Letter Cite This: Nano Lett. XXXX, XXX, XXX−XXX
pubs.acs.org/NanoLett
Real-Time Detection of Markers in Blood Jukwan Na,† Min-Ho Hong,∥ Jun Shik Choi,† Hankyul Kwak,‡ Seungwoo Song,§ Hyoseok Kim,† Youngcheol Chae,§ Eunji Cheong,‡ Ju Hee Lee,⊥ Yong-beom Lim,*,† and Heon-Jin Choi*,†
Downloaded via UNIV OF TEXAS AT DALLAS on March 12, 2019 at 21:45:57 (UTC). See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles.
†
Department of Materials Science and Engineering, ‡Department of Biotechnology, §Department of Electrical and Electronic Engineering, Yonsei University, Seoul 03722, Republic of Korea ∥ Nature Inspired Materials Processing Research Center, Department of Energy Science, Sungkyunkwan University, Suwon 16419, Republic of Korea ⊥ Department of Dermatology, Cutaneous Biology Research Institute, Yonsei University College of Medicine, Seoul, 03722, Republic of Korea S Supporting Information *
ABSTRACT: The real-time selective detection of disease-related markers in blood using biosensors has great potential for use in the early diagnosis of diseases and infections. However, this potential has not been realized thus far due to difficulties in interfacing the sensor with blood and achieving transparent circuits that are essential for detecting of target markers (e.g., protein, ions, etc.) in a complex blood environment. Herein, we demonstrate the real-time detection of a specific protein and ion in blood without a skin incision. Complementary metal-oxide-semiconductor technology was used to fabricate silicon micropillar array (SiMPA) electrodes with a height greater than 600 μm, and the surface of the SiMPA electrodes was functionalized with a self-assembling artificial peptide (SAP) as a receptor for target markers in blood, i.e., cholera toxin (CTX) and mercury(II) ions (Hg). The detection of CTX was investigated in both in vitro (phosphate-buffered saline and human blood serum, HBO model) and in vivo (mouse model) modes via impedance analysis. In the in vivo mode, the SiMPA pierces the skin, comes into contact with the blood system, and creates comprehensive circuits that include all the elements such as electrodes, blood, and receptors. The SiMPA achieves electrically transparent circuits and, thus, can selectively detect CTX in the blood in real time with a high sensitivity of 50 pM and 5 nM in the in vitro and in vivo modes, respectively. Mercury(II) ions can also be detected in both the in vitro and the in vivo modes by changing the SAP. The results illustrate that a robust sensor that can detect a variety of molecular species in the blood system in real time that will be helpful for the early diagnosis of disease and infections. KEYWORDS: Real-time detection, blood, sensor, comprehensive circuit, micropillar
T
biochemical and immunological assays, which are often timeconsuming and have limitations with regard to real-time detection. Biosensors are a promising candidate for the real-time detection of disease-related proteins, toxic molecular species, or ions related to living organisms. While many real-time sensors are currently available, they are limited to detecting
he real-time detection of molecular species inside a living organism such as in blood and intra- or extracellular spaces is important but challenging. In particular, the in vivo detection of proteinaceous materials related to diseases and associated with or secreted by bacteria and viruses with high sensitivity and selectivity is critical for the early diagnosis of diseases and infections. The real-time in vivo detection of other hazardous species, such as heavy-metal ions, is also critical for understanding the causes of various diseases and infections. Current detection methods typically use blood samples or biopsy specimens taken from the body and used in ex vivo © XXXX American Chemical Society
Received: November 29, 2018 Revised: February 6, 2019
A
DOI: 10.1021/acs.nanolett.8b04775 Nano Lett. XXXX, XXX, XXX−XXX
Letter
Nano Letters
Figure 1. (a) Schematic image of comparison between a field-effect transistor (FET) sensor on the skin and the silicon micropillar array (SiMPA), which forms a comprehensive circuit in the blood vessel. (b) The fabrication process of SiMPA. (c) The height of SiMPA according to the deep reactive ion etch (DRIE) process for contact blood from the stratum corneum. (d) SiMPA with the height of 600 μm through DRIE and the wetetching process.
which ensures a high aspect ratio (see section S1 in the Supporting Information). The typical morphologies of a SiMPA are shown in Figure 1c. The skin contains three layers (epidermis, dermis, and subcutaneous), each having its own sublayers, and the blood vessels are located 600 μm under the skin surface in the dermis layer.4 For contact between a SiMPA and blood vessels in the dermis layer, the height of the SiMPA must be >600 μm. The tip of the SiMPA was sharpened for piercing the skin easily by etching with a 45% KOH aqueous solution. The Si (111) plane was etched slowly compared to the (100) and (110) planes.5 The etch rates of the (100) and (110) planes using KOH were 300−600 times higher than that of the (111) plane; therefore, the SiMPA was sharpened after the wet-etch process using KOH (Figure 1d). After the etching processes, the diameter of each SiMPA was 200−250 μm. To create a circuit with the fabricated SiMPA, a complementary metal-oxide-semiconductor (CMOS) process was applied to form Au electrodes and passivation.6,7 The second passivation layer of the upper part of the SiMPA was then selectively etched using a buffered oxide etchant, and, finally, Au on the upper part of the SiMPA was exposed as part of the detection electrodes. Because the process with the SiMPA necessitated piercing of the skin for direct contact with blood vessels, a biocompatibility test was performed to verify the safety of its application. The tests were performed with a tracking test and H&E staining for a mouse model and micro computed tomography (μ-CT) imaging and a TUNEL assay for an explanted human skin model (section S2 and Figure S1). All mice were managed under a 12 h/12 h light−dark cycle (lights on 7:00 A.M.) and had ad libitum access to food and water. Also, every mouse was cared for and handled in accordance with the guidelines of the Institutional Animal Care and Use Committee at Yonsei University (Seoul, Korea). The tracking test was used to determine the free movement of the mouse after the insertion of the SiMPA to confirm the damage to the mouse by the SiMPA. The hematoxylin and eosin (H&E) staining of the mouse skin punctured by the SiMPA did not indicate any inflammatory response. Additionally, the SiMPA did not result in apoptosis of the epidermis and dermis in the explanted human skin model after insertion. Our preliminary findings
physical changes (e.g., body temperature and heart rate) and molecules excreted along with sweat (e.g., glucose and lactate) or breath (e.g., volatile organic compounds). There has been limited development of sensors that can directly detect markers, molecules, and ions in body fluids1,2 because of the difficulties in interfacing sensors with blood. The blood vessels are located ∼600 μm under the skin and are thus difficult for conventional planar type sensors to access without dissection. It is also difficult to detect a target material selectively in complex blood environments even if the sensor can successfully blood. Most biosensors use a field-effect transistor (FET)-type current change in the channel, in which the signal change is induced by the molecular recognition between the target and receptor interactions on the surface of the channels. In complex blood environments, however, various ions or proteins interfere with target and receptor reactions in many ways, disrupt the gating effects, and make it difficult to detect the targets. This study aims to (1) directly access the blood system and (2) detect the targets in blood in real-time with high sensitivity and selectivity. To achieve these goals, we designed a threedimensional (3D) sensor with Si micropillar array (SiMPA) electrodes that can pierce the skin and directly contact blood, as shown in Figure 1a. The SiMPA creates a “comprehensive circuit” that includes blood, electrodes, and the receptor as circuit components. Target detection is achieved by monitoring the impedance change induced by peptide and protein receptor interactions at the surface of the SiMPA electrodes. This design is unique because the targets are detected by a change in the circuit impedance, not by the indirect gating effect. This can provide a reliable sensing mechanism in complex conditions such as in a blood system. Figure 1b shows the fabrication process of the SiMPA using a Si (111) wafer. The SiMPA were fabricated using a deep reactive-ion etching (DRIE) technique.3 The etching process was carried out using high-density plasma with two alternating etching and passivation steps. Both the etching and the passivation steps lasted for a few seconds, and the process continued in a cyclic manner for minutes or hours depending on the height of the SiMPA. The scallop features on the sidewalls of the SiMPA are a signature of the DRIE process, B
DOI: 10.1021/acs.nanolett.8b04775 Nano Lett. XXXX, XXX, XXX−XXX
Letter
Nano Letters
Figure 2. (a) Multivalent galactose-self-assembling artificial peptide (G-SAP) formation with triethylene glycol mono-11-meraptoundecyl ether (TGME) onto SiMPA for selective recognition of cholera toxin-B (CTXB). (b) Experimental proof of G-SAP binding onto Au nanoparticle (AuNP) and the importance of TGME presence for the steric stabilization of G-SAP/AuNP suspension. (c) Determination of optimal G-SAP with the tetramethylrhodamine (TG-SAP)-to-TGME ratio for highest-affinity CTX-B recognition. (d) Bright-field (left) and fluorescence microscope (right) images in two different regions of the same SiMPA after TG-SAP decoration.
showed that the SiMPA can be applied to humans and animals without inducing any inflammatory response. A cholera toxin (CTX) was selected as a test bed to test the SiMPA as a sensor for protein in blood. Cholera causes an acute diarrheal infection, owing to the ingestion of food or water contaminated with the bacterium Vibrio cholera. There were more than 100 000 reported cases of cholera, including 1304 deaths, in 2015.8 The short incubation period of 2 h to 5 days intensifies the potentially explosive pattern of outbreaks, and therefore, early detection is critical. CTX, which is produced from V. cholera is the actual cause of the disease. It is composed of a single toxic A-subunit (CTX-A) and five nontoxic B-subunits (CTX-B), which constitute a pentameric receptor for the GM1 glycolipid found on the surface of intestinal epithelial cells.9,10 CTX-B exhibits multivalency, which is a common feature of cell-surface adhesion, and this feature has been the primary focus of creating inhibitors of CTX-B.11−14 For example, Branson et al. reported a multivalent inhibitor for CTX based on an inactive mutant CTX-B protein modified with GM 1 oligosaccharide ligands. 15 Although the design and fabrication of the defined complexes exhibiting multivalency are difficult and optimizing a multivalent interaction is a highly challenging issue, CTX-B is a suitable target for a first test bed to verify the performance of the receptor. In this study, self-assembling artificial peptides (SAP) was used as a receptor for detection targets, owing to their advantages in this SiMPA. For example, compared to a typical receptor, SAPs are more flexible in terms of recognizing targets by controlling the structures. For example, the SAP in this study was designed to recognize the pentameric binding sites
of CTX through multivalent interactions using galactose and triethylene glycol mono-11-meraptoundecyl ether (TGME) as a ligand and a spacer, respectively (Figure 2a). The multivalent interactions can result in strongly binding for recognition and thus, they can be used to precisely detect targets in complex blood environments. The SAP (number of amino acid sequences) is also smaller than proteins such as antibodies; hence, they can be used to create more electrically transparent circuit interfaces between electrodes and in vivo (i.e., blood) in our comprehensive circuits because peptides and proteins are typically not good conductors. The small sizes of SAPs are also helpful in piercing the skin without damaging either the skin or the SAP itself. Importantly, the thermal stability of an SAP is better than that of a typical protein receptor, which is critical for maintaining function in the blood at body temperature for both short- and long-term detection. To apply a SiMPA and detect CTX, which is composed of a pentavalent protein with five carbohydrate binding regions on the CTX-B, the SAP was designed to consist of a flexible hydrophilic linker, a carbohydrate segment (galactose), a cysteine segment for Au binding, and a hydrophobic segment for self-assembled monolayer (SAM) formation via noncovalent interactions (galactose-self-assembling artificial peptide; G-SAP). To confirm a binding interaction between the synthesized G-SAP and the Au surface, the G-SAP was applied to a citrate-removed Au nanoparticle (AuNP) colloid.16 Because the G-SAP has a positive charge, the solution in which AuNP aggregated exhibited a blue color as evidence of AuNP binding with the G-SAP (Figure 2b).17 However, when TGME binds with AuNP, which is bound with G-SAP, it could sterically inhibit charge interactions between the G-SAP and C
DOI: 10.1021/acs.nanolett.8b04775 Nano Lett. XXXX, XXX, XXX−XXX
Letter
Nano Letters
Figure 3. Circuit of the hybrid-SiMPA (H-SiMPA) and measurement of impedance change through the circuit. (a) Schematic diagram of a comprehensive circuit. (b) Impedance analysis data shows the impedance change depends on the composition of a solution on the H-SiMPA and (c) the difference of impedance before and after binding G-SAP on the SiMPA in phosphate-buffered saline (PBS).
the surface of the AuNPs, thereby stabilizing the suspension. The binding of G-SAP with AuNPs was tested using fluorescence quenching to determine whether the stability of the AuNP suspension was compromised by factors such as an increasing concentration of salt. Energy transfer from a dye to AuNPs occurs if both are close to each other (distance of up to 22 nm).18 This transfer occurs via a surface energy-transfer process and it follows a 1/d4 distance dependence.19 G-SAP was synthesized using a fluorescent probe, tetramethylrhodamine (TAMRA), in addition to galactose (TG-SAP). As shown in Figure S3a, when TG-SAP was added to an AuNP solution, there was strong fluorescence quenching against TGSAP in deionized (DI) water. This proves that the synthesized G-SAP bound well with AuNP. In addition, the maximum number of G-SAPs that could attach themselves to a single AuNP was quantified. Increasing concentrations of TG-SAP were added to a fixed amount of AuNP solution (0.56 pmol), and the resulting mixtures were subjected to fluorescence measurements (Figure S3b). As shown in Figure S3b, fluorescence emission from TG-SAPs in the G-SAP/AuNP mixture were effectively quenched up to 36 pmol of the GSAP, after which the emissions began to increase. Thus, up to
36 pmol of TG-SAP can be attached to 0.56 pmol of AuNPs (64 peptide molecules per AuNP particle). Because multivalent interactions are more effective than monovalent interactions when the spacing between multiple ligands and G-SAP is correctly adjusted, the space between GSAPs were controlled using TGME as a spacer to achieve an increase in affinity and selectivity.20 The flexibility of the GSAP made spacing control possible, and hence, the critical ratio of G-SAP to TGME could be studied with TGME as a spacer to help maintain the critical ligand placement and orientation required for CTX binding with the highest affinity. Herein, the hydrophobic domain of G-SAP resulted in the effective coassembly of G-SAP and TGME as a monolayer on Au film via G-SAP-G-SAP or G-SAP-TGME interactions. G-SAP/ TGME mixtures were added to AuNP solutions with increasing G-SAP concentrations to the critical ratio of GSAP: TGME for optimized multivalent recognition of pentavalent CTX-B. Then CTX-B was added to the stable G-SAP/TGME-AuNP hybrid suspension and incubated for more than 1 h. The supernatant containing unbound CTX-B and the pellet containing the CTX-B-hybrid complex were separated, and each sample was subjected to gel electroD
DOI: 10.1021/acs.nanolett.8b04775 Nano Lett. XXXX, XXX, XXX−XXX
Letter
Nano Letters
Figure 4. Impedance analysis data for different chemical conditions by using the H-SiMPA (at 1 kHz). (a) The difference of the impedance between the pristine-SiMPA (P-SiMPA) and the H-SiMPA from different concentration for CTX-B in PBS and human blood serum (HBS). (b) The different sensitivity results between the P-SiMPA and the H-SiMPA using the positive and negative control group (asterisks indicate the statistically significant difference with respect to CTX-B, p < 0.001, n = 10).
the SiMPA in 50 nM of the TG-SAP solution and removing the excess TG-SAP by washing. The binding of TG-SAP onto the SiMPA was visually confirmed via fluorescence microscopic observation (Figure 2d). The above-described experiments confirmed the optimized multivalent interactions between GSAP and CTX-B and proved G-SAP’s affinity to the SiMPA. The comprehensive circuit that exists when the SiMPA is coated with SAP is shown in Figure 3a. As the targets are recognized by the receptor (i.e., SAP), the current path in the circuits changes from electrode A/SAP/blood/SAP/electrode B to electrode A/SAP-target/blood/SAP-target/electrode B. This leads to a change in the capacitance in the circuit owing to the target-receptor recognition at the surface of the electrodes, which results in a change in the impedance. The impedance between the SiMPA electrodes was measured using an impedance analyzer through a flexible printed circuit board (FPCB) and a printed circuit board (PCB) connected to the SiMPA. Each independent channel of the SiMPA was connected to the FPCB separately and chosen for measuring the impedance of the circuit using a switch on the PCB. This minimized external noise and increased the stability and reliability of the circuit. The impedance change in the circuit is represented by:
phoresis under denaturing conditions (sodium dodecyl sulfate polyacrylamide gel electrophoresis, SDS-PAGE; Figure S5). This assay revealed that the optimal G-SAP-to-TGME ratio for the CTX-B binding with the highest affinity ranges from 4.8:1 to 9.5:1. The optimal G-SAP-to-TGME ratio was further corroborated using a colorimetric assay in which the TGME concentration was a variable (Figure 2c). A total of 10 pmol of TG-SAP was mixed with increasing concentration of TGME and then was added to 10 μL of citrate-removed AuNP (56 fmol). After a 3 h incubation period, increasing amounts of CTX-B were added to the solution and left overnight. The color of the solution changed from red to blue as the quantity of CTX-B increased. The most significant blue-colored aggregate formation was observed at a TG-SAP-to-TGME ratio of 1:8. This result was in accordance with the SDS-PAGE experiment (Figure S5). Thus, this ratio (1:8) was used during SiMPA fabrication. Transmission electron microscopy (TEM) was used to verify the G-SAP coat formation around the AuNP core (Figure S4). Consistent with the formation of the G-SAP/ TGME shell, a corona with a reduced electron density was observed on the outer surface of the AuNP core. A solution of TG-SAP was added to the SiMPA to confirm the binding of GSAP to the Au surface of the SiMPA. This was done by soaking
ÄÅ ÉÑ−1 ÅÅ 1 ÑÑ R ct Å + jω(Cdl + CCTX‐B)ÑÑÑÑ = R e + R s + Z = R e + R s + ÅÅÅ ÅÅÇ R ct ÑÑÖ 1 + jωR ct(Cdl + CCTX‐B) j = imaginary number, j =
− 1 , ω = angular frequency, ω = 2πf , s−1
where Z is a meaningful factor including capacitance and resistance in the comprehensive circuit; Re is the resistance of electrode on the SiMPA; Rs is the resistance of the solution, Rct
(1)
and Cdl are the charge-transfer resistance and double layer capacitance, respectively, on the surface of the SiMPA; and CCTX‑B is the capacitance due to CTX-B. E
DOI: 10.1021/acs.nanolett.8b04775 Nano Lett. XXXX, XXX, XXX−XXX
Letter
Nano Letters
Figure 5. Measurement set-up of in vivo test and impedance analysis data. (a) H-SiMPA connected to flexible printed circuit board (FPCB) for real-time recording. (b) Impedance analysis data in different chemical environments [blood, concanavalin A (Con A), and CTX-B] using the HSiMPA at 1 kHz. (c) Comparison of impedance with injected CTX-B (50 nM) between H-SiMPA and improved H-SiMPA with TGME (8:1) at 1 kHz. (d) Impedance change ratio from blood to blood with injected different concentrations (5 and 50 nM) of CTX-B using H-SiMPA with TGME. Data were analyzed by t test (single asterisks indicate p < 0.05 and double asterisks indicate p < 0.001, n = 3). (e) A chemical structure of the mercury binding SAP (M-SAP) containing dendritic cysteinerich segments (the left side of the molecule) and mercury ion binding segments (the right side of the molecule). (f) Comparison of the impedance change of several ions solution including heavy ions from the mercury(II) ion solution in 10 mM of KCl solution. Data were analyzed by t test (single asterisks indicate p < 0.05 and double asterisks indicate p < 0.001, n = 4). (g) Changes of measured impedance by using the H-SiMPA decorated with M-SAP when calcium ion solution and mercury(II) ion solution were injected (50 μL at a concentration of 4 nM) into mice. Data were analyzed by t test (single asterisks indicate p < 0.05, and double asterisks indicate p < 0.001; n = 5).
As per eq 1, the measured impedance, which comprises the charge-transfer resistance in parallel to a double layer capacitance and the capacitance of CTX-B, corresponds to the measuring frequency.21 The measured capacitance of the double layer and CTX-B tends toward zero in the lowfrequency range, and a change of impedance is rapidly decreased with an increase of measuring frequency. Therefore, the proper range of the measuring frequency was required to achieve effective change of the capacitance and, in turn, the impedance. To investigate the impedance change according to the frequency, the impedance was measured through a frequency sweep in a broad frequency range from 20 Hz to 1 MHz. After the frequency sweep and analysis of the tendency, the optimized frequency range that the biggest signal-to-noise ratio (SNR) was found to be from 0.1 to 10 kHz. For CTX-B detection, the SiMPA was tested in phosphatebuffered saline (PBS) and human blood serum (HBS) in the optimized frequency range. PBS and HBS were used to create an environment similar to that of human blood. Figure 3b shows the results obtained using the SiMPA without G-SAP (pristine-SiMPA; P-SiMPA) to measure the difference in impedance for each solution, i.e., the SiMPA’s comprehensive circuit responses to changes of environment. Furthermore, the impedance change caused by G-SAP is shown in Figure 3c, which proves that the SiMPA coated with G-SAP (hybridSiMPA; H-SiMPA) reacted to changes in the surface conditions. The surface condition and environment of the H-
SiMPA influenced the impedance towing to changes in the charge-transfer resistance and double layer capacitance. The reason for the impedance change, as shown in Figure 3c, is that G-SAP created SAMs on the Au surface, which acted as a porous insulating film. Hence, the impedance increased with increasing charge-transfer resistance of the H-SiMPA. A sensitivity test for the H-SiMPA was carried out by increasing the concentration of CTX-B in the PBS and HBS solutions. The PBS solution was used to create an environment with abundant ions near the H-SiMPA, and the HBS solution was used to create an environment similar to that of blood. The CTX-B was added to each PBS and HBS solution during the impedance analysis of the H-SiMPA. The impedance was analyzed by injecting a 5.0−5.0 × 103 pM concentration of CTX-B into the solution. As shown in Figure 4a, the impedance for the P-SiMPA tended to slightly increase, while that of H-SiMPA decreased following an increase in the CTX-B concentration in both the PBS and HBS solutions. These tendencies depend on whether CTX-B was bound to the G-SAP on the surface of the SiMPA. For the P-SiMPA, CTX-B hardly affected its surface of the P-SiMPA and this directly resulted in a change in this circuit because of the absence of a binding site on the surface of the P-SiMPA. However, the impedance increased slightly, because CTX-B, which is a protein, worked as a resistor in the solutions. Meanwhile, CTX-B was bound to the G-SAP on the surface of the H-SiMPA, and this resulted in an additional capacitance parallel to the double layer capacitance in Figure 3a. As F
DOI: 10.1021/acs.nanolett.8b04775 Nano Lett. XXXX, XXX, XXX−XXX
Letter
Nano Letters
confirmed critical ratio, thus controlling the space between GSAP. By binding G-SAP and TGME to the Au surface of the SiMPA, the binding between the G-SAP and the pentamer structured CTX-B was maximized (Figure 5c). The results of the impedance change ratio are shown in Figure 5d. They indicate that the improved H-SiMPA can detect much lower concentrations of CTX-B in vivo than the H-SiMPA without TGME. To determine whether the SiMPA can be used for sensing targets other than proteins, a new type of SAP with mercury(II) ion binding capabilities was designed (Figure 5e). SAPs have high flexibility; hence, the new type of SAP can be designed based on the G-SAP via a sequence change. This SAP consisted of a mercury(II) ion binding segment23 and tetrapod-type multiple thiols for strong attachment to an Au surface. Instead of G-SAP or TG-SAP, the SiMPA was covered with a mercury-binding SAP (M-SAP). To confirm the attachment of multiple thiols onto the surface of the SiMPA and AuNPs, as in the case of the TG-SAP, M-SAP including TAMRA at the branch point (TM-SAP) was synthesized and tested (section S6). A sensing experiment showed that the HSiMPA could detect mercury(II) ions in both the in vitro and in vivo modes with high sensitivity and selectivity. The HSiMPA recognized mercury(II) ions owing to a noticeable impedance change compared to that in the case of the other ions in the in vitro test (Figure 5f). The selectivity test for H-SiMPA was carried out in the in vitro mode with several ions, including heavy-metal ions. The impedance changes for each ion were different from that for the mercury(II) ions at the same concentration (5 nM). This test was performed in 10 mM KCl solution to determine the performance of H-SiMPA with abundant ions in solution. The mercury(II) ions resulted in an impedance change that was up to 30% more than that for other ions. The in vitro test produced a noticeable result compared with that of other mercury(II) ion sensors because of the high sensitivity of the H-SiMPA even in a solution with abundant ions.24,25 Figure 5g shows the ion sensing capability of the HSiMPA in circulating blood in the in vivo mode. This test was set and performed in the same manner as the CTX test in Figure 5a−d. The concentration of the injected calcium and mercury ion solutions was adjusted to 100 pM by sequentially adding 4 nM of calcium ion solution and mercury(II) ion solution for 1.8 mL of total blood of the mouse.26 The calcium ion solution caused the impedance to drop by ∼0.05 kΩ at 1 kHz, whereas the mercury(II) ion solution resulted in an impedance drop 5 times more than this value. The results were similar to those of the CTX test, which confirmed the reliability of the H-SiMPA. Our results show the feasibility of using a SiMPA as a protein and ion sensor in in vitro and in vivo tests. This feasibility was established via direct blood system access and real-time target detection in blood with high sensitivity and selectivity. Therefore, a SiMPA can be used as a general platform for real-time, ultrasensitive, and ultraselective protein or ion detection in blood vessels without the need for an incision. A SiMPA can also be used to create comprehensive circuits in other systems; thus, it can be used in many applications not only in blood systems but also in other organ systems.
described by eq 1, the capacitance of CTX-B added in parallel to the double-layer capacitance reduces the total impedance. For these reactions, the transparent circuit instantly detected changes in the base solution condition and the surface state of the SiMPA, and these changes were reflected in the impedance as shown in Figure 4a. Figure 4b shows the results of the selectivity test based on the HBS solution in the in vitro mode of the same H-SiMPA. The selectivity test was designed to confirm the real-time detection ability of the H-SiMPA in a situation in which multiple species coexist like in the actual blood.22 The results showed that the H-SiMPA could recognize CTX-B selectively compared to other control solutions. Botulinum toxin (BTX) and immunoglobulin (IgG) were used as a negative control group, and concanavalin (Con A) was used as a positive control group for the selectivity test. It should be noted that all the targets, i.e., CTX-B, BTX, IgG, and Con A are proteins. The concentration of CTX-B and the control groups was 50 pM in the HBS. First, the H-SiMPA was immersed in 200 μL of HBS as a base solution, and then, 20 μL of each solution was added onto the H-SiMPA in the following order: Con A, BTX, IgG, and CTX-B. For each addition, 20 μL of the base solution was removed and 20 μL of each solutions was added to maintain the total amount of the mixed base solution as 200 μL. The impedance of the H-SiMPA rapidly decreased when CTX-B was added compared to that in the cased of the control groups. However, for P-SiMPA, there was no significant difference in the impedance for the control groups and CTX-B as expected. The highly significant (p < 0.001) effect of CTX-B on the impedance was observed with the positive control group as well as the negative control groups when the HSiMPA was used for impedance analysis. Moreover, the selectivity test proved the capability of the H-SiMPA for realtime detection in a multiple species solution. Real-time analysis was performed to determine the feasibility of using the H-SiMPA for detecting CTX-B in the blood system in the in vivo mode (mouse model) (Figure 5a). The H-SiMPA was connected to an FPCB for recording the impedance change in the in vivo model in real-time and sensing the presence of the toxin in blood vessels. CTX-B and Con A (50 nM) were injected into a real blood system for selectivity testing. The other control groups (IgG and BTX) were excluded because there were already plenty of physiological proteins, such as immunoglobulins in the body. The results of the in vitro selectivity test showed that Con A (positive control group) is appropriate for verifying the selectivity. It also should be noted that because the target CTX-B is a nontoxic subunit of the cholera toxin, it works as a protein similar to Con A in the living mouse model. The impedance was measured after the injection and circulation of CTX-B and Con A in the real blood system. Each protein was injected intravenously into a tail vein. As shown in Figure 5b, the impedance decreased by 5 kΩ at 1 kHz for Con A, while it decreased by more than 10 kΩ for CTX-B, clearly demonstrating in vivo selectivity. If the total volume of blood in the mouse is assumed to be ∼1.5 mL, the concentration of transfused CTX-B and Con A in the blood is predicted to be 1.6 nM or less. Because of the loss of CTX-B and Con A in vivo due to the immune system and blood circulation system, a higher concentration of CTX-B and Con A was used than the in vitro test (50 pM). In addition, it was found that the sensitivity of the H-SiMPA can be improved by using G-SAP mixed with TGME in the G
DOI: 10.1021/acs.nanolett.8b04775 Nano Lett. XXXX, XXX, XXX−XXX
Letter
Nano Letters
■
(13) Pieters, R. J. Org. Biomol. Chem. 2009, 7 (10), 2013−2025. (14) Wittmann, V.; Pieters, R. J. Chem. Soc. Rev. 2013, 42 (10), 4492−4503. (15) Xu, S.; Yan, Z.; Jang, K. I.; Huang, W.; Fu, H. R.; Kim, J.; Wei, Z.; Flavin, M.; McCracken, J.; Wang, R.; Badea, A.; Liu, Y.; Xiao, D. Q.; Zhou, G. Y.; Lee, J.; Chung, H. U.; Cheng, H. Y.; Ren, W.; Banks, A.; Li, X. L.; Paik, U.; Nuzzo, R. G.; Huang, Y. G.; Zhang, Y. H.; Rogers, J. A. Science 2015, 347 (6218), 154−159. (16) Choi, J. S.; Han, S. H.; Kim, H.; Lim, Y. B. Bioconjugate Chem. 2016, 27 (3), 799−808. (17) Khan, S. A.; DeGrasse, J. A.; Yakes, B. J.; Croley, T. R. Anal. Chim. Acta 2015, 892, 167−174. (18) Park, S. Y.; Lee, S. M.; Kim, G. B.; Kim, Y. P. Gold Bull. 2012, 45 (4), 213−219. (19) Sen, T.; Sadhu, S.; Patra, A. Appl. Phys. Lett. 2007, 91, 043104. (20) Mammen, M.; Choi, S. K.; Whitesides, G. M. Angew. Chem., Int. Ed. 1998, 37 (20), 2755−2794. (21) Orazem, M. E.; Tribollet, B. Electrochemical impedance spectroscopy; Wiley: Hoboken, NJ, 2008; p 523. (22) Artigues, M.; Abella, J.; Colominas, S. Sensors 2017, 17, 2620. (23) DeSilva, T. M.; Veglia, G.; Porcelli, F.; Prantner, A. M.; Opella, S. J. Biopolymers 2002, 64, 189−197. (24) Chen, K.; She, S.; Zhang, J.; Bayaguud, A.; Wei, Y. Sci. Rep. 2015, 5, 16316. (25) Han, K. N.; Choi, J. S.; Kwon, J. Sci. Rep. 2017, 7, 2806. (26) Ko, K. K.; Lee, H. W.; Kang, K. Journal of the Korean Neurological Association 2013, 31, 196−198.
ASSOCIATED CONTENT
S Supporting Information *
The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.nanolett.8b04775. Additional details on the fabrication of the SiMPA, synthesis of SAP, and analysis of the synthesized SAP (PDF)
■
AUTHOR INFORMATION
Corresponding Authors
*E-mail:
[email protected]. *E-mail:
[email protected]. ORCID
Min-Ho Hong: 0000-0001-9268-9906 Yong-beom Lim: 0000-0001-6590-7373 Heon-Jin Choi: 0000-0001-9427-0252 Author Contributions
J.N., M.-H.H., and J.S.C. contributed equally to this work. Notes
The authors declare no competing financial interest.
■
ACKNOWLEDGMENTS This work was supported by a National Research Foundation of Korea (NRF) grant funded by 2017R1A2A2A05069773; the Ministry of Science, ICT & Future Planning (grant no.490 2018M3C7A1024654); the Korea government (MSIP) (grant no. 2017R1A2B3011586); the Ministry of Education, Science, and Technology (2014M3A7B4051594); the Yonsei University Yonsei-SNU Collaborative Research Fund of 2014.
■
REFERENCES
(1) Martin, A.; Kim, J.; Kurniawan, J. F.; Sempionatto, J. R.; Moreto, J. R.; Tang, G.; Campbell, A. S.; Shin, A.; Lee, M. Y.; Liu, X.; Wang, J. ACS Sensors 2017, 2 (12), 1860−1868. (2) Al-Omari, M.; Liu, G.; Mueller, A.; Mock, A.; Ghosh, R. N.; Smith, K.; Kaya, T. J. Appl. Phys. 2014, 116, 183102. (3) Madou, M. J. Fundamentals of microfabrication: the science of miniaturization. 2nd ed.; CRC Press: Boca Raton, FL, 2002; p 723. (4) Rosell, J.; Colominas, J.; Riu, P.; Pallasareny, R.; Webster, J. G. IEEE Trans. Biomed. Eng. 1988, 35 (8), 649−651. (5) Wilke, N.; Morrissey, A. J. Micromech. Microeng. 2007, 17 (2), 238−244. (6) Kim, I.; Lee, H. Y.; Kim, H.; Lee, E.; Jeong, D. W.; Kim, J. J.; Park, S. H.; Ha, Y.; Na, J.; Chae, Y.; Yi, S.; Choi, H. J. Nano Lett. 2015, 15 (8), 5414−5419. (7) No, Y.; Gao, R.; Mankin, M. N.; Day, R. W.; Park, H.; Lieber, C. M. Nano Lett. 2016, 16 (7), 4713−4719. (8) Harris, M.Wkly Epidemiol Rec. 2016, 91, 13169−176. (9) Vanden Broeck, D.; Horvath, C.; De Wolf, M. J. Int. J. Biochem. Cell Biol. 2007, 39 (10), 1771−5. (10) Branson, T. R.; McAllister, T. E.; Garcia-Hartjes, J.; Fascione, M. A.; Ross, J. F.; Warriner, S. L.; Wennekes, T.; Zuilhof, H.; Turnbull, W. B. Angew. Chem., Int. Ed. 2014, 53 (32), 8323−8327. (11) Branson, T. R.; Turnbull, W. B. Chem. Soc. Rev. 2013, 42 (11), 4613−22. (12) Bernardi, A.; Jimenez-Barbero, J.; Casnati, A.; De Castro, C.; Darbre, T.; Fieschi, F.; Finne, J.; Funken, H.; Jaeger, K. E.; Lahmann, M.; Lindhorst, T. K.; Marradi, M.; Messner, P.; Molinaro, A.; Murphy, P. V.; Nativi, C.; Oscarson, S.; Penades, S.; Peri, F.; Pieters, R. J.; Renaudet, O.; Reymond, J. L.; Richichi, B.; Rojo, J.; Sansone, F.; Schaffer, C.; Turnbull, W. B.; Velasco-Torrijos, T.; Vidal, S.; Vincent, S.; Wennekes, T.; Zuilhof, H.; Imberty, A. Chem. Soc. Rev. 2013, 42 (11), 4709−4727. H
DOI: 10.1021/acs.nanolett.8b04775 Nano Lett. XXXX, XXX, XXX−XXX
