Electroaddressing Agarose Using Fmoc-Phenylalanine as a

LETTER pubs.acs.org/Langmuir

Electroaddressing Agarose Using Fmoc-Phenylalanine as a Temporary Scaffold Yi Liu,† Yi Cheng,‡ Hsuan-Chen Wu,§ Eunkyoung Kim,† Rein V. Ulijn,^ Gary W. Rubloff,‡,|| William E. Bentley,†,§ and Gregory F. Payne*,†,§ Center for Biosystems Research, ‡Institute for Systems Research, §Fischell Department of Bioengineering, and Department of Materials Science and Engineering, University of Maryland, College Park, Maryland 20742, United States ^ WestCHEM Department of Pure and Applied Chemistry, The University of Strathclyde, 295 Cathedral Street, Glasgow G1 1XL, Scotland, U.K.

)

†

ABSTRACT: Electroaddressing, the use of imposed electrical stimuli to guide assembly, is attractive because electrical stimuli can be conveniently applied with high spatial and temporal resolution. Several electroaddressing mechanisms have been reported in which electrode-induced pH gradients trigger stimuli-responsive materials to undergo localized sol
gel transitions to form hydrogel matrices. A common feature of existing hydrogel electrodeposition mechanisms is that the deposited matrix retains residual charged, acidic, or basic (macro)molecules. Here, we report that pH-responsive fluorenyl-9-methoxycarbonyl-phenylalanine (Fmoc-Phe) can be used to codeposit the neutral and thermally responsive polysaccharide agarose. Upon cooling, an agarose network is generated and Fmoc-Phe can be removed. The Fmoc-Phe-mediated codeposition of agarose is simple, rapid, spatially selective, and allows for the electroaddressing of a bioactive matrix.

’ INTRODUCTION Methods to direct the assembly of bioactive matrices at specific device locations are required for various applications in bioelectronics and nanobiotechnology. Electroaddressing is an attractive approach to functionalizing surfaces because electrical signals can be applied conveniently and with high fidelity. Several mechanisms have been identified for electroaddressing,1
7 and recent studies have shown that pH-responsive hydrogel-forming materials can be triggered to electrodeposit in response to electrochemically generated pH gradients.8
17 Electrodeposition with stimuli-responsive and hydrogel-forming materials is particularly attractive because it is rapid, reversible, and reagentless, and the resulting network is generally compatible with labile biological components. Agarose is a neutral, thermally responsive polysaccharide that is widely used in biological sciences (e.g., for microbial cultivation, tissue engineering, and gel electrophoresis).18,19 Although it would be desirable to electroaddress agarose-based matrices, we are unaware of mechanisms that allow agarose’s electrodeposition. Previous studies demonstrated that agarose can be codeposited with calcium-responsive polysaccharide alginate.14 However, the migration of large proteins through the codeposited alginate
agarose network was suppressed, presumably in part because of electrostatic repulsions between the protein and the alginate. Recently, it was discovered that Fmoc-conjugated peptides and amino acids are small-molecule gelators that can be triggered to self-assemble into fibrous hydrogels in response to reductions in pH (typically below 6).20
26 These hydrogels appear to be biocompatible and are being investigated as scaffolds for 3D cell r 2011 American Chemical Society

culturing.22,27 Recently, it was shown that Fmoc-dipeptides and Fmoc-amino acids can be electrodeposited in response to a low pH generated at a gold anode.28,29 Here, we extend these observations and show that Fmoc-phenylalanine (Fmoc-Phe) can codeposit agarose and that after the agarose network has set (by cooling) Fmoc-Phe can be “extracted” from the deposited hydrogel. To illustrate the versatility of Fmoc-Phe-mediated agarose deposition, we codeposited protein G-functionalized microparticles and demonstrated that these agarose-entrapped microparticles retained their bioactivity for antibody binding.

’ EXPERIMENTAL SECTION Materials. The following materials were purchased from SigmaAldrich: Fmoc-D-phenylalanine (Fmoc-Phe, g98%), DMSO (99.94þ%), hydroquinone (HQ, g99%), sodium chloride (99.5þ%), indium tin oxide (ITO)-coated glass slides (25  25  1 mm3, surface resistivity 8
12 Ω/square), IgG-FITC from human serum, and albumin from bovine serum (BSA). Low-melting-point agarose was purchased from Promega (Madison, WI). Labeled polystyrene microparticles (0.7
0.9 μm, 1% w/v) were purchased from Spherotech Inc. (Lake Forest, IL). FITC-streptavidin was obtained from BD Biosciences. Methods. An Fmoc-Phe solution was prepared by first dissolving Fmoc-Phe in DMSO (100 mg mL
1) and then diluting this concentrate in DI water (with or without HQ and NaCl) with 0.5 M NaOH. After vortex mixing, the solution was filtered using a syringe filter (0.45 μm). Received: April 26, 2011 Revised: May 16, 2011 Published: May 20, 2011 7380

dx.doi.org/10.1021/la201541c | Langmuir 2011, 27, 7380–7384

Langmuir

LETTER

Figure 1. Fmoc-Phe-mediated codeposition of agarose. (a) Schematic showing codeposition, agarose gel formation, and Fmoc-Phe “extraction”. (b) Chemical (Raman) evidence of agarose codeposition and Fmoc-Phe extraction. (c) Physical (ex situ QCM) evidence of codeposition and Fmoc-Phe extraction. Solution thus prepared has a pH of ∼7. An agarose solution was prepared by dissolving agarose in DI water (2% w/v) at 70 C and then cooling to 37 C before use. Deposition solutions were prepared prior to codeposition by mixing corresponding amounts of solutions of Fmoc-Phe, agarose, and/or labeled microparticles. Gold-coated glass slides and silicon wafer chips were prepared and patterned by standard photolithographic methods.17 Electrodeposition was performed using a dc power supply (2400 Sourcemeter, Keithley). Instrumentation. Raman spectra were obtained from a Jobin Yvon LabRamHR Raman microscope. Ex situ quartz crystal microbalance (QCM) measurements were made with a CHI420a electrochemical analyzer (CH Instruments, Inc., Austin, TX) as described elsewhere.30 Profilometric measurements were performed on a Veeco Dektak 6 M stylus profilometer. Fluorescence micrographs were obtained using an Olympus microscope, and the images were analyzed using ImageJ (http://rsb.info.nih.gov/ij/).

’ RESULTS AND DISCUSSION The codeposition of Fmoc-Phe and agarose was demonstrated using a gold-coated silicon wafer as illustrated in Figure 1a. First, the chip was partially immersed in a warm deposition solution (37 C) containing Fmoc-Phe (2.5 mM, 0.2%), agarose (1%), hydroquinone (50 mM), and NaCl (0.1 M), and then the gold was biased to serve as the anode (0.5 A m
2) and a platinum wire served as the cathode. Previous studies demonstrated that the addition of HQ facilitated gelation by allowing the pH gradient to be established at lower anodic potentials and thus limited damage to the gold electrode.15,28,29 After electrodeposition, the chip was removed from the deposition solution, rinsed and soaked in cool water for 30 min, and air dried at room temperature. Visually, a deposited film was observed on the gold electrode.

The mechanistic depictions in Figure 1a indicate that electrodeposition induces Fmoc-Phe to undergo its sol
gel transition to form a fibrous network. Upon cooling, Figure 1a suggests that the disordered agarose chains initially entrapped within the Fmoc-Phe network undergo a separate sol
gel transition to form polysaccharide network junctions. Chemical evidence of codeposition was provided by Raman spectroscopy. The bottom spectra in Figure 1b are controls that show characteristic peaks for Fmoc-Phe (1022, 1295, 1482, 1612 cm
1)31 and agarose (840, 893, 1082 cm
1).32 The spectrum for the codeposited hydrogel film shows peaks for both Fmoc-Phe and agarose. These Raman measurements provide direct spectroscopic evidence that Fmoc-Phe allows for the codeposition of agarose. Next, the codeposited film was treated with pH 8 buffer for 15 min to extract Fmoc-Phe as illustrated in Figure 1a. The Raman spectrum of the extracted film in Figure 1b indicates that the characteristic peaks of Fmoc-Phe have disappeared but the characteristic peaks of agarose are retained. These results provide direct chemical evidence that Fmoc-Phe can be removed from the codeposited agarose network. Physical evidence for agarose codeposition and subsequent Fmoc-Phe extraction is provided by ex situ QCM measurements. Similar to the experimental procedure depicted in Figure 1a, a gold-coated QCM crystal was immersed in the deposition solution (2.5 mM Fmoc-Phe, 1% agarose, 50 mM HQ, and 0.1 M NaCl); a constant anodic current (0.5 A m
2) was applied to the gold electrode for a specified time; the crystal was removed from the solution, washed with cool water, and dried; and then the resonance frequency was measured and compared to the initial frequency measured before deposition. Separate crystals 7381

dx.doi.org/10.1021/la201541c |Langmuir 2011, 27, 7380–7384

Langmuir

Figure 2. Spatial selectivity of the Fmoc-Phe-mediated codeposition of agarose. (a) Profilometric measurements of films codeposited onto patterned gold electrodes (1-mm-wide gold lines patterned on a glass slide). (b) Fluorescence photomicrographs of FITC-streptavidin binding onto the codeposited biotin-functionalized microparticles. See the text for experimental details.

were used for each deposition-time measurement. The observed frequency change was converted to the mass of the dried deposit using the Sauerbrey equation.15,33 The QCM results in Figure 1c show a linear increase in dried mass with deposition time for up to 8 min. Two controls are also shown at the lower right in Figure 1c. One control examined the requirement for Fmoc-Phe: deposition was performed for 8 min using the same conditions except that Fmoc-Phe was absent from the deposition solution. The second control examined the requirement for an applied voltage: the gold-coated crystal was incubated in the deposition solution (containing Fmoc-Phe) for 8 min without applying an anodic potential. No frequency (or mass) change was observed for either control. These QCM results indicate that Fmoc-Phe is required for electrodeposition and that deposition can be controlled by the deposition conditions (e.g., deposition time). To assess Fmoc-Phe extraction, the individual film-coated QCM crystals were immersed in the basic buffer (pH 8, 22 C) and incubated for 15 min, after which they were rinsed and dried and the resonance frequency measured. As illustrated by the arrows in Figure 1c, a loss of mass is observed after base treatment, consistent with the removal of Fmoc-Phe from the deposited film. These results indicate that Fmoc-Phe can serve as a temporary scaffold for the electrodeposition of thermally responsive neutral polysaccharide agarose. In previous studies, the electrodeposition of Fmoc-Phe was observed to be spatially selective in the lateral dimensions.29 Here, we extend this observation to the codeposition of agarose using a glass chip patterned with individually addressable gold electrodes (1-mm-wide lines separated by 3 mm spaces) as illustrated in Figure 2a. Codeposition was performed by partially immersing this patterned chip in the deposition solution (4 mM

LETTER

Fmoc-Phe, 1% agarose, 50 mM HQ, and 0.1 M NaCl) and biasing the individual electrodes (1 A m
2) for different times (2, 4, and 6 min). After deposition, the chip was rinsed with cool water, dried, and then measured using profilometry. The thickness measurements in Figure 2a indicate good spatial selectivity for the codeposition of Fmoc-Phe and agarose. Furthermore, the thickness measurements indicate a systematic growth of the film with deposition time, which is consistent with the QCM results of Figure 1c. The deposits on the patterned chip were next extracted by immersing the chip in the basic buffer (pH 8) and incubated for 15 min. After drying, the films were measured by profilometry, and Figure 2a shows a reduction in thickness. This observation is again consistent with the removal of Fmoc-Phe. To analyze the spatial selectivity of the agarose codeposition further and to explore its potential for electroaddressing biofunctional components, we incorporated biotin-labeled microparticles (0.1%) in the deposition solution. Similar to the experiments in Figure 2a, deposition was induced by anodically biasing (1 A m
2) the individual electrodes for different times (2, 4, and 6 min). After deposition, the chip was rinsed with cool water, air dried at room temperature for 20 min, and then incubated in an FITC-streptavidin solution (4 μg mL
1 in PBS buffer containing 1% BSA) for 3 h. Then, the chip was rinsed three times (10 min each) with PBS buffer and observed using fluorescence microscopy. Three observations are apparent from the images and image analysis in Figure 2b. First, microparticle electroaddressing is controllable; the fluorescent images indicate spatial selectivity in the lateral dimensions, and image analysis suggests quantitative control based on the deposition conditions (i.e., deposition time). Second, the granularity of the images suggests that the fluorescence of FITC-streptavidin is localized to the microparticles. Third, the control (far-right image) that lacked microparticles showed no fluorescence, which indicates that FITCstreptavidin can be readily rinsed out of the hydrogel network. (Note that although Fmoc-Phe was not purposely extracted from the deposited hydrogel, visual evidence indicates that it is partially removed during PBS incubation.) As a final proof of concept, we demonstrate that Fmocmediated agarose deposition allows the electroaddressing of a bioactive matrix. For this, we addressed protein G-functionalized microparticles and demonstrated that they retained the ability to bind IgG. As illustrated in Figure 3a, we sequentially deposited three hydrogels at individual electrode addresses of a goldpatterned silicon chip (1 mm gold lines spaced 1 mm apart). Initially, the particle-free control was codeposited (0.5 A m
2 for 2 min) on the left-most electrode, after which it was rinsed with cool water and dried. Next, protein G was addressed to the middle electrode by incorporating protein G microparticles (0.1%) in the deposition solution and biasing this electrode (0.5 A m
2 for 2 min). After being rinsed and dried, a BSAparticle control was assembled at the right electrode by incorporating the BSA-functionalized microparticles (0.1%) into the deposition solution and depositing under the same conditions. After the three films were assembled, the chip was incubated at room temperature with FITC-labeled human IgG (80 μg mL
1 in 10 mM PBS buffer containing 1% BSA) for 3 h and then rinsed with PBS buffer for 30 min. The bright-field images in Figure 3a show that microparticles are present in the middle (protein G) and right (BSA control) films but are absent in the left film (particle-free control). The 7382

dx.doi.org/10.1021/la201541c |Langmuir 2011, 27, 7380–7384

Langmuir

LETTER

Figure 3. Electroaddressing of protein G-functionalized microparticles to form a bioactive matrix. (a) Bright-field and fluorescence photomicrographs of a codeposited films with no particles (left electrode), protein G microparticles (middle electrode), and BSA microparticles (right electrode). (b) Highmagnification photomicrographs of deposited protein G microparticles on an ITO slide showing the colocalization of microparticles (bright field) and fluorescence associated with FITC-IgG (fluorescence).

fluorescence photomicrograph shows that only the middle electrode (with protein G microparticles) exhibits fluorescence, which is consistent with the binding of FITC-IgG. The absence of fluorescence in the two controls suggests that the nonspecific binding of IgG to the network or the microparticles is limited. To analyze IgG binding to the entrapped protein G microparticles further, we codeposited these microparticles with Fmoc-Phe and agarose onto a transparent indium tin oxide (ITO)-coated glass slide. Because ITO is more stable (vs gold) under the anodic potentials (∼2.5 V), HQ is not needed for deposition because a pH gradient can be generated at the anode by water electrolysis reactions. (Note that the elimination of HQ may be beneficial in some cases, although we did not observe a benefit for protein G.) For experiments with the ITO-coated slide, the deposition solution was modified to include protein G-functionalized microparticles (0.1%), Fmoc-Phe (4 mM), and agarose (1%), and deposition was performed at 1 A m
2 for 4 min. After deposition, the film was rinsed with cool water, air dried, and then incubated with FITC-IgG (80 μg mL
1 in 10 mM PBS buffer containing 1% BSA) for 3 h. The films were then rinsed and examined using high magnification. The bright-field images in Figure 3b show the location of the microparticles, and the fluorescent images show signals associated with IgG. The colocalization of these images indicates that IgG binds to the protein G microparticles. These results demonstrate that the Fmoc-Phe-mediated electrodeposition of agarose allows for the electroaddressing of functionally active biological components.

’ CONCLUSIONS We investigated Fmoc-Phe as a temporary scaffold to codeposit the thermally responsive polysaccharide agarose. Fmoc-Phe (and probably other Fmoc-peptides) has several attractive features: it is pH-responsive, it forms a fibrous hydrogel, its reversible sol
gel transition occurs in aqueous solution at nearneutral pH, and its low molecular weight allows it to be readily removed after the agarose network has formed. The Fmoc-Phemediated codeposition of agarose is simple, rapid, and spatially selective. A demonstration study with protein G-functionalized

microparticles indicates that Fmoc-Phe-mediated agarose electrodeposition allows for the electroaddressing of a bioactive matrix. We believe the ability to codeposit agarose should extend the capabilities for hydrogel electroaddressing because of agarose’s broad utility in the life sciences.

’ AUTHOR INFORMATION Corresponding Author

*E-mail: [email protected]. Phone: 301-405-8389. Fax: 301-3149075.

’ ACKNOWLEDGMENT We gratefully acknowledge financial support from the R. W. Deutsch Foundation, the National Science Foundation (NSF, EFRI-0735987), the Defense Threat Reduction Agency (W91B9480520121), and the Office of Naval Research (N000141010446). We acknowledge the Maryland NanoCenter’s FabLab for chip fabrication and thank Jordan Betz for assistance with Raman measurements and Peter Dykstra for assistance with profilometry. ’ REFERENCES (1) Corgier, B. P.; Marquette, C. A.; Blum, L. J. J. Am. Chem. Soc. 2005, 127, 18328. (2) Kaji, H.; Tsukidate, K.; Matsue, T.; Nishizawa, M. J. Am. Chem. Soc. 2004, 126, 15026. (3) Haddour, N.; Cosnier, S.; Gondran, C. J. Am. Chem. Soc. 2005, 127, 5752. (4) Mendes, P. M.; Christman, K. L.; Parthasarathy, P.; Schopf, E.; Ouyang, J.; Yang, Y.; Preece, J. A.; Maynard, H. D.; Chen, Y.; Stoddart, J. F. Bioconjugate Chem. 2007, 18, 1919. (5) Pavlovic, E.; Quist, A. P.; Gelius, U.; Nyholm, L.; Oscarsson, S. Langmuir 2003, 19, 4217. (6) Kim, K.; Yang, H.; Jon, S.; Kim, E.; Kwak, J. J. Am. Chem. Soc. 2004, 126, 15368. (7) Yousaf, M. N.; Mrksich, M. J. Am. Chem. Soc. 1999, 121, 4286. (8) Wu, L. Q.; Gadre, A. P.; Yi, H. M.; Kastantin, M. J.; Rubloff, G. W.; Bentley, W. E.; Payne, G. F.; Ghodssi, R. Langmuir 2002, 18, 8620. 7383

dx.doi.org/10.1021/la201541c |Langmuir 2011, 27, 7380–7384

Langmuir

LETTER

(9) Kurzawa, C.; Hengstenberg, A.; Schuhmann, W. Anal. Chem. 2002, 74, 355. (10) Cheong, M.; Zhitomirsky, I. Colloid Surf., A 2008, 328, 73. (11) Pang, X.; Zhitomirsky, I. Mater. Chem. Phys. 2005, 94, 245. (12) Redepenning, J.; Venkataraman, G.; Chen, J.; Stafford, N. J. Biomed. Mater. Res. A 2003, 66, 411. (13) Shi, X. W.; Tsao, C.-Y.; Yang, X.; Liu, Y.; Dykstra, P.; Rubloff, G. W.; Ghodssi, R.; Bentley, W. E.; Payne, G. F. Adv. Funct. Mater. 2009, 19, 2074. (14) Yang, X. H.; Kim, E.; Liu, Y.; Shi, X. W.; Rubloff, G. W.; Ghodssi, R.; Bentley, W. E.; Pancer, Z.; Payne, G. F. Adv. Funct. Mater. 2010, 20, 1645. (15) Zhou, Q. M.; Xie, Q. J.; Fu, Y. C.; Su, Z. H.; Jia, X.; Yao, S. Z. J. Phys. Chem. B 2007, 111, 11276. (16) Liu, Y.; Kim, E.; Ghodssi, R.; Rubloff, G. W.; Culver, J. N.; Bentley, W. E.; Payne, G. F. Biofabrication 2010, 2, 022002. (17) Cheng, Y.; Luo, X. L.; Betz, J.; Buckhout-White, S.; Bekdash, O.; Payne, G. F.; Bentley, W. E.; Rubloff, G. W. Soft Matter 2010, 6, 3177. (18) Stellwagen, N. C. Electrophoresis 2009, 30, S188. (19) de Lange, V.; Binkert, A.; Voros, J.; Bally, M. ACS Appl. Mater. Interfaces 2011, 3, 50. (20) Yang, Z.; Gu, H.; Fu, D.; Gao, P.; Lam, J. K.; Xu, B. Adv. Mater. 2004, 16, 1440. (21) Ulijn, R. V.; Smith, A. M. Chem. Soc. Rev. 2008, 37, 664. (22) Jayawarna, V.; Ali, M.; Jowitt, T. A.; Miller, A. F.; Saiani, A.; Gough, J. E.; Ulijn, R. V. Adv. Mater. 2006, 18, 611. (23) Mahler, A.; Reches, M.; Rechter, M.; Cohen, S.; Gazit, E. Adv. Mater. 2006, 18, 1365. (24) Orbach, R.; Adler-Abramovich, L.; Zigerson, S.; Mironi-Harpaz, I.; Seliktar, D.; Gazit, E. Biomacromolecules 2009, 10, 2646. (25) Tang, C.; Smith, A. M.; Collins, R. F.; Ulijn, R. V.; Saiani, A. Langmuir 2009, 25, 9447. (26) Sutton, S.; Campbell, N. L.; Cooper, A. I.; Kirkland, M.; Frith, W. J.; Adams, D. J. Langmuir 2009, 25, 10285. (27) Liebmann, T.; Rydholm, S.; Akpe, V.; Brismar, H. BMC Biotechnol. 2007, 7, 88. (28) Johnson, E. K.; Adams, D. J.; Cameron, P. J. J. Am. Chem. Soc. 2010, 132, 5130–5136. (29) Liu, Y.; Kim, E.; Ulijn, R. V.; Bentley, W. E.; Payne, G. F. Adv. Funct. Mater. 2011, 21, 1575. (30) Kim, E.; Liu, Y.; Shi, X. W.; Yang, X.; Bentley, W. E.; Payne, G. F. Adv. Funct. Mater. 2010, 20, 2683. (31) Larsen, B. D.; Christensen, D. H.; Holm, A.; Zillmer, R.; Nielsen, O. F. J. Am. Chem. Soc. 1993, 115, 6247. (32) Pereira, L.; Sousa, A.; Coelho, H.; Amado, A. M.; Ribeiro-Claro, P. J. A. Biomol. Eng. 2003, 20, 223. (33) Buttry, D. A.; Ward, M. D. Chem. Rev. 1992, 92, 1355.

7384

dx.doi.org/10.1021/la201541c |Langmuir 2011, 27, 7380–7384