Terminal-Functionality Effect of Poly(N‑isopropylacrylamide) Brush Surfaces on Temperature-Controlled Cell Adhesion/Detachment Naoki Matsuzaka,†,‡,§ Masamichi Nakayama,‡ Hironobu Takahashi,‡ Masayuki Yamato,‡ Akihiko Kikuchi,† and Teruo Okano*,‡ †

Department of Materials Science and Technology, Graduate School of Industrial Science and Technology, Tokyo University of Science, 6-3-1 Niijuku, Katsushika, Tokyo 125-8585, Japan ‡ Institute of Advanced Biomedical Engineering and Science, Tokyo Women’s Medical University (TWIns), 8-1 Kawada-cho, Shinjuku, Tokyo 162-8666, Japan § Research Fellow, Japan Society for the Promotion of Science (JSPS), Tokyo, Japan S Supporting Information *

ABSTRACT: Terminally functionalized poly(N-isopropylacrylamide) (PIPAAm) brush grafted glass surfaces were prepared by a surface-initiated reversible addition-fragmentation chain transfer radical (SI-RAFT) polymerization. SIRAFT mediated PIPAAm chains possessed terminal dodecyl trithiocarbonate groups which can be substituted with various functional groups. In this study, dodecyl groups were substituted with hydrophilic maleimide groups for controlling the thermoresponsive character of PIPAAm brushes. PIPAAm brushes exhibited reversible temperature-dependent surface wettability changes around PIPAAm’s lower critical solution temperature. Phase transition of dodecyl-terminated PIPAAm brushes clearly shifted to lower temperature than that of maleimide-terminated PIPAAm brushes, and this shift was attributed to promoted PIPAAm dehydration via terminal hydrophobes. By using this feature, the specific adhesion temperatures of bovine carotid artery endothelial cells (BAECs) on the PIPAAm brush surfaces were successfully controlled. BAECs were initiated to adhere on dodecyl-PIPAAm surfaces at 31 °C, while their adhesion was significantly suppressed on maleimide-PIPAAm surfaces under 33 °C. In contrast, terminal functionality scarcely affected the thermoresponsive behavior of PIPAAm brushes in the polymer rehydration process by reducing temperatures, and thus, the difference in spontaneous cell detachment from different PIPAAm-brush surface was negligible. Consequently, confluently cultured cells were able to be harvested as contiguous cell sheets from individual surfaces with comparable periods at 20 °C.

group has developed a unique concept, called “cell sheet engineering”, as a new approaching strategy for regenerative medicine using thermoresponsive PIPAAm-grafted surfaces.10 Tissue-like cellular monolayers, “cell sheets”, are harvested from confluently cultured cells on thermoresponsive culture surfaces by reducing temperatures.11 The intact cellular architectures maintain their associated extracellular matrix (ECM) and native-tissue functions due to the nonuse of digestive enzyme,12 and therefore, cell sheets can be transplanted to damaged host tissues without using conventional biodegradable scaffolds. Cell sheet-based human clinical applications have already been initiated for reconstructing various organs such as the cornea, heart, and periodontal ligament, and several clinical approaches are now in progress.10


Over the past three decades, stimuli-responsive polymers have attracted great attention to create intelligent materials, which possess on−off switching sequences on the molecular level in response to external stimuli.1,2 Especially, special attention is paid to thermoresponsive materials using poly(N-isopropylacrylamide) (PIPAAm) for various biomedical applications such as controlled drug delivery,2,3 enzyme bioconjugation,4 and bioseparation.5,6 PIPAAm fully hydrates and becomes watersoluble at a specific temperature below its lower critical solution temperature (LCST) (approximately 32 °C in water), while the polymer shows a thermal coil-to-globule phase transition to hydrophobic aggregates above LCST.7 Therefore, the introduction of PIPAAm molecules gives thermoresponsive surface hydrophobic/hydrophilic properties around typical physiological temperature onto material interfaces. This alteration of PIPAAm-grafted surfaces can be used to modulate their interactions with biomolecules in aqueous chromatographic separation8 and with cells.9 In the recent decade, our research © 2013 American Chemical Society

Received: May 31, 2013 Revised: July 30, 2013 Published: August 2, 2013 3164 | Biomacromolecules 2013, 14, 3164−3171



Health Science Research Resources Bank (Osaka, Japan). Glass coverslips (size, 24 × 50 mm; thickness, 0.2 mm) were purchased from Matsunami Glass (Osaka, Japan). Glass beads (the average diameter, 50 μm) were obtained from Toshinriko (Tokyo, Japan). Water used in this study was purified by a Milli-Q A10 (Millipore, Billerica, MA), unless otherwise mentioned. Preparation of PIPAAm Brush Surfaces. Glass coverslips were cleaned by oxygen plasma in a plasma dry cleaner PX-1000 (SAMCO, Kyoto, Japan). In this study, all reactions and polymerizations were performed under a nitrogen atmosphere. A silane coupling agent, APTES, reacted with the surface of glass coverslips through silane coupling reaction (Scheme 1), as previously reported.33 4-Cyano-4-

Thermoresponsive culture surfaces are conventionally prepared with electron beam-induced and grafted PIPAAm for constructing nanoscale thermoresponsive polymer layers with a cross-linked structure.13,14 In the recent years, newly designed intelligent surfaces grafted with densely packed linear PIPAAm chains (PIPAAm brushes) are prepared by various surface-initiated living radical polymerization methods.15−17 Furthermore, the grafting density and molecular weight of PIPAAm chain are reported to be key factors for precisely controlled cell adhesion/detachment via a temperature switch.16,18 Interestingly, the terminal functionalities of PIPAAm brushes significantly affect cell adhesion and subsequent cell sheet fabrication, even though the molecular weights of grafted PIPAAm chains are equivalent. For example, negatively charged carboxyl-terminated PIPAAm brush surfaces demonstrated a good performance in the simultaneous cell culture and cell sheet harvest of human smooth muscle cells.19 Reversible addition-fragmentation chain transfer radical (RAFT) polymerization is one of valuable methods for fabricating terminally functionalized PIPAAm with controlled molecular weights.20−23 RAFT-mediated polymers can be substituted with various functional groups through the reduction of chain transfer agent (CTA)-derived terminal groups and subsequent coupling reaction.24−29 By using the unique function, our previous works proposed the unexpected thermal phase transitions of PIPAAm derivatives by varying terminal hydrophobic/hydrophilic properties.24,30 Hydrophobically terminated PIPAAm exhibits a lower LCST shift compared with native PIPAAm’s LCST, because the freely mobile terminal hydrophobe promotes the dehydration of proximal IPAAm units and further disrupt the polymer hydration.31,32 Moreover, close-packed thermoresponsive polymer structures (e.g., the outer coronas of polymeric micelles) enhance the peculiar LCST effect due to the hydrophobic cluster formation of concentrated terminal hydrophobes.24 This study focused on the preparation of PIPAAm brushes having terminal hydrophobic/hydrophilic functionalities on glass substrates by surface-initiated RAFT (SI-RAFT) polymerization and terminal substitution methods. In addition, this study further investigated the effect of terminal chemistry on the thermoresponsive behavior of PIPAAm brushes for discussing relationships with thermoresponsive cell adhesion/ detachment profiles as well as cell sheet fabrication.

Scheme 1. Synthetic Scheme of Preparation of Terminally Functionalized Poly(N-isopropylacrylamide) (PIPAAm) Brush Grafted Surfaces

[(dodecylsulfanylthiocarbonyl)sulfanyl]pentanoic acid as the CTA was introduced on the amino-functionalized glass coverslips through an amide-bond formation reaction. Functionalized glass coverslips were immersed in 250 mL of DCM containing CTA (1.25 mmol), NHS (1.88 mmol), and DCC (1.88 mmol) and allowed to react with DCM solution by stirring the reaction solution for 24 h at 25 °C in the dark. The resultant CTA-immobilized glass coverslips (sCTA) were rinsed with DCM and acetone adequately, followed by drying under vacuum. PIPAAm-grafted glass surfaces were prepared through SI-RAFT polymerization using sCTA. The prepared sCTAs were immersed in 250 mL of 1,4-dioxane containing IPAAm monomer (0.30 mol), CTA (0.25 mmol), and V-501 (0.05 mmol) in a 300 mL separable flask, which was deoxygenated by N2 gas bubbling for 1 h, and then the polymerization was performed at 70 °C for 6 h. After the polymerization, glass coverslips grafted with PIPAAm chains having terminal dodecyl trithiocarbonate groups derived from CTA (sIP-D) were thoroughly washed with acetone and dried in vacuo. Substitution of terminal moieties of PIPAAm brushes was adopted a one-pot reaction of both terminal reduction and coupling reaction with maleimide for obtaining the highly efficient modification of polymer termini. The sIP-Ds were immersed in 200 mL of 0.1 mol/L carbonate-bicarbonate buffer solution (pH 9.0) containing maleimide (6 mmol), NaBH4 (100 mmol), and sodium hydrosulfite (0.2 mmol) for 4 h at 20 °C. Subsequently, glass coverslips grafted with maleimideterminated PIPAAm brushes (sIP-M) were washed with Milli-Q water and acetone, followed by drying under reduced pressure. In this study, SI-RAFT polymerization of IPAAm on glass beads was also performed for determining the molecular weight of grafted PIPAAm chains.34 Glass beads (the average diameter, 50 μm; 60.0 g) were washed with concentrated hydrochloric acid for 3 h at 90 °C and rinsed with a large amount of water repeatedly until the washing-water pH became neutral, followed by thorough drying in a vacuum oven at 110 °C for overnight. The acid-treated glass beads were reacted again with APTES (0.5 v/v%) in 200 mL of toluene in a 500 mL round-


Materials. N-Isopropylacrylamide (IPAAm) was kindly gifted by Kohjin (Tokyo, Japan) and recrystallized twice from n-hexane. 3Aminopropyltriethoxysilane (APTES; Shin-Etsu Chemical, Tokyo) was used as received. 4,4′-Azobis(4-cyanovaleric acid) (V-501), Nhydroxysuccinimide (NHS), N,N′-dicyclohexylcarbodiimide (DCC), 1,4-dioxane, dehydrated dichloromethane (DCM), dehydrated toluene, diethyl ether, sodium hydrosulfite, sodium hydroxide, hydrochloric acid, acetone, and methanol were obtained from Wako Pure Chemicals (Osaka, Japan) and used without further purification. Sodium borohydride (NaBH4) was supplied from Kanto Chemical (Tokyo, Japan). 4-Cyano-4-[(dodecylsulfanylthiocarbonyl)sulfanyl]pentanoic acid, maleimide, Dulbecco’s modified Eagle’s medium (DMEM), Mg2+- and Ca2+-free Dulbecco’s phosphate-buffered saline (PBS), and trypsin-EDTA were purchased from Sigma-Aldrich (St. Louis, MO). Penicillin-streptomycin solution (100 unit/mL penicillin and 100 μg/mL streptomycin) was obtained from Gibco BRL (Gland Island, NY). Fetal bovine serum (FBS) was supplied from Japan Bioserum (Hiroshima, Japan). Rhodamine-labeled fibronectin (FN) from bovine plasma was purchased from Cytoskeleton (Denver, CO). Bovine carotid artery endothelial cells (BAECs) were provided from 3165 | Biomacromolecules 2013, 14, 3164−3171



bottom flask at 20 °C for 20 h. After the reaction, aminofunctionalized glass beads were washed with toluene and methanol and then were dried in a vacuum oven at 25 °C. The glass beads were immersed in reaction solution containing CTA with carboxyl-activated ester at 25 °C for 24 h, and then the resultant CTA-immobilized glass beads were rinsed with DCM and acetone, followed by drying under vacuum. The prepared CTA-immobilized glass beads (20.0 g) were immersed in 200 mL of 1,4-dioxane containing IPAAm (0.24 mol), CTA (0.20 mmol), and V-501 (0.04 mmol). The polymerization procedure and terminal substitution with maleimide were carried out as a procedure similar to that of the preparation of PIPAAm-grafted coverslips. Characterization of PIPAAm Brush Surfaces. For investigating the surface elemental compositions of various surfaces, X-ray photoelectron spectroscopy (XPS) measurement was performed by an XPS instrument (K-Alpha, Thermo Fisher Scientific, Waltham, MA) with a monochromatic Al Kα1,2 radiation devices. Elemental compositions of individual surfaces were determined at a takeoff angle of 10°. Amount of grafted PIPAAm was determined by an attenuated total reflection-Fourier transform infrared (ATR/FT-IR) spectroscopy system (Nicolet 6700; Thermo Scientific, Waltham, MA) equipped with a germanium ATR crystal (Harrick Scientific Corporation, Pleasantville, NY).16,35 The peak intensity at near 1650 cm−1 originated from the amide carbonyl group of IPAAm was normalized by the intensity of peak observed at 1000 cm−1 from the Si−O−Si bonds of glass substrates. The grafted amounts of PIPAAm on the glass surfaces were determined from the intensity ratio (I1650/I1000) using a calibration curve that was made from PIPAAm with known amount cast on sCTA.16,36 Data are expressed as the mean of three separate substrates with standard deviation (SD). Wettability changes on thermoresponsive polymer-grafted surfaces were characterized by a static water contact angle meter (DSA100; KRÜ SS, Hamburg, Germany) by a captive bubble method.16,33 Samples were immersed in PBS for at least 1 h at specific temperatures, and then air bubbles were put on the surfaces. PBS temperature was thermostated by a circulating water bath (Lauda Re104; Lauda, Lauda-Königshofen, Germany). Data are averaged from three separate samples and shown with SD. Characterization of PIPAAm. PIPAAm-grafted glass beads were treated with 10 mol/L sodium hydroxide aqueous solution for overnight to retrieve the grafted PIPAAm. After being neutralized by adding hydrochloric acid, the solution was filtered and dialyzed against Milli-Q water using dialysis membrane (Spectra/Por standard regenerated cellulose dialysis membrane; number of membrane, 6; molecular weight cut off, 1000; Spectrum Laboratories, Rancho Dominguez, CA) for 1 week with a daily water exchange. Finally, retrieved PIPAAm was recovered by freeze-drying. Molecular weight of PIPAAm was analyzed by a gel permeation chromatography (GPC) system (HLC-8320GPC; Tosoh, Tokyo) with three sequentially connected columns (TSKgel Super AW2500, TSKgel Super AW3000, and TSKgel Super AW4000; Tosoh) at 40 °C using DMF containing 50 mmol/L LiCl as an eluent at a flow rate of 0.6 mL/min. Polymer molecular weights and their polydispersities (Mw/Mn) were calculated from a calibration curve prepared using poly(ethylene oxide) standards (Polysciences, Warrington, PA). Graft density of PIPAAm on glass surfaces was estimated by the follow equation:34 graft density =

cooling process at a rate of 0.1 °C/min. A sample cuvette thermostat was a Peltier-effect cell holder (EHC-477S, JASCO). LCSTs of PIPAAm solutions were defined as a specific temperature producing a 50% decrease in optical transmittance. Cell Adhesion and Detachment Assay. BAECs were cultured in DMEM supplemented with 10% FBS, 100 units/mL penicillin, and 100 μg/mL streptomycin on tissue culture polystyrene (TCPS) dishes at 37 °C under a humidified atmosphere with 5% CO2. Either PIPAAm-brush surfaces or sCTA were cut into half size (24 × 25 mm), and were placed on 35 mm Petri dishes (Falcon 351008, BD Biosciences, Franklin Lakes, NJ). After being recovered by 0.25% trypsin-EDTA treatment from the dish, cells were seeded at a density of 5 × 103 cells/cm2 onto PIPAAm-grafted surfaces and sCTA at room temperature (grafted PIPAAm was in hydrated state) and incubated at various temperatures. Adhering cells were observed by a phase contrast microscope (ECLIPSE TE2000-U; Nikon, Tokyo) for evaluating temperature-dependent cell adhesion/proliferation properties. To investigate thermoresponsive cell detachment behavior from the surfaces, cells cultured at 37 °C were incubated at 20, 28, and 29 °C, and the number of adhering cells was counted at various time periods. Cell adhesion profiles were investigated as the mean of three separate experiments with SD. Protein Adsorption Assay. Fibronectin (FN) adsorption on various surfaces was also observed as incubation temperature at 29, 31, and 33 °C. PIPAAm-grafted surfaces and CTA-immobilized surfaces were incubated with Rhodamine-labeled FN in PBS (5 μg/mL) for 3 h. After being washed the surfaces with PBS incubated three times at specific temperatures, the surfaces were examined by a fluorescence microscopy (ECLIPSE TE2000-U), and the image data were collected with Axio version (version 3.1; Carl Zeiss, Inc., NY). Cell Sheet Preparation. For harvesting cell sheets, BAECs were seeded at 1 × 105 cells/cm2 on each PIPAAm-grafted surface (sIP-D and sIP-M), followed by incubation at 37 °C for allowing them to reach to a confluent condition. The adhering cells were incubated at 20 °C for initiating the detachment of cell sheets. Harvesting cell sheets were observed microscopically and visually at various time periods. Recovery periods of BAEC sheets from the surfaces were determined from at least three reproducible results. Statistical Analysis. Experimental data were expressed as the mean with SD of sample number, n (n is at least 3). Analysis of variance followed by two-tailed student’s t-test was used to evaluate significant differences among experimental groups. Results were considered statistically significant if p < 0.05.

RESULTS Characterization of Terminal-Functionalized Thermoresponsive Polymer Brushes. PIPAAm brushes having terminal dodecyl trithiocarbonate groups on glass surfaces (sIPD) were prepared by SI-RAFT polymerization. Maleimideterminated PIPAAm brushes (sIP-M) were obtained through the reduction of terminal trithiocarbonate moieties, which were converted to thiol groups and were reacted with maleimide (Scheme 1). XPS measurements were performed for determining the elemental composition of the prepared surfaces (Table 1). Sulfur peak derived from trithiocarbonate moiety was detected for CTA-immobilized surfaces (sCTA). Peaks of carbon and nitrogen for the surfaces after conducting IPAAm polymerization were much larger than those of sCTA. ATR/FT-IR study also provided information about PIPAAmgrafted surfaces. As shown in Table 1, the grafted amounts of PIPAAm for sIP-D and sIP-M were 1.02 ± 0.10 and 0.95 ± 0.12 μg/cm2, respectively. Furthermore, PIPAAm recovered from glass substrate was collected and evaluated by GPC analysis for characterizing the surface grafted polymers. Retrieved PIPAAm possessed Mn of 3.8 × 104 and a comparatively narrow distribution value (Mw/Mn) of 1.52. From the above-mentioned results, the graft PIPAAm density

mpNA Mn


where mp is the weight of grafted PIPAAm per square centimeter (μg/ cm2), NA is Avogadro’s number, and Mn is the number average molecular weight of the grafted PIPAAm. Dodecyl- and maleimide-terminated PIPAAms, which possessed equivalent molecular weights to grafted PIPAAm, were individually dissolved in PBS at a polymer concentration of 10 mg/mL. Optical transmittances of PIPAAm aqueous solutions at various temperatures were measured at 600 nm using a UV−vis spectrometer (V-530; JASCO, Tokyo). Temperature scanning was performed in heating/ 3166 | Biomacromolecules 2013, 14, 3164−3171



around PIPAAm’s LCST, while the surface wettability of sCTA was temperature-independent. Namely, cos θ values were gradually decreased from 25 °C and drastically changed around PIPAAm’s LCST. In addition, maleimide-terminated PIPAAm brush (sIP-M) possessed a slight hydrophilic property in comparison with the surfaces having hydrophobic dodecyl termini (sIP-D) below 33 °C (in most points, p < 0.05). Interestingly, terminal functionality affected the thermoresponsive phase transition behavior of PIPAAm brushes in the heating process for proceeding polymer dehydration. The phase transition profile of sIP-D was clearly shifted to lower temperature (approximately 2 °C) than that of sIP-M (Figure 1A). Especially, a large magnitude contact angle difference was observed between terminally functionalized PIPAAm brushes at 31 °C (cos θ for sIP-D, 0.79 ± 0.02; and cos θ for sIP-M, 0.84 ± 0.02). On the other hand, in the cooling process for the rehydration of grafted PIPAAm chains, sIP-D and sIP-M exhibited comparable temperature-dependent behavior without the shift of phase transition profile via terminal functionality (Figure 1B). To clearly describe the effect of polymer termini on the thermoresponse of PIPAAm brushes, the hydration/dehydration behavior of linear semitelechelic (one-end functionalized) (dodecyl- and maleimide-) PIPAAms having equivalent molecular weights to grafted polymer chains was investigated by a solution turbidity method (Table 2). In the heating

Table 1. Amount of Grafted Poly(N-isopropylacrylamide) (PIPAAm) on Glass Surfaces elementb (atomic%)







grafted PIPAAm amountc (μg/cm2)


27.6 65.4 61.9

3.7 8.5 7.3

44.0 17.9 20.9

1.1 0.1 0.0

23.6 8.1 9.9

1.02 ± 0.10 0.95 ± 0.12

graft density (chains/ nm2) 0.16 0.15


The samples used in this study were CTA-immobilized surface (sCTA), PIPAAm brush surface with terminal dodecyl group (sIP-D) and PIPAAm brush surface with terminal maleimide group (sIP-M). b Determined by X-ray photoelectron spectroscopy (XPS); takeoff angle, 10°. cDetermined by attenuated total reflection-Fourier transform infrared (ATR/FT-IR), mean ± SD (n = 3)

on glass substrates was calculated to be 0.16 chains/nm2, indicating that high-density brush was created on glass coverslips. Temperature-dependent surface wettability changes of polymer-grafted surfaces were investigated by static water contact angle measurements using captive bubble method at temperatures across PIPAAm’s LCST in the range of 20 to 37 °C. Unlike PIPAAm brush grafted surfaces (sIP-D and sIP-M), CTA-immobilized surfaces (sCTA) exhibited hydrophobic properties (cos θ: 0.68 ± 0.02; Figure 1). As a notable surface-wettability characteristic, individual PIPAAm brush grafted surfaces showed significant thermoresponsive behavior

Table 2. Characterization of Linear Semitelechelic Poly(Nisopropylacrylamide)s (PIPAAm). LCSTb (°C) sample


dodecyl-PIPAAm maleimide-PIPAAm

4.0 × 10 4.0 × 104 4




1.62 1.69

29.0 29.6

28.0 27.9

a Determined by gel permeation chromatography (GPC) using N, Ndimethylformamide with 50 mmol/L LiCl. bLower critical solution temperature (LCST) profiles were determined by optical transmittance changes at 600 nm in phosphate-buffered saline (PBS) at a heating/cooling rate of 0.1 °C/min and a polymer concentration of 10 mg/mL.

process for polymer dehydration, hydrophobic dodecylterminated PIPAAm in PBS exhibited a lower LCST at 29.0 °C than that of maleimide-terminated PIPAAm (at 29.6 °C). On the other hand, the hysteresis of thermal phase transition for both semitelechelic polymers was observed between heating and cooling treatment.1 Furthermore, the aqueous solutions of PIPAAms having heterogeneous terminal groups showed almost the same LCST at 28.0 °C in the cooling process for polymer rehydration. Temperature-Dependent Cellular Behavior on PIPAAm Brush Surfaces. To investigate the effect of terminal functionality on cell-adhesion temperature on PIPAAm brush surfaces, BAEC adhesion profiles on various surfaces at specific incubation temperatures were observed microscopically. BAEC adhesion profiles on PIPAAm brush surfaces (sIP-D and sIPM), and sCTA at various incubation temperatures at 24 h after cell seeding were shown in Figure 2 and Figure 3A. Remarkable adhesion of BAECs was observed on sCTA with a slight temperature effect in the range of 25 to 37 °C. In contrast, PIPAAm terminal properties significantly affected temperaturedependent cell adhesion behavior. BAECs adhered on both PIPAAm brush surfaces at the same level as sCTA surfaces at

Figure 1. Temperature-dependent static contact angle changes of various surfaces in the heating process (A) and the cooling process (B). sCTA indicates CTA-immobilized surfaces; sIP-D, dodecylterminated poly(N-isopropylacrylamide) (PIPAAm) brush grafted surfaces; sIP-M, maleimide-terminated PIPAAm brush grafted surfaces. *p < 0.05. 3167 | Biomacromolecules 2013, 14, 3164−3171



temperature than that of sIP-D. Cell adhesion was drastically suppressed on sIP-M at 31 °C and promoted upon heating above 33 °C. These unique cell adhesion characters strongly corresponded with the temperature-dependent surface-property alteration of PIPAAm brush surfaces having terminal hydrophobic/hydrophilic functionality, from which was confirmed by the surface wettability study. Temperature-dependent protein adhesion was also investigated for the possible mechanism of cell adhesion properties of various surfaces by using FN, which is one of the critical components of ECM proteins for mediating cell adhesion. Rhodamine-labeled FN was used to visualize protein adsorption on PIPAAm-grafted surface and sCTA at various incubation temperatures (i.e., 29, 31, and 33 °C; Figure 3B). In fluorescence microscopic study, both terminally functionalized PIPAAm brush surfaces showed scarcely stained with Rhodamine-labeled FN at 29 °C, unlike temperature-independent sCTA surfaces, which were stained in ocher due to the adsorption of Rhodamine-labeled FN. On the other hand, although sIP-D surface was fluorescently stained above 31 °C, FN adsorption on sIP-M surface was clearly observed at a higher temperature of 33 °C. Effect of terminal chemistry on temperature-induced cell detachment kinetics was also investigated by incubating cells at various temperatures (20, 28, and 29 °C), as shown in Figure 4.

Figure 2. Temperature-dependent cell adhesion behavior on individual surfaces for after 24 h incubation at various temperatures. sCTA indicates CTA-immobilized surfaces; sIP-D, dodecyl-terminated poly(N-isopropylacrylamide) (PIPAAm) brush grafted surfaces; sIPM, maleimide-terminated PIPAAm brush grafted surfaces.

Figure 4. Temperature-dependent cell detachment behavior from individual terminally functionalized poly(N-isopropylacrylamide) (PIPAAm) brush grafted surfaces at various temperatures: open symbols, dodecyl-terminated PIPAAm brush grafted surfaces (sIP-D), and close symbols, maleimide-terminated PIPAAm brush grafted surfaces (sIP-M).

Figure 3. (A) Microscopic photographs of adhering cells at various temperatures 29, 31, and 33 °C for 24 h after cell seeding on the surfaces. Cell seeding density: 5.0 × 103 cells/cm2. (B) Fluorescent microscopic images Rhodamine-labeled fibronectin adsorbed to the surfaces at 29, 31, and 33 °C for 3 h. sCTA indicates CTAimmobilized surfaces; sIP-D, dodecyl-terminated brush poly(Nisopropylacrylamide) (PIPAAm) grafted surfaces; sIP-M, maleimideterminated PIPAAm brush grafted surfaces. Scale bar 100 μm.

Cell detachment from both PIPAAm surfaces was quite rapid and completed within 2 h at 20 °C incubation. Upon increasing incubation temperature to 28 °C, slow cell detachment profiles from the individual PIPAAm brush surfaces were observed, and approximately 80% adhering cells were detached spontaneously within 24 h. On the other hand, 29 °C incubation demonstrated the negligible amounts of detached cells from thermoresponsive surfaces after 24 h. Furthermore, temperature-dependent cell detachment was scarcely influenced by PIPAAm termini. For investigating the efficiency of cell sheet fabrication, BAECs were seeded at 1 × 105 cells/cm2 on PIPAAm brush surfaces, and then cell culture was performed for 5 days at 37 °C. In addition, adherent cells proliferated and reached to confluent at equivalent rates within 5 days, regardless of their terminal functional groups (see Figure S1 in Supporting Information). After reaching to confluency, cells were incubated

37 °C regardless of their terminal hydrophobic/hydrophilic properties. However, adhering cells on both sIP-D and sIP-M surfaces were extremely decreased below 29 °C. Interestingly, the numbers of adhering cells were quite different between sIPD and sIP-M at incubation temperature of 31 °C. BAEC adhesion on dodecyl-terminated PIPAAm brush surface (sIPD) was initiated above 31 °C, and the population of adhering cells was comparable to that of sCTA as the control. However, the cell adhesion-initiating temperature of sIP-M having terminal maleimide groups was clearly shifted to higher 3168 | Biomacromolecules 2013, 14, 3164−3171



at 20 °C, and the detachment of cell sheet was observed. As shown in Figure 5, square-shaped cellular monolayers were

densely packed PIPAAm grafted surfaces (chain density: 0.16 chains/nm2) successfully and further investigated the influence of PIPAAm’s terminal hydrophobic/hydrophilic functionality on the thermoresponsive surface properties and corresponding temperature-dependent cell adhesion characters of the surfaces. We have developed PIPAAm-grafted surfaces with 10−20 nm thick polymer layer for temperature-controlled cell adhesion/detachment, which is caused by reversible surface hydrophobic/hydrophilic alternation and polymer conformational changes (coil-to-globule transition) across PIPAAm’s LCST.14 In this study, terminal hydrophobic/hydrophilic functionality was found to affect cell adhesion temperatures on PIPAAm brush surfaces (Figure 2). BAECs adhered on sIPD surfaces above 31 °C, while cell adhesion on sIP-M surfaces initiated was promoted above 33 °C, as illustrated in Figure 6A. These results indicated that temperature-induced cell-adhesion behavior was attributed to the switchable surface property of hydrophobic/hydrophilic states across the LCST of grafted PIPAAm chains. Therefore, this study investigated temperaturedependent surface wettability changes of polymer-grafted surfaces. Terminally functionalized PIPAAm brush surfaces exhibited reversible thermoresponsive static contact angle changes in the range of 20 to 37 °C, indicating that grafted PIPAAm chains showed switchable hydrophobic/hydrophilic properties through their dehydration/rehydration processes. Both PIPAAm brush surfaces showed continuous changes in contact angles. These results indicated that thermoresponsive surface property changes were attributed to the polymer dehydration condition in various temperatures. Interestingly, hydrophobic dodecyl-terminated PIPAAm brush surface (sIPD) shifted its phase transition behavior to lower temperature by 2 °C than that of hydrophilic maleimide-terminated PIPAAm brushes (sIP-M) in the polymer dehydration process via heating. On the other hand, a slight shift in phase transition temperature up to 0.6 °C via terminal functionality was also determined by the optical transmittance changes of linear semitelechelic PIPAAm aqueous solutions. Our research group and others independently reported the effects of terminal hydrophobic moieties (e.g., alkyl and phenyl groups) on the thermoresponsive behavior of PIPAAm.24,31,32 Terminal hydrophobes enhances the dehydration of proximal IPAAm units and further disrupt polymer hydration, resulting in a typical LCST shift to lower temperature. The terminal effect on PIPAAm’s

Figure 5. Spontaneous cell sheets detachment from poly(Nisopropylacrylamide) (PIPAAm) brush surfaces. Cell sheet detachments from dodecyl-terminated PIPAAm brush grafted surfaces (sIPD) and maleimide-terminated PIPAAm brush grafted surfaces (sIP-M) completed after 29 ± 2 and 23 ± 2 min, respectively (sample number n = 3). White dashed lines shows the edges of PIPAAm brush grafted glass surfaces. Scale bars: 1 cm.

successfully harvested from sIP-D and sIP-M surfaces by reducing temperature. Detached cell sheets shrunk their sizes, which were smaller than the original areas of PIPAAm brush grafted glass surfaces. In addition, cell sheets were completely detached from both thermoresponsive surfaces at comparable time scales (sIP-D, 29 ± 2 min; sIP-M, 23 ± 2 min; p < 0.05).

DISCUSSION SI-RAFT polymerization of IPAAm gave to construct highly dense-packed thermoresponsive polymer brushes on glass surfaces. In addition, CTA-derived dodecyl trithiocarbonate groups of PIPAAm chains were successfully substituted with hydrophilic maleimide groups through terminal reduction and maleimide coupling reactions. In previous studies, polymer brush is defined as closely packed polymer chains at interface at a high density of more than 0.1 chains/nm2, and limited space gives the strongly extended conformation of grafted polymer chains.37,38 In the past decade, polymer brushes have attracted much attention due to their physical properties (highly stretched and extended polymer architectures) and unique interfacial properties including liquid wetting and inhibition of protein adsorption.39 This study fabricated well-defined and

Figure 6. Schematic illustrations of temperature-dependent cell adhesion/detachment behavior on terminally functionalized poly(Nisopropylacrylamide) (PIPAAm)-grafted surfaces (sIP-D, dodecyl-terminated PIPAAm brush grafted surfaces; sIP-M, maleimide-terminated PIPAAm brush grafted surfaces). 3169 | Biomacromolecules 2013, 14, 3164−3171



effectively in a manner similar to that of nonpolymer grafted surfaces (sCTA). Discussion of spontaneous cell detachment behavior from PIPAAm brushes via the cooling treatment was schematically represented in Figure 6B. Kinetics of cell detachment from the thermoresponsive surfaces was dramatically affected by incubation temperature below PIPAAm’s LCST. In addition, cell adhesion protein was attached to basal cell surface.40 Cell detachment rate at 20 °C was quite rapid regardless of terminal chemistry. However, cell detachment was remarkably delayed by increasing temperatures to 28 °C, and the incubation at 29 °C demonstrated the negligible amounts of cells detached from the PIPAAm grafted surfaces. This was probably due to the difference of hydration states of grafted PIPAAm chains at individual temperatures. Temperature-dependent hydration states directly affected surface hydrophobic characters and PIPAAm chain conformation, which were key factors for promoting spontaneous cell detachment. Furthermore, temperatures for initiating cell detachment were found to be lower than those for cell adhesion due to the hysteresis of thermal polymer phase transition. Cell sheets were successfully harvested from terminally functionalized PIPAAm brush grafted surfaces by reducing temperature to 20 °C. Harvested cell sheets had biologically intact cellular functions including inherent cell-cell junctions, and the size of cell sheet shrank because of the cytoskeletal rearrangement of individual cell in the cell sheet. The cytoskeletal rearrangement gives that morphologic change of cells from spreading shapes in adherent states to round shapes in nonadherent states. The integrated cellular shrinkage force derived from cell−cell connection accelerates the detachment rate of sheet-like cells. Thermoresponsive phase transition of PIPAAm-based brushes has been conventionally controlled by introducing hydrophobic/hydrophilic comonomers in PIPAAm main chains.41 However, the copolymerization methods often change the hydrophobicity and extension/aggregation states of thermoresponsive polymer grafted surfaces, and thus protein adsorption and cell adhesion to the surfaces are dramatically varied.41,42 In this study, thermoresponsive behavior (especially, dehydration process) of PIPAAm brushes was able to be regulated by only varying terminal functionality with a scarce characteristic alternation of PIPAAm chains. Consequently, specific cell adhesion temperatures on to PIPAAm brushes was successfully controlled even though molecular weight and chemical composition of grafted polymers were equivalent except for terminal groups. By using the unique regulation system, the construction of PIPAAm brushes having micropatterned terminal functional groups can be applied to twodimensional heterogeneous cell coculture systems via a multistep cell seeding at various temperatures and subsequently fabricated high-functional cell sheets.

LCST greatly depends on polymer molecular weight, and the LCST shift gradually reduces with increasing molecular weight. Especially, PIPAAms having their molecular weight (Mn) more than 3.0 × 104 show the slight shifts of their LCST values.24 Therefore, in this study, the difference of LCST between dodecyl-terminated and maleimide-terminated PIPAAms (Mn: 4.0 × 104) was a small value of 0.6 °C. Moreover, our previous work also demonstrated the unique terminal hydrophobe effect on thermoresponse of PIPAAm brushes using PIPAAm-based block copolymer micelles.24 Highly concentrated hydrophobes on PIPAAm brush interface totally enhance their hydrophobic effect and significantly promote polymer dehydration, and therefore PIPAAm’s LCST shift to lower temperatures rather than that of dispersed linear polymers. In the present PIPAAm brush surface system, concentrated hydrophobic dodecyl groups on PIPAAm brush surfaces probably resulted in totally enhancing the dehydration of densely packed grafted polymer chains and amplifying the shift of PIPAAm’s phase transition temperatures. In contrast, the polymer rehydration process via cooling across the temperature of PIPAAm’s LCST was found to be similar to the thermoresponsive profiles of individual PIPAAm brush surfaces without a significant temperature shift, which resembled linear semitelechelic PIPAAms. These results indicated that the rehydration of PIPAAm chains was scarcely affected by terminal hydrophobic/hydrophilic characters, because the adsorption of water molecules (polymer rehydration) was speculated to be initiated at random IPAAm units in the polymer main chains with a scarce terminal effect. This study also investigated temperature-dependent FN adhesion on PIPAAm brush grafted surfaces for describing the relationship between cell adhesion behavior and adhesionmediating proteins. Protein adsorption is often concerned for designing biointerfaces for enhancing cell adhesion. Especially, the quantitative adhesion analyses of cell adhesion-mediating proteins (e.g., FN) is a valuable way for understanding cell adhesive property on target surfaces.14 Temperature-dependent FN adsorption on individual PIPAAm brush surfaces was found to closely relate to thermoresponsive surface wettability profiles (Figure 1A and Figure 3B). Although FN adsorption on each PIPAAm brush surface was negligible at low temperatures below 29 °C, the adsorption was greatly enhanced above LCST because of increase in surface hydrophobicity. In our previous studies, protein adhesion on PIPAAm-related surfaces is thermally regulated through the alternation of surface wettability and grafted polymer conformation.14 At a temperature below PIPAAm’s LCST, highly hydrated grafted PIPAAm chains with hydrophilic and extending chain conformation suppress their interaction with proteins. Upon heating above LCST, proteins interact with dehydrated grafted polymer chains in hydrophobic and aggregated states. As important finding, differences in the specific temperatures of FN adsorption among PIPAAm brush surface were obviously attributed to the phase transition shifts of PIPAAm brushes caused by terminal functionalities. Therefore, the temperaturedependent adsorption behavior of cell adhesion-mediating proteins (e.g., FN) possibly resulted in varying specific cell adhesion temperatures on heterogeneously terminated-PIPAAm brush surfaces. In addition, individual PIPAAm brush surfaces showed equivalent surface wettabilities in the completely dehydrated condition, independent of terminal functionalities. Therefore, BAECs adhered on both hydrophobic PIPAAm brush surfaces at 37 °C and proliferated

CONCLUSIONS In this study, densely packed end-functionalized thermoresponsive-polymer-brush surfaces were successfully prepared through SI-RAFT polymerization and subsequent terminal substitution reaction. Terminal functionality induced the alternation of thermoresponsive surface properties in polymer dehydration process via heating across PIPAAm’s LCST. Hydrophobic terminal groups promoted the dehydration of grafted PIPAAm chains and gave the lower LCST shift of PIPAAm in comparison to hydrophilic maleimide-terminated 3170 | Biomacromolecules 2013, 14, 3164−3171



(13) Yamada, N.; Okano, T.; Sakai, H.; Karikusa, F.; Sawasaki, Y.; Sakurai, Y. Makromol. Chem., Rapid Commun. 1990, 11, 571−576. (14) Akiyama, Y.; Kikuchi, A.; Yamato, M.; Okano, T. Langmuir 2004, 20, 5506−5511. (15) Barbey, R.; Lavanant, L.; Paripovic, D.; Schuwer, N.; Sugnaux, C.; Tugulu, S.; Klok, H. A. Chem. Rev. 2009, 109, 5437−5527. (16) Takahashi, H.; Nakayama, M.; Yamato, M.; Okano, T. Biomacromolecules 2010, 11, 1991−1999. (17) Nagase, K.; Watanabe, M; Kikuchi, A.; Yamato, M.; Okano, T. Macromol. Biosci. 2011, 11, 400−409. (18) Li, L.; Zhu, Y.; Li, B.; Gao, C. Langmuir 2008, 24, 13632− 13639. (19) Takahashi, H.; Matsuzaka, N.; Nakayama, M.; Kikuchi, A.; Yamato, M.; Okano, T. Biomacromolecules 2012, 13, 253−260. (20) McCormick, C. L.; Sumerlin, B. S.; Lokitz, B. S.; Stempka, J. E. Soft Matter 2008, 4, 1760−1773. (21) Ganachaud, F.; Monterio, M. J.; Gilbert, R. G.; Dourges, M. A.; Thang, S. H.; Rizzardo, E. Macromolecules 2000, 33, 6738−6745. (22) Moad, G.; Rizzardo, E.; Thang, S. H. Aust. J. Chem. 2005, 58, 379−410. (23) Baum, M.; Brittain, W. J. Macromolecules 2002, 35, 610−615. (24) Nakayama, M.; Okano, T. Biomacromolecules 2005, 6, 2320− 2327. (25) Lowe, A. B.; Sumerlin, B. S.; Donovan, M. S.; McCormick, C. L. J. Am. Chem. Soc. 2002, 124, 11562−11563. (26) Sumerlin, B. S.; Lowe, A. B.; Stroud, P. A.; Zhang, P.; Urban, M. W.; McCormick, C. L. Langmuir 2003, 19, 5559−5562. (27) Scales, C. W.; Converctine, A. J.; McCormick, C. L. Biomacromolecules 2006, 7, 1389−1392. (28) York, A. W.; Scales, C. W.; Huang, F.; McCormick, C. L. Biomacromolecules 2007, 8, 2337−2341. (29) Li, M.; De, P.; Gondi, S. R.; Sumerlin, B. S. J. Polym. Sci., Part A: Polym. Chem. 2008, 46, 5093−5100. (30) Nakayama, M.; Okano, T. Macromolecules 2008, 41, 504−507. (31) Chung, J. E.; Yokoyama, M.; Suzuki, K.; Aoyagi, T.; Sakurai, Y.; Okano, T. Colloids Surf., B 1997, 9, 37−48. (32) Furyk, S.; Zhang, Y.; Ortiz-Acosta, D.; Cremer, P. S.; Bergbreiter, D. E. J. Polym. Sci., Part A: Polym. Chem. 2006, 44, 1492−1501. (33) Matsuzaka, N.; Takahashi, H.; Nakayama, M.; Kikuchi, A.; Okano, T. J. Biomater. Sci., Polym. Ed. 2012, 23, 1301−1314. (34) Nagase, K.; Kobayashi, J.; Kikuchi, A.; Akiyama, Y.; Kanazawa, H.; Okano, T. Appl. Mater. Interfaces 2012, 4, 1998−2008. (35) Tsuda, Y.; Kikuchi, A.; Yamato, M.; Nakao, A.; Sakurai, Y.; Umezu, M.; Okano, T. Biomaterials 2005, 26, 1885−1893. (36) Mandal, T. K.; Woo, E. M. Macromol. Chem. Phys. 1999, 200, 1143−1149. (37) Yamamoto, S.; Ejaz, M.; Tsuji, Y.; Matsumoto, M.; Fukuda, T. Macromolecules 2000, 33, 5602−5607. (38) Yamamoto, S.; Ejaz, M.; Tsuji, Y.; Fukuda, T. Macromolecules 2000, 33, 5608−5612. (39) Yoshikawa, C.; Goto, A.; Tsuji, Y.; Fukuda, T.; Kimura, T.; Yamamoto, K.; Kishida, A. Macromolecules 2006, 39, 2284−2290. (40) Yamato, M.; Konno, C.; Kushida, A.; Hirose, M.; Utsumi, M.; Kikuchi, A.; Okano, T. Biomaterials 2000, 21, 981−986. (41) Tamura, A.; Nishi, M.; Kobayashi, J.; Nagase, K.; Yajima, H.; Yamato, M.; Okano, T. Biomacromolecules 2012, 13, 1765−1773. (42) Ebara, M.; Yamato, M.; Nagai, S.; Aoyagi, T.; Kikuchi, A.; Sakai, K.; Okano, T. Surf. Sci. 2004, 570, 134−141.

polymers. Using this unique surface feature, specific cell adhesion temperatures of PIPAAm brush surface was regulated by changing terminal hydrophilic/hydrophobic moieties. On the other hand, low temperature-induced cell detachment through PIPAAm rehydration was scarcely affected by terminal functional groups. Furthermore, cell sheets were harvested from both PIPAAm brush surfaces within 30 min at relatively comparable rates. In the future work, the unique regulation of cellular behavior via the terminal chemistry of PIPAAm brushes can be used for performing two-dimensional coculture systems and functional cell sheet fabrication by micropatterning of terminal functional groups.


S Supporting Information *

Cell proliferation profiles. This material is available free of charge via the Internet at


Corresponding Author

*Tel.: +81-3-5367-9945 (6201). Fax: +81-3-3359-6046. E-mail: [email protected]. Notes

The authors declare no competing financial interest.

ACKNOWLEDGMENTS This work was partially supported by JSPS KAKENHI (Grant Number 24-7141) from Japan Society for Promotion of Science (JSPS), “Creation of Innovation Centers for Advanced Interdisciplinary Research Areas Program in the Project for Developing Innovation Systems” and “Grant-in-Aid for Scientific Research (Grant Number 23106009) on Innovative Areas “Bio Assembler” (Area Number 2305)” from the Ministry of Education, Culture, Sports, Science and Technology (MEXT), Japan. We also acknowledge Dr. Kenichi Nagase of Tokyo Women’s Medical University for supporting the preparation of PIPAAm-grafted glass beads and recovery of grafted PIPAAm.


(1) Okano, T. Biorelated Polymers and Gels; Academic Press: San Diego, CA, 1998. (2) Okano, T.; Bae, Y. H.; Jacobs, H.; Kim, S. W. J. Controlled Release 1990, 11, 255−265. (3) Hoffman, A. S. Adv. Drug Delivery Rev. 2002, 54, 3−12. (4) Hoffman, A. S.; Stayton, P. S. Prog. Polym. Sci. 2007, 32, 922− 932. (5) Mori, T.; Umeno, D.; Maeda, M. Biotechnol. Bioeng. 2001, 72, 261−268. (6) Becker, K.; Hallgren, E.; Carredano, E.; Palmgren, R.; Bulow, L. J. Mol. Recognit. 2009, 22, 104−109. (7) Heskins, M.; Guillet, J. E. J. Macromol. Sci. Chem. 1968, A2, 1441−1455. (8) Kanazawa, H.; Nishikawa, M.; Mizutani, A.; Sakamoto, C.; Morita-Murase, Y.; Nagata, Y.; Kikuchi, A.; Okano, T. J. Chromatogr., A 2008, 1191, 157−161. (9) Okano, T.; Yamada, N.; Okuhara, M.; Sakai, H.; Sakurai, Y. Biomaterials 1995, 16, 297−303. (10) Yamato, M.; Okano, T. Mater. Today 2004, 7, 42−47. (11) Kikuchi, A.; Okuhara, M.; Karikusa, F.; Sakurai, Y.; Okano, T. J. Biomater. Sci., Polym. Ed. 1998, 9, 1331−1348. (12) Kushida, A.; Yamato, M.; Konno, C.; Kikuchi, A.; Sakurai, Y.; Okano, T. J. Biomed. Mater. Res. 2000, 51, 216−223. 3171 | Biomacromolecules 2013, 14, 3164−3171


Terminally functionalized poly(N-isopropylacrylamide) (PIPAAm) brush grafted glass surfaces were prepared by a surface-initiated reversible addition-f...
4MB Sizes 0 Downloads 0 Views