Bioconjugate Chem. 2005, 16, 1088−1097
1088
Micropattern Printing of Adhesion, Spreading, and Migration Peptides on Poly(tetrafluoroethylene) Films To Promote Endothelialization Virginie Gauvreau† and Gae´tan Laroche*,†,‡ Unite´ de Biotechnologie et de Bioinge´nierie, Centre de Recherche de l'Hoˆpital Saint-Franc¸ ois d’Assise, C.H.U.Q., 10 rue de l’Espinay, Que´bec, Que´bec, Canada, G1L 3L5, and De´partement de Ge´nie des Mines, de la Me´tallurgie et des Mate´riaux, Centre de Recherche en Science et en Inge´nierie des Macromole´cules, Faculte´ des sciences et ge´nie, Universite´ Laval, Que´bec, Que´bec, Canada, G1K 7P4. Received November 26, 2004; Revised Manuscript Received May 6, 2005
We report here the development of an original multistep micropatterning technique for printing peptides on surfaces, based on the ink-jet printer technology. Contrary to most micropatterning methods used nowadays, this technique is advantageous because it allows displaying 2D-arrays of multiple biomolecules. Moreover, this low cost procedure allies the advantages of computer-aided design with high flexibility and reproducibility. A Hewlett-Packard printer was modified to print peptide solutions, and Adobe Illustrator was used as the graphic-editing software to design high-resolution checkerboardlike micropatterns. In a first step, PTFE films were treated with ammonia plasma to introduce amino groups on the surface. These chemical functionalities were reacted with heterobifunctional crosslinker sulfo-succinimidyl 4-(N-maleimidomethyl)cycloexane-1-carboxylate (S-SMCC) to allow the subsequent surface covalent conjugation of various cysteine-modified peptides to the polymer substrate. These peptidic molecules containing RGD and WQPPRARI sequences were selected for their adhesive, spreading, and migrational properties toward endothelial cells. On one hand, our data demonstrated that the initial cell adhesion does not depend on the chemical structure and combination of the peptides covalently bonded either through conventional conjugation or micropatterning. On the other hand, spreading and migration of endothelial cells is clearly enhanced while coconjugating the GRGDS peptide in conjunction with WQPPRARI. This behavior is further improved by micropatterning these peptides on specific areas of the polymer surface.
INTRODUCTION
The surface of cells displays a complex combination of receptors and ligands to mediate cell adhesion, spreading, migration, proliferation, and communication. These receptors and ligands are protein molecules that clasp one another like a lock and a key. Typically, receptors are proteins integrated in the cell membrane with domains extending beyond the membrane either or both inside and outside the cell. Ligands are proteins integrated in a neighboring cell membrane, part of the extracellular matrix or freely diffusing in the cell environment. Via receptor and ligand links, cells attach to each other and to the matrix to form tissue structure. Ligands from the milieu may code for a cell to adhere, to migrate, to grow and proliferate, or even to die. A well-known example of this is the R5β1 integrin receptor that attaches with fibronectin proteins of the extracellular matrix, more specifically at the RGD binding site (1). A multitude of different receptors are present on the surface of a single cell at any given moment. Likewise, an impressive diversity of ligands are present on the extracellular matrix. For over a decade, various proteins have been immobilized on polymer surfaces to enhance endothelial cells adhesion (2-5). Multiple strategies have been * Corresponding author. Phone: (418) 656-2131 ext. 7983, Fax: (418) 656-5343, E-mail:
[email protected]. † Centre de Recherche de l'Ho ˆ pital Saint-Franc¸ ois d’Assise. ‡ Universite ´ Laval.
employed to that end, the most common being adsorption or grafting of fibronectin (6-8) and RGD peptides (9, 10). The common point between these strategies is that they all aim to mimic the extracellular matrix normally supporting and interacting with cells, hence possibly leading to a better integration and a more natural host response to the biomaterial. It is known that vascular grafts that could support a layer of endothelial cells would have better resistance to thrombosis and graft stenosis in long-term applications (11, 12). In this study, RGD type peptides were grafted on the PTFE surface because this sequence is widely acknowledged as a cell-binding signal (13, 14), the goal being to promote in vivo endothelialization of ePTFE arterial prosthesis. RGD was originally identified as the fibronectin receptor sequence of the integrin R5β1, but nowadays it has been identified as a recognition motif for various ligands of several different integrins (15). It is found in numerous plasma and extracellular matrix proteins such as fibronectin, vitronectin, and type I collagen (1). Massia et al. in the 1990s have investigated human umbilical vein endothelial cells (HUVECs) attachment, spreading, and cytoskeletal organization on substrates covalently grafted with RGD peptides (4). They observed that this strategy enhanced cell adhesion on substrates such as glycophase glass, poly(ethylene terephthalate) (PET), and poly(tetrafluoroethylene) (PTFE) that were otherwise repellent to cells. Fundamental to the development of biologically integrated biomaterials is the ability to display a micropat-
10.1021/bc049717s CCC: $30.25 © 2005 American Chemical Society Published on Web 09/02/2005
Micropattern Printing of Peptides on PTFE Films
terned array of several specifically targeted ligands to mimic the extracellular matrix on the polymer surfaces. Since 1978, with MacAlear and Wehrung, who applied the photoresist technology from the semiconductor industry to create patterns on an underlying compressed proteinaceous layer (16, 17), several other techniques have been examined in the fields of micro- and nanopatterning to create two-dimensional arrays of proteins or other biologically active molecules on surfaces. These techniques include photoresist lithography, photochemistry, and self-assembled monolayers (18). While each of these micropatterning techniques may be useful for some applications, each also has inherent limitations, for the most part concerning multiple protein binding, nonspecific binding, and immobilization of proteins while retaining their maximum biological activity (18). In most cases, protein micropatterning is simply about immobilization through physical adsorption of proteins to a surface (19, 20). Entropy drives protein adsorption which always results from attractive forces such as ionic, hydrophobic, or van der Waals interactions (21). A more stable means to immobilize proteins to surfaces is to covalently bind the proteins to the surfaces via the use of heterobifunctional cross-linkers such as silanes (22), which bind at one end to a glass surface via a silanol group and to a protein at the other end via a variable pendant group. Finally, another way to provide a stable immobilization of proteins to surfaces is through the use of high-affinity ligand pairs such as avidin-biotin (23). The central hypothesis of this study was that adhesion, spreading, and migration signals should be simultaneously present on the surface of arterial grafts to promote complete integration and endothelialization. Thus, in addition to RGD adhesion peptides, spreading and migration peptides should be present on the surface. The amino acid sequence WQPPRARI originating from the 33kDa fragment of the A chain of fibronectin was recognized in the literature (24, 25) for enhancing adhesion, spreading and migration of endothelial cells. It was therefore selected as the second signal to be grafted on the PTFE surface for the purpose of this study. In this work, surface modifications were performed on Teflon (PTFE) films since small diameter arterial grafts are in general made of ePTFE, i.e., microporous or expanded PTFE. The first step of our approach consisted in an ammonia plasma treatment allowing the incorporation of amino groups on the PTFE surface. These highly reactive functionalities were conjugated, in a second step, with the heterobifunctional cross-linker Sulfo-SMCC, thus introducing maleimide functionalities on the PTFE surface. To exploit the very straightforward reaction between these maleimido groups and thiolic groups, the various peptides were synthesized with a cysteine at their terminal end. This study was thus performed with the peptide sequences CRGD, CGRGDS, and CWQPPRARI. Next, to graft this variety of specifically targeted peptides simultaneously on the polymer surface, we chose to develop a novel multistep micropatterning technique based on the ink-jet printing technology. This very advantageous ink-jet microprinting technique is an ultralow cost procedure that allows the flexibility of computer-aided design. The color cartridge of a commercial Hewlett-Packard printer (HP DeskJet 5500) was simply modified to print the biological peptide solutions according to a defined micropattern designed using Adobe Illustrator as the graphic-editing software. This micro-
Bioconjugate Chem., Vol. 16, No. 5, 2005 1089
patterning technique enabled rapid and precise deposition of both adhesion and migration peptides on PTFE films. EXPERIMENTAL SECTION
Materials. Poly(tetrafluoroethylene) (Teflon or PTFE) films 0.05 mm thick were purchased from Goodfellow (Huntingdon, England). Vapor-phase chemical derivatization was performed with 5-bromosalicylaldehyde purchased from Sigma, St Louis, MO. The cross-linker sulfosuccinimidyl-4-(N-maleidomethyl)cyclohexane-1-carboxylate (S-SMCC) to be grafted on the PTFE surface was also obtained from Sigma. The various peptide sequences (CRGD, CGRGDS, CRGE, CRGDY, and CWQPPRARI) grafted on the polymer surface were synthesized by Service de synthe`se de peptides de l′est du Que´bec, CHUQ, Que´bec, QC, Canada. Peptide solutions were prepared in phosphate-buffered saline (PBS) and glycerol that were purchased from VWR Canlab, Ville MontRoyal, QC, Canada and Laboratoire MAT, Montre´al, QC, Canada, respectively. The buffer solutions were prepared in filtered, deionized water (Nanopure system) while fine adjustments of the pH were made by adding either 0.1 M of HCl or NaOH to reach the desired value. CRGDY peptides were iodated using Iodo-Beads Iodination Reagent from Pierce, Brockville, ON, Canada, and sodium iodide from Sigma. All chemicals were of commercial grades of highest purity and were used without further purification. The cell culture experiments were performed with human umbilical vein endothelial cells (HUVECs) recolted and cultured at our lab. The culture medium M199 supplemented with 0.09 g/L of heparin sodium salt and the porcine gelatin were purchased from Sigma. This medium was also supplemented with fetal bovine serum (FBS) provided by Wiscent, St-Bruno, Qc, Canada, and endothelial cell growth supplement (ECGS) provided by VWR Canlab. Cells were cultured in 25 cm2 tissue culture-treated polystyrene flasks or plated in 48-well culture plates from BD Falcon. Trypsin-EDTA used to detach the cells from the culture surfaces was purchased at Invitrogen Canada inc., Burlington, ON, Canada. Rezasurine purchased from Sigma was used for fluorescence emission. Products required for HUVECs fixation and staining procedures including formaldehyde, bovine serum albumin (BSA), Saponine, 4′,6-diamidino-2-phenylindole (DAPI), Rhodamine-Phalloidine, and Tween 20 were also acquired from Sigma. Methods. Ammonia Plasma Treatment. The ammonia radiofrequency (rf) plasma treatment and system apparatus were previously described elsewhere (26, 27). Briefly, the treatments were performed at a rf power of 20 W and a pressure of 300 mTorr for 100 s using high purity ammonia gas. The 3 cm × 3 cm PTFE films were held by a Teflon-made circular sample holder and treated in a cylindrical plasma chamber. Using such a plasma configuration and geometry allows to modify only the internal side of the film. Surface Functionalization. The cross-linker S-SMCC was reacted with the ammonia plasma-treated film in 3 mg/mL of reagent in PBS 0.2 M at pH 7.4. The reaction was then allowed to proceed for 2 h in a nitrogen environment and under low light intensity to protect the reagents against UV-induced degradation. The characterization of this reaction has been presented elsewhere (28). Briefly, this reaction is complete and allows the introduction of maleimido groups on the PTFE surface. These functions were used to subsequently conjugate the
1090 Bioconjugate Chem., Vol. 16, No. 5, 2005
peptides through a sulfhydryl containing N-terminal Cystein residue. Peptide Solutions. The peptide sequences CRGD, CGRGDS, CRGE, CRGDY, and CWQPPRARI were synthesized and solutions of 10 µg peptide/µL of PBS 0.2M at pH 7.4 containing 7.5% glycerol were prepared. Glycerol was added to prevent rapid evaporation of the solution when the peptides were printed and to adjust the peptide solution viscosity to a value similar to that of the ink normally used by the printer. These solutions were used either to print the peptides or homogeneously conjugate them on the polymer surface. The occurrence of the peptide graftings on the SMCC-conjugated surfaces was ascertained by iodating a tyrosine-containing peptide (CRGDY) using Iodo-Beads and sodium iodide (29) using one bead to iodinate 0.2 mg of peptides. CRGDY peptides were first iodated and then grafted to the PTFE surface. The I3d XPS signal was used to confirm the peptide conjugation. Ink-Jet Printer. A commercial HP 5500 Deskjet printer was modified to print the peptide solutions. We carefully chose this printer because it offered an optimized resolution up to 4800 dots per inch (dpi) according to the manufacturer and most importantly because the cartridges were not monocoque, as are most of the new generation cartridges. Therefore, these cartridges could be refilled easily. First, the top of the tricolor cartridge HP#57 was cut so as to leave the contour in place, since it was essential to click the cartridge back in the cradle; the internal sponges made of polyurethane foam were then removed as well as the little grids and filter-papers covering the wells. The cartridges were then extensively washed with deionized water and ethanol. The peptide solutions were then deposited with micropipets directly into the wells situated very next to the printhead. Calibration of the Printer. The ink-jet printer drivers interpret the colors drawn on the computer screen that intrinsically work in RGB (red, green, blue) color-space, or additive light, to print them with a CMYK (cyan, magenta, yellow, black) color cartridge, or subtractive light. Therefore, it is essential to carefully calibrate the printer so that it prints pure color, or in this case, pure peptide and not a mixture of the peptide solutions. It is recommended to work in RGB color space if the artwork will only be displayed on-screen in the case of a web-site for example, and in CMYK if the work will be printed with a commercial printer (30), as the HP DeskJet 5500 used in this study. However, we tried both modes, looking at every printed surface with an optical microscope, and surprisingly it appeared that pure colors could be achieved more easily by working in RGB. Since the HP Deskjet printer used is targeted at the mass market, its drivers are probably not as trustwothy, in regard to color-space translation, as those of a real four-color-process commercial CMYK laser printer. Adobe Illustrator graphicediting software was used to design the desired patterns and define the pure colors. In the RGB mode, fine color adjustments are obtained by adjusting the intensity of each color, or wavelength, on a scale from 0 to 255 (8 bits). The full spectrum of colors is then achieved by combination of the three scales. Since we are in the additive color space, 0, 0, 0 of red, green, blue would give black, and 255, 255, 255 would give white. On the basis of this knowledge, and verifying the printed surfaces with an optical microscope, we obtained pure cyan with 0, 255, 255, pure magenta with 255, 0, 255, and pure yellow with 255, 255, 0. However, printing 255, 255, 0 for example, meaning 100% yellow, deposited too much ink on the surface, and independent droplets combined together. For
Gauvreau and Laroche
that reason, we reduced the quantity of ink for each droplet by incrementing the third factor to 100 (ex.: 255, 255, 100), therefore obtaining a lighter color. Grafting Peptides in Solution and Printing Peptides. The same peptide solutions were use for control, i.e., peptides homogeneously conjugated on the polymer surface, and for printing. Conventional in-solution grafting was performed using 1 mL of the peptide solutions for each 1 cm2 of plasma-treated SMCC-grafted PTFE films. The reaction was performed for 120 min at room temperature under mechanical agitation. To graft the peptides through the micropatterning technique, 30 µL of the solution of the individual peptide were micropipetted directly into the adjacent wells of the printer cartridge. The peptides were then printed on the ammonia plasma-treated SMCC-grafted PTFE films according to the micropattern designed using Adobe Illustrator software. Squares of 300 µm width of both adhesion peptides and spreading and migration peptides were printed next to the other just like a checkerboard. Concerns about the viscosity and the wetting properties of the peptide solutions affecting the quality of the arrays were considered. Moreover, to ensure the covalent grafting of the peptides to SMCC-grafted PTFE, the individual peptide spots had to remain in the aqueous form and not evaporate too rapidly. The viscosity of the solution was therefore adjusted with 7.5% glycerol (31). Printed surfaces were maintained in a humidification chamber for 120 min at room temperature. X-ray Photoelectron Spectroscopy (XPS) Characterization. Immediately following the ammonia plasma treatment, samples of the aminated PTFE film were cut in a glovebox purged with dry nitrogen. These samples were analyzed by XPS without delay to assess the exact atomic composition of the surface following the plasma treatment. SMCC grafted surfaces, as well as all the peptidegrafted surfaces, were also analyzed by XPS after having been vacuum-dried overnight at ambient temperature. The XPS spectra were recorded using a PHI 5600-ci spectrometer (Physical Electronics, Eden Prairie, MN). A monochromatic aluminum X-ray source (1486.6 eV) at 400 W with neutralizer was used to record the survey spectra while high-resolution spectra were obtained using the monochromatic magnesium X-ray source (1253.6 eV) at 400 W without charge neutralization. The detection was performed at 45° with respect to the surface normal. Amine Quantification. Samples of ammonia plasmatreated films were used to quantify the amine surface concentration through vapor-phase chemical derivatization using 5-bromosalicylaldehyde as described elsewhere (26). Briefly, the reaction was performed at 85 °C for 2 h in a sealed glass tube in which a 1 cm-thick bed of sodalime glass beads was used to separate the reagent from the reactive surface. The surfaces were then vacuumdried overnight at 40 °C and analyzed by XPS. Endothelial Cell Culture. HUVECs were cultured in a 25 cm2 tissue culture-treated polystyrene flasks covered with porcine gelatin. The culture medium M199 with heparin was supplemented with FBS and ECGS, and the cells were grown in a 37 °C incubator with a humidified atmosphere of 5% CO2 and 95% air. For cell culture experiments, modified PTFE films of 27 mm2 (5.86 mm of diameter) were deposited in a 48-well microplate and HUVECs were seeded at 50 000 cells/well for adhesion tests and 25 000 cells/well for proliferation tests. Adhesion measurements were performed for 2 h in a serumfree M199 medium to ensure that RGD peptides grafted to the surface, and not the serum protein contained in the medium, favored HUVEC adhesion to PTFE. For
Micropattern Printing of Peptides on PTFE Films
spreading and migration measurements, the serum-free medium was removed from the wells after 2 h of adhesion and replaced by a fresh culture medium supplemented with serum and growth factor. Spreading and migration tests were performed for 72 h. Counting the Number of HUVECs Adhered to Surfaces. After 2 h of adhesion, HUVECs seeded on different surfaces were detached in order to evaluate the number of cells capable of adhesion on the various peptide-grafted PTFE surfaces. The culture medium and nonadhered cells were first aspirated from the wells. The surfaces were then rinsed twice with PBS and the rinse transferred to new wells to prevent detachment and counting of cells adhered to the well walls. Cells were then detached from the surface using 90 µL of trypsin-EDTA per well and incubated at 37 °C for 3 min. Trypsin activity was then inhibited with 10 µL of FBS per well. Aliquots were withdrawn from each well, and cells were counted the usual way using a hemacytometer. In a different way, the number of HUVECs adhered to the various surfaces were also estimated by monitoring fluorescence emission of cells incubated with resazurine. The cell dehydrogenase enzyme converts resazurine into resorufine, the concentration of which was monitored by measuring the intensity of fluorescence for each surface and comparing with a blank. First, the culture medium and the cells in suspension were aspirateded from the wells, and the surfaces were rinsed with PBS. A solution of 1/10 of resazurine 250µg/mL and 9/10 of culture medium M199 supplemented with heparin and FBS was prepared and maintained at 37 °C. Aliquots of 100 µL were added to each wells, and the surfaces were stored in the incubator. Fluorescence measurements were made on a Bio-Tek FL600 luminescence spectrometer with settings of excitation and emission wavelengths at 485 nm and 590 nm, respectively. All experiments were performed in triplicate. HUVEC Fixation and Staining. After the 2 h of adhesion or the 72 h of spreading and migration, the culture medium was aspired from the wells and HUVECs were fixed for 30 min at room temperature with 3.7% formaldehyde in PBS. The formaldehyde solution was then removed from the wells, and cells were washed three times with PBS. Cell permeability was increased with a solution of PBS with 3% BSA and 0.1% saponine for 1 h 30 min at room temperature. This solution was removed, and cells were stained using 1:3000 DAPI and 1:400 Rhodamine-Phalloidin in PBS and maintained in the 37 °C incubator for 1 h. Cells were then washed six times with PBS and 0.05% Tween 20 before the surfaces were fixed between blade and lamella to be observed under the microscope. Microscopy and Image Analysis. Microscopic pictures of every HUVECs seeded surfaces were taken using a Nikon E800 confocal microscope equipped with a Hamamatsu Orca ER digital camera (Nikon Canada, Mississauga ON). Pictures taken in microscopy were analyzed using Adobe Photoshop histogram and color selection pixels count functions. Image color selection allowed to distinguish between background and cellular bodies. A ratio between the number of pixels corresponding to cellular bodies and the total number of pixels provided a percentage of cellular coverage of the surface. The total number of pixels was of course similar for all images (all microscope images compared in this study being the same size).
Bioconjugate Chem., Vol. 16, No. 5, 2005 1091 RESULTS AND DISCUSSION
The various peptides were grafted on the surface through simple chemical reactions. First, an ammonia plasma treatment allowed the introduction of about 13% of nitrogen-containing species, one-third of which were amino groups according to vapor-phase chemical derivatization using 5-bromosalicylaldehyde (26, 28). S-SMCC was then allowed to react with the highly reactive amino groups to provide maleimido groups on the polymer surface. According to previous results (28), this reaction is complete. The maleimide functionalities were useful to covalently conjugate peptides through terminal end thio-containing cysteine residues. A schematic example of these chemical reactions and the XPS spectra monitored after each step of the conjugation procedure are shown in Figure 1. As can be seen in Figure 1, the peptide graftings on the SMCC-conjugated surfaces did not bring important changes from a XPS point of view. Therefore, to ascertain the covalent conjugation of these peptides on the PTFE surface, a CRGDY peptide was iodinated using the simple chemical reaction between the tyrosine residue and Iodobeads (Figure 2). Iodine has a high XPS sensitivity index, and therefore its detection on the PTFE surface confirmed the covalent binding of the peptides on the polymer surface (Figure 3) through the reaction between the thiolic group of the terminal cysteine and the maleimido group previously introduced on the polymer surface. All peptide-grafted surfaces, both in solution and homogeneously printed (meaning that the peptide solution was printed everywhere on the PTFE surface), were analyzed by XPS. Similar results were obtained for each grafting procedure (Table 1), indicating that the printing technique did not alter the capacity of the peptides to react with the maleimide functionalities previously grafted on the PTFE surface. The two most common ink jet dispensing technologies are thermal and piezoelectric. Both modes have been tested in the literature, and the thermal dispensing technology appears to be more reliable for printing biological solutions (32). This technology is based on the expulsion of picoliter-sized droplets from the capillary nozzles of the printhead which is assembled directly on the cartridge and consists of a nozzle plate with multiple capillaries or ink-ejection orifices surrounded by resistive heater elements. When a current pulse is received by an individual heater element, the rapid transfer of heat to a small portion of the sample solution located in a capillary causes its vaporization, creating a vapor bubble. The vapor bubble creates a volume displacement that leads to the expulsion of a droplet. Therefore, it was important that the conductivity of the original ink and our peptide solutions be identical (0.08 µS/cm or a resistance of 12 Mohm-cm). The local temperature applied to a small sample volume by the heater elements can reach 200 °C. Concerns about the integrity of the peptides have been addressed, and no degradation has been observed, as relative surface concentrations obtained through XPS analyses were similar for printed peptides as compared to conventional in-solution grafted peptides. This result is in agreement with various studies who have addressed this question concerning other biomolecules such as DNA, enzymes, and proteins (31, 33). Furthermore, we did not observed clogging problems caused by heat-aggregation of peptides in the printhead. Therefore, the commercial tricolor cartridge can be successfully washed and refilled with peptide solutions for
1092 Bioconjugate Chem., Vol. 16, No. 5, 2005
Gauvreau and Laroche
Figure 1. XPS spectra monitoring each reaction step of the CRGD grafting procedure: a. Virgin PTFE. b. Ammonia plasmatreated PTFE. c. SMCC-grafted PTFE. d. CRGD conjugated to plasma-treated SMCC-grafted PTFE. e. Synthetic scheme of CRGD peptide grafting on a PTFE film.
precise microdeposition of the surfaces. This procedure did not compromise the performance of the printer nor did it damage the printhead or alter the integrity of the various peptides. The use of the tricolor cartridge instead of the black one allowed the deposition of arrays containing up to three different peptides at the time. The quantity of peptides deposited in each printed spot was calculated by weighing the cartridge before and after printing a large known number of spots (33). In the present study, it was found that each 300 µm wide spot printed on the polymer contained 0.18 nL of the peptide solutions (0.18 nL/spots of 90 000 µm2 or 2.0 fL/µm2). Roda et al. (33), using a HP DeskJet 600 and printing horseradish peroxidase (HRP) on cellulose paper, found that 17 nL/spots of 1 mm in diameter (21.6 fL/µm2) were deposited on the surface of their substrate. This large difference is mainly attributed to the properties of the surface to be printed on. Cellulose paper is absorbent, and the authors used the highest quality printer settings to deposit the HRP, whereas PTFE is impermeable and printing with the highest quality printer mode deposited a volume of solution too large to maintain independent spots and retain the micropatterned design. Okamoto and al. (31) have printed DNA on glass surfaces using Bubble Jet printers from Canon (BJC-600 and BJC-700J). The resolution of their printer was 360 × 720 dpi, and they
printed 24 pL of their solution per droplet (9.6 fL/µm2) onto a glass surface. Again, the quantity of solution deposited is higher than the quantity we deposited on PTFE, and this can probably be explained by the more hydrophilic property of the glass substrate compared to ammonia plasma-treated PTFE. To determine whether it was easier to print pure color in RGB color-space mode or in CMYK mode, we printed individual colors on premium picture paper and observed the prints with an optical microscope. The pure-color prints were obtained in the RGB color-scale mode with codes 0, 255, 255 for pure cyan, 255, 0, 255 for pure magenta, and 255, 255, 0 for pure yellow. Droplets on clean PTFE film presented a diameter of 10 µm or a surface area of 80 µm2 (Figure 4a). Unsurprisingly, because plasma-treated PTFE is hydrophilic and still maintains its nonabsorbance property, droplets of this size could not be obtained on this surface. With this printer, the smallest micropatterned spots obtained were squares of 300 µm of width since smaller independent droplets combined together (Figure 4b). To evaluate the peptides’ capacity to promote adhesion, PTFE surfaces conjugated in solution with each individual peptide and with combination of CRGD/CWQPPRARI and CGRGDS/CWQPPRARI were seeded with HUVECs. After 2 h of HUVECs adhesion, the surfaces
Micropattern Printing of Peptides on PTFE Films
Bioconjugate Chem., Vol. 16, No. 5, 2005 1093
Figure 2. Synthetic scheme of iodination of CRGDY peptide and subsequent conjugation to plasma-treated SMCC-grafted PTFE film.
Figure 3. XPS spectrum of PTFE film conjugated with CRGDY-I.
Figure 4. a. Ink spots of 10 µm on virgin PTFE. b. Spots of 300 µm on plasma-treated PTFE.
were rinsed with sterile PBS, and cells were harvested using trypsin-EDTA and counted (Figure 5). From these results, some important trends need to be pointed out. First, there is a significant difference in the capacity of the peptide-grafted surfaces to favor adhesion compared to the control surfaces. Surfaces grafted with peptides were more attractive to cells than control surfaces. Another important point to notice is that, although CRGE peptide is stated in the literature (34) as a negative control of CRGD, in the sense that it does not support adhesion of cells, in this part of the study the short adhesion period of 2 h was insufficient at demonstrating
Figure 5. The number of HUVECs harvested with trypsinEDTA on PTFE surfaces after 2 h of adhesion; aliquots were withdrawn from each well for the cells to be counted with a hemacytometer.
a difference between these two peptides. Probably for the same reason, CGRGDS, which is known to better enhance adhesion of cells on surfaces (35), did not stand out from CRGD. Finally, the peptide CWQPPRARI, which was shown (24, 25) to favor adhesion, spreading and migration of endothelial cells, did favor adhesion of HUVECs to PTFE in a comparable way to CRGD. Similar trends were noticed while the cells attached to the surfaces were counted after incubating the cells with resazurine (Figure 6); that is to say that no individual peptide appeared more adhesive than one other, but peptide-grafted surface were more favorable to adhesion of HUVECs than control surfaces. Therefore, the initial HUVEC adhesion does not depend on the chemical structure of the peptides nor does it depend on the combination of peptides covalently bonded to the PTFE surface. Adhesion, spreading, and migration of HUVECs on surfaces either grafted with individual peptides in solution, peptide mixtures in solution, or combination of peptides micropatterned was also examined in microscopy (Table 2). These images confirmed the results presented in Figures 5 and 6, as they also demonstrated that the initial cellular adhesion did not depend on the
1094 Bioconjugate Chem., Vol. 16, No. 5, 2005
Gauvreau and Laroche
Figure 6. The number of HUVECs adhered to the different surfaces was estimated by monitoring fluorescence emission of cells incubated with resazurine after 2 h of adhesion on various PTFE surfaces.
chemical structure of the peptides conjugated to the surface. In addition, they also revealed that similar cellular adhesion was observed for combination of peptides either coconjugated in solution or micropatterned on the PTFE surface, in accordance with our previous statement that the printing technique did not alter neither the peptides integrity nor their capacity to promote adhesion of cells. Typical examples of these images are presented in Table 2. From these images, qualitative conclusions can be drawn. First, clean PTFE film does not support adhesion of endothelial cells, although some cells were observed on the surface. These cells were very small in size, and the actin cytoskeleton was disrupting. This cytoskeletal disruption alone presets
apoptosis (36), and therefore no cells were observed on the virgin PTFE surface after 72 h. Microscopic images of CRGE- and CRGD-grafted surfaces in adhesion tests were very similar with small round cells, indicating poor adhesion and spreading, and other larger spreading cells. Besides these early adhesion results, there was a major difference between these two types of modified surfaces after 72 h. Cells on the CRGE-conjugated surfaces were deprived of any attachment signal, and most of them did not survive over a three-day period of time. Hence, almost no cells were observed on the CRGE surfaces in the spreading and migration tests, and the ones that were present looked abnormal with fragmented nuclei and distorted cytoskeletons. On the contrary, cells on the CRGD surfaces were larger, some were still roundshaped, but many were spreading and their nuclei were intact. These observations are in accordance with what we know from the literature; cellular functions are regulated both by the adhesion of cells to the surface via the integrin receptors (1) and by their subsequent spreading (37, 38). Whitesides and colleagues have also demonstrated that reduction of cell spreading is associated with a decrease in cell survival, as the more retracted and round the cells, the greater the apoptotic rate (39). PTFE surfaces conjugated with CGRGDS did not exhibit many more adhered cells than CRGD surfaces after 2 h of adhesion, but the cells on the CGRGDS surfaces were wider and their cytoskeleton was more developed. Again in the 72-h spreading and migration tests, there were not more cells on the CGRGDS surfaces compared to the CRGD surface, but cell spreading was obvious. Some cells on the CWQPPRARI surface, after 2
Table 1. XPS Analysis of Plasma-Treated SMCC-Grafted PTFE Surfaces Conjugated with Various Peptides Either in Solution or Printed CRGE %C %F %N %O
CRGD
CGRGDS
CWQPPRARI
solution
printed
solution
printed
solution
printed
solution
printed
56.8 25.6 7.5 10.1
56.5 24.6 8.5 10.4
56.8 25.9 7.7 9.6
56.2 26.5 7.6 9.7
56.5 27.6 8.2 7.7
55.2 28.7 8.5 7.6
61.2 18.4 10.3 10.1
59.6 19.8 10.2 10.4
Table 2: Typical Microscopic Images (100×) of the Various Surfaces without Distinction between the Grafting Technique for Individual Peptides and with Distinction between the Grafting Procedure Performed Either in Solution or via the Microprinting Technique for the Surfaces Conjugated with a Combination of Peptidesa
a Cells were fixed and colored with DAPI and Rhodamine-Phalloidin after 2 h in the adhesion experiments and after 72 h in the proliferation experiments.
Micropattern Printing of Peptides on PTFE Films
h of adhesion, were round-shaped and did not present much of a cytoskeleton, but in general they seemed welladhered to the surface and undergoing spreading. After 72 h of spreading and migration on these surfaces, the cells were fully spread and some appeared to be in motion with projections in similar directions. This is in agreement with results in the literature presenting this amino acid sequence as an adhesion signal (40) and spreading and migration signal (24, 25). Microscopic images captured for the surfaces conjugated with a combination of peptides, either CRGD or CGRGDS and CWQPPRARI, did not present noteworthy differences between the substrates conjugated with the peptides in solution or via the micropatterning technique after 2 h of adhesion. On the other hand, there was a significant difference between the two combinations of peptides. The simultaneous conjugation of adhesion signal CRGD and spreading and migration signal CWQPPRARI did seem to define a more adhesive surface than CWQPPRARI peptide alone but not to a great extent. Concurrently, the conjugation of CGRGDS and CWQPPRARI did definitely define a more adhesive surface than individual peptides alone. Quantitatively, there were not a greater number of cells adhered to the surface, but from the glimpse of an eye, it was obvious that the cells were larger and spreading out. After 72 h of spreading and migration, there was a major difference between the peptide-combined surfaces grafted in solution and via the printing technique. On the CRGD/CWQPPRARI surfaces, cells were spreading and the cytoskeleton was effusive. There were more cells on the micropatterned surface than on the surface with the combination of peptides conjugated in solution. Moreover, there seems to be a certain directional organization in the spreading of these cells, indicating that cells might be communicating with one another. Surfaces conjugated with both CGRGDS and CWQPPRARI fully supported endothelial cell attachment and spreading. The microscopic images captured for these surfaces indicated that this combination of adhesion and proliferation signal was the key combination in this study to promote endothelialization of PTFE surfaces, the ultimate goal being to develop ePTFE vascular grafts able to promote and support an endothelial cell layer at the luminal face. CGRGDS and CWQPPRARI peptides grafted in solution did enhance attachment and spreading of HUVECs to a great extent. Nevertheless, the micropatterning printing technique seems to be advantageous in promoting endothelialization since the surfaces were nearly fully covered with cells. These results are in agreement with Whitesides and colleagues who have demonstrated the influence of micropatterned surfaces on the control of cell shape, position, and function (41). To draw some quantitative conclusions from these microscopic images, they were processed through Adobe Photoshop histogram function. Values corresponding to the percentage of cell coverage of the surfaces were calculated from ratios between the number of pixels corresponding to cell bodies and the total number of pixels. Results for the adhesion and the spreading and migration experiments are presented in Figures 7 and 8. Unsurprisingly, from these results it can be concluded that virgin PTFE is not suitable at promoting and supporting adhesion of endothelial cells over a period of 72 h. Thus, even though some cells were present on this surface after 2 h of adhesion, none were capable of surviving over a 3-day period of time. In the same way, CRGE did not support adhesion, spreading, and migra-
Bioconjugate Chem., Vol. 16, No. 5, 2005 1095
Figure 7. Percentage of cellular coverage of PTFE surfaces after 2 h of adhesion calculated from a ratio between the number of pixels corresponding to cellular bodies/total number of pixels for microscopic pictures. PTFE surfaces were grafted with various peptides and combinations of peptides, either in solution or printed in micropatterns of 300 µm.
Figure 8. Percentage of cellular coverage of PTFE surfaces after 72 h of spreading and migration calculated from a ratio between the number of pixels corresponding to cellular bodies/ total number of pixels for microscopic pictures. PTFE surfaces were grafted with various peptides and combinations of peptides, either in solution or printed in micropatterns of 300 µm.
tion of HUVECs over the same period of time. The ability of CRGD surfaces to promote adhesion of endothelial cells and the inability of CRGE to do so is in agreement with previous studies in the literature (42-44). Again, there was only a slight difference between CRGD and CGRGDS in their capacity of promoting and supporting adhesion of HUVECs. Although CWQPPRARI alone did no better than RGD-type peptides at promoting adhesion, spreading, and migration of HUVECs, the combination of CRGD/CWQPPRARI and most appreciably the combination of CGRGDS/CWQPPRARI did enhance adhesion, spreading, and migration of endothelial cells significantly. Therefore, it appeared that a combination of both adhesion and, spreading, and migration peptides is better than individual peptide at promoting and supporting endothelialization of PTFE films. Although CGRGDS was not markedly better than CRGD at enhancing HUVECs adhesion, in the presence of CWQPPRARI, CGRGDS was more effective than CRGD for endothelialization of PTFE. In addition, our data clearly demonstrated that micropatterning CWQPPRARI and CGRGDS side by side on the polymer surface rather than coconjugating them led to an improvement of the HUVEC migration and spreading after 72 h.
1096 Bioconjugate Chem., Vol. 16, No. 5, 2005 CONCLUSIONS
In this study, PTFE surfaces were covalently conjugated with various peptides containing RGD and WQPPRARI sequences which were selected for their adhesive, spreading, and migration properties toward endothelial cells. Our data indicate that the initial cell adhesion does not depend on the chemical structure and combination of the peptides covalently bonded to the PTFE surface either through conventional in-solution conjugation or through the micropatterning printing technique. Despite these initial adhesion results, our data clearly indicate that spreading and migration of endothelial cells are enhanced while coconjugating RGD-type peptide and WQPPRARI peptides. Outstanding results for adhesion, spreading, and migration of endothelial cells on the modified PTFE surface were obtained with the combination of GRGDS and WQPPRARI. This behavior was further improved by micropatterning these peptides on specific areas of the polymer surface. ACKNOWLEDGMENT
Special thanks to Dr. Pascale Chevallier for XPS analyses, Richard Janvier for technical assistance with HUVEC culture, Dr. Stephane Turgeon for multidisciplinary technical assistance, and Louis Gagne´ for undergraduate work. We would also like to extend our gratitude to Dr. Madeleine Carreau for providing the microscope facilities and Dr. Rene´ C.-Gaudreault for advice and providing the HUVECs. We kindly acknowledge Gilles Brassard from ARC/PreciJet for excellent advice concerning printers and cartridges, as well as providing some cartridges. Finally, special thanks to Nicolas Dube´ for computer engineering support and numerous advice. We acknowledge grants from the National Science and Engineering Research Council (NSERC) of Canada and the Fonds pour la Recherche en Sante´ du Que´bec (FRSQ) (G.L). V. Gauvreau acknowledges a graduate scholarship from NSERC. LITERATURE CITED (1) Hynes, R. O. (1992). Integrins: versatility, modulation, and signaling in cell adhesion. Cell. 69, 11-25. (2) Sipehia, R. (1993). Enhanced attachment and growth of human endothelial cells derived from umbilical veins on ammonia plasma modified surfaces on PTFE and ePTFE synthetic vascular graft biomaterials. Biomater. Artif. Cells Immobilization Biotechnol. 21, 455-468. (3) Dekker, A., Reitsma, K., Beugeling, T., Bantjes, A., Feijen, J., and van Aken, W. G. (1991). Adhesion of endothelial cells and adsorption of serum proteins on gas plasma-treated polytetrafluoroethylene. Biomaterials 12, 130-138. (4) Massia, S. P., and Hubbell, J. A. (1991). Human endothelial cell interactions with surface-coupled adhesion peptides on a nonadhesive glass substrate and two polymeric biomaterials. J. Biomed. Mater. Res. 25, 223-42. (5) Ramires, P. A., Mirenghi, L., Romano, A. R., Palumbo, F., and Nicolardi, G. (2000). Plasma-treated PET surfaces improve the biocompatibility of human endothelial cells. J. Biomed. Mater. Res. 51, 535-539. (6) Koenig, A. L., Gambillara, V., and Grainger, D. W. (2003). Correlating fibronectin adsorption with endothelial cell adhesion and signaling on polymer substrates. J. Biomed. Mater. Res. 64A, 20-37. (7) Klueh, U., Goralnick, S., Bryers, J. D., and Kreutzer, D. L. (2001). Binding and orientation of fibronectin on surfaces with collagen-related peptides. J. Biomed. Mater. Res. 56, 307323. (8) Kowalczynska, H. M., Nowak-Wyrzykowska, M., Dobkowski, J., Kolos, R., Kaminski, J., Makowska-Cynka, A., and Mar-
Gauvreau and Laroche ciniak, E. (2002). Adsorption characteristics of human plasma fibronectin in relationship to cell adhesion. J. Biomed. Mater. Res. 61, 260-269. (9) Reinhart-King, C. A., Dembo, M., and Hammer, D. A. (2003). Endothelial cell traction forces on RGD-derivatized polyacrylamide substrata. Langmuir 19, 1573-1579. (10) Maheshwari, G., Brown, G., Lauffenburger, D. A., Wells, A., and Griffith, L. G. (2000). Cell adhesion and motility depend on nanoscale RGD clustering. J. Cell Sci. 113, 16771686. (11) Salacinski, H. J., Punshon, G., Krijgsman, B., Hamilton, G., and Seifalian, A. M. (2001). A Hybrid Compliant Vascular Graft Seeded with Microvascular Endothelial Cells Extracted from Human Omentum. Artif. Organs 25, p 974-982. (12) Tiwari A, Punshon, S. H., Hamilton G., and Seifalian A. M. (2002). Development of a hybrid cardiovascular graft using a tissue engineering approach. FASEB J. 16, 791-796. (13) Massia, S. P., and Stark, J. Immobilized RGD peptides on surface-grafted dextran promote biospecific cell attachment. (2001). J. Biomed. Mater. Res. 56, 390-399. (14) Porte-Durrieu, M. C., Labrugere, C., Villars, F., Lefebvre, F., Dutoya, S., Guette, A., Bordenave, L., and Baquey, C. (1999). Development of RGD peptides grafted onto silica surfaces: XPS characterization and human endothelial cell interactions. J. Biomed. Mater. Res. 46, 368-375. (15) Plow, E. F., Haas, T. A., Zhang, L., Loftus, J., and Smith, J. W. Ligand binding to integrins. (2000). J. Biol. Chem. 275, 21785-21788. (16) MacAlear, J., and Wehrung, J. (1978). Microsubstrate and method for making micropattern devices. US Patent 4,103,073. (17) MacAlear, J., and Wehrung, J. (1978). Microdevice substrate and method for making micropattern devices. US Patent 4,103,064. (18) Blawas, A. S., and Reichert, W. M. Protein patterning. (1998). Biomaterials 19, 595-609. (19) Lin, J. N., Andrade, J. D., and Chang, I. N. (1989). The influence of adsorption of native and modified antibodies on their activity. J. Immunol. Methods 125, 67-77. (20) Soderquist, M. E., and Walton, A. G. (1980). Structural changes in proteins adsorbed on polymer surfaces. J. Colloid Interface Sci. 75, 386-397. (21) Norde, W., and Lyklema, J. (1991). Why proteins prefer interfaces. J. Biomater. Sci. Polym. Ed. 2, 183-202. (22) Weetall, H. H. (1993). Preparation of immobilized proteins covalently coupled through silane coupling agents to inorganic supports. Appl. Biochem. Biotechnol. 41, 158-188. (23) Park, J. P., Lee, S. J., Park, T. J., Lee, K.-B., Choi, I. S., Lee, S. Y., Kim, M.-G., and Chung, B. H. (2004). Microcontact printing of biotin for selective immobilization of streptavidinfused proteins and SPR analysis. Biotechnol. Bioprocess Eng. 9, 137-142. (24) Mooradian, D. L., McCarthy, J. B., Skubitz, A. P., Cameron, J. D., and Furcht, L. T. (1993). Characterization of FN-C/ H-V, a novel synthetic peptide from fibronectin that promotes rabbit corneal epithelial cell adhesion, spreading, and motility. Investig. Ophthalmol. Vis. Sci. 34, 153-164. (25) Huebsch, J. C., McCarthy, J. B., Diglio, C. A., and Mooradian, D. L. (1995). Endothelial cell interactions with synthetic peptides from the carboxyl-terminal heparin-binding domains of fibronectin. Circ. Res. 77, 43-53. (26) Chevallier, P., Castonguay, M., Turgeon, S., Dubrulle, N., Mantovani, D., McBreen, P. H., Wittmann, J. C., and Laroche, G. (2001). Ammonia RF-Plasma on PTFE Surfaces: Chemical Characterization of the Species Created on the Surface by Vapor-Phase Chemical Derivatization. J. Phys. Chem. B 105, 12490-12497. (27) Mantovani, D., Castonguay, M., Pageau, J. F., Fiset, M., and Laroche, G. (1999). Ammonia RF-plasma treatment of tubular ePTFE vascular prostheses. Plasmas Polym. 4, 207228. (28) Gauvreau, V., Chevallier, P., Vallie`res, K., Petitclerc, E Ä ., C.-Gaudreault, R., and Laroche, G. (2004). Engineering Surfaces for Bioconjugation: Developing Strategies and Quantifying the Extent of the Reactions. Bioconjugate Chem. 15, 1146-1156.
Bioconjugate Chem., Vol. 16, No. 5, 2005 1097
Micropattern Printing of Peptides on PTFE Films (29) Hermanson, G. T., Bioconjugate Techniques. (1996), San Diego: Academic Press. 785. (30) Stever, S. (2002). The Illustrator 10 Wow! Book, p 407, Peachpit Press, Berkeley. (31) Okamoto, T., Suzuki, T., and Yamamoto, N. (2000). Microarray fabrication with covalent attachment of DNA using bubble jet technology. Nature Biotechnol. 18, 438-441. (32) Allain, L. R., Askari, M., Stokes, D. L., and Vo-Dinh, T. (2001). Microarray sampling-platform fabrication using bubblejet technology for a biochip system. Fresenius J. Anal. Chem. 371, 146-150. (33) Roda, A., Guardigli, M., Russo, C., Pasini, P., and Baraldini, M. (2000). Protein microdeposition using a conventional ink-jet printer. BioTechniques 28, 492-496. (34) Fussell, G. W., and Cooper, S. L. (2004). Endothelial cell adhesion on RGD-containing methacrylate terpolymers. J. Biomed. Mater. Res. 70A, 265-273. (35) Pfaff, M., McLane, M., Beviglia, L., Niewiarowski, S., and Timpl, R. (1994). Comparison of disintegrins with limited variation in the RGD loop in their binding to purified integrins alpha IIb beta 3, alpha V beta 3 and alpha 5 beta 1 and in cell adhesion inhibition. Cell Adhes. Commun. 2, 491-501. (36) Flusberg, D. A., Numaguchi, Y., and Ingber, D. E. (2001). Cooperative control of Akt phosphorylation, bcl-2 expression, and apoptosis by cytoskeletal microfilaments and microtubules in capillary endothelial cells. Mol. Biol. Cell 12, 30873094. (37) Ingber, D. E. (2002). Mechanical Signaling and the Cellular Response to Extracellular Matrix in Angiogenesis and Cardiovascular Physiology. Circ. Res. 91, 877-887.
(38) Ingber, D. E. (2003). Tensegrity II. How structural networks influence cellular information processing networks. J. Cell Sci. 116, 1397-1408. (39) Chen, C. S., Mrksich, M., Huang, S., Whitesides, G. M., and Ingber, D. E. (1997). Geometric control of cell life and death. Science 276, 1425-1428. (40) Woods, A., McCarthy, J. B., Furcht, L. T., and Couchman, J. R. (1993). A synthetic peptide from the COOH-terminal heparin-binding domain of fibronectin promotes focal adhesion formation. Mol. Biol. Cell. 4, 605-613. (41) Chen, C. S., Mrksich, M., Huang, S., Whitesides, G. M., and Ingber, D. E. (1998). Micropatterned surfaces for control of cell shape, position, and function. Biotechnol. Prog. 14, 356-363. (42) Irvine, D. J., Mayes, A. M., and Griffith, L. G. (2001). Nanoscale clustering of RGD peptides at surfaces using Comb polymers. 1. Synthesis and characterization of Comb thin films. Biomacromolecules 2, 85-94. (43) Lin, H. B., Sun, W., Mosher, D. F., Garcia-Echeverria, C., Schaufelberger, K., Lelkes, P. I., and Cooper, S. L. (1994). Synthesis, surface, and cell-adhesion properties of polyurethanes containing covalently grafted RGD-peptides. J. Biomed. Mater. Res. 28, 329-342. (44) Harrison, D., Johnson, R., Tucci, M., Puckett, A., Tsao, A., Hughes, J., and Benghuzzi, H. (1997). Interaction of cells with UHMWPE impregnated with the bioactive peptides RGD, RGE or Poly-L-lysine. Biomed. Sci. Instrum. 34, 41-46.
BC049717S