Disseminated Trichosporonosis in a Murine Model of Chronic Granulomatous Disease

Home , Coccidioides, Trichosporonosis, Trichosporon beigelii

Shannon H. Lacy,1 Donald J. Gardner, DVM,1 Leonard C. Olson, DVM,1 Li Ding, MD,2 Steven M. Holland, MD,2 and Mark A. Bryant, DVM1,*

Over a period of ten months, five mice submitted to our service (the Pathology Section of the Veterinary Re- sources Program, Office of Research Services at the National Institutes of Health, Bethesda, Md.) were diagnosed with disseminated trichosporonosis. These mice had pyogranulomatous inflammation in multiple organs, including lung, liver, lymph nodes, salivary gland, and skin. Fungal elements in many of the lesions were identified, using special histochemical stains, and Trichosporon beigelii was obtained by use of culture of specimens at affected sites. This saprophytic fungus has caused disseminated disease in immunosuppressed humans. However, despite widespread use of immunosuppressed rodents in research, to the authors’ knowledge, this organism had not previously been reported to cause spontaneous disseminated disease in laboratory mice. All affected mice had a geneti- cally engineered defect in p47phox, a critical component of the nicotinamide dinucleotide phosphate (NADPH) oxidase, the enzyme responsible for generating the phagocyte oxidative burst. These animals are used as a murine model of human chronic granulomatous disease. We discuss the lesions, differential diagnosis, identification of the organism, and the role of NADPH oxidase in protecting against disseminated trichosporonosis.

Trichosporon species are ubiquitous soil saprophytes present ism (1, 2, 6, 8, 12, 18, 27, 28). To our knowledge, there have been no in small numbers as normal flora of human skin and gastrointes- published reports of spontaneous disseminated trichosporonosis in tinal mucosa. In immunocompetent human hosts, these yeast- laboratory mice. like fungi can cause a nodular infection of the hair shafts called white piedra (T. beigelii, formerly T. cutaneum [3]), and summer- Case Reports type hypersensitivity pneumonitis (T. asahii [22]). The organism Animals. Between August 2000 and May 2001, disseminated is now emerging as an important opportunist; deep-seated infec- trichosporonosis was diagnosed in five mice at postmortem ex- tions in immunocompromised hosts are reported with increasing amination. There were two female and three male mice, with frequency (14, 18). Disseminated trichosporonosis is characterized age at necropsy ranging from 18 to 68 weeks. All mice were by a wide variety of disease manifestations, including endocarditis C57BL/6Nai-[KO] P47PHOX, a line that is homozygously defi- (4), endophthalmitis (21), brain abscess (25), pneumonia, fever, cient for p47phox (10). The p47phox is a critical component of the fungemia, funguria (24), and high mortality (23, 24). Granulocy- NADPH oxidase, and deficiency of p47phox leads to CGD (10). As topenia is the most commonly recognized risk factor for this in- early as 1997, there was inapparent and undetected genetic con- fection (18). Disseminated trichosporonosis is most commonly tamination of the C57BL/6Nai strain with the C57BL/6IFN-γ reported in patients receiving cytotoxic therapy for hematologic knockout strain. Polymerase chain reaction (PCR) analysis de- malignancy (8), but is also seen in intravenous drug users and termined the presence of one interferon-gamma (IFN-γ) knockout HIV patients. Recipients of organ transplants, prosthetic limbs, allele in three of the five p47phox mice submitted for necropsy. and artificial heart valves are also at increased risk for dissemi- Therefore, all animals were homozygous for p47phox deficiency, nated trichosporonosis (6). The infection has also been reported and three were heterozygous for IFN-γ deficiency. in patients with chronic granulomatous disease (CGD), a he- Care and husbandry. The animal holding room was on a reditary defect in neutrophil function (20). 14/10-h light/dark cycle, with ambient temperature range of 20 Trichosporon spp. are rarely mentioned as animal pathogens. to 23°C. Room air pressure was negative to the corridor, with 15 Trichosporon beigelii was recently reported as a cause of masti- to 30 air exchanges of 100% non-circulated air per hour. The tis in dairy cows (7), and was previously implicated in a case of relative humidity of the air in the animal room was maintained cystitis in a cat (11). Murine models to simulate human dis- between 35 and 55%. The mice were group housed on autoclaved seminated trichosporonosis typically require profound immuno- hardwood bedding in Micro-Isolator™ cages, located on conven- suppression with cyclophosphamide and/or corticosteroids, tional racks. After the animal cages were washed, bedding and followed by intravenous inoculation of large doses of the organ- feed were added, a wire cage top and Micro-Isolator™ top were put on the cages, then the entire cage unit was autoclaved. The Received: 9/24/02. Revision requested: 11/12/02. Accepted: 1/22/03. animals received acidified water within a pH range of 2.8 to 3.0 1Veterinary Pathology Section, Veterinary Resources Program, Office of Research in sterile bottles with added trimethoprim/sulfamethoxazole 2 Services and Laboratory of Host Defenses, National Institutes of Allergy and (7.6 ml/480 ml), and an autoclaved rodent diet (Zeigler Bros., Infectious Diseases, National Institutes of Health, Bethesda, Maryland 20892. *Corresponding author. Inc., Gardners, Pa.) was available ad libitum. Serum samples 303 Vol 53, No 3 Comparative Medicine June 2003

Table 1. Postmortem findings in five mice with disseminated trichosporonosis Animal information Gross findings Histologic findings Culture results ID Accession Sex Age Skin Splenomegaly Lung Lung Hepatitis Lymphadenitis Splenitis Sialoadenitis (wk) Lesion Lesion Lesion f281 MS0007786 F 18 X X, N PG P* S PG* Trichosporon sp. Skin Pyogranuloma f354 MS0103625 M 30 X X, N X NS BP* NS* PG PG* Trichosporon sp. Spleen f283 MS0007035 M 13 X X X PG P* PG* PG* T. beigelii Skin Skin Pyogranuloma e749 MS0105320 F 68 X X PG PL* PG* PG* T. beigelii Lung Pyogranuloma f558 MS0105320 M 28 X X T. beigelii Skin Skin Pyogranuloma

X = lesion present; P = pneumonia; PG = pyogranulomatous; N = nodule; BP = bronchopneumonia; NS = necrosuppurative; * = intralesional fungi; PL = pleuropneumonia; S = suppurative; F = female; M = male from sentinel animals in the animal colony were tested quar- arthroconidia along with septate hyphae and yeast-like cells in terly; twice yearly, a full panel of tests was performed, and an the tissue sections are histologic findings consistent with abbreviated panel of tests was performed during the other two trichosporonosis (3). These histologic findings allowed one to quarters. The full panel of serologic tests included evaluation rule out the higher bacteria as the sole cause of the lesions and for the presence of antibodies to: cilia-associated respiratory other organisms were presumptively ruled out on the basis of bacillus, ectromelia virus, epizootic diarrhea of infant mice vi- their morphologic appearance in tissue sections. However, in rus, Theiler’s mouse encephalomyelitis virus, mouse adenovirus, this study, definitive diagnosis of trichosporonosis relied on iso- lymphocytic choriomeningitis, mouse cytomegalovirus, mouse lation of the organism in culture. hepatitis virus, minute virus of mice, Mycoplasma pulmonis, Candida sp. was the principal suspect in the differential diag- parvovirus, polyoma virus, pneumonia virus of mice, reovirus 3, nosis. Candida sp. has caused disseminated infection in and Sendai virus. Results of the serologic tests were negative immunocompromised mice (6, 28), and is frequently falsely diag- during the period of study. The institute’s animal care- and-use nosed in humans with disseminated trichosporonosis (9). Pre- committee approved the animal study protocol. The animal care sumptive differentiation between the two relies on the presence and husbandry practices in the facility were consistent with the of arthroconidia and yeast-like Trichosporon cells that are larger guidelines set forth in the Guide for the Care and Use of Labora- and more pleiomorphic than those of Candida sp. (3). Unlike tory Animals (15). Candida sp., Trichosporon spp. produce arthroconidia by fragmen- Postmortem findings. A summary of the postmortem find- tation of the septate hyphae. However, these are often not seen in ings is provided in Table 1. All five mice were submitted alive; tissue, leaving yeast-like and hyphal elements similar to those of one died prior to examination. The other four mice were hu- Candida sp. Direct immunofluorescent testing and culturing are manely euthanized in a carbon dioxide gas chamber. On gross valuable tools for differentiating the two organisms (9). In addi- examination, four mice had ulcerated, purulent skin lesions tion, a nested PCR assay has recently been developed to achieve about the ventral aspect of the neck. Lung lesions were seen early and accurate diagnosis of disseminated trichosporonosis (23). grossly in three of the mice; one mouse had pleural adhesions Aspergillus spp. are ubiquitous saprophytes that cause oppor- and a soft white exudate within the pulmonary parenchyma, tunistic infections in immunocompromised hosts. In a national and the other two had multifocal-to-coalescing, tan-gray consoli- registry of human patients with chronic granulomatous disease, dated areas in the lungs. Splenomegaly was present in all five Aspergillus spp. was isolated from 41% of 290 pneumonic processes mice, and two had splenic abscesses. (19, 26). It produces inflammatory lesions similar to those found in On histologic examination, lung lesions were found in four mice; two the mice reported here; however, it does not appear as yeast-like had pyogranulomatous pneumonia, one had pyogranulomatous pleu- cells, but as large colonies of radiating, 3- to 6-µm-wide hyphae (in ropneumonia, and one had necrotizing bronchopneumonia. deep mycoses) or as randomly dispersed hyphae (in granulomas) Pyogranulomatous dermatitis was evident in three mice, and (11). This morphologic distinction allowed Aspergillus sp. to be three mice had pyogranulomatous lymphadenitis. Hepatitis was ruled out in these cases. detected in three of the mice, and in one mouse, it was character- The yeast-like cells were somewhat evocative of a Blastomyces ized as pyogranulomatous. Two mice had pyogranulomatous infection; however, Blastomyces cells are larger (8 to 20 µm in di- splenitis, and one had diffuse pyogranulomatous sialoadenitis in ameter), have a double wall, reproduce by broad-based budding, a submandibular salivary gland. and do not produce hyphae in tissue. Histoplasma capsulatum Identification of the organism. In each case, the extensive could also be ruled out on the basis of morphology; this organism pyogranulomatous inflammation was suggestive of a fungal is much smaller in diameter (2 to 4 µm), oval in shape, and present (Aspergillus, Blastomyces, Histoplasma, Sporothrix, Coccidio- within the cytoplasm of the epithelioid and multinucleated giant ides, Paecilomyces spp.), yeast (Candida, Cryptococcus spp.), or cells that proliferate extensively in histoplasmosis lesions (11). higher bacterial (Actinomyces, Nocardia, Dermatophilus, Myco- The gross skin lesions were not unlike those associated with bacterium spp.) cause. In many of these lesions, staining with sporotrichosis, which is caused by another dimorphous fungus, Gomori methenamine silver (GMS) and periodic acid-Schiff Sporothrix sp. This organism can also disseminate to the lungs, (PAS) revealed 8- to 12-µm-diameter yeast-like cells with sep- liver, lymph nodes, and kidneys. However, like Histoplasma and tate hyphae (Fig. 1) and a few arthroconidia. The presence of Sporothrix spp., yeast cells are more likely to be located within 304 Disseminated trichosporonosis in murine granulomatous disease

Figure 1. Photomicrographs of Trichosporon fungal elements in mouse tissue lesions. (A) Multiple lung granulomas. Notice organisms at the center of one granuloma (solid arrow). Airways indicated by open arrow. Periodic acid-Schiff (PAS) stain; × 100. (B) Higher magnification of Fig. A, with open arrows pointing to yeast-like fungal cells. PAS stain; × 1,000. (C) Yeast form of the organism. Gomori methenamine silver (GMS) stain; × 1,000. (D) Septate hyphae (solid arrow) and yeast-like forms (open arrows) of the organism within a liver granuloma. GMS stain, × 1,000. the cytoplasm of macrophages and multinucleated giant cells. yeast forms also present. However, the hyphae have regular In addition, Sporothrix yeast cells are much smaller (4 to 6 µm bulbous swellings and a few conidiophores (11), neither of which in length) than were those observed in these cases, and have a was seen in these cases. characteristic elliptic-to-elongate (cigar) shape. Coccidioides Multiple specimens were collected for fungal culture at immitis causes deep mycoses in many species of animals, and necropsy and were submitted to the VRP Bacteriology Labora- reproduces by formation of arthroconida in the soil. These tory. For each specimen, two plates of Sabouraud dextrose agar arthroconidia are infective, but once in the host, reproduction is were inoculated, with one incubated at 37°C and the other incu- by endosporulation, and large spherules (up to 50 µm in diam- bated at room temperature. Blood agar (trypticase soy agar with eter) are found in tissues (11). 5% sheep blood) and thioglycollate broth also were inoculated Cryptococcus neoformans is a yeast-like fungus that can infect and incubated at 37°C. Specimens from skin pyogranulomata the lungs, with dissemination rarely reported. Size and morphol- (three mice), spleen (one mouse), and lung (one mouse) were cul- ogy allowed us to rule it out in this case. The organism character- ture positive for Trichosporon sp. Colonies grew within one week, istically has a wide, gelatinous capsule, bringing its total diameter initially appearing as cream-colored, smooth, and shiny before to 30 µm. Disseminated infection with Paecilomyces sp. has been becoming dry and membranous. The organism was identified on reported in humans with CGD, as well as in the murine model for the basis of morphology, and identification was confirmed, using the disease (10, 20, 26). It produces granulomatous inflammation two commercial identification systems: RapID Yeast Plus Sys- and appears in tissue as septate, branching hyphae, with a few tem (Innovative Diagnostic Systems, L.P., Norcross, Ga.), and 305 Vol 53, No 3 Comparative Medicine June 2003

API 20 C AUX (bioMerieux, Hazelwood, Mo.). Culture of three of tion of M-CSF as an adjunct to standard anti-fungal treatment the specimens yielded Trichosporon beigelii; however, the other increased survival in bone marrow transplant patients suffer- two isolates were not speciated. ing invasive mycoses (16). The critical importance of granulocyte function in resistance Discussion to fungi is best evidenced by development of infection in human Trichosporon spp. are ubiquitous soil saprophytes and minor patients with CGD (20). As indicated (Fig. 2), NADPH oxidase components of normal flora of the human skin and gastrointes- catalyzes the transfer of an electron from NADPH to molecular tinal tract. Serious systemic infections with this organism are oxygen to form superoxide radical, a key substrate for the gen- well documented in immunosuppressed humans. It was previ- eration of antimicrobial oxidants. Four major reactive oxidants ously believed that T. beigelii was the most common cause of dis- are formed, either directly or indirectly, from superoxide radical: seminated trichosporonosis (1, 3); however, recent taxonomic peroxynitrite anion, hydroxyl anion, hypochlorous acid, and nit- revisions have suggested that T. asahii and T. mucoides are the ryl chloride. In patients with CGD, a defect in any one of the most common causative agents of opportunistic disseminated four structural proteins that comprise the NADPH oxidase re- trichosporonosis (27). sults in a functional defect in oxidant production that impairs To the authors’ knowledge, there have been no previous re- killing of bacterial and fungal pathogens. Seventy percent of ports of spontaneous disseminated Trichosporon infection in human cases of CGD are X-linked and result from a mutation in laboratory mice, despite the vast number of immunosuppressed the heavy chain of the membrane-bound cytochrome, gp91phox. mice used in research. Perhaps this is due in part to the similar Mutations in the other structural proteins are autosomal reces- morphologic appearance of Trichosporon and Candida spp. in tis- sive, including mutation of p47phox (25% of human cases), which sue sections (3, 9). Trichosporon misdiagnosis has been blamed is altered in the mouse strain of this report (10). In humans for a number of human deaths because of its delayed diagnosis, with CGD and in the murine models of the disease, fatal oppor- and because medical treatment for trichosporonosis differs from tunistic infections typically ensue (10, 19, 20, 26). Relevant to that for candidiasis (9). Trichosporon sp. is more susceptible to the spontaneous occurrence of T. beigelii in the p47phox-deficient the azoles, whereas Candida sp. is also susceptible to amphoteri- mouse are reports of spontaneous Trichosporon (Sarcinosporin) cin B (14). Murine models suggest that fluconazole is superior to inkin infection in patients with CGD (13, 17), confirming the amphotericin B and that the combination of amphotericin B identity of infection susceptibility between the mouse and hu- deoxycholate plus fluconazole is superior to either agent alone in man forms of this immunodeficiency. the treatment of trichosporonosis (1). Even with antifungal These cases of spontaneous Trichosporon infection in human therapy, fungemia associated with Trichosporon sp. is fatal in 65 and murine cases of CGD indicate that the NADPH oxidase is a to 90% of human cases (9). critical pathway for Trichosporon killing. Although there was Granulocytopenia is the most commonly recognized risk factor genetic contamination of this mouse line with an IFN-γ-defi- for disseminated trichosporonosis (18). Absolute neutrophil counts cient line, and several of these animals were heterozygous for are thought to play a paramount role in normal host defense the IFN-γ defect, this does not explain our findings. Only three against disseminated trichosporonosis (8), and resolution of neu- of the five mice carried a heterozygous defect in IFN-γ. Also, tropenia improves prognosis for survival (6). Drugs that induce spontaneous infection with Candida or Trichosporon species neutropenia (e.g., cyclophosphamide) or cause neutrophil dysfunc- has not been seen in patients with complete homozygous defects tion (e.g., corticosteroids) have been used extensively by investiga- of the IFN-γ gene (5). Therefore, although it is possible that loss tors developing murine models of trichosporonosis (1, 2, 8, 12, 18, of one IFN-γ allele may have slightly modified disease presenta- 27, 28). However, this ubiquitous organism has not caused sponta- tion, the finding of spontaneous Trichosporon infection in ani- neous disease in these susceptible animals; instead, the organism mals that were IFN-γ intact but NADPH oxidase deficient is typically administered intravenously in large doses. By contrast, clearly indicates that the NADPH oxidase is a critical pathway spontaneous disseminated trichosporonosis in humans probably responsible for control of these emerging pathogens. results from a modest infective dose, with the respiratory and gas- trointestinal tracts believed to be the most common ports of entry (27). A model of progressive disseminated trichosporonosis was cre- Acknowledgments ated by inducing latent trichosporonemia with a small infective We thank Dan Paré, Michael Eckhaus, and Georgina Miller for pro- dose in neutropenic mice, followed by re-immunosuppression at vision of the animals reported here and for critical review of the manu- script. We also thank Fran Benedetti for microbiological support and variable time points after infection (27). In this model, small Annie Merriweather for histology support, without which, this manu- numbers of fungi persist without causing signs of disease until script would not have been possible. immunosuppression is exacerbated or repeated up to three weeks later (27). This more accurately reflects what may happen clinically, when a patient is episodically immunosuppressed by References use of myeloablative chemotherapy. 1. Anaissie, E. J., R. Hachem, N. C. Karyotakis, A. Gokaslan, Mononuclear phagocyte function is necessary for resistance to M. C. Dignani, L. C. Stephens, and C. K. Tin-U. 1994. Com- systemic mycoses. Macrophage colony-stimulating factor (M-CSF) parative efficacies of amphotericin B, triazoles, and combination of both as experimental therapy for murine trichosporonosis. enhanced the fungicidal activity of mononuclear phagocytes Antimicrob. Agents. Chemother. 38:2541-2544. against Trichosporon asahii in neutropenic mice (18), by up- 2. Aratani, Y., F. Kura, H. Watanabe, H. Akagawa, Y. Takano, regulation of tumor necrosis factor-alpha (TNF-α). There was a K. Suzuki, N. Maeda, and H. Koyama. 2000. Differential host marked decrease in survival when TNF-α blocking antibody susceptibility to pulmonary infections with bacteria and fungi in was co-administered. Consistent with these results, administra- mice deficient in myeloperoxidase. J. Infect. Dis. 182:1276-1279. 306 Disseminated trichosporonosis in murine granulomatous disease

Figure 2. Role of NADPH oxidase in reactive oxidant generation. The NADPH oxidase enzyme consists of four structural proteins: two cytosolic proteins, p47phox and p67phox, and two membrane bound proteins, p22phox, and heavy chain gp91phox. The enzyme transfers an electron from cyto- - plasmic NADPH to extracellular or vacuolar oxygen to form superoxide radical (O2 ), which is converted to hydrogen peroxide (H2O2) by superox- - - ide dismutase. Peroxynitrite anion (OOON ) may form in the presence of NO. H2O2 is converted to hydroxyl anion (OH ) by the Fenton reaction, and to hypochlorous acid (HOCl) by myeloperoxidase. HOCl can then be converted to nitryl chloride (NO2Cl). A defect in any one of the structural proteins of NADPH oxidase results in chronic granulomatous disease. Adapted from Segal et al (22).

3. Chandler, F. W. and J. C. Watts. 1987. Trichosporonosis, p. 113- 7. Gonzalez, R. N., D. J. Wilson, S. A. Sickles, M. J. Zurakowski, 116. In F. W. Chandler and J. C. Watts, Pathologic diagnosis of P. M. Weybrecht, and A. K. Walsh. 2001. Outbreaks of clinical fungal infections. ASCP Press, Chicago. mastitis caused by Trichosporon beigelii in dairy herds. J. Am. 4. Chaumentin, G., A. Boibieux, M. A. Piens, C. Douchet, P. Vet. Med. Assoc. 218:238-242. Buttard, J. L. Bertrand, and D. Peyramond. 1996. Trichos- 8. Hospenthal, D., T. Belay, P. Lappin, A. Rogers, and M. poron inkin endocarditis: short-term evolution and clinical re- Kennedy. 1993. Disseminated trichosporonosis in a neutropenic port. Clin. Infect. Dis. 23:396-397. murine model. Mycopathologia 122:115-122. 5. Dorman, S. E. and S. M. Holland. 2000. Interferon-gamma and 9. Itoh, T., H. Hosokawa, U. Kohdera, N. Toyazaki, and Y. interleukin-12 pathway defects and human disease. Cytokine Asada. 1996. Disseminated infection with Trichosporon asahii. Growth Factor Rev. 11:321-333. Mycoses 39:195-199. 6. Gokaslan, A. and E. Anaissie. 1992. A novel murine model of 10. Jackson, S. H., J. I. Gallin, and S. M. Holland. 1995. The p47phox disseminated trichosporonosis. Infect. Immun. 60:3339-3344. mouse knock-out model of chronic granulomatous disease. J. Exp. Med. 182:751-758.

307 Vol 53, No 3 Comparative Medicine June 2003

11. Jones, T. C., R. D. Hunt, and N. W. King. 1996. Diseases caused 19. Segal, B. H., E. S. DeCarlo, K. J. Kwon-Chung, H. L. Malech, by fungi, p. 505-547. In T. C. Jones, R. D. Hunt, and N. W. King, J. I. Gallin, and S. M. Holland. 1998. Aspergillus nidulans in- Veterinary pathology. Williams & Wilkins, Baltimore. fection in chronic granulomatous disease. Medicine 77:345-354. 12. Kahn, Z. U., V. C. Misra, H. S. Randhawa, and V. N. Damodaran. 20. Segal, B. H., T. L. Leto, J. I. Gallin, H. L. Malech, and S. M. 1980. Pathogenicity of some ordinarily harmless yeasts for corti- Holland. 2000. Genetic, biochemical, and clinical features of sone-treated mice. Sabouraudia 18:319-327. chronic granulomatous disease. Medicine 79:170-200. 13. Kenney, R. T., K. J. Kwon-Chung, F. G. Witebsky, D. A. 21. Sheikh, H. A., S. Mahgoub, and K. Badi. 1974. Postoperative Melnick, H. L. Malech, and J. I. Gallin. 1990. Invasive infec- endophthalmitis due to Trichosporon cutaneum. Br. J. Ophthalmol. tion with Sarcinosporon inkin in a patient with chronic granulo- 58:591-594. matous disease. Am. J. Clin. Pathol. 94:344-350. 22. Shimazu, K., M. Ando, T. Sakata, K. Yoshida, and S. Araki. 14. Moretti-Branchini, M. L., K. Fukushima, A. Z. Schreiber, 1984. Hypersensitivity pneumonitis induced by Trichosporon K. Nishimura, P. M. O. Papaiordanou, P. Trabasso, R. cutaneum. Am. Rev. Respir. Dis. 130:407-411. Tanaka, and M. Miyaji. 2001. Trichosporon species infection in 23. Sugita, T., M. Nakajima, R. Ikeda, Y. Niki, T. Matsushima, bone marrow transplanted patients. Diagn. Microbiol. Infect. Dis. and T. Shinoda. 2001. A nested PCR assay to detect DNA in 39:161-164. sera for the diagnosis of deep-seated trichosporonosis. Microbiol. 15. National Research Council. 1996. Guide for the care and use Immunol. 45:143-148. of laboratory animals. National Academy Press, Washington, D.C. 24. Walsh, T. J. 1989. Trichosporonosis. Infect. Dis. Clin. North Am. 16. Nemunaitis, J., K. Shannon-Dorcy, F. R. Appelbaum, J. 3:43-51. Meyers, A. Owens, R. Day, D. Ando, C. O’Neill, D. Buckner, 25. Watson, K. C. and S. Kallichurum. 1970. Brain abscess due to and J. Singer. 1993. Long-term follow-up of patients with inva- Trichosporon cutaneum. J. Med. Microbiol. 3:191-193. sive fungal disease who received adjunctive therapy with recom- 26. Winkelstein, J. A., M. C. Marino, R. B. Johnston, Jr., J. Boyle, binant human macrophage colony-stimulating factor. Blood J. Curnutte, J. I. Gallin, H. L. Malech, S. M. Holland, H. 82:1422-1427. Ochs, P. Quie, R. H. Buckley, C. B. Foster, S. J. Chanock, 17. Piwoz, J. A., G. J. Stadtmauer, E. J. Bottone, I. Weitzman, and H. Dickler. 2000. Chronic granulomatous disease: report E. Shlasko, and C. Cunningham-Rundles. 2000. Trichosporon on a national registry of 368 patients. Medicine 79:155-169. inkin lung abscesses presenting as a penetrating chest wall mass. 27. Yamagata, E., P. Kamberi, Y. Yamakami, A. Hashimoto, and Pediatr. Infect. Dis. J. 19:1025-1027. M. Nasu. 2000. Experimental model of progressive disseminated 18. Sasaki, E., T. Tashiro, M. Kuroki, M. Seki, Y. Miyazaki, S. trichosporonosis in mice with latent trichosporonemia. J. Clin. Maesaki, K. Tomono, J. Kadota, and S. Kohno. 2000. Effects Microbiol. 38:3260-3266. of macrophage colony-stimulating factor (M-CSF) on anti-fungal 28. Yamamoto, K., K. Makimura, T. Sudo, K. Shibuya, K. Uchida, activity of mononuclear phagocytes against Trichosporon asahii. and H. Yamaguchi. 1997. Experimental disseminated Clin. Exp. Immunol. 119:293-298. trichosporonosis in mice: tissue distribution and therapy with antifungal agents. J. Med. Vet. Mycol. 35:411-418.

308