Surface patterned hydrogel film as a flexible scaffold for 2D and 3D cell co-culture

Abstract

Herein we report a facile route for the preparation of surface patterned dynamic hydrogel films, which are not only a matrix to encapsulate one type of cell in 3D but also a substrate to support aligned aggregates of magnetic silica rods to adhere another type of cell in 2D. This enables the composite hydrogel films to be flexible scaffolds for engineering multi-cellular tissues.

---

Chemical cues and topography of a scaffold are able to affect cell behaviour and hence the quality of tissue regeneration.^1–4 For example, cell-growth direction, such as the alignment of cells in vessel endothelium, is recognized as a key issue in controlling the structure and composition of the particular tissue.^5,6 Thus the construction of extracellular substrates is of great interest to render the engineering of an oriented cell layer. Many approaches have been developed for the fabrication of a surface with fibril like pattern in order to regulate the orientation of cells during their proliferation.^7–12 However, in most of the microfabrications, the substrate for supporting the patterned surface has no additional bio-function rather than providing mechanical properties. Natural tissues are normally multi-cellular systems of two or more types of cell. Therefore in recent years, hierarchical structured double-layered scaffolds were developed,¹³ for example by electrospun technique, which contain a top layer formed by aligned fibres and a bottom matrix by randomly packed fibres for the co-culture of two cell lines, and thereby duplicate the architecture of native tissue like the peripheral nerve.⁶ Nevertheless, it is still a challenge to develop flexible scaffolds for simultaneously control growth of different cells.

Hydrogel is known as an ideal extracellular matrix (ECM) for culturing cell in three dimensions (3D cell culture).^3,14,15 The soft nature of hydrogel can more favourably mimic the property of soft tissues. Besides, hydrogel has the capacity of incorporation of bioactives, such as growth factors, to influence the embryonic development, proliferation and even the differentiation of the encapsulated cells.^16–18 Hydrogel for cell co-culture was also reported.^19–21 Bi-layered and multilayered bulk hydrogels have been synthesized to mediate the formation of distinctive structure of osteochondral tissue and articular cartilage.^22,23 However, in some cases it requires the proliferation of cell in 2D and 3D concurrently instead of a 3D/3D co-culture, such as for the regeneration of tissues containing endothelial system. Recently, the modification of hydrogel surface with cell-adhesive patterns for directional 2D culture has been documented in literatures.^7,24–26 Nevertheless, the fabrication involves organic solvent or other toxic additives. Therefore cells must be seeded after the fabrication and thus can no longer be entrapped inside the gel substrate and proliferated in 3D.

In previous work, we developed in situ forming hydrogel based on glycol chitosan (GC) and benzaldehyde capped poly(ethylene oxide) derivatives (OHC–PEO–CHO) through the formation of benzoic-imine linkage (ESI, Scheme S1†).²⁷ The GC and benzaldehyde functionalized polymers are proven to be non-toxic to cells. By Schiff's reaction, imine bond forms as a dynamic covalent bond which makes the sol–gel transition reversible and allows the hydrogel as an injectable matrix for cell implantation,²⁸ giving a promising example of using dynamic interactions on the construction of hydrogel based ECM. In this work, we further exploited additional dynamic forces, e.g. electrostatic interaction, to conjugate cell-adhesive silica rods onto the surface of the dynamic hydrogel. Besides, the of magnetic responsiveness further managed the orientation of the rods and thus achieved an aligned pattern on the substrate. By this methodology, we obtained composite hydrogel films as a scaffold that enables 2D/3D compartmentalized co-culture and meanwhile the guidance of cell growth.

Scheme 1 illustrates the fabrication procedure. MC3T3-E1 and L929 cells were used as models for testing the capacity of cell encapsulation and co-culture. The preparation details including the synthesis of silica rods are given in ESI.† Briefly, MC3T3-E1 encapsulated GC/OHC–PEO–CHO hydrogel film was first prepared (Scheme 1). And the maximum compress stress of the hydrogel is 130 kPa at a stain value of 30% which can be adjusted by changing the polymer composition (Fig. S1†). Then the magnetic silica rods, functionalized by cyclic (Arg-Gly-Asp-D-Phe-Lys) (RGD), were immobilized onto the hydrogel surface and aligned in magnetostatic field for the following loading of L929 cells. The dynamic nature of imine bond endows the self-healing ability to the GC/OHC–PEO–CHO hydrogel (Scheme S1†). Pipe shaped scaffold could be made from the hydrogel films, by rolling a rectangular shaped film and fusing the met edges together for about 5 min (Scheme 1). The characteristic will simplify the engineering of personalized scaffolds with desired dimensions.

Scheme 1 Illustrative route for the fabrication of the surface patterned GC/OHC–PEG–CHO composite film for 2D and 3D cell co-culture.

The magnetic silica rods have an average length of 1.5 μm and an average diameter of 150 nm. Both amine group and RGD peptide were functionalized to the rods (Fig. S1†), in order to enhance the affinity to either hydrogel or the surface cells. The orientation of the rods can be seen in optical images (Fig. 1), which was induced by the application of magnetostatic field. And the density of the rod covered surface can be adjusted by the loading concentration of the rod dispersion (Fig. 1a and b). Enlarged SEM image reveals the aggregation state of the rods. While end-to-end assembly leads to the directional alignment, side-to-side assembly of the rods forms the fibril like structure (Fig. 1d). Unlike covalent bonding, the rod immobilization on gel surface is considered via dynamic interactions, majorly attributed to the electrostatic force, generated by the presence of a polyanion, i.e. poly(acrylic acid) (PAA), which could bind the interface between the substrate and the pattern, by bridging the amino groups in both the GC and the silica rods. Rod-like material was selected to pattern the hydrogel surface due to it can increase the area of the interface, compared to round particles, in case lying on the substrate, to gain tighter conjugation. The firmness of the pattern, i.e. the fibrils, on the hydrogel surface was examined by washing the surface with PBS for several times, after which no obvious change of the morphology was viewed (Fig. S4†).

Fig. 1 Optical images of aligned rods on GC/OHC–PEO–CHO hydrogel surface with different density, i.e. 0.65 μg cm⁻² (a) and 2.6 μg cm⁻² (b) (scale bar 20 μm). SEM on the surface of freeze-dried composite hydrogel firm (scale bar 20 μm) (c). Enlarged SEM of aggregated silica rods (scale bar 1 μm) (d) and TEM image (inset) of single silica rod (scale bar 100 nm).

The safety of the RGD functionalized silica rods was checked by CCK-8 assay on L929 cell line (Fig. S2†). The rods didn't appear cytotoxicity with concentration up to 1.6 mg mL⁻¹, with a criteria of >90% viability after 24 h incubation. Besides, a novel characteristic of the silica rods was discovered, that is, they cannot be favourably taken up by the cells. As shown in Fig. S3,† major portion of the rods were distributed on the membrane of the cells, while few of them were found in the cytoplasm. It is known that RGD peptide can help the interaction of nano- or micro-particles with cell membrane during culture.²⁹ However, the internalization was not facilitated in this regime which is probably due to the non-spherical shape of the rods.³⁰ Large length-to-diameter ratio of the rods might have caused anisotropic interactions in the axial and the radial directions which influenced the inward folding of cell membrane.³¹ Nevertheless, the rationale of this phenomenon is to be further clarified. And whether the reducing of cell uptake would affect the long-term safety also need a deep study.^32,33

The cell adhesion on the hydrogel surface was examined using confocal laser scattering microscopy (CLSM). Fig. 2 shows that the L929 cells stayed on the modified surface sooner after they were seeded, e.g. within 0.5 h, whereas poor cell affinity is observed for the neat hydrogel surface. The result proves that the silica rods provided the necessary affinity in the adhesion of cells. Thus the orientation of the cells was critically determined by the patterning of the rods. On the surface with an aligned pattern, major portion of the cells are oriented adhered, with an elongated morphology along the direction of the fibrils (Fig. 2c and d). Comparably, fusiform shaped cells together with triangle shaped cells are seen distributed on the surface with randomly oriented rods, fitting the topologies formed by the aggregation of the rods (Fig. 2b). In addition to the alignment, the number of fibrils also influences the level of the orientation and the aspect ratio of the cells. Statistical analysis shows that a surface covered by compact fibril array can lead nearly 45% of the cells to have an orientation of angle within 10° parallel to the direction of the fibrils, while the value is down to 27% in case the pattern of fibrils becomes sparse and loose (Fig. 2e). Meanwhile, the elongation of the cells is larger on the compact pattern (Fig. 2d).

Fig. 2 Merged CLSM images of L929 cells cultured on the surface of neat GC/OHC–PEO–CHO hydrogel (a), hydrogel surface modified by randomly distributed silica rods (2.6 μg cm⁻²) (b) and the surface patterned by the aligned silica rods (0.65 μg cm⁻² and 2.6 μg cm⁻²) (c, d). The pictures was taken within 0.5 h after surface seeding. The bright field images were taken after the focus was set on the rod layer between the L929 cells and hydrogel surface to visualize the status of the silica rods. The scale bar is 50 μm. Orientation angle distribution of the L929 cells on hydrogel surfaces with different patterns which was measured after an incubation of 4 h (counted as the percentage of total measured cells) (e). Dish curves illustrate the profile of the distributions.

It is worth noticing that the adhesion of cells led to the deformation of the pattern on the hydrogel surface, when comparing the image plotted in Fig. 1b with 2d. A possible explanation is that the rods were not permanently grafted to the hydrogel surface via covalent bonding. Therefore, the cell enable to “move” the rods by the stretching force generated from the surface protein such as integrin.³⁴ It can be expected that compared to solid substrates, such movement of substrate could reduce the tensile force loaded on the cell membrane. However, the effect of such force relaxation on cell behaviour is almost unknown. Considering there are many tissues can be recognized as soft materials, further understanding in this aspect might be interesting.

The cell growth on the patterned hydrogel surface was monitored for 7 days of incubation (Fig. 3a, c and e). It can be seen that on day 4, the proliferation of L929 cells already made the cells firmly connected with each other and covered most of the surface area, resulting in the formation of a monolayer (Fig. 3c). By staining with acridine orange (AO) and propidium iodide (PI), the emission in AO channel (500–580 nm) confirms that the cells are in good condition, while no emission from PI channel (600–650 nm) were recorded during the period of experiment. However, it was noted that portion of cells detached from the hydrogel surface and dispersed in the medium after prolonged incubation for 7 days. At this time point, the proliferation also decreases from 217% on day 4 to 195% (Fig. 3h), clearly indicating the occurrence of cell apoptosis during the time interval. Furthermore, the packing of the cells becomes slightly disordered because of the overlap of the cells after the incubation of 7 days (Fig. 3e). Considering high coverage of cells was already viewed on the surface after 4 days of incubation (Fig. 3c), the later on decrease of cell population can be attributed to the density-dependent inhibition of cell growth.³⁵ Control experiment by incubating the cells in a normal culture dish was performed and gained faster proliferation for the first 4 days, but subsequently more serious decrease of cell number on the 7th day of the incubation (Fig. 3h), which confirms the density effect on the cell growth. It can be presumed that the apoptosis made the dead cells no longer adhesive to the substrate and therefore they were not monitored by the CLSM. Similar proliferation trend was gained from the growth of the L929 cells on the hydrogel surface with randomly oriented pattern (Fig. 3h, S5†), implying that in this work the morphology of pattern is not critical factor of the cell growth of L929.

Fig. 3 Merged CLSM images of L929 cells on the surface (a, c, e) and MC3T3-E1 cells in the hydrogel matrix (b, d, f) of the GC/OHC–PEO–CHO composite hydrogel film after the cells were incubated for 4 h (a, b), 4 days (c, d) and 7 days (e, f). The scale bar is 50 μm. Images from individual channels of the 4 h incubated film are given in the ESI (Fig. S6†). Side view of cell loaded composite film from CLSM after incubated for 4 days (g). Proliferation of the L929 and MC3T3-E1 cells co-cultured in the composite hydrogel film for 4 and 7 days (h). The proliferation of L929 on culture dish was performed as a reference.

Meanwhile, CLSM also reveals the morphology of MC3T3-E1 inside the polymer matrix (Fig. 3b, d and f). Round shape cells were viewed attributed to the hydrophilic nature of the ECM, i.e. the GC/OHC–PEO–CHO hydrogel.³⁶ The proliferation of the MC3T3-E1 was confirmed within an incubation of 7 days, gaining 134% and 156% of cell growth on day 4 and 7, not as fast as that of the fibroblast cell, mainly due to the instinctive slower growth rate of the osteoblastic cell (Fig. 3h).³⁷ From a side view of the CLSM image, one can easily distinguish the layers as well as the interface of the two types of cell by seeing the difference on the thickness and the density of the cells (Fig. 3g). The results again support that the composite film enables the compartmentive loading and proliferation of different cell lines simultaneously.

With an objective of tissue regeneration, handling property of the scaffold needs to be considered for the operation. During the transplantation, the scaffold may have to leave the surrounding medium. Hydrogel could protect the encapsulated cells from dehydration and prolongs the duration for the coadaptation. However, whether a hydrogel will help the survival of cells on its surface has not been investigated. To clarify this, we removed the surrounding medium of the composite film with L929 cells seeded on the patterned surface, and exposed it to air for up to 4 h (Fig. S7†). 99.6 ± 2.7% of cell viability was obtained (p > 0.05 vs. normal cultured cells). Comparably, only 6.7 ± 0.4% of cells survived after the media were removed if they were seeded on a culture dish.

Conclusions

In conclusion, we report a facile route to pattern a hydrogel surface while cells have been encapsulated in the gel matrix for the co-culture of different types of cell in two layers. As an example, the composite hydrogel film allows the directional growth of L929 cells in 2D on surface and meanwhile a 3D proliferation of MC3T3-E1 cells inside the gel matrix. Besides, the GC/OHC–PEO–CHO hydrogel also plays as a reservoir of ECM to prolong the cell viability in case the scaffold is out of medium. The results imply that the composite hydrogel is promising as a kind of compartmentalized scaffold for building multi-cellular structures. Further works are warranted on the investigation of their potential utilities such as for the construction of blood vessel.

Acknowledgements

This work is financially supported by the National Natural Science Foundation of China (NSFC) (21274155, 51233007, 51473169) and MOST (2012CB933200).

Notes and references

1. C. J. Bettinger, R. Langer and J. T. Borenstein, Angew. Chem., Int. Ed., 2009, 48, 5406.
2. X. Yao, R. Peng and J. Ding, Adv. Mater., 2013, 25, 5257.
3. M. Verhulsel, M. Vignes, S. Descroix, L. Malaquin, D. M. Vignjevic and J. L. Viovy, Biomaterials, 2014, 35, 1816.
4. R. J. Wade, E. J. Bassin, W. M. Gramlich and J. A. Burdick, Adv. Mater., 2015, 27, 1356.
5. Y. Wang, S. Jiang, H. Shi, W. Zhang, J. Qiao, M. Wu, Y. Tian, Z. Niu and Y. Huang, Chem. Commun., 2013, 49, 10421.
6. J. Xie, M. R. MacEwan, W. Liu, N. Jesuraj, X. Li, D. Hunter and Y. Xia, ACS Appl. Mater. Interfaces, 2014, 6, 9472.
7. L. Y. Jiang and Y. Luo, Soft Matter, 2013, 9, 1113.
8. M. Guvendiren and J. A. Burdick, Adv. Healthcare Mater., 2013, 2, 155.
9. J. Ge, H. Lee, L. He, J. Kim, Z. Lu, H. Kim, J. Goebl, S. Kwon and Y. Yin, J. Am. Chem. Soc., 2009, 131, 15687.
10. J. H. Slater, J. S. Miller, S. S. Yu and J. L. West, Adv. Funct. Mater., 2011, 21, 2876.
11. J. Lu, F. Zheng, Y. Cheng, H. Ding, Y. Zhao and Z. Gu, Nanoscale, 2014, 6, 10650.
12. X. Zan, S. Feng, E. Balizan, Y. Lin and Q. Wang, ACS Nano, 2013, 7, 8385.
13. K. G. Battiston, J. W. C. Cheung, D. Jain and J. P. Santerre, Biomaterials, 2014, 35, 4465.
14. S. Y. Choh, D. Cross and C. Wang, Biomacromolecules, 2011, 12, 1126.
15. P. M. Kharkar, K. L. Kiick and A. M. Kloxin, Chem. Soc. Rev., 2013, 42, 7335.
16. K. Ladewig, Expert Opin. Drug Delivery, 2011, 8, 1175.
17. S. Fu, P. Ni, B. Wang, B. Chu, L. Zheng, F. Luo, J. Luo and Z. Qian, Biomaterials, 2012, 33, 4801.
18. T. T. Lau and D. A. Wang, Nanomedicine, 2013, 8, 655.
19. Y. Aizawa and M. S. Shoichet, Biomaterials, 2012, 33, 5198.
20. X. Guo, J. Liao, H. Park, A. Saraf, R. M. Raphael, Y. Tabata, F. K. Kasper and A. G. Mikos, Acta Biomater., 2010, 6, 2920.
21. E. J. Sheehy, T. Vinardell, C. T. Buckley and D. J. Kelly, Acta Biomater., 2013, 9, 5484.
22. K. Kim, J. Lam, S. Lu, P. P. Spicer, A. Lueckgen, Y. Tabata, M. E. Wong, J. A. Jansen, A. G. Mikos and F. K. Kasper, J. Controlled Release, 2013, 168, 166.
23. L. H. Nguyen, A. K. Kudva, N. S. Saxena and K. Roy, Biomaterials, 2011, 32, 6946.
24. P. Liu, J. Sun, J. Huang, R. Peng, J. Tang and J. Ding, Nanoscale, 2010, 2, 122.
25. R. Peng, X. Yao and J. Ding, Biomaterials, 2011, 32, 8048.
26. Z. Guo, K. Hu, J. Sun, T. Zhang, Q. Zhang, L. Song, X. Zhang and N. Gu, ACS Appl. Mater. Interfaces, 2014, 6, 10963.
27. C. Ding, L. Zhao, F. Liu, J. Cheng, J. Gu, S. Dan, C. Liu, X. Qu and Z. Yang, Biomacromolecules, 2010, 11, 1043.
28. B. Yang, Y. Zhang, X. Zhang, L. Tao, S. Li and Y. Wei, Polym. Chem., 2012, 3, 3235.
29. C. Cruje and B. D. Chithrani, J. Nanosci. Nanotechnol., 2015, 15, 2125.
30. S. Mitragotri and J. Lahann, Nat. Mater., 2009, 8, 15–23.
31. J. B. Gilbert, J. S. O'Brien, H. S. Suresh, R. E. Cohen and M. F. Rubner, Adv. Mater., 2013, 25, 5948.
32. C. Fu, T. Liu, L. Li, H. Liu, D. Chen and F. Tang, Biomaterials, 2013, 34, 2565.
33. A. Tarantini, R. Lanceleur, A. Mourot, M. T. Lavault, G. Casterou, G. Jarry, K. Hogeveen and V. Fessard, Toxicol. In Vitro, 2015, 29, 398.
34. A. G. Karakecili, T. T. Demirtas, C. Satriano, M. Gumusderelioglu and G. Marletta, J. Biosci. Bioeng., 2007, 104, 69.
35. A. I. McClatchey and A. S. Yap, Curr. Opin. Cell Biol., 2012, 24, 685.
36. Y. Wu, Z. Zhao, Y. Guan and Y. Zhang, Acta Biomater., 2014, 10, 1965.
37. H. M. Elgendy, M. E. Norman, A. R. Keaton and C. T. Laurencin, Biomaterials, 1993, 14, 263.

---

Footnote

† Electronic supplementary information (ESI) available: Experimental details and supplementary figures. See DOI: 10.1039/c6ra11249h

---