- Email: [email protected]
Surface microstructuring and protein patterning using hyaluronan derivatives
Microelectronic Engineering 106 (2013) 21–26
Contents lists available at SciVerse ScienceDirect
Microelectronic Engineering journal homepage: www.elsevier.com/locate/mee
Surface microstructuring and protein patterning using hyaluronan derivatives M.C. Márquez-Posadas a,⇑, J. Ramiro a,1, J. Becher b,2, Y. Yang c,3, A. Köwitsch c,3, I. Pashkuleva d, R. Díez-Ahedo a,1, M. Schnabelrauch b,2, R.L. Reis d, T. Groth c,3, S. Merino a,1 a
IK4-Tekniker, Micro and Nano Manufacture Unit, Polo Tecnológico De Eibar, C/ Iñaki Goenaga 5, 20600 Eibar, Gipuzkoa, Spain Innovent e.V., Biomaterials Department, Prüssingstraße 27 B, D-07745 Jena, Germany c Biomedical Materials Group, Martin Luther University, Kurt-Mothes-Strasse 1, Halle, 06120, Germany d 3B’s Research Group, University of Minho, Headquarters of the European Institute of Excellence on Tissue Engineering and Regenerative Medicine, AvePark, 4806-909 Taipas, Guimarães, Portugal e ICVS/3B’s - PT Government Associate Laboratory, Braga/Guimarães, Portugal b
a r t i c l e
i n f o
Article history: Received 13 September 2012 Received in revised form 1 December 2012 Accepted 12 January 2013 Available online 8 February 2013 Keywords: Microstructuring Hyaluronan Glycosaminoglycan Hydrogel Photocrosslinking Microcontact printing
a b s t r a c t Natural polymers, such as hyaluronan, are considered as good candidates to substitute synthetic materials in many biological applications due to their intrinsic biocompatibility for in vitro and living cell experiments. This work describes surface modifications performed over modified hyaluronan derivatives which could help to understand the physical and biochemical cues on the cell-behaviour. A photocrosslinkable methacrylated hyaluronan was microstructured using soft lithography techniques to obtain a microenvironment suitable for cell-behaviour experiments, which might mimic the extra-cellular matrix. In addition, sulphated and non-sulphated oxidized hyaluronan were immobilized on bare substrates and micrometric features of cell adhesive and non-adhesive proteins were patterned by microcontact printing on top of them. The obtained structures were characterized by optical microscopy, profilometry and atomic force microscopy. The stability of structures was tested by immersion in physiological salt solutions. The observed results prove the suitability of the materials and protocols described. Ó 2013 Elsevier B.V. All rights reserved.
1. Introduction Natural polymers have been extensively used in biological research, tissue engineering, diagnostics and screening [1,2]. Among them, glycosaminoglycans (GAGs) are large complex carbohydrate molecules related to many physiological processes. These molecules are present on all animal cell surfaces in the extracellular matrix (ECM) [3] and they include two main types: non-sulphated GAGs, such as hyaluronan (HA) and sulphated GAGs (e.g. chondroitin sulphate, heparin). Since they are found throughout the body, these GAGs offer much better biocompatibility than synthetic polymers, a prerequisite for in vitro cell-based studies and devices. In fact, HA has been approved and used clinically [4] and it is considered a key biomaterial for the development of tissue engineering [2]. ⇑ Corresponding author. Tel.: +34 943 20 67 44; fax: +34 943 20 27 57. E-mail addresses: [email protected] (M.C. Márquez-Posadas), [email protected] (J. Ramiro), [email protected] (J. Becher), yuan.yang@ pharmazie.uni-halle.de (Y. Yang), [email protected] (I. Pashkuleva), [email protected] (R. Díez-Ahedo), [email protected] (M. Schnabelrauch), [email protected] (R.L. Reis), [email protected] (T. Groth), [email protected] (S. Merino). 1 Tel.: +34 943 20 67 44. 2 Tel.: +49 3641 2825 12. 3 Tel.: +49 345 5528461. 0167-9317/$ - see front matter Ó 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.mee.2013.01.039
Traditionally, the interactions of GAGs with proteins and cells have been studied in solution. However, since GAGs in the ECM are immobilized (covalently or physically), in vitro immobilized GAGs may better mimic the in vivo environment for cell culture experiments. Therefore, the interaction of cells with substrates is a vast field of intense research due to its relevance for very different applications [5]. Since the ECM possesses different topographical and adhesive features, the response of cells to surface topography, such as grooves, pillars, etc. has been studied extensively [6]. The ECM also exposes adhesive and non-adhesive features which have been mimicked in vitro by surface-patterning of proteins [7]. However, the way the micro or nanotopography regulates cell behaviour is not well understood [8]. Due to this lack, there is a need for developing adequate functional surface patterns for in vitro cell research. Topographical patterns for cell studies are considered a powerful mechanism to alter, direct and control in vitro cell behaviour from cell adhesion to gene expression [8–10]. Despite some promising results, these structures still do not completely mimic the microenvironment found in the ECM [7]. Since it is known that GAGs have a biochemical effect on the cells, the combination of mechanical effects by microstructuring and biochemical cues induced by GAGs is an interesting field of research. Unfortunately, the high solubility of these materials is a problem when cellculture experiments are carried out [11]. However, by using
22
M.C. Márquez-Posadas et al. / Microelectronic Engineering 106 (2013) 21–26
photocrosslinkable HA derivatives, whose properties can be modified by adjusting the crosslinking density of the resulting hydrogel, suitable microenvironments have been already created [2,12–14]. The present work describes surface structuring on modified HA derivatives. Methacrylated hyaluronan (MAHA) has been microstructured by photocrosslinking and soft lithography techniques to obtain a microenvironment emulating the ECM. In another approach, two modified GAGs, oxidized hyaluronan and oxidized sulphated hyaluronan (ox-HA and ox-HAS), have been covalently immobilized onto bare substrates giving rise to a self-assembled monolayer. Resistant micrometric features of cell adhesive (fibronectin, FN) and non-adhesive (bovine serum albumin, BSA) proteins have been transferred to these substrates by microcontact printing (lCP) [15]. The resulting structures were characterized by optical microscopy, profilometry and atomic force microscopy (AFM) and their resistance to immersion was tested showing that they could be used for cell-behaviour studies. 2. Materials and methods MAHA with a degree of substitution (average number of methacrylated groups per disaccharide repeating unit of HA) of 0.9 was prepared starting from native high-molecular weight HA (MW = 923.6 kDa) according to a reported procedure [16]. Ox-HA with 10% oxidation percentage (104 kg/mol, polydispersity index of 1.66) and ox-HAS with 11.3% oxidation percentage and a sulphate degree of 2.4 were also synthesized. 3-(Trimethoxysily)propyl methacrylate (TMSPM), 11-amino-1-undecanethiol hydrochloride, 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid (HEPES) 1 M buffering agent, sodium cyanoborohydride (NaBH3CN), triethylene glycol dimethacrylate (TEGDM), Dulbecco’s phosphate buffered saline (PBS) and hexamethyldisilazane 98% (HMDS) were purchased from Sigma–Aldrich. 3-Aminopropyl trimethoxysilane (APTMS) and 1H, 1H, 2H, 2H-perfluorooctyl trichlorosilane (FOTS) were obtained from ABCR GmbH. A Sylgard 184 poly(dimethyl siloxane) (PDMS) was purchased from Dow Corning. Irgacure 369 was obtained from CIBA. Alexa Fluor 488-conjugated BSA (BSA Alexa Fluor 488), fibronectin (FN) were provided by Fisher Scientific SL. Hilyte 488-conjugated fibronectin (FN 488) was purchased from Cytoskeleton. Ultrapure water was used for all dilutions and washings and all reagents were of analytical grade. 2.1. Fabrication of silicon masters Silicon masters, for PDMS stamps fabrication, with microstructures consisting of grooves from 4 to 200 lm of periodicity were fabricated by UV photolithography and deep reactive ion etching (DRIE). Four inch silicon wafers were first dehydrated at 180 °C for 30 min in a convection oven. A first layer of an adhesion promoter, HMDS, and a positive photoresist (maP-1205) were spin coated and processed using a conventional UV-Lithography process. These patterns were transferred to the silicon by DRIE (Oxford Plasmalab System 80) using a combined SF6/C4F8 plasma. Silicon stamps with two different depths (240 nm for micro-structuration studies and 5 lm for lCP experiments) were obtained depending on the etching time. Finally, FOTS was evaporated over the silicon stamp in a vacuum desiccator during 30 min as anti-adhesive to facilitate the PDMS stamp peeling off. 2.2. Fabrication of PDMS stamps PDMS stamps were fabricated using the silicon masters described in the previous section. To cure the PDMS prepolymer, a mixture of 10:1 silicone elastomer and the curing agent was cast over silicon masters and placed at 60 °C overnight. PDMS was then
carefully peeled off from the silicon master and cut into 10 mm 10 mm stamps. The protruding ridges in the masters resulted in PDMS replicas with grooves. Finally, the stamps were cleaned with 70% ethanol, sonicated for 5 min and let dry prior to use. 2.3. Substrate preparation 2.3.1. Glass substrates Four inch glass wafers were diced into 10 mm 10 mm substrates and cleaned prior to use by copious rinsing with acetone, isopropyl alcohol and ultrapure water. Glass substrates were methacrylated after 5 min O2 plasma treatment by deposition of a drop of TMSPM. The substrates were kept in an oven at 100 °C for 1 h followed by 15 min at 110 °C [2]. Substrates were rinsed with ultrapure water and dried under a N2 flow. Amination of glass substrates for subsequent ox-HA and ox-HAS deposition was performed by evaporation of APTMS. 2.3.2. Gold substrates Another set of glass wafers was cleaned as described in Section 2.3.1. Thereafter, a 40 nm thick gold layer was deposited onto the wafers by electron beam physical vapour deposition (ATC Orion series UHV Evaporation system, AJA International Inc.). A previous evaporation step of a 3 nm thick layer of titanium was carried out to improve adhesion to the glass. These wafers were also diced into 10 mm 10 mm substrates. These substrates were cleaned using a piranha solution (H2SO4:H2O2, 3:1) for 10 min, rinsed with ethanol and dried using N2. Subsequently, substrates were immersed in a 20 lM solution of 11-amino-1-undecanethiol hydrochloride in pure ethanol for 48 h. After amination, substrates were washed with ethanol and dried under a N2 flow to be ready for GAG immobilization. 2.4. Microstructuring of MAHA Two percent and 5% w/v MAHA solutions in water were filtered using 0.2 lm diameter porous filters (Millex, Millipore) and blended with 0.3% m/v Irgacure 369 in N-vinyl pyrrolidone and with TEGDM as an additional crosslinker using a MAHA:TEGDM 1:1 ratio. One hundred and thirty microlitres of these mixtures were deposited onto the microstructured PDMS stamps and the methacrylated glass substrates were carefully placed over them. Photopolymerization was performed under UV light from a 350 W short arc lamp (Advanced Radiation Corporation, 350– 450 nm, 10 mW/cm2) for 2 h under a N2 atmosphere. After UV exposure, the PDMS stamp was left overnight before demoulding to ensure the stabilization of the hydrogel microstructures (Fig. 1a). After demoulding of the stamps from the microstructured MAHA, a Veeco Dektak 8 (Veeco Instruments, Plainview, NY) mechanical profilometer was used for profile measurements. 2.5. Immobilization of ox-HA and ox-HAS GAGs over aminated gold/ glass substrates Ox-HA and ox-HAS were immobilized onto substrates as follows: each well of a 12-well plate containing a single aminated gold or glass substrate was filled with 1 ml of 4 mg/ml ox-HA or ox-HAS solution in PBS. The plates were kept shaking gently for 24 h. One millilitre of NaBH3CN (3 mg/ml in PBS) was added into each well and kept at 4 °C for another 24 h. After the ox-GAG immobilization, the coated surface was once more rinsed with water and dried under a N2 stream.
M.C. Márquez-Posadas et al. / Microelectronic Engineering 106 (2013) 21–26
23
Fig. 1. Schematic diagram of the surface modifications techniques performed over modified HA, (a) microstructuring of MAHA: MAHA is dropped over the PDMS stamp and the substrate is placed over it and exposure to UV light. After the PDMS stamp release microstructured hydrogel is obtained, (b) lCP of proteins over ox-HA/ox-HAS: ox-HA/ ox-HAS are immobilized covalently over an aminated gold/glass substrate. Protein solutions (BSA or FN) are incubated for 1 h over a PDMS stamp and afterwards the stamp is rinsed, dried and placed onto the ox-HA/ox-HAS substrates applying a gently pressure for getting a conformal contact for 1 min. After detachment of the PDMS stamp, a protein patterning is obtained.
2.6. lCP of proteins over immobilized ox-HA and ox-HAS GAGs
lCP of BSA Alexa Fluor 488 and a 1:100 mixture of FN 488 (fluorescent) and FN (non-fluorescent) was carried out over ox-HA and ox-HAS GAGs substrates. Hundred microlitres of 100 lg/ml BSA or FN were incubated over microstructured PDMS stamps for 1 h for surface protein absorption. Afterwards, the stamps were rinsed with water and dried under a N2 flow. The lCP was performed by placing the stamps over the substrates, applying a gentle pressure during 1 min to get a conformal contact with the surfaces and then carefully peeling them off (Fig. 1b). The stamps were examined after demoulding to check a possible deterioration which could imply a deterioration of subsequent patterns. Characterization of the patterns was performed with a NT-MDT AFM and an AXIO Imager.A1m microscope (Carl Zeiss) equipped with a CCD camera and adequate fluorescence filters.
3. Results and discussion 3.1. MAHA microstructuring HA is a cell surface-associated polysaccharide, ubiquitous in the ECM, which is known to have biochemical and mechanical effects on the cells [17]. Therefore, microstructures made of HA could offer an in vitro microenvironment resembling that found in the ECM. In the present study, we have microstructured a photocrosslinkable HA derivative, MAHA (Fig. 2). The crosslinking density and the polysaccharide concentration have been adjusted for optimal results. The solubility of the microstructured material has also been tested to ensure their ability to be used in cell-behaviour experiments.
As mentioned in Section 2.3.1, the glass substrates had to be methacrylated, prior to the microstructuring of MAHA, in order to enhance the adhesion of the MAHA. Methacrylate groups on the substrate react with those of the MAHA hydrogel substantially improving substrate adhesion. It was experimentally tested that the lack of this glass methacrylation step gave rise to a poor adhesion of the MAHA, causing detachment of the MAHA hydrogel layer from the glass substrates during the immersion step. Taking into account that, due to their low crosslinking density, non modified HA hydrogels are often readily degraded in aqueous solutions such as a cell culture medium [16], the use of an additional crosslinker, TEGDM, was tested in order to increase crosslinking density and, therefore the immersion resistance and mechanical stability of the resulting hydrogels. Finally, in order to minimize oxygen diffusion into the samples during polymerization, as molecular oxygen is a radical scavenger and a radical quencher that can readily terminate initiating and propagating radicals [18], UV light exposure was performed in a N2 atmosphere. Fig. 3 shows optical microscopy images of the microstructures present on the MAHA hydrogel surface. The replication quality was high and the homogeneity over the whole area was remarkable for substrates with and without additional crosslinker (see Fig. 3a and b, respectively). In order to characterize the filling of PDMS microstructures by the hydrogels, profile measurements of both the stamp and the microstructured hydrogel were performed and are shown in Fig. 4. In this figure, it can be clearly seen that the microstructures underwent an anisotropic shrinkage, maintaining the period size but showing a lower depth than the PDMS stamp. The depth difference was approximately 90 nm which could be attributed to the evaporation of water from the hydrogel (as the PDMS stamp is left over the hydrogel overnight before detachment). Another possible reason could be a not complete filling of
Fig. 2. Scheme of hyaluronan methacrylation reaction using glycidyl methacrylate.
24
M.C. Márquez-Posadas et al. / Microelectronic Engineering 106 (2013) 21–26
Fig. 3. Optical micrograph of microstructured MAHA after demoulding. (a) 4 lm periodic grooves of 2% w/v without the use of additional crosslinker and (b) 4 lm periodic grooves of 5% w/v with additional crosslinker.
be convenient for microstructuring of polysaccharides (such as HA) due to their high hydrophilicity [11], the use of stamps without this plasma treatment has been successful in the present study. It has also been suggested that, due to their solubility in water, a coating with non-modified HA hydrogel would be washed away during immersion leading to a chemisorbed layer only [11]. However, probably due the use of an additional crosslinker, this phenomenon was not observed in this study. The results of the present study could pave the way to future cell-behaviour studies using microstructured substrates composed of MAHA.
3.2. lCP of BSA and FN over ox-HA and ox-HAS GAGs substrates
Fig. 4. Profile of the microstructured MAHA (black) obtained by mechanical profilometry. The PDMS stamp profile is also shown as comparison (dark grey). The microstructures have a period of 50 lm.
the PDMS stamp grooves. A similar behaviour has been previously described and attributed to the stresses created during the crosslinking process [19]. Microstructured MAHA glass substrates were immersed in 10 mM HEPES buffer solution for 24 h to test the material resistance under salt concentration or pH conditions similar to those found in cell-culture studies. Fig. 5 shows pictures of substrates during immersion. As can be seen, the surface of MAHA 5% w/v samples withstood the immersion (Fig. 5a). However, due to its lower concentration, in the case of MAHA 2% w/v samples, only some structured small areas remained (Fig. 5b). MAHA samples without additional crosslinker did not withstand the immersion test. These results show that an increase of the concentration causes an improvement of the resistance of the structured surfaces to immersion, although a compromise is necessary. Khademhosseini et al. [2] described that a high concentration of HA, although desirable, could affect negatively the moulding process because of the excessive viscosity of solutions with high concentrations of HA or HA derivatives. They established a 5% concentration as optimal. It is worth mentioning that although it can be found in literature that a previous oxygen plasma treatment of the stamps could
The degree of sulfation of GAGs affects cells adhesion, growth and differentiation [20–22]. In addition, sulphated GAGs bind a wide variety of ECM proteins and growth factors. Hence a combination of GAGs with proteins in spatial well-designed arrangements, like micro patterns to obtain adhesive and non-adhesive areas, could provide useful information on cell behaviour and means to control cell activity. Hyaluronan and its sulphated derivatives possess sugar rings with vicinal hydroxyl groups which were oxidized with sodium periodate to generate free aldehyde groups (see Fig. 6). More details of the procedure are described in a previous work [20]. FN and BSA are proteins, with well-known cell adhesive and non-adhesive properties, respectively, which can be printed onto a wide variety of substrates to obtain patterns suitable for cellculture studies [7]. Since ox-GAGs may also bind proteins covalently via cross-linking of GAG-aldehyde amino side-groups of proteins, either FN or BSA were transferred by lCP to the immobilized ox-GAG layers. In order to carry out lCP of proteins, gold and glass substrates had to be aminated using an amino-thiol or amino-silane reagent, so that the aldehydes of ox-HA and ox-HAS could react with the amino groups of substrates via Schiff’s base reaction. The addition of NaBH3CN solution reduced the intermediate imines to stable amines [20]. Subsequently proteins, like BSA or FN were immobilized on ox-HA and ox-HAS modified surfaces using lCP. After the lCP of fluorescent BSA and FN, the transferred patterns were characterized by fluorescence microscopy (see Fig. S1) and AFM. The binding of proteins was, on average, reduced on HA surfaces when compared to glass [11]. However, from the fluorescence microscopy images (Fig. 7) it can be seen that the obtained patterns of both proteins were well defined on the ox-GAGs substrates.
M.C. Márquez-Posadas et al. / Microelectronic Engineering 106 (2013) 21–26
25
Fig. 5. Optical microscopy photograph of microstructured MAHA during immersion in HEPES 10 mM. (a) 4 and 50 lm periodic grooves of 5% w/v with additional crosslinker and (b) 4 lm periodic grooves of 2% w/v with additional crosslinker. The insets images show a detail with 4 lm periodic grooves.
Fig. 6. Chemical structures of, (a) oxidized hyaluronan (ox-HA) and (b) oxidized hyaluronan sulphate (ox-HAS).
Fig. 7. Fluorescence microscopy images of microcontact printed proteins. (a) BSA Alexa Fluor 488 (100 lg/ml concentration) patterns of 50 lm periodicity over immobilized ox-HAS. (b) FN Hilyte 488-conjugated (100 lg/ml concentration) patterns of 100 lm periodicity over immobilized ox-HA. The inset images show the same areas after 24 h immersion in PBS.
Fig. 7a shows a fluorescence image of the lCP over ox-HAS substrates which was performed with a concentration of 100 lg/ml of Alexa Fluor 488 conjugated BSA solution. Some fluorescent spots can be noticed between fluorescent patterns before PBS immersion, which could be caused by small drops of protein solution that remain in the stamp grooves after the drying process and before the lCP. After PBS immersion, a decrease of the fluorescence intensity can be appreciated. It might be attributed to a small loss of weakly bound proteins that after rinsing and drying were spread over the patterned surface (see inset in Fig. 7a and Fig. S1). Fig. 7b shows the lCP over ox-HA substrates which was carried out with a 100:1 ratio of FN and Hilyte 488-conjugated FN. The fainter-fluorescence of the FN pattern and the less homogeneous fluorescent patterning could be caused by the lower concentration of fluorophore (as mentioned in Section 2.6).
The AFM image (Fig. 8) shows the topography of the proteinprinted and non-printed areas. A 4–5 nm step was measured by AFM in semi-contact mode, which is consistent with the height of a protein layer [23]. The stability of the protein pattern was confirmed by the immersion of printed surfaces in PBS for 24 h. The effects of topographical modifications in cell-behaviour experiments are usually studied before reaching confluence, a process which takes place in approximately 24–48 h, depending on the cell type. The samples were rinsed with ultrapure water and dried with N2 to remove salt aggregates. The samples were again studied by fluorescence microscopy (Fig. 7 insets) finding that the homogeneity of the patterns was kept and also proving that the protein adsorption to oxHA and ox-HAS was strong enough to allow future cell-behaviour studies.
26
M.C. Márquez-Posadas et al. / Microelectronic Engineering 106 (2013) 21–26
Fig. 8. AFM image of a BSA pattern of 4 lm periodicity over ox-HAS. (a) Topography, (b) profile of the marked line shows BSA height around 4–5 nm.
4. Conclusions In this study we have described surface modifications (structuring and protein patterning) performed over modified novel HA derivatives (MAHA, ox-HA and ox-HAS) deposited on solid substrates such as gold and glass. Soft lithography and microstructuring approaches have been successfully used here to obtain the desired microstructures. The surface topography obtained for HA-based natural polysaccharides can be used to achieve in vitro conditions similar to those found in the ECM. The feasibility of using the obtained substrates in, for instance, future cell-behaviour experiments was checked by immersion resistance showing the stability of the immobilized GAG layers. Acknowledgments This work was supported and carried out under the scope of the European Project Find and Bind (NMP4-SL-2009-229292). Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.mee.2013.01.039. References [1] N.A. Peppas, J.Z. Hilt, A. Khademhosseini, R. Langer, Adv. Mater. 18 (2006) 1345–1360. [2] A. Khademhosseini, G. Eng, J. Yeh, J. Fukuda, J. Blumling III, R. Langer, J.A. Burdick, J. Biomed. Mater. Res. A 79 (3) (2006) 522–532. [3] N.S. Gandhi, R.L. Mancera, Chem. Biol. Drug Des. 72 (2008) 455–482.
[4] J.W. Kuo, Practical Aspects of Hyaluronan Based Medical Products, CRC/Taylor & Francis, Boca Raton, 2006. [5] L.E. Dickinson, C.C. Ho, G.M. Wang, K.J. Stebe, S. Gerecht, Biomaterials 31 (2010) 5472–5478. [6] Kenneth.E. Gonsalves, Craig.R. Halberstadt, Cato.T. Laurencin, Lakshmi.S. Nair, Biomedical Nanostructures, John Wiley & Sons, Hoboken, 2008. [7] L. Kam, S.G. Boxer, J. Biomed. Mater. Res. 55 (2001) 487–495. [8] D.H. Kim, P.P. Provenzano, C.L. Smith, A. Levchenko, J. Cell Biol. 197 (3) (2012) 351–360. [9] D. Guduru, M.S. Niepel, J. Vogel, T. Groth, Int. J. Artif. Organs 34 (10) (2011) 963–985. [10] M.S. Kim, A.Y. Kim, K.J. Jang, J.H. Kim, J.B. Kim, K.Y. Suh, Nanotechnology 22 (2011) 494017–494025. [11] K.Y. Suh, A. Khademhosseini, J.M. Yang, G. Eng, R. Langer, Adv. Mater. 16 (7) (2004) 584–588. [12] J.A. Burdick, C. Chung, X. Jia, M.A. Randolph, R. Langer, Biomacromolecules 6 (2005) 386–391. [13] J.B. Leach, C.E. Schmidt, Biomaterials 26 (2005) 125–135. [14] K.S. Masters, D.N. Shah, L.A. Leinwand, K.S. Anseth, Biomaterials 26 (2005) 2517–2525. [15] S.M. Ruiz, C.S. Chen, Soft Matter 3 (2007) 168–177. [16] S. Möller, J. Weisser, S. Bischoff, M. Schnabelrauch, Biomol. Eng. 24 (2007) 496–504. [17] R.A.A. Muzzarellia, F. Grecoa, A. Busilacchia, V. Sollazzo, A. Gigante, Carbohydr. Polym. 89 (2012) 723–739. [18] S.J. Bryant, K.S. Anseth, Photopolimerization of Hydrogel Scaffolds, in: P.X. Ma, J. Elisseeff (Eds.), Scaffolding in Tissue Engineering, Taylor and Francis Group, 2006, pp. 71–90. [19] S. Rodriguez Vilches, C. Séverac, C. Thibaut, L. Laplatine, C. Vieu, J. Fitremann, A.F. Mingotaud, P. Martinoty, D. Collin, Colloid Polym. Sci. 289 (2011) 1437– 1449. [20] A. Köwitsch, Y. Yang, N. Ma, J. Kuntsche, K. Mäder, T. Groth, Biotechnol. Appl. Biochem. 58 (2011) 376–389. [21] D. Peschel, K. Zhang, S. Fischer, T. Groth, Acta Biomater. 8 (2012) 183–193. [22] D.S. Costa, R.A. Pires, A.M. Frias, R.L. Reis, I. Pashkuleva, J. Mater. Chem. 22 (2012) 7172–7178. [23] H.D. Inerowicz, S. Howell, F.E. Regnier, R. Reifenberger, Langmuir 18 (2002) 5263–5268.
Contents lists available at SciVerse ScienceDirect
Microelectronic Engineering journal homepage: www.elsevier.com/locate/mee
Surface microstructuring and protein patterning using hyaluronan derivatives M.C. Márquez-Posadas a,⇑, J. Ramiro a,1, J. Becher b,2, Y. Yang c,3, A. Köwitsch c,3, I. Pashkuleva d, R. Díez-Ahedo a,1, M. Schnabelrauch b,2, R.L. Reis d, T. Groth c,3, S. Merino a,1 a
IK4-Tekniker, Micro and Nano Manufacture Unit, Polo Tecnológico De Eibar, C/ Iñaki Goenaga 5, 20600 Eibar, Gipuzkoa, Spain Innovent e.V., Biomaterials Department, Prüssingstraße 27 B, D-07745 Jena, Germany c Biomedical Materials Group, Martin Luther University, Kurt-Mothes-Strasse 1, Halle, 06120, Germany d 3B’s Research Group, University of Minho, Headquarters of the European Institute of Excellence on Tissue Engineering and Regenerative Medicine, AvePark, 4806-909 Taipas, Guimarães, Portugal e ICVS/3B’s - PT Government Associate Laboratory, Braga/Guimarães, Portugal b
a r t i c l e
i n f o
Article history: Received 13 September 2012 Received in revised form 1 December 2012 Accepted 12 January 2013 Available online 8 February 2013 Keywords: Microstructuring Hyaluronan Glycosaminoglycan Hydrogel Photocrosslinking Microcontact printing
a b s t r a c t Natural polymers, such as hyaluronan, are considered as good candidates to substitute synthetic materials in many biological applications due to their intrinsic biocompatibility for in vitro and living cell experiments. This work describes surface modifications performed over modified hyaluronan derivatives which could help to understand the physical and biochemical cues on the cell-behaviour. A photocrosslinkable methacrylated hyaluronan was microstructured using soft lithography techniques to obtain a microenvironment suitable for cell-behaviour experiments, which might mimic the extra-cellular matrix. In addition, sulphated and non-sulphated oxidized hyaluronan were immobilized on bare substrates and micrometric features of cell adhesive and non-adhesive proteins were patterned by microcontact printing on top of them. The obtained structures were characterized by optical microscopy, profilometry and atomic force microscopy. The stability of structures was tested by immersion in physiological salt solutions. The observed results prove the suitability of the materials and protocols described. Ó 2013 Elsevier B.V. All rights reserved.
1. Introduction Natural polymers have been extensively used in biological research, tissue engineering, diagnostics and screening [1,2]. Among them, glycosaminoglycans (GAGs) are large complex carbohydrate molecules related to many physiological processes. These molecules are present on all animal cell surfaces in the extracellular matrix (ECM) [3] and they include two main types: non-sulphated GAGs, such as hyaluronan (HA) and sulphated GAGs (e.g. chondroitin sulphate, heparin). Since they are found throughout the body, these GAGs offer much better biocompatibility than synthetic polymers, a prerequisite for in vitro cell-based studies and devices. In fact, HA has been approved and used clinically [4] and it is considered a key biomaterial for the development of tissue engineering [2]. ⇑ Corresponding author. Tel.: +34 943 20 67 44; fax: +34 943 20 27 57. E-mail addresses: [email protected] (M.C. Márquez-Posadas), [email protected] (J. Ramiro), [email protected] (J. Becher), yuan.yang@ pharmazie.uni-halle.de (Y. Yang), [email protected] (I. Pashkuleva), [email protected] (R. Díez-Ahedo), [email protected] (M. Schnabelrauch), [email protected] (R.L. Reis), [email protected] (T. Groth), [email protected] (S. Merino). 1 Tel.: +34 943 20 67 44. 2 Tel.: +49 3641 2825 12. 3 Tel.: +49 345 5528461. 0167-9317/$ - see front matter Ó 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.mee.2013.01.039
Traditionally, the interactions of GAGs with proteins and cells have been studied in solution. However, since GAGs in the ECM are immobilized (covalently or physically), in vitro immobilized GAGs may better mimic the in vivo environment for cell culture experiments. Therefore, the interaction of cells with substrates is a vast field of intense research due to its relevance for very different applications [5]. Since the ECM possesses different topographical and adhesive features, the response of cells to surface topography, such as grooves, pillars, etc. has been studied extensively [6]. The ECM also exposes adhesive and non-adhesive features which have been mimicked in vitro by surface-patterning of proteins [7]. However, the way the micro or nanotopography regulates cell behaviour is not well understood [8]. Due to this lack, there is a need for developing adequate functional surface patterns for in vitro cell research. Topographical patterns for cell studies are considered a powerful mechanism to alter, direct and control in vitro cell behaviour from cell adhesion to gene expression [8–10]. Despite some promising results, these structures still do not completely mimic the microenvironment found in the ECM [7]. Since it is known that GAGs have a biochemical effect on the cells, the combination of mechanical effects by microstructuring and biochemical cues induced by GAGs is an interesting field of research. Unfortunately, the high solubility of these materials is a problem when cellculture experiments are carried out [11]. However, by using
22
M.C. Márquez-Posadas et al. / Microelectronic Engineering 106 (2013) 21–26
photocrosslinkable HA derivatives, whose properties can be modified by adjusting the crosslinking density of the resulting hydrogel, suitable microenvironments have been already created [2,12–14]. The present work describes surface structuring on modified HA derivatives. Methacrylated hyaluronan (MAHA) has been microstructured by photocrosslinking and soft lithography techniques to obtain a microenvironment emulating the ECM. In another approach, two modified GAGs, oxidized hyaluronan and oxidized sulphated hyaluronan (ox-HA and ox-HAS), have been covalently immobilized onto bare substrates giving rise to a self-assembled monolayer. Resistant micrometric features of cell adhesive (fibronectin, FN) and non-adhesive (bovine serum albumin, BSA) proteins have been transferred to these substrates by microcontact printing (lCP) [15]. The resulting structures were characterized by optical microscopy, profilometry and atomic force microscopy (AFM) and their resistance to immersion was tested showing that they could be used for cell-behaviour studies. 2. Materials and methods MAHA with a degree of substitution (average number of methacrylated groups per disaccharide repeating unit of HA) of 0.9 was prepared starting from native high-molecular weight HA (MW = 923.6 kDa) according to a reported procedure [16]. Ox-HA with 10% oxidation percentage (104 kg/mol, polydispersity index of 1.66) and ox-HAS with 11.3% oxidation percentage and a sulphate degree of 2.4 were also synthesized. 3-(Trimethoxysily)propyl methacrylate (TMSPM), 11-amino-1-undecanethiol hydrochloride, 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid (HEPES) 1 M buffering agent, sodium cyanoborohydride (NaBH3CN), triethylene glycol dimethacrylate (TEGDM), Dulbecco’s phosphate buffered saline (PBS) and hexamethyldisilazane 98% (HMDS) were purchased from Sigma–Aldrich. 3-Aminopropyl trimethoxysilane (APTMS) and 1H, 1H, 2H, 2H-perfluorooctyl trichlorosilane (FOTS) were obtained from ABCR GmbH. A Sylgard 184 poly(dimethyl siloxane) (PDMS) was purchased from Dow Corning. Irgacure 369 was obtained from CIBA. Alexa Fluor 488-conjugated BSA (BSA Alexa Fluor 488), fibronectin (FN) were provided by Fisher Scientific SL. Hilyte 488-conjugated fibronectin (FN 488) was purchased from Cytoskeleton. Ultrapure water was used for all dilutions and washings and all reagents were of analytical grade. 2.1. Fabrication of silicon masters Silicon masters, for PDMS stamps fabrication, with microstructures consisting of grooves from 4 to 200 lm of periodicity were fabricated by UV photolithography and deep reactive ion etching (DRIE). Four inch silicon wafers were first dehydrated at 180 °C for 30 min in a convection oven. A first layer of an adhesion promoter, HMDS, and a positive photoresist (maP-1205) were spin coated and processed using a conventional UV-Lithography process. These patterns were transferred to the silicon by DRIE (Oxford Plasmalab System 80) using a combined SF6/C4F8 plasma. Silicon stamps with two different depths (240 nm for micro-structuration studies and 5 lm for lCP experiments) were obtained depending on the etching time. Finally, FOTS was evaporated over the silicon stamp in a vacuum desiccator during 30 min as anti-adhesive to facilitate the PDMS stamp peeling off. 2.2. Fabrication of PDMS stamps PDMS stamps were fabricated using the silicon masters described in the previous section. To cure the PDMS prepolymer, a mixture of 10:1 silicone elastomer and the curing agent was cast over silicon masters and placed at 60 °C overnight. PDMS was then
carefully peeled off from the silicon master and cut into 10 mm 10 mm stamps. The protruding ridges in the masters resulted in PDMS replicas with grooves. Finally, the stamps were cleaned with 70% ethanol, sonicated for 5 min and let dry prior to use. 2.3. Substrate preparation 2.3.1. Glass substrates Four inch glass wafers were diced into 10 mm 10 mm substrates and cleaned prior to use by copious rinsing with acetone, isopropyl alcohol and ultrapure water. Glass substrates were methacrylated after 5 min O2 plasma treatment by deposition of a drop of TMSPM. The substrates were kept in an oven at 100 °C for 1 h followed by 15 min at 110 °C [2]. Substrates were rinsed with ultrapure water and dried under a N2 flow. Amination of glass substrates for subsequent ox-HA and ox-HAS deposition was performed by evaporation of APTMS. 2.3.2. Gold substrates Another set of glass wafers was cleaned as described in Section 2.3.1. Thereafter, a 40 nm thick gold layer was deposited onto the wafers by electron beam physical vapour deposition (ATC Orion series UHV Evaporation system, AJA International Inc.). A previous evaporation step of a 3 nm thick layer of titanium was carried out to improve adhesion to the glass. These wafers were also diced into 10 mm 10 mm substrates. These substrates were cleaned using a piranha solution (H2SO4:H2O2, 3:1) for 10 min, rinsed with ethanol and dried using N2. Subsequently, substrates were immersed in a 20 lM solution of 11-amino-1-undecanethiol hydrochloride in pure ethanol for 48 h. After amination, substrates were washed with ethanol and dried under a N2 flow to be ready for GAG immobilization. 2.4. Microstructuring of MAHA Two percent and 5% w/v MAHA solutions in water were filtered using 0.2 lm diameter porous filters (Millex, Millipore) and blended with 0.3% m/v Irgacure 369 in N-vinyl pyrrolidone and with TEGDM as an additional crosslinker using a MAHA:TEGDM 1:1 ratio. One hundred and thirty microlitres of these mixtures were deposited onto the microstructured PDMS stamps and the methacrylated glass substrates were carefully placed over them. Photopolymerization was performed under UV light from a 350 W short arc lamp (Advanced Radiation Corporation, 350– 450 nm, 10 mW/cm2) for 2 h under a N2 atmosphere. After UV exposure, the PDMS stamp was left overnight before demoulding to ensure the stabilization of the hydrogel microstructures (Fig. 1a). After demoulding of the stamps from the microstructured MAHA, a Veeco Dektak 8 (Veeco Instruments, Plainview, NY) mechanical profilometer was used for profile measurements. 2.5. Immobilization of ox-HA and ox-HAS GAGs over aminated gold/ glass substrates Ox-HA and ox-HAS were immobilized onto substrates as follows: each well of a 12-well plate containing a single aminated gold or glass substrate was filled with 1 ml of 4 mg/ml ox-HA or ox-HAS solution in PBS. The plates were kept shaking gently for 24 h. One millilitre of NaBH3CN (3 mg/ml in PBS) was added into each well and kept at 4 °C for another 24 h. After the ox-GAG immobilization, the coated surface was once more rinsed with water and dried under a N2 stream.
M.C. Márquez-Posadas et al. / Microelectronic Engineering 106 (2013) 21–26
23
Fig. 1. Schematic diagram of the surface modifications techniques performed over modified HA, (a) microstructuring of MAHA: MAHA is dropped over the PDMS stamp and the substrate is placed over it and exposure to UV light. After the PDMS stamp release microstructured hydrogel is obtained, (b) lCP of proteins over ox-HA/ox-HAS: ox-HA/ ox-HAS are immobilized covalently over an aminated gold/glass substrate. Protein solutions (BSA or FN) are incubated for 1 h over a PDMS stamp and afterwards the stamp is rinsed, dried and placed onto the ox-HA/ox-HAS substrates applying a gently pressure for getting a conformal contact for 1 min. After detachment of the PDMS stamp, a protein patterning is obtained.
2.6. lCP of proteins over immobilized ox-HA and ox-HAS GAGs
lCP of BSA Alexa Fluor 488 and a 1:100 mixture of FN 488 (fluorescent) and FN (non-fluorescent) was carried out over ox-HA and ox-HAS GAGs substrates. Hundred microlitres of 100 lg/ml BSA or FN were incubated over microstructured PDMS stamps for 1 h for surface protein absorption. Afterwards, the stamps were rinsed with water and dried under a N2 flow. The lCP was performed by placing the stamps over the substrates, applying a gentle pressure during 1 min to get a conformal contact with the surfaces and then carefully peeling them off (Fig. 1b). The stamps were examined after demoulding to check a possible deterioration which could imply a deterioration of subsequent patterns. Characterization of the patterns was performed with a NT-MDT AFM and an AXIO Imager.A1m microscope (Carl Zeiss) equipped with a CCD camera and adequate fluorescence filters.
3. Results and discussion 3.1. MAHA microstructuring HA is a cell surface-associated polysaccharide, ubiquitous in the ECM, which is known to have biochemical and mechanical effects on the cells [17]. Therefore, microstructures made of HA could offer an in vitro microenvironment resembling that found in the ECM. In the present study, we have microstructured a photocrosslinkable HA derivative, MAHA (Fig. 2). The crosslinking density and the polysaccharide concentration have been adjusted for optimal results. The solubility of the microstructured material has also been tested to ensure their ability to be used in cell-behaviour experiments.
As mentioned in Section 2.3.1, the glass substrates had to be methacrylated, prior to the microstructuring of MAHA, in order to enhance the adhesion of the MAHA. Methacrylate groups on the substrate react with those of the MAHA hydrogel substantially improving substrate adhesion. It was experimentally tested that the lack of this glass methacrylation step gave rise to a poor adhesion of the MAHA, causing detachment of the MAHA hydrogel layer from the glass substrates during the immersion step. Taking into account that, due to their low crosslinking density, non modified HA hydrogels are often readily degraded in aqueous solutions such as a cell culture medium [16], the use of an additional crosslinker, TEGDM, was tested in order to increase crosslinking density and, therefore the immersion resistance and mechanical stability of the resulting hydrogels. Finally, in order to minimize oxygen diffusion into the samples during polymerization, as molecular oxygen is a radical scavenger and a radical quencher that can readily terminate initiating and propagating radicals [18], UV light exposure was performed in a N2 atmosphere. Fig. 3 shows optical microscopy images of the microstructures present on the MAHA hydrogel surface. The replication quality was high and the homogeneity over the whole area was remarkable for substrates with and without additional crosslinker (see Fig. 3a and b, respectively). In order to characterize the filling of PDMS microstructures by the hydrogels, profile measurements of both the stamp and the microstructured hydrogel were performed and are shown in Fig. 4. In this figure, it can be clearly seen that the microstructures underwent an anisotropic shrinkage, maintaining the period size but showing a lower depth than the PDMS stamp. The depth difference was approximately 90 nm which could be attributed to the evaporation of water from the hydrogel (as the PDMS stamp is left over the hydrogel overnight before detachment). Another possible reason could be a not complete filling of
Fig. 2. Scheme of hyaluronan methacrylation reaction using glycidyl methacrylate.
24
M.C. Márquez-Posadas et al. / Microelectronic Engineering 106 (2013) 21–26
Fig. 3. Optical micrograph of microstructured MAHA after demoulding. (a) 4 lm periodic grooves of 2% w/v without the use of additional crosslinker and (b) 4 lm periodic grooves of 5% w/v with additional crosslinker.
be convenient for microstructuring of polysaccharides (such as HA) due to their high hydrophilicity [11], the use of stamps without this plasma treatment has been successful in the present study. It has also been suggested that, due to their solubility in water, a coating with non-modified HA hydrogel would be washed away during immersion leading to a chemisorbed layer only [11]. However, probably due the use of an additional crosslinker, this phenomenon was not observed in this study. The results of the present study could pave the way to future cell-behaviour studies using microstructured substrates composed of MAHA.
3.2. lCP of BSA and FN over ox-HA and ox-HAS GAGs substrates
Fig. 4. Profile of the microstructured MAHA (black) obtained by mechanical profilometry. The PDMS stamp profile is also shown as comparison (dark grey). The microstructures have a period of 50 lm.
the PDMS stamp grooves. A similar behaviour has been previously described and attributed to the stresses created during the crosslinking process [19]. Microstructured MAHA glass substrates were immersed in 10 mM HEPES buffer solution for 24 h to test the material resistance under salt concentration or pH conditions similar to those found in cell-culture studies. Fig. 5 shows pictures of substrates during immersion. As can be seen, the surface of MAHA 5% w/v samples withstood the immersion (Fig. 5a). However, due to its lower concentration, in the case of MAHA 2% w/v samples, only some structured small areas remained (Fig. 5b). MAHA samples without additional crosslinker did not withstand the immersion test. These results show that an increase of the concentration causes an improvement of the resistance of the structured surfaces to immersion, although a compromise is necessary. Khademhosseini et al. [2] described that a high concentration of HA, although desirable, could affect negatively the moulding process because of the excessive viscosity of solutions with high concentrations of HA or HA derivatives. They established a 5% concentration as optimal. It is worth mentioning that although it can be found in literature that a previous oxygen plasma treatment of the stamps could
The degree of sulfation of GAGs affects cells adhesion, growth and differentiation [20–22]. In addition, sulphated GAGs bind a wide variety of ECM proteins and growth factors. Hence a combination of GAGs with proteins in spatial well-designed arrangements, like micro patterns to obtain adhesive and non-adhesive areas, could provide useful information on cell behaviour and means to control cell activity. Hyaluronan and its sulphated derivatives possess sugar rings with vicinal hydroxyl groups which were oxidized with sodium periodate to generate free aldehyde groups (see Fig. 6). More details of the procedure are described in a previous work [20]. FN and BSA are proteins, with well-known cell adhesive and non-adhesive properties, respectively, which can be printed onto a wide variety of substrates to obtain patterns suitable for cellculture studies [7]. Since ox-GAGs may also bind proteins covalently via cross-linking of GAG-aldehyde amino side-groups of proteins, either FN or BSA were transferred by lCP to the immobilized ox-GAG layers. In order to carry out lCP of proteins, gold and glass substrates had to be aminated using an amino-thiol or amino-silane reagent, so that the aldehydes of ox-HA and ox-HAS could react with the amino groups of substrates via Schiff’s base reaction. The addition of NaBH3CN solution reduced the intermediate imines to stable amines [20]. Subsequently proteins, like BSA or FN were immobilized on ox-HA and ox-HAS modified surfaces using lCP. After the lCP of fluorescent BSA and FN, the transferred patterns were characterized by fluorescence microscopy (see Fig. S1) and AFM. The binding of proteins was, on average, reduced on HA surfaces when compared to glass [11]. However, from the fluorescence microscopy images (Fig. 7) it can be seen that the obtained patterns of both proteins were well defined on the ox-GAGs substrates.
M.C. Márquez-Posadas et al. / Microelectronic Engineering 106 (2013) 21–26
25
Fig. 5. Optical microscopy photograph of microstructured MAHA during immersion in HEPES 10 mM. (a) 4 and 50 lm periodic grooves of 5% w/v with additional crosslinker and (b) 4 lm periodic grooves of 2% w/v with additional crosslinker. The insets images show a detail with 4 lm periodic grooves.
Fig. 6. Chemical structures of, (a) oxidized hyaluronan (ox-HA) and (b) oxidized hyaluronan sulphate (ox-HAS).
Fig. 7. Fluorescence microscopy images of microcontact printed proteins. (a) BSA Alexa Fluor 488 (100 lg/ml concentration) patterns of 50 lm periodicity over immobilized ox-HAS. (b) FN Hilyte 488-conjugated (100 lg/ml concentration) patterns of 100 lm periodicity over immobilized ox-HA. The inset images show the same areas after 24 h immersion in PBS.
Fig. 7a shows a fluorescence image of the lCP over ox-HAS substrates which was performed with a concentration of 100 lg/ml of Alexa Fluor 488 conjugated BSA solution. Some fluorescent spots can be noticed between fluorescent patterns before PBS immersion, which could be caused by small drops of protein solution that remain in the stamp grooves after the drying process and before the lCP. After PBS immersion, a decrease of the fluorescence intensity can be appreciated. It might be attributed to a small loss of weakly bound proteins that after rinsing and drying were spread over the patterned surface (see inset in Fig. 7a and Fig. S1). Fig. 7b shows the lCP over ox-HA substrates which was carried out with a 100:1 ratio of FN and Hilyte 488-conjugated FN. The fainter-fluorescence of the FN pattern and the less homogeneous fluorescent patterning could be caused by the lower concentration of fluorophore (as mentioned in Section 2.6).
The AFM image (Fig. 8) shows the topography of the proteinprinted and non-printed areas. A 4–5 nm step was measured by AFM in semi-contact mode, which is consistent with the height of a protein layer [23]. The stability of the protein pattern was confirmed by the immersion of printed surfaces in PBS for 24 h. The effects of topographical modifications in cell-behaviour experiments are usually studied before reaching confluence, a process which takes place in approximately 24–48 h, depending on the cell type. The samples were rinsed with ultrapure water and dried with N2 to remove salt aggregates. The samples were again studied by fluorescence microscopy (Fig. 7 insets) finding that the homogeneity of the patterns was kept and also proving that the protein adsorption to oxHA and ox-HAS was strong enough to allow future cell-behaviour studies.
26
M.C. Márquez-Posadas et al. / Microelectronic Engineering 106 (2013) 21–26
Fig. 8. AFM image of a BSA pattern of 4 lm periodicity over ox-HAS. (a) Topography, (b) profile of the marked line shows BSA height around 4–5 nm.
4. Conclusions In this study we have described surface modifications (structuring and protein patterning) performed over modified novel HA derivatives (MAHA, ox-HA and ox-HAS) deposited on solid substrates such as gold and glass. Soft lithography and microstructuring approaches have been successfully used here to obtain the desired microstructures. The surface topography obtained for HA-based natural polysaccharides can be used to achieve in vitro conditions similar to those found in the ECM. The feasibility of using the obtained substrates in, for instance, future cell-behaviour experiments was checked by immersion resistance showing the stability of the immobilized GAG layers. Acknowledgments This work was supported and carried out under the scope of the European Project Find and Bind (NMP4-SL-2009-229292). Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.mee.2013.01.039. References [1] N.A. Peppas, J.Z. Hilt, A. Khademhosseini, R. Langer, Adv. Mater. 18 (2006) 1345–1360. [2] A. Khademhosseini, G. Eng, J. Yeh, J. Fukuda, J. Blumling III, R. Langer, J.A. Burdick, J. Biomed. Mater. Res. A 79 (3) (2006) 522–532. [3] N.S. Gandhi, R.L. Mancera, Chem. Biol. Drug Des. 72 (2008) 455–482.
[4] J.W. Kuo, Practical Aspects of Hyaluronan Based Medical Products, CRC/Taylor & Francis, Boca Raton, 2006. [5] L.E. Dickinson, C.C. Ho, G.M. Wang, K.J. Stebe, S. Gerecht, Biomaterials 31 (2010) 5472–5478. [6] Kenneth.E. Gonsalves, Craig.R. Halberstadt, Cato.T. Laurencin, Lakshmi.S. Nair, Biomedical Nanostructures, John Wiley & Sons, Hoboken, 2008. [7] L. Kam, S.G. Boxer, J. Biomed. Mater. Res. 55 (2001) 487–495. [8] D.H. Kim, P.P. Provenzano, C.L. Smith, A. Levchenko, J. Cell Biol. 197 (3) (2012) 351–360. [9] D. Guduru, M.S. Niepel, J. Vogel, T. Groth, Int. J. Artif. Organs 34 (10) (2011) 963–985. [10] M.S. Kim, A.Y. Kim, K.J. Jang, J.H. Kim, J.B. Kim, K.Y. Suh, Nanotechnology 22 (2011) 494017–494025. [11] K.Y. Suh, A. Khademhosseini, J.M. Yang, G. Eng, R. Langer, Adv. Mater. 16 (7) (2004) 584–588. [12] J.A. Burdick, C. Chung, X. Jia, M.A. Randolph, R. Langer, Biomacromolecules 6 (2005) 386–391. [13] J.B. Leach, C.E. Schmidt, Biomaterials 26 (2005) 125–135. [14] K.S. Masters, D.N. Shah, L.A. Leinwand, K.S. Anseth, Biomaterials 26 (2005) 2517–2525. [15] S.M. Ruiz, C.S. Chen, Soft Matter 3 (2007) 168–177. [16] S. Möller, J. Weisser, S. Bischoff, M. Schnabelrauch, Biomol. Eng. 24 (2007) 496–504. [17] R.A.A. Muzzarellia, F. Grecoa, A. Busilacchia, V. Sollazzo, A. Gigante, Carbohydr. Polym. 89 (2012) 723–739. [18] S.J. Bryant, K.S. Anseth, Photopolimerization of Hydrogel Scaffolds, in: P.X. Ma, J. Elisseeff (Eds.), Scaffolding in Tissue Engineering, Taylor and Francis Group, 2006, pp. 71–90. [19] S. Rodriguez Vilches, C. Séverac, C. Thibaut, L. Laplatine, C. Vieu, J. Fitremann, A.F. Mingotaud, P. Martinoty, D. Collin, Colloid Polym. Sci. 289 (2011) 1437– 1449. [20] A. Köwitsch, Y. Yang, N. Ma, J. Kuntsche, K. Mäder, T. Groth, Biotechnol. Appl. Biochem. 58 (2011) 376–389. [21] D. Peschel, K. Zhang, S. Fischer, T. Groth, Acta Biomater. 8 (2012) 183–193. [22] D.S. Costa, R.A. Pires, A.M. Frias, R.L. Reis, I. Pashkuleva, J. Mater. Chem. 22 (2012) 7172–7178. [23] H.D. Inerowicz, S. Howell, F.E. Regnier, R. Reifenberger, Langmuir 18 (2002) 5263–5268.