Bone marrow transplantation (BMT) is an important therapeutic option for a variety of malignant and nonmalignant disorders. Unfortunately, BMT recipients are at increased risk of infection, and in particular, pulmonary complications occur frequently. Although the risk of infection is greatest during the neutropenic period immediately following transplant, patients are still vulnerable to pulmonary infections even after neutrophil engraftment. We evaluated the risk of infection in this postengraftment period by using a well-established mouse BMT model. Seven days after syngeneic BMT, B6D2F1 mice are no longer neutropenic, and by 3 wk, they demonstrate complete reconstitution of the peripheral blood. However, these mice remain more susceptible throughout 8 wk to infection after intratracheal administration of Pseudomonas aeruginosa; increased mortality in the P. aeruginosa-infected BMT mice correlates with increased bacterial burden in the lungs as well as increased systemic dissemination. This heightened susceptibility to infection was not secondary to a defect in inflammatory cell recruitment to the lung. The inability to clear P. aeruginosa in the lung correlated with reduced phagocytosis of the bacteria by alveolar macrophages (AMs), but not neutrophils, decreased production of TNF-α by AMs, and decreased levels of TNF-α and IFN-γ in the bronchoalveolar lavage fluid following infection. Expression of the β2 integrins CD11a and CD11c was reduced on AMs from BMT mice compared with wild-type mice. Thus, despite restoration of peripheral blood count, phagocytic defects in the AMs of BMT mice persist and may contribute to the increased risk of infection seen in the postengraftment period.
Bone marrow transplantation (BMT) 3 is increasingly used in the treatment of various malignant and nonmalignant disorders. Unfortunately the application of BMT is limited by a number of serious side effects. Recipients of BMT are often plagued with life-threatening infections. Infectious complications occur in 60–80% of BMT recipients, usually occur within the first several weeks after BMT, and account for over half of transplant-related mortality (1, 2). The lung is a common target of infection after BMT, and bacterial pneumonias with Gram-negative rods and Gram-positive cocci predominate (1, 3, 4). The reported incidence of infectious pneumonia in BMT recipients varies from 11–30% (2, 5, 6). Associated mortality has been estimated at ∼22% in one study, with nosocomial pathogens leading to a significantly worse outcome (5).
Pseudomonas aeruginosa is an important cause of nosocomial pneumonia. It is especially virulent in the immunocompromised patient with structural or functional immune defects (4, 5, 7, 8). Immunodeficiency in the immediate post-transplant period is usually ascribed to the accompanying neutropenia. The degree and duration of neutropenia directly correlate with the risk of acquiring infections. Patients with severe neutropenia (<100 granulocytes/mm3) are at greatest risk (9, 10), and the mortality rate for patients developing pneumonia in this setting can be as high as 67% (4). With the advent of stem cell transplantation and the use of colony-stimulating factors, the duration of neutropenia post-BMT has been significantly shortened (11, 12, 13, 14, 15). Despite this, immunodeficiency can persist, and patients remain at increased risk for pulmonary infections. The mechanisms responsible for the increased susceptibility to infections beyond the time of complete engraftment are poorly understood.
Using an established syngeneic BMT model, we have evaluated the susceptibility of BMT recipient mice to the nosocomial pathogen P. aeruginosa in comparison to nontransplant controls at a time when engraftment is complete and cellular numbers are restored. Our data show that recipients of syngeneic BMT have an increased susceptibility to P. aeruginosa pneumonia at inoculum doses that are readily cleared by nontransplanted mice. We also demonstrate that this increased susceptibility is associated with defects in alveolar macrophage phagocytosis and alterations in the cytokine milieu.
Materials and Methods
Specific pathogen-free B6D2F1/J mice (age 8 wk) were obtained from The Jackson Laboratory (Bar Harbor, ME) and used for the majority of experiments. In some experiments B6Ly5.2 mice, purchased from the Frederick Cancer Research Facility (Frederick, MD), were used as BMT donors for irradiated B6Ly5.1 (The Jackson Laboratory) recipients so that donor vs host leukocytes could be distinguished by staining for the CD45.1 and CD45.2 alleles using Abs commercially available from BD PharMingen (San Diego, CA). Mice were housed under specific pathogen-free conditions and were monitored daily by the veterinary staff. These experiments were approved by the University of Michigan committee on the use and care of animals.
Bone marrow transplantation
The protocol for syngeneic BMT has been previously described (16, 17). Briefly, in preparation for transplantation, recipient mice received 13 Gy of total body irradiation (137Cs source) delivered in two fractions separated by 3 h. This dose of irradiation causes no histologically detectable pulmonary injury in mice receiving syngeneic BMT (17). Bone marrow was obtained from the femurs of donor mice. Cell mixtures of 5 × 106 bone marrow cells and 1 × 106 nylon wool-purified T cells were resuspended in Leibovitz’s L-15 medium (Life Technologies, Grand Island, NY) and transplanted into syngeneic recipients via tail vein infusion (0.25 ml total volume). Post-transplantation, mice were housed in sterilized microisolator cages; they were fed normal chow and autoclaved hyperchlorinated water for the first 2 wk post-BMT, and filtered water thereafter.
Intratracheal (i.t.) inoculation of P. aeruginosa
Ten microliters of P. aeruginosa PAO1 frozen stock was grown in 10 ml of tryptic soy broth (Difco, Detroit, MI) for 18 h at 37°C on a rocker at 225 rpm. Bacterial concentration was determined by measuring the amount of absorbance at 600 nm and was compared with a predetermined standard curve. Bacteria were then diluted to the desired concentration for inoculation. Animals were anesthetized by i.p. administration of 150 μl of a sterile solution of ketamine (1 ml of a 100 mg/ml solution), xylazine (0.33 ml of a 20 mg/ml solution), and saline (8.67 ml). The trachea was exposed in sterile fashion, and 30 μl of P. aeruginosa inoculum was administered i.t. using a sterile 26-gauge needle. The skin edges were then carefully reapposed with surgical glue (Nexaband; Abbott Laboratories, Chicago, IL). The solution of P. aeruginosa inoculum was then plated on blood agar plates in serial 5-fold dilutions and incubated at 37°C, and bacterial colonies were counted at 24 h to confirm the actual dose of bacteria administered.
Quantification of lung and blood bacterial burden
Following i.t. inoculation of P. aeruginosa, mice were euthanized at 24 h, and the thoracic cavity was exposed. Blood was collected by puncture of the right ventricle using syringes prelubricated with 50 μl of 1000 U/ml heparin. The pulmonary circulation was then perfused with saline, and lungs were excised, taking care not to include hilar tissues or proximal bronchi. Lungs were suspended in 1 ml of normal saline and homogenized. Ten microliters of each specimen (blood or lung) was then plated on blood agar plates using serial 5-fold dilutions, and plates were incubated at 37°C. Bacterial colonies were counted at 24 h as CFU per milliliter of blood or CFU per whole lung.
Pulmonary leukocyte recruitment
At designated time points following i.t. inoculation of P. aeruginosa or saline, mice were euthanized; the pulmonary circulation was perfused with PBS via the right ventricle; lungs were harvested, then minced to a slurry; and each specimen was suspended in 15 ml of a digest solution containing collagenase (15 mg), DNase I (0.25 ml or 250 KU units), penicillin, streptomycin, and RPMI 1640. The suspension was incubated on a rocker for 30 min at 37°C. The cells were dispersed in solution by repetitive suction (20 times) with a 10-ml syringe, then centrifuged at 500 × g for 10 min. After decanting the supernatant, each pellet was briefly suspended for 3 min in 3 ml of cold NH4Cl RBC lysing solution, then neutralized by adding 10 ml of RPMI, and again centrifuged at 500 × g for 10 min. Cell pellets were resuspended in 5 ml of RPMI containing 5% FCS. The samples were again dispersed in similar fashion as described above, then passed through a NITEX nylon screen (no. 103-100/32; Sefar America, Kansas City, MO), which was rinsed with an additional 5 ml of medium to obtain a single-cell suspension. Ten milliliters of 40% Percoll (Sigma-Aldrich, St. Louis, MO) was added to the cell suspension, and the solution was centrifuged at 2000 × g for 20 min. The supernatant was gently decanted, and the soft pellet of cells was resuspended in 10 ml of complete medium. Cells were enumerated by counting on a hemocytometer.
Cytospins of collagenase digestions were made by centrifuging 50,000 cells onto microscope slides using a Cytospin 3 (Shandon, Astmoore, U.K.). The slides were allowed to air-dry and were stained using a modified Wright-Giemsa (WG) stain. For WG staining, the slides were fixed/prestained for 2 min with a one-step, methanol-based, WG stain (Harleco; EM Diagnostics, Gibbstown, NJ). Slides were rinsed in water, then dipped eight times in Diff-Quick Solution I, followed by five dips in Diff-Quick Solution II (Diff-Quick; Baxter Scientific, Miami, FL). Slides were rinsed in water and allowed to air-dry. This modification of the Diff-Quick staining procedure improves the resolution of eosinophils from neutrophils in the mouse. A total of 300 cells were counted from randomly chosen high power microscopic fields for each sample. The differential percentage was multiplied by the total leukocyte number to derive the absolute number of monocyte/macrophages, neutrophils, and eosinophils per sample.
Alveolar macrophage and neutrophil in vivo phagocytosis assay
Mice were injected with 2 × 107 CFU of P. aeruginosa at time zero. At 2 or 4 h postinfection, bronchoalveolar lavage (BAL) was performed to obtain alveolar leukocytes. The trachea was exposed and intubated with 1.7-mm outer diameter polyethylene catheter attached to a three-way stop cock. BAL was performed by instilling 1 ml of normal saline, followed by gentle suction with an ∼0.8 ml volume return. This was repeated three times with a total return of ∼2.5 ml lavage fluid. BAL leukocytes were counted on a hemocytometer to obtain a total cell count. Assuming an average of 2.5 × 105 leukocytes in the alveolar space of a naive animal, this challenge represents a bacteria to phagocyte ratio of ∼80:1. The lavage fluid was centrifuged at 1500 rpm for 4 min, and the pellet was resuspended. Fifty thousand cells were then spun onto a glass slide, which was air-dried, stained with Diff-Quick, and examined under oil immersion. The initial 300 alveolar macrophages (AMs) or polymorphonuclear cells (PMNs) were counted to determine the number of whole intracellular bacteria in each cell (total number of bacteria). Then the number of cells that contained at least one bacterium was determined, and the percent phagocytosis was calculated: (number of cells containing bacteria/number of cells counted) × 100%. The phagocytic index (PI) was calculated as: PI = (total number of bacteria in all cells/number of cells that contained at least one bacteria) × percent phagocytosis.
Ex vivo fluorescent phagocytosis assay
AMs were lavaged from the lungs of uninfected wild-type and BMT mice as described above. AMs from 10 mice/group were pooled, and cell counts and cellular differential were obtained. Based on the total number of cells and the percentage of AMs by differential staining, 100,000 AMs were plated per well of a 96-well plate and allowed to adhere for 1 h in serum-free medium. To isolate PMNs, mice were given an i.p. injection of 3 ml of 3% thioglycolate (Sigma-Aldrich) dissolved in sterile saline. Five hours later, the peritoneal cavity was lavaged five times with 2–3 ml of 1× PBS, and cells were collected by centrifugation. The percentage of PMNs in the lavage was determined by differential staining analysis (the percentage ranged from 55–80% in separate experiments). For PMNs, 300,000 cells were plated per well of a 96-well plate, and cells were allowed to adhere for 1 h in serum-free medium. Following this incubation, medium was removed, 100 μl of fluorescently labeled Escherichia coli bioparticles (1 mg/ml corresponding to 3 × 107 bacteria/ml) was added in HBSS for 2 h, and a phagocytosis assay was performed according to the manufacturer’s instructions (Vybrant phagocytosis assay; Molecular Probes, Eugene, OR). This ratio of bacteria to phagocytes is 300:1 for AMs and 100:1 for PMNs. In this assay, fluorescent bacteria that are ingested by phagocytes register as fluorescent when read with a fluorometer, whereas extracellular bacterial fluorescence is quenched by the addition of trypan blue. Fluorescence intensity is linear with respect to a standard curve of labeled bacteria and correlates with phagocytic ability. Controls were treated with HBSS alone (no bacteria). In preliminary experiments it was determined that this control was equivalent to bacteria added to phagocytes pretreated for 45 min with 5 μg/ml cytochalasin D (Sigma-Aldrich) to inhibit phagocytosis.
AM and neutrophil in vitro phagocytosis assay
AMs were isolated from BAL and PMNs elicited from the peritoneal cavity as described above. For phagocytosis assays, 100,000 BAL or peritoneal lavage cells were plated onto eight-well Titer-Tek slides in serum-free medium and allowed to adhere for 1 h. Cells were then rinsed, and P. aeruginosa was added in serum-free medium at a bacteria to phagocyte ratio of 40:1. Cells and bacteria were incubated for 45 min before being washed three times. Plastic wells were removed from the slides, and the slides were stained with the Diff-Quick procedure. Slides were examined under oil emersion (×400 magnification), and the first 300 AMs or PMNs were examined. Cells were scored individually for the presence of whole intracellular bacteria, and the phagocytic index was determined as mentioned above. Control experiments used cells that had been pretreated for 45 min with a 5 μg/ml solution of cytochalasin D. Phagocytosis was inhibited by >95% in these control cultures, indicating that the bacteria identified in this assay are intracellular and not merely bound to the phagocyte cell surface.
ELISA measurements in supernatants of isolated AMs
BAL was performed on either wild-type or BMT mice on day 21 post-transplant. BAL fluid (BALF) from 10 individual mice was combined. BAL cells were collected by centrifugation and counted by hemocytometer to obtain total cell counts. Differential analysis was performed on an aliquot of the total cells to determine the percentage of AMs. The total number of AMs was determined by multiplying the percentage of AMs from the differential analysis by the total number of cells. AMs were resuspended at a concentration of 0.5 × 106 cells/ml in serum-free medium, and 200 μl of this suspension was plated in each well of a 96-well plate (100,000 cells/well). AMs were allowed to adhere to tissue culture plates for 1 h before wells were washed, and fresh medium was added (200 μl/well). AMs were then cultured for 24 h in the presence or the absence of 10 μg/ml LPS. After 24 h, cell-free supernatants were collected and analyzed by specific ELISA (Opti-EIA; BD PharMingen) for production of TNF-α.
ELISA measurements of BALF
BAL was performed on either wild-type or BMT mice at 24 h after P. aeruginosa inoculation. BALF was assayed for the presence of TNF-α, IFN-γ, IL-10, and monocyte chemotactic protein 1 (MCP-1) using commercially available ELISA kits (Opti-EIA; BD PharMingen) according to the manufacturer’s instructions.
Flow cytometric analysis of β2 integrins
AMs were harvested and pooled from BALF of 10 mice. In these experiments >90% of the recovered cells from the BAL were AMs based on differential analysis. Cells were first blocked with anti-murine FcγIII/II (CD16/CD32) and then were incubated with labeled primary rat anti-murine CD11a, rat anti-murine CD11b, or hamster anti-murine CD11c. The control stain for CD11a was rat IgG2a, κ; for CD11b the control was rat IgG2b, κ; and for CD11c it was hamster IgG1. All Abs were directly labeled with PE or FITC and were purchased from BD PharMingen. Following staining, cells were washed twice in FA buffer (Difco, Detroit, MI) and then were fixed with 0.5% paraformaldehyde. Flow cytometry was performed on a FACScan flow cytometer (BD Biosciences, Mountain View, CA). Data were analyzed using the CellQuest software package (BD Biosciences). Thresholds for positive staining were determined from the isotype-matched control samples.
Statistical significance was analyzed using the InStat 2.01 program (GraphPad, San Diego, CA) on a Power Macintosh G3. Student’s t tests or Mann-Whitney tests were run to determine p values when comparing two groups. When comparing more than two groups, ANOVA was performed with a post hoc Bonferroni test to determine significance. A value of p < 0.05 was considered significant.
Low dose P. aeruginosa infection is cleared in normal mice
To first characterize the course of P. aeruginosa pneumonia in normal hosts, we inoculated mice i.t. with varying doses of P. aeruginosa and analyzed survival up to 120 h. In additional experiments, bacterial burden in the lung and blood 24 h postinoculation was determined for each i.t. dose. In normal hosts, the injection of P. aeruginosa at doses ≥5 × 106 CFU was uniformly fatal within 48 h (Fig. 1,A). The bacterial burden in the lungs (Fig. 1,B) of mice treated within this dose range was also several-fold greater than the initial inoculum and showed systemic dissemination (Fig. 1,C). By contrast, i.t. doses of ≤1 × 106 CFU of P. aeruginosa are nonlethal (Fig. 1,A), are effectively cleared in the lung by the host (Fig. 1,B), and result in minimal systemic dissemination (Fig. 1 C). Thus, doses in the range of 1 × 106 CFU were used for subsequent experiments comparing the response to i.t. P. aeruginosa infection between syngeneic BMT recipients and nontransplant controls.
Hemopoietic reconstitution of the peripheral blood of syngeneic BMT mice is complete by day 21 post-transplant
Human BMT recipients are at increased risk of infection post-transplant despite peripheral blood cell reconstitution. The reasons for increased susceptibility at these later time points are not known. To study this phenomenon in animals, we chose a time point following syngeneic BMT that reflected complete hemopoietic reconstitution of the peripheral blood. By day 21 after syngeneic BMT, total blood leukocyte counts in BMT recipients were equivalent to those in nontransplant controls (7.13 ± 0.95 × 106 vs 8.95 ± 2.45 × 106; p = NS). Furthermore, differential analysis confirmed that there were no differences in the percentage or absolute counts of lymphocytes, macrophages, neutrophils, or eosinophils in the blood at this time (data not shown).
BMT mice have increased mortality to P. aeruginosa infection at day 21 post-BMT compared with wild-type mice
To determine whether mice receiving syngeneic BMT are more susceptible to P. aeruginosa infection despite peripheral blood engraftment, we inoculated wild-type or BMT mice on day 21 post-transplant with 2 × 106 CFU of P. aeruginosa i.t. and analyzed mortality over 96 h. The dose of 2 × 106 CFU caused 30% mortality at 48 h in the wild-type mice (Fig. 2,A, ▾), similar to that shown in Fig. 1,A. In contrast, this dose was lethal to all BMT mice by 48 h postinfection (Fig. 2 A, ▿). Thus, BMT mice at day 21 post-transplant are significantly (p < 0.05) more susceptible to pulmonary infection with P. aeruginosa.
P. aeruginosa infection in BMT mice results in increased bacterial burden in both lung and peripheral blood compared with wild-type mice
To determine whether BMT mice are more susceptible to a lower dose infection with P. aeruginosa, BMT and control mice were injected i.t. with 5 × 105 (5.7 in log10 scale) CFU P. aeruginosa. Twenty-four hours later, blood and lung tissue were harvested for quantification of pulmonary and systemic bacterial burdens. Fig. 2,B shows that the mean lung CFUs were significantly greater (p < 0.02) in BMT recipients compared with wild-type controls. Similarly, Fig. 2,C demonstrates that BMT mice display greater systemic dissemination (blood CFU, p < 0.0001) of P. aeruginosa following i.t. infection. These results demonstrate that at 24 h postinfection, mice receiving BMT 21 days earlier have a profound inability to clear pulmonary P. aeruginosa infection even when peripheral blood reconstitution is complete. Furthermore, following this small inoculum, the lung bacterial burden in the BMT mice (mean lung CFU log10 = 7.56) is close to the range previously shown to be fatal in normal mice (mean lung CFU log10 = 7.60–11.48; see Fig. 1).
We tested the susceptibility of BMT mice to P. aeruginosa at later time points as well. At 6 wk post-transplant, wild-type and BMT mice were given an i.t. inoculum of 5.89 log10 CFU. Twenty-four hours later, lungs of wild-type mice had a bacterial burden of 2.86 ± 0.24 CFU, whereas BMT mice had a higher lung bacterial burden (4.47 ± 0.37 CFU; n = 8; p = 0.003). Similarly, at 8 wk post-BMT, the mice were given an i.t. injection of 4.95 log10 CFU P. aeruginosa. Twenty-four hours later, the bacterial burden in the lungs of wild-type mice was 2.0 ± 1.43. In contrast, the bacterial burden in the lungs of BMT mice was much higher (4.71 ± 4.5 log10 CFU; n = 6; p = 0.0006). Thus, the enhanced susceptibility to infection seen in BMT mice persists at least through 8 wk post-transplant.
Pulmonary leukocyte recruitment in response to P. aeruginosa is similar in BMT and untransplanted mice
Effective host defense against Gram-negative infection involves recruitment of inflammatory cells to the site of infection. Abnormalities in leukocyte recruitment result in increased mortality from P. aeruginosa infection (18). We examined pulmonary leukocyte recruitment by enumerating total and differential pulmonary leukocyte counts 24 h after inoculation with saline or P. aeruginosa to determine whether the increased susceptibility after BMT could be explained on this basis (Fig. 3). Following saline injection, no differences were observed in total pulmonary leukocyte counts between BMT and non-BMT mice (1.25 ± 0.24 × 107 vs 1.50 ± 0.22 × 107 total cells, respectively; p = NS; Fig. 3,A). Twenty-four hours after inoculation with 5 × 105 CFU P. aeruginosa, total pulmonary leukocyte counts from both BMT recipients and wild-type animals were comparable and significantly increased from those obtained with saline injection (p < 0.05 for both groups). Similarly, the magnitude of change in pulmonary leukocyte numbers from baseline was not different between groups, indicating that robust cellular recruitment in response to infection occurred in each case (Fig. 3,A). In each experimental group, macrophages constituted the predominant cell type in the lungs at baseline (saline challenge), whereas a significant influx of neutrophils to the lungs was noted at 24 h following P. aeruginosa infection (Fig. 3 B). Collectively, these data indicate that the enhanced susceptibility to infection seen after BMT is not secondary to impaired leukocyte recruitment.
Alveolar macrophage phagocytic activity is impaired in BMT mice
The fact that BMT mice were more susceptible to P. aeruginosa infection despite equivalent leukocyte recruitment to the lung suggested that leukocytes in the BMT mice may have functional defects. We examined the ability of lung leukocytes from wild-type and BMT mice to phagocytose bacteria in vivo and ex vivo as described in Materials and Methods. At 4 h postinfection, the mean in vivo phagocytic index of AMs (Fig. 4 A, □) from BMT mice was 19.3 ± 8 vs 89.3 ± 20 in non-BMT mice (p < 0.05). Thus, AMs from BMT mice display greatly reduced phagocytic capacity in vivo for P. aeruginosa.
These in vivo results were further investigated using an ex vivo phagocytosis assay using AMs purified from uninfected wild-type and BMT mice as described in Materials and Methods. As shown in Fig. 4,B (□), the phagocytic activity of AMs from BMT mice was significantly less than that observed in untransplanted, wild-type controls (p = 0.03). However, AMs from BMT mice were not totally deficient in their phagocytic activity, as they did display activity above the negative control (cells without bacteria added; Fig. 4, ▩). Collectively, these data demonstrate that AMs in BMT mice have significantly reduced phagocytic activity for Gram-negative bacteria both in vivo and ex vivo.
Neutrophil phagocytic activity is not impaired in BMT mice
In contrast to the results for AMs, the in vivo phagocytic activity of neutrophils (Fig. 4,A, ▪) from BMT mice was somewhat increased compared with that in wild-type controls (29.6 ± 5.3 vs 15.5 ± 4.5; p < 0.05). To confirm that PMNs from BMT mice were not deficient in their phagocytic ability, we used thioglycolate-elicited PMNs from wild-type and BMT mice in a 45-min in vitro phagocytosis assay with a bacteria to phagocyte ratio of 40:1 as described in Materials and Methods. In this assay, wild-type PMNs had a PI of 54 ± 6.42, whereas BMT PMNs had a PI of 70.75 ± 9.9. Additionally, elicited PMNs were tested in the ex vivo assay to determine their ability to phagocytose fluorescently labeled E. coli at a 100:1 ratio. Fig. 4,B demonstrates that PMNs elicited from wild-type and BMT mice have similar phagocytic indexes for E. coli (Fig. 4 B, ▪; p = NS). These data confirm the in vivo results and suggest that PMN phagocytic function is not defective in BMT mice 3 wk post-transplant.
AMs are of donor origin on day 21 post-BMT
To confirm that the lung leukocytes isolated on day 21 post-BMT were of donor origin, we performed transplants using B6Ly5.2 mice as donors and B6Ly5.1 mice as recipients. Donor leukocytes could be distinguished by the expression of the CD45.1 allele present in B6Ly5.2 mice. Recipient cells were distinguished with an Ab to the CD45.2 allele present in B6Ly5.1 mice. Flow cytometric analysis of infected mice indicated that 99% of neutrophils (identified by the expression of GR-1) present in a collagenase digest of lung tissue from BMT mice were of donor origin by day 7 post-transplant (data not shown). Analysis of AMs purified by BAL from uninfected mice at day 21 post-BMT revealed that 82% of cells were donor-derived in one experiment, and 97% were donor-derived in a second experiment (data not shown). Greater than 85% of the cells in the BALF at this time point were macrophages by differential analysis. Furthermore, BAL cells had a forward scatter vs side scatter profile consistent with AM morphology. Thus, the AMs found in the lung on day 21 postsyngeneic BMT display phagocytic defects despite being of donor origin.
AMs from BMT mice are defective in the production of TNF-α
TNF-α is a critical cytokine necessary for effective clearance of P. aeruginosa by AMs (19, 20). We compared the ability of AMs from BMT and wild-type mice to produce TNF-α both basally and in response to endotoxin stimulation. AMs were adherence-purified from uninfected wild-type and BMT (day 21) BAL and were cultured for 24 h in the presence or the absence of 10 μg/ml LPS (Fig. 5). Supernatants from wild-type AMs produced 903 pg/ml of TNF-α basally compared with only 262 pg/ml in the same number of AMs harvested from BMT recipients (Fig. 5,A; p < 0.002). Following LPS stimulation, wild-type AMs produced 72.8 ± 7.6 ng/ml compared with only 5.4 ± 0.5 ng/ml of TNF-α in the LPS-stimulated BMT AMs (Fig. 5 B; p < 0.001). Thus, AMs from BMT mice have severe defects in the ability to produce TNF-α.
BMT mice produce reduced levels of TNF-α, IFN-γ, and IL-10 in response to P. aeruginosa infection
The results in Fig. 5 demonstrate that AMs from BMT mice are deficient in TNF-α production before infection. We wanted to determine whether there were deficiencies in the in vivo cytokine milieu within the alveolar space following infection. Therefore, we next measured BALF levels of TNF-α, IFN-γ, and IL-10 at 24 h after P. aeruginosa infection. As shown in Fig. 6, levels of TNF-α and IFN-γ were significantly decreased in BALF of BMT mice compared with wild-type mice (p < 0.03). By contrast, BMT mice had slightly elevated levels of MCP-1 (76.5 ± 8.4 vs 34.5 ± 14 pg/ml; p < 0.05), suggesting that cells from these animals were not universally deficient in inflammatory cytokine production. Furthermore, we found that IL-10 levels (Fig. 6) were decreased in BAL from infected BMT mice, indicating that the decrease in inflammatory cytokines (TNF-α and IFN-γ) was not attributable to the overzealous production of IL-10.
AMs from BMT mice display an immature pattern of β2 integrin staining
The defects seen in phagocytic function and TNF-α production by AMs from BMT mice suggested that the cells may be immature. Recently, we noted reduced expression of the β2 integrins CD11a and CD11c on immature AMs from GM-CSF−/− mice (21, 22). Therefore, we characterized the expression of the β2 integrins CD11a, CD11b, and CD11c by flow cytometry on AMs from wild-type and BMT mice. Table I demonstrates that AMs from BMT mice display reduced expression of CD11a and CD11c, but not CD11b, compared with wild-type cells.
|.||% Positive Cells .||.|
|.||Wild-type .||BMT .|
|.||% Positive Cells .||.|
|.||Wild-type .||BMT .|
This staining pattern is consistent with an immature phenotype. Results shown are from one experiment, representative of two.
We have established a murine model of syngeneic BMT that results in complete peripheral blood reconstitution by day 21 post-BMT. Despite hemopoietic reconstitution, BMT mice are more susceptible to pulmonary infection with the nosocomial Gram-negative pathogen, P. aeruginosa. The defective immunity to P. aeruginosa is associated with reduced surface expression of the β2 integrins CD11a and CD11c on AMs and a diminished capacity of these cells to secrete TNF-α and to phagocytose bacteria both in vivo and in vitro. In addition, TNF-α, IFN-γ, and IL-10 levels are all reduced in the BALF of P. aeruginosa-infected BMT mice. These results suggest the increased susceptibility of BMT mice to P. aeruginosa may be related to the immature phenotype of the AMs.
Life-threatening infections are common after BMT, and the increased susceptibility to infections in BMT recipients is often ascribed to the severity and duration of neutropenia (10). Neutropenia commonly results from myelosuppression induced by the conditioning regimen, delayed engraftment, or poor graft function. Failure of engraftment inevitably leads to mortality (23). However, infectious complications, specifically nosocomial pneumonias, significantly contribute to mortality after BMT even after hemopoietic engraftment has occurred. The mechanisms responsible for this delayed susceptibility are not known. Our findings are particularly relevant to this time interval after BMT; there is evidence of engraftment within the first week following BMT, and cell numbers are restored to normal by day 21. Despite reconstitution of the peripheral blood, syngeneic BMT recipients are highly susceptible to P. aeruginosa pneumonia for at least 8 wk post-transplantation. Transplanted animals have significantly increased mortality (Fig. 2) and increased lung and systemic bacterial burdens (Fig. 2) after i.t. inoculation of P. aeruginosa compared with untransplanted controls. Furthermore, the lung CFUs resulting from low dose infection in the BMT mice on day 21 are in the range that cause mortality in uncompromised hosts, suggesting that this low dose inoculum would be lethal by 48 h in the BMT mice. In contrast, when inoculated with this same low dose of P. aeruginosa, wild-type mice cleared the infection; lung bacterial burdens at 24 h were less than the initial inoculum. Additionally, systemic dissemination was minimal, and the control mice showed no clinical signs of disease.
The innate immune response in the lungs is critical for clearance of P. aeruginosa pneumonia. Recruitment of leukocytes in response to P. aeruginosa peaks at 12–24 h postinfection (18). The AM is vital in the initial host response to pulmonary infections (24, 25, 26), and together with neutrophils that are recruited to the lungs, AMs mediate the phagocytic response to Gram-negative bacteria. We explored whether abnormalities in recruitment of inflammatory cells to the lung mediated the increased susceptibility in BMT recipients and found no differences between groups with respect to the number of total cells or the number of specific leukocyte subpopulations in the lung at baseline or after infection. These data suggest that functional, rather than quantitative, defects are responsible for the enhanced susceptibility of BMT mice.
We next evaluated the in vivo and ex vivo functions of pulmonary leukocytes of wild-type and BMT mice. Our data demonstrate that AMs from BMT mice show defective phagocytosis of P. aeruginosa in vivo and in vitro. The phagocytic defect in the BMT AMs was related to the fact that fewer cells took up bacteria as well as the observation that there were fewer bacteria per cell, on the average, in BMT AMs. In addition, AMs from BMT mice display defective ex vivo phagocytosis of another Gram-negative bacteria, E. coli (Fig. 4). This defective AM phagocytic function correlated with a severe defect in the ability of the AMs to secrete TNF-α, both constitutively and in response to LPS stimulation (Fig. 5). TNF-α is a critical cytokine for the elimination of P. aeruginosa (19, 20). Previous data have demonstrated that the immunosuppression associated with cecal ligation and puncture (CLP) models of sepsis is associated with defective production of TNF-α by AMs. As a result, CLP animals are extremely susceptible to infection with P. aeruginosa. The transient transgenic expression of TNF-α within the lungs of CLP animals improves AM phagocytic function and protects from P. aeruginosa infection (19, 20). Further experimental evidence for the importance of AMs and TNF-α production in the clearance of P. aeruginosa come from studies examining chronic P. aeruginosa infection in resistant (BALB/c) vs susceptible (C57BL/6) strains of mice (19, 20). These experiments demonstrated that TNF-α production by AMs was significantly higher in resistant BALB/c mice on day 7 postinfection than in susceptible C57BL/6 mice. Furthermore, the inflammatory response in the susceptible C57BL/6 mice was largely comprised of neutrophils, whereas in the resistant BALB/c mice, macrophages predominated. Thus, these data suggest that AMs are crucial for host defense against P. aeruginosa, and that TNF-α production is critical for clearance of the organism. Our data support these results and demonstrate that persistent susceptibility to infection after BMT is also associated with defective AM production of TNF-α.
While there were clear defects in AM phagocytic function in BMT mice, our data do not show phagocytic defects in neutrophils of these animals. The in vivo assay was performed at 4 h postinfection, a point when neutrophils could readily be identified within the BAL. At this time point, numerous intracellular bacteria could be seen in neutrophils from BMT mice, but were less obvious in neutrophils from wild-type mice. These data could be interpreted in two ways. First, it is possible that the phagocytic defect seen in BMT mice is specific for AMs only and not for neutrophils. It is also possible, however, that the neutrophils in wild-type mice are so efficient in their killing that the intracellular bacteria cannot be identified at the 4 h point. Lavages performed at 2 h postinfection did not contain enough neutrophils to accurately determine the phagocytic index. Unfortunately, lung neutrophil phagocytosis could not be assessed ex vivo due to the paucity of these cells in the lungs of uninfected hosts. Therefore, to address this question we obtained thioglycolate-elicited PMNs from wild-type and BMT mice. When tested in a short term (45-min) in vitro phagocytosis assay, there was no significant difference in the phagocytic index measurements of PMNs from wild-type and BMT mice. In fact, the BMT PMNs showed slightly better phagocytosis than wild-type PMNs, confirming our in vivo results. Additionally, there were no differences in the ex vivo phagocytosis of E. coli at a 90 min point between PMNs from wild-type and BMT mice (Fig. 4 B). Thus, the defect in phagocytosis in BMT mice appears to be confined to AMs, and not PMNs.
Our data demonstrate that the cytokine milieu of BMT mice is altered both before and after bacterial infection. The data presented in Fig. 5 demonstrate that AMs from uninfected BMT mice have a significantly reduced capacity to make TNF-α under basal or LPS-stimulated conditions. The data in Fig. 6 demonstrate that levels of the activating cytokines, TNF-α and IFN-γ, are reduced within the alveolar space following P. aeruginosa infection as well. IFN-γ has been reported to be important in mediating microbicidal activity (27, 28, 29), and diminished production of this protein is associated with defective killing of intracellular bacteria by both AMs and PMNs (30, 31, 32). Our data suggest that impaired phagocytic function after BMT may be secondary to deficient production of IFN-γ as well as TNF-α, although the cellular source that is defective in this capacity (e.g., T cells, NK cells, or AMs) remains to be determined.
Thus, our work supports the hypothesis that a lack of stimulatory factors in the BMT lung contribute to the immunosuppression seen in these animals. Alternatively, increased susceptibility to infection could be secondary to the presence of suppressive factors in the in vivo milieu of BMT mice. Enhanced expression of Th2 cytokines, such as IL-10, has been shown to inhibit macrophage function in mice (33, 34) and to increase the susceptibility to fungal infections after allogeneic T cell-depleted BMT (35). In our experiments, however, BALF IL-10 levels are reduced after BMT (Fig. 6), ruling out this possibility. The cytokine/chemokine responses in BMT mice were not generally deficient, however, since levels of MCP-1 were slightly elevated in BALF from P. aeruginosa-infected BMT mice.
Table I demonstrates that expression of the β2 integrins CD11a and CD11c are reduced, whereas levels of CD11b are increased on BMT AMs compared with wild-type AMs. Interestingly, these same alterations in β2 integrin staining have previously been reported in AMs from GM-CSF−/− mice (22), and may be indicative of impaired functional activity associated with immaturity. We have demonstrated that the AMs in day 21 BMT mice are donor-derived. Given our data showing that BMT mice are more susceptible to P. aeruginosa even at 8 wk post-transplant, it seems unlikely that AMs are functionally and phenotypically immature because they represent a population of newly recruited cells to the lung that have not yet had time to mature. Rather, it may be that BMT AMs are defective due to the persistent lack of supportive factors from structural/resident cells within the BMT lung. It is possible that the conditioning regimen (irradiation) alters the ability of the pulmonary parenchymal cells, especially alveolar epithelial cells, to support AM differentiation in the lung (36).
In summary, we have found that mice undergoing syngeneic BMT remain extremely susceptible to infection with the nosocomial pathogen P. aeruginosa even after complete reconstitution of peripheral blood leukocyte counts. Furthermore, our data suggest that this increased susceptibility is probably due to impaired production of activating cytokines, such as TNF-α and IFN-γ, resulting in defective AM phagocytosis. Experiments are planned to determine whether augmentation of AM function using exogenous cytokines may represent an effective strategy to reduce the risk of infection after BMT.
We thank Susan Morris and Steve Wilcoxen for help with the flow cytometry analyses, and Dave Aronoff for help with the Vybrant assay for phagocytosis.
This work was supported by National Institutes of Health Grant CA79046 (to B.B.M.), HL071586 (to B.B.M.), and P50HL60289. K.R.C. is an Amy Strelzer-Manasevit Scholar of the National Marrow Donor Program and a fellow of the Robert Wood Johnson Medical Minority Faculty Development Program.
Abbreviations used in this paper: BMT, bone marrow transplantation; AM, alveolar macrophage; BAL, bronchoalveolar lavage; BALF, BAL fluid; CLP, cecal ligation and puncture; i.t., intratracheally; MCP-1, monocyte chemotactic protein 1; PI, phagocytic index; PMN, polymorphonuclear cell; WG, Wright-Giemsa.