Layer-by-layer self-assembled laminin/fucoidan films: towards better hemocompatibility and endothelialization

Abstract

Rapid endothelialization is an effective solution to thrombosis and restenosis in vascular-implanting materials. Surface modification is a concise and impactful approach to improve biocompatibility including endothelialization and hemocompatibility. In this paper, we report on a layer-by-layer self-assembly of laminin/fucoidan multilayers for the improvement of hemocompatibility and endothelialization. The chemical changes and wettability of the surface assembled films were investigated by static water contact angle measurement, X-ray photoelectron spectroscopy, colorimetric and immunosorbent assay. The real time assembly process was in situ monitored by quartz crystal microbalance with dissipation. The platelets adhesion/activation test showed that significantly fewer platelets were adherent and activated on the surface of assembled three layer laminin/fucoidan samples with the outermost layer of fucoidan. Furthermore, cell culture tests indicated that the adhesion of human umbilical vein endothelial cells and endothelial progenitor cells were promoted on the surface of an assembled five layer laminin/fucoidan sample with the outermost layer of laminin. It is concluded that layer-by-layer self-assembled laminin/fucoidan multilayers are a promising candidate for a biofunctional coating applied to vascular implants.

---

1. Introduction

Cardiovascular diseases are the primary cause of death in humans. Drug-eluting stents (DES), widely used to treat artery atherosclerosis, have achieved an initial success since they can reduce inflammation and restenosis rate.^1–4 However, in-stent restenosis and late stent thrombosis are still observed after DES has been implanted.^5–8 The endothelium plays a crucial role in the prevention of in-stent thrombosis. Inoue et al.⁸ and Waksman et al.⁹ emphasized that delayed re-endothelialization and impaired endothelial function are linked to stent thrombosis and restenosis. Consequently, rapid endothelialization^10,11 and functionalization on the stent surface are the strategic processes for a new generation of coronary stent systems.¹²

Up to now, various approaches have been performed to improve hemocompatibility and accelerate endothelialization for devices implanted in vascular systems. For instance, heparin with excellent antithrombotic property has been immobilized commonly to improve hemocompatibility.^13,14 Yao et al.¹⁵ fabricated a hybrid small-diameter vascular graft, releasing heparin constantly, by the co-electrospinning technique, to promote the growth of human umbilical vein endothelial cells (HUVECs). Kang et al.¹⁶ fabricated a stent with EPC specificity and an antibody against vascular endothelial cadherin for vascular re-endothelialization. Laminin is one of the best candidates for endothelialization; ss a major component of basement membrane, it plays a significant role in cellular adhesion, growth,¹⁷ migration,¹⁸ and anchorage for cells to the basement surface.¹⁹ For example, laminin has been used to regulate vascular cells behavior,²⁰ and accelerated neovascularization and endothelialization.²¹ In terms of the hemocompatibility, low molecular weight fucoidan, a heparin-like compound,²² showed significant anticoagulant²³ and antithrombotic activities^24–26 which is a favorable candidate for blood contact material modification as a coating, however, it has not been extensively investigated.

To introduce bioactive molecules to biomaterials and medical devices, covalent binding and physical adsorption technologies, as two main approaches, have been most commonly used for surface modification. The layer-by-layer (LBL) self-assembly technique, a common physical adsorption method, is an eminent candidate for surface biological modification to construct multilayer membranes with oppositely charged biomacromolecules, such as polysaccharides, proteins and enzymes.^27,28 Heparin,²⁹ fibronectin³⁰ and chitosan^31,32 are widely used to improve anticoagulation and endothelialization via LBL self-assembly technique. Furthermore, the LBL self-assembly technique, based on electrostatic interaction³⁰ enabling many layers to be formed, is a simple and effective approach to introduce a significant amount of biomolecules to the multiple biomaterial surface and maximise the biomolecule activity.

In this study, we fabricated laminin/fucoidan (Ln/F) multilayer membranes by an LBL self-assembly technique. The Ln/F multilayer membranes create an alternate interface with positive and negative charges. The cytocompatibility and anticoagulation properties of the multilayers were evaluated in vitro, which is used to confirm their potential in biological surface modification for the blood-contacting materials.

2. Experimental

2.1 Materials

Glass plates, acetone, ethanol, hydrogen peroxide (H₂O₂) and concentrated sulfuric acid (H₂SO₄) were purchased from Sigma-Aldrich. Phosphate buffer saline (PBS, 0.067 M, pH 7.4), laminin (from human fibroblasts, Sigma-Aldrich, USA, molecular weight about 805 kDa) and fucoidan (from fucus vesiculosus-crude, Sigma-Aldrich, USA, molecular weight about 20–200 kDa) were diluted to a concentration of 50 μg mL⁻¹ and 100 μg mL⁻¹ with PBS solution respectively. 3-Aminopropyltriethoxysilane (APTE), 1-ethyl-3-(3-dimethylaminopropyl)carbodiimide (EDC), N-hydroxysuccinimide (NHS), toluidine blue O (TBO) and acid orange 7 were supplied by Sigma-Aldrich Chemical Co. All the other reagents were analytical reagent grade.

2.2 Preparation of the Ln/F multilayer membranes

The Ln/F films were fabricated on glass plates. First, the glass plates (1 × 1 cm) were sonicated with acetone, ethanol, deionized water and finally dried at 80 °C. The cleaned glass plates were then activated in “piranha solution” (mixture of H₂SO₄/H₂O₂ solution, 7 : 3) for 2 h at 80 °C, then rinsed thoroughly three times with deionized water. The “piranha solution” activated glass plates were denoted as G@OH. Subsequently G@OH were immersed in a 3% (v/v) ethanol solution of (3-aminopropyl)triethoxysilane (APTE) for 12 h at 50 °C with gentle shaking. After that, the carriers were washed with deionized water and kept in a 120 °C oven for 2 h to enhance the binding of APTE with the carrier (sample denoted as G@OH@APTE or G@APTE). The aminolyzed substrates were then dipped in a mixed solution A–B (1 : 10 v/v) for 2 h at 37 °C, where A is a crosslinking solution of EDC/NHS/MES (0.1 : 0.05 : 0.05 mol%) and B is a solution of laminin (50 μg mL⁻¹). The carrier plate was then subsequently rinsed with PBS, and the sample is denoted as G@A@1Ln. The laminin substrates were then placed into a fucoidan solution (100 μg mL⁻¹) for 15 min at 37 °C (sample denoted as G@A@1F, an abbreviation of G@A@1L@1F), then rinsed thoroughly with deionized water. Several bilayers of laminin and fucoidan were manufactured on glass by repeating the assembly procedure mentioned above, and in this way, samples of G@A@2Ln, G@A@2F, G@A@3Ln, G@A@3F, G@A@5Ln and G@A@5F were prepared (Scheme 1).

Scheme 1 Schematic view of formation of bilayer structures G@A@1Ln, G@A@1F, G@A@5Ln and G@A@5F from G@APTE

2.3 Quartz crystal measurement of the Ln/F multilayer membranes

Quartz crystal microbalance with dissipation (QCM-D, Q-Sense AB, Sweden) is a reliable tool for real-time monitoring of LBL multilayer buildup.³³ A SiO₂ quartz crystal was initially treated with UV-ozone for at least 10 min to clean and sterilize the surface, then settled in the measurement chamber and ethanol was injected as a buffer for equilibrium. The temperature was maintained at 25 ± 0.2 °C in all experiments. A 3% ethanol solution of APTE was injected at 20 μL min⁻¹ continuously until the adsorption reached equilibrium and stabilized, then rinsed with ethanol and PBS. Subsequently, laminin solution was injected until no variation appeared in the adsorption curves. PBS was then pumped in again and fucoidan solution was injected thereafter at the same speed for the next equilibrium. Laminin and fucoidan were then alternately pumped into the chamber for the buildup of multilayers on the quartz crystal surface. Adsorbed mass change ΔM during the film assembly was calculated by the Sauerbrey equation:³³

ΔM = −CΔf/n (1)

where the constant C is 17.7 ng cm⁻² Hz⁻¹, Δf is the frequency change, and n is the overtone (n = 5). Mass changes vs. time (ΔM–t) curves were recorded to monitor the assembly process of Ln/F multilayer membranes.

2.4 Chemical characterization of Ln/F films

2.4.1 Surface characterization of amino groups. Amino group density on the surface of the silicon alkylated surface was characterized quantitatively by acid orange II³⁵ (Sigma-Aldrich, USA) as follows: amino groups can form compounds at a pH 3 with acid orange II, which can be dissolved by 1 mM NaOH solution, and the absorbance of the supernatant can be measured. Finally, according to the standard curve method, surface amino density (nmol cm⁻²) can be obtained.

2.4.2 Chemical compositions. The chemical structures of the Ln/F films were investigated by Fourier transform infrared spectroscopy (FTIR, Philips, Netherlands) in the range of 4000 to 1000 cm⁻¹ (FT80 accessory). The surface chemical compositions were characterized by X-ray photoelectron spectroscopy (XPS, Perkin-Elmer 16PC) with a monochromatic Al-Kα excitation radiation (1486.6 eV). Binding energies were calibrated by using adventitious carbon (C 1s = 284.7 eV). Peaks were fitted using Xpspeak 4.1 to obtain the high resolution information.

2.4.3 Wettability. The contact angle analysis with double-distilled water in this study was performed using a contact angle goniometer (JY-82, Tianjin, China). More than three different points were measured for each sample in order to obtain statistical averages. All the measurements were performed at ambient temperature.

2.4.4 Surface characterization of laminin. The amount of laminin was determined by enzyme-linked immunosorbent assay (ELISA) as follows: laminin modified glass samples were blocked with sheep serum (from Sigma-Aldrich, 1/100 dilution in PBS) for 30 min, and then washed with PBS three times and incubated with mouse anti-human laminin antibody (1/250 dilution in PBS) for 1 h. Subsequently, the samples were rinsed with PBS again and incubated with horseradish peroxidase labelled goat anti-mouse antibody (1/100 dilution in PBS) for another 1 h. Finally, after washing and dying with TMB reagent, the absorbance was obtained at 450 nm.

2.4.5 Surface characterization of fucoidan. The sulfonic groups on the surface can form complexes with toluidine blue O (TBO) dye (purchased from Sigma-Aldrich), as described by Smith et al.³⁶ The structure of fucoidan is similar to heparin with sulfonic groups. To determine the density of fucoidan immobilized on the surface, the modified glass samples with Ln/F were immersed in 1 mL TBO with a concentration of 50 mg L⁻¹ for 2 h, n-hexane (3 mL) was then added and the mixture was shaken well to ensure uniformity of the dye. The absorbance at 530 nm was then measured by UV spectrophotometry (BIO-TEK instruments, USA) and the amount of immobilized fucoidan was calculated from the calibration curve of free fucoidan.

2.5 In vitro hemocompatibility evaluation

2.5.1 Activated partial thromboplastin time. The anticoagulation property of Ln/F assembled samples was determined by means of an APTT assay, which is a frequently used method to evaluate the intrinsic coagulation system. For the APTT test, the samples were first immersed in 500 μL PPP (poor platelet plasma) and incubated at 37 °C for 30 min. Then 100 mL incubated PPP was transferred to a test tube and 100 mL APTT reagent was added to the same test tube and incubated at 37 °C for 3 min. Subsequently, 100 mL of an aqueous 0.025 M CaCl₂ solution was added. The suspension was stirred magnetically and the coagulation time was determined at 37 °C using a coagulation instrument (Hospitex Diagnostics, Italy).

2.5.2 Platelet adhesion. In vitro platelet adhesion assay was carried out to further evaluate the anticoagulant effect of Ln/F modified samples.²⁰ First, fresh human venous blood was centrifuged at 1500 rpm for 15 min to obtain platelet-rich plasma (PRP). Then, 60 μL PRP was dropped onto each sample (1 × 1 cm) surface and incubated for 2 h at 37 °C in humidified air. After thoroughly washing the sample with PBS, 2.5% glutaraldehyde solution was added onto each sample surface to fix the adherent platelets for 4 h. Finally, the samples were rinsed three times with PBS, dehydrated and dealcoholized. The morphology and density of adhered platelets were observed by scanning electron microscopy (SEM, Quanta 200; FEI, Holland).

2.6 Cellular compatibility evaluation

2.6.1 HUVECs and EPCs isolation and culture. HUVECs were isolated from human umbilical vein by the explanted technique using enzymatic digestion according to Jaffe et al.³⁴ Following isolation, HUVECs were suspended in Dulbecco’s Modified Eagle Medium/Nutrient Mixture F-12 containing 20% fetal bovine serum (FBS), transferred to a culture flask and cultured in a humidified atmosphere (95%) containing 5% CO₂ at 37 °C. Each cell test was carried out using HUVEC passage number 1. The medium was changed every 2 days.

EPCs were isolated from the bone marrow of Sprague-Dawley rats according to Wang et al.²⁰ The bone marrow extracted from Sprague-Dawley rats was cultured in α-Modified Eagle’s Medium (α-MEM) supplemented with 10% FBS at 37 °C under 5% CO₂. When high purity bone marrow stem cells (BMSCs) were obtained, 10 ng mL⁻¹ vascular endothelial growth factor (VEGF) were mixed to α-MEM, containing 10% FBS, for controlling the oriented differentiation of BMSCs to EPCs. After 2 weeks, the cells were considered as EPC according to the identification of cell morphology and specific markers.

2.6.2 Cell seeding. Cytocompatibility of the Ln/F films were evaluated by adhesion and proliferation of HUVECs and EPCs in vitro. The samples were placed inside 24-well polystyrene culture dishes, and ECs and EPCs were seeded on the surfaces of the samples, respectively at an identical density of 3 × 10⁴ cells mL⁻¹ and incubated at 37 °C under 5% CO₂ for 1 day. The medium was removed and the samples were gently washed twice with PBS. Subsequently, the adherent cells were fixed with 4% glutaraldehyde for 12 h. After that, the cells on the sample surfaces were stained with 50 μL rhodamine at room temperature for 20 min under dark condition, rinsed gently with PBS and observed by fluorescence microscopy (Olympus IX51).

2.7 Data analysis

The data were analyzed by the software SPSS 11.5 (Chicago, IL). The data were expressed as mean ± standard deviation (SD). The probability value (p) < 0.05 was considered to be statistically significant.

3. Results and discussion

3.1 Amino groups characterization

Fig. 1 shows the amino density on the surface of glass, G@OH and G@OH@APTE. A significant increase of amino density was seen on the G@OH@APTE surface compared to glass and G@OH surfaces, which suggested glass surface was successfully modified by APTES.

Fig. 1 Amino density of the samples of glass, G@OH and G@OH@APTE (means ± SD, n = 3, **P < 0.01).

3.2 Chemical characterization of Ln/F film

Fig. 2 shows the FT-IR spectra of the laminin and the fucoidan films. After the glass platelets were immersed in “piranha solution” and subsequently (3-aminopropyl)triethoxysilane solution, stretching vibrations of –OH and –NH₂ were observed at 3443 and 3366 cm⁻¹. After laminin and fucoidan assembly, a substantial difference in the chemical structures were observed. An amide II absorption peak (C=O stretching vibrations) at 1666 cm⁻¹ appeared for the sample of G@A@1Ln, which is due to immobilization of laminin, while for G@OH@A@1F, the presence of antisymmetric stretching vibrations of O–S–O at 1415–1390 cm⁻¹, and the strengthened stretching vibrations of –OH at 3050–3570 cm⁻¹ revealed that immobilization of fucoidan had occurred.

Fig. 2 FTIR spectra of glass, glass@OH, G@OH@APTE, G@OH@A@1Ln (also denoted G@A@1Ln) and G@OH@A@1F (also denoted as G@A@1F).

The Ln/F films assembled on glass plates were analyzed by XPS. Fig. 3(a) and (c) shows that a N 1s peak appeared in the survey spectra of G@OH@A@5F, for which the Si 2s and Si 2p peaks were weaker than samples without assembling laminin or fucoidan (see Fig. 3(d)), which is due to that laminin with nitrogen chemical groups, were assembled on the glass surface. The Si peaks are weaker since a film covered the glass surface. Fig. 3(b) shows O 1s XPS scan spectra and reveals large differences of binding energy. For the glass sample, the O 1s signal is from the O–Si bond (532.0 eV), and is shifted to a higher binding energy about 532.7 eV, this corresponds to hydroxyl oxygen (O–H) of glass modified by “piranha” solution. Then it shifted to a lower binding energy of 532.45 eV on binding to APTE, corresponding to the Si–O–Si bond of G@OH@APTE. For G@OH@A@5F, the observed binding energy of 531.5 eV is for C=O and S–O bonds of laminin and fucoidan molecules, respectively. As shown in Fig. 4, the water contact angle (WCA) values of glass was 37.4° and increased to 53.2° on the G@OH@APTE sample, which is attributed to the effect of abundant amino-NH₂ on the surface treated by APTES. WCA values of the samples with immobilized Ln/F is increased significantly compared to the glass sample. In particular, WCA values of the G@A@5F sample is the largest among all modified samples. Laminin is composed of various combinations of α, β and γ chains,¹⁹ each of which possess a different wettability due to their differing structure, with increased hydrophilicity being the result of exposed hydrophilic group domains. Also, the alteration of surface roughness is another reason for the change of WCA values after Ln/F immobilization.³⁷

Fig. 3 (a) X-Ray photoelectron spectra of glass, glass@OH, G@OH@APTE and G@OH@A@5F; (b–d) show detailed X-ray photoelectron spectra.

Fig. 4 Water contact angles (mean ± SD, n = 3, *P < 0.05).

3.3 QCM-D for monitoring assembly process

QCM-D was applied to monitor in real-time the assembly process of laminin and fucoidan on a quartz crystal surface. Mass shift–time curves (Fig. 5) clearly demonstrated the LBL buildup process of the Ln/F coating. Upon injection of the laminin or fucoidan solution alternatively, the mass on the crystal increased, indicating the adsorption of the laminin or fucoidan molecules onto the crystal surface. The formation of the ladder-like QCM-D traces suggested that layer-by-layer structures of the Ln/F membranes were constructed on the glass surface.

Fig. 5 Mass changes with the number of assembled layers obtained from QCM-D.

3.4 Determination of immobilized laminin and fucoidan

The amount of laminin was determined by ELISA. As shown in Fig. 6(a), no significant changes of O.D values could be observed among the samples assembled with different layers of laminin, which suggests the amount of laminin that could be assembled per layer on the samples is almost constant. However, an obvious increase of O.D value was seen on the samples assembled with laminin, compared to sample G@APTE.

Fig. 6 Surface characteristics; (a) the O.D value (at 450 nm) of laminin on the surface (laminin as the outermost layer); (b) the amount of sulfonic groups on the surface (fucoidan as the outermost layer) (mean ± SD, n = 3, *P < 0.05, **P < 0.01).

The sulfonic groups of fucoidan can form complexes with TBO dye. According to Fig. 6(b), the concentration of sulfonic groups increases with the increase of number of fucoidan layers (G@A@1F, G@A@3F, G@A@5F), compared to glass and G@APTE. However, no significant changes could be observed when the fucoidan layers were increased from 5 to 7. This indicates that the amount of fucoidan that could be immobilized onto the surface reached a maximum value.

3.5 Hemocompatibility

3.5.1 Activated partial thromboplastin time. In order to evaluate anticoagulant activity comprehensively an APTT test was applied (Fig. 7). It can be concluded that with the increase of the number of Ln/F layers, the APTT value becomes larger. The APTT value of G@A@5F was significantly larger compared with glass and G@A@1F. Fucoidan has been widely proved its excellent anticoagulant activity,^38–40 and the APTT test could partly characterize the hemocompatibility of laminin/fucoidan multilayers.

Fig. 7 APTT test of all samples.

3.5.2 Platelet adhesion and activation. In vitro platelet adhesion and activation test is used to investigate the blood compatibility of Ln/F surface. As shown in Fig. 8, the glass and G@APTE surfaces facilitated much more platelet adhesion and almost of the platelets were in the aggregation and pseudopodium state (Fig. 8(a)). Besides, for the same number of layers assembled, it is noted that platelets adhered more on the surfaces with laminin as the outermost layer than for surfaces with fucoidan as the outermost layer. This is attributed to that laminin, as one of three major components of the subendothelial matrix, will support platelets adhesion but not activation by integrin VLA-6.⁴¹ On the surfaces with fucoidan as the outermost layer, there was the least (*P < 0.05) platelet adhesion and activation on the surface of G@A@3F among all samples (Fig. 8(b)). According to Zhu et al.²⁴ and Pereira et al.⁴² fucoidan can inhibit thrombin catalyzed fibrinogen cleavage by a complex mechanism, and further inhibit platelet adhesion.^43,44 Furthermore, a small amount of platelets were adhered and spread on the surfaces of G@A@5Ln, compared to the surfaces of G@A@1Ln and G@A@3Ln, which was due to the influence of fucoidan under the laminin layer. It is noteworthy that fucoidan is not totally covered by laminin, as in the film of assembled Ln/F, fucoidan or laminin molecules crossed with each other, instead of being a compact coating, and the activity of the Ln/F film thus results from the synergetic effect of laminin and fucoidan.

Fig. 8 Platelets adhered on different samples; (a) SEM morphology; (b) platelet count.

3.6 Cytocompatibility of HUVECs and EPCs

For application of films in vascular stents or grafts, rapid regeneration of the endothelium is crucial to the success of their implantation.⁴⁵ There is convincing evidence in vivo that the vascular homing of endothelial progenitor cells contributes to rapid endothelial regeneration.¹¹ HUVECs and EPCs were seeded on different sample surfaces to investigate the influence of self-assembled Ln/F on the proliferation behaviors. A significant increase (*P < 0.05) of HUVECs proliferation on G@A@5Ln was observed after incubation of 1 day (Fig. 9(a) and (b)), which confirms the promoting of endothelial cells adhesion when laminin and fucoidan are assembled to a certain amount. In the same way, the amounts of EPCs on G@A@5Ln were significantly more (*P < 0.05) than that on glass, G@APTE surface and the other Ln/F samples, after incubation for 1 day (Fig. 9(c) and (d)). There is more EPCs adhesion and fully spreading on the samples for an outermost layer of laminin, compared with that for the outermost layer of fucoidan. That could be explained by that laminin shows specific binding sites to cells,^18,46,47 thereby facilitating HUVECs and EPCs adhesion,^17,48,49 which make laminin essential for devices applied in vascular applications.

Fig. 9 Cellular compatibility evaluation results; (a) and (c): immunofluorescence staining of ECs and EPCs after 1 day culture; (b) and (d): related cell counting results of each sample (mean ± SD, n = 3, *P < 0.5).

4. Conclusion

In this study, Ln/F multilayered films were constructed on a glass surface by the layer-by-layer assembly technique. Such Ln/F assembled films were found to show differing anticoagulation and cytocompatibility responses with the alteration of layers and top molecules: generally, the films with fucoidan as the outermost layer displayed better anticoagulation properties and lower cytocompatibility than that with the laminin as the outermost layer. Especially, the G@A@3F sample with fucoidan as the outermost layer showed the best effect of inhibiting platelet adhesion. On the other hand, the increase of laminin layers enhances the cytocompatibility of the film. Especially, the G@A@5Ln sample showed both anticoagulation properties and cytocompatibility of HUVECs and EPCs significantly, which provides a possible solution for the rapid endothelialization and anticoagulation simultaneously. Thus, such a film may have a great potential for vascular-implanting applications in the future.

Acknowledgements

The authors gratefully acknowledge the assistances from Mr Xiaohua Zhu and Ms Xin Li. This work was financially supported by National Natural Science Foundation of China (No. 81330031 and No. 81401522), and Key Basic Research Program (No. 2011CB606204).

References

1. A. T. Ong, E. P. McFadden, E. Regar, P. P. de Jaegere, R. T. van Domburg and P. W. Serruys, J. Am. Coll. Cardiol., 2005, 45, 2088–2092.
2. M. Cardona, S. Mirabet, V. Martin, V. Marti, M. Sabate and A. Bayes-Genis, Med. Clin., 2012, 138, 562–564.
3. Y. Q. Zhu, L. Edmonds, X. Zhao, X. L. Chen, C. M. Hu, Y. S. Cheng and W. G. Cui, RSC Adv., 2014, 4, 34405–34411.
4. A. Santos, M. S. Aw, M. Bariana, T. Kumeria, Y. Wang and D. Losic, J. Mater. Chem. B, 2014, 2, 6157–6182.
5. G. Nakazawa, M. Vorpahl, A. V. Finn, J. Narula and R. Virmani, JACC: Cardiovasc. Imaging, 2009, 2, 625–628.
6. A. Giraldo, A. Dannoura, I. Pereira, P. Dash, J. Gibbins, G. Brooks and K. Bicknell, Heart, 2013, 99, A105–A106.
7. A. T. L. Ong, E. P. McFadden, E. Regar, P. P. T. de Jaegere, R. T. van Domburg and P. W. Serruys, Eur. Heart J., 2005, 26, 640.
8. T. Inoue, K. Croce, T. Morooka, M. Sakuma, K. Node and D. I. Simon, JACC: Cardiovasc. Interventions, 2011, 4, 1057–1066.
9. R. Waksman, B. Bhargava, G. S. Mintz, R. Mehran, A. J. Lansky, L. F. Satler, A. D. Pichard, K. M. Kent and M. B. Leon, J. Am. Coll. Cardiol., 2000, 36, 65–68.
10. X. K. Ren, Y. K. Feng, J. T. Guo, H. X. Wang, Q. Li, J. Yang, X. F. Hao, J. Lv, N. Ma and W. Z. Li, Chem. Soc. Rev., 2015, 44, 5745.
11. J. L. Chen, Q. L. Li, J. G. Xu, L. Zhang, M. F. Maitz and J. Li, J. Mater. Chem. B, 2015, 3, 2615–2623.
12. F. Otsuka, A. V. Finn, S. K. Yazdani, M. Nakano, F. D. Kolodgie and R. Virmani, Nat. Rev. Cardiol., 2012, 9, 439–453.
13. K. Christensen, R. Larsson, H. Emanuelsson, G. Elgue and A. Larsson, Biomaterials, 2001, 22, 349–355.
14. J. Zhao, Y. Chen, S. Yang, S. Wu, R. Zeng, H. Wu, J. Zhang, Z. Zha and M. Tu, Mater. Sci. Eng., C, 2016, 58, 133–141.
15. Y. Yao, J. Wang, Y. Cui, R. Xu, Z. Wang, J. Zhang, K. Wang, Y. Li, Q. Zhao and D. Kong, Acta Biomater., 2014, 10, 2739–2749.
16. C.-K. Kang, W.-H. Lim, S. Kyeong, W.-S. Choe, H.-S. Kim, B.-H. Jun and Y.-S. Lee, Colloids Surf., B, 2013, 102, 744–751.
17. S. N. Ge, J. Y. Chen, Y. X. Leng and N. Huang, Key Eng. Mater., 2007, 342–343, 305–308.
18. T. Cui and J. Li, Wound Repair Regen., 2013, 21, A19.
19. J. Tzu and M. P. Marinkovich, Int. J. Biochem. Cell Biol., 2008, 40, 199–214.
20. J. Wang, Y. Chen, T. Liu, X. Wang, Y. Liu, Y. Wang, J. Chen and N. Huang, Appl. Surf. Sci., 2014, 317, 776–786.
21. S. K. Williams, L. B. Kleinert and V. Patula-Steinbrenner, J. Biomed. Mater. Res., Part A, 2011, 99A, 67–73.
22. T. C. de Azevedo, M. E. Bezerra, G. Santos Mda, L. A. Souza, C. T. Marques, N. M. Benevides and E. L. Leite, Biomed. Pharmacother., 2009, 63, 477–483.
23. M. T. Ale and A. S. Meyer, RSC Adv., 2013, 3, 8131–8141.
24. Z. Zhu, Q. Zhang, L. Chen, S. Ren, P. Xu, Y. Tang and D. Luo, Thromb. Res., 2010, 125, 419–426.
25. S. D. Anastyuk, N. M. Shevchenko, S. P. Ermakova, O. S. Vishchuk, E. L. Nazarenko, P. S. Dmitrenok and T. N. Zvyagintseva, Carbohydr. Polym., 2012, 87, 186–194.
26. .
27. S. Meng, Z. Liu, L. Shen, Z. Guo, L. L. Chou, W. Zhong, Q. Du and J. Ge, Biomaterials, 2009, 30, 2276–2283.
28. P. Zhang, J. Qian, Y. Yang, Q. An, X. Liu and Z. Gui, J. Membr. Sci., 2008, 320, 73–77.
29. T. Taniguchi, K. H. Kyung and S. Shiratori, RSC Adv., 2015, 5, 107488–107496.
30. G. Li, P. Yang and N. Huang, Phys. Procedia, 2011, 18, 112–121.
31. J. Fu, J. Ji, W. Yuan and J. Shen, Biomaterials, 2005, 26, 6684–6692.
32. Y. Liu, T. He and C. Gao, Colloids Surf., B, 2005, 46, 117–126.
33. H. Deligöz and B. Tieke, Colloids Surf., A, 2014, 441, 725–736.
34. E. A. Jaffe, R. L. Nachman, C. G. Becker and C. R. Minick, J. Clin. Invest., 1973, 52, 2745–2756.
35. P. Hamerli, Biomaterials, 2003, 24, 3989–3999.
36. P. Smith, A. Mallia and G. Hermanson, Anal. Biochem., 1980, 109, 466–473.
37. E. A. Vogler, Adv. Colloid Interface Sci., 1998, 74, 69–117.
38. J. Wang, Q. Zhang, Z. Zhang, H. Song and P. Li, Int. J. Biol. Macromol., 2010, 46, 6–12.
39. J.-G. Koo, Y.-S. Choi and J.-K. Kwak, Korean J. Fish. Aquat. Sci., 2001, 34, 515–520.
40. S. Colliec, A. Fischer, J. Tapon-Bretaudiere, C. Boisson, P. Durand and J. Jozefonvicz, Thromb. Res., 1991, 64, 143–154.
41. A. Sonnenberg, P. W. Modderman and F. Hogervorst, Nature, 1988, 336, 487–489.
42. M. S. Pereira, B. Mulloy and P. A. S. Mourao, J. Biol. Chem., 1999, 274, 7656–7667.
43. L. S. E. P. W. Castro, T. D. Pinheiro, A. J. G. Castro, M. D. N. Santos, E. M. Soriano and E. L. Leite, J. Appl. Phycol., 2015, 27, 1315–1325.
44. J. Dürig, T. Bruhn, K.-H. Zurborn, K. Gutensohn, H. D. Bruhn and L. Béress, Thromb. Res., 1997, 85, 479–491.
45. Z. Yang, Q. Tu, M. F. Maitz, S. Zhou, J. Wang and N. Huang, Biomaterials, 2012, 33, 7959–7971.
46. K. M. Malinda, M. Nomizu, M. Chung, M. Delgado, Y. Kuratomi, Y. Yamada, H. K. Kleinman and M. L. Ponce, Thromb. Res., 1999, 13, 53–62.
47. T. Cui, L. Tang and J. Li, Wound Repair Regener., 2011, 19, 19.
48. M. Gonzales, S. B. Hopkinson and J. C. R. Jones, Mol. Biol. Cell, 2001, 12, 194a.
49. F. P. He, X. P. Wang, O. Maruyama, R. Kosaka, Y. Sogo, A. Ito and J. D. Ye, J. R. Soc. Interface, 2013, 10, 20130014.

---

Footnote

† Co-first author.

---