Our cytotoxicity findings are consistent with several other studies using GO layer cultured with mesenchymal stromal cells and GO/chitosan scaffolds seeded with human adipose-derived stem cells [53, 54]. In our study, LDH measurements showed significantly higher levels of DPSCs toxicity for uncoated SA scaffolds which can be attributed to poor ability and lack of efficient sites to support cell adhesion and proliferation. This result is supported by another study utilizing other materials mixed with alginate to create biocomposites [55]. However, the incorporation of GO into GOSA scaffolds did not elicit significant differences in cytotoxic effects after 2 days of DPSCs culture. Overall, the GO-enriched scaffolds exhibited cytotoxicity of 15–27% after 48h of culture, suggesting that these materials are biocompatible with DPSCs. Therefore, GO has no apparent cytotoxic effect but exhibits positive effects on cell function in long-term DPSCs culture.

The use of serum-free or serum-rich culture media for biological assays is a contentious issue [56]. The associated clinical uncertainties with the use of FBS include immune rejection, batch-to-batch variations, ethical concerns, interfering effects of unidentified growth factors and proteins, and viral contamination [57]. Regarding DPSCs stability, there is clear evidence about the positive effects of serum-free culture media on retaining stemness [58, 59]. Furthermore, serum-free cultured DPSCs within chitosan scaffolds expressed stem cell markers (Nestin and Sox2) and survived successfully after transplantation into the rodent spinal cord [60]. The present study investigated whether serum depletion facilitates the attachment of DPSCs onto hybrid scaffolds. Thus, the effects of serum on the cytotoxicity of DPSCs were tested using 2D and 3D culture systems across three different coating conditions at a single cell density of 4 × 10⁴cells/sample. We have indicated that the presence of serum caused significantly higher cytotoxic behaviour of coated 3D scaffolds which could be due to suboptimal surface-cell attachment [20]. This was also shown for the combination of PLL and LAM coatings in the presence of serum on 3D structures. It is surmised that coatings lower the available surface area of 3D scaffold biomaterials for the attachment of cell receptors, which subsequently causes poor adhesion. In contrast, coated 2D surface was found to result in better attachment of DPSCs which is due to the unfavourable bare tissue culture surface. Moreover, our study found that the cytotoxicity of DPSCs cultured in serum-rich media on 3D scaffolds increased significantly in comparison with the 2D control. The increased level of cytotoxicity in serum-containing medium could be attributed to the surface oxygen content of GO and RGO, which favours FBS adsorption. This could be due to the formation of protein corona on the GO surface which in turn influence the toxicity of GO-based materials [61, 62]. Furthermore, studies have indicated that serum proteins can have an interfering effect on interactions between nanoparticles, cells and biological molecules [63-65]. Accordingly, Lesniak et al. [64] demonstrated that various protein corona formed on silica nanoparticles modified cell adhesion, cellular uptake, and toxicity, which is determined by the serum protein concentration. It was also shown that serum-containing media resulted in lower cell adhesion and internalization efficiency of silica nanoparticles. The present study found DPSCs can be seeded in serum-free media onto GOSA and RGOSA scaffolds with no cytotoxic effects, showing promising potential for clinical translation as cellular transplants are typically serum-free.