Characterization of a Unique Corticosterone-binding Protein in Candida albicans*

This paper further characterizes a protein we have demonstrated in Candida albicans which has the ability to bind corticosterone and related steroid hormones. Fungal cells are disrupted and cytosol is incubated with [3H]corticosterone for 3 h at which time peak steady state binding is achieved. Bound hormone is separated from free using Sephadex G-50 minicolumns or dex-tran-coated charcoal. Binding was found to be a linear function of protein concentration. The bound hormone co-migrates with authentic corticosterone in thin layer chromatographic systems indicating no metabolism of the radioprobe. Scatchard analysis of the binding in the pseudohyphal form of C. albicans yielded values of 6.3 n~ for the Kd and a binding capacity of about 650 fmol/ mg of cytosol protein; both determinations are comparable to our findings in the yeast form of this orga- nism. A series of sterols were tested for their ability to displace [3H]corticosterone from the yeast binder, and the results show that the binder is remarkably selective and stereo specific. Physical-chemical studies show the binder to be degraded at high temperatures and that binding is destroyed by trypsin and sulfhydryl block- ers. The protein sediments at 4 S on sucrose gradients and does not exhibit ionic dependent aggregation. The molecular weight is estimated to be -43,000 daltons by gel chromatography. We hypothesize that this intra- cellular protein may represent a primitive form of either the mammalian glucocorticoid receptor or the plasma corticosteroid-binding globulin.

This paper further characterizes a protein we have demonstrated in Candida albicans which has the ability to bind corticosterone and related steroid hormones. Fungal cells are disrupted and cytosol is incubated with [3H]corticosterone for 3 h at which time peak steady state binding is achieved. Bound hormone is separated from free using Sephadex G-50 minicolumns or dextran-coated charcoal. Binding was found to be a linear function of protein concentration. The bound hormone co-migrates with authentic corticosterone in thin layer chromatographic systems indicating no metabolism of the radioprobe. Scatchard analysis of the binding in the pseudohyphal form of C. albicans yielded values of 6.3 n~ for the Kd and a binding capacity of about 650 fmol/ mg of cytosol protein; both determinations are comparable to our findings in the yeast form of this organism. A series of sterols were tested for their ability to displace [3H]corticosterone from the yeast binder, and the results show that the binder is remarkably selective and stereo specific. Physical-chemical studies show the binder to be degraded at high temperatures and that binding is destroyed by trypsin and sulfhydryl blockers. The protein sediments at 4 S on sucrose gradients and does not exhibit ionic dependent aggregation. The molecular weight is estimated to be -43,000 daltons by gel chromatography. We hypothesize that this intracellular protein may represent a primitive form of either the mammalian glucocorticoid receptor or the plasma corticosteroid-binding globulin.
We have recently demonstrated the presence of a corticosterone-binding protein in the eukaryotic fungus Candida albicans and hypothesized that it may represent a steroid hormone receptor (1). It is not yet clear that this unicellular yeast actually utilizes hormones in the same sense that more complex multicellular organisms do. However, molecules functioning in a fashion analogous to mammalian hormones have been demonstrated in several fungi. Gametogenesis in the yeast Saccharomyces cereuisiae is under the control of two mating-type specific peptides (2). In addition, Fusarium, a more complex fungus, has been shown to produce estrogeniclike molecules (3) which have the ability to bind to the mammalian estrogen receptor (4). Also, evidence has been * This work was supported by a grant (GM 28825) from the National Institutes of Health. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "aduertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. +This work was submitted as partial fulfillment of the requirements for a doctoral degree. adduced that the watermold Achyla bisexualis secretes sterols which regulate its reproductive processes in a manner similar to vertebrate hormones (5). However, to our knowledge a hormone receptor system has not previously been described in unicellular eukaryotes. In this paper we provide further biochemical characterization of the corticosteronebinding protein we previously identified in C. albicans (1). The protein exhibits many properties compatible with it being a receptor molecule; however, it is not surprising to find that there are major differences compared to mammalian glucocorticoid receptors. C. albicans also produces a lipid-extractable product which we hypothesize is the endogenous ligand for this binding protein and which will be described in a subsequent publication.'

EXPERIMENTAL PROCEDURES
Materials-Radioactive steroids were purchased from Amersham Corp. with the exception of R-5020 (Promegestone) which was from New England Nuclear. l4C-1abeled proteins, used for markers in sucrose gradients, were products of New England Nuclear. Most of the radioinert steroids were obtained from Steraloids (Wilton, NH). Antheridiol was a generous gift from Dr. Trevor McMorris (University of California, San Diego). Dexamethasone was a gift from Merck, and 1,25(OH)2D3 and 25(OH)D3 were gifts from Hoffmann-LaRoche. Solvents used for extraction and chromatography were high performance liquid chromatography grade purchased from J . T. Baker Chemical Co. Sephadex (G-50 fine) and Sephacryl (S-200) were products of Pharmacia (Piscataway, NJ). Trayslol was obtained from Mobay Chemical Corp. (New York, NY). Other reagents were purchased from Sigma unless noted.
Yeast Cultures-The strain of C. albicans employed in these studies was derived from a clinical isolate kindly supplied by D. Schurman and identified serologically as type A in the laboratory of D. Stevens (both a t Stanford University). Cultures were periodically examined using standard serological techniques to ensure that the cells used were unchanged. C. albicans were grown on nutrient agar (Difco) in 10-cm Petri dishes (Falcon, Oxnard, CA). Routinely, 12 dishes were streaked for heavy growth and cells were grown for 24 h at 37 "C before harvesting. Although most experiments employed the yeast form of the organism, the pseudohyphal form was also examined. Greater than 99% pseudohyphae were generated by growing the yeast in the presence of 5% newborn calf serum for 18 h a t 37 "C. The serum was heated to 56 "C for 30 min to destroy corticosteroidbinding globulin (6) and was also charcoal stripped to eliminate endogenous steroids (7).
Steroid-binding Assays-C. albicans yeast forms were harvested by centrifugation at 1000 X g, and the packed cell pellet (-1.0 m l ) was resuspended in 2 ml of homogenizing medium containing 250 mM sucrose, 10 m~ Tris-HCI, 12 mM monothioglycerol, 1.5 mM EDTA, and 10 mM Na2Mo04, pH 7.8. An 0.5-ml aliquot of cells was added to 0.7 ml of 250-300 pm glass beads (Sigma) in 1.5-ml conical plastic centrifuge tubes. Cells were lysed by vigorous agitation on a Vortex mixer with repeated 15-s periods of cooling of the tubes in an ice slurry. Cytosol was then prepared by ultracentrifugation a t 204,000 X g for 30 min at 4 "C. In typical binding experiments, cytosol samples were incubated with [3H]corticosterone for 3 h at 0 "C. Preliminary experiments revealed that equilibrium was obtained under these 4926 Corticosterone-binding Protein in C. albicans conditions. Nonspecific binding was assessed in all experiments by parallel incubations performed in the presence of a 250-fold molar excess of radioinert corticosterone. Bound steroid was separated from free steroid by chromatography of samples on 4 ml of Sephadex G-50 minicolumns as previously described (8). Specific binding was calculated by subtracting nonspecific binding from total binding. In some experiments removal of free steroid was accomplished with a dextrancoated charcoal technique. Cytosol samples containing [JH]steroid were agitated for 15 s with a volume of 0.5% Norit A charcoal (Sigma) with 0.05% T-70 dextran (Phannacia) equal to the incubation volume and centrifuged at 3000 X g to remove charcoal (9). Chromatographic Analysis of the Ligand to Evaluate Steroid Metabolism-["H]Corticosterone was examined for metabolic conversion after incubation both in the presence of cytosol and with intact C. albicans. Cytosol was prebound with 26 nM ["Hlcorticosterone for 3 h at 0 "C. Sephadex chromatography was performed as usual, and the protein peak was collected and extracted with methylene chloride. A second approach employed intact C. albicans cells incubated in serum-free Dulbecco's medium (Grand Island Biological Co., Grand Island, NY) with 7 X lo4 dpm/ml of ["H]corticosterone. After 48 h of incubation, yeast were removed by centrifugation, and the medium was extracted with methylene chloride. Both types of extract were dried under nitrogen and applied to 13-cm instant thin layer chromatography strips (Gelman Sciences, Inc., Ann Arbor, MI) and developed in hexane:chloroform:methanol, 8020:2. Strips were then assessed for radioactivity with a radiochromatogram scanner (Packard Instrument Co.).
Sucrose Gradient Analysis-Cytosol was obtained by lysing cells in sucrose-free homogenizing medium in the absence (hypotonic) or presence of 0.3 M KC1 (hypertonic). Cytosol samples (200 p1) were applied to 5-2C% sucrose gradients which were prepared in either hypo-or hypertonic buffer. The gradients were centrifuged for 18 h at 337,000 X gmax, and 0.15-ml fractions were collected. Each fraction was incubated with 26 nM [3H]corticosterone f 250-fold molar excess of radioinert corticosterone. Binding was determined with the dextran-coated charcoal technique.
Molecular Weight Estimation-Cytosol samples were applied to a calibrated Sephacryl S-200 column (SO X 1.6 cm) and eluted at a flow rate of 0.75 ml/min with a buffer consisting of 50 m~ Tris and 1.5 mM EDTA, pH 7.6. Fractions (2 m l ) were divided and incubated with 26 nM ["H]corticosterone f 250-fold radioinert corticosterone to determine specific binding using the dextran-coated charcoal technique.
Enzymatic Studies-Cytosol was prepared in modified homogenizing medium which contained no monothioglycerol or NazMoOl. Trypsin (1:200, Difco), neuraminidase (Millipore Corp., Freehold, NJ), and N-ethylmaleimide (Sigma) were tested in this buffer; RNase and DNase I (Worthington) were tested in the presence of 6 m~ MgC12; phospholipase A2 (Sigma) was tested in the presence of 2.0 mM CaC12. AU samples were incubated for 30 min at 37 "C, and then specific ["H]corticosterone binding was determined by the column chromatography method.
Other Assays-Protein concentration was measured by the Coomassie dye binding technique (10) employing a mixture of 80% human y-globulin and 20% bovine serum albumin (Reheis Chemical, Phoenix, AZ) as the standard.
All data are expressed as mean -t S.E of specific binding.

RESULTS
Existence of Specific Steroid-binding Sites in C. albicans-Initial experiments to determine whether high affinity saturable steroid-binding sites were present in cytosol from C. albicans involved incubation of fungal cytosol with a variety of 3H-ligands. As described previously, specific binding was found with ["H]corticosterone and ['Hlprogesterone (1). The following tritiated probes failed to reveal specific binding: testosterone, dihydrotestosterone, estradiol, diethylstilbestrol, R-5020, dexamethasone, and triamcinolone acetonide, all tested at 26 n~ and 1,25(OH)2D3 a t 1.3 nM and 25(OH)D3 at 26 nM. As noted, both [3H]corticosterone and ["Hlprogesterone exhibited specific binding. Cross-competition experiments revealed that both ligands were labeling the same binding site. As somewhat higher levels of specific binding were detected with ["H]corticosterone we have employed this steroid as our radioprobe.
Binding Assay Conditions- Fig. 1 illustrates the time course of [JH]corticosterone binding to C. albicans cytosol at 0 "C. Binding is rapid, reaching maximal levels in 2 h, and remaining stable for several hours. Experiments at temperatures between 0 and 37 "C give similar results with similar levels of maximum binding detected although degradation occurs more rapidly at elevated temperatures. Most experiments were, therefore, performed at 0 "C after 3 h of incubation.
In further experiments to validate the binding assay we examined [3H]corticosterone binding as a function of cytosol protein concentration, and the results are shown in Fig. 2. Binding was found to be linearly related to protein concentration above 0.25 mg of protein/ml, and in subsequent studies we used this or a higher concentration of protein.
The binding reaction displayed quite a broad pH maximum, with little change in binding seen between pH 6.0 and 8.0 (data not shown). Buffer conditions were selected such that upon homogenization at 0 "C, cytosol pH was approximately 7.3-7.5.
Proof that the Bound ["€€]Corticosterone Is Unmetabolized-It has been reported (11) that C. albicans possesses the ability to metabolize C-21 steroids. To ascertain whether the radioactive ligand bound in C. albicans cytosol was unmetabolized [3H]corticosterone, we performed a series of chromatographic analyses. The results of one such experiment are presented in Fig. 3. To assess the ability of the intact fungus  Fig. 3 4 . As can be seen in Fig. 3b almost all the 'H-material extracted from the medium co-migrated with ['H]corticosterone. A small unresolved peak (-4% of the total, determined by peak area) migrates as a more polar entity. Based on the results of others (11) this material is probably a C-20 hydroxy derivative of corticosterone. In further studies ["H]corticosterone was incubated with cytosol from C. albicans and chromatographed over a Sephadex column. The protein peak was collected, and the protein-bound steroid was extracted and chromatographed in the same instant thin layer chromatography system. The radioactive material extracted from the cytosol migrates with authentic ['H]corticosterone (Fig.  3c). Although these results do not absolutely identify the bound steroid, they are consistent with minimal corticosterone metabolism and strongly suggest that the bound hormone in o w studies is intact corticosterone.
After establishing the appropriate conditions for the binding assay, we next examined various properties of the binding site.  radioactive material recovered from the growth medium after incubation of 7 X lo4 dpm/ml of [YH]corticosterone with C. albicans for 48 h at 20 "C. c, radioactive steroid recovered from C. albicans cytosol after a typical binding experiment. Cytosol and growth medium were extracted with methylene chloride, dried down under nitrogen, and redissolved in ethanol. Extracts and authentic [3H]corticosterone were then applied to 13-cm instant thin layer chromatography plates and developed in hexane:chloroform:methanol, 80202. Radioactivity was detected with a radiochromatogram scanner.  Equilibrium analysis of specific [3€&orticosterone binding in cytosol prepared from the pseudohyphal form of C. albicans. Cytosol was prepared and incubated with [3H] corticosterone (5.2 X IO"" to 5.2 x IO-* M ) for 3 h at 0 "C. a, binding isotherm. 0, total binding; A, nonspecific binding; 0, and specific binding. b, specific binding plotted by the method of Scatchard. Units on the ordinate are (dpm/mg)/ (dpm/ml). Protein in C. albicans Binding Kinetics-The binding of [3H]corticosterone to the fungal binder is rapidly reversible at 0 "C as shown in Fig. 5. The dissociation of the steroid-binder complex is apparently first order, as expected, and the rate constant calculated from the slope of the logarithmic plot is 0.043 min", the tIl2 for dissociation then being about 16 min. Given this rapid dissociation, our technique of separating bound hormone from free hormone underestimates the binding capacity of C. albicans   binder by -40% (given that the average column separation takes 12-14 min). If the dissociation is a simple first order reaction there should be no effect on our estimate of the apparent equilibrium constant.

Comparison of the Binding Protein in Pseudohyphal and
Steroid Specificity of the Binding Protein-Detailed anal- ysis of the steroid specificity of the binding in the yeast form is presented in Table I. Corticosterone and progesterone are potent and essentially equal in their ability to compete for ["H]corticosterone-binding sites. Amongst glucocorticoids, cortisol and prednisolone are moderately good competitors, whereas lla-cortisol (the inactive stereoisomer of cortisol), dexamethasone, and triamcinolone acetonide are without binding activity at the concentrations tested. Note that R-5020, the synthetic progestin with high affinity for mammalian progesterone receptors (12), is only about 2% as potent a competitor as progesterone. The other compounds assayed had minimal or no competitive activity at the concentrations tested. Ergosterol, the major sterol synthesized by this organism, is a very weak competitor. Mineralocorticoids, sex steroids, vitamin D metabolites, and ergocalciferol are all without activity. Antheridiol, a putative fungal hormone in Achyla bisexualis (5, 13) fails to compete at the concentrations tested. Taken together, these data indicate that the binder in C. albicans is remarkably steroid selective and stereospecific in nature.
Binding Protein Stability-We next focused on some of the properties of this binder using physical, chemical, and biochemical techniques. First, thermal stability and susceptibility to enzymatic digestion were explored. As shown in Fig. 6 the fungal binder is stable a t 37 "C for 30 min but destroyed at 56 "C. After 30 min of incubation at 37 "C with 100 p g / d of trypsin essentially all binding was abolished. Similar results followed exposure to 6 mM N-ethylmaleimide. DNase, RNase, phospholipase AP, and neuraminidase had little or no effect on binding. These results suggest that the binder is a protein with free sulfhydryl groups required for binding. Sedimentation a n d Chromatographic Analysis-Sucrose gradients have frequently been used in the study of mammalian steroid receptors, and the results of this technique used with the C. albicans protein are depicted in Fig. 7. The binder migrated a t -4 S in 5-20% gradients whether prepared in hyper-or hypotonic buffers. No disaggregation to smaller forms occurred in hypertonic gradients, a characteristic frequently noted with mammalian steroid receptors (14). In experiments not shown we did not see any changes in sedimentation coefficient when cytosol was prepared with protease inhibitors such as phenylmethylsulfonyl fluoride or tra-To further assess the molecular size of this protein we employed chromatography over a Sephacryl S-200 exclusion gel column. These results are shown in Fig. 8. The broad based peak suggests heterogeneity of the binder. Since the binding assay was performed on the eluate fractions, this finding is not due to ligand dissociation during the chromatography procedure. Based on several experiments we estimate the molecular weight of the major binding protein peak to be -43,000 daltons in hypotonic buffer. Taken with the sylol.
sedimentation value of -4 S, the fungal binding protein is somewhat asymmetrical (15).

DISCUSSION
In this report we have described an intracellular protein found in C. albicans which binds corticosterone and certain other steroid hormones with high affinity, selectivity, and stereospecificity. We hypothesize that this protein represents a primitive form of either the mammalian glucocorticoid receptor or CBG.' Even if this speculation does not prove to be correct, the protein is still of considerable interest.
Intense investigation into the evolution of steroid endocrine systems has occurred (for review see Ref. 16). Binding of mammalian hormones to a transport protein-enzyme has been documented in Pseudomonas testosteroni (17) and seen after abiotic synthesis (16). In addition, a 5.1 S progesterone-binding protein has been reported in the bacterium Streptomyces hydrogenans (18). However, an intracellular steroid-binding protein with receptor-like properties has not previously been demonstrated in primitive eukaryotes. The fact that unicellular organisms may use sterols as message molecules appears to be well demonstrated in the fungus Achyla bisexualis where a C-29 sterol, antheridiol, seems to function via interaction with the genome and requires de novo protein synthesis for activity (5). In fact, this heterothallic organism appears to employ two sterols to govern its reproductive cycle (13,19). The data presented here indicate that some unicellular fungi also possess intracellular steroid-binding proteins. We interpret this finding to imply that these simple organisms use such proteins as sterol/steroid receptors. However, we have examined several other simple organisms including Saccharomyces cereviseae, Neurospora crassa, and Paracoccidioides brasiliensis and have not successfully demonstrated this corticosterone-binding site.3 Thus all fungi do not appear to possess a similar glucocorticoid binder.
The binding protein we have demonstrated in C. albicans exhibits some characteristics suggestive of mammalian receptors or specific plasma transport proteins; however, several properties of the fungal binder are clearly different. Table I1 compares the properties of the C. albicans binder with two well studied mammalian steroid-binding proteins, CBG, and the classical intracellular glucocorticoid receptor. Of the properties examined, the fungal binder resembles CBG somewhat more closely than the receptor although it is clearly not identical to either. Localization of the binder in the cytosol and the absence of secretion into the medium is suggestive of a receptor rather than a transport role for this binder. It would of course be interesting to establish whether the fungal binder has structural homology to mammalian receptors or CBG but  The affinity exhibited for corticosterone is similar in all three proteins. This glucocorticoid is the native ligand for both CBG and the receptor in rodents but is probably not the native ligand in C. albicans. We have described a lipid-extractable substance present in C. albicans and also released into the medium which has the ability to reversibly displace ["H]corticosterone from the fungal binding site (1). We believe this material represents the natural ligand and it may prove to have a higher affinity than corticosterone at the fungal binding site. Until the active agent is available in pure form and radiolabeled the properties of the binding protein in the presence of the endogenous ligand cannot be evaluated.
The finding of a protein in a pathogenic fungus with the ability to interact with mammalian hormones raises the possibility that such organisms may respond to the host endocrine milieu via this binder. We are currently investigating this possibility.