This study demonstrates for the first time that hPDLSCs provide a significant therapeutic benefit in sustaining retinal function for up to 8 weeks and delay photoreceptor degeneration along most of the retina after implantation into the subretinal space of RCS rats. Our data suggest that hPDLSCs may represent a promising therapeutic alternative for RD.

hPDLSCs are thought to originate from the cranial neural crest. They abundantly express MSC markers and exhibit the unique features of neuroectodermal stem cells [25, 26]. Consistently, our immunocytochemical staining and flow cytometry analysis revealed that hPDLSCs presented a high rate of positivity (> 90%) for conventional cell surface markers of MSCs, including CD44 and CD90, indicating their mesenchymal characteristics. A considerable amount of hPDLSCs express neural cell markers, including PKCα (99.7%) and Nestin (98.04%), which may be related to the neural crest origin of PDL. Our study also provides evidence that hPDLSCs represent a population of postnatal stem cells capable of multilineage differentiation into adipogenic, osteogenic, and chondrogenic lineages when grown in their respective induction media.

ERG recordings at 2, 4, and 8 weeks post transplantation showed substantial preservation with hPDLSC treatment compared with that of controls, indicating the therapeutic efficacy of hPDLSC transplantation. However, the b-wave amplitudes decreased between 4 and 8 weeks post treatment, suggesting that the degeneration process was delayed rather than halted by the cell treatment. We proposed that poor cell survival may be an important factor influencing the therapeutic efficacy of hPDLSC treatment following transplantation. Immunohistochemical staining with human specific marker TRA-1-85 revealed a substantial number of transplanted hPDLSCs in the subretinal space at 2 weeks post transplantation. However, the number of transplanted cells dramatically declined with time, with a negligible amount of cells remaining in the host retina at 4 weeks post transplantation. Previous studies have reported a similar fallout of donor cells with time and limitations of the therapeutic effects using subretinal transplants of human MSCs [4, 27], NSCs [4], ADSCs [23], and human umbilical cord tissue-derived stem cells (UTCs) [28] in retinal degeneration models. One likely explanation is that transplanted cell survival may have been compromised because of the hostile environment of the diseased retina. Immune-mediated xenograft rejection may also pose a major obstacle to the survival duration of graft cells. Whether graft cell survival and therapeutic effects could be improved by utilizing a bioscaffold or a second transplantation warrants further study. It is also conceivable that autologous hPDLSC transplantation may have the advantage of reducing the risk of rejection and present prolonged therapeutic benefits for the treatment of retinal degeneration.

It is now widely recognized that adult stem cells exert a beneficial influence on retinal repair predominantly by paracrine action through the secretion of trophic and anti-inflammatory factors [27, 29, 30]. Like other MSCs, the extensive neurotrophic secretome of DPSCs has been documented widely [11, 31, 32]. Furthermore, DPSCs were found to secrete VEGF, NGF, BDNF, and NT-3 at higher concentrations than BMSCs [11, 33]. Mead et al. [11, 12] reported a more pronounced paracrine-mediated RGC survival and neurite outgrowth by transplanted human DPSCs compared with these variables after BMSC transplantation in animal models of optic nerve injury and glaucoma. In contrast to DPSCs, few data exist to date describing the neurotrophic factor secretion and neuroregenerative/neuroprotective activities of hPDLSCs. Here, we report for the first time the expression by hPDLSCs of a variety of neurotrophic factors, including VEGFA, bFGF, BDNF, NT-3, IGF-1, NGF, and GDNF, as determined by PCR and ELISA measurements. In the eye, VEGF, NGF, bFGF, BDNF, NT-3, IGF-1, and GDNF are widely considered neuroprotective, possibly by an anti-apoptotic mechanism, for photoreceptor survival [23, 3439]. We reported recently that hADSCs preserved retinal structure and visual function in RCS rats, at least in part by secreting VEGF [23]. In this study, similar quantities of VEGF expression were detected in hPDLSCs (1255.14 ± 35.70 pg/ml) and hADSCs (1268.72 ± 334.67 pg/ml). In this scenario, hPDLSCs may also serve as a source for supply of these neurotrophic and neuroprotective factors to the diseased retina for photoreceptor survival and therefore delay retinal degeneration in RCS rats.

Apoptotic cell death has been shown to be the common endpoint of photoreceptors in most RD [40]. Undigested POS accumulates as toxic debris in the subretinal space of RCS rats, which instigates photoreceptor apoptosis [21]. Our data showed slowed retinal pathogenesis in RCS rats with an observed reduction in the photoreceptor debris zone and the TUNEL-positive ONL cells in the hPDLSC transplantation group compared with those in the PBS group and the untreated group. This result was consistent with the findings of Tzameret et al. [41], which indicated reduced DZ and retinal structure improvement by human BMSC transplantation. These findings further corroborate the hypothesis that paracrine-mediated anti-apoptotic effects could be partially responsible for the beneficial effects of hPDLSCs on the degenerating retina.

Histological analysis and ERG recordings demonstrated the obvious delay of photoreceptor degeneration and maintenance of retinal function that persisted up to 8 weeks post transplantation. In contrast, there were few hPDLSCs remaining in the retina. Our findings support the notion that achieving a therapeutic benefit does not require the presence or long-term survival of engraftment, leading to the hypothesis that trophic factors secreted by transplanted hPDLSCs during the first 2 weeks may be sufficient to delay retinal degeneration for a longer duration.

Because hPDLSCs constitutively express PKCα and Tuj1, even in an undifferentiated state, the differentiation of these cells into bipolar cells or RGCs after engraftment is contentious. To determine whether the grafted hPDLSCs have the potential to differentiate into photoreceptor cells in the retina, double labeling of the grafted cells, as indicated by TRA-1-85-positive staining with recoverin, was performed. Our data showed evidence of a few grafted hPDLSCs migrating into the host retina but failed to express recoverin at 8 weeks post transplantation (Fig. 

8

). The disparity between the reported successful photoreceptor differentiation of hPDLSCs in vitro in our previous study [

18

] and the lack of differentiation in vivo could be due in part to the pathological microenvironment of the diseased retina, which presents a vastly different environment from the carefully controlled in-vitro setting. Another possible explanation is that the retinal neuronal differentiation of hPDLSCs may take a much longer time to develop. The poor survival of hPDLSCs after engraftment may diminish their chance for subsequent differentiation. Further studies confirming the neuronal differentiation potential of hPDLSCs by developing methods to improve graft cell survival have not yet been conducted.

Fig. 8

Fluorescence microscope images of retinal sections injected with hPDLSCs at 8 weeks post transplantation. Retinal sections were double-stained with human antigen TRA-1-85 (b) and recoverin (c), with DAPI counterstaining (a) which showed no colocalization (d). Arrowheads indicate migrating hPDLSCs. ONL outer nuclear layer, INL inner nuclear layer, GCL ganglion cell layer

To the best of our knowledge, this is the first demonstration of the preservation of photoreceptors and the maintenance of retinal function by the subretinal transplantation of hPDLSCs. Here, we found no evidence of untoward pathological manifestations or tumor-like transformations up to at least 8 weeks after transplantation during the experimental period. The advantages of being an ethical cell source, being relatively easy to access from teeth, having low immunogenicity, and having the potential for an off-the-shelf product without a predifferentiation process before transplantation, along with the nontumorigenic and distinctive trophic characteristics, make autologous and allogeneic hPDLSC transplantation a promising paracrine-mediated therapy for retinal degenerative diseases.

In summary, this study provides a proof of principle for the use of hPDLSCs to treat retinal degeneration. For translational implications, it remains critical to investigate the cells’ long-term safety and optimize strategies to enhance the therapeutic benefits of hPDLSC transplantation in future work.