Colloids and Surfaces B: Biointerfaces 112 (2013) 51–60
Contents lists available at ScienceDirect
Colloids and Surfaces B: Biointerfaces journal homepage: www.elsevier.com/locate/colsurfb
The optimal combination of substrate chemistry with physiological fluid shear stress Yan Li a,b , Yanfeng Luo a,b,∗ , Zhao Xie c , Juan Xing a,b , Manping Lin a,b , Li Yang a,b , Yuanliang Wang a,b , Ke Huang c a
Key Laboratory of Biorheological Science and Technology, Chongqing University, Ministry of Education, Chongqing 400030, China Research Center of Bioinspired Materials Science and Engineering, College of Bioengineering, Chongqing University, Chongqing 400030, China c National & Regional United Engineering Laboratory of Tissue Engineering, Department of Orthopedics, Southwest Hospital, Third Military Medical University, Chongqing 400038, China b
a r t i c l e
i n f o
Article history: Received 13 March 2013 Received in revised form 28 June 2013 Accepted 2 July 2013 Available online 25 July 2013 Keywords: Self-assembled monolayers Substrate chemistry Physiological fluid flow shear stress The optimal combination Osteoblasts Bone tissue engineering
a b s t r a c t Osteoblasts on implanted biomaterials sense both substrate chemistry and mechanical stimulus. The effects of substrate chemistry alone and mechanical stimulus alone on osteoblasts have been widely studied. This study investigates the optimal combination of substrate chemistry and 12 dyn/cm2 physiological flow shear stress (FSS) by examining their influences on primary rat osteoblasts (ROBs), including the releases of ATP, nitric oxide (NO), and prostaglandin E2 (PGE2 ). Self-assembled monolayers (SAMs) on glass slides with –OH, –CH3 , and –NH2 were employed to provide various substrate chemistries, whereas a parallel-plate fluid flow system produced the physiological FSS. Substrate chemistry alone exerted no observable effects on the releases of ATP, NO, and PGE2 . Nevertheless, when ROBs were exposed to both substrate chemistry and FSS, the ATP releases of NH2 were upregulated about 12-fold compared to substrate chemistry alone, while the ATP releases of CH3 and OH was similarly increased 7-fold at the peak. Similar trends were observed for the releases of NO and PGE2 . The expressions of ATP, NO, and PGE2 followed the pattern of NH2 -FSS > Glass-FSS > CH3 -FSS ≈ OH-FSS. ROBs on NH2 produced the optimal combination of substrate chemistry with the physiological FSS. The F-actin organization and focal adhesion (FA) formation of ROBs on various SAMs without FSS were examined. NH2 produced the best results whereas CH3 and OH produced the worst ones. Inhibition of FAs and/or disruption of F-actin significantly decreased the releases of FSS-induced PGE2 , NO, and/or ATP. Consequently, a mechanism was proposed that the best F-actin organization and FA formation of ROBs on NH2 lead to the optimal combination of substrate chemistry with the 12 dyn/cm2 physiological FSS. This mechanism gives guidance for the design of implanted biomaterials and bioreactors for bone tissue engineering. © 2013 Elsevier B.V. All rights reserved.
1. Introduction Scaffolds and seeded cells such as osteoblasts are two important components of bone tissue engineering. Scaffolds provide physical support and living microenvironment for seeded cells [1]. The properties of scaffolds, including scaffold chemistry [2,3], stiffness [4] and architecture [5], determine cellular responses, tissue growth, and even clinical success of bone tissue engineering. The design and selection of scaffold chemistry is the first step for scaffold applications. To provide guidance for designing and selecting scaffold chemistry, the effect of substrate chemistry on osteoblasts has been widely studied by using 2D [6–8] or 3D [2,9] systems and
∗ Corresponding author at: College of Bioengineering, Chongqing University (Campus A), Chongqing 400030, China. Tel.: +86 23 65102509; fax: +86 23 65102507. E-mail address: yfl
[email protected] (Y. Luo). 0927-7765/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.colsurfb.2013.07.001
some important results were reported. For instance, Healy et al. [7] demonstrated enhanced attachment of osteosarcoma cells to NH2 surfaces compared to CH3 surfaces and they were able to guide spatial distribution of cells by controlling the distribution of NH2 and CH3 . Similarly, Garcia et al. [8] observed different osteoblasts adhesion and mineralization on well-defined NH2 , COOH, CH3 and OH surfaces. In 3D scaffolds, Reis et al. found that sulfonic and phosphonic groups grafted on the starch and polycaprolactone (SPCL) scaffolds significantly promoted osteoblast proliferation compared to virgin SPCL [9]. In addition to substrate chemistry of scaffolds, mechanical loading is indispensable for mechanosensitive osteoblasts to produce normal bone tissue and maintain normal bone functions [1,10]. Generally, the mechanical loading is applied to osteoblasts either through scaffolds stretch [11], compression [12] or fluid flow through scaffolds [13]. Especially, fluid shear stress (FSS) is regarded as the principal mechanical stimulus responsible for bone
52
Y. Li et al. / Colloids and Surfaces B: Biointerfaces 112 (2013) 51–60
adaption and remodeling [14–16]. In bone tissue in vivo, the physiological FSS caused by interstitial fluid flow is 8–30 dyn/cm2 on the basis of the model presented by Weinbaum et al. [17]. In this study, 12 dyn/cm2 were employed to represent physiological FSS. Accordingly, appropriate substrate chemistry and 12 dyn/cm2 FSS are both required to realize bone tissue engineering. The osteoblast responses to substrate chemistry under 12 dyn/cm2 FSS seem more significant and instructive for scaffold design than the osteoblast responses to substrate chemistry alone. The aim of this study is to examine the osteoblast responses to various substrate chemistries under 12 dyn/cm2 physiological FSS, including ATP, nitric oxide (NO), and prostaglandin E2 (PGE2 ) releases, and find an optimal combination of substrate chemistry with the physiological FSS. The productions of NO and PGE2 are early responses of osteoblasts to FSS and play important roles in bone formation and remodeling [18–20]. Thus, NO and PGE2 productions were selected as parameters to indicate the early responses of ROBs. Besides, ATP, as a key energy provider, is the precondition for other follow-up cellular responses including the productions of NO and PGE2 , cell proliferation and differentiation. Moreover, ATP release is one of the fastest responses of ROBs to FSS and can be employed to indicate the sensitiveness of ROBs to FSS [21]. Therefore, ATP release was considered as well. Self-assembled monolayers (SAMs) on glass slides were employed to provide a variety of surface chemistries, including –OH, –CH3 and –NH2 , and a parallel-plate fluid flow system was applied to produce 12 dyn/cm2 physiological FSS. The results demonstrated that ROBs on NH2 surfaces had better responses than on OH and CH3 surfaces, indicating that NH2 is the optimal combination of substrate chemistry with 12 dyn/cm2 FSS. The different cytoskeleton organization and focal adhesion formation on various substrate chemistries were responsible for this optimal combination.
2.2. Osteoblasts culture Primary rat osteoblasts (ROBs) cultures were described in our previous publication [18]. Briefly, cells were isolated from minced rat calvarial chips, followed by 2 h of 0.125% collagenase and 0.25% trypsin at 37 ◦ C shaking. Cultures were initiated in DMEM (Gibico, USA) supplemented with 10% heat-inactivated fetal calf serum (FCS, Sijiqing, China), penicillin (100 U/mL), streptomycin (100 g/mL) and 0.05% l-glutamine, and maintained in a humidified atmosphere of 5% CO2 /95% air at 37 ◦ C. The medium was changed every two days. After confluence, the cells were subcultured and identified by von Kossa staining method according to previously reported procedure [23]. The forth, fifth, and sixth passage ROBs were seeded on slides for experiments at a density of 2 × 105 cells/slide. 2.3. Fluid shear stress A parallel-plate flow chamber apparatus was employed to provide steady FSS [24]. The dimension of the flow chamber was 7.50 cm (length, L) × 2.50 cm (width, W) × 0.03 cm (height, H). FSS was produced by circulating 10 mL DMEM via a peristaltic pump (JieHeng, China). The produced FSS (, dyn/cm2 ) was calculated according to equation = 6Q/H 2 W , where is the dynamic viscosity of the perfusate and Q is the flow rate. The employed FSS in this study was steady 12 dyn/cm2 . All components of the apparatus except the pump were maintained in a 37 ◦ C incubator during the experiment, and the medium was continuously saturated with 5% CO2 /95% air. For FSS-loaded samples (denoted by X-FSS, X represents surface chemistry), slides with attached ROBs were mounted on the flow chamber and then exposed to FSS for a predetermined time. Other samples without FSS exposure (denoted by X, X represents surface chemistry) were kept in Petri dishes at 37 ◦ C in a humidified atmosphere of 5% CO2 /95% air for the same predetermined time. 2.4. Determination of ATP, NO and PGE2 releases
2. Materials and methods 2.1. Preparation and characterization of SAMs on glass slides Blank glass slides were cleaned by acetone and ethanol, and then rinsed with deionized water (DIW). All cleaned slides were stored in DIW before introduction of chemical groups. To introduce –OH groups, the cleaned slides were dipped into freshly prepared Piranha solution (concentrated H2 SO4 : 30% H2 O2 = 7:3, v/v) at 80 ◦ C for 1 h, and then were rinsed with excessive DIW and blown dry with nitrogen. The obtained slides were labeled as OH slides. To get NH2 slides, the OH slides were dipped into 1% solution of (3-aminopropyl) triethoxysilane (Alfa Aesar, USA) in acetone, and then treated by refluxing 1 mL distilled water for 30 min. The obtained slides were rinsed with ethanol and DIW, and blown dry with nitrogen [22]. For CH3 slides, the OH slides were dipped into 5% solution of chloro(dimethyl) octadecylsilane (Sigma–Aldrich, USA) in hexane for 1 h, and then rinsed with ethanol and DIW, blown dry with nitrogen [9]. Finally, all slides were placed in Petri dishes (Corning, USA), soaked with 75% alcohol overnight and rinsed with sterile phosphate-buffered saline (PBS). Water contract angles of the slides were characterized by using a Model 200 video-based optical system (Future Scientific Ltd. Co., Taiwan, China). Six specimens were measured for each kind of SAMs, and three different points were selected for each specimen. All measurements were performed at room temperature by dropping 5 L of ultra-pure water.
After seeded on SAMs for 48 h, ROBs (reaching ∼80% confluence) were employed to examine the releases of ATP, NO and PGE2 . When ROBs on slides were exposed to FSS for a predetermined time (0, 1, 2, 3, 4, 5, 10, and 15 min for ATP determination, 0, 5, 10, 15, 30, 45, and 60 min for NO and PGE2 determination), 2 mL of medium was withdrawn and replenished with an equal volume of fresh culture medium to maintain a constant circulating fluid volume. The same procedure was performed for samples without FSS exposure. 2.4.1. Release of ATP The ATP release was detected by using an ATP Assay Kit (Beyotime, China) based on bioluminescence technology. Briefly, 500 L of the withdrawn medium and 200 L of luciferin–luciferase reagent were mixed for 5 s; thereafter the fluorescence intensity was measured by using Multifunctional Microplate Reader (Biotek Synergy HT, USA). The concentration of ATP was obtained from the standard curve plotted by using the supplied ATP standards and normalized to the total cellular protein using a BCA Protein Assay Kit (Biotech, China). To ensure that the detected ATP was all extracellular ATP, the plasma membrane integrity was confirmed through surveying the lactate dehydrogenase (LDH) levels of the conditioned media by using a LDH Assay Kit (Jiangcheng, China). 2.4.2. Releases of NO The concentration of NO was detected by using Nitric Oxide Assay Kit (Beyotime, China) on the basis of Griess Reagent. The withdrawn medium was allowed to react with Griess Reagent for
Y. Li et al. / Colloids and Surfaces B: Biointerfaces 112 (2013) 51–60
10 min, and then the absorbance was measured at 540 nm using a Microplate Reader (Bio-Rad, USA). The concentration of NO was obtained according to the standard curve plotted by the supplied NaNO2 standards and normalized to the total cellular protein.
53
Table 1 Water contact angles of various SAMs with terminal NH2 , OH, CH3 , and blank glass slides. The results were shown as mean ± SD. Contact angle (◦ ) NH2 Blank CH3 OH
61.7 28.6 97.9 11.4
± ± ± ±
2.4 1.6 1.8 1.2
2.4.3. Release of PGE2 The releases of PGE2 were measured by using Rat PGE2 ELISA Assay Kit (Yuan-Ye Chemical reagents, China). The absorbance was detected at 450 nm and the concentration of PGE2 was obtained from the standard curve plotted by the supplied PGE2 standards and normalized to the total cellular protein.
The results were shown as mean ± SD.
2.5. Focal adhesion formation and F-actin organization
According to the Berg limit of 65◦ [25], only CH3 surfaces are hydrophobic.
After seeded on SAMs for 2 h, 12 h, and 48 h (∼80% confluence, equivalent to osteoblasts before FSS exposure), ROBs were stained for focal adhesion and cytoskeleton visualization. Before stained, ROBs were fixed through incubation in 4% (w/v) paraformaldehyde in PBS for 30 min and permeabilized for 10 min with 0.25% (v/v) Triton X-100 in PBS. 1% BSA in PBS was added to block the nonspecific binding sites by incubating with ROBs for 1 h. To observe focal adhesion formation, vinculin was stained through incubating with a mouse monoclonal anti-vinculin antibody (Abcam, UK) at room temperature for 60 min, and then incubated with a FITC-conjugated anti-mouse immunoglobulin secondary antibody (ZSGB-Bio, China) at room temperature for another 60 min. F-actin was stained using Rhodamine-Phalloidin (Molecular Probes, Invitrogen, USA) at room temperature for 60 min. Finally, the nuclei were stained by incubation with DAPI (Sigma–Aldrich, USA) at room temperature for 10 min. The stained specimens were washed, mounted in glycerol and examined by Confocal Laser Scanning Microscope (CLSM; TCS SP5, Leica, Germany). The number of vinculin and the area of F-actin were calculated using Image J. 2.6. Inhibition of focal adhesions and disruption of F-actin To inhibit the cell attachment and indirectly inhibit FA formation of ROBs, suspended ROBs were firstly incubated under rotation for 1 h at 37 ◦ C in DMEM containing 0% fetal calf serum and 500 g/mL RGDS peptide (Glbiochem, China). Thereafter, ROBs were permitted to seed on various SAMs for 48 h in the continued presence of RGDS. To disrupt F-actin organization, ROBs were seeded on SAMs for 47 h and then cultured in DMEM containing 0% fetal calf serum and 1 M cytochalasin B (Sigma–Aldrich, USA) for another 1 h. The FAs and F-actin organization before FSS exposure and the releases of ATP, NO, and PGE2 after FSS exposure were determined according to the methods described in 2.4, respectively. 2.7. Statistical analysis Data were expressed as means ± SD (n ≥ 6 for all experiments). Single factor analysis of variance (ANOVA) technique with OriginPro (version 8.0) was used to assess the statistical significance of results. The confidence interval was set to 0.05. 3. Results 3.1. Measurements of water contact angle for SAMs Various SAMs with terminal –NH2 , –CH3 , and –OH were prepared on glass slides. The static water contact angles were shown in Table 1. The contact angle of CH3 was the highest (97.9 ± 1.8◦ ), while that of OH was the lowest (11.4 ± 1.2◦ ). The water contact angles of NH2 and blank glass were 61.7 ± 2.4◦ and 28.6 ± 1.6◦ , respectively. The hydrophilicity decreased following OH > Glass > NH2 > CH3 .
3.2. ROBs on NH2 surfaces demonstrate highest releases of ATP, NO and PGE2 under 12 dyn/cm2 physiological FSS In this study, a steady physiological FSS should be more useful to understand the combined mechanism of substrate chemistry with FSS, although oscillatory FSS has been reported to have greater effects on osteoblasts. Therefore, 12 dyn/cm2 steady physiological FSS was employed in this study, which was produced by using a parallel-plate flow chamber system designed by Nauman [24]. First, the effects of 12 dyn/cm2 FSS on the ATP, NO and PGE2 releases of ROBs with virgin glass slides as the substrate (labeled as glass-FSS) were examined. To find the optimal combination of substrate chemistry and 12 dyn/cm2 FSS, the virgin glass slides were modified and replaced with CH3 , OH and NH2 SAMs. The corresponding samples were labeled as NH2 -FSS, CH3 -FSS and OH-FSS. The releases of ATP, NO, PGE2 and ROBs were also examined. To ensure that the detected ATP is actively released by ROBs, the plasma membrane integrity was monitored by detecting and comparing the LDH levels in the conditioned media with FSS and without FSS. The comparison did not yield a significant difference (data not shown), suggesting that the ROBs plasma membrane was not disrupted by the substrate chemistry and loaded FSS, and the detected ATP was actively released by ROBs. When ROBs received substrate chemistry stimulus alone, that is, when ROBs were cultured on various SAMs without FSS exposure, the concentrations of released ATP on NH2 , Glass, CH3 , and OH showed no observable change with incubation time (Fig. 1A). This implies that substrate chemistry has negligible effects on ATP releases of ROBs. When 12 dyn/cm2 FSS were loaded, the corresponding ATP releases were significantly increased on all SAMs. This is consistent with the reported results that FSS enhances ATP releases [21,26]. However, different ATP releases responding to the FSS were observed on different SAMs (Fig. 1A). NH2 -FSS led to the highest ATP releases, while CH3 -FSS and OH-FSS led to the similarly lowest ones. To more clearly see the combined effects of substrate chemistry and the FSS, the ATP ratios of X-FSS to the corresponding X (X-FSS/X) were calculated (Fig. 1B). As can be seen in Fig. 1B, the ATP ratios of NH2 -FSS/NH2 were the highest with a peak value of about 12-fold, while those of CH3 -FSS/CH3 and OH-FSS/OH were the lowest with a peak value of about 7-fold. This suggests that NH2 surfaces with 12 dyn/cm2 physiological FSS have the optimal releases of ATP. The NO and PGE2 releases of ROBs were further detected. When exposed to substrate chemistry alone, ROBs on various SAMs produced similarly small amounts of NO (Fig. 2A) and PGE2 (Fig. 3A). When ROBs were exposed to both substrate chemistry and the FSS, their releases of NO and PGE2 were significantly higher than those
54
Y. Li et al. / Colloids and Surfaces B: Biointerfaces 112 (2013) 51–60
Fig. 1. (A) ATP release of ROBs receiving both substrate chemistry (NH2 , OH, CH3 , or glass) and FSS (12 dyne/cm2 ) stimulus compared to those receiving substrate chemistry stimulus alone. (B) The ratios between the ATP releases of ROBs exposed to both substrate chemistry and FSS (X-FSS) and those exposed to substrate chemistry (X) alone. X represents NH2 , OH, CH3 , and Blank glass slides (glass). FSS is 12 dyn/cm2 . The results were shown as mean ± SD. Statistical analysis results are shown in Table 3.
Fig. 2. (A) NO production of ROBs receiving both substrate chemistry (NH2 , OH, CH3 , or glass) and FSS (12 dyn/cm2 ) stimulus compared to those receiving substrate chemistry stimulus alone. (B) The ratios between the NO releases of ROBs exposed to both substrate chemistry and FSS (X-FSS) and those exposed to substrate chemistry (X) alone. X represents NH2 , OH, CH3 or Blank glass slides (glass). FSS is 12 dyn/cm2 . The results were shown as mean ± SD. Statistical analysis results are shown in Table 4.
Fig. 3. (A) PGE2 production of ROBs receiving both substrate chemistry (NH2 , OH, CH3 , or glass) and FSS (12 dyn/cm2 ) stimulus compared to those receiving substrate chemistry stimulus alone. The ratios between the PGE2 releases of ROBs exposed to both substrate chemistry and FSS (X-FSS) and those exposed to substrate chemistry (X) alone. X represents NH2 , OH, CH3 or Blank glass slides (Glass). FSS is 12 dyn/cm2 . The results were shown as mean ± SD. The “*” represents that it has significant difference at the confidence of 0.05.
Y. Li et al. / Colloids and Surfaces B: Biointerfaces 112 (2013) 51–60
55
Fig. 4. Focal adhesion and F-actin formation of ROBs receiving substrate chemistry stimulus alone without FSS exposure after incubation for 2 h, 12 h, and 48 h. The cells were fixed, permeabilized, and double labeled for vinculin (green), F-actin (red), and nuclei (blue). The yellow areas are vinculin, partially colocalized with F-actin in the focal adhesion plaques. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of the article.)
of ROBs exposed to substrate chemistry alone (Figs. 2A and 3A). This is consistent with the reported results that FSS enhances NO and PGE2 releases [27,28]. Similar to ATP releases, NH2 -FSS produced the highest NO and PGE2 releases, whereas CH3 -FSS and OH-FSS produced the lowest ones (Figs. 2A and 3A). The NO ratios and PGE2 ratios between X-FSS and the corresponding X are shown in Figs. 2B and 3B, respectively. The peak value of NO releases for NH2 -FSS were upregulated about 3-fold while those for CH3 -FSS and OH-FSS were similarly increased 2fold compared to their corresponding substrate chemistry alone. The peak value of PGE2 releases were upregulated about 2.4-fold for NH2 -FSS and 1.7-fold for both CH3 -FSS and OH-FSS. Again, these results suggest that NH2 with 12 dyn/cm2 FSS has the optimal releases of NO and PGE2 .
3.3. Focal adhesion formation and F-actin organization of ROBs on various SAMs before FSS exposure To elucidate the possible mechanism by which NH2 with 12 dyn/cm2 physiological FSS has the optimal responses of ROBs, including the releases of ATP, NO, and PGE2 , the FA formation and F-actin organization in ROBs receiving substrate chemistry stimulus alone were examined using a CLSM. The pictures of ROBs after cultured on SAMs for 2 h, 12 h, and 48 h are shown in Fig. 4, where the blue, green, and red represent nuclei, vinculin, and F-actin, respectively. After ROBs were cultured for 2 h (Fig. 4), some FAs formed and F-actin filaments appeared on NH2 slides, whereas rare FAs or Factin filaments were observed on CH3 and OH slides. On blank
56
Y. Li et al. / Colloids and Surfaces B: Biointerfaces 112 (2013) 51–60
Table 2 The number of vinculin (number/cell) and the area of F-actin (m2 /cell). The results were shown as mean ± SD. Time
2h
12 h
Vinculin Glass NH2 CH3 OH
96 153 23 18
± ± ± ±
F-actin
13 21 5 9
61 98 35 33
± ± ± ±
48 h
Vinculin
8 12 14 9
143 227 78 75
± ± ± ±
F-actin
26 34 19 10
± ± ± ±
87 110 42 52
Vinculin
12 27 15 24
263 372 150 144
± ± ± ±
F-actin
29 32 23 29
132 168 79 86
± ± ± ±
23 25 21 19
The results were shown as mean ± SD. Table 3 The statistical analysis results of Fig. 1. The “*” represents that it has significant difference at the confidence of 0.05. The “NS” represents that it has no significant difference at the confidence of 0.05. NH2 -FSS (NH2 -FSS/NH2 )
Glass-FSS (glass-FSS/glass)
OH-FSS (OH-FSS/OH)
CH3 -FSS (CH3 -FSS/CH3 )
Time (min)
1
1 2 3 4 5 10 15
*
1 2 3 4 5 10 15
*
1 2 3 4 5 10 15
*
2
3
4
5
CH3 -FSS (CH3 -FSS/CH3 ) 10
15
1
2
3
OH-FSS (OH-FSS/OH)
4
5
10
15
1
*
2
3
4
5
10
15
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
NS *
NS *
NS *
NS *
NS *
NS *
NS
* * * * * *
glass slides, only a few fuzzy F-actin filaments formed with a few FAs. After culturing for 12 h (Fig. 4), short linear focal adhesion plaques and longitudinal F-actin stress fibers were observed on both NH2 and Glass slides, whereas only very small or punctuate FAs started to develop and no F-actin stress
fibers formed on OH slides. On CH3 slides, neither FAs nor F-actin stress fibers formed. After ROBs were cultured for 48 h, all ROBs on NH2 contained a large number of FAs and clear Factin stress fibers, whereas ROBs on OH and CH3 displayed few FAs and vague F-actin stress fibers (Fig. 4). With respect to ROBs
Table 4 The statistical analysis results of Fig. 2. The “*” represents that it has significant difference at the confidence of 0.05. The “NS” represents that it has no significant difference at the confidence of 0.05. NH2 -FSS
CH3 -FSS
(NH2 -FSS/NH2 )
Glass-FSS (glass-FSS/glass)
OH-FSS (OH-FSS/OH)
CH3 -FSS (CH3 -FSS/CH3 )
Time (min)
5
5 10 15 30 45 60
*
5 10 15 30 45 60
*
5 10 15 30 45 60
*
10
15
OH-FSS
(CH3 -FSS/CH3 ) 30
45
60
5
10
(OH-FSS/OH) 15
30
45
60
* * * * * * NS NS NS NS
* * * * *
60
* *
NS
*
45
* *
*
30
* *
*
15
* *
*
10
* *
*
5
NS
*
Y. Li et al. / Colloids and Surfaces B: Biointerfaces 112 (2013) 51–60
57
Fig. 5. Inhibition of focal adhesions and disruption of F-actin of ROBs receiving substrate chemistry stimulus alone without FSS exposure after incubation for 48 h. The cells were fixed, permeabilised, and double labeled for vinculin (green), F-actin (red), and nuclei (blue). The yellow areas are vinculin, partially colocalized with F-actin in the focal adhesion plaques. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of the article.)
on Glass slides, the number of FAs was fewer and the formation of F-actin stress fibers was not significantly different comparing to those on NH2 slides (Fig. 4). The statistical results of FAs and F-actin using Image J were shown in Table 2. In a brief summary, FA formation and F-actin organization in ROBs evidently depended on substrate chemistry following the pattern NH2 > Glass > CH3 ≈ OH. Garcia et al. [8] also reported that wet-cleaving/biochemical quantification of assembled FAs displayed significant differences among surface chemistries (NH2 = COOH > OH > CH3 ). In addition, the same trend of NH2 > OH > CH3 for FAs were reported by Faucheuxa et al. [22].
3.4. Focal adhesion inhibition and F-actin disruption of ROBs To further elucidate whether the different FA formation and Factin organization on various SAMs are responsible for the different cellular responses of ROBs to FSS on various SAMs, the FA formation was inhibited and/or F-actin organization was disrupted before FSS exposure and then FSS-induced ATP, NO, and PGE2 releases were examined. RGDS is the primary integrin binding sequence within fibronectin. Treatment of cells with soluble RGDS peptide may compete with the RGDS sequence in fibronectin for integrin binding and prevent FA formation. Therefore, RGDS peptide was used to
58
Y. Li et al. / Colloids and Surfaces B: Biointerfaces 112 (2013) 51–60
Fig. 6. FSS-induced response of ROBs treated with RGDS to inhibit focal adhesions, and treated with cytochalasin B to disrupt F-actin on various SAMs. The results were shown as mean ± SD.
indirectly inhibit FA formation. The concentration of RGDS, 500 g/mL, was chosen through Ponik’s research, which can damage the FA formation without causing cellular cytotoxicity [29]. When FA formation was inhibited (Fig. 5), both ATP and NO releases decreased to some extent on all SAMs and maintained a similar pattern of NH2 -FSS > glass-FSS > CH3 -FSS ≈ OH-FSS (Fig. 6) compared to that without FA inhibition (Figs. 1 and 2). On the other hand, PGE2 release declined dramatically and demonstrated no statistical difference among various SAMs (Fig. 6). Cytochalasin B can disrupt the F-actin specifically. The setting concentration of cytochalasin B was based on a previous study [30]. When F-actin was disrupted (Fig. 5), the releases of ATP and PGE2 slightly decreased and maintained a similar pattern of NH2 -FSS > Glass-FSS > CH3 -FSS ≈ OH-FSS (Fig. 6) compared to that without F-actin disruption (Fig. 3). However, the disruption of Factin resulted in a considerable reduction in the release of NO on all SAMs (Fig. 6). When both FA was inhibited and F-actin was disrupted (Fig. 5), FSS-induced responses of ROBs including the releases of ATP, NO, and PGE2 nearly disappeared on all SAMs (Fig. 6).
4. Discussion The aim of this study was to find the optimal combination of substrate chemistry with 12 dyn/cm2 physiological FSS by investigating the synergistic role of the physiological FSS and substrate chemistry in osteoblast responses, and to understand the possible mechanism of these responses. The substrate chemistry was represented by SAMs on glass slides with terminal NH2 , CH3 , and OH. Blank glass slides (glass) were used as the control of SAMs. FSS was provided by a parallel plate flow chamber and a steady FSS of 12 dyn/cm2 was employed. Since the preparation of SAMs on glass slides is well established, only the water contact angles were determined to indicate the formation of NH2 , CH3 , Glass, and OH SAMs. The detected water contact angles are similar to those reported in previous studies [31,32]: CH3 is hydrophobic, OH and Glass are hydrophilic, and NH2 is moderate, which indicates that the preparation of the SAMs is successful. The wettability of a surface has a significant effect on cell adhesion. Cells adhered well on moderately wettable surfaces with water contact angles of 40–60◦ [33]. Our subsequent results also verified this rule (Fig. 4).
Y. Li et al. / Colloids and Surfaces B: Biointerfaces 112 (2013) 51–60
The short-term responses of ROBs to substrate chemistry and physiological FSS were examined, including the releases of ATP as well as the releases of NO and PGE2 . In comparison with substrate chemistry alone, FSS induced quick changes in the release of ATP, NO, and PGE2 (Figs. 1A, 2A and 3A). The sensitiveness of ROBs to FSS, characterized by ATP release, was strongest on the NH2 substrate and obviously weakened on CH3 and OH surfaces (Fig. 1A). As a result, different releases of NO and PGE2 were observed following the pattern: NH2 -FSS > Glass-FSS > CH3 FSS ≈ OH-FSS (Figs. 2A and 3A). Therefore, one conclusion could be made that NH2 is the optimal combination of substrate chemistry with 12 dyn/cm2 physiological FSS because of the highest releases of ATP, NO, and PGE2 . Initial adhesion and spreading of osteoblasts, including integrin-medicated cytoskeleton rearrangement [34–38] and initial formation of FA [39–42], have been regarded as critical factors that influence FSS-related mechanotransduction. Pavalko et al. proved that FSS-induced mechanical signaling in MC3T3-E1 osteoblasts requires cytoskeleton-integrin interactions [43]. However, McGarry et al. disrupted the cytoskeleton of osteoblasts and found that FSS-induced NO releases need intact actins, whereas PGE2 releases are independent on actin cytoskeleton reorganization [44]. In addition, Ponik and Pavalko used bovine serum albumin and RGDS to inhibit formation of focal adhesions in different degrees and confirmed that the releases of FSS-induced osteoblast PGE2 are controlled by focal adhesion [29]. On the other hand, it has been proved that adhesion formation and cytoskeleton rearrangement at the early stage of matrix–cell interactions are regulated by substrate chemistry [8]. Therefore, FAs and cytoskeleton are the common components by which both substrate chemistry and FSS influence the behavior of osteoblasts. Taking all above information in mind, one hypothesis could be made that FA and F-actin organization should be responsible for the optimal combination of NH2 and 12 dyn/cm2 FSS. To verify this hypothesis, FA and F-actin on various SAMs before FSS exposure were pictured (Fig. 4). The trend of F-actin is NH2 > Glass > CH3 ≈ OH, while that of the FAs is NH2 > Glass > OH ≈ CH3 (Table 2). These patterns are consistent with previously reported cases and different hydrophilicity and electrostatic effects are believed to be responsible for these patterns [8,22]. Comparing the patterns of FSS-induced releases of ATP, NO, and PGE2 (Figs. 1–3) revealed that the releases of ATP, NO, and PGE2 followed the similar patterns of FAs or F-actin organization. On the other hand, inhibition of FAs, disruption of F-actin, or simultaneous inhibition of FAs and disruption of F-actin led to a considerable decrease in the releases of FSS-induced PGE2 and NO (Fig. 6). These observations are in consistent with the previous findings that FSS-induced releases of ATP, NO, and PGE2 of osteoblasts are mediated by FAs and F-actin [29,41,42,44]. The FSSinduced release of PGE2 of ROBs followed the pattern of FAs (see Figs. 3 and 4 without FA inhibition; see Figs. 5 and 6 with FA inhibition), whereas the NO release followed the pattern of F-actin (see Figs. 2 and 4 without F-actin disruption; see Figs. 5 and 6 with Factin disruption). These results are consistent with the observed facts that FSS-induced release of NO requires intact actin [44] and the release of PGE2 does not. These results imply that substrate chemistry regulates the responses of ROBs to FSS by controlling the formation of FA and the organization of F-actin. NH2 surfaces produced the best FAs and F-actin organization, so that the optimal combination of substrate chemistry with 12 dyn/cm2 physiological FSS was observed on NH2 surfaces. 5. Conclusions In summary, this is the first study to demonstrate that the physiological FSS-induced responses of osteoblasts,
59
including the releases of ATP, NO, and PGE2 , follow the pattern NH2 > Glass > CH3 ≈ OH, and NH2 surfaces are the optimal combination of substrate chemistry with 12 dyn/cm2 physiological FSS. Substrate chemistry differentially regulates the role of FSS by controlling F-actin organization and FA formation of osteoblasts, which further mediate the short-term and long-term cellular responses to FSS. These findings provide valuable guidance for the design of scaffold chemistry to synergize with the in vivo FSS to produce normal bone tissues. However, the surfaces with mixed chemical groups may be better than those surfaces with single chemical group to study the optimal combination of substrate chemistry with physiological FSS. This work will be studied in the future. Acknowledgements This work was supported by Grants from the National Natural Science Foundation of China (Nos. 30970700, 11032012 and 30973065). Appendix A. Supplementary data Supplementary data associated with cle can be found, in the online http://dx.doi.org/10.1016/j.colsurfb.2013.07.001.
this artiversion, at
References [1] D.W. Hutmacher, Biomaterials 24 (2000) 2529. [2] Y. Hu, S.R. Winn, L. Krajbich, J.O. Hollinger, J. Biomed. Mater. Res. A (2003) 583. [3] J.A. Barry John, D. Howard, K.M. Shakesheff, S.M. Howdle, M.R. Alexander, Adv. Mater. 18 (2006) 1406. [4] K. Schlichting, H. Schell, R.U. Kleemann, A. Schill, A. Weiler, G.N. Duda, D.R. Epari, Am. J. Sport Med. 36 (2008) 2379. [5] G. Kumar, C.K. Tison, K. Chatterjee, S.P. Pine, J.H. McDaniel, M.L. Salit, M.F. Young Jr., C.G. Simon, Biomaterials 32 (2011) 9188. [6] M. Padial-Molina, P. Galindo-Moreno, J.E. Frandez-Barbero, F. O’Valle, A.B. Jodar-Reyes, J.L. Ortega-Vinuesa, P.J. Ramon-Torregrosa, Acta Biomater. 7 (2011) 771. [7] K.E. Healy, B. Lorn, P.E. Hockberger, Biotechnol. Bioeng. 43 (1994) 792. [8] B.G. Keselowsky, D.M. Collard, A.J. Garcia, Biomaterials 25 (2004) 5947. [9] P.M. Lopez-Perez, R.M.P. da Silva, R.A. Sousa, I. Pashkuleva, R.L. Reis, Acta Biomater. 6 (2010) 3704. [10] J.R. Mauney, S. Sjostorm, J. Blumberg, R. Horan, J.P. O’Leary, G. VunjakNovakovic, V. Volloch, D.L. Kaplan, Calcif. Tissue Int. 74 (2004) 458. [11] A. Sittichockechaiwut, A.M. Scutt, A.J. Ryan, L.F. Bonewald, G.C. Reilly, Bone 44 (2009) 822. [12] B. Rath, J. Nam, T.J. Knobloch, J.J. Lannutti, S. Agarwal, J. Biomech. 41 (2008) 1095. [13] G.N. Bancroft, V.I. Sikavitsas, J. Dolder, T. Sheffield, C.G. Ambrose, J.A. Jansen, A.G. Mikos, PNAS 99 (2002) 12600. [14] J.F.A.M. Hillsley, Biotechnol. Bioeng. 43 (1993) 573. [15] E.H. Burger, J.K. Nulen, Adv. Dent. Res. 13 (1999) 93. [16] N. Basso, J.N.M. Heersche, Bone 30 (2002) 347. [17] S. Weinbaum, S.C. Cowin, Zeng. Yu, J. Biomech. 27 (1994) 339. [18] K.M. Reich, J.A. Frangos, Am. J. Physiol. 261 (1991) C428. [19] P. Collin-Osdoby, G.A. Nickols, P. Osdoby, J. Cell. Biochem. 57 (1995) 399. [20] I. Owan, D.B. Burr, C.H. Turner, J. Qui, Y. Tu, J.E. Onyia, R.L. Duncan, Am. J. Physiol. Cell Physiol. 42 (1997) C810. [21] D.C. Genetos, D.J. Geist, D. Liu, H.J. Donahue, R.L. Duncan, J. Bone Miner. Res. 20 (2005) 41. [22] N. Faucheuxa, R. Schweiss, K. Lutzowa, C. Werner, T. Groth, Biomaterials 25 (2004) 2721. [23] Y.F. Luo, Y.L. Wang, X.F. Niu, J.F. Shang, Eur. Polym. J. 44 (2008) 1390. [24] E.A. Nauman, K.J. Risic, T.M. Keaveny, R.L. Satcher, Ann. Biomed. Eng. 27 (1999) 194. [25] J.M. Berg, L.G.T. Erikkson, P.M. Claesson, K.G.N. Borve, Langmuir 10 (1994) 1225. [26] R.C. Riddle, A.F. Taylor, J.R. Rogers, H.J. Donahue, J. Bone Miner. Res. 22 (2007) 589. [27] R.G. Bacabac, T.H. Smit, M.G. Mullender, S.J. Dijcks, JJWAV. Loon, J.K. Nulend, Biochem. Biophys. Res. Commun. 315 (2004) 823. [28] T.N. Mcallister, J.A. Frangos, J. Bone Miner. Res. 14 (1999) 930. [29] S.M. Ponik, F.M. Pavalko, J. Appl. Physiol. 97 (2004) 135. [30] N.E. Ajubi, J. Klein-Nulend, P.J. Nijweide, T. Vrijheid-Lam-mers, M.J. Alblas, E.H. Burger, Biochem. Biophys. Res. Commun. 225 (1996) 62. [31] H.G. Hong, M. Jiang, S.G. Sligar, P.W. Bohn, Langmuir 10 (1994) 153.
60
Y. Li et al. / Colloids and Surfaces B: Biointerfaces 112 (2013) 51–60
[32] Y.J. Ren, H. Zhang, H. Huang, X.M. Wang, Z.Y. Zhou, F.Z. Cui, Y.H. An, Biomaterials 30 (2009) 1036. [33] Y. Arima, H. Iwata, Biomaterials 28 (2007) 3074. [34] D.E. Ingber, Proc. Natl. Acad. Sci. 100 (2003) 1472. [35] K.A. Myers, J.B. Rattner, N.G. Shrive, D.A. Hart, Biochem. Biophys. Res. Commun. 364 (2007) 214. [36] A.W. Orr, B.P. Helmke, B.R. Blackman, M.A. Schwartz, Dev. Cell 10 (2006) 11. [37] K. Hayakawa, H. Tatsumi, M. Sokabe, J. Cell Sci. 121 (2008) 496. [38] N. Wang, J.D. Tytell, D.E. Ingber, Nat. Rev. Mol. Cell Biol. 10 (2009) 75. [39] T. Shemesh, B. Geiger, A.D. Bershadsky, M.M. Kozlov, Proc. Natl. Acad. Sci. 102 (2005) 12383.
[40] S.R.L. Young, R.G. O’Riley, J.B. Kim, F.M. Pavalko, J. Bone Miner. Res. 24 (2009) 411. [41] S.M. Norvell, S.M. Ponik, D.K. Bowen, R. Gerard, F.M. Pavalko, J. Appl. Physiol. 96 (2004) 957. [42] H. Kamioka, Y. Sugawara, S.A. Murshid, Y. Ishihara, T. Honjo, T.T. Yamamoto, J. Bone Miner. Res. 21 (2006) 1012. [43] F.M. Pavalko, N.X. Chen, C.H. Turner, D.B. Burr, S. Atkinson, Y.F. Hsieh, J. Qiu, R.L. Duncan, Am. J. Physiol. Cell Physiol. 275 (1998) C1591. [44] J.G. McGarry, J.K. Nulend, P.J. Prendergast, Biochem. Biophys. Res. Commun. 330 (2005) 341.