Alveolar Macrophages from Human ... - Semantic Scholar

Alveolar Macrophages from Human Immunodeficiency Virus–Infected Persons Demonstrate Impaired Oxidative Burst Response to Pneumocystis carinii In Vitro Henry Koziel, Xiuhong Li, Martine Y. K. Armstrong, Frank F. Richards, and Richard M. Rose* Division of Pulmonary and Critical Care Medicine, Beth Israel Deaconess Medical Center and Harvard Medical School, Boston, Massachusetts; and MacArthur Center for Molecular Parasitology, Yale University School of Medicine, New Haven, Connecticut

The alveolar macrophage (AM) oxidative burst response is an important component of microbicidal effector cell function against a variety of potential pathogens in the lungs, although the role against Pneumocystis carinii has not been fully investigated. The goals of this study were to characterize the P. carinii–mediated oxidative burst of AMs from healthy individuals, and to examine the oxidative burst of AMs from human immunodeficiency virus (HIV)–infected persons. For healthy individuals, the AM oxidative burst (measured as hydrogen peroxide [H2O2] production) increased in a time- and concentration-dependent manner in response to P. carinii or to the major surface glycoprotein of P. carinii, gp-A (0.01 to 10 !g/ml), required physical contact of P. carinii with AMs, and was not dependent on organism viability. Enzymatic removal of the surface-associated molecules of P. carinii reduced the oxidative burst to 43% of control (P ! 0.01). Blocking the AM mannose receptor reduced the P. carinii–mediated oxidative burst response to 37% of control (P ! 0.01). Compared with AMs from healthy individuals, P. carinii–mediated H2O2 production was significantly reduced in AMs from asymptomatic HIV-positive (HIV") persons with CD4" counts # 200 cells/mm3 (249 # 43 relative fluorescence units [RFU] versus 130 # 44 RFU; mean # standard error of the mean, P $ 0.038) and HIV" persons with active P. carinii pneumonia (78 # 40 RFU; P $ 0.014), but preserved for HIV" persons with CD4" counts $ 200 cells/mm3. Importantly, H2O2 production in response to phorbol myristate acetate or serum-opsonized zymosan particles was preserved in all groups studied. Thus, AM oxidative burst, mediated in part via P. carinii gp-A and AM mannose receptor may represent an important host response to P. carinii. A specific impairment of P. carinii–mediated AM oxidative burst in persons with advanced HIV infection may contribute to the pathogenesis of P. carinii pneumonia.

The pulmonary host factors accounting for susceptibility to Pneumocystis carinii pneumonia are only partially understood (1). Alveolar macrophages (AMs) are the most abundant immune cell in the alveolar air space, and mediate important antimicrobial functions against invading patho(Received in original form December 30, 1999 and in revised form May 3, 2000) *Current address: Stem Cells, Inc., Sunnyvale, CA. Address correspondence to: Henry Koziel, M.D., Div. of Pulmonary and Critical Care Medicine, KSB-23, Beth Israel Deaconess Medical Center, 330 Brookline Ave., Boston, MA 02215. E-mail: [email protected]. edu Abbreviations: alveolar macrophage, AM; bronchoalveolar lavage, BAL; BAL fluid, BALF; bovine serum albumin, BSA; dichlorodihydrofluorescein diacetate, H2DCFDA; hydrogen peroxide, H2O2; Hanks’ balanced salt solution containing calcium and magnesium, HBSS"; heat-inactivated, HI; human immunodeficiency virus, HIV; immunoglobulin, Ig; the major surface glycoprotein of P. carinii, P. carinii gp-A; phosphatidylinositol, PI; phorbol myristate acetate, PMA; relative fluorescence units, RFU. Am. J. Respir. Cell Mol. Biol. Vol. 23, pp. 452–459, 2000 Internet address: www.atsjournals.org

gens via oxidative and nonoxidative mechanisms (2). Several observations support an important role for AMs in the successful host response to P. carinii (3). Rats selectively depleted of AMs demonstrate impaired clearance of P. carinii pneumonia (4), resolution of pneumonia in various animal models of P. carinii pneumonia (5–7) is associated with an influx of AMs and release of macrophagederived cytokines (8), and in vitro exposure of P. carinii to macrophages results in phagocytosis and rapid digestion of the organisms (9, 10). However, the mechanism(s) by which AMs effectively clear P. carinii are not completely known. The macrophage oxidative burst response represents an important component of microbicidal effector cell function against a variety of potential pathogens (11). Macrophagemediated phagocytosis of microbes is associated with release of oxygen radicals into the phagosome and the extracellular environment (12). Evidence suggests that the oxidative burst may be important in the host response to P. carinii. In vitro, P. carinii induces an oxidative burst response in a rodent NR8383 macrophage cell line and rodent AMs (13), and in monocytes and monocyte-derived macrophages from healthy individuals (14). Inasmuch as products of the oxidative burst response (such as superoxide anion and hydrogen peroxide [H2O2]) may be directly toxic to P. carinii (15), the macrophage oxidative burst response may represent an important mechanism for elimination of P. carinii in the healthy host. Previous observations in our laboratory demonstrated evidence for oxidative burst response by AMs from healthy individuals exposed to P. carinii in vitro (16), although this was not fully characterized. Further, for persons infected with human immunodeficiency virus (HIV) type-1 (representing the risk group with the highest incidence of P. carinii pneumonia [17]), AMs have an impaired ability to phagocytose P. carinii in vitro (18), although the capacity to release reactive oxygen species (ROS) in response to P. carinii is not known. The purposes of the present study were: (1) to characterize the P. carinii–mediated oxidative burst response in AMs from healthy individuals; and (2) to examine the oxidative burst response of AMs from HIV-infected individuals to P. carinii challenge in vitro.

Materials and Methods Study Subjects Study subjects recruited for research bronchoscopy included healthy individuals and asymptomatic HIV-seropositive individuals. All study subjects were without clinical or radiographic evidence for active pulmonary disease and had normal spirometry.

Koziel, Li, Armstrong, et al.: Impaired HIV Alveolar Macrophage Oxidative Burst Response to P. carinii

HIV-seropositive individuals were characterized according to low clinical risk for P. carinii pneumonia (peripheral blood CD4" T-lymphocyte count $ 200 cells/mm3) or high clinical risk for P. carinii pneumonia (peripheral blood CD4" T-lymphocyte count # 200 cells/mm3). In addition, bronchoscopy specimens from HIV-seropositive individuals with active P. carinii pneumonia were available for this study. All subjects were confirmed HIV-seropositive or HIV-seronegative.

Bronchoscopy Bronchoalveolar cells were obtained from study subjects by bronchoalveolar lavage (BAL) using standard technique (18). All BAL procedures were performed on consenting adults under the auspices of the Institutional Review Board of the Beth Israel Deaconess Medical Center. Briefly, after topical 1% lidocaine anesthesia, a fiberoptic bronchoscope was wedged in a subsegment of the right middle lobe, and BAL performed by instilling 4 to 6 50-ml aliquots of warm normal saline (0.9%), followed by gentle suction after the instillation of each aliquot. For the HIV-seropositive individuals with active P. carinii pneumonia, BAL cells were obtained from clinical specimens of individuals undergoing diagnostic bronchoscopy for high clinical suspicion of P. carinii pneumonia, and only that portion of the BAL specimen which remained after diagnostic clinical use was available for this study. All cases demonstrated P. carinii by immunofluorescent staining of BAL specimens.

AMs The BAL cell pellet was isolated from the pooled BAL fluid (BALF) by centrifugation at 100 % g for 10 min and washed in cold media RPMI 1640 (Cellgro; Mediatech, Washington, DC) supplemented with 100 U/ml penicillin and 100 mg/ml streptomycin (Sigma, St. Louis, MO), and cells were counted on a hemacytometer. For the fluorescent microplate and cytochrome c reduction assays, AMs were isolated by adherence in 24-well plastic tissue culture plates (7.5 % 105 cells/well) (Falcon) for 2 h at 37&C in 5% CO2, then washed and maintained overnight in complete RPMI1640 medium supplemented with 10% heat-inactivated fetal calf serum (JRH Biosciences; Lanexa, KS), 100 U/ml penicillin, 100 mg/ml streptomycin, and 2 mM L-glutamine (Sigma) at 37&C in 5% CO2. By this method, the remaining adherent cells were $ 98% viable as determined by trypan blue dye exclusion, demonstrated $ 95% positive nonspecific esterase staining, and were $ 95% actively phagocytic of 1.1-'m latex beads. For the flow cytometry assay, BAL cells were maintained in suspension.

P. carinii Organisms Organisms were isolated from the lungs of the immunosuppressed rat model of P. carinii pneumonia (19). Briefly, barrierraised, male Sprague–Dawley rats (Hilltop Laboratory Animals, Scottdale, PA) were maintained on dexamethasone (1 mg/liter in the drinking water) and low-protein (8%) normocaloric diets. After 9 to 10 wk when the animals developed moderate-to-severe P. carinii pneumonia, the animals were killed and the lungs aseptically removed, homogenized, and filtered through sterile wire mesh. The P. carinii was isolated from the supernatant by differential centrifugation, then maintained in short-term culture with a feeder layer of mink lung cell line MV 1 Lu to disperse the native P. carinii clusters. After 8 d, culture supernatants containing the organisms were centrifuged and washed, morphology was verified by modified Giemsa stain (which stains the trophozoiteand cyst-form nuclei) and by Toluidine blue O stain (which stains the cyst-forms), and suspensions were counted as previously described (20). These mixed life-cycle preparations used in all experiments contained approximately 90% trophozoite and 10% cyst forms, with viability $ 85% (21). These preparations were $ 98%

453

free of contaminating host cells. P. carinii preparations visibly contaminated by other microorganisms or as determined by routine microbiologic culture were excluded. P. carinii preparations were endotoxin-free (# 1.0 endotoxin units/ml) as determined by E-toxate Limulus polyphemus assay (Sigma). For select experiments, viable P. carinii organisms were compared with heatkilled (100&C % 10 min) organisms.

P. carinii Surface Molecules To examine the effect of P. carinii surface molecules in the generation of an oxidative burst, select experiments were performed with organisms pretreated with zymolyase, an enzyme preparation that cleaves ((1,3)-glycosidic linkages and removes the outer cell surface of P. carinii (22). As previously described (20), organisms were divided equally and one portion was preincubated with zymolyase (ICN Biomedicals, Aurora, OH), 0.33 mg/ml in phosphate-buffered saline at 21&C for 30 min, then washed. The oxidative burst was measured comparing zymolyase-pretreated organisms and organisms without enzymatic pretreatment. To determine the role of the major surface glycoprotein, experiments compared the oxidative burst response with increasing concentrations of human-derived P. carinii gp-A (the major surface glycoprotein of P. carinii; the generous gift of Drs. B. Lundgren and J. Lundgren, University of Copenhagen and Hvidovre University Hospital, Hvidovre, Denmark) with intact P. carinii organisms. The humanderived P. carinii gp-A was endotoxin-free (# 1.0 endotoxin units/ ml) as determined by E-toxate L. polyphemus assay (Sigma).

Oxidative Burst Response by Fluorescence Microplate Assay AM oxidative burst response (measured as the production of H2O2) was performed by a fluorescence microplate assay (23) using dichlorodihydrofluorescein diacetate (H2DCFDA) (Molecular Probes, Eugene, OR) (24). H2DCFDA is a cell-permeant fluorogenic probe that passively diffuses into cells, and in the presence of ROS is oxidized to a fluorescent moiety. AMs were incubated with H2DCFDA (10 'M final) at 37&C in 5% CO2 in balanced salt solution containing calcium and magnesium (BSS"), in the presence and absence of particles or stimulatory molecules. H2DCFDA and the appropriate particles or molecules were added to the wells containing AMs at time 0, and at the appropriate time points fluorescence for each well was measured at an excitation wavelength of 485 nm and emission wavelength of 530 nm (Cytofluor 2300 Measurement System; Millipore, Bedford, MA). Measurements are expressed as relative fluorescence units (RFU). For the time-dependence experiments, measurements were performed at time 0, 30, 60, 90, 120, 150, and 180 min. Scan time for the microplate assay was # 10 s. Background AM autofluorescence and spontaneous conversion of H2DCFDA by unstimulated AMs were subtracted from experimental measurements at each time point except where noted. As control conditions, P. carinii were incubated with H2DCFDA in the absence of AMs (to examine the contribution of P. carinii to the generation of H2O2), and P. carinii were separated from AMs by the use of 0.4-'m-pore-diameter membrane inserts (Costar Transwell, Cambridge, MA) in the tissue culture wells, which physically separated the organisms from the AMs but allowed interaction of soluble components through a semipermeable membrane filter.

Alternative Stimulants for Oxidative Burst Response To evaluate the specificity of the P. carinii–mediated oxidative burst response, phorbol myristate acetate (PMA) (Sigma) and heat-inactivated (HI) serum-opsonized zymosan particles (Molecular Probes) were used. For opsonized zymosan, particles were incubated with 10% HI human serum for 20 min at 37&C,

454

AMERICAN JOURNAL OF RESPIRATORY CELL AND MOLECULAR BIOLOGY VOL. 23 2000

washed and resuspended in Hanks’ BSS" (HBSS"), and added to experimental conditions at a multiplicity of 20 particles per macrophage. Zymosan particles were sonicated immediately before use.

TABLE 1

Demographic characteristics of study subjects HIV"

Detection of Oxidative Burst by Cytochrome c Reduction As an independent validation of the fluorescence microplate assay, select specimens of adherent AMs were analyzed for oxidative burst response as measured by the reduction of cytochrome c as previously described (25). Replicate wells of AMs containing 80 'M cytochrome c (Sigma) in the presence and absence of 10 'g/ml superoxide dismutase (SOD) (Sigma) were incubated with P. carinii for 90 min at 37&C in 5% CO2. The concentration of SOD-inhibitable O2) (nmol O2)/ml) in 1-ml aliquots of cultured supernatants was determined by absorbance at 550 nm (OD550) using spectrophotometry. Results were compared with unstimulated AMs. As controls, blanks for OD550 were prepared with complete reaction mixtures maintained for 90 min at 37&C in 5% CO2 in the absence of AMs.

Subjects, n Age, yr mean * standard deviation Female gender, n (%) Tobacco use, n (%) CD4" T-cell count, mean, cells/mm3 Antiretroviral treatment, n (%)

Healthy

CD4 $ 200

CD4 # 200

PCP

20

15

11

4

38 * 10 5 (25) 6 (30)

37 * 9 4 (33) 5 (42)

39 * 7 5 (45) 6 (55)

38 * 5 1 (25) 1 (25)

NA

344 * 173

87 * 59

15 * 28

6 (50)

9 (82)

4 (100)

None

NA ! not available; PCP ! Pneumocystis carinii pneumonia.

Oxidative Burst Response by Flow Cytometry Analysis To validate the fluorescence microplate assay, and to examine the oxidative response of macrophages in suspension, samples were independently analyzed by flow cytometry. Oxidative burst response determinations were performed on select specimens as described (26). Briefly, suspensions of AMs (5 % 105 BAL cells) in 0.5 ml HBSS" were incubated with H2DCFDA (200 'M final) in HBSS" with or without appropriate particles or molecules for 30 min at 37&C in the dark with end-to-end rotation. The reaction was terminated by placing tubes on ice, proprydium iodide (PI) (10 'g/ml final) was added to each tube, and the samples were analyzed on a Coulter Epics Profile II flow cytometer (Coulter Electronics, Inc., Hialeah, FL). Cells were first examined on a forwardversus-PI bivariate histogram to gate out PI-positive (nonviable) cells. The PI-negative (viable) population was then identified on the basis of characteristic forward and right-angle (side) scatter properties, and analyzed by univariate histogram for fluorescein isothiocyanate. As control conditions, AMs in duplicate samples were preincubated with antimycin A (10)7 M final; inhibitor of mitochondrial respiration) for 10 min before adding H2DCFDA. Samples were prepared and analyzed in duplicate, and a minimum of 5,000 cells were counted for each sample. For each sample, the proportion of cells staining positive and a mean log fluorescence value were determined for the population of cells and the results recorded as mean RFU. For the purpose of reducing nonspecific binding, all solutions contained 0.1% bovine serum albumin (BSA).

infected persons were similar to those previously reported (18) (data not shown). BALF return and BAL cell differential counts were not available for the persons with active P. carinii pneumonia. Time Dependence of AM Oxidative Burst Response to P. carinii Initial experiments examined the oxidative response to P. carinii of AMs from healthy individuals. Incubation of AMs with P. carinii (multiplicity of 10:1, organisms to AMs) resulted in a time-dependent release of H2O2 as determined by fluorescence microplate analysis. H2O2 increased linearly over the course of incubation up to 180 min (Figure 1). For wells containing only H2DCFDA (without AMs), there was no production of H2O2 as fluorescent measurements remained at background levels (data not shown). On the basis of time-dependence experiment results, subsequent fluorescence microplate assays were measured at 90 min

Statistical Analysis All experiments were performed in duplicate on at least three different occasions using AMs from at least three different individuals, except where noted. The results of each experiment are presented as means * standard error of the mean (SEM) or as percent change compared with control conditions. Statistical data for comparison of groups were calculated by analysis of variance, and parametric data were compared using Student’s t test, using INSTAT2 statistical package (Graphpad Software, San Diego, CA) on an IBM PS/2 386 computer. Statistical significance was accepted for P # 0.05.

Results Clinical Characteristics of the Study Subjects The clinical characteristics of the study subjects are presented in Table 1. BALF return and BAL cell differential determinations comparing healthy with asymptomatic HIV-

Figure 1. Time-dependent H2O2 production by AMs from healthy individuals to mixed life-cycle P. carinii (solid circles; multiplicity 10:1). Incubations were performed at 37&C in 5% CO2. Spontaneous H2O2 production by AMs is indicated by the open circles (n ! 6).


because this represented the time of maximum interaction of P. carinii with AMs (20). Concentration Dependence for AM Oxidative Burst Response to P. carinii and P. carinii gp-A H2O2 production by AMs from healthy individuals demonstrated a concentration-dependent response to P. carinii over a range of P. carinii/AM ratios, 1:1, 2:1, 5:1, and 10:1 (Figure 2A), incubated at 37&C in 5% CO2 for 90 min. To determine the effect of the major surface glycoprotein of P. carinii gp-A on AM oxidative burst, AMs from healthy individuals were incubated with P. carinii gp-A at 37&C in 5% CO2 for 90 min. Production of H2O2 demonstrated a concentration-dependent response to P. carinii gp-A over a range of 0.1 to 10 'g/ml (Figure 2B). As a control for a possible effect of rat-derived proteins in the P. carinii preparations, there was no response to lung homogenate from normal control rat preparation (data not shown). As a control for the human-derived P. carinii gp-A, no effect on H2DC-

Figure 2. Concentration-dependent H2O2 production by AMs from healthy individuals for (A) whole mixed life-cycle P. carinii organisms; and (B) the major component of the organism surface glycoprotein P. carinii gp-A. P. carinii (10:1 P. carinii/AM ratio) or P. carinii gp-A were incubated with AMs for 90 min at 37&C in 5% CO2 (n ! 4).

455

FDA conversion was observed when AMs were incubated with concentrated (15% to 20%) BALF from healthy human subjects (data not shown). Characterization of AM Oxidative Burst Response to P. carinii To further characterize the oxidative burst response to P. carinii, AMs from healthy individuals were incubated with P. carinii at a multiplicity of 10:1 at 37&C in 5% CO2 for 90 min. To normalize for biologic variability of P. carinii organisms and human AMs, data are expressed as a change from control (representing AMs incubated with viable P. carinii organisms). Compared with unstimulated cells, H2O2 release was increased after incubation of AMs with viable P. carinii organisms (Figure 3). Compared with unstimulated cells, AM oxidative burst response was increased after incubation with heat-killed (nonviable) P. carinii, and the value was similar to viable organisms (control; P $ 0.05). Enzymatic removal of the outer surface molecules of P. carinii reduced the respiratory burst response to 43% of control (P ! 0.01). Physical separation of the organisms from the AMs reduced the macrophage oxidative burst response to 22% of control (P ! 0.001). In the presence of a competitive inhibitor of the AM mannose receptor (mannosyl-BSA, 1,000 'g/ml), the oxidative burst response was reduced to 37% of control (Figure 3; P ! 0.01), whereas BSA alone had no effect (data not shown). To examine the relative contribution of P. carinii organisms to H2O2 production, incubation of P. carinii with H2DCFDA in the

Figure 3. The characterization of healthy AM H2O2 production in response to mixed life-cycle P. carinii. The production of H2O2 by AMs from healthy individuals was determined after incubation with unopsonized P. carinii, and compared with heat-killed P. carinii organisms, P. carinii organisms pretreated with zymolyase (to remove surface molecules), and P. carinii organisms physically separated from the AM monolayers. Additional conditions measured H2O2 production after incubation with P. carinii in the presence of a competitive inhibitor of the macrophage mannose receptor (1,000 'g/ml mannosyl-BSA), and H2O2 production associated with P. carinii in the absence of AMs. Spontaneous H2O2 release by AMs was measured in the absence of P. carinii. P. carinii (10:1 P. carinii/AM ratio) were incubated with AMs for 90 min at 37&C in 5% CO2 (n ! 5). *P + 0.01.

456


absence of AMs reduced H2O2 production to 8% of control (P ! 0.001). Oxidative Burst Response to P. carinii of AMs from HIV-Infected Individuals After challenge with P. carinii for 90 min at 37&C in 5% CO2, AMs from healthy individuals (n ! 9) revealed a value of 249 * 43 RFU (mean * SEM). A similar value was observed for asymptomatic HIV" persons with CD4" counts $ 200 cells/mm3 (n ! 7), 180 * 33 RFU (P ! 0.17); whereas H2O2 was reduced for asymptomatic HIV" persons with CD4" counts # 200 cells/mm3 (n ! 3), 130 * 44 RFU (P ! 0.038), and persons with active P. carinii pneumonia (n ! 4), 78 * 40 RFU (P ! 0.014) (Figure 4). As an independent validation of the fluorescence microplate assay, select specimens were analyzed for oxidative burst response by measurement of superoxide (O2)) generation. Similar to results observed with the fluorescence microplate assay, superoxide production for AMs from healthy individuals (n ! 4) was 24.4 * 6.9 nmol O2)/ ml and for asymptomatic HIV" persons with CD4 $ 200 cells/mm3 (n ! 4) was 18.7 * 7.0 nmol O2)/ml (P $ 0.05). Compared with AMs from healthy individuals, superoxide production was 12.1 * 6.2 nmol O2)/ml (P # 0.05) for asymptomatic HIV" persons with CD4 # 200 cells/mm3 (n ! 5), and 5.6 * 1.6 nmol O2)/ml (P # 0.01) for HIV" individuals with active P. carinii pneumonia (n ! 3). Specificity of P. carinii–Mediated Oxidative Burst Response of AMs from Healthy and HIV-Infected Individuals To determine whether differences in the oxidative burst response to P. carinii represented a specific or generalized

Figure 4. Impaired oxidative burst response in AMs from subjects with advanced HIV disease. Comparison of H2O2 production by AMs from healthy individuals (n ! 9) with asymptomatic HIV-infected subjects with CD4" T-lymphocyte count $ 200 cells/ mm3 (n ! 4), CD4" T-lymphocyte count # 200 cells/mm3 (n ! 3), and persons with active P. carinii pneumonia (n ! 4). Mixed life-cycle P. carinii (10:1 P. carinii/AM ratio) were incubated with AMs for 90 min at 37&C in 5% CO2. (*P # 0.05 compared with healthy individuals; **P # 0.05 compared with healthy individuals and compared with the HIV" group with CD4" T-lymphocyte count $ 200 cells/mm3).

alteration in oxidative burst capacity, AMs from healthy and from HIV-infected individuals were examined for the oxidative burst response to other stimuli. For healthy individuals, AMs exhibited a time-dependent oxidative burst response to PMA (10)7 M final) and HI serum-opsonized zymosan (20:1) (Figure 5A). The oxidative burst response to PMA and HI serum-opsonized zymosan was similar for asymptomatic HIV-infected persons with peripheral blood CD4" T-lymphocyte counts $ 200 cells/mm3 (Figure 5B) and peripheral blood CD4" T-lymphocyte counts # 200 cells/mm3 (Figure 5C). AMs from patients with active P. carinii pneumonia were not available for flow cytometry analysis. Flow Cytometry Analysis of AM Respiratory Burst Response to Stimulation As an independent validation of the fluorescence microplate assay, and to examine the respiratory burst of AMs in suspension, select specimens were analyzed by flow cytometry. Quantitative determinations demonstrated that the oxidative burst response to PMA or HI serum-opsonized zymosan was preserved for all groups investigated (Table 2), similar to results observed with the fluorescence microplate assay.

Discussion These data demonstrate that AMs from healthy individuals release H2O2 in a time- and concentration-dependent manner in response to P. carinii in vitro. The oxidative burst requires physical contact, is independent of P. carinii viability, and is mediated in part through the major surface glycoprotein P. carinii gp-A, as the enzymatic removal of the outer surface of P. carinii resulted in reduced respiratory burst response and exposure of AM to exogenous P. carinii gp-A was sufficient for an oxidative burst response. Together, these observations suggest that the oxidative burst may represent an important host-cell response to this oxygen-sensitive, opportunistic pathogen. In comparison with healthy individuals, AMs from asymptomatic HIV-infected individuals with CD4-T-lymphocyte counts # 200 cells/mm3 (considered at high clinical risk for P. carinii pneumonia) or active P. carinii pneumonia demonstrated a significant reduction in H2O2 release to P. carinii in vitro. Importantly, H2O2 production in response to PMA or HI serum-opsonized zymosan particles was preserved in AMs from all groups studied, suggesting a specific defect in the P. carinii–mediated response of AMs from HIV-infected persons rather than global impairment in the AM oxidative burst response. Recognizing the toxic effect of H2O2 on P. carinii, an impaired AM oxidative burst in HIV-infected persons may in part contribute to host susceptibility and accumulation of this opportunistic pathogen in the lungs (27). The oxidative response of human AMs to P. carinii challenge has not been investigated previously. P. carinii induced an oxidative response in rodent NR8383 macrophage cell line and rodent AMs (13), and in human monocytes and monocyte-derived macrophages (14). Previous observations in our laboratory demonstrated evidence for oxidative response of human AMs to P. carinii, although this


457

TABLE 2

Flow cytometry determination of AM oxidative burst response to PMA or opsonized (IgG) zymosan particles H2O2 Production (RFU) Group

Unstimulated

PMA

IgG-Zymosan

Healthy

0.4 * 0.1 0.4 * 0.1*

10.7 * 9.9 2.7 * 1.6*

11.4 * 7.3 2.7 * 1.7*

HIV" CD4 $ 200

0.5 * 0.1 0.5 * 0.1*

7.4 * 5.5 2.3 * 1.7*

8.8 * 3.3 2.6 * 1.9*

HIV" CD4 # 200

0.6 * 0.2 0.6 * 0.2*

16.6 * 20 5.0 * 2.0*

16.8 * 22 3.6 * 3.4*

*AMs pretreated with antimycin A.

Figure 5. Fluorescence microplate assay for time-dependent H2O2 production by AMs from (A) healthy individuals (n ! 9); and from asymptomatic HIV-infected subjects with (B) CD4" T-lymphocyte counts $ 200 cells/mm3 (n ! 4) and (C) CD4" T-lymphocyte counts # 200 cells/mm3 (n ! 3) in response to PMA (10)7 M; open triangles) and HI serum-opsonized zymosan particles (10:1 zymosan/AM ratio; solid triangles). Incubations were performed at 37&C in 5% CO2. Spontaneous H2O2 production by AMs is indicated by the open circles (n ! 4).

qualitative assessment was not fully characterized (16). The current study further defines the oxidative burst response of human AMs to P. carinii in vitro, and suggests an important role for the major surface glycoprotein P. carinii gp-A. The P. carinii–mediated oxidative burst response of AMs from HIV-infected persons has not been investigated previously. Prior investigations examining the oxidative burst response in monocytes or monocyte-derived macrophages from HIV-1–infected persons demonstrated either preserved (28) or impaired (29) function, whereas in vitro HIV-1 infection of healthy cells had no effect (30) in response to zymosan or PMA. Similarly, no significant differences in PMA- or HI serum-opsonized zymosan-mediated oxidative burst response were noted in the current study comparing AMs from healthy individuals with those from HIV-infected persons. However, the current study identifies a relatively specific impaired AM oxidative burst response to P. carinii. The observation that competitive inhibition of the AM mannose receptor significantly reduced the oxidative burst response to P. carinii suggests that the mannose receptor– mediated H2O2 production represents an important response to this opportunistic pathogen. Recognizing that P. carinii gp-A binds to the macrophage mannose receptor (31), the observations that P. carinii gp-A was sufficient to induce an oxidative burst response and that zymolyase treatment of P. carinii reduced the oxidative burst response support an important role for the interaction of the mannose-rich P. carinii gp-A with the AM mannose receptor. The precise mechanism accounting for the observed reduction in AM H2O2 production by AMs from HIVinfected persons was not established in the current study. The observation that immunoglobulin (Ig) receptor (FcR)– mediated (as determined by HI serum-opsonized zymosan) and PMA-mediated H2O2 production were preserved suggests that the nicotinamide adenine dinucleotide phosphate oxidase–mediated pathway is intact in AMs from asymptomatic HIV-infected individuals. The recent observation that macrophage mannose receptor–mediated P. carinii phagocytosis is impaired in AMs from HIV-infected persons at risk for P. carinii pneumonia (18) raises the possibility that altered mannose receptor–mediated oxidative burst response may account for the findings in the current study. As local HIV-1 replication occurs in the lungs of

458


asymptomatic patients or persons with active P. carinii pneumonia (32), the observed defect in oxidative burst may represent direct or indirect effects of HIV structural or regulatory proteins on signal transduction pathways, although measurements of HIV-1 infection are not available for these subjects. Alternatively, the observed defect in oxidative response may reflect P. carinii–associated factors. Studies to define the mechanism of HIV-associated defect in P. carinii–mediated AM oxidative burst response are currently the focus of active investigation. The oxidative response of AMs to P. carinii was less vigorous when compared with PMA or opsonized zymosan. Similar results were noted by other investigators (13), suggesting that undefined P. carinii factors mediate or modulate the cellular response after binding and/or ingestion. These factors may represent differences in the contributions of various life forms of P. carinii, and/or the effect of P. carinii surface-associated Ig complement, surfactant proteins (SPs), and other lung factors, although the cultured organisms used in these experiments have minimal levels of lung-associated molecules such as SP-A (20). The finding that P. carinii organisms generated detectable levels of H2O2 above background is consistent with that reported by other investigators (13), suggests a certain level of intrinsic oxidative metabolism or esterase activity, and supports the viability measurements for the organisms used. Other factors mediating macrophage oxidative response to P. carinii, such as IgG (13, 33), were not specifically investigated. Although fibronectin facilitates P. carinii binding to AMs, this interaction does not result in an oxidative response (34). The use of rat-derived P. carinii organisms in the current study was necessary because no reliable sources of human-derived P. carinii are available. Although the use of rat-derived organisms may limit the applicability of the results to human disease, the observation that P. carinii gp-A from human-derived organisms induced a similar oxidative burst response in part validates the use of these organisms. Inasmuch as P. carinii viability after exposure to AMs was not directly determined, the precise mechanisms of P. carinii killing by AMs remains to be determined, although prior investigations suggest that H2O2 and superoxide anion may be directly toxic to P. carinii (15). Differences in the oxidative burst response to P. carinii trophozoites, cysts, and other intermediate forms (13), and comparison of P. carinii binding versus phagocytosis was not determined. Finally, these in vitro findings may not represent functions in vivo. In summary, this study demonstrates that in vitro P. carinii challenge stimulates an oxidative burst response in AMs from healthy individuals, and this response may be mediated predominantly through the the interaction of the major surface glycoprotein molecule P. carinii gp-A with the AM mannose receptor. The oxidative response may represent an important mechanism for the elimination of P. carinii by AMs from healthy individuals. Importantly, AMs from asymptomatic HIV" individuals at high clinical risk for P. carinii pneumonia and HIV" individuals with active P. carinii pneumonia demonstrate reduced oxidative responses, whereas HIV" individuals with earlier stages of HIV infection demonstrate preserved oxidative burst response. Impairment of this critical microbicidal ef-

fector cell mechanism may in part contribute to host susceptibility to this opportunistic pathogen, and may contribute to the inability to effectively clear this organism in the setting of CD4" T-lymphocyte depletion associated with HIV infection. Acknowledgments: The authors give special thanks to Russell Morin, Cheryl Arena, Theresa Kelley, Nichole Pelletier, and Robert Garland for their dedicated and excellent technical assistance. These data were presented in part at the 1998 ATS/ALA International conference. This work was supported in part by NIH Public Health Service Grants HL43510 and HL63655 and a Massachusetts Thoracic Society/American Lung Association Research Grant, and by support from the Parker B. Francis Foundation to one author (H.K.).

References 1. Martin, W. J. 1993. Pathogenesis of Pneumocystis carinii pneumonia. Am. J. Respir. Cell Mol. Biol. 8:356–357. 2. Fels, A., and Z. Cohn. 1986. The alveolar macrophage. J. Appl. Physiol. 60: 353–369. 3. Vassallo, R., C. F. Thomas, Jr., Z. Vuk-Paklovic, and A. H. Limper. 1999. Alveolar macrophage interactions with Pneumocystis carinii. J. Lab. Clin. Med. 133:535–540. 4. Limper, A. H., J. S. Hoyte, and J. E. Standing. 1997. The role of alveolar macrophages in Pneumocystis carinii degradation and clearance from the lung. J. Clin. Invest. 99:2110–2117. 5. Barton, E., and W. Campbell. 1969. Pneumocystis carinii in lungs of rats treated with cortisone acetate. Am. J. Pathol. 54:209–236. 6. Shellito, J. E., V. V. Suzara, W. Blumenfeld, J. M. Beck, H. J. Steger, and T. H. Ermak. 1990. A new model of Pneumocystis carinii infection in mice selectively depleted of helper T lymphocytes. J. Clin. Invest. 85:1686–1693. 7. Harmsen, A., and M. Stankiewicz. 1990. Requirement for CD4" cells in resistance to Pneumocystis carinii pneumonia in mice. J. Exp. Med. 172:937– 945. 8. Harmsen, A. G., and W. Chen. 1992. Resolution of Pneumocystis carinii pneumonia in CD4" lymphocyte-depleted mice given aerosols of heattreated Escherichia coli. J. Exp. Med. 176:881–886. 9. Masur, H., and T. C. Jones. 1978. The interaction in vitro of Pneumocystis carinii with macrophages and L-cells. J. Exp. Med. 147:157–170. 10. von Behren, L. A., and E. L. Pesanti. 1978. Uptake and degradation of Pneumocystis carinii by macrophages in vitro. Am. Rev. Respir. Dis. 118:1051– 1059. 11. Yves, S., and H. Y. Reynolds. 1990. Macrophages and polymorphonuclear neutrophils in lung defense and injury. Am. Rev. Respir. Dis. 141:471–501. 12. Johnston, R. B., C. A. Godzik, and Z. A. Cohn. 1978. Increased superoxide anion production by immunologically activated and chemically elicited macrophages. J. Exp. Med. 148:115–127. 13. Hidalgo, H. A., R. J. Helmke, V. F. German, and J. A. Mangos. 1992. Pneumocystis carinii induces an oxidative burst in alveolar macrophages. Infect. Immun. 60:1–7. 14. Laursen, A. L., B. Moller, J. Rungby, C. M. Petersen, and P. L. Andersen. 1994. Pneumocystis carinii-induced activation of the respiratory burst in human monocytes and macrophages. Clin. Exp. Immunol. 98:196–202. 15. Pesanti, E. 1984. Pneumocystis carinii: oxygen uptake, antioxidant enzymes, and susceptibility to oxygen-mediated damage. Infect. Immun. 44:7–11. 16. Ezekowitz, R. A. B., D. J. Williams, H. Koziel, M. Y. K. Armstong, A. Warner, F. F. Richards, and R. M. Rose. 1991. Uptake of Pneumocystis carinii mediated by the macrophage mannose receptor. Nature 351:155–158. 17. Fauci, A. S., A. M. Macher, D. L. Longo, H. C. Lane, A. H. Rook, H. Masur, and E. P. Gelmann. 1984. NIH Conference. Acquired immunodeficiency syndrome: epidemiologic, clinical, immunologic, and therapeutic considerations. Ann. Intern. Med. 100:92–106. 18. Koziel, H., Q. Eichbaum, B. A. Kruskal, P. Pinkston, R. A. Rogers, M. Y. K. Armstrong, F. F. Richards, R. M. Rose, and R. A. B. Ezekowitz. 1998. Reduced binding and phagocytosis of Pneumocystis carinii by alveolar macrophages from persons infected with HIV-1 correlates with mannose receptor downregulation. J. Clin. Invest. 102:1332–1344. 19. Williams, D. J., J. A. Radding, A. Dell, K.-H. Khoo, M. E. Rogers, F. F. Richards, and M. Y. K. Armstrong. 1991. Glucan synthesis in Pneumocystis carinii. J. Protozool. 38:427–437. 20. Koziel, H., D. S. Phelps, J. A. Fishman, M. Y. K. Armstrong, F. F. Richards, and R. M. Rose. 1998. Surfactant protein-A reduces binding and phagocytosis of Pneumocystis carinii by human alveolar macrophages in vitro. Am. J. Respir. Cell Mol. Biol. 18:834–843. 21. Armstrong, M. Y. K., H. Koziel, R. M. Rose, C. Arena, and F. F. Richards. 1991. Indicators of Pneumocystis carinii viability in short-term culture. J. Protozool. 38:88S–90S. 22. DeStefano, J., M. Cushion, R. Sleight, and P. Walzer. 1990. Analysis of Pneumocystis carinii cyst wall: I. Evidence for an outer surface membrane. J. Protozool. 37:428–435. 23. Wan, C. P., E. Myung, and B. H. S. Lau. 1993. An automated micro-fluoro-


24.

25. 26.

27.

28.

metric assay for monitoring oxidative burst activity of phagocytes. J. Immunol. Methods 159:131–138. Bass, D. A., J. W. Parce, L. R. Dechatelet, P. Szejda, M. C. Seeds, and M. Thomas. 1983. Flow cytometric studies of oxidative product formation by neutrophils: a graded response to membrane stimulation. J. Immunol. 130:1910–1917. Johnston, R. 1981. Secretion of superoxide anion. In Methods for Studying Mononuclear Phagocytes. D. Adams, P. Edelson, and H. Koren, editors. Academic Press, New York. 489–497. Imrich, A., and L. Kobzik. 1998. Flow cytometric analysis of macrophage oxidative metabolism using DCFH. In Methods in Molecular Biology: Flow Cytometry Protocols. M. J. Jaroszeski and R. Heller, editors. Humana Press, Totowa, NJ. 97–108. Limper, A. H., K. P. Offord, T. F. Smith, and W. J. Martin, II. 1989. Pneumocystis carinii pneumonia: differences in lung parasite number and inflammation in patients with and without AIDS. Am. Rev. Respir. Dis. 140:1204–1209. Nielsen, H., A. Kharazmi, and V. Faber. 1986. Blood monocyte and neutrophil functions in the acquired immune deficiency syndrome. Scand. J. Immunol. 24:291–296.

459

29. Muller, F., H. Rollag, and S. S. Froland. 1990. Reduced oxidative burst responses in monocytes and monocyte-derived macrophages from HIV-infected subjects. Clin. Exp. Immunol. 82:10–15. 30. Dukes, C. S., T. J. Matthews, and J. B. Weinberg. 1993. Human immunodeficiency virus type-1 infection of human monocytes and macrophages does not alter their ability to generate oxidative burst. J. Infect. Dis. 168:459– 462. 31. O’Riordan, D. M., J. E. Standing, and A. H. Limper. 1995. Pneumocystis carinii glycoprotein A binds macrophage mannose receptors. Infect. Immun. 63:779–784. 32. Koziel, H., S. Kim, C. Reardon, X. Li, R. Garland, P. Pinkston, and H. Kornfeld. 1999. Enhanced in vivo HIV-1 replication in the lungs of HIVinfected persons with Pneumocystis carinii pneumonia. Am. J. Respir. Crit. Care Med. 160:2048–2050. 33. Laursen, A. L., B. Moller, J. Rungby, C. M. Petersen, and P. L. Andersen. 1994. Pneumocystis carinii-induced activation of the respiratory burst in human monocytes and macrophages. Clin. Exp. Immunol. 98:196–202. 34. Pottratz, S. T., and W. J. Martin. 1990. Mechanism of Pneumocystis carinii attachment to cultured rat alveolar macrophages. J. Clin. Invest. 86:1678– 1683.