An improved model for exploring the effect of physicochemical properties of alginate-based microcapsules on their fibrosis formation in vivo

Abstract

In vivo microcapsule implantation suffers from varying degrees of inflammatory responses. Considering the complications of the implantation process and the individual variation in laboratory animals, a feasible model, constructed in vitro by culturing fibroblasts on the surfaces of different types of microcapsules, is established in the present study. A high consistency is shown between the cell adhesion on beads or microcapsules based on the developed model in vitro and their fibrosis formation after implantation in vivo, which can be used to evaluate the biocompatibility of the microcapsules. Moreover, the relationship between the surface properties of the microcapsules and the cell adhesion behavior was explored using this developed model. It was found that cell adhesion was aggravated with chitosan reaction, but was reduced with alginate neutralization. This was closely related to the variation trend of surface charges, but not surface roughness and the stiffness of alginate-based microcapsules during the process of chitosan coating and alginate neutralization. This model could therefore provide a rapid evaluation and guidance to tailor the optimal surface properties for long-term transplantation.

---

1. Introduction

Microcapsules have been extensively employed to encapsulate therapeutic cells and tissues for allogenic and xenogenic transplantation, owing to their advantage of semi-permeability.^1–3 The unique membrane on the surface of microcapsules not only allows the efficient exchange of small molecules between the transplanted cells and the native tissue,⁴ but also protects the encapsulated cells or tissues from being attacked by the immune cells and proteins in the host.^5–7 However, it is worth pointing out that the biocompatibility of the microcapsules, another critical issue determining the outcomes of the transplanted cells or tissues, remains a challenge, due to the complicated interactions between the microcapsules and the biological systems.⁸

It is well known that upon transplantation, the host responses to microcapsules usually occur in various degrees, including protein adsorption, cell adhesion, and even fibrosis formation on their surfaces.⁹ These reactions in turn produce some adverse influences on the encapsulated cells or tissues, leading to the failure of transplantation therapy.^10,11 Several factors, such as the purity of materials,^2,11 the size of the microcapsule,¹² surface properties,^13,14 the volume of the injection¹⁰ and implantation site,¹⁵ have been reported to be strongly related to the multistages of fibrosis formation in vivo. In particular, a growing body of investigation is being conducted to focus on the relationship between the surface properties of microcapsules, including their surface chemical composition, hydrophobicity, surface charge, topography and mechanical strength,^16–20 and fibrosis formation after implantation in vivo.^15,21,22 Taken together, animal experiments are still generally applied for the evaluation of the biocompatibility of the microcapsules, and a number of biologically relevant data have been achieved by taking advantage of in vivo transplantation in the past decades.^5,21,22 However, it is worth noting that some detailed mechanisms are difficult to interpret in animal models and many ambiguous questions still remain unanswered, due to the difference in species and the complex environment in vivo. For example, one study by Tam et al. has suggested that the outer alginate layer produces non-significant effects on the biocompatibility of microcapsules.²¹ In contrast, several other reports have highlighted the advantage of alginate coating, based on the fact that the direct contact of polycations with cells or tissues on the surface of microcapsules is the main pro-inflammatory cue to trigger down-stream inflammatory reactions.^1,22,23 Although these in vivo experiments have provided more biologically relevant data during the related investigations, the complexity, difficulty and even morality of the experiments often preclude their widespread use in this research field. For instance, there are numerous potential factors to influence the process from injection to recovery during the in vivo transplantation of microcapsules, such as the individual differences, especially the lack of a standard evaluation system and the difference in first-hand experience.^9,24 These all cause difficulty in making the results more adequately reproducible and intercomparable. Therefore, to address this issue, it is essential to establish an improved model system that is capable of better evaluating the biocompatibility of microcapsules in vitro.

In the present study, an in vitro model based on fibroblasts co-cultured with beads or microcapsules has been established, in order to better evaluate the effectiveness of bead or microcapsule transplantation in vivo, under more defined and reproducible conditions. By using this model system, the relationship between the surface properties of microcapsules and the behavior of cell adhesion is explored to combine their performance after transplantation in vivo. It is expected that this in vitro model system will provide a robust tool to gain a deeper insight into the interaction between surface physicochemical properties of alginate-based microcapsules and cell adhesion. This would greatly contribute to the better design of microcapsules with more biocompatible surfaces for long-term transplantation.

2. Experimental section

2.1. Materials

Sodium alginate (molecular weight 460 kDa; molar ratio of mannuronic acid to guluronic acid = 2/1) was obtained from Qingdao Crystal Salt Bioscience and Technology Corporation (Qingdao, Shandong, China). Chitosan (molecular weight 41 kDa; degree of deacetylation 90%) was degraded from the raw material (Yu Huan Chemical Plant, Zhejiang, China) in our laboratory. All other chemical reagents were analytical grade and used as received.

2.2. Preparation of alginate-based microcapsules

Both the calcium alginate beads and alginate–chitosan microcapsules were prepared as described in our previous study.²⁵ Briefly, 1.5% w/v sodium alginate solution was firstly extruded through a 0.5 mm needle into 0.1 M CaCl₂ solution by using an electrostatic droplet generator (YD-04, Dalian Institute of Chemical Physics, Chinese Academy of Sciences, China) to form calcium alginate gel beads (named “A”) with a diameter of 300 μm. After gelling for 30 min, calcium alginate gel beads were reacted with chitosan solution (0.5% w/v) for 12 min to form alginate–chitosan microcapsules (named “AC”) with 15 μm thickness on the surface. To prepare alginate–chitosan–alginate microcapsules (named “ACA”), AC microcapsules were immersed in the diluted alginate (0.15% w/v) for 10 min to form the outer layer of alginate.

2.3. Implantation and explantation of empty microcapsules

All experiments were conducted according to the guidelines of the Committee on Care and Use of Laboratory Animals, Liaoning province, China. This study was approved by the animal ethics committee of Dalian Medical University. Male C57BL/6 mice (purchased from the Laboratory Animal Center of Dalian Medical University, China) were maintained in the Animal Center of Dalian Medical University, and they were housed in 12 h darkness and 12 h light cycles at 24–26 °C, in a controlled, specific-pathogen free environment and fed standard chow and water.

To evaluate the inflammatory reaction caused by the transplantation of different types of beads or microcapsules, in vivo, twenty C57BL/6 mice in total (male, 6–8 weeks of age, weighing 18 to 22 g) were divided into 4 groups as follows: (1) the A bead group, (2) the AC microcapsule group, (3) the ACA microcapsule group and (4) the negative control group, with normal saline. Particularly, different groups of beads or microcapsules (12.5 ml kg⁻¹) were gently implanted into the peritoneal cavities of the mice, with a 20 G cannula via a syringe, while the mice treated with 0.9% NaCl aqueous solution (12.5 ml kg⁻¹) served as the negative control. These beads or microcapsules were retrieved after 14 days of transplantation through peritoneal lavage, by infusing 50 ml of 0.9% NaCl into the mouse peritoneal cavity. The recovery ratio (RR_f) and cell coverage ratio (CR_f) of the freely floating beads or microcapsules after transplantation were calculated using eqn (1) and (2) with a reference to the methodology described by Li et al.¹¹

where N_a represents the number of beads or microcapsules before transplantation, N_f represents the number of freely floating microcapsules retrieved after transplantation.

2.4. Histological examination

The retrieved microcapsules were collected and fixed in 4% neutral buffered paraformaldehyde solution overnight. The samples were then dehydrated in an ethanol series and embedded in paraffin. Paraffin-embedded sections were prepared at 5 mm of thickness and used for all staining. After rehydration with xylene and an ethanol series (from 100 to 70 vol%), the samples were stained with hematoxylin and eosin (H&E, Sigma-Aldrich). Slides were visualized and captured by a Nikon Eclipse TE 2000 Inverted Research Microscope (Nikon Corp. Japan).

2.5. Cell culture

Mouse fibroblast cell line L929 cells (RCB 0081, Cell Bank, Japan), were cultivated in RPMI-1640 medium supplemented with 10% fetal bovine serum (FBS) and 1% penicillin/streptomycin (P/S, all from Invitrogen, Carlsbad, CA) at 37 °C and 5% CO₂. The medium was replaced every other day and cells were sub-cultured regularly when they reached over 90% confluence.

2.6. Cellular adhesion on the surface of beads or microcapsules

L929 cells were harvested after trypsinization by 0.25% trypsin solution (Invitrogen) and resuspended in the culture medium to achieve single cell suspension with a final concentration of 1 × 10⁶ cells per ml. Cell density (1 × 10⁶ cells per well) used in the model was determined according to the highest level of the immune cell provoked by microcapsule implantation in the abdominal cavity. They were then seeded into 12-well plates, in which each well was filled with 60 μL of beads or microcapsules. All the 12-well plates used for this experiment were pre-coated with a thin layer of agarose gel (2% w/v) to prevent cellular adhesion to the bottom. The 12-well plate loaded with both the cells and beads or microcapsules was cultured at 37 °C and 5% CO₂ for 24 hours. For the initial 4 hours of cultivation, the cultures were gently shaken (for 5 minutes every half an hour). The morphology of cells that adhered to the surface of the beads or microcapsules was observed under a phase contrast microscope (Nikon Eclipse TE 2000, Nikon Corp. Japan). The samples were incubated with live/dead staining working solution composed of 2 μM calcein AM (Sigma-Aldrich) and 4 μM ethidium homodimer-1 (ED-1, Sigma-Aldrich) at 37 °C for 1 h. Preparations were scanned with a confocal laser scanning microscope (CLSM, Leica SP2, Heidelberger, Germany).

2.7. SEM observation

After 24 h of incubation, the beads or microcapsules with cell adhesion were washed in saline solution and fixed using paraformaldehyde (4% w/v) for 4 h at room temperature. Afterwards, the samples were rinsed with saline three times and then dehydrated with different concentrations of ethanol (30%, 50%, 70%, 80%, 90%, 100%, v/v), for 15 min at each concentration. At last, the samples were dried by vacuum drying. The morphology of cell adhesion on beads or microcapsules was observed by FE-SEM (JSM-7800F, JEOL, Japan).

2.8. DNA content assay

Cellular adhesion on the surface of the beads or microcapsules was quantitatively determined by DNA content analysis. After harvesting all the samples at indicated time points (store at −80 °C), the DNA content was measured by PicoGreen DNA Assay, following the protocol provided by the manufacturer (Molecular Probes, Eugene, OR). The fluorescence was measured with a microplate reader (BioTek, Synergy H1, USA) at excitation and emission wavelengths of 480 nm and 520 nm, respectively.

2.9. Surface roughness measurement

Surface roughness and surface morphology of the wet beads or microcapsules were measured by a noncontact, three-dimensional white-light optical interferometer (New-View 5020, ZYGO, USA). Freshly prepared beads or microcapsules were firstly loaded onto the object stage and their surface free water was gently removed by absorbent paper. The samples were vertically scanned and the interference fringes were recorded by CCD. The optical path difference could then be calculated according to the intensity change of the interference fringe, to precisely reflect the spherical surface roughness. Afterwards, the authentic surface roughness of the beads or microcapsules was obtained by spherical aberration correction, as described in our previous study.²⁶ The average value was obtained from at least three beads or microcapsules for each sample.

2.10. Stiffness measurement

The mechanical resistance of beads or microcapsules was characterized by ElectroForce (ELF3100, BOSE, USA), a high precision biomechanical test instrument mounted with a 250 g load cell. Tests were performed to compress a single bead or microcapsule with the speed of 0.05 mm s⁻¹, and set the probe from the position of −1.0 mm to −1.4 mm. A curve of force–displacement was then obtained. During the compression, the max deformation and force were at the position of −1.4 mm. The resistance force of the sample at 50% deformation was used to delegate the stiffness. All the average values were calculated using at least three beads or microcapsules of each sample.

2.11. Zeta potential measurement

Zeta potentials of samples were determined using a SurPASS Electrokinetic analyzer (Anton Paar GmbH, Austria) equipped with a cylindrical cell. The streaming potential was measured by using Ag/AgCl electrodes, a power measuring cell and the electrolyte circuit. For each measurement, approximately 1.0 ml of beads or microcapsules was first transferred into the glass cylinder of the measuring cell. Before starting the measurement, the beads or microcapsules were then rinsed with doubly distilled water to completely remove NaCl. The electrolyte (1 mM KCl solution) was forced through the measuring cell containing the sample. The pH value of the electrolyte solution was kept at ∼7.0 to simulate the physiological pH in vivo. The temperature was kept at 25 °C during the assessment of the streaming potential.²⁷ KCl solution (1 mM) served as the background electrolyte. The value of the zeta potential (ζ, mV) was obtained from the streaming potential measurements, based on the Smoluchowski equation, as follows: eqn (3):where U_str represents the streaming potential, p represents the pressure drop across the streaming channel, ε₀ represents the vacuum permittivity (8.854 × 10⁻¹² J⁻¹ C² m⁻¹), ε_r represents the dielectric constant of aqueous solution (78.3), η represents the electrolyte solution viscosity (0.8902 mPa s), and K_B represents the electrical conductivity of the bulk solution. Each sample was measured three times.

2.12. X-ray photoelectron spectroscopy (XPS)

XPS was applied to quantitatively analyze the elemental composition of the bead or microcapsule surface, except for hydrogen. To prepare the sample for XPS, the beads or microcapsules (∼1 ml) were firstly rinsed with sterile water to remove excess salt and then were dehydrated through gradient concentration of ethanol (30%, 50%, 70%, 90%, 100%, v/v).²¹ After being dried by a critical point dryer overnight, the elemental compositions of the samples were analyzed under a high resolution X-ray photoelectron spectrometer (ESCALAB250, Thermo VG, USA) with an Al Kα source operating at a take-off angle of 45° and the approximate depth of profile was 2–10 nm. Charge shift correction was conducted by setting the C_1s peak value of saturated hydrocarbons (C–C) to 284.6 eV. The surface atomic percentage was calculated using the XPS Peak Software 4.1 and the sensitivity factor provided by the manufacturer. High-resolution N_1s spectra were curve-fitted with 80% Gaussian – 20% Lorentzian peak shape, and peak integration was calculated with the subtraction of the Shirley background.²⁸ For each sample, three separate batches were analyzed (n = 3).

2.13. Statistical analysis

All reported values were averaged (n = 3 repeats, except for the specific experiments where explanations are provided) and expressed as mean ± standard deviation (SD). Statistical differences were determined by Student's two-tailed test and differences were considered statistically significant at p < 0.05.

3. Results

3.1. In vivo evaluation of cell coverage and fibrosis formation on the surface of beads or microcapsules

After 14 days of implantation in vivo, each type of bead or microcapsule was retrieved from the mouse peritoneal cavity, and their distinct morphologies were photographed as shown in Fig. 1A–C. Few cells were found to be attached to the surface of the retrieved A beads, contributing to their clean and smooth surfaces after implantation. Relatively smooth surfaces were also observed in the retrieved ACA microcapsules; however, their surfaces were found to be partially covered by some cells in vivo. Moreover, most of those ACA microcapsules look much darker than the A beads, which might be due to the fact that those chitosan molecules were competitively bound by the low molecular electrolytes in peritoneal fluid, such as HCO₃⁻ and HPO₄²⁻/H₂PO₄⁻.⁶ In contrast to these two groups, a large number of inflammatory cells was observed on the surface of the retrieved AC microcapsules, leading to their dirty appearance and preference for adherence to each other.

Fig. 1 Morphology of the retrieved A beads, AC and ACA microcapsules after 14 days of transplantation (A–C). (D) and (E) show their recovery ratio and the cell coverage ratio, respectively (*p < 0.05). Scale bar = 100 μm.

To quantitatively evaluate the retrieved beads or microcapsules after implantation in vivo, the recovery ratio (RR_f) and the cell coverage ratio (CR_f) of the freely floating retrieved beads or microcapsules were calculated after 14 days of transplantation. As shown in Fig. 1D and E, a highest recovery ratio and the lowest cell coverage ratio were found in the group of A beads, which might suggest a minor inflammatory reaction induced by the A beads, in comparison to the other two groups. In contrast, AC microcapsules exhibited a most evident inflammatory response, evidenced by their lowest retrieval rate and the highest cell coverage rate after explantation. This indicates that chitosan binding on the surface of the A beads might potentially increase the risk of proinflammatory response. However, when the surfaces of the AC microcapsules were further coated with alginate (namely ACA microcapsule), a decreased level of cell coverage was observed (in the ACA microcapsules), compared to that of the AC group. This data suggests that the positive effect of alginate neutralization could alleviate the inflammatory reaction after implantation.

In addition to the analysis on those retrievable beads or microcapsules above, a histological analysis was conducted in order to learn more details about those beads or microcapsules with fibrous overgrowth after implantation. As shown in Fig. 2, the formation of granulation/fibrous tissue over the surface of the beads or microcapsules was observed. Compared to either the A beads or ACA microcapsule group, where only a few immune-cells were attached, or slight cellular infiltration occurred after explantation, a thicker layer of fibrous tissue, predominantly consisting of fibrocytes and collagen, was found to cover around the surface of the AC microcapsules. This further indicated that chitosan reaction could increase the fibrosis formation, but alginate coating could reduce the cellular infiltration on the surface of the microcapsules.

Fig. 2 Representative images of H&E staining of A beads (A), AC (B) and ACA (C) microcapsules after 14 days implantation. Scale bar = 100 μm.

3.2. In vitro evaluation of the cell adhesion model

A fibroblast–microcapsule co-culture model in vitro was first established in the present study. After 24 hours of co-culture in vitro, a few cells were found to be adherent on the surface of the A beads, leaving most of the cells evenly suspended in the culture medium. All of the beads exhibited smooth surfaces, but obviously swelled morphology (compared to both the AC and the ACA groups) (Fig. 3A). A similar profile of cell attachment was also observed on the surface of the ACA microcapsules (Fig. 3C). However, when the fibroblasts were incubated with AC microcapsules, a large number of cells was found on the surface of the AC microcapsules (Fig. 3B).

Fig. 3 Morphology of cell adhesion on the A beads, AC and ACA microcapsules (A–C) after they were co-cultured with L929 cells for 24 hours. The DNA content in each experimental group was further measured after 24 hours of co-cultivation (D) (*p < 0.05). Scale bar = 100 μm.

To quantify the amount of cells that adhered to the surface of the beads or microcapsules, the DNA content in each experimental group was further analyzed using the PicoGreen DNA assay. As shown in Fig. 3D, the highest level of DNA concentration (nearly 60-fold higher) was exclusively observed in the AC microcapsules, when compared to that of the A beads and ACA microcapsules. However, no statistical difference in DNA content exists between the A bead group and the ACA group. These data are somewhat consistent with our observation of fibrosis formation on the surface of beads or microcapsules in vivo, suggesting the desirable potential of this fibroblast–microcapsule co-culture system in exploring the interfacial interactions between beads or microcapsules and cells in vivo.

In addition, the viability and morphology of the fibroblasts co-cultured with beads or microcapsules were further revealed by the CLSM and SEM images as shown in Fig. 4A–D. The majority of the cells adhering to the AC microcapsules displayed green fluorescence, suggesting their good viability during the co-cultivation. Furthermore, these cells exhibited a round shape and were tightly attached to the surface of AC microcapsules through several pseudopodia that formed during the co-cultivation. We analyzed that these pseudopodia on the cell-materials interface were indispensable to mediate or promote cell adhesion on the surface of the microcapsules. Also, immunofluorescence staining of the F-actin and vinculin protein, two important markers for focal adhesion,²⁹ provided further evidence to support the cell adhesion on the AC microcapsules (Fig. 4E and F), which suggested that the cells were able to adhere to the surface of AC microcapsules stably.

Fig. 4 CLSM images (A and B) show the viability of the adherent cell on the surface of the AC microcapsules (green denotes live cells and red denotes dead cells in live/dead staining, scale bar = 150 μm). The morphology of cell adhesion on the AC microcapsules (scale bar = 10 μm) is shown in SEM images (C), in which the pseudopodia formed by the fibroblasts (denoted by white arrows) are observed on the cell-material interface (D) (scale bar = 1 μm). Immunofluorescence staining of the F-actin (green, E) and vinculin (red, F) of cells adhered to AC microcapsules, scale bar = 10 μm.

3.3. Characterization of the surface physicochemical properties of the beads and microcapsules

3.3.1. Surface roughness of beads and microcapsules. As shown in Fig. 5, a similar surface roughness (around 30 nm) was observed in each group, including A beads, AC and ACA microcapsules. Furthermore, the two processes of polyelectrolyte complexing reaction, including chitosan reaction and alginate coating, exerted little influence on the surface roughness. It was suggested that the surface roughness of beads or microcapsules showed no significant difference and would not display many changes during the process of preparation, for microcapsules. Thus, it is unlikely that the surface roughness of alginate-based beads or microcapsules plays a role in determining the distinct outcomes of cell adhesion after implantation.

Fig. 5 Surface roughness of beads and microcapsules (A beads, AC and ACA microcapsules); determined by optical interferometry.

3.3.2. Stiffness of beads and microcapsules. To clarify whether the stiffness of beads or microcapsules has some influence on cell adhesion, the stiffness was further characterized through compression testing. As shown in Fig. 6A, with the compression by the probe, the pressure increased sharply. It was found that both the beads and microcapsules started to deform from the position of −1.2 mm, and this deformation of beads or microcapsules was up to 50% at the position of −1.4 mm. Hence, the resistance force responsible for 50% deformation was used to represent the stiffness of beads or microcapsules. The A beads were found to display the maximum force, up to 30.1 g, which is remarkably higher than that of AC microcapsules. However, no significant difference in the resistance force was observed between the AC and ACA microcapsules, in which the value of their resistance was only up to ∼6.0 g (Fig. 6B).

Fig. 6 The force–displacement curves of the beads and microcapsules (A) and their resistance force at 50% deformation (B) (*p < 0.05).

To further characterize whether or not the process of coating with chitosan solution (for AC microcapsule preparation) affects the inner hydrogel network of the microcapsules, FITC-labeled alginate was utilized to prepare each type of bead or microcapsule and the images were captured by CLSM. As shown in Fig. 7, a relatively homogeneous structure was observed in the A bead group, evidenced by the even distribution of fluorescence within the beads. In contrast, the fluorescence distribution within the AC microcapsules exhibited a unique pattern of “dense surroundings and sparse center”, indicating the hydrogel network formed within the AC microcapsules became more heterogeneous, as compared to the A beads. This phenomenon might be explained by the strong electrostatic interaction between the alginate and the chitosan. We determined that this kind of interaction promotes the liberation of some alginate molecules from the calcium alginate hydrogel network and leads to their migration from the inner part to the outer part of gel network. Thus, it is easy to understand that the outward migration of alginate might in turn be able to cause a decreased rigidity in the core of the AC microcapsules, when compared with that of the original A beads. With respect to the ACA microcapsule, the alginate coating exerts little influence on the inner hydrogel structure, and thereafter, no difference in the resistance force was observed between AC and ACA microcapsules.

Fig. 7 CLSM images show the distribution profile of FITC-alginate within the A beads (A) and AC microcapsules (B). Scale bar = 150 μm.

3.3.3. Surface charge of beads and microcapsules. To explore the surface charge of the beads or microcapsules, the surface potential was firstly measured by using a SurPASS Electrokinetic analyzer. As shown in Fig. 8A, a significantly distinct profile of zeta-potentials was observed in different types of beads or microcapsules. Specifically, the A beads displayed a zeta-potential of −4.67 ± 0.65 mV, while the zeta-potential of the AC microcapsules was much higher (−1.93 ± 0.11 mV). Compared to the AC microcapsules, a remarkably decreased zeta-potential was observed in the ACA microcapsules, where the value of their zeta-potential was down to −6.17 ± 0.72 mV. This data suggests that neutralization of those unbound chitosan molecules in the AC microcapsules with alginate molecules (for ACA preparation) could alter the surface potential of the microcapsules significantly. Moreover, the zeta-potential of three groups of beads or microcapsules all showed negative charge, which was significantly distinct from the performance of LBL self-assembly reported before.³⁰ It was supposed that not only are the chitosan molecules complexed and inside diffusion occurs, but alginate molecules are also complexed and outward migration takes place during the membrane formation (Fig. 7).

Fig. 8 (A) Zeta potential and (B) positively charged, surface amino groups of beads and microcapsules (A beads, AC and ACA microcapsules) (*p < 0.05).

When the cells adhere to the surface of the microcapsules, the local surface charge, especially the surface positive charge, plays an important role. Considering the fact that the zeta-potential only denotes the net charge of the beads or microcapsules, high resolution X-ray photoelectron spectrometry was then employed to analyze the surface positive charge distribution. The percentage of protonated amino groups was quantified by analyzing the N_1s spectrum, which was curve-fitted with 80% Gaussian – 20% Lorentzian peak shapes. As shown in Fig. 8B, no amino group was detected in the A beads because no chitosan molecules were involved in this system. However, the percentage of protonated amino groups increased significantly after their reaction with chitosan molecules, evidenced by the improved protonated amino content on the surface of AC microcapsules, when compared to that of the A beads. When these AC microcapsules were further neutralized by alginate molecules (resultant ACA microcapsules), the percentage of protonated amino groups significantly decreased (p < 0.05). The data suggest that the outermost layer of the ACA microcapsules, i.e. the alginate layer, is capable of forming a shielding layer to reduce the surface exposure of protonated amino groups.

4. Discussion

In the present study, a simple and feasible in vitro model that could be employed to better evaluate the biocompatibility of microcapsules in vivo was established through co-culturing fibroblasts with different types of alginate-based beads or microcapsules. The reason that fibroblasts were chosen is mainly due to their critical role during the process of fibrosis formation after transplantation in vivo, since a number of investigations have demonstrated that the recruitment of immune-cells and fibroblasts would finally lead to overgrowth of the capsules as a consequence.³¹ Here, we found that a large number of fibroblasts preferred to adhere to the surface of AC microcapsules, which is distinct from either A beads or the ACA microcapsule group. Since the only major difference between the AC and ACA microcapsule group is the outer layer coating by chitosan or alginate molecules, we therefore concluded that it is chitosan, but not alginate that plays the dominant role in promoting cell attachment on the surface of beads or microcapsules. These results in vitro are highly consistent with the results of cell coverage ratio and retrieved ratio in vivo. Therefore, to some extent, the cell adhesion model in vitro established in this study could effectively simulate the fibrosis formation in vivo, and might be potentially used to evaluate and predict the effectiveness of microcapsule transplantation in vivo.

To better understand the underlying mechanism that affects cell adherence to the surface of the alginate-based microcapsules, in the present study, surface properties, including surface roughness, stiffness and surface charge, were further characterized. Particularly, when the processing parameters change, more than one property would be changed simultaneously in the microcapsule system, which could therefore be the main reason impeding the mechanism between the surface properties and cell interactions on the microcapsules. However, according to our quantitative analysis, surface roughness had no significant change during the process of chitosan reaction and alginate neutralization. Our previous cognition about the effect of surface roughness on protein adsorption and cell adhesion has also been updated, due to our recent research on the novel characterization of surface morphology of beads or microcapsules in wet and in situ conditions.^25,26 Therefore, it is unlikely that surface roughness caused the different cell behaviors on the beads or microcapsules, since each group retained similar surface roughness during the preparation process of the microcapsules.

Regarding the effect of stiffness on cell adhesion, many studies have demonstrated that an enhanced stiffness of the substrate is advantageous to cell adhesion and cell shape,^19,32 especially for fibroblasts.³³ It is generally considered that cells are able to sense and respond to the stiffness of their substrate, via an integrin–focal adhesion–cytoskeleton mediated pathway (also called mechanotransduction).¹³ Here, we did observe the enhanced stiffness of A beads when compared to the AC or ACA group; however, no significant difference in stiffness was found between the AC and ACA microcapsule group. Thus, it is obvious that the stiffness cannot be used to account for their distinct profiles of cell adherence on beads or microcapsules. Combining the outcomes of cell responses both in vivo and in vitro, it seems that the stiffness for alginate-based beads or microcapsules produced a negligible effect on cell adhesion during transplantation.

However, in contrast to the negligible effect produced by either roughness or stiffness, the surface positive charge, another important property of alginate-based microcapsules was found to play a dominant role in determining cell adhesion on the surface of beads or microcapsules. Although the zeta potential of beads or microcapsules showed a net negative charge, the –NH₃⁺ groups exposed on the surface should be considered, since cell–microcapsule interactions were always influenced by the local surface positive charge.³⁴ As the evidence of the comparatively similar variation trend between surface positive charge and the cell adhesion suggests, the surface positive charge plays a significant role in cell adhesion. To explain this phenomenon, we analyzed the electrostatic interactions between the cells and extracellular microenvironment, which might take part in the modulation of cell behaviors.³⁴ It is known that the cell membrane displays net negative charges.¹³ Thereafter, the microcapsules with protonated amino groups might be able to directly cause electrostatic attraction to the cell membrane and promote cell attachment. During the transplantation, more –NH₂ groups on the surface of the microcapsules were transformed into –NH₃⁺ groups, since the pH value was demonstrated to decrease when the inflammation response was triggered.²⁷

In addition, many reports have demonstrated that protein adsorption could promote the integrin aggregation and focal adhesion formation. Also, it is important to note that the surface positive charge could facilitate protein adsorption (showed in Fig. S1†).^35,36 Therefore, it was supposed that cell adhesion on microcapsules was not only directly mediated by the electrostatic attraction, but might also be indirectly affected via the protein adsorption, due to the surface positive charge.

Altogether, our study demonstrates that polycations play dominant roles in guiding cell adherence on microcapsules, because they contribute directly to the surface positive charge, and this would in turn trigger an inflammatory response, such as cell coverage and fibrosis formation during transplantation. We propose that two strategies could be adopted to design microcapsules with decreased surface positive charge, which might reduce cell adhesion and improve the biocompatibility of microcapsules. One is to reduce polycation reactions directly during the preparation of microcapsules, which could be achieved by decreasing the reaction time or polycation concentration. However, it should be kept in mind that the fundamental performances of microcapsules, including their stability and immune isolation properties, cannot be altered during the adjustment of polycation reactions. Another strategy is to generate a shield layer in order to decrease the percentage of protonated amino, such as coating the alginate outer layer or grafting polyethylene glycol (PEG) onto the surface of microcapsules in situ, to improve the protein repellence and biocompatibility.²⁵ Therefore, they all could be able to reduce the protein adsorption and cell adhesion to improve the biocompatibility appropriately.

5. Conclusions

In this article, a simple but effective in vitro co-cultured model was established through co-culturing fibroblasts with different types of alginate-based beads or microcapsules. This model has been demonstrated to be capable of partially simulating cell coverage and fibrosis formation on the surface of microcapsules in vivo. By using this model, we found that it was the surface positive charge of the alginate-based microcapsules, rather than their surface roughness or stiffness that exerted profound influence on cell adhesion and the inflammatory response after microcapsule transplantation. Thus, the co-culture model established here might provide a robust tool to better evaluate the biocompatibility of the transplanted alginate-based microcapsules in vitro. This study should greatly enlighten our future research on the in-depth mechanism exploration of the complex interactions between surface properties and cell adhesion.

Acknowledgements

This work was partially funded by the National Basic Research Program of China (grant 2012CB720801), the National Natural Science Foundation of China (NSFC 81173125), and the Ocean Public Welfare Scientific Research Project of China (grant 201405015-3).

References

1. P. de Vos, M. M. Faas, B. Strand and R. Calafiore, Biomaterials, 2006, 27, 5603–5617.
2. P. de Vos, C. G. van Hoogmoed, J. van Zanten, S. Netter, J. H. Strubbe and H. J. Busscher, Biomaterials, 2003, 24, 305–312.
3. P. de Vos, M. Bucko, P. Gemeiner, M. Navratil, J. Svitel, M. Faas, B. L. Strand, G. Skjak-Braek, Y. A. Morch, A. Vikartovska, I. Lacik, G. Kollarikova, G. Orive, D. Poncelet, J. L. Pedraz and M. B. Ansorge-Schumacher, Biomaterials, 2009, 30, 2559–2570.
4. P. L. Lam, K. K. H. Lee, Y. W. Ho, R. S. M. Wong, S. W. Tong, C. H. Cheng, K. H. Lam, J. C. O. Tang, Z. X. Bian, R. Gambari, S. H. L. Kok and C. H. Chui, RSC Adv., 2014, 4, 14109.
5. J. T. Wilson, W. Cui, X.-L. Sun, C. Tucker-Burden, C. J. Weber and E. L. Chaikof, Biomaterials, 2007, 28, 609–617.
6. L. Chen, Y. Zhang, S. Li, X. Wang, N. Li, Y. Wang, X. Guo, S. Zhao, W. Yu, G. Sun, Y. Liu and X. Ma, J. Biomed. Mater. Res., Part A, 2014, 102, 2408–2416.
7. S. M. Naqvi, S. Vedicherla, J. Gansau, T. McIntyre, M. Doherty and C. T. Buckley, Adv. Mater., 2016, 28, 5662–5671.
8. J. T. Wilson and E. L. Chaikof, Adv. Drug Delivery Rev., 2008, 60, 124–145.
9. S. Franz, S. Rammelt, D. Scharnweber and J. C. Simon, Biomaterials, 2011, 32, 6692–6709.
10. R. M. Hernandez, G. Orive, A. Murua and J. L. Pedraz, Adv. Drug Delivery Rev., 2010, 62, 711–730.
11. S. Li, Y. Zhang, L. Chen, N. Li, H. Xie, X. Guo, S. Zhao, W. Yu, Y. Lv, G. Lv, H. Wu and X. Ma, J. Biomed. Mater. Res., Part A, 2015, 103, 2333–2343.
12. S. Sakai, C. Mu, K. Kawabata, I. Hashimoto and K. Kawakami, J. Biomed. Mater. Res., Part A, 2006, 78, 394–398.
13. E. Kaivosoja, G. Barreto, K. Levon, S. Virtanen, M. Ainola and Y. T. Konttinen, Ann. Med., 2012, 44, 635–650.
14. S. Ponce, G. Orive, R. Hernandez, A. R. Gascon, J. L. Pedraz, B. J. de Haan, M. M. Faas, H. J. Mathieu and P. de Vos, Biomaterials, 2006, 27, 4831–4839.
15. G. J. Bakeine, A. Bertolotti, M. Latina, T. Congiu, U. Prati, L. Roveda, F. Trotta, M. Tormen, E. D. Fabrizio, G. Carlini, A. Facoetti and R. Nano, J. Biomed. Mater. Res., Part A, 2007, 83, 965–969.
16. N. M. Alves, I. Pashkuleva, R. L. Reis and J. F. Mano, Small, 2010, 6, 2208–2220.
17. L. Bacakova, E. Filova, M. Parizek, T. Ruml and V. Svorcik, Biotechnol. Adv., 2011, 29, 739–767.
18. X. Huang, N. Lin, R. Hang, X. Wang, X. Zhang and B. Tang, Carbohydr. Polym., 2013, 92, 1064–1070.
19. T. Yeung, P. C. Georges, L. A. Flanagan, B. Marg, M. Ortiz, M. Funaki, N. Zahir, W. Ming, V. Weaver and P. A. Janmey, Cell Motil. Cytoskeleton, 2005, 60, 24–34.
20. X. Guo, L. Yu, L. Chen, H. Zhang, L. Peng, X. Guo and W. Ding, J. Mater. Chem. B, 2014, 2, 1760.
21. S. K. Tam, S. Bilodeau, J. Dusseault, G. Langlois, J. P. Halle and L. H. Yahia, Acta Biomater., 2011, 7, 1683–1692.
22. P. de Vos, M. Spasojevic, B. J. de Haan and M. M. Faas, Biomaterials, 2012, 33, 5552–5559.
23. A. King, B. Strand, A.-M. Rokstad, B. Kulseng, A. Andersson, G. Skjåk-Bræk and S. Sandler, J. Biomed. Mater. Res., Part A, 2003, 64A, 7.
24. S. A. Safley, H. Cui, S. Cauffiel, C. Tucker-Burden and C. J. Weber, J. Diabetes Sci. Technol., 2008, 2, 8.
25. J. Zheng, H. Xie, W. Yu, M. Tan, F. Gong, X. Liu, F. Wang, G. Lv, W. Liu, G. Zheng, Y. Yang, W. Xie and X. Ma, Langmuir, 2012, 28, 13261–13273.
26. H. Zheng, H. Xie, H. Wu, F. Wang, X. Liu, W. Yu and X. Ma, Colloids Surf., A, 2015, 484, 457–462.
27. P. de Vos, B. J. de Haan, J. A. Kamps, M. M. Faas and T. Kitano, J. Biomed. Mater. Res., Part A, 2007, 80, 813–819.
28. H. Y. Yeh and J. C. Lin, J. Biomater. Sci., Polym. Ed., 2008, 19, 291–310.
29. P. Kanchanawong, G. Shtengel, A. M. Pasapera, E. B. Ramko, M. W. Davidson, H. F. Hess and C. M. Waterman, Nature, 2010, 468, 580–584.
30. C. Peng, Q. Zhao and C. Gao, Colloids Surf., A, 2010, 353, 132–139.
31. M. D. Swartzlander, A. D. Lynn, A. K. Blakney, T. R. Kyriakides and S. J. Bryant, Biomaterials, 2013, 34, 952–964.
32. D. E. Discher, P. Janmey and Y. L. Wang, Science, 2005, 310, 1139–1143.
33. I. Hopp, A. Michelmore, L. E. Smith, D. E. Robinson, A. Bachhuka, A. Mierczynska and K. Vasilev, Biomaterials, 2013, 34, 5070–5077.
34. J. Kim, D. H. Kim, K. T. Lim, H. Seonwoo, S. H. Park, Y. R. Kim, Y. Kim, Y. H. Choung, P. H. Choung and J. H. Chung, Tissue Eng., Part C, 2012, 18, 913–923.
35. Y. H. Chen, Y. C. Chung, I. J. Wang and T. H. Young, Biomaterials, 2012, 33, 1336–1342.
36. C. R. Wittmer, J. A. Phelps, W. M. Saltzman and P. R. Van Tassel, Biomaterials, 2007, 28, 851–860.

---

Footnotes

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c6ra19294g

‡ The authors contributed equally.

---