Defective Candidacidal Activity of Alveolar Macrophages and Peripheral Blood Monocytes from Patients with Chronic Obstructive Pulmonary Disease1, 2
ANNA VECCHIARELLI, MAURIZIO DOTTORINI, MANUELA PULITI, TOMMASO TODISCO, ELiO CENCI, and FRANCESCO BISTONI
Introduction
Alveolar macrophages (AM) are involved in the local defense of the lung against foreign microorganisms or cancer cells; therefore, their functional activities are obviously crucial for the pulmonary tissue. Moreover, AM activity appears to be direct against some challengers and indirect through regulation exerted on other cell types such as polymorphonuclear cells (PMN) (1)and lymphokine-activated killer cells (LAK) (2). Most AM derive from bone marrow and circulate as peripheral blood monocytes (PBM) (3) which, subsequently, settle in the lung and differentiate. Recent evidence indicates that AM can proliferate in the lung in situ as well (4). A number of studies comparing AM with PBM have been carried out (5, 6). Surprisingly, these two closely related cell populations showed a very different behavior under some circumstances (5). However, relatively little is known about the functionality of AM in some pulmonary disorders because of the obvious problems in the recovery of these cells when the lungs are affected by an inflammatory process. Recently, it was shown that AM from patients with idiopathic pulmonary fibrosis display defective killing of Listeria monocytogenes (7). We have shown that AM and PBM are able to kill Candida albicans (8), a well-known opportunistic pathogen that is often the cause of serious infections in immunocompromised hosts (9, 10).Followingthis line of research, we focused our interest on the evaluation of anti-Candida activity of AM and PBM in a group of patients with chronic obstructive pulmonary disease (COPD). It has been reported that patients with COPD frequently show an impairment in some immunologic responses, including defective chemotaxis as well as phagocytic and microbicidal activity of circulating
SUMMARY We investigated the in vitro candidacidal activity of alveolar macrophages (AM) and peripheral blood monocytes (PBM) from normal subjects or from patients with chronic obstructive pulmonary disease (COPO)displaying defective skin test delayed-type hypersensitivity (OTH)reactivity to seven antigens including Candida albicans. The results showed that cells from patients with COPOwere significantly less effective than cells from control sUbjects in the killing of C. albi· cans. Toexplore whether the observed functional impairment could be reversed, interferon-gamma (IFN-y) was added to AM and PBM from patients with COPO, alone or In the presence of lipopolysaccharide (LPS) as a suboptimal stimulus. The cells were cultured for 24 h and then assayed for antiCandida activity. After IFN-y treatment, the fungicidal activity of cells from patients with COPOwas comparable to that of unstimulated AM or PBM from healthy donors. Treatment with IFN-y plus LPS resulted in a further enhancement in the killing of C. albicans. Togain more insight into the mechanisms involved in the modulation of killing, we evaluated the possible stimulating activity of IFN-y plus LPS treatment on the secretion of tumor necrosis factor (TNF) and interleukin-1 (IL-1), two cytokines produced by activated macrophages and capable of stimulating natural effectors. The results showed that IFN-y plus LPS can indeed stimulate TNF and IL-1 secretion by AM and PBM from patients with COPO. Therefore, a precise role can reasonably be ascribed to these soluble factors in the observed augmentation of candidacidal activity as ascertained by treatment with IFN-y AM REV RESPIR DIS 1991; 143:1049-1054 plus LPS.
monocytes and neutrophils (11, 12). In the present study, we selected patients with COPO showing hypoergy in a cutaneous delayed test with seven different antigens, including C. albicans. Our results show that decreased killing was exhibited by AM or PBM from patients with COPD with respect to that in healthy control subjects. A well-defined role in macrophage activation has been described for interferon-gamma (lFN-y) (13, 14). In particular, this cytokine is able to enhance the anti-Candida activity of peritoneal macrophages (13) and human AM (6, 8, 15) and potentiate killing of Salmonella by murine macro phages (16). Furthermore, IFN-y induces antitumoral activity and expression of Class I and II histocompatibility molecules on macrophages (17). Stimulation with IFN-y in vitro enhances anti-Candida activity of AM and PBM. The addition of a suboptimal concentration of LPS to IFN-y resulted in an even greater enhancement in the killing of C. albicans and primed the cells to secrete tumor necrosis factor-alpha (TNF-a) and
interleukin-l (IL-l). Thus, IFN-yappears to act on AM and PBM both as a direct stimulating factor and as an indirect mediator through the induction of TNFa and IL-l secretion. Methods Reagents and Media RPM I 1640medium and fetal bovine serum (FBS) were obtained from Eurobio Laboratories (Paris, France). Triton X 100, Phytohemagglutinin (PHA), and Concanavalin A (Con A) were provided by Sigma Chemical Co. (St. Louis, MO). Human recombinant
(Received in original form November 7, 1989 and in revised form December 11, 1990) 1 From the Department of Experimental Medicine and Biochemical Sciences, Microbiology Section, and the Pulmonary Unit, University of Perugia, Perugia, Italy. 2 Correspondence and requests for reprints should be addressed to Dr. Anna Vecchiarelli, Department of Experimental Medicine and Biochemical Sciences, Microbiology Section, University of Perugia, Via del Giochetto, 06100Perugia, Italy.
1049
1050
VECCHIARELLI, DOTTORINI, PULlTI, TODISCO, CENCI, AND BISTONI
TABLE 1 CHARACTERISTICS OF THE STUDY POPULATION* Patients with COPD (n = 11) Age, yr Sex, M/F Smoking status Current smokers < 10 cig/day Ex-smokers Ex-smokers cigarette, pack-years Interval since smoking cessation, yr
Control Subjects (n = 13)
65.8 ± 5.9
67.3 ± 5.6 9M2F
9M4F
4 7
13
40.5 ± 6.2 3.2 ± 1.4
Lung function VC, % pred. FEV1NC, % RVITLC, % Pa02, mm Hg Paco 2 , mm Hg
68 41 56 61 47
Circulating lymphocyte subsets CD4, % CD8, %
45 ± 10 27 ± 8
50 ± 9 25 ± 6
56,216 ± 10,200t
84,840 ± 13,800
3.9 ± 2.6t 0.8 ± 0.5t
17.2 ± 6.5 4.1 ± 2.3
Proliferative response to mitogens PHA, cpm Cutaneous reactivity Multitest score, mm C. albicans score, mm
± ± ± ± ±
2 9
101 ± 7 97 ± 6
1 7
29 ± 2
5
• Results are expressed as mean values ± SO. with COPO versus control subjects using Student's t test.
t p < 0.001, patients
interferon-y, human recombinant interleukinln, and rabbit antihuman IL-l were provided by Janssen Biochimica (Beerse, Belgium). TNF-a and rabbit antihuman TNF-a monoclonal antibodies were purchased from Genzyme Corp. (Boston, MA). Anti-OKT3, antiOKT4, and anti-OKT8 monoclonal antibodies were obtained from Ortho Diagnostics (Raritan, NJ). Endotoxin lipopolysaccharide from Escherichia coli 055:135 was obtained from Difco Laboratories (Detroit, MI). PH]Thymidine (sp. act. 88 Ci/mmol) was purchased from Amersham International (Buckinghamshire, UK).
Study Population Our study population included 11 patients with COPD (nine men and two women). An established diagnosis of COPD (Stage III according to the Medical Research Council) (18) was required for admission to the study. The smoking status, lung function, and immunologic parameters are reported in table 1. The patients with COPD had depressed delayedtype hypersensitivity (DTH).to a multipuncture seven-antigen test (tetanus and diphtheria toxoids, streptococcal, tuberculin, C albicans, Trichophyton, Proteus antigens, and as a negative control 70070 glycerine-saline diluent obtained from the Institute Merieux, Lyon, France). Eligible patients showed no evidence of present viral or bacterial infections or other conditions that might affect host defenses, and they had received no medication possibly influencing immune reactivity during the 3 months prior to the study. In particular, no antibiotics were being used, and systemic or inhaled steroids or theophillines were not being administered. Thirteen healthy
volunteers, all smokers (nine men and four women) in the same age range, served as control subjects; their characteristics are reported in table 1.
Preparation and Purification of Human AM AM werecollected as described elsewhere(19). Briefly, after laryngeal anesthesia with 2070 lidocaine, bronchoalveolar lavage (BAL) was performed through a fiberoptic bronchoscope (Model BF-type 10;Olympus Co., Tokyo, Japan). Three 50-ml aliquots of sterile saline (0.9070 NaCI) warmed to 37 0 C were infused into one of the segments of the right or left lobe and then removed by gentle suction. Lavage fluid specimens were filtered through coarse gauze, and cells were separated by centrifugation at 500 g for 8 min at 4 0 C. The number of cells harvested was approximately 10to 15 x 106 • AM, evaluated as cells positive on nonspecific esterase staining, were > 90070 in healthy smokers and 70 to 80070 in patients with COPD (15 to 20070 being polymorphonuclear cells). Thereafter, these cells were plated in cell culture petri dishes (Nunc Inter Med, Roskilde, Denmark) at a concentration of2 to 3 x 106/ml in RPMI 1640 plus 5070 FBS (hereafter referred to as cRPMI) and incubated at 37 0 C in 5070 CO 2 for 1 h. The nonadherent cells were removed by washing the dishes three to fivetimes with warm RPMI 1640 medium; adherent cells were carefully removed using a rubber policeman. These latter cells were 95 to 98070 esterase-positive, and cell viability was assessed by the trypan blue dye exclusion test (more than 98070 viable). The harvested cells were plated into 96-well, flat-bottom tissue culture microtiter plates
(Falcon; Becton Dickinson, Oxnard, CAl at lOS cells/well stimulated or not with IFN-y for 24 h and then tested for anti-Candida activity. The cultured cells (mean viability, 93 to 95070) were then handled as freshly harvested AM in order to test their candidacidal activity.
Preparation of PBM Heparinized venous blood, obtained from the donors at the time of BAL, was diluted with cRPMI. Then, mononuclear cells were separated by Ficoll-Hypaque density gradient centrifugation (20). Cells were washed twice in cRPMI and plated into cellculture petri dishes (Nunc) and incubated for 1 h at a concentration of 2 to 3 X 106 ml. The adherent cells were recoveredand treated as described above. Determination of Lymphocyte Subsets Peripheral blood leukocytes were mixed with anti-OKT4 or OKT8 monoclonal antibodies. After a 30-min incubation at 4°C the cells were washed with phosphate-buffered saline and then mixed with fluorescein-conjugated goat antihuman IgG. The cells were then washed three times with phosphate-buffered saline and examined by fluorescence microscopy. At least 100 cells were counted. Proliferation Assay Mitogen-induced blastogenesis was obtained by culturing 1 x lOS cells/OJ ml of cRPMI in flat-bottom 96-well plates in the presence of 5 ug/ml of PHA. Cultures werepulsed with pH]thymidine, and the isotope incorporation was determined by (3-scintillation counting. Candida albicans C albicans (PCA-2) was kindly supplied by D. Kerridge (Department of Biochemistry,
University of Cambridge, Cambridge, UK). This is an agerminative strain and grows as a pure yeast form in vitro at 28 0 C or at 37° C in conventional mycological media. Under conditions that promote germ-tube formation in vitro (37 0 C), this strain grows as a yeast and short pseudomycelium. The yeast was grown at 28 0 C with slight agitation in low-glucose Winge medium composed of 0.2070 (wt/vol) glucose and 0.3070 (wt/vol) yeast extract (BBL Microbiology Systems, Cockeysville, MD) until a stationary phase was reached (approximately 3 x 108 yeasts/ml) and the cells had the morphologic features described elsewhere (21). Cells were harvested by low-speed centrifugation (100 x g), washed twice in saline, and diluted to the desired concentration.
Colony-forming Units Inhibition Assay Plate counts. AM or PBM (1 x lOS) in 0.1 ml of suspension per well were incubated in flat-bottom, 96-well microtiter tissue culture plates (Falcon) with C albicans (2 x 104 ) in 0.1 ml of RPMI. The cells mixed with C albicans were incubated for 6 h. After incubation at 37° C under 5070 CO 2 , the plates were vigorously shaken, monolayers were lysed by adding Triton X 100(0.1070 in distilled water),
1051
DEFECTIVE CANDIDACIDAL ACTIVITY OF AM OR PBM FROM PATIENTS WITH COPD
TABLE 2 CANOIOACIOAL ACTIVITY OF AM OR PBM FROM A PATIENT WITH COPO OR A CONTROL SUBJECT" Colony-forming Units ± SO x 10 (n)
Colony-forming Units Growth Inhibition (%)
None AM from control subject AM from patient with COPO
20.2 ± 0.15 11.4 ± 0.61 14.6 ± 0.52t
43.5 27.7
PBM from control subject PBM from patient with COPO
11.2 ± 0.42 14.1 ± 0.54t
44.5 30.1
Effector Cells
Definition of abbreviations: AM = alveolar macrophages; PBM = peripheral blood monocytes. • C. alb/cans cells (2 x 104 ) were incubated in medium alone or with 10' AM or PBM. After a 6-h
incubation effector cells were lysed by adding Triton x 100, and the samples (triplicate cultures) were plated on Sabouraud dextrose agar after dilution in distilled water. t p < 0.01 (patient with COPD versus control subject) according to the Student's t test.
and serial dilutions were prepared in distilled water from each well. Plates (triplicate samples) were made by spreading each sample on Sabouraud dextrose agar, and colony-forming units (cfu) were visually evaluated after 24 to 48 h of incubation at 37° C. Control cultures consisted of C albicans incubated without effector cells. All reagents used in this study did not directly affect microbial growth. 90 80 70 60
:::: :z 0 ~
50
Data Analysis Differences between the mean values of pooled determinations were analyzed using Student's t test. Comparison of candidacidal activity between pairs of unstimulated and stimulated cells were performed using Wilcoxon's matched-pairs signed-rank test. Differences in IL-l or TNF activity in culture supernatants were evaluated according to the Mann-Whitney V test. Data on candidacidal activity of AM or PBM stimulated with different doses of IFN-y were analyzed using ANOVA.
~
40
~
u
30 20 10
NORMAL STIMULUS -
90 80 70 60 ~ :z
co ~
Production of AM or PBM Culture Supernatants and Assaying of IL-l or TNF Activity Supernatants from AM or PBM cultures were obtained after incubation of AM for 24 h, with or without stimuli, according to a previously described method (22). IL-l activity was measured on freshly isolated thymocytes from C3H/HeJ male mice, 4 to 8 wk of age, obtained from the Jackson Laboratory (Bar Harbor, ME), incubated with 1 ug/rnl Con A (23). IL-l titers were calculated by comparing the dilutions of the test supernatants that gave 50070 maximal stimulation with a standard recombinant human IL-l preparation in a tritiated thymidine uptake test. TNF cytotoxic activity in supernatants was determined as described (24), using actinomycinD-treated L-929 cells as targets. The determination of lymphokine activity was made in comparison with commercially available preparations with known titers, and the results were expressed as Ll/ml.
periment in which the ability of AM or PBM to kill Candida cells in vitro is shown in table 2, as evaluated by the number of cfu recovered after 6 h of incubation. The data reported in figure 1 show that the killing rate of AM from patients with COPO was significantly lower than that in normal control subjects. Our previous data (8) showed that candidacidal activity of AM from normal subjects can be enhanced by IFN-y treatment. In an attempt to restore the candidacidal activity of cells from patients with COPD, we added IFN-y (300 U) for 24 h before the in vitro assay. This treatment enhanced the killing activity of AM from patients with COPO. This enhancement reached a level comparable to that of unstimulated AM from normal control subjects (figure 1). We also performed experiments to determine whether PBM from patients with capo wereimpaired in C. albicans killing ability with respect to PBM from normal control subjects. The results in figure 2
COPD
+
0
E co
50
~
:z
...
40
~
u
30 20 10
Results
+
Fig. 1. Candidacidal activity of AM from patients with COPO or from normal donors. Each point represents the percentage of cfu inhibition at an effector to target cell ratio of 5:1. AM from normal donors unstimulated (open circles) or stimulated with IFN-y (300 Ulml) (closed circles). AM from patients with COPO unstimulated (open triangles) or stimulated with IFN-y (300 Ulml) (closed triangles). Horizontal bars represent the median values (p < 0.Q1 normal AM versus COPO AM; p < 0.Q1 normal stimulated AM versus unstimulated; p < 0.05 COPO stimulated AM versus unstimulated; p = NS normal AM versus stimulated COPO AM according to Wilcoxon's matched-pairs signed-rank test).
Candidacidal Activity of AM and PBM from Patients with COPD and from Healthy Donors
Previously,we haveshown that AM from normal subjects are able to phagocytize and kill different microorganisms, including C albicans (8). In order to assess the ability of AM or PBM from patients with COPO to kill C albicans, we performed experiments comparing their candidacidal activity with that of normal AM or PBM. A representative ex-
NORMAL STIMULUS -
COPD +
+
Fig. 2. Candidacidal activity of PBM from patients with COPO or from normal donors. Each point represents the percentage of cfu inhibition at an effector to target ratio of 5:1.PBM from normal donors unstimulated (open circles) or stimulated with IFN-y (300 U/ml) (closed circles). PBM from patients witih COPO unstimulated (open triangles) or stimulated with IFN-y (300 U/ml) (closed triangles). Horizontal bars represent the median values (p < 0.001 normal PBM versus COPO PBM; P < 0.001 normal stimulated PBM versus unstimulated; p = NS normal PBM versus stimulated COPO PBM, according to Wilcoxon's matched-pairs signed-rank test).
1052
VECCHIARELLI, DOTTORINI, PULlTI, TODISCO, CENCI, AND BISTONI
TABLE 3 EFFECT OF DIFFERENT DOSES OF IFN-y IN VITRO ON CANDIDACIDAL ACTIVITY OF AM AND PBM FROM PATIENTS WITH COPD Effector Cells AM
PBM
Subjects (n)
8 8 8 8 8 8 8 8
Paired t Test
IFN-y in vitro (Ulm/)
Candidacidal Activity *
None
21.62 20.75 29.50 36.00
± ± ± ±
6.32 5.59 5.21 7.11
0.49 3.23t 7.84+
27.56 27.50 35.62 42.45
± ± ± ±
5.21 5.23 5.82 5.31
0.20 3.15t 6.42+
3 30 300 None
3 30 300
For definition of abbreviations, see table 2. • Candidacidal activity is expressed as percentage of colony-forming units inhibition ± SD at an effector to target cell ratio of 5: 1. The means are significantly different according to ANaVA: F = 11.11 for AM (p < 0.001) and F = 13.87 for PBM (p < 0.01). t p < 0.05 (IFN-y = stimulated cells versus unstimulated). :j: p < 0.Q1 (IFN-y-stimulated cells versus unstimulated cells).
show that anti-Candida activity of PBM from patients with COPO was significantly lowerthan that of control subjects. IFN-y treatment resulted in increased killing potential of PBM from patients with COPO, thus bringing the killing activity to levels similar to those observed in unstimulated PBM from normal control subjects.
Effect of the Addition of Different Doses of IFN-y on Candidacidal Activity of AM and PBM To better define the effect of IFN-y treatment on candidacidal activity of AM and PBM from patients with COPO, differr-P
ANNA VECCHIARELLI, MAURIZIO DOTTORINI, MANUELA PULITI, TOMMASO TODISCO, ELiO CENCI, and FRANCESCO BISTONI
Introduction
Alveolar macrophages (AM) are involved in the local defense of the lung against foreign microorganisms or cancer cells; therefore, their functional activities are obviously crucial for the pulmonary tissue. Moreover, AM activity appears to be direct against some challengers and indirect through regulation exerted on other cell types such as polymorphonuclear cells (PMN) (1)and lymphokine-activated killer cells (LAK) (2). Most AM derive from bone marrow and circulate as peripheral blood monocytes (PBM) (3) which, subsequently, settle in the lung and differentiate. Recent evidence indicates that AM can proliferate in the lung in situ as well (4). A number of studies comparing AM with PBM have been carried out (5, 6). Surprisingly, these two closely related cell populations showed a very different behavior under some circumstances (5). However, relatively little is known about the functionality of AM in some pulmonary disorders because of the obvious problems in the recovery of these cells when the lungs are affected by an inflammatory process. Recently, it was shown that AM from patients with idiopathic pulmonary fibrosis display defective killing of Listeria monocytogenes (7). We have shown that AM and PBM are able to kill Candida albicans (8), a well-known opportunistic pathogen that is often the cause of serious infections in immunocompromised hosts (9, 10).Followingthis line of research, we focused our interest on the evaluation of anti-Candida activity of AM and PBM in a group of patients with chronic obstructive pulmonary disease (COPD). It has been reported that patients with COPD frequently show an impairment in some immunologic responses, including defective chemotaxis as well as phagocytic and microbicidal activity of circulating
SUMMARY We investigated the in vitro candidacidal activity of alveolar macrophages (AM) and peripheral blood monocytes (PBM) from normal subjects or from patients with chronic obstructive pulmonary disease (COPO)displaying defective skin test delayed-type hypersensitivity (OTH)reactivity to seven antigens including Candida albicans. The results showed that cells from patients with COPOwere significantly less effective than cells from control sUbjects in the killing of C. albi· cans. Toexplore whether the observed functional impairment could be reversed, interferon-gamma (IFN-y) was added to AM and PBM from patients with COPO, alone or In the presence of lipopolysaccharide (LPS) as a suboptimal stimulus. The cells were cultured for 24 h and then assayed for antiCandida activity. After IFN-y treatment, the fungicidal activity of cells from patients with COPOwas comparable to that of unstimulated AM or PBM from healthy donors. Treatment with IFN-y plus LPS resulted in a further enhancement in the killing of C. albicans. Togain more insight into the mechanisms involved in the modulation of killing, we evaluated the possible stimulating activity of IFN-y plus LPS treatment on the secretion of tumor necrosis factor (TNF) and interleukin-1 (IL-1), two cytokines produced by activated macrophages and capable of stimulating natural effectors. The results showed that IFN-y plus LPS can indeed stimulate TNF and IL-1 secretion by AM and PBM from patients with COPO. Therefore, a precise role can reasonably be ascribed to these soluble factors in the observed augmentation of candidacidal activity as ascertained by treatment with IFN-y AM REV RESPIR DIS 1991; 143:1049-1054 plus LPS.
monocytes and neutrophils (11, 12). In the present study, we selected patients with COPO showing hypoergy in a cutaneous delayed test with seven different antigens, including C. albicans. Our results show that decreased killing was exhibited by AM or PBM from patients with COPD with respect to that in healthy control subjects. A well-defined role in macrophage activation has been described for interferon-gamma (lFN-y) (13, 14). In particular, this cytokine is able to enhance the anti-Candida activity of peritoneal macrophages (13) and human AM (6, 8, 15) and potentiate killing of Salmonella by murine macro phages (16). Furthermore, IFN-y induces antitumoral activity and expression of Class I and II histocompatibility molecules on macrophages (17). Stimulation with IFN-y in vitro enhances anti-Candida activity of AM and PBM. The addition of a suboptimal concentration of LPS to IFN-y resulted in an even greater enhancement in the killing of C. albicans and primed the cells to secrete tumor necrosis factor-alpha (TNF-a) and
interleukin-l (IL-l). Thus, IFN-yappears to act on AM and PBM both as a direct stimulating factor and as an indirect mediator through the induction of TNFa and IL-l secretion. Methods Reagents and Media RPM I 1640medium and fetal bovine serum (FBS) were obtained from Eurobio Laboratories (Paris, France). Triton X 100, Phytohemagglutinin (PHA), and Concanavalin A (Con A) were provided by Sigma Chemical Co. (St. Louis, MO). Human recombinant
(Received in original form November 7, 1989 and in revised form December 11, 1990) 1 From the Department of Experimental Medicine and Biochemical Sciences, Microbiology Section, and the Pulmonary Unit, University of Perugia, Perugia, Italy. 2 Correspondence and requests for reprints should be addressed to Dr. Anna Vecchiarelli, Department of Experimental Medicine and Biochemical Sciences, Microbiology Section, University of Perugia, Via del Giochetto, 06100Perugia, Italy.
1049
1050
VECCHIARELLI, DOTTORINI, PULlTI, TODISCO, CENCI, AND BISTONI
TABLE 1 CHARACTERISTICS OF THE STUDY POPULATION* Patients with COPD (n = 11) Age, yr Sex, M/F Smoking status Current smokers < 10 cig/day Ex-smokers Ex-smokers cigarette, pack-years Interval since smoking cessation, yr
Control Subjects (n = 13)
65.8 ± 5.9
67.3 ± 5.6 9M2F
9M4F
4 7
13
40.5 ± 6.2 3.2 ± 1.4
Lung function VC, % pred. FEV1NC, % RVITLC, % Pa02, mm Hg Paco 2 , mm Hg
68 41 56 61 47
Circulating lymphocyte subsets CD4, % CD8, %
45 ± 10 27 ± 8
50 ± 9 25 ± 6
56,216 ± 10,200t
84,840 ± 13,800
3.9 ± 2.6t 0.8 ± 0.5t
17.2 ± 6.5 4.1 ± 2.3
Proliferative response to mitogens PHA, cpm Cutaneous reactivity Multitest score, mm C. albicans score, mm
± ± ± ± ±
2 9
101 ± 7 97 ± 6
1 7
29 ± 2
5
• Results are expressed as mean values ± SO. with COPO versus control subjects using Student's t test.
t p < 0.001, patients
interferon-y, human recombinant interleukinln, and rabbit antihuman IL-l were provided by Janssen Biochimica (Beerse, Belgium). TNF-a and rabbit antihuman TNF-a monoclonal antibodies were purchased from Genzyme Corp. (Boston, MA). Anti-OKT3, antiOKT4, and anti-OKT8 monoclonal antibodies were obtained from Ortho Diagnostics (Raritan, NJ). Endotoxin lipopolysaccharide from Escherichia coli 055:135 was obtained from Difco Laboratories (Detroit, MI). PH]Thymidine (sp. act. 88 Ci/mmol) was purchased from Amersham International (Buckinghamshire, UK).
Study Population Our study population included 11 patients with COPD (nine men and two women). An established diagnosis of COPD (Stage III according to the Medical Research Council) (18) was required for admission to the study. The smoking status, lung function, and immunologic parameters are reported in table 1. The patients with COPD had depressed delayedtype hypersensitivity (DTH).to a multipuncture seven-antigen test (tetanus and diphtheria toxoids, streptococcal, tuberculin, C albicans, Trichophyton, Proteus antigens, and as a negative control 70070 glycerine-saline diluent obtained from the Institute Merieux, Lyon, France). Eligible patients showed no evidence of present viral or bacterial infections or other conditions that might affect host defenses, and they had received no medication possibly influencing immune reactivity during the 3 months prior to the study. In particular, no antibiotics were being used, and systemic or inhaled steroids or theophillines were not being administered. Thirteen healthy
volunteers, all smokers (nine men and four women) in the same age range, served as control subjects; their characteristics are reported in table 1.
Preparation and Purification of Human AM AM werecollected as described elsewhere(19). Briefly, after laryngeal anesthesia with 2070 lidocaine, bronchoalveolar lavage (BAL) was performed through a fiberoptic bronchoscope (Model BF-type 10;Olympus Co., Tokyo, Japan). Three 50-ml aliquots of sterile saline (0.9070 NaCI) warmed to 37 0 C were infused into one of the segments of the right or left lobe and then removed by gentle suction. Lavage fluid specimens were filtered through coarse gauze, and cells were separated by centrifugation at 500 g for 8 min at 4 0 C. The number of cells harvested was approximately 10to 15 x 106 • AM, evaluated as cells positive on nonspecific esterase staining, were > 90070 in healthy smokers and 70 to 80070 in patients with COPD (15 to 20070 being polymorphonuclear cells). Thereafter, these cells were plated in cell culture petri dishes (Nunc Inter Med, Roskilde, Denmark) at a concentration of2 to 3 x 106/ml in RPMI 1640 plus 5070 FBS (hereafter referred to as cRPMI) and incubated at 37 0 C in 5070 CO 2 for 1 h. The nonadherent cells were removed by washing the dishes three to fivetimes with warm RPMI 1640 medium; adherent cells were carefully removed using a rubber policeman. These latter cells were 95 to 98070 esterase-positive, and cell viability was assessed by the trypan blue dye exclusion test (more than 98070 viable). The harvested cells were plated into 96-well, flat-bottom tissue culture microtiter plates
(Falcon; Becton Dickinson, Oxnard, CAl at lOS cells/well stimulated or not with IFN-y for 24 h and then tested for anti-Candida activity. The cultured cells (mean viability, 93 to 95070) were then handled as freshly harvested AM in order to test their candidacidal activity.
Preparation of PBM Heparinized venous blood, obtained from the donors at the time of BAL, was diluted with cRPMI. Then, mononuclear cells were separated by Ficoll-Hypaque density gradient centrifugation (20). Cells were washed twice in cRPMI and plated into cellculture petri dishes (Nunc) and incubated for 1 h at a concentration of 2 to 3 X 106 ml. The adherent cells were recoveredand treated as described above. Determination of Lymphocyte Subsets Peripheral blood leukocytes were mixed with anti-OKT4 or OKT8 monoclonal antibodies. After a 30-min incubation at 4°C the cells were washed with phosphate-buffered saline and then mixed with fluorescein-conjugated goat antihuman IgG. The cells were then washed three times with phosphate-buffered saline and examined by fluorescence microscopy. At least 100 cells were counted. Proliferation Assay Mitogen-induced blastogenesis was obtained by culturing 1 x lOS cells/OJ ml of cRPMI in flat-bottom 96-well plates in the presence of 5 ug/ml of PHA. Cultures werepulsed with pH]thymidine, and the isotope incorporation was determined by (3-scintillation counting. Candida albicans C albicans (PCA-2) was kindly supplied by D. Kerridge (Department of Biochemistry,
University of Cambridge, Cambridge, UK). This is an agerminative strain and grows as a pure yeast form in vitro at 28 0 C or at 37° C in conventional mycological media. Under conditions that promote germ-tube formation in vitro (37 0 C), this strain grows as a yeast and short pseudomycelium. The yeast was grown at 28 0 C with slight agitation in low-glucose Winge medium composed of 0.2070 (wt/vol) glucose and 0.3070 (wt/vol) yeast extract (BBL Microbiology Systems, Cockeysville, MD) until a stationary phase was reached (approximately 3 x 108 yeasts/ml) and the cells had the morphologic features described elsewhere (21). Cells were harvested by low-speed centrifugation (100 x g), washed twice in saline, and diluted to the desired concentration.
Colony-forming Units Inhibition Assay Plate counts. AM or PBM (1 x lOS) in 0.1 ml of suspension per well were incubated in flat-bottom, 96-well microtiter tissue culture plates (Falcon) with C albicans (2 x 104 ) in 0.1 ml of RPMI. The cells mixed with C albicans were incubated for 6 h. After incubation at 37° C under 5070 CO 2 , the plates were vigorously shaken, monolayers were lysed by adding Triton X 100(0.1070 in distilled water),
1051
DEFECTIVE CANDIDACIDAL ACTIVITY OF AM OR PBM FROM PATIENTS WITH COPD
TABLE 2 CANOIOACIOAL ACTIVITY OF AM OR PBM FROM A PATIENT WITH COPO OR A CONTROL SUBJECT" Colony-forming Units ± SO x 10 (n)
Colony-forming Units Growth Inhibition (%)
None AM from control subject AM from patient with COPO
20.2 ± 0.15 11.4 ± 0.61 14.6 ± 0.52t
43.5 27.7
PBM from control subject PBM from patient with COPO
11.2 ± 0.42 14.1 ± 0.54t
44.5 30.1
Effector Cells
Definition of abbreviations: AM = alveolar macrophages; PBM = peripheral blood monocytes. • C. alb/cans cells (2 x 104 ) were incubated in medium alone or with 10' AM or PBM. After a 6-h
incubation effector cells were lysed by adding Triton x 100, and the samples (triplicate cultures) were plated on Sabouraud dextrose agar after dilution in distilled water. t p < 0.01 (patient with COPD versus control subject) according to the Student's t test.
and serial dilutions were prepared in distilled water from each well. Plates (triplicate samples) were made by spreading each sample on Sabouraud dextrose agar, and colony-forming units (cfu) were visually evaluated after 24 to 48 h of incubation at 37° C. Control cultures consisted of C albicans incubated without effector cells. All reagents used in this study did not directly affect microbial growth. 90 80 70 60
:::: :z 0 ~
50
Data Analysis Differences between the mean values of pooled determinations were analyzed using Student's t test. Comparison of candidacidal activity between pairs of unstimulated and stimulated cells were performed using Wilcoxon's matched-pairs signed-rank test. Differences in IL-l or TNF activity in culture supernatants were evaluated according to the Mann-Whitney V test. Data on candidacidal activity of AM or PBM stimulated with different doses of IFN-y were analyzed using ANOVA.
~
40
~
u
30 20 10
NORMAL STIMULUS -
90 80 70 60 ~ :z
co ~
Production of AM or PBM Culture Supernatants and Assaying of IL-l or TNF Activity Supernatants from AM or PBM cultures were obtained after incubation of AM for 24 h, with or without stimuli, according to a previously described method (22). IL-l activity was measured on freshly isolated thymocytes from C3H/HeJ male mice, 4 to 8 wk of age, obtained from the Jackson Laboratory (Bar Harbor, ME), incubated with 1 ug/rnl Con A (23). IL-l titers were calculated by comparing the dilutions of the test supernatants that gave 50070 maximal stimulation with a standard recombinant human IL-l preparation in a tritiated thymidine uptake test. TNF cytotoxic activity in supernatants was determined as described (24), using actinomycinD-treated L-929 cells as targets. The determination of lymphokine activity was made in comparison with commercially available preparations with known titers, and the results were expressed as Ll/ml.
periment in which the ability of AM or PBM to kill Candida cells in vitro is shown in table 2, as evaluated by the number of cfu recovered after 6 h of incubation. The data reported in figure 1 show that the killing rate of AM from patients with COPO was significantly lower than that in normal control subjects. Our previous data (8) showed that candidacidal activity of AM from normal subjects can be enhanced by IFN-y treatment. In an attempt to restore the candidacidal activity of cells from patients with COPD, we added IFN-y (300 U) for 24 h before the in vitro assay. This treatment enhanced the killing activity of AM from patients with COPO. This enhancement reached a level comparable to that of unstimulated AM from normal control subjects (figure 1). We also performed experiments to determine whether PBM from patients with capo wereimpaired in C. albicans killing ability with respect to PBM from normal control subjects. The results in figure 2
COPD
+
0
E co
50
~
:z
...
40
~
u
30 20 10
Results
+
Fig. 1. Candidacidal activity of AM from patients with COPO or from normal donors. Each point represents the percentage of cfu inhibition at an effector to target cell ratio of 5:1. AM from normal donors unstimulated (open circles) or stimulated with IFN-y (300 Ulml) (closed circles). AM from patients with COPO unstimulated (open triangles) or stimulated with IFN-y (300 Ulml) (closed triangles). Horizontal bars represent the median values (p < 0.Q1 normal AM versus COPO AM; p < 0.Q1 normal stimulated AM versus unstimulated; p < 0.05 COPO stimulated AM versus unstimulated; p = NS normal AM versus stimulated COPO AM according to Wilcoxon's matched-pairs signed-rank test).
Candidacidal Activity of AM and PBM from Patients with COPD and from Healthy Donors
Previously,we haveshown that AM from normal subjects are able to phagocytize and kill different microorganisms, including C albicans (8). In order to assess the ability of AM or PBM from patients with COPO to kill C albicans, we performed experiments comparing their candidacidal activity with that of normal AM or PBM. A representative ex-
NORMAL STIMULUS -
COPD +
+
Fig. 2. Candidacidal activity of PBM from patients with COPO or from normal donors. Each point represents the percentage of cfu inhibition at an effector to target ratio of 5:1.PBM from normal donors unstimulated (open circles) or stimulated with IFN-y (300 U/ml) (closed circles). PBM from patients witih COPO unstimulated (open triangles) or stimulated with IFN-y (300 U/ml) (closed triangles). Horizontal bars represent the median values (p < 0.001 normal PBM versus COPO PBM; P < 0.001 normal stimulated PBM versus unstimulated; p = NS normal PBM versus stimulated COPO PBM, according to Wilcoxon's matched-pairs signed-rank test).
1052
VECCHIARELLI, DOTTORINI, PULlTI, TODISCO, CENCI, AND BISTONI
TABLE 3 EFFECT OF DIFFERENT DOSES OF IFN-y IN VITRO ON CANDIDACIDAL ACTIVITY OF AM AND PBM FROM PATIENTS WITH COPD Effector Cells AM
PBM
Subjects (n)
8 8 8 8 8 8 8 8
Paired t Test
IFN-y in vitro (Ulm/)
Candidacidal Activity *
None
21.62 20.75 29.50 36.00
± ± ± ±
6.32 5.59 5.21 7.11
0.49 3.23t 7.84+
27.56 27.50 35.62 42.45
± ± ± ±
5.21 5.23 5.82 5.31
0.20 3.15t 6.42+
3 30 300 None
3 30 300
For definition of abbreviations, see table 2. • Candidacidal activity is expressed as percentage of colony-forming units inhibition ± SD at an effector to target cell ratio of 5: 1. The means are significantly different according to ANaVA: F = 11.11 for AM (p < 0.001) and F = 13.87 for PBM (p < 0.01). t p < 0.05 (IFN-y = stimulated cells versus unstimulated). :j: p < 0.Q1 (IFN-y-stimulated cells versus unstimulated cells).
show that anti-Candida activity of PBM from patients with COPO was significantly lowerthan that of control subjects. IFN-y treatment resulted in increased killing potential of PBM from patients with COPO, thus bringing the killing activity to levels similar to those observed in unstimulated PBM from normal control subjects.
Effect of the Addition of Different Doses of IFN-y on Candidacidal Activity of AM and PBM To better define the effect of IFN-y treatment on candidacidal activity of AM and PBM from patients with COPO, differr-P
