Virus-Enabled Biosensor for Human Serum Albumin

Abstract

The label-free detection of human serum albumin (HSA) in aqueous buffer is demonstrated using a simple, monolithic, two-electrode electrochemical biosensor. In this device, both millimeter-scale electrodes are coated with a thin layer of a composite containing M13 virus particles and the electronically conductive polymer poly(3,4-ethylenedioxy thiophene) or PEDOT. These virus particles, engineered to selectively bind HSA, serve as receptors in this biosensor. The resistance component of the electrical impedance, Z_re, measured between these two electrodes provides electrical transduction of HSA binding to the virus-PEDOT film. The analysis of sample volumes as small as 50 μL is made possible using a microfluidic cell. Upon exposure to HSA, virus-PEDOT films show a prompt increase in Z_re within 5 s and a stable Z_re signal within 15 min. HSA concentrations in the range from 100 nM to 5 μM are detectable. Sensor-to-sensor reproducibility of the HSA measurement is characterized by a coefficient-of-variance (COV) ranging from 2% to 8% across this entire concentration range. In addition, virus-PEDOT sensors successfully detected HSA in synthetic urine solutions.

Figure 1

Schematic diagrams of the biosensor and the flow cell described in this study. (a) An assembled flow cell attached to a gold-electrode device consists of two gold contacts connected to a potentiostat for EIS measurements. (b) A gold-electrode device and detailed structure of a single PMMA flow cell; (c) a single device with a red box showing the two planar gold electrodes used for sensing. The two gold electrodes have a length (L) of 2 mm, width (w) of 0.85 mm, and are separated by a 50 µm gap. (d) Dimensions of the first PMMA flow cell layer which creates a cell holding 6 µL of solution over the gold electrodes; (e) Top view representation of assembled flow cell. Solution flows from the inlet port (right), through the cell, and exits through the outlet port (left) into a reservoir with a 75 µL capacity.

Figure 2

Electrodeposition and SEM characterization of virus-PEDOT bioaffinity coatings. (a) Electrodeposition of a virus-PEDOT film by cyclic voltammetry. Film prepared by two cycles in aqueous EDOT solution (2.5 mM EDOT, 12.5 mM LiClO₄) followed by eight cycles in a virus-EDOT solution (2.5 mM EDOT, 12.5 mM LiClO₄, 8 nM HSA virus). Virus-EDOT solution was replenished every two cycles. All scan rates were 20 mV/s. Optical image of: (b) bare gold electrodes and (c) gold electrodes after electrodeposition of virus-PEDOT film. (d,f,e,g) Scanning electron microscopy images of uncoated films. (d) PEDOT film prepared by ten consecutive cycles of deposition in aqueous EDOT solution (2.5 mM EDOT, 12.5 mM LiClO₄). (e) PEDOT edge showing film height of approximately 220 nm. (f) Virus-PEDOT film prepared as described in (a) showing dense incorporation of virus bundles on the surface. (g) Virus-PEDOT edge showing primer layer of PEDOT with thickness of approximately 160 nm and PEDOT-coated virus on top.

Figure 3

Atomic force microscopy of virus-PEDOT bioaffinity films and AFM line scans shown at the bottom. (a) PEDOT-only film prepared by ten cycles of deposition in EDOT solution (2.5 mM EDOT, 12.5 mM LiClO₄). Topography of the middle (left) and the edge (right) of films imaged by atomic force microscopy. The film-edge height shown in line scans includes the gold electrode layer (60 nm). (b) Virus-PEDOT film prepared by two cycles of deposition in EDOT solution followed by eight cycles in virus-EDOT solution (2.5 mM EDOT, 12.5 mM LiClO₄, 8 nM HSA virus); virus-EDOT solution replenished every two cycles. The rms roughness for PEDOT and virus-PEDOT films is ≈10 nm and ≈150 nm, respectively.

Figure 4

Detection of HSA binding using electrochemical impedance spectroscopy (EIS). The EIS response of virus-PEDOT biosensors upon exposure to 500 nM BSA (blue) and 500 nM HSA (red) is compared. No redox species are added to the solution in these measurements. Error bars represent the standard deviation, ±1σ, of five consecutive EIS measurements on a single electrode. (a,b) Nyquist plots for virus-PEDOT films in solutions of run buffer (black) and 500 nM BSA or HSA. Plots of (c) ΔZ_re and (d) ΔZ_im versus frequency, where ΔZ is defined as Z_analyte − Z_buffer. Corresponding (e) ΔZ_re and (f) ΔZ_im signal-to-noise ratio, defined as ΔZ/σ, as a function of frequency.

Figure 5

Real-time HSA biosensing. Plot of ΔZ_re verus time, of a single virus-PEDOT, using a control virus that did not bind HSA (blue) and HSA virus (green), electrode when exposed to three concentrations of HSA. A freshly electrodeposited virus-PEDOT film was first immersed in run buffer (PBS-casein-tween) until reaching an equilibration signal. The time scan was then paused and five EIS spectra were acquired in rapid succession. Immediately following this, the virus-PEDOT film was exposed to 100 nM HSA in run buffer and the time scan was restarted within 5 seconds of exposure. This procedure was repeated for exposures to 500 nM and 5000 nM HSA.

Figure 6

Sensor-to-sensor reproducibility of HSA detection. Calibration plot of (a) ΔZ_re, and (b) ΔZ_im versus frequency for multiple virus-PEDOT films exposed to varying concentrations of HSA. ΔZ values for n = 3 independent virus-PEDOT electrodes were averaged to obtain each curve, errors bars indicate, ±1σ. Corresponding coefficient of variation, defined as the relative standard deviation for n=3 virus-PEDOT electrodes of (c) ΔZ_re and (d) ΔZ_im versus frequency plots for each HSA concentration. ΔZ_re shows regions of COV values < 20 % while, ΔZ_im COV values are too high for reliable measurements. At each frequency, ΔZ_re versus [HSA] was fitted to the Hill equation and the square of the regression coefficient, R², versus frequency plot (e). R² = 1 represents the best fit of the Hill equation to the data. The highlighted interval in (a,c,e) indicates the frequency range where ΔZ_re signal is largest, COV is at a minimum, and the peak for goodness of fit occur, respectively. (f) Calibration plot of ΔZ_re, measured at 340 Hz, versus concentration. Each data point represents a different virus-PEDOT electrode with error bars defined as the standard deviation, ±1σ, of five consecutive impedance measurements. Impedance data for HSA exposures to virus-PEDOT films containing HSA (red) virus are fitted to the hill equation (red line). Three controls to confirm specific binding to HAS are shown: BSA exposure to virus-PEDOT films containing HSA binding virus (blue), HSA exposure to virus-PEDOT films containing a control virus having no affinity for HSA (green), and HSA exposure to pure PEDOT films containing no virus (black).

Figure 7

Virus-PEDOT sensors in synthetic urine. Calibration plot of (a) ΔZ_re versus frequency for multiple virus-PEDOT films exposed to varying concentrations of HSA. ΔZ values for n = 3 independent virus-PEDOT electrodes were averaged to obtain each curve, errors bars indicate, ±1σ. (b) Corresponding coefficient of variation (COV), defined as the relative standard deviation for n=3 virus-PEDOT electrodes of ΔZ_re versus frequency plots for each HSA concentration. (c) Calibration plot of ΔZ_re, measured at 136 Hz, versus concentration. Each data point represents an independent virus-PEDOT electrode with error bars defined as the standard deviation, ±1σ, of five consecutive impedance measurements. Impedance data for HSA exposures to virus-PEDOT films containing HSA (red) virus are fitted to the hill equation (red line). Three controls to confirm specific binding to HSA are shown: BSA exposure to virus-PEDOT films containing HAS virus (blue), HSA exposure to virus-PEDOT films containing a control virus that did not bind HSA (green), and HSA exposure to pure PEDOT films containing no virus (black).