A monoclonal antibody to Candida albicans enhances mouse neutrophil candidacidal activity.

A monoclonal antibody (MAb) to Candida albicans (MAb B6.1) that protects against candidiasis and the nonprotective MAb B6 were compared for ability to support neutrophil (polymorphonuclear leukocyte [PMN]) candidacidal activity. Both MAbs are immunoglobulin M, and each recognizes distinct C. albicans mannan cell wall determinants. PMN candidacidal activity was assessed by transmission electron microscopy and by an in vitro killing assay. The results indicated that MAb B6.1, but not MAb B6, enhances ingestion and killing of yeast cells by PMN in the presence of serum complement.     

Interferon-gamma activates the oxidative killing of Candida albicans by human granulocytes.

Although granulocytes are essential for the resistance against infections with Candida albicans, these cells do not kill the ingested yeast optimally. Various cytokines can enhance functional activities of granulocytes, but until now only interferon-gamma (IFN-gamma) has been applied more widely, namely in patients with chronic granulomatous disease. Since it is not certain whether IFN-gamma is able to enhance the candidacidal activity of granulocytes the present study was undertaken. Human granulocytes incubated with various concentrations of recombinant human IFN-gamma (rIFN-gamma) were studied for the phagocytosis and intracellular killing of C. albicans and their oxygen metabolism after stimulation with opsonized Candida. Results showed a small increase in the rate of phagocytosis and a dose-dependent increase of the intracellular killing of C. albicans and the production of H2O2. The increased candidacidal activity and H2O2 production by rIFN-gamma-stimulated granulocytes were inhibited by diphenylene iodonium (DPI). From these results it is concluded that the increased candidacidal activity of granulocytes activated by rIFN-gamma is caused by the increased production of reactive oxygen radicals.     

Effect of mouse phagocytes on Candida albicans in in vivo chambers.

Mouse phagocytic cells in in vivo diffusion chambers had either candidacidal or candidastatic activity depending on the kind of phagocyte studied, the type of mouse whence the phagocytes came, and whether the chamber inoculum of Candida albicans consisted primarily of yeasts or hyphae. Killing of C. albicans occurred when yeasts were placed into chambers with membranes with 3.0-micrometer pores and implanted intraperitoneally into normal mice or thymus-deficient (nude) mice. Although C. albicans remains in chambers with 3.0-micrometer pores, host phagocytic cells can migrate into the chambers. Killing also occurred when yeasts were combined with normal or nude mouse neutrophils in chambers made with membranes with 0.45-micrometer pores, which restrict migration of host cells, but not diffusion of soluble factors. Populations of cells rich in macrophages were candidastatic for yeasts when the phagocytes came from normal mice but candidacidal when obtained from nude mice. Results of gradient fractionations of peritoneal exudates indicated that more than one cell type may be responsible for candidacidal activity by nude mouse macrophage-rich cells. Hyphal-phase cells of C. albicans appear to be more resistant than yeast-phase cells to killing by normal and nude mouse phagocytic cells.     

Candidacidal mechanisms of peritoneal macrophages activated with lymphokines or gamma-interferon

The mechanisms by which resident peritoneal macrophages, activated in vitro by lymphokines (LK) or recombinant gamma-interferon (IFN), kill Candida parapsilosis or C. albicans were studied. Resident non-activated peritoneal macrophages killed C. parapsilosis (55.5% SD 6.8%), but not C. albicans. This killing was completely inhibited by superoxide dismutase (SOD), partially by dimethyl sulphoxide (DMSO), but not by catalase or azide. Killing correlated with a brisk lucigenin-dependent chemiluminescence (CL) response by macrophages interacting with C. parapsilosis. No enhanced luminol-dependent CL response was observed in this system. This suggests that C. parapsilosis is killed by resident macrophages via a mechanism dependent on the presence of superoxide anion. By contrast, killing of C. parapsilosis by activated macrophages (49.0% SD 5.9%) was not inhibited by SOD or DMSO, suggesting the induction of a non-oxidative candidacidal mechanism. C. albicans was killed only by macrophages activated with IFN (52.0% SD 3.7%) or LK (55.7% SD 2.8%). Inhibition of killing by SOD was greater in IFN- than in LK-activated macrophages. Conversely, killing by LK-, but not IFN-, activated macrophages was significantly inhibited by catalase, DMSO or azide. The killing by LK-activated macrophages, and its inhibition by scavengers, correlated with the luminol-dependent CL response. The non-killing resident macrophages interacting with C. albicans made lucigenin-dependent CL responses similar to those of activated macrophages. The mechanisms enabling killing of C. albicans induced by activation appear to be different for LK and IFN, and appear to depend upon the myeloperoxidase systems and superoxide respectively.     

Modulation of anti-Candida activity of human alveolar macrophages by interferon-gamma or interleukin-1-alpha

The fungicidal and bactericidal activities of human alveolar macrophages (AM) and peripheral blood monocytes (PBM) from 18 healthy volunteers were evaluated. The results showed that AM were able to phagocytize and kill Candida albicans, Pseudomonas aeruginosa, and Staphylococcus aureus. However, killing of the bacteria was already complete in 2 h, whereas killing of Candida required 4 to 6 h despite an early phagocytosis of yeast cells. The fungicidal activity of freshly collected AM and PBM was also tested after effector cell exposure to interferon-gamma (IFN-gamma), interleukin-1-alpha (IL-1 alpha), endotoxin lipopolysaccharide (LPS), or interleukin 2 (IL-2). It was found that treatment with IFN-gamma, IL-1 alpha, or LPS significantly augmented macrophage and PBM candidacidal activity, whereas the addition of IL-2 was ineffective. We also evaluated killing of C. albicans by AM cultured in vitro for different times. While phagocytosis was apparently unaffected, the candidacidal activity progressively decreased over the in vitro culture period, an effect that was largely reversed by cell exposure to IFN-gamma, IL-1 alpha, or LPS. In an experimental model in which mice infected with an agerminative C. albicans strain (PCA-2) resisted lethal microbial challenge, freshly harvested AM showed increased cytotoxic activity to Aspergillus fumigatus in vitro as well as enhanced IL-1 production. In conclusion, present data confirm the crucial role of AM in the surveillance of bacterial and fungal infections and indicate that treatment of these cells with IFN-gamma or IL-1 alpha is able to enhance their antimicrobial capability.     

Candidacidal activity of myeloperoxidase: mechanisms of inhibitory influence of soluble cell wall mannan

We have previously demonstrated the ability of human neutrophil myeloperoxidase to bind to mannan isolated from Candida albicans. Mannan may therefore be a primary component of the yeast cell wall which provides for binding of myeloperoxidase, a requirement potentially important for the candidacidal activity of the enzyme. In this report, we describe experiments to consider the relationship of the mannan-binding activity of myeloperoxidase to its candidacidal activity and the possibility that free mannan may inhibit myeloperoxidase-mediated candidacidal activity. We observed that binding of myeloperoxidase to the target yeasts was required for killing of C. albicans. We also observed that addition of soluble mannan significantly reduced myeloperoxidase-mediated killing of the yeasts in a dose-dependent manner by antagonizing binding of myeloperoxidase. Soluble mannan was demonstrated to have a similar dose-dependent inhibitory effect on neutrophil-mediated candidacidal activity without influencing phagocytosis of the organism. On the basis of these observations, we speculate that mannan solubilized in plasma and tissue fluid may interfere with neutrophil-mediated host defense against Candida infection.     

Poly(I.C)-induced interferons enhance susceptibility of SCID mice to systemic candidiasis.

In the absence of any demonstrable T- or B-cell responses, gnotobiotic CB-17 SCID (severe combined immunodeficient) mice not only show innate resistance to acute systemic (intravenous challenge) candidiasis but also manifest innate resistance to systemic candidiasis of endogenous (gastrointestinal tract) origin. Poly(I. C), a potent inducer of interferons (IFNs) in vivo, enhanced the susceptibility of CB-17 SCID mice to acute systemic candidiasis and to systemic candidiasis of endogenous origin, as demonstrated by increased numbers of viable Candida albicans in internal organs after poly(I. C) treatment. The poly(I. C)-enhanced susceptibility of mice to candidiasis was abrogated by in vivo treatment with antibodies to IFN-alpha, -beta, and -gamma. In vivo depletion of natural killer cells from SCID mice did not significantly enhance their susceptibility to systemic candidiasis or abrogate poly(I. C)-enhanced susceptibility. In vivo and in vitro, treatment with poly(I. C) impaired the candidacidal and phagocytic activity of thioglycollate-elicited macrophages from SCID mice. Antibody to IFN-alpha/beta or IFN-beta alone interfered with the ability of poly(I. C) to impair the candidacidal activity of macrophages from SCID mice in vitro. These data suggest that poly(I. C)-induced interferons can impair the candidacidal activity of macrophages in SCID mice and decrease their innate resistance to acute systemic candidiasis and to systemic candidiasis of endogenous origin.     

The Role of BCG/PPD-Activated Macrophages in Resistance against Systemic Candidiasis in Mice

The main conclusions of this study are that BCG/PPD-activated macrophages, in contrast to macrophages from control mice, exhibit an increased PMA-induced production of H2O2, kill about one-third of the phagocytosed Candida albicans, and cause more than 50% inhibition of the intracellular formation of germ tubes by C. albicans. Peritoneal macrophages from mice that were colonized post-natally with C. albicans do not show increased production of H2O2 upon stimulation with PMA and the intracellular outgrowth of germ tubes is inhibited to only a limited degree. These macrophages are capable of killing about 20% of the ingested C. albicans. In vivo, the number of Candida in the kidney, spleen and liver after intravenous injection of Candida albicans is significantly lower in BCG-treated mice than in control mice. Post-natal colonization with C. albicans has only a limited effect on the outgrowth of intravenously injected C. albicans in the spleen and liver but does not influence growth in the kidney. These results indicate that acquired immunity against a systemic Candida infection involves both oxidative and non-oxidative mechanisms of intracellular killing and that these mechanisms may have different effects on the yeast and hyphal forms of C. albicans.     

Suppression of type 2 NO-synthase activity in macrophages by Candida albicans

Macrophages (Mphi) are important for the defence against experimental disseminated candidiasis. Nitric oxide (NO) generated by the inducible isoform of NO-synthase (iNOS or NOS2) is thought to contribute to candidacidal effector functions by activated Mphi. In vitro, however, Mphi cannot control the growth and hyphal formation of Candida (C.) albicans. Using mouse peritoneal exudate Mphi stimulated with IFN-gamma and LPS, we examined the effect of C. albicans on NO synthesis, NOS2 enzyme activity and macrophage survival. C. albicans effectively inhibited the production of NO via suppression of total NOS2 protein and enzyme activity. Hyphal formation of C. albicans and direct interaction with host cells was required for maximum inhibition of NO production, whereas non-filamentous C. albicans mutants released soluble products that effected only partial inhibition. Ultimately, Mphi underwent apoptotic cell death after infection with C. albicans wild-type strains capable of hyphal formation, indicated by loss of the mitochondrial membrane potential and onset of chromatin degradation. NO suppression and Mphi killing are potent activities of C. albicans that may augment virulence of C. albicans.     

Mechanism for candidacidal activity in macrophages activated by recombinant gamma interferon.

Candidacidal activity in macrophages activated by recombinant gamma interferon was examined kinetically in relation to acidification of phagolysosomes. In resident peritoneal macrophages (PMPs) of BALB/c mice, enhanced killing activity against Candida albicans was demonstrated after incubation with 100 U of gamma interferon per ml for 24 h but not after incubation for 48 to 72 h. Conversely, increased generation of H2O2 was exhibited in PMPs incubated from 48 to 72 h but not in PMPs incubated for 24 h. In normal PMPs, fusion of lysosomes to candida-containing phagosomes was readily accomplished and phagosome-lysosome fusion was not enhanced further by activation. The candidacidal substance was extracted from granule-rich fractions of either normal or activated PMPs by using citric acid (pH 2.7) in equal amounts; the substance showed a noncationic, heat-stable protein nature. In addition, when phagolysosomal pH was determined by flow cytometry of intraphagolysosomal fluorescein isothiocyanate-labeled C. albicans, phagolysosomes with low pH (less than 4.0) were detected in about 40% of PMPs activated for 24 h but not in those activated for 72 h or in normal PMPs. Moreover, increasing the intralysosomal pH with NH4Cl resulted in a significant reduction of candidacidal activity in activated PMPs. These results indicate that the candidacidal activity of gamma interferon-activated PMPs correlates well with enhanced acidification of their phagolysosomes and suggest that the candidacidal activity of activated PMPs is independent from reactive oxygen molecules and is mediated by proteinaceous substance(s) generated only in a strong acidic milieu of phagolysosomes by activation.     

Nitric oxide production does not directly increase macrophage candidacidal activity.

Some activated murine macrophages produced nitrite but were unable to kill Candida albicans. Furthermore, a nitric oxide (NO) generator inhibited C. albicans growth but was not candidacidal. Our results suggest that NO is candidastatic and that No is not directly involved but is associated with or induces other macrophage candidacidal mechanisms.     

CD40/CD40 ligand interactions in the host defense against disseminated Candida albicans infection: the role of macrophage-derived nitric oxide

CD40L interaction with CD40 is required for normal cellular immune responses such as T cell-mediated activation of monocytes/macrophages, proinflammatory cytokine production, and leukocyte extravasation. We investigated the role of CD40/CD40 ligand (L) interactions during disseminated candidiasis in CD40L knockout (CD40L-/-) mice. While early during infection there were no differences in the Candida albicans outgrowth in the organs of wild-type and knockout mice, the CD40L-/- mice had a significantly increased yeast load in the kidneys compared to CD40L+/+ mice late during infection. Similar effects were observed in CD40L+/+ mice in which CD40 ligation was blocked by a neutralizing anti-CD40 antibody. The peak TNF-alpha plasma concentrations were significantly lower in the CD40L-/- mice than in CD40L+/+ mice. C. albicans-stimulated production of nitric oxide (NO) by peritoneal macrophages from CD40L-/- in vitro was significantly lower than that of control mice, and this was responsible for a reduced candidacidal activity of CD40L-/- macrophages. The role of endogenous NO synthesis induced by CD40 ligation for the defense against disseminated candidiasis was further demonstrated by the absence of these effects in knockout mice deficient in inducible NO-synthase. In conclusion, absence of CD40/CD40L interactions results in increased susceptibility to disseminated infection with C. albicans through decreased NO-dependent killing of Candida by macrophages.     

Differential role of MyD88 in macrophage-mediated responses to opportunistic fungal pathogens

Toll-like receptors mediate macrophage recognition of microbial ligands, inducing expression of microbicidal molecules and cytokines via the adapter protein MyD88. We investigated the role of MyD88 in regulating murine macrophage responses to a pathogenic yeast (Candida albicans) and mold (Aspergillus fumigatus). Macrophages derived from bone marrow of MyD88-deficient mice (MyD88(-/-)) demonstrated impaired phagocytosis and intracellular killing of C. albicans compared to wild-type (MyD88(+/+)) macrophages. In contrast, ingestion and killing of A. fumigatus conidia was MyD88 independent. Cytokine production by MyD88(-/-) macrophages in response to C. albicans yeasts and hyphae was substantially decreased, but responses to A. fumigatus hyphae were preserved. These results provide evidence that MyD88 signaling is involved in phagocytosis and killing of live C. albicans, but not A. fumigatus. The differential role of MyD88 may represent one mechanism by which macrophages regulate innate responses specific to different pathogenic fungi.     

Candidacidal activity of mouse macrophages in vitro.

Mouse peritoneal macrophages were infected in vitro with Candida albicans, and the phagocytic and candidacidal activities were estimated by microscopic examination of Giemsa-stained cells. Activated macrophages obtained from either BCG-vaccinated animals or by in vitro exposure of normal macrophages to phytohemagglutinin-induced lymphokines exhibited higher phagocytic and candidacidal activities than did normal macrophages. However, activated macrophages obtained by in vitro exposure of macrophages to candida-induced lymphokines exhibited the highest phagocytic and candidacidal activities. The incorporation of immune mouse serum into the culture medium also enhanced the phagocytic and candidacidal activities of the normal macrophages but failed to improve the function of the activated macrophages. These results suggest that both activated macrophages and antibodies may be required for controlling candida infections in mice.     

Peroxynitrite contributes to the candidacidal activity of nitric oxide-producing macrophages.

Nitric oxide (NO) is associated with functions as diverse as peristalsis, blood flow, neuroendosecretion, visual transduction, smooth muscle relaxation, and microbial killing (H. H. W. H. Schmidt and V. Walter, Cell 78:919-925, 1994). Despite the well-established role of NO in macrophage candidacidal activity (E. Cenci, L. Romani, A. Mancacci, R. Spaccapelo, E. Schiaffella, P. Puccetti, and F. Bistoni, Eur. J. Immunol. 23:1034-1038, 1993; J. Jones-Carson, A. Vazquez-Torres, H. Van der Heide, R. D. Wagner, T. Warner, and E. Balish, Nature Med. 1:552-557, 1995; and A. Vazquez-Torres, J. Jones-Carson, T. Warner, and E. Balish, J. Infect. Dis. 172:192-198, 1995), NO is not directly candidacidal for Candida albicans (A. Vazquez-Torres, J. Jones-Carson, and E. Balish, Infect. Immun. 63:1142-1144, 1995). Because macrophages can produce both NO and superoxide anion (02-), we postulated that peroxynitrite (ONOO-), a product of the dilution-limited reaction of NO and O2-, is the candidacidal molecule of activated macrophages. We now report that ONOO-, in addition to being candidacidal in vitro, is responsible for the candidacidal activity of NO-producing macrophages. ONOO- synthesis by NO-producing macrophages was triggered by two independent mechanisms: one was nonopsonic and dependent on fungal cell wall glucan moieties, and the other was dependent on opsonic antibodies. As we have demonstrated for the pathogenic fungus C. albicans, ONOO- may also be the molecule that enables macrophages to kill other microbes that are resistant to both O2- and NO.     

Effect of Friend leukemia virus infection on susceptibility to Candida albicans.

Previous studies have demonstrated that Friend leukemia virus (FLV) induces a profound immunosuppression in susceptible mice. The studies described in this report indicate that mice infected with FLV have an increased susceptibility to subsequent infection with the opportunistic pathogen Candida albicans, as measured by increased numbers of C. albicans CFU in the kidneys of FLV-infected mice relative to uninfected controls. Experiments in which the NB-tropic and N-tropic strains of FLV were used suggest that virus replication or the resulting virus burden may be important in the observed increased susceptibility to C. albicans. Since neutrophils are believed to be important in the response of mice to systemic Candida infections, the effect of FLV infection on neutrophil candidacidal activity was investigated. The percentage of neutrophils present in unfractionated Proteose Peptone-elicited peritoneal exudates of mice infected with FLV for 14 days was significantly lower than in uninfected control mice or mice infected with FLV for 6 or 10 days. When neutrophils from FLV-infected and control mice were purified, adjusted to equal concentrations, and tested for in vitro candidacidal activity, neutrophils from mice infected with FLV for 14 days were deficient in their ability to kill C. albicans relative to normal controls and mice infected with FLV for 6 or 10 days. Addition of normal mouse serum increased killing in all groups but did not restore candidacidal activity of neutrophils from mice infected with FLV for 14 days to levels of control neutrophils or neutrophils from mice infected for 6 or 10 days with the virus. These results suggest a defect in neutrophil function, at the later stages of FLV infection, involving in vitro candidacidal activity. In addition, neutrophils from FLV-infected mice may be deficient in in vivo chemotactic activity. These defects in neutrophil function could account, at least in part, for the observed increased susceptibility of FLV-infected mice to C. albicans.     

Role of L3T4+ lymphocytes in protective immunity to systemic Candida albicans infection in mice.

Protective immunity to lethal Candida albicans challenge in vivo and activation of splenic macrophages with highly candidacidal activity in vitro were detected in mice infected with low-virulence agerminative yeast cells of the variant strain PCA-2, at a time when a strong delayed-type hypersensitivity (DTH) reaction to C. albicans occurred in the footpads of PCA-2-treated mice. The DTH reaction was transferable with spleen cell populations from these animals, and enrichment of splenic lymphocytes in L3T4+ cells significantly increased the footpad swelling. The reactivity transferred by L3T4+ cells was a radiosensitive (2,500 rads in vitro) phenomenon that required collaboration with radioresistant, silica-sensitive syngeneic cells in the host and was inhibited by treatment of recipient mice with antibodies to the L3T4 antigen or murine gamma interferon. In vitro, the PCA-2-immune L3T4+ cells produced various lymphokine activities upon incubation with C. albicans, including gamma interferon and granulocyte-macrophage colony-stimulating factor. Anti-L3T4 monoclonal antibody treatment of PCA-2-infected mice significantly impaired their footpad reaction and resistance to C. albicans, as shown by increased recovery of yeast cells from the kidneys of anti-L3T4-treated mice. These results suggested that the mechanisms of anti-Candida resistance induced by PCA-2 may involve specific induction of a DTH response mediated by inflammatory L3T4+ T cells and lymphokine-activated phagocytic effectors. However, the survival rate of the PCA-2-immune mice challenged with C. albicans was not significantly modified by administration of the anti-L3T4 antibody, thus allowing for the conclusion that compensatory mechanisms lead to considerable anti-Candida resistance when the activity of L3T4+ cells is deficient.     

Antibody response that protects against disseminated candidiasis.

We previously showed that surface mannans of Candida albicans function as adhesins during yeast cell attachment to mouse splenic marginal zone macrophages. The mannan adhesin fraction was encapsulated into liposomes and used to vaccinate mice over a 5- to 6-week period. Circulating agglutinins specific for the fraction correlated with increased resistance to disseminated candidiasis. Antiserum from vaccinated animals protected naive BALB/cByJ mice against C. albicans serotype A and B strains and Candida tropicalis. Antiserum also protected SCID mice against disseminated disease. The serum protective ability was stable at 56 degrees C, but this ability was adsorbed by C. albicans cells. The antiserum was divided into three fractions after separation by high-performance liquid chromatography. One fraction contained all of the agglutinin activity and transferred resistance to naive mice. A second fraction also transferred resistance. Two monoclonal antibodies (MAbs) specific for candidal surface determinants were obtained. MAb B6.1 is specific for a mannan epitope in the adhesin fraction, and MAb B6 is specific for a different epitope in the fraction. Both MAbs are immunoglobulin M, and both strongly agglutinate candidal cells, but only MAb B6.1 protected both normal and SCID mice against disseminated candidiasis. In one experiment, 10 normal mice were given MAb B6.1 and challenged with yeast cells. Six mice survived the 67-day observation period; 4 of the survivors were cured as evidenced by the lack of CFU in the kidney and spleen. Our studies show that antibodies against certain cell surface antigens of C. albicans help the host resist disseminated candidiasis.     

Fungicidal activity of rabbit alveolar and peritoneal macrophages against Candida albicans.

We tested the ability of rabbit macrophages to kill Candida albicans in vitro. Resident (unstimulated) alveolar macrophages killed 28.1 +/- 1.9% of ingested organisms in 4 h, whereas resident peritoneal macrophages killed only 15.2 +/- 1.3% (mean +/- standard error of the mean, P < 0.01). Peritoneal macrophages obtained from rabbits treated 3 weeks earlier with complete Freund adjuvant showed enhanced candidacidal activity relative to normally resident peritoneal cells (28.2 +/- 3.1%, P < 0.01). Candidacidal activity by alveolar macrophages recovered from such treated animals was slightly enhanced relative to untreated alveolar macrophages (32.9 +/- 2.3%). Candidacidal activity by peritoneal and alveolar macrophages was not decreased by several agents (cyanide, azide, sulfadiazine, and phenylbutazone) that inhibit the ability of human blood monocytes to kill C. albicans. In contrast, candidacidal activity by alveolar macrophages was greatly diminished by iodoacetate, an ineffective inhibitor of this function in human monocytes. We conclude that rabbit macrophages kill C. albicans by a fungicidal mechanism distinct from the peroxidase-H2O2 mechanism of human granulocytes and monocytes, and that the fungicidal properties of peritoneal and alveolar macrophage populations are enhanced after nonspecific stimulation with complete Freund adjuvant.     

Augmentation of Human Macrophage Candidacidal Capacity by Recombinant Human Myeloperoxidase and GranulocyteMacrophage Colony-Stimulating Factor

Phagocyte myeloperoxidase (MPO) is believed to be particularly important in defense against candida infection. We reported earlier that monocytes, rich in MPO, killed Candida albicans at a significantly higher rate and extent than did monocyte-derived macrophages, known to lack MPO, and that C. albicans is less resistant to MPO-dependent oxidants than less pathogenic Candida species. We hypothesized, therefore, that the capacity of macrophages to kill C. albicans might be improved in the presence of MPO. In this study, we evaluated the ability of recombinant human MPO (rhMPO) to augment the killing of C. albicans by resident macrophages and macrophages activated by recombinant human granulocyte-macrophage colony-stimulating factor. Addition of rhMPO (concentration range, 0.8 to 6.4 U/ml) to suspensions of resident and activated macrophages and opsonized C. albicans resulted in concentration-dependent and significant increases in candida killing. This enhancement was particularly pronounced with activated macrophages, whether C. albicans was opsonized or unopsonized and ingested through the macrophage mannose receptor. rhMPO did not affect the killing of C. albicans by monocytes, nor did it affect phagocytosis of opsonized or unopsonized C. albicans. These results indicate that exogenous rhMPO can augment the candidacidal capacity of both resident and activated macrophages, with a more profound effect on activated cells. We suggest that rhMPO may be effective in the treatment of invasive candidiasis.