Discussion

Significant complexity has been uncovered in the interaction between stem cells and engineered scaffolds [38]. In this study, we report the profiles of DPSCs cultured on fabricated 3D GOSA and RGOSA scaffolds with 2D, coating and media controls.
Our in vitro biodegradation data revealed an inverse relationship between GO concentrations and weight loss. This relationship is explained by accessibility of water molecules to GO composites as a function of GO concentration. As discussed in the water contact angle measurements in our previous paper [24], the incorporation of GO increases the interaction of composite scaffolds with water and this effect is GO concentration-dependent. Therefore, as hydrophilicity accelerates with higher GO concentration, water-mediated scaffold degradation increases [39]. Notably, a reduced degradation rate is assumed to be beneficial for tissue regeneration [40]. Collectively, our results confirm that the degradability of composite 3D R/GOSA scaffolds is adjustable and controlled by graphene content.
It is noteworthy that when cultured onto 3D SA and GOSA scaffolds, DPSCs viability was enhanced in relation to 2D culture plates. This relationship clearly signifies the optimal condition of initial cell adhesion to the scaffold surface to promote subsequent cell proliferation and infiltration. Our observation of an increase in the total metabolic activity of cell-seeded scaffolds provides proof-of-principle support for cell growth and proliferation in a 3D matrix which can act as a delivery system for seeded cells. Thus, it is reasonable to conclude that an artificial 3D scaffold is an acceptable approach to mimic the natural architecture of the native tissue and crease a microenvironment conducive to DPSCs engraftment.
The cell-cell interaction within the matrix of 3D cell culture systems has a profound influence on cellular functions including viability, migration and proliferation in contrast to 2D culture [41]. For example, one report showed that 3D polymer-based scaffolds seeded with hepatic cells had less cytotoxic effects than those cultured in 2D [42]. In another study, the 3D culture of dental stem cells was found to support their neuronal characteristics and maintain cell phenotypes [43]. Extending this, we have confirmed that a superior proliferative ability of DPSCs, as measured by metabolic activity, can be obtained when cells are cultured on 3D porous scaffolds.
Regarding cell seeding density, we have shown that when cells are seeded on 3D scaffolds at all densities, the cell proliferation rate is significantly increased in comparison to 2D. Moreover, we found that increased seeding density in 3D scaffolds could be achieved without inducing cytotoxic effects, as determined by LDH assay. In this study, we also showed that the addition of graphene to 3D composite scaffolds improved cellular behaviours and this was seen across all DPSCs seeding densities examined. Notably, the degree of cellular metabolic activity did not differ significantly between the different cell densities tested. This result is important because the implication is that seeding efficiency can be achieved even at high cell densities, at least within a 48-hour period. More work would be required to determine if high seeding densities might have different effects on longer-term culture.
Assessment of coating reagents revealed that PLL+LAM coating did not affect cell viability as indicated by AB reduction percentage. After 48 hours of DPSCs culture, both laminin coating and no coating conditions decreased cell viability on both SA and GOSA scaffolds. Possible explanations involve interactions between coating properties and DPSCs adherence or cell aggregation which can decrease proliferation [44]. Interestingly, PLL was identified as the coating reagent that enhanced cell-matrix adherence. This enhancement might be due to the larger number of cationic sites offered by PLL coating on the 3D surface. In agreement with another study [44], our results show that PLL is superior to laminin coating. In addition, the effect of all three coating conditions on DPSCs was irrespective of cell seeding density.
The AB assessment of metabolic activity shows that biomaterial composition can modulate DPSCs responses to fabricated scaffolds. Our no scaffold controlled LDH results also indicate that SA, GOSA and RGOSA scaffolds materials are nontoxic to DPSCs in short-term culture. Based on previous work, it might be inferred that the composition of a scaffold material can have a direct effect on the biodistribution of secreted factors that in turn influence the stem cell fate. The results appear to suggest that different scaffolds with varying material properties (such as blend ratio, swelling index, or microstructure) elicit diverse DPSCs behaviours. The increase in cell viability observedin vitro upon the incorporation of graphene in composite scaffolds is consistent with other studies [36, 45]. Published data suggest that the outstanding surface properties and adsorption capacity of graphene-based nanomaterials are the main contributors to the observed DPSCs responses.
Our results showed a strong influence of pore size, material composition and substrate dimensionality on cell viability, in accordance with the findings of Domingos et al. [46]. Comparisons of AB reduction in SA (97.2%) and graphene-based scaffolds including GOSA0.5 (97.5%), GOSA1 (98.0%), RGOSA0.5 (99.05%), and RGOSA1 (99.18%) revealed the differences in DPSCs proliferation markers across various graphene-based scaffolds. These differences appear to be explained by variations in scaffold porosity (%) such that scaffolds with higher porosity (RGOSA ≈ 99%) are able to accommodate higher numbers of viable cells. Furthermore, it was shown that scaffolds with smaller mean pore sizes induce relatively less toxicity. This can be explained by the available surface area of scaffolds for cultured cells or applying the principle that the mean pore size and specific surface area are inversely proportional. In consideration of the specific surface area and mean pore size of a scaffold, biophysical properties can affect cell adhesion [47]. It follows that the low levels of cell adhesion are observed on scaffolds with larger pore size and less specific surface area [48, 49]. As a result, the available specific surface area per unit volume for cell adhesion of each fabricated scaffolds can be calculated using mean pore sizes [50]. Accordingly, the normalized specific surface area of GOSA and RGOSA scaffolds, as shown in Table 1, can be obtained by dividing the mean pore size of each scaffold by mean pore size of SA scaffold (3D control). Thus, the higher AB reduction observed in our RGOSA1 scaffold can be explained by the higher specific surface area in comparison with GOSA1 and SA scaffolds. These data indicate that higher pore size facilitates increased DPSCs migration and proliferation, in agreement with a previous report on culturing DPSCs into 3D poly-L-lactic acid-based scaffolds [51]. In addition, the mechanical properties of scaffolds with overly large pores are compromised, whereas higher cellular proliferation within large pore sizes can have implications for differentiation [52].
Table 1. Estimates of the specific surface area of graphene-based scaffolds relative to SA scaffold