Advancements in Label-free Biosensing Using Field-Effect Transistors and Aided by Molecular Dynamics Simulations

Thumbnail Image
Publication or External Link
Guros, Nicholas
Klauda, Jeffery B
Balijepalli, Arvind
Biosensors are used to characterize or measure concentrations of physiologically or pathologically significant biomarkers that indicate the health status of a patient, for example, a biomarker associated with a specific disease or cancer. Presently, there is a need to improve the capabilities of biosensors, which includes their rate of detection, limit of detection, and usability. With respect to usability, it is advantageous to develop biosensors that can detect a biomarker that is not labeled, such as with a conventional fluorescent, magnetic, or radioactive label, prior to characterization or measurement by that biosensor. Such biosensors are known as label-free biosensors and are the primary focus of this work. Biosensors are principally evaluated by two standards: their sensitivity to detect a target biomarker at physiologically relevant concentrations and their specificity to detect only the target biomarker in the presence of other molecules. The elements of biosensing critical to improving these two standards are: biorecognition of the biomarker, immobilization of the biorecognition element on the biosensor, and transduction of biomarker biorecognition to a measurable signal. Towards the improvement of sensitivity, electrostatically sensitive field-effect transistors (FET) were fabricated in a dual-gate configuration to enable label-free biosensing measurements with both high sensitivity and signal-to-noise ratio (SNR). This high performance, quantified with several metrics, was principally achieved by performing a novel annealing process that improved the quality of the FET’s semiconducting channel. These FETs were gated with either a conventional oxide or an ionic liquid, the latter of which yielded quantum capacitance-limited devices. Both were used to measure the activity of the enzyme cyclin-dependent kinase 5 (Cdk5) indirectly through pH change, where the ionic-liquid gated FETs measured pH changes at a sensitivity of approximately 75 times higher than the conventional sensitivity limit for pH measurements. Lastly, these FETs were also used to detect the presence of the protein streptavidin through immobilization of a streptavidin-binding biomolecule, biotin, to the FET sensing surface. To study the biomolecular factors that govern the specificity of biomarker biorecognition in label-free biosensing, molecular dynamics (MD) simulations were performed on several proteins. MD simulations were first performed on the serotonin receptor and ion channel, 5-HT3A. These simulations, which were performed for an order of magnitude longer than any previous study, demonstrate the dynamic nature of serotonin (5-HT) binding with 5-HT3A. These simulations also demonstrate the importance of using complex lipid membranes to immobilize 5-HT3A for biosensing applications to adequately replicate native protein function. The importance of lipid composition was further demonstrated using MD simulations of the ion channel alpha-hemolysin (αHL). The results of these simulations clearly demonstrate the lipid-protein structure-function relationship that regulates the ionic current though a lipid membrane-spanning ion channel. Finally, to demonstrate the impact of MD simulations to inform the design of FET biosensing, a strategy to use FETs to measure the ultra-low ionic currents through the ion channel 5-HT3A is outlined. This strategy leverages critical elements of 5-HT biorecognition and ion channel immobilization extracted from MD simulations for the design of the proposed FET sensing surface interface.