Full text of DOI:10.1080/mmy.39.5.423.428

Medical Mycology 2001, 39, 423±428
Accepted 27 November 2000
Levels of speci®c antigen (gp43), speci®c antibodies, and
antigen±antibody complexes in saliva and serum of
paracoccidioidomycosis patients
Ä
C. S. N. MIURA,* D. ESTEVAO, J. D. LOPES & E. N. ITANO
y
z
y
y
*CoordenacËaÄo de AperfeicËoamento de Pessoal de NõÂvel Superior (CAPES); Department of Pathology Science, CCB, State University
Â
of Londrina, Parana (PR), Brazil; Department of Immunology, Federal University of SaÄo Paulo, SaÄo Paulo (SP), Brazil
z
The present study analyses human immunoglobulin G (IgG) antibodies directed
Paracoccidioides brasiliensis
against the
exoantigen, gp43, as well as the presence of
gp43–IgG immune complexes (ICs) in 31 samples of saliva and serum from
19 patients with paracoccidioidomycosis (PCM) and 12 normal donors. Additional
analysis of secretory IgA (sIgA) was performed on the same saliva samples.
Consistent with previous Žndings, a signiŽcant increased speciŽc IgG level was
P
¢
observed in PCM patients’ saliva and serum (
0 05). The analysis of serum gp43
<
P
¢
and gp43–IgG IC demonstrated a higher level in patients with PCM (
0 05);
<
however, this difference was not statistically signiŽcant with regard to gp43 and
gp43–IgG in saliva when compared to the healthy donors. A high level of sIgA in
saliva of PCM patients compared to that of normal donors was also observed
P
¢
(
0 05). Patients exhibiting low levels of serum IgG but with high titres of IC were
<
observed, thus strengthening the idea of the necessity to use more than one marker
for diagnosis and treatment monitoring of PCM. This is the Žrst report of sIgA in
PCM patients’ saliva and may be indicative of a protective role in neutralizing
antigens on mucosal surfaces.
Paracoccidioides brasiliensis
, paracoccidioi-
Keywords gp43, immune complexes,
domycosis
Introduction
observation of the characteristic multiple-budding yeast
forms in clinical specimens or by culture [4]. More
frequently, diagnosis is inferred from indirect evidence
obtained via serological procedures, mainly immuno-
Paracoccidioidomycosis (PCM) is one of the most
important systemic mycoses in Latin America. It is
especially prevalent in Brazil. The causal agent is
Paracoccidioides brasiliensis
human pathogen [1]. The disease is thought to be
acquired when airborne propagules produced by
P. brasi-
diffusion and immunoenzymatic assays [5,6].
,
a thermally dimorphic
liensis
exoantigens have been widely used to detect
in vitro
et al
speciŽc antibodies
[6]. Puccia
. [7] puriŽed
P. brasiliensis
the speciŽc
antigen gp43, a glycoprotein
P. brasiliensis
saprobic phase are inhaled into the host’s
with a molecular mass of 43 kDa. In all cases of PCM
studied, anti-gp43 antibodies [8], as well as speciŽc IgG,
IgM and IgA antibodies to gp43 have been detected [9].
The classical enzyme immunoassay (ELISA) and capture
enzyme immunoassay (EIA) both detect speciŽc anti-
gp43 antibodies. However, the capture method using
monoclonal antibody against gp43 proved to be a
signiŽcantly more sensitive and speciŽc test [10]. Both
lungs, where they convert to the yeast form [2]. The
infection can cause overt disease involving many body
sites including the lymph nodes, the skin and the oral and
rectal mucosa [3]. PCM can be diagnosed by direct
ˆ
Correspondence: Eiko Nakagawa Itano, Departamento de Ciencias
´
Patologicas, CCB, Universidade Estadual de Londrina. Campus
P. brasi-
techniques were capable of distinguishing anti-
´
Universitario. 86.051-970.Londrina, PR, Brazil. Tel.: ‡ 43 371 4469;
liensis
antibodies from those formed in response to other
fax: ‡ 43 371 4207; e-mail: itano@sercomtel.com.br
ã
2001 ISHAM, Medical Mycology, 39, 423±428

424 Miura et al.
o
o
fungal infections and from immunoglobulins obtained
from healthy controls [10].
Decreases in the antibody response to gp43 during
therapy indicate that gp43 is a good marker for follow-up
of patients with acute or chronic paracoccidioidomycosis
[11].
The production of anti-gp43 antibodies is related to
the severity of the disease [12], as are levels of circulating
antigen [13] and circulating immune complexes (IC) [14].
There is, however, a lack of available data about soluble
antigens, IC and secretory immunoglobulin A (sIgA)
antibodies in PCM patients’ saliva. The present research
was directed toward analysing gp43, gp43–IgG IC and
at 37 C and then overnight at 4 C. The plates were
¢
washed with phosphate-buffered saline containing 0 05%
Tween 20 (PBS-T), blocked with PBS-T 5% skim milk
for 1 h at 37 C, and then incubated for 1 h with sera
(1/400) or saliva (1/10) diluted in PBS-T. The threshold
was previously determined by titration of positive and
negative serum samples. The plates were washed with
o
£
PBS-T (5 ) and incubated with mouse anti-human IgG
labelled with peroxidase (100 l well¡1, 1/1700 dilution)
m
¡1
m
and 100 l well of substrate solution was added (5 mg
ortho-phenylenediamine, 10 ml of 0 1 citrate buffer,
pH 4 5 and 10 l H2O2). The reaction was stopped with
50 l of 4 N H2SO4 per well and the absorbance was
¢
M
¢
m
m
P. brasiliensis
sIgA directed against
of patients with PCM. Additionally, we determined
P. brasiliensis
exoantigens in saliva
read in a Multiskan EX reader (Labsystems, Helsinki,
Finland) at 492 nm. For salivary sIgA determination
after incubation with saliva samples, mouse anti-human
secretory component (1/10 000; Sigma Chemical Co., St.
Louis, Missouri, USA) and anti-mouse IgG labelled with
peroxidase were employed.
levels of anti-
IgG in sera and saliva and of
gp43 and IC in patients’ sera.
Material and methods
This study was approved by the Internal ScientiŽc
Commission and the Bioethics in Research Committee
Monoclonal antibodies anti-gp43
´
of the State University of Londrina, Parana, Brazil.
Anti-gp43 monoclonal antibodies (Mab) [10] (provided
´
by Dr Jose D. Lopes, Department of Immunology,
Serum and saliva samples
˜
Federal University of Sao Paulo, SP, Brazil) were
puriŽed from ascitic uids in a Sepharose-protein G
column (Sigma). The protein concentration was esti-
mated using the Folin phenol method [15].
Serum and saliva samples were obtained from 19
patients with the chronic form of PCM. Of these
patients, ten had unifocal lesions (Žve oropharyngeal
including laryngeal, four pulmonary and one intestinal)
and nine had multifocal lesions. All were receiving
antifungal treatment (3 for less than 1 year, 4 for less
than 2 years, and 12 for more than 2 years) from the
Clinical Hospital of the State University of Londrina.
Sera and saliva from 12 healthy control subjects were
obtained from blood donors at the Hematology Institute
of Londrina. Both groups consisted entirely of persons
between 30 and 59 years of age. Informed consent was
obtained from all subjects. Saliva was collected without
exogenous stimulation, then clariŽed by centrifugation at
gp43 puri®cation
PuriŽed gp43 was obtained by afŽnity chromatography
P. brasiliensis
of the crude exoantigen of
B-339 on
AfŽgel-10 (Bio-Rad Laboratories, Hercules, CA, USA)
with anti-gp43 Mab as a speciŽc ligand. The gp43 was
¢
M
¢
eluted with 0 1 glycine–HCl, pH 2 8, and immediately
M
¢
¢
neutralized with 2
Tris, pH 9 0. Fractions (1 0 ml)
were collected and read in a spectrophotometer at
280 nm. The fractions with the highest absorbance were
¢
M
mixed. The resulting pool was dialysed against 0 15
PBS, and the protein concentration was determined.
g
400 for 1 min. All serum and saliva specimens were
o
¡
divided into aliquots and stored at 20 C.
Preparation of anti-gp43 rabbit IgG
P. brasiliensis exoantigen
P. brasiliensis
m
gp43 (200 g) was subcuta-
The puriŽed
A lyophilized exoantigen was prepared from a yeast-
neously injected into rabbits. The Žrst immunization was
performed with complete Freund’s adjuvant (CFA;
P. brasiliensis
. [6].
phase culture of
et al
strain B-339 according to
Camargo
m
m
800 l CFA and 800 l gp43) and the second and third
with incomplete Freund’s adjuvant (IFA) at successive
intervals of two weeks. After bleeding, the hyperimmune
ELISA for antibodies
exoantigen (1 mg ml¡1 dry weight) was
serum was passed through a Sepharose-protein G
P. brasiliensis
¢
diluted in carbonate bicarbonate buffer (Na2CO3 1 59 g,
column and the protein concentration was determined
using the Folin phenol method with bovine serum
albumin as a standard.
¢
¢
NaHCO3 2 93 g, distilled water qsp 1000 ml, pH 9 6),
then used to coat immunoplates (100 l well¡1) for 1 h
m
ã
2001 ISHAM, Medical Mycology, 39, 423±428

Immune complexes in paracoccidioidomycosis patients 425
ELISA for antigenemia determination
¡1
¢
m
Immunoplates were coated with 1 2 g ml of rabbit
¢
anti-gp43 IgG in carbonate bicarbonate buffer (pH 9 6)
o
o
for 1 h at 37 C and then overnight at 4 C. After
blocking and washing, the plates were incubated with
o
sera or saliva (1/10 in PBS), for 1 h at 37 C and then
with anti-gp43 Mab (300 ng ml¡1) for 1 h at 37 C. Plates
o
were washed and incubated with anti-mouse IgG
labelled with peroxidase (1/18 000 in PBS). Processing
was then done as described in the previous section.
Immune complexes determination
Plates were sensitized with anti-gp43 Mab at a concen-
¡1
m
tration of 20 g ml and then sera or saliva was added,
followed by peroxidase conjugated anti-immunoglobulin
(sIgA or IgG). Processing was then carried out as
described in the previous section.
Statistical analysis
Data were analysed statistically using Statistica for
Windows Release 4, 3D, StatSoft, Inc. Tulsa, USA
Fig. 1 Levels of speciŽc serum IgG, salivary IgG and secretory
IgA (sIgA) antibodies reacting with Paracoccidioides brasiliensis
exoantigens in paracoccidioidomycosis (PCM) patients (*) and
t
1993 software; the test for independent samples was
P
¢
considered signiŽcant if
0 05.
<
*
normal donors ( ), in optical densities (OD). ELISA plates coated
with P. brasiliensis exoantigens were incubated with serum (1/400)
or saliva (1/10) diluted in phosphate-buffered saline containing
0¢05% Tween 20 (PBS-T). Plates were washed, incubated with
peroxidase-conjugatedanti-human IgG (1/18 000 in PBS-T) for IgG
study or with mouse anti-secretory component of IgA (1/10 000)
and then with peroxidase-conjugated anti-mouse IgG labelled with
peroxidase for sIgA. Orthophenylenediamine solution was added,
and the reactions were read at 492 nm. Results are expressed in
optical densities (OD) at 492 nm and horizontal bars represent
mean values. (1) PCM patients’ serum IgG; (2) normal human
serum IgG; (3) PCM patients’ salivary IgG; (4) normal salivary IgG;
(5) PCM patients’ salivary sIgA; (6) normal salivary sIgA. (1, 2)
Results
ELISA analysis of serum and salivary IgG reacting with
P. brasiliensis exoantigens
The results of ELISA expressed as optical densities
(OD) were higher in PCM patients’ serum
¢
§ ¢
¢
§ ¢
(0 192 0 088) than in normal serum (0 092 0 036),
P
¢
0 05. ODs were also higher in PCM patients’ saliva
<
¢
§ ¢
¢
§ ¢
(0 165 0 086) than in normal saliva (0 069 0 014);
P
¢
0 05 (Fig. 1). Fourteen of 19 patients had positive
<
P
0¢05; (3, 4): P 0¢05; (5, 6): P 0¢05.
<
<
<
serum IgG, while eighteen of 19 had positive salivary
IgG.
¢
§ ¢
P
¢
(0 260 0 96),
0 05. Salivary immune complexes
<
¢
§ ¢
were slightly higher in PCM patients (0 199 0 249)
Antigenemia
¢
§ ¢
than in controls (0 149 0 112), but this difference was
not statistically signiŽcant (
P
¢
0 05) (Fig. 3).
>
Soluble gp43 levels measured as OD in ELISA were
¢
§ ¢
signiŽcantly increased in patients’ sera (0 84 0 013 vs.
Salivary sIgA
¢
§ ¢
P
¢
0 146 0 083 for controls,
0 05) but not in saliva
<
¢
§ ¢ § ¢
¢
P
¢
(0 171 0 23 vs. 0 167 0 021 for controls,
0 05)
>
Salivary sIgA values were signiŽcantly higher in patients
(Fig. 2). All the PCM cases exhibited positive anti-
genemia in serum, but the three cases exhibiting antigens
in saliva all had buccal lesions noted prior to treatment,
although not during the salivary sample collection.
¢
§ ¢
¢
§ ¢
P
¢
(0 196 0 082) than controls (0 102 0 024),
OD ranged from 0 097 to 0 420. Fourteen of 19 patients
0 001.
<
¢
¢
were positive reactors (Fig. 1).
Discussion
Circulating immune complexes
As in previous Žndings [9], our results demonstrate an
P. brasiliensis
IgG antibody level in PCM
Circulating IC in patients’ serum, read as OD, were also
elevated anti-
¢
§ ¢
increased (0 609 0 191) as compared to controls
patients’ sera. Sera of 14/19 patients showed positive
ã
2001 ISHAM, Medical Mycology, 39, 423±428

426 Miura et al.
Fig. 3 Levels of serum and salivary immune complexes (IC), in
*
PCM patients (*) and normal donors ( ), in optical densities
(OD). ELISA plates coated with Mab anti-gp43 were incubated
with serum or saliva diluted 1/10 in PBS-T. Plates were washed,
incubated with peroxidase-conjugated anti-mouse IgG. Orthophe-
nylenediamine solution was added, and the reactions were read at
492 nm. Results are expressed in OD and horizontal bars represent
mean values. (1) PCM patients’ serum IC; (2) normal serum IC; (3)
Fig. 2 Levels of serum (a) and salivary (b) free gp43 antigenemia,
*
in PCM patients (*) and normal donors ( ), in optical densities
(OD). ELISA plates coated with rabbit IgG anti gp43 were
incubated with serum or saliva diluted 1/10 in PBS-T. Plates were
washed, incubated with mouse monoclonal antibodies (Mab) anti-
gp43 and then with peroxidase-conjugated anti-mouse IgG.
Orthophenylenediamine solution was added, and the reactions
were read at 492 nm. Results are expressed in OD and horizontal
bars represent mean values. (a) (1) PCM patients’serum; (2) normal
PCM patients’ salivary IC; (4) normal salivary IC. (1, 2) P 0¢05; (3,
<
4) P 0¢05.
>
levels of free speciŽc IgG were present in serum. It
would thus seem prudent to use more than one marker
for routine serodiagnosis and monitoring of treatment.
Anemia is a common complication of PCM, and is
similar to that seen in other chronic inammatory
conditions. It may be mild to moderate in intensity.
Antifungal drugs such as sulphonamides, ketoconazole,
and amphotericin B are also associated with anemia
[18,19]. Erythrocytes have an important role in binding
circulating IC and carrying these complexes to the liver
and spleen [20]. The reduction of the level of erythro-
cytes could hinder the system in removing circulating IC.
The lungs are the usual portal of entry of
serum (P 0.05). (b) (1) PCM patients’ saliva; (2) normal saliva
(P 0.05).
>
<
results by classical ELISA. The high threshold titre could
be due to some positive reactors in the normal controls,
who were from endemic areas [16]. The serum gp43 level
was also signiŽcantly different between PCM patients
and controls.
Circulating IC, consisting of two polypeptides with
apparent masses of 43 and 62 kDa, were previously
demonstrated through western blot assay in PCM
patients’ serum [17]. The present study showed increased
amounts of circulating IC containing gp43 molecules in
sera. In this study, patients had active chronic type PCM
and showed response to treatment, so that serum IgG
titres may have decreased with the death of
P. brasiliensis
. The infecting inoculum could stimulate
the mucosally associated lymphoid tissue and induce the
P. brasiliensis
production of sIgA to
. sIgA can efŽciently
eliminate microorganisms in the intestinal lumen or in
the mucus-rich environment of other mucosal surfaces
[21]. The activated plasmocytes producing the IgA may
migrate to other regions such as salivary glands [22].
The saliva has an important role in protecting the
mouth against endogenous microorganisms. It inhibits
microbial growth and adherence, neutralizes toxins and
P. brasiliensis.
This may have liberated soluble Ag and
consequently increased the levels of circulating IC.
IC were detected in 17/19 (89%) patients, whereas
speciŽc IgG was detected in only 13/19 (68%). In Žve of
the cases with negative IgG, IC levels appeared to be
elevated. This suggested that the imunoglobulins pro-
duced were complexed with gp43, and that only low
ã
2001 ISHAM, Medical Mycology, 39, 423±428

Immune complexes in paracoccidioidomycosis patients 427
agglutinates foreign cells, and ushes them to the gut.
Inhibition of adhesion is a major function of sIgA in
saliva [23].
for their excellent technical assistance and Robert Boyle
for the careful correction of the English text.
Because antigen–sIgA complexes are transported
across M-cells, specialized cells in the epithelia over
organized mucosa-associated lymphoid tissue in the
intestine, the bronchi, and the nasal and oral cavities, it
References
1 San-Blas G. Paracoccidioidomycosis and its etiological agent
Paracoccidioides brasiliensis. J Med Vet Mycol 1993; 31: 99–113.
2 McEwen JG, Garcia AM, Ortiz BL, et al. In search of natural
habitat of Paracoccidioides brasiliensis. Arch Med Res 1995; 26:
305–306.
3 Mendes RP. The Gamut of clinical manifestations. In: Franco
M, Lacaz CS, Restrepo-Moreno A, Del-Negro G, eds. Para-
coccidioidomycosis. 1st edn. Boca Raton, Florida: CRC Press,
1994: 233–258.
a
is possible that receptors speciŽc for IgA, Fc -receptors,
on antigen-presenting cells (B-cells or macrophages)
may enhance uptake and processing of antigens, thus
enhancing the immune response. In this way, reuptake of
secreted IgA may be important in maintenance of
antigenic stimulation and may contribute to secondary
responses [24,25].
The high level of anti-exoantigen sIgA observed in 14/
19 PCM patients’ saliva could be due to the migration of
activated IgA-producing plasma cells to other mucosal
regions such as oral mucosa. Alternatively, it could
reect the existence of focal infections in the buccal
mucosa inducing the local production of sIgA. The level
4 Manns BJ, Baylis BW, Urbanski SJ, et al. Paracoccidioidomy-
cosis: case report and review. Clin Infect Dis 1996; 23: 1026–
1032.
5 Mendes-Giannini MJS, Camargo ME, Lacaz CS, Ferreira AW.
Immunoenzymatic absorption test for serodiagnosis of para-
coccidioidomycosis. J Clin Microbiol 1984; 20: 103–108.
6 Camargo ZP, Unterkircher C, Campoy SP, Travassos LR.
Production of Paracoccidioides brasiliensis exoantigens for
immunodiffusion tests. J Clin Microbiol 1988; 26: 2147–2151.
7 Puccia R, Schenckman S, Gorin PAJ, Travassos LR. Exocellular
components of Paracoccidioides brasiliensis: identiŽcation of a
speciŽc antigen. Infect Immun 1986; 53: 199–206.
8 Mendes-Giannini MJ, Bueno JP, Shikanai-Yasuda MA, et al.
Antibody response to the 43 kDa glycoprotein of Paracocci-
dioides brasiliensis as a marker for the evaluation of patients
under treatment. Am J Trop Med Hyg 1990; 43: 200–206.
9 Bueno JP, Mendes-Giannini MJS, Del-Negro GMB, Assis CM,
Taniguti CK, Shikanai-Yasuda MA. IgG, IgM and IgA antibody
response for the diagnosis and follow-up of paracoccidioidomy-
cosis: comparison of counterimnunoelectrophoresis and com-
plement Žxation. J Med Vet Mycol 1997; 35: 213–217.
10 Camargo ZP, Gesztesi JL, Saraiva EC, Taborda CP, Vicentini
AP, Lopes JD. Monoclonal antibody capture enzyme immu-
noassay for detection of Paracoccidioides brasiliensis antibodies
in paracoccidioidomycosis. J Clin Microbiol 1994; 32: 2377–2381.
11 Blotta MHS, Camargo ZP. Immunological response to cell-free
antigens of Paracoccidioides brasiliensis: relationship with
clinical forms of paracoccidioidomycosis. J Clin Microbiol
1993; 31: 671–676.
12 Mendes-Giannini MJS, Bueno JP, Shikanai-Yasuda MA,
Ferreira AW, Masuda A. Detection of the 43,000-molecular-
weight glycoprotein in sera of patients with paracoccidioidomy-
cosis. J Clin Microbiol, 1989; 27: 2842–2845.
13 Sugizaki MF, Perac¸oli MT, Mendes-Giannini MJ, et al.
Correlation between antigenemia of Paracoccidioides brasilien-
sis and inhibiting effects of plasma in patients with paracocci-
dioidomycosis. Med Mycol 1999; 37: 277–284.
14 Cherquer-Bou-Habib D, Oliveira-Neto MP, Oliveira-da-Cruz
MF, Galva˜o-Castro B. The possible role of circulating immune
complexes in the deŽciency of cell-mediated immunity in
paracoccidioidomycosis. Braz J Med Biol Res 1989; 22: 205–212.
15 Lowry OH, Rosobrough MJ, Farr AL, Randall RJ. Protein
measurement with the Folin phenol reagent. J Biol Chem 1971;
246: 1889–1894.
P. brasiliensis
of salivary IgG reacting with
exoantigens
can also reect both circulating IgG and IgG produced in
response to local stimulation.
Levels of salivary gp43 were elevated, but no
statistically signiŽcant differences between patients and
controls were found. Using the mean plus the standard
deviation as the cut-off value for determining normal
levels of reaction with anti-gp43 Mab in saliva, we found
only three positive cases out of 19. Some positive
reactors among the normal controls were from endemic
areas and this could have contributed to the high
threshold titre [16]. The possibility of cross-reactions
with an antigen or antigens from the normal oral
microora cannot be excluded. Additional study is
required to determine whether gp43 is dispersed through
the oral mucosa or if its presence is due to the release of
P. brasiliensis
the
via local infection.
The cases of patients presenting a low level of serum
IgG but a high titre of IC reinforce the need, already
stressed in the literature, to use more than one marker
for diagnosis and treatment monitoring of PCM [26]. The
presence of sIgA in patient’s saliva suggests a protecting
role for neutralizing antigens on mucosal surfaces. This
will be the object of future investigations.
Acknowledgements
˜
This work was supported by Coordenac¸ao de Aperfei-
16 Lacaz CS, Passos-Filho MCR, Fava-Neto C, Macarron B.
´
õ
c¸oamento de Pessoal de N vel Superior (CAPES) and
˜
˜
Contribuic¸ao para o estudo da ‘‘blastomicose-infec¸cao’’. In-
´
˜
Coordenadoria de Pos-Graduac¸ao da Universidade
õ
´
´
´
querito com a paracoccidioidina. Estudo sorologico e cl nico-
Estadual de Londrina (CPG-UEL). The authors thank
´
radiologico de paracoccidioidino-positivos. Rev Inst Med Trop
´
Mari Sumigawa Kaminami and Nilson de Jesus Carlos
Sa˜o Paulo 1959; 1: 245–259.
ã
2001 ISHAM, Medical Mycology, 39, 423±428

428 Miura et al.
17 Unterkircher CS, Yazaki SC, Shimizu MT, Jorge AOC,
Camargo ZP. SpeciŽc components found in circulating immune
complexes (CIC) in paracoccidioidomycosis. J Med Vet Mycol
1996; 34: 273–277.
22 Kugler J, Hess M, Haake D. Secretion of salivary immunoglo-
bulin A in relation to age, saliva ow, mood states, secretion of
albumin, cortisol, and catecholamines in saliva. J Clin Immunol
1992; 12: 45–49.
18 Martinez R. Other Laboratory tests: hematologic and biochem-
ical changes. In: Franco M, Lacaz CS, Restrepo-Moreno A, Del
Negro G, eds. Paracoccidioidomycosis. 1st edn. Boca Raton,
Florida: CRC Press, 1994: 365–372.
19 Barraviera B, Mendes RP, Pereira PC, Machado JM, Curi PR,
Meira DA. Measurement of glucose-6-phosphate dehydrogen-
ase and glutathione reductase activity in patients with para-
coccidioidomycosis treated with ketoconazole. Mycopathologia
1988; 104: 87–91.
23 Tenovuo, J. Antimicrobial function of human saliva – how
important is it for oral health? Acta Odontol Scand 1998; 56:
250–256.
24 Kraehenbuhl J-P, Neutra MR. Molecular and cellular basis of
immune protection of mucosal surfaces. Physiol Rev 1992; 72:
853–879.
25 Neutra MR, Kraehenbuhl J-P. Mucosal immunization via M
cells for production of protective secretory IgA antibodies. Am J
Med Hyg 1994; 50: 10–13.
20 Goldsby RA, Kindt TJ, Osborne BA. The Complement System.
Kuby Immunology. 4th edn. New York: W.H. Freeman and
Company, 2000: 329–350.
21 Tomasi TB, Jr. Mechanisms of immune regulation at mucosal
surfaces. Rev Infect Dis 1983; 5: 784–792.
26 Martins R, Marques S, Alves M, Fecchio D, Franco M.
Serological follow-up of patients with paracoccidioidomycosis
treated with itraconazole using dot blot, ELISA and Western
Blot. Rev Inst Med Trop Sa˜o Paulo 1997; 39: 261–269.
ã
2001 ISHAM, Medical Mycology, 39, 423±428