ARTICLE IN PRESS
Biomaterials 27 (2006) 68–78 www.elsevier.com/locate/biomaterials
Protein bonding on biodegradable poly(L-lactide-co-caprolactone) membrane for esophageal tissue engineering Yabin Zhua, Kerm Sin Chianb, Mary B. Chan-Parkb,c,, Priyadarshini S. Mhaisalkara, Buddy D. Ratnerd a Biomedical Engineering Research Centre, Nanyang Technological University, Singapore 639798, Singapore School of Mechanical and Aerospace Engineering, Nanyang Technological University, Singapore 639798, Singapore c School of Chemical and Biomedical Engineering, Nanyang Technological University, Singapore 639798, Singapore d Department of Bioengineering, University of Washington, Bagley 461, Seattle, WA 98195, USA
b
Received 23 November 2004; accepted 27 May 2005 Available online 11 July 2005
Abstract A biodegradable and flexible poly(L-lactide-co-caprolactone) (PLLC) copolymer was synthesized and surface modification has been performed aiming at application as a scaffold in esophageal tissue engineering. The PLLC membrane surface was aminolyzed by 1,6-hexanediamine to introduce free amino groups. Using these amino groups as bridges, fibronectin and collagen were subsequently bonded with glutaraldehyde as a coupling agent. The presence of free amino groups on the aminolyzed PLLC surface was quantified using fluorescamine analysis method, which revealed that the surface NH2 density increased and eventually saturated with increasing 1,6-hexanediamine concentration or reaction time. X-ray photoelectron spectroscopy (XPS) confirmed the presence of both proteins separately on the modified PLLC surface. Water contact angle measurements evaluate the wettability of modified and unmodified PLLC surfaces. Protein-bonded surface presented more hydrophilic and homogeneous, yet PLLC can also adsorb some protein molecules. In vitro long-term (12 d) culture of porcine esophageal cells proved that fibronectin- and collagen-modified PLLC surface (denoted PLLC–Fn and PLLC–Col, respectively) can more effectively support the growth of smooth muscle cells and epithelial cells; both modified and unmodified PLLC support fibroblasts growth. Mitochondrial activity assay and cell morphology observation indicate that the PLLC–Fn surface is more favorable to epithelium regeneration than PLLC–Col. These culture results provide much valuable information for our subsequent research on the construction of artificial scaffolds with esophageal function. Fibronectin-integrated PLLC will be a good candidate scaffold to support the growth of all types of esophageal cells. r 2005 Elsevier Ltd. All rights reserved. Keywords: Poly(L-lactide-co-caprolactone) copolymer; Fibronectin; Collagen; Surface modification; Esophageal tissue engineering
1. Introduction Esophageal cancer is one of the most difficult gastrointestinal malignancies to treat because of the poor regenerative capacity of the esophagus tissue [1]. Excision of a substantial portion followed either by Corresponding author. School of Mechanical and Aerospace Engineering, Nanyang Technological University, Singapore 639798. Tel.: +65 6790 6064; fax: +65 6792 4062. E-mail address:
[email protected] (M.B. Chan-Park).
0142-9612/$ - see front matter r 2005 Elsevier Ltd. All rights reserved. doi:10.1016/j.biomaterials.2005.05.069
reattachment of the shortened esophagus to stomach or replacement of the excised portion with some form of substitute is normally clinical treatment. Many kinds of non-biodegradable tubes made of rubber, Tteflon and the like have been investigated as possible esophagus substitutes, but all these tubes tend to present problems of anastomotic leakage or stenosis. Esophageal tissue engineering is deemed to be a potential way for replacement of lost or malignant esophagus. Thus far, some scaffolds have been studied for substitute esophagus. For example, Sato and Natsume et al. have
ARTICLE IN PRESS Y. Zhu et al. / Biomaterials 27 (2006) 68–78
proposed to use poly(glycolic acid) (PGA) or silicone with collagen coating as a tubular scaffold [2–4]. Badylak has investigated small intestinal submucosa (SIS) as scaffolds for replacement of tubular organs such as esophagus and large-diameter vascular grafts [5–7]. However, none has overcome the problems of leakage, extrusion and stenosis in a long-term implant. The esophagus is a muscular canal extending from pharynx to stomach. Transporting food and water from mouth to stomach is the main function of the esophagus. Therefore, an effective artificial esophagus must be implantable without rejection and must maintain a luminal structure without stenosis in vivo [3]. Moreover, it must be soft and elastomeric in order to maintain the transporting function. Three kinds of cells, i.e., smooth muscle cell, fibroblast and epithelial cell, constitute the four layers of this tissue: mucosa, submucosa, muscularis externa and adventitia. An ideal tissue-engineered esophagus substitute would reproduce these layers using these cell types, which calls for a tissue scaffold which is elastomeric and broadly cytocompatible. A random copolymer of L-lactic acid (LLA) and caprolactone (CL), i.e. PLLC, is a good candidate scaffold material because these materials have been widely used in biodegradable scaffolds, and the copolymer manufactured in our laboratory is elastomeric at room temperature due to its low glass transition temperature (T g , 9 1C). However, PLLC generally has poor cytocompatibility due to its hydrophobicity and lack of bioactive domains. Surface modification to introduce recognition factors is necessary to improve the interaction between cells and substrate membrane, because the initial response of cells to the implanted materials mostly depends on the surface properties. Due to the lack of chemical functionalities, it is usually difficult for these synthetic polyesters to favor polymer–protein interaction. The method of aminolysis between diamine and polyester matrix introduces functional amino groups, through which protein can be bonded onto polymer surface. This method has previously been shown to promote the growth of vascular endothelial cells and chondrocytes on polyester substrates like poly(L-lactic acid), polycaprolactone and polyurethane [8–11]. In this paper, the same method is used to bond adhesion proteins like fibronectin (Fn) and collagen (Col) onto PLLC membranes to promote the adhesion and growth of all three types of esophageal cells, i.e. smooth muscle cell (SMC), fibroblast and epithelial cell (EC). PLLC surface was first aminolyzed with 1,6-hexanediamine to introduce free amino groups. Using these amino groups as active sites, Fn or Col was separately bonded with glutaraldehyde as a coupling agent. The cytocompatibility to SMC, fibroblast and EC was evaluated for esophagus tissue engineering scaffold.
69
2. Materials and methods 2.1. Materials Poly(L-lactide-co-caprolactone) copolymer was polymerized in our laboratory according to a published procedure [12]. The monomers (LLA and CL) were added simultaneously into a reactor in the presence of stannous 2-ethylhexanoate at 160 1C for 3 h, and then reacted at 140 1C for 24 h. Nuclear magnetic resonance spectrometry ( 1H NMR) determined the ratio of LLA:CL of the copolymer to be 84:16. T g was measured to be 9 1C by differential scanning calorimetry (DSC). Collagen (type I from calf skin) was purchased from Sigma and human plasma fibronectin was purchased from Invitrogen Co. (GibcoTM). Collagen was dissolved in 0.1 N acetic acid to produce a collagen solution with pH 3.5, and Fn was dissolved in distilled water at 1.0 mg/ml (pH 11.5). CellTiter 96s aqueous cell proliferation assay kit was purchased from Promega Co. for cell mitochondrial activity measurement. All other items were purchased from SigmaAldrich Co. and used as received if not otherwise stated. All water used in the experiment was doubly distilled. 2.2. Aminolysis and protein bonding on PLLC membrane PLLC membrane was prepared by casting 0.06 g/ml PLLC/ 1,4-dioxane solution onto a glass mold and evaporating the solvent at 40 1C for 24 h. The membrane was subsequently immersed in 1,6-hexanediamine/propanol solution for predetermined periods at room temperature (21 1C) and then rinsed with large quantity of water to remove free 1,6-hexanediamine. Aminolysis is the reaction between amine of the 1,6-hexenediamine and ester of the PLLC surface. In order to avoid protein aggregation, two-step glutaraldehyde cross-linking was employed. Firstly, the aminolyzed PLLC membrane was immersed in 1 wt% glutaraldehyde solution at room temperature for 24 h and then rinsed with a large amount of water to remove free glutaraldehyde. Secondly, the membrane was incubated in 1 mg/ml Col or Fn solution for 24 h at 2–4 1C. The Col-bonded membrane (denoted PLLC–Col) was first rinsed with 0.1 N acetic acid solution and then with water for 24 h to remove free Col. The Fnbonded membrane (denoted PLLC–Fn) was rinsed with water to remove free Fn. 2.3. Determination of the amino groups on PLLC surface The fluorescamine method [13] was employed to quantify the density of NH2 groups after PLLC was aminolyzed with 1,6-hexanediamine. A 2 cm2 piece of aminolyzed PLLC membrane was immersed in 1 ml fluorescamine (4-phenylspiro[furan-2(3 H, 1 phthalan)-3,30 dione])/acetone solution (1 mg/ ml) for 5 min to dissolve the membrane. 1 ml acetone and 1 ml propanol were then added to make the final volume 3 ml. For every sample, 200 ml of this solution was loaded in duplicate into a white 96-well plate for fluorescence intensity measurement (excitation wavelength 395 nm, maximum emission wavelength 480 nm) on a microplate reader (TECAN spectrophotometer). A calibration curve was obtained with 1,6hexanediamine in the same solution.
ARTICLE IN PRESS 70
Y. Zhu et al. / Biomaterials 27 (2006) 68–78
2.4. Contact angle measurement The water contact angle (WCA) of the modified and unmodified PLLC membranes was measured by the sessile drop method using a Contact Angle System OCA20 (Germany). Five independent determinations at different sites of one sample were averaged for each membrane type. 2.5. X-ray photoelectron spectroscopy (XPS) measurement XPS spectra were recorded on an AXIS ULTRA spectrometer (Kratos Analytical Ltd., Surface Analysis Product Group, England) employing the excitation of a non-monochromatized Ka X-ray source (1486.6 eV). The operating pressure in the analysis chamber was maintained at 3.0 10–9 Torr or lower during the measurements. The corelevel spectra were obtained at a photoelectron take-off angle of 901 (measured with respect to the sample surface). The charging shift was referred to the C1 s line emitted from saturated hydrocarbon at a binding energy (BE) of 285 eV. Wide scan (0–1000 eV BE) and high-resolution (C1s, N1s, O1s, Si2p) spectra were acquired. C1s spectra were best fitted with five peaks corresponding to: C–C (C0, BE ¼ 285.0 eV, reference), C–N (C1, BE ¼ 285.9 eV), C–O (C2, BE ¼ 286.9 eV), N–CQO (C3, BE ¼ 287.7 eV) and O–CQO (C4, BE ¼ 289.1 eV). 2.6. Porcine esophageal cell culture All three types of cells important to esophageal tissue, i.e., SMC, fibroblast and EC, were digested from porcine esophagus. For SMC culture, two digestion steps were employed. Firstly, the muscle layer was minced and digested in a solution consisting of 5 ml Dulbecco’s Modification of Eagle’s Medium (DMEM), 0.625 ml fetal bovine serum (FBS), 3.2 mg collagenase (type I), 100 U penicillin/100 mg streptomycin/0.25 mg fungizone/ml (P/S/F) at 37 1C for 20 min [14]. The digested solution was centrifuged at 500 rpm for 1 min and washed with DMEM containing P/S/F, and then resuspended in a second digestion solution: 5 ml DMEM, 0.625 ml FBS, 3.2 mg collagenase type I, 1.2 mg elastase type III and P/S/F at 37 1C for 2 h. The digestion procedure was stopped by diluting with an equal volume of DMEM containing 20% FBS (v/v, same as following except statement). This digested solution was centrifuged at 1000 rpm for 5 min and the supernatant was completely removed. The cells were resuspended in the culture medium containing 10 ml DMEM, 20% FBS, 2 ml/ml kanamycin and P/S/F, seeded in 75 cm2 T-flask (IWAKI) and cultured in an incubator (SANYO CO2 Incubator, 37 1C, 5% CO2). After incubation for 24 h, the medium was changed and changed thereafter every 2 d. After a confluent cell layer was formed (usually 1 week), the cells were detached using 0.25 wt% trypsin–ethylene diamine tetraacetic acid (EDTA) in phosphate-buffered saline (PBS, pH 7.4), resuspended in complete medium as described above except using 10% FBS and employed in the following experiments. Immunostaining with anti-smooth muscle a-actin confirmed the SMC obtained by this digestion procedure. Fibroblasts were digested using a simpler method due to their strong extending and attachment capability. Mucosa and
submucosa cubes were washed three times with PBS containing 1000 U/ml penicillin, 1 mg/ml kanamycin and 25 mg/ml amphotericin (P/K/A) and centrifuged at 1000 rpm for 5 min. The resulting tissue pellets were seeded in a culture plate with DMEM containing 10% FBS and P/S/F. After 2 d, fibroblasts had extended from the pellets and attached to the culture plate. These primary fibroblasts were collected and subcultured for further applications. For EC culture, the mucosal-submucosal layers were separated from the underlying muscularis externa and rinsed in PBS supplemented with 100 U/ml penicillin, 100 mg/ml streptomycin, 100 mg/ml kanamycin and 0.25 mg/ml amphotericin B. Small pieces of the tissue were then incubated in dispase solution (10 U/ml) at 4 1C overnight. The epithelial layer was then separated from the underlying lamina propia and further incubated in 10 ml trypsin–EDTA (0.5 wt%) at 37 1C for 30 min. The suspension was mixed with an equal volume of medium containing 10% FBS for inactivation of the trypsin. The cell pellet was strained through a cell strainer and then centrifuged at 1000 rpm for 5 min. The epithelial cells thus obtained were cultured in a T-flask for further applications. The culture medium was high-glucose-DMEM supplemented with 10% FBS, epidermal growth factor (10 ng/ml), insulin (10 u/ml) and antibiotics P/S/F [15]. Immunostaining with anti-keratin-13 confirmed the epithelial cells obtained by this digestion procedure. All these cells were used from fourth to sixth passages to measure the cytocompatibility of modified and unmodified PLLC membrane. In order to observe the different response of unmodified and protein-modified PLLC surface, cells were characterized by attachment ratio, methylthiazoletetrazolium (MTT) activity, proliferation (DNA assay) and morphology observation. Cell attachment ratio was determined at 24 h with a seeding density of 12 104 cells/cm2. Before harvesting the adhered cells by trypsinization, three gentle washings with PBS were performed. The cells were counted using a haemocytometer. The attachment ratio was defined as the percentage of the number of the cells adhered on the sample to the number adhered on a tissue culture polystyrene (TCPS) well. For the morphology observation, the cells were washed with PBS three times and fixed with 2.5 wt% glutaraldehyde solution for 30 min, and then washed with water followed by overnight desiccation in a freeze-dryer (Alpha 1-2, Germany). After gold metallization with a JEOL auto fine coater (JFC-1600), the cells were observed under a JEOL scanning electron microscope (SEM, JSM-5600LV, Japan).
2.7. DNA assay The DNA content was quantitatively analyzed with Hoechst dye H33258 after cells were cultured for some days on modified and unmodified PLLC membranes [16]. Samples were washed with PBS and frozen at 20 1C for storage. For analysis, samples with seeded cells were dehydrated at room temperature in a biological safety cabinet (Gelman, BH class II series), and then digested in 3 mg/ml papain (from papaya latex, Sigma) overnight at 60 1C. H33258 in TRIS-EDTA.2Na (TNE) buffer (0.1 mg/ml) was used to dye the digested solution which was loaded in duplicate in white 96-well plates. Blanks
ARTICLE IN PRESS Y. Zhu et al. / Biomaterials 27 (2006) 68–78
and a series of DNA standards (Calf thymus DNA, Sigma) were also loaded in order to permit calibration of the fluorescence reading. Fluorescence intensity was measured on a microplate reader (TECAN spectrophotometer) at excitation and emission wavelengths of 360 and 465 nm, respectively. A calibration curve was obtained from DNA standard solutions with known concentrations. The DNA concentration is proportional to the cell number because every cell has a fixed DNA content. The DNA concentrations, thus, provide reliable proxies for comparison of cell numbers grown on the modified membranes and control surfaces. 2.8. Mitochondrial function measurement (MTT activity) To examine mitochondrial function, CellTiter 96s aqueous cell proliferation assay kit was used. After cells were cultured on different samples for some time, 20 ml of CellTiter 96s aqueous one solution reagent was added into each well of a 96well assay plate. The plate was then incubated for 4 h at 37 1C in a humidified, 5% CO2 atmosphere. The absorbance at 490 nm was recorded using a 96-well microplate reader. The value was compared against and expressed as the percentage of the data for cells grown on TCPS well. Five parallels for every sample were averaged. 2.9. Statistics Data are expressed as mean7standard deviation. Statistical analysis was performed using MATLAB 7.0 software. Statistical comparisons were made by analysis of variance (ANOVA). The t-test was used for evaluations of differences between groups. P values less than 0.05 were considered to be significant.
3. Results 3.1. Surface modification of PLLC membrane Under appropriate conditions, one amino group of 1,6-hexanediamine can react with an ester group on the PLLC membrane surface to form a covalent bond, while the other amino group is unreacted and free. These free
71
NH2 groups can be used as active sites through which proteins like Fn or Col can be bonded to the surface using glutaraldehyde as a cross-linking agent [8,9]. Glutaraldehyde has been widely used for coupling amine-containing molecules to proteins [17]. However, to avoid aggregation, a two-step procedure was employed in our process. The existence of free amino groups (NH2) on PLLC surface is, therefore, a prerequisite for protein bonding in this modification method. It is important to confirm the existence of amino groups before protein is further introduced. Fluorescamine reacts instantaneously with primary amines at room temperature, producing a fluorophor which emits at 480 nm under 395 nm excitation [13]. The products are highly fluorescent whereas the reagent and its degradation products are non-fluorescent. Herein, it is used as an indicator to confirm and quantify the –NH2 moiety on the aminolyzed PLLC surface. The NH2 density on PLLC membrane surface was influenced by the concentration of 1,6-hexanediamine, aminolyzing time and temperature. Because the Tg of the present PLLC is 9 1C, aminolyzation which should preferably be done above the polymer T g was thus conducted at room temperature (21 1C). Fig. 1a shows that the surfacebound NH2 (expressed in fluorescence units) increased with increasing 1,6-hexanediamine concentration and saturated at high concentration. A similar saturation effect was seen in the reaction-time dependence of the surface-bound NH2 concentration (Fig. 1b). The fluorescence intensity increased rapidly with reaction time, reaching a maximum value at about 8 min, and then saturated or decreased slightly. The decrease at the longer aminolyzing time may be caused by a further reaction of the free amino group on the terminal chain with another ester group, or by the degradation of the superficial layer. Both of these situations will reduce the density of the surface amino group and thus the fluorescence intensity. The aminolyzation reaction parameters adopted for subsequent cell/PLLC interaction experiments (0.06 g/ml 1,6-hexanediamine/propanol, 2 min, 21 1C) were chosen in order to avoid the
18000 20000
Fluorescence unit
Fluorescence unit
16000 14000 12000 10000
18000
16000
14000
8000 12000
6000
(a)
0.00 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 1,6-Hexanediamine concentration (g/ml)
0
(b)
5
10 15 20 25 Aminolyzing time (min)
30
Fig. 1. The fluorescence intensity as a function of (a) 1,6-hexanediamine concentration (for a 2 min reaction) and (b) aminolyzing time (with 0.06 g/ ml 1,6-hexanediamine solution) at 21 1C for PLLC membrane.
ARTICLE IN PRESS Y. Zhu et al. / Biomaterials 27 (2006) 68–78
72
Table 1 Atomic concentration (%) of C1s, N1s, O1s and Si2p for modified and unmodified PLLC membranes. The atomic concentration of fibronectin and collagen were herein used as reference Sample
C1s
N1s
O1s
Si2p
O1s/C1s
PLLC PLLC–NH2 PLLC–Fn PLLC–Col Fn Col
65.57 66.97 66.78 67.14 63.38 67.38
0 0.72 6.45 1.56 13.61 12.73
33.32 31.42 25.53 30.30 22.97 19.89
1.11 0.88 1.24 1.01 0 0
0.50 0.46 0.38 0.45 0.36 0.30
possible damage to the membrane. According to the calibration curve obtained with 1,6-hexanediamine of known concentration, the NH2 density on the PLLC membrane aminolyzed under these conditions was 4.0 108 mol/cm2. The surface alteration after the PLLC membrane was aminolyzed and protein immobilized was also studied by XPS. Table 1 shows the alteration of atomic concentrations and O1s/C1s ratios after PLLC membrane was modified. There is no nitrogen and the O1s/C1s ratio is 0.50 for the control PLLC. With aminolysis, nitrogen appeared on the PLLC membrane, though the concentration was not high (Table 1, PLLC–NH2). The ratio of O1s/C1s decreased from the control 0.50 to 0.46 due to absence of oxygen in 1,6-hexanediamine molecules. Correspondingly, C–N (C1) and N–CQO (C3) peaks in the C1s spectra were displayed (Fig. 2b). The percentage of C1 and C3 is 13.6 and 6.1, respectively (Table 2). The O–CQO (C4) percentage decreased as expected. In contrast, no C1 and C3 peaks appeared on the control PLLC as expected (Table 2). These results confirmed the occurrence of the aminolysis reaction and the existence of the amino group on aminolyzed PLLC membrane surface. After Fn and Col were bonded onto the aminolyzed PLLC surface (PLLC–Fn and PLLC–Col in Tables 1 and 2), these surfaces showed presence of nitrogen and the O1s/ C1s ratios became lower than that of the control PLLC, which is attributed to the lower O1s/C1s ratios of the original fibronectin and collagen (Table 1, Fn and Col). These surfaces also show significant C1 and C3 peaks for both PLLC–Fn and PLLC–Col membranes (Table 2) resulting in the obvious decrease of C2 and C4 peaks. (Fig. 2(c) shows a representative C1s spectrum of PLLC–Fn.) However, it seems that the Fn density is higher than Col density on the modified PLLC surface from the present XPS data using the same procedure and reaction condition. Further fundamental study about bonding efficiency of different proteins would be carried out. 3.2. Surface wettability of the modified PLLC membrane Fn and Col were bonded to separate aminolyzed PLLC samples using glutaraldehyde as a cross-linking
Fig. 2. C1s XPS spectra of the control PLLC (a), aminolyzed PLLC (b) and PLLC–Fn (c). C0, C–C BE ¼ 285.0 eV, reference; C1, C–N BE ¼ 285.9 eV; C2, C–O BE ¼ 286.9 eV; C3, N–CQO BE ¼ 287.7 eV and C4, O–CQO BE ¼ 289.1 eV.
agent. Table 3 shows the surface wettability of unmodified, aminolyzed, and Fn- or Col-bonded PLLC. The WCA measured by the sessile drop method decreased slightly from 83.41 to 78.51 after the membrane was aminolyzed, but decreased substantially after the protein was bonded. That is, the hydrophilicity of the PLLC membrane was increased after it was modified by protein. Dent A. and Aslam M. stated that
ARTICLE IN PRESS Y. Zhu et al. / Biomaterials 27 (2006) 68–78
73
Table 2 C1s percentages of modified and unmodified PLLC membranes Sample
C0 (285.0 eV)
C1 (285.9 eV)
C2 (286.9 eV)
C3 (287.7 eV)
C4 (289.1 eV)
PLLC PLLC–NH2 PLLC–Fn PLLC–Col
48.99 41.54 46.75 45.99
0 13.56 14.77 10.47
25.17 16.67 12.98 17.56
0 6.06 8.69 6.19
25.84 22.17 16.80 19.79
Table 3 Water contact angles of the control and modified PLLC membranes
120
Attachment ratio (%) MTT activity (%)
Water contact angle/degree (Sessile drop)
Control PLLC PLLC–NH2a PLLC–Fna PLLC–Cola PLLC–ad-Fnb PLLC–ad-Colb
83.471.1 78.571.4 66.572.3 69.271.3 68.476.9 71.873.6
100
SMC behavior
Samples
3.3. Cell response to the modified PLLC All three cell types in porcine esophagus, SMC, fibroblast and EC, were seeded separately on unmodified and modified PLLC to evaluate their cytocompatibility. Figs. 3–6 present the attachment and pro-
60 40
a The PLLC membrane was prepared in 0.06 g/ml 1,6-hexanediamine solution at 21 1C for 2 min. b The PLLC membrane was immersed in 1 mg/ml fibronectin and collagen solutions for 24 h, respectively, followed by the rinsing process same as PLLC–Fn and PLLC–Col.
20 0 PLLC
PLLC-Fn
PLLC-Col
TCPS
Fig. 3. The SMC attachment ratio and MTT activity (%, relative to TCPS) seeded for 24 h and 4 d at 37 1C in humidified air with 5% CO2, respectively. Cells were seeded at a density of 12 104/ml. Same culture conditions were employed in all following culture.
1.2
Absorbance Absorbance of every cell (xE-4)
1 SMC MTT absorbance
physisorption processes are virtually always involved in a coating procedure, even if the intended mechanism of immobilization is covalent in nature [18]. Ref. [19] also presented that Col can adsorb on untreated poly(ethylene terephthalate) (PET) surface. In our case, we had tried physical coating of Fn and Col on control PLLC surface instead of aminolysis and glutaraldehyde crosslinking. WCA measurements showed physisorbed protein-coated PLLC surfaces (Table 3, PLLC–ad-Fn and PLLC–ad-Col) have lower contact angles than that of the control PLLC but higher values than covalently bonded with protein surfaces. However, it is worth noting that both of them displayed high standard deviations, much higher than those of their covalent counterpoints, which indicated the inhomogeneous surface caused by the aggregation of Fn and Col, or the conformational heterogeneity [18]. Both are not beneficial to grow homogenous cell layer. Contrarily, the covalent PLLC–Fn and PLLC–Col displayed a more homogeneous and hydrophilic surface. We can expect that this kind of a surface will be more favorable for cell growth.
80
0.8 0.6 0.4 0.2 0 PLLC
PLLC-Fn
PLLC-Col
TCPS
Fig. 4. The MTT absorbance of the SMC in whole well and every cell ( 104) in a 96-well plate after cultured for 4 d.
liferation behaviors of esophageal SMC. All unmodified and modified PLLC have high cell attachment ratio and mitochondrial viability (Fig. 3). Protein-modified PLLC displayed higher MTT activity and relative individual cell activity (high absorbance in Fig. 4). However, from DNA assay (Fig. 5) and morphology observation (Fig. 6) after 12 d culture interval, it is apparent that
ARTICLE IN PRESS Y. Zhu et al. / Biomaterials 27 (2006) 68–78
74
&1,2=p
