We investigated putative mechanisms by which human surfactant protein A (SP-A) effects killing ofKlebsiella pneumoniae by human alveolar macrophages (AMs) isolated from bronchoalveolar lavagates of patients with transplanted lungs. Coincubation of AMs with human SP-A (25 μg/ml) andKlebsiella resulted in a 68% decrease in total colony forming units by 120 min compared with AMs infected withKlebsiella in the absence of SP-A, and this SP-A-mediated effect was abolished by preincubation withN G-monomethyl-l-arginine. Incubation of transplant AMs with SP-A increased intracellular Ca2+concentration ([Ca2+]i) by 70% and nitrite and nitrate (NOx) production by 45% (from 0.24 ± 0.02 to 1.3 ± 0.21 nmol · 106AMs−1 · h−1). Preincubation with 1,2-bis(2-aminophenoxy)ethane-N,N,N′,N′-tetraacetic acid-acetoxymethyl ester inhibited the increase in [Ca2+]i and abrogated the SP-A-mediatedKlebsiella phagocytosis and killing. In contrast, incubation of AMs from normal volunteers with SP-A decreased both [Ca2+]i and NOx production and did not result in killing of Klebsiella. Significant killing of Klebsiella was also seen in a cell-free system by sustained production of peroxynitrite (>1 μM/min) at pH 5 but not at pH 7.4. These findings indicate that SP-A mediates pathogen killing by AMs from transplant lungs by stimulating phagocytosis and production of reactive oxygen-nitrogen intermediates.
- innate immunity
- lung transplant
klebsiella pneumoniae is an important cause of gram-negative nosocomial bacteremia (56) because it accounts for 10–13% of ventilator-associated pneumonias (31) and is one of the top three complicating pathogens isolated from patients with acute respiratory distress syndrome (2). Innate pulmonary immunity against bacterial pathogens involves resident phagocytes and a number of inducible proteins and peptides produced by these leucocytes and the lung epithelial cells (10). As part of this innate immune system, surfactant protein A (SP-A) is synthesized by a variety of airway cells, is a member of the family of collagenous Ca2+-dependent lectins or collectins, and is important for the clearance of many important lung pathogens (26, 34, 35, 55,61).
SP-A has been shown to bind to the capsular polysaccharide ofKlebsiella and to enhance phagocytosis and killing of this pathogen by rat and guinea pig alveolar macrophages (AMs) (28). SP-A has also been reported to modulate the production of nitric oxide (NO) by a variety of cell types (5,29), and recently SP-A stimulation of NO production and inducible nitric oxide synthase (iNOS) upregulation were related to cell state and mechanism of activation (52).
The importance of NO for the clearance of bacteria from the lung has been well documented (4, 25), and the requirement of NO specifically for Klebsiella clearance was previously demonstrated utilizing a mouse model (56). Although NO itself does not possess microbicidal activity, it may react with iron or thiol groups on proteins forming iron-nitrosyl complexes that inactivate enzymes important for bacterial DNA or mitochondrial replication (15). Additionally, at high inspired NO concentrations, autoxidation of NO may form toxic reactive species such as NO and N2O3 or it may react with O · to form peroxynitrite (ONOO−), which is highly bactericidal (9,25). To this extent, SP-A may enhance killing by stimulating the production of superoxide and NO and/or enhancing phagocytosis, thus exposing pathogens to intense localized production of reactive oxygen-nitrogen intermediates in phagosomes.
Stimulation of NO production by rat and mouse macrophages has become a standard part of cell activation protocols, while normal human monocytes remain refractile to a variety of stimuli (27). In contrast, AMs isolated from the lungs of patients with tuberculosis produce NO (53). Thus the mechanisms responsible for the production of NO by human AMs are complex and poorly understood. The interaction of SP-A with human AMs and its effects on NO production and pathogen killing may depend on preexposure of AMs to activating stimuli in the alveolar milieu. To address these issues, we isolated human AMs from bronchoalveolar lavagates from transplanted lungs and lungs of normal human volunteers and measured the ability of human SP-A to effect production of NO and killing of K. pneumoniae. Our results are the first to delineate the differential ability of SP-A to modulate innate immune responses of primary human AMs and to show that SP-A may modulate killing by enhancing phagocytosis in addition to NO production.
PBS, DMEM with l-arginine and 4.5 g/l of glucose, and Hanks' balanced salt solution (HBSS containing Ca2+ and Mg2+) were from Cellgro (Atlanta, GA). Diff Quik stain kits were obtained from Baxter Healthcare (McGraw Park, IL). 3-Morpholinosydnonimine (SIN-1), xanthine oxidase (XO), 1-propanamine-3-(2-hydroxy-2-nitroso-1-propylhydrazine) (PAPANONOate), and 1,2-bis(2-aminophenoxy)ethane-N,N,N′, N′-tetraacetic acid (BAPTA)-acetoxymethyl ester were from Calbiochem. Dihydrorhodamine (DHR)-123 and fura 2-acetoxymethyl ester came from Molecular Probes (Eugene, OR). Unless otherwise specified all other chemicals were from Sigma (St. Louis, MO).
Isolation of AMs.
Bronchoalveolar lavage (BAL) fluid was obtained either from lung transplant patients undergoing routine surveillance lavage [Univ. of Alabama at Birmingham (UAB)] or normal volunteers [Environmental Protection Agency (EPA)]. The BAL was centrifuged at 800g for 10 min to pellet cells. Cells were resuspended in DMEM supplemented with 1% l-glutamine, 2.5% HEPES, 0.2% low endotoxin BSA, and penicillin/streptomycin antibiotic. Cells were counted with a hemocytometer, and sample viability tested by trypan blue exclusion was >90%. Samples that could not be maintained as “sterile” in the presence of antibiotics and utilizing standard primary tissue culture techniques were discarded. Furthermore, in order for samples to be utilized, they were >90% AMs as determined by examination of cytospin preparations stained with Diff Quik. Examination of cell content of BAL from transplant patients has demonstrated increased neutrophils during infection and increased lymphocytes during rejection (16). For this reason, samples containing >10% lymphocytes or neutrophils were not utilized. Identification of infection or rejection by the UAB hospital as reported in Table 1 was determined according to different criterion than BAL cellular content, i.e., rejection was identified according to standard histological examination of transbronchial biopsy and infection by culture of BAL fluid and examination of fluid for fungal hyphae.
All BAL samples were maintained as anonymous. General demographic information was requested at the end of all experiments and only in regard to particular samples of interest. The study was approved by the Institutional Review Board at UAB and by the Human Studies Facility at the United States EPA.
Purification of SP-A.
SP-A is purified sterilely from the BAL of patients with alveolar proteinosis by N-butanol extraction as previously described (21). PAGE and Western blot analysis of SP-A were carried out to ensure the purity of SP-A preparations (23). SP-A was stored at −20°C in 5 mM HEPES, pH 7.4. Aliquots were cultured for aerobic bacteria in BBL brain heart infusion broth (Becton Dickinson, Franklin Lakes, NJ), and only culture-negative aliquots were used in experiments.
Each lot of SP-A was tested for endotoxin, and only batches of SP-A with <0.5 U/ml of endotoxin were used in experiments. PBS, DMEM, HBSS+, and saline were tested and certified to contain <0.5 U/ml of endotoxin. Low endotoxin BSA was used in all tissue culture experiments. Endotoxin testing was performed by the Media Preparation Shared Facility at UAB.
Measurement of NO and NO .
Levels of NO and NO were measured in the supernatant using either the Greiss reaction or by fluorescence utilizing 2,3-diaminonaphthalene (DAN) (38). NO was first converted to NO withEscherichia coli reductase. For the Greiss reaction, 100 μl of sample was incubated in duplicate with equal volumes of 1% sulfanilamide and 0.1% N-1-naphthylethylene-diamine dihydrochloride for 10 min, and the absorbance was read at 550 nm. For fluorescence measurements, 100 μl of sample was incubated in duplicate with 25 μl of freshly prepared DAN (0.05 mg/ml in 0.62 M HCl) and incubated for 10 min. The reaction was stopped by the addition of 25 μl of 2.8 N NaOH, and the signal was measured using a fluorescent plate reader with excitation at 360 nm, emission at 450 nm, and a gain setting of 100%. NO concentration for both methods was determined using a NaNO2 standard. Results are expressed as concentrations of NO per 106 cells/h.
For the purposes of these studies, we utilized K. pneumoniae[American Type Culture Collection (ATCC) 43816, type 2]. Bacterial stocks were maintained at −80°C until used. For in vitro experiments, stocks were thawed, inoculated into broth, and grown to log phase. All bacteria were washed free of growth media before infection studies. Colony forming units (CFUs) were determined by enumeration after serial dilution and inoculation onto agar plates.K. pneumoniae was grown in BBL brain heart infusion broth, and nutrient agar plates (DIFCO; Becton Dickinson) incubated in room air at 37°C for 18–24 h.
Binding assays were performed as previously described (26). Briefly, siliconized microcentrifuge tubes (Fisher Scientific, Pittsburgh, PA) were filled with HBSS containing 1% BSA for 24 h at 4°C to block nonspecific binding of SP-A to plastic (55). Log-phase K. pneumoniae were resuspended to a concentration of 1010 CFUs/100 μl in HBSS containing 1% BSA, pH 7.4. Bacterial numbers were confirmed by serial dilution and culture. Bacteria were mixed with SP-A in the presence of 1 mM Ca2+ or 1 mM EGTA in the absence of Ca2+ and incubated for 30 min at room temperature. Solutions were then centrifuged to pellet bacteria. Pellets were washed twice and resuspended in 200 μl of sterile water. Solutions were sonicated to disperse pellets. Control samples were incubated in the absence of bacteria to quantify nonspecific binding of SP-A to plastic. Bacteria-associated SP-A was then measured by ELISA.
Concentrations of SP-A in samples were determined by indirect ELISA using a polyclonal rabbit anti-human SP-A (Dr. David Phelps, Univ. of Pennsylvania, Hershey, PA) as the primary antibody and horseradish peroxidase-conjugated goat anti-rabbit immunoglobulin G as the secondary antibody. SP-A was quantified using a standard curve generated with SP-A diluted to 80 ng/ml applied to Immulon 2 ELISA plates (Dynatech Laboratories, Chantilly, VA), serially diluted and allowed to bind for 1 h at 37°C. The wells were blocked with BSA, incubated with each antibody for 1 h at 37°C, the reaction product generated using O-phenylenediamine dihydrochloride (2.2 mM) as substrate, and the optical density read at 490 nm (23). Results of the SP-A ELISA was expressed as nanograms of SP-A per total bacteria.
Cell culture assay. Bacterial in vitro killing assay.
5 × 105 AMs were allowed to adhere to plastic in 24-well tissue culture plates for 2 h, washed, and resuspended in DMEM supplemented as described previously. For all experiments, AMs were washed and resuspended in HBSS+ (containing Ca2+ and magnesium) and 0.1% low endotoxin BSA. AMs were then infected withK. pneumoniae in a concentration ratio of 10:1–100:1 bacteria per AM in the presence or absence of SP-A. After 30 min, infected AM cultures were washed to remove unattached bacteria and resuspended in HBSS+ and 0.1% low endotoxin BSA. AMs were scraped at 0, 1, and 2 h, and aliquots were cultured for determination of viable organisms by quantitative culture. In some experiments, 3–6 μg/ml of gentamycin sulfate was added after the 30-min incubation with SP-A to remove extracellular bacteria (57).
Generation of reactive oxygen and nitrogen species.
The aim of these experiments was to expose Klebsiella to a variety of reactive oxygen and nitrogen species at pH 7.4 and 5, the phagosomal pH. K. pneumoniae (108) were suspended in 10 ml of HEPES (25 mM) and exposed to the following reactants in autoclaved 130-ml centrifuge tubes: SIN-1 (0.5 or 1 mM), pH 7.4; PAPANONOate (100 μM), pH 7.4 or 5; and 10 mU/ml XO, 500 μM xanthine, and 100 μM FeCl3-EDTA. Because SIN-1 does not decompose at pH 5, in a subsequent series of experiments we generated ONOO− at pH 5 by coincubating 10 milliunits/ml XO, 500 μM xanthine, 100 μM FeCl3-EDTA, and 100 μM PAPANOate. Tubes were agitated in a shaking water bath at 37°C (25). Aliquots were taken at 0, 15, 30, 45, 60, and 90 min for determination of CFUs. The generation of ONOO− by SIN-1 was calculated from the oxidation of DHR-123 (22,32). NO was generated with 100 μM PAPANONOate in 25 mM HEPES buffer, pH 7.4 or pH 5. NO concentration was measured with an ISO-NO electrochemical probe (World precision Instruments, Sarasota, FL). Because the rate of PAPANONOate decomposition increases markedly at low pH, additional PAPANONOate was added every 10 min to maintain [NO] at a constant level based on the measured depletion of substrate at pH 5. Reactive oxygen species were generated by 10 mU/ml XO, 500 μM xanthine, and 100 μM FeCl3-EDTA in 25 mM HEPES buffer. Based on calculated depletion of the substrate, additional xanthine (500 μM) was also added every 15 min to maintain constant substrate levels. Reactivity of XO at pH 5 was adjusted according to the production of urate as measured spectrophotometrically at 400 nm.
Cytosolic Ca2+ in AMs was measured with dual-excitation wavelength fluorescence microscopy using the fluorescent probe fura 2-acetoxymethyl ester. Fura 2 fluorescence was measured at an emission wavelength of 510 nm in response to excitation wavelengths of 340 and 380 nm and alternated at a rate of 25 Hz by a computer-controlled chopper assembly. AMs attached to glass coverslips in DMEM were loaded with 10 μM of dye dissolved in DMSO for 45 min at 37°C with gentle agitation every 10 min. AMs were washed and placed in normal Ringer solution for fluorescence measurements. Increases in cytosolic Ca2+ resulted in an increase in the fura 2 340-nm signal and a decrease in the 380-nm signal. The 340:380 ratios were converted into Ca2+ values using the equation of Grynkiewicz et al. (20). In situ calibration was accomplished by permeabilizing AMs with 5 μM ionomycin and measuring fluorescence at both wavelengths under Ca2+-free (2 mM EGTA) or saturating (1.5 mM CaCl2) Ca2+ conditions.
In additional experiments, the buffering action of the intracellular Ca2+ chelator BAPTA-acetoxymethyl ester on changes in AM [Ca2+]i was tested in paired AM samples. Baseline [Ca2+]i and the initial rate of elevation in [Ca2+]i in response to 1 μM ionomycin were measured in control AMs loaded with fura 2 and in the absence and presence of 2 μM BAPTA-acetoxymethyl ester.
Labeling of bacteria with fluorescein.
Frozen stocks of K. pneumoniae were thawed and grown to log phase. Bacteria were washed four times in sterile PBS and resuspended in 15 ml of 0.1 M NaCO3 buffer, pH 9.2. 5- (and-6-)Carboxyfluorescein succinymidyl ester (Molecular Probes) was added to a final concentration of 100 μg/ml, and the suspension was stirred at room temperature in the dark for 2 h. LabeledKlebsiella were washed four times in sterile PBS to remove unconjugated fluorophore, resuspended in 10% glycerol in sterile H2O, and stored at −20°C.
Phagocytosis was assayed by flow cytometry as described previously (18, 49). Briefly, 5 × 105 AMs were suspended in DMEM, supplemented as described previously in the presence or absence of 2 μM BAPTA-acetoxymethyl ester, and incubated for 30 min at 37°C. AMs were infected with fluorescein-labeled K. pneumoniae at a ratio of 10 bacteria:1 AM in the presence or absence of 25 μg/ml SP-A and incubated for 1 h at 37°C with gentle shaking. Cells were centrifuged at 500 g for 10 min at 4°C to pellet, the supernatant was discarded, and 500 μl of ice-cold PBS plus 0.02% EDTA was added to stop phagocytosis. To identify extracellular bacteria, 10 μg/ml of ethidium bromide (EtBr) was added. EtBr immediately quenches the green fluorescence of extracellular bacteria by fluorescent resonance energy transfer but does not penetrate live cells. AMs with bound extracellular bacteria can therefore be distinguished from those containing phagocytosed bacteria because the former will fluoresce red and the latter will fluoresce green. Immediately after the addition of EtBr, samples were analyzed by flow cytometry.
Samples were analyzed with a FACScan flow cytometer (Becton Dickinson). The AM population was identified based on forward scatter (FSC) and side scatter (SSC) characteristics. Instrument settings and red/green fluorescence compensation were adjusted using appropriate unlabeled and single color control samples: AMs plus fluorescein-labeledKlebsiella and AMs plus EtBr. For each patient sample set, positive fluorescence was determined by gating on an unstained matched control AM population. Ten thousand events within the macrophage gate were analyzed per sample using histogram plots generated with CELLQuest analysis software (Becton Dickinson). Percentages of fluorescein- and EtBr-positive AM were determined together with mean channel fluorescence (MCF) of fluorescein- and EtBr-positive cells (an indication of mean number of phagocytosed/bound bacteria, respectively).
All experiments were performed three times in duplicate withn ≥ 2. Duplicate measurements were averaged and did not contribute to n. For human sample data experiments,n refers to total culture samples containing 5 × 105 cells with the total number of donors reported separately for each experiment. Data were analyzed by ANOVA followed by Tukey's multigroup comparison of the means for parametric data and by the Kruskal-Wallis ANOVA and Pearson's multigroup comparison of the means for nonparametric data. Klebsiella CFUs were converted to common logarithms for statistical analysis. Flow cytometry data were analyzed by paired Student's t-test. Results are expressed as means ± SE. P ≤ 0.05 was considered significant.
SP-A- and NO-mediated killing of K. pneumoniae.
SP-A has been shown to regulate the uptake of bacteria to which it is capable of binding (55), and there is a difference in the ability of SP-A to bind different strains of K. pneumoniae(28). We found that K. pneumoniae ATCC 43816, type 2, bound to SP-A in a dose-dependent fashion and that this binding was partially inhibited by the addition of EGTA (Fig.1).
We isolated AMs from the BAL of transplant patients and normal volunteers, maintained the AMs in primary culture for 18 h, and simultaneously infected them with K. pneumoniae in the presence or absence of SP-A (25 μg/ml). To control replication of noncell-associated bacteria within the media, gentamycin sulfate was added at the incubation stage of these experiments. AMs from normal volunteers were unable to kill Klebsiella at any time point tested, even in the presence of SP-A (Fig.2 A). In contrast, incubation of AMs from transplant patients withKlebsiella and SP-A resulted in a 45% decrease in CFUs by 60 min and a 68% (P < 0.01) decrease by 120 min (Fig.2 B). Incubation of Klebsiella with gentamycin sulfate (3–6 μg/ml) for 120 min in the absence of AMs did not result in decreased CFUs as evidenced by control samples incubated simultaneously in the absence of AMs (data not shown).
To determine the role of NO in SP-A-mediated Klebsiellakilling, AMs from transplant patients were pretreated with the NO synthase (NOS) inhibitorN G-monomethyl-l-arginine (l-NMMA; 1 mM for 30 min) and then infected withKlebsiella in the presence or absence of SP-A. SP-A mediated a 50% decrease in Klebsiella CFUs by 60 min compared with untreated media controls and a 68% decrease compared withl-NMMA-treated samples (P = 0.035). SP-A did not affect killing by l-NMMA-pretreated AMs (Fig.2 C).
NO and phagolysosomal involvement in bacterial killing.
To determine which specific reactive nitrogen species was involved inKlebsiella killing, we exposed K. pneumoniae to chemical generators of reactive species in the absence of AMs in 25 mM HEPES buffer. We utilized SIN-1 (1 mM or 500 μM), which is stable at pH 5 but decomposes at pH 7.4, to release NO and O ·, which then combine to form ONOO−. One millimolar SIN-1 produces ∼1 μM/min ONOO− in this buffer at 37°C (25). In the presence of SIN-1,Klebsiella CFUs continued to increase over the 90-min incubation period (Fig.3 A). The NO donor PAPANONOate (100 μM), which generates ∼4–5 μM NO, or PAPANONOate plus H2O2 (1 mM) also had no effect on Klebsiella growth (Fig. 3 B). Combinations of xanthine (500 μM), XO (10 mU), and Fe-EDTA (10 μM), which produce O ·, H2O2, and ·OH, had no effect on Klebsiella growth at pH 7.4 or pH 5 (data not shown). However, coincubation of Klebsiellawith xanthine and XO in combination with the NO donor PAPANONOate at pH 5 (lysosomal pH) resulted in a significant decrease inKlebsiella CFUs (Fig. 3 C). The combination of xanthine, XO, and PAPANONOate produced ∼2.8 ± 0.19 μM/min ONOO− at pH 5 as measured by the oxidation of DHR-123.
To determine whether exposure to low pH predisposedKlebsiella to killing by ONOO−, we incubatedKlebsiella at pH 5 for 120 min, washed, and then exposed them to 1 mM SIN-1 at pH 7.4. Preexposure of Klebsiella to pH 5 did not increase susceptibility to bacterial killing by SIN-1 (data not shown).
Ca2+ release from intracellular stores is thought to play a role in transcription, cell motility, and secretion. Previous studies have concluded that SP-A activates a phosphoinositide/Ca2+signaling pathway in rat AMs that may be responsible for enhanced serum-independent phagocytosis of bacteria (42). We measured changes in cytosolic Ca2+ levels within AMs from transplant patients and normal volunteers in response to SP-A. Baseline cytosolic Ca2+ levels were higher in AMs from normal volunteers (110 ± 26, n = 9, 3 separate patient samples) compared with transplant AMs (29 ± 10, n= 12, 8 separate transplant patient samples); means ± SE. In response to SP-A, intracellular Ca2+ levels increased by 70% above baseline in eight out of eight batches of transplant AMs. Examination of transplant patient demographics demonstrated that four male and four female patient samples had been tested and that there were no differences as to age, length of time from transplant, drug therapy, or infection or rejection status (data not shown). In contrast, SP-A decreased Ca2+ levels by 7% in AMs from normal volunteers (P < .01; Fig.4, A and B).
To determine the contribution of intracellular Ca2+ changes to Klebsiella killing, we inhibited intracellular Ca2+ changes with the Ca2+ chelator BAPTA-acetoxymethyl ester. Paired samples of transplant patient AMs were loaded with fura 2-acetoxymethyl ester or fura 2-acetoxymethyl ester plus 2 μM BAPTA-acetoxymethyl ester. BAPTA-acetoxymethyl ester-loaded AMs had a lower cytosolic Ca2+ baseline than BAPTA-acetoxymethyl ester-free controls, and the initial rate of rise in [Ca2+]i elicited by ionomycin was also significantly decreased by the buffering action of BAPTA-acetoxymethyl ester (Fig. 5, A–C). Preloading of transplant AMs with BAPTA-acetoxymethyl ester abrogated SP-A-mediated Klebsiella killing at 120 min (Fig. 6).
Phagocytosis of Klebsiella by human AMs.
SP-A has been shown to enhance phagocytosis of a variety of bacteria by rat AMs (36, 37, 45, 55). To determine the contribution of SP-A and intracellular Ca2+ changes to phagocytosis ofKlebsiella by human AMs, we loaded paired samples of transplant patient AMs with 2 μM BAPTA-acetoxymethyl ester and infected them with fluorescein-labeled Klebsiella in the presence or absence of 25 μg/ml of SP-A. To discriminate between adherent and internalized Klebsiella, 10 μg/ml of EtBr was added to samples immediately before analysis in the flow cytometer. EtBr transforms fluorescein-green fluorescence into red fluorescence by resonance energy transfer between the two fluorochromes (49, 54). However, EtBr does not enter living cells; therefore, ingested fluorescein-labeled Klebsiella retain their green fluorescence while extracellular Klebsiellafluoresce red. SP-A significantly increased phagocytosis of fluorescein-labeled Klebsiella by 32% compared with controls in the absence of SP-A (P < 0.01), while BAPTA-acetoxymethyl ester decreased phagocytosis by 22% compared with controls in the absence of BAPTA-acetoxymethyl ester (P= 0.01). The addition of SP-A to BAPTA-acetoxymethyl ester-treated samples protected against BAPTA-acetoxymethyl ester inhibition of phagocytosis but did not increase phagocytosis above control untreated levels (Fig. 7,A–C).
Importantly, neither BAPTA nor SP-A treatment significantly affected MCF of fluorescein-positive AMs (i.e., the number of intracellular bacteria per cell), suggesting that phagocytosis may be the rate-limiting step in bacterial killing (data not shown). Moreover, neither treatment altered either the percentage of AMs with bacteria bound to their surface (i.e., the percentage of EtBr-positive AM) or the number of bacteria bound per AM (i.e., the MCF of EtBr-positive AM; data not shown). This suggests that SP-A has a greater effect on phagocytosis than it does on cell-surface binding and indicates both that the effects of SP-A are specific and that the assay can clearly discriminate between extracellular and intracellular bacteria.
SP-A and NO production by human AMs.
We exposed AMs from transplant patients to human interferon (IFN)-γ (100–1,000 U/ml), rough (J5) and smooth (055:B5) forms of lipopolysaccharide (LPS; 10–200 ng), BSA (1 mg/ml), SP-A (25 μg/ml), and combinations of LPS and IFN-γ as well as live K. pneumoniae and measured NO and NO (NOx) in the media as an indication of NO production. The ability of AMs to “respond” to stimulus was defined as a statistically significant increase in measurable NOx compared with control unstimulated samples from the same patient. In all cases, >90% of the NOx was in the form of NO , in agreement with our previous measurements (50). SP-A stimulated increased NOx production in 10 out of 20 samples of AMs from transplant patients. SP-A “responder” AMs fell into two separate categories. In the first group (n = 3), NOxincreased by ∼2.2 nmol · h−1 · 106AMs−1, and in the second group (n = 7), NOx increased by ∼0.3 nmol · h−1 · 106AMs−1 compared with the corresponding media control (Fig.8). IFN-γ, LPS, and BSA had no effect on NOx levels. On the other hand, coincubation of AMs withKlebsiella resulted in a significant increase in NOx in all samples tested (1.31 ± 0.652 nmol · h−1 · 106AMs−1; n = 16, 7 separate transplant patient samples, Fig. 9 A).
Unstimulated AMs from transplant patients produced 244.5 ± 24.8 pmol · h−1 · 106AMs−1. In contrast, unstimulated AMs from normal volunteers produced 57.2 ± 5.48 pmol · h−1 · 106AMs−1 of NOx, and these levels were significantly decreased by the addition of SP-A. As in AMs from transplant patients, IFN-γ had no effect on NOxproduction by AMs from normal volunteers and failed to reverse the NO-suppressive effect of SP-A (Fig. 9 B).
We examined a select set of transplant patient demographics in an attempt to determine whether these might affect AM production of NO in response to SP-A. All patients were similar in age (51.7 ± 1.8 yr), nonsmokers (BAL was performed on transplanted lung only), and maintained on a standard three-drug regimen for immune suppression: prednisone plus either cyclosporine or tacrolimus plus either azathioprine or mycophenolate, with the most common immunosuppressive protocol being prednisone plus cyclosporine plus azathioprine. There were no discernible differences in drug regimen between responders and nonresponders. Patients were all Caucasian and were transplanted for a variety of reasons, listing from most to least common: chronic obstructive pulmonary syndrome > Eisenmenger's syndrome = α1-antitrypsin protease deficiency > idiopathic pulmonary fibrosis = cystic fibrosis = pulmonary hypertension. Neither the reason for lung transplant nor the drug regimen appeared to have any affect on the ability of these cells to produce NO in response to SP-A.
However, we determined that of the responder patients, 8 were female (80%) with an average of 11.4 mo from the time of transplant to BAL procedure. In contrast, of the nonresponder patient samples, there were 6 males and 4 females with an average length of time from transplant for males of 10.8 mo and females of 29.7 mo (Table 1). All normal volunteers were healthy nonsmokers.
The major findings of this study are 1) AMs from transplant patients are stimulated by SP-A or Klebsiella to produce NOx; 2) killing of Klebsiellaby AMs from transplant patients requires SP-A and NOx;3) incubation of AMs from transplant patients with SP-A increased [Ca2+]i and phagocytosis;4) abrogation of the increase in [Ca2+]i by preincubation with BAPTA-acetoxymethyl ester decreases phagocytosis and eliminates killing of Klebsiella; 5) neither SP-A norKlebsiella increases NOx or [Ca2+]i in AMs from normal volunteers; and6) AMs from normal volunteers incubated with SP-A do not kill Klebsiella.
The ability of human AMs to produce NO and effect bacterial killing has been alluded to in a number of clinical studies (1, 33, 41, 47,59); however, the majority of NO bacterial pathogenesis studies has relied on information obtained by utilizing rodent models (15, 39, 40). Although those studies have provided a vast amount of information regarding the role of NO in bacterial pathogenesis, caution must be exercised when extrapolating these results to human cells because iNOS expression and production of NO by AMs differ dramatically among species (27). Likewise, information regarding the immunomodulatory functions of SP-A in the regulation of NO production (5, 29) and in bacterial clearance (25, 28, 34, 35, 55) relies heavily on rat and mouse models.
We found that SP-A binding to K. pneumoniae ATCC 43816, serotype 2, was Ca2+ dependent. Previous reports have indicated that other strains of Klebsiella bind to SP-A much more tightly than serotype 2 (28); however,Klebsiella serotype 2 is one of the most virulent of theKlebsiella strains and has been utilized in a variety of pathogenesis studies (7, 48, 56), making it a logical choice for this series of experiments.
Consistent with in vitro studies utilizing rat AMs (28), in the presence of SP-A, transplant AMs effected a 50% decrease inKlebsiella CFUs by 60 min and a 68% decrease in CFUs by 120 min. SP-A was unable to stimulate Klebsiella killing by AMs from normal volunteers, a fact that was not surprising, considering the inability of SP-A to stimulate NO production in these cells.
Previous studies in mice utilized the NOS inhibitor NG-nitro-l-arginine methyl ester to inhibit the clearance of respiratory K. pneumoniae infection (56). The addition of the iNOS inhibitorl-NMMA to infected transplant AM cultures abrogated SP-A-mediated Klebsiella killing, confirming the importance of reactive oxygen-nitrogen intermediates for this effect. The fact that Klebsiella alone significantly upregulated NO production in AMs from transplant patients demonstrates that live bacteria are more effective in generating a response from these cells than isolated LPS. However, despite significant NO production in the absence of SP-A, the addition of SP-A stimulated a significant increase in Klebsiella killing, which indicates that SP-A may be additionally stimulating phagocytosis. These data differ from a previous report in which human AMs from patients with pulmonary fibrosis were shown to kill the intracellular pathogen bacillus Calmette-Guerin by a NO-mediated mechanism in the absence of any immunomodulatory agent such as SP-A (41).
In the absence of AMs, Klebsiella was only susceptible to killing by generation of O · and NO at pH 5. NO was not toxic at pH 5, indicating that acidified NO was not responsible for Klebsiella killing. O · and NO react at near diffusion rates to form the highly reactive compound ONOO− (3), which has been shown to be extremely bactericidal (9, 11, 12,25). At pH 5, most ONOO− is in the protonated form of peroxynitrous acid (ONOOH), which has been demonstrated to cross membranes and thereby damage intracellular components by passive diffusion (13). Klebsiella have prominent capsules of complex polysaccharides that form thick bundles of fibrillous structures covering the bacterial surface and protecting the bacterium from phagocytosis and the bactericidal effects of complement (46). While the microenvironment of the phagolysosome allows for the concentration of such toxic radicals as ONOO− (12), the low pH may also allow ONOOH to penetrate the thick capsule of Klebsiella and thus damage intracellular lipids, proteins, or DNA (15, 43). Previous studies have demonstrated that SP-A activates a phosphoinositide/Ca2+ signaling pathway in rat AMs, leading to enhanced serum-independent phagocytosis of bacteria (42). Alternatively, translocation of p47phoxfor the respiratory burst production of superoxide is Ca2+dependent (60), which may provide an additional mechanism by which Ca2+ is important for pathogen killing. Consistent with rat studies, SP-A stimulated a significant increase in intracellular Ca2+ levels in AMs from transplant patients; however, SP-A decreased intracellular Ca2+ levels in AMs from normal volunteers. These differences in Ca2+production between transplant patients and normal volunteers in response to SP-A were not consistent with the differential effects of SP-A for NO production since all transplant patient AMs reacted similarly with an increase in [Ca2+]i in response to SP-A. Ca2+ is known to be important for NO production by the constitutive form of NOS; however, studies with mouse AMs have indicated that this contribution to overall NO production by the constitutive form of NOS is very slight (∼0.12 nm · h−1 · 105AMs−1) (57). Inhibition of intracellular Ca2+ changes abrogated SP-A-mediatedKlebsiella killing by transplant AMs, demonstrating the requirement for Ca2+ and implicating the importance of phagocytosis and/or superoxide production for this effect.
We utilized flow cytometry to determine the ability of SP-A to stimulate phagocytosis of Klebsiella. We found that SP-A increased the total number of transplant patient AMs that phagocytoseKlebsiella by 32% above untreated AMs. This method also allowed us to determine that SP-A did not increase the total amount of surface-bound Klebsiella per AM (MCF for red-positive AMs), which suggests that phagocytosis is the rate-limiting step, i.e., adherent bacteria are more likely to be phagocytosed in the presence of SP-A. Moreover, we did not see an increase in MCF for fluorescein-positive AMs (expected if Klebsiella accumulate within the cell) in the presence of SP-A, which suggests that onceKlebsiella are phagocytosed, they are rapidly degraded. In contrast, preincubation of AMs from transplant patients with BAPTA-acetoxymethyl ester decreased Klebsiella phagocytosis without affecting adherence. This finding supports the premise that SP-A-stimulated intracellular Ca2+ fluxes are important for phagocytosis of Klebsiella.
SP-A stimulated a significant increase in NO production by AMs isolated from 10 out of the 20 transplant patients tested. BALs were performed only on the transplanted lung, which raises the question of the source of the AMs utilized in these studies, i.e., donor vs. recipient. Previous studies have utilized microsatellite technology (30) and human leukocyte antigen immunocytochemistry (44) to determine that transplant recipients rapidly repopulate the transplanted organ with host cells; however, some donor cells may remain in excess of 6 mo posttransplant. The average length of time post transplant for responder AMs was 11.4 ± 4 mo for females and 6.5 mo for males, while the time post transplant for nonresponder AMs was 29.7 ± 10.3 mo for females and 10.8 ± 5.9 mo for males. These data indicate that donor cells may be contributing to any of the responses to SP-A, either positive or negative, except for the nonresponder female patients. Additionally, 8 out of 10 of the transplant responder patients were female, a finding that correlates well with previous reports of increased NO production by AMs from female rats (51).
Previous studies also have attempted to identify the emergence of an inflammatory phenotype for AMs from transplant patients (19); however, inflammatory surface antigen was only weakly expressed on AMs from patients exhibiting normal cellular content (>90% AMs), as were all the samples utilized for these studies. It should be considered that the inability of transplant AMs to respond to LPS may be attributed to the lack of LPS-binding protein because all of the AMs were cultured in the absence of serum. Serum proteins interfere with SP-A interaction with AMs, and, therefore, serum was omitted from all cell activation protocols.
AMs isolated from normal volunteers produced significantly lower baseline levels of NO compared with transplant AMs, and exposure to SP-A significantly decreased NO production further. SP-A has been shown to differentially regulate NO production by rat AMs (52), and, recently, heterogeneous response patterns of NO production were reported in IFN-γ-stimulated AMs from lung cancer patients (14). Our results indicate that SP-A can stimulate human AMs to produce completely different responses dependent on the cell source. The ability of SP-A to differentially regulate NO production and yet uniformly regulate Klebsiella killing indicates that while SP-A and NO are both necessary for the killing of this pathogen, SP-A may not be the primary trigger for NO production involved inKlebsiella killing.
These data further support the hypothesis that the ability of SP-A to modulate cell responses is dependent on the activation state of the cell (52) and that the cellular environment is of extreme importance when attempting to understand cellular responses to external stimuli (e.g., bacteria). However, these data also seem to present a dilemma. AMs from transplant patients produce significantly larger amounts of NO and kill Klebsiella in vitro and yet transplant patients are immunosuppressed and extremely susceptible to infections. Several explanations exist for this apparent dichotomy:1) AMs isolated from immunosuppressed patients and cultured in the absence of drugs recover quickly (17);2) transplanted organs are by definition foreign and, therefore, inflammatory for the host; and 3) constant inflammation caused by the transplanted organ in vivo will not only activate AMs but may stimulate the production of excess reactive oxygen-nitrogen species by a variety of cell types, causing damage to the epithelial lining, SP-A, and the innate immune response. Nitration of SP-A in vitro abrogates its ability to mediate binding and phagocytosis of Pneumocystis carinii (61), and nitrated SP-A has been isolated from the edema fluid of patients with Adult Respiratory Distress Syndrome (62). Although NO defends the host against infectious agents, its effects are nonspecific, and overproduction of NO may be cytotoxic not only for microbes but also for the cells and tissues that produce it (24).
We thank Dr. J. Russell Lindsey for intellectual input and Christine Miskall and Julie Gibbs-Erwin for excellent technical support.
This work was supported by National Institutes of Health Grants RR-00149 (to J. M. Hickman-Davis) and HL-31197 and HL-51173 (to S. Matalon).
Address for reprint requests and other correspondence: S. Matalon, Dept. of Anesthesiology, Univ. of Alabama at Birmingham, 619 S. 19th St., THT 940, Birmingham, AL 35294 (E-mail).
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
First published November 2, 2001;10.1152/ajplung.00216.2001
- Copyright © 2002 the American Physiological Society