Optimizing Donor Cellular Dissociation and Subretinal Injection Parameters for Stem Cell-Based Treatments

Abstract

Subretinal delivery of stem cell-derived retinal cells as a strategy to treat retinal degenerative blindness holds great promise. Currently, two clinical trials are underway in which human fetal retinal progenitor cells (RPCs) are being delivered to patients by intravitreal or subretinal injection to preserve or restore vision, respectively. With the advent of the induced pluripotent stem cell (iPSC), and in turn three-dimensional derivation of retinal tissue, it is now possible to generate autologous RPCs for cell replacement. The purpose of this study was to evaluate the effect of commonly used cell isolation and surgical manipulation strategies on donor cell viability. iPSC-RPCs were subjected to various conditions, including different dissociation and isolation methods, injection cannula sizes, and preinjection storage temperatures and times. The effects of commonly used surgical techniques on both host and donor cell viability were evaluated in Yucatan mini-pigs (n = 61 eyes). We found a significant increase in cell viability when papain was used for RPC isolation. In addition, a significant decrease in cell viability was detected when using the 41G cannula compared with 31G and at storage times of 4 hours compared with 30 minutes. Although 96.4% of all eyes demonstrated spontaneous retinal reattachment following injection, retinal pigment epithelium (RPE) abnormalities were seen more frequently in eyes receiving injections via a 31G cannula; interestingly, eyes that received cell suspensions were relatively protected against such RPE changes. These findings indicate that optimization of donor cell isolation and delivery parameters should be considered when developing a subretinal cell replacement strategy. Stem Cells Translational Medicine 2019;8:797&809.

Keywords: Induced pluripotent stem cells; Inherited retinal dystrophy; Pig model; Retinal progenitor cells; Stem cell transplantation; Subretinal transplantation.

Figure 1

Isolation of retinal progenitor cells from three‐dimensional (3D) retinal organoids. (A): Retinal progenitor cells were isolated from 3D retinal organoids (via the protocol outlined in the schematic) and plated on laminin‐coated culture surfaces at 60 days postdifferentiation via papain dissociation. At 1 day postplating, a mix of pigmented RPE and nonpigmented neural retinal progenitor cells were present. By 4 days postplating, cells began to spread over the culture surface (a loss of pigment was noted at this time point). Scale bar: 400 μm. (B–F): Immunocytochemical analysis targeted against Pax6 (B, D), Sox2 (C, D), MITF (E), and OTX2 (F) revealed the presence of retinal progenitor and RPE cell precursors, respectively. Scale bar: 50 μm. Abbreviations: DAPI, 4′,6‐diamidino‐2‐phenylindole; MITF, melanogenesis‐associated transcription factor; OTX2, orthodenticle homeobox 2; Pax6, paired box 6; Sox2, SRY (sex determining region Y)‐box 2.

Figure 2

Temperature, storage time, and cannula gauge effect retinal progenitor cell viability. (A): Induced pluripotent stem cell‐derived retinal progenitor cells (RPCs) were incubated at various temperatures (0°C, 21°C, 37°C, and 50°C). Cell viabilities (mean ± SEM) were measured using trypan blue staining and an automatic cell counter after 1 hour (black) and 4 hours (checkered). Two‐way ANOVA was performed (factors: temp and time) with Sidak's comparisons testing. **, p < .01; ***, p < .001; ****, p < .0001 comparing temperature‐matched samples. ⁺⁺, p < .01 comparing 37° to 0° at 4 hours. ^†, p < .05; ^††, p < .01; ^†††, p < .001 comparing samples to 50° at 4 hours. (B): RPCs were injected through either a larger cannula (31G) or a smaller cannula (41G) and collected for in vitro analysis of the percentage of live cells compared with preinjection cell counts. Cell viabilities (mean ± SEM) were measured immediately using an automatic cell counter using trypan blue staining (black) or an MTS assay (blue). *, p < .05; **, p < .01; ****, p < .0001 compared with no cannula. ⁺⁺, p < .01 comparing 31G to 41G. Abbreviations: G, gauge of needle‐cannula; MTS, tetrazolium assay.

Figure 3

Statistical analysis of injection conditions in association with pseudo‐geographic atrophy or RPE changes based on ophthalmoscopic assessment. Indirect ophthalmoscopy was performed on all eyes (n = 50) that received subretinal injections of either iPSC‐derived retinal progenitor cells (n = 16) or vehicle (n = 34) via a 31G (n = 32) or 41G cannula (n = 18), and the number of eyes with pseudo‐GA or RPE changes without GA were compared between the two cannula groups (A–C) or the two injection groups (D–F). For (A)–(C), comparisons between the cannula groups were performed for all eyes that received subretinal injections (A), eyes that only received iPSC‐derived RPC injections (B), and eyes that only received subretinal vehicle injections (C). For (D)–(F), the number of eyes with pseudo‐GA or RPE changes were compared between the two treatment groups for all eyes that received subretinal injections (E), eyes that only received 41G cannula injections (F), and eyes that only received 31G cannula injections (G). For each graph, bars represent the percentage of affected eyes for a specific finding. Data were analyzed by chi‐square test and then Fisher's exact test using GraphPad prism software. p < .05 was considered statistically significant. All eyes that were examined and found to have a dense cataract, total retinal detachment, and/or dense vitreous opacities were excluded as the presence or absence of pseudo‐GA and RPE could not be assessed. Abbreviations: G, gauge of needle‐cannula; GA, geographic atrophy; RPE, retinal pigment epithelium. *, p < .05; ***, p < .001.

Figure 4

Histological and clinical evidence of pseudo‐GA and retinal folds after subretinal vehicle injections in a pig model. (A–C): Representative images taken outside the bleb region (i.e., noninjected retina). (D–I): Representative images taken of the injected, bleb region of these eyes show evidence of either pseudo‐GA (D–F) or retinal folds (G–I). The images in (A), (B), (D), and (E) represent the same eye. H&E staining (A and D) shows depigmentation of the retinal pigment epithelium within the bleb region only (D), and immunohistochemistry (B and E) demonstrates that this same bleb region has a significant loss (E, white arrows) of RPE65 signal (green). For (B) and (E), DAPI was used to counterstain nuclei. Indirect ophthalmoscopy demonstrates a normal appearance of the retina outside the bleb region (C) and RPE changes with pseudo‐GA within the bleb region (F, white arrows). (G–I): A prominent retinal fold within the bleb region is seen using H&E staining (G), ocular coherence tomography imaging (H, white arrows), and indirect ophthalmoscopy (I, white arrows). Scale bar = 100 μm. All panels are from postoperative week 2 except (C) (postoperative week 1), (F) (week 5) and (I) (week 1). Abbreviations: DAPI, 4′,6‐diamidino‐2‐phenylindole; GA, geographic atrophy; INL, inner nuclear layer; ONL, outer nuclear layer; RPE, retinal pigment epithelium; RPE65, retinal pigment epithelium marker.

Figure 5

Differential effects of injection rate on loss of RPE65 expression in cadaver pig eyes. (A–C): RPE65 retinal pigment epithelial cell marker‐immunolabeled (red) histological sections show higher expression of RPE65 in noninjected vitrectomized pig cadaver eyes (A; n = 4) compared with 1.8 ml/minute injected eyes (C; n = 8) but not to 0.18 ml/minute injected eyes (B; n = 6). Nuclei were counterstained with DAPI (blue). (D–F): Corresponding H&E stains on adjacent sections (bottom panels) show no substantial change in pigmentation of RPE in the 1.8 ml/minute group (F) compared with noninjected vitrectomized control sections (D) and the lower 0.18 ml/minute injection group (E). Scale bar: 100 μm. (G): There was a significant loss of RPE65 signal in the pig retinas injected with at a high speed (1.8 ml/minute) compared with the retinas injected at a lower speed (0.18 ml/minute) and the retinas that were not injected. Each point represents the percentage of RPE65 signal loss for the bleb region of an individual eye, and inset horizontal lines show results as mean ± SEM. p < .05 was considered statistically significant. For all histological images, representative images are shown. Abbreviations: DAPI, 4′,6‐diamidino‐2‐phenylindole; RPE65, retinal pigment epithelium marker.

Figure 6

Upregulated inflammatory cytokine levels in pig eyes with vitreous floaters/opacities after subretinal vehicle injections. (A–E): Analysis of inflammatory cytokines by multiplex analysis. Supernatant was obtained from pig vitreous samples at euthanization 2–5 weeks after subretinal injections. These samples were analyzed on a 13‐plex panel multiplex plate, which confirmed that at least five inflammatory mediators (IL‐1RA, IFN‐γ, IL‐12, IL‐8, and IL‐6) are upregulated in the vitreous at the protein level for eyes with ophthalmoscopic evidence of vitreous floaters. Each point represents the average of duplicate readings of the cytokine concentration in an individual eye, and inset horizontal lines show results as median with range. Concentrations lower than the low limit of detection were defined as nonmeasurable, and data were analyzed by chi‐square test and then Fisher's exact test. Logistic regression analyses were also performed and there were no significant covariates. p < .05 was considered statistically significant. Abbreviations: IL, interleukin; IFN, interferon.