Estimating Detection Limits of Potentiometric DNA Sensors Using Surface Plasmon Resonance Analyses

As the signals of potentiometric-based DNA ion-selective field effect transistor (ISFET) sensors differ largely from report to report, a systematic revisit to this method is needed. Herein, the hybridization of the target and the probe DNA on the sensor surface and its dependence on the surface probe DNA coverage and the ionic strength were systematically investigated by surface plasmon resonance (SPR). The maximum potentiometric DNA hybridization signal that could be registered by an ISFET sensor was estimated based on the SPR measurements, without considering buffering effects from any side interaction on the sensing electrode. We found that under physiological solutions (200 to 300 mM ionic strength), the ISFET sensor could not register the DNA hybridization events on the sensor surface due to Debye screening. Lowering the salt concentration to enlarge the Debye length would at the same time reduce the surface hybridization efficiency, thus suppressing the signal. This adverse effect of low salt concentration on the hybridization efficiency was also found to be more significant on the surface with higher probe coverage due to steric hindrance. With the method of diluting buffer, the maximum potentiometric signal generated by the DNA hybridization was estimated to be only around 120 mV with the lowest detection limit of 30 nM, occurring on a surface with optimized probe coverage and in the tris buffer with 10 mM NaCl. An alternative method would be to achieve high-efficiency hybridization in the buffer with high salt concentration (1 M NaCl) and then to perform potentiometric measurements in the buffer with low salt concentration (1 mM NaCl). Based on the characterization of the stability of the hybridized DNA duplexes on the sensor surface in low salt concentration buffer solutions, the estimated maximum potentiometric signal could be significantly higher using the alternative method. The lowest detection limit for this alternative method was estimated to be around 0.6 nM. This work can serve as an important quantitative reference for potentiometric DNA sensors.