High-Performance Conducting Polymer Nanotube-based Liquid-Ion Gated Field-Effect Transistor Aptasensor for Dopamine Exocytosis

In this study, ultrasensitive and precise detection of a representative brain hormone, dopamine (DA), was demonstrated using functional conducting polymer nanotubes modified with aptamers. A high-performance aptasensor was composed of interdigitated microelectrodes (IMEs), carboxylated polypyrrole nanotubes (CPNTs) and DA-specific aptamers. The biosensors were constructed by sequential conjugation of CPNTs and aptamer molecules on the IMEs, and the substrate was integrated into a liquid-ion gating system surrounded by pH 7.4 buffer as an electrolyte. To confirm DA exocytosis based on aptasensors, DA sensitivity and selectivity were monitored using liquid-ion gated field-effect transistors (FETs). The minimum detection level (MDL; 100 pM) of the aptasensors was determined, and their MDL was optimized by controlling the diameter of the CPNTs owing to their different capacities for aptamer introduction. The MDL of CPNT aptasensors is sufficient for discriminating between healthy and unhealthy individuals because the total DA concentration in the blood of normal person is generally determined to be ca. 0.5 to 6.2 ng/mL (3.9 to 40.5 nM) by high-performance liquid chromatography (HPLC) (this information was obtained from a guidebook “Evidence-Based Medicine 2018 SCL “ which was published by Seoul Clinical Laboratory). The CPNTs with the smaller diameters (CPNT2: ca. 120 nm) showed 100 times higher sensitivity and selectivity than the wider CPNTs (CPNT1: ca. 200 nm). Moreover, the aptasensors based on CPNTs had excellent DA discrimination in the presence of various neurotransmitters. Based on the excellent sensing properties of these aptasensors, the DA levels of exogeneous DA samples that were prepared from PC12 cells by a DA release assay were successfully measured by DA kits, and the aptasensor sensing properties were compared to those of standard DA reagents. Finally, the real-time response values to the various exogeneous DA release levels were similar to those of a standard DA aptasensor. Therefore, CPNT-based aptasensors provide efficient and rapid DA screening for neuron-mediated genetic diseases such as Parkinson’s disease.


Results and Discussion
CPNT-based aptasensor for DA exocytosis. We designed an electrical methodology for real-time exogeneous DA release detection. Figure 1 shows a representative illustration of the cell-related DA exocytosis process and an FET aptasensor for exogeneous DA monitoring. The DA in cells was released by Ca 2+ transport through the calcium ion channel, and the Ca 2+ influx was controlled by the KCl concentration via DA release assay. The advantage of KCl addition is that the rate of DA release is rapid compared to the rate in the presence of only Ca 2+ owing to the enhanced Ca 2+ influx effect 44 . The exogeneous DA samples were prepared and monitored by liquid-ion gated FET aptasensors consisting of an IME array, CPNTs and aptamers. First, the IME array was prepared by a microelectromechanical system (MEMS) (Fig. 2a). Specifically, the photoresistor was spin-coated onto a glass wafer and exposed to UV light using a patterned photomask. Subsequently, a pattern was developed by the developer, and Au was deposited on the substrate with Cr as a binder by thermal evaporation. A final pattern was obtained after a simple lift-off. A photograph and scanning electron microscopy (SEM) images of the IME array are also presented in Fig. 2b. The Au lines bridged electrode pads, and active IMEs were passivated with a photoresistor to protect against nonspecific binding with nontarget biomolecules and liquid-surrounding contaminants.
Generally, the nanomaterials functioning as transistors in biosensors were adsorbed on the IME array because the drop-casting and spin-coating processes are simple and easy. However, the adsorption of nanomaterials on the IME array is easily affected by the surrounding environment, such as liquid fluctuations, and the nanomaterials are detached on the IME array during cleaning or washing processes for introducing aptamers, leading to low reproducibility and sensitivity. Compared to conventional adsorption processes of nanomaterials, in this study, the CPNTs were immobilized on the IME array, which was functionalized with the -NH group by silane coupling agents 45 . The overall construction protocol of the aptasensor is illustrated in Fig. 3. CPNTs were chemically immobilized on the silane-treated IME arrays by a condensation reaction with DMT-MM, resulting in the stable electrical contact between CPNTs and IMEs. The extended illustration in Fig. 3 shows the ideal aptasensor consisting of immobilized CPNT bridges between the source and drain electrodes. The aptamer, which was benchmarked by a previous study, was chemically attached on the surface of CPNTs by the same reaction 46,47 . This benchmarked aptamer consists of DNA sequences, not RNA [48][49][50] . DNA aptamers have attractive advantages compared to RNA aptamers: 1) they are more stable under various environmental conditions, and 2) they have improved affinity owing to superior DA binding sites. Additionally, DMT-MM, which is representative of chemical reagents used to chemically bind the carboxylic group of nanomaterials and the amine group of bioprobes, was used as a conjugation reagent.
To confirm the formation of the aptasensors, successful production of CPNTs ca. 200 nm in diameter (CPNT1) and ca. 120 nm in diameter (CPNT2) on the IME array was confirmed by the SEM images in Fig. 4a,b. While the Figure 1. Schematic illustration of cell exocytosis for DA release from PC12 cells via rapid Ca 2+ reflux accelerated by K + ions (upper) and liquid-ion gated FET aptasensors using aptamer-conjugated CPNTs for exogeneous DA detection (down). (2020) 10:3772 | https://doi.org/10.1038/s41598-020-60715-x www.nature.com/scientificreports www.nature.com/scientificreports/ reactivity of the Py-COOH monomer was significantly low for oxidative polymerization due to the presence of functional groups such as COOH, the addition of Py co-monomer accelerated the electrochemical polymerization 51 . Once the initiation step was triggered by iron chloride salt, the following reaction could be spontaneous and fast. The transmission electron microscopy (TEM) images in Fig. 4 clearly show that the morphology of the produced CPNTs is tubular and that each diameter is different (insets in Fig. 4a,b). It was obvious that a sufficient amount of CPNTs with different diameters was obtained by simple polymerization. This was important because an ample number of sites existed for the specific interactions between nanotubes (NTs), condensing agent, and aptamer molecules. It was predictable that the performance of the sensor was strongly dependent on the uniformity and dimension of the CPNTs. As the shape of the CPNTs was dominantly one-directional, the spacing between these NTs on the substrate was distant. Therefore, we supplied a sufficient amount of NTs to the electrode to guarantee intimate contact between NTs and the electrode. This was advantageous because the capacity of the CPNTs enabled to determine the sensitivity and selectivity of aptasensor.
To confirm the integration of DA aptamer on the CPNTs covalently, Fourier-transform infrared spectroscopy (FT-IR) and fluorescence image were introduced. First, the chemical characterization of aptamer-CPNTs was  www.nature.com/scientificreports www.nature.com/scientificreports/ carried out using FT-IR(Alpha-p) spectroscopy (Fig. 4c). The absorption peaks of the CPNTs were characterized with the absorption peaks in the range of 1725 cm −1 to 1663 cm −1 which are contributed to stretching vibrations of acid group (RCOOH) on CPNT. The broad OH stretch of carboxylic acids on CPNT is also attributed around 3000 cm −1 . However, the amide groups, which are formed by covalent bonding between CPNT and aptamer, enabled to be characterized by amide I (at 1620 cm −1 ) and II (at 1500 cm −1 ) bonds 52,53 . Generally, aptamer is composed of DNA which has a phosphate backbone. Therefore, X-ray photoelectron spectroscopy (XPS) analysis was introduced to compare the P2p ratios (ca. 133 eV) in aptamer-conjugated CPNT1 and CPNT2 (Fig. 4d). The P2p ratios of aptamer-CPNT2 is higher than one of the aptamer-CPNT1, resulted into the presence of more amount of the aptamer on the CPNT2 54 . Furthermore, the aptamer-CPNTs on IME were characterized by fluorescent images with following samples: aptamer-CPNTs and fluorescein isothiocyanate (FITC)-conjugated aptamer-CPNTs. The aptamer-CPNTs showed no significant fluorescent images (left in Fig. 4e), while the aptamer-CPNTs conjugated with FITC were clearly observed by fluorescence microscope (EVOS M5000) with emission spectrum peak wavelengths of approximately 519 nm (right in Fig. 4e).
The first step to characterizing the electrical properties of aptasensors was examining the electrical contact between sensing material aptamer-conjugated CPNTs and IME arrays. Figure 5a,b displays I-V curves of the CPNTs on the arrays obtained before and after the introduction of aptamer under ambient conditions. The slope variation (dV/dI) for CPNT2 (from 7 to 25) is larger than that for CPNT1 (from 19 to 38), which is caused by the different amounts of aptamer attachment on the CPNTs. A linear correlation over a voltage range of −10 to +10 mV (scan rate: 0.1 mVs −1 ) showed a stable ohmic contact. A slight decrease in the value of slope was www.nature.com/scientificreports www.nature.com/scientificreports/ attributed the increase in resistance due to the introduction of aptamer. It was also clear that the aptasensor could maintain reliable electrical contact to provide abundant conductive pathways. To measure electrical responses to DA, a liquid-ion gated aptasensor was constructed. The schematic illustration in Fig. 5c indicates that the sensing medium, aptamer-CPNTs, was surrounded with phosphate-buffered saline (PBS, pH 7.4) as the electrolyte. Preventing the leakage of liquid gate was critical for biosensor performance, which was overcome by adequate sealing with grease. This strategy was advantageous because it achieved both subtle contact with the CPNTs and signal amplification 55 . Figure 5d presents the output characteristics of the liquid-ion gated FET aptasensor with CPNT2 operated under ambient conditions with E G varying from 0 to −100 mV in −20 mV steps (source-drain voltage sweep rate = 1 mV −1 ). The drain to source current (I DS ) increased with negatively increasing gate potential, indicating charge transport behavior. This observation was important because the binding of target molecules to the aptamer could be detected according to a change in the drain to source current under this p-type configuration. It could also be inferred that the aptamer-CPNTs were suitable as sensing devices for DA detection.
Aptasensor performance. As the successful fabrication of FET-type aptasensors was confirmed, the sensor performance, such as sensitivity and selectivity of the detection of real DA samples obtained from cells, was measured. First of all, although the crystal structure of the DA-binding aptamer is still unresolved, the sensing mechanism can be suggested with previous biochemical data; the major binding site between DA and aptamer is two major stem-loop domains (Fig. 6a) 50 . The aptamer benchmarked has a pairing interaction between the two major stem-loop domain, which are consisted of five nucleotides (C-G-T-G-T and G-C-A-C-A), to form the DA binding pocket. The binding pocket is supposed to interact with the hydroxyl groups of the position 3 in DA structure by hydrogen bonding [48][49][50] . The binding cites enable contact to be highly compact and enclosed (Fig. 6a  inset) 56 . Therefore, the sensing mechanisms can be explained with two suggestions: 1) the DA molecules closely interact with two major stem-loop domains of negatively charged aptamers, and then, the DA-aptamer complex approaches the CPNT surface by physical or environmental movements. The close access of the negatively charged aptamer to the carrier, a hole, of CPNTs increases, resulting in an indirect doping effect. 2) The amount of complex increases with the negative charge intensity on the surface of CPNTs at pH 7.2 because the pK a of DA is 6.2, increasing the negative charge intensity from the catechol side of DA. Figure 6b shows real-time responses from CPNT aptasensors toward various DA concentrations. First, pristine CPNT1 and 2 were introduced as control experiments and showed no significant responses. The sensitivity values of the CPNT2 aptasensor were higher than those of the CPNT1 aptasensor owing to the former's enhanced surface area for the introduction of aptamers. The MDL of the CPNT2 aptasensor was ca. 100 pM, which is 100 times higher than that of the www.nature.com/scientificreports www.nature.com/scientificreports/ CPNT1 aptasensor. The CPNT aptasensor enables discrimination between healthy individuals and patients with DA-related illnesses, such as Parkinson's disease and schizophrenia, because the patients have been determined to have DA concentrations of ca. 0.6 to 6.2 ng/mL (3.9 to 40.5 nM) by HPLC.
To quantify the sensing behaviors of CPNT aptasensors, the normalized sensitivity to various DA concentrations were evaluated, and dose-dependent DA concentrations were obtained, as shown Fig. 6c. The dose-dependent values indicate that the CPNT2 aptasensor has higher sensitivity changes than one of the CPNT1 aptasensor at a fixed concentration. In addition, the CPNT aptasensors were characterized by excellent sensitivity in the presence of similar neurotransmitters, such as norepinephrine (NE), serotonin (ST), and phenethylamine  Fig. 6b,d, were calculated as standard deviation (n = 5).
(PEA), and even at a high concentration of DA (10 mM) (Fig. 6d). The response from the CPNT2 aptasensor was also monitored at a higher sensitivity level than the CPNT1 aptasensor. Moreover, the extended selectivity of aptasensors was also measured using various neurotransmitters, including NE, ST, PEA, ascorbic acid (AA), catechol (CT), epinephrine (EP) and DA (Fig. 6e). There were no significant responses without DA, and the graph showing sensitivity levels clearly indicates that the aptasensors enable the discrimination of DA in mixtures.
To confirm exogeneous DA release from the PC12 cells, target samples with DA release were prepared by a DA release assay (see the experimental section for details). Various KCl concentrations (12.5 mM to 200 mM) were sequentially introduced to promote Ca 2+ influx owing to membrane depolarization induced by K + charge, leading to the rapid DA release from PC12 cell (Fig. 7a). Therefore, the amount of the active Ca 2+ ion channel in the cell with K + is larger than one from the cell without K + . The concentrations of DA in target and control samples were calculated by commercial DA ELISA (Enzyme-Linked Immunosorbent Assay) kits (KA 1887, Abnova), resulted into the increasing DA concentrations dependent on the KCl concentrations (Fig. 7b). Based on the results with DA exocytosis, the increasing KCl concentration induced rapid and high-level DA release under a constant KCl treatment time. To investigate DA exocytosis with the sensitivity and selectivity of CPNT aptasensors, the prepared exogeneous DA samples were characterized. Figure 7c displays KCl concentration-dependent real-time DA responses with the addition of various exogeneous DA samples. The response levels from CPNT2 aptasensor increased with increasing KCl concentration, showing similar standard DA concentration-dependent results. Moreover, the CPNT2 aptasensor clearly displayed excellent selectivity toward DA in DA release samples, even in mixtures (Fig. 7d).

conclusion
In this study, a DA exocytosis system using CPNTs was firstly demonstrated by a liquid-ion gated FET system. The aptasensors were prepared by conjugating CPNTs with designed diameters (ca. 120 nm and ca. 200 nm) and DA-selective aptamers. Introducing a liquid-ion gated FET system resulted in high-performance DA aptasensors. The aptamer on the CPNTs was operated as a gate-modulator toward DA in this system. Therefore, the gate-modulated stimulation from DA was monitored and showed highly sensitive (ca. 100 pM) and selective DA detection. Moreover, the CPNT aptasensors also demonstrated excellent analysis of DA exocytosis by detecting exogeneous DA samples prepared by a DA release assay. Based on the excellent sensing properties of FET-type aptasensors, the aptasensors can be used for clinical diagnostic studies by overcoming i) DA extraction from the patient samples such as blood and saliva without additional pre-treatments and ii) cost-efficiency.

Fabrication of carboxylated polypyrrole nanotubes (CPNTs). CPNTs with different diameters (ca.
120 and 200 nm) were prepared by benchmarks in our previous works 51 . A surfactant, AOT (15 mmol), was dissolved in hexane (40 mL), and the solution was stirred for 30 minutes to reach equilibrium. Subsequently, aqueous FeCl 3 solution (7 M, 1 mL) was introduced into the AOT/hexane solution to generate reverse-cylindrical micelles containing iron cations 55 . Finally, a pyrrole (Py) (3.75 and 7.5 mmol) and carboxylated pyrrole (Py-COOH) (0.25 mmol) mixture was added dropwise into the reverse-cylindrical micelle phase. The chemical oxidation polymerization of Py/Py-COOH mixture monomers proceeded for 3 h at 18 °C to room temperature 51 . The resulting product was thoroughly washed with excess ethanol to remove the surfactant and other residual reagents. The final products were obtained after drying under vacuum at room temperature.
Instruments and procedures for IME array preparation. The photoregistor (DNR L300) was coated by a spincoater (ACE-200) on a glass wafer. Aligner (MA-6 II) was used to develop the photoregistor. To design the electrodes, Au/Cr was deposited on the photoregistor-coated glass wafer with a thickness 25/5 nm by Thermal Evaporator (MHS-1800). The Au/Cr deposited on glass wafer was developed into the developer, resulted into the IME array electrodes. Finally the glass wafer was cutted into each electrode substrates by Dicing Saw (DAD 3350). All processes were conducted by ourselves at the Inter-university Semiconductor Research Center (ISRC).

Construction of FET-type aptasensors.
The substrate was treated with a 1 wt% aq. amino silane (3-aminopropyltrimethoxysilane, APS from Aldrich) solution for 12 h to introduce amino groups on the IME glass substrate and washed with distilled water. The amino group-functionalized IME substrate was exposed to a mixture solution containing 3 wt% CPNT solution (40 µL) and 1 wt% aq. DMT-MM solution (40 µL) for 12 h. The IME array substrate conjugated with CPNTs was washed with distilled water several times. Subsequently, the aptamer was then immobilized onto the CPNT surface on IME substrate with same reaction: the mixture of aptamer and DMT-MM solution (40 µL) was dropped on the CPNT-IME substrate for 12 h. Afterwards, the final aptasensor was rinsed with distilled water and dried at room temperature under glass container. The characterization of the aptasensor was carried out using FT-IR (Alpha-p), TEM (MTE20), SEM (MSE20), SEM-EDX (MSE20), XPS (PHI 5000 VersaProbe (Ulvac-PHI)), fluorescence image (EVOS M5000) and I-V curve (Keithley 2612 A Dopamine detection experiments. The aptasensor was placed on the stage of the probe station con- DA release assay. Rat pheochromocytoma PC12 cell lines were obtained from the American Type Culture Collection (ATCC). PC12 cells were cultured in RPMI-1640 in 100 Φ dishes using 10% horse serum (v/v) (heat-inactivated, Gibco), 5% fetal bovine serum (v/v) (heat-inactivated, Gibco) and 1% penicillin-streptomycin (Gibco) solution. Then, the cells were separated by incubation at 37 °C for 2 minutes using 1 mL of Trypsin-EDTA 0.25% solution (Gibco). The detached cells were transferred to 24-well plates at 5 ×10 4 in each well and incubated at 37 °C.
The DA was pre-equilibrated in PC12 cells: the cells were confirmed to be well adhered by microscope, and the medium was removed. One milliliter of loading buffer (Hank's Balanced Salt Solution with 10 mM HEPES) was added to each well and incubated at 37 °C for 2 minutes. The supernatant was discarded, and then DA hydrochloride (20 nM) was prepared in loading buffer. One milliliter of DA hydrochloride solution was added to each well. The samples were incubated for 20 minutes at 37 °C, and the DA was pre-equilibrated with PC12 cells. The samples were washed with loading buffer (0.5 mL) 3 times.
KCl stimulation and supernatant acquisition: KCl buffer was prepared at 5 concentrations (25 mM, 50 mM, 100 mM, 150 mM and 200 mM) in loading buffer. PC12 cells were exposed to 1 mL of KCl buffer (25 mM, 50 mM,