A Reinterpretation of Cell Transplantation: GFP Transfer From Donor to Host Photoreceptors

Abstract

The utilization of fluorescent reporter transgenes to discriminate donor versus host cells has been a mainstay of photoreceptor transplantation research, the assumption being that the presence of reporter1 cells in outer nuclear layer (ONL) of transplant recipients represents the integration of donor photoreceptors. We previously reported that GFP 1 cells in the ONL of cone-GFP transplanted retinas exhibited rod-like characteristics, raising the possibility that GFP signal in recipient tissue may not be a consequence of donor cell integration. To investigate the basis for this mismatch, we performed a series of transplantations using multiple transgenic donor and recipient models, and assessed cell identity using nuclear architecture, immunocytochemistry, and DNA prelabeling. Our results indicate that GFP 1 cells in the ONL fail to exhibit hallmark elements of donor cells, including nuclear hetero/euchromatin architecture. Furthermore, GFP signal does not appear to be a consequence of classic donor/host cell fusion or transfating post-transplant, but is most likely due to material exchange between donor and host photoreceptors. This transfer can be mediated by rods and cones, is bidirectional between donor and host cells, requires viable photoreceptors, occurs preferentially at sites of outer limiting membrane disruption and can be detected in second-order retinal neurons and M€ uller glia. Collectively, these data warrant re-evaluation of the use of lineage tracing fluorescent reporters in transplantation studies involving the retina and other CNS tissues. Furthermore, the reinterpretation of previous functional rescue data, based on material exchange, rather than cell integration, may offer a novel approach to vision rescue. STEM CELLS 2017;35:932-939 SIGNIFICANCE STATEMENTGFP labeled photoreceptors are observed in the outer nuclear layer of recipient retinas following transplantation of GFP-tagged photoreceptors. The historical interpretation of this observation has centered around the migration of donor cells into recipient retina, and maturation of these cells into functional photoreceptors. This study challenges this interpretation by showing that there is almost no donor cell integration into intact retinal tissue, and provides evidence that the origin of the GFP signal in the recipient retina is due to exchange of GFP signal between donor and host retinal cells. This work reveals that the adult retina is not as receptive to donor cell integration as was previously thought and deepens our understanding of how photoreceptor therapy, via material exchange, could work therapeutically.

Differentiation and Transplantation of Embryonic Stem Cell-Derived Cone Photoreceptors into a Mouse Model of End-Stage Retinal Degeneration

“…Our optimized methodology resulted in a high success rate, with 26 out of 29 eyes (89%) supporting the survival of the transplanted cell mass (Figures S6E–S6H), without immune suppression. Reporter-positive photoreceptors were present in the host outer nuclear layer (Figure S6H), but similar to previous studies demonstrated rod-like morphology and were likely the result of uptake of cellular components by host photoreceptors from the subretinal graft (Decembrini et al., 2017, Ortin-Martinez et al., 2016). In the subretinal space, the cells were negative for rod markers RHODOPSIN and GNAT1 (Figures S6E and S6F) and continued to develop in vivo, forming peanut agglutinin (PNA)-labeled extracellular matrix and expressing ARRESTIN3 (Figures S6G and S6H).

Figure 6Isolation and Transplantation of Purified mESC-Derived Cone Precursors(A) Reporter fluorescence driven by AAV2/9-2.1.GFP vector following subretinal injection into adult mouse retina.

…”

“…Unsurprisingly, given the advanced degenerative state of the recipient retina, the morphology of transplanted cones was compromised. This contrasts with apparently mature features of GFP-labeled cells observed in cone transplants into non-degenerative retina, which are now understood to arise from cytoplasmic material transfer from the subretinal graft to host rods (Decembrini et al., 2017, Ortin-Martinez et al., 2016). Nonetheless, cones in the Aipl1 − / − recipients appeared to make physical contact with inner retinal neurons and expressed components associated with synaptic transmission, alongside phototransduction-related proteins, suggesting advanced differentiation and maturation.…”

Cell lineage tracing in the retina: Could material transfer distort conclusions?

“…Similar mismatch between nuclear morphologies of the grafted cells versus host cells was observed after GFP‐labeled rod precursors were grafted in an S‐cone‐like cell dominated retina ( Nrl ‐/‐). In complementary experiments, they found that transplanting P3–P5 retinal cells isolated from a membrane‐tethered tdTomato mouse line into adult Nrl ‐GFP retinas led to tdTomato+/GFP+ cells in the ONL (Ortin‐Martinez et al, ), similar to what was observed when transplanting GFP+ cells in DsRed retinas. Altogether, these results suggested that some fusion event occurred between the grafted cells and the host photoreceptors.…”

“…It is still unclear whether only photoreceptors in the host retina engage in material transfer with the grafted cells. While three groups did not report GFP‐labeled cells outside of the ONL after transplantation of rod precursors (Pearson et al, ; Santos‐Ferreira et al, ; Singh et al, ), another observed low levels of GFP in both bipolar and Müller cells located in the INL (Ortin‐Martinez et al, ). These cells were localized next to areas with high density of material transfer‐labeled host photoreceptors in Nrl ‐/‐ retinas.…”

hESC-derived photoreceptors survive and integrate better in immunodeficient retina

“…E.g., Barnard et al (27) Recently, several laboratories have demonstrated that recipient photoreceptors incorporate GFP label from transplanted photoreceptor precursor cells that were injected into the subretinal space. This means that GFP label alone is insufficient to tell whether donor cells really integrated into the host retina (28)(29)(30). However, Zhu et al showed that GFP-labeled cells stain for human specific markers and do not co-express a mouse-specific MHC class I marker which clearly determined that they were all of human origin.…”