Tunable Hydrogel Thin Films from Reactive Synthetic Polymers as

Increasing the APM content in PAPMx from 10 to 100% led to apparent Young's moduli from 300 to 700 kPa while retaining sufficient anhydride groups to ...
0 downloads 3 Views 5MB Size
Download PDF
Article pubs.acs.org/Langmuir

Tunable Hydrogel Thin Films from Reactive Synthetic Polymers as Potential Two-Dimensional Cell Scaffolds Laurent J. Goujon,† Santosh Hariharan,‡ Bahareh Sayyar,† Nicholas A. D. Burke,† Emily D. Cranston,§ David W. Andrews,‡ and Harald D. H. Stöver*,† †

Department of Chemistry and Chemical Biology, McMaster University, 1280 Main Street West, Hamilton, Ontario L8S 4M1, Canada ‡ Biological Sciences, Sunnybrook Research Institute and Department of Biochemistry, University of Toronto, Toronto, Ontario M5S 1A8, Canada § Department of Chemical Engineering, McMaster University, Hamilton, Ontario L8S 4L7, Canada S Supporting Information *

ABSTRACT: This article describes the formation of crosslinked 10−200-nm-thick polymer hydrogel films by alternating the spin-coating of two mutually reactive polymers from organic solutions, followed by hydrolysis of the resulting multilayer film in aqueous buffer. Poly(methyl vinyl ether-altmaleic anhydride) (PMM) was deposited from acetonitrile solution, and poly(N-3-aminopropylmethacrylamide-co-N-2hydroxypropylmethacrylamide) (PAPMx, where x corresponds to the 3-aminopropylmethacrylamide content ranging from 10 to 100%) was deposited from methanol. Multilayer films were formed in up to 20 deposition cycles. The films cross-linked during formation by reaction between the amine groups of PAPMx and the anhydride groups of PMM. The resulting multilayer films were covalently postfunctionalized by exposure to fluoresceinamine, decylamine, D-glucamine, or fluorescently labeled PAPMx solutions prior to the hydrolysis of residual anhydride in aqueous PBS buffer. This allowed tuning the hydrophobicity of the film to give static water contact angles ranging from about 5 to 90°. Increasing the APM content in PAPMx from 10 to 100% led to apparent Young’s moduli from 300 to 700 kPa while retaining sufficient anhydride groups to allow postfunctionalization of the films. This allowed the resulting (PMM/PAPMx) multilayer films to be turned into adhesionpromoting or antifouling surfaces for C2C12 mouse myoblasts and MCF 10A premalignant human mammary epithelial cells.

■

INTRODUCTION 1,2

have been developed to covalently cross-link such PEMs by thermal16−18 or UV17,19,20 treatment or by using a chemical cross-linking agent.21,22 Such post-treatments also help to control the mechanical properties of the film, which is a crucial parameter in cell attachment and development.5,7 A complementary approach uses dip-coating of substrates with two mutually reactive polymers to form cross-linked multilayer films.23−29 “Click” chemistries have often been used for this purpose, such as the formation of triazole linkages between poly(acrylic acid) modified with either azide or alkyne groups.27 Reactions between polymeric nucleophiles and electrophiles can also be used for cross-linking, and polyamines bearing primary amine groups are a common type of reactive polymer employed in this kind of system. They have been used to form amide bonds between carbon nanotubes functionalized with branched polyethylenimine (PEI) and poly(methyl vinyl ether-

3,4

The surface chemistry, topography, and mechanical properties5−7 of biomaterials are critical to their use as cell supports and scaffolds. Stem cell research has a need for fully defined, scalable extracellular matrixes (ECMs) that encourage cell attachment while maintaining pluripotency8−10 and that may allow the development of quantitative differentiation pathways.11 Conversely, antifouling surfaces help prevent immune rejection of implanted biomaterials.12 Therefore, it is important to develop new synthetic hydrogels with controllable surface physicochemical and mechanical properties. Layer-by-layer (LbL) assembly is a versatile method of fabricating such synthetic ECMs. The alternating deposition of oppositely charged polyelectrolytes on a substrate by spincoating or related methods results in polyelectrolyte multilayer films (PEM) held together by electrostatic interactions.12,13 One challenge in this approach is that changes in temperature, pH, and ionic strength14,15 may weaken the ionic or hydrogen-bonding interactions within the PEM, possibly causing film deterioration. In response, approaches © 2015 American Chemical Society

Received: January 29, 2015 Revised: April 9, 2015 Published: April 23, 2015 5623

DOI: 10.1021/acs.langmuir.5b00376 Langmuir 2015, 31, 5623−5632

Article

Langmuir

Figure 1. Schematic overview of (PMM/PAPMx)n.5 film assembly and postfunctionalization.

alt-maleic anhydride) (PMM)25 or between poly(vinylamineco-N-vinylisobutyramide) and poly(acrylic acid).26 Similarly, Lynn’s group described systems combining branched PEI with poly(2-vinyl-4,4-dimethylazlactone) (PVDMA)28 or smallmolecule pentacrylate species (5-Ac).29 A notable feature in some of this work is that the multilayers were assembled with polymers dissolved in organic solvents rather than water. This was done for solubility reasons but also to preserve watersensitive groups such as anhydride or azlactone for subsequent reactions. One additional benefit of preparing LbL systems from reactive polymers is that residual reactive groups in the last layer may be postmodified to change the film surfaces.25,29,28 For instance, PEI/5-Ac29 and PEI/PVDMA30 multilayers have been postfunctionalized with D-glucamine or decylamine to obtain hydrophilic or hydrophobic surfaces that were shown to, respectively, prevent and promote mammalian cell adhesion and bacterial biofilm growth. The work presented here focused on thin, cross-linked PEMs by spin-assisted LbL deposition of a series of mutually reactive polymers that allow both postfunctionalization and control of the mechanical properties to prevent or promote cell attachment. The multilayer films are formed from PMM and poly(3-aminopropylmethacrylamide-co-2-hydroxypropylmethacrylamide) (P(APM-co-HPM), or PAPMx, where x corresponds to the APM content in the copolymer) in organic solvents. This approach utilizes the high reactivity of the anhydride groups of PMM with nucleophiles,31 both to drive the spontaneous crosslinking during LbL assembly and for postfunctionalization. Residual anhydride groups may be rapidly converted to carboxylate anions by hydrolysis in aqueous buffer, preventing any undesired covalent binding of biomolecules in subsequent cell experiments. This is, to our knowledge, the first example of the spin-assisted LbL deposition of cross-linked networks from mutually reactive organic-soluble polymers. Benefits include the rapid deposition of polymer and wash solutions and very smooth films. PMM is commercially available and has already been shown to support long-term self-renewal and proliferation of human pluripotent stem cells.8 As reviewed by Popescu et al.,32 PMM and related copolymers containing maleic anhydride groups

have been used in many biomedical applications, including directly as therapeutic agents, in controlled release systems, the preparation of conjugates or biomaterials, in dental applications, or as a support for bioactive molecules in tissue engineering. PAPMx with low APM content was first described by Kopeček and co-workers,33,34 where the APM units were further reacted to obtain functional versions of PHPM, a polymer with excellent biocompatibility. Our group has prepared analogous PAPMx copolymers with higher APM content (x = 10−100 mol %) as potential building blocks in cell-encapsulation systems.35 The current article focuses on (PMM/PAPMx) PEM films with controlled chemical and mechanical properties through postfunctionalization with amine-bearing small molecules, biomolecules, or polymers (Figure 1). The first part of this article describes the LbL assembly, thickness, roughness, and wettability of (PMM/PAPM10) multilayer films formed by spin-coating onto amino-silanized glass coverslips. Residual maleic anhydride groups were then used to postfunctionalize the (PMM/PAPM10) films with various amine-containing small molecules or polymers in an organic solvent or aqueous buffer. The ability to control the cross-link density and mechanical properties of the multilayer films by adjusting the APM content within the PAPMx copolymer was further demonstrated. Finally, (PMM/PAPM10) hydrogel films (native or postfunctionalized with decylamine and/or Dglucamine) were seeded with C2C12 mouse myoblast cells and MCF 10A human epithelial cells to show the capacity of the films to promote or prevent cell adhesion, according to the surface composition and hydrophobicity.

■

EXPERIMENTAL SECTION

N-(3-Aminopropyl)methacrylamide hydrochloride (APM, ≥98%) and N-(2-hydroxypropyl)methacrylamide (HPM) were obtained from Polysciences (Warrington, PA). 2,2′-Azobis(2-methylpropionamidine) dihydrochloride (AAPH, 97%), cysteamine hydrochloride, fluorescein isothiocyanate, isomer I (FITC, 90%), poly(methyl vinyl ether-altmaleic anhydride) (PMM, Mn ≈ 80 kg mol−1), 3-aminopropyltriethoxysilane (APTES, ≥98%), and fluoresceinamine isomer I (FA) were purchased from Sigma-Aldrich (Oakville, ON). Acetone (reagent 5624

DOI: 10.1021/acs.langmuir.5b00376 Langmuir 2015, 31, 5623−5632

Article

Langmuir

Postfunctionalization of (PMM/PAPM 10 ) 4.5 Films with PAPMx. (PMM/PAPM10)4.5 thin films were also postfunctionalized by the deposition of a single layer of methanolic PAPMx containing 10, 25, 50, 75, or 100 mol % APM, using the same process as described above for the deposition of a PAPM10 layer by spin-coating. Postfunctionalization of (PMM/PAPM10)5.5 Thin Films with PAPM10f in Aqueous Solution. Drops of PAPM10f dissolved at 1 mg·mL−1 in PBS (150 mM NaCl, 10 mM phosphate buffer, pH 7.4), were placed by means of a capillary tube onto (PMM/PAPM10)5.5 films supported on a glass coverslip. After 10 min, the films were washed 10 times with PBS and analyzed by fluorescence microscopy. Film Hydrolysis. (PMM/PAPM10)n.5 films, both as formed and postfunctionalized, were hydrolyzed by immersion in PBS (pH 7.4) at 37 °C for 3 days. The resulting hydrolyzed films were soaked twice in deionized water for 10 min to remove the buffer salts, rinsed with MeOH, and finally dried in a stream of air. Characterizations of PEM Films and Postfunctionalized Films. Ellipsometry. The thickness of dry (PMM/PAPMx)n films coated on silicon wafers was measured at four different positions across each sample surface by ellipsometry (Nanofilm ellipsometer with EP3 software, Accurion). Measurements were made at 10× magnification with a 658 nm laser using a 42° angle of incidence. Ellipsometry data were processed using multilayer modeling software (EP4, Accurion) to calculate the film thickness. Optical Profilometry. The thickness and the roughness (root mean square, rms) of dry films coated onto aminated glass coverslips were measured by optical profilometry (New View 500 profiler with MetroPro 8.1.5 software, Zygo Instruments). Films were scratched with a metal hook (0.50 mm) from a microscopist microtool set (Ted Pella) in order to measure the step height between the film surface and the substrate. Measurements were performed using a 10× magnification objective. The film roughness was typically calculated for 200 μm × 200 μm regions on the film surfaces. Atomic Force Microscopy Measurements (AFM). AFM measurements were carried out with a BioScope Catalyst BioAFM (Bruker) equipped with silicon probes having nominal spring constants of 0.06 to 0.35 N/m (SNL-10, Bruker). Bruker’s PeakForce QNM imaging mode was used to measure the thickness and the roughness of dry or swollen films (immersed in PBS) over scan areas of 20 μm × 20 or 100 μm × 100 μm (scan rate, 1 Hz; 256 lines with 256 samples/line). Films were scratched with a metal hook to allow thickness measurements. AFM data were analyzed by NanoScope Analysis software (Bruker), and the rms roughness was typically calculated for 10 μm × 10 μm surface areas. To determine the Young’s moduli of the films immersed in PBS at 20°, force curves were obtained using the ramp experiment (500 nm· s−1 tip velocity perpendicular to the film surface, with a total force-pull distance of 400 nm) in PeakForce QNM mode in fluid. The Young’s modulus was calculated by the Sneddon model for a conical probe38,39

grade), toluene (reagent grade), methanol (MeOH, reagent grade), acetonitrile (MeCN, HPLC grade), N,N-dimethylformamide (DMF, 99.8%), and sodium chloride (reagent grade) were purchased from Caledon Laboratories Ltd. (Caledon, ON), and ethanol (EtOH, 95%) was purchased from Green Field Ethanol Inc. Decylamine (98%) and D-glucamine (≥95%) were purchased from TCI America (Portland, OR). Sodium hydroxide was obtained from EMD Canada, and sodium dihydrogen orthophosphate was from BDH. All chemicals were used as received. Synthesis of PAPMx and Fluorescent Labeling of PAPM10 (PAPM10f). PAPMx with 10, 25, 50, 75, and 100 mol % APM had been prepared previously in our laboratory35 using a procedure similar to that described by Kopecek.33,34 PAPMx and PAPM10f were prepared again during the course of the current project, using the same procedure which is given in the Supporting Information. Characterization data for all copolymers are shown in Table S1. Aminosilanization of Substrate Surfaces by 3-Aminopropyltriethoxysilane (APTES). Round glass coverslips (12 and 25 mm diameter) and square silicon wafers (1 cm × 1 cm) were cleaned by sonication in ethanol and acetone (10 min each) and dried in a stream of air. The substrates then were immersed in an APTES solution in toluene (1 vol %) for 1 h and rinsed by dipping into two successive baths of toluene and one bath of ethanol. After the removal of residual ethanol in an air stream, the substrates were dried in an oven at 100 °C for 1 h and stored in multiwell plates at room temperature for up to several weeks until use. Polyelectrolyte Multilayer (PMM/PAPMx)n Assembly by LbL Spin-Coating. PMM and PAPMx were dissolved in MeCN and MeOH, respectively, to obtain 1, 5, or 10 mg·mL−1 stock solutions. Before use, the solutions were shaken for 30 min and filtered with a syringe filter (Chromafil PET-20/15 MS, pore size 0.2 μm, from Macherey-Nagel). After the aminated substrate was placed in a stream of air to remove any dust particles, it was mounted in the spin coater (model EC101, Headway Research Inc.). The substrate was rinsed with MeCN and then MeOH by placing a few drops of the solvent onto the stationary substrate, followed by spinning at 2000 rpm for 25 s. A (PMM/PAPM10) bilayer was typically formed by the following four-step process: (1) 30 μL of PMM solution was placed on the glass coverslip surface before spinning at 2000 rpm for 25 s. (2) Excess PMM was removed by adding 30 μL of MeCN to cover the substrate before spinning as described in step 1. (3) The step 1 procedure was used, except with 30 μL of PAPM10 solution. (4) The step 2 procedure was used, except with MeOH. This cycle was repeated n times until the desired bilayer number was reached. In some cases, an additional PMM layer was deposited to yield (PMM/PAPMx)n.5 films containing PMM as the final layer. During film fabrication, the spin-coater enclosure was flushed with a 1 L·min−1 flow of dry nitrogen to avoid PMM hydrolysis and atmospheric water condensation on the films during drying that could otherwise lead to a modification of the surface topography.36,37 Before characterization, the dry (PMM/PAPMx)n films were stored under vacuum to avoid the hydrolysis of PMM by atmospheric moisture. Postfunctionalization of (PMM/PAPM10)5.5 Thin Films with Amines. (PMM/PAPM10)5.5 films deposited on glass coverslips were postfunctionalized by spin-coating with primary amines (FA, decylamine, D-glucamine, or decylamine/D-glucamine mixtures) immediately after film formation: (1) 30 μL of the amine solution (11.5 mM in MeOH) was placed on the film before spinning at 2000 rpm for 25 s. (2) Excess amine was removed by the addition of about 0.5−1 mL of MeOH before and during spinning at 2000 rpm for 25 s. This twostage process was repeated for up to 20 cycles. The postfunctionalized films were then soaked briefly in MeOH and dried in a stream of air. Alternatively, (PMM/PAPM10)5.5 films were immersed in 11.5 mM amine solutions in MeOH for 5 min to 24 h, followed by soaking in two separate MeOH baths and drying in a stream of air. The postfunctionalized films were stored under vacuum to avoid the hydrolysis of residual PMM by atmospheric moisture.

F=

2 E tan(α) δ2 π 1 − υ2

where F is the applied loading force, α is the semiopening angle of the cone-shaped tip, E is the Young’s modulus, υ is Poisson’s ratio, and δ is the tip indentation depth. Poisson’s ratio was assumed to be 0.5 for these soft materials. The spring constants (k) were experimentally determined for each cantilever via Bruker’s thermal tuning procedure. To avoid irreversible deformation of the films and undue influence of the hard substrate during AFM force measurements, the indentation depth was limited to 10 to 20% of the sample thickness. Therefore, the indentation depth was close to 10 nm, which approaches the sensitivity limit of the AFM. The Young’s modulus values determined are thus described as apparent in recognition of the extremely thin films tested and the limitations of the method; the moduli are considered to be approximate and within the correct order of magnitude. The Young’s moduli of the samples were corrected using a correction factor determined by measuring the Young’s modulus of a PDMS standard (PDMS-soft-1-12M, track 0 from Bruker with a nominal Young’s modulus at 2.5 ± 0.7 MPa) under the same experimental conditions 5625

DOI: 10.1021/acs.langmuir.5b00376 Langmuir 2015, 31, 5623−5632

Article

Langmuir fcor =

EPDMSexp

links were not detectable and were presumably masked by the much stronger amide signals arising from the PAPM backbone. The reaction of the anhydride groups with MeOH was tested in a control experiment: 5 mol equiv of MeOH (wrt anhydride groups) was added to 5.6% (w/v) PMM (0.36 M anhydride) in MeCN-d3, and the methanolysis was followed by 1H NMR. Only about 10% of the anhydride groups had been consumed after 24 h at room temperature, and thus the brief exposure of PMM to MeOH during the spin-coating process should not cause a significant loss of anhydride groups. In a bulk control experiment designed to show covalent cross-linking, mixing a PMM solution (10−20 wt % in MeCN) with a PAPM10 solution (10−20% in MeOH) led to the rapid formation of bulk, cross-linked gels whereas mixtures of corresponding solutions of PMM and PAPM0 (poly-2-hydroxypropylmethacrylamide) remained liquid even after several days. Although relative amounts of both polymers deposited were not determined, the high anhydride content and MW of the PMM suggests that most films would carry excess anhydride groups available for postfunctionalization. Given the typically incomplete nature of the polymer−polymer reaction, this would include even films incorporating PAPM100. The spin-coating process was repeated n or n.5 times to obtain multilayer films with PAPMx or PMM, respectively, as the last layer. Film growth was studied for (PMM/PAPM10)n.5 dry films using optical profilometry, and in many cases, the results were confirmed by ellipsometry and/or AFM measurements. The film thickness increased linearly with the bilayer number, e.g., from 7 ± 2 to 130 ± 10 nm when the number of bilayers increased from 1.5 to 20.5 for films made with 1 mg· mL−1 solution (Figure 2). A linear increase in film thickness is

EPDMSnominal

where fcor is the correction factor, EPDMSexp is the experimentally determined Young’s modulus of the PDMS standard, and EPDMSnominal is the Young’s modulus given for the PDMS standard. Fluorescence Microscopy Imaging. Film morphologies were examined using fluorescence microscopy (Olympus BX51 optical microscope fitted with a Q-Imaging Retiga EXi digital camera and ImagePro software) with a fluorescein filter cube (λex 470−490 nm, λem >520 nm). Fluorescence intensities were measured using ImageJ software. Film Stability in Solution. Glass coverslips bearing (PMM/ PAPM10f)5.5 films made from 5 mg·mL−1 polymer solutions were immersed in 10 mL of PBS (pH 7.4), 2 M NaCl, or 0.1 M NaOH in 20 mL glass vials maintained at 37 °C. The films were examined by fluorescence microscopy at different times. Static Water Contact Angle Measurements. The contact angle measurements were performed using a Krüss contact angle instrument running Drop Shape Analysis (DSA) 1.80.0.2 software. Typically, a 1.5 μL water drop was placed on the film surface, and the average static contact angle was determined by the software during the first 30 s following the drop deposition. For each sample, eight measurements were obtained on different areas, and two samples were characterized for each film composition. Degree of Swelling. The degree of swelling of films in PBS at 20 °C was defined as

SD =

ts − td × 100 td

where ts and td are the film thicknesses in the swollen state and in the dry state, respectively. Because the films are extremely thin and covalently bonded to the glass coverslips, swelling is limited to one dimension, with no observed wrinkling. Typically, three to five measurements each were made on two to four samples from different sets, using standard deviations for the error bars. Characterization of Cell Adhesion on Film-Coated Coverslips. Cell adhesion was investigated on unmodified (PMM/PAPM10)5 and (PMM/PAPM10)5.5 films and on (PMM/PAPM10)5.5 films postfunctionalized with decylamine, D-glucamine, and decylamine/Dglucamine mixtures. The multilayer films were formed using 5 mg· mL−1 polymer solutions, and the postfunctionalizations were carried out by spin coating 5 aliquots of the corresponding amine solution. The films were then hydrolyzed in PBS for 3 days at 37 °C before use. C2C12 mouse myoblast cells and human mammary epithelial cell line MCF 10A were seeded and cultured in specific growth mediums as reported in the Supporting Information.

■

Figure 2. Thickness of dry (PMM/PAPM10)n.5 multilayer films as a function of the bilayer number and the polymer concentrations (10 mg·mL−1, squares; 5 mg·mL−1, diamonds; and 1 mg·mL−1, triangles). The thickness was measured by optical profilometry.

RESULTS AND DISCUSSION Film Fabrication and Growth. (PMM/PAPMx)n films were formed on aminated glass coverslips or silicon wafers using an LbL assembly approach. Amine groups were introduced onto the surface of the silicon or glass substrates by treatment with APTES, which allowed covalent bonding between the substrates and the first PMM layer and prevented film delamination. The films were assembled by alternatively spin-coating solutions of PMM in MeCN and PAPMx in MeOH (Figure 1). The covalent cross-linking reaction between successive layers is assumed to be nearly instantaneous, aided by the rapid evaporation of solvent during spin-coating. After each deposition, the excess polymer was removed by rinsing with the corresponding solvent. The use of organic solvents limits reactions of the anhydride groups of PMM to crosslinking with PAPMx during the assembly. FTIR and ATR-FTIR were used to try to directly track the formation of amide crosslinks in the formed films, but the signals from the amide cross-

usually seen where there is limited diffusion of the newly deposited polymer into the previously deposited material. In contrast, exponential growth is observed when the film can swell with the newly introduced polymer solution and allow polymer diffusion throughout the film.12 In this study, each polymer solvent is a poor solvent for the other polymer, which, together with the cross-linking, limits interdiffusion and results in linear film growth. The dry film thickness increases by 5−15 nm per bilayer, with higher polymer concentrations leading to thicker films but also to higher variability, particularly when a polymer concentration of 10 mg·mL−1 was used. For 5.5 bilayer films, the use of 1, 5, and 10 mg·mL−1 solutions gave films with thicknesses of 35 ± 7, 60 ± 10, and 85 ± 25 nm, respectively. 5626

DOI: 10.1021/acs.langmuir.5b00376 Langmuir 2015, 31, 5623−5632

Article

Langmuir

PBS at 37 °C. After 85 days, films were scratched to confirm their presence on the substrate. One sample was immersed in 2 M NaCl to confirm that the film integrity was due to covalent cross-linking because this ionic strength is high enough to break electrostatic interactions (Figure 4). Films made with 1 mg· mL−1 solutions and with a higher bilayer number showed similar results (data not shown). In a harsher test of film stability and cross-linking, the films were exposed to 0.1 M NaOH (final solution pH 12.7). At this pH, the PAPM is uncharged whereas PMM is highly negatively charged, and thus the PEM films would dissolve if not covalently cross-linked. The films formed on amine-functionalized glass wrinkled quickly (Figure 4), delaminated from the substrate after 4 h, and broke into small pieces after 10 days because of charge repulsion within the polyanionic film. However, the ability of the film to survive the high-pH treatment, albeit in the form of small fragments, confirmed that it was covalently cross-linked.40 Films formed on unmodified glass or silicon delaminated from the substrate in PBS and 2 M NaCl solutions but remained otherwise intact. In 0.1 M NaOH solutions, the films delaminated and again broke into small pieces. Films made with PAPM10 have a ca. 10-fold excess of anhydrides over amines and will hence be highly anionic once hydrolyzed. Amine functionalization of glass or silicon substrates is important for keeping these PMM/PAPMx films attached, in particular for PAPMx with low APM content. The deposition of an initial layer of a highly charged polycation such as PAPM100 or poly(ethylenimine) may also serve to keep the hydrogel attached to the substrate, at least in PBS. Film polarity was studied by measuring static water contact angles (Table S2). Before hydrolysis, films with PAPM10 as the top layer ((PMM/PAPM10)5) were more hydrophilic than those with PMM as the top layer ((PMM/PAPM10)5.5), with static water contact angles of 25 ± 9 and 56 ± 4°, respectively. After exposure to PBS and drying in air, the static water contact angles on all films were below 10° due to the presence of the COOH/COO− groups derived from anhydride hydrolysis. These results demonstrate that LbL spin-coating of PMM and PAPM10 from organic solvents can rapidly generate smooth, stable, covalently cross-linked hydrogel films of 100− 200 nm thickness with as few as five bilayer depositions. Film Postfunctionalization. Control of surface physicochemical properties is key in 2D-ECM development for cell studies, including those aimed at maintaining stem cell pluripotency8−10 and the development of quantitative differentiation pathways.11 Below we describe the chemical postmodification of (PMM/PAPM10)5.5 films with methanolic solutions of decylamine (hydrophobic), D-glucamine (hydrophilic), and their mixtures, as previously explored by Lynn’s group.29,30 The amine depositions involved one or more spincoating cycles, following a small model study comparing dipcoating and spin-coating (Supporting Information and Figure 5). Starting from a contact angle of 55° for (PMM/PAPM10)5.5 films, five or more D-glucamine spin-coatings were needed to decrease the contact angle to 5−10° while only one decylamine spin coating deposition was needed to increase the static water contact angle to 95°. The need of a higher deposition number for the surface functionalization with D-glucamine than with decylamine can be explained by a lower D-glucamine reactivity due to a higher steric hindrance effecting the PMM amidation.31 With the exception of the unfunctionalized films and those with a single D-glucamine deposition, subsequent

The root-mean-square roughness (rms) of the (PMM/ PAPM10)n.5 films was also measured by optical profilometry and was found to be independent of both the bilayer number and the polymer solution concentration (Figure S1). Similar roughness values were determined by AFM. The roughness of the dry films (3 ± 2 nm) remains in the same size range as that of the aminated coverslips, indicating that smooth films were obtained. Fluorescence microscopy of (PMM/PAPM10f)5.5 films made with fluorescently labeled PAPM10f at 5 mg·mL−1 showed a uniform fluorescence distribution across the whole film surface at both low and high resolution (Figure S2). This further confirms that the PAPM10f is homogeneously distributed within the films, at least down to micrometer scales. The film was scratched to prove that the fluorescence was due to the film. Similar observations were obtained for (PMM/PAPM10f)n.5 films with bilayer numbers of up to 20.5 and for films made from 1 or 10 mg·mL−1 polymer solutions. Film Swelling, Hydrolysis, and Stability in Aqueous Solutions. The spin-coating process produces dry multilayer films. When the films were immersed in PBS buffer (pH 7.4) at 37 °C, they became hydrated and swelled, resulting in a roughly 3-fold increase in film thickness. Figure 3 shows AFM images of

Figure 3. Three-dimensionally rendered AFM height image of a scratched (PMM/PAPM10)5.5 film immersed in PBS (after 3 days of swelling at 37 °C). Films made with solutions of 1 mg·mL−1 (a) or 5 mg·mL−1 (b).

a (PMM/PAPM10)5.5 film made with 1 or 5 mg·mL−1 polymer solutions after 3 days of swelling in PBS, showing that the film thickness had increased from 31 ± 2 to 105 ± 4 nm and from 60 ± 10 to 185 ± 10 nm, respectively. Swelling also slightly increased the rms for these films from 1.5 ± 0.5 to 3.6 ± 2 nm (1 mg·mL−1) and from 1.5 ± 1 to 4 ± 2 nm (5 mg·mL−1). The absence of wrinkling in the swollen films indicates that the covalent attachment to the aminated support prevents lateral swelling. Film stability was monitored with fluorescence microscopy, for (PMM/PAPM10f)5.5 films assembled from 5 mg·mL−1 solutions on amine-functionalized glass or silicon (Figure 4). The films retained their integrity during storage for 85 days in 5627

DOI: 10.1021/acs.langmuir.5b00376 Langmuir 2015, 31, 5623−5632

Article

Langmuir

Figure 4. Fluorescence microscopy images of (PMM/PAPM10)5.5 films, from 5 mg·mL−1 solutions, immersed at 37 °C in PBS, NaCl, or NaOH solutions after 30 min or 1 h 30 min and after 86 days. *The brightness of the images taken after 86 days was increased by 50% because of the fluorescein photobleaching over time.

Figure 5. Evolution of static water contact angles on dry (PMM/PAPM10)5.5 films, from 5 mg·mL−1 solutions, as a function of the number of depositions by spin-coating decylamine (squares) or D-glucamine (triangles), before (solid marker) and after (hollow marker) hydrolysis. Solid and hollow diamonds on the y axis are for the unfunctionalized film before and after hydrolysis. Pictures show a water drop on an unfunctionalized film (left) and on films after five decylamine depositions (top right) or five D-glucamine depositions (bottom right) before hydrolysis.

exposure of these films to PBS for 3 days at 37 °C did not affect the contact angles, in agreement with efficient attachment of the modifiers to a cross-linked network that prevented reorientation of the film surface. The roughness of functionalized samples (RMS = 4 ± 3 nm for both dry and PBS swollen films) was similar to that of nonfunctionalized surfaces. The hydrophobicity of the film surface could be more finely controlled by postfunctionalization with decylamine/D-glucamine mixtures. The static water contact angle, before and after hydrolysis, remained close to 100° for mixtures with 0 to 50% D-glucamine (Figure 6) and then decreased progessively to about 10° as the D-glucamine content increased through 75 to 100%. Another way to control the surface properties of the multilayer films is to change the polymer used for the last layer. Here, a series of (PMM/PAPM10)4.5 films were terminated by spin-coating an additional PAPMx layer containing varying amounts of APM. The resulting films were hydrolyzed in PBS for 3 days, rinsed with water, and dried before measuring the static water contact angle. The static

Figure 6. Evolution of the static water contact angle of (PMM/ PAPM10)5.5 films postfunctionalized using five spin-coating/wash cycles involving 5 mg·mL−1 amine solutions in MeOH, as a function of the D-glucamine percentage in mixed decylamine/D-glucamine solution, before (full diamond) and after (hollow diamond) hydrolysis for 3 days in PBS.

5628

DOI: 10.1021/acs.langmuir.5b00376 Langmuir 2015, 31, 5623−5632

Article

Langmuir water contact angle was found to increase with the APM content in the PAPMx top layer (Figure 7). This increase in

Figure 8. (PMM/PAPMx)5.5 degree of swelling (hollow diamonds, left vertical scale) and apparent Young’s modulus (hollow squares, right vertical scale) in PBS at 20 °C according to the APM content (mol %) in PAPMx for films made with 1 mg·mL−1 solutions.

Figure 7. Evolution of the static water contact angle (diamonds, left vertical scale) and the rms roughness (squares, right vertical scale) of (PMM/PAPM10)4.5 + PAPMx according to the APM% in the top layer (hydrolyzed films, formed from 1 mg·mL−1 solutions).

the PAPMx (Figure 8). Picart et al. showed recently that C2C12 myoblast cells showed better adhesion and differentiation on poly(L-lysine)/hyaluronan PEMs with a Young’s modulus above 320 kPa, whereas softer films were not favorable for cell anchoring, spreading, or proliferation.41 Measuring the mechanical properties of such thin films by AFM is near the limit of the technique, and the apparent Young’s modulus values determined in the present work should best be taken as indicating a trend of increasing modulus with increasing crosslink density. Cell Studies. The adhesion of C2C12 mouse myoblast cells on various unmodified and postfunctionalized (PMM/ PAPM10)n films was studied because these cells are known to be sensitive to the mechanical properties40 and surface chemistry of the substrate.2 Previous research in our group had shown that polymer hydrogel capsules formed from PMM and poly-L-lysine (15−30 kDa) were very compatible with both encapsulated C2C12 cells.42 We also recently used MTT assays to show the cytocompatibility of C2C12 cells with PMM/ PAPMx combinations.43 Films differing in the composition of the last layer [PAPM10 (n = 5) and PMM (n = 5.5)] and those that had been postfunctionalized [D-glucamine (n = 5.5), decylamine/Dglucamine (25/75) (n = 5.5), and decylamine (n = 5.5)] were investigated. All films were made using 5 mg·mL−1 polymer solutions and were hydrolyzed after formation. After 48 h of incubation in cell media (DMEM + 10 vol % FBS) and C2C12 cells, there was poor (less than 20% coverage) cell attachment on the unmodified films, whether terminated by PAPM10 (Figure 9b) or PMM (Figure 9c), probably due to the hydrophilic nature (contact angles