Regenerability and Specificity of Thrombin IDA Aptasensor

In this section, an impedimetric aptasensor for detection of thrombin is fabricated on IDA electrode chips. After fabrication, 20nM thrombin or 15μM HSA is added to

compare their specificity against HTDQ29-SH. A regeneration step is applied following each protein binding step to thoroughly break the binding between HTDQ29-SH and the target protein and wash off all non-specific binding conditions. This can restore the

functionality of the aptasensor for the next detection step. For every protein binding step, 80μL target protein diluted in binding buffer is added and the chips are held upside down for 10min at RT. For every regeneration step, 100μL regeneration buffer (2M NaCl in

ddH2O) is added and the chips are placed upright for 10min at RT. The chips are washed with binding buffer twice and characterized using EIS after each protein binding or regeneration process.

EIS data for the first six steps is shown in Figure 4-15. Qualitatively interpreting, the Rct (the diameter of the imperfect semi-circle) becomes smaller after MCH blocking.

Due to the fact that the immobilization buffer has high ionic strength that cancels out the negative charge on the ssDNA, there might be strong non-specific binding of the aptamer on the electrode after HTDQ29-SH immobilization (Studier, 1969). After immersing in a comparatively higher concentration of MCH in binding buffer, the non-specific bound aptamers are somehow plucked off from the electrode surface with MCH occupying their

original sites. This will lead to the decrease in Rct. After adding 2M NaCl, the Rct becomes even smaller. This might be caused by further washing of the electrode, removing non-specifically bound MCH or even HTDQ29-SH (Goode et al., 2015). Addition of 20nM thrombin increases the Rct, and then addition of 2M NaCl (second time) decreases the Rct

to a point even lower than the first regeneration step. This might be caused by the fact that the non-specifically bound molecules aren’t thoroughly removed after the first regeneration step. For clarification and examination of regenerability, repeated steps of protein binding and regeneration procedures are performed in the following paragraph.

Note the down-bending of impedance at low frequencies. It is clear from Chapter 3 that this is caused by finite diffusion of redox species between two band electrodes, and there aren’t any previous literatures that propose an adequate model to fit this region that can yield the geometrical parameters. It can be seen that the shapes of impedances at low frequencies are not influenced by the immobilized materials on the electrode. This is

because the diffusion path from the ICCB (cf. section 3.2.1) to the electrode surface (typically a few to tens of μm) is comparatively longer that the tip of the self-assembled

monolayer (SAM) to the electrode surface (typically tens of nm), making its specific diffusion impedance unchanged.

Figure 4-15 Nyquist plot for characterizing the fabrication, detection and regeneration process of thrombin impedimetric aptasensor using IDA chips. (a) Chip #1 and (b) chip

#2. The first six steps are shown in the sequence from “Bare” to “2M NaCl 2^nd” in the legend.

The circuit fitting program is a self-written program that is capable of fitting several elements including the IDA diffusion impedance. The average MSE is 2.547 with a standard deviation of 1.794 using a Randles circuit with an IDA diffusion element for fitting. This error is small enough and can yield precise fitted parameters (cf. section 3.4.6 for comparison of different diffusion impedance elements used for circuit fitting of IDA electrode EIS data).

A total of four thrombin binding steps and two HSA binding steps are applied on the same electrode consecutively. Figure 4-16a plots the fitted Rct after each step. One can see that after almost each regeneration step, the Rct drops a little compared with its previous regeneration step. Taking the non-specific bound molecules still existing on the electrode in account, the baseline for Rct seems to decay after each step. This baseline is arbitrarily defined as the Rct which is affected by the immobilized SAM layer and the non-specifically bound molecules. If the steps are numbered starting from 1, then on the k^th step, the amount of temporarily and non-specifically bound molecules is arbitrarily assumed to decay exponentially. This consequent cause of transient change of Rct is defined as

Δ𝑅𝑅𝑐𝑐𝑢𝑢,𝑢𝑢𝑡𝑡𝑖𝑖𝑢𝑢𝑠𝑠𝑑𝑑𝑒𝑒𝑢𝑢𝑢𝑢(𝑘𝑘) = 𝑗𝑗𝑢𝑢𝑒𝑒^{− 𝑘𝑘𝑘𝑘0}

, where Bt ( [=] Ω ) is the proportional constant that is related to the amount of transient non-specific binding and k0 (dimensionless) is the constant that can characterize the speed (52)

an effect on the Rct (e.g. electrode area), aptamer and MCH immobilization amount … etc, and k0 might be affected by the regeneration conditions (e.g. ionic strength, regeneration time). Note that eqn. (52) is analogous to the first order exponential decay of receptor-ligand dissociation kinetics (Cima, 1994), where the dissociation rate constant (koff) is analogous to 1/ k0. The difference is k0 is dimensionless, and is affected by the regeneration condition, while koff ( [=] s^-1 ) characterizes the dissociation rate. Also, the number of steps k is concerned rather than the time.

The steady-state baseline BSS ( [=] Ω ) is defined as the Rct only affected by the immobilized aptamer and MCH without the non-specific bound molecules. Therefore, the overall baseline can be written as

𝑅𝑅𝑐𝑐𝑢𝑢,𝑏𝑏𝑖𝑖𝑠𝑠𝑒𝑒𝑐𝑐𝑑𝑑𝑢𝑢𝑒𝑒(𝑘𝑘) = 𝑗𝑗𝑆𝑆𝑆𝑆+ Δ𝑅𝑅𝑐𝑐𝑢𝑢,𝑢𝑢𝑡𝑡𝑖𝑖𝑢𝑢𝑠𝑠𝑑𝑑𝑒𝑒𝑢𝑢𝑢𝑢(𝑘𝑘) = 𝑗𝑗𝑆𝑆𝑆𝑆+ 𝑗𝑗𝑢𝑢𝑒𝑒^{− 𝑘𝑘𝑘𝑘0}

Note that eqn. (53) doesn’t account for the change of Rct cause by target binding. Thus, only the regeneration steps can be used to determine the parameters. Accordingly, all the even number of steps after step 3 are used for non-linear curve fitting of eqn. (53). The baseline of the two chips are calculated and plotted in Figure 4-16a. The three parameters can characterize the regenerability of this aptasensor. It can be observed that after each regeneration step, the Rct drops to approximately the value of baseline, and BSS is almost reached after the 12^th step. This proves that the fabricated IDA impedimetric aptasensor is able to be regenerated up to six times for six detections.

(53)

Figure 4-16 (a) Rct vs procedure steps for regenerability test using the IDA chips. (b) ΔRct

The method for determining a baseline for characterizing the regenerability can be

further used for obtaining the binding amount of different targets using the same IDA chip.

The ΔRct for each binding event is obtained by

Δ𝑅𝑅_{𝑐𝑐𝑢𝑢}(𝑘𝑘) = 𝑅𝑅_{𝑐𝑐𝑢𝑢}(𝑘𝑘) − 𝑅𝑅𝑐𝑐𝑢𝑢,𝑏𝑏𝑖𝑖𝑠𝑠𝑒𝑒𝑐𝑐𝑑𝑑𝑢𝑢𝑒𝑒(𝑘𝑘)

Figure 4-16b plots the calculated values. A 5.1-fold (chip #1) and 5.5-fold (chip #2)

difference between the two target proteins is calculated. Even if the concentration of HSA is significantly higher than the concentration of thrombin, a smaller ΔRct indicates that

the binding of HTDQ29-SH and thrombin is quite specific. At last, it is proved that a regenerable impedimetric aptasensor for specific detection of thrombin is successfully fabricated on IDA electrode chips.

(54)

4.5 Summary

In summary, specific EIS detection of the thrombin and MUC1 is performed using two symmetric electrode setups. The theory of this study can serve as a method for analyzing symmetric electrode bio-sensing applications. For asymmetric conditions (10mM Fe(CN)63-), the optimal operating sensing potential for impedimetric aptasensing is identified at 0.7V (vs Au). Specific EIS detection of MUC1 (3 ~ 200 nM) is performed and KD values of 16.31±1.44nM is obtained. Validity of the two symmetric Au electrode equivalent circuit model is confirmed by comparing Rct values in different electrode setups. Sequence specificity of the binding event is confirmed by comparing S2.2 with 25mer random ssDNA sequences. Fabricated impedimetric aptasensors using IDA chips

have the regenerability for six times of detection and the specificity is confirmed (over 5-fold between 20nM thrombin and 15μM HSA).

Though further improvement on sensitivity and selectivity are needed to be realized, the label free and simple characteristics of this system widens its potential for portable, real-time and multiplex developments.