Abstract

Blastomyces dermatitidis is a dimorphic fungus that invades the lungs via inhalation of conidia.  The acute pulmonary disease blastomycosis results after a temperature-induced conversion to a yeast form.  Although much about the factors that contribute to the pathogenicity of B. dermatitidis strains remains unknown, the presence of the cell surface antigens BAD1 and α-(1,3)-glucan have been implicated as possible virulence factors in previous research.  The purpose of this study was to utilize monoclonal antibodies to detect BAD1 and α-(1,3)-glucan in yeast cell lysates produced from six different B. dermatitidis strains using an indirect enzyme-linked immunosorbent assay.  Each lysate’s ability to bind anti-B. dermatitidis polyclonal rabbit serum was also measured.  The six B. dermatitidis isolates were capable of  binding all three antibody preparations and exhibited mean residual absorbance values ranging from 1.301-1.802 for the detection of BAD1 and 0.031-0.199 for the detection of α-(1,3)-glucan.  Antigen detection utilizing rabbit serum yielded mean residual absorbance values ranging from 0.847-1.470.  The data from this study provide insight into the use of BAD1 and α-(1,3)-glucan detection as diagnostic tools for establishing the presence of B. dermatitidis.

Introduction

Blastomyces dermatitidis is a dimorphic fungus that causes the potentially serious disease blastomycosis.  This disease is endemic to North America and cases of blastomycosis have been reported in both humans and canines (1, 6). Blastomycosis typically presents itself as a mild, chronic respiratory infection.  B. dermatitidis invades the lungs via inhalation of the organism’s conidia, which often leads to a pulmonary infection (5, 16).  If the disease is left untreated, the infection can disseminate to other parts of the body, resulting in cutaneous lesions, bone and joint infections, additional organ complications, or death (6). 

Two cell surface molecules of B. dermatitidis, α-(1,3)-glucan and the protein Blastomyces adhesin 1 (BAD1), have been implicated as factors associated with strain virulence in previous studies by Klein et al. (4, 12, 13). Data suggest that the establishment of a successful infection may be correlated with the amounts of BAD1 and α-(1,3)-glucan present on the surface of B. dermatitidis yeast cells.  The compound α-(1,3)-glucan is a cell surface carbohydrate that has been cited as a virulence factor or virulence-associated factor in the dimorphic fungal pathogens Histoplasma capsulatum and Paracoccidioides brasiliensis and the opportunistic fungal pathogen Cryptococcus neoformans (8, 17, 20, 21, 22).  The carbohydrate is also a component of the yeast cell surface of B. dermatitidis, comprising 95% of total yeast phase glucan content (10).  It is thought that this molecule serves a role in masking BAD1 on the surface of B. dermatitidis yeast cells (2, 3, 7, 11, 13).  BAD1 (formerly WI-1) is a 120-kDa major surface protein adhesin that is an essential virulence factor in B. dermatitidis and a major target of humoral and cell-mediated immune responses (8, 15). Studies suggest that more virulent strains of B. dermatitidis exhibit less BAD1 on their yeast cell surface and/or shed BAD1 from their surfaces more readily in culture during in vitro growth (8, 11, 13).

The purpose of this study was to utilize specific monoclonal antibodies to evaluate the presence of BAD1 and α-(1,3)-glucan in six different B. dermatitidis yeast lysate antigens using an indirect enzyme-linked immunosorbent assay (ELISA, peroxidase system).  The six antigens were also tested with polyclonal preparations of anti-B. dermatitidis rabbit serum for comparison.  The objective of this project was to obtain insight into the identification of B. dermatitidis based on BAD1 and α-(1,3)-glucan detection and to detect variations in the abundance of these two compounds in different strains of B. dermatitidis.

Materials and Methods

Antigens

Yeast phase lysate antigens were prepared from six different strains of B. dermatitidis (T58: dog, Tennessee; T27: polar bear, Tennessee; ERC-2: dog, Wisconsin; SOIL: soil, Canada; B5896: human, Minnesota; 48938: bat, India). The yeast phase lysates were prepared using a method modified from the one previously described by Johnson and Scalarone for the preparation of yeast lysate reagents from H. capsulatum (9, 18, 19, 23).  Mycelial phase isolates were converted to yeast cells in brain heart infusion medium and grown in a chemically defined medium for 7 days at 37˚C with shaking.  Harvested cells were washed 5 times with sterile deionized water using centrifugation (5 min at 700 x g).  Cells were incubated in sterile deionized water for 7 days at 37°C with shaking to promote lysis.  The recovered suspension was centrifuged (30 min at 700 x g) to remove debris and sterilized via filtration through a 0.2μm Nalgene filter (Nalge Company, Rochester, NY). Merthiolate (1:10000) was added to each solution as a preservative. Protein concentrations of the preparations were determined using the BCA assay kit (Sigma Chemical Company, St. Louis, MO) and appropriate antigen dilutions were prepared based on these concentrations.  Lysed yeast cell preparations were stored at 4°C.

Indirect ELISA Procedure

Yeast phase lysate antigens were diluted to 100 ng/ml in a carbonate-bicarbonate buffer (pH 9.6) and placed into the wells of an Immunomaxi 96-well modified flat bottom high binding microdilution plate (100 μl/well; TPP, Switzerland).  Thirty-six samples of each antigen were run.  The plate was incubated in a humid chamber overnight at 4°C and washed three times with 0.15% Tween 20 in phosphate buffered saline (PBS-T; pH 7.4).

To detect any BAD1 antigen present in the yeast lysates, murine anti-BAD1 monoclonal antibody diluted 1:2000 in PBS-T was added to each well (100 μl/well; DD5CB4, provided by Bruce Klein).  Murine monoclonal antibody MOPC 104e (Sigma Chemical Company, St. Louis, MO) was diluted 1:100 in PBS-T and placed in the wells to detect any α-(1,3)-glucan present in the yeast lysates (100 μl/well). Murine ascites fluid was utilized with the same six yeast lysate antigens as controls for both anti-BAD1 (MOPC 21 clarified ascites; Sigma Chemical Company, St. Louis, MO) and MOPC 104e (MOPC 104e clarified ascites; Sigma Chemical Company, St. Louis, MO) monoclonal antibody binding.  General detection of B. dermatitidis-specific immunoglobulins was performed utilizing pooled sera collected from rabbits immunized with yeast lysate preparations of the six B. dermatitidis strains diluted 1:5000 in PBS-T (100 µl/well).  Sera from uninoculated rabbits diluted 1:5000 in PBS-T was utilized as a control (100 µl/well).  After the appropriate antibody preparation was added to each well, the microdilution plate was incubated in a humid chamber for 30 min at 37°C and washed as previously described.  A 1:2000 dilution of goat-anti mouse IgG conjugated antibody (peroxidase; Kirkegaard and Perry Laboratories, Gaithersburg, MD) was added to wells (100 μl/well) to detect bound anti-BAD1 monoclonal antibodies.  A 1:500 dilution of goat anti-mouse IgM conjugated antibody (peroxidase; Sigma Chemical Company, St. Louis, MO) was added to wells (100 μl/well) to detect α-(1,3)-glucan-bound MOPC 104e  monoclonal antibodies.  A 1:2000 dilution of goat anti-rabbit IgG conjugated antibody (peroxidase; Kirkegaard and Perry Laboratories, Gaithersburg, MD) was added to wells to detect bound ant­i-B. dermatitidis rabbit immunoglobulins (100 µl/well).  The microdilution plate was incubated in a humid chamber for 30 min at 37°C and washed as previously described.  Peroxidase substrate (1-Step Ultra TMB-ELISA, Pierce Chemical Company, Rockford, IL) was added to each well (100 μl/well).  Reactions were allowed to proceed at room temperature for 2 min and 2N H₂SO₄ was added to each well to stop the reactions (100 μl/well).  The absorbance value for each well at 450 nm was determined using a BIO-RAD model 2550 EIA reader (BIO-RAD Laboratories, Hercules, CA).

Statistical Analysis

The original statistical design of this study was a two-factor analysis of variance (ANOVA) with the two factors defined as the six different strains and the three types of antibody preparations.  The response variable was the difference in absorbance between each antibody preparation and its respective control (“residual absorbance”).  This residual absorbance was used as a response variable after initial analysis determined that different controls (i.e. ascites fluid or normal rabbit sera) elicited different background absorbance values.  After this analysis was performed, the residuals were examined for normality.  Strong positive skewness was found to exist in the residuals, so consequent analyses were performed nonparametrically.  The Kruskal-Wallis test was used to compare the three antibody preparations with each strain using a Bonferroni correction of the p-values to account for the six tests performed.  When a significant Kruskal-Wallis test was found, the appropriate pairwise comparisons were made using a Mann-Whitney U test with a Bonferroni correction.

Results

The range of reactivity for both monoclonal antibody preparations in the detection of either BAD1 or α-(1,3)-glucan varied, as did antigen detection using anti-B. dermatitidis rabbit serum (Fig. 1).  Statistical analysis was used to determine if the average difference in residual absorbance values for the three different types of antibody preparations different significantly from one another.  It was found that MOPC 104e absorbance values were significantly lower than anti-BAD1 monoclonal antibody and rabbit serum absorbance values for each strain (p<0.001).  Anti-BAD1 monoclonal antibody absorbance values differed significantly from rabbit serum absorbance values in the strains T58, ERC-2, SOIL, B5896, and 48938 (p<0.001).  This difference was not significant for the strain T27 (p>0.05).






Figure 1. Comparison of indirect ELISA absorbance values for the six B. dermatitidis lysate antigens used for the detection of the yeast cell surface antigens BAD1 and α-(1,3)-glucan.  Absorbance values obtained using anti-B. dermatitidis rabbit serum are also presented for comparison.  Bars represent mean residual absorbance values obtained after absorbance values for control serum or ascites fluid were subtracted.  Error bars represent two standard deviations.  Asterisks are used to indicate that the mean residual absorbance values for the anti-BAD1 monoclonal antibody are significantly higher than the other two antibody preparations in five of the six strains.

The absorbance data indicates that antigen B5896 exhibited the highest amounts of detectable BAD1 antigen (mean residual absorbance value: 1.802).  In contrast, antigen T27 elicited the lowest sensitivity for BAD1 detection (mean residual absorbance value: 1.301).  The yeast lysate antigen T27 was shown to exhibit the most detectable α-(1,3)-glucan based on absorbance values for MOPC 104e monoclonal antibody binding (mean residual absorbance value: 0.199).  The SOIL and B5896 antigens produced the lowest absorbance values for α-(1,3)-glucan detection (mean residual absorbance value: 0.031 for both).  The mean absorbance values obtained for each isolate using anti-B. dermatitidis rabbit serum also displayed variation.  The yeast lysate antigen B5896 exhibited the highest absorbance values for B. dermatitidis antibody detection (mean residual absorbance value: 1.470).  The yeast lysate antigen 48938 produced the lowest absorbance values for detection of these same antibodies (mean residual absorbance value: 0.847).     

Discussion

The indirect ELISA procedure established that the monoclonal antibodies employed for this project were capable of detecting levels of BAD1 and α-(1,3)-glucan in all six B. dermatitidis yeast lysate antigens.  Treatment with polyclonal anti-B. dermatitidis rabbit serum also yielded positive results.  Indirect ELISA absorbance values for all six lysates indicate that variations exist in the detection abilities possessed by the three antibody preparations.  The MOPC 104e monoclonal antibody, which was utilized to detect α-(1,3)-glucan in the lysates, consistently yielded mean residual absorbance values that were lower than those obtained with anti-BAD1 monoclonal antibody or rabbit serum (mean residual absorbance value range: 0.031-0.199).  Although MOPC 104e is routinely used in research to detect α-(1,3)-glucan, its antigen specificity is for α(1→3) glucose (7, 15, 24).  This cross-reaction yielded consistently low absorbance values for this experiment, indicating that using this antibody as a diagnostic tool may be a poor choice.   The anti-BAD1 monoclonal antibody yielded the highest mean residual absorbance values for all of the strains, with statistically significant differences observed in five of the six strains (mean residual absorbance value range: 1.301-1.802).  Given the three choices of antibody preparations, the anti-BAD1 monoclonal antibody is the best choice for use as a diagnostic tool.  Not only does this antibody preparation yield the highest absorbance values in an indirect ELISA, it is also designed to target the novel BAD1 protein (14).  This makes the antibody a good candidate as a sensitive and specific tool for B. dermatitidis detection.

Based on these results, there appears to be a relationship between the amounts of BAD1 and α-(1,3)-glucan detected in each strain.  Antigen T27 exhibited the lowest absorbance reading for detection of BAD1 while antigen B5896 exhibited the highest absorbance reading for BAD1 detection.  Further variations are observed in absorbance values for the four additional yeast lysates (Fig. 1).  The T58, 48938, and T27 antigens exhibited high absorbance readings for the detection of α-(1,3)-glucan.  Antigens ERC-2, SOIL, and B5896 exhibited lower absorbance readings.  These data suggest a possible correlation between the detection of BAD1 and α-(1,3)-glucan in various B. dermatitidis strains that warrants further investigation.

Acknowledgements

This project was made possible by NIH Grant # P20 RRO16454 from the INBRE Program of the National Center for Research Resources.  The Idaho State University Department of Biological Sciences and the ISU Undergraduate Research and Scholarship Committee also supported this research. Teri Peterson provided guidance on statistical analysis.  Mike Chester is acknowledged for his assistance with this project.  A.M. Legendre (University of Tennessee School of Veterinary Medicine, Knoxville, TN), D.J. Baumgardner (University of Wisconsin Medical School, Milwaukee, WI) and The Centers for Disease Control (CDC; Atlanta, GA) provided the B. dermatitidis isolates used in this study. Anti-BAD1 monoclonal antibodies were generously provided by Bruce Klein (University of Wisconsin School of Medicine, Madison, WI). 

References

1.       Bradsher, R.W., S.W. Chapman, and P.G. Pappas. 2003. Blastomycosis. Infect. Dis. Clin. N. Am. 17: 21-40.

2.       Brandhorst, T.T., P.J. Rooney, T.D. Sullivan, and B.S. Klein. 2002. Molecular genetic analysis of Blastomyces dermatitidis reveals new insights about pathogenic mechanisms. Int. J. Med. Microbiol. 292: 363-371.

3.       Brandhorst, T.T., P.L. Rooney, T.D. Sullivan, and B.S. Klein. 2002. Using new genetic tools to study the pathogenesis of Blastomyces dermatitidis. Trends Microbiol. 10: 25-30.

4.       Brandhorst, T.T., M. Wüthrich, T. Warner, and B. Klein. 1999. Targeted gene disruption reveals an adhesin indispensable for pathogenicity of Blastomyces dermatitidis. J. Exp. Med. 189: 1207-1216.

5.       Causey, W.A., and G.D. Campbell. 1992. Clinical aspects of blastomycosis, p. 165-188. In Y. Al-Doory and A.F. DiSalvo (ed.), Blastomycosis. Plenum Medical Book Co., New York, NY.

6.       Davies, S.F.,  and G.A. Sarosi. 1997. Epidemiological and clinical features of pulmonary blastomycosis. Semin. Respir. Infect. 12: 206-218.

7.       Hogan, L.H., and B.S. Klein. 1994. Altered expression of surface alpha-1,3-glucan in genetically related strains of Blastomyces dermatitidis that differ in virulence. Infect. Immun. 62: 3543-3546.

8.       Hogan, L.H., B.S. Klein, and S.M. Levitz. 1996. Virulence factors of medically important fungi. Clin. Microbiol. Rev. 9: 469-488.

9.       Johnson, S.M., and G.M. Scalarone. 1988. Preparation and ELISA evaluation of Blastomyces dermatitidis yeast phase lysate antigens. Diagn. Microbiol. Infect. Dis. 11: 81-86.

10.   Kanetsuna, F., and L.M. Carbonell. 1971. Cell wall composition of the yeastlike and mycelial forms of Blastomyces dermatitidis. J. Bacteriol. 106: 946-948.

11.   Klein, B. 2000. Molecular basis of pathogenicity in Blastomyces dermatitidis: The importance of adhesion. Curr. Opin. Microbiol. 3: 339-343.

12.   Klein, B.S. 1997. Role of cell surface molecules of Blastomyces dermatitidis in the pathogenesis and immunobiology of blastomycosis. Semin. Respir. Infect. 12: 198-205.

13.   Klein, B.S., L.H. Hogan, and S.L. Newman. 1997. Cell surface molecules of Blastomyces dermatitidis. ASM News. 63: 140-145.

14.   Klein, B.S., and J.M. Jones. 1990. Isolation, purification, and radiolabeling of a novel 120-kD surface protein on Blastomyces dermatitidis yeasts to detect antibody in infected patients. J. Clin. Invest. 85: 152-161.

15.   Klein B.S., and S.L. Newman. 1996. Role of cell-surface molecules of Blastomyces dermatitidis in host-pathogen interactions. Trends Microbiol. 4: 246-251.

16.   Klein, B.S., R.A. Squires, J.K. Lloyd, D.R. Ruge, and A.M. Legendre. 2000. Canine antibody response to Blastomyces dermatitidis WI-1 antigen. Am. J. Vet. Res. 61: 554-558.

17.   Klimpel, K.R., and W.E. Goldman. 1998. Cell walls from avirulent variants of Histoplasma capsulatum lack α-(1,3)-glucan. Infect. Immun. 56: 2997-3000.

18.   Levine, H.B., G.M. Scalarone, G.D. Campbell, P.C. Kelly, and S.D. Chaparas. 1979. Histoplasmin-CYL, a yeast phase reagent in skin test studies in humans. Am. Rev. Respir. Dis. 119: 629-636.

19.   Levine, H.B., G.M. Scalarone, and S.D. Chaparas.  1977. Preparation of fungal antigens and vaccines: Studies on Coccidioides immitis and Histoplasma capsulatum. Contrib. Microbiol. Immunol. 3: 106-125.

20.   Rappleye, C.A., J.T. Engle, and W.E. Goldman. 2004. RNA interference in Histoplasma capsulatum demonstrates a role for α-(1,3)-glucan in virulence. Mol. Microbiol. 53: 153-165.

21.   Reese, A.J., and T.L. Doering. 2003. Cell wall α-1,3-glucan is required to anchor the Cryptococcus neoformans capsule. Mol. Microbiol. 50: 1401-1409.

22.   San Blas, G., and F. San Blas. 1977. Paracoccidioides brasiliensis: Cell wall structure and virulence. A review. Mycopathologia. 62: 77-86.

23.   Scalarone, G.M., H.B. Levine, and S.D. Chaparas. 1978. Delayed hypersensitivity responses of experimental animals to histoplasmin from the yeast and mycelial phases of Histoplasma capsulatum. Infect. Immun. 21: 705-713.

24.   Sigma-Aldrich.com [homepage on the Internet]. St. Louis: Sigma-Aldrich Co.; c2005 [cited 12 July 2005]. Available from: <http://www.sigma-aldrich.com/>.