HIV-1 Envelope Protein (gp120) Inhibits the Activity of Human Bronchoalveolar Macrophages against Cryptococcus neototmens:»
RANDALL P. WAGNER, STUART M. LEVITZ,3 ABDULMONEIM TABUNI, and HARDY KORNFELD4 Introduction
Cryptococcus neoformans infections are a major cause of morbidity and mortality for HIV-infected persons (1). Environmental exposure to Cryptococcus organisms is common, with the usual site of inoculation being the lower respiratory tract through the inhalation of aerosolized infectious material (2). Despite colonization of the tracheobronchial tree, dissemination outside the lung in the normal host is rare (3, 4). Several mechanisms contribute to the containment of the fungus to the lung, including the spontaneous fungistatic activity of bronchoalveolar macrophages. When cryptococcosis does develop, control of the disease usually requires the action of cell-mediated immunity. The innocuous behavior of this yeast in the normal host is in sharp contrast to the severity of the disease in patients with AIDS. Between 5 and 10070 of all patients with AIDS in the United States present with cryptococcal meningitis, and the case mortality rate remains high despite improvements in therapy (1). Although it is clear that the pulmonary defenses are altered in AIDS, the basis for this loss of protection against Cryptococcus has not been determined. The bronchoalveolar macrophage (BAM) is the principal phagocyte and major inflammatory cell in the normal lung (5). It is a target for HIV-l infection via the surface protein CD4, which binds the viral surface glycoprotein gp120 (6). The extent and consequences of BAM infection in vivo remain poorly characterized. In contrast to the decline in circulating CD4 lymphocytes, the number of macrophages recovered by bronchoalveolar lavage (BAL) remains normal even in late stages of AIDS. Although HIV infection of a small percent of BAM can usually be detected by sensitive techniques, viral expression appears limited (7). If BAM dysfunction occurs in patients with AIDS, it may result from an 1434
SUMMARY Cryptococcus neoformans Infections are a major cause of morbidity and mortality for HIV-Infected persons. Containment of the initial respiratory inoculation to the lung appears defective in patients with AIDS despite the low burden of HIV In bronchoalveolar macrophages. We have studied the fungistatic activity of human bronchoalveolar macrophages (BAM) cultured with an encapsulated strain of C. neoformans In the presence of pooled human serum. We observed 51.6% fungistasls after 24 h of culture. Funglstasls was diminished If the pooled human serum was heatinactivated but was not affected by anticryptococcal capsular IgG. HIV envelope protein (gp120) has been shown to Interfere with lymphocyte activation in vitro. We studied the effects of gp120 on BAM function and found that fungistatic activity was Inhibited 25% (p < 0.001).Although binding of yeasts was not affected, gp120 Inhibited the Internalization of bound yeasts by 46% (p 0.025). These experiments Indicate that gp120 decreases the Internalization and funglstasis of C. neofor· mans by human BAM, and they suggest a mechanism to explain how a small number of HIV-1-infected cells in the lung could impair the containment of C. neoformans.
=
AM REV RESPIR DIS 1992; 146:1434-1438
indirect consequence of HIV-l infection such as a biologicallyactiveviral product. The HIV-l virus surface glycoprotein (gp120) is the prototype of such a viral product. A growing body of literature describes innate activities of gp120(8-11) in lymphocytes and monocytes through interaction of gp120 with CD4. The low viral burden in the lung, along with the speculation that viral proteins may play a role in the pathogenesis of HIV disease, prompted us to examine the effects of gp120 on the anticryptococcal activity of BAM. Methods Bronchoalveolar Lavage Cells BAL was performed on healthy, nonsmoking volunteers with no risk factors for HIV infection. Using a standard protocol (12)approved by the Human Studies Committees of Boston City Hospital and University Hospital, a total of 240 ml of normal saline was instilled and 50 to 80070 was recovered by gentle suction. The recovered fluid was strained through a single layerof sterile gauze and centrifuged (at 500 g for 10 min), and the cells were resuspended for 20 s in iced sterile water to lyse erythrocytes. Cells were washed twice in phosphate-buffered saline (PBS), counted, and resuspended in RPMI 1640(GIBCO Laboratories, Grand Island, NY). The average viability of BAL cells (BALe) processed in this fashion was 89% (range, 75 to 95%).
Cytospin preparations (Shandon Southern Instruments, Sewickley, PAl of the fully processed BALe were stained with a modified Wright stain (Leukostat; Fisher Diagnostics, Orangeburg, NY), and differential counts were performed by examining 300 cells. Processed BALe averaged greater than 91% BAM (range, 83 to 99%) and less than 2% neutrophils. For certain experiments, bronchoalveolar lymphocytes wereremoved from BALe by a modified sheep RBC rosetting technique (13).Sheep blood alsevers (SRBC) (Organon Teknika Corp., West Chester, PAl werewashed twice in Hanks' balanced salt solution (HBSS), incubated with 1unit of neuraminidase (Sigma Chemical Co., St. Louis, MO) for 1 hat 37° C, and washed three times in HBSS. BALe were incubated with treated (Received in original form October 21, 1991 and in revised form August 5, 1992) 1 From the Pulmonary Center, Evans Department of Clinical Research, and the Department of Medicine, Boston University School of Medicine, Boston, Massachusetts. 2 Correspondence and requests for reprints should be addressed to Dr. Hardy Kornfeld, Pulmonary Center, Boston UniversitySchool of Medicine, 80 East Concord Street, Boston, MA 02118. 3 Supported by Grant ROI AI-25780 from the National Institute of Allergy and Infectious Diseases. 4 Recipient of Grant ROI HL-44846-0l from the National Heart, Lung and Blood Institute and of a Career Investigator Award from the American Lung Association.
1435
gp120 INHIBITS THE ANTICRVPTOCOCCAL ACTIVITY OF BRONCHOALVEOLAR MACROPHAGES
SRBC for 5 min at 37 0 C followed by 45 min at 4 0 C with continuous gentle agitation. Cells were then subjected to Ficoll-Paque (Pharmacia LKB, Uppsala, Sweden) densitygradient centrifugation. Purified BAM obtained in this manner were> 90070 viable by trypan blue exclusion and> 99% pure as determined by morphologic examination and a-naphthyl acetate esterase staining (Sigma). Cryptococcus Neoformans A serotype D encapsulated strain, B3501, was obtained from Dr. Eric Jacobson (Richmond, VA). The culture was maintained by serial passage on asparagine minimal essential medium (14)and harvested by suspending a single colony in 1.0 ml of RPMI 1640 (growth inhibition assay) or PBS (binding and internalization assays). Repeated aspiration through a 25-gauge needle and vortexing for 30 s was used to produce a suspension of yeasts; 95% of organisms appeared as single cells when counted on a hemacytometer. Reagents Pooled human serum (PHS) was obtained by combining serum from more than 10healthy volunteers under conditions preserving complement activity. Recombinant gp120, recombinant soluble CD4 (rsCD4), and rabbit antigp120IgO weregifts from American BioTechnologies (Cambridge, MA). The recombinant proteins are baculovirus products purified by immunoaffinity chromatography. All reagents were negative for lipopolysaccharide (LPS) by Limulus amebocyte lysate assay. Biologic activity of gp120 was confirmed by lymphocyte chemotaxis assay (8). The gp120was used at 2 ug/ml (17 nM), except where otherwise noted; rsCD4 was used at 3.5 ug/ml (70 nM). Rabbit antiserotype D cryptococcal capsular antibody (anti-CN IgO), a gift of Dr. John Bennett (Bethesda, MD), was used at a dilution of 1:10,000; no clumping of yeasts was seen at this antibody dilution. Fungistasis and Killing of C. Neoformans BALC or BAM and C neoformans were cultured in 0.5 ml RPMI 1640 with 10% PHS in sterile polypropylene tubes. After incubation with reagents as described below,200,000 BALC or BAM were challenged with 40,000 colony-forming units (CFU) of C neoformans and gently agitated for 60 min at 37 0 C. Experiments performed with lower ratios of BAM to C neoformans, as described by Weinberg and coworkers (15), yielded similar results. Cultures were then incubated 18 to 24 h at 37 0 C, 5% CO 2 , 90% humidity, and the cells were lysed in 0.1% Triton X-IOO (Fisher Scientific, Fair Lawn, NJ). Preliminary experiments confirmed the efficiency of lysis and lack of effect of 0.1% Triton X-IOO on cryptococcal viability. After lysis, cultures werediluted 1:20in sterile water with 50 ug/ml chloramphenicol (Sigma), vortexed, and spread onto Sabouraud dextrose agar plates (REMEL, Lenexa, KS). Plates wereincubated at room temperature for 3 days, and the colonies were counted. All conditions were per-
formed in triplicate, and all tubes were plated in duplicate. Control cultures were made with C neoformans alone in media with 10% PHS, with or without anti-CN Igo. One set of cultures was plated immediately to establish inoculum size;the other was handled identically to cultures containing BALC or BAM to establish the growth of C neoformans in the absence of these cells. On average, cultures doubled 3.6 times during the incubation period. Preliminary experiments showed that gp120 and rsCD4 had no effect on the growth characteristics of the C neoformans. The rsCD4 and gp120, individually and in combination, were incubated at 4 0 C for 4 h immediately prior to their addition to the culture tubes. Reagents wereadded to the BALCcontaining culture tubes using the following schedule: anti-CN IgO was added immediately prior to the addition of C neoformans; gp120 or gp120 plus rsCD4 was added to BALC 60 min prior to the addition of C neoformans; in crosslinking experiments, antigp120 was added 30 min after the addition of gp120 and 30 min before the addition of C neoformans. The incubations were arranged so that C neoformans was added to all tubes at the same time. Percent inhibition was calculated as {[(CFU growth control) (CFU experimental group)]I(CFU growth control)} x 100. Binding Assay Binding of C neoformans to BAM adherent to plastic was performed as previously described (16). Briefly, 50,000 BAM were plated in microwells. Lymphocytes and nonadherent BAM were removed by washing with RPMI 1640. Wells were challenged with 2 x 106 heat-killed yeasts (30 min at 37 0 C) labeled with fluorescein isothiocyanate (FITC). Unbound organisms werewashed off, and the monolayer was fixed with 1070 formaldehyde in PBS. Using an inverted microscope at x 200, BAM were randomly selected under bright field and then examined under epifluorescence to determine the number of yeasts associated per BAM. Binding index represents the number of cell-associatedyeasts per 100 BAM. Internalization Assay A previously described assay (17) using diaethanol (Uvitex 2B; Ciba-Giegy, Basel, Switzerland) was used to distinguish surfacebound from internalized organisms. BAM were plated in an eight-chamber tissue culture slide (Lab-Tek; Miles Scientific, Naperville, IL) and challenged for 30 min with heatkilled organisms labeled with rhodamine isothiocyanate (RITC). Plates were washed and incubated for 1 min with 0.1% Uvitex at 23 0 C, then washed again and incubated for 1 min with 0.1% Uvitex at 23 0 C, then washed again and fixed with 1% formaldehyde for epifluorescent microscopy. Organisms that are not fully internalized were stained with Uvitex 2B, which appears blue under UV excitation. Yeasts were identified by RITe fluorescence at 546 nm excitation
and then examined under UV excitation to determine if the yeast was fully internalized. At least 100cryptococci per wellweresequentially examined at 546 nm and UV illumination with the observer blinded to the experimental variables. Results are expressed as the percent RITC-stained organisms that escaped counterstaining with Uvitex. Immunofluorescent Analysis BALC were washed twice in PBS with 0.1% sodium azide and incubated at 4 0 C with 50 ug of nonimmune mouse IgO to block nonspecific and FcR-mediated binding oflabeled antibody. Equal portions of cells were then incubated with OKT4-phycoerythrin conjugate (Ortho Diagnostic Systems, Raritan, NJ) or control IgO-phycoerythrin conjugate for 30 min at 4 0 C, washed three times in iced PBS, resuspended in PBS with 10% formalin, and maintained on ice until analysis with a FACScan cell sorter (Becton-Dickinson, Sunnyvale, CA). Statistics Comparison of means for experimental data was made using Student's t test for paired samples. For those experiments in which more than one comparison was made, the p value to determine significance was adjusted using the Bonferroni correction. Control conditions and conditions of blocking were evaluated at the less stringent levelof significance to avoid type II errors. Significance was confirmed using Wilcoxon's signed rank test (18).There wereno discrepancies in the results from these two methods of statistical analysis. Results
Initial experiments were performed to verify the ability of BALC in vitro to inhibit the growth of C neoformans. The comparison of the BALe versus the 370 C growth control demonstrated a 51.6% reduction of C neoformans growth by the BALC (figure 1). The substitution of heat 70
en -c len c;
iii
60 50 40
Z ::J
30
'#.
20 10
II..
l
0 BALC
BALC + aCN-lgG
Fig. 1. Anticryptococcal activity of BALC. BALC (2 x 105) were incubated with 4 x 104 CFU of C. neoformans in medium containing 10% PHS for 18 to 24 h in the presence (BALC + aCN-lgG) or absence (BALC)of rabbit anticryptococcal capsular antibody. Fungistasis was determined by comparing the CFU of C. neoformans in tubes containing BALC with control tubes incubated in an identical manner without BALC. Results represented means ± SEM of six triplicate experiments. Antibody had no significant effect on fungistasis.
1436
WAGNER, LEVITZ, TABUNI, AND KORNFELD
inactivated serum decreased the anticryptococcal activity of the BALe (data not shown), and the addition of anti-CN IgG did not enhance fungistasis (figure I). To determine if the anticryptococcal activity of the BALC requires the presence of lymphocytes recovered during BAL, a portion of the BALC was set aside, and the remainder was purified to BAM by SRBC rosetting. No difference was detected in the anticryptococcal activity of these two populations of cells (figure 2). The presence of the CD4 receptor on monocytes and macrophages is widelyaccepted. Conflicting reports of high (19, 20) and low (21, 22) levels of expression probably reflects the difficulty of using fluorescence-labeled antibodies to detect CD4 on mononuclear phagocytes, which have a significant degree of autofluorescence. Salahuddin and coworkers (23) reported that only 2 to 4070 of adherencepurified BAM expressedCD4 after 5 days of culture. Our data suggest that the majority of freshly isolated BAM expresses at least low levelsof CD4, the usual receptor for gp120 (figure 3). Experiments performed with BALe in the presence of 2 ug/ml of gp120 led to a 25% reduction in fungistasis (51.9 to 38.8%, p < 0.001) as illustrated in figure 4. In dose-response experiments using 2, 5, and 10 ug/ml of gp120, stepwise decrements of fungistasis were seen that reached 40% at 10 ug/rnl (data not shown). Preincubation of gp120 with rsCD4 partially blocked the effect of the gp120; a statistically significant difference could not be demonstrated between the blocked gp120 condition and either the control or the gp120condition. This sug70 (j)
en
60
-c
50
(j)
40
Z ::J IL
30
I-
RANDALL P. WAGNER, STUART M. LEVITZ,3 ABDULMONEIM TABUNI, and HARDY KORNFELD4 Introduction
Cryptococcus neoformans infections are a major cause of morbidity and mortality for HIV-infected persons (1). Environmental exposure to Cryptococcus organisms is common, with the usual site of inoculation being the lower respiratory tract through the inhalation of aerosolized infectious material (2). Despite colonization of the tracheobronchial tree, dissemination outside the lung in the normal host is rare (3, 4). Several mechanisms contribute to the containment of the fungus to the lung, including the spontaneous fungistatic activity of bronchoalveolar macrophages. When cryptococcosis does develop, control of the disease usually requires the action of cell-mediated immunity. The innocuous behavior of this yeast in the normal host is in sharp contrast to the severity of the disease in patients with AIDS. Between 5 and 10070 of all patients with AIDS in the United States present with cryptococcal meningitis, and the case mortality rate remains high despite improvements in therapy (1). Although it is clear that the pulmonary defenses are altered in AIDS, the basis for this loss of protection against Cryptococcus has not been determined. The bronchoalveolar macrophage (BAM) is the principal phagocyte and major inflammatory cell in the normal lung (5). It is a target for HIV-l infection via the surface protein CD4, which binds the viral surface glycoprotein gp120 (6). The extent and consequences of BAM infection in vivo remain poorly characterized. In contrast to the decline in circulating CD4 lymphocytes, the number of macrophages recovered by bronchoalveolar lavage (BAL) remains normal even in late stages of AIDS. Although HIV infection of a small percent of BAM can usually be detected by sensitive techniques, viral expression appears limited (7). If BAM dysfunction occurs in patients with AIDS, it may result from an 1434
SUMMARY Cryptococcus neoformans Infections are a major cause of morbidity and mortality for HIV-Infected persons. Containment of the initial respiratory inoculation to the lung appears defective in patients with AIDS despite the low burden of HIV In bronchoalveolar macrophages. We have studied the fungistatic activity of human bronchoalveolar macrophages (BAM) cultured with an encapsulated strain of C. neoformans In the presence of pooled human serum. We observed 51.6% fungistasls after 24 h of culture. Funglstasls was diminished If the pooled human serum was heatinactivated but was not affected by anticryptococcal capsular IgG. HIV envelope protein (gp120) has been shown to Interfere with lymphocyte activation in vitro. We studied the effects of gp120 on BAM function and found that fungistatic activity was Inhibited 25% (p < 0.001).Although binding of yeasts was not affected, gp120 Inhibited the Internalization of bound yeasts by 46% (p 0.025). These experiments Indicate that gp120 decreases the Internalization and funglstasis of C. neofor· mans by human BAM, and they suggest a mechanism to explain how a small number of HIV-1-infected cells in the lung could impair the containment of C. neoformans.
=
AM REV RESPIR DIS 1992; 146:1434-1438
indirect consequence of HIV-l infection such as a biologicallyactiveviral product. The HIV-l virus surface glycoprotein (gp120) is the prototype of such a viral product. A growing body of literature describes innate activities of gp120(8-11) in lymphocytes and monocytes through interaction of gp120 with CD4. The low viral burden in the lung, along with the speculation that viral proteins may play a role in the pathogenesis of HIV disease, prompted us to examine the effects of gp120 on the anticryptococcal activity of BAM. Methods Bronchoalveolar Lavage Cells BAL was performed on healthy, nonsmoking volunteers with no risk factors for HIV infection. Using a standard protocol (12)approved by the Human Studies Committees of Boston City Hospital and University Hospital, a total of 240 ml of normal saline was instilled and 50 to 80070 was recovered by gentle suction. The recovered fluid was strained through a single layerof sterile gauze and centrifuged (at 500 g for 10 min), and the cells were resuspended for 20 s in iced sterile water to lyse erythrocytes. Cells were washed twice in phosphate-buffered saline (PBS), counted, and resuspended in RPMI 1640(GIBCO Laboratories, Grand Island, NY). The average viability of BAL cells (BALe) processed in this fashion was 89% (range, 75 to 95%).
Cytospin preparations (Shandon Southern Instruments, Sewickley, PAl of the fully processed BALe were stained with a modified Wright stain (Leukostat; Fisher Diagnostics, Orangeburg, NY), and differential counts were performed by examining 300 cells. Processed BALe averaged greater than 91% BAM (range, 83 to 99%) and less than 2% neutrophils. For certain experiments, bronchoalveolar lymphocytes wereremoved from BALe by a modified sheep RBC rosetting technique (13).Sheep blood alsevers (SRBC) (Organon Teknika Corp., West Chester, PAl werewashed twice in Hanks' balanced salt solution (HBSS), incubated with 1unit of neuraminidase (Sigma Chemical Co., St. Louis, MO) for 1 hat 37° C, and washed three times in HBSS. BALe were incubated with treated (Received in original form October 21, 1991 and in revised form August 5, 1992) 1 From the Pulmonary Center, Evans Department of Clinical Research, and the Department of Medicine, Boston University School of Medicine, Boston, Massachusetts. 2 Correspondence and requests for reprints should be addressed to Dr. Hardy Kornfeld, Pulmonary Center, Boston UniversitySchool of Medicine, 80 East Concord Street, Boston, MA 02118. 3 Supported by Grant ROI AI-25780 from the National Institute of Allergy and Infectious Diseases. 4 Recipient of Grant ROI HL-44846-0l from the National Heart, Lung and Blood Institute and of a Career Investigator Award from the American Lung Association.
1435
gp120 INHIBITS THE ANTICRVPTOCOCCAL ACTIVITY OF BRONCHOALVEOLAR MACROPHAGES
SRBC for 5 min at 37 0 C followed by 45 min at 4 0 C with continuous gentle agitation. Cells were then subjected to Ficoll-Paque (Pharmacia LKB, Uppsala, Sweden) densitygradient centrifugation. Purified BAM obtained in this manner were> 90070 viable by trypan blue exclusion and> 99% pure as determined by morphologic examination and a-naphthyl acetate esterase staining (Sigma). Cryptococcus Neoformans A serotype D encapsulated strain, B3501, was obtained from Dr. Eric Jacobson (Richmond, VA). The culture was maintained by serial passage on asparagine minimal essential medium (14)and harvested by suspending a single colony in 1.0 ml of RPMI 1640 (growth inhibition assay) or PBS (binding and internalization assays). Repeated aspiration through a 25-gauge needle and vortexing for 30 s was used to produce a suspension of yeasts; 95% of organisms appeared as single cells when counted on a hemacytometer. Reagents Pooled human serum (PHS) was obtained by combining serum from more than 10healthy volunteers under conditions preserving complement activity. Recombinant gp120, recombinant soluble CD4 (rsCD4), and rabbit antigp120IgO weregifts from American BioTechnologies (Cambridge, MA). The recombinant proteins are baculovirus products purified by immunoaffinity chromatography. All reagents were negative for lipopolysaccharide (LPS) by Limulus amebocyte lysate assay. Biologic activity of gp120 was confirmed by lymphocyte chemotaxis assay (8). The gp120was used at 2 ug/ml (17 nM), except where otherwise noted; rsCD4 was used at 3.5 ug/ml (70 nM). Rabbit antiserotype D cryptococcal capsular antibody (anti-CN IgO), a gift of Dr. John Bennett (Bethesda, MD), was used at a dilution of 1:10,000; no clumping of yeasts was seen at this antibody dilution. Fungistasis and Killing of C. Neoformans BALC or BAM and C neoformans were cultured in 0.5 ml RPMI 1640 with 10% PHS in sterile polypropylene tubes. After incubation with reagents as described below,200,000 BALC or BAM were challenged with 40,000 colony-forming units (CFU) of C neoformans and gently agitated for 60 min at 37 0 C. Experiments performed with lower ratios of BAM to C neoformans, as described by Weinberg and coworkers (15), yielded similar results. Cultures were then incubated 18 to 24 h at 37 0 C, 5% CO 2 , 90% humidity, and the cells were lysed in 0.1% Triton X-IOO (Fisher Scientific, Fair Lawn, NJ). Preliminary experiments confirmed the efficiency of lysis and lack of effect of 0.1% Triton X-IOO on cryptococcal viability. After lysis, cultures werediluted 1:20in sterile water with 50 ug/ml chloramphenicol (Sigma), vortexed, and spread onto Sabouraud dextrose agar plates (REMEL, Lenexa, KS). Plates wereincubated at room temperature for 3 days, and the colonies were counted. All conditions were per-
formed in triplicate, and all tubes were plated in duplicate. Control cultures were made with C neoformans alone in media with 10% PHS, with or without anti-CN Igo. One set of cultures was plated immediately to establish inoculum size;the other was handled identically to cultures containing BALC or BAM to establish the growth of C neoformans in the absence of these cells. On average, cultures doubled 3.6 times during the incubation period. Preliminary experiments showed that gp120 and rsCD4 had no effect on the growth characteristics of the C neoformans. The rsCD4 and gp120, individually and in combination, were incubated at 4 0 C for 4 h immediately prior to their addition to the culture tubes. Reagents wereadded to the BALCcontaining culture tubes using the following schedule: anti-CN IgO was added immediately prior to the addition of C neoformans; gp120 or gp120 plus rsCD4 was added to BALC 60 min prior to the addition of C neoformans; in crosslinking experiments, antigp120 was added 30 min after the addition of gp120 and 30 min before the addition of C neoformans. The incubations were arranged so that C neoformans was added to all tubes at the same time. Percent inhibition was calculated as {[(CFU growth control) (CFU experimental group)]I(CFU growth control)} x 100. Binding Assay Binding of C neoformans to BAM adherent to plastic was performed as previously described (16). Briefly, 50,000 BAM were plated in microwells. Lymphocytes and nonadherent BAM were removed by washing with RPMI 1640. Wells were challenged with 2 x 106 heat-killed yeasts (30 min at 37 0 C) labeled with fluorescein isothiocyanate (FITC). Unbound organisms werewashed off, and the monolayer was fixed with 1070 formaldehyde in PBS. Using an inverted microscope at x 200, BAM were randomly selected under bright field and then examined under epifluorescence to determine the number of yeasts associated per BAM. Binding index represents the number of cell-associatedyeasts per 100 BAM. Internalization Assay A previously described assay (17) using diaethanol (Uvitex 2B; Ciba-Giegy, Basel, Switzerland) was used to distinguish surfacebound from internalized organisms. BAM were plated in an eight-chamber tissue culture slide (Lab-Tek; Miles Scientific, Naperville, IL) and challenged for 30 min with heatkilled organisms labeled with rhodamine isothiocyanate (RITC). Plates were washed and incubated for 1 min with 0.1% Uvitex at 23 0 C, then washed again and incubated for 1 min with 0.1% Uvitex at 23 0 C, then washed again and fixed with 1% formaldehyde for epifluorescent microscopy. Organisms that are not fully internalized were stained with Uvitex 2B, which appears blue under UV excitation. Yeasts were identified by RITe fluorescence at 546 nm excitation
and then examined under UV excitation to determine if the yeast was fully internalized. At least 100cryptococci per wellweresequentially examined at 546 nm and UV illumination with the observer blinded to the experimental variables. Results are expressed as the percent RITC-stained organisms that escaped counterstaining with Uvitex. Immunofluorescent Analysis BALC were washed twice in PBS with 0.1% sodium azide and incubated at 4 0 C with 50 ug of nonimmune mouse IgO to block nonspecific and FcR-mediated binding oflabeled antibody. Equal portions of cells were then incubated with OKT4-phycoerythrin conjugate (Ortho Diagnostic Systems, Raritan, NJ) or control IgO-phycoerythrin conjugate for 30 min at 4 0 C, washed three times in iced PBS, resuspended in PBS with 10% formalin, and maintained on ice until analysis with a FACScan cell sorter (Becton-Dickinson, Sunnyvale, CA). Statistics Comparison of means for experimental data was made using Student's t test for paired samples. For those experiments in which more than one comparison was made, the p value to determine significance was adjusted using the Bonferroni correction. Control conditions and conditions of blocking were evaluated at the less stringent levelof significance to avoid type II errors. Significance was confirmed using Wilcoxon's signed rank test (18).There wereno discrepancies in the results from these two methods of statistical analysis. Results
Initial experiments were performed to verify the ability of BALC in vitro to inhibit the growth of C neoformans. The comparison of the BALe versus the 370 C growth control demonstrated a 51.6% reduction of C neoformans growth by the BALC (figure 1). The substitution of heat 70
en -c len c;
iii
60 50 40
Z ::J
30
'#.
20 10
II..
l
0 BALC
BALC + aCN-lgG
Fig. 1. Anticryptococcal activity of BALC. BALC (2 x 105) were incubated with 4 x 104 CFU of C. neoformans in medium containing 10% PHS for 18 to 24 h in the presence (BALC + aCN-lgG) or absence (BALC)of rabbit anticryptococcal capsular antibody. Fungistasis was determined by comparing the CFU of C. neoformans in tubes containing BALC with control tubes incubated in an identical manner without BALC. Results represented means ± SEM of six triplicate experiments. Antibody had no significant effect on fungistasis.
1436
WAGNER, LEVITZ, TABUNI, AND KORNFELD
inactivated serum decreased the anticryptococcal activity of the BALe (data not shown), and the addition of anti-CN IgG did not enhance fungistasis (figure I). To determine if the anticryptococcal activity of the BALC requires the presence of lymphocytes recovered during BAL, a portion of the BALC was set aside, and the remainder was purified to BAM by SRBC rosetting. No difference was detected in the anticryptococcal activity of these two populations of cells (figure 2). The presence of the CD4 receptor on monocytes and macrophages is widelyaccepted. Conflicting reports of high (19, 20) and low (21, 22) levels of expression probably reflects the difficulty of using fluorescence-labeled antibodies to detect CD4 on mononuclear phagocytes, which have a significant degree of autofluorescence. Salahuddin and coworkers (23) reported that only 2 to 4070 of adherencepurified BAM expressedCD4 after 5 days of culture. Our data suggest that the majority of freshly isolated BAM expresses at least low levelsof CD4, the usual receptor for gp120 (figure 3). Experiments performed with BALe in the presence of 2 ug/ml of gp120 led to a 25% reduction in fungistasis (51.9 to 38.8%, p < 0.001) as illustrated in figure 4. In dose-response experiments using 2, 5, and 10 ug/ml of gp120, stepwise decrements of fungistasis were seen that reached 40% at 10 ug/rnl (data not shown). Preincubation of gp120 with rsCD4 partially blocked the effect of the gp120; a statistically significant difference could not be demonstrated between the blocked gp120 condition and either the control or the gp120condition. This sug70 (j)
en
60
-c
50
(j)
40
Z ::J IL
30
I-
