isolation of paracoccidioides brasiliensis from

Am. J. Trop. Med. Hyg., 58(4), 1998, pp. 505–512 Copyright q 1998 by The American Society of Tropical Medicine and Hygiene

ISOLATION OF PARACOCCIDIOIDES BRASILIENSIS FROM ARMADILLOS (DASYPUS NOVEMINCTUS) CAPTURED IN AN ENDEMIC AREA OF PARACOCCIDIOIDOMYCOSIS EDUARDO BAGAGLI, AYAKO SANO, KUNIE IABUKI COELHO, SISLAINE ALQUATI, MAKOTO MIYAJI, ZOILO PIRES DE CAMARGO, GLAUCE MARY GOMES, MARCELLO FRANCO, AND MARIO RUBENS MONTENEGRO Departamento de Microbiologia e Imunologia, Instituto de Biocieˆncias, Universidade Estadual Paulista (UNESP), Botucatu, Sao Paulo, Brazil; Research Center for Pathogenic Fungi and Microbial Toxicoses, Chiba University, Chiba, Japan; Departamento de Patologia, Faculdade de Medicina, UNESP, Botucatu, Sao Paulo, Brazil; Disciplina de Biologia Celular, Universidade Federal de Sao Paulo, Sao Paulo, Brazil

Abstract. Paracoccidioides brasiliensis, the causative agent of paracoccidioidomycosis (PCM), was first isolated from armadillos from the Amazonian region where the mycosis is uncommon. In the present study, we report on the high incidence of PCM infection in armadillos from a hyperendemic region of the disease. Four nine-banded armadillos (Dasypus novemcinctus) were captured in the endemic area of Botucatu, Sao Paulo, Brazil, killed by manual cervical dislocation and autopsied under sterile conditions. Fragments of lung, spleen, liver, and mesenteric lymph nodes were processed for histology, cultured on Mycosel agar at 378C, and homogenized for inoculation into the testis and peritoneum of hamsters. The animals were killed from week 6 to week 20 postinoculation and fragments of liver, lung, spleen, testis, and lymph nodes were cultured on brain heart infusion agar at 378C. Paracoccidioides brasiliensis was isolated from three armadillos both by direct organ culture and from the liver, spleen, lung, and mesenteric lymph nodes of hamsters. In addition, one positive armadillo presented histologically proven PCM disease in a mesenteric lymph node. The three armadillos isolates (Pb-A1, Pb-A2, and Pb-A4) presented thermodependent dimorphism, urease activity, and casein assimilation, showed amplification of the gp43 gene, and were highly virulent in intratesticularly inoculated hamsters. The isolates expressed the gp43 glycoprotein, the immunodominant antigen of the fungus, and reacted with a pool of sera from PCM patients. Taken together, the present data confirm that armadillos are a natural reservoir of P. brasiliensis and demonstrate that the animal is a sylvan host to the fungus. by manual cervical dislocation and autopsied under sterile conditions. Histopathologic observations. Fragments of the lung, spleen, mesenteric lymph nodes, and liver were fixed in 10% formalin, processed for histology, and stained with hematoxylin and eosin, periodic acid–Schiff, and Gomori-Grocott methanamine silver stain. Direct organ culture. Six to nine 8–60 mm3 fragments of the lung, spleen, mesenteric lymph node, and liver were placed on 15 3 90 mm Mycosel agar (BBLt; Becton Dickinson and Co., Cockeysville, MD) plates supplemented with gentamicin (40 mg/ml) at 378C for three weeks. The plates were checked daily for the detection of growing cerebriform yeast colonies on the fragments. Animal inoculation. The remaining liver, lung, and spleen (about 10 g each) were homogenized together in a ceramic mortar with 150 ml of phosphate-buffered saline (PBS) (5 ml/ g of tissue). After thorough mixing and filtration through gauze, 0.2 ml or 1.0 ml of the suspension was inoculated with a hypodermic needle into the right testis (IT) or intraperitoneally (IP) in two groups of 20 eight-week-old adult male hamsters, respectively. Groups of animals were killed from week 6 to week 20 postinoculation and small fragments of liver, lung, spleen, testis, and lymph nodes collected aseptically were plated onto 1% brain heart infusion dextrose agar (BHIDA) containing chloramphenicol (50 mg/ml), gentamicin (50 mg/ml), and cyclohexamide (400 mg/ ml) and supplemented with horse serum (4%) and growthpromoting factor for P. brasiliensis (5%), obtained from the supernatant of old liquid cultures, at 35–378C.16 Samples of all the cerebriform yeast colonies obtained from direct organ cultures or from cultures of the organs of inoculated hamsters were subjected to mycologic, physio-

The ecologic niche of Paracoccidioides brasiliensis, the etiologic agent of paracoccidioidomycosis (PCM), has not been determined and has proved to be a difficult task for mycologists.1 Several factors have contributed to this difficulty, such as the rare isolation of the fungus from the environment,2–6 the prolonged latency period of the disease,7 and the lack of reports of the disease in domestic or wild animals8, 9 and of outbreaks.10, 11 Naiff and others reported on the isolation of P. brasiliensis from nine-banded armadillos (20% of the studied animals) in Tucurui, Para, Brazil, which is located in the Amazon region, considered to be a nonendemic area of the disease.12 In the present study, we confirm these findings and describe the frequent isolation of the fungus from nine-banded armadillos from a hyperendemic region of the disease. We believe that armadillos, which are animals that live in welldefined environmental conditions with no apparent migration habits,13, 14 may represent an important biological indicator for the elucidation of the ecologic niche of P. brasiliensis and as a harbinger of PCM. MATERIALS AND METHODS

Armadillos and sites of capture. Two adult male (armadillos 1 and 2) and two adult female (armadillos 3 and 4) nine-banded armadillos were studied. Armadillos 1, 2, and 3 were captured in Pratania County between January and May 1996, and armadillo 4 was captured in Manduri County in May 1996. Pratania and Manduri are located 100 km apart and belong to the same hyperendemic area of Botucatu (22859’S, 48826’W, average altitude 5 800 m).15 The animals were captured under a license from the Brazilian Federal Environmental Protection Agency. The animals were killed

505

506

BAGAGLI AND OTHERS

logic, immunologic, and genetic studies and also tested for virulence. Mycologic studies. The criteria for the identification of the isolates included colonial and cellular morphology (mycelium and yeast cultures, giant colony, slide culture), yeastto-mycelium-to-yeast transformation, and physiologic features (casein assimilation, urease and gelatinase activity). Fragments of the colonies were stained with lactophenol cotton blue and observed under the light microscope for the detection of multiple budding cells (MB). The MB-positive sprouts were transferred onto BHIDA slants at 378C and Mycosel agar, 2% Sabouraud dextrose agar (SDA), and potato dextrose agar (PDA) slants at room temperature (approximately 258C) for gross observations of both yeast and mycelial form colonies. Giant colonies were evaluated on SDA and PDA plates at room temperature for two months. Fourweek-old slide cultures on SDA and PDA at room temperature stained with lactophenol cotton blue were observed under the light microscope. Mycelial cultures on PDA or SDA slants were inoculated on BHIDA slants at 378C for 14 days for yeast form conversion, and yeast cultures on BHIDA slants were transferred onto SDA and PDA slants at room temperature for 14 days to revert to the mycelial form. The isolates were cultured at 378C for one week on Christensen urea agar slants (Eiken, Tokyo, Japan) to determine urease activity. We also evaluated casein assimilation on skim milk agar plates (50 g/L of skim milk and 15 g/L of agar) and gelatinase activity in gelatin tubes (40 g/L of gelatin) at room temperature for two months. Immunologic studies. Antigen preparation. Samples of isolates were cultured in PYG medium (5 g of bactopeptone, 5 g of yeast extract, and 20 g of glucose in 1,000 ml of deionized water) for seven days under constant shaking (150 rpm) at 358C. The fungal suspensions, maintained overnight at 48C in merthiolate (0.5 g/L), were then washed three times in sterile saline by centrifugation at 3,000 rpm for 5 min. The fungal mass was resuspended in an equal volume of sterile saline containing 50 mM phenylmethylsulfonyl fluoride (20 mg/ml) (Sigma, St. Louis, MO) as a protease inhibitor, and subjected to overnight to cell disruption using glass beads with magnetic agitation at 48C. After centrifugation at 10,000 rpm for 40 min, the supernatant was dialyzed against distilled water for 48 hr, lyophilized, and stored at 2208C. At the time of use, the protein concentration of the samples was determined.17 Human isolate Pb-18 of P. brasiliensis was used as a control. Sodium dodecyl sulfate–polyacrylamide gel electrophoresis. The antigens were subjected to electrophoresis on a 10% polyacrylamide gel in the presence of 0.2% sodium dodecyl sulfate (SDS) in a discontinuous buffer system.18 Lyophilized samples were dissolved in buffer containing 2% SDS, 0.7 M b-mercaptoethanol, 0.002% bromphenol blue, 10% glycerol, and 60 mM Tris-HCl, pH 6.8 and then heated at 1008C for 2 min. Approximately 125 mg of protein from each sample was applied to the gel. Electrophoresis was performed at constant voltage (120 V) at room temperature for 4 hr. The polypeptide banding pattern was visualized by staining the gel with Coomassie blue and compared with a molecular mass standard in kilodaltons (Sigma). Western blotting. After separation, the protein components were electrotransferred to nitrocellulose paper, which was

then cut into strips.19 The paper strips were saturated in 10 mM PBS, pH 7.2, with 5% skim milk (PBS-M) and incubated at room temperature with a pool of human sera from patients with proven cases of active PCM (pool of PCM sera), obtained from the University Hospital of the Botucatu Medical School, diluted in PBS-M. After the strips were washed with PBS, they were incubated with peroxidase-conjugated anti-human IgG (Sigma) diluted 1:4,000 in PBS-M for 5 hr at room temperature, and then washed with PBSM. The peroxidase activity was developed with a preparation of 5 mg of 3,3’-diaminobenzidine (Sigma) in 30 ml of PBS, pH 7.2, and 150 ml of 30% H2O2 (Sigma). After color development, the strips were rinsed in distilled water, dried, and photographed. Immunodiffusion. The agar-gel immunodiffusion test (ID) was performed using 1% agar (Merck, Darmstadt, Germany) in PBS.20 The antigens prepared from the armadillo isolates were used at a concentration of 5 mg/ml, and the pool of PCM sera was tested at a 1:16 dilution. Slides were incubated overnight in a moist chamber at room temperature and washed for 1 hr in distilled water and then for 24 hr in PBS. After drying, they were stained for 10 min in 0.15% Coomassie blue (Sigma) in ethanol:acetic acid:water (4:2:4) and destained in the solvent mixture whenever necessary. Precipitation bands were recorded. Genetic studies. Preparation of genomic DNAs. Genomic DNAs were extracted by adding 1 ml of TES buffer (50 mM EDTA, 20% glucose, 10 mM Tris, pH 8.0), 5 ml of b-mercaptoethanol, and 5 mg of novozyme (Novo Nordisk, Bagsvaerd, Denmark) to yeast cell pellets.21 The mixture was vortexed using six 30-sec bursts and incubated at 308C for 2 hr and at 2708C for 10 min in dry ice plus ethanol. The mixture was centrifuged at 4,000 3 g for 5 min at room temperature. The pellet was added to 0.5 ml of TES buffer plus 0.5 ml of 10% SDS, incubated at 658C for 30 min, and 200 ml of 3 M sodium acetate and 50 ml of RNAse (10 mg/ml) were added. The mixture was incubated at 378C for 30 min and then centrifuged at 10,000 3 g at 48C for 10 min. The final supernatant was mixed with the same volume of phenolchloroform. The mixture was incubated at 48C for 15 min with shaking, centrifuged at 12,000 3 g at 48C for 10 min, and the supernatant was transferred to a new microcentrifuge tube, and further extracted with phenol-chloroform. The final supernatant was added (1:10 [v/v]) to 3 M sodium acetate plus 2.5 volumes of absolute ethanol (08C), and incubated at 2708C for 4 hr and then centrifuged at 12,000 3 g at 48C for 2 min to precipitate the nucleic acids. The DNA was washed with 1.0 ml of 70% alcohol (08C) and centrifuged at 12,000 3 g at 48C for 10 min. The pellet (DNA) was added to 100 of ml TE buffer (1M Tris-HCl, pH 8.0, 0.5 M EDTA, pH 8.0) and incubated at 658C for 5 min; an aliquot was removed to determine nucleic acid purity and concentration by spectroscopy at 260 and 280 nm. Polymerase chain reaction (PCR) profile. The primers (30 bases each) were obtained from the P. brasiliensis gp43 gene sequence.22 The PCR was carried out as follows. Every PCR mixture contained 20 ng of genomic DNA, 12.5 pmoles of each primer, 0.2 mM of dNTP mixture (Pharmacia, Uppsala, Sweden), and 1 U of Taq polymerase (Pharmacia) in the buffer provided by the manufacturer. The final volume of the reaction mixture was 25 ml. After mixing, the tubes were

ISOLATION OF P. BRASILIENSIS FROM ARMADILLOS

507

FIGURE 1. Histopathology of a mesenteric lymph node from armadillo 4. Note the epithelioid granulomas containing yeast-like cells consistent with Paracoccidioides brasiliensis (arrows). a, hematoxylin and eosin stained, magnification 3 200. b, Gomori-Grocott methenamine silver stained, magnification 3 400.

placed in a thermal cycler (Perkin-Elmer Cetus, Norwalk, CT) and programmed as follows: one cycle at 948C for 3 min followed by 25 cycles at 948C for 1 min, 508C for 1 min, and 728C for 1.5 min, followed by a cycle at 728C for 7 min. The reaction product was analyzed by electrophoresis on 0.8% (w/v) agarose (agarose M, Mr 5 0.13; 17-0422202; Pharmacia) gel slabs (10 cm 3 16 cm 3 6 mm) with TAE buffer (40 mM Tris base, 5.7% acetic acid, 1 mM EDTA, pH 8.0) at 100 V for 1 hr. Gels were stained with ethidium bromide, placed over a source of UV light, and photographed (667 film; Polaroid, Cambridge, MA). The molecular size of the DNA fragment was determined in relation to a molecular size standard derived from a 100-basepair ladder (Pharmacia). Virulence in hamsters. Groups of five eight-week-old male outbred golden hamsters were inoculated IT with 2 3 106 colony-forming units of each of the cerebriform yeast colonies isolated from the armadillos. The animals were killed at week 4 postinoculation; fragments of the testis, liver, lung, heart, kidney, and spleen were cultured on Mycosel agar plates at 378C for three weeks and processed for histopathologic analysis as described above. RESULTS

Environmental features of the sites where the armadillos were captured. The sites of capture of armadillos 1 and 2 were separated by a distance of 150 m and consisted

of a 15–25 m-wide riparian forest around a lake, which contained a large amount of submerged decaying wood. The sites were located near an agricultural area for cotton, corn and beans, and a Pinus wood processing sawmill. The site of capture of armadillo 3 was covered with low vegetation consisting of grasses, herbs, and sparse scrubs close to an artificial lake with submerged decaying wood. It was close to a pasture field and an agricultural area for corn, beans, and sunflowers, and was located 3 km from the above sites. Sandy acid soil (pH 5.3) was present in all locations. Armadillo 4 was captured in a site located along a riverbank, surrounded by a highly disturbed riparian forest in which natural trees were mixed with a newly planted eucalyptus forest. Pasture fields and sugar cane and coffee plantations were located nearby. This site was located 100 km from the previous sites and presented an alfisol type of soil. Autopsy findings. Small whitish lesions were observed in the lungs of armadillos 1, 2, and 4, and enlarged mesenteric lymph nodes were observed in armadillos 2 and 4. The autopsy of armadillo 3 was unremarkable. Histopathology. Epithelioid granulomas were observed in the liver of armadillos 2 and 4 and in the spleen of armadillo 2; no fungal structures were observed on hematoxylin and eosin–, periodic acid–Schiff–, or Gomori-Grocott methenamine– stained slides. There were epithelioid granulomas containing fungal cells morphologically similar to P. brasiliensis (Figure 1) in the mesenteric lymph nodes of armadillo 4. No granulomas were observed in the organs of armadillos 1 and 3.

508

BAGAGLI AND OTHERS

TABLE 1 Isolation of Paracoccidioides brasiliensis from spleen, liver, lung, and lymph nodes from four armadillos by direct organ culture* Number of colonies/Number of fragments cultured Armadillo

Liver

Spleen

Lung

Lymph node

1 2 3 4

0/306 4/320 0/656 12/296

1/165 6/136 0/80 23/152

0/206 ND 0/248 0/184

ND 12/20 ND ND

* ND 5 not done.

Isolation of P. brasiliensis. Paracoccidioides brasiliensis isolates were obtained from three armadillos (Pb-A1, Pb-A2, and Pb-A4) by direct organ culture. The number of fragments plated from each organ and the number of P. brasiliensis colonies isolated are presented in Table l. Colonies were isolated from the spleen of all three animals, from the liver of two animals, and from the mesenteric lymph nodes of one armadillo. Armadillo 4 was the most heavily infected animal. Paracoccidioides brasiliensis was also isolated from organs of the hamsters inoculated IT or IP with organ macerates from armadillos 1, 2, and 4. The two routes of inoculation resulted in good fungal recovery from spleen and liver, and less frequently from the lungs. The fungus was frequently isolated from the testis, which was evaluated only in the IT group. Mycologic studies. The three P. brasiliensis isolates cultured on SDA slants at room temperature showed colonies with a white cotton-like surface with or without central fissures; the reverses were brownish in color. On PDA slants,

the cultures showed a white to yellow cotton-like surface and the reverses were dark brown or dark wine red. The yeast form cultures on BHIDA slants were cream colored with a typical cerebriform aspect (Figure 2). Mycelial growth was slow on both SDA and PDA plates, with mean diameters of giant colonies ranging from 10 to 21 mm on SDA and from 23 to 32 mm on PDA. Chlamydospores, arthroconidia, and aleuroconidia were observed in material cultured on SDA and PDA at room temperature for four weeks (Figure 3). Multiple budding consisting of mother cells (10–50 mm in diameter) with daughter cells (1–20 mm in diameter) attached through a narrow base was observed in yeast forms cultured on BHIDA slants at 378C for three days in all three isolates (Figure 4). Thermodependent conversion from mycelium to yeast and the reverse, urease activity, and casein assimilation were observed in all three isolates. Only the Pb-A1 isolate was positive for gelatinase activity. Immunodiffusion. The antigens of the three armadillo isolates and of a control human isolate (Pb-18) reacted with a pool of human PCM sera. Immunoblotting. The pool of human PCM sera strongly recognized the gp43-kD band in all the armadillo isolates and in Pb-18 used as control. Furthermore, several other bands were observed in the armadillo isolates (Figure 5). Genetic studies. The three armadillo isolates showed positive amplification of the gp43 gene in a pattern similar to that of the control isolate (Figure 6). Pathogenicity in hamsters. No mortality occurred among the hamsters inoculated with the three armadillo isolates dur-

FIGURE 2. Colony morphology of Paracoccidioides brasiliensis isolates Pb-A1 (a), Pb-A2 (b), and Pb-A4 (c) cultured on brain heart infusion dextrose agar at 378C (left) and Sabouraud dextrose agar (right) at room temperature. Note the dimorphic features of the isolates.


509

FIGURE 3. Slide cultures of Paracoccidioides brasiliensis isolate Pb-A1 cultured on potato dextrose agar at room temperature for four weeks: chlamydospores (a) and arthroconidia (b) (arrows). (Lactophenol cotton blue stained, magnification 3 200.)

ing the observation period. Mild-to-severe testicular enlargement and small white spots in the liver were observed in all infected animals. Histology showed that all inoculated testes developed confluent epithelioid granulomata with a necrotic center and numerous fungal yeast cells with multiple budding. Disseminated lesions were detected in lymph nodes (Figure 7), liver, lungs, spleen, and adrenal glands. DISCUSSION

The present study confirms the occurrence of PCM infection in armadillos.12 Since we studied animals captured in a hyperendemic region of PCM, the frequency of wild infection was higher (75%) compared with the data reported by Naiff and others (20%).12 The fungus was isolated from the

liver and spleen of the animals, as previously reported,12 as well as from the mesenteric lymph nodes, further documenting the systemic nature of the infection. The fungus was recovered from all of these viscera by animal inoculation of organ macerates, as also used by Naiff and others,12 and by direct culture of small organ fragments. The three new armadillo isolates were identified as P. brasiliensis by mycologic, immunologic, and molecular techniques. In addition, they were virulent to hamsters as were the Amazonian isolates of Vidal and others.23 The three isolates presented the gp43 gene and expressed the corresponding protein, considered to be the immunodominant antigen of the fungus. An additional interesting finding was the histopathologic evidence of active PCM disease, involving a mesenteric

FIGURE 4. Multiple budding yeast cells in Paracoccidioides brasiliensis isolates Pb-A1 (a), Pb-A2 (b), and Pb-A4 (c) (lactophenol cotton blue stained, original magnification 3 400).

510

BAGAGLI AND OTHERS

FIGURE 5. Imunoblotting of Paracoccidioides brasiliensis antigens recognized by a pool of paracoccidioidomycosis human sera. Lane A, Pb-18 control isolate; lanes B, C, and D, isolates (Pb-A1, Pb-A2, and Pb-A4, respectively). kD 5 kilodaltons.

lymph node in one of the armadillos. This indicates that the animals are not only healthy carriers of the fungus, as previously demonstrated,12 but can also develop PCM disease. This finding was recently confirmed by a report on an armadillo captured in a coffee farm close to our endemic area that presented paracoccidioidal granulomas in the lungs (Silva MLV, University of Sao Paulo, unpublished data). Paracoccidioides in armadillos comes as no surprise since these animals have a low body temperature,14 weak cell-mediated immunity,24 and have been previously indicated as model organisms for several other infectious disease.25 The fungus has never been regularly isolated from other animals, although evidence of infection, as detected by paracoccidoidin skin testing,8, 9 was presumed to occur both in domestic and wild animals. The previous reports of isolation of P. brasiliensis from the feces of an Antarctic penguin,26 from the intestinal tract of three fructivorous bats (Artibeus lituratus),3 and from a biopsy specimen of a squirrel monkey (Saimiri sciureus)27 have never been confirmed by additional animal experiments or field work. The confirmation of armadillos as a natural reservoir of PCM may provide the first valuable clues in the search for the natural habitat of P. brasiliensis. Nine-banded armadillos can be found from the southcentral United States to Argentina west of the Andean ridge, a distribution that has many points in common with that of human PCM.11 The armadillos have a deep contact with soil due to their constant digging activity, both to look for food and to construct burrows to a depth of 4–6 m in which they convey plant material to build

FIGURE 6. Specific amplification of a 1.03-kilobase (kb) fragment from genomic DNAs of Paracoccidioides brasiliensis isolates Pb-A1 (B), Pb-A2 (C), and Pb-A4 (D) compared with the control Pb B-339 (E). Lane A: size markers (100-basepair ladder, Pharmacia).

nests.14, 15 The climatic features of armadillo holes are similar to those areas in which PCM is endemic; the temperature at this depth is relatively stable (approximately 208C) and humidity approaches 70–80% throughout the year.11 Armadillos wander, mainly during the night, along riverbank trails located in disturbed riparian forests. They are also found in areas with low vegetation, pastures, and crop plantations.14, 15 Furthermore, these animals apparently do not migrate from their habitats. These simple behavioral habits make it easy to circumscribe the area where the animals acquire the fungus in nature. The environmental features of the places where the animals were captured strongly support previous hypotheses11, 28, 29 that the natural habitat of P. brasiliensis is located in vegetation sites disturbed by humans near water sources. Blastomyces dermatitidis, considered to be a species related to P. brasiliensis both in terms of ecology and epidemiology,30, 31 has a microniche also associated with bodies of water.32, 33 Armadillos, similar to dogs for blastomycosis,34 may be considered good sentinel animals for locating places in which P. brasiliensis is present, since they are frequently infected when living in areas endemic for PCM. Armadillos, like humans, are probably infected by airborne P. brasiliensis conidia. It is unlikely that the disease is transmitted to humans by the animals since the yeast form of the fungus, as found in armadillo tissues, is of low infectivity. However, rural workers, the exposed population in endemic areas, have close contact with this animal, which is

511


FIGURE 7. Histopathology of a hamster infected with the Paracoccidioides brasiliensis Pb-A1 isolate. Note the confluent epithelioid granuloma in the lymph node. a, hematoxylin and eosin stained, magnification 3 200. Typical P. brasiliensis multiple budding cells in necrotic areas of the testis. b, periodic acid–Schiff stained, magnification 3 400. c, Gomori-Grocott methenamine silver stained, magnification 3 200.

considered to be excellent game food. Both humans and armadillos are probably infected by a common environmental source, most likely located in disturbed gallery forests. The ecosystem inhabited by armadillos, an underground protected niche, with soil and plant debris, may have played a relevant role in the evolution and natural adjustment to a condition of pathogenicity of P. brasiliensis. The same role has been attributed to other soil-inhabiting animals such as rodents for Penicillium marneffei.35 Paracoccidioides brasiliensis presents some features of a species well adjusted to living in animal tissues, such as a strong tendency to induce chronic infection in the host and a small conidia production in the saprophytic phase,7, 10, 11 similar to what has been observed in other fungi, such as zoophilic and anthropophilic dermatophytic species.36 This possible zoophilic condition of P. brasiliensis in armadillos could also explain its low resistance to dryness37 and low ability to survive as free organisms in natural unsterilized soil,1 which might explain the great difficulty in recovering the fungus from nature. We are carrying out further investigations by studying armadillos from different endemic regions and comparing the animal isolates with those obtained from patients. In addition, we will use primers of the gp43 gene to detect fungal DNA in soil and plant debris samples collected in armadillo holes. Acknowledgments: We express our special thanks to farmers, rural workers, and technicians involved in the capture of the armadillos and the laboratory procedures. Financial support: This work was partly supported by Fundac¸a˜o de Amparo a` Pesquisa do Estado de Sao Paulo and Conselho Nacional de Desenvolvimento Cientifico e Technolo´gico. Ayako Sano was a

Visiting Professor at the Faculdade de Medicina de Botucatu-Universidade Estadual Paulista (UNESP) during the project. Authors’ addresses: Eduardo Bagagli and Sislaine Alquati, Departamento de Microbiologia e Imunologia, Instituto de Biocieˆncias, UNESP, Botucatu, Sao Paulo, Brazil. Ayako Sano and Makoto Miyaji, Research Center for Pathogenic Fungi and Microbial Toxicoses, Chiba University, Chiba, Japan. Kunie Iabuki Coelho, Marcello Franco, and Mario Rubens Montenegro, Departamento de Patologia, Faculdade de Medicina, UNESP, Botucatu, Sao Paulo, Brazil. Zoilo Pires de Camargo and Glauce Mary Gomes, Disciplina de Biologia Celular, Universidade Federal de Sao Paulo, Sao Paulo, Brazil. REFERENCES

1. Borelli D, 1971. Some ecological aspects of paracoccidioidomycosis. Pan Am Health Organ Scientific Publication 245: 59–64. 2. Shome SK, Batista AC, 1963. Occurrence of Paracoccidioides brasiliensis in the soil of Recife, Brazil. Rev Fac Med Fed Ceara 3: 90–94. 3. Grose E, Tramsitt JR, 1965. Paracoccidioides brasiliensis recovered from the intestinal tract of three bats (Artibeus lituratus) in Colombia, S.A. Sabouraudia 4: 124–125. 4. Negroni P, 1966. El Paracoccidioides brasiliensis vive saprofiticamente en el suelo Argentino. Pren Med Argent 53: 2381– 2382. 5. Albornoz MCB, 1971. Isolation of Paracoccidioides brasiliensis from rural soil in Venezuela. Sabouraudia 9:248–253. 6. Ferreira MS, Freitas LH, Lacaz CS, Del Negro GM, Aielo NT, Garcia MN, Assis CM, Salebian A, Heins-Vaccari EM, 1990. Isolation and characterization of a Paracoccidioides brasiliensis strains from dog food probably contaminated with soil in Uberlandia, Brazil. J Med Vet Mycol 28: 253–256. 7. Montenegro MR, Franco M, 1994. Pathology. Franco M, Lacaz CS, Restrepo-Moreno A, Del Negro G, eds. Paracoccidioidomycosis. Boca Raton, FL: CRC Press, 131–149. 8. Costa EO, Diniz LSM, Fava Netto C, 1995. The prevalence of

512

9.

10.

11.

12.

13. 14.

15.

16.

17. 18. 19.

20.

21.

BAGAGLI AND OTHERS

positive intradermal reactions to paracoccidioidin in domestic and wild animals in Sao Paulo, Brazil. Vet Res Commun 19: 127–130. Costa EO, Diniz LSM, Fava Neto C, Arruda C, Dagli MLZ, 1995. Delayed hypersensitivity test with paracoccidioidin in captive Latin American wild mammals. J Med Vet Mycol 33: 39–42. Wanke B, Londero AT, 1994. Epidemiology and paracoccidioidomycosis infection. Franco M, Lacaz CS, Restrepo-Moreno A, Del Negro G, eds. Paracoccidioidomycosis. Boca Raton, FL: CRC Press, 109–119. Restrepo-Moreno A, 1994. Ecology of Paracoccidioides brasiliensis. Franco M, Lacaz CS, Restrepo-Moreno A, Del Negro G, eds. Paracoccidioidomycosis. Boca Raton, FL: CRC Press, 121–128. Naiff RD, Ferreira LCP, Barrte TV, Naiff MF, Arias JR, 1986. Paracoccidioidomicose enzootica em tatus (Dasypus novemcinctus) no Estado do Para. Rev Inst Med Trop Sao Paulo 28: 19–27. Taber FW, 1945. Contribution on the life story and ecology of the nine-banded armadillo. J Mammal 26: 211–216. Talmage RV, Buchanan GD, 1954. The armadillo Dasypus novemcinctus - a review of its natural history, ecology, anatomy and reproductive physiology. Rice Inst Pamphlet, Houston 41: 1–135. Marques SA, Franco MF, Mendes RP, Silva NCA, Baccili C, Curcelli ED, Feracin ACM, Oliveira CS, Tagliarini JV, Dillon NL, 1983. Aspectos epidemiologicos da paracoccidioidomicose na area endemica de Botucatu, Sao Paulo, Brasil. Rev Inst Med Sao Paulo 25: 87–92. Singer-Vermes LM, Ciavaglia MC, Kashino SS, Burger E, Calich VL, 1992. The source of growth-promoting factor(s) affects the plating-efficiency of Paracoccidioides brasiliensis. J Med Vet Mycol 30: 261–264. Lowry OH, Rosebrough NJ, Farr AL, Randall RJ, 1951. Protein measurement with the Folin phenol reagent. J Biol Chem 193: 265–275. Laemmli UK, 1970. Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227: 680– 685. Towbin H, Staehelin T, Gordon J, 1979. Electrophoretic transfer of proteins from polyacrylamide gels to nitrocellulose sheets: Procedure and some applications. Proc Natl Acad Sci USA 76: 4350–4354. Ouchterlony O, 1968. Diffusion-in-gel methods for immunological analysis. Ouchterlony O, ed. Handbook of Immunodiffusion and Immunoelectrophoresis. Ann Arbor, MI: Ann Arbor Science Publishers, Inc., 1–132. Sambrook J, Fritsch EF, Maniatis T, 1989. Molecular Cloning. A Laboratory Manual. Second edition. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory Press.

22. Cisalpino PS, Puccia R, Yamauchi LM, Cano MIN, Silveira JF, Travassos LR, 1996. Cloning, characterization, and epitope expression of the major diagnostic antigen of Paracoccidioides brasiliensis. J Biol Chem 271: 4553–4560. 23. Vidal MS, De Melo NT, Garcia NM, Del Negro GM, De Assis CM, Heins Vaccari EM, Naiff RD, Mendes RP, Lacaz CS, 1995. Paracoccidioides brasiliensis. A mycologic and immunochemical study of a sample isolate from an armadillo (Dasypus novemcinctus). Rev Inst Med Trop Sao Paulo 37: 43–49. 24. Ulrich M, Convit J, Centeno M, Rapetti M, 1976. Immunological characteristics of the armadillo, Dasypus sabanicola. Clin Exp Immunol 25: 170–176. 25. Storrs EE, Walsh GP, Burchfield HP, Binford CH, 1974. Leprosy in the armadillo: new model for biomedical research. Science 183: 851–852. 26. Camargo ZP, Taborda CP, 1993. Antigenic relationship between Paracoccidioides brasiliensis isolated from faeces of a penguin and a human isolate of P. brasiliensis. J Med Vet Mycol 31: 347–352. 27. Johnson WD, Lang CM, 1977. Paracoccidioidomycosis (South American blastomycosis) in a squirrel monkey (Saimiri sciureus). Vet Pathol 14: 368–371. 28. Conti-Diaz I, Rilla FD, 1989. Hipotesis sobre el nicho ecologico de Paracoccidioides brasiliensis. Rev Med Urugua 5: 97–103. 29. Cadavid D, Restrepo A, 1993. Factors associated with Paracoccidioides brasiliensis infection among permanent residents of three endemic areas in Colombia. Epidemiol Infect 111: 121–133. 30. Ajello L, 1967. Comparative ecology of respiratory mycotic disease agents. Bacteriol Rev 31: 6–24. 31. Schwarz J, 1986. Comparative epidemiology of four deep mycoses–a review. Mykosen 29: 5–9. 32. Klein BS, Vergeront JM, Disalvo AF, Kaufman L, Davis JP, 1987. Two outbreaks of blastomycosis along rivers in Wisconsin. Isolation of Blastomyces dermatitidis from riverbank and evidence of its transmission along waterways. Am Rev Respir Dis 136: 1333–1338. 33. Baumgardner DJ, Paretsky SP, Yopp AC, 1995. The epidemiology of blastomcycosis in dogs: north central Wisconsin, USA. J Med Vet Mycol 30: 171–176. 34. Sarosi GA, Eckman MR, Davies SF, Laskey WK, 1979. Canine blastomycosis as a harbinger of human disease. Ann Intern Med 91: 733–755. 35. Deng, G, Yun MA, Ajello L, 1986. Human penicilliosis marneffei and its relation to the bamboo rat (Rhizomys pruinosus). J Med Vet Mycol 24: 383–389. 36. Rippon JW, 1988. Medical Mycology. Philadelphia, W.B. Saunders Co. 37. Conti-Diaz IA, Mackinnon JE, Furcolow ML, 1971. Effect of drying on P. brasiliensis. Sabouraudia 9: 69–78.