| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Abstracts of Oak Ridge Posters |
[1 Department of Pharmaceutical Sciences, University of Bologna, Bologna, Italy;2 Center for Applied Biomedical Research (CRBA), S. Orsola-Malpighi Hospital, Bologna, Italy;3 Department of Biochemistry, University of Turku, Turku, Finland;4 Tampere University of Technology, Institute of Environmental Engineering and Biotechnology, Tampere, Finland;
aaddress correspondence to this author at: Department of Pharmaceutical Sciences, University of Bologna, Via Belmeloro 6, 40126 Bologna, Italy; fax 39-051-343398, e-mail aldo.roda{at}unibo.it]
Androgens regulate several physiologic processes, such as normal prostate development and maintenance of male sexual function in adult life, and they are also involved in pathologic conditions, including prostate cancer, prepubertal gynecomastia, and premature pubarche. Androgens enter the target cell and bind to androgen receptor (AR), a ligand-dependent transcription factor in the nuclear receptor superfamily comprising receptors for vitamin D3 and thyroid and steroid hormones (1)(2). After binding of the hormone, AR enters the nucleus and binds to the regulatory region of the target gene as a homodimer.
Recent findings suggest that some environmental chemicals disrupt the endocrine system and cause adverse effects such as cancers and sexual abnormalities in humans and wildlife (3). These endocrine-disrupting compounds include several xenoestrogens, androgens, and antiandrogens, which present a variety of chemical structures and mechanisms of action (4)(5).
Conventional detection methods for androgenic and antiandrogenic compounds cannot evaluate hormonal bioactivity, which is important in a wide range of clinical conditions, and no alternative direct and simple methods for measurement of plasma androgen bioactivity are available. In addition, the increased use of androgenic substances as therapeutic drugs and their abuse to enhance athletic performance require sensitive and rapid screening tests. The very low amounts of single analytes in anabolic cocktails and nutritional supplements are difficult to detect with standard techniques such as gas chromatography–mass spectrometry. Thus, bioassays that reveal the overall androgen bioactivity of complex mixtures would be more appropriate tools with great implications in antidoping analysis (6)(7)(8).
We developed and validated a yeast-based bioassay to measure androgen bioactivity in clinical samples, including human serum, for the detection of androgen-like and antiandrogenic compounds (9). The bioassay is based on recombinant Saccharomyces cerevisiae BMA64-1A strain genetically engineered with the introduction of 2 plasmids: an expression plasmid (pG1AR), in which the expression of human androgen receptor (hAR) is regulated by a constitutive yeast promoter; and a reporter plasmid (YipLuc-ARE), in which the ARE sequences drive the expression of a truncated form of Photinus pyralis luciferase, used as reporter gene. The latter plasmid was integrated into the yeast genome to provide greater stability. In the presence of androgenic compounds, hAR moves into the nucleus and binds ARE sequences, leading to luciferase expression directly proportional to the androgenic activity of the sample (Fig. 1A
). The use of a luciferase without the peroxisomal targeting sequence (luc-skl) avoided luciferase importation into peroxisomes and produced higher light emission without interfering with normal host physiology (10).
|
A recombinant yeast strain constitutively expressing the same luciferase in the plasmid PRS316-Luc was used as control to correct the bioluminescent signal according to cell vitality and nonspecific matrix effects (11).
The 2 strains were routinely grown in selective synthetic complete (sc) medium containing 6.7 g/L yeast nitrogen base without amino acids, 1.4 g/L yeast synthetic drop-out media supplement, and 40 mL/L of a 500 g/L D-glucose solution.
The control strain was grown in sc medium without uracil, whereas the androgen-responsive strain was grown in sc medium without uracil and tryptophan. Glycerol stocks of the 2 recombinant strains were prepared and kept at –80 °C for long-term storage. Stock solutions of the compounds to be tested, all purchased from Sigma, were prepared in ethanol at a concentration of 10 mmol/L and stored at –20 °C.
The bioassay procedure was as follows: a 2-mL overnight culture was used to inoculate 18 mL of fresh selective sc medium. The A600 nm was monitored until it reached 0.4, at which point 90-µL volumes of the cell suspension were transferred to wells in a sterile white 96-well microtiter plate with 10 µL of sample or calibrator solution.
Plates were sealed to prevent evaporation and shaken for 1 min before incubation. After 2.5 h of incubation at 30 °C, luminescence was measured with a Luminoskan microplate reader as follows: 100 µL of 1 mmol/L D-luciferin (supplied by Synchem) in 0.1 mol/L sodium citrate buffer (pH 5.0) was injected with a built-in reagent dispenser. After brief shaking, luminescence was measured with 5-s integration. Light emissions were expressed as relative light units.
Dose–response curves for testosterone were produced in each plate for testosterone concentrations ranging from 0.01 to 104 nmol/L. The analytical signal was plotted as either light emission against log[testosterone] (uncorrected curves) or as the ratio of biosensor light emission over control light emission against log[testosterone] (corrected curves).
The corrected light signal was proportional to the testosterone concentration in the concentration range from 0.1 to 100 nmol/L. The detection limit of the bioassay, defined as the testosterone concentration that produced a luciferase activity 2 SD above the mean of 10 samples of charcoal-stripped serum (used to reduce the contaminating steroids from the serum), was 0.05 nmol/L. Increasing sample volumes did not yield an improvement of the assay sensitivity. The intraassay (estimated by assaying the same serum pools 6 times in a single assay) and the interassay (estimated by assaying duplicate samples from a serum pool on 6 separate days) variations were evaluated at 3 testosterone concentrations: low (1 nmol/L), middle (100 nmol/L), and high (1 x 104 nmol/L). The intra- and interassay CVs were <13% and <22%, respectively. In particular, for the lowest concentration (1 nmol/L), the intraassay and interassay CVs were 11% and 21%, respectively.
We compared the results obtained by the yeast bioassay with results obtained by commercial androgen enzyme immunoassays. Serum androgen activity correlated well with serum testosterone concentration measured with standard enzyme immunoassays (r = 0.91; P <0.0001; n = 10; see Fig. 1S in the Data Supplement that accompanies the online version of this poster abstract at http://www.clinchem.org/content/vol51/issue10/). Other hormones, such as 4-androstene-3ß,17ß-dione (or androstenedione) and dihydrotestosterone (DHT), were tested to verify the biosensor specificity. Different amounts of testosterone and other androgenic compounds were added to charcoal-stripped fetal calf serum, and the corresponding dose–response curves were obtained and compared with that of testosterone (Fig. 1B
). DHT was the most active androgen: fetal calf serum containing 0.1 nmol/L DHT induced a light signal equal to 
1 nmol/L testosterone. 4-Androstene-3ß,17ß-dione was slightly less active than testosterone. These results agree with reported data (7). For analysis of samples with unknown composition, androgenic bioactivity was expressed in terms of testosterone equivalents.
We investigated the effect of antiandrogens on the hAR by performing competitive assays. Testosterone was first added to charcoal-stripped FCS at a subsaturating concentration (10 nmol/L), after which increasing amounts of the tested compounds (bisphenol A, a plasticizer used in the production of epoxy resins and polycarbonate plastics, and vinclozolin, an antiandrogenic fungicide) were added; the resulting solution was subjected to the bioassay. The median inhibitory concentrations (IC50s) were 5 µmol/L for bisphenol A and 10 µmol/L for vinclozolin (data not shown). In addition, hydroxyflutamide and diethylstilbestrol (DES) were also shown to block androgen stimulation in a dose-dependent way. Hydroxyflutamide also displayed some agonistic activity at high concentrations, which is in agreement with previous findings (see Fig. 2S in the online Data Supplement) (12). Hydroxyflutamide had an IC50 of 0.5 nmol/L in the presence of 10 nmol/L testosterone. The difference between the IC50 values reported here and those reported earlier in other cell lines can be attributed to the presence of coactivators in different cell lines and to interference through other signal transduction pathways (13). The antiandrogenic activity of DES, together with other xenoestrogens, could explain the occurrence of several pathologies, such as hypospadias and cryptorchidism, associated with maternal exposure to DES and pesticides (14).
To test a clinical application of the bioassay, we performed a preliminary screening of human serum samples, which were assayed without any sample pretreatment. The bioassay allowed differentiation between samples from males, females, and prepubertal boys, and all of the tested samples were above the assay detection limit.
For clinical applications, this bioassay appears to be superior to conventional assays for steroid hormones based on immunologic detection. This assay allows evaluation of the overall androgenic effect instead of measuring single androgenic compounds; thus, it can be used to estimate circulating androgen bioactivity. Another advantage of this yeast-based bioassay is the absence of expression of endogenous receptors, such as glucocorticoid or progesterone receptor, which might interfere with the screening. Such interference is present in cell-based AR assays that use the Chinese hamster ovary (CHO) cell line and breast cancer cell lines (14)(15). Hence, the developed bioassay is a more robust model for screening both androgenic and antiandrogenic compounds. Because of the short incubation time (2.5 h), the assay should be less affected by toxic effects from the matrix and thus can be used for different sample matrices without sample pretreatment. Because preanalytical procedures required in bioassays using mammalian cells or other reporter genes are avoided, less time is required (16). The use of modified luciferase simplifies the assay procedure, allowing 1-step measurement and increasing the bioassay sensitivity with respect to previously reported bioassays.
In view of the potentially extensive application of this assay, lyophilized strains could also be used to avoid the need for routine cultivation of the recombinant strains. This feature, associated with the 96-well plate format, makes the test even more suitable for different laboratories and for high-throughput screening.
Further applications of this bioassay include measurement of serum and urine androgen bioactivity in children and adults to investigate the relationship between androgenic status and endocrine diseases as well as the detection of illegally used synthetic androgens for human and animal doping.
Acknowledgments
The work was partly supported by the Italian Ministry for University and Research (MIUR) and by Fondazione Cassa di Risparmio di Bologna.
References
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
, DHT; 
, androstenedione. The recombinant yeast cells were incubated with increasing concentrations of various androgens to evaluate the transactivation capacity of each steroid. Each point represents the mean (SD; error bars) of at least 3 replicates. The dose–response curves were obtained from 5 independent experiments.