Human Alveolar and Peritoneal Macrophages Mediate Fungistasis Independently of L-Arginine Oxidation to Nitrite or Nitrate1-4
MIRIAM L. CAMERON, DONALD L. GRANGER, J. BRICE WEINBERG, WALTER J. KOZUMBO, and HILLEL S. KOREN Introduction SUMMARY Human alveolar macrophages (HAM) from 28 normal volunteers ware found to Inhibit
Macrophages play an important role in replication of Cryptococcus neoformans. Conditions under which funglstasls occurred ware differthe killing and inhibition of replication ent than those required for mouse peritoneal macrophage-mediated funglstasls. Inhibition of fungal of microorganisms and tumor cells. There replication by mouse peritoneal macrophages (MPM) requires that the macrophages are activated has been extensive work studying the oxyand that the cocultures of C. neoform.ns and macrophages be done In the presence of serum, gen-dependent mechanisms of host resisL-arginine, and endotoxin. During MPM-medlated funglstasls and tumor cell killing, L-arginine Is oxidized to NO,·, NO.·, and L-citrulllne. In addition, MPM have arginase activity that converts Ltance, but more recently interest has arginine to L-ornlthine and urea. turned to the role of L-arginine nitrogen HAM-mediated funglstasls was similar to that mediated by MPM In terms of the serum requireoxidation metabolism in microbiostasis ment, but HAM did not require L-arginine or endotoxin. HAM did not produce NO,· or NO.· detectable and tumor cell injury (1-3). L-arginineby colorimetric and bioassay, nor did HAM produce L-citrulllne or L-ornlthlne from "C-radlolabeled dependent mouse peritoneal macrophageL-arginine as detectable by reverse-phase Ion-pairing HPLC of macrophage-c. neoformans coculmediated microbiostasis and tumoristature supernatants. HAM had no detectable arginase actiVity, hence there was no evidence for Lsis has been studied. Stuehr and MarIetta arginine nitrogen metabolism In HAM. HAM-mediated funglstasls was not enhanced by endotoxin have shown that MPM infected in vivo or by recombinant human Interferon-y (rHIFN-y). The combination of endotoxin and rHIFN-y Inhibitby BCG produce N0 2 - and N0 3 - (4). ed the fungistatic effect of HAM. Human peritoneal macrophages (HPM) from women undergoing laparoacopy ware tested for funglstasls and L-arginine nitrogen oxidation. Partial Inhibition of crypHibbs and colleagues showed that actitococcal replication occurred; howaver, there was no evidence of L-arginine metabolism to NO,· vated MPM were cytotoxic to tumor cells or NO.·. The absence of L-arginlne-dependent nitrogen oxidation In HAM and HPM, compared to via an L-arginine-dependent pathway that MPM, during conditions under which funglstasls occurs suggests that this phenomenon Is species produces N02 - and L-citrulline (1, 2). specific rather than specific to the tissue origin of the macrophages. Further work by Granger and coworkers AM REV RESPIR DIS 1990; 142:1313-1319 showed that microbiostasis of Cryptococcus neojormans by MPM also occurred via the L-arginine-dependent mechanism fed a low-nitrate diet show that endogein original form April 9, 1990 and in and that this pathway and fungistasis nous nitrate biosynthesis occurs (8). (Received revised form June 13, 1990) could be specifically inhibited by the The site of endogenous nitrate biosynL-arginine analog N-guanidomonomethyl thesis was not known when the metabolFrom the Division of Infectious Diseases, Duke L-arginine (3). Although murine macro- ic studies were done, but nitric oxide syn- University Medical Center, and the Division of phages require L-arginine nitrogen oxida- thesis has been found in neutrophils (9), Hematology-Oncology, Veterans Administration tion to nitrite or nitrate to mediate tumor cells (to), hepatocytes (11), central and Duke University Medical Center, Durham; the Center. for Environmental Medicine and Lung Bimicrobiostasis or tumoristasis, neither ni- nervous system tissue (12), and endotheli- ology, University of North Carolina, Chapel Hill; trite or nitrate seemed to be the mediator al cells (13). Hence the possibility exist- and the Clinical Research Branch, Human Studies molecule of cytostasis (Granger, unpub- ed that human macrophages, in a manner Division, Health Effects Research Laboratory, Unitlished data and reference 5). Stuehr and similar to MPM, also utilized L-arginine ed States Environmental Protection Agency, Reltiangle Park, North Carolina. coworkers have shown that nitric oxide nitrogen oxidation to generate NO to me- search , Supported by Grant No. AI-26188, CA-41987, (NO) seems to be the important effector diate microbiostasis. Results of this study POI AI-23308, POI 32682, and P50 AR-39162 from molecule produced by murine macro- demonstrated that HAM and human the National Institutes of Health, by R.J. Reyphages during the L-arginine-dependent peritoneal macrophage (HPM) inhibit- nolds-Nabisco Industries, by the Veterans AdriUnisinhibition of tumor cell replication (5). In ed C neojormans replication in vitro in tration Research Service, and by Cooperative Agreement No. CR8 12738 from the Environmental Proaddition, they have shown that the meta- . an L-arginine-independent manner and tection Agency. bolic product of L-arginine oxidation by that the macrophages did not metabo3 The research described in this report has been murine macrophages has the same activ- lize L-arginine to nitrate or nitrite either reviewed by the Health Effects Research Laboraityas endothelial relaxation factor (6). in their basal state or after treatment with tory, U.S. Environmental Protection Agency, and The purpose of this study was to de- endotoxin or recombinant human inter- approved for publication. Approval does not signify that the contents necessarily reflect the views termine whether L-arginine oxidation feron-y (rHIFN-y). of the agency, nor does mention of trade names was required for human macrophageor commercial products constitute endorsement or recommendation for use. mediated fungistasis. Human alveolar Methods 4 Correspondence and requests for reprints should macrophages (HAM) inhibit fungal replibe addressed to Miriam L. Cameron, M.D., Box Cell Culture Media cation in vitro (7). In addition, metabol3524, Duke University Medical Center, Durham, ic studies in normal human volunteers RPMI 1640 (JR Scientific, Woodland, CA) NC 27710. 1
1313
1314
was used to prepare alveolar macrophages immediately after explantation; otherwis~, all experiments were performed in Dulbecco's modified Eagle medium (DMEM). The DMEM was prepared in the laboratory as previously described (3) from stock solutions. L-arginine in a final concentration of 500 J.1M was added to the medium at each experiment unless a comparison to arginine-free media was done. Several experiments were carried out in DMEM that was modified by excluding all amino acids and replacing them with either 500 J.1M L-asparagine or 500 J.1M L-citrulline as a source of the nitrogen required by C neojormans (14). Experiments in L-arginine-free media containing L-asparagine or L-citrulline were carried out with human serum that had been dialyzed (molecular weight cutoff 6,000 to 8,000 D) extensively against phosphatebuffered saline (PBS) to remove free amino acids. In a separate group of experiments, L-arginine was replaced by L-arginine analogs (Sigma Chemical Co., St. Louis, MO). All the L-arginine analogs were at a 500 J.1M final concentration except for poly-L-arginine, which was 5 J.1M, and Ala-gly-arginine vasopressin, which was 70 J.1M. In some experiments, I J.1g/ml of endotoxin (phenol extract of Escherichia coli O.l28:B12 from Sigma Chemical Co.) and/or 100 V/m1 of recombinant human interferon-y (generously donated by Genentech, Inc., San Francisco, CA) was added to the cultures.
Alveolar Macrophages Human alveolar macrophages (HAM) were obtained by bronchoscopy and bronchoalveolar lavage as previously described (15) from 28 normal volunteers aged 18 to 35 yr who gave informed consent under a protocol approved by the Investigational Review Board. The HAM were washed three times with RPMI 1640 and then resuspended in RPMI 1640 at a concentration of 1 x 1()6/ml. TIypan blue exclusion showed a mean viability of 850/0 (range 50 to 100%). Previous work has shown that 95% ofthe cells were maerophages by morphologic criteria (7). Two different treatments of the HAM occurred after their resuspension. A pilot experiment comparing the two treatments revealed no difference in HAM viability or functional activity. In most experiments the HAM were incubated overnight at 4° C in 10% fetal bovine serum (FBS) in RPMI 1640. The day after explantation, these cells were washed three times in PBS containing 1.2 mM CaC1 2 and 0.8 mM MgC1 2. The cells were then resuspended in DMEM without serum. The cells were plated into 16-mm Costar tissue culture wells (Costar, Cambridge, MA) at 0.5 x 10" cells/well and incubated in 5% CO 2 and 95% humidified air at 37° C for I h. Nonadherent cells were removed by gentle washing with PBS. To increase the density of the adherent cells, the wells were reseeded with another 0.5 x 1()6 cells/well, which were allowed to adhere under the same conditions for an hour. The remaining nonadherent cells were
CAMERON, GRANGER, WEINBERG, KOZUMBO, AND KOREN
gently washed off the plates three times with PBS. After the final wash, 0.5 ml DMEM with fresh-frozen human serum was added. Alternatively, HAM were plated into the 16-mm Costar culture wells the day of explantation. Into each well was placed 0.5 x 10" cells in RPMI 1640, and then the cells were incubated under the same conditions as before for 1 h. Nonadherent cells were washed off three times with PBS, and the media were replaced with 10% FBS and RPMI 1640. The macrophage monolayers were incubated overnight at 37° C in humidified 5% CO2 and 95% air. The day after explantation, the monolayers were washed three times with PBS and then 0.5 ml of the appropriate experimental medium was added.
Peritoneal Macrophages Several experiments were carried out using human peritoneal macrophages. Human peritoneal macrophages were obtained from women undergoing laparoscopy for evaluation of infertility (16). The explanted cells were stored overnight at 4° C in 10% autologous unheated serum in DMEM. The cells were centrifuged at 200 x g for 10 min and then the supernatant was removed. The pellet was resuspended in DMEM and layered onto Ficoll-Hypaque (Histopaque®; Sigma Chemical Co.) and centrifuged at 400 x g for 25 min. The interface that contained the HPM was removed, and the cells were washed three times with DMEM (200 x g for 10 min) and resuspended to a concentration of 2 x 10" cells/ml. A single seeding of 0.7 ml (1.4 x 10" cells/ml) was applied to 16-mm Costar tissue culture wells and allowed to adhere under the same conditions as for the HAM. Nonadherent cells were removed by washing with PBS three times, and the media were replaced with the experimental media (5% fresh human serum and DMEM). Morphologic studies have shown that> 95% of these cells were macrophages by Wright stain and nonspecific esterase (17). Cryptococcus neoformans A previously described (18), thinly encapsulated clone of C neojormans (serotype A) designated H99/C3D, which originated from the cerebrospinal fluid of a patient, was used in all fungistasis assays. The yeasts were maintained in DMEM suspension cultures in an environment of humidified 5% CO2and 95% air at 37° C. A fresh suspension of log-phase C neojormans was prepared by washing the yeast cells three times in PBS and then resuspending the washed yeasts in the modified DMEM. The resuspended yeasts were counted manually on a hemacytometer and electronically on a Coulter counter (19) and then diluted with media to a concentration of 1 x 105/ml before use in each assay.
Fungistasis Measurements The fungistasis assays were done using a modification of the methods of Granger and coworkers (19). After confluent macrophage monolayers and C neojormans were pre-
pared, 0.5 ml of the yeasts at a concentration of 1 x 105/ml was added to each of the macrophage monolayer wells (HAM-cryptococci, 10:1 to 20:1) and to control wells containing 0.5 ml medium alone. Macrophage controls were cultured in media without the addition of C neojormans. The macrophage-yeast cocultures, macrophage controls, and yeast controls were incubated in humidified 5% CO 2 and 95% air at 37° C for various time periods. The assay was terminated at 48 h except for the time course assays, for which various time points were determined. At the appropriate time, 0.1 ml of 10% sodium dodecyl sulfate (SDS; Sigma Chemical Co.) was added to each well to lyse the macrophages and terminate C neojormans replication (3). If the supernatants were to be assayed for N0 2- or N0 3-, 0.8 ml supernatant was removed before SDS treatment and replaced with 0.8 ml PBS. The supernatant was saved and processed as described subsequently. If quantitative plate counts were to be done, triplicate HAM-C neojormans cocultures and C neojormans controls were treated with 0.1 ml of 10% deoxycholate instead of SDS. After lysis of macrophage monolayers, 0.5 ml each of the lysates and controls was mixed in 10 ml of 10% hypochlorite in Isoton buffer (Curtin Matheson Scientific, Inc., Marietta, GA), and then they were counted electronically in a ZBI Coulter counter (Coulter Electronics, Inc., Hialeah, FL) in the manner previously described (16). All cultures except macrophage-alone controls were duplicates. The results of macrophage-cryptococci cocultures were expressed by two methods. First, cryptococcal cell counts versus time were plotted directly on semilogarithmic coordinates. Second, fungistasis was expressed as the difference between cryptococcal replication in the absence and presence of macrophages. This difference was expressed by the formula (20) Fungistasis = log2 (fungi-alone cultures) log, (macrophage + fungi cocultures) The doubling time of H99/C3D cryptococci in DMEM containing human serum is constant at about 2.5 h after an initial seeding of 5 x 104 yeast cells/ml.
Assessment oj Phagocytosis C neojormanswas cultured overnight at 5% CO" 37° C, in 5 ml DMEM containing 250 J.1Ci 35S (ICN Radiochemicals, Irvine, CA) (specific activity 771 J.1Ci/ml). Confluent macrophage monolayers were infected with 2 x 105 cryptococci/ml, and at the appropriate time points the cells were washed gently three times with PBS and then lysed with 0.5 ml of 0.5% SDS. To 10 ml Biofluor® (DuPont, Wilmington, DE) 0.5 mllysate was added, and this was counted in a liquid scintillation counter (19). Results were expressed as the percentage uptake of 35S-labeled C neojormans using the formula
1315
HUMAN MACROPHAGES INHIBIT CRYPTOCOCCAL REPLICATION INDEPENDENTLY OF L-ARGININE Fig. 1. HAM require serum to inhibit fungal replication. There were 9 experiments utilizing 25% serum and 20 experiments utilizing 5% serum. Three experiments utilizing no serum were run in parallel with serum-containing fungistasis assays. All experiments contained duplicate samples. Results are expressed as the mean fungistasis ± standard error of the mean (SEM). There is no significant difference in fungistasis between 5 and 25% serum assays, but there is a significant difference between fungistasis in the absence of serum compared to the 5% serum assays (Studenfs ttest, p < 0.0009). Cryptococci have a dOUbling time of 3.5 to 4.0 h in the absence of serum and 2 to 2.5 h in the presence of serum.
od that measures "CO, trapping after incubation of the samples with urease (22, 25).
5
~
4
< ..
3
-'" "c z3
2
en!!! _:I: en!
Statistical Analysis Most results for fungistasis and NO,- and N0 3 - are expressed as the mean of comparable experiments ± standard error of the mean (SEM). Statistical analysis was made using Student's t test to compare the mean of media controls versus supernatants for NO,- and N03 - or comparing fungistasis occurring under various treatments.
I- Z enS!
::::llli II. a: ~
0/0 35S-labeled cryptococci uptake = DPM sample/DPM cryptococci inoculum x 100
Assessment of Fungal Viability Colony-forming units (cfu) were determined by culturing dilutions of lysates from macrophage-cryptococci cocultures and C neoformans controls onto Sabouraud modified agar (Difco, Detroit, MI). Cultured lysates were incubated at 35° C in humidified air for 5 days. Colony-forming units were counted manually.
Nitrite Assay Nitrite concentration in the supernatants of macrophage-cryptococci cocultures control macrophage cultures, control cryptococci cultures, and media alone was measured as previously described using Greiss reagents (21). Concentration was calculated using a standard curve for NO,- concentration, which is linear in the range 10 to 200 IlM (22). The lower threshold of NO,- detection in this assay is ~ 5 IlM. Nitrite concentration is expressed in micromolars.
o
'V({«(J
0%
1({(frO
vcr//cd
5%
25%
Results
Inhibition of Fungal Replication by HAM HAM from normal volunteers required the presence of fresh human serum to inhibit C neoformans replication (figure 1). There was no difference in fungistasis in the presence of 5 or 25OJo human serum, but fungistasis was significantly less in cultures without serum compared to fungistasis in the presence of 5OJo serum (p < 0.0009, Student's t test). In the absence of serum, HAM remained viable but were unable to inhibit fungal replication. The serum requirement was not dependent on total hemolytic complement activity. Human serum heated at 56° C for 30 min, which was devoid of total hemolytic complement activity, and dialyzed serum, which maintained its total hemolytic complement activity, were able to support fungistasis (figure 2). There was a trend for decreased antifungal activity in the presence of heated serum. This effect appeared to be due to decreased phagocytosis of organisms as determined by visual inspection of cultures. Resident HAM inhibited cryptococcal replication in the absence of endotoxin or rHIFN-y. Neither endotoxin nor rHIFN-y
SERUM CONCENTRATION
centration of NO,- after bioreduction minus the concentration of NO,- before bioreduction. Nitrate standards were run to show that the N0 3 - was not reduced to ammonia.
Analysis of Amino Acid Metabolites of L-Arginine "C-radiolabeled L-arginine and its metabolites were assayed utilizing the previously described method of reversed-phase ion-pairing HPLC (22,24). 14C-radiolabeled L-arginine was added to the culture media before the fungistasis assay. The conditions of the assays used to analyze L-arginine metabolites were the same as those conditions used for fungistasis assays occurring without radiolabeled L-arginine. After separation of Larginine and its metabolites by reversed-phase ion-pairing HPLC, radiolabeled sample effluents were counted in a scintillation counter. Results were expressed as degenerations per minute per sample.
Analysis for Urea Urea was assayed for in supernatants from experiments containing 14C-radiolabeled Larginine utilizing a previously described meth-
Nitrate Assay Nitrate concentration in the media and supernatants was measured utilizing a bioassay using a bacterial nitrate reductase to convert N03 - to NO,-. The original method of Bartholomew (23) was modified by using Pseudomonas oleovorans instead of Escherichia coli as the source of the nitrate reductase (Granger, unpublished method). Briefly, 0.5 ml of stock P. oleovorans was diluted 1: 10 with modified DMEM, and then 0.15 ml of the diluted P. oleovorans was added to 0.3 ml supernatant or media samples. This was followed by the addition of 0.15 ml of 0.2 M inorganic sodium phosphate buffer, pH 7.25. The reaction mixture was incubated at 37° C in humidified air for 30 min. The samples were microfuged for 2 min, and then 0.4 ml supernatants was used in the nitrite assay to determine the concentration of NO,- + N0 3 reduced to NO,-. Nitrate concentration expressed in micromolars is equal to the con-
Fig. 2. HAM-mediated fungistasis is not dependent on total hemoly1ic comple· ment activity. The mean fungistasis ± SEM is shown for three experiments. All experiments contain duplicate samples. Fungistasis was measured in the presence of fresh human serum (control), serum heated 30 min at 56° C, dialyzed serum, and serum that had been both heated and dialyzed. No total hemoly1ic complement activity was detected in heated serum or heated and dialyzed serum. Normal total hemoly1ic complement activity was found in the control serum and the dialyzed serum. There was no significant difference between fungistasis mediated in control serum and heated serum (p = 0.2, Student's t test), between control serum and dialyzed serum (p = 0.7), or between control serum and heated and dialyzed serum (p = 0.2).
5
4 lI) lI)
MIRIAM L. CAMERON, DONALD L. GRANGER, J. BRICE WEINBERG, WALTER J. KOZUMBO, and HILLEL S. KOREN Introduction SUMMARY Human alveolar macrophages (HAM) from 28 normal volunteers ware found to Inhibit
Macrophages play an important role in replication of Cryptococcus neoformans. Conditions under which funglstasls occurred ware differthe killing and inhibition of replication ent than those required for mouse peritoneal macrophage-mediated funglstasls. Inhibition of fungal of microorganisms and tumor cells. There replication by mouse peritoneal macrophages (MPM) requires that the macrophages are activated has been extensive work studying the oxyand that the cocultures of C. neoform.ns and macrophages be done In the presence of serum, gen-dependent mechanisms of host resisL-arginine, and endotoxin. During MPM-medlated funglstasls and tumor cell killing, L-arginine Is oxidized to NO,·, NO.·, and L-citrulllne. In addition, MPM have arginase activity that converts Ltance, but more recently interest has arginine to L-ornlthine and urea. turned to the role of L-arginine nitrogen HAM-mediated funglstasls was similar to that mediated by MPM In terms of the serum requireoxidation metabolism in microbiostasis ment, but HAM did not require L-arginine or endotoxin. HAM did not produce NO,· or NO.· detectable and tumor cell injury (1-3). L-arginineby colorimetric and bioassay, nor did HAM produce L-citrulllne or L-ornlthlne from "C-radlolabeled dependent mouse peritoneal macrophageL-arginine as detectable by reverse-phase Ion-pairing HPLC of macrophage-c. neoformans coculmediated microbiostasis and tumoristature supernatants. HAM had no detectable arginase actiVity, hence there was no evidence for Lsis has been studied. Stuehr and MarIetta arginine nitrogen metabolism In HAM. HAM-mediated funglstasls was not enhanced by endotoxin have shown that MPM infected in vivo or by recombinant human Interferon-y (rHIFN-y). The combination of endotoxin and rHIFN-y Inhibitby BCG produce N0 2 - and N0 3 - (4). ed the fungistatic effect of HAM. Human peritoneal macrophages (HPM) from women undergoing laparoacopy ware tested for funglstasls and L-arginine nitrogen oxidation. Partial Inhibition of crypHibbs and colleagues showed that actitococcal replication occurred; howaver, there was no evidence of L-arginine metabolism to NO,· vated MPM were cytotoxic to tumor cells or NO.·. The absence of L-arginlne-dependent nitrogen oxidation In HAM and HPM, compared to via an L-arginine-dependent pathway that MPM, during conditions under which funglstasls occurs suggests that this phenomenon Is species produces N02 - and L-citrulline (1, 2). specific rather than specific to the tissue origin of the macrophages. Further work by Granger and coworkers AM REV RESPIR DIS 1990; 142:1313-1319 showed that microbiostasis of Cryptococcus neojormans by MPM also occurred via the L-arginine-dependent mechanism fed a low-nitrate diet show that endogein original form April 9, 1990 and in and that this pathway and fungistasis nous nitrate biosynthesis occurs (8). (Received revised form June 13, 1990) could be specifically inhibited by the The site of endogenous nitrate biosynL-arginine analog N-guanidomonomethyl thesis was not known when the metabolFrom the Division of Infectious Diseases, Duke L-arginine (3). Although murine macro- ic studies were done, but nitric oxide syn- University Medical Center, and the Division of phages require L-arginine nitrogen oxida- thesis has been found in neutrophils (9), Hematology-Oncology, Veterans Administration tion to nitrite or nitrate to mediate tumor cells (to), hepatocytes (11), central and Duke University Medical Center, Durham; the Center. for Environmental Medicine and Lung Bimicrobiostasis or tumoristasis, neither ni- nervous system tissue (12), and endotheli- ology, University of North Carolina, Chapel Hill; trite or nitrate seemed to be the mediator al cells (13). Hence the possibility exist- and the Clinical Research Branch, Human Studies molecule of cytostasis (Granger, unpub- ed that human macrophages, in a manner Division, Health Effects Research Laboratory, Unitlished data and reference 5). Stuehr and similar to MPM, also utilized L-arginine ed States Environmental Protection Agency, Reltiangle Park, North Carolina. coworkers have shown that nitric oxide nitrogen oxidation to generate NO to me- search , Supported by Grant No. AI-26188, CA-41987, (NO) seems to be the important effector diate microbiostasis. Results of this study POI AI-23308, POI 32682, and P50 AR-39162 from molecule produced by murine macro- demonstrated that HAM and human the National Institutes of Health, by R.J. Reyphages during the L-arginine-dependent peritoneal macrophage (HPM) inhibit- nolds-Nabisco Industries, by the Veterans AdriUnisinhibition of tumor cell replication (5). In ed C neojormans replication in vitro in tration Research Service, and by Cooperative Agreement No. CR8 12738 from the Environmental Proaddition, they have shown that the meta- . an L-arginine-independent manner and tection Agency. bolic product of L-arginine oxidation by that the macrophages did not metabo3 The research described in this report has been murine macrophages has the same activ- lize L-arginine to nitrate or nitrite either reviewed by the Health Effects Research Laboraityas endothelial relaxation factor (6). in their basal state or after treatment with tory, U.S. Environmental Protection Agency, and The purpose of this study was to de- endotoxin or recombinant human inter- approved for publication. Approval does not signify that the contents necessarily reflect the views termine whether L-arginine oxidation feron-y (rHIFN-y). of the agency, nor does mention of trade names was required for human macrophageor commercial products constitute endorsement or recommendation for use. mediated fungistasis. Human alveolar Methods 4 Correspondence and requests for reprints should macrophages (HAM) inhibit fungal replibe addressed to Miriam L. Cameron, M.D., Box Cell Culture Media cation in vitro (7). In addition, metabol3524, Duke University Medical Center, Durham, ic studies in normal human volunteers RPMI 1640 (JR Scientific, Woodland, CA) NC 27710. 1
1313
1314
was used to prepare alveolar macrophages immediately after explantation; otherwis~, all experiments were performed in Dulbecco's modified Eagle medium (DMEM). The DMEM was prepared in the laboratory as previously described (3) from stock solutions. L-arginine in a final concentration of 500 J.1M was added to the medium at each experiment unless a comparison to arginine-free media was done. Several experiments were carried out in DMEM that was modified by excluding all amino acids and replacing them with either 500 J.1M L-asparagine or 500 J.1M L-citrulline as a source of the nitrogen required by C neojormans (14). Experiments in L-arginine-free media containing L-asparagine or L-citrulline were carried out with human serum that had been dialyzed (molecular weight cutoff 6,000 to 8,000 D) extensively against phosphatebuffered saline (PBS) to remove free amino acids. In a separate group of experiments, L-arginine was replaced by L-arginine analogs (Sigma Chemical Co., St. Louis, MO). All the L-arginine analogs were at a 500 J.1M final concentration except for poly-L-arginine, which was 5 J.1M, and Ala-gly-arginine vasopressin, which was 70 J.1M. In some experiments, I J.1g/ml of endotoxin (phenol extract of Escherichia coli O.l28:B12 from Sigma Chemical Co.) and/or 100 V/m1 of recombinant human interferon-y (generously donated by Genentech, Inc., San Francisco, CA) was added to the cultures.
Alveolar Macrophages Human alveolar macrophages (HAM) were obtained by bronchoscopy and bronchoalveolar lavage as previously described (15) from 28 normal volunteers aged 18 to 35 yr who gave informed consent under a protocol approved by the Investigational Review Board. The HAM were washed three times with RPMI 1640 and then resuspended in RPMI 1640 at a concentration of 1 x 1()6/ml. TIypan blue exclusion showed a mean viability of 850/0 (range 50 to 100%). Previous work has shown that 95% ofthe cells were maerophages by morphologic criteria (7). Two different treatments of the HAM occurred after their resuspension. A pilot experiment comparing the two treatments revealed no difference in HAM viability or functional activity. In most experiments the HAM were incubated overnight at 4° C in 10% fetal bovine serum (FBS) in RPMI 1640. The day after explantation, these cells were washed three times in PBS containing 1.2 mM CaC1 2 and 0.8 mM MgC1 2. The cells were then resuspended in DMEM without serum. The cells were plated into 16-mm Costar tissue culture wells (Costar, Cambridge, MA) at 0.5 x 10" cells/well and incubated in 5% CO 2 and 95% humidified air at 37° C for I h. Nonadherent cells were removed by gentle washing with PBS. To increase the density of the adherent cells, the wells were reseeded with another 0.5 x 1()6 cells/well, which were allowed to adhere under the same conditions for an hour. The remaining nonadherent cells were
CAMERON, GRANGER, WEINBERG, KOZUMBO, AND KOREN
gently washed off the plates three times with PBS. After the final wash, 0.5 ml DMEM with fresh-frozen human serum was added. Alternatively, HAM were plated into the 16-mm Costar culture wells the day of explantation. Into each well was placed 0.5 x 10" cells in RPMI 1640, and then the cells were incubated under the same conditions as before for 1 h. Nonadherent cells were washed off three times with PBS, and the media were replaced with 10% FBS and RPMI 1640. The macrophage monolayers were incubated overnight at 37° C in humidified 5% CO2 and 95% air. The day after explantation, the monolayers were washed three times with PBS and then 0.5 ml of the appropriate experimental medium was added.
Peritoneal Macrophages Several experiments were carried out using human peritoneal macrophages. Human peritoneal macrophages were obtained from women undergoing laparoscopy for evaluation of infertility (16). The explanted cells were stored overnight at 4° C in 10% autologous unheated serum in DMEM. The cells were centrifuged at 200 x g for 10 min and then the supernatant was removed. The pellet was resuspended in DMEM and layered onto Ficoll-Hypaque (Histopaque®; Sigma Chemical Co.) and centrifuged at 400 x g for 25 min. The interface that contained the HPM was removed, and the cells were washed three times with DMEM (200 x g for 10 min) and resuspended to a concentration of 2 x 10" cells/ml. A single seeding of 0.7 ml (1.4 x 10" cells/ml) was applied to 16-mm Costar tissue culture wells and allowed to adhere under the same conditions as for the HAM. Nonadherent cells were removed by washing with PBS three times, and the media were replaced with the experimental media (5% fresh human serum and DMEM). Morphologic studies have shown that> 95% of these cells were macrophages by Wright stain and nonspecific esterase (17). Cryptococcus neoformans A previously described (18), thinly encapsulated clone of C neojormans (serotype A) designated H99/C3D, which originated from the cerebrospinal fluid of a patient, was used in all fungistasis assays. The yeasts were maintained in DMEM suspension cultures in an environment of humidified 5% CO2and 95% air at 37° C. A fresh suspension of log-phase C neojormans was prepared by washing the yeast cells three times in PBS and then resuspending the washed yeasts in the modified DMEM. The resuspended yeasts were counted manually on a hemacytometer and electronically on a Coulter counter (19) and then diluted with media to a concentration of 1 x 105/ml before use in each assay.
Fungistasis Measurements The fungistasis assays were done using a modification of the methods of Granger and coworkers (19). After confluent macrophage monolayers and C neojormans were pre-
pared, 0.5 ml of the yeasts at a concentration of 1 x 105/ml was added to each of the macrophage monolayer wells (HAM-cryptococci, 10:1 to 20:1) and to control wells containing 0.5 ml medium alone. Macrophage controls were cultured in media without the addition of C neojormans. The macrophage-yeast cocultures, macrophage controls, and yeast controls were incubated in humidified 5% CO 2 and 95% air at 37° C for various time periods. The assay was terminated at 48 h except for the time course assays, for which various time points were determined. At the appropriate time, 0.1 ml of 10% sodium dodecyl sulfate (SDS; Sigma Chemical Co.) was added to each well to lyse the macrophages and terminate C neojormans replication (3). If the supernatants were to be assayed for N0 2- or N0 3-, 0.8 ml supernatant was removed before SDS treatment and replaced with 0.8 ml PBS. The supernatant was saved and processed as described subsequently. If quantitative plate counts were to be done, triplicate HAM-C neojormans cocultures and C neojormans controls were treated with 0.1 ml of 10% deoxycholate instead of SDS. After lysis of macrophage monolayers, 0.5 ml each of the lysates and controls was mixed in 10 ml of 10% hypochlorite in Isoton buffer (Curtin Matheson Scientific, Inc., Marietta, GA), and then they were counted electronically in a ZBI Coulter counter (Coulter Electronics, Inc., Hialeah, FL) in the manner previously described (16). All cultures except macrophage-alone controls were duplicates. The results of macrophage-cryptococci cocultures were expressed by two methods. First, cryptococcal cell counts versus time were plotted directly on semilogarithmic coordinates. Second, fungistasis was expressed as the difference between cryptococcal replication in the absence and presence of macrophages. This difference was expressed by the formula (20) Fungistasis = log2 (fungi-alone cultures) log, (macrophage + fungi cocultures) The doubling time of H99/C3D cryptococci in DMEM containing human serum is constant at about 2.5 h after an initial seeding of 5 x 104 yeast cells/ml.
Assessment oj Phagocytosis C neojormanswas cultured overnight at 5% CO" 37° C, in 5 ml DMEM containing 250 J.1Ci 35S (ICN Radiochemicals, Irvine, CA) (specific activity 771 J.1Ci/ml). Confluent macrophage monolayers were infected with 2 x 105 cryptococci/ml, and at the appropriate time points the cells were washed gently three times with PBS and then lysed with 0.5 ml of 0.5% SDS. To 10 ml Biofluor® (DuPont, Wilmington, DE) 0.5 mllysate was added, and this was counted in a liquid scintillation counter (19). Results were expressed as the percentage uptake of 35S-labeled C neojormans using the formula
1315
HUMAN MACROPHAGES INHIBIT CRYPTOCOCCAL REPLICATION INDEPENDENTLY OF L-ARGININE Fig. 1. HAM require serum to inhibit fungal replication. There were 9 experiments utilizing 25% serum and 20 experiments utilizing 5% serum. Three experiments utilizing no serum were run in parallel with serum-containing fungistasis assays. All experiments contained duplicate samples. Results are expressed as the mean fungistasis ± standard error of the mean (SEM). There is no significant difference in fungistasis between 5 and 25% serum assays, but there is a significant difference between fungistasis in the absence of serum compared to the 5% serum assays (Studenfs ttest, p < 0.0009). Cryptococci have a dOUbling time of 3.5 to 4.0 h in the absence of serum and 2 to 2.5 h in the presence of serum.
od that measures "CO, trapping after incubation of the samples with urease (22, 25).
5
~
4
< ..
3
-'" "c z3
2
en!!! _:I: en!
Statistical Analysis Most results for fungistasis and NO,- and N0 3 - are expressed as the mean of comparable experiments ± standard error of the mean (SEM). Statistical analysis was made using Student's t test to compare the mean of media controls versus supernatants for NO,- and N03 - or comparing fungistasis occurring under various treatments.
I- Z enS!
::::llli II. a: ~
0/0 35S-labeled cryptococci uptake = DPM sample/DPM cryptococci inoculum x 100
Assessment of Fungal Viability Colony-forming units (cfu) were determined by culturing dilutions of lysates from macrophage-cryptococci cocultures and C neoformans controls onto Sabouraud modified agar (Difco, Detroit, MI). Cultured lysates were incubated at 35° C in humidified air for 5 days. Colony-forming units were counted manually.
Nitrite Assay Nitrite concentration in the supernatants of macrophage-cryptococci cocultures control macrophage cultures, control cryptococci cultures, and media alone was measured as previously described using Greiss reagents (21). Concentration was calculated using a standard curve for NO,- concentration, which is linear in the range 10 to 200 IlM (22). The lower threshold of NO,- detection in this assay is ~ 5 IlM. Nitrite concentration is expressed in micromolars.
o
'V({«(J
0%
1({(frO
vcr//cd
5%
25%
Results
Inhibition of Fungal Replication by HAM HAM from normal volunteers required the presence of fresh human serum to inhibit C neoformans replication (figure 1). There was no difference in fungistasis in the presence of 5 or 25OJo human serum, but fungistasis was significantly less in cultures without serum compared to fungistasis in the presence of 5OJo serum (p < 0.0009, Student's t test). In the absence of serum, HAM remained viable but were unable to inhibit fungal replication. The serum requirement was not dependent on total hemolytic complement activity. Human serum heated at 56° C for 30 min, which was devoid of total hemolytic complement activity, and dialyzed serum, which maintained its total hemolytic complement activity, were able to support fungistasis (figure 2). There was a trend for decreased antifungal activity in the presence of heated serum. This effect appeared to be due to decreased phagocytosis of organisms as determined by visual inspection of cultures. Resident HAM inhibited cryptococcal replication in the absence of endotoxin or rHIFN-y. Neither endotoxin nor rHIFN-y
SERUM CONCENTRATION
centration of NO,- after bioreduction minus the concentration of NO,- before bioreduction. Nitrate standards were run to show that the N0 3 - was not reduced to ammonia.
Analysis of Amino Acid Metabolites of L-Arginine "C-radiolabeled L-arginine and its metabolites were assayed utilizing the previously described method of reversed-phase ion-pairing HPLC (22,24). 14C-radiolabeled L-arginine was added to the culture media before the fungistasis assay. The conditions of the assays used to analyze L-arginine metabolites were the same as those conditions used for fungistasis assays occurring without radiolabeled L-arginine. After separation of Larginine and its metabolites by reversed-phase ion-pairing HPLC, radiolabeled sample effluents were counted in a scintillation counter. Results were expressed as degenerations per minute per sample.
Analysis for Urea Urea was assayed for in supernatants from experiments containing 14C-radiolabeled Larginine utilizing a previously described meth-
Nitrate Assay Nitrate concentration in the media and supernatants was measured utilizing a bioassay using a bacterial nitrate reductase to convert N03 - to NO,-. The original method of Bartholomew (23) was modified by using Pseudomonas oleovorans instead of Escherichia coli as the source of the nitrate reductase (Granger, unpublished method). Briefly, 0.5 ml of stock P. oleovorans was diluted 1: 10 with modified DMEM, and then 0.15 ml of the diluted P. oleovorans was added to 0.3 ml supernatant or media samples. This was followed by the addition of 0.15 ml of 0.2 M inorganic sodium phosphate buffer, pH 7.25. The reaction mixture was incubated at 37° C in humidified air for 30 min. The samples were microfuged for 2 min, and then 0.4 ml supernatants was used in the nitrite assay to determine the concentration of NO,- + N0 3 reduced to NO,-. Nitrate concentration expressed in micromolars is equal to the con-
Fig. 2. HAM-mediated fungistasis is not dependent on total hemoly1ic comple· ment activity. The mean fungistasis ± SEM is shown for three experiments. All experiments contain duplicate samples. Fungistasis was measured in the presence of fresh human serum (control), serum heated 30 min at 56° C, dialyzed serum, and serum that had been both heated and dialyzed. No total hemoly1ic complement activity was detected in heated serum or heated and dialyzed serum. Normal total hemoly1ic complement activity was found in the control serum and the dialyzed serum. There was no significant difference between fungistasis mediated in control serum and heated serum (p = 0.2, Student's t test), between control serum and dialyzed serum (p = 0.7), or between control serum and heated and dialyzed serum (p = 0.2).
5
4 lI) lI)
