CELLULAR
IMMUNOLOGY
134,65-76
(199 I)
ln Vitro Production of Tumor Macrophages Stimulated of Candida
Necrosis Factor with Mannoprotein albicans
by Murine Splenic Constituents
Cell Wall
ANNA VECCHIARELLI,* MANUELA PULITI, * ANTONELLA TOROSAN’~LJWI,-F ANTONIOCASSONE,'.~ ANDFRANCESCOBISTONI*
Mannoprotein components from Chnclidu u/hw~~.s were investigated for their ability to induce production of tumor necrosis factor (TNF) by cultured splenocytes from naive or Candida-infected mice. Two chromatographically separated mannoproteins preparations, designated Fl and F2. were as able as the heat-inactivated Candida cells to induce the production ofTNF from splenocytes of naive animals. In addition. they caused a significant augmentation of basic TNF secretion by splenocytes ofcandida-infected animals. Experiments using plastic and/or nylon wool adherence. as well as treatments with antibodies depleting T or NK cells. consistently indicated that most If not all TNF was produced by splenic macrophages. In cultures of splenocytes from Candidainfected mice. mannoprotein addition also stimulated interferon-y (IF&) production by Thy I .Z positive cells. Depletion of these cells or addition of anti-IFN-7 antibodies abolished IFN production and reduced TNF secretion by adherent cells to the levels found in the cultures of mannoprotein-stimulated spleen cells from naive mice. These data add further evidence to the immunomodulatory properties possessed by some cell wall constituents ofthe human commensal microorganism (‘ a/hican.s and suggest that IFN-y is endowed with a regulatory role in TNF production by mouse macrophages in v//m c IYY I Academic Presr. Inc
INTRODUCTION
Cundida ulbicuns is a dimorphic fungus whose importance as opportunistic agent of human disease is increasingly considered (1, 2). The microorganism is usually harbored as harmless commensal in the gastrointestinal tract (3) but candidosis occurs in severely debilitated host undergoing aggressive chemotherapy or other immunosuppressive treatments or conditions, mostly affecting T cells or phagocytes (4-8). Recently. strong evidence has been obtained by our groups showing that c’. ulbicuns cells do not only prime the animal toward specific immune responses (9) but can modulate the overall activity of the immune system as well (10). In fact, injection of whole fungal cells or constituents of its cell wall conferred antitumor resistance in murine models ( I 1) and modulated NK activity ( 12- 14). Candida-mediated immunosuppressive activities on T cells ( 15, lb), immunoadjuvant ’ To whom
correspondence
should
he addressed.
FAX:
39.6.4’)
1723
65 OOW8749/9
I $3.00
CopyrIght Cc: 1991 by Academic Press. Inc. All rights ~Tr~productton tn any form rcwved
66
VECCHIARELLI
ET AL.
activity on B cells ( 17) and induction of LAK-like activity in human peripheral blood lymphocytes (18) have been reported. Hence, C. albicans cell or its cell wall extracts behave as powerful biological response modifier (BRM)‘. In previous studies, we showed that a low pathogenic strain of C. albicuns (PCA2) incapable of mycelial conversion is able to confer protection against a variety of microbial challenges ( 19). This resistance correlated with the establishment of a chronic infection (20), during which a massive production of several cytokines occurs ( 10). Now, we focused upon the immunomodulatory effects of mannoproteins fractions of C. afbicans, with the intent to ascertain whether these cell wall constituents could reproduce the effects exerted by the whole yeast cells. As previous data suggested that tumor necrosis factor (TNF) could play an important role in the establishment of PCA-2-induced antimicrobial state, (10) we particularly evaluated the effect of mannoproteins on TNF production by murine splenocytes. The results reported here clearly indicate that mannoproteins can indeed stimulate splenic macrophages from PCA-2infected or naive mice to produce elevated amounts of TNF. In addition, our data suggest that abundant TNF production by splenocytes of infected mice may require the production of interferon (@N--r). MATERIALS
AND METHODS
Animals Male hybrid (Balb/c Cr X DBA/2 Cr) F 1 (CD2F 1: H-2d/H-2d) mice, 14 to 16 weeks old, were obtained from Charles River Breeding Laboratories Inc. (Calco, Milan, Italy). Chemical and Reagents RPM1 1640 medium was purchased from Eurobio Labs, (Paris, France). Fetal bovine serum (FBS) was purchased from GIBCO (Grand Island, NY). Actinomycin-D was obtained from Sigma Chemicals Co. (St. Louis, MO). TNF-a! and rabbit anti-mouse TNF-a monoclonal antibodies were provided by Genzyme (Boston). The cloned mutine IFN-7 used in these studies was provided by Genentech Inc. (South San Francisco). Anti-mouse IFN--y monoclonal antibody was obtained as culture supernatant of R46A2 hybridoma cell line (American Type Culture Collection HB170). Monoclonal antibodies to the Thy 1.2 antigen were purchased from New England Nuclear Corp. (Boston) (Lot LK 104). Anti-Asia10 GM1 antiserum, previously shown to react selectively with mouse NK cells (2 l), was purchased from Wako Chemicals Gmbh (Dusseldorf, FRG). Yeasts The characteristic growth conditions of the two strains of C. albicans used have been described in detail elsewhere (20). The virulent strain CA-6 was isolated from a clinical specimen (22). The low pathogenic agerminative strain PCA-2 was an echino* Abbreviations used: TNF, tumor necrosis factor; IFN, interferon; NK, natural killer; PBMC, peripheral blood mononuclear cells; BRM, biological response modifier; PMN, polymorphonucleate cells; IL1 interleukin 1; C, complement; LAK, lymphokine-activated killer cells; SDS-PAGE, sodium dodecyl sulfate poliacrylamide gel electrophoresis.
TNF
INDUCTION
BY
CANDIDA
MANNOPROTEIN
candin-resistant mutant of the parental strain 3 153 (23). Heat inactivation (HCA) cells was performed at 12 1“C for 20 min.
Mannoprotein
67 of CA-6
Separation
The mannoprotein fractions of the cell wall of C. albicuns were obtained and analyzed by a combination of methods previously described by Cassone (24) and Torosantucci et al. (25,26). Briefly, washed yeast cells were autoclaved in a neutral buffer (140°C 2 hr) and the cell lysate was precipitated in cold acid ethanol, then extracted with a chloroform-butanol mixture. The aqueous phase was extensively dyalyzed against H20 and lyophilized (crude mannoprotein extract). This extract was subjected to ion-exchange chromatography using a 14 I X 2.6 cm column of DEAE-Sephadex A-50 (Pharmacia, Uppsala, Sweden) in 0.05 M Tris-HCl buffer, pH 7.5. 200 mg of the extract were loaded on the column and, after a prolonged wash with the equilibrating buffer to remove unbound material, retained fractions were eluted with two consecutive linear gradients of increasing NaCl concentrations (from 0.1 to 0.4 M and from 0.4 to 2 M, respectively). The eluted fractions were assayed for polysaccharide. protein, and phosphorus (25-28) content and absorbance at 260/280 nm: those corresponding to peaks of polysaccharide content were pooled, extensively dialyzed against water and lyophilized. The chromatographic fractions were dissolved at a concentration of 1 mg/ml in PBS, sterilized by filtration through a 0.45-pm filter, aliquoted, and kept frozen at -80°C. When needed, aliquots were thawed and diluted in complete medium (see below). All fractions were assayed with the limulus lysate gelihcation test for endotoxin detection.
Production of Culture Supernatants Containing Cytokine Activit~l Supernatants were produced as previously described ( 10, 29). Briefly, spleens were aseptically removed and splenocytes were resuspended in RPM1 1640 medium supplemented with 10% (v/v) FCS, 25 mM Hepes (N-2-hydroxyethylpiperazine-N-2-ethanesulphonic acid) buffer (Eurobio) and 0.1% (w/v) gentamicin sulphate, hereafter referred to as cRPM1. Red cells were osmotically lysed, spleen cells were washed three times, and then 5 X lo6 cells/ml were incubated at 37°C under 5% CO2 atmosphere in the presence of various doses of HCA or mannoprotein fractions in 25 cm3 tissue culture flasks (Falcon, CA) for 24 hr. Supernatants were harvested, filtered, and stored at -80°C until assayed.
Cell Fractionation
Procedures
Plastic adherence. Splenocytes (4 X 107), suspended in a volume of 10 ml of cRPM1 were incubated for 3 hr at 37°C in 5% COz atmosphere in 93-mm petri dishes (Nunc Inter Med, Roskilde, Denmark). After incubation, the dishes were extensively washed with cRPM1 to remove the nonadherent cells. The adherent cells were recovered by scraping with a rubber policeman, and then they were washed and suspended (viability 80 to 90%) in cRPMI. More than 98% of the recovered cells had the morphology of macrophages on Giemsa stain. Macrophages were then incubated and tested for cytokine production (see above). Nylon wool column. Plastic nonadherent cells were passed over a nylon fiber column as previously described (30).
68
VECCHIARELLI
ET
AL.
Treatment with anti-Thy 1.2 plus complement. Spleen cells (3 X 107) were preincubated with anti-Thy 1.2 antiserum diluted 1:lOO in cRPM1 for 30 min at room temperature, washed once in cRPMI, suspended in a 1:4 dilution of low-toxic-M rabbit complement (Cedarlane Laboratories, Hornby, Ontario, Canada), and incubated for 45 min at 37°C. The surviving cells were then washed twice, counted and stimulated for cytokine production as described above. In all experiments a complement control was performed in which the first incubation was in medium alone and the second was with complement. Cell recovery was between 50 and 60%. Treatment with anti-asialo GM, antiserum plus complement. Spleen cells were treated with a 1:200 dilution of antiserum in cRPM1 for 30 min at room temperature, washed twice, suspended in a 1:4 dilution of low-toxic-M rabbit complement and incubated for 1 hr at 37°C. The surviving cells were then washed, counted and stimulated for cytokine production as described above. Assay for TNF Activity This procedure has been previously described in detail (3 1). Briefly, TNF cytotoxic activity in supernatants was determined using actinomycin D-treated L-929 cells as targets. All determinations were made in comparison to commercially available preparations with known titers, and the results were expressed as U/ml. In selected experiments, anti TNF-a monoclonal antibodies were also added to the splenocytes cultures. The addition of HCA or mannoprotein fractions in the assay mixture for cytokine determination did not produce any effect. Assay for IFN Activity Interferon activity in supematants was measured as protection of mouse L-929 cells from the cytopathic effect of vesicular stomatitis virus as previously described (32). Selected samples were incubated with R46A2 monoclonal antibody before performing the assay for IFN-7 neutralization. Statistical Analysis Each experiment concerning cytokine assays was repeated three to five times, and the resulting data were treated statistically by the Student’s t test. RESULTS Separation and Identification
of Mannoproteins
A mannoprotein extract ( 18) was separated into three major, (Fl , F2, and F5) well resolved components by ion-exchange chromatography on DEAE-Sephadex and elution with NaCl gradient. Slightly more than 80% of the total polysaccharide material applied to the column was recovered into these fractions. Table 1 shows the chemical composition of the fractions obtained from two independent batches of the initial crude mannoprotein extract. Fl and F2 were essentially mannoprotein material as judged by the high mannose content and the amount of protein and phosphorus, in close analogy to previous reports by ourselves (24-26) and others (33, 34) and to the gross chemical composition of the initial crude extract (18). Fraction F5 was mostly composed of nucleoprotein material (25) and was used in this study only as “negative”
TNF
INDUCTION
BY
CANDIDA TABLE
Chemical Material Fl F2 F5
Composition
69
MANNOPROTEIN I
of Mannoprotein
Constitutent”
Total”
Mannan’
Promind
94.0-96.0 93.0-98.0 18.7-22.0
90.7-92.9 9 I .O-96.0 0.3-1.5
2.3-2.6 4.0-5.0 2.5-2.9
Phosphorus” 0.0 I 0.03 3.30
’ Values are reported as range of determinations performed on two different batches of chromatographically separated fractions. * As determined by the method of Dubois cl al. (27) and expressing the value as ‘5 of lyophilized material. ’ As determined by mannose content after gas-liquid chromatographic analysis (I 2. 25). and expressed as % of the total saccharide. ‘As determined by the method of Lowry n al. (35). ” As determined by the method of Chen (‘I (I/. (28) and expressed as c% of Iyophilized material. Only one batch was assayed.
control impurity. Fl and F2 were negative, while F5 was slightly positive in the limulus lysate gelification test for endotoxin detection. Both Fl and F2 contained high M,, polydisperse mannoprotein molecules as shown in SDS-PAGE 5-10% gradient gels and immunoblots with a polyclonal anti-Candida serum raised in rabbit to whole heat-inactivated cells of C. albicans (HCA), and with a monoclonal antibody (mAbAF1) recognizing an oligosaccharide (probably, an oligomannoside) epitope of Candida cell surface (25, 26). Minor batch to batch variations in the chemical composition of the crude extract and consequently its fractions, were irrelevant for their immunological activity (data not shown).
TNF Production by Splenocytes Treuted with Candida Materials The ability of inactivated intact cells of C. albicans (HCA) or cell wall mannoproteins to modulate TNF production by splenocytes from naive or CA-2-infected mice (3 days after infection) was assayed in several preliminary experiments in which each material was added to splenocyte cultures at doses ranging from 5 to 200 pg/ml. These experiments consistently indicated that the concentration of 25 pg/ml was the optimal one for TNF induction by mannoprotein fractions, and that the inducing activity of HCA was optimal at a ratio of 1: 1 between Candida cells and splenocytes. Table 2 reports the data of a typical experiment using optimal doses of candidal materials. Confirming previous reports (lo), the unstimulated splenocytes of PCA-2-infected animals produced an appreciable amount of TNF. The table also shows that, upon in vitro stimulation with Fl and F2 mannoprotein fractions, the amount of TNF produced by the splenocytes augmented (about 3-4 times) reaching a level comparable to that obtained following splenocyte exposure to whole Candida cells. Splenocytes from naive animals, whose basic TNF production was negligible, produced appreciable quantities of TNF when cultured in the presence of mannoprotein, similar to those of splenocytes from PCA-2-infected mice not treated with candidal materials (Table 2). In all cases, the addition of anti-TNF-antibodies in the assay abolished (
IMMUNOLOGY
134,65-76
(199 I)
ln Vitro Production of Tumor Macrophages Stimulated of Candida
Necrosis Factor with Mannoprotein albicans
by Murine Splenic Constituents
Cell Wall
ANNA VECCHIARELLI,* MANUELA PULITI, * ANTONELLA TOROSAN’~LJWI,-F ANTONIOCASSONE,'.~ ANDFRANCESCOBISTONI*
Mannoprotein components from Chnclidu u/hw~~.s were investigated for their ability to induce production of tumor necrosis factor (TNF) by cultured splenocytes from naive or Candida-infected mice. Two chromatographically separated mannoproteins preparations, designated Fl and F2. were as able as the heat-inactivated Candida cells to induce the production ofTNF from splenocytes of naive animals. In addition. they caused a significant augmentation of basic TNF secretion by splenocytes ofcandida-infected animals. Experiments using plastic and/or nylon wool adherence. as well as treatments with antibodies depleting T or NK cells. consistently indicated that most If not all TNF was produced by splenic macrophages. In cultures of splenocytes from Candidainfected mice. mannoprotein addition also stimulated interferon-y (IF&) production by Thy I .Z positive cells. Depletion of these cells or addition of anti-IFN-7 antibodies abolished IFN production and reduced TNF secretion by adherent cells to the levels found in the cultures of mannoprotein-stimulated spleen cells from naive mice. These data add further evidence to the immunomodulatory properties possessed by some cell wall constituents ofthe human commensal microorganism (‘ a/hican.s and suggest that IFN-y is endowed with a regulatory role in TNF production by mouse macrophages in v//m c IYY I Academic Presr. Inc
INTRODUCTION
Cundida ulbicuns is a dimorphic fungus whose importance as opportunistic agent of human disease is increasingly considered (1, 2). The microorganism is usually harbored as harmless commensal in the gastrointestinal tract (3) but candidosis occurs in severely debilitated host undergoing aggressive chemotherapy or other immunosuppressive treatments or conditions, mostly affecting T cells or phagocytes (4-8). Recently. strong evidence has been obtained by our groups showing that c’. ulbicuns cells do not only prime the animal toward specific immune responses (9) but can modulate the overall activity of the immune system as well (10). In fact, injection of whole fungal cells or constituents of its cell wall conferred antitumor resistance in murine models ( I 1) and modulated NK activity ( 12- 14). Candida-mediated immunosuppressive activities on T cells ( 15, lb), immunoadjuvant ’ To whom
correspondence
should
he addressed.
FAX:
39.6.4’)
1723
65 OOW8749/9
I $3.00
CopyrIght Cc: 1991 by Academic Press. Inc. All rights ~Tr~productton tn any form rcwved
66
VECCHIARELLI
ET AL.
activity on B cells ( 17) and induction of LAK-like activity in human peripheral blood lymphocytes (18) have been reported. Hence, C. albicans cell or its cell wall extracts behave as powerful biological response modifier (BRM)‘. In previous studies, we showed that a low pathogenic strain of C. albicuns (PCA2) incapable of mycelial conversion is able to confer protection against a variety of microbial challenges ( 19). This resistance correlated with the establishment of a chronic infection (20), during which a massive production of several cytokines occurs ( 10). Now, we focused upon the immunomodulatory effects of mannoproteins fractions of C. afbicans, with the intent to ascertain whether these cell wall constituents could reproduce the effects exerted by the whole yeast cells. As previous data suggested that tumor necrosis factor (TNF) could play an important role in the establishment of PCA-2-induced antimicrobial state, (10) we particularly evaluated the effect of mannoproteins on TNF production by murine splenocytes. The results reported here clearly indicate that mannoproteins can indeed stimulate splenic macrophages from PCA-2infected or naive mice to produce elevated amounts of TNF. In addition, our data suggest that abundant TNF production by splenocytes of infected mice may require the production of interferon (@N--r). MATERIALS
AND METHODS
Animals Male hybrid (Balb/c Cr X DBA/2 Cr) F 1 (CD2F 1: H-2d/H-2d) mice, 14 to 16 weeks old, were obtained from Charles River Breeding Laboratories Inc. (Calco, Milan, Italy). Chemical and Reagents RPM1 1640 medium was purchased from Eurobio Labs, (Paris, France). Fetal bovine serum (FBS) was purchased from GIBCO (Grand Island, NY). Actinomycin-D was obtained from Sigma Chemicals Co. (St. Louis, MO). TNF-a! and rabbit anti-mouse TNF-a monoclonal antibodies were provided by Genzyme (Boston). The cloned mutine IFN-7 used in these studies was provided by Genentech Inc. (South San Francisco). Anti-mouse IFN--y monoclonal antibody was obtained as culture supernatant of R46A2 hybridoma cell line (American Type Culture Collection HB170). Monoclonal antibodies to the Thy 1.2 antigen were purchased from New England Nuclear Corp. (Boston) (Lot LK 104). Anti-Asia10 GM1 antiserum, previously shown to react selectively with mouse NK cells (2 l), was purchased from Wako Chemicals Gmbh (Dusseldorf, FRG). Yeasts The characteristic growth conditions of the two strains of C. albicans used have been described in detail elsewhere (20). The virulent strain CA-6 was isolated from a clinical specimen (22). The low pathogenic agerminative strain PCA-2 was an echino* Abbreviations used: TNF, tumor necrosis factor; IFN, interferon; NK, natural killer; PBMC, peripheral blood mononuclear cells; BRM, biological response modifier; PMN, polymorphonucleate cells; IL1 interleukin 1; C, complement; LAK, lymphokine-activated killer cells; SDS-PAGE, sodium dodecyl sulfate poliacrylamide gel electrophoresis.
TNF
INDUCTION
BY
CANDIDA
MANNOPROTEIN
candin-resistant mutant of the parental strain 3 153 (23). Heat inactivation (HCA) cells was performed at 12 1“C for 20 min.
Mannoprotein
67 of CA-6
Separation
The mannoprotein fractions of the cell wall of C. albicuns were obtained and analyzed by a combination of methods previously described by Cassone (24) and Torosantucci et al. (25,26). Briefly, washed yeast cells were autoclaved in a neutral buffer (140°C 2 hr) and the cell lysate was precipitated in cold acid ethanol, then extracted with a chloroform-butanol mixture. The aqueous phase was extensively dyalyzed against H20 and lyophilized (crude mannoprotein extract). This extract was subjected to ion-exchange chromatography using a 14 I X 2.6 cm column of DEAE-Sephadex A-50 (Pharmacia, Uppsala, Sweden) in 0.05 M Tris-HCl buffer, pH 7.5. 200 mg of the extract were loaded on the column and, after a prolonged wash with the equilibrating buffer to remove unbound material, retained fractions were eluted with two consecutive linear gradients of increasing NaCl concentrations (from 0.1 to 0.4 M and from 0.4 to 2 M, respectively). The eluted fractions were assayed for polysaccharide. protein, and phosphorus (25-28) content and absorbance at 260/280 nm: those corresponding to peaks of polysaccharide content were pooled, extensively dialyzed against water and lyophilized. The chromatographic fractions were dissolved at a concentration of 1 mg/ml in PBS, sterilized by filtration through a 0.45-pm filter, aliquoted, and kept frozen at -80°C. When needed, aliquots were thawed and diluted in complete medium (see below). All fractions were assayed with the limulus lysate gelihcation test for endotoxin detection.
Production of Culture Supernatants Containing Cytokine Activit~l Supernatants were produced as previously described ( 10, 29). Briefly, spleens were aseptically removed and splenocytes were resuspended in RPM1 1640 medium supplemented with 10% (v/v) FCS, 25 mM Hepes (N-2-hydroxyethylpiperazine-N-2-ethanesulphonic acid) buffer (Eurobio) and 0.1% (w/v) gentamicin sulphate, hereafter referred to as cRPM1. Red cells were osmotically lysed, spleen cells were washed three times, and then 5 X lo6 cells/ml were incubated at 37°C under 5% CO2 atmosphere in the presence of various doses of HCA or mannoprotein fractions in 25 cm3 tissue culture flasks (Falcon, CA) for 24 hr. Supernatants were harvested, filtered, and stored at -80°C until assayed.
Cell Fractionation
Procedures
Plastic adherence. Splenocytes (4 X 107), suspended in a volume of 10 ml of cRPM1 were incubated for 3 hr at 37°C in 5% COz atmosphere in 93-mm petri dishes (Nunc Inter Med, Roskilde, Denmark). After incubation, the dishes were extensively washed with cRPM1 to remove the nonadherent cells. The adherent cells were recovered by scraping with a rubber policeman, and then they were washed and suspended (viability 80 to 90%) in cRPMI. More than 98% of the recovered cells had the morphology of macrophages on Giemsa stain. Macrophages were then incubated and tested for cytokine production (see above). Nylon wool column. Plastic nonadherent cells were passed over a nylon fiber column as previously described (30).
68
VECCHIARELLI
ET
AL.
Treatment with anti-Thy 1.2 plus complement. Spleen cells (3 X 107) were preincubated with anti-Thy 1.2 antiserum diluted 1:lOO in cRPM1 for 30 min at room temperature, washed once in cRPMI, suspended in a 1:4 dilution of low-toxic-M rabbit complement (Cedarlane Laboratories, Hornby, Ontario, Canada), and incubated for 45 min at 37°C. The surviving cells were then washed twice, counted and stimulated for cytokine production as described above. In all experiments a complement control was performed in which the first incubation was in medium alone and the second was with complement. Cell recovery was between 50 and 60%. Treatment with anti-asialo GM, antiserum plus complement. Spleen cells were treated with a 1:200 dilution of antiserum in cRPM1 for 30 min at room temperature, washed twice, suspended in a 1:4 dilution of low-toxic-M rabbit complement and incubated for 1 hr at 37°C. The surviving cells were then washed, counted and stimulated for cytokine production as described above. Assay for TNF Activity This procedure has been previously described in detail (3 1). Briefly, TNF cytotoxic activity in supernatants was determined using actinomycin D-treated L-929 cells as targets. All determinations were made in comparison to commercially available preparations with known titers, and the results were expressed as U/ml. In selected experiments, anti TNF-a monoclonal antibodies were also added to the splenocytes cultures. The addition of HCA or mannoprotein fractions in the assay mixture for cytokine determination did not produce any effect. Assay for IFN Activity Interferon activity in supematants was measured as protection of mouse L-929 cells from the cytopathic effect of vesicular stomatitis virus as previously described (32). Selected samples were incubated with R46A2 monoclonal antibody before performing the assay for IFN-7 neutralization. Statistical Analysis Each experiment concerning cytokine assays was repeated three to five times, and the resulting data were treated statistically by the Student’s t test. RESULTS Separation and Identification
of Mannoproteins
A mannoprotein extract ( 18) was separated into three major, (Fl , F2, and F5) well resolved components by ion-exchange chromatography on DEAE-Sephadex and elution with NaCl gradient. Slightly more than 80% of the total polysaccharide material applied to the column was recovered into these fractions. Table 1 shows the chemical composition of the fractions obtained from two independent batches of the initial crude mannoprotein extract. Fl and F2 were essentially mannoprotein material as judged by the high mannose content and the amount of protein and phosphorus, in close analogy to previous reports by ourselves (24-26) and others (33, 34) and to the gross chemical composition of the initial crude extract (18). Fraction F5 was mostly composed of nucleoprotein material (25) and was used in this study only as “negative”
TNF
INDUCTION
BY
CANDIDA TABLE
Chemical Material Fl F2 F5
Composition
69
MANNOPROTEIN I
of Mannoprotein
Constitutent”
Total”
Mannan’
Promind
94.0-96.0 93.0-98.0 18.7-22.0
90.7-92.9 9 I .O-96.0 0.3-1.5
2.3-2.6 4.0-5.0 2.5-2.9
Phosphorus” 0.0 I 0.03 3.30
’ Values are reported as range of determinations performed on two different batches of chromatographically separated fractions. * As determined by the method of Dubois cl al. (27) and expressing the value as ‘5 of lyophilized material. ’ As determined by mannose content after gas-liquid chromatographic analysis (I 2. 25). and expressed as % of the total saccharide. ‘As determined by the method of Lowry n al. (35). ” As determined by the method of Chen (‘I (I/. (28) and expressed as c% of Iyophilized material. Only one batch was assayed.
control impurity. Fl and F2 were negative, while F5 was slightly positive in the limulus lysate gelification test for endotoxin detection. Both Fl and F2 contained high M,, polydisperse mannoprotein molecules as shown in SDS-PAGE 5-10% gradient gels and immunoblots with a polyclonal anti-Candida serum raised in rabbit to whole heat-inactivated cells of C. albicans (HCA), and with a monoclonal antibody (mAbAF1) recognizing an oligosaccharide (probably, an oligomannoside) epitope of Candida cell surface (25, 26). Minor batch to batch variations in the chemical composition of the crude extract and consequently its fractions, were irrelevant for their immunological activity (data not shown).
TNF Production by Splenocytes Treuted with Candida Materials The ability of inactivated intact cells of C. albicans (HCA) or cell wall mannoproteins to modulate TNF production by splenocytes from naive or CA-2-infected mice (3 days after infection) was assayed in several preliminary experiments in which each material was added to splenocyte cultures at doses ranging from 5 to 200 pg/ml. These experiments consistently indicated that the concentration of 25 pg/ml was the optimal one for TNF induction by mannoprotein fractions, and that the inducing activity of HCA was optimal at a ratio of 1: 1 between Candida cells and splenocytes. Table 2 reports the data of a typical experiment using optimal doses of candidal materials. Confirming previous reports (lo), the unstimulated splenocytes of PCA-2-infected animals produced an appreciable amount of TNF. The table also shows that, upon in vitro stimulation with Fl and F2 mannoprotein fractions, the amount of TNF produced by the splenocytes augmented (about 3-4 times) reaching a level comparable to that obtained following splenocyte exposure to whole Candida cells. Splenocytes from naive animals, whose basic TNF production was negligible, produced appreciable quantities of TNF when cultured in the presence of mannoprotein, similar to those of splenocytes from PCA-2-infected mice not treated with candidal materials (Table 2). In all cases, the addition of anti-TNF-antibodies in the assay abolished (
