Effects of macrophage colony-stimulating factor (M-CSF) on anti-fungal activity of mononuclear phagocytes against Trichosporon asahii

Abstract

Trichosporon asahii is an emerging opportunistic pathogen in immunocompromised patients. Little is known about the mechanisms of host defence against T. asahii. We investigated the fungicidal activity of human peripheral blood monocytes and murine peritoneal macrophages against T. asahii isolates, and the effects of M-CSF on the anti-fungal activity of mononuclear phagocytes. We also established a neutropenic mouse model of disseminated trichosporonosis with T. asahii. M-CSF enhanced the phagocytic fungicidal activity of mononuclear cells, and infected mice treated with human M-CSF at 10 × 10⁶ U/kg showed a significant improvement in survival rate, with fewer fungal colony counts in the lung compared with control mice. Mice treated with human M-CSF showed higher concentrations of tumour necrosis factor-alpha (TNF-α) in the lung and plasma compared with control mice. The survival rate was significantly reduced in mice treated with anti-mouse TNF-α. Our results showed that M-CSF enhanced the fungicidal activity of mononuclear phagocytes partly by production of TNF-α, and suggest that the administration of M-CSF to patients with disseminated trichosporonosis may be a useful adjunct to conventional anti-microbial therapy and prophylaxis.

INTRODUCTION

Disseminated trichosporonosis is an uncommon, but increasingly reported and frequently life-threatening infection in immunocompromised patients [1,2]. Trichosporon beigelii (formerly T. cutaneum) has been considered to be the aetiological pathogen of disseminated trichosporonosis. However, the revised taxonomic classification based on molecular data shows the genus Trichosporon to consist of a large number of species, and that T. asahii is the main agent of disseminated trichosporonosis and summer-type hypersensitivity pneumonitis [3,4]. Trichosporon asahii may be resistant in vitro to amphotericin B and such resistance correlates with refractory, disseminated trichosporonosis in granulocytopenic patients [5–7].

The majority of patients with disseminated trichosporonosis receive cytotoxic anti-cancer chemotherapy for the associated cancers or as a preparative regimen for bone marrow transplantation. Thus, granulocytopenia is the most common apparent risk factor of disseminated trichosporonosis, and the resolution of infection in neutropenic patients is related primarily to bone marrow recovery. Although neutrophils are important components of the defence system against this infection [8], mononuclear cells serve as the primary source of phagocytic defence in patients with profound and prolonged neutropenia. Macrophages play an important role in host resistance to acute systemic candidiasis and to disseminated candidiasis of endogenous origin [9]. Thus, mononuclear phagocytes may be more important in the control of Trichosporon infection than previously shown [10].

In the present study, we investigated the fungicidal activity of mononuclear phagocytes and the effects of M-CSF using a murine model of disseminated trichosporonosis.

MATERIALS AND METHODS

Fungicidal activity of human peripheral monocytes

Isolation of peripheral blood monocytes

Peripheral blood monocytes were isolated from heparinized venous blood from consenting healthy human donors. Monocytes were separated by dextran sedimentation followed by Ficoll–Hypaque centrifugation. Cell viability was consistently > 95% as determined by trypan blue exclusion, and the majority of the isolated cell population were monocytes (> 90%) as determined by morphology and non-specific esterase staining. After separation, monocytes were suspended in Ca²⁺- and Mg²⁺-free modified Hanks' balanced salt solution (HBSS) and stored on ice until use. Before incubation, cells were equilibrated in Ca²⁺- and Mg²⁺-containing HBSS.

Preparation of organisms

Five strains of T. asahii used in the present study were clinical isolates from blood samples taken from infected patients. They were identified as T. asahii using the DNA-DNA re-association technique by Prof. T. Shinoda (Meiji College of Pharmacy, Tokyo, Japan). Organisms were stored in a skimmed milk suspension at −80°C. Before use, they were cultured on Sabouraud dextrose agar (SDA) plate for 24–48 h, and re-cultured in Sabouraud dextrose broth (SDB) at 30°C for 12–15 h. All isolates cultured in SDB contained > 95% of the organisms in conidial forms (both blastconidia and arthroconidia). Organisms were washed twice with 0.15 m NaCl, and the concentration in each inoculum was adjusted based on manual counting using a haemocytometer. Organisms were opsonized with 50% pooled human AB serum for 30 min at 37°C, washed once with cold modified HBSS, and resuspended in a final concentration of 1 × 10⁵/ml in HBSS. Organisms were kept on ice until use.

M-CSF

Recombinant human M-CSF with a specific activity of 5.6 × 10⁵ U/mg (Genzyme Corp., Cambridge, MA) contained < 0.2 ng/μg of endotoxin. Monocytes were cultured in RPMI 1640 with 10% bovine serum with M-CSF (100 or 500 U/ml) for 72 h at 37°C in 7% CO₂ in an attempt to enhance their fungicidal activity against the organisms. Control cells were cultured in medium alone. Before testing, cells were harvested, washed with modified HBSS, and recounted. Trichosporon asahii OMU-239, a virulent clinical isolate from the blood of a patient with disseminated trichosporonosis and chronic myelogenous leukaemia, was used for this series of experiments.

Fungicidal assay

The fungicidal activity of monocytes was assessed using the colony-forming unit (CFU) assay. Serum-opsonized organisms were mixed with monocytes in a final effector cell-to-target (E:T) ratio of 10:1 in HBSS containing 0.1% bovine serum albumin (BSA). Samples were incubated on a rotator at 37°C, and aliquots were obtained at 60 min and 120 min. Phagocytes were lysed in sterile water; serial 10-fold dilutions were prepared and plated in duplicate on SDA plates. CFU were determined after incubation for 24 h at 37°C. The percentage of killing was calculated as follow: killing (%) = (1 − sample CFU/control CFU) × 100.

Fungicidal activity of murine peritoneal macrophages

Animals

Male ddY, 6-week-old, specific pathogen-free mice (body weight 20–25 g) were purchased from Shizuoka Agricultural Cooperative Association Laboratory Animals (Shizuoka, Japan). They were housed in cages and given laboratory chow and tap water ad libitum. Animal studies were performed according to the Guidelines of the Ethical Committee for Animal Experiments, Nagasaki University, Nagasaki, Japan.

Isolation of peritoneal macrophages

Human M-CSF was injected subcutaneously into mice at a dose of 10 × 10⁶ U/mg per day for 5 consecutive days. Control mice received an equivalent volume of sterile saline. Murine peritoneal macrophages were harvested 24 h after the last injection. The peritoneal macrophages were incubated at 37°C in RPMI 1640 for 1 h. The cells were washed with PBS at 37°C. Cell viability was consistently > 98% as determined by trypan blue exclusion, and the majority of the isolated cell population were macrophages (> 99%) as determined by morphology and non-specific esterase staining.

Fungicidal assay

Serum-opsonized T. asahii OMU-239 were mixed with murine peritoneal macrophages at a final E:T ratio of 10:1 in RPMI 1640 containing 0.1% BSA. Samples were incubated on a rotator at 37°C and aliquots were obtained at 60, 120 and 180 min. The fungicidal activity of murine peritoneal macrophages was assessed as described above.

Neutropenic murine model of disseminated trichosporonosis

Organism and preparation of inoculum

Trichosporon asahii OMU-239 was used in this series of experiments. Organisms were cultured on SDA plate for 48 h, then cultured in SDB at 30°C for 12–15 h. On the day of infection they were washed twice with sterile distilled water. The resulting suspension was manually counted using a haemocytometer, and then adjusted using sterile distilled water to produce the desired inoculum size. Dilutions of the original suspension were prepared and cultured on SDA plates to confirm the haemocytometer count. Suspensions contained > 95% conidial forms.

Experimental design

Male ddY, 6-week-old, specific pathogen-free mice (body weight 20–25 g) were employed. Cyclophosphamide (CY; Shionogi Co., Osaka, Japan) was injected intraperitoneally at a dose of 100 mg/kg body weight at 3 days and 1 day before infection. Various concentrations of infectious inocula of T. asahii were used in preliminary experiments to determine the optimal dose for the following experiments. Thus, 2 × 10⁵, 1 × 10⁶ or 2 × 10⁶ CFU/mouse were injected into the lateral tail vein after CY administration. In order to evaluate the effects of human M-CSF, it was injected subcutaneously at a dose of 1 × 10⁶ or 10 × 10⁶ U/mg per day for 5 consecutive days starting 2 days before inoculation with 2 × 10⁶ CFU of T. asahii. Control mice received an equivalent volume of sterile saline before infection. Ten animals in each group were monitored for 30 days after infection. To investigate the effects of anti-tumour necrosis factor-alpha (TNF-α), a dose of 200 μg/kg of rat anti-mouse TNF-α (Upstate Biotechnology Inc., Lake Placid, NY) was injected subcutaneously into mice treated with M-CSF and infected with 2 × 10⁶ CFU/mouse of T. asahii at 0 and 2 days after infection. Control infected mice were given an equivalent volume of sterile saline and normal rat serum. Ten animals in each group were monitored for 30 days after infection.

Organ clearance

Four randomly selected mice infected with 2 × 10⁵ CFU/mouse were killed 12, 24, 72 and 96 h and 7 and 14 days after infection, and 12, 24, 72 and 96 h after infection in another group of mice infected with 2 × 10⁶ CFU/mouse. Another five randomly selected mice treated with M-CSF and infected with 2 × 10⁶ CFU/mouse of T. asahii were killed 12, 24, 48 and 72 h after infection. The lung, kidney, liver, spleen, and heart of each animal were dissected and removed aseptically, and homogenized in 1 ml of sterile saline. The homogenate was diluted with saline and incubated on SDA at 37°C for 48 h. The colonies were then counted and the CFU/organ calculated.

TNF-a in lung and plasma

Homogenized lungs were centrifuged and the supernatants were used. The concentrations of TNF-α were determined by the solid-phase ELISA, using TNF-α mouse ELISA kits (Amersham International plc, Aylesbury, UK). Absorbance at 450 nm was measured with a microplate reader (NovaPath; BioRad, Milan, Italy). A standard curve was constructed by plotting the concentration of cytokine standard versus the respective absorbance. The concentration of cytokine in each sample was determined by interpolation from the standard curve.

Statistical analysis

All biochemical values were expressed as mean ± s.e.m. of three or more samples. Differences between groups were examined for statistical significance using Student's t-test. P < 0.05 denoted the presence of a statistically significant difference.

RESULTS

Human peripheral monocyte fungicidal activity

All T. asahii strains tested were resistant to killing by human monocytes at 1:1 and 10:1 E:T ratios after a 60-min incubation. The percentage killing slightly increased at 10:1 E:T ratio after 120 min incubation time, although the killing activity was still relatively low. However, the addition of human M-CSF enhanced the percentage killing of five strains of serum-opsonized T. asahii at a 10:1 E:T ratio after incubation for 60 min. A more potent fungicidal activity was noted when M-CSF was used at a concentration of 500 U/ml (Fig. 1). There was no difference in monocyte fungicidal activity after 60 or 120 min incubation (data not shown).

Fig. 1.

M-CSF enhanced the fungicidal activity of human peripheral monocytes against Trichosporon asahii after 60 min incubation. Effector cell-to-target ratio used was 10:1. Data are mean ± s.e.m. of five experiments.

Murine peritoneal macrophage fungicidal activity

Figure 2 shows the percentage killing of serum-opsonized T. asahii by murine peritoneal macrophages harvested from mice treated with human M-CSF (10 × 10⁶ U/kg per day). There was no difference in peritoneal macrophage fungicidal activity at a 10:1 E:T ratio after incubation for 60 min between the control and M-CSF-treated groups. However, M-CSF enhanced the percentage killing of T. asahii after incubation for 120 and 180 min.

Fig. 2.

The percentage killing of Trichosporon asahii by murine peritoneal macrophages harvested from mice treated with human M-CSF (10 × 10⁶ U/kg per day) after incubation for 60, 120 and 180 min. Effector cell-to-target ratio used was 10:1. Data are mean ± s.e.m. of five experiments. ▪, M-CSF; □, control.

Induction of neutropenia

Neutropenia was induced, with a maximum at 4 days after the initial injection of CY, and continued at least 6 days. The peripheral leucocyte count fell to 1371 ± 422/mm³ from a baseline of 6850 ± 552/mm³, and the percentage of neutrophils was decreased from 27.3 ± 4.4% to 2.3 ± 1.9%. Administration of M-CSF did not affect the number or percentage of peripheral blood neutrophils.

Animal survival and inoculum size

Trichosporon asahii OMU-239 was associated with 90% survival when inoculated at 2 × 10⁵ CFU in immunocompromised mice (mean survival time 27.6 days, range 5–30 days). In contrast, all mice infected with a higher inoculum size (1 × 10⁶ CFU or 2 × 10⁶ CFU) died, with a median survival time of 6.3 days and 3.3 days, respectively (range 2–13 days and 2–10 days, respectively). All mice infected with < 2 × 10⁵ CFU survived throughout the 30-day observation period (data not shown). Thus, in the present study, mice infected with 2 × 10⁵ CFU and with 2 × 10⁶ CFU were used as the survival and lethal models, respectively.

Organ clearance

Trichosporon asahii, but no other pathogen, was isolated from various body organs. In the survival model, organisms were noted in all organs examined 12 h after infection, and their numbers increased until 96 h after infection in the kidney, spleen and liver, but not in the lung and the heart. The number of fungi in the lung was 6.8 × 10⁴ ± 4.1 × 10³ CFU and 5.5 × 10⁴ ± 1.3 × 10³ CFU at 12 h and 96 h after infection, respectively. Then, the number of fungi decreased gradually and disappeared, except in the kidney, by 14 days after infection (Fig. 3a). In the lethal model, the fungal counts increased in all organs for the 96 h after infection. The fungal count in the lung was 3.0 × 10⁵ ± 8.4 × 10⁴ CFU and 3.7 × 10⁷ ± 7.6 × 10⁶ CFU at 12 h and 96 h after infection, respectively (Fig. 3b). In all organs the number of fungi was about 1000 times the respective number in the survival model 96 h after infection.

Fig. 3.

Organ clearance of trichosporonosis murine survival model (a) and lethal model (b). The survival model was inoculated intravenously with Trichosporon asahii at 2 × 10⁵ colony-forming units (CFU)/mouse. Four mice were killed at 12, 24, 72, 96 h and 7 and 14 days after infection. The lethal model was inoculated intravenously with T. asahii at 2 × 10⁶ CFU/mouse, and four mice were killed at 12, 24, 72, 96 h after infection. The lung (•), kidney (○), liver (Δ), heart (▴) and spleen (□) of each animal were removed aseptically and homogenized in sterile saline. The homogenate was diluted with sterile saline and incubated on Sabouraud dextrose agar (SDA) at 37°C. Data are mean ± s.e.m. of four experiments.

Effects of M-CSF

The effects of M-CSF were investigated using the lethal model. The number of fungi in the lung of M-CSF-treated mice was significantly less than in control mice. The fungal counts were 1.2 × 10⁷ ± 2.1 × 10⁶ CFU in M-CSF-treated mice and 7.7 × 10⁷ ± 5.7 × 10⁶ CFU in control mice 72 h after infection (Fig. 4). A high dose of M-CSF (10 × 10⁶ U/kg per day), but not a low dose (1 × 10⁶ U/kg per day), significantly increased the survival rate (Fig. 5).

Fig. 4.

Organ clearance of human M-CSF-treated trichosporonosis murine model. Five mice were killed at 12, 24, 48, 72 h after infection. In the human M-CSF-treated model (•) and control model (○), data are mean ± s.e.m. of five experiments. *P < 0.05 versus human M-CSF at 1 × 10⁶ U/kg per day group and control group.

Fig. 5.

Effect of M-CSF on survival rates of mice with disseminated trichosporonosis. Mice immunosuppressed with two injections of 100 mg of cyclophosphamide were treated with M-CSF at 10 × 10⁶ U/kg per day (○), 1 × 10⁶ U/kg per day (•) or sterile saline (□) for 5 consecutive days starting 2 days before infection with Trichosporon asahii (2 × 10⁶ colony-forming units (CFU)/mouse). Treatment with M-CSF at 10 × 10⁶ U/kg per day significantly increased the survival rate. *P < 0.05 versus M-CSF at 1 × 10⁶ U/kg per day group and control group.

TNF-α concentrations in lung and plasma

TNF-α concentrations in the lung and plasma of M-CSF-treated and control mice are shown in Table 1. TNF-α was produced in the lung and plasma of infected mice, and M-CSF increased the levels of TNF-α. The concentrations of TNF-α in the lung of M-CSF-treated mice were significantly higher than those of control mice at 12, 24 and 48 h after infection. TNF-α levels in the plasma of M-CSF-treated mice were also significantly higher at 12h and 48 h after infection.

Table 1.

Tumour necrosis factor-alpha in lung and plasma of trichosporon murine model after inoculation†

Effects of anti-TNF-α

Finally, we examined the effects of anti-TNF-α on survival rate and time (Fig. 6). The lethal model with human M-CSF (10 × 10⁶ U/kg per day) was used in these experiments. The survival rate of M-CSF-treated mice was 60%, but all mice treated with anti-TNF-α died within 4 days. There was no difference in the survival rate between anti-TNF-α-injected mice and control mice.

Fig. 6.

Effect of anti-mouse tumour necrosis factor-alpha (TNF-α) on survival rates of trichosporonosis murine model. Every 10 mice immunosuppressed with two injections of 100 mg/kg of cyclophosphamide were treated with human M-CSF at 10 × 10⁶ U/kg per day (○), anti-mouse TNF-α at 200 μg/kg on 0 and 2 days after infection and human M-CSF at 10 × 10⁶ U/kg per day (•) or saline (□). *P < 0.01 versus anti-TNF-α with M-CSF group and M-CSF alone group.

DISCUSSION

Disseminated trichosporonosis is a lethal opportunistic infection in immunocompromised patients. However, little is known about the mechanisms of host defence against this infection. The port of entry is usually the respiratory or gastrointestinal tract, and the lung is the most commonly involved organ. Pulmonary alveolar macrophages may ingest and destroy conidia that have been inhaled in the alveolar spaces, preventing their germination to hyphae, while circulating phagocytes (polymorphonuclear leucocytes and monocytes) move to the site of infection, where they participate in host defence against the fungi. During the early stages of infection, mononuclear cells phagocytose yeasts, but fungi that escape phagocytosis rapidly convert into the hyphal form and invade different organs. As hyphae are too large to be engulfed by mononuclear phagocytes, they are more likely to be killed by extracellular mechanisms. During prolonged periods of granulocytopenia induced by anti-cancer chemotherapy or primary bone marrow failure, hyphae can escape these lines of defence, and disseminated trichosporonosis can occur.

We have demonstrated that human peripheral monocytes and murine peritoneal macrophages do not exhibit fungicidal activity against T. asahii, but that the killing activity of these cells can be induced and enhanced by M-CSF. M-CSF has been reported to up-regulate anti-fungal activity of mononuclear phagocytes against other organisms including Candida species [11], Aspergillus species [12], Cryptococcus neoformans [13] and Histoplasma capitatum [14].

M-CSF augments the ability of mature macrophages and monocytes to kill microorganisms by enhancing their production of superoxides and cytokines, as well as their phagocytic and chemotactic properties [15]. TNF-α, interferon-gamma (IFN-γ), granulocyte-macrophage CSF (GM-CSF), granulocyte-CSF (G-CSF), IL-1, IL-6, IL-8 and IL-12 are the main cytokines produced by mononuclear phagocytes [16–18]. These cytokines may further enhance the cellular killing properties against pathogenic organisms. For example, TNF-α enhances chemotaxis of macrophages and neutrophils and augments neutrophil oxidative respiratory burst and fungicidal activity [19]. In addition, TNF-α stimulates natural killer (NK) cells, and the stimulated NK cells in turn may exhibit some direct anti-fungal effects and produce TNF-α, IFN-γ and GM-CSF, which enhance the anti-fungal effects of neutrophils and macrophages [20,21].

In clinical trials, Nemunaitis et al. [22,23] demonstrated that the combined use of M-CSF with standard anti-fungal treatment in patients with invasive fungal infections after bone marrow transplantation resulted in an improvement of the survival rate compared with the control. Ohno et al. [24] reported that the administration of M-CSF significantly reduced the incidence and duration of febrile neutropenia during the intensive consolidation therapy, and shortened the period required to complete three courses of consolidation chemotherapy in acute myelogenous leukaemia.

Our results demonstrate that the administration of M-CSF prolongs the survival time, reduces the number of Trichosporon organisms in the lung, and increases the concentrations of TNF-α in the lung and plasma in a murine model of disseminated trichosporonosis. The role of TNF-α in this infection was further demonstrated in experiments using anti-TNF-α antibody; the administration of the antibody markedly reduced the survival rate. Louie et al. [25] described that the administration of TNF-α antibody in a murine model of systemic candidiasis neutralized serum TNF-α, shortened survival, and increased fungal colony counts in the spleen, kidney, liver and lung. Nagai et al. [20] also reported that TNF-α exhibits a protective role in invasive aspergillosis.

Thus, TNF-α produced during an infection may augment host resistance against fungi and prolong survival, and the mechanism of the protective effect of M-CSF in disseminated trichosporonosis is, at least in part, due to the production of TNF-α. While TNF-α may enhance the host defence mechanisms against the invading microorganisms, previous studies have shown that large amounts of TNF-α released during the acute stages of infection may actually worsen the condition. In this regard, Muranaka et al. [26] suggested that TNF-α overproduction in the lung has an important role in the prognosis of trichosporonosis. The peak concentrations of TNF-α in the plasma and bronchoalveolar lavage fluid in their model at 24 h after infection were 742 ± 104 pg/ml and 60.1 ± 6.3 ng/ml, respectively. These levels are higher than the concentrations measured in the plasma and lung in our model. These differences are probably due to differences in the virulence of strains, size of inocula, and the animal model used in each study.

In conclusion, our study shows that the administration of M-CSF to a murine model of disseminated trichosporonosis significantly increased survival. This effect was caused by the enhancement of the fungicidal activity of mononuclear phagocytes, partly due to the production of TNF-α. These results suggest that M-CSF may be a useful adjunct to conventional anti-microbial therapy and prophylaxis.