Formation of MG-63 cell spheroids on parylene N film
Parylene N film was deposited thermally on a polystyrene surface. As shown in Fig. 1a, the chemical structure of the parylene N film was similar to that of polystyrene, lacking polar chemical functional groups. To characterize the surface properties of the parylene N film, the contact angle was measured and compared with those of a conventional cell culture plate, polystyrene surface, and UV-treated parylene N film.
The surface of UV-treated parylene N film contains several types of oxygen-bearing functional groups that increase its hydrophilicity4. The UV-treated parylene N was used for the comparison with other surfaces because this surface showed a significant influence on the cell proliferation of osteoblast- and neuron-like cells as well as the cell differentiation of neuron-like cells in the previous work5,6. As shown in Fig. 1b, the contact angles of the polystyrene surface, conventional cell culture plate, parylene N film, and UV-treated parylene N film were 86.7 ± 3.5°, 80.1 ± 2.8°, 81.1 ± 2.2°, and 68.6 ± 2.6° (n = 5), respectively. These results show that UV-treatment rendered parylene N’s surface hydrophilic, while the surface of the parylene N film was similarly hydrophobic to that of conventional cell culture matrices.
The surface roughness of polystyrene surfaces, conventional cell culture dishes, parylene N, and UV-treated parylene N was estimated using atomic force microscopy. As shown in Fig. 1c, the estimated average roughness ranged between 2.2 and 5.6 nm, demonstrating a similarly smooth surface for the four kinds of surfaces, in comparison with the size of cells in the microscale. As reported in previous work5, after UV-irradiation, the addition of oxygen species to the surface of parylene-N and parylene-C films was analyzed by XPS analysis, and the oxygen-bearing functional groups were characterized to be hydroxyl, carbonyl, and carboxylic acids, using FT-IR (Fig. 1d). These results showed that the hydrophilicity of UV-treated parylene N film was generated from the addition of such functional groups4.
As mentioned previously, spheroids form when the cell-to-cell interaction is higher than the interaction between the cell and matrix surface. A strong interaction between the cell and matrix surface can generally be observed on hydrophilic matrix surfaces with oxygen-bearing functional groups, such as hydroxyl, formyl, and carboxylic acid13,14. MG-63 cells were cultured on parylene N film, UV-treated parylene N film, conventional cell culture plates, and polystyrene surfaces. MG-63 cells on the parylene N film began to aggregate after incubation for 12 h, and small spheroids were observed after incubation for 24 h (Fig. 2a). Uniform-sized spheroids with diameters of ~ 100 µm were observed after incubation for 96 h. The shapes of the cells on parylene N, polystyrene, and conventional cell culture plates were compared after incubation for 24, 48, and 72 h. As shown in Fig. 2b, cells on the conventional cell culture plate, UV-treated parylene N film and polystyrene were well-attached to the surface, and spheroids did not form, even after incubation for 72 h. In the case of highly hydrophobic polystyrene, the formation of small spheroids was rare after incubation for 6 days (Sup Fig. S1).
A live video showed the formation of spheroids on the parylene N film (Supplementary Material 1) and sustained a monolayer on the conventional cell culture plate (Supplementary Material 2) and polystyrene (Supplementary Material 3) for 18 h incubation time after 24 h of seeding. In the case of parylene N film, the cell edges were round, and the cores of the spheroids were moving on the surface (Fig. 3a). This morphology, i.e., round cell edges, is considered to be due to a weaker interaction between the cell and surface compared with conventional cell culture plates. This relatively stronger cell-to-cell interaction results in cell aggregation and spheroid formation. These results show that the hydrophilic surfaces of UV-treated parylene N film and conventional culture plates yield better cell attachment compared with the less hydrophilic surfaces of polystyrene and parylene N film. These results also indicate that the formation of spheroid cells can be managed by increasing the cell-to-cell interaction by controlling the hydrophilicity of the culture matrix.
Based on these preliminary findings, we carried out the same experiment with GFP-labeled MG-63 cells (Fig. 3b and Sup Fig. S2). Figure 3b shows the start of spheroid formation in a single layer of GFP-labeled MG-63 cells. The GFP signal was maintained, indirectly showing that the cells inside the spheroids were alive. Thus, we confirmed the chemical environment due to the parylene N effect influences the self-assembly of 3D spheroids from osteoblast-like MG-63 cells.
Differentiation of MG-63 cells cultured on parylene N film
The formation of spheroids alters cell properties. According to previous studies, stem cells can be differentiated by the formation of 3D spheroids15. Because we studied spheroid formation on parylene N-coated film, we determined whether the cell properties changed due to the formation of spheroids. For this purpose, we performed a comparison of the global gene expression profiles of MG-63 spheroids formed on parylene N-coated plates and monolayers on the control (polystyrene) plate. The parylene N film was coated on a plate made of polystyrene. Thus, an uncoated polystyrene plate was used as a negative control for comparison of gene expression during spheroid formation. The 8,076 probe sets from spheroids cultured on parylene N that were compared with monolayers on the control plate were dysregulated (3,789 upregulated and 4,287 downregulated, Fig. S3a).
The significantly altered probes were used for hierarchical clustering and gene ontology (GO) analyses. Specifically, to gain insight into the biological significance of the 8,076 probe sets identified in our microarray analysis, we performed GO and Kyoto Encyclopedia of Genes and Genomes pathway enrichment analyses using the DAVID GO online analysis tool. The functional categories of significantly affected genes were determined. The DAVID analysis indicated that dysregulated genes primarily affected bone remodeling and regulators of bone mineralization (Sup Fig. S3b). Sonic hedgehog is also a major morphogen involved in osteoblast differentiation16,17. The GO analysis showed that spheroids on the parylene N-coated plate induced osteoblast differentiation.
Gene array analysis indicated that significantly dysregulated gene-encoding proteins were involved in osteoblast cell differentiation and bone morphogenesis of spheroids on parylene-N-coated film (Fig. S4). These included RUNX2 (log2 fold change = + 1.8) and ALPL (log2 fold change = + 9.8). Osteoblast differentiation is regulated by key transcription factors, such as RUNX2, accompanied by the upregulation of bone matrix proteins, including ALP (encoded by ALPL)18. Thus, to validate the gene expression profiles, qRT-PCR was performed for selected genes. Statistically significant upregulation of ALPL and RUNX2 mRNA in spheroids from the parylene N-coated plate compared with the control polystyrene plate were observed (Fig. 4a).
Osteocalcin (BGLAP), preferentially expressed by osteoblasts, is often used as a late marker of bone formation19. We observed the induction of BGLAP mRNA in cells cultured on parylene N, suggesting that spheroids from MG-63 cells include heterogeneous cells at various stages of osteogenic differentiation. Figure 4b,c show Alizarin Red S staining of MG-63 cells cultured on a parylene-N-coated plate and control plate. The spheroids’ staining intensity notably increased after incubation, revealing mineralization of the osteoblast-like cells20. Indeed, quantification of alizarin red S was significantly higher in MG-63 cells cultured on parylene N-coated plates compared to controls (Fig. 4c). Overall, these results indicate that the microenvironment of parylene N triggers the differentiation of osteoblast-like cells. Interestingly, in our study, the hydrophobicity of the surface may have been an important factor that enabled differentiation and spheroid formation of osteoblast-like cells and human mesenchymal stem cells. Analysis of the global gene expression profiles of MG-63 spheroids formed on parylene N-coated plates and monolayers cultured on a control plate indicated that dysregulated genes primarily affected calcium transport (Sup Fig. S5). For instance, TrpC4 (+ 69.94-fold change vs. control) is a transient receptor potential cation channel protein with distinct channel properties, such as altered calcium permeability (Sup Fig. S5)21. The CACNAIH gene encodes Cav3.2, a T-type member of the α1 subunit family, a protein in the voltage-dependent calcium channel complex and mediates influx of calcium ions into cells22,23. Overall, genes involved in calcium import, calcium channel activity, and calcium ion homeostasis were dysregulated in spheroids formed on parylene-N-coated plates as compared to controls. Thus, transport efficiency of cations, including calcium, might be increased in spheroid formed on parylene N-coated plates. However, further studies are required to investigate the mechanisms involved in dysregulation of cation channels.
Spheroid formation of MSCs on parylene N film
Increased osteoblast proliferation or induced osteoblast differentiation are important keys to the enhancement of bone formation during normal bone remodeling24. To investigate the growth of spheroids from MG-63 cells on parylene N-coated plates, the expression of Ki67, a marker of cell proliferation, was determined by immunofluorescence staining of MG-63 cells cultured on a parylene N-coated and control coverslip (Fig. 5a and Sup Fig. S6). Cells positive for Ki67 were maintained after the spheroid formation of MG-63 cells on parylene N (Fig. 5a). Consistent with the Ki67 staining, statistically significant upregulation of the proliferation-related genes, MKi67 and PCNA, mRNA was observed in the 3D spheroids compared with the control plate (Fig. 5b). Thus, in addition to differentiation, parylene-N induced the proliferation of MG-63 cells.
Enhanced proliferation during bone formation plays a vital role in evaluating parylene N-coated plates as future injectable scaffolds for clinical bone defect repair applications. Because the microenvironment formed by parylene N promotes osteogenic differentiation of MG-63 cells, we used human MSCs that can differentiate into mesenchymal tissues, such as bone and cartilage. Similar to MG-63 cells, there were interesting morphological changes of MSCs on parylene N-coated plates. MSC morphology on the control plate was characterized by a homogenous monolayer of adherent cells. In contrast, MSCs on parylene N formed incomplete spheroids but distinct MSC colonies after 18 days of culturing (Fig. 6a). In addition, MSCs on parylene N exhibited significantly increased expression of ALPL, RUNX2, BGLAP, and MKi67 mRNAs, while the control showed no change or even a reduction in the expression of these mRNAs after incubation (Fig. 6b). These results confirm that the parylene N film could effectively drive osteogenic differentiation and spheroid formation. Such results could be achieved by the control of chemical environments, through the controlled hydrophobic properties of the parylene N film. Usually, hydrophilic surfaces functionalized with oxygen species have been used for the osteogenic differentiation and spheroid formation; however, this work demonstrated that surfaces with controlled hydrophobicity could effectively drive osteogenic differentiation and spheroid formation in the absence of scaffolds or physical treatments such as hanging-drop methods.