Steroid binding sites with high affinity for progesterone (Kd= 40±14 nM determined by binding, and Kd= 71±22 nM determined by displacement studies) and lower affinity for 21-hydroxyprogesterone and for testosterone, but no affinity for estradiol-17β, onapristone and α-naphthoflavone were detected in the enriched plasma membrane fraction of the fungus Rhizopus nigricans. The amount of steroid binding sites is in accordance with the value of Bmax= 744±151 fmol (mg protein)−1. In the membrane fraction, progesterone induced about 30% activation of G proteins over basal level, as determined by GTPase activity (EC50= 32±8 nM) and by the guanosine 5′-O-(3-thiotriphosphate) (GTPγS) binding rate (EC50= 61±21 nM). The affinity of receptors for progesterone was substantially decreased in the presence of GTPγS and of cholera toxin. Our results suggest the existence of progesterone receptors in the membrane of Rhizopus nigricans and their coupling to G proteins.
It was reported some decades ago that the mammalian gonadal hormones progesterone, 21-hydroxyprogesterone, and dihydrotestosterone inhibit growth of bacteria and some pathogenic fungi [1,2]. In addition to mammalian pathogenic fungi, growth inhibition by mammalian steroids, especially progesterone and 21-hydroxyprogesterone, was observed also in the saprophytic fungus Rhizopus nigricans. In response to progesterone action this fungus has evolved defense mechanisms including heat shock protein synthesis, sugar epimerase gene expression, and induction of progesterone hydroxylating enzyme(s) which transform progesterone into less toxic and more easily removable hydroxy-products. In this investigation we have studied the interaction of progesterone with R. nigricans plasma membrane, the first site of mycelial contact with toxic steroids from the environment.
Plasma membrane initiated events leading, together with classical genomic steroid action, to diverse cellular responses are well known in vertebrate steroid responsive cells [7,8]. Some recent studies suggest the association of the mammalian plasma membrane steroid receptors with G proteins [9,10]. In fungi, however, only steroid signaling analogous to the classical mechanism of steroid action has been observed [2,11,12]. In this context we found and partially characterized progesterone receptors in the cytosol from R. nigricans. So far, fungal plasma membrane steroid receptors have only been detected in the ascomycetous fungus Cochliobolus lunatus. On the other hand, the presence of G proteins in fungi is well documented, in particular in phyla Ascomycetes and Basidiomycetes where they respond via receptors to various non-steroid signals such as pheromones, UV-irradiation, and sources of glucose and nitrogen. The cascade(s) resulting from G protein activation in fungal cells regulate development, mating and virulence. The activation of G proteins by steroid-binding molecules in the plasma membrane has not, to our knowledge, been reported in fungi.
2 Materials and methods
2.1 Chemicals, microorganism and preparation of subcellular fractions
Radioactive chemicals were obtained from NEN Research Products (Boston, MA, USA); non-labeled steroids (except onapristone, Schering, Berlin, Germany), α-naphthoflavone, pertussis toxin (PTX), cholera toxin (CTX), guanosine 5′-O-(3-thiotriphosphate) (GTPγS), Millipore G/F filters from Sigma Chemical Co. (St. Louis, MO, USA); and Sephadex LH-20 from Pharmacia (Uppsala, Sweden).
Filamentous fungus R. nigricans ATCC 6227 from the phylum Zygomycetes obtained from MZKI (Ljubljana, Slovenia) was grown in nutrient medium in a rotary shaker for 18 h at 28°C. Fungal mycelia were homogenized in TEGMP buffer (10 mM Tris–HCl, 1 mM EDTA, 10% glycerol, 2 mM monothioglycerol, 1 mM phenylmethylsulfonyl fluoride, pH 7.4) as described in. Plasma membrane and cytosol fractions were prepared by centrifuging the homogenate at 40 000×g for 40 min and at 105 000×g for 1 h, respectively. Membrane sediment was resuspended in TEGMP buffer (2–3 mg protein ml−1) and characterized by specific enzyme activity of H+-ATPase.
2.2 Progesterone binding and progesterone competition assay
In the binding membrane samples were incubated with an equal volume of [3H]progesterone (0.5–100 nM) in TEGMP buffer for 45 min at 22°C. Non-specific binding was determined by parallel incubations in the presence of a 125-fold excess of unlabeled progesterone. Non-bound steroid was removed by Sephadex LH-20 gel exclusion chromatography.
In competition experiments membrane fraction or cytosol was incubated for 45 min at 22°C with equal volume of 40 nM [3H]progesterone in TEGMP buffer alone and with [3H]progesterone in the presence of different excesses of selected competitors.
To examine the linkage of receptors to G proteins in the membrane fraction progesterone-displacement experiments were performed either in the presence of G protein-modifying agents (GTPγS, CTX, PTX) or without them. In vitro ADP ribosylation of G proteins by CTX and PTX was performed according to McKenzie.
2.3 Assay of GTPase activity and of GTPγS binding
The effect of progesterone (1 nM to 10 μM) on GTPase activity and on the rate of GTPγS binding to G proteins in membrane fractions was tested radiometrically. In GTPase activity assay plasma membranes (about 0.5 mg protein ml−1) were incubated at 25°C for 10 min with trace amounts of [γ-32P]GTP to give 50 000–100 000 cpm in an aliquot of the reaction cocktail and cold GTP to give the required total concentration of GTP of 0.5 μM. Subsequently, the unreacted GTP was removed by activated charcoal and the remaining radioactivity of the released phosphate was determined. In GTPγS binding assay plasma membranes (0.5 mg protein ml−1) in the reaction cocktail were incubated with [35S]GTPγS (50 000 cpm per assay) at 25°C for 2 min, followed by rapid vacuum-filtering through glass fiber filters. After washing, the remaining radioactivity of the filters was counted.
2.4 Other procedures
The radioactivity of the samples was determined in LKB 1214 Rackbeta liquid scintillation counter. Sf9 cells were grown and transfected with recombinant baculovirus in order to express G proteins (Gα5, Gαi1, Gαo or Gα11, together with β1γ2) as described by Näsman et al.. Results were analyzed by the Prism 2 computer package (Graphpad, San Diego, CA, USA).
3 Results and discussion
3.1 Progesterone binding and steroid specificity of membrane receptors
Binding of progesterone to the enriched plasma membrane fraction was characterized by a Kd of 35±14 nM and Bmax of 744±151 fmol (mg protein)−1, from the saturation binding curve (Fig. 1A), and by Kd of 71±22 nM obtained from displacement studies (Fig. 1B). Compared to results obtained on rat brain and Xenopus oocytes, the affinity of progesterone for R. nigricans membrane receptors is fairly high. It is worth noting that a high level of non-specific progesterone binding to fungal membrane particles was observed (Fig. 1B). This is in accordance with the results obtained by others in various tissues where non-specific intercalation of steroids in the lipid bilayer of the cell membrane has been demonstrated.
Specific progesterone binding to plasma membranes of R. nigricans. A: Saturation analysis; plasma membrane fraction was incubated with [3H]progesterone at multiple concentrations (total binding). To assess the non-specific binding, parallel samples were incubated with [3H]progesterone in the presence of a 125-fold excess of [1H]progesterone. The experimental points, representing specifically bound progesterone (a difference between total and non-specific binding), were fitted by a one-site binding equation by a non-linear regression procedure using the Prism 2 computer package (Graphpad). The corresponding Scatchard plot is shown in the inset. Each point is the average of three independent determinations. B: Competition analysis; plasma membrane fraction was incubated with 10 nM [3H]progesterone in the presence of different concentrations of unlabeled steroids. The value of 100% binding ranged from 0.77 to 1.85 nM. Each point is the average of four independent determinations. Standard errors of the mean never exceeded 15% and are omitted for clarity.
In competition experiments radiolabeled progesterone was displaced effectively by unlabeled progesterone, followed by moderately active 21-hydroxyprogesterone and testosterone, whereas estradiol-17β and the mammalian progesterone antagonist onapristone did not displace labeled progesterone (Fig. 1B). Blackmore et al. demonstrated that progesterone binds to human sperm membrane receptor across the β face of the steroid C/D-ring upper edge and that the methyl group at C-21 is important for the binding. As for sperm progesterone membrane receptors the β face of the steroid might be a site of interaction also with fungal membrane receptors since onapristone with a bulky group at C-11 in the β position was not able to displace progesterone. In addition, it seems likely that the steroid A-ring plays a significant role in ligand recognition by fungal membrane receptor since estradiol-17β with aromatic A-ring was not able to displace progesterone. It appears that the required structural characteristics of a ligand for R. nigricans membrane binding sites are 3-keto, 4-ene steroid, possibly methyl group at C-21 and free C-11 β position.
In the search for an antagonist of membrane progesterone receptors we examined the competitive ability of α-naphthoflavone, previously shown to inhibit progesterone induced steroid-hydroxylase activity. α-Naphthoflavone competed efficiently for cytosolic but not for membrane-bound progesterone receptors (Fig. 2). This result together with different affinities of cytosolic (two types of binding sites Kd1= 4 nM, Kd2= 30 nM; see) and of membrane receptors for progesterone (single binding site, Kd= 35–70 nM; see Fig. 1) are strong evidence of two different progesterone receptors in R. nigricans. Both types of receptor may contribute to steroid action in mycelia. According to Brann et al. a non-classical mechanism of action via plasma membrane steroid receptors is not a challenge to the classical mechanism but an additional, complementing layer of diversity.
Displacement of [3H]progesterone by α-naphthoflavone from membrane-bound and cytosolic receptors of R. nigricans. Samples of cytosol and membrane fraction were incubated with 40 nM [3H]progesterone in the presence of different concentrations of α-naphthoflavone. 100% of [3H]progesterone bound ranged from 2.8 to 3.2 nM in the cytosol and from 0.65 to 0.92 nM in the membranes. Each point represents the mean value of four independent determinations. The competition curves for a one-site competition equation were fitted to the experimental points using the Prism 2 computer package (Graphpad).
We also tested the possibility that plasma membrane receptors are designed primarily to recognize some natural ligand(s) and that toxic steroids just use this natural signaling pathway. A well known model of sexual recognition by steroid pheromones oogoniol and antheridiol has been thoroughly studied in oomycetous fungus Achlya ambisexualis, where specific pheromone receptors were detected in the cytosol. We examined the zymomycetous pheromone trisporic acid for competition for plasma membrane progesterone receptors of R. nigricans but it did not displace radiolabeled progesterone from membrane receptors (results not shown). The plant hormone brassinosteroid 24-epicastasterone, which is converted into the hydroxy-derivative by the fungus C. lunatus, was also tested, but no affinity for R. nigricans membrane progesterone receptors was detected (data not shown). Thus, progesterone appears to be the most appropriate candidate for the in vivo ligand of membrane-bound receptors in R. nigricans. The latter are likely to be involved in a process leading to progesterone being made less toxic, possibly by hydroxylation by induced enzymes.
3.2 Activation of G proteins in the presence of progesterone
G proteins in the membrane fraction from R. nigricans were activated in the presence of progesterone (Fig. 3). Due to the rather low degree of maximal activation of G proteins observed, we used two independent methods for detecting G protein activation, GTPase activity assay and determination of GTPγS binding rate. The low activation of G proteins on binding of ligands to membrane-bound receptors is not unusual and has frequently been observed in tissues and cells with low levels of G proteins in the membrane. In R. nigricans plasma membranes progesterone increased GTPase activity up to 30±9% over basal and a value of EC50 of 32±8 nM was determined (Fig. 3). By the analogous procedure for the rate of GTPγS binding we obtained the maximal activation of G proteins by progesterone of 35±4% over basal and EC50 of 61±21 nM (Fig. 3). Both values of EC50 are reasonably close to the values of Kd for progesterone–receptor binding (Fig. 1), indicating that G proteins were activated as a consequence of progesterone binding to the plasma membrane steroid receptors.
Effect of progesterone on the GTPase activity and on the rate of GTPγS binding to the plasma membranes of R. nigricans. GTPase and GTPγS binding measurements were performed as described in Section 2. Experimental points were fitted by a single-phase dose–response equation by a non-linear regression procedure using the Prism 2 computer package (Graphpad). Each point is the average of three independent determinations. 100% of basal GTPase activity represents 8±2 pmol min−1 mg−1 protein and 100% of basal rate of GTPγS binding was 39±6 fmol min−1 mg−1 protein.
We also checked the possibility that the activation of G proteins by progesterone was a consequence of the direct binding of the steroid to G proteins. We determined the rate of [35S]GTPγS binding to membrane from Sf9 cells with overexpressed Gαs, Gαi1, Gαo or Gα11, together with β1γ2, in the absence and presence of progesterone. No effect of progesterone on the rate of [35S]GTPγS binding to the membranes of Sf9 cells transfected with any of the G proteins listed above could be detected (results not shown).
3.3 Coupling of membrane progesterone receptors to G proteins
To shed more light on the linkage between progesterone receptors and G proteins we treated fungal membrane fraction with GTPγS and with toxins PTX and CTX. The presence of GTPγS in the incubation medium changed the progesterone binding profile in the membrane fraction by decreasing its affinity for receptors (Fig. 4A), indicating the coupling of steroid receptors to G proteins. In the case of toxins, no alteration of the binding affinity in the presence of PTX (results not shown), but a substantial reduction of progesterone binding affinity in the presence of CTX was observed (Fig. 4B). These results, summarized in Table 1, strongly suggest the involvement of G proteins in signal transduction via membrane-bound progesterone receptors in the fungus and are in accordance with observations by Caldwell et al. who, using GTPγS and toxins, demonstrated that G proteins were associated with progesterone and estradiol membrane binding sites in the medial preoptic area–anterior hypothalamus of rat.
Effect of GTPγS and CTX on the binding of progesterone to fungal plasma membrane. Experiments were performed as described in Section 2. 100% of [3H]progesterone bound in different experiments ranged from 2.2 to 2.5 nM. The competition curves were fitted to the experimental points in accordance with the one-site competition equation using the Prism 2 computer package (Graphpad). A: Competition of unlabeled progesterone for labeled membrane binding sites in the presence and absence of GTPγS. Each point represents the mean value of three independent determinations. B: Competition of unlabeled progesterone for labeled membrane binding sites in the presence and absence of CTX. Each point represents the mean value of six independent determinations.
Effect of GTPγS, CTX and PTX on the binding of progesterone to membrane-bound progesterone receptors from R. nigricans
3.4 Concluding remarks
We present here evidence that the filamentous fungus R. nigricans contains in the plasma membrane progesterone receptors that are coupled to G proteins. Possible downstream signaling routes, such as the activation of adenylyl cyclase, phospholipase C, and mitogen-activated kinase cascade are under investigation. At present, the biological role of R. nigricans plasma membrane progesterone receptors coupled to G proteins is not known but it can be speculated that these signaling molecules are involved in processes rendering progesterone less toxic.
The authors are indebted to Marjan Kužnik for excellent technical assistance, and would like to thank Prof. Herman van den Ende (University of Amsterdam, The Netherlands), Dr. Ladislav Kohout (Academy of Sciences of Czech Republic), and Prof. Ülo Langel (Stockholm University, Sweden) for the gift of trisporic acid, 24-epicastasterone, and the recombinant baculoviruses, respectively. We are also grateful for the support of our work by Prof. Katja Breskvar. The work was supported by grants from the Ministry of Education, Science and Sport of Slovenia.
(1999) Radioligand assays for oestradiol and progesterone conjugated to protein reveal evidence for a common membrane binding site in the medial preoptic area-anterior hypothalamus and differential modulation by cholera toxin and GTPγS. J. Neuroendocrinol. 11, 409–417.