Sensing Estrogen with Electrochemical Impedance Spectroscopy.

While rapid industrialization in the 20th century has introduced a significant degree of convenience to human life, the global population is being exposed to a polluted environment that includes contaminated air such as carbon monoxide, sulfur dioxide, dioxin, formaldehyde, chlorine, bisphenol, and estradiol [ 1 ]. Among the pollutants, 4-nonylphenol, bisphenol, and estradiol are classified as endocrine disrupting chemicals (EDCs) and these EDCs are known to be related to various diseases of infertility, spontaneous abortions, birth defects, endometriosis, breast cancer, and so forth [ 2 – 8 ]. Several papers have been published on the detection of EDCs in solution phase using methods such as fluorimetry, NMR spectroscopy, chromatography, enzyme linked receptor assays, and linear sweep voltammetry [ 9 – 17 ]. Recently, a few researchers studied the detection of EDCs using gas chromatography-mass spectrometry and found 4~8 ng/L of EDCs in gas phase [ 18 ]. Therefore, the development of EDC detection and measurement technique is in a great demand, particularly, at very low concentration (0.2~141 ng/L of 17 β -estradiol exists in environment) in gas phase since the extremely low amount of EDCs may cause the fatal diseases to human [ 19 ]. Nevertheless, most people are not aware of the seriousness of EDCs in air. Notably, inhalation in an EDC exposed environment is an easy means of being affected by EDCs. Therefore, one of the main challenges in research is to develop a sensitive and reliable method for determination of EDCs in gas phase. In the present study, a highly sensitive electrochemical biosensor for the detection of the EDC, 17 β -estradiol, in gas phase has been developed using estrogen receptor- α immobilized on a gold electrode. 17 β -estradiol in gas phase was detected by electrochemical impedance spectroscopy of Fe(CN) 6 3− /Fe(CN) 6 4− at the fabricated biosensor. The present biosensor gives a linear response ( R 2 = 0.999) for the 17 β -estradiol vapor concentration from 3.7 ng/L to 3.7 × 10 −4  ng/L. This study is aimed at developing a technique capable of detecting low EDC concentration (>3.7 × 10 −4  ng/L) in gas phase.

Estrogen receptor- α (ER- α , >85.0%) was purchased from Calbiochem (San Diego, USA). Estrogen hormone (17 β -estradiol, >98%), 3-mercaptopropionic acid (3-MPA, >99%), 3-(3-dimethylaminopropyl)-1-ethylcarbodiimide hydrochloride (EDC, >98%), N-hydroxysuccinimide (NHS, >98%), and bovine serum albumin (BSA, >96%) were purchased from Aldrich. A phosphate buffered saline (PBS) was purchased from Gibco (NY, USA). For the gas phase experiment, the estrogen biosensor was prepared as previously reported [ 20 , 21 ]. The fabricated estrogen biosensor was applied to the detection of 17 β -estradiol in gas phase. In the experiment, a freshly pretreated gold electrode was incubated in 40 mM 3-MPA solution for 4 h to form a carboxylate terminated self-assembled monolayer (SAM) on the electrode. The electrode was then rinsed using distilled water three times to remove any nonphysically adsorbed 3-MPA molecules before it was stored in distilled water. The 3-MPA modified gold electrode was immersed in a mixture of EDC (1.0 wt.%) and NHS (1.0 wt.%) in 0.05 M PBS (pH 7.4) for 2 h to activate the terminal carboxylic groups, before it was washed using distilled water to remove any nonbonded residual EDC/NHS. Next, a 15  μ L aliquot of 0.67  μ M estrogen receptor- α solution was applied to the surface of the EDC-NHS/3-MPA modified gold electrode. Any nonbonded estrogen receptor- α on the EDC-NHS/3-MPA modified gold electrode was removed by rinsing using PBS. Finally, the electrode was incubated in 3.5 wt.% BSA for 4 h to block any nonspecific binding sites. Then the biosensor was dried at 25°C for 4 h and then fixed in the empty headspace of a 10 mL vial which contained only 2 mL of 17 β -estradiol solution (0.01~100 mM) at the bottom of the vial in order to avoid direct contact between the biosensor and the solution. The vial was stirred at 80 rpm for 90 min (25°C, pH 7). The process was conducted in a sealed vial [ 22 – 24 ]. The schematic diagram of the biosensor fabrication and 17 β -estradiol binding is illustrated in Scheme 1 .

Electrochemical impedance experiments were performed using an EG&G 263A potentiostat (Princeton, NJ, USA). A 15 mL electrochemical cell that accommodates a platinum wire counter electrode, the estrogen biosensor as the working electrode, and a Ag/AgCl (3 M NaCl) reference electrode was used. AC impedance spectra were recorded in 0.05 M PBS (pH 7.4) containing 5.0 mM Fe(CN) 6 3− and 5.0 mM Fe(CN) 6 4− . The impedance spectra were obtained in the frequency range from 100 MHz to 100 kHz at a bias potential of 0.2 V that was superimposed by an ac potential of 5.0 mV peak-to-peak amplitude. The impedance spectra were plotted in the form of impedance plane plots (Nyquist plots). Zsimpwin program (Princeton Applied Research, Oak Ridge, TN, USA) was used to fit theoretical impedance plots generated based on an equivalent circuit (depicted in Figure 1(c) ) to those obtained experimentally. In this way, the diameter of a semicircle in the high frequency range of Nyquist plot was estimated [ 20 , 21 ].

For the detection of 17 β -estradiol in gas phase, the estrogen biosensor was exposed to 17 β -estradiol vapor generated by solutions of concentration between 0.01 and 100 mM for 90 min. Based on Henry's Law, the concentration of 17 β -estradiol vapor is proportional to that of aqueous 17 β -estradiol. Henry's Law constant ( k H ) of 17 β -estradiol is known to be 3.64 × 10 −11  L·atm/mol [ 25 ]. SI unit of Henry's Law constant ( k H ) was converted to unit of M/atm and the value is calculated to be 2.7 × 10 10  M/atm, and the corresponding concentration of gaseous 17 β -estradiol was estimated to be within the range between 0.00037 ng/L and 3.7 ng/L. As shown in Figure 1(a) , the diameter of Nyquist plots, which corresponds to the electron transfer resistance, linearly increases with 17 β -estradiol concentration. This is because, with increasing 17 β -estradiol concentration, a corresponding increased quantity of 17 β -estradiol in the vapor was expected to bind to the immobilized estrogen receptor- α , which would in turn gradually hinder the diffusion of [Fe(CN) 6 ] 3− /[Fe(CN) 6 ] 4− to the biosensor surface. We assume that the 17 β -estradiol vapor was well bonded to the ER- α on the biosensor surface. In order to normalize the data, the impedance change, Δ R et = R et ( i ) − R et (0), was used in the data analysis, where R et ( i ) and R et (0) represent the electron transfer resistance after and before the binding of 17 β -estradiol to the estrogen biosensor, respectively. As shown in Figure 1(b) , the calibration plot was obtained by graphing the impedance change as a function of the 17 β -estradiol concentration in gas phase. A linear regression equation was obtained as Δ R et = 2195 log⁡[17 β -estradiol] + 4048 ( R 2 = 0.999, S / N = 3, n = 3). The limit of quantification of the present estrogen biosensor was determined to be 3.7 × 10 −4  ng/L of 17 β -estradiol. The resistance of biosensor was measured with a CPE of electrode since the component has incorporated both the Helmholtz double layer and surface roughness or heterogeneity of the electrode. The values of R s were fixed at 80 ohm as shown in Table 1 . CPE and W were determined from the fitting program (ZsimpWin fitting software). The values of R et were compensated by the fitting software because the semicircle did not reach to the axis of abscissa in the low frequency region. The fitted values were presented in Table 1 . Association constant ( K a ) of the binding reaction between 17 β -estradiol and estrogen receptor- α was determined using a Langmuir isotherm approach as follows [ 26 ]: (1) θ = K a × C 1 + K a × C , where θ is the fractional coverage of the surface and C is the concentration of 17 β -estradiol vapor. Also, θ is expressed by the following equation between R et ( i ) and R et (0) [ 27 ]: (2) θ = R e t i − R e t 0 R e t i = 1 − R e t 0 R e t i , where R et ( i ) and R et (0) represent the electron transfer resistance after and before the binding of 17 β -estradiol to the estrogen biosensor, respectively. Above equations are further rearranged to (3) C R e t i − R e t 0 = 1 R e t i C + 1 K a × R e t i . Accordingly, the ordinate intercept of a linear C /Δ R et versus C plot (as shown in Figure 2 ) will provide an estimate for K a × R et ( i ), through which K a can be evaluated. Figure 2 reveals that C /Δ R et linearly changes with the concentration, which results in a linear equation of C /Δ R et = 8.56 × 10 −5 C + 2.63 × 10 −6 ( R 2 = 0.999, N = 3). From the fitting of the equation, the association constant was observed to be ca. 32.5 L/ng (8.84 × 10 12  M −1 ). The value of K a is larger than the reported value of 5.2 × 10 8  M −1 [ 28 ]. The result implied that the binding sites on the ER- α were well maintained after the introduction of ER- α to the surface of modified electrode. The identical biosensing measurements were carried out with corticosterone and dexamethasone that have the similar chemical structures without the bonding characteristics to 17 β -estradiol [ 29 ]. The selectivity values were obtained by comparing the impedance values at the high concentrations (3.7 × 10 −1  ng/L) of corticosterone and dexamethasone with those for 17 β -estradiol. As shown in Figure 3 , the selectivity values for corticosterone and dexamethasone are 9.44 ± 2.2% and 5.05 ± 2.7% due to nonspecific adsorption to the biosensor [ 30 ]. The result indicates that the fabricated estrogen biosensor in this study can be used for the selective estimation of 17 β -estradiol in gas phase.

In this study, the electrochemical impedance biosensor was successfully developed by immobilizing ER- α on a Au electrode and demonstrated for the detection of the gas phase of 17 β -estradiol. It is considered that such study may open the possibility of the detector development for the detection of EDCs in gas phase.