CLINICAL AND VACCINE IMMUNOLOGY, June 2007, p. 700–709
Vol. 14, No. 6
Production and Evaluation of Reagents for Detection of
Histoplasma capsulatum Antigenuria by
Mark D. Lindsley,1* Heather L. Holland,1† Sandra L. Bragg,1Steven F. Hurst,1
Kathleen A. Wannemuehler,2and Christine J. Morrison1‡
Mycotic Diseases Branch1and Office of the Director, Division of Food-borne, Bacterial, and Mycotic Diseases,2
Centers for Disease Control and Prevention, Atlanta, Georgia
Received 12 February 2007/Returned for modification 19 March 2007/Accepted 29 March 2007
The detection of urinary Histoplasma capsulatum polysaccharide antigen (HPA) by enzyme immunoassay
(EIA) has proven useful for the presumptive diagnosis of histoplasmosis in AIDS patients. Assay limitations
include (i) detection of a largely uncharacterized antigen and (ii) difficulty in reproducibly generating anti-
bodies for use in the EIA. To improve antibody production for use in this test and to better understand the
antigen being detected, we compared rabbit antibodies elicited using various immunization schedules, routes,
and H. capsulatum-derived antigens. Antibodies were evaluated by EIA for their ability to detect purified H.
capsulatum C antigen (C-Ag) and antigenuria. Reported as enzyme immunoassay (EI) units (the A450with
antigen divided by the A450without antigen), results demonstrated that intravenous immunization of rabbits
with whole, killed yeast-phase cells (yeast-i.v. regimen) produced antibodies giving the highest EI values in the
C-Ag EIA (mean EI units ? standard deviation, 14.9 ? 0.6 versus 6.4 ? 0.4 for rabbits immunized with C-Ag
versus 2.4 ? 0.3 for all other regimens combined). Yeast-i.v. antibodies were highly sensitive for the detection
of antigenuria in patients with histoplasmosis, as shown by the following results: 12/12 patients compared to
10/12, 6/12, 3/12, and 3/12, respectively, for antibodies from rabbits immunized with (i) C-Ag; (ii) whole, killed
yeast-phase cells administered subcutaneously and intramuscularly; (iii) yeast-phase culture filtrates; and (iv)
HPA-positive urine. Rabbits immunized using the yeast-i.v. regimen also gave higher peak antibody titers than
rabbits immunized by any other regimen (P < 0.03), and their antibodies were most comparable in reactivity
to antibodies produced for use in the standard HPA-EIA test (P < 0.001). Therefore, rabbits immunized using
the yeast-i.v. regimen produced the most sensitive antibodies with the highest titers for detection of C-Ag and
antigenuria in histoplasmosis patients.
Histoplasma capsulatum is a thermally dimorphic fungus that
is worldwide in distribution. Endemic to the Mississippi and
Ohio River valleys of North America, H. capsulatum conidia
are most often found in soil enriched with bird or bat guano (1,
7, 22). Disruption of contaminated soil causes aerosolization of
fungal conidia which can then enter the body via inhalation
(20). The resulting disease is usually self-limited in healthy
individuals but can cause serious, disseminated disease in those
with underlying immunosuppression (17, 30). A definitive di-
agnosis of histoplasmosis is obtained by positive culture from a
clinical specimen or by histopathologic evidence of infection in
tissues. However, recovery of H. capsulatum from clinical ma-
terials requires up to 4 weeks for growth to occur, and his-
topathologic testing, involving invasive procedures to obtain
tissue, is insensitive and requires expertise for interpretation
(36). Therefore, serologic tests to detect circulating anti-H.
capsulatum antibodies are commonly relied upon as an aid to
Complement fixation and immunodiffusion are standard
tests for the serologic diagnosis of histoplasmosis (21). Com-
bined results from both tests help to improve the overall sen-
sitivity and specificity for the diagnosis of histoplasmosis in
immunocompetent patients (24). However, antibody titers are
often negative or equivocal early in infection, and a second
specimen, obtained 3 to 4 weeks later, is required for confir-
mation, thereby delaying diagnosis. Furthermore, immunosup-
pressed individuals, who are most at risk for the development
of disseminated histoplasmosis, may be antibody deficient,
leading to falsely negative serology results (27, 30, 43). For
example, it has been reported that complement fixation titers
of 1:32 or greater occurred in 83% (25 of 30) of nonimmuno-
compromised histoplasmosis patients but in only 50% (16 of
32) of immunocompromised patients (P ? 0.05) (27). A diag-
nostic method that does not rely upon an antibody response is
therefore especially valuable in such cases. Detection of H.
capsulatum polysaccharide antigen (HPA) (13, 42) in body
fluids, especially urine, has been useful in the presumptive
diagnosis of H. capsulatum infections in patients with dissem-
inated disease. For example, histoplasmosis antigenuria was
detected in 92% of patients with disseminated disease, in 39%
of patients with self-limited disease, and in 21% of patients
with the chronic pulmonary form (31, 40).
* Corresponding author. Mailing address: 1600 Clifton Road, NE,
Mailstop G-11, Atlanta, GA 30333. Phone: (404) 639-4340. Fax: (404)
639-3546. E-mail: email@example.com.
† Present address: Strategic Science and Program Unit, Coordinat-
ing Center for Infectious Diseases, Mailstop A-02, Centers for Disease
Control and Prevention, 1600 Clifton Road, NE, Atlanta, GA 30333.
‡ Present address: Office of Public Health Research, Office of the
Director, Mailstop D-72, Centers for Disease Control and Prevention,
1600 Clifton Road, NE, Atlanta, GA 30333.
?Published ahead of print on 11 April 2007.
The standard format of the HPA detection assay is a double
antibody sandwich enzyme immunoassay (HPA-EIA) (13)
modified from the original radioimmunoassay format (42). An-
tibodies raised for both the capture and detection of HPA were
produced by immunizing rabbits subcutaneously with whole,
killed H. capsulatum yeast-phase cells (in an adjuvant emul-
sion), followed by intravenous booster immunizations with live
organisms (13, 42). Various modifications of the test format
have been implemented over time (16, 42a), but the urinary
antigen detected has never been purified to homogeneity or
fully characterized. Studies to date have indicated that the
antigen is primarily (?95%) carbohydrate by weight, is stable
to boiling, and is not destroyed by pronase treatment (42).
Unfortunately, the efficiency of antibody production against
this antigen (42) in rabbits has been very poor; immunization
of as many as 40 rabbits can be required to obtain a single
rabbit with a sufficiently robust antibody response to be useful
in the HPA-EIA.
This lack of a robust antibody response to the HPA antigen
may be caused by any number of factors, including the follow-
ing: (i) a poorly immunogenic antigen, (ii) an immunologically
inaccessible antigen, (iii) an insufficient amount of specific
antigen, (iv) ineffective or insufficient adjuvant to facilitate an
adequate immune response, and/or (v) a nonoptimal immuni-
zation schedule and/or immunization route. Therefore, the
present study was undertaken (i) to determine the best anti-
genic preparation to employ as well as the optimum immuni-
zation regimen to follow to reproducibly generate the maximum
quantity of antibodies to detect H. capsulatum antigenuria and
(ii) to obtain a better understanding of the nature of the anti-
gen being detected. To accomplish these objectives, five different
immunization regimens and several different H. capsulatum-
derived antigen preparations were examined. Antibodies were
then evaluated and compared to one another and to antibodies
produced for use in the standard commercial HPA-EIA for
their capacity to detect chromatographically purified H. cap-
sulatum polysaccharide C antigen (C-Ag) and to detect anti-
genuria in histoplasmosis patients.
MATERIALS AND METHODS
Microorganism. A clinical isolate of H. capsulatum (Thon strain) was kindly
provided by L. Joseph Wheat, MiraVista Laboratories, Indianapolis, IN, and was
used throughout this study to produce antigens for rabbit immunizations.
Chemicals. All chemicals were obtained from Sigma-Aldrich, Co., St. Louis,
MO, unless otherwise indicated.
Patient urine. Banked patient urine samples (with direct identifiers removed)
from 12 patients with histoplasmosis were obtained from L. Joseph Wheat
(MiraVista Laboratories). Normal human urine was voluntarily obtained from
healthy adult subjects at the Centers for Disease Control and Prevention (CDC).
Preparation of antigens for rabbit immunizations. Rabbits were immunized
by one of five different regimens using one of the four antigen preparations
(i) Whole, killed yeast cells. H. capsulatum yeast-phase cells were inoculated
into Pine’s liquid medium for yeast-phase H. capsulatum (23), and cultures were
placed at 37°C on a platform shaker rotating at 150 rpm. After incubation for
72 h, thimerosal was added to the culture to a final concentration of 0.02%.
Cultures were then incubated in the presence of thimerosal for an additional 48 h
at 37°C with shaking. Yeast cells were harvested by centrifugation at 2,100 ? g
for 10 min and washed three times in 0.01 M phosphate-buffered saline (PBS; 8.1
mM Na2HPO4, 1.9 mM KH2PO4, 0.15 M NaCl), pH 7.2. Yeast cells were then
suspended in PBS containing 0.01% (wt/vol) thimerosal to a final concentration
of 30% (vol/vol) packed yeast cells. Suspensions were frozen at ?20°C until used.
Immediately before use, yeast-phase H. capsulatum cells were thawed and
diluted in PBS-thimerosal to a 0.2% (vol/vol) suspension of packed yeast cells for
intravenous immunizations and to a 20% (vol/vol) suspension of packed yeast
cells for subcutaneous immunizations. Sterility checks were performed by incu-
bating 100-?l aliquots of the thimerosal-treated yeast cells on brain heart infu-
sion agar containing 5% sheep erythrocytes (BBL-Becton Dickinson, Sparks,
MD) and on Sabouraud dextrose agar (BBL) for 8 days at both 25°C and 37°C.
(ii) C-Ag. Culture filtrates from mycelial-phase growth of H. capsulatum (45)
were purified chromatographically using a CM-Sepharose column as described
previously (44). The resulting C-Ag preparation contained no detectable H or M
antigens as determined by immunodiffusion assay against H. capsulatum refer-
ence antisera (24). The C-Ag preparation contained ?95% carbohydrate, as
measured by the micro phenol-sulfuric acid procedure (12), and ?1% protein, as
determined by the Bradford assay (4). The final C-Ag preparation contained 10
mg of carbohydrate per ml in 0.01 M borate-buffered saline (9.5 mM H3BO3, 0.5
mM Na2B4O7, and 0.15 M NaCl, pH 8.0).
(iii) HPA-positive urine. HPA-positive urine from a patient with proven dis-
seminated histoplasmosis was kindly provided by L. Joseph Wheat. Urine had
been autoclaved and concentrated 20-fold before receipt. Analysis of the con-
centrated, HPA-positive urine by the phenol-sulfuric acid and Bradford assays
demonstrated that the final carbohydrate and protein concentrations for this
antigen preparation were 19.9 mg/ml and 5.67 mg/ml, respectively.
(iv) Yeast-phase culture filtrate antigen. Pine’s liquid medium for yeast-phase
H. capsulatum (23) was inoculated with H. capsulatum yeast-phase cells and
incubated at 37°C for 7 days on a rotating platform (150 rpm). Culture super-
natants were harvested by centrifugation at 10,000 ? g for 20 min and were
concentrated 10-fold by pressure ultrafiltration using a YM-10 membrane (Ami-
con, Inc., Danvers, MA). The resulting 10-fold concentrate was further concen-
trated 2.6-fold by using a centrifugal filtration unit (Centricon instrument with a
10,000-molecular-weight cutoff; Millipore Corp., Billerica, MA). The 26-fold
concentrated yeast-phase culture filtrate antigen demonstrated final carbohy-
drate and protein concentrations of 20 mg/ml and 1.5 mg/ml, respectively.
Immunization regimens. Female, New Zealand White rabbits, 2 to 3 kg in
weight, were used in the following experiments. Rabbits were fed and watered ad
libitum, and all Institutional Animal Care and Use Committee recommendations
were followed. Twenty rabbits were divided into five groups such that each group
was subjected to one of the following immunization regimens: protocol 1, intra-
venous injection followed by subcutaneous booster injections with whole, killed
H. capsulatum yeast-phase cells (yeast-i.v. group); protocol 2, subcutaneous and
intramuscular injection followed by subcutaneous booster injections with H.
capsulatum-derived C-Ag (C-Ag group); protocol 3, subcutaneous and intramus-
cular injection followed by subcutaneous booster injections with whole, killed H.
capsulatum yeast-phase cells (yeast-s.c. group); protocol 4, subcutaneous and
intramuscular injection followed by subcutaneous booster injections with con-
centrated HPA-positive patient urine (H. capsulatum antigen-positive urine
group); and protocol 5, subcutaneous and intramuscular injection followed by
subcutaneous booster injections with concentrated H. capsulatum yeast-phase
culture filtrate antigen (culture filtrate group). One rabbit immunized using
protocol 3 died during the immunization process, prior to the generation of a
significant antibody response, and was eliminated from further evaluation.
(i) Protocol 1. Four rabbits were injected intravenously with 1 ml each of
whole, killed H. capsulatum yeast-phase cells on days 0, 1, 4, 8, 10, 15, 18, 23, 29,
and 32. After day 56, rabbits were boosted every 4 to 6 weeks by injecting a total
of 1 ml of whole, killed H. capsulatum yeast-phase cells subcutaneously at four
dorsal sites. Rabbits were bled from the central ear artery on day 0 (before
immunization) and on days 14, 28, 42, 49, and 56 and then weekly beginning 7 to
10 days after each booster injection.
(ii) Protocols 2 through 5. All antigen preparations for immunization proto-
cols 2 through 5 were made by emulsifying equal volumes of the given antigen
with an adjuvant (TiterMax USA, Inc., Norcross, GA). Emulsification was facil-
itated by repeated passage of the antigen-adjuvant mixture through a double-
hubbed syringe. Initially, each rabbit received 0.2 ml of the emulsion in four
different inoculation sites: subcutaneously over each shoulder and intramuscu-
larly in each hind quarter. After day 56, rabbits were boosted subcutaneously
every 4 to 6 weeks at the original immunization sites, but antigen was suspended
in saline rather than in adjuvant. Each rabbit was bled on day 0 (before immu-
nization) and on days 14, 28, 42, 49, and 56 and then weekly beginning 7 to 10
days after each booster injection.
Indirect EIA to screen rabbit antisera. Chromatographically purified C-Ag
(44) was diluted to 0.1 ?g/ml in PBS containing 0.01% (wt/vol) sodium azide.
One hundred microliters of this solution was then added to each well of a
flat-bottom microtiter plate (Immulon 2 HB; Thermo Electron Corp., Milford,
MA). Plates were covered with plastic film and held at 4°C for 48 to 72 h.
Microtiter plate wells were then washed three times with distilled water, and 100
?l of PBS containing 0.025% (wt/vol) sodium azide was added to each well.
VOL. 14, 2007REAGENT OPTIMIZATION FOR H. CAPSULATUM ANTIGEN EIA701
Covered plates were stored at 4°C for up to 7 days before use. Immediately
before use, plates were washed three times with PBS containing 0.05% (vol/vol)
Tween 20 (PBS-T), and any remaining fluid was removed by vigorous inversion
onto paper towels.
Antiserum from each of the immunized rabbits was serially diluted from
1:1,000 to 1:512,000 in PBS-T containing 0.1% (wt/vol) bovine serum albumin
(BSA). Serially diluted antiserum was then added to the wells of the micro-
titer plates. Plates were covered with plastic film and incubated for 1 h at
ambient temperature. Microtiter plate wells were then washed three times
with PBS-T (without BSA) and vigorously inverted onto paper towels to
remove any remaining fluid. One hundred microliters of horseradish perox-
idase-labeled goat anti-rabbit immunoglobulin G (IgG) (H?L; Bio-Rad Lab-
oratories, Hercules, CA), diluted 1:2,000 in PBS-T, was then added to each
well. Plates were covered with plastic film and incubated for 1 h at ambient
temperature. The wells were washed four times with PBS-T before addition
of 100 ?l of a colorimetric substrate mixture (50:50 [vol/vol] mixture of 3,3?,5,
5?-tetramethylbenzidine [TMB] colorimetric substrate and H2O2, supplied in
a reagent kit; KPL, Inc., Gaithersburg, MD). Plates were incubated for 30
min, and results were obtained spectrophotometrically at A650(SpectraMax
250; Molecular Devices Corp., Sunnyvale, CA).
Antibody purification. Purification of rabbit Igs was initially performed using
standard ammonium sulfate precipitation methods (19). Briefly, equal volumes
of rabbit antiserum and 70% saturated ammonium sulfate solution were com-
bined and incubated at room temperature for 4 h with continuous gentle stirring.
The preparation was then centrifuged at 10,000 ? g for 5 min, and the pellet was
resuspended in distilled water to the original volume of antiserum. An equal
volume of 70% saturated ammonium sulfate was added with continuous gentle
stirring. The mixture was centrifuged at 10,000 ? g for 5 min, and the ammonium
sulfate precipitation procedure was repeated once again. The sample was cen-
trifuged as before, and the pellet was resuspended to one-half of the original
volume of antiserum using distilled water. The ammonium sulfate-precipitated
antibodies were then dialyzed overnight at 4°C against 1,000 volumes of PBS
(Slide-A-Lyzers; Pierce Chemical Co., Rockford, IL). The dialyzed sample was
assayed for protein content using a bicinchoninic acid protein assay reagent kit
(Pierce). All antibody preparations were suspended to a final concentration of 10
mg/ml and were stored at ?20°C until used.
Dot blot enzyme immunoassay. One microliter of each of the following was
pipetted onto strips cut from nitrocellulose membrane (Optitran, 0.45-?m po-
rosity; Schleicher & Schuell BioScience, Inc., Keene, NH) to produce discrete
dots: C-Ag (1.6, 8, and 40 ?g/ml), normal human urine, and PBS. The strips were
allowed to dry at ambient temperature for approximately 20 min before place-
ment into individual lanes of a strip incubation tray (Bio-Rad). The following
incubation and wash steps were performed at ambient temperature on a rocker
platform unless otherwise indicated. To each lane of the tray, 5 ml of 5% (wt/vol)
nonfat dried milk in PBS-T was added as a blocking agent, and strips were
incubated in this solution for 5 min. The blocking solution was removed by
aspiration, and each strip was washed three times with PBS-T for 5 min. Purified
rabbit antibodies (3.5 ml), obtained from each of the five immunization proto-
cols, were diluted to 25 ?g/ml in PBS-T and added to a given lane of the tray.
Strips were incubated for 1 h and then washed for 5 min with PBS-T that had
been heated to 50°C. This wash was followed by three 5-min washes with PBS-T
at ambient temperature. After aspiration of the wash solution, 5 ml of a 1:4,000
dilution of horseradish peroxidase-labeled goat anti-rabbit IgG (Bio-Rad) in
PBS-T was added to each lane, and trays were incubated for 1 h. Strips were then
washed four times for 5 min each with PBS-T followed by a single wash with PBS
only for 5 min. Wash fluid was removed and replaced with 5 ml of a colorimetric
substrate solution, made immediately before use by mixing 7.5 ?l of a 30%
solution of H2O2with 75 ml of a 0.05% (wt/vol) solution of DAB (3?3?-diami-
nobenzidine tetrachloride dehydrate) made in PBS. Strips were incubated in the
H2O2-DAB solution for 20 min before being washed three times for 5 min each
with distilled H2O. Strips were allowed to dry overnight before the intensity of
the colorimetric reaction was recorded.
Biotinylation of purified antibodies. One-half of the ammonium sulfate-puri-
fied antibodies were labeled with biotin for use in a double-antibody EIA.
Biotinylation of purified antibodies was conducted using the EZ-Link biotin
hydrazide kit (Pierce) according to the manufacturer’s instructions. Five hundred
microliters of ammonium sulfate-purified antibodies (10 mg/ml) was placed in a
Slide-A-Lyzer 2K Dialysis Cassette (Pierce) and dialyzed overnight at 4°C
against 200 volumes of 0.1 M sodium acetate coupling buffer (86 mM sodium
acetate [C2H3NaO2], 14 mM glacial acetic acid [C2H4O2], pH 5.50). The anti-
body was removed from the dialysis cassette and diluted with fresh coupling
buffer to a final concentration of 2 mg/ml and placed at 4°C. Once the solution
reached 4°C, an equal volume of sodium meta-periodate (dissolved in coupling
buffer to a concentration of 20 mM and equilibrated to 4°C) was added, and the
mixture was incubated in the dark for 30 min at 4°C. Glycerol was then added to
a final concentration of 15 mM, and the mixture was held at 4°C for 5 min. The
solution was then dialyzed overnight at 4°C against 1,000 volumes of coupling
buffer. One part of 50 mM biotin hydrazide, dissolved in dimethyl sulfoxide, was
added to nine parts of the antibody preparation. The mixture was agitated by
rotation for 2 h at 60 rpm in the dark at ambient temperature. The sample was
then dialyzed overnight at 4°C against 1,000 volumes of PBS. An equal volume
of glycerol was added to the dialyzed sample to yield a final antibody concen-
tration of 0.5 mg/ml. This solution was stored at ?20°C until used.
Double-antibody sandwich EIA. Ammonium sulfate-purified anti-H. capsula-
tum antibodies, produced from each of the immunization protocols 1 to 5, or
anti-H. capsulatum antibodies (anti-HPA antibodies) provided by L. Joseph
Wheat were diluted to 20 ?g/ml in 0.01 M Tris-HCl buffer, pH 7.0. One hundred
microliters of purified antibodies was then added to each well of a flat-bottom
microtiter plate (Immulon 2 HB; Thermo Electron Corp., Milford, MA). The
plate was covered with a plastic lid and incubated at 37°C for 1 h. Plates were
then washed six times with PBS-T using an ELX50 Autostrip washer (BioTek
Instruments, Inc., Winooski, VT). Residual wash buffer was removed from the
microtiter plate wells by vigorous inversion of the plate onto paper towels. Two
hundred microliters of 5% (wt/vol) BSA solution in 0.01 M Tris-HCl buffer, pH
7.0, was then added to each well. The plate was covered with a plastic lid,
incubated at 37°C for 1 h, and then washed as described above. Urine samples
from patients and healthy individuals were heated at 100°C for 5 min and allowed
to cool before 100 ?l of each sample was added to duplicate wells of the
precoated microtiter plate. Positive and negative control wells, respectively,
contained 100 ?l of a 0.32 ?g/ml solution of purified C-Ag and 100 ?l of 0.01 M
Tris-HCl buffer, pH 7.0. After addition of the test samples, the plate was covered
with a plastic lid, incubated at 37°C for 1 h, and washed as before. One hundred
microliters of biotinylated rabbit anti-H. capsulatum antibody in 0.1 M Tris-HCl
buffer, pH 8.0, was then added to each well, and plates were incubated at 37°C
for 1 h. Plates were washed as before, and 100 ?l of a 1:1,000 dilution of
streptavidin-horseradish peroxidase (Pierce), diluted in 0.1 M Tris-HCl buffer
(pH 8.0) containing 5% (wt/vol) BSA, was added per well. The plate was again
covered with a plastic lid, incubated at 37°C for 1 h, and washed as before. One
hundred microliters of a colorimetric substrate solution (50:50 [vol/vol] mixture
of TMB and H2O2; KPL) was then added to each well, and plates were incubated
at ambient temperature for 15 min. The reaction was stopped by the addition of
100 ?l of 1.0 M H2SO4to each well, and plates were read spectrophotometrically
at A450using a SpectraMax 250 microtiter plate reader (Molecular Devices). The
coefficient of variance for within-run testing ranged from 1.7 to 6.4%, and the
coefficient of variance for between-run testing ranged from 3.9 to 7.7%, depend-
ing on the concentration of C-Ag tested.
The enzyme immunoassay index (EI) values were calculated by dividing the
mean A450value obtained for wells containing patient urine by the mean A450
value obtained for wells containing urine from healthy controls. EI values for the
positive antigen control wells were calculated by dividing the mean A450value
obtained for wells containing purified C-Ag by the mean A450value obtained for
wells that received 0.01 M Tris-HCl buffer, pH 7.0. A positive EIA result using
patient urine was defined as an EI value of ?1.
Limit of sensitivity for the detection of serially diluted antigen. The lower limit
of assay sensitivity was determined using fivefold serial dilutions of chromato-
graphically purified C-Ag (range, 0.1 to 8,000 ng/ml) in a double-antibody sand-
wich EIA using antibodies produced by the yeast-i.v. immunization regimen.
Statistical analyses. Differences between antibody titers obtained for each of
the immunization regimens were determined using the exact Wilcoxon-2 sample
test. Differences between immunization groups in the mean EI values obtained
using the indirect antibody EIA and the double-antibody sandwich EIA were
determined using the paired Student t test. Pearson’s r correlation coefficient was
calculated to determine correlations between test groups. Differences between
comparison groups as determined using any of the statistical tests employed were
considered to be significant when the P value was ?0.05.
Screening rabbit antisera in an indirect EIA. To evaluate
and compare sera produced by each of five different immuni-
zation regimens for their potential use in the detection of H.
capsulatum urinary antigen, 180 sera from 19 rabbits (repre-
senting three to four rabbits from each of the five regimens
examined) were collected at regular intervals throughout the
702LINDSLEY ET AL.CLIN. VACCINE IMMUNOL.
immunization process. Sera were then tested in an indirect
EIA against chromatographically purified H. capsulatum C-Ag
(44). C-Ag was used to coat microtiter plates for the initial
screening assay because the HPA urinary antigen detected by
commercial anti-HPA antibodies has never been purified to
homogeneity or fully characterized. Therefore, HPA was not
available in purified form for use as a screening antigen. For
this reason and because C-Ag shares many of the properties
attributed to HPA (e.g., HPA is primarily [?95%] carbohy-
drate by weight, is stable to boiling, is not destroyed by pronase
or nuclease treatment, is destroyed by mixed glycosidases or
periodate treatment, and is removed by binding to concanava-
lin A) (42), C-Ag served as a logical antigen for screening
rabbit antisera for use in the detection of H. capsulatum anti-
genuria. Further, the commercial HPA-EIA has also charac-
teristically demonstrated cross-reactivity among specimens ob-
tained from patients with diseases caused by fungi known to
share common cell wall carbohydrate antigens such as the
C-Ag (2, 11, 26, 33, 35, 44).
Figure 1 depicts the peak anti-C-Ag antibody titers obtained
for each of the various immunization regimens employed when
tested in the indirect EIA. Rabbits immunized using the yeast-
i.v. protocol most consistently demonstrated high peak anti-
body titers. All four rabbits in this group demonstrated anti-
body titers of ?1:512,000 (Fig. 1). Rabbits immunized using
the C-Ag protocol demonstrated peak titers that varied from
1:128,000 to ?1:512,000 among the four rabbits immunized.
Peak antibody titers ranged from 1:8,000 to 1:32,000 among
the four rabbits immunized with H. capsulatum yeast-phase
culture filtrate antigens (culture filtrate) (Fig. 1). Three rabbits
immunized with whole, killed H. capsulatum yeast-phase cells,
administered subcutaneously and intramuscularly (yeast-s.c.),
produced peak antibody titers of 1:64,000. In contrast, rabbits
immunized with concentrated urine from a patient with dis-
seminated histoplasmosis (H. capsulatum antigen-positive urine)
displayed the lowest antibody response (i.e., all four rabbits
produced peak antibody titers of only 1:1,000).
Pairwise comparisons between immunization groups indi-
cated that sera from rabbits immunized intravenously with
whole, killed H. capsulatum yeast-phase cells (yeast-i.v.), as
well as rabbits immunized with purified C-Ag, demonstrated
peak antibody titers that were significantly greater (P ? 0.03)
than those obtained using any of the remaining three immu-
nization regimens (Fig. 1). It was further shown that use of the
yeast-s.c. protocol resulted in peak antibody titers that were
significantly greater (P ? 0.03) than those for both the culture
filtrate and the H. capsulatum antigen-positive urine protocols.
Peak antibody titers for rabbits immunized by the culture fil-
trate protocol were also found to be significantly greater (P ?
0.03) than those for rabbits immunized by the H. capsulatum
antigen-positive urine protocol (Fig. 1). Therefore, rabbits im-
munized using the yeast-i.v. protocol demonstrated the highest
peak antibody titers and also the most consistently high peak
antibody titers (i.e., all four rabbits gave antibody titers of
The mean time required for rabbits that had been immu-
FIG. 1. Evaluation of antisera to detect C-Ag in an indirect EIA. Rabbits were either immunized intravenously using whole, killed H.
capsulatum yeast-phase cells (yeast-i.v.) or immunized subcutaneously and intramuscularly using one of the following immunogens: chromato-
graphically purified C-antigen (C-Ag); whole, killed H. capsulatum yeast-phase cells (yeast-s.c.); filtrates from H. capsulatum yeast-phase cell
cultures (culture filtrate); or concentrated urine from a patient with disseminated histoplasmosis (H. capsulatum antigen-positive urine). Sera were
screened against chromatographically purified C-Ag in an indirect EIA (see inset). Maximum EIA titers obtained for three to four rabbits in each
immunization group are shown. Each symbol (}) represents one rabbit. The geometric mean antibody titer obtained for each immunization group
is represented by a horizontal bar; maximum titers varied among immunization groups from as high as ?1:512,000 (yeast-i.v. and C-Ag) to as low
as 1:1,000 (H. capsulatum antigen-positive urine). Hc, H. capsulatum.
VOL. 14, 2007 REAGENT OPTIMIZATION FOR H. CAPSULATUM ANTIGEN EIA 703
nized by the yeast-i.v. regimen to reach peak antibody titers
was 32 days (range, 28 to 42 days). This compares to 49 days
(H. capsulatum antigen-positive urine group), 53 days (culture
filtrate group), 71 days (C-Ag group), and 79 days (yeast-s.c.
group) to reach peak antibody titers when rabbits were immu-
nized by other methods. Peak antibody titers remained stable
for 32 to 58 days (32 days for the yeast-i.v. group and 58 days
for the C-Ag group; rabbits in all other immunization groups
sustained peak antibody levels for intervals between 32 and 58
days). These data suggest that the yeast-i.v. immunization reg-
imen resulted in the most consistently high peak antibody titers
and that peak titers for this group were attained in a little over
1 month following initial immunization. Interestingly, the high-
est peak EIA titers obtained for antibodies from the yeast-
i.v. immunization group were obtained before subcutaneous
booster immunizations were initiated. Therefore, these data
indicate that the intravenous immunization of rabbits with
whole, killed H. capsulatum yeast-phase cells alone was respon-
sible for the majority of the immune response observed and
that subsequent subcutaneous booster immunizations are not
Reactivity of purified antibodies, including anti-HPA anti-
bodies, with C-Ag in a dot blot enzyme immunoassay. Sera
from two rabbits from each immunization group that demon-
strated the highest titers in the indirect EIA were selected,
purified by ammonium sulfate precipitation, and adjusted to a
25 ?g/ml concentration. These antibodies, along with anti-
HPA antibodies obtained from L. Joseph Wheat, were evalu-
ated in an indirect dot immunoblot assay to detect C-Ag.
Replicate nitrocellulose membrane strips were dotted with
three concentrations of serially diluted, chromatographically
purified C-Ag (40, 8, and 1.6 ?g/ml, respectively) and reacted
with antibodies from each of the immunization groups or with
anti-HPA antibodies (Fig. 2). Antibodies from rabbits that had
been immunized using the yeast-i.v. protocol demonstrated the
strongest reactivity to C-Ag, and this reactivity was dose de-
pendent. Reactivity of a similar, dose-dependent intensity was
observed using rabbit anti-HPA antibodies (Fig. 2, HPA).
These data indicate that the antibodies used in the standard
HPA-EIA urinary antigen test (13) detect the C-Ag of H.
capsulatum in a dose-dependent manner and suggest that the
antigen detected in the HPA-EIA may be, at least in part, the
C-Ag of H. capsulatum.
Antibodies produced by rabbits in the remaining immuniza-
FIG. 2. Reactivity of purified antibodies, including anti-HPA anti-
bodies, with C-Ag in a dot blot enzyme immunoassay. C-Ag was dotted
onto nitrocellulose membrane strips at concentrations of 40, 8, or 1.6
?g/ml and reacted with purified antibodies from each of the immuni-
zation groups described in the legend of Fig. 1. In addition, antibodies
produced for use in the commercial HPA-EIA (HPA) were tested in
parallel. Membrane strips that received PBS instead of primary anti-
body (no 1° Ab) or instead of all reagents (PBS) served as negative
controls and did not react with any antibody tested. Pooled normal
human urine also did not react with any antibody tested (not shown).
In contrast, all antibodies reacted with C-Ag in a dose-dependent
manner, although to various degrees. Hc, H. capsulatum.
FIG. 3. Comparison of antibodies from each immunization group using a double-antibody sandwich EIA. The most reactive antibodies within
each immunization group were used to capture C-Ag onto microtiter plates in a double-antibody sandwich EIA (see inset). Immunization groups
were as noted in the legend of Fig. 1. Biotinylated anti-HPA antibodies were used throughout as detector antibodies. The bars and corresponding
standard deviations represent the mean and standard deviations of five replicate EIA index measurements. Antibodies produced by the yeast-i.v.
immunization regimen demonstrated the highest reactivity to C-Ag compared to all other immunization groups. Reactivity of anti-HPA antibodies
(not shown) was intermediate to that of the yeast-i.v. and C-Ag group antibodies. Hc, H. capsulatum.
704 LINDSLEY ET AL.CLIN. VACCINE IMMUNOL.
tion groups also reacted with purified C-Ag in a dose-depen-
dent manner; however, the intensity of this reactivity varied
considerably among the different immunization groups. For
example, antibodies produced using H. capsulatum antigen-
positive urine as the immunogen demonstrated weak reactivity
with C-Ag (Fig. 2). Antibodies from rabbits immunized by the
yeast-s.c. and C-Ag regimens gave reactions that were robust,
although less intense than those for the yeast-i.v. immunization
group. Lastly, antibodies from rabbits immunized using the
culture filtrate immunization regimen reacted even less
strongly than antibodies from the yeast-s.c. and C-Ag groups.
No C-Ag reactivity was observed using normal human urine
(not shown), PBS as a negative control (Fig. 2, PBS), or in the
absence of primary antibody (Fig. 2, no 1° Ab).
Comparison of the anti-C-Ag reactivities of antibodies from
each immunization group using a double-antibody sandwich
EIA. Antibodies from the most C-Ag-reactive rabbit from each
immunization group were then tested in a colorimetric double-
antibody sandwich EIA to measure the capacity of each to
capture purified C-Ag onto microtiter plates (Fig. 3). The same
biotinylated anti-HPA detector antibody was used throughout
so that the only variable among test groups was the capture
antibody employed. Five (replicate) tests were performed us-
ing antibodies from each of the five immunization groups.
Antibodies obtained from the yeast-i.v. immunization regimen
demonstrated the greatest capacity to capture C-Ag (mean
EI value ? standard deviation [SD], 14.9 ? 0.6) (Fig. 3)
among antibodies from all immunization groups. Anti-HPA
antibodies demonstrated lower reactivity when employed as
the capture antibodies (mean EI value ? SD, 11.3 ? 0.6; data
not shown in Fig. 3) than those from the yeast-i.v. immuniza-
tion group. Antibodies obtained from the C-Ag immunization
group demonstrated substantially lower reactivity than anti-
bodies from the yeast-i.v. immunization group (mean EI value
? SD, 6.4 ? 0.4) but displayed higher reactivity than antibod-
ies from all remaining immunization groups (mean EI value ?
SD for all remaining immunization groups combined, 2.4 ?
Limit of sensitivity to detect antigen. Fivefold serial dilu-
tions of purified C-Ag (range, 0.1 to 8,000 ng/ml) were tested
in the double-antibody sandwich EIA format using antibodies
produced by the yeast-i.v. immunization regimen. The results
are shown in Fig. 4. Using one EI unit as the cutoff value for a
positive reaction, the assay could detect between 0.1 and 0.5
ng/ml of C-Ag.
Evaluating the double-antibody sandwich EIA using urine
from patients with histoplasmosis. To ensure that the C-Ag
reactivity of the antibodies produced by each of the various
immunization regimens correlated with their capacity to detect
antigenuria in patients, antibodies from each of the immuni-
zation groups were used in a double-antibody sandwich EIA to
detect antigenuria in patients with confirmed histoplasmosis
(Fig. 5). Antibodies produced using the yeast-i.v. protocol de-
tected antigen in the urine of 12 of 12 patients (100%) with
histoplasmosis as did anti-HPA antibodies (Fig. 5, HPA). EIA
reactivity was higher using antibodies derived from the yeast-
i.v. immunization regimen (mean EI value ? SD, 2.32 ? 0.65;
n ? 12) than using anti-HPA antibodies (mean EI value ? SD,
1.74 ? 0.36; n ? 12; P ? 0.02). Nonetheless, antibodies from
both sources detected all cases of histoplasmosis (12 of 12
patients). Antibodies produced using any of the other immu-
nization regimens also detected antigenuria in histoplasmosis
patients to various degrees: 10 of 12 (83%, C-Ag regimen), 6 of
12 (50%, yeast-s.c. regimen), 3 of 12 (25%, culture filtrate
regimen), and 3 of 12 (25%, H. capsulatum antigen-positive
A direct correlation (r ? 0.97, P ? 0.001) was observed
between antibody reactivity to C-Ag and the detection of an-
tigenuria by enzyme immunoassay. For example, antibodies
produced using the yeast-i.v. immunization regimen demon-
strated the greatest mean reactivity to C-Ag (mean EI value ?
SD, 14.9 ? 0.6) and also demonstrated the greatest mean
reactivity to patient urine (mean EI value ? SD, 2.32 ? 0.6).
Mean reactivity to C-Ag was lowest for antibodies produced by
the culture filtrate, H. capsulatum antigen-positive urine, and
yeast-s.c. immunization regimens, and these antibodies gave
the lowest mean reactivity to patient urine.
Correlation of results between the HPA-EIA and the double-
antibody EIA using antibodies from each of five immunization
regimens for the detection of histoplasmosis antigenuria. The
correlation between results obtained for the HPA-EIA and
those for the double-antibody EIA to detect antigenuria in
histoplasmosis patients was strongest when antibodies from
intravenously immunized rabbits (yeast-i.v.) were employed
(Pearson’s r ? 0.85, P ? 0.001) (Fig. 6). These data indicate
that the double-antibody sandwich EIA using antibodies from
the yeast-i.v. immunization regimen correlates favorably with the
HPA-EIA for the detection of histoplasmosis antigenuria. The
correlation between the results obtained for the HPA-EIA and
the double-antibody sandwich EIA using antibodies from the
other immunization groups was less strong (Pearson’s r ranged
from 0.24 to 0.67; P ? 0.05 to P ? 0.02).
Of the major endemic mycoses, histoplasmosis is responsible
for the greatest number of hospitalizations in the United States
and has a mortality rate of 7.5% in this setting (9). In human
immunodeficiency virus (HIV)-infected persons, institution of
highly active antiretroviral therapy has been associated with a
FIG. 4. Limit of sensitivity for the detection of antigen. Antibodies
produced by the yeast-i.v. immunization regimen were used to deter-
mine the lower limit of assay sensitivity. Fivefold serial dilutions of
purified C-Ag (range, 0.1 to 8,000 ng/ml) were tested in the double-
antibody sandwich EIA format. The results represent the mean and
standard deviation of results from three independent experiments. The
horizontal line represents a positive cutoff value defined as one EI unit.
VOL. 14, 2007 REAGENT OPTIMIZATION FOR H. CAPSULATUM ANTIGEN EIA 705
reduced risk of developing the most severe, life-threatening
form of the disease (15). However, access to highly active
antiretroviral therapy is limited in many developing countries
in which histoplasmosis is endemic, and incidence rates have
been reported to be as high as 21% (6) and mortality rates to
be as high as 32% (10). Early detection and implementation of
appropriate therapy have been shown to reduce morbidity and
mortality (32). Therefore, a rapid, effective, and low-cost diag-
nostic test is needed in countries where both HIV and his-
toplasmosis are endemic. In order to produce a cost-effective
test, an improved method must be developed to efficiently
produce antibodies for the detection of H. capsulatum antigen.
The primary objective of the research reported here was to
compare and contrast the use of a variety of immunization
routes, schedules, and H. capsulatum-derived antigens to de-
termine the optimum method for the production of antibodies
useful for the detection of H. capsulatum antigenuria. Anti-
bodies were then compared to those produced for use in a
standard commercial HPA-EIA antigenuria test. We found
that rabbits immunized by the intravenous route with whole,
killed H. capsulatum yeast-phase cells produced the most con-
sistently reactive antibodies for the detection of chromato-
graphically purified C-Ag as well as for the detection of anti-
genuria in patients with histoplasmosis. Reactivity of these
antibodies with C-Ag and with patient urine gave results com-
parable to those obtained using antibodies produced for the
Intravenous immunization of rabbits with whole, killed
yeast-phase cells produced antisera with consistently high titers
in 100% of rabbits immunized (4 of 4 rabbits immunized gave
titers of ?1:512,000). This contrasts to methods used to pro-
duce antibodies for the HPA-EIA, where immunization of as
many as 40 rabbits was required to produce a single, high-
responding rabbit. Production of antibodies for the HPA-EIA
used antigen in incomplete Freund’s adjuvant as part of the
immunization regimen as well as intravenous injection of live
H. capsulatum yeast-phase cells (42). This immunization regi-
FIG. 5. Comparison of anti-HPA antibodies and antibodies from five different immunization regimens to detect antigen in the urine of patients
with histoplasmosis by double-antibody sandwich EIA. Immunization groups are as noted in the legend of Fig. 1. Antibodies produced using the
yeast-i.v. immunization regimen detected 12 of 12 histoplasmosis patients (100%) as did anti-HPA antibodies (HPA) in the double-antibody
sandwich EIA format (inset). Antibodies produced using any of the other immunization regimens also reacted with patient urines to various
degrees, although all were significantly less sensitive (P ? 0.02) than antibodies produced by the yeast-i.v. and C-Ag immunization regimens. A
direct correlation (r ? 0.97, P ? 0.001) was observed between antibody reactivity to C-Ag and the detection of antigenuria by double-antibody
sandwich EIA. The bars and corresponding standard deviations shown represent the mean and standard deviations of two replicate EIA index
measurements for urines from each of 12 histoplasmosis patients (patient numbers 1 to 12) and a healthy control (specimen number 13). Hc, H.
FIG. 6. Direct correlation between results from the HPA-EIA and
the double-antibody EIA to detect antigenuria in histoplasmosis pa-
tients. EI values obtained using the HPA-EIA test to detect antige-
nuria directly correlated with those from the double-antibody EIA test
when antibodies from the yeast-i.v. immunization regimen were em-
ployed (Pearson’s r ? 0.85, P ? 0.001).
706 LINDSLEY ET AL.CLIN. VACCINE IMMUNOL.
men was later modified so that intravenous immunizations
were no longer given; intramuscular and subcutaneous injec-
tions of antigen in Freund’s complete adjuvant, followed by
booster immunizations in Freund’s incomplete adjuvant, were
used. In contrast, rabbits producing antisera with consistently
high titers in our study were immunized by the intravenous
route, and no adjuvant was necessary to obtain peak antibody
titers. Indeed, although an adjuvant (TiterMax) was employed
for subcutaneous booster injections of the yeast-i.v. immuni-
zation group, antibody titers peaked in these rabbits before
booster injections were initiated. This lack of dependence on
adjuvant for the production of highly reactive antibodies is an
especially important feature of this method not only because it
saves time, labor, and expense but also because of animal
welfare considerations. Although there is some controversy
regarding whether judicious use of newer formulations of
Freund’s adjuvant may be acceptable for the augmentation of
polyclonal antibody production in rabbits (18, 28), historical
evidence indicates that deleterious side effects such as focal
necrosis and ulceration of the skin can occur in a dose-depen-
dent manner (5). Finally, the amount of time required to ob-
tain peak antibody titers using the intravenous immunization
method was only 32 days, and titers remained at peak levels for
at least an additional 32 days. In contrast, rabbits were not
reported to have reached peak antibody titers until 104 or
more days after initial immunization in the HPA-EIA antibody
production regimen (although no data were presented regard-
ing test bleeds or antibody titers at earlier time points).
A second objective of the research reported here was to
better understand the nature of the antigen being detected by
anti-HPA antibodies and by antibodies raised by our own im-
munization regimens. The intravenous immunization method
used whole, formalin-killed H. capsulatum yeast-phase cells as
the immunogen. This was the same immunogen employed to
produce antibodies for use in the HPA-EIA test (29, 42). The
rationale for using C-Ag to screen rabbit antisera for use in our
test was based on the following considerations. One of the
limitations of the HPA-EIA test has been its detection of a
largely unknown and poorly characterized antigen that has not
been purified to homogeneity. Studies to date have indicated
that this antigen is primarily carbohydrate in nature, based on
its stability during boiling, resistance to degradation by various
proteases and nucleases, destruction by mixed glycosidases or
periodate treatment, and removal by binding to concanavalin
A (42). The HPA-EIA has also characteristically demonstrated
cross-reactivity when specimens from patients with diseases
caused by fungi known to share common cell wall carbohydrate
antigens (e.g., blastomycosis and paracoccidioidomycosis) (2,
14, 35) are tested. It was therefore postulated that the antigen
detected in the HPA-EIA test may be, at least in part, the
C-Ag of H. capsulatum. The true identity of the H. capsulatum
polysaccharide antigen may never be determined, primarily
because it is found in such low concentrations in patient urine
(i.e., in nanogram-per-milliliter concentrations in urine) (ref-
erences 14 and 42a and the present study) that purification to
homogeneity would be difficult. Concentration of patient urine
before purification may be helpful, as a 20-fold concentration
of HPA-positive patient urine resulted in a total carbohydrate
content of 19.9 mg/ml (this study). However, it is uncertain
how much of the carbohydrate content measured in the con-
centrated HPA-positive urine was contributed by H. capsula-
tum antigen and how much came from dietary or other sources.
It would be interesting to determine if anti-HPA antibodies
might be used to affinity purify the antigen of interest from
patient urine. Nonetheless, obtaining sufficient amounts of
pure antigen for specific structural analyses would be challeng-
Results from the present study and from studies using
monoclonal antibodies directed against chromatographically
purified C-Ag (25; S. Das, L. Benjamin, M. D. Lindsley, B.
Samayoa, E. Arathoon, J. Morgan, and C. J. Morrison, unpub-
lished data) indicate that C-Ag may, at least in part, be respon-
sible for the observed reactivity of patient urine in the HPA-
EIA. In addition, the C-Ag may be partially responsible for test
cross-reactivity with specimens from patients living in areas
where fungal diseases other than histoplasmosis are endemic.
Indeed, it has been suggested that the carbohydrate antigen
common to H. capsulatum, Paracoccidioides brasiliensis, and
Blastomyces dermatitidis is a galactomannan and that this car-
bohydrate is responsible for test cross-reactivity (2, 11, 26, 33).
In any case, when tested in an indirect dot blot immunoassay,
anti-HPA antibodies gave positive reactions with purified
C-Ag in a dose-dependent manner.
Another advantage to the use of chromatographically puri-
fied C-Ag as a control antigen is the capacity to generate a
standard curve against which the amount of urinary antigen
could be measured and quantified. The standard HPA-EIA
urinary antigen test has long relied on the use of HPA-positive
patient urine as a positive control and normal urine as a neg-
ative control against which to measure test results (13, 39).
Obtaining sufficiently large quantities of uniformly positive and
negative control urine to use as a test standard over time can
be difficult. Storage of such samples so that they do not gain or
lose reactivity can be problematic. In contrast, chromatograph-
ically purified C-Ag can be produced in large quantities (44)
and can be standardized for carbohydrate content. Although a
preliminary report (42a) indicated that a standard curve has
now been developed for use in the HPA-EIA, the nature of the
antigen being detected has not been revealed for proprietary
reasons (patent pending). Now that we have established that
the antibodies produced by our yeast-i.v. immunization regi-
men are equivalent to those developed for the HPA-EIA for
detecting H. capsulatum antigenuria, future studies can employ
a C-Ag standard curve to better quantify the actual antigen
concentrations found in patient urine.
A second-generation HPA-EIA test has recently been de-
scribed (37). In this format, the HPA-EIA has been modified
to reduce cross-linking of capture and detector antibodies by
human anti-rabbit antibodies (HARA) which have been re-
ported to occur in 16% of patients receiving rabbit anti-thy-
mocyte globulin for the reduction of allograft rejection (38).
Sera containing HARA have been demonstrated to produce
false-positive HPA-EIA results (38). The nature of the HPA-
EIA modification to produce the second-generation test is
unknown (proprietary; patent pending) but may involve an
alteration of the detector antibody or incorporation of proteo-
lytic enzymes to reduce the activity of the HARA cross-linking
antibodies in serum. However, our test was designed to detect
H. capsulatum antigenuria rather than antigenemia, given re-
ports that HPA detection is more sensitive in urine than in
VOL. 14, 2007 REAGENT OPTIMIZATION FOR H. CAPSULATUM ANTIGEN EIA 707
serum (41); therefore, interference in our antigenuria test by
HARA would not be expected.
Concurrent with the present work, we have developed ani-
mal infection models of histoplasmosis, blastomycosis, penicil-
liosis marneffei, aspergillosis, cryptococcosis, and candidiasis
(9a) in order to obtain well-characterized urine specimens to
test in our EIA. We are also prospectively collecting urine
from HIV-seropositive patients with histoplasmosis, coccid-
ioidomycosis, cryptococcosis, and other respiratory infections
to fully evaluate our H. capsulatum antigenuria EIA. In addi-
tion, test parameters are being optimized (27a) and monoclo-
nal as well as polyclonal detector antibodies are being evalu-
ated for use in the detection of H. capsulatum antigenuria by
EIA (25; Das et al., unpublished).
Although it is expected that the antibodies produced in this
study will cross-react with urinary antigens produced during
other fungal infections, particularly those caused by B. derma-
titidis and P. brasiliensis, in a manner much like that reported
for the HPA-EIA test, our test should nonetheless be useful as
a screening tool to rule out respiratory infections from other
causes. The United States is not an area of endemicity for
paracoccidioidomycosis, and although blastomycosis has the
same areas of endemicity as histoplasmosis, it is a much rarer
disease (3, 9). In addition, both histoplasmosis and blastomy-
cosis can be successfully treated with similar antifungal drug
regimens (8, 34).
To our knowledge, we are the first to conduct a systematic
assessment of immunization schedules, routes, and antigens to
optimize the production of antibodies for use in the detection
of H. capsulatum polysaccharide antigenuria. As a result, we
developed an immunization regimen that is faster, easier, more
humane, and more consistently reliable than regimens em-
ployed previously to produce antibodies to detect HPA anti-
genuria. Antibodies produced by the yeast-i.v. immunization
regimen were equivalent to those produced for the standard
HPA-EIA test in detecting chromatographically purified C-Ag
as well as in detecting H. capsulatum antigenuria in histoplas-
mosis patients. Further studies are currently under way to
improve and optimize test parameters.
Finally, there is an urgent public health need for the avail-
ability of reagents that can be used for the detection of signif-
icant fungal diseases outside the United States, particularly in
resource-poor countries lacking the means to identify patients
with these infections. Outside the United States, histoplasmo-
sis cannot be readily discriminated from a variety of bacterial
and parasitic diseases, especially when observed in the dissem-
inated form in HIV-infected individuals in the regions of Cen-
tral and South America where it is endemic. A rapid urine
antigen test will be of great value in these countries so that
patients who require specific antifungal therapy can be readily
identified and treated. These studies will hopefully contribute
to the development and distribution of such materials in re-
source-poor countries throughout the world.
We thank L. Joseph Wheat for providing the antisera and urine that
were used in the evaluation of our reagents.
1. Ajello, L. 1964. Relationship of Histoplasma capsulatum to avian habitats.
Public Health Rep. 79:266–270.
2. Azuma, I., F. Kanetsuna, Y. Tanaka, Y. Yamamura, and L. M. Carbonell.
1974. Chemical and immunological properties of galactomannans obtained
from Histoplasma duboisii, Histoplasma capsulatum, Paracoccidioides brasil-
iensis and Blastomyces dermatitidis. Mycopathol. Mycol. Appl. 54:111–125.
3. Baumgardner, D. J., E. M. Knavel, D. Steber, and G. R. Swain. 2006.
Geographic distribution of human blastomycosis cases in Milwaukee, Wis-
consin, USA: association with urban watersheds. Mycopathologia 161:275–
4. Bradford, M. M. 1976. A rapid and sensitive method for the quantitation of
microgram quantities of protein utilizing the principle of protein-dye bind-
ing. Anal. Biochem. 72:248–254.
5. Broderson, J. R. 1989. A retrospective review of lesions associated with the
use of Freund’s adjuvant. Lab. Anim. Sci. 39:400–405.
6. Cahn, P., W. H. Belloso, J. Murillo, and G. Prada-Trujillo. 2000. AIDS in
Latin America. Infect. Dis. Clin. N. Am. 14:185–209.
7. Chamany, S., S. A. Mirza, J. W. Fleming, J. F. Howell, S. W. Lenhart, V. D.
Mortimer, M. A. Phelan, M. D. Lindsley, N. J. Iqbal, L. J. Wheat, M. E.
Brandt, D. W. Warnock, and R. A. Hajjeh. 2004. A large histoplasmosis
outbreak among high school students in Indiana, 2001. Pediatr. Infect. Dis.
8. Chapman, S. W., R. W. Bradsher, Jr., G. D. Campbell, Jr., P. G. Pappas, and
C. A. Kauffman. 2000. Practice guidelines for the management of patients
with blastomycosis. Clin. Infect. Dis. 30:679–683.
9. Chu, J. H., C. Feudtner, K. Heydon, T. J. Walsh, and T. E. Zaoutis. 2006.
Hospitalizations for endemic mycoses: a population-based national study.
Clin. Infect. Dis. 42:822–825.
9a. Das, S., M. D. Lindsley, and C. J. Morrison. 2005. Abstr. 105th Gen. Meet.
Am. Soc. Microbiol., abstr. F-016.
10. de Francesco Daher, E., F. A. de Sousa Barros, G. B. da Silva, Jr., C. F.
Takeda, R. M. Mota, M. T. Ferreira, J. C. Martins, S. A. Oliveira, and O. A.
Gutierrez-Adrianzen. 2006. Risk factors for death in acquired immunodefi-
ciency syndrome-associated disseminated histoplasmosis. Am. J. Trop. Med.
11. Domer, J. E. 1971. Monosaccharide and chitin content of cell walls of
Histoplasma capsulatum and Blastomyces dermatitidis. J. Bacteriol. 107:870–
12. Dubois, M., H. Gillis, J. Hamilton, A. Rebers, and R. Smith. 1956. Colori-
metric method for the determination of sugars and related substances. Anal.
13. Durkin, M. M., P. A. Connolly, and L. J. Wheat. 1997. Comparison of
radioimmunoassay and enzyme-linked immunoassay methods for detection
of Histoplasma capsulatum var. capsulatum antigen. J. Clin. Microbiol. 35:
14. Garringer, T. O., L. J. Wheat, and E. J. Brizendine. 2000. Comparison of an
established antibody sandwich method with an inhibition method of His-
toplasma capsulatum antigen detection. J. Clin. Microbiol. 38:2909–2913.
15. Goldman, M., R. Zackin, C. J. Fichtenbaum, D. J. Skiest, S. L. Koletar, R.
Hafner, L. J. Wheat, P. M. Nyangweso, C. T. Yiannoutsos, C. T. Schnizlein-
Bick, S. Owens, J. A. Aberg, and AIDS Clinical Trials Group A5038 Study
Group. 2004. Safety of discontinuation of maintenance therapy for dissem-
inated histoplasmosis after immunologic response to antiretroviral therapy.
Clin. Infect. Dis. 38:1485–1489.
16. Hage, C. A., T. E. Davis, L. Egan, M. Parker, D. Fuller, A. M. Lemonte, M.
Durkin, P. Connelly, L. J. Wheat, D. Blue-Hnidy, and K. S. Knox. 2007.
Diagnosis of pulmonary histoplasmosis and blastomycosis by detection of
antigen in bronchoalveolar lavage fluid using an improved second-generation
enzyme-linked immunoassay. Respir. Med. 101:43–47.
17. Hajjeh, R. A., P. G. Pappas, H. Henderson, D. Lancaster, D. M. Bamberger,
K. J. Skahan, M. A. Phelan, G. Cloud, M. Holloway, C. A. Kauffman, L. J.
Wheat, and National Institute of Allergy and Infectious Diseases Mycoses
Study Group. 2001. Multicenter case-control study of risk factors for his-
toplasmosis in human immunodeficiency virus-infected persons. Clin. Infect.
18. Halliday, L. C., J. E. Artwohl, R. M. Bunte, V. Ramakrishnan, and B. T.
Bennett. 2004. Effects of Freund’s complete adjuvant on the physiology,
histology, and activity of New Zealand white rabbits. Contemp. Top. Lab.
Anim. Sci. 43:8–13.
19. Hebert, G. A. 1976. Improved salt fractionation of animal serums for immu-
nofluorescence studies. J. Dent. Res. 55:A33–37.
20. Kauffman, C. A. 2001. Pulmonary histoplasmosis. Curr. Infect. Dis. Rep.
21. Lindsley, M. D., D. W. Warnock, and C. J. Morrison. 2006. Serological and
molecular diagnosis of fungal infections, p. 569–605. In B. Detrick, R. G.
Hamilton, and J. D. Folds (ed.), Manual of molecular and clinical laboratory
immunology, 7th ed. ASM Press, Washington, DC.
22. Lyon, G. M., A. V. Bravo, A. Espino, M. D. Lindsley, R. E. Gutierrez, I.
Rodriguez, A. Corella, F. Carrillo, M. M. McNeil, D. W. Warnock, and R. A.
Hajjeh. 2004. Histoplasmosis associated with exploring a bat-inhabited cave
in Costa Rica, 1998–1999. Am. J. Trop. Med. Hyg. 70:438–442.
23. Pine, L. 1957. Studies of the growth of Histoplasma capsulatum. III. Effects
of thiamin and other vitamins on the growth of the yeast and mycelial phases
of Histoplasma capsulatum. J. Bacteriol. 74:239–245.
708LINDSLEY ET AL.CLIN. VACCINE IMMUNOL.
24. Reiss, E., L. Kaufman, J. A. Kovacs, and M. D. Lindsley. 2002. Clinical
immunomycology, p. 559–583. In N. R. Rose, R. G. Hamilton, and B.
Detrick (ed.), Manual of clinical laboratory immunology, 6th ed. ASM Press,
25. Reiss, E., J. B. Knowles, S. L. Bragg, and L. Kaufman. 1986. Monoclonal
antibodies against the M-protein and carbohydrate antigens of histoplasmin
characterized by the enzyme-linked immunoelectrotransfer blot method.
Infect. Immun. 53:540–546.
26. Reiss, E., W. O. Mitchell, S. H. Stone, and H. F. Hasenclever. 1974. Cellular
immune activity of a galactomannan-protein complex from mycelia of His-
toplasma capsulatum. Infect. Immun. 10:802–809.
27. Sathapatayavongs, B., B. E. Batteiger, J. Wheat, T. G. Slama, and J. L. Wass.
1983. Clinical and laboratory features of disseminated histoplasmosis during
two large urban outbreaks. Medicine (Baltimore) 62:263–270.
27a.Scheel, C., L. Benjamin, J. Morgan, S. Hurst, B. Samayoa, M. Lindsley, B.
Arthington-Skaggs, and E. Arathoon. 2006. Abstr. 16th Congr. Int. Soc.
Hum. Anim. Mycol., abstr. P-0422.
28. Stills, H. F., Jr. 2005. Adjuvants and antibody production: dispelling the
myths associated with Freund’s complete and other adjuvants. ILAR J.
29. Tewari, R. P., D. K. Sharma, and A. Mathur. 1978. Significance of thymus-
derived lymphocytes in immunity elicited by immunization with ribosomes or
live yeast cells of Histoplasma capsulatum. J. Infect. Dis. 138:605–613.
30. Tobon, A. M., C. A. Agudelo, D. S. Rosero, J. E. Ochoa, C. De Bedout, A.
Zuluaga, M. Arango, L. E. Cano, J. Sampedro, and A. Restrepo. 2005.
Disseminated histoplasmosis: a comparative study between patients with
acquired immunodeficiency syndrome and non-human immunodeficiency
virus-infected individuals. Am. J. Trop. Med. Hyg. 73:576–582.
31. Wheat, J. 1996. Histoplasmosis in the acquired immunodeficiency syndrome.
Curr. Top. Med. Mycol. 7:7–18.
32. Wheat, J. 1997. Histoplasmosis. Experience during outbreaks in Indianapolis
and review of the literature. Medicine (Baltimore) 76:339–354.
33. Wheat, J., M. L. French, S. Kamel, and R. P. Tewari. 1986. Evaluation of
cross-reactions in Histoplasma capsulatum serologic tests. J. Clin. Microbiol.
34. Wheat, J., G. Sarosi, D. McKinsey, R. Hamill, R. Bradsher, P. Johnson, J.
Loyd, and C. Kauffman. 2000. Practice guidelines for the management of
patients with histoplasmosis. Clin. Infect. Dis. 30:688–695.
35. Wheat, J., H. Wheat, P. Connolly, M. Kleiman, K. Supparatpinyo, K. Nelson,
R. Bradsher, and A. Restrepo. 1997. Cross-reactivity in Histoplasma capsu-
latum variety capsulatum antigen assays of urine samples from patients with
endemic mycoses. Clin. Infect. Dis. 24:1169–1171.
36. Wheat, L. J. 2003. Current diagnosis of histoplasmosis. Trends Microbiol.
37. Wheat, L. J., P. Connolly, M. Durkin, B. K. Book, and M. D. Pescovitz. 2006.
Elimination of false-positive Histoplasma antigenemia caused by human anti-
rabbit antibodies in the second-generation Histoplasma antigen assay.
Transpl. Infect. Dis. 8:219–221.
38. Wheat, L. J., P. Connolly, M. Durkin, B. K. Book, A. J. Tector, J. Fridell, and
M. D. Pescovitz. 2004. False-positive Histoplasma antigenemia caused by
antithymocyte globulin antibodies. Transpl. Infect. Dis. 6:23–27.
39. Wheat, L. J., P. Connolly-Stringfield, R. B. Kohler, P. T. Frame, and M. R.
Gupta. 1989. Histoplasma capsulatum polysaccharide antigen detection in
diagnosis and management of disseminated histoplasmosis in patients with
acquired immunodeficiency syndrome. Am. J. Med. 87:396–400.
40. Wheat, L. J., P. A. Connolly-Stringfield, R. L. Baker, M. F. Curfman, M. E.
Eads, K. S. Israel, S. A. Norris, D. H. Webb, and M. L. Zeckel. 1990.
Disseminated histoplasmosis in the acquired immune deficiency syndrome:
clinical findings, diagnosis and treatment, and review of the literature. Med-
icine (Baltimore) 69:361–374.
41. Wheat, L. J., T. Garringer, E. Brizendine, and P. Connolly. 2002. Diagnosis
of histoplasmosis by antigen detection based upon experience at the his-
toplasmosis reference laboratory. Diagn. Microbiol. Infect. Dis. 43:29–37.
42. 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.
42a.Wheat, L. J., E. J. Hackett, and P. A. Connolly. 2006. Abstr. 46th Intersci.
Conf. Antimicrob. Agents Chemother., abstr. M-1611.
43. Williams, B., M. Fojtasek, P. Connolly-Stringfield, and J. Wheat. 1994.
Diagnosis of histoplasmosis by antigen detection during an outbreak in
Indianapolis, Ind. Arch. Pathol. Lab. Med. 118:1205–1208.
44. Zancope-Oliveira, R. M., S. L. Bragg, S. F. Hurst, J. M. Peralta, and E.
Reiss. 1993. Evaluation of cation exchange chromatography for the isolation
of M glycoprotein from histoplasmin. J. Med. Vet. Mycol. 31:29–41.
45. Zancope-Oliveira, R. M., S. L. Bragg, E. Reiss, and J. M. Peralta. 1994.
Immunochemical analysis of the H and M glycoproteins from Histoplasma
capsulatum. Clin. Diagn. Lab. Immunol. 1:563–568.
VOL. 14, 2007REAGENT OPTIMIZATION FOR H. CAPSULATUM ANTIGEN EIA 709