F), in contrastto their circulating precursors, demonstrated increasedCXCR2 expression, suggesting that pneumococcal infectiontriggered increased CXCR2 expression on the cell surface ofrecruited exudate macrophages but not circulating inflamma-tory monocytes.

We next determined the effect of graded CXCR2 ablationsin peripheral blood on the lung host defense against pneumo-coccal infection. As shown in Fig. 4A, we found that more than

90% of irradiated wild-type mice reconstituted with 100%wild-type bone marrow (serving as transplantation controls)survived a low-dose infection with serotype 19 S. pneumoniae,which was nonsignificantly reduced compared to mock-in-fected wild-type mice (Fig. 4A versus 1A). Unexpectedly, how-ever, we found that just a 10% CXCR2 ablation in the hema-topoietic system (90:10 WT:KO) led to a 20% reduced survivalof mice, relative to transplantation controls (P � 0.05) (Fig.4A). Moreover, mice reconstituted with a 75:25 WT:KO bonemarrow mixture demonstrated a sharp decline in survival, re-sulting in 50% mortality by day 4 of infection relative totransplantation controls and further progressing toward 80%mortality at day 7 postinfection. In addition, mice transplantedwith a 50:50 or 25:75 WT:KO bone marrow mixture demon-strated an overall mortality of 100% on days 5 and 3, respec-tively, in response to low-dose infection with S. pneumoniae(Fig. 4A).

Analysis of pneumococcal loads in BAL fluids and lungparenchymal tissue of chimeric wild-type mice on days 1 and 3after infection with S. pneumoniae showed increased bacterialcounts starting on day 1 and particularly by day 3 postinfection(Fig. 4B and C) relative to transplantation controls. Morespecifically, a less pronounced, but detectable, bacterial out-growth was observed in the 10% and 25% chimeric wild-typemice, whereas 50% and 75% chimeric wild-type mice demon-strated a significantly increased bacterial outgrowth both inBAL fluids and lung parenchymal tissue relative to transplan-tation controls (Fig. 4B and C).

We next examined the effect of partial CXCR2 ablation onnumbers of alveolar macrophages, exudate macrophages, andneutrophils in the various experimental groups on days 1 and 3of infection with S. pneumoniae. As shown in Fig. 4D, alltreatment groups demonstrated reduced numbers of alveolarmacrophages at day 1 postinfection, consistent with recentreports (27), which were found to again increase toward base-line levels in the transplantation controls and the 10% and25% CXCR2-ablated chimeric wild-type mice. In contrast, the50% and 75% CXCR2-ablated chimeric wild-type mice dem-onstrated highly reduced numbers of alveolar macrophages intheir BAL fluids by day 3 of infection (Fig. 4D), demonstratingthat a 50% CXCR2 ablation is sufficient to perturb alveolarmacrophage repopulation in response to S. pneumoniae.

We observed that exudate macrophage recruitment, whichtypically develops after 3 to 4 days of bacterial infection (31),increased both in transplantation controls as well as the 10%and 25% CXCR2-ablated chimeric mice from day 1 until day3 of infection, whereas strongly and significantly reduced num-bers of exudate macrophages were noted in the 50% and 75%CXCR2-ablated mice by days 1 and 3 postinfection (Fig. 4E).Analysis of neutrophil recruitment patterns in mice of thevarious treatment groups revealed a nearly linear inverse re-lationship between the degree of CXCR ablation and the num-bers of BAL fluid neutrophils in S. pneumoniae-infected chi-meric mice (Fig. 4F). Transplantation controls and 10%CXCR2-ablated chimeric wild-type mice demonstrated nearlysimilar degrees of alveolar neutrophil recruitment in responseto pneumococcal lung infection, whereas the 50% and 75%CXCR2-ablated chimeric wild-type mice, in particular, showeda nearly 70% decreased alveolar neutrophil recruitment atdays 1 and 3 of infection with S. pneumoniae (Fig. 4F).

2624 HERBOLD ET AL. INFECT. IMMUN.

on June 9, 2020 by guesthttp://iai.asm

.org/D

ownloaded from

FIG. 3. CXCR2 gene and protein expression patterns in wild-type mice, chimeric wild-type mice, and CXCR2 KO mice. (A) Bone marrow cellswere isolated from wild-type mice or chimeric wild-type mice (50:50 KO:WT) or CXCR2 KO mice and subjected to analysis of CXCR2 geneexpression by real-time RT-PCR analysis. (B to I) Flow cytometric analysis of CXCR2 expression on blood neutrophils of wild-type mice (B andE), CXCR2 KO mice (G), or chimeric wild-type mice (50:50, KO:WT) (I), or on blood monocytes (C) or alveolar recruited exudate macrophagesfrom S. pneumoniae-infected wild-type mice (48 h) (F). (D and H) CCR2 protein expression on Gr-1pos inflammatory monocytes of untreatedwild-type mice (D) or CXCR2 KO mice (H). Gray histograms, negative controls; lined overlays, leukocyte-specific chemokine receptor expression(except for panel I, where the gray histogram peaks represent CXCR2-negative and -positive neutrophils from chimeric wild-type mice, asindicated). Real-time PCR and fluorescence-activated cell sorting profiles are representative of at least five independent determinations per group.

VOL. 78, 2010 CXCR2 IN PNEUMOCOCCAL LUNG INFECTION 2625

on June 9, 2020 by guesthttp://iai.asm

.org/D

ownloaded from

2626 HERBOLD ET AL. INFECT. IMMUN.

on June 9, 2020 by guesthttp://iai.asm

.org/D

ownloaded from

Finally, we determined the effect of a selective nonpeptideCXCR2 antagonist on bacterial loads and lung leukocyte re-cruitment profiles in mice infected with S. pneumoniae. Asshown in Fig. 4G to J, application of the CXCR2 inhibitorSB-225002 in mice dose dependently reduced numbers of bothalveolar recruited neutrophils and exudate macrophages andled to significantly increased bacterial loads as early as 24 hpost-pneumococcal challenge (Fig. 4I and J). Collectively,these data add to the findings in chimeric mice and support theconcept that interfering with CXCR2 signaling may severelyperturb the lung host defense against inhaled bacterial patho-gens.

Effect of a gradual CXCR2 ablation on inflammatory cyto-kine release in S. pneumoniae-infected mice. We finally deter-mined inflammatory CXC (KC [CXCL1] and MIP-2 [CXCL2])and CC chemokine (MCP-1 [CCL2] and MCP-5 [CCL12])responses in BAL fluids of wild-type and CXCR2 KO mice aswell as the various groups of chimeric wild-type mice. Figure5A and B demonstrates that relative to wild-type mice, CXCR2KO mice had drastically increased neutrophil chemokine re-lease in their BAL fluids (up to 10,000 pg/ml) on days 1 to 3 ofinfection with S. pneumoniae (Fig. 5A and B). In addition, wealso found highly increased monocyte chemokines (CCL2 andCCL12) (Fig. 5C and D) in BAL fluids of CXCR2 KO mice butnot wild-type mice on days 1 to 3 of infection, reflective ofuncontrolled CXC and CC chemokine storms in CXCR2-de-ficient mice in response to S. pneumoniae. As observed inwild-type mice, transplant controls also demonstrated low lev-els of neutrophil and monocyte chemoattractants in their BALfluids upon infection with S. pneumoniae (Fig. 5E to H).

We found a nearly linear relationship between the degree ofhematopoietic CXCR2 ablation and the amount of both neu-trophil and monocyte chemoattractants released into the bron-choalveolar compartment: 10% CXCR2-deficient chimericwild-type mice demonstrated low levels of both CXC (KC andMIP-2) and CC (CCL2 and CCL12) chemokines in their BALfluids, whereas 50% and 75% CXCR2-ablated chimeric wild-type mice responded with a strongly increased release of CXCand CC chemokines upon infection with S. pneumoniae (Fig.5E to H). Importantly, the strongest chemotactic responses in50% and 75% chimeric wild-type mice were observed just onday 3 postinfection and coincided with the highest bacterialloads, lowest neutrophil and exudate macrophage responses,and a rapid decline in survival rates in these groups. Together,these data demonstrate that successive deletion of hematopoi-etic CXCR2 leads to a stepwise increased release of not only

neutrophil but also monocyte chemotactic activities in the BALfluids of mice infected with S. pneumoniae, with the highestchemokine responses noted in the mutant mice upon infectionwith S. pneumoniae. These data illustrate for the first timeimportant effects of CXCR2 on both neutrophil and monocytechemotactic and cellular responses to infection with S. pneu-moniae.

DISCUSSION

The purpose of the current study was to determine whetherpartial CXCR2 ablation would reduce neutrophil burdens inthe lungs of mice without immediately impairing lung protec-tive immunity to challenge with inhaled S. pneumoniae. Weinfected wild-type mice and CXCR2 KO mice and a series ofincreasingly CXCR2-ablated chimeric wild-type mice with alow dose of S. pneumoniae. Our results indicated that CXCR2is absolutely required for protective immunity against S. pneu-moniae. Deletion of CXCR2 resulted in an excessive earlymortality by 3 to 4 days after infection, accompanied by adrastic overgrowth of pneumococci in lung distal airspaces anda strongly decreased alveolar neutrophil and alveolar exudatemacrophage recruitment that resulted in a markedly reducedresident pulmonary macrophage pool size. Ablation of CXCR2in the hematopoietic system was sufficient to trigger substantialmortality, with levels of just 10 to 25% hematopoietic CXCR2ablation causing strongly increased bacterial loads and earlymortality in the chimeric WT/CXCR2 KO mice. Together,these results for the first time demonstrate the critical andnonredundant role of CXCR2 in lung innate immunity againsta prototypic, Gram-positive bacterial pathogen and demon-strate that even partial deletion of CXCR2 is sufficient tointerrupt recruitment profiles of all major professional phago-cyte subsets, including neutrophils, exudate macrophages, andresident alveolar macrophages, with fatal consequences for theinfected host.

There are several as-yet-unanswered questions regarding therole of CXCR2 in lung leukocyte subset recruitment in S.pneumoniae-induced lung infection. First, is CXCR2 essentialfor lung exudate macrophage recruitment in response to S.pneumoniae or is this simply secondary to lung neutrophilrecruitment? Second, what degree of abundance of lung neu-trophil recruitment is required to contain and eradicate bac-terial burden in S. pneumoniae pneumonia, and is there athreshold level of reduced neutrophil recruitment that doesnot impair lung protective immunity against inhaled S. pneu-

FIG. 4. Effects of successive CXCR2 ablation or inhibition on survival, bacterial pathogen elimination, and professional phagocyte recruitmentprofiles in chimeric wild-type mice infected with S. pneumoniae. (A to F) Irradiated wild-type mice were reconstituted with various mixtures ofCXCR2 KO and wild-type bone marrow cells, as indicated. Six weeks later, mice were infected with S. pneumoniae (3 � 106 CFU/mouse), andsurvival (A) and bacterial loads in BAL fluids (B) and lung parenchymal tissue (C), as well as recruitment profiles of professional phagocyte subsets(alveolar macrophages [D], alveolar exudate macrophages [E], and alveolar neutrophils [F]) were determined over time, as indicated. (G to K)Mice were treated with vehicle (veh) or SB-225002 (inh) at 1.5 �g/g of body weight (left black bars) or 15 �g/g (right black bars) immediately priorto infection with S. pneumoniae (3 � 106 CFU/mouse). Twenty-four hours later, mice were euthanized and their BAL fluid neutrophil (G) andexudate macrophage (H) counts and bacterial loads in BAL fluid (I) and lung tissue (J) were determined. Survival curves in panel A arerepresentative of 12 mice per experimental group. The data in panels B to F are means SEM of seven to nine mice per time point and treatmentgroup. Values in panels G to J are means SEM of four mice per time point and treatment group. Asterisks indicate a significant differencecompared to transplanted (B to F) or the respective vehicle controls (G to J) (*, P � 0.05; **, P � 0.01; ***, P � 0.001). Plus signs indicate asignificant difference compared to the respective day zero value ( , P � 0.05; , P � 0.01; , P � 0.001).

VOL. 78, 2010 CXCR2 IN PNEUMOCOCCAL LUNG INFECTION 2627

on June 9, 2020 by guesthttp://iai.asm

.org/D

ownloaded from

moniae? Our results suggest that both alveolar neutrophil andalveolar exudate macrophage recruitment levels are attenu-ated in S. pneumoniae-infected CXCR2 KO mice, thus dem-onstrating the critical role of CXCR2 in regulating the recruit-ment of both neutrophils and exudate macrophages.

Recent reports have demonstrated a role for neutrophil- andendothelial/epithelial CXCR2 expression in contributing to al-veolar neutrophil recruitment in a model of E. coli-inducedacute lung inflammation (20), but monocyte recruitment wasnot a prominent feature. Notably, Gram-negative as opposedto Gram-positive bacteria and their bacterial toxins (E. coliversus S. pneumoniae) are known to trigger different molecularsignaling and recruitment pathways that mediate lung neutro-phil recruitment in response to bacterial infection (10, 11, 15),thus making a direct comparison between the previous studyand our report difficult. However, a weak neutrophilic re-sponse was still detectable in the mutant mice, thus indicatingpossible involvement of chemokine receptor systems otherthan CXCR2 in such residual neutrophilic responses.

The current study addressed whether partial ablation ofCXCR2 reduces the neutrophil burden of infected lungs with-out affecting lung protective immune responses against inhaledbacteria. We found that even a 10% ablation of CXCR2 ex-pression resulted in impairment of the host defense as mani-fested by increased bacterial loads and reduced survival inchimeric wild-type mice without significantly affecting neutro-phil or exudate macrophage counts in the lungs of these mice.Classically, G-protein-coupled chemokine receptors, includingCXCR2 or CCR2, on macrophages and neutrophils are wellknown to mediate chemotactic recruitment of leukocyte sub-sets toward the site of inflammation. Against this background,the phagocytic capacity of professional phagocytes such as neu-trophils per se is not expected to be altered by lack of thereceptor. However, binding of a chemokine to its receptor isknown to trigger a series of cellular events in leukocytes andother chemokine receptor-expressing cells like endothelial andepithelial barrier cells, such as intracellular calcium release,induction of a respiratory burst, and burst-dependent activa-tion of antibacterial proteolytic activities, thereby affecting thekilling potential of chemokine-activated leukocytes (20, 21).Also, changes in affinity profiles of cell surface adhesion mol-ecules on circulating leukocytes and sessile barrier cells, whichare needed for coordinated cellular transmigration, have beendescribed as dependent on chemokine receptor signaling (12).All of these steps are integral components of the overall hostinflammatory response to infection. In this view, even partialdeletion of selected chemokine receptors in hematopoietic

FIG. 5. Effects of partial or total CXCR2 deletion on the release ofneutrophils compared to monocyte chemoattractants in bronchoalveolarlavage fluids of wild-type mice, CXCR2 KO mice, and chimeric wild-typemice challenged with S. pneumoniae. Wild-type mice, CXCR2 KO mice,transplantation controls (wild-type bone marrow transplantation onto

wild type [E to H]), and chimeric wild-type mice were infected with S.pneumoniae, and BAL fluid neutrophil chemokines (MIP-2, KC) (A,B, E, and F) and monocyte chemoattractants CCL2 and CCL12 (C, D,G, and H) were analyzed by ELISA, as indicated. The data aremeans SEM of seven to nine mice per time point and treatmentgroup. Asterisks indicate a significant difference compared to the wildtype (A to D) or a significant difference for the indicated chimericwild-type mice relative to transplantation controls (E to H) (*, P �0.05; **, P � 0.01; ***, P � 0.001). Plus signs indicate a significantdifference compared to the respective day zero value ( , P � 0.05; , P � 0.01; , P � 0.001).

2628 HERBOLD ET AL. INFECT. IMMUN.

on June 9, 2020 by guesthttp://iai.asm

.org/D

ownloaded from

cells, as reported in the current study, will inevitably affectmany different facets of the lung host response to bacterialchallenge beyond the recruitment process itself.

Only a 25% ablation of hematopoietic CXCR2 expressionreduced neutrophil and exudate macrophage recruitment andresulted in a fatal outcome in mice, which was also confirmedin experiments employing selective CXCR2 antagonists. Thesedata, in whole, indicate that partial CXCR2 interference is notbeneficial in limiting lung injury due to uncontrolled neutro-philic responses in mice with severe Gram-positive bacterialinfections. At the same time, these observations may also beimportant and relevant in view of currently emerging clinicaltrials to evaluate CXCR2 antagonists for the treatment ofchronic pulmonary disorders (6). Our data are biologicallyimportant and relevant to these emerging clinical trials that areevaluating small-molecule inhibitors against CXCR2 for thetreatment of noninfectious chronic neutrophilic pulmonarydisorders, such as COPD, ARDS, and IPF. It is also possiblethat this approach could be applied to postviral neutrophilicairway disease and neutrophilic asthma. Based on our study ina mouse model, caution is needed with regard to interruptionof CXCR2 signaling in patients with either acute or chronicpulmonary infections like pneumonia and bronchiectasis, andwhether this treatment would be safe in patients with latentpulmonary infections, such as tuberculosis and cytomegalovi-rus (CMV) infection, will require careful monitoring in ourview.

Another unexpected finding of the current study was theobservation that progressive 50% and 75% CXCR2 ablation,as well as complete CXCR2 deletion, severely affected thealveolar exudate macrophage recruitment into the lungs of S.pneumoniae-infected mice. Of note, the main chemokine re-ceptor/ligand system mediating alveolar exudate macrophagerecruitment is binding of the major monocyte chemoattractantCCL2 (JE, MCP-1) to its CC chemokine receptor CCR2 (13,21, 30). Importantly, we observed here that CCR2 was abun-dantly expressed on Gr-1-positive inflammatory monocytesacting as the critical precursors of alveolar recruited exudatemacrophages (13), and we found greatly increased monocytechemotactic activities in BAL fluids of CXCR2 KO mice andof 50% and 75% chimeric wild-type mice on day 3 postinfec-tion. Although these inflammatory mediators usually drive in-flammatory monocyte recruitment, we observed in mutantmice that there is a lack of a coordinated exudate macrophagemobilization. These data illustrate the critical importance ofCXCR2 in regulating monocyte/macrophage recruitment path-ways in pneumococcal lung infection.

Several possibilities may explain the heavily increased neu-trophil and monocyte chemokine responses observed in chi-meric mice and also the mutant mice. First, increased chemo-kine responses may reflect the severe disease progression inCXCR2 KO mice and chimeric mice, leading to exaggeratedchemokine storms. Of note, baseline levels of monocyte andneutrophil chemoattractants in BAL fluids of mutant micewere similar to those in wild-type mice. Second, we previouslyobserved that CCR2 KO mice challenged with E. coli LPSresponded with greatly reduced exudate macrophage recruit-ment despite increased alveolar CCL2 protein levels, thus sup-porting the view that chemokine receptors on the cell surfaceof recruited leukocytes may internalize their ligands and thus

downregulate local chemokine levels (17). This interpretationmay explain increased neutrophil and monocyte chemokinelevels in BAL fluids of mutant and chimeric wild-type mice, inwhich both numbers of BAL fluid neutrophils and exudatemacrophages were significantly reduced. However, another is-sue is how CXCR2 deletion may cause reduced alveolar exu-date macrophage recruitment in response to pneumococcalchallenge. Another possible explanation was provided by arecent in vitro study which demonstrated that the amount ofmonocyte chemotaxis to suboptimal concentrations of mono-cyte chemokines CCL2 and CCL7 was synergistically increasedby the binding of CXC chemokines interleukin-8 (CXCL8) andstromal cell-derived factor 1 (SDF-1; CXCL12) to their respec-tive chemokine receptors, CXCR2 and CXCR4, on humanmonocytes and monocytic THP-1 cells (8).

In the current study, we also observed CXCR2 chemokinereceptor expression on the cell surface of alveolar exudatemacrophages but not circulating blood monocytes of wild-typemice. Such CXCR2 expression by recruited exudate macro-phages in S. pneumoniae-challenged wild-type but not mutantmice provides a molecular basis for inflammatory monocytechemotaxis toward CXCL1 to -3 chemokines. Along this line,Hartl and colleagues also demonstrated that recruited but notcirculating neutrophils from patients with chronic inflamma-tory lung diseases demonstrated increased chemokine receptorexpression profiles in vivo and chemokine responsiveness invitro (9). In addition, other reports have demonstrated a rolefor enhanced CXCR2-dependent monocyte migration in thepathogenesis of COPD (28). Thus, our data support the con-cept that CXCR2 and its respective ligands, together with CCchemokine receptors such as CCR2, synergistically regulatethe overall inflammatory recruitment and/or local activation ofa given leukocyte subset, such as exudate macrophages, in bothhumans and mice (8).

In summary, these data demonstrate a critical and nonre-dundant role of CXCR2 in lung protective immunity to infec-tion with the prototype Gram-positive bacterial pathogen S.pneumoniae. Just 10 to 25% ablation of hematopoietic CXCR2expression was sufficient to reduce survival in mice. Moreover,successive CXCR2 ablation (stepwise) decreased alveolar neu-trophil recruitment but also severely impaired the alveolarexudate macrophage recruitment in response to S. pneu-moniae. These data have identified a previously unrecognizedrole of CXCR2 in the regulation of inflammatory monocyte/macrophage recruitment in acute infections of the lungs in-duced by Gram-positive bacterial pathogens and may be ofsignificant clinical impact in view of emerging clinical trials toevaluate CXCR2 antagonists for the treatment of chronic neu-trophilic pulmonary disorders.

ACKNOWLEDGMENT

This study was supported by the German Research Foundation,grant SFB 587 to U.A.M. and T.W.

REFERENCES

1. Abraham, E. 2003. Neutrophils and acute lung injury. Crit. Care Med.31:S195–S199.

2. Albiger, B., A. Sandgren, H. Katsuragi, U. Meyer-Hoffert, K. Beiter, F.Wartha, M. Hornef, S. Normark, and B. H. Normark. 2005. Myeloid differ-entiation factor 88-dependent signalling controls bacterial growth duringcolonization and systemic pneumococcal disease in mice. Cell. Microbiol.7:1603–1615.

VOL. 78, 2010 CXCR2 IN PNEUMOCOCCAL LUNG INFECTION 2629

on June 9, 2020 by guesthttp://iai.asm

.org/D

ownloaded from

3. Bento, A. F., D. F. Leite, R. F. Claudino, D. B. Hara, P. C. Leal, and J. B.Calixto. 2008. The selective nonpeptide CXCR2 antagonist SB225002 ame-liorates acute experimental colitis in mice. J. Leukoc. Biol. 84:1213–1221.

4. Briles, D. E., S. K. Hollingshead, J. C. Paton, E. W. Ades, L. Novak, F. W.van Ginkel, and W. H. Benjamin, Jr. 2003. Immunizations with pneumococ-cal surface protein A and pneumolysin are protective against pneumonia ina murine model of pulmonary infection with Streptococcus pneumoniae.J. Infect. Dis. 188:339–348.

5. Chapman, R. W., M. Minnicozzi, C. S. Celly, J. E. Phillips, T. T. Kung, R. W.Hipkin, X. Fan, D. Rindgen, G. Deno, R. Bond, W. Gonsiorek, M. M. Billah,J. S. Fine, and J. A. Hey. 2007. A novel, orally active CXCR1/2 receptorantagonist, Sch527123, inhibits neutrophil recruitment, mucus production,and goblet cell hyperplasia in animal models of pulmonary inflammation.J. Pharmacol. Exp. Ther. 322:486–493.

6. Chapman, R. W., J. E. Phillips, R. W. Hipkin, A. K. Curran, D. Lundell, andJ. S. Fine. 2009. CXCR2 antagonists for the treatment of pulmonary disease.Pharmacol. Ther. 121:55–68.

7. Dockrell, D. H., H. M. Marriott, L. R. Prince, V. C. Ridger, P. G. Ince, P. G.Hellewell, and M. K. Whyte. 2003. Alveolar macrophage apoptosis contrib-utes to pneumococcal clearance in a resolving model of pulmonary infection.J. Immunol. 171:5380–5388.

8. Gouwy, M., S. Struyf, S. Noppen, E. Schutyser, J. Y. Springael, M. Parmen-tier, P. Proost, and J. Van Damme. 2008. Synergy between coproduced CCand CXC chemokines in monocyte chemotaxis through receptor-mediatedevents. Mol. Pharmacol. 74:485–495.

9. Hartl, D., S. Krauss-Etschmann, B. Koller, P. L. Hordijk, T. W. Kuijpers, F.Hoffmann, A. Hector, E. Eber, V. Marcos, I. Bittmann, O. Eickelberg, M.Griese, and D. Roos. 2008. Infiltrated neutrophils acquire novel chemokinereceptor expression and chemokine responsiveness in chronic inflammatorylung diseases. J. Immunol. 181:8053–8067.

10. Issekutz, A. C., H. E. Chuluyan, and N. Lopes. 1995. CD11/CD18-indepen-dent transendothelial migration of human polymorphonuclear leukocytesand monocytes: involvement of distinct and unique mechanisms. J. Leukoc.Biol. 57:553–561.

11. Issekutz, A. C., and T. B. Issekutz. 1993. Quantitation and kinetics of bloodmonocyte migration to acute inflammatory reactions, and IL-1 alpha, tumornecrosis factor-alpha, and IFN-gamma. J. Immunol. 151:2105–2115.

12. Maus, U., S. Henning, H. Wenschuh, K. Mayer, W. Seeger, and J. Lohmeyer.2002. Role of endothelial MCP-1 in monocyte adhesion to inflamed humanendothelium under physiological flow. Am. J. Physiol. Heart Circ. Physiol.283:H2584–H2591.

13. Maus, U., K. von Grote, W. A. Kuziel, M. Mack, E. J. Miller, J. Cihak, M.Stangassinger, R. Maus, D. Schlondorff, W. Seeger, and J. Lohmeyer. 2002.The role of CC chemokine receptor 2 in alveolar monocyte and neutrophilimmigration in intact mice. Am. J. Respir. Crit. Care Med. 166:268–273.

14. Maus, U. A., M. Backi, C. Winter, M. Srivastava, M. K. Schwarz, T. Ruckle,J. C. Paton, D. Briles, M. Mack, T. Welte, R. Maus, R. M. Bohle, W. Seeger,C. Rommel, E. Hirsch, J. Lohmeyer, and K. T. Preissner. 2007. Importanceof phosphoinositide 3-kinase gamma in the host defense against pneumo-coccal infection. Am. J. Respir. Crit. Care Med. 175:958–966.

15. Maus, U. A., M. Srivastava, J. C. Paton, M. Mack, M. B. Everhart, T. S.Blackwell, J. W. Christman, D. Schlondorff, W. Seeger, and J. Lohmeyer.2004. Pneumolysin-induced lung injury is independent of leukocyte traffick-ing into the alveolar space. J. Immunol. 173:1307–1312.

16. Maus, U. A., K. Waelsch, W. A. Kuziel, T. Delbeck, M. Mack, T. S. Blackwell,J. W. Christman, D. Schlondorff, W. Seeger, and J. Lohmeyer. 2003. Mono-cytes are potent facilitators of alveolar neutrophil emigration during lunginflammation: role of the CCL2-CCR2 axis. J. Immunol. 170:3273–3278.

17. Maus, U. A., S. Wellmann, C. Hampl, W. A. Kuziel, M. Srivastava, M. Mack,

M. B. Everhart, T. S. Blackwell, J. W. Christman, D. Schlondorff, R. M.Bohle, W. Seeger, and J. Lohmeyer. 2005. CCR2-positive monocytes re-cruited to inflamed lungs downregulate local CCL2 chemokine levels. Am. J.Physiol. Lung Cell. Mol. Physiol. 288:L350–L358.

18. Moore, T. A., M. W. Newstead, R. M. Strieter, B. Mehrad, B. L. Beaman, andT. J. Standiford. 2000. Bacterial clearance and survival are dependent onCXC chemokine receptor-2 ligands in a murine model of pulmonary Nocar-dia asteroides infection. J. Immunol. 164:908–915.

19. Reutershan, J. 2006. CXCR2: the receptor to hit? Drug News Perspect.19:615–623.

20. Reutershan, J., M. A. Morris, T. L. Burcin, D. F. Smith, D. Chang, M. S.Saprito, and K. Ley. 2006. Critical role of endothelial CXCR2 in LPS-induced neutrophil migration into the lung. J. Clin. Investig. 116:695–702.

21. Rollins, B. J., A. Walz, and M. Baggiolini. 1991. Recombinant human MCP-1/JE induces chemotaxis, calcium flux, and the respiratory burst in humanmonocytes. Blood 78:1112–1116.

22. Rosseau, S., P. Hammerl, U. Maus, H. D. Walmrath, H. Schutte, F. Grim-minger, W. Seeger, and J. Lohmeyer. 2000. Phenotypic characterization ofalveolar monocyte recruitment in acute respiratory distress syndrome. Am. J.Physiol. Lung Cell. Mol. Physiol. 279:L25–L35.

23. Russo, R. C., R. Guabiraba, C. C. Garcia, L. S. Barcelos, E. Roffe, A. L.Souza, F. A. Amaral, D. Cisalpino, G. D. Cassali, A. Doni, R. Bertini, andM. M. Teixeira. 2009. Role of the chemokine receptor CXCR2 in bleomycin-induced pulmonary inflammation and fibrosis. Am. J. Respir. Cell Mol. Biol.40:410–421.

24. Sibille, Y., and H. Y. Reynolds. 1990. Macrophages and polymorphonuclearneutrophils in lung defense and injury. Am. Rev. Respir. Dis. 141:471–501.

25. Srivastava, M., S. Jung, J. Wilhelm, L. Fink, F. Buhling, T. Welte, R. M.Bohle, W. Seeger, J. Lohmeyer, and U. A. Maus. 2005. The inflammatoryversus constitutive trafficking of mononuclear phagocytes into the alveolarspace of mice is associated with drastic changes in their gene expressionprofiles. J. Immunol. 175:1884–1893.

26. Sun, K., S. L. Salmon, S. A. Lotz, and D. W. Metzger. 2007. Interleukin-12promotes gamma interferon-dependent neutrophil recruitment in the lungand improves protection against respiratory Streptococcus pneumoniae in-fection. Infect. Immun. 75:1196–1202.

27. Taut, K., C. Winter, D. E. Briles, J. C. Paton, J. W. Christman, R. Maus, R.Baumann, T. Welte, and U. A. Maus. 2008. Macrophage turnover kinetics inthe lungs of mice infected with Streptococcus pneumoniae. Am. J. Respir.Cell Mol. Biol. 38:105–113.

28. Traves, S. L., S. J. Smith, P. J. Barnes, and L. E. Donnelly. 2004. SpecificCXC but not CC chemokines cause elevated monocyte migration in COPD:a role for CXCR2. J. Leukoc. Biol. 76:441–450.

29. Tsai, W. C., R. M. Strieter, B. Mehrad, M. W. Newstead, X. Zeng, and T. J.Standiford. 2000. CXC chemokine receptor CXCR2 is essential for protec-tive innate host response in murine Pseudomonas aeruginosa pneumonia.Infect. Immun. 68:4289–4296.

30. Winter, C., W. Herbold, R. Maus, F. Langer, D. E. Briles, J. C. Paton, T.Welte, and U. A. Maus. 2009. Important role for CC chemokine ligand2-dependent lung mononuclear phagocyte recruitment to inhibit sepsis inmice infected with Streptococcus pneumoniae. J. Immunol. 182:4931–4937.

31. Winter, C., K. Taut, M. Srivastava, F. Langer, M. Mack, D. E. Briles, J. C.Paton, R. Maus, T. Welte, M. D. Gunn, and U. A. Maus. 2007. Lung-specificoverexpression of CC chemokine ligand (CCL) 2 enhances the host defenseto Streptococcus pneumoniae infection in mice: role of the CCL2-CCR2axis. J. Immunol. 178:5828–5838.

32. Zarbock, A., M. Allegretti, and K. Ley. 2008. Therapeutic inhibition ofCXCR2 by Reparixin attenuates acute lung injury in mice. Br. J. Pharmacol.155:357–364.

Editor: J. N. Weiser

2630 HERBOLD ET AL. INFECT. IMMUN.

on June 9, 2020 by guesthttp://iai.asm

.org/D

ownloaded from