Thermo-Responsive PNiPAAm-g

Biomacromolecules 2003, 4, 1733-1739

1733

Thermo-Responsive PNiPAAm-g-PEG Films for Controlled Cell Detachment Dirk Schmaljohann,* Joachim Oswald, Birgitte Jørgensen, Mirko Nitschke, Detlev Beyerlein,† and Carsten Werner* Institute of Polymer Research Dresden & Max Bergmann Center of Biomaterials Dresden, Hohe Str. 6, 01069 Dresden, Germany

Downloaded via KAOHSIUNG MEDICAL UNIV on July 4, 2018 at 09:30:18 (UTC). See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles.

Received May 23, 2003; Revised Manuscript Received July 29, 2003

A series of graft copolymers consisting of either poly(N-isopropylacrylamide) (PNiPAAm) or poly(N,Ndiethylacrylamide) (PDEAAm) as a thermo-responsive component in the polymer backbone and poly(ethyleneglycol) (PEG) were immobilized as thin films and cross-linked on a fluoropolymer substrate using low-pressure argon plasma treatment. The surface-immobilized hydrogels exhibit a transition from partially collapsed to completely swollen, which is in the range of 32-35 °C and corresponds to the lower critical solution temperature of the soluble polymers. The hydrogels were used as cell carriers in culture experiments with L929 mouse fibroblast cells to probe for cell adhesion, proliferation, and temperature-dependent detachment of cell layers. The fibroblast cells adhere, spread, and proliferate on the hydrogel layers at 37 °C and become completely detached after reducing the temperature by 3 K. The cell release characteristics were further correlated to the swelling and collapsing behavior of the hydrogel films and the polymer solutions as measured in PBS solution and RPMI cell cultivation medium. It could be shown that, long before the swelling has completed upon temperature reduction, the cells detach. This can be attributed to the large content of PEG present in the hydrogel, which weaken the cell adhesion strength to the hydrogel layers. Introduction Surface-immobilized and stimuli-responsive hydrogel films gain increasing attention because of their potential as in vitro carriers for regenerative medicine.1,2 The advantage of hydrogels for this application lays in its mechanical properties in the swollen state. The softness allows cellular matrix reorganization, but the hydrogels still provide enough mechanical strength and elasticity, which is needed for cell anchorage to the substrate. Besides these properties, the meshwork of the gel permits diffusion and delivery of nutrients and growth factors. Stimuli-responsive hydrogels with changing surface properties offer the advantage of controlling cell adhesion and detachment upon the external physical signal. This allows one to harvest the cells without enzymatic treatment.3,4,5,6 It is related to the possibility of exposing or hiding surface functionalities and the change in hydrophilicity, which accompanies the variation in the degree of swelling. In general, stimuli-responsive polymers (SRP) are of great interest in biomedical sciences, because slight changes in the environmental conditions as pH, ionic strength, or temperature can cause a dramatic change in the polymer properties.7 Thus, the changes in solubility or the degree of swelling is due to a balance of competing interactions such as electrostatic forces and hydrophobic dehydration. Finally, * To whom correspondence should be addressed. (D.S.) E-mail: [email protected]. Phone: +49-351-4658446. Fax: +49-3514658565. (C. W.) E-mail: [email protected]. Phone: +49-351-4658532. Fax: +49-351-4658533. † Present address: Thu ¨ ringisches Institut fu¨r Textil- und KunststoffForschung e.V., Breitscheidstrasse 97, 07407 Rudolstadt, Germany.

enthalpic and entropic contributions shift the minimum of the free energy and causes the volume phase transition. This has been utilized for applications such as drug delivery,1,8,9 gene delivery,10,11 biosensors,12 besides the cell-surface adhesion control,3,6 respectively. Thermo-responsive materials are of particular interest for cell culture technologies, because the volume phase transition has to occur within the settings of the physiological conditions. Thermo-responsive polymers based on N-alkylacrylamides such as N-isopropylacrylamide (NiPAAm) have been studied intensively.1,13 PNiPAAm has a lower critical solution temperature (LCST) of 32 °C and the LCST can be easily increased by incorporation of polar comonomers or reduced by adding hydrophobic comonomers, respectively. In the field of cell cultivation supports based on thermoresponsive cell cultivation supports, Okano et al. have introduced NiPAAm-based hydrogels.1,3,6,14 The group of Okano also used copolymers of NiPAAm and 2-carboxyisopropylacrylamide to further accelerate the cell detachment.4 Recently, Ratner et al. reported work on plasmapolymerized NiPAAm.5 Both groups are using PNiPAAm based hydrogels, which show the desired change from a collapsed to a swollen state upon lowering the temperature from 37 °C. This swelling is accompanied by a change from a hydrophobic to a hydrophilic surface. The hydrogels based on PNiPAAm homopolymer almost completely expel their containing water upon rising the temperature above the transition temperature Ttr, and therefore, they also loose the gel properties. Our approach to this problem is to create a hydrogel, which responds to temperature changes, but it should also maintain

10.1021/bm034160p CCC: $25.00 © 2003 American Chemical Society Published on Web 08/23/2003

1734

Biomacromolecules, Vol. 4, No. 6, 2003

certain gel properties above Ttr. The thermo-responsive behavior can be enhanced with accelerated swelling and complemented by incorporation of matrix polymers and growth factors.2,15-17 The addition of poly(ethyleneglycol) (PEG) as a second component to a thermo-responsive hydrogel was of interest, because PEG is known for its inert behavior toward biosystems in general and to protein adsorption in particular.18,19 We expected an additional effect for the release of the embedded biomacromolecules, when the temperature is decreased below Ttr, which subsequently would trigger the cell detachment. Recently, we were able to present data on hydrogels based on graft copolymers of either NiPAAm or N,N-diethylacrylamide (DEAAm) with PEG. The use of low-pressure argon plasma as immobilization and cross-linking step has been proved to be a versatile tool, which is nearly independent of the applied hydrogel forming polymer and the used (polymeric) substrate.20-23 The surface-immobilized hydrogels showed a thermo-reversible swelling and collapsing, and the transition temperature Ttr could be adjusted to 35-38 °C in deionized water, when the PEG content was kept at 15-20 wt. %.24,25 Immersing the hydrogels in cell culture medium or phosphate buffer saline solution (PBS) shifted Ttr slightly to 32-37 °C.26 These hydrogels have been selected here to study their cell adhesion and detachment properties as a function of temperature. Experimental Section Materials. The synthesis of the polymers and its characterization have been reported elsewhere.27,28 The hydrogel coated cell cultivation carriers have been prepared by lowpressure argon plasma immobilization and cross-linking. The experimental details of the plasma process and the characterization of the surface-immobilized hydrogels are also given in a separate paper.24,25 Deionized (DI) water was obtained from a Barnstead Easypure RF system. The phosphate buffer saline solution (PBS) was prepared by dissolving a tablet (Aldrich) in DI water, RPMI was purchased from Sigma and FCS from Biochrom. RPMI and fetal calve serum (FCS) were mixed prior to the measurement or cell cultivation, respectively. UV/vis Turbidity Measurements. UV-vis spectra were obtained from a Varian Cary 100. The polymers were dissolved in deionized water, PBS solution, or RPMI medium + 10% FCS with cpolymer ) 1 mg/mL. Measurements were started after 5 min temperature equilibration. The transmittance at two wavelengths, 400 and 600 nm, was evaluated. For the measurements in RPMI, only the transmittance at 600 nm was evaluated because of its absorption band below 600 nm. The LCST was determined as the inflection point of the transmittance vs temperature curve. Ellipsometry. A M-44 ellipsometer (Woollam Co. Inc., USA) with a rotating analyzer and a detector array of 44 wavelengths between 428 and 763 nm was used. The instrument is equipped with an automated goniometer, a horizontally mounted sample stage, and 75 W xenon short arc lamp. Radiation below 400 nm was filtered off by a WG 360 Schott glass filter before it reached the polarizer. After reflection from the sample and passing through the analyzer,

Schmaljohann et al.

Figure 1. Synthesis, chemical structure, and notation of the graft copolymers.

the light was dispersed onto a detector array, which permits to perform fast measurements (in some seconds) simultaneously at 44 wavelengths. For modeling, a five layer system (Figure 3) was assumed. The optical constants of Si and SiO2 were taken from literature.29,30 The optical properties of the PTFE-like fluoropolymer substrate layer were estimated from a series of 20 immobilized layers. The values yield in a Cauchy function with a refractive index of n ) 1.359 at 632.8 nm. For every substrate, the exact thickness of the PTFElike layer were estimated. For the swollen layers of hydrogels, two different types of modeling were tested. The consideration as one homogeneous layer makes it possible to fit the thickness and the effective optical constants of the mixed layer. Although the thickness is reasonable, the refractive index of the completely collapsed layer does not fit the dependence on the wavelength. An effective medium approximation with polymer as Cauchy function (n ) 1.550 at 632.8 nm) and water gives information about the layer thickness and the composition of the mixed layer.31,32 The use of this model for the mixed layer up to water content over 90% seems to be doubtful. The thicknesses for both models correspond within the error of 2 nm. The measurements in dry state were done at 22 °C and 40-50% relative humidity with angles of incident of 65°, 70°, and 75°. For the temperature-dependent swelling experiments, the samples were placed in a home-built cell with 74.8° as angle of incident.29 Temperature Scan of the Ellipsometry Measurements. The sample was immersed into the flow-cell containing deionized water or PBS solution at 23 °C. Then, the flowcell was heated at a rate of ca. 0.25 K/min, and the ellipsometry data was collected continuously. A rate of -0.25 K/min was applied for the cooling curve. The LGTT was determined as the inflection point of the film thickness vs temperature curve. Cell Cultivation. L929 mouse fibroblasts were obtained from DSMZ, Braunschweig, Germany and cultivated in RPMI medium containing 10% fetal calf serum (FCS) and antibiotics. Cells grow as a monolayer and were passaged when they reached confluence using trypsin (0,25% w/v in PBS). All media and supplements were optained from Biochrom, Berlin, Germany. Cells were normally cultivated at 37 °C and 5% CO2. The temperature-controlled cell detachment was monitored with a Zeiss Cell Observer, a special controlled incubator mounted on a heated scanning optical microscope table. The heating of the table and the air stream was adjusted manually. The air stream was supplemented with 5% CO2. The

Thermo-Responsive PNiPAAm-g-PEG Films

Biomacromolecules, Vol. 4, No. 6, 2003 1735

Figure 2. UV/vis transmittance vs temperature at 1 mg/mL for (a) P1, (b) P2, (c) P3, and (d) P4.

Figure 3. Layer system of the surface-immobilized hydrogel.

microscope was connected to a digital camera (Zeiss AxioCam Color) in order to record the cell behavior in the incubator during the experimental course. Images were taken every 30 s and subsequently analyzed with the software AxioVision (Zeiss). Results and Discussion Thermo-Reversible Aggregation of the Soluble Polymer Precursor in PBS and RPMI. A series of graft copolymers have been synthesized and characterized. Particular interest was laid on the lower critical solution temperature (LCST), which correlates to the lower gel transition temperature of the corresponding hydrogels. The graft copolymers served as model compounds to identify the desired composition. The same polymers were also later applied for the plasma immobilization. Figure 1 shows the synthetic scheme and the chemical structure of the four selected polymers. Polymers P1 and P2 contain NiPAAm as the thermoresponsive monomer, whereas P3 and P4 have DEAAm as

the thermo-responsive monomer. Both homopolymers (PNiPAAm and PDEAAm) have a LCST of 32-33 °C, and the addition of a polar comonomer like ethyleneglycol renders an increase in the transition temperature. Thus, there is an upper limit, how much PEG can be added to the system while still keeping the LCST below 37 °C. A detailed discussion of the thermo-reversible properties of the soluble polymers in deionized (DI) water has been given elsewhere,27,28 but it has to be noted that certain errors in the assigned composition are possible because of the signal overlap in the NMR spectra. Nevertheless, it was found that the overall PEG content should not exceed 20 wt. % for keeping the transition temperature in the range of 37 °C. This section investigates the changes occurring when DI water is replaced by PBS solution or cell cultivation medium RPMI + 10% FCS.33 Fortunately, the addition of electrolytes to a solution of a LCST polymer decreases the observed transition temperature because of the ionic interactions, which allows one to add a higher PEG content to the reaction mixture without exceeding 37 °C with the LCST.34 Figure 2 demonstrates this effect of the electrolytes for the four polymers. Figure 2 indicates that in all cases the shift toward a lower LCST ranges from 2 to 4 K. Furthermore, there is no significant difference between PBS and RPMI, except for P1 with ∆PBS/RPMI ) 1.7 K. Because of the small differences between the transition in PBS or in RPMI, respectively, one can conclude that the addition of the electrolytes is the major

1736

Biomacromolecules, Vol. 4, No. 6, 2003

Schmaljohann et al.

Table 1. LCST Determination via UV/vis Turbidity Measurements of P1-P4a polymer

DI water

PBS solution

RPMI + 10% FCS

P1 P2 P3 P4

40.7 °C 35.1 °C 38.8 °C 37.7 °C

38.0 °C 34.4 °C 34.3 °C 34.4 °C

36.3 °C 34.3 °C 33.7 °C 34.4 °C

a

Data from 400 and 600 nm evaluated, LCST ) inflection point.

effect in this particular system to cause the shift, whereas the further addition of glucose, amino acids, and other minor components (as in RPMI) does not have any substantial impact on the position of the LCST. A change in pH or the addition of other solvents would also shift the LCST. The effect of salt addition on the position of the LCST has also been explored by other groups; in all cases, a downshift in the LCST was observed after salt addition.13,34 Table 1 summarizes the results of the LCST determination. The fourth column is important, when studying biological systems, and in all cases, the LCST is below 37 °C. (Plasma immobilization will further reduce the transition temperature.25) Based on these results, the polymers P1-P4 were selected for the plasma immobilization and cross-linking procedure. Preparation and Thermo-Responsive Properties of the Surface-Immobilized Hydrogels. Si wafer or glass slides have been coated with a ∼50 nm layer of a PTFE-like fluoropolymer with a very smooth surface.25 Subsequently, the hydrogels have been prepared by low-pressure argon plasma treatment of a 10-20 nm thick layer of the spincoated precursor polymer. After plasma treatment, the glass slides were rinsed with chloroform in order to remove all nonimmobilized polymer. The layer system is depicted in Figure 3. It could be shown by ATR-FTIR spectroscopy that the plasma treatment does not change the monomer composition or the functionality. Thus, the data of the polymer solution and that of the surface-immobilized hydrogels should be comparable. The swelling and collapsing behavior of the hydrogel in DI water as a function of temperature is given in Figure 4. It is noteworthy that the swelling and collapsing in Figure 4 is obtained under dynamic conditions with a heating/ cooling rate of ca. (0.25 K/min. The initial swelling as well as the swelling equilibrium deviate from this situation. After the first heating, the hydrogels have a thermal history, which gives reproducible data for the dynamic scans.25 The temperature gradient applied in the cell detachment experiment comes close that of Figure 4. Table 2 summarizes the results of the dynamic swelling studies. The absolute film thicknesses as well as the degrees of swelling are effected by the immobilization and cross-linking efficiency. Even though the plasma treatment was the same for each polymer sample, there are some differences, especially P4 deviates from other samples. It should be mentioned further that even in the collapsed state the hydrogels have a film thickness of 30-50 nm, which translates to a swelling ratio of 2 or larger. This means that there is still a certain degree of hydrogel properties, which enables the nutrients to permeate through the gel. This is a

Figure 4. Film thickness vs temperature during the 2nd heating/ cooling cycle in DI water for (a) P1 and P2 and (b) P3 and P4. Table 2. Lower Gel Transition Temperature (LGTT) and Degree of Swelling (DS) for P1-P4 in Deionized Water film film LGTTa at LGTTa at thickness DSb at thickness DSb at hydrogel heating cooling at 25 °C 25 °C at 42 °C 42 °C P1 P2 P3 P4 a

37 °C 35 °C ∼35 °C ∼34 °C

35 °C 33 °C ∼32 °C ∼33 °C

93 nm 77 nm 61 nm 84 nm

6.1 6.8 3.9 12.8

30 nm 39 nm 35 nm 52 nm

2.0 3.5 2.2 8.0

LGTT ) inflection point. b DS ) dswollen/ddry.

major advantage compared to hydrogels consisting only of PNiPAAm or other thermo-responsive polymers, which will act more like a hydrophobic substrate above the LCST without any hydrogel properties. It could have been shown by inverse contact angle measurements that the change in surface polarity for hydrogels based on P1-P4 is still observable but less pronounced compared to hydrogels based on NiPAAm homopolymers.24 In this sense, the addition of PEG to the system also allows one to fine-tune the surface polarity. The PNiPAAm containing hydrogels (P1, P2) show a sharper transition compared to the PDEAAm containing hydrogel (P3, P4). This should be favorable when studying the response of cells toward changes in the surface properties. P1 was then selected for cell cultivation studies, because it showed the larger difference between swollen and collapsed state. Thus, the swelling behavior of the surface-immobilized

Thermo-Responsive PNiPAAm-g-PEG Films

Figure 5. Film thickness vs temperature during the 2nd heating/ cooling cycle for P1 in DI water and PBS solution, respectively.

hydrogel P1 was also studied in PBS solution. (Ellipsometry in RPMI was not possible because of the absorption characteristics of this medium.) As shown in Figure 5, the transition temperatures of P1 in PBS solution are 34 °C for the heating cycle and 32 °C for the cooling cycle, respectively. The addition of electrolyte to the system gives another 3 K downshift in the transition temperature similar to the results of the soluble polymer samples. This allows us to seed and cultivate cells at 37 °C in the collapsed state. Furthermore, the degree of swelling in the collapsed state is larger compared to that for DI water (PBS: DS ) 2.4; DI water: DS ) 2.0; both at 42 °C), whereas there is no difference in the swollen state. The higher degree of swelling in PBS indicates that there are some attractive interaction between the salt, PEG and water, which reduces the extrusion of water above the transition temperature.

Biomacromolecules, Vol. 4, No. 6, 2003 1737

Cell Cultivation and Detachment via Thermal Stimulus. L929 mouse fibroblasts were cultivated on hydrogel P1 for several days in RPMI medium. The cells adhere well, spread, and proliferate on the surface at 37 °C, which showed that the applied hydrogels are suitable as cell cultivation carriers. The gels were confirmed to be nontoxic and did not show impaired cell adhesion because of their PEG contents (data not shown). In a second experiment, the cell detachment was studied upon a thermal stimulus. After cultivation overnight, cells were placed at 37 °C into the Zeiss Cell Observer, and the temperature was reduced at a rate of ca. 0.1 K/min to follow the response of the cell toward the changes in film thickness and surface polarity. Already 1-2 K below 37 °C, the cells round up. Furthermore, after decreasing the temperature by 3 K to 34 °C, complete cell detachment was observed. Figure 6 shows micrographs of the cells up to the point, when the cells finally go off the substrate. Figure 6 also indicates how the changes in cells are correlated to the changes of the hydrogel film. Long before the swelling is completed, the cells already detach from the surface. We attribute the immediate response of the cells toward the temperature to the substantial amount of PEG in the hydrogel. Upon swelling, the PEG gains further mobility, which allows it to diffuse to the surface. The cell detachment occurred within 20 min. Thus, we prepared a fast responding hydrogel, which again can be attributed to the presence of the hydrophilic PEG. Okano et al. also reported that the incorporation of ionic groups or PEG accelerated the cell detachment of NiPAAm containing hydrogels.4 Nevertheless, our results are in contrast to those reported by Okano, where they state that incorporation of 0.5 wt. % of PEG dramatically decreases the cell adhesion. The hydrogels, discussed in this

Figure 6. Micrographs of mouse fibroblast during cell detachment during decreasing temperature and correlation to the film thickness of the hydrogel.

1738

Biomacromolecules, Vol. 4, No. 6, 2003

Schmaljohann et al.

Figure 7. Micrographs of mouse fibroblast on P1 hydrogel (left) and PTFE-like control (right), and at 37 °C (top) and 31 °C (bottom), respectively.

paper, contain ca. 19 wt. % PEG and they show at 37 °C good cell adhesion behavior even without any precoating with proteins of the extracellular matrix. However, Okano et al. used bovine aortic endothelial cells in culture experiments which may behave different as compared with the L929 mouse fibroblasts applied in our study. The cell number on the PEG-containing hydrogel films was found to be roughly similar as on the polystyrene (PS) culture dish control. The growth kinetics and vitality of the cells will be presented in a separate publication. This unusual behavior at a relatively high PEG content can possibly be attributed to the graft architecture of the precursor polymer, which exhibits a strong interaction of the ethyleneglycol units with the NiPAAm units above the transition temperature and in a less pronounced fashion also with the DEAAm units above Ttr.27,28 The micrographs in Figure 7 compare the behavior of the cells on two substrates (P1 hydrogel and PTFE-like polymer as control) at the two temperatures (above and below Ttr of P1 hydrogel). During the polymer immobilization, one part of the glass slide with the PTFE-like layer (compare Figure 3) was covered in order to avoid plasma exposure. When rinsing the glass slide with chloroform after plasma treatment, the non-crosslinked polymer hydrogel was removed completely from this covered area of the sample. This region on the glass slide was taken as the control in Figure 7. The fibroblasts continue to adhere well on the uncoated fluoropolymer even upon reducing the temperature to 31 °C, whereas they are efficiently removed from the thermoresponsive hydrogel at this temperature. After temperature reduction to 31 °C, the cells on the control show light shrinking in shape, but this completely reversed after going back 37 °C. Conclusion The control of cell adhesion and the potential to detach cells from a surface is an important issue in biomedical

science. The use of stimuli-responsive hydrogels is very attractive because of the balance of mechanical support and softness for reorganization and because of the mobility of water, proteins, and nutrients within the hydrogels. Further, as a major advantage, these properties can be tuned when the transition point is crossed. The combination of NiPAAm as thermo-responsive units and EG giving reduced cellsurface-interaction turned out to be advantageous for the aim of controlling cell adhesion. The thermo-responsive NiPAAm changes the surface properties upon going through the volume phase transition, and it provides a more hydrophobic surface above the transition, allowing for cell attachment, spreading, and proliferation, and a hydrophilic surface below Ttr, which initiates the cell detachment. The PEG provides properties, which enhance and accelerate the detachment of the cell layers. Furthermore, the PEG makes the hydrogels hydrophilic enough that they stay partially swollen even above the transition. Thus, the transition, which is accompanying the cell detachment, goes from partially swollen to completely swollen state, when lowering the temperature. The transition temperature of the hydrogels in comparison to the soluble polymers were always shifted 2-4 K toward lower temperatures and also the addition of electrolytes as present in cell cultivation media causes a downshift of Ttr. The complete cell detachment could be achieved within 20 min and reduction of the temperature from 37 to 34 °C. The NiPAAm containing hydrogel P1 was selected, because the transition is much sharper compared to DEAAm containing gels. The fast cell detachment after a small temperature change of only 3 K is advantageous to minimize cell response to the temperature change. It has been shown that the cell detachment works well in cell culture medium which offers an alternative to the enzymatic cell detachment using, e.g., trypsin. The fast and sharp response may be attributed to the fact that the NiPAAM units are interacting with the EG units above the transition temperature, whereas below Ttr, the complexation to the PEG chains is replaced by solubilization with water.

Thermo-Responsive PNiPAAm-g-PEG Films

Acknowledgment. The authors thank Grit Eberth for help with the ellipsometry measurements. Supporting Information Available. Details of the synthesis of the polymers and the plasma immobilization procedure. This material is available free of charge via the Internet at http://pubs.acs.org. References and Notes (1) Okano, T., Ed.; Biorelated Polymers and Gels; Academic Press: San Diego, CA, 1998. (2) Kwon, O. H.; Kikuchi, A.; Yamato, M.; Okano, T. Biomaterials 2003, 24, 1223-1232. (3) Yamada, N.; Okano, T.; Sakai, H.; Karikusa, F.; Sawasaki, Y.; Sakurai, Y. Macromol. Rap. Comm. 1990, 11, 571-576. (4) Ebara, M.; Yamato, M.; Hirose, M.; Aoyagi, T.; Kikuchi, A.; Sakai, K.; Okano T. Biomacromolecules 2003, 4, 344-349. (5) Ratner, B. D.; Cheng, X.; Wang, Y.; Hanein, Y.; Bo¨hringer, K. F. Polym. Prepr. 2003, 44, 198-199. (6) Yamato, M.; Konno, C.; Kushida, A.; Hirose, M.; Utsumi, M.; Kikuchi, A.; Okano, T. Biomaterials 2000, 21, 981-986. (7) McCormick, C. L., Ed.; Stimuli-responsiVe Water-soluble and Amphiphilic Polymers; ACS Symposium Series No. 780; American Chemical Society: Washington, DC, 1999. (8) Ramkissoon-Ganorkar, C.; Liu, F.; Baudys, M.; Kim, S. W. J. Controlled Release 1999, 59, 287-298. (9) Yu, H.; Grainger, D. W. J. Appl. Polym. Sci 1993, 49, 1553-1563. (10) Kurisawa, M.; Yokoyama, M.; Okano, T. J. Controlled Release 2000, 69, 127-137. (11) Kyriakides, T. R.; Cheung, C. Y.; Murthy, N.; Bornstein, P.; Stayton, P. S.; Hoffman, A. S. J. Controlled Release 2002, 78, 295-303. (12) Ding, Z.; Fong, R. B.; Long, C. J.; Stayton, P. S.; Hoffman, A. S. Nature 2001, 411, 59-62. (13) Schild, H. G. Prog. Polym. Sci. 1992, 17, 163. (14) Okano, T.; Yamada, N.; Okuhara, M.; Sakai, H.; Sakurai, Y. Biomaterials 1995, 16, 297-303. (15) Van Recum, H.; Okano, T.; Wan Kim, S. J. Controlled Release 1998, 55, 121-130.

Biomacromolecules, Vol. 4, No. 6, 2003 1739 (16) Shimizu, T.; Yamato, M.; Akutsu, T.; Shibata, T.; Isoi, Y.; Kikuchi, A.; Umezu, M.; Okano, T. J. Biomed. Mater. Res. 2002, 60, 110117. (17) Harimoto, M.; Yamato, M.; Hirose, M.; Takahashi, C.; Isoi, Y.; Kikuchi, A.; Okano, T. J. Biomed. Mater. Res. 2002, 62, 464-470. (18) Harder, P.; Grunze, M.; Dahint, R.; Whitesides, G. M.; Laibinis, P. E. J. Phys. Chem. B 1998, 102, 426-436. (19) Harris, J. M., Ed.; Poly(ethyleneglycol) chemistry: biotechnical and biomedical applications; Plenum Press: New York, 1992. (20) Nitschke, M.; Menning, A.; Werner, C. J. Biomed. Mater. Res. 2000, 50, 340-343. (21) Terlingen, J. G. A.; Feijen, J.; Hoffman, A. S. J. Colloid Interface Sci. 1993, 155, 55-65. (22) Terlingen, J. G. A.; Brenneisen, L. M.; Super, H. T. J.; Pijpers, A. P.; Hoffman, A. S.; Feijen, J. J. Biomater. Sci. Polymer Ed. 1993, 4, 165. (23) Sheu, M. S.; Hoffman, A. S.; Feijen, J. J. Adhesion Sci. Technol. 1992, 6, 995. (24) Schmaljohann, D.; Beyerlein, D.; Nitschke, M.; Zschoche, S.; Werner, C. PMSE Prepr. 2003, 88, 551-552. (25) Schmaljohann, D.; Beyerlein, D.; Nitschke, M.; Werner, C. Langmuir submitted. (26) Schmaljohann, D.; Nitschke, M.; Beyerlein, D.; Werner, C. Polym. Prepr. 2003, 44, 196-197. (27) Schmaljohann, D.; Gramm, S. Polym. Prepr. 2002, 43, 758. (28) Gramm, S.; Komber, H.; Ha¨ussler, L.; Tarek El Tahan, N.; Schmaljohann, D. Macromolecules submitted. (29) Werner, C.; Eichhorn, K.-J.; Grundke, K.; Simon, F.; Gra¨hlert, W.; Jacobasch, H.-J. Colloids Surf. Part A: Physicochem. Eng. Aspects 1999, 156, 3. (30) Woollam, J. A. User Manual VASE and M-44 Ellipsometers, WVASE 32; J. A. Woollam Co., Inc.: Lincoln, Nebraska. (31) Mathe, G.; Gege, C.; Neumaier, K. R.; Schmidt, R. R.; Sackmann, E. Langmuir 2000, 16, 3835. (32) Murphy, E. F.; Lu, J. R.; Lewis, A. L.; Brewer, J.; Russell, J.; Stratford, P. Macromolecules 2000, 33, 4545. (33) In the following, the cell cultivation medium RPMI + 10% FCS (fetal calf serum) is denoted as RPMI. (34) Zhang, X.; Hu, Z.; Li, Y. J. Appl. Polym. Sci. 1997, 63, 1851-1856.

BM034160P