LETTER pubs.acs.org/ac
Biomolecule-Recognition Gating Membrane Using Biomolecular Cross-Linking and Polymer Phase Transition Hidenori Kuroki,† Taichi Ito,‡ Hidenori Ohashi,† Takanori Tamaki,† and Takeo Yamaguchi*,† †
Chemical Resources Laboratory, Tokyo Institute of Technology, 4259 Nagatsuta-cho, Midori-ku, Yokohama, Kanagawa, 226-8503, Japan ‡ Center for Disease Biology and Integrative Medicine, The University of Tokyo, 7-3-1 Hongo, Bunkyo-ku, Tokyo 113-8656, Japan
bS Supporting Information ABSTRACT: We present for the first time a biomoleculerecognition gating system that responds to small signals of biomolecules by the cooperation of biorecognition cross-linking and polymer phase transition in nanosized pores. The biomolecule-recognition gating membrane immobilizes the stimuliresponsive polymer, including the biomolecule-recognition receptor, onto the pore surface of a porous membrane. The pore state (open/closed) of this gating membrane depends on the formation of specific biorecognition cross-linking in the pores: a specific biomolecule having multibinding sites can be recognized by several receptors and acts as the cross-linker of the grafted polymer, whereas a nonspecific molecule cannot. The pore state can be distinguished by a volume phase transition of the grafted polymer. In the present study, the principle of the proposed system is demonstrated using poly(N-isopropylacrylamide) as the stimuli-responsive polymer and avidinbiotin as a multibindable biomolecule-specific receptor. As a result of the selective response to the specific biomolecule, a clear permeability change of an order of magnitude was achieved. The principle is versatile and can be applied to many combinations of multibindable analyte-specific receptors, including antibodyantigen and lectinsugar analogues. The new gating system can find wide application in the bioanalytical field and aid the design of novel biodevices.
C
ross-linking based on biomolecular recognition plays various important roles in biosystems. In the coagulation cascade mediated by small-signal biomolecular recognition, fibrin is made from fibrinogen and is then polymerized and cross-linked to form a hemostatic plug over a wound site.1 Another example is cross-linking of actin filaments, which plays an important role in cell morphogenesis and migration.2 In this way, cross-linking can amplify the signal of biomolecular recognition to be stronger by converting into other kinds of signal. Inspired by such biological cross-linking functions, several researchers39 have developed synthetic hydrogels that are responsive to a biomolecular signal; these are functionalized by conjugated polymers and biomolecule-recognition receptors such as antibodies and lectin. On the other hand, when synthetic stimuli-responsive polymers are placed in the restricted space of nano- or microsized pores, the simple swellingshrinking behavior because of a volume phase transition is converted to a clear pore gating (opening and closing) function.1021 Some of these gating membranes can convert a specific molecular signal at a concentration ranging from tens to hundreds of millimolar into a pore gating using a polymer phase transition.14,15,17 However, a gating membrane responsive to trace amounts of biomolecules has not yet been achieved. By combining the useful function of biological cross-linking and polymer phase transition in the restricted space of the membrane pore, it should be possible to develop a novel gating system that enables the conversion of biomolecular signals into strong signals accompanied by the pore r 2011 American Chemical Society
gating. It is expected that such a gating membrane can aid in the design of novel biodevices for bioanalytical applications. In this letter, we present a biomolecule-recognition gating system that uses, for the first time, the cooperation of biorecognition cross-linking and a polymer phase transition in nanosized pores. The concept of this study is shown in Figure 1a. A biomolecule-recognition gating membrane immobilizes the stimuli-responsive polymer, including the biomolecule-recognition receptor, onto the pore surface of a porous membrane. In this gating membrane, the biorecognition cross-linking in nanosized pores can control the opening and closing of the pore via a two-step protocol. The first step is the biomolecule-recognition step, which involves the recognition of a biomolecule by biomoleculerecognition receptors immobilized in the membrane pore. In this step, the specific biomolecule is bound by immobilized receptors, and binding points work as biorecognition cross-linking when one specific biomolecule can bind several receptors. Conversely, a nonspecific biomolecule in a membrane pore does not produce biorecognition cross-linking. Many biomolecules have specific multibinding sites for one biomolecule-recognition receptor, such as lectinsugar binding and antibodyantigen binding; therefore, our proposed membrane can be used in the detection of various biomolecules, such as antigen, glycoprotein, allergen, and virus. Received: October 4, 2011 Accepted: November 18, 2011 Published: November 18, 2011 9226
dx.doi.org/10.1021/ac202629h | Anal. Chem. 2011, 83, 9226–9229
Analytical Chemistry
Figure 1. (a) Conceptual illustration describing this study: (i) specific biomolecule and (ii) nonspecific biomolecule. (b) Illustration of an avidin-recognition gating membrane.
In the second step, the stimulus is added to shrink the stimuliresponsive grafted polymer in the membrane pore. For a specific biomolecule, the closed pore can be preserved because shrinkage of the grafted polymer is prevented by the cross-linking. However, if the biomolecule is nonspecific, biomolecular cross-linking is not formed and thus the pore opens. In other words, in the second step, the pore state, which depends on the specificity of the biomolecule, can be definitely distinguished by a volume phase transition of the grafted polymer in the membrane pore. Therefore, a biomolecular signal via the control of a pore gating can be converted into a strong visual signal or a physical force such as the generation of osmotic pressure with the change of solute (e.g., colored nanoparticles and macromolecules) permeability. In addition, the gating membrane is thin; that is, the permeation length is short. Therefore, the injection of an analyte and a solute for a signal conversion using a syringe to the membrane would allow the rapid formation of biorecognition crosslinking and signal conversion, leading to quick and easy detection. In this study, avidinbiotin binding was utilized as a model to demonstrate the concept of a biomolecule-recognition gating membrane, because avidinbiotin binding has suitable properties for a novel gating system in which one avidin can specifically bind four biotins. Thus, we fabricated an avidin-recognition gating membrane, as shown in Figure 1b, in which the copolymer of N-isopropylacrylamide (NIPAM) and biotin-PEG2-acrylamide (NIPAM/biotin-PEG2-acrylamide = 95:5 mol %) was grafted onto the surface of the membrane pores using the plasma graft polymerization technique15,22 (see the Supporting Information, section S-2). A porous, high-density polyethylene (PE) membrane (maximum pore size, 0.15 μm; thickness, 27 μm; porosity, 50%) was used as a substrate membrane. NIPAM is a thermosensitive polymer and acts as an actuator via its volume phase transition at the lower critical solution temperature (LCST). The grafted amount of the fabricated membrane was 0.2 mg cm2. The fabricated membrane was analyzed using field emission-scanning electron
LETTER
Figure 2. Characterization of an avidin-recognition gating membrane. (a,b) SEM images of (a) PE substrate and (b) avidin-recognition gating membrane (scale bar = 1 μm). (c) FT-IR spectra of (black line) PE(NIPAM) membrane and (red line) avidin-recognition gating membrane. (d,e) The color test with p-DACA of (d) PE(NIPAM) membrane and (e) avidin-recognition gating membrane.
microscopy (FE-SEM; S-5200, Hitachi, Japan), FT-IR (FT/IR 6200; JASCO, Japan), and the color test with p-dimethylaminocinnamaldehyde (p-DACA), which reacts with biotin to give a color.23 The SEM images of the dry state in Figure 2a,b confirm that the pores of the fabricated membrane were still open after the grafting; that is, there would be enough space in the pores for swelling and shrinking of the grafted polymer. Figure 2c shows the FT-IR spectra of the PE(NIPAM) membrane and the avidin-recognition gating membrane. The FT-IR spectrum of the avidin-recognition gating membrane shows amide I (1650 cm1) and II (1550 cm1) derived from NIPAM and biotin-PEG2-acrylamide. In addition, the small peak of the CdO stretching vibration (1696 cm1) derived from biotin-PEG2-acrylamide were also observed. In the color test after immersion in p-DACA solution, while the color of the PE(NIPAM) membrane remains white (Figure 2d), the color of the avidinrecognition gating membrane changes from white to pale pink (Figure 2e), resulting from the existence of biotin in the grafted polymer. These results indicate that the avidin-recognition gating membrane with the copolymer of NIPAM and biotin-PEG2acrylamide, as shown in Figure 1b, was successfully fabricated. Subsequently, the permeation tests were conducted to evaluate the pore state of the avidin-recognition gating membrane before and after recognizing the biomolecule solutions whose concentration was constant at 1.5 μM. Here, the permeability coefficient Lp was calculated from the amount of permeated solution (phosphatebuffered saline) through the membrane in order to eliminate the effect of the applied pressure and media viscosity on the permeation flux J [m3/(s cm2)]. Lp [m3 m2 s1/(kgf cm2)] is defined in eq 1: LP ¼ 9227
J ΔPðμ20°C =μT Þ
ð1Þ
dx.doi.org/10.1021/ac202629h |Anal. Chem. 2011, 83, 9226–9229
Analytical Chemistry
Figure 3. (a) Permeation test using an avidin-recognition gating membrane: (i) avidin (specific) and (ii) albumin (nonspecific), (left) without biomolecules at 37 °C (above LCST), (middle) with biomolecules at 20 °C (below LCST), (right) with biomolecules at 37 °C (above LCST). (b) Validation of the protein selectivity by permeation tests using an avidin-recognition gating membrane.
where ΔP [kgf/m2] is the pressure applied across the membrane and μ20°C [Pa s] and μT [Pa s] represent the media viscosity at 20 °C and at the measurement temperature, respectively (see the Supporting Information, section S-3 for details). The results for the permeability coefficient Lp in various conditions are shown in Figure 3a. These results clearly indicate that the avidin-recognition gating membrane as newly proposed in Figure 1a functioned successfully in response to micromolar levels of biomolecules. This response sensitivity was 103105 times higher than that of molecule-recognition gating membranes in previous studies.14,15,17 The explanation for the results in Figure 3a is given below. At 37 °C (above the LCST) and without biomolecules (left column of Figure 3a), the grafted polymer in the membrane pore shrank, which left the pores open. After the biomolecules were recognized in the first step, Lp was measured at 20 °C, where the grafted polymer was swollen below the LCST (middle column of Figure 3a). In this condition, the values of Lp for specific avidin and nonspecific albumin were almost the same as that without biomolecules, and thus the permeation flux was little affected by the bound and/or adsorbed proteins and the amount of avidin or albumin in the pores was low. In other words, it was considered that the resistance of the swollen NIPAM polymers in the pores was dominant in the permeation flux, regardless of the binding and/or adsorption of proteins. Subsequently, in the second step, a thermal stimulus
LETTER
was added by increasing the temperature from 20 to 37 °C (right column of Figure 3a) to shrink the grafted polymer in the pores. After the thermal stimulus, the value of Lp for recognizing avidin was dramatically different from that for albumin. The value of Lp for recognizing avidin remained low by forming biorecognition crosslinking in the membrane pore. In contrast, the value of Lp with albumin became as high as that without biomolecules at 37 °C. In addition, Figure 3b shows the permeability coefficient in the temperature range 1545 °C, using an avidin-recognition gating membrane before and after immersion in various biomolecule solutions to investigate the specificity of the gating membrane. For the entire temperature range, the values of Lp with any nonspecific biomolecules were almost the same as that without biomolecules. In contrast, the value of Lp after binding avidin dramatically decreases above the LCST. This result is attributed to the same reason as shown in Figure 3a. Thus, we succeeded in selectively converting specific/nonspecific biomolecular signals into an order of magnitude change in permeability. In addition, Figure 3b also indicates that biomolecules can be detected under mild condition in our gating system because the permeability changes around physiological temperature were selective and clear. We also found that the permeability of the membrane after dissociating avidinbiotin binding by guanidinium chloride was not the same as that before avidin binding; that is, the response of the avidin-recognition gating membrane was irreversible. In addition, it was observed that the pores in the membrane with avidin binding were distorted in shape and almost closed, even in the dry state. (see the Supporting Information, section S-4 for details). Considering the observation, the irreversible response would be because avidin was bound by immobilized biotins with a high binding constant, and the pores of the polyethylene substrate could be deformed by the binding on closing. Here, it is considered that the irreversible and stable locking of pores by binding specific biomolecules is desirable for a system involving a disposable biosensor such as point-of-care testing because the stable pore state (opening and closing) results in robust detection for a biosensor. Finally, we show the potential performance in our gating system. Many biomolecules have specific multibinding sites for one biomolecule-recognition receptor, such as lectinsugar binding and antibodyantigen binding. Therefore, it would be possible to develop a gating membrane with the functional gate responding to a wide variety of biomolecules based on the concept proposed in this study. Here, the response sensitivity to a biomolecular signal was estimated by calculating the equilibrium state of antigenantibody bindings (typical dissociation constant, Kd = 107∼1010 [M]) in a nanosized pore (see the Supporting Information, section S-5 for calculation details). The calculation result indicates that an antigen-recognition gating membrane with ultrahigh sensitivity (up to 1015 M = 1 fM) can be achieved by controlling the solute permeability with thin layers cross-linked by a small number of recognized antigens. In addition, even with a much higher dissociation constant (Kd = 107), sufficiently high sensitivity (up to 1012 M = 1 pM) would be realizable. It is worth noting that the receptor (antibody) concentration (∼103 M) in the restricted space of nanosized pores can be very high compared with that in commonly used solution systems such as ELISA. Therefore, even if a ligand (antigen) concentration in a solution is ultralow, biomolecular bindings (ligandreceptor complexes) would be formed in pores. In summary, we have succeeded in developing a biomoleculerecognition gating membrane, as newly proposed in Figure 1a, that can control the opening and closing of the pore responding 9228
dx.doi.org/10.1021/ac202629h |Anal. Chem. 2011, 83, 9226–9229
Analytical Chemistry to a biomolecular signal with much higher sensitivity than previously reported gating membranes. In addition, the response of the fabricated gating membrane to a biomolecular signal was selective and clear. The cooperation of biorecognition crosslinking and polymer phase transition in nanosized pores enabled this advanced response. The control of a clear pore gating can be converted into a strong visual signal (e.g., colored nanoparticles) or a physical force such as the generation of osmotic pressure with the change of solute permeability. Moreover, the determination of the equilibrium state of biomolecule-recognition bindings in a nanosized pore indicates the possibility of developing a more advanced gating membrane with an ultrahigh sensitivity. The above-described features have never been achieved by the previously reported responsive hydrogels and gating membranes, and thus our proposed gating system should provide a new direction in designing novel devices in the bioanalytical field.
LETTER
(21) Friebe, A.; Ulbricht, M. Macromolecules 2009, 42, 1838. (22) Yamaguchi, T.; Nakao, S.; Kimura, S. Macromolecules 1991, 24, 5522. (23) Wilchek, M.; Bayer, E. A. Methods Enzymol. 1990, 184, 5.
’ ASSOCIATED CONTENT
bS
Supporting Information. Additional information as noted in text. This material is available free of charge via the Internet at http://pubs.acs.org.
’ AUTHOR INFORMATION Corresponding Author
*Phone: +81-45-924-5254. Fax: +81-45-924-5253. E-mail:
[email protected].
’ REFERENCES (1) Davie, E. W.; Fujikawa, K.; Kisiel, W. Biochemistry 1991, 30, 10363. (2) Alberts, B.; Johnson, A.; Lewis, J.; Raff, M.; Roberts, K.; Walter, P. Molecular Biology of the Cell, 5th ed.; Garland Science: New York, 2008. (3) Miyata, T.; Asami, N.; Uragami, T. Nature 1999, 399, 766. (4) Lutolf, M. P.; Raeber, G. P.; Zisch, A. H.; Tirelli, N.; Hubbell, J. A. Adv. Mater. 2003, 15, 888. (5) Ehrick, J. D.; Deo, S. K.; Browning, T. W.; Bachas, L. G.; Madou, M. J.; Daunert, S. Nat. Mater. 2005, 4, 298. (6) Murakami, Y.; Maeda, M. Biomacromolecules 2005, 6, 2927. (7) Thornton, P. D.; Mart, R. J.; Ulijn, R. V. Adv. Mater. 2007, 19, 1252. (8) Zhang, R.; Bowyer, A.; Eisenthal, R.; Hubble, J. Biotechnol. Bioeng. 2007, 97, 976. (9) Yuan, W. W.; Yang, J. Y.; Kopeckova, P.; Kopecek, J. J. Am. Chem. Soc. 2008, 130, 15760. (10) Ishihara, K.; Kobayashi, M.; Ishimaru, N.; Shinohara, I. Polym. J. 1984, 16, 625. (11) Mika, A. M.; Childs, R. F.; Dickson, J. M.; McCarry, B. E.; Gagnon, D. R. J. Membr. Sci. 1995, 108, 37. (12) Ito, Y.; Ochiai, Y.; Park, Y. S.; Imanishi, Y. J. Am. Chem. Soc. 1997, 119, 1619. (13) Park, Y. S.; Ito, Y.; Imanishi, Y. Macromolecules 1998, 31, 2606. (14) Yamaguchi, T.; Ito, T.; Sato, T.; Shinbo, T.; Nakao, S. J. Am. Chem. Soc. 1999, 121, 4078. (15) Ito, T.; Hioki, T.; Yamaguchi, T.; Shinbo, T.; Nakao, S.; Kimura, S. J. Am. Chem. Soc. 2002, 124, 7840. (16) Chu, L. Y.; Li, Y.; Zhu, J. H.; Chen, W. M. Angew. Chem., Int. Ed 2005, 44, 2124. (17) Ito, T.; Yamaguchi, T. Angew. Chem., Int. Ed. 2006, 45, 5630. (18) Lee, D.; Nolte, A. J.; Kunz, A. L.; Rubner, M. F.; Cohen, R. E. J. Am. Chem. Soc. 2006, 128, 8521. (19) Tokarev, I.; Orlov, M.; Minko, S. Adv. Mater. 2006, 18, 2458. (20) Ulbricht, M. Polymer 2006, 47, 2217. 9229
dx.doi.org/10.1021/ac202629h |Anal. Chem. 2011, 83, 9226–9229
