- Email: [email protected]
Glucose microfluidic biosensors based on reversible enzyme immobilization on photopatterned stimuli-responsive polymer
Biosensors and Bioelectronics 50 (2013) 229–234
Contents lists available at SciVerse ScienceDirect
Biosensors and Bioelectronics journal homepage: www.elsevier.com/locate/bios
Glucose microfluidic biosensors based on reversible enzyme immobilization on photopatterned stimuli-responsive polymer Meng Xiong a, Bin Gu b, Jia-Dong Zhang a, Jing-Juan Xu a, Hong-Yuan Chen a,n, Hui Zhong b,nn a
State Key Laboratory of Analytical Chemistry for Life Science, School of Chemistry and Chemical Engineering, Nanjing University, Nanjing 210093, China Jiangsu Key Laboratory for Chemistry of Low-dimensional Materials, School of Chemistry and Chemical Engineering, Huaiyin Normal University, Huaian 223300, China
b
art ic l e i nf o
a b s t r a c t
Article history: Received 12 March 2013 Received in revised form 15 June 2013 Accepted 17 June 2013 Available online 24 June 2013
In this paper, we demonstrate a new strategy for replaceable enzymatic microreactor based on a switchable wettability interface of poly(N-isopropylacrylamide) (PNIPAAm). PNIPAAm porous polymer monolith (PPM) with 3D macroporous framework is photopolymerized in glass microchip within 30 s. The PNIPAAm PPM not only shows its reversible swelling/shrinking property at the different temperature around the lower critical solution temperature (LCST), but also shows reversible hydrophilicity/ hydrophobicity corresponding to its swelling/shrinking status. Based on these properties, a biocompatible and replaceable on-chip enzymatic microreactor has been successfully built by means of the reversible adsorption and release of glucose oxidase (GOx) on the robust and stable matrix. Coupled with a carbon fiber microelectrode as electrochemical detector, the microreactor has been successfully employed for detection of glucose with a linear range from 0.05 to 5 mM. This approach may provide a promising way for high efficient and renewable microreactors that will find wide application in clinical diagnosis, biochemical synthesis/analysis, and proteomic research. & 2013 Elsevier B.V. All rights reserved.
Keywords: Photopatterning Poly(N-isopropylacrylamide) Enzyme immobilization Microfluidics Amperometry
1. Introduction Recently, research in enzyme-based microchip devices has received significant attention due to their powerful functions in biochemistry, bioanalysis, and clinical diagnosis (Chen et al., 2010; Chen et al., 2011, 2012; Li et al., 2010; Mersal and Bilitewski, 2005; Nomura et al., 2004; Sheng et al., 2012; Zhang et al., 2006a, 2006b, 2008). Up to date, enzyme immobilization methods in these devices can be summarized by three formats: (1) surface attachment, (2) beads packing, and (3) monoliths immobilization. Among these approaches, monolithic method has emerged as an attractive orientation to pattern enzymes in the network that can enhance the immobilization capacity, catalytic efficiency and avoid the low bioactivity from covalent binding (Miyazaki and Maeda, 2006). Even, it can use spatially controlled photopatterning technology to control the position of immobilized enzymes and improve the spatial resolution or precision (Tentori and Herr, 2011). However, the main limitation of porous polymer monolith (PPM) applied in enzyme immobilization is that the enzyme cannot be replaced (Mersal and Bilitewski, 2005). Until now, to
n
Corresponding author. Tel./fax: +86 25 83594862. Corresponding author. Tel.: +86 517 83525083; fax: +86 517 83525369. E-mail addresses: [email protected] (H.-Y. Chen), [email protected] (H. Zhong). nn
0956-5663/$ - see front matter & 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.bios.2013.06.030
the best of our knowledge, few works related to the renewable monoliths are reported. Stimuli-responsive polymers (Russell, 2002) exhibit reversible phase and surface structure changes in response to external stimuli in environmental factors such as pH (Xia et al., 2007), temperature (Sun et al., 2004), electric fields (Prins et al., 2001) and light (Ichimura et al., 2000). Usually, the properties of volume and surface wettability changes have wide applications in microfluidic systems (Beebe et al., 2000), drug delivery (Kiser et al., 1998), protein separation (Mu et al., 2007), cell culture (Akiyama et al., 2004), etc. Among these stimuli-responsive polymers, thermally responsive polymers may be one of the most commonly studied materials. And poly(N-isopropylacrylamide) (PNIPAAm) can be taken as a representative of a class of temperaturesensitive polymers characterized by a lower critical solution temperature (LCST) of 32 1C that is attributed to alterations in hydrogen-bonding interactions of the amide group, exhibiting not only the volume transition but also the surface wettability changes (Tokarev and Minko, 2009). Below LCST it is swollen, hydrated, and hydrophilic. Above LCST it is shrunken, dehydrated, and hydrophobic. Although the covalent immobilization (Klis et al., 2009), gel entrapment of enzymes (Hoffman and Stayton, 2007), and the reversible capture and release of proteins have been reported (Huber et al., 2003), it is still rare to study the controllable immobilization of enzymes in catalytic reaction via hydrophobic/ hydrophilic interaction (Choi et al., 2007). Sometimes, scientists
230
M. Xiong et al. / Biosensors and Bioelectronics 50 (2013) 229–234
focused too much on volumetric changes of PNIPAAm that can be designed to microfluidic valves and ignored their wettability conversions (Luo et al., 2003; Hisamoto et al., 2005; Yu et al., 2003) or the opposites individually. Thus researches of the two features combination may be a challenge for practical application, especially in microfluidic systems which offer many benefits, like high designation and operability, small quantities of samples and reagents, short times and high sensitivity for analysis (Whitesides, 2006). In this work, PNIPAAm polymers were selected as the substrate to fabricate the monoliths due to the macropores control and their two fantastic properties listed above. High spatial regulation and resolution of PNIPAAm PPM with 3D macroporous framework has been photopolymerized in glass microchip rapidly. The two special features of PNIPAAm PPM with reversible swelling/shrinking and hydrophilicity/hydrophobicity are combined to build a biocompatible and replaceable enzymatic microreactor-on-a-chip in short length with excellent performance that can be engineered for proteomics, medical diagnostics and biochemistry reactions in microfluidic systems.
Fig. 1. The photograph of the combined glass microchip (A) and the microscope images of the amplified cross-sections in the glass microchip (B,C).
2. Experimental 2.1. Fabrication of glass microchips Standard photolithography and wet chemical etching technique were used to fabricate glass microchips (Jia et al., 2004). Briefly, a design on a photomask with microchannels was transferred onto the glass plate (chromium and photoresist coating, SG2506, Shaoguang Microelectronics Corp., Changsha, China) by an UVlamp exposure (12 W, 365 nm, for 2 min), then following a developing step in 0.5% NaOH solution and an etching step in 1 M HF/ NH4F/HNO3 (1:1:0.5) solution at 40 1C. 2 mm diameter access holes were drilled on the etched plate at channel terminals using an emery drill bit to form reservoirs. In the room temperature bonding process of glass microchips, the glass substrates and polished cover plates were thoroughly scrubbed using acetone, and stored in a piranha solution (1:3H2O2:H2SO4) for half an hour. After that, the plates were washed sequentially with household dishwashing detergent, tap water and deionized water. Under the continuous stream of deionized water, etched substrates and cover plates were combined each other and placed on a horizontal hot platform without disturbing overnight. Two repeat patterns could be obtained in one glass plate (Fig. 1A). 2.2. Fabrication of carbon fiber microelectrode The carbon fiber electrode was fabricated as previously reported (Xu et al., 2004). Firstly, a glass capillary with inner diameter of 0.5 mm was pulled under a multifunctional glass microelectrode puller (Shanghai Biological Institute, China) to form a fine tip. Then a single carbon fiber with diameter of 8 μm was carefully inserted into the tip and fixed with dry glue. A copper wire was connected with the carbon fiber through carbon powder on the other end of capillary and then fastened with UV glue. Before use, the tip of the carbon fiber was cut with a clean scalpel to form a 2 mm long cylindrical electrode under a stereoscopic microscope. 2.3. In situ photopatterning macroporous polymer monolith Two glass microchips could be obtained in one glass plate by cutting it into two pieces (Fig. 1A). The microchip consists of one channel (separation channel, 20 μm deep, 130 μm wide and 26 mm long) with a double T structure for sample injection (20 μm deep,
90 μm wide and 5 mm long, Fig. 1B) and two branch channels for enzyme injection (20 μm deep, 130 μm wide and 5 mm long, Fig. 1C). The distance between the two vertical sample channels was 200 μm, and distance between the two branch channels could be changed to 5 mm, 6 mm and 8 mm in order to match the different length of photopatterned polymer monoliths, such as 1 mm, 2 mm, 4 mm and so on. The layout of the microfluidic device was shown in Scheme SI1. Then the surface chemistry of the channel wall was coated with acrylate-terminated selfassembled monolayer, and PPM fabrication was accomplished using in situ photopolymerization techniques. Briefly, the channels were conditioned with 0.2 M NaOH for 30 min, washed with deionized water and acetone for 15 min, and dried with N2. A 30 vol/vol% 3-(tri-methoxysilyl)-propyl-methacrylate (TMSPMA) acetone solution was loaded into the channels, and the reservoirs were sealed using 5 wt/vol% 2-hydroxyethyl cellulose (HEC) drops to prevent the solution from drying. The glass microchips were then incubated with the solution at room temperature in the dark for 24 h. The modified microchips were washed with acetone and dried with N2. At last, the prepolymer solution for fabricating PNIPAAm PPM was created by mixing 0.8 g N-isopropylacrylamide (Monomer, NIPAAm), 0.024 g N,N-methylene-bis-arylamide (Crosslinker, MBAAm), 18.6 μL 2-hydroxy-2-methylpropiophenone (Photoinitiator, PI), 2 mL CH3CH2OH and 2 mL deionized water. Noting that, the photoinitiator was added shortly before the fabrication step. The prepolymer solution purged with nitrogen gas for 5 min was loaded via capillary action. 5 wt/vol% HEC drops were used to cover all the outlets to stop hydrodynamic flow in the microchannel. The glass microchip was placed at 4 1C for 10 min. A photomask with 2 mm wide pattern was aligned to the channel immediately, then exposed to collimated UV light with 320–500 nm wavelength at an intensity of 5.6 mW cm−2 for 30 s. After PPM polymerization, unreacted chemicals were washed with phosphate buffer solution (PBS) and maintained in PBS via vacuum. The PPM microchips were stored in PBS at 4 1C for use. In order to achieve the microchip recycling and avoid the repetitive works in fabrication, the used PPM microchips were thoroughly immersed in a dissolution mixture (volume ratio, 1:2 H2O2:HClO4) container which was placed on a hot plate in a fume hood and incubated at 80 1C for 24 h (He and Herr, 2010). After taking out of the container, the microchips were immediately flushed with large amounts of water, and the channels were washed with deionized water using syringe. Noting that, before
M. Xiong et al. / Biosensors and Bioelectronics 50 (2013) 229–234
231
soaking, the buffer solution in the channel and polymer network must be replaced with the dissolution mixture through a syringe pump carefully.
3. Results and discussion
2.4. Preparation of monolithic enzymatic microreactor and procedure for electrophoresis
In order to make the monolith tightly anchor to the channel wall and tolerate the generally high voltage and high pressure, the glass chip was firstly silanized by silane coupling agents (TMSPMA) whose vinyl groups could be copolymerized with the monomers. The results of high voltage (≥ 300 V cm−1) test showed that the monolith in the unmodified channel was seriously damaged by Joule heating and electroosmotic flow (EOF) (Fig. SI1A). In contrast, the monolith almost kept intact (Fig. SI1B) with the covalent linkage at the same conditions, but just only turned opaquely. Therefore, vinyl silanization was necessary to strengthen the PPM stability in the later electrophoresis experiments (Li et al., 2010a). Meanwhile, the pumping pressure test showed that 780 μm PPM could stand for varied flow rate up to 10 μL min−1 (equivalent to a linear flow velocity of ∼64 mm/s) without any structural damage or dislocation of the plug, which was significantly greater than the commonly used velocity (0–5 μL min−1) in the microfluidic analysis above LCST temperature (Li et al., 2010b; Yu et al., 2001). As shown in Fig. SI2, static bubbles were difficult to pump in swell state of PPM at temperature below LCST (Fig. SI2A). After heating to 37 1C for a few seconds, the bubbles began to move rapidly through the monolith (Fig. SI2B) and arrived at the other end (Fig. SI2C). In addition, this test also illustrated the feasibility that microfluids could pass through the monolith above LCST, it was very useful to construct enzymatic microreactor. Because of the advantages of the spatially controlled photopatterning technology that enables the created regions to have shape controllability, arbitrary location and high spatial resolution structures (Tentori and Herr, 2011), a 2 mm PNIPAAm PPM have been photopolymerized between the two branch channels by 30 s with light intensity of 5.6 mW cm−2 at a distance of 10 cm (Fig. 2A), the defined edges and homogeneous body could be clearly seen in the magnified images, it might attribute to the collimated UV light which was used to avoid uneven polymerization and eliminate the expansion. Open and macroporous polymers with honeycomb cell morphology were formed (Fig. 2B), and the higher magnification SEM image showed that the adjacent macropores were interconnected to each other, which confirmed the 3D framework structure. High swelling/shrinking ratio of the polymer is beneficial for sample driving. It was reported that PNIPAAm synthesized in ethanol/water solvent with 2-hydroxy-2-methylpropiophenone as photoinitiator showed high swelling/shrinking ratio (Singh et al., 2006). Thus they are chosen to fabricate PPM in this work. To obtain the robust and well-dispersed 3D macropore framework for enzyme immobilization, various concentrations of monomer and crosslinker in the polymerization mixture listed in Table SI1 were optimized. As shown in Fig. SI3, broken cavities and aggregative scraps were observed in polymers A (Fig. SI3A) and D (Fig. SI3D) at lower crosslinker content. When dissolved in water, they were collapsed with bad strength. Jelly like polymers B and C were obtained as the crosslinker content increased, which was corresponding to decreased pore size (Fig. SI3B and Fig. SI3C). Although polymer C has a well morphology, the degree of shrinkage is significantly limited owing to the over crosslinking. So as expected, polymer E is suitable to fabricate PPM for robust, permeable structure and high swelling/shrinking ratio. In addition, porous properties of polymer D and E with the same monomer content were also investigated by mercury intrusion porosimetry and the results were shown in Fig. SI4 and Table SI2. A greater crosslinker content resulted in a smaller pore size as well as the larger porosity, pore volume and specific surface area. Moreover, polymer D could not endure higher pressure in the
The microchip experimental setup was illustrated in Scheme SI1. Prior to preparing enzymatic microreactor, the glass microchip was placed on a plexiglass holder platform attached with a thin PDMS sheet, and a small amount of silicon grease was used to prevent leakage. Then the tip of the carbon fiber electrode which was activated at a working potential of 1.5 and −1.0 V for 200 s was carefully plugged in the outlet of the separation channel with a distance ca. 40 μm under a stereoscopic microscope. Electrophoresis experiments were carried out using a programmable highvoltage supply (0–2000 V). Amperometry was performed in a three-electrode system on an electrochemical workstation 1030 A. The detailed information for detection system could be found in Supporting information (SI2). To fabricate the enzymatic microreactor in microchip, 10 mg mL−1 enzyme solutions were prepared in buffer and injected into the microchannel at a high voltage of 300 V between the two branch channels while the PPM was heated with Peltier to 37 1C for adsorbing the enzymes via hydrophobic interaction. After 15 min, the enzyme solutions in the reservoir and waste reservoir were exchanged with PBS, and run in another 10 min to clear away the unadsorbed enzyme. Prior to amperometric detection, sample reservoir (SR) and other reservoirs were replaced with fresh sample and PBS, respectively. Afterward, the injection was carried out by applying a high voltage of 300 V to the sample reservoir for 8 s, with the sample waste reservoir (SWR) grounded and the other reservoirs floated. Once the injection step was finished, the programmed power supply was switched to separation. The results of electrophoretic analysis were recorded on an electrochemical workstation 1030 A that was set in “i-t” mode at a working potential of 0.7 V and temperature of 37 1C. For human serum analysis, the proteins were removed by acetonitrile since they can be strongly adsorbed on PNIPAAm polymers surface. The procedure is as follows: 300 μL serum was mixed with 900 μL acetonitrile and centrifuged at a speed of 9000 r min−1 for 10 min, then the supernatant serum was collected, dried, dissolved in 600 μL PBS and finally filtered through a 0.22 μm syringe filter. 2.5. Enzyme immobilization and release In order to perform the adsorption and release behavior of enzyme on the PNIPAAm film, PNIPAAm polymers were synthesized on glass plates (see SI3). 200 μL of 10 mg mL−1 glucose oxidase (GOx) (25 mM PBS, pH 8.5) and 12.5 μL of fluorescein isothiocyanate isomer I (FITC) (10 mM, dissolved in DMSO) were added to 800 μL PBS (25 mM, pH 8.5) in a 1.5 mL vial. After incubation at room temperature for 24 h, the FITC-labeled GOx was purified by dialysis and stored in a refrigerator at 4 1C. PNIPAAm grafted glass slides were placed onto a heat plate with constant temperature of 37 1C, and then a PDMS frame matched with slide was fastened to form a little pool where FITC-labeled GOx solution was poured. After 15 min adsorption, the GOx-PNIPAAm grafted slide was rinsed with warm PBS (37 1C) and dried in vacuum at 37 1C. For enzyme release, this slide was immersed into PBS solution at 25 1C and rinsed carefully. All samples were dried by N2. The behavior of adsorption and release was evaluated with the images taken by an inverted fluorescence microscope equipped with a cooled CCD camera and analyzed by IPP software.
3.1. Characterization of photopatterning PPM
232
M. Xiong et al. / Biosensors and Bioelectronics 50 (2013) 229–234
Fig. 2. Microscope images of 2 mm photopatterned macroporous polymer monolith with high-definition edges (insets) in the microchannel (A) and the typical SEM of internal morphology of monolith prepared from polymer E shown in Table SI1 under the optimal irradiation condition, the inset shows a higher resolution image (B).
mercury intruded process and was compressed, thus the porosity data calculated for pore sizes between 0.02 and 187.1 μm was sharply distorted. From the data which is calculated for pore sizes between 0.1 and 10 μm, polymer E has a mode pore diameter of 2.67 μm, porosity of 83.8%, pore volume of 8.55 mL g−1 and specific surface area of 13.7 m2 g−1, which were much higher than previous reports (Peterson et al., 2002), suggesting that it should provide a good performance for enzyme immobilization, sample diffusion and catalytic reaction. The transverse section of polymer E above LCST for morphology was compact, rough and rigid. There were no pores in the PPM (Fig. SI5). 3.2. Characterization of the reversible properties of PNIPAAm In order to study the reversible enzyme immobilization on 3D macroporous framework of PNIPAAm via hydrophobic/hydrophilic interaction, the surface wettability was performed on a PNIPAAmgrafted slide using static water-contact angles measurements. During the temperature-dependent experiments, the responsive surface exhibited a water-contact angle of about 461 below 25 1C, whereas it was about 831 above 37 1C (Fig. 3, inset photographs), thus indicating a thermally responsive switching between hydrophilicity and hydrophobicity. Moreover, the change in contact angles reached as high as 371, significantly larger than that of other grafting (251) (Liang et al., 1998). The transition temperature of the polymer was estimated around 34 1C from the contact angle curve in Fig. 3. Bunker’s group utilized the excellently reversible wettability of uncrosslinked PNIPAAm to adsorb and release proteins, such as human serum albumin (HSA), bovine serum albumin (BSA), cytochrome C, myoglobin and hemoglobin (Huber et al., 2003). Here, a crosslinked PNIPAAm film grafted on a slide were fabricated to control the adsorption/release of enzyme. To demonstrate this reversible process, GOx was used as a model. Fouriertransform infrared absorption by attenuated total reflection spectra (ATR-FTIR) and fluorescence image were applied to characterize the GOx immobilization. Fig. SI6A displayed the typical characteristics of native GOx, the broad band at 3283 cm−1 belonged to the N–H stretching vibrations for amide A with a shoulder at 2880 cm−1 assigned to the symmetric and asymmetric vibrations of methyl group. The C ¼O stretching vibrations coupled to the stretching of C–N for amide I located at 1640 cm−1, amide II with N–H bending vibrations and C–N stretching appeared at 1536 cm−1 (Liu et al., 2006). The bands attributed to C–H bending vibration of methyl and methylene group were located at 1458 cm−1 and 1387 cm−1, whereas the band at 1097 cm−1 was related to the C–O stretching vibrations. Fig. SI6C also showed a standard spectrum of crosslinked PNIPAAm, which gave a similar absorption peak positions of amide I (1640 cm−1), amide II (1536 cm−1), methyl group (2880 cm−1 and 1458 cm−1) and
Fig. 3. Temperature dependence of water-contact angle for PNIPAAm thin film on polished glass substrate. The insets are the water drop profiles at low (25 1C), medium (35 1C) and high (45 1C) temperature.
methylene group (1387 cm−1). Besides, the peak ranging from 1360–1400 cm−1 corresponded to the bending vibrations of isopropyl group (Lin et al., 1999). After adsorption of GOx shown in Fig. SI6B, the new peaks corresponding to GOx appeared which were signed with black triangles, indicating that GOx was successfully immobilized. Oppositely, the blue arrows also pointed out the difference between curve A and B, indicating the existence of PNIPAAm. The reversible process was studied by fluorescence analysis (shown in Fig. 4). Obviously, there was no background fluorescence in PNIPAAm matrix (Fig. 4A), whereas a clearly green emission was observed after adsorption of fluorescein-labeled GOx above LCST (Fig. 4B), indicating the hydrophobic chains of PNIPAAm interacted strongly with enzyme. Following by washing steps at 25 1C, the labeled enzymes were released from the PNIPAAm film with the decreasing of intensity (Fig. 4C), indicating the hydrophilic conversion of PNIPAAm molecules below LCST. In general, these reversible properties are controlled by the interaction between intermolecular and intramolecular hydrogen bonding (Lin et al., 1999). The predominantly intermolecular hydrogen bonding between the PNIPAAm chains and water molecules below LCST leads to the hydrophilicity and swelling of PNIPAAm. At 37 1C, intramolecular hydrogen bonding between C ¼O and N–H in the PNIPAAm chains occurs and results in a hydrophobicity and shrinking of PNIPAAm. In addition, to investigate the reproducibility of this process, it was repeated for up to 9 cycles with a small decrease in the amount of adsorbed enzyme for the same PNIPAAm film (Fig. 4D). The loading amount of GOx on the PNIPAAm film is calculated to be ca. 1.065 10−3 mg mm−2 from UV–visible absorption spectroscopy data (see SI3, Fig. SI7).
M. Xiong et al. / Biosensors and Bioelectronics 50 (2013) 229–234
233
Fig. 4. Fluorescence microscopy images of PNIPAAm thin film (A), fluorescein-labeled GOx adsorbed on the PNIPAAm thin film at 37 1C (B) and released from PNIPAAm thin film at 25 1C (C). Intensity analysis of cyclic adsorption and release experiments (D). Conditions: exposure time 1.5 s, 10 objective lens. (For interpretation of the references to color in this figure caption, the reader is referred to the web version of this article.)
3.3. Preliminary test on enzymatic microreactor performance When glucose passed through the 3D macroporous framework that adsorbed GOx via hydrophobic interaction, they were enzymatically oxidized with dissolved oxygen and formed hydrogen peroxide, which could be electrochemically detected at the end of the separation channel. The detection potential of 0.70 V to H2O2 was optimized in Fig. SI9. The separation voltage also had a large influence on PPM microreactor. As discussed in the previous sections, the thermally responsive PNIPAAm PPM was susceptive to Joule heat at a voltage higher than 300 V cm−1. Although this problem could be relieved by covalent linkage, repeated electrophoresis for a long time might damage the 3D porous structure more or less. On the other hand, the increased voltage would accelerate the flow rate of sample, thus reduced the reaction time between enzyme and substrate. So the optimum separation voltage was finally selected as 600 V (equivalent to 225 V cm−1) in electrophoresis experiments. Fig. 5 showed the peak current of hydrogen peroxide as a function of the concentration of glucose on a 2 mm length of PPM microreactor. The current response increased linearly with the glucose concentration lower than 10 mM. The results exhibited a sensitive response to glucose with a linear range from 0.05 to 5 mM (r ¼0.998) and a detection limit of 20 μm. At higher glucose concentration, the response leveled off which demonstrates that glucose did not have enough time to react with oxygen (only 2 mm length), or there was no enough oxygen in the buffer solution to support the enzymatic reaction. However, compared with other reports, the linear range (Wang et al., 2010; Zhang et al., 2006a) and the detection limit (Zhang
Fig. 5. The electrochemical response to glucose concentration in the enzymatic microreactor. Inset: amplification signal of the linear range. Conditions: separation voltage: 600 V; sample injection: 300 V for 8 s; detection potential: 0.7 V; running buffer: phosphate buffer (25 mM, pH 7.4); temperature: 37 1C.
et al., 2012) were excellent in microchip bioanalysis for very short length of microreactor. The good performance could be explained that the larger number of enzymes immobilized and the faster mass transport led to higher catalytic efficiency in 3D macroporous framework. According to our group’s previous work (Chen et al., 2010; Zhang et al., 2006b), the linear ranges and detection limits could be further improved by increasing the microreactor length to prolong the reaction time.
234
M. Xiong et al. / Biosensors and Bioelectronics 50 (2013) 229–234
Furthermore, based on the advantages of PNIPAAm PPM, the enzymatic microreactor could be renewed. Once the enzymes lost their activity, they could be easily electrophoretically removed from the PPM by cooling to room temperature under the voltage of 100–300 V, and then fresh enzymes were loaded using the same steps in Section 2.4 at 37 1C. So the release and restoration of enzymatic activity of glucose measurement was provided to confirm this renewal process in Fig. SI9, and there were almost no obvious activity differences after restoration of enzymes. The operational stability of the immobilized enzyme was investigated by consecutively injecting 1.0 mM glucose solution over 20 times, and there was no obvious activity loss of the immobilized enzymes (Fig. SI10). The microreactor-tomicroreactor reproducibility to 1.0 mM glucose under the same conditions was 5.6% (n ¼6). Benefiting from the separation ability of this device, the response from the interferent such as ascorbic acid could be split from that of glucose (Fig. SI11). Finally, the microfluidic biosensor was used to determine the glucose concentration in human serum to investigate its feasibility for real sample analysis. The results were shown in Table SI3, which were consistent with those obtained by enzymatic photometric test on automatic biochemical analyzer. 4. Conclusions A smart 3D macroporous monolith was successfully fabricated by in situ spatially controlled photopatterning technology for constructing a replaceable on-chip enzymatic microreactor. It showed high resolution edges and homogeneous body. The 3D porous interconnected framework provides a high surface area and biocompatible microenvironment for enzyme loading, bioactivity maintenance and makes for the diffusion of substrate. The enzymatic microreactor with replaceable function has been achieved by the process of reversible adsorption/release under softly external stimuli. Therefore, it opens a new perspective for the development of highly efficient microreactors on microchips. Acknowledgments This work was supported by the 973 Program (Grant 2012CB932600), the National Natural Science Foundation (Grants 21135003 and 20975043), and the National Natural Science Funds for Creative Research Groups (Grant 21121091). The authors also wish to thank Professor Ning Bao, Nantong University, for his valuable comments and suggestions. Appendix A. Supporting information Supplementary data associated with this article can be found in the online version at http://dx.doi.org/10.1016/j.bios.2013.06.030.
References Akiyama, Y., Kikuchi, A., Yamato, M., Okano, T., 2004. Langmuir 20, 5506–5511. Beebe, D.J., Moore, J.S., Bauer, J.M., Yu, Q., Liu, R.H., Devadoss, C., Jo, B.-H., 2000. Nature 404, 588–590. Chen, L., Wu, Z.Q., Wang, C., Ouyang, J., Xia, X.H., 2012. Analytical Methods 4, 2831–2837. Chen, S.P., Yu, X.D., Xu, J.J., Chen, H.Y., 2010. Analyst 135, 2979–2986. Chen, S.P., Yu, X.D., Xu, J.J., Chen, H.Y., 2011. Biosensors and Bioelectronics 26, 4779–4784. Choi, H.G., Lee, J.H., Yoon, H.C., Lee, D.-S., 2007. BioChip Journal 1, 253–261. He, M., Herr, A.E., 2010. Nature Protocols 5, 1844–1856. Hisamoto, H., Funano, S.-i., Terabe, S., 2005. Analytical Chemistry 77, 2266–2271. Hoffman, A.S., Stayton, P.S., 2007. Progress in Polymer Science 32, 922–932. Huber, D.L., Manginell, R.P., Samara, M.A., Kim, B.-I., Bunker, B.C., 2003. Science 301, 352–354. Ichimura, K., Oh, S.-K., Nakagawa, M., 2000. Science 288, 1624–1626. Jia, Z.J., Fang, Q., Fang, Z.L., 2004. Analytical Chemistry 76, 5597–5602. Kiser, P.F., Wilson, G., Needham, D., 1998. Nature 394, 459–462. Klis, M., Karbarz, M., Stojek, Z., Rogalski, J., Bilewicz, R., 2009. Journal of Physical Chemistry B 113, 6062–6067. Li, S.J., Wang, C., Wu, Z.Q., Xu, J.J., Xia, X.H., Chen, H.Y., 2010. Chemistry—A European Journal 16, 10186–10194. Li, Z.M., He, Q.H., Ma, D., Chen, H.W., Soper, S.A., 2010a. Analytical Chemistry 82, 10030–10036. Li, Z.M., He, Q.H., Ma, D., Chen, H.W., 2010b. Analytica Chimica Acta 665, 107–112. Liang, L., Feng, X.D., Liu, J., Rieke, P.C., Fryxell, G.E., 1998. Macromolecules 31, 7845–7850. Lin, S.Y., Chen, K.S., Liang, R.C., 1999. Polymer 40, 2619–2624. Liu, A.L., Zhou, T., He, F.Y., Xu, J.J., Lu, Y., Chen, H.Y., Xia, X.H., 2006. Lab on Chip 6, 811–818. Luo, Q.Z., Mutlu, S., Gianchandani, Y.B., Svec, F., Fréchet, J.M.J., 2003. Electrophoresis 24, 3694–3702. Mersal, G.A.M., Bilitewski, U., 2005. Electrophoresis 26, 2303–2312. Miyazaki, M., Maeda, H., 2006. Trends in Biotechnology 24, 463–470. Mu, L., Liu, Y., Cai, S.Y., Kong, J.L., 2007. Chemistry—A European Journal 13, 5113–5120. Nomura, A., Shin, S., Mehdi, O.O., Kauffmann, J.-M., 2004. Analytical Chemistry 76, 5498–5502. Peterson, D.S., Rohr, T., Svec, F., Fréchet, J.M.J., 2002. Analytical Chemistry 74, 4081–4088. Prins, M.W.J., Welters, W.J.J., Weekamp, J.W., 2001. Science 291, 277–280. Russell, T.P., 2002. Science 297, 964–967. Sheng, J., Zhang, L., Lei, J.P., Ju, H.X., 2012. Analytica Chimica Acta 709, 41–46. Singh, D., Kuckling, D., Choudhary, V., Adler, H.-J., Koul, V., 2006. Polymers for Advanced Technologies 17, 186–192. Sun, T.L., Wang, G.J., Feng, L., Liu, B.Q., Ma, Y.M., Jiang, L., Zhu, D.B., 2004. Angewandte Chemie International Edition 43, 357–360. Tentori, A.M., Herr, A.E., 2011. Journal of Micromechanics and Microengineering 21, 054001–054014. Tokarev, I., Minko, S., 2009. Soft Matter 5, 511–524. Wang, Y., He, Q.H., Dong, Y.Y., Chen, H.W., 2010. Sensors and Actuators B—Chemical 145, 553–560. Whitesides, G.M., 2006. Nature 442, 368–373. Xia, F., Ge, H., Hou, Y., Sun, T.L., Chen, L., Zhang, G.Z., Jiang, L., 2007. Advanced Materials 19, 2520–2524. Xu, J.J., Bao, N., Xia, X.H., Peng, Y., Chen, H.Y., 2004. Analytical Chemistry 76, 6902–6907. Yu, C., Davey, M.H., Svec, F., Fréchet, J.M.J., 2001. Analytical Chemistry 73, 5088–5096. Yu, C., Mutlu, S., Selvaganapathy, P., Mastrangelo, C.H., Svec, F., Fréchet, J.M.J., 2003. Analytical Chemistry 75, 1958–1961. Zhang, L., Qu, P., Sheng, J., Lei, J.P., Ju, H.X., 2012. Chinese Journal of Chemistry 30, 2145–2150. Zhang, Q., Xu, J.J., Chen, H.Y., 2006a. Electrophoresis 27, 4943–4951. Zhang, Q., Xu, J.J., Chen, H.Y., 2006b. Journal of Chromatography A 1135, 122–126. Zhang, Q., Xu, J.J., Liu, Y., Chen, H.Y., 2008. Lab on Chip 8, 352–357.
Contents lists available at SciVerse ScienceDirect
Biosensors and Bioelectronics journal homepage: www.elsevier.com/locate/bios
Glucose microfluidic biosensors based on reversible enzyme immobilization on photopatterned stimuli-responsive polymer Meng Xiong a, Bin Gu b, Jia-Dong Zhang a, Jing-Juan Xu a, Hong-Yuan Chen a,n, Hui Zhong b,nn a
State Key Laboratory of Analytical Chemistry for Life Science, School of Chemistry and Chemical Engineering, Nanjing University, Nanjing 210093, China Jiangsu Key Laboratory for Chemistry of Low-dimensional Materials, School of Chemistry and Chemical Engineering, Huaiyin Normal University, Huaian 223300, China
b
art ic l e i nf o
a b s t r a c t
Article history: Received 12 March 2013 Received in revised form 15 June 2013 Accepted 17 June 2013 Available online 24 June 2013
In this paper, we demonstrate a new strategy for replaceable enzymatic microreactor based on a switchable wettability interface of poly(N-isopropylacrylamide) (PNIPAAm). PNIPAAm porous polymer monolith (PPM) with 3D macroporous framework is photopolymerized in glass microchip within 30 s. The PNIPAAm PPM not only shows its reversible swelling/shrinking property at the different temperature around the lower critical solution temperature (LCST), but also shows reversible hydrophilicity/ hydrophobicity corresponding to its swelling/shrinking status. Based on these properties, a biocompatible and replaceable on-chip enzymatic microreactor has been successfully built by means of the reversible adsorption and release of glucose oxidase (GOx) on the robust and stable matrix. Coupled with a carbon fiber microelectrode as electrochemical detector, the microreactor has been successfully employed for detection of glucose with a linear range from 0.05 to 5 mM. This approach may provide a promising way for high efficient and renewable microreactors that will find wide application in clinical diagnosis, biochemical synthesis/analysis, and proteomic research. & 2013 Elsevier B.V. All rights reserved.
Keywords: Photopatterning Poly(N-isopropylacrylamide) Enzyme immobilization Microfluidics Amperometry
1. Introduction Recently, research in enzyme-based microchip devices has received significant attention due to their powerful functions in biochemistry, bioanalysis, and clinical diagnosis (Chen et al., 2010; Chen et al., 2011, 2012; Li et al., 2010; Mersal and Bilitewski, 2005; Nomura et al., 2004; Sheng et al., 2012; Zhang et al., 2006a, 2006b, 2008). Up to date, enzyme immobilization methods in these devices can be summarized by three formats: (1) surface attachment, (2) beads packing, and (3) monoliths immobilization. Among these approaches, monolithic method has emerged as an attractive orientation to pattern enzymes in the network that can enhance the immobilization capacity, catalytic efficiency and avoid the low bioactivity from covalent binding (Miyazaki and Maeda, 2006). Even, it can use spatially controlled photopatterning technology to control the position of immobilized enzymes and improve the spatial resolution or precision (Tentori and Herr, 2011). However, the main limitation of porous polymer monolith (PPM) applied in enzyme immobilization is that the enzyme cannot be replaced (Mersal and Bilitewski, 2005). Until now, to
n
Corresponding author. Tel./fax: +86 25 83594862. Corresponding author. Tel.: +86 517 83525083; fax: +86 517 83525369. E-mail addresses: [email protected] (H.-Y. Chen), [email protected] (H. Zhong). nn
0956-5663/$ - see front matter & 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.bios.2013.06.030
the best of our knowledge, few works related to the renewable monoliths are reported. Stimuli-responsive polymers (Russell, 2002) exhibit reversible phase and surface structure changes in response to external stimuli in environmental factors such as pH (Xia et al., 2007), temperature (Sun et al., 2004), electric fields (Prins et al., 2001) and light (Ichimura et al., 2000). Usually, the properties of volume and surface wettability changes have wide applications in microfluidic systems (Beebe et al., 2000), drug delivery (Kiser et al., 1998), protein separation (Mu et al., 2007), cell culture (Akiyama et al., 2004), etc. Among these stimuli-responsive polymers, thermally responsive polymers may be one of the most commonly studied materials. And poly(N-isopropylacrylamide) (PNIPAAm) can be taken as a representative of a class of temperaturesensitive polymers characterized by a lower critical solution temperature (LCST) of 32 1C that is attributed to alterations in hydrogen-bonding interactions of the amide group, exhibiting not only the volume transition but also the surface wettability changes (Tokarev and Minko, 2009). Below LCST it is swollen, hydrated, and hydrophilic. Above LCST it is shrunken, dehydrated, and hydrophobic. Although the covalent immobilization (Klis et al., 2009), gel entrapment of enzymes (Hoffman and Stayton, 2007), and the reversible capture and release of proteins have been reported (Huber et al., 2003), it is still rare to study the controllable immobilization of enzymes in catalytic reaction via hydrophobic/ hydrophilic interaction (Choi et al., 2007). Sometimes, scientists
230
M. Xiong et al. / Biosensors and Bioelectronics 50 (2013) 229–234
focused too much on volumetric changes of PNIPAAm that can be designed to microfluidic valves and ignored their wettability conversions (Luo et al., 2003; Hisamoto et al., 2005; Yu et al., 2003) or the opposites individually. Thus researches of the two features combination may be a challenge for practical application, especially in microfluidic systems which offer many benefits, like high designation and operability, small quantities of samples and reagents, short times and high sensitivity for analysis (Whitesides, 2006). In this work, PNIPAAm polymers were selected as the substrate to fabricate the monoliths due to the macropores control and their two fantastic properties listed above. High spatial regulation and resolution of PNIPAAm PPM with 3D macroporous framework has been photopolymerized in glass microchip rapidly. The two special features of PNIPAAm PPM with reversible swelling/shrinking and hydrophilicity/hydrophobicity are combined to build a biocompatible and replaceable enzymatic microreactor-on-a-chip in short length with excellent performance that can be engineered for proteomics, medical diagnostics and biochemistry reactions in microfluidic systems.
Fig. 1. The photograph of the combined glass microchip (A) and the microscope images of the amplified cross-sections in the glass microchip (B,C).
2. Experimental 2.1. Fabrication of glass microchips Standard photolithography and wet chemical etching technique were used to fabricate glass microchips (Jia et al., 2004). Briefly, a design on a photomask with microchannels was transferred onto the glass plate (chromium and photoresist coating, SG2506, Shaoguang Microelectronics Corp., Changsha, China) by an UVlamp exposure (12 W, 365 nm, for 2 min), then following a developing step in 0.5% NaOH solution and an etching step in 1 M HF/ NH4F/HNO3 (1:1:0.5) solution at 40 1C. 2 mm diameter access holes were drilled on the etched plate at channel terminals using an emery drill bit to form reservoirs. In the room temperature bonding process of glass microchips, the glass substrates and polished cover plates were thoroughly scrubbed using acetone, and stored in a piranha solution (1:3H2O2:H2SO4) for half an hour. After that, the plates were washed sequentially with household dishwashing detergent, tap water and deionized water. Under the continuous stream of deionized water, etched substrates and cover plates were combined each other and placed on a horizontal hot platform without disturbing overnight. Two repeat patterns could be obtained in one glass plate (Fig. 1A). 2.2. Fabrication of carbon fiber microelectrode The carbon fiber electrode was fabricated as previously reported (Xu et al., 2004). Firstly, a glass capillary with inner diameter of 0.5 mm was pulled under a multifunctional glass microelectrode puller (Shanghai Biological Institute, China) to form a fine tip. Then a single carbon fiber with diameter of 8 μm was carefully inserted into the tip and fixed with dry glue. A copper wire was connected with the carbon fiber through carbon powder on the other end of capillary and then fastened with UV glue. Before use, the tip of the carbon fiber was cut with a clean scalpel to form a 2 mm long cylindrical electrode under a stereoscopic microscope. 2.3. In situ photopatterning macroporous polymer monolith Two glass microchips could be obtained in one glass plate by cutting it into two pieces (Fig. 1A). The microchip consists of one channel (separation channel, 20 μm deep, 130 μm wide and 26 mm long) with a double T structure for sample injection (20 μm deep,
90 μm wide and 5 mm long, Fig. 1B) and two branch channels for enzyme injection (20 μm deep, 130 μm wide and 5 mm long, Fig. 1C). The distance between the two vertical sample channels was 200 μm, and distance between the two branch channels could be changed to 5 mm, 6 mm and 8 mm in order to match the different length of photopatterned polymer monoliths, such as 1 mm, 2 mm, 4 mm and so on. The layout of the microfluidic device was shown in Scheme SI1. Then the surface chemistry of the channel wall was coated with acrylate-terminated selfassembled monolayer, and PPM fabrication was accomplished using in situ photopolymerization techniques. Briefly, the channels were conditioned with 0.2 M NaOH for 30 min, washed with deionized water and acetone for 15 min, and dried with N2. A 30 vol/vol% 3-(tri-methoxysilyl)-propyl-methacrylate (TMSPMA) acetone solution was loaded into the channels, and the reservoirs were sealed using 5 wt/vol% 2-hydroxyethyl cellulose (HEC) drops to prevent the solution from drying. The glass microchips were then incubated with the solution at room temperature in the dark for 24 h. The modified microchips were washed with acetone and dried with N2. At last, the prepolymer solution for fabricating PNIPAAm PPM was created by mixing 0.8 g N-isopropylacrylamide (Monomer, NIPAAm), 0.024 g N,N-methylene-bis-arylamide (Crosslinker, MBAAm), 18.6 μL 2-hydroxy-2-methylpropiophenone (Photoinitiator, PI), 2 mL CH3CH2OH and 2 mL deionized water. Noting that, the photoinitiator was added shortly before the fabrication step. The prepolymer solution purged with nitrogen gas for 5 min was loaded via capillary action. 5 wt/vol% HEC drops were used to cover all the outlets to stop hydrodynamic flow in the microchannel. The glass microchip was placed at 4 1C for 10 min. A photomask with 2 mm wide pattern was aligned to the channel immediately, then exposed to collimated UV light with 320–500 nm wavelength at an intensity of 5.6 mW cm−2 for 30 s. After PPM polymerization, unreacted chemicals were washed with phosphate buffer solution (PBS) and maintained in PBS via vacuum. The PPM microchips were stored in PBS at 4 1C for use. In order to achieve the microchip recycling and avoid the repetitive works in fabrication, the used PPM microchips were thoroughly immersed in a dissolution mixture (volume ratio, 1:2 H2O2:HClO4) container which was placed on a hot plate in a fume hood and incubated at 80 1C for 24 h (He and Herr, 2010). After taking out of the container, the microchips were immediately flushed with large amounts of water, and the channels were washed with deionized water using syringe. Noting that, before
M. Xiong et al. / Biosensors and Bioelectronics 50 (2013) 229–234
231
soaking, the buffer solution in the channel and polymer network must be replaced with the dissolution mixture through a syringe pump carefully.
3. Results and discussion
2.4. Preparation of monolithic enzymatic microreactor and procedure for electrophoresis
In order to make the monolith tightly anchor to the channel wall and tolerate the generally high voltage and high pressure, the glass chip was firstly silanized by silane coupling agents (TMSPMA) whose vinyl groups could be copolymerized with the monomers. The results of high voltage (≥ 300 V cm−1) test showed that the monolith in the unmodified channel was seriously damaged by Joule heating and electroosmotic flow (EOF) (Fig. SI1A). In contrast, the monolith almost kept intact (Fig. SI1B) with the covalent linkage at the same conditions, but just only turned opaquely. Therefore, vinyl silanization was necessary to strengthen the PPM stability in the later electrophoresis experiments (Li et al., 2010a). Meanwhile, the pumping pressure test showed that 780 μm PPM could stand for varied flow rate up to 10 μL min−1 (equivalent to a linear flow velocity of ∼64 mm/s) without any structural damage or dislocation of the plug, which was significantly greater than the commonly used velocity (0–5 μL min−1) in the microfluidic analysis above LCST temperature (Li et al., 2010b; Yu et al., 2001). As shown in Fig. SI2, static bubbles were difficult to pump in swell state of PPM at temperature below LCST (Fig. SI2A). After heating to 37 1C for a few seconds, the bubbles began to move rapidly through the monolith (Fig. SI2B) and arrived at the other end (Fig. SI2C). In addition, this test also illustrated the feasibility that microfluids could pass through the monolith above LCST, it was very useful to construct enzymatic microreactor. Because of the advantages of the spatially controlled photopatterning technology that enables the created regions to have shape controllability, arbitrary location and high spatial resolution structures (Tentori and Herr, 2011), a 2 mm PNIPAAm PPM have been photopolymerized between the two branch channels by 30 s with light intensity of 5.6 mW cm−2 at a distance of 10 cm (Fig. 2A), the defined edges and homogeneous body could be clearly seen in the magnified images, it might attribute to the collimated UV light which was used to avoid uneven polymerization and eliminate the expansion. Open and macroporous polymers with honeycomb cell morphology were formed (Fig. 2B), and the higher magnification SEM image showed that the adjacent macropores were interconnected to each other, which confirmed the 3D framework structure. High swelling/shrinking ratio of the polymer is beneficial for sample driving. It was reported that PNIPAAm synthesized in ethanol/water solvent with 2-hydroxy-2-methylpropiophenone as photoinitiator showed high swelling/shrinking ratio (Singh et al., 2006). Thus they are chosen to fabricate PPM in this work. To obtain the robust and well-dispersed 3D macropore framework for enzyme immobilization, various concentrations of monomer and crosslinker in the polymerization mixture listed in Table SI1 were optimized. As shown in Fig. SI3, broken cavities and aggregative scraps were observed in polymers A (Fig. SI3A) and D (Fig. SI3D) at lower crosslinker content. When dissolved in water, they were collapsed with bad strength. Jelly like polymers B and C were obtained as the crosslinker content increased, which was corresponding to decreased pore size (Fig. SI3B and Fig. SI3C). Although polymer C has a well morphology, the degree of shrinkage is significantly limited owing to the over crosslinking. So as expected, polymer E is suitable to fabricate PPM for robust, permeable structure and high swelling/shrinking ratio. In addition, porous properties of polymer D and E with the same monomer content were also investigated by mercury intrusion porosimetry and the results were shown in Fig. SI4 and Table SI2. A greater crosslinker content resulted in a smaller pore size as well as the larger porosity, pore volume and specific surface area. Moreover, polymer D could not endure higher pressure in the
The microchip experimental setup was illustrated in Scheme SI1. Prior to preparing enzymatic microreactor, the glass microchip was placed on a plexiglass holder platform attached with a thin PDMS sheet, and a small amount of silicon grease was used to prevent leakage. Then the tip of the carbon fiber electrode which was activated at a working potential of 1.5 and −1.0 V for 200 s was carefully plugged in the outlet of the separation channel with a distance ca. 40 μm under a stereoscopic microscope. Electrophoresis experiments were carried out using a programmable highvoltage supply (0–2000 V). Amperometry was performed in a three-electrode system on an electrochemical workstation 1030 A. The detailed information for detection system could be found in Supporting information (SI2). To fabricate the enzymatic microreactor in microchip, 10 mg mL−1 enzyme solutions were prepared in buffer and injected into the microchannel at a high voltage of 300 V between the two branch channels while the PPM was heated with Peltier to 37 1C for adsorbing the enzymes via hydrophobic interaction. After 15 min, the enzyme solutions in the reservoir and waste reservoir were exchanged with PBS, and run in another 10 min to clear away the unadsorbed enzyme. Prior to amperometric detection, sample reservoir (SR) and other reservoirs were replaced with fresh sample and PBS, respectively. Afterward, the injection was carried out by applying a high voltage of 300 V to the sample reservoir for 8 s, with the sample waste reservoir (SWR) grounded and the other reservoirs floated. Once the injection step was finished, the programmed power supply was switched to separation. The results of electrophoretic analysis were recorded on an electrochemical workstation 1030 A that was set in “i-t” mode at a working potential of 0.7 V and temperature of 37 1C. For human serum analysis, the proteins were removed by acetonitrile since they can be strongly adsorbed on PNIPAAm polymers surface. The procedure is as follows: 300 μL serum was mixed with 900 μL acetonitrile and centrifuged at a speed of 9000 r min−1 for 10 min, then the supernatant serum was collected, dried, dissolved in 600 μL PBS and finally filtered through a 0.22 μm syringe filter. 2.5. Enzyme immobilization and release In order to perform the adsorption and release behavior of enzyme on the PNIPAAm film, PNIPAAm polymers were synthesized on glass plates (see SI3). 200 μL of 10 mg mL−1 glucose oxidase (GOx) (25 mM PBS, pH 8.5) and 12.5 μL of fluorescein isothiocyanate isomer I (FITC) (10 mM, dissolved in DMSO) were added to 800 μL PBS (25 mM, pH 8.5) in a 1.5 mL vial. After incubation at room temperature for 24 h, the FITC-labeled GOx was purified by dialysis and stored in a refrigerator at 4 1C. PNIPAAm grafted glass slides were placed onto a heat plate with constant temperature of 37 1C, and then a PDMS frame matched with slide was fastened to form a little pool where FITC-labeled GOx solution was poured. After 15 min adsorption, the GOx-PNIPAAm grafted slide was rinsed with warm PBS (37 1C) and dried in vacuum at 37 1C. For enzyme release, this slide was immersed into PBS solution at 25 1C and rinsed carefully. All samples were dried by N2. The behavior of adsorption and release was evaluated with the images taken by an inverted fluorescence microscope equipped with a cooled CCD camera and analyzed by IPP software.
3.1. Characterization of photopatterning PPM
232
M. Xiong et al. / Biosensors and Bioelectronics 50 (2013) 229–234
Fig. 2. Microscope images of 2 mm photopatterned macroporous polymer monolith with high-definition edges (insets) in the microchannel (A) and the typical SEM of internal morphology of monolith prepared from polymer E shown in Table SI1 under the optimal irradiation condition, the inset shows a higher resolution image (B).
mercury intruded process and was compressed, thus the porosity data calculated for pore sizes between 0.02 and 187.1 μm was sharply distorted. From the data which is calculated for pore sizes between 0.1 and 10 μm, polymer E has a mode pore diameter of 2.67 μm, porosity of 83.8%, pore volume of 8.55 mL g−1 and specific surface area of 13.7 m2 g−1, which were much higher than previous reports (Peterson et al., 2002), suggesting that it should provide a good performance for enzyme immobilization, sample diffusion and catalytic reaction. The transverse section of polymer E above LCST for morphology was compact, rough and rigid. There were no pores in the PPM (Fig. SI5). 3.2. Characterization of the reversible properties of PNIPAAm In order to study the reversible enzyme immobilization on 3D macroporous framework of PNIPAAm via hydrophobic/hydrophilic interaction, the surface wettability was performed on a PNIPAAmgrafted slide using static water-contact angles measurements. During the temperature-dependent experiments, the responsive surface exhibited a water-contact angle of about 461 below 25 1C, whereas it was about 831 above 37 1C (Fig. 3, inset photographs), thus indicating a thermally responsive switching between hydrophilicity and hydrophobicity. Moreover, the change in contact angles reached as high as 371, significantly larger than that of other grafting (251) (Liang et al., 1998). The transition temperature of the polymer was estimated around 34 1C from the contact angle curve in Fig. 3. Bunker’s group utilized the excellently reversible wettability of uncrosslinked PNIPAAm to adsorb and release proteins, such as human serum albumin (HSA), bovine serum albumin (BSA), cytochrome C, myoglobin and hemoglobin (Huber et al., 2003). Here, a crosslinked PNIPAAm film grafted on a slide were fabricated to control the adsorption/release of enzyme. To demonstrate this reversible process, GOx was used as a model. Fouriertransform infrared absorption by attenuated total reflection spectra (ATR-FTIR) and fluorescence image were applied to characterize the GOx immobilization. Fig. SI6A displayed the typical characteristics of native GOx, the broad band at 3283 cm−1 belonged to the N–H stretching vibrations for amide A with a shoulder at 2880 cm−1 assigned to the symmetric and asymmetric vibrations of methyl group. The C ¼O stretching vibrations coupled to the stretching of C–N for amide I located at 1640 cm−1, amide II with N–H bending vibrations and C–N stretching appeared at 1536 cm−1 (Liu et al., 2006). The bands attributed to C–H bending vibration of methyl and methylene group were located at 1458 cm−1 and 1387 cm−1, whereas the band at 1097 cm−1 was related to the C–O stretching vibrations. Fig. SI6C also showed a standard spectrum of crosslinked PNIPAAm, which gave a similar absorption peak positions of amide I (1640 cm−1), amide II (1536 cm−1), methyl group (2880 cm−1 and 1458 cm−1) and
Fig. 3. Temperature dependence of water-contact angle for PNIPAAm thin film on polished glass substrate. The insets are the water drop profiles at low (25 1C), medium (35 1C) and high (45 1C) temperature.
methylene group (1387 cm−1). Besides, the peak ranging from 1360–1400 cm−1 corresponded to the bending vibrations of isopropyl group (Lin et al., 1999). After adsorption of GOx shown in Fig. SI6B, the new peaks corresponding to GOx appeared which were signed with black triangles, indicating that GOx was successfully immobilized. Oppositely, the blue arrows also pointed out the difference between curve A and B, indicating the existence of PNIPAAm. The reversible process was studied by fluorescence analysis (shown in Fig. 4). Obviously, there was no background fluorescence in PNIPAAm matrix (Fig. 4A), whereas a clearly green emission was observed after adsorption of fluorescein-labeled GOx above LCST (Fig. 4B), indicating the hydrophobic chains of PNIPAAm interacted strongly with enzyme. Following by washing steps at 25 1C, the labeled enzymes were released from the PNIPAAm film with the decreasing of intensity (Fig. 4C), indicating the hydrophilic conversion of PNIPAAm molecules below LCST. In general, these reversible properties are controlled by the interaction between intermolecular and intramolecular hydrogen bonding (Lin et al., 1999). The predominantly intermolecular hydrogen bonding between the PNIPAAm chains and water molecules below LCST leads to the hydrophilicity and swelling of PNIPAAm. At 37 1C, intramolecular hydrogen bonding between C ¼O and N–H in the PNIPAAm chains occurs and results in a hydrophobicity and shrinking of PNIPAAm. In addition, to investigate the reproducibility of this process, it was repeated for up to 9 cycles with a small decrease in the amount of adsorbed enzyme for the same PNIPAAm film (Fig. 4D). The loading amount of GOx on the PNIPAAm film is calculated to be ca. 1.065 10−3 mg mm−2 from UV–visible absorption spectroscopy data (see SI3, Fig. SI7).
M. Xiong et al. / Biosensors and Bioelectronics 50 (2013) 229–234
233
Fig. 4. Fluorescence microscopy images of PNIPAAm thin film (A), fluorescein-labeled GOx adsorbed on the PNIPAAm thin film at 37 1C (B) and released from PNIPAAm thin film at 25 1C (C). Intensity analysis of cyclic adsorption and release experiments (D). Conditions: exposure time 1.5 s, 10 objective lens. (For interpretation of the references to color in this figure caption, the reader is referred to the web version of this article.)
3.3. Preliminary test on enzymatic microreactor performance When glucose passed through the 3D macroporous framework that adsorbed GOx via hydrophobic interaction, they were enzymatically oxidized with dissolved oxygen and formed hydrogen peroxide, which could be electrochemically detected at the end of the separation channel. The detection potential of 0.70 V to H2O2 was optimized in Fig. SI9. The separation voltage also had a large influence on PPM microreactor. As discussed in the previous sections, the thermally responsive PNIPAAm PPM was susceptive to Joule heat at a voltage higher than 300 V cm−1. Although this problem could be relieved by covalent linkage, repeated electrophoresis for a long time might damage the 3D porous structure more or less. On the other hand, the increased voltage would accelerate the flow rate of sample, thus reduced the reaction time between enzyme and substrate. So the optimum separation voltage was finally selected as 600 V (equivalent to 225 V cm−1) in electrophoresis experiments. Fig. 5 showed the peak current of hydrogen peroxide as a function of the concentration of glucose on a 2 mm length of PPM microreactor. The current response increased linearly with the glucose concentration lower than 10 mM. The results exhibited a sensitive response to glucose with a linear range from 0.05 to 5 mM (r ¼0.998) and a detection limit of 20 μm. At higher glucose concentration, the response leveled off which demonstrates that glucose did not have enough time to react with oxygen (only 2 mm length), or there was no enough oxygen in the buffer solution to support the enzymatic reaction. However, compared with other reports, the linear range (Wang et al., 2010; Zhang et al., 2006a) and the detection limit (Zhang
Fig. 5. The electrochemical response to glucose concentration in the enzymatic microreactor. Inset: amplification signal of the linear range. Conditions: separation voltage: 600 V; sample injection: 300 V for 8 s; detection potential: 0.7 V; running buffer: phosphate buffer (25 mM, pH 7.4); temperature: 37 1C.
et al., 2012) were excellent in microchip bioanalysis for very short length of microreactor. The good performance could be explained that the larger number of enzymes immobilized and the faster mass transport led to higher catalytic efficiency in 3D macroporous framework. According to our group’s previous work (Chen et al., 2010; Zhang et al., 2006b), the linear ranges and detection limits could be further improved by increasing the microreactor length to prolong the reaction time.
234
M. Xiong et al. / Biosensors and Bioelectronics 50 (2013) 229–234
Furthermore, based on the advantages of PNIPAAm PPM, the enzymatic microreactor could be renewed. Once the enzymes lost their activity, they could be easily electrophoretically removed from the PPM by cooling to room temperature under the voltage of 100–300 V, and then fresh enzymes were loaded using the same steps in Section 2.4 at 37 1C. So the release and restoration of enzymatic activity of glucose measurement was provided to confirm this renewal process in Fig. SI9, and there were almost no obvious activity differences after restoration of enzymes. The operational stability of the immobilized enzyme was investigated by consecutively injecting 1.0 mM glucose solution over 20 times, and there was no obvious activity loss of the immobilized enzymes (Fig. SI10). The microreactor-tomicroreactor reproducibility to 1.0 mM glucose under the same conditions was 5.6% (n ¼6). Benefiting from the separation ability of this device, the response from the interferent such as ascorbic acid could be split from that of glucose (Fig. SI11). Finally, the microfluidic biosensor was used to determine the glucose concentration in human serum to investigate its feasibility for real sample analysis. The results were shown in Table SI3, which were consistent with those obtained by enzymatic photometric test on automatic biochemical analyzer. 4. Conclusions A smart 3D macroporous monolith was successfully fabricated by in situ spatially controlled photopatterning technology for constructing a replaceable on-chip enzymatic microreactor. It showed high resolution edges and homogeneous body. The 3D porous interconnected framework provides a high surface area and biocompatible microenvironment for enzyme loading, bioactivity maintenance and makes for the diffusion of substrate. The enzymatic microreactor with replaceable function has been achieved by the process of reversible adsorption/release under softly external stimuli. Therefore, it opens a new perspective for the development of highly efficient microreactors on microchips. Acknowledgments This work was supported by the 973 Program (Grant 2012CB932600), the National Natural Science Foundation (Grants 21135003 and 20975043), and the National Natural Science Funds for Creative Research Groups (Grant 21121091). The authors also wish to thank Professor Ning Bao, Nantong University, for his valuable comments and suggestions. Appendix A. Supporting information Supplementary data associated with this article can be found in the online version at http://dx.doi.org/10.1016/j.bios.2013.06.030.
References Akiyama, Y., Kikuchi, A., Yamato, M., Okano, T., 2004. Langmuir 20, 5506–5511. Beebe, D.J., Moore, J.S., Bauer, J.M., Yu, Q., Liu, R.H., Devadoss, C., Jo, B.-H., 2000. Nature 404, 588–590. Chen, L., Wu, Z.Q., Wang, C., Ouyang, J., Xia, X.H., 2012. Analytical Methods 4, 2831–2837. Chen, S.P., Yu, X.D., Xu, J.J., Chen, H.Y., 2010. Analyst 135, 2979–2986. Chen, S.P., Yu, X.D., Xu, J.J., Chen, H.Y., 2011. Biosensors and Bioelectronics 26, 4779–4784. Choi, H.G., Lee, J.H., Yoon, H.C., Lee, D.-S., 2007. BioChip Journal 1, 253–261. He, M., Herr, A.E., 2010. Nature Protocols 5, 1844–1856. Hisamoto, H., Funano, S.-i., Terabe, S., 2005. Analytical Chemistry 77, 2266–2271. Hoffman, A.S., Stayton, P.S., 2007. Progress in Polymer Science 32, 922–932. Huber, D.L., Manginell, R.P., Samara, M.A., Kim, B.-I., Bunker, B.C., 2003. Science 301, 352–354. Ichimura, K., Oh, S.-K., Nakagawa, M., 2000. Science 288, 1624–1626. Jia, Z.J., Fang, Q., Fang, Z.L., 2004. Analytical Chemistry 76, 5597–5602. Kiser, P.F., Wilson, G., Needham, D., 1998. Nature 394, 459–462. Klis, M., Karbarz, M., Stojek, Z., Rogalski, J., Bilewicz, R., 2009. Journal of Physical Chemistry B 113, 6062–6067. Li, S.J., Wang, C., Wu, Z.Q., Xu, J.J., Xia, X.H., Chen, H.Y., 2010. Chemistry—A European Journal 16, 10186–10194. Li, Z.M., He, Q.H., Ma, D., Chen, H.W., Soper, S.A., 2010a. Analytical Chemistry 82, 10030–10036. Li, Z.M., He, Q.H., Ma, D., Chen, H.W., 2010b. Analytica Chimica Acta 665, 107–112. Liang, L., Feng, X.D., Liu, J., Rieke, P.C., Fryxell, G.E., 1998. Macromolecules 31, 7845–7850. Lin, S.Y., Chen, K.S., Liang, R.C., 1999. Polymer 40, 2619–2624. Liu, A.L., Zhou, T., He, F.Y., Xu, J.J., Lu, Y., Chen, H.Y., Xia, X.H., 2006. Lab on Chip 6, 811–818. Luo, Q.Z., Mutlu, S., Gianchandani, Y.B., Svec, F., Fréchet, J.M.J., 2003. Electrophoresis 24, 3694–3702. Mersal, G.A.M., Bilitewski, U., 2005. Electrophoresis 26, 2303–2312. Miyazaki, M., Maeda, H., 2006. Trends in Biotechnology 24, 463–470. Mu, L., Liu, Y., Cai, S.Y., Kong, J.L., 2007. Chemistry—A European Journal 13, 5113–5120. Nomura, A., Shin, S., Mehdi, O.O., Kauffmann, J.-M., 2004. Analytical Chemistry 76, 5498–5502. Peterson, D.S., Rohr, T., Svec, F., Fréchet, J.M.J., 2002. Analytical Chemistry 74, 4081–4088. Prins, M.W.J., Welters, W.J.J., Weekamp, J.W., 2001. Science 291, 277–280. Russell, T.P., 2002. Science 297, 964–967. Sheng, J., Zhang, L., Lei, J.P., Ju, H.X., 2012. Analytica Chimica Acta 709, 41–46. Singh, D., Kuckling, D., Choudhary, V., Adler, H.-J., Koul, V., 2006. Polymers for Advanced Technologies 17, 186–192. Sun, T.L., Wang, G.J., Feng, L., Liu, B.Q., Ma, Y.M., Jiang, L., Zhu, D.B., 2004. Angewandte Chemie International Edition 43, 357–360. Tentori, A.M., Herr, A.E., 2011. Journal of Micromechanics and Microengineering 21, 054001–054014. Tokarev, I., Minko, S., 2009. Soft Matter 5, 511–524. Wang, Y., He, Q.H., Dong, Y.Y., Chen, H.W., 2010. Sensors and Actuators B—Chemical 145, 553–560. Whitesides, G.M., 2006. Nature 442, 368–373. Xia, F., Ge, H., Hou, Y., Sun, T.L., Chen, L., Zhang, G.Z., Jiang, L., 2007. Advanced Materials 19, 2520–2524. Xu, J.J., Bao, N., Xia, X.H., Peng, Y., Chen, H.Y., 2004. Analytical Chemistry 76, 6902–6907. Yu, C., Davey, M.H., Svec, F., Fréchet, J.M.J., 2001. Analytical Chemistry 73, 5088–5096. Yu, C., Mutlu, S., Selvaganapathy, P., Mastrangelo, C.H., Svec, F., Fréchet, J.M.J., 2003. Analytical Chemistry 75, 1958–1961. Zhang, L., Qu, P., Sheng, J., Lei, J.P., Ju, H.X., 2012. Chinese Journal of Chemistry 30, 2145–2150. Zhang, Q., Xu, J.J., Chen, H.Y., 2006a. Electrophoresis 27, 4943–4951. Zhang, Q., Xu, J.J., Chen, H.Y., 2006b. Journal of Chromatography A 1135, 122–126. Zhang, Q., Xu, J.J., Liu, Y., Chen, H.Y., 2008. Lab on Chip 8, 352–357.