**3. Discussion**

The present study demonstrates that there were high levels of SphK1 expression after wounding and that high endogenous production of S1P via SphK1 plasmid transfection accelerated wound closure and induced less scarring. It also suggested that the scarless healing induced by the SphK1 plasmid is due to the anti-fibrotic e ffect of S1P-S1PR signaling in the early phase of wound healing. This is supported by the fact that S1P treatment of dermal fibroblasts suppressed their production of collagens and that this e ffect was reversed by S1PR inhibitors. These observations together sugges<sup>t</sup> that since the SphK1 plasmid transfection increases S1P levels, it may prolong the anti-fibrotic e ffects of S1P-S1PR signaling during the proliferative phase of wound healing. This in turn limits the deposition of extracellular matrix, thus generating minimal granulation tissue and very thin scars when epithelialization is complete.

We found that SphK1 knockout significantly decreased the recruitment to the wound of CD3a+ T cells but not CD4+CD3a+ and CD8a+CD3a+ T cells. By contrast, topical overexpression of SphK1 significantly increased the CD4+CD3a+ and CD8a+CD3a+ T cell populations as well as the total CD3a+ T cell population in the wound. This apparent discrepancy may reflect the di fferent methods involved: SphK1 knockout has systemic e ffects whereas the SphK1 ointment has direct local e ffects. It will be of interest to further examine the profile of the T cells that are recruited to the wound (including NK cells and γδ T cells) when SphK1 is systemically knocked out: this may help elucidate the mechanisms by which the SphK1/S1P axis recruits T cells. Our finding that altering SphK1/S1P signaling shapes not just wound healing but also T cell recruitment suggests that T cells participate in wound healing. This notion is supported by multiple lines of evidence. For example, CD3 T cell numbers in the wound bed increase during the proliferation phase, particularly in the regenerating epidermis [35], and total T cell depletion impairs wound healing [36]. Moreover, burn injuries activate γδ T cells at the injury site. This initiates extensive infiltration by αβ T cells, which facilitate the transition from the inflammatory phase to the proliferative phase [37]. In addition, wound healing is associated with activation of epidermal T cells and their production of growth factors [38].

Overexpression of SphK1 increases recruitment of lymphocytes and macrophages while at the same time enhancing retention; this strengthens subsequent fibroblast activation. Therefore, granulation is promoted. However, fibroblasts activated by TGFβ-1 find it di fficult to respond to S1P stimulation as S1PRs are down-regulated (Figure 7D). This phenomenon should decrease as wound healing proceeds towards epithelialization. The anti-fibrotic e ffect mediated by S1PR signaling occurs with epithelialization. Treatment with the SphK1 plasmid strengthens the dynamic transition from the proliferative phase to the remodeling phase, resulting in inhibited scarring (Figure 7F).

TGF-β1 is likely to play a key role in S1P-S1PR1 signaling-induced less scarring. This cytokine is produced during the proliferative phase of wound healing and induces granulation [33,34]. Fibroblasts activated by TGFβ-1 find it di fficult to respond to S1P stimulation as S1PRs are down-regulated (Figure 7A,D). This phenomenon should decrease as wound healing proceeds towards epithelialization. The anti-fibrotic e ffect mediated by S1PR signaling occurs with epithelialization. Treatment with the SphK1 plasmid strengthens the dynamic transition from the proliferative phase to the remodeling phase, resulting in inhibited scarring (Figure 7F). The expression of TGF-β1 during the proliferative phase of wound healing postpones S1PR2 expression in the wound. Thus, when high S1P levels are generated by SphK1 plasmid transfection, the S1PR2 that is eventually produced is less able to suppress anti-fibrotic S1P-S1PR1 signaling. This limits the production of extracellular matrix and thereby inhibits excessive scar formation.

Immune responses play important roles in skin wound healing [39], and the present study suggests that increased expression of SphK1 also improves wound healing by strengthening these responses. These responses result in the production of multiple wound healing factors; they also have an anti-bacterial e ffect [40,41]. Since all currently available wound healing agents either target angiogenesis or fibroblast function [42–44], SphK1 plasmid treatment constitutes a novel approach to wound healing. Moreover, SphK1 plasmid treatment is simple. By contrast, other wound healing agents must be applied in a stepwise fashion: first, infections or necrotic tissues must be removed, followed by treatment with agents that support fibroblast function and/or angiogenesis. Notably, macrophage polarization was suggested recently to participate in wound healing [45,46]. Further studies exploring the effect of the systemic or local SphK1/S1P axis on macrophage polarization may expand the potential usefulness of these molecules in wound management.

In summary, we found that the SphK1/S1P axis accelerated wound healing by increasing angiogenesis and the recruitment of T cells and macrophages, which secreted various wound-related factors. It was also involved in inhibiting excessive scarring because it promoted the anti-fibrotic effect of S1P signaling. This finding suggests that the SphK1/S1P axis may be a novel therapeutic agen<sup>t</sup> that could help limit scarring after surgery and trauma and aid chronic wound management. However, studies that further elucidate the role of the SphK1/S1P axis in wound healing are needed to determine its full clinical potential.
