Biosensors and Bioelectronics 21 (2006) 2078–2085
A novel estrogen sensor based on recombinant Arxula adeninivorans cells Thomas Hahn a , Kristina Tag a , Klaus Riedel a , Steffen Uhlig b , Keith Baronian c , Gerd Gellissen d , Gotthard Kunze a,∗ a
Institut f¨ur Pflanzengenetik und Kulturpflanzenforschung, Corrensstr. 3, D-06466 Gatersleben, Germany b quo data GmbH, Siedlerweg 20, D-01465 Dresden-Langebr¨ uck, Germany c Christchurch Polytechnic Institute of Technology, City Campus, Madras Street, PO Box 540, Christchurch, New Zealand d PharmedArtis GmbH, Forckenbeckstr. 6, 52074 Aachen, Germany Received 7 June 2005; received in revised form 28 September 2005; accepted 27 October 2005 Available online 20 January 2006
Abstract A novel yeast cell-based assay was developed for the detection of estrogenic activity in wastewater. Recombinant Arxula adeninivorans strains were engineered to co-express the human estrogen receptor ␣ (hER␣) and a Klebsiella-derived phytase (phyK) reporter gene under the control of an A. adeninivorans-derived glucoamylase (GAA) promoter which had been modified by the insertion of estrogen-responsive elements (EREs). In the presence of estrogenic compounds, hER␣ dimerizes and binds to the estrogen. Reporter gene expression is induced by subsequent binding of the hER␣-dimer/estrogen complex to estrogen responsive elements (ERE) in the promoter. The insertion of different numbers of EREs in three alternative promoter positions and its effect on reporter gene expression were assessed. In one of the constructs, a detection limit of 5 ng l−1 and a determination limit of 10 ng l−1 for 17-estradiol-like activity was achieved. The photometric assay used enabled estrogen determination in sewage samples within 30 h. © 2005 Elsevier B.V. All rights reserved. Keywords: Arxula adeninivorans; Estrogen; Estrogen receptor ␣; Phytase; Yeast assay
1. Introduction Endocrine disruptors (EDs) are compounds with estrogenic, anti-estrogenic, androgenic or anti-androgenic effects. A prevalence of estrogenic compounds can disrupt reproduction of organisms, especially in aquatic environments. It is known that low concentrations of 17-estradiol induce vitellogenin synthesis in primary hepatocytes and promote feminization in fish species (Rankouhi et al., 2004). Furthermore, estrogens are considered to be weak carcinogens for mammalian species (Liehr, 2001; Yager and Liehr, 1996). EDs with estrogenic activity include phytoestrogens such as coumestrol and xenoestrogens such as bisphenol A. Despite their relatively low affinity for the human estrogen receptor ␣ (hER␣), they can elicit undesired effects when present in higher concentrations (Yamamoto et al., 2001). In wastewater of paper mills for example, bisphenol A ∗
Corresponding author. Tel.: +49 39482 5247; fax: +49 39482 5366. E-mail address:
[email protected] (G. Kunze).
0956-5663/$ – see front matter © 2005 Elsevier B.V. All rights reserved. doi:10.1016/j.bios.2005.10.019
can occur in high amounts (F¨urhacker et al., 2000; Mellanen et al., 1999) and monitoring of river water samples (Tamamoto et al., 2001) has revealed both the presence of high estrogenic activity and large seasonal variation. The vast majority of anthropogenic EDs can be found in domestic wastewater of sewage treatment plants (STP). Concentrations usually range between 2.5 and 25 ng l−1 of 17-estradiol equivalent activity in effluent wastewater (Korner et al., 1999) and can even be higher (Hegemann and Busch, 2000). Moreover synthetic estrogens such as 17␣-ethynylestradiol are relatively recalcitrant. Competition experiments have shown that 17␣-ethynylestradiol has 10 times the affinity for hER␣ than does 17-estradiol (Barkheim et al., 1998). Hohenblum et al. (2004) report levels of up to 0.94 ng l−1 17␣-ethynylestradiol in surface water and groundwater samples in Austria. These findings emphasise the need to monitor estrogenic activity in STPs. Analyses of estrogens in wastewater monitoring are based on chemical or biological assays. Estrogen screens based on
T. Hahn et al. / Biosensors and Bioelectronics 21 (2006) 2078–2085
MCF-7 breast cancer cells (E-screen) exhibit the highest sensitivity (Korner et al., 1999). This screening technique however, is not useful for measuring the influent from STPs since growth parameters of the individual cancer cells are variable. In addition the E-screen (MCF-7 cells) requires more than 4 days to complete. Other methods such as the enzyme-linked receptor assays (ELRA) are only useful in a laboratory setting and for detection of 17-estradiol concentrations greater than 20 ng l−1 (Seifert, 2004). One of the recently developed assays is based on recombinant Saccharomyces cerevisiae strain (S-YES). In this strain, the DNA sequence of the human estrogen receptor (hER) is integrated into the yeast genome and an expression plasmid carrying an E. coli-derived lacZ reporter gene controlled by a promoter containing estrogen-responsive sequences (EREs) is present in multiple copies. The production of the encoded galactosidase is induced in the presence of estrogens (Routledge and Sumpter, 1997). However, use of -galactosidase as a reporter enzyme is relatively problematic as the protein is not actively transported through the cytoplasmic membrane and the S. cerevisiae host is known to be sensitive to environmental fluctuations. To overcome these problems, an estrogen assay system was developed based on the more robust yeast strain Arxula adeninivorans LS3 (A-YES). This species is a non-pathogenic, ascomycetous, haploid, dimorphic yeast with unusual biochemical characteristics (Middelhoven et al., 1984, 1991, 1992; Van der Walt et al., 1990; Gienow et al., 1990). It is able to assimilate and ferment a very broad range of substrates as sole energy and carbon sources, for example, starch, tannic acid, adenine and ethanol. In addition this yeast is a thermoand osmotolerant organism and can be cultured at temperatures of up to 48 ◦ C in media containing up to 17.5% NaCl (Yang et al., 2000; Wartmann and Kunze, 2000). In the present study transgenic A. adeninivorans strains were constructed containing the human estrogen receptor (hER␣) gene and the phyK gene from Klebsiella sp. ASR1 encoding a secretory phytase gene product (Sajidan et al., 2004) as the reporter gene. 2. Materials and methods 2.1. Strains and media E. coli TOP 10 [F− mcrA (mrr-hsdRMS-mcrBC) 80lacZM15 lacX74 recA1 araD139 (ara leu) 7697 galU galK rspL (Str® ) endA1 nupG] from Invitrogen (USA), served as the host strain for bacterial transformation and plasmid isolation. LB medium supplemented with ampicillin (50 g ml−1 ; AppliChem, Germany) was used for the selection of transformants. A. adeninivorans G1211 (aleu2), an auxotrophic mutant, was used as the yeast host in this study (Samsonova et al., 1996). Yeast strains were grown either under non-selective conditions in a complex medium (YEPD) or under selective conditions in a yeast minimal medium (YMM) supplemented with 2% glucose (Tanaka et al., 1967; Rose et al., 1990). Cultivation was
2079
performed for 20–70 h at 30 ◦ C. Agar plates were prepared by adding 1.6% (w/v) agar. 2.2. Transformation procedures and isolation and characterization of nucleic acid E. coli strains were transformed according to the procedure developed by Hanahan (1983), and A. adeninivorans cells according to the procedure outlined by R¨osel and Kunze (1998). Plasmid DNA and restriction fragment isolation, labeling of fragments and Southern hybridization were carried out as previously described (Wartmann et al., 2002). 2.3. Plasmid constructions Inserting the hERα gene into A. adeninivorans G1211 involved flanking the ORF directly with EcoRI and BamHI cleavage sites by in vitro mutagenesis. The oligonucleotides 5 -GAATTCATGACCATGACCCTCCAC-3 (nucleotide positions −6–+18 with the EcoRI restriction site in bold type) and 5 -GGATCCGCCAGGGAGCTCTCAGA-3 (nucleotide positions 1790–1812 with the BamHI restriction site in bold type) were used as primers and the pHEGO-Hyg (ATCC) pDNA containing the hERα gene served as the template in the PCR. The EcoRI–BamHI-flanked hERα gene fragment was inserted into the plasmid pBS-TEF-PHO5 between the strong constitutive A. adeninivorans-derived TEF1 promoter and the S. cerevisiae-derived PHO5 terminator. The expression cassette TEF1 promoter-hERα gene-PHO5 terminator thus constructed was then inserted into the basic pAL-HPH1 plasmid and the resulting construct, pAL-HPH-hER␣, was used to transform A. adeninivorans G1211 (R¨osel and Kunze, 1998; Wartmann et al., 2002). The phyK gene from Klebsiella sp. ASR1 (Sajidan et al., 2004) was selected as a reporter gene for A-YES. The vector was constructed by replacing the SalI–EcoRI flanked TEF1 promoter of the plasmid pBS-TEF-phyK-PHO5 (obtained from Borriss) with a SalI–EcoRI-flanked A. adeninivorans-derived GAA promoter (Bui et al., 1996) to form pBS-GAA-phyKPHO5. Subsequently the expression cassette GAA promoterphyK gene-PHO5 terminator of this plasmid was inserted into the plasmid pAL-ALEU2m to form pAL-ALEU2m-GAA-phyK and this construct was used to transform A. adeninivorans G1211 and A. adeninivorans G1211/pAL-HPH-hER␣ (Wartmann et al., 2003). 2.4. Modification of the GAA promoter EREs with the nucleotide sequence AGGTCAGAGTGACCT were selectively retrieved from chicken vitellogenin gene (Klinge, 2001) and integrated into positions −107, −150 and −203 of the GAA promoter by PCR and slicing by overlap extension (SOE) mutagenesis using the primers shown in Table 1 (Horton et al., 1989). The resulting DNA fragments with one or two EREs at position −107, one ERE at position −150 and one, two or three EREs at position −203 were digested with SalI–EcoRI, gel-purified,
2080
T. Hahn et al. / Biosensors and Bioelectronics 21 (2006) 2078–2085
Table 1 Oligonucleotide primers of PCR and SOE mutagenesis for construction of modified GAA promoters with ERE region inserted at positions −107, −150 and −203 Primer
Sequence
Position number
Comments
GAASER1 GAASER2 GAA107a GAA107b GAA150a GAA150b GAA203a GAA203b MFERE107a MFERE107b MFERE150a MFERE150b MFERE203a MFERE203b
5 -GTCGACCACAGACCAGACGG-3
−638 to −618 −1 to −23 −108 to −142 −122 to −91 −151 to −185 −165 to −141 −204 to −241 −218 to −184 −107 to −143 −137 to −101 −150 to −186 −180 to −144 −203 to −165 −233 to −196
5 -end of the GAA-pro. 3 -end of the GAA-pro. Insert ERE by −107 Insert ERE by −107 Insert ERE by −150 Insert ERE by −150 Insert ERE by −203 Insert ERE by −203 Insert 2xERE or 3xERE by −107 Insert 2xERE or 3xERE by −107 Insert 2xERE or 3xERE by −150 Insert 2xERE or 3xERE by −150 Insert 2xERE or 3xERE by −203 Insert 2xERE or 3xERE by −203
5 -GAATTCCATTCTGTTTCGTCCAC-3 5 -AGGTCACTCTGACCTTCCCTTACACTTCCAAGTCG-3 5 -AGGTCAGAGTGACCTGGGGATCCAACTCGGCC-3 5 -AGGTCACTCTGACCTTGCATGTTCCAGGGTCCAAA-3 5 -AGGTCAGAGTGACCTTAATGCGGAGATCTCACTCG-3 5 -AGGTCACTCTGACCTTGAGATAACAGTGAAAACGTTG-3 5 -AGGTCAGAGTGACCTTTTATCCCCGGATTGGCCAA-3 5 -AGGTCACTCTGACCTAGGTCACTCTGACCTTCCCTT-3 5 -AGGTCAGAGTGACCTAGGTCAGAGTGACCTGGGGAT-3 5 -AGGTCACTCTGACCTAGGTCACTCTGACCTTGCATG-3 5 -AGGTCAGAGTGACCTAGGTCAGAGTGACCTTAATGC-3 5 -AGGTCACTCTGACCTAGGTCACTCTGACCTTGAGATAA-3 5 -AGGTCAGAGTGACCTAGGTCAGAGTGACCTTTTATCC-3
EREs are underlined.
and subcloned into the SalI- and EcoRI-digested plasmid pBS-GAA-phyK-PHO5 to replace the existing GAA promoter with the GAA(xERE) promoter. Subsequently the expression cassettes containing the various GAA(xERE) promoters— phyK gene–PHO5 terminator fusions were integrated into the SalI–ApaI sizes of plasmid pAL-ALEU2m (Wartmann et al., 2003). 2.5. Protein analysis Preparation of protein extracts, SDS-PAGE, and Western blot analysis of hER␣-producing strains were carried out as described by Wartmann et al. (2003). The dye-binding method of Bradford (1976) was used for quantitative determination of the protein concentration in cell extracts with bovine serum albumin serving as the standard protein. 2.6. Estrogenic activity assay A. adeninivorans G1211 transformed with the plasmids pAL-HPH-hER␣ and pAL-ALEU2m-GAA(xERE)-phyK (G1211/pAL-HPH-hER␣-pAL-ALEU2m-GAA(xERE)-phyK) was the bio component of the A-YES bio-assay. These cells were cultured in YMM with 2% glucose at 30 ◦ C for 48 h and stored in 50 l aliquots. To detect estrogenic activity, 950 l of the sample containing 2% maltose (final concentration) and 50 l stationary phase yeast cell suspension (final OD600 nm = 1) was incubated for 30 h at 30 ◦ C and 140 rpm orbital shaking. At the end of incubation, all samples were centrifuged for 10 min at 5000 × g to separate the cells from the cultivation medium. Extracellular phytase activity was assessed using a modification of the method described by B¨uttner et al. (1991) and Sajidan et al. (2004). Aliquots (25 l) of the supernatant were incubated with 25 l 0.1 M sodium-citrate (pH 3.9) containing 3.5 mM 4-nitrophenyl phosphate for 60 min at 37 ◦ C in microtitre plates. The reaction was stopped by adding 50 l of 15% TCA. After the addition of 100 l 1N sodium hydroxide,
the absorbance of the solution was measured at 405 nm with a “Sunrise” reader (Tecan Trading AG, Z¨urich, Switzerland). 2.7. Statistical analysis In order to compare eight different estrogen responsive transformants, the phytase activity was monitored in three separate experiments with at least three measurements per experiment. Statistical analysis was performed using an analysis of variance (ANOVA) and pairwise t-tests. Differences of phytase activity were significant at significance level 0.01. 3. Results 3.1. Expression of the hERα gene in A. adeninivorans The plasmid pAL-HPH-hER␣ (Fig. 1A) containing the hERα gene under control of the strong constitutive TEF1 promoter was linearized and used to transform the auxotrophic mutant A. adeninivorans strain, G1211. The resulting recombinant strain G1211/pAL-HPH-hER␣ contains a single copy of the heterologous DNA integrated into the 25S rDNA (data not shown). Transformed cells were cultured in YMM glucose medium for 48 h at 30 ◦ C to provide appropriate conditions for high TEF1 promoter-controlled gene expression. Cell extracts as well as culture medium samples were analysed for the presence of hER␣. The recombinant protein was found in the intracellular fraction with similar levels of hER␣ in all selected transformants (data not shown). 3.2. Expression of the phyK gene in A. adeninivorans Recombinant A. adeninivorans carrying the Klebisielladerived phyK reporter gene (Sajidan et al., 2004) was constructed as follows. The plasmid pAL-ALEU2m-GAA-phyK (Fig. 1B) containing the inducible GAA promoter for expression control was linearized and used to transform A. adeninivorans strain G1211. The transformants, G1211/pAL-ALEU2m-GAA-phyK,
T. Hahn et al. / Biosensors and Bioelectronics 21 (2006) 2078–2085
2081
Fig. 1. Physical maps of the expression/integration vectors: (A) pAL-HPH-hER␣ and (B) pAL-ALEU2m-GAA-phyK. The vectors contain the 25S rDNA sequence (rDNA) and the selection markers ALEU2m (pAL-ALEU2m) and hph (pAL-HPH1). Both vectors further contain unique ApaI and SalI restriction sites for the insertion of expression cassettes, and unique BplI and MluI sites within the rDNA sequence for linearization. The hERα and phyK expression cassettes used in this study were inserted as TEF1/GAA promoter-hERα/phyK gene-PHO5 terminator elements.
contained a single copy of the recombinant DNA integrated into the 25S rDNA (data not shown). Transformed cells were cultured in YMM glucose and YMM maltose medium for up to 48 h at 30 ◦ C and cell extracts and culture medium samples were analysed for the presence of phytase. The YMM maltose culture medium contained up to 21 kat ml−1 recombinant enzyme whereas cell extracts from cells grown in YMM maltose had only trace amounts (
