ABSTRACT
In this study we found that serum inhibitory activity against Blastomyces dermatitidis was principally mediated by albumin. This was confirmed in experiments using albumin from several mammalian species. Analbuminemic rat serum did not inhibit B. dermatitidis growth in vivo; however, the addition of albumin restored inhibitory activity. Inhibitory activity does not require albumin domain III and appears to involve binding of a low-molecular-weight yeast-derived growth factor.
Blastomyces dermatitidis is a thermally dimorphic fungal pathogen that infects humans, dogs, and other mammals (2, 4, 11, 26, 27). Blastomycosis occurs following inhalation of B. dermatitidis conidia, which change into a pathogenic yeast form in the lungs. The resulting pulmonary infection can range in severity from mild to acute and can lead to disseminated disease (3, 4, 11, 25). Relatively little is known about components of the innate immune system that limit the multiplication of the yeast form of B. dermatitidis during the early stages of infection.
Innate immunity is mediated by several mechanisms that serve as important first lines of host defense against microbial pathogens. Human serum contains several factors (i.e., transferrin, lactoferrin, lysozyme, and complement) that can inhibit or kill pathogenic microorganisms. The inhibitory effect of serum on most pathogenic fungi is dependent on the iron-binding activity of transferrin, which is reported to inhibit the growth of Candida albicans, Histoplasma capsulatum, Cryptococcus neoformans, and Penicillium marneffei in vitro (1, 19, 20, 27, 31-33, 35). In a previous study, we demonstrated that serum inhibits B. dermatitidis yeast form growth in vitro via a mechanism that is independent of transferrin (12). Other prominent serum factors, such as complement components and natural antibodies, do not inhibit B. dermatitidis yeast form growth in vitro (13). In the present study, we sought to elucidate the identity of the serum factor that mediates inhibitory activity against B. dermatitidis.
Preliminary experiments indicated that protein fractions smaller than 100 kDa possessed inhibitory activity against B. dermatitidis (data not shown). The most abundant protein present in these fractions was albumin. Although it seemed unlikely, we assessed whether albumin could exert inhibitory activity against B. dermatitidis. To do so, B. dermatitidis yeast cells were suspended in RPMI 1640 medium supplemented with fraction V bovine serum albumin (BSA; at 1 mg/ml, which is equivalent to the albumin concentration of 5% fetal bovine serum [FBS]), and viable yeast cells were quantified as described previously (12). To our surprise, we found that BSA inhibited the growth of B. dermatitidis yeast cells to an extent similar to that of 5% FBS (Fig. 1A). This activity was not restricted to BSA since human (Fig. 1B), mouse (Fig. 1C), and canine (Fig. 1D) fraction V serum albumin proteins also significantly (P < 0.01) inhibited B. dermatitidis growth.
These unexpected results prompted us to explore the novel possibility that albumin could exert fungistatic activity against B. dermatitidis. The most compelling evidence was provided through the use of serum from analbuminemic Nagase rats. Nagase rats grow at a normal rate and have normal reproductive capabilities but have a mutation in their albumin gene that prevents the production of albumin (22, 31). We found that analbuminemic Nagase rat serum (added to RPMI 1640 medium [5%, vol/vol]) lacks inhibitory activity against B. dermatitidis yeast cells (Fig. 2). Furthermore, the addition of fraction V rat serum albumin (at 1.28 mg/ml, which is equivalent to the albumin concentration present in 5% [vol/vol] rat serum) to analbuminemic rat serum restored fungistatic activity (P < 0.01). These results demonstrate a relationship between the presence of albumin in serum and serum inhibitory activity against B. dermatitidis.
We next sought to exclude the possibility that the inhibitory activity was due to impurities present in fraction V albumin preparations. To address this concern, RPMI 1640 tissue culture medium was supplemented with immunoaffinity chromatography-purified BSA (1 mg/ml; Sigma), fraction V BSA (1 mg/ml: Sigma), FBS (5%, vol/vol), fatty acid-free human serum albumin (HSA) (2.275 mg/ml; Sigma), fraction V HSA (2.275 mg/ml; Sigma), or human serum (5%, vol/vol), and yeast growth was quantified. We found that all of these reagents exerted equivalent inhibitory activities against B. dermatitidis, suggesting that albumin inhibitory activity was mediated by albumin and not by protein or lipid impurities (data not shown).
Albumin reversibly binds a variety of diverse ligands. These include amino acids, divalent cations, fatty acids, lipoteichoic acid (Streptococcus pyogenes), several streptococcal proteins (M12 protein, protein G, and protein H), the N-terminal peptide of gp41 of human immunodeficiency virus type 1, and surface components of hepatitis B virus and echovirus 7 (5, 8, 10, 14-16, 21, 24, 28, 29, 34). Furthermore, these interactions can interfere with host-pathogen interactions. BSA can inhibit the adsorption of respiratory syncytial virus to Madin-Darby bovine kidney cells (9), and HSA can inhibit the uncoating of echovirus 7 and impair its ability to infect rhabdomyosarcoma cells (34).
We postulated that albumin inhibitory activity might be mediated either by direct interaction with B. dermatitidis or by binding a growth factor secreted by B. dermatitidis. To discern between these two potential mechanisms, we utilized dialysis membranes to physically separate albumin from B. dermatitidis. Immunoaffinity chromatography-purified BSA was placed within a dialysis membrane (50,000-molecular-weight cutoff) that was then placed into a flask containing RPMI 1640 medium and B. dermatitidis yeast cells. The dialysis membrane prevented albumin from escaping into the medium in which the yeast cells resided (verified by Western blotting) but allowed passage of small molecules through the membrane, where they could interact with BSA. We found similar fungistatic activities against B. dermatitidis yeast cells whether the BSA was contained within a dialysis membrane or added directly to the RPMI 1640 medium in which the yeast cells were suspended (Fig. 3A). Growth inhibition was confirmed by measuring optical density (data not shown), as described previously (12). One would predict that coincubation of albumin with B. dermatitidis would result in the formation of an albumin-yeast factor complex that would migrate slower than native albumin on a nondenaturing polyacrylamide gel electrophoresis (PAGE) gel. To assess this possibility, BSA was incubated in RPMI 1640 medium with B. dermatitidis yeast cells or in RPMI 1640 medium alone for 72 h at 37°C. The media were collected and centrifuged (2,000 × g), and the supernatants were filtered through a 0.2-μm-pore-size syringe filter and visualized on a discontinuous nondenaturing gel. We found that BSA incubated with yeast cells did not migrate as far as native BSA (Fig. 3B). In addition, albumin collected from the dialysis tubes used to generate the results presented in Fig. 3A gave similar results (data not shown). These results suggest that albumin can bind a growth factor secreted by B. dermatitidis but do not rule out the possibility that a direct interaction between albumin and B. dermatitidis might also occur.
We also investigated the possibility that albumin might exert its effects by sequestering essential nutrients from the medium. To test this possibility, BSA was placed within a dialysis membrane (50,000-molecular-weight cutoff) and dialyzed against RPMI 1640 tissue culture medium for 48 h at 37°C. We found that the BSA dialysate supported the growth of B. dermatitidis yeast cells as well as fresh tissue culture medium did (data not shown), which suggests that albumin inhibitory activity is not due to depletion of an essential nutrient from the RPM1 1640 medium.
The crystal structure of albumin has been resolved into three homologous structural domains (domains I, II, and III), which contain two well-described drug binding sites (5, 7, 16, 29, 30). Drug binding site 1 is located in domains I and II, and drug binding site 2 is located in domain III (Fig. 4A) (5, 16). To identify the structural features required for inhibitory activity, we assessed the ability of recombinant HSA (rHSA) proteins that lacked either domains I and II (drug binding site 1) or domain III (drug binding site 2) to inhibit B. dermatitidis growth (6). Because the availability of the rHSA proteins was limited, all serum albumin proteins were added to RPMI 1640 tissue culture medium at molar concentrations equivalent to that of albumin in 2.5% (vol/vol) human serum. We found that rHSA domains I and II, but not rHSA domain III, exerted inhibitory activity against B. dermatitidis (Fig. 4B). These results suggest that drug binding site 1, or other portions of domains I and II, are required for inhibitory activity.
The results of this study provide evidence that serum inhibitory activity against B. dermatitidis is mediated by a novel mechanism in which albumin binds a small yeast-derived factor that appears to be involved in regulating yeast growth in vitro. There are other examples of factors secreted by pathogenic fungi that are involved in growth regulation and pathogenesis. For example, B. dermatitidis, C. albicans, Sporothrix schenckii, H. capsulatum, and Aspergillus niger produce siderophores that facilitate the acquisition and assimilation of iron, which is an essential nutrient (18). The dimorphic pathogen H. capsulatum, whose teleomorph stage is in the same genus (Ajellomyces) as B. dermatitidis, secretes a calcium binding protein (CBP1) that is important for the intracellular parasitism of macrophages (26). H. capsulatum mutants lacking CBP1 cannot kill macrophages in vitro or cause pulmonary disease in a mouse model of histoplasmosis (26). C. albicans regulates its growth via the secretion of a quorum-sensing molecule (farnesol) that facilitates yeast-to-mycelium transformation (17). Recently, Noverr et al. reported that several fungal pathogens, including B. dermatitidis yeast cells, produce eicosanoids (i.e., prostaglandins and leukotrienes) and that production of eicosanoids appears to be essential for the growth of these fungi (23). Treatment of C. albicans or C. neoformans yeast cells with indomethacin, which is a cyclooxygenase inhibitor, inhibited the growth of these fungi by 95 and 99%, respectively (23). Although the nature of the B. dermatitidis factor is not known at this time, and its isolation was beyond the scope of the present study, we have found in preliminary experiments that the addition of yeast culture supernatants (10%, vol/vol) promoted more rapid growth of B. dermatitidis yeast cells in fresh medium and that the growth-promoting activity was retained within a 3,000-molecular-weight-cutoff Centricon tube (data not shown). These observations are consistent with the proposed binding of a low-molecular-weight growth factor by albumin. Although we cannot yet provide direct evidence that albumin contributes to host defense against B. dermatitidis in vivo, the ubiquitous distribution of albumin throughout the bloodstream and extravascular fluids raises intriguing possibilities regarding its potential role in innate immunity against infection by B. dermatitidis. Future studies will focus on the identification and characterization of the B. dermatitidis factor.
Fraction V albumin from several animal species inhibits the growth of the B. dermatitidis yeast form. B. dermatitidis yeast cells (105) were suspended in RPMI 1640 medium supplemented with the following: fraction V BSA (1 mg/ml) or whole serum (FBS; 5%, vol/vol) (A), fraction V HSA (2.28 mg/ml) or whole serum (HS; 5%, vol/vol) (B), fraction V murine serum albumin (MSA; 1.5 mg/ml) or whole serum (MS: 5%, vol/vol) (C), or fraction V canine serum albumin (CSA; 1.48 mg/ml) or whole serum (CS; 5%, vol/vol) (D). B. dermatitidis yeast cells (105) suspended in RPMI 1640 medium alone served as a control in all four panels. The concentrations of the various fraction V albumins were chosen because they are equivalent to the estimated albumin concentrations in the respective sera. The data illustrated are the means ± standard errors of the means of results of three separate experiments (asterisks indicate results significantly different from those of the medium control at a P value of <0.001).
Analbuminemic Nagase rat serum does not inhibit the growth of the B. dermatitidis yeast form. B. dermatitidis yeast cells (105) were grown in RPMI 1640 medium supplemented with analbuminemic Nagase rat serum (ARS; 5%, vol/vol), fraction V rat serum albumin (RSA; 1.28 mg/ml), normal rat serum (RS; 5%, vol/vol), or analbuminemic Nagase rat serum supplemented with fraction V rat albumin (ARS + RSA). Analbuminemic Nagase rat serum had no inhibitory effect on the growth of the B. dermatitidis yeast form. However, rat serum (5%, vol/vol), rat serum albumin, and analbuminemic Nagase rat serum supplemented with rat serum albumin significantly inhibited the growth of B. dermatitidis yeast cells at 48 h (P < 0.001; indicated by asterisks) relative to the growth of B. dermatitidis in RPMI 1640 medium alone (control). The data illustrated are the means ± standard errors of the means of the results of four separate experiments.
Albumin inhibitory activity is mediated by a mechanism that does not require direct contact between the B. dermatitidis yeast form and albumin. (A) B. dermatitidis yeast cells (107) were grown in RPMI 1640 medium (40 ml) in which a dialysis membrane containing immunoaffinity chromatography-purified BSA (BSA in dialysis tubing) was suspended. Controls included B. dermatitidis grown in RPMI 1640 medium containing a control dialysis membrane (no BSA) (dialysis tubing control), RPMI 1640 medium containing immunoaffinity chromatography-purified BSA, or RPMI 1640 medium alone (medium). Immunoaffinity chromatography-purified BSA that was contained within the dialysis membrane inhibited the growth of B. dermatitidis to an extent comparable to that of BSA added directly to the medium (results of both were significantly different from those of the medium control at a P of <0.01; indicated by asterisks). The data illustrated are the means ± standard errors of the means of results of three separate experiments. (B) Nondenaturing PAGE analysis was performed with culture supernatants from flasks that contained medium plus BSA and B. dermatitidis (lanes 1 to 3) or medium plus BSA without B. dermatitidis (lanes 4 to 6). The BSA incubated with B. dermatitidis did not migrate as far on a 10% PAGE gel as did BSA incubated with medium alone. Results are for one representative experiment of three performed.
Albumin inhibitory activity does not require the drug binding site 2 on domain III. (A) Diagram of the structure of serum albumin indicating the three structural domains and the two known drug binding sites. (B) B. dermatitidis yeast cells (105) were grown in RPMI 1640 medium supplemented with fraction V HSA (HSA), whole rHSA, rHSA domains I and II (contains drug binding site 1), or rHSA domain III (contains drug binding site 2). rHSA proteins and fraction V HSA were added at a final concentration of 1 mg/ml, which is equivalent to the molar concentrations of albumin in 2.5% human serum. B. dermatitidis yeast cells (105) were grown in RPMI 1640 medium alone as a control (medium). B. dermatitidis growth was significantly inhibited by fraction V HSA, rHSA, and rHSA domains I and II at 48 h (P < 0.01; indicated by asterisks). The data illustrated are the means ± standard deviations of results from one representative experiment of three that were performed.
ACKNOWLEDGMENTS
This study was supported by the Barbara Rettgen Blastomycosis Fund, the University of Wisconsin—Madison School of Veterinary Medicine, and a Robert D. Watkins fellowship from the American Society for Microbiology (to S.S.G.).
We thank Lauren Trepanier and Jackie Miller for providing human serum and plasma and Ronald Schultz for providing canine serum (University of Wisconsin—Madison). Serum from analbuminemic Nagase rats was generously provided by George Kaysen (University of California, Davis), and serum from healthy Sprague-Dawley rats was provided by Noah Zimmerman (University of Wisconsin—Madison). We are especially grateful to F. Rucker (Vienna, Austria) for generously providing the recombinant HSA fragments.
FOOTNOTES
- Received 23 June 2003.
- Returned for modification 9 July 2003.
- Accepted 11 August 2003.
- Copyright © 2003 American Society for Microbiology
