Visualization of Flow-Aligned Type I Collagen Self-Assembly in

Dec 14, 2006 - Max Planck Institute for Dynamics and Self-Organization, Göttingen, Germany, ... method to investigate the dynamics of collagen self-a...
6 downloads 0 Views 96KB Size
Langmuir 2007, 23, 357-359

357

Visualization of Flow-Aligned Type I Collagen Self-Assembly in Tunable pH Gradients Sarah Ko¨ster,†,‡ Jennie B. Leach,‡,| Bernd Struth,§ Thomas Pfohl,*,† and Joyce Y. Wong*,‡ Max Planck Institute for Dynamics and Self-Organization, Go¨ttingen, Germany, Department of Biomedical Engineering, Boston UniVersity, Boston, Massachusetts 02215, and European Synchrotron Radiation Facility, Grenoble, France ReceiVed August 22, 2006. In Final Form: NoVember 13, 2006 Collagen is a major component of the extracellular matrix that exhibits unique hierarchical organization at multiple length scales ranging from nano to macroscale. Despite numerous methods to create collagen-based biomaterials, the self-assembly process of collagen ex vivo is poorly understood. Here, we describe a system that uses a microfluidic method to investigate the dynamics of collagen self-assembly. A main inlet stream of semidilute soluble collagen-I is hydrodynamically focused by two side inlet streams, which gradually increases the pH in the main stream. This enables dynamic nonequilibrium investigation of the self-assembly process simultaneously at different positions and therefore different stages in the assembly process within the same system. The device is designed for in situ monitoring and characterization of collagen assembly using polarization microscopy and X-ray diffraction: the continuous extensional flow provides highly ordered phases of the macromolecules over a large distance in the outlet microchannel and allows for data collection without material damage. We further demonstrate that finite element method simulations provide a good description of our experimental results regarding the diffusive phenomena, flow profile, and pH distribution. Our approach has broad impact, since it provides a powerful means of controlling and investigating the dynamic self-assembly process of biomacromolecules.

Introduction In mammals, collagen is the most abundant protein and consists of more than 20 different types, of which collagen-I is the largest fraction. It is a fibril-forming protein that self-assembles hierarchically at the nano, micro-, and macroscales. The macroscale structure and organization of collagen-I fibrils are key contributors to the mechanical properties of soft tissue, bone, tendon, and ligaments. Furthermore, the microscale organization of collagen profoundly influences cell morphology, migration, proliferation, and gene expression.1 Whereas collagen gels and substrates modified with collagen have been used extensively to investigate cell behavior,2-4 the ability to assemble collagen into its unique hierarchical organization ex vivo remains a significant challenge. The control of collagen assembly ex vivo has broad applications in engineering functional tissue replacements such as orthopedic soft tissues,5 heart valves,6 and blood vessels.7 * Corresponding authors. Joyce Y. Wong, Boston University, Department of Biomedical Engineering, 44 Cummington St., Boston, MA 02215, USA; [email protected]. Tel: (617) 353-2374. Fax: (617) 353-6766. Thomas Pfohl, MPI for Dynamics and Self-Organization, Bunsenstrasse 10, 37073 Go¨ttingen, Germany; [email protected]. Tel. +49 551 5176 240. Fax +49 551 5176 202. † Max Planck Institute for Dynamics and Self-Organization. ‡ Boston University. § European Synchrotron Radiation Facility. | Current address: Dept. of Chemical & Biochemical Engineering, Univ. of Maryland, Baltimore County. (1) Thakar, R. G.; Ho, F.; Huang, N. F.; Liepmann, D.; Li, S. Biochem. Biophys. Res. Commun. 2003, 307 (4), 883-890. (2) Gaudet, C.; Marganski, W. A.; Kim, S.; Brown, C. T.; Gunderia, V.; Dembo, M.; Wong, J. Y. Biophys. J. 2003, 85 (5), 3329-3335. (3) Haga, H.; Irahara, C.; Kobayashi, R.; Nakagaki, T.; Kawabata, K. Biophys. J. 2005, 88 (3), 2250-2256. (4) O’Connor, S. M.; Stenger, D. A.; Shaffer, K. M.; Ma, W. Neurosci. Lett. 2001, 304 (3), 189-193. (5) Freytes, D. O.; Badylak, S. F.; Webster, T. J.; Geddes, L. A.; Rundell, A. E. Biomaterials 2004, 25 (12), 2353-2361. (6) Stamm, C.; Khosravi, A.; Grabow, N.; Schmohl, K.; Treckmann, N.; Drechsel, A.; Nan, M.; Schmitz, K. P.; Haubold, A.; Steinhoff, G. Ann. Thorac. Surg. 2004, 78 (6), 2084-2092; discussion 2092-2093. (7) Boland, E. D.; Matthews, J. A.; Pawlowski, K. J.; Simpson, D. G.; Wnek, G. E.; Bowlin, G. L. Front. Biosci. 2004, 9, 1422-1432.

The alignment of collagen fibrils in the direction of highest strain reinforces the tissue structure.8 Aligned collagen fibers can also assist cell migration through contact guidance.3 Thus, there has been much effort devoted to preparing aligned collagen networks using diverse techniques such as intense magnetic9,10 or electric fields, cell-generated traction forces,11 hydrodynamic flow,12 dip-pen nanolithography,13 and interfacial orientation on the surface of other biopolymers.14 However, most of these methods lack microscale control of the self-assembly process, and none are amenable to dynamic in situ investigations. Here, we describe a system that uses a microfluidic method to investigate the dynamics of collagen self-assembly in a tunable pH gradient (Figure 1). This enables dynamic nonequilibrium investigation of the self-assembly process simultaneously at different positions and therefore different stages in the assembly process within the same system. The device is designed for in situ monitoring and characterization of collagen assembly using polarization microscopy and X-ray diffraction: the continuous extensional flow provides highly ordered phases of the macromolecules over a large distance in the outlet microchannel and allows for data collection without material damage. We further demonstrate that finite element method simulations provide a good description of our experimental results regarding the diffusive phenomena, flow profile, and pH distribution. Experimental Section A “cross” configuration of poly(dimethylsiloxane) (PDMS) microchannels (35 µm deep, 100 µm wide) is fabricated (Figure 1) (8) Fung, Y. C. Biomechanics: Mechanical Properties of LiVing Tissues; Springer: New York, 1993; Vol. p. (9) Murthy, N. S. Biopolymers 1984, 23 (7), 1261-1267. (10) Tranquillo, R. T.; Girton, T. S.; Bromberek, B. A.; Triebes, T. G.; Mooradian, D. L. Biomaterials 1996, 17 (3), 349-357. (11) Guido, S.; Tranquillo, R. T. J. Cell. Sci. 1993, 105 (Pt 2), 317-331. (12) Jiang, F.; Ho¨rber, H.; Howard, J.; Mu¨ller, D. J. J. Struct. Biol. 2004, 148 (3), 268-278. (13) Wilson, D. L.; Martin, R.; Hong, S.; Cronin-Golomb, M.; Mirkin, C. A.; Kaplan, D. L. Proc. Natl. Acad. Sci. U.S.A. 2001, 98 (24), 13660-13664. (14) Knight, J. B.; Vishwanath, A.; Brody, J. P.; Austin, R. H. Phys. ReV. Lett. 1998, 80, 3863-3866.

10.1021/la062473a CCC: $37.00 © 2007 American Chemical Society Published on Web 12/14/2006

358 Langmuir, Vol. 23, No. 2, 2007

Figure 1. Schematic representation of collagen-I hydrodynamically focused by a NaOH solution. The self-assembly of collagen-I is initiated in the pH gradient which is controlled by diffusive material transport from the side streams into the center stream. using standard soft lithography techniques .15,16 Briefly, silicon wafers are cleaned with isopropanol and spin-coated with a 35 µm layer of SU-8 50 photoresist (Micro Resist Technology GmbH, Berlin, Germany). The coated wafers are then selectively exposed to UV light through a high-resolution lithography transparency and developed. The three-dimensional structures (perpendicularly crossed channels) are cast in PDMS, and these replicas are used for the experiments. Holes are punched into the filling areas at the ends of the channels, and the PDMS microstructures are plasma-treated and irreversibly bound to glass coverslips to provide a tight seal. The microfluidic devices are connected via polyethylene and Teflon tubing to custom-made syringe pumps, which in turn are driven by programs written in LabView (National Instruments Corporation, Austin, TX). A 10 mg/mL solution of collagen-I (calf skin, USB Corporation, Cleveland, OH) in 0.075 M acetic acid (pH 3.7) is injected into the main channel. The collagen solution obtained from the manufacturer contains trimers of collagen peptide chains (Mw ≈ 285 kDa). These triple helical collagen molecules are stable at acidic pH, and the assembly process is initiated at neutral to basic pH. This process is often referred to as “fibrillogenesis”17 or “gelation”.18 All studies were conducted at room temperature. In order to stabilize the collagen flow before initiating the collagen self-assembly, we first inject ultrapure water (Millipore Milli-Q) into the side channels. Then, the flow is switched to a 0.075 M NaOH solution (pH 13) using a T-valve. The fluid velocity in the main channel (1.35-8.1 mm/s) is slower than that in the side channels (40.5 mm/s), leading to a hydrodynamically focused collagen stream.14 In this configuration, the pH of the collagen solution gradually increases along the length of the outlet channel resulting from diffusive mixing with the NaOH (Figure 1).19,20 An Olympus BX61 microscope (Hamburg, Germany) equipped with crossed polarizers, a 10× objective, and a halogen lamp is used to obtain polarization images during the collagen assembly process in the microfluidic device. Images of the flowing collagen stream are captured using a SensiCam CCD camera (PCO, Kelheim, Germany). To obtain predictions of the pH profile within the microchannels, FEM simulations are carried out with the commercial software Femlab (Comsol, Inc., Burlington, MA). Using about 20 000 elements, the incompressible Navier-Stokes equation is solved in two dimensions to obtain the stationary solution (low Reynolds number, i.e.,