Detection of Paracoccidioides brasiliensis gp70 Circulating Antigen ...

1 downloads 0 Views 107KB Size Report
Mar 22, 2004 - Silvia Helena Marques da Silva,1 Daniela de Mattos Grosso,1 José ...... Ferreira-da-Cruz, M. F., B. Galva˜o-Castro, and C. T. Daniel-Ribeiro.
JOURNAL OF CLINICAL MICROBIOLOGY, Oct. 2004, p. 4480–4486 0095-1137/04/$08.00⫹0 DOI: 10.1128/JCM.42.10.4480–4486.2004 Copyright © 2004, American Society for Microbiology. All Rights Reserved.

Vol. 42, No. 10

Detection of Paracoccidioides brasiliensis gp70 Circulating Antigen and Follow-Up of Patients Undergoing Antimycotic Therapy Silvia Helena Marques da Silva,1 Daniela de Mattos Grosso,1 Jose´ Daniel Lopes,1 Arnaldo Lopes Colombo,2 Maria Heloisa Souza Lima Blotta,3 Fla´vio Queiroz-Telles,4 and Zoilo Pires de Camargo1* Department of Microbiology, Immunology and Parasitology1 and Department of Infectious Disease,2 Federal University of Sa ˜o Paulo, Sa ˜o Paulo, and Department of Clinical Pathology, Medical School, State University of Campinas, Campinas,3 Sa ˜o Paulo, and Department of Community Health, Medical School, Federal University of Parana ´, Curitiba, Parana ´,4 Brazil Received 22 March 2004/Returned for modification 5 April 2004/Accepted 18 June 2004

Paracoccidioidomycosis (PCM), one of the most important systemic mycoses in Central and South America, is caused by the dimorphic fungus Paracoccidioides brasiliensis and has a high prevalence in Brazil. Glycoproteins of 43 and 70 kDa are the main antigenic compounds of P. brasiliensis and are recognized by Western blotting by 100 and 96% of PCM patient sera, respectively. In the present study, an inhibition enzyme-linked immunosorbent assay (ELISA) was used to detect gp70 in different biological samples from patients with PCM. gp70 was detected in 98.76% of 81 serum samples, with an average concentration of 8.19 ␮g/ml. The test was positive for 100% of the patients with the acute and chronic unifocal forms of PCM and 98.43% of the patients with the multifocal chronic form, with average concentrations of 11.86, 4.83, and 7.87 ␮g/ml, respectively. Bronchoalveolar lavage fluid from 23 patients with pulmonary unifocal PCM and 14 samples of cerebrospinal fluid from patients with neurological PCM were also tested for gp70 detection, with the test showing 100% sensitivity and 100% specificity, with mean gp70 concentrations of 7.5 and 6.78 ␮g/ml, respectively. To investigate the potential of gp70 detection by inhibition ELISA for the follow-up of PCM patients during antimycotic therapy with itraconazole (ITZ), the sera of 23 patients presenting with the chronic multifocal form of PCM were monitored at regular intervals of 1 month for 12 months. The results showed a decrease in circulating gp70 levels during treatment which paralleled the reduction in anti-P. brasiliensis antibody levels. The detection of P. brasiliensis gp70 from the biological fluids of patients suspected of having PCM proved to be a promising method for diagnosing infection and evaluating the efficacy of ITZ treatment. passive hemagglutination inhibition (24), inverted linear immunoelectrophoresis (23), and immunoblotting (28); but all of these assays presented low sensitivities. Go ´mez et al. (19) were the first to use monoclonal antibodies (MAbs) to detect the 87-kDa circulating antigen in PCM patients by inhibition ELISA (inh-ELISA), which had 80.4% sensitivity. More recently, the same technique was used to detect gp43 in serum, cerebrospinal fluid (CSF), and bronchoalveolar lavage (BAL) fluid of PCM patients with 95.1% sensitivity and 100% specificity (25) and to monitor PCM patients during azole treatment (26). The gp43 glycoprotein from P. brasiliensis is the main antigen used for the diagnosis of PCM (32, 33, 40), being recognized by virtually all sera from infected patients by different serological methods (25, 27). Another important glycoprotein expressed by the fungus is gp70, recognized by 96% of PCM patient sera (3). gp70 is predominantly composed of polysaccharides (27) and is able to induce proliferation of lymphocytes from PCM patients (2). gp70 is also detected in the urine of patients with the acute form of PCM (36). The treatment of mice by injection of anti-gp70 MAb abolishes lung infection, suggesting that gp70 may facilitate fungal installation and the progression of lesions during the primary infection (27). Detection of circulating antigens in patients with PCM has proved to be useful not only for diagnostic purposes but also as a tool to evaluate clearance of the fungal burden during treatment. Go ´mez et al. (19) reported on the use of the inh-ELISA

Paracoccidioidomycosis (PCM), the most prevalent systemic fungal mycosis in Latin America, is caused by Paracoccidioides brasiliensis , a thermally dimorphic fungus (5, 34). Different clinical forms of the disease result from the interaction of different P. brasiliensis strains with the host, a phenomenon that may be related to the virulence of the infecting strain. Some investigators (21, 37, 38, 41) believe that the mechanism of virulence is determined by chemical components present on the fungal wall, but others contest this explanation (44). Diagnosis of PCM is usually based on the demonstration of multibudding yeast cells in different biological specimens or the detection of specific antibodies by histopathology or culture. During the past decade, substantial progress has been made in the development of innovative approaches and methods for the serological diagnosis of PCM. Clinically relevant antigens have been identified and adapted for use in immunoassays for the detection of specific antibodies (12). Some investigators have tried to detect circulating antigen in PCM patients using polyvalent antigens or antibodies in different assays such as competition enzyme-linked immunosorbent assay (ELISA) (16), immunoradiometric assay (13), immunoelectrophoresisimmunodiffusion (17), counterimmunoelectrophoresis (35), * Corresponding author. Mailing address: Universidade Federal de Sa˜o Paulo (UNIFESP), Disciplina de Biologia Celular, 04023-062, Rua Botucatu, 862, 8° andar, Sa˜o Paulo, SP, Brazil. Phone: 55 11 5576 45 23. Fax: 55 11 5571 58 77. E-mail: [email protected]. 4480

DETECTION OF P. BRASILIENSIS gp70 CIRCULATING ANTIGEN

VOL. 42, 2004

to detect gp87 circulating antigen in sera from patients with active disease and in patients receiving antimycotic therapy (20). A previous study that used the same methodology was able to detect gp43 in 96.29% of PCM patients, mainly in those with the acute form of the disease (100%) (25). Moreover, patients cleared this antigen during itraconazole (ITZ) therapy, suggesting that this molecule may be used to assess the host response to antifungal therapy. When it is considered that, in addition to gp43, gp70 is an important molecule in the P. brasiliensis system, in the present study we investigated the application of gp70 antigen detection in different biological specimens to obtain a diagnosis of PCM as well as to assess antigen clearance during antifungal treatment for disseminated disease. MATERIALS AND METHODS Clinical samples for PCM diagnosis. Blood specimens from a total of 81 patients with active PCM were taken at the time of the primary diagnosis. The diagnosis of PCM was established for all patients by direct KOH examination, isolation or the fungus by culture, and/or conventional positive serological tests. Samples were collected from patients at different times between March 1990 and December 2001. The samples were from patients seen at Hospital Sa˜o Paulo, Federal University of Sa˜o Paulo, Sa˜o Paulo, Brazil; Hospital das Clínicas, Federal University of Parana´, Curitiba, Brazil; and Hospital das Clínicas, University of Campinas, Campinas, Brazil. The 81 patients tested (73 males and 8 females) were classified by their clinical presentations, as follows: 11 had the acute form of PCM (mean age, 18.8 years) and 70 had the chronic form (mean age, 46.7 years), including 64 patients with multifocal disease and 6 with unifocal disease. Biological material from 14 patients with neurological PCM (neuro-PCM) were also tested, including 14 CSF samples and 11 serum samples, all of which were obtained before treatment. In addition, 13 BAL fluid samples from patients with pulmonary PCM were tested. Serum samples from patients (n ⫽ 71) with other mycologically and/or serologically confirmed mycoses (51 serum samples from individuals with active histoplasmosis and 20 individuals with cryptococcosis), obtained at the time of diagnosis, were also evaluated. Ninety-three serum samples from healthy volunteers (blood donors), 6 CSF samples from patients with no fungal disease, and 10 BAL fluid samples from individuals with tuberculosis were included as negative controls. Clinical samples for monitoring of therapy. In order to evaluate the decrease in the antigen concentration in response to antifungal therapy, a total of 23 patients with active PCM were evaluated. Sequential serum samples were obtained from all selected patients, including one sample collected at the time of diagnosis (baseline) and at least seven others collected during 8 to 12 months of treatment. Samples were collected from patients attending the Infectious Diseases Division, Hospital das Clínicas, Federal University of Parana´, between 2001 and 2002. The patients were males (mean age, 47.3 years) with the chronic multifocal form of the disease who were successfully treated with ITZ. The good clinical response to antifungal therapy exhibited by all selected patients was characterized by the cessation of all clinical signs and symptoms related to the infection, significant improvement in the chest X-ray pattern of pulmonary lesions, the absence of fungi in biological specimens, and negative serology or the presence of low titers of specific antibodies (as determined by the immunodiffusion [ID] test). Fungal isolates, antigen preparations, and gp70 purification. P. brasiliensis 113 (which was originally from the Faculdade de Medicina da Universidade de Sa˜o Paulo and which was a gift of C. S. Lacaz) was transformed to the yeast phase, and exoantigen was produced as described by Camargo et al. (7). P. braziliensis 113 was selected, as previous studies have shown that it produces large amounts of gp70. The yeast cells were harvested from this growth, washed three times with lysing buffer (0.05 mM Tris-HCl, 5 mM EDTA), and resuspended in 50 ml of the same buffer. Yeast cells were broken with glass beads (diameter, 425 to 600 ␮m; Sigma) in a vortex mixer. A mixture of protease inhibitors containing 1 mM phenylmethylsulfonyl fluoride (Sigma) and 1 mM EDTA was added four times during the process. The homogenate was then centrifuged at 13,000 ⫻ g for 30 min at 4°C. The supernatant (which contained the cytoplasmic antigen) was recovered and used for gp70 purification. The cytoplasmic yeast antigen was fractionated by gel filtration chromatography with Sephadex G-50 resin (Pharmacia). The eluted fractions containing the gp70

4481

molecule were concentrated and further purified by gel filtration on a HiTrap S-200 column (Pharmacia) to eliminate high-molecular-mass contaminants. The protein concentration was determined by the method of Bradford (4). To verify the real purification of gp70, the material was monitored by sodium dodecyl sulfate (SDS)-polyacrylamide gel electrophoresis (PAGE), silver staining, and immunoblotting assays. MAb production and immunization protocols. In the first schedule of immunization, 6-week-old BALB/c mice were injected intraperitoneally with a macerated polyacrylamide gel containing gp70 (gp70 was cut from the SDS-polyacrylamide gel) at 2-week intervals for 4 months. In a second schedule, mice were immunized subcutaneously with 50 ␮g of gp70 (purified through MAbs obtained in the first fusion) in phosphate-buffered saline (PBS) every 2 weeks. gp70 was incorporated in Freund’s complete adjuvant in the first injection and in incomplete Freund’s adjuvant in the subsequent ones. Injections were always made at four different sites in the axillary and inguinal regions with a final volume of 100 ␮l per site. Before each immunization, the mice were bled through the ocular plexus, and the serum was separated by centrifugation and stored at ⫺20°C. The final immunization was made intravenously 2 days before cell fusion. Fusion protocols. Cells of murine myeloma line sp2 were fused with spleen cells from immunized mice by the method of Lopes and Alves (22). For the first fusion, hybridomas were distributed on 48-well plates (Costar), and at 10 days postfusion the colonies were screened by immunoblotting, as described below. For the second fusion, hybridomas were distributed on 96-well plates (Costar), and screening was performed by an enzyme immunoassay, as described elsewhere (18). After the hybridomas were cloned by limiting dilution and expansion of the positive clones, large amounts of antibodies were obtained from the ascitic fluid induced in BALB/c mice injected with Pristane (Sigma) before inoculation of hybridoma cells. Antibody screening by immunoblotting. Exoantigens were submitted to SDSPAGE with 10% acrylamide under reducing conditions, and proteins were then transferred to nitrocellulose membranes (Sigma). Fragments of nitrocellulose membranes containing gp70 were cut into small pieces and distributed in 48-well plates (Costar). After the free sites were blocked with PBS containing 5% skim milk, 200 ␮l of the supernatants from the hybridoma cultures was added to each well. After 1 h of incubation at room temperature, the wells were washed five times for 5 min each time with PBS containing 0.1% Tween 20 (Sigma) (PBS-T) and treated with affinity-purified peroxidase-conjugated goat anti-mouse immunoglobulin (Bio-Rad) for 1 h at room temperature. The wells were again washed with PBS-T five times. The presence of reactive immunoglobulin G (IgG) was tested for by the addition of 4-chloro-naphthol in 50 mM Tris buffer (pH 6.8)– methanol–H2O2, and the IgG was finally washed with distilled water. Pretreatment of immune sera for use in inh-ELISA. Aliquots of immune serum, CSF, or BAL fluid (200 ␮l) were mixed with an equal volume of 0.1 M EDTA (pH 7.2; Sigma) and boiled at 100°C for 3 to 5 min. The tubes were cooled and centrifuged at 13,000 ⫻ g for 30 min, and the supernatant was used for the test. Inh-ELISA. The inh-ELISA was performed as described previously (19, 25). A standard inhibition curve was first prepared by adding known concentrations of gp70 to pooled normal human serum (NHS), CSF, or BAL fluid (a standard curve was prepared for each biological material) (25). The inhibition reaction occurred when constant aliquots of anti-gp70 MAb were mixed with the inhibition standards, PCM patient serum, CSF and BAL fluid samples, and NHS control samples. These samples were then plated on previously blocked microtiter plates (inhibition plates) and incubated overnight at 4°C. The reaction plates were coated with gp70, incubated overnight at 4°C, and blocked; and samples from each well in the inhibition plates (containing a mixture of MAb bound to circulating antigen and free MAb) were transferred to the respective wells in the reaction plates. The plates were washed and probed with goat anti-mouse IgG-peroxidase conjugate and developed with a chromogenic substrate, as described previously (19, 25). The optical density (OD) readings at 492 nm were then plotted on a standard curve constructed from the data derived from MAb titration with the inhibition standards. The antigen concentrations in serum, BAL fluid, and CSF samples were calculated by using a regression model constructed with the reciprocal values of the fixed concentrations of gp70 and the OD values. All of the standards, samples, and controls were tested in duplicate. The cutoff point was established as the receiver operator characteristic (ROC) curve. Statistical analysis. Data were analyzed statistically with Stata software (version 7.0 for Windows 98/95/NT, 2001; Stata Corporation, College Station, Tex.), and specificity and sensitivity were determined by use of the ROC curve. The inhibition standard curves were constructed in duplicate for at least four independent assays. A regression model was constructed with the reciprocal values of

4482

MARQUES

DA

SILVA ET AL.

FIG. 1. Detection of circulating antigen in sera from patients with PCM or other mycoses and in NHS by inh-ELISA. Groups studied: 1, patients with PCM (n ⫽ 81); 2, patients with cryptococcosis (n ⫽ 20); 3, patients with histoplasmosis (n ⫽ 51); 4, NHS (n ⫽ 93). Bars represent the mean antigen concentration for each group. The long fine line represents the cutoff point (1.0 ␮g/ml).

the antigen concentrations and the OD values obtained. Comparisons were made by one-way analysis of variance.

RESULTS MAb production. The anti-gp70 MAb belongs to the IgG1 subclass and recognizes an antigenic determinant of P. brasiliensis with a relative molecular mass of 70 kDa. This MAb did not recognize antigens from Histoplasma capsulatum or Cryptococcus neoformans. Detection of P. brasiliensis gp70 in serum samples. The standard curve was obtained by adding increasing concentrations of gp70 to a pool of control NHS samples. The sensitivity ranged from 0.001 to 30 ␮g/ml. The cutoff point was established from the ROC curve and was based on the antigen concentration in the sera of patients with PCM and NHS. Samples with antigen concentrations greater than 1.0 ␮g/ml were considered positive. gp70 was detected in 80 of 81 patients (98.76%) with active PCM (mean concentration, 8.19 ␮g/ml) (Fig. 1). Among this group, 11 patients presented with the acute form, 6 presented with the chronic unifocal form, and 63 presented with the chronic multifocal form; and the mean antigen concentrations in these three groups were 11.86, 4.83, and 7.87 ␮g/ml, respectively (Fig. 2). Positive results were observed for all patients with the acute or chronic unifocal form and 98.43% of the patients with the chronic multifocal form (Table 1). No crossreactivity with sera from patients with histoplasmosis or cryptococcosis or NHS was observed. Overall, the sensitivity and specificity were 98.8 and 100%, respectively. Detection of P. brasiliensis gp70 in CSF and sera from patients with neuro-PCM. A standard inhibition curve was constructed with known concentrations of gp70 diluted in CSF from patients with no fungal diseases. The cutoff value was determined to be 0.21 ␮g/ml for CSF. gp70 was detected in all CSF samples, and the mean antigen concentration was 6.78 ␮g/ml, showing 100% sensitivity and 100% specificity. Ten of 11 serum samples from patients with neuro-PCM had detectable antigen (mean concentration, 3.97 ␮g/ml; cutoff value,

J. CLIN. MICROBIOL.

FIG. 2. Detection of circulating antigen in sera from patients with different forms of PCM by inh-ELISA. Groups studied: 1, patients with acute form of PCM (n ⫽ 11); 2, patients with chronic unifocal form of PCM (n ⫽ 6); 3, patients with chronic multifocal form of PCM (n ⫽ 64). Bars represent the mean antigen concentration for each group. The long fine line represents the cutoff point (1.0 ␮g/ml).

0.97 ␮g/ml) (Table 2 and Fig. 3). No false-positive reactions with control samples were observed. Detection of P. brasiliensis gp70 in BAL fluid. A standard inhibition curve was constructed with known quantities of gp70 diluted in BAL fluid samples from patients with no fungal diseases. The cutoff level for positivity was established to be 0.46 ␮g/ml. All BAL fluid samples were positive for gp70, and the mean antigen concentration for this group was 7.5 ␮g/ml. BAL fluid samples from individuals with no known infectious diseases were negative for antigen detection (Table 2; Fig. 3). gp70 antigen clearance during treatment with ITZ. Table 3 shows the initial clinical manifestations of all 23 patients with PCM as well as the gp70 antigen concentrations obtained for serum samples collected at the baseline and at the end of treatment. All 23 serum samples from patients with PCM had levels of circulating gp70 antigen above the cutoff point at the time of diagnosis, with a mean antigen concentration of 8.23 ␮g/ml. Antigen levels in serum samples collected at the first visit after the start of therapy (30 days after the start of antifungal therapy) dropped to a mean concentration of 4.0 ␮g/ml and further decreased by the 12th month, when the mean antigen concentration was 0.30 ␮g/ml. Of interest, all but four TABLE 1. Detection of P. brasiliensis 70-kDa circulating antigen in sera from patients with PCM and other fungal infections and sera from healthy controls by inh-ELISA No. of serum samples Serum group

PCM Acute Chronic unifocal Chronic multifocal Histoplasmosis Cryptococcosis NHS

Total

Reactivea

Nonreactive

11 6

11 6

0 0

64

63

1

51 20 93

0 0 0

51 20 93

% Reactive samples

100 100 98.43 0 0 0

Mean antigen concn (␮g/ml)

11.86 4.83 7.87 0 0 0

a A sample was considered positive (i.e., reactive) if the antigen concentration was ⬎1.0 ␮g/ml.

DETECTION OF P. BRASILIENSIS gp70 CIRCULATING ANTIGEN

VOL. 42, 2004

4483

TABLE 2. Detection of P. brasiliensis gp70 antigen by inh-ELISA in various specimens from patients with PCM and their respective controls Specimen group

BAL fluid (from patients with PCM)a CSF (from patients with PCM)b BAL fluid (control) CSF (control) Serum from patients with neuro-PCMc a b c

No. of serum samples Total

Reactive

Nonreactive

23 14 10 6 11

23 14 0 0 10

0 0 10 6 1

% Reactive samples

Mean antigen concn (␮g/ml)

100 100 0 0 90.9

7.5 6.78 0 0 3.97

A sample was considered positive (i.e., reactive) if the antigen concentration was ⬎0.46 ␮g/ml. A sample was considered positive (i.e., reactive) if the antigen concentration was ⬎0.21 ␮g/ml. A sample was considered positive (i.e., reactive) if the antigen concentration was ⬎0.97 ␮g/ml.

serum samples collected at the end of antifungal treatment showed complete antigen clearance and exhibited negative results (Fig. 4). Overall, the patients showed a significant reduction (P ⬍ 0.0001) in the levels of circulating gp70 by the third month after the initiation of therapy, and this reduction was maintained until the end of the follow-up period (Fig. 5). All patients improved clinically and mycologically with therapy. The antibody titers determined by the ID test at the time of diagnosis varied from 1:2 to 1:16, and in almost all patients the titers decreased during the course of treatment. DISCUSSION Since the demonstration of the antigenemia produced in animals experimentally infected with Streptococcus pneumoniae (9), there have been numerous reports on the detection of antigens in blood, urine, and CSF of patients and animals infected with bacteria, fungi, viruses, and parasites. In the case of fungi, with the exception of the capsular polysaccharide of C. neoformans, which has been detected since 1950 (10), advances have required sensitive immunoassay procedures that have become widely applied only recently. In immunocompro-

FIG. 3. Detection of gp70 antigen in CSF from patients with neuroPCM and BAL fluid from patients with pulmonary PCM by inhELISA. Groups studied: 1, CSF of patients with PCM (n ⫽ 14); 2, CSF of patients with no fungal disease (n ⫽ 6); 3, BAL fluid from patients with PCM (n ⫽ 23); 4, BAL fluid from patients with no fungal pulmonary disease (n ⫽ 10); 5, serum from patients with neuro-PCM (n ⫽ 11). Bars represent the average antigen concentration for each group.

mised patients with cryptococcosis, histoplasmosis, or aspergillosis, the detection of circulating antigens can provide a rapid diagnosis and prompt treatment in the first days of infection, when antibodies are not yet produced or are present at low levels (11, 12). Detection of circulating H. capsulatum antigen has proved to be useful for the diagnosis and follow-up of patients with histoplasmosis (42, 43). Very few techniques have shown high sensitivities and specificities for antigen detection in biological specimens for the diagnosis and follow-up of PCM patients. In our previous studies, we successfully detected gp43 in sera, CSF, and BAL fluid samples from PCM patients by inh-ELISA, which proved to be suitable for the monitoring of patients receiving antimycotic therapy. Go ´mez et al. (19, 20) used the inh-ELISA to detect the 87-kDa molecule in the sera of patients with PCM for both diagnostic and follow-up purposes. Antibody detection systems have also been used to assay other molecules for the diagnosis of PCM, such as a 22- to 25-kDa protein (14), a 58-kDa glycoprotein (15), and an 87-kDa protein (8). In the present study, the overall sensitivity of gp70 detection among the 81 PCM patients was 98.8% in the presence of a mean antigen concentration of 8.19 ␮g/ml and reached 100% in patients with the acute form of the disease, with a mean antigen concentration of 11.86 ␮g/ml. Among patients with the chronic form of PCM, antigenemia due to gp70 was observed in 98.43% of serum samples from patients with the multifocal form (mean antigen level, 7.87 ␮g/ml); on the other hand, the sera of 100% of the patients with the chronic unifocal form had gp70, with a mean antigen concentration of 4.83 ␮g/ml. gp70 was detected in 100% of the CSF and BAL fluid samples, with mean antigen concentrations of 6.78 and 7.5 ␮g/ml, respectively. No cross-reactions were observed when sera from patients with other mycoses were tested. The majority (90.9%) of serum samples from patients with neuro-PCM were positive for gp70, with a mean antigen concentration of 3.97 ␮g/ml. The inh-ELISA proved to be helpful in detecting P. brasiliensis gp70 in biological samples, including BAL fluid and CSF, with 100% sensitivity. Considering the limitations of the conventional methods for the diagnosis of neuro-PCM, our findings suggest that the gp70 antigen test may be useful for confirmation of the diagnosis in such patients. In addition, the inhELISA has the advantage of being able to process large numbers of samples at the same time and presents a high sensitivity and a high specificity. For diagnostic purposes, tests based on the ID test are generally considered to have high specificities; however, false-

4484

MARQUES

DA

SILVA ET AL.

J. CLIN. MICROBIOL.

TABLE 3. Clinical manifestations and serology results for 23 patients evaluated before and after antifungal treatment Patient no. and lesion sitea

1. L, M, LN 2. L, M, LN 3. L, M, LN 4. L, ORP, LN 5. L, S, M, LN 6. L, M, LN 7. L, LN, Lx 8. L, M, LN 9. L, M, LN 10. L, M 11. L, M, LN 12. L, M, LN 13. L, M 14. L, M 15. L, M 16. L, M, LN, Lx 17. L, M 18. L, M 19. L, Lx 20. L, M 21. L, SNC 22. L, M 23. L, M, LN Mean a b c

No. of samples tested

8 7 7 9 7 7 6 7 7 7 7 6 6 5 6 6 8 7 7 8 6 8 7 6.91

Antibody titer atb:

Antigen level (␮g/ml) at: Diagnosis

First follow-up

End of treatment

6.75 9.75 5.64 14.25 13.5 9.0 11.25 10.5 4.16 7.5 9.75 4.53 7.12 6.94 3.57 13.5 4.94 3.57 8.25 10.5 12.75 4.16 7.5

5.01 15.0 3.24 9.75 3.39 3.59 1.77 3.8 7.26 2.67 2.74 2.37 3.20 3.69 1.76 2.86 2.76 1.57 4.18 4.09 3.69 1.60 2.13

0 (11) 0 (9) 0 (9) 2.16 (12) 0 (11) 0 (9) 0 (8) 1.69 (9) 0 (11) 0 (9) 0 (10) 0 (9) 1.03 (9) 0 (9) 0 (8) 2.24 (9) 0 (10) 0 (10) 0 (10) 0 (10) 0 (12) 0 (12) 0 (12)

8.23

4.0

c

Diagnosis

End of treatment

1:8 1:4 1:2 1:16 1:16 1:16 1:4 1:16 NR 1:16 1:8 1:2 1:4 1:2 1:4 1:16 1:2 1:4 1:16 1:16 1:16 1:4 1:4

NDS NR NR 1:4 1:2 1:4 NDS 1:4 NR 1:4 1:2 NR 1:2 NDS NDS 1:4 NDS NDS 1:2 1:16 1:2 1:2 NDS

0.30

L, lung; M, mucosae; LN, lymph nodes; Lx, Larynx; ORP, oropharynx; S, skin; CNS, central nervous system. NR, nonreactor; NDS, reactive only with nondiluted serum. Values in parentheses are the length of follow-up (in months).

negative results may also occur, and this may be related to the production of low-avidity IgG2 antibodies directed against carbohydrate epitopes (31). The sensitivities of ID test-based tests range from 65 to 100% (1, 6, 29), depending on the antigen preparation used. On the other hand, tests with higher sensitivities, such as immunoenzymatic assays, present problems associated with specificity due to cross-reactivity with heterologous sera (5, 19). For many years, assessment of the clearance of the fungal burden during treatment for PCM has been based on the

FIG. 4. Mean circulating antigen concentration in sera from patients with PCM at the time of diagnosis and during the period of ITZ treatment. Error bars, standard deviation; m, months; n, number of patients.

detection of antibodies against P. brasiliensis crude or purified antigens (5, 29, 30, 39). Although healing of apparent fungal lesions may occur within a short time after antifungal treatment is initiated, long therapeutic courses are desirable in order to prevent relapses. In this case, the status of the fungal infection and the host response are usually monitored by indirect methods, in which serology provides information about the prognosis (5). Our experience with the follow-up of patients receiving treatment for PCM has shown that many times the antibody titers obtained by the ID test do not correlate with the clinical status of the patient. For example, in some patients elevated antibody titers (1:64) were observed until the end of treatment, when the patients were clinically cured. On the other hand, although for most patients low antibody titers are related to the absence of clinical symptoms (1:2 or 1:4), in some cases low titers are present for months and are present in concert with clinical symptoms. Such discrepancies between the clinical status of the host and specific antibody detection are probably related to the fact that the cell response and not the humoral response is the main immunologic mechanism able to contain P. brasiliensis in the infected organism. For these reasons, the use of serology as a single criterion of cure in patients with PCM is controversial. In clinical practice, clinical, radiological, mycological, and serological aspects must be evaluated over a long period of observation in order to assess the results of treatment. Regarding the detection of antigen during treatment, Mendes-Giannini et al. (28), working with a pool of sera from patients with PCM using a Western blot assay, observed that

VOL. 42, 2004

DETECTION OF P. BRASILIENSIS gp70 CIRCULATING ANTIGEN

4485

FIG. 5. Correlation between the levels of antigenemia due to gp70 and anti-P. brasiliensis antibody titers at the time of diagnosis (A) and at the end of treatment (B). NR, nonreactor; NDS, reactive only with not diluted sera.

gp43 started to disappear from the circulation after 10 months of treatment and was undetectable after 2 years of treatment. More recently, the same group of investigators (36) detected P. brasiliensis antigens in 75% of patient urine samples tested by an indirect competition enzyme immunoassay and by an immunoblot test for monitoring the response to therapy. One of the main goals of the present study was to assess gp70 antigen clearance during treatment of PCM with ITZ by comparing the clearance from serum samples collected at the baseline and after 8 to 12 months of therapy. All 23 patients selected for the study showed a good clinical response to therapy, with no relapse of fungal disease documented over the follow-up period. Overall, decreasing levels of P. brasiliensis gp70 were detected in patients during successful antifungal therapy. In contrast, the ID test titers of these patients varied over the follow-up period (1:16 to 1:2). Table 3 shows the clinical manifestations of PCM in the patients and the correlation between the levels of gp70 antigenemia and the levels of anti-P. brasiliensis antibody at the time of diagnosis, 1 month after treatment, and the end of treatment. At the time of diagnosis, all except one of the patients had detectable gp70 antigen and anti-P. brasiliensis antibodies (one patient had a negative ID test result). A significant decrease in gp70 titers occurred during the first month of therapy in 22 patients; at the end of therapy, only 4 (17.39%) patients (patients 4, 8, 13, and 16) had detectable gp70 antigen (antigen levels, ⬍2.5 ␮g/ml) and positive titers by the ID test. By the end of the follow-up period, antibody titers were negative in three patients (patients 2, 3, and 12), and one patient (patient 9) had undetectable antibodies (by the ID test) from the time of diagnosis to the end of treatment. However, the antigen level was also negative until the end of follow-up. Once the levels of gp70 decreased, gp70 persisted at low levels in serum until the end of therapy. On the other hand, in various patients, the antibody titers seemed to decrease a little later, after about 6 months of therapy. Of particular note is the common persistence of antibodies for a considerable period of time even after the cessation of an apparently successful course of treatment.

Decreased levels of antigenemia were observed in two patients after 6 months of ITZ treatment, but their anti-P. brasiliensis antibody titers remained high (1:16) and did not change throughout the period of treatment (patient 20). On the other hand, patient 9 had undetectable antibodies from the time of diagnosis to month 11 of treatment but had a substantial decrease in antigenemia levels during treatment with ITZ. All PCM patients showed clinical improvement after the period of antifungal therapy (8 to 12 months). In conclusion, our results indicate that the detection and quantification of P. brasiliensis gp70 antigen in serum by inhELISA is a sensitive method for the diagnosis of PCM and may be used to evaluate patients for the clearance of the fungal burden during treatment. Moreover, the high sensitivity of this assay for the detection of gp70 in different biological fluids opens a new possibility for the diagnosis of PCM in patients for whom conventional methods present limitations. ACKNOWLEDGMENTS This work was supported by Fundac¸˜ao de Amparo `a Pesquisa do Estado de Sa˜o Paulo (FAPESP). We thank Elettra Grene for correction of the English. REFERENCES 1. Alves, J. R. 1996. Comparac¸˜ao entre treˆs me´todos sorolo ´gicos no seguimento de pacientes com paracoccidioidomicose. Master’s thesis. Universidade Federal de Sa˜o Paulo, Sa˜o Paulo, Brazil. 2. Benard, G., M. J. S. Mendes-Giannini, M. Juvenale, E. T. Miranda, and A. J. S. Duarte. 1997. Immunosuppression in paracoccidioidomycosis: T cell hyporesponsiveness to two Paracoccidioides brasiliensis glycoproteins that elicit strong humoral immune response. J. Infect. Dis. 175:1263–1267. 3. Blotta, M. H. S. L., and Z. P. Camargo. 1993. Immunological response to cell-free antigens of Paracoccidioides brasilienis: relationship with clinical forms of paracoccidioidomycosis. J. Clin. Microbiol. 31:671–676. 4. Bradford, M. 1976. A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal. Biochem. 72:248–254. 5. Brummer, E., E. Castan ˜ eda, and A. Restrepo. 1993. Paracoccidioidomycosis: an update. Clin. Microbiol. Rev. 6:89–117. 6. Camargo, Z. P., C. S. Unterkircher, and L. R. Travassos. 1989. Identification of antigenic polypeptides of Paracoccidioides brasiliensis in serological tests. J. Med. Vet. Mycol. 27:407–412. 7. Camargo, Z. P., R. Berzaghi, C. C. Amaral, and S. H. Marques da Silva. 2003. Simplified method for producing Paracoccidioides brasiliensis exoantigens for use in immunodiffusion tests. Med. Mycol. 41:539–542.

4486

MARQUES

DA

SILVA ET AL.

8. Díez, S., B. L. Go ´mez, A. Restrepo, R. J. Hay, and A. Hamilton. 2002. Paracoccidioides brasiliensis 87 kilodalton antigen, a heat shock protein useful in diagnosis: characterization, purification and detection via immunohistochemistry. J. Clin. Microbiol. 40:359–365. 9. Dochez, A. R., and O. T. Avery. 1917. The elaboration of specific soluble substances by Pneumococcus during growth. J. Exp. Med. 26:477–479. 10. Drouhet, E. 1989. Overview of fungal antigens, p. 3–38. In E. Drouhet, G. T. Cole, L. Repentigny, J. P. Latge´, and B. Dupont (ed.), Fungal antigen: isolation, purification, and detection. Plenum Publishing Corporation, New York, N.Y. 11. Drutz, D. J. 1986. Antigen detection in fungal infections. N. Engl. J. Med. 314:115–117. 12. Ferreira-da-Cruz, M. F., B. Galva ˜o-Castro, and C. T. Daniel-Ribeiro. 1992. Isolation and antigenicity of a 45-kilodalton Paracoccidioides brasiliensis immunodominant antigen. Infect. Immun. 60:2667–2671. 13. Ferreira-da-Cruz, M. F., B. Galva ˜o-Castro, and C. T. Daniel-Ribeiro. 1991. Sensitive immunoradiometric assay for the detection of Paracoccidioides brasiliensis antigens in human sera. J. Clin. Microbiol. 29:1202–1205. 14. Figueroa, J., A. J. Hamilton, M. H. Allen, and J. Hay. 1994. Immunohistochemical detection of a novel 22- to 25-kilodalton glycoprotein of Paracoccidioides brasiliensis in biopsy material and partial characterization by using species-specific monoclonal antibodies. J. Clin. Microbiol. 32:1566–1574. 15. Figueroa, J., A. J. Hamilton, M. H. Allen, and J. Hay. 1995. Isolation and partial characterization of a Paracoccidioides brasiliensis 58 kDa extracellular glycoprotein which is recognized by human immune sera. Trans. R. Soc. Trop. Med. Hyg. 89:566–572. 16. Freitas-da-Silva, G., and M. C. Roque-Barreira. 1992. Antigenemia in paracoccidioidomycosis. J. Clin. Microbiol. 30:381–385. 17. Garcia, N. M., G. B. Del Negro, H. P. Martins, and C. S. Lacaz. 1987. Detection of Paracoccidioides brasiliensis circulating antigens by immunoelectrophoresis-immunodiffusion technique. Preliminary report. Rev. Med. Trop. Sa˜o Paulo 29:327–328. 18. Gesztesi, J.-L., R. Puccia, L. R. Travassos, A. P. Vicentini, J. Z. Moraes, M. F. Franco, and J. D. Lopes. 1996. Monoclonal antibodies against the 43,000 kDa glycoprotein from Paracoccidioides brasiliensis modulate laminin-mediated fungal adhesion to epithelial cells and pathogenesis. Hybridoma 15:415–422. 19. Go ´mez, B. L., J. L. Figueroa, A. J. Hamilton, B. Ortiz, M. A. Robledo, R. J. Hay, and A. Restrepo. 1997. Use of monoclonal antibodies in diagnosis of paracoccidioidomycosis: new strategies for detection of circulating antigens. J. Clin. Microbiol. 35:3278–3283. 20. Go ´mez, B. L., J. I. Figueroa, A. J. Hamilton, S. Diez, M. Rojas, A. M. Tobo ´n, R. J. Hay, and A. Restrepo. 1998. Antigenemia in patients with paracoccidioidomycosis: detection of the 87-kilodalton determinant during and after antifungal therapy. J. Clin. Microbiol. 36:3309–3316. 21. Kurokawa, C. S., M. F. Sugisaki, and M. T. S. Perac¸oli. 1998. Virulence factors in fungi of systemic mycosis. Rev. Inst. Med Trop. Sa˜o Paulo 40:125– 135. 22. Lopes, J. D., and, M. G. M. Alves. 1983. Production of monoclonal antibodies by somatic cells hibridization, p. 386–398. In C. M. Morel (ed.), Genes and parasites: a laboratory manual. Fundac¸˜ao Oswaldo Cruz, Rio de Janeiro, Brazil. 23. Magaldi, S. W., and D. W. R. Mackenzie. 1986. Deteccio ´n de antigenemia y anticuerpos de Paracoccidioides mediante procedimentos electrofore´ticos invertidos, p. 80. In Proceedings Coloquio Internacional sobre la Paracoccidioidiomicosis. Corporacio ´n para Investigaciones Biolo ´gicas, Medellín, Colombia. 24. Magaldi, S. W., D. W. R. Mackenzie, and M. B. Albornoz. 1989. Detection of Paracoccidioides brasiliensis circulating antigen by the passive hemagglutination inhibition in patients sera, p. 18. In Proceedings Encuentro International sobre Paracoccidioidiomicosis. Instituto Venezolano de Investigaciones Cientificas, Caracas, Venezuela. 25. Marques da Silva, S. H., A. L. Colombo, M. H. S. L. Blotta, L. D. Lopes, F. Queiroz-Telles, and Z. P. de Camargo. 2003. Detection of circulating gp43 antigen in serum, cerebrospinal fluid and bronchoalveolar lavage of patients with paracoccidioidomycosis. J. Clin. Microbiol. 41:3675–3680. 26. Marques da Silva, S. H., F. Queiroz-Telles, A. L. Colombo, M. H. S. L. Botta, J. D. Lopes, and Z. P. Camargo. 2004. Monitoring gp43 antigenemia in

J. CLIN. MICROBIOL.

27. 28.

29.

30. 31.

32.

33. 34. 35.

36.

37. 38. 39. 40. 41.

42. 43.

44.

paracoccidioidomycosis patients during therapy. J. Clin. Microbiol. 42:2419– 2424. Mattos Grosso, D., S. R. de Almeida, M. Mariano, and J. D. Lopes. 2003. Characterization of gp70 and anti-gp70 monoclonal antibodies in Paracoccidioides brasiliensis pathogenesis. Infect. Immun. 71:6534–6542. Mendes-Giannini, M. J. S., J. P. Bueno, M. A. Shikanai-Yasuda, A. W. Ferreira, and A. Masuda. 1989. Detection of the antigen 43,000-molecularweight glycoprotein in sera of patients with paracoccidioidomycosis. J. Clin. Microbiol. 27:2842–2845. Mendes-Giannini, M. J. S., J. P. Bueno, M. A. Shikanai-Yasuda, A. M. S. Stolf, A. Masuda, V. Amato-Neto, and A. W. Ferreira. 1990. Antibody response to the 43kDa glycoprotein of Paracoccidioides brasiliensis as a marker for the evaluation of patients under treatment. Am. J. Trop. Med. Hyg. 43:200–206. Mendes-Giannini, M. J. S., G. B. Del Negro, and A. M. Siqueira. 1994. Serodiagnosis, p. 345–363. In M. Franco, C. S. Lacaz, A. Restrepo, and G. Del Negro (ed.), Paracoccidioidomycosis. CRC Press, Inc., Boca Raton, Fla. Neves, A. R., R. L. Mamoni, C. L. Rossi, Z. P. Camargo, and M. H. S. L. Blotta. 2003. Negative immunodiffusion test results obtained with sera of paracoccidioidomycosis patients may be related to low-avidity immunoglobulin G2 antibodies directed against carbohydrate epitopes. Clin. Diagn. Lab. Immunol. 10:802–807. Puccia, R., and L. R. Travassos. 1991. 43-Kilodalton glycoprotein from Paracoccidioides brasiliensis: immunochemical reactions with sera from patients with paracoccidioidomycosis, histoplasmosis, or Jorge Lobo’s disease. J. Clin. Microbiol. 29:1610–1615. Puccia, R., S. Schenkman, P. A. Gorin, and L. R. Travassos. 1986. Exocellular components of Paracoccidioides brasiliensis: identification of a specific antigen. Infect. Immun. 53:199–206. Restrepo, A. 2000. Paracoccidioides brasiliensis, p. 2768–2772. In G. L. Mandell, J. E. Bennett, and R. Rolin (ed.), Principles and practice of infectious disease, 5th ed. Churchill Livingstone, Philadelphia, Pa. Rodrigues, M. C., C. M. Cassaguerra, and C. S. Lacaz. 1984. Antigenemia in paracoccidioidomycosis. Probable demonstration of circulating antigen by counterimmunoelectrophoresis test. Rev. Inst. Med. Trop. Sa˜o Paulo 26: 285–287. Salina, M. A., M. A. Shikanai-Yasuda, R. P. Mendes, B. Barravieira, and M. J. S. Mendes-Giannini. 1998. Detection of circulating Paracoccidioides brasiliensis antigen in urine of paracoccidioidomycosis patients before and during treatment. J. Clin. Microbiol. 36:1723–1728. San-Blas, G., F. San-Blas, and L. E. Serrano. 1977. Host-parasite relationship in the yeast-like form of Paracoccidioides brasiliensis strain IVIC Pb9. Infect. Immun. 15:343–346. San-Blas, G., and F. San-Blas. 1982. Variability of cell wall composition in Paracoccidioides brasiliensis: a study of two strains. Sabouraudia 20:31–40. Siqueira, A. M. 1982. Diagno ´stico imunolo ´gico, p. 253–264. In G. Del Negro, C. S. Lacaz, and A. M. Forillo (ed.), Paracoccidioidomicose. Sarvier-Editora da Universidade de Sa˜o Paulo, Sa˜o Paulo, Brazil. Stambuk, B. U., R. Puccia, M. L. C. Almeida, L. R. Travassos, and S. Schenkman. 1988. Secretion of the 43 kDa glycoprotein antigen by Paracoccidioides brasiliensis. J. Med. Vet. Mycol. 26:367–373. Svidzinski, T. I. E., M. H. Miranda-Neto, R. G. Santana, O. Fischman, and A. L. Colombo. 1999. Paracoccidioides brasiliensis isolates obtained from patients with acute and chronic disease exhibit morphological differences after animal passage. Rev. Inst. Med. Trop. Sa˜o Paulo 41:279–283. Wheat, L. J., R. B. Kohler, and R. P. Tewari. 1986. Diagnosis of disseminated histoplasmosis by detection of Histoplasma capsulatum antigen in serum and urine specimens. N. Engl. J. Med. 314:83–88. Wheat, L. J., H. Wheat, P. Connoly, M. Kleiman, K. Supparatpinyo, K. Nelson, R. Bradsher, and A. Restrepo. 1997. Cross-antigen in Histoplasma capsulatum variety capsulatum antigen assays of urine samples from patients with endemic mycoses. Clin. Infect. Dis. 24:1169–1171. Zacharias, D., A. Ueda, M. Moscardi-Bacchi, M. Franco, and G. San-Blas. 1986. A comparative histopathological, immunological, and biochemical study of experimental intravenous paracoccidioidomycosis induced in mice by three Paracoccidioides brasiliensis isolates. J. Med. Vet. Mycol. 24:445– 454.