|Home | About | Journals | Submit | Contact Us | Français|
Alveolar macrophages and neutrophils mediate innate immune defense against the opportunistic fungal pathogen Aspergillus fumigatus and are believed to be essential for host survival following inhalation of fungal spores (conidia). Although alveolar macrophages are postulated to kill inhaled conidia and neutrophils are believed to act against hyphae, the relative contribution of alveolar macrophages and neutrophils to early defense against A. fumigatus remain incompletely defined. To more precisely characterize the contributions of alveolar macrophages and neutrophils in antifungal host defense, we selectively depleted each cell population at different times following pulmonary challenge with conidia. Mice depleted of alveolar macrophages prior to pulmonary A. fumigatus infection recruited neutrophils normally and restricted hyphal tissue invasion. In contrast, neutrophil depletion prior to or within 3 h after infection was associated with high mortality. Neutrophil depletion at later time points, however, was associated with nearly normal survival rates. Our studies suggest that neutrophils, but not alveolar macrophages, provide essential anticonidial defense and that a brief period of influx into the respiratory tree is sufficient to prevent conidial germination and invasive disease.
Invasive aspergillosis is a major cause of infectious morbidity and mortality in immunocompromised patients [1-4]. Development of invasive aspergillosis signifies host failure to kill inhaled spores (conidia) and to prevent hyphal tissue invasion. Although loss of or deficiency in myeloid function is common among high-risk persons, the precise in vivo role of and temporal requirements for specific myeloid cells in host defense against Aspergillus fumigatus, the most common cause of invasive aspergillosis, remain controversial. The timing between conidial exposure and myeloid dysfunction or injury remains poorly understood but likely influences the outcome of infection.
Although neutropenia is an important risk factor for invasive aspergillosis [5, 6], animal and in vitro studies support a key role for alveolar macrophages in the early stages of host defense . This tenet is based in part on the finding that corticosteroids impair macrophage killing of conidia [8, 9] and enable hyphal escape from intracellular compartments [10, 11]. Corticosteroid-treated mice are more susceptible to resting than to swollen, germinating conidia , findings that were interpreted to signify that macrophages form a primary defense against inhaled conidia. Conversely, induction of neutropenia with nitrogen mustard renders mice more susceptible to germinating than to resting conidia, supporting the notion that neutrophils act primarily against products of germination . In addition, neutrophil contact with hyphae induces degranulation and the respiratory burst .
Animal models of invasive aspergillosis traditionally rely on the administration of cytotoxic agents and corticosteroids. Whereas this strategy is successful in recapitulating many aspects of human disease, extrapolating the role of individual cell types in host defense is difficult because these regimens exert pharmacologic effects on multiple cell types . For example, corticosteroids impair not only macrophage  but also neutrophil function by inducing defects in trafficking, adherence, and the release of antimicrobial effector molecules .
Recent approaches to examine the role of innate immune cells or effector systems against A. fumigatus rely on knockout mice [15-17] and antibodies that deplete leukocyte subsets [18, 19] or interfere with their trafficking [18, 20]. Mice deficient in the chemokine receptor CXCR2 or that are defective in CXCR2 signaling develop invasive aspergillosis [17, 18]. Delayed airway neutrophil influx in CXCR2-/- mice leads to conidial germination, a result that supports an early role for neutrophils in host defense . Conidial exposure induces alveolar macrophages to transcribe CXCR2 ligands (ie, macrophage inflammatory protein-2 [murine chemokine CXCL2; MIP-2] and keratinocyte-derived chemokine [murine chemokine CXCL1; KC]) in vivo and secrete these chemokines in vitro, consistent with a key role for alveolar macrophages in the orchestration of neutrophil recruitment [21-23].
In this study, we examined the temporal requirements for alveolar macrophages and neutrophils following murine pulmonary A. fumigatus infection with use of approaches that target these cells with specificity and temporal control. Our results suggest that alveolar macrophages are dispensable for survival. Alveolar macrophage-depleted mice were capable of limiting hyphal tissue invasion, likely because of preserved lung inflammatory responses and neutrophil airway recruitment. In contrast, mice developed invasive aspergillosis if neutrophil depletion occurred prior to or within ~3 h after infection. Neutrophil depletion at later time points was not associated with mortality, even when conidial inocula were standardized for different experimental conditions. These results support an early, nonredundant role for neutrophils, but not alveolar macrophages, in conidial defense.
Dichloromethylenebisphosphonate (clodronate) was a gift of Roche Diagnostics GmbH and was incorporated into liposomes as described elsewhere . Mice received 100 μL phosphate-buffered saline (PBS) or clodronate liposomes delivered intratracheally. The monoclonal antibody RB6-8C5 (rat IgG2b) was purified by protein G chromatography and quantified by Bradford assay with use of bovine serum albumin as a standard . For depletion studies, mice received 0.1 mg of RB6-8C5 in PBS via intraperitoneal injection.
C57BL/6 mice were purchased from The Jackson Laboratory. All in vivo studies were approved by the Memorial Sloan-Kettering Cancer Center Institutional Review Board. For survival studies, mice were maintained on Sulfatrim (Research Diets). A. fumigatus strain AF293 conidial suspensions were administered to anesthesized mice via the intratracheal route as described elsewhere .
Following euthanasia, 3 mL of bronchoalveolar lavage (BAL) fluid was collected in PBS and 5% fetal calf serum. The lungs were perfused with PBS and homogenized in 4 mL of PBS containing 0.025% Tween 20, for colony forming unit (CFU) determination. For ELISA studies, the BAL fluid and lung homogenates were clarified by centrifugation and analyzed for tumor necrosis factor (TNF), interferon (IFN)-γ (BD Biosciences), MIP-2, and KC (R&D Systems) according to the manufacturers’ instructions.
For flow cytometric analysis, finely minced lung tissue was digested at 37°C in PBS and 5% fetal calf serum containing 3 mg/mL collagenase IV (Worthington Biochemical Corporation) and 20 U/mL DNAse I (Roche) for 60 min, was ejected 5 times through a 20-gauge syringe, had red blood cells lysed, and was passed though a 100-μm filter.
BAL and lung cells were enumerated using a Z2 Coulter Counter (7-15 μm window; Beckman Coulter) and were stained with the following antibodies: anti-Ly6C-FITC (clone AL-21), anti-Ly6G-PE (clone A18), anti-CD11b-PERCP-Cy5.5, anti-CD11c-APC, anti-CD45-APC-Cy7 (BD Biosciences), anti-7/4-Alexa 647, anti-F4/80-FITC, and mouse anti-rat IgG2b-FITC (AbD Serotec). Alveolar macrophages were identified as SSChighCD45+CD11c+ BAL cells, and neutrophils were identified as CD45+CD11b+Ly6C+Ly6G+ cells. In specific experiments, 7/4 and Ly6G expression was used to define neutrophils. Flow cytometric data was collected on a BD LSR II flow cytometer and was analyzed with FlowJo, version 8.7.1 (Treestar).
Mice were euthanized either 3 days after infection or following the final manipulation (ie, RB6-8C5 administration). Perfused lungs were fixed in 10% formalin, embedded in paraffin, cut in 8-μm slices, stained with Gomori methenamine silver, and analyzed by microscopy.
Statistical analyses were performed using Stata, version 10.0 (StataCorp). For 2-group comparisons, the Mann-Whitney U test was used. For 3-group comparisons, the Kruskal-Wallis 1-way analysis of variance was calculated and, in cases of statistical significance, pairwise comparisons were analyzed using the Mann-Whitney test. For multiple comparisons, the P values were corrected by the Bonferroni method.
Survival data were plotted by the Kaplan-Meier method and analyzed using log-rank testing. Fisher’s exact method was used to analyze differences in 21-day survival among experimental groups. Statistical significance was defined by a P value of <.05.
To examine the role of alveolar macrophages in defense against A. fumigatus infection, C57BL/6 mice underwent alveolar macrophage depletion via intratracheal instillation of clodronate liposomes [24, 26]. Alveolar macrophage depletion was assessed by flow cytometric analysis (figure 1A) and enumeration of BAL cells (figure 1B). Consistent with prior studies [26, 27], intratracheal administration of clodronate liposomes on consecutive days led to >90% alveolar macrophage depletion 3 days later (figure 1B). Depletion of pulmonary-resident CD103+ dendritic cells , CD11b+ dendritic cells, or CD11b+Ly6C+Ly6G- monocytes was not observed (figure 2). Alveolar macrophage depletion was not observed in mice treated with PBS liposomes (figure 1B) or in mice treated intraperitoneally with clodronate liposomes (data not shown), a route of administration that targets splenic macrophage populations .
Alveolar macrophage depletion did not result in murine susceptibility to intratracheal challenge with 2.5 × 107 to 4 × 107 conidia. The 21-day survival rate approached 100% in 3 experiments pooled for analysis (figure 1C). Similar results were observed in mice treated intraperitoneally with clodronate or intratracheally with PBS liposomes. Histologic analysis of Gomori methenamine silver-stained lung sections showed that alveolar macrophage depletion did not interfere with host ability to limit conidial germination (figure 1D). Abundant conidia were visualized within the parenchyma, surrounded by neutrophil-rich infiltrates. Hyphae invading the lung parenchyma were not observed (figure 1D), though in rare instances, individual hyphae within large airways were noted (data not shown). Thus, C57BL/6 mice control conidial germination and survive A. fumigatus challenge in the absence of alveolar macrophages.
Because the alveolar macrophage depletion strategy per se may induce inflammation, BAL and lung neutrophil counts and inflammatory cytokine and chemokine levels were examined in mice treated with clodronate, PBS, or no liposomes. Neutrophil counts ( ± standard deviation) in BAL fluid samples were marginally higher in mice treated with clodronate liposomes (26.6 × 103 ± 34.0 × 103 neutrophils/mL), compared with mice treated with PBS liposomes (4.6 × 103 ± 3.8 × 103 neutrophils/mL) or no liposomes (1.3 × 103 ± 0.5 × 103 neutrophils/mL) (figure 3A), although statistical significance was reached only when comparing the clodronate and no liposome groups (P = .01, by Mann-Whitney test). The neutrophil airway influx associated with clodronate liposome treatment only constituted ~1% of the influx associated with A. fumigatus infection (figure 3E).
TNF, MIP-2, KC, and IFN-γ were chosen for analysis because of their role in A. fumigatus host defense [18, 29-31]. Intratracheal administration of clodronate or PBS liposomes did not trigger TNF, MIP-2, KC, or IFN-γ release in the lungs of naive mice (figures 3B-3D 4).4). In all groups, TNF, MIP-2, and KC levels in the airway were below the limit of detection (15 pg/mL; data not shown). Thus, the alveolar macrophage depletion strategy is highly unlikely to exert proinflammatory effects that influence the experimental outcome.
Because alveolar macrophages secrete inflammatory mediators in response to A. fumigatus challenge [21-23, 32, 33], the effect of alveolar macrophage depletion on airway neutrophil recruitment and lung and BAL cytokine and chemokine levels was examined 24 h after infection, a time point that is associated with peak TNF levels . Alveolar macrophage depletion did not reduce airway neutrophil recruitment (figure 3E). In addition, alveolar macrophage depletion was associated with preserved TNF, MIP-2, and KC release in the lungs and airways of infected animals (figure 3F-3H). Although lung MIP-2 levels were higher in infected mice treated with clodronate liposomes, this finding failed to achieve statistical significance after correction for multiple comparisons (P = .054). These results suggest that alveolar macrophage depletion neither reduces lung or airway TNF, MIP-2, and KC production nor impairs neutrophil airway influx 24 h after infection.
In summary, these data indicate that alveolar macrophages are dispensible for control of A. fumigatus infection in otherwise immunocompetent mice. Alveolar macrophage depletion is overcome because of preserved inflammatory responses and neutrophil airway influx, resulting in host ability to prevent the formation of tissue-invasive hyphae.
Because alveolar macrophages appear redundant for defense against A. fumigatus infection, the timing of neutrophil depletion on the development of invasive aspergillosis and survival was examined. Although invasive aspergillosis develops in mice that are neutropenic at the time of infection [7, 18, 19], the temporal requirement for neutrophils in host defense remains unknown. We observed that C57BL/6 mice reared in pathogen-free conditions and infected with 1 × 107 to 4 × 107 conidia had reduced CFUs in lung tissue by 90%-95% by 48 h after infection, consistent with results published elsewhere [7, 11]. However, we routinely cultured 104-105 CFUs from lung samples at 30 days after infection (data not shown), suggesting that infected mice may remain vulnerable to neutrophil depletion for prolonged periods.
To determine the effect of neutrophil depletion at specific time points in relation to infection, mice were treated with a single 0.1 mg intraperitoneal dose of monoclonal antibody RB6-8C5 (anti-Gr-1; anti-Ly6G/Ly6C). This method of transient neutrophil depletion was chosen because it is well-characterized and specific compared with other methods (eg, administration of cyclophosphamide  or irradiation). RB6-8C5-induced neutropenia enables the establishment of invasive aspergillosis in the lungs of infected mice , and tissue damage is exacerbated by neutrophil recovery . The dose used for experiments led to >99% neutrophil depletion in the circulation (data not shown) and in single cell lung suspensions but to no significant depletion in CD11b+Ly6C+Ly6G- monocytes (Gr-1+ or inflammatory monocytes; figure 5). C57BL/6 mice administered RB6-8C5 24 h prior to infection were highly susceptible to intratracheal A. fumigatus challenge, with an observed mortality rate of 72% (figure 6A).
RB6-8C5-mediated neutrophil depletion initiated 1 or 3 h after infection resulted in a similar mortality to that observed among mice treated with RB6-8C5 24 h before infection (figure 6A). In contrast, no deaths were observed among mice that received RB6-8C5 at 6 or 24 h after infection (figure 6A) or at later time points (data not shown). Thus, infected mice remain sensitive to neutrophil depletion for 3 h after infection.
Examination of Gomori methenamine silver-stained lung sections revealed that murine susceptibility to RB6-8C5 administration correlated with the presence of tissue-invasive hyphae. Mice treated with RB6-8C5 either 24 h before (figure 6B) or 1 h after (figure 6C) conidial challenge exhibited hyphal lesions in airways and terminal air spaces with similar patterns of hyphal tissue invasion. In contrast, mice treated with RB6-8C5 at 24 h after infection demonstrated numerous conidia within the lungs without evidence of hyphal tissue invasion (figure 6D).
A trivial explanation for differences in survival among experimental groups is a difference in fungal burden at the onset of neutrophil depletion. To exclude this possibility, the inoculum was adjusted to achieve the same lung fungal burden at the time of RB6-8C5 administration (figure 7A and 7B). Despite a trend toward higher CFU counts in lung samples of mice that received RB6-8C5 at 24 h after infection (P = .08, by Mann-Whitney rank-sum test), this group had a greater 21-day survival rate than did mice that received RB6-8C5 at 1 h after infection (P = .001, by Fisher’s exact test). This result indicates that differences in survival are not attributable to differences in lung fungal burden at the time of neutrophil depletion.
To examine the effects of combining neutrophil and macrophage depletion, mice were treated with RB6-8C5 as well as clodronate liposomes via the intratracheal or intraperitoneal route. Mice treated intratracheally with clodronate liposomes and RB6-8C5 demonstrated a similar susceptibility to A. fumigatus infection as did mice treated with RB6-8C5 alone (figure 8A), consistent with a predominant role for neutrophils over alveolar macrophages in host defense in the experimental model examined.
As a control, 2 groups of mice received intraperitoneal clodronate liposomes (with or without intratracheal clodronate liposomes) and were subsequently treated with RB6-8C5. Intraperitoneal administration of clodronate liposomes rendered both groups resistant to conidial challenge (figure 8A). Single cell lung suspensions from these mice were analyzed to examine the cell populations affected by intratracheal or intraperitoneal administration of clodronate liposomes and RB6-8C5. In this experiment, 7/4 and Ly6G were used to define neutrophils to avoid the use of antibodies with the same isotype (IgG2b) as RB6-8C5. As expected, RB6-8C5 depleted circulating neutrophils (data not shown) and neutrophils in perfused lungs (figure 8B, top row vs. second row) but not CD11b+ F4/80+ macrophages. Intratracheal administration of clodronate liposomes followed by RB6-8C5 led to depletion of lung macrophage and neutrophil populations (figure 8B, third row). Intraperitoneal administration of clodronate liposomes followed by RB6-8C5 did not deplete lung macrophages (figure 8B, bottom row) and reduced the cell frequency in the neutrophil gate. In addition, a 7/4+Ly6Gint cell population was observed (figure 8B, red arrow).
Neutrophils that bind RB6-8C5 are rapidly removed from the circulation and other organs, including the lungs . After administration of 125I-labelled RB6-8C5, radiolabeled neutrophils are sequestered in the spleen and liver , suggesting that the reticuloendothelial system mediates RB6-8C5-dependent neutrophil clearance. To determine whether systemic administration of clodronate liposomes influences neutrophil clearance, mice were treated intraperitoneally with clodronate liposomes followed by RB6-8C5, and single cell lung suspensions were stained with FITC-labeled anti-rat IgG2b to detect bound RB6-8C5.
Administration of PBS liposomes did not interfere with RB6-8C5-dependent neutrophil clearance in the lung (figure 8C, black gate), and 7/4+Ly6Gint cells were not observed (figure 8C, red gate). In contrast, mice treated intraperitoneally with clodronate liposomes accumulated 7/4+Ly6Gint cells with bound anti-rat IgG2b antibody (figure 8C). This result indicates that 7/4+Ly6Gint cells represent neutrophils with bound RB6-8C5. Thus, mice treated intraperitoneally with clodronate liposomes followed by RB6-8C5 are resistant to intratracheal A. fumigatus challenge because of defective neutrophil depletion rather than direct effects of intraperitoneal clodronate liposome administration.
Mice that undergo neutrophil depletion within 3 h after infection are as susceptible to invasive aspergillosis as neutropenic mice at the time of infection (figure 6). To determine whether the presence of alveolar macrophages can alter murine susceptibility to A. fumigatus in the context of neutrophil depletion, murine mortality was compared in alveolar macrophage-depleted or alveolar macrophage-sufficient mice that underwent neutrophil depletion 24 h after infection (figure 9A). There was no difference in survival rate among experimental groups (P = .75, by log-rank test), consistent with the notion that loss of alveolar macrophages does not render mice more susceptible to neutrophil depletion after infection.
The objectives of this study were to examine the essential role and temporal requirements for alveolar macrophages and neutrophils in host defense against A. fumigatus. An important finding is the redundancy of alveolar macrophages for murine survival, neutrophil recruitment, and inflammatory responses in a pulmonary infection model. This conclusion is based on a compartmentalized macrophage depletion strategy rather than systemic impairment with a broadly immunosuppressive agent. A second finding relates to the brief period of host susceptibility to neutrophil depletion after infection. Neutrophil influx to the lungs in the first 6 h after infection is sufficient to protect mice from invasive disease.
The finding that alveolar macrophage-depleted mice sustain neutrophil recruitment and the synthesis of inflammatory mediators may appear surprising. However, this finding has been reported in other pulmonary infection models, such as those for infection due to Klebsiella pneumoniae and Pseudomonas aeruginosa [36, 37]. Preserved neutrophil recruitment may reflect a role for epithelial cells and endothelial cells in phagocytic and inflammatory responses against A. fumigatus [38-40]. In addition, natural killer cells and CCR6+ myeloid dendritic cells participate in inflammatory responses in neutrophil-depleted mice [20, 41]. The saturable uptake of conidia by alveolar macrophages in vitro  and the dispersal of conidia observed in lung sections suggest that it is highly unlikely that the small number of alveolar macrophages that escape in vivo depletion can compensate for the profound loss of these cells in the generation of inflammatory responses. More likely, alveolar macrophage depletion may result in a transient reduction in the synthesis of inflammatory mediators in the period immediately after infection (<24 h). However, any alveolar macrophage-dependent reduction is readily compensated within a time frame that permits intact neutrophil airway recruitment and control of conidial germination.
Although these studies argue against an essential, nonredundant role for alveolar macrophages against A. fumigatus, this conclusion does not preclude a role for these cells in optimal responses against A. fumigatus. Because these studies did not employ broadly immunosuppressive or myelotoxic strategies to render mice susceptible to low conidial inocula (at the cost of damaging multiple leukocyte subsets), we do not exclude the possibility that alveolar macrophages have the capacity to clear small inocula without participation of other cell subsets. Consistent with this possibility, there is evidence that mice administered cytotoxic chemotherapy may become susceptible to targeted macrophage depletion . Thus, in patients with significant myeloid injury, the slow turnover and relative resistance of alveolar macrophages to irradiation  may enable these cells to maintain a cellular defense against inhaled conidia after the production, trafficking, or function of short-lived leukocytes (eg, neutrophils) has been impaired.
The requirement for neutrophils in the period immediately after infection period is consistent with an essential role for these cells in preventing conidial germination at the earliest stage after infection . Our results suggest that there is a critical time window for neutrophil recruitment for protection to occur. Induction of neutropenia beyond 6 h after infection did not result in murine mortality, even when inocula were standardized for the time point of depletion. Neutrophils inactivate conidia and prevent their germination by phagocytic and nonphagocy tic mechanisms [17, 44, 45]. Although extravasated neutrophils undergo apoptosis rapidly, the rapid establishment of neutrophil-dependent extracellular defense mechanisms (eg, lactoferrin release, the formation of BAL aggregates, and DNA extracellular traps) may serve to control conidial germination and hyphal tissue invasion, even with subsequent interruption in neutrophil influx.
The results presented in this study lend a cautionary note to cell depletion strategies that target macrophages and neutrophils. We found that clodronate liposomes administered systemically eliminate phagocytic cells that promote the clearance of antibody-coated neutrophils. Neutrophil and macrophage depletion strategies are often combined in animal models of inflammatory and infectious disease [46, 47]. Therefore, it is imperative to verify that the depletion strategy of a particular cell type does not unintentionally interfere with depletion of a second cell subset.
Our findings relate to infection control practices in high-risk patient populations. A major clinical goal is to prevent opportunistic infections in these populations, in part through reduction in exposure and administration of antimicrobial prophylaxis. Acute infection in high-risk hosts is a likely route to the development of invasive aspergillosis. However, a second possibility is that exposure prior to the onset of immune suppression or myeloblation may enable conidia to persist long enough in a viable form to establish invasive disease in the context of incipient or progressive myeloid injury . The results presented in this study are most consistent with the first possibility, because of the rapid resistance of mice to transient neutrophil depletion after infection, despite a significant viable burden of spores in the lung. Although a prior history of invasive aspergillosis has a deleterious effect on the outcome of allogeneic stem cell transplantation [49, 50], the window of susceptibility following conidial exposure to administration of corticosteroids or cytotoxic agents remains unknown. Understanding both the specific requirements for and the critical timing of innate immune cell action in relation to exposure is highly relevant to the formulation of strategies to protect vulnerable hosts during periods of highest risk.
The A. fumigatus strain AF293 and the RB6-8C5 hybridoma were generous gifts from Michael Anderson (University of Manchester; Manchester, England) and Robert Coffmann (DNAX Research Institute; Palo Alto, CA), respectively. We thank Ingrid Leiner for help with antibody preparation and technical support and Ying Taur for help with statistical analysis.
Financial support: National Institutes of Health (grants K08 AI071998 to T.M.H., T32 AI055409 to M.M., and R01 AI067359 to E.G.P.).
Potential conflicts of interest: none reported.
Presented in part: 46th Annual Meeting of the Infectious Disease Society of America, Washington, DC, 25-28 October 2008 (presentation B-3708).