It’s About Time: Time-Dependent Tissue Damage in the Adult Porcine Retina After Enucleation

The isolated adult retina is extensively used in contemporary research for explorations of pathological disease mechanisms and tissue engineering, as well as in treatment paradigms ranging from pharmacological intervention to experimental retinal transplantation [for a review, see Rettinger and Wang, 2018]. Most experiments have explored retinal tissue from small animals, but research involving large animal models, including porcine and bovine eyes, has gained interest in recent years due to an enhanced similarity to the human retina.

When working with large animal models, the tissue used for retinal experiments is frequently harvested from local abattoirs, which limits the possibility of a standardized handling process, and the time after enucleation is often extended for several hours [Schnichels et al., 2019]. The retina degenerates postmortem after enucleation due to global ischemia as well as axotomy of ganglion cell axons. To limit tissue damage, enucleated eyes can be kept under hypothermic conditions, which in a few studies has been shown to, at least partly, attenuate ganglion cell death [Reinhard et al., 2016; Schultheiss et al., 2016]. However, the influence of time from enucleation on overall retinal tissue damage, as well as any impact on subsequent investigations in the large animal model eye, has not yet been thoroughly explored.

For the present study, we wanted to expand the knowledge on time after enucleation as a relevant factor for experimental research involving the adult retina in a large animal model, with the aim of systematic characterization of time-dependent tissue damage per se and exploration of any extended effects in vitro. Previous in vivo and clinical investigations have indicated that the retinal tolerance time to ischemia is approximately 90 min and that extensive tissue destruction is evident after 240 min [Hayreh et al., 2004]. These time points were therefore chosen for the present experiment in which we investigated retinal tissue damage in porcine eyes 90 or 240 min after enucleation, with the addition of experiments in which we cultured retinal explants after these time points to explore whether an increased time after enucleation has an extended impact in vitro.

Adult pigs aged between 4 and 6 months were transported from a local breeder to the Biomedical Center at Lund University where all experiments were performed. Animals were sacrificed after sedation by an intravenous overdose of sodium pentobarbital (Apoteket, Umeå, Sweden), after which both eyes were enucleated and placed in a vial containing CO₂-independent culture medium (Invitrogen, Paisley, UK) at 4°C on ice. Eyes were immediately transported to the laboratory and dissected at room temperature 90 or 240 min after enucleation as follows. The anterior segment was removed by sharp incision in the pars plana 360°, and the vitreous carefully removed using sterilized tissue paper. The neuroretinas were gently dissected free from the pigment epithelium with micro-forceps, and the optic nerve was cut using micro-scissors. Six full-thickness retinal pieces measuring approximately 5 × 5 mm were dissected from the centrally located area on the temporal and nasal side of the optic nerve head between the 3 major vascular arcades as previously described [Taylor et al., 2013]. After dissection, retinal pieces were either immediately fixed in 4% paraformaldehyde in 0.1 M phosphate buffer, pH 7.2 for 2 h at room temperature, or cultured as described below.

For exploration of time-dependent tissue damage after enucleation, retinal pieces were immediately fixed in 4% paraformaldehyde in 0.1 M phosphate buffer, 90 or 240 min (±10 min) after enucleation (n = 12 in each group).

For the culture experiment, additional pieces from each time point (n = 12 in each group) were explanted onto Millicell–PCF 0.4 mm culture plate inserts (Millipore, Billerica, MA, USA), with the inner retina facing the membrane. This orientation of explants was chosen to optimize tissue survival in vitro [Taylor et al., 2014]. The 24 retinal explants were cultured in Dulbecco’s modified Eagle’s medium F12 (DMEM/F12; Invitrogen). The medium was supplemented with 10% fetal calf serum (Sigma-Aldrich, St Louis, MA, USA) and 1% antibiotics (2 mML–glutamine, 100 U/mL penicillin, and 100 ng/mL streptomycin; Sigma Aldrich). Explants were maintained at 37°C (95% humidity and 5% CO₂ + 95% air) for 48 h after which they were fixed as described above. The culture time points were chosen based on previous experiments in which the tissue was relatively well preserved [Åkerström et al., 2017].

After several rinses in phosphate buffered saline (PBS), the fixed tissue explants were infiltrated with 0.1 M Sörensens solution with increasing concentrations of sucrose up to 25%. They were then embedded in egg albumin/gelatine medium, cryosectioned at 12 µm, and transferred to glass slides. For light microscopy, slide 1, 10, 20, and 25 were stained with hematoxylin and eosin (H&E). For immunohistochemical labeling of ganglion cells, slides were rinsed 3 times with PBS and then incubated with PBS buffer containing 0.25% Triton-X and 1% bovine serum albumin (BSA), for 20 min at room temperature. The specimens were then incubated overnight at 4°C with an antibody raised against RNA-binding protein with multiple splicing (RBPMS; PhosphoSolutions, Aurora, CO, USA), a specific ganglion cell marker [Rodriguez et al., 2014]. The specimens were then rinsed in PBS and incubated for 45 min with a rhodamine red-conjugated secondary antibody (Jackson ImmunoResearch, West Grove, PA, USA), and finally mounted in Vectashield mounting medium with DAPI (Vector Laboratories Inc., Burlingame, CA, USA). Negative control experiments were performed as above, replacing the primary antibody with PBS containing 0.25% Triton-X and 1% BSA.

Terminal deoxynucleotidyl transferase dUTP nick end labeling (TUNEL) staining was performed using the TMR red In Situ Cell Death Detection kit supplied by Roche Diagnostics GmbH, Mannheim, Germany.

For assessment of overall cell viability, 240 µL of medium derived from retinal explants cultured 90 and 240 min after enucleation, was sampled after 2, 24, and 48 h and stored at −80°C. After thawing, the samples were diluted (1:3) with PBS and analyzed for lactate dehydrogenase (LDH) activity using the CytoTox 96® Non-Radioactive Cytotoxicity Assay (Promega, Madison, WI, USA), according to the instructions from the manufacturers.

All morphological data from retinas examined 90 and 240 min after enucleation were compared with in vivo controls and analyzed using one-way ANOVA. In the culture experiment, morphological and LDH data from the 2 groups (cultured 48 h, 90 or 240 min after enucleation) was analyzed using Student’s t-test. GraphPad Prism version 8.4.2 for Mac (GraphPad Software, La Jolla, CA, USA) was used for all statistical analyses, and p < 0.05 was considered significant.

For morphological analysis (H&E, TUNEL, and RBPMS), 4 tissue sections from each retinal specimen were photographed in a masked manner under the microscope using the ×20 objective (Olympus BX53; Olympus, Münster, Germany). For each section, 3 photographs (1920 × 1200 pixels) were taken using a digital camera system (Olympus DP74), one from the very center of the section and one on each side, leaving approximately one image frame of space between photographs. Thus, for each staining/labeling, each retinal specimen generated 12 images, and each group 144 images to be analyzed. Furthermore, 24 images per staining/labeling derived from 2 normal porcine eyes were used as in vivo controls.

For quantification of overall retinal tissue damage (H&E staining), images were graded according to a tissue damage score (Table 1): 0 representing normal morphology, 1 mild, 2 moderate, and 3 severe damage. The following parameters were evaluated: pyknosis, lamination, and vacuolization with the inner and outer retina graded separately. The average score for the inner and outer retina was calculated from the 3 parameters above and analyzed statistically. The tissue damage score was graded individually in a masked manner by 2 observers.

Since examination of the ganglion cell layer is precarious in H&E staining, the specific ganglion cell marker RBPMS mentioned above was used to ascertain the number of remaining ganglion cells. Labeled cells on each image were counted in a masked manner. Only completely labeled cells, positioned in the ganglion cell layer were counted. Debris and fragmented cellular structures were excluded.

TUNEL labeling was quantified in a masked manner using ImageJ (US National Institutes of Health, Bethesda, MD, USA; https://imagej.nih.gov/ij/). On each image, the inner (from the inner nuclear layer to the inner limiting membrane) and outer retina (from the outer plexiform layer to the outer segments) were delineated using the polygon selection tool, and TUNEL-positive cells automatically counted using the ImageJ-plugin, Macro cell counter [Maidana et al., 2015].

The in vivo controls, representing the normal retina, displayed the expected laminated appearance with no pyknotic cells in the cellular layers (Fig. 1A). Small vacuoles were present in the ganglion cell layer (GCL), as well as the in the outer part of the inner nuclear layer (INL). No vacuoles were detected in the outer nuclear layer (ONL).

Retinas examined 90 min after enucleation, in general, displayed well-defined lamination, but revealed tissue damage in the inner and outer retina in the form of loss of cell organization in the INL and ONL, an increase in vacuolization in the INL, and an increased amount of pyknotic cells when compared with the in vivo controls (Fig. 1B, D; tissue damage score inner retina 0.12 ± 0.25 mean ± SD, p < 0.05; outer retina 0.18 ± 0.22, p < 0.001).

Retinas examined 240 min after enucleation displayed a highly significant increase in tissue damage compared with 90-min counterparts in the form of increased vacuolization, especially in the GCL and INL, increased pyknosis, and loss of cellular organization (Fig. 1C, D; inner retina 0.67 ± 0.32, p < 0.0001; outer retina 0.34 ± 0.23, p < 0.0001).

TUNEL staining of in vivo retinas displayed background fluorescence only with no cellular labeling (Fig. 1E). Occasional TUNEL-labeled cells could be seen in retinas examined 90 and 240 min after enucleation, however, only background fluorescence was present in the majority of images (Fig. 1F, G). Statistical analysis of the number of TUNEL-labeled cells revealed no significant difference when comparing the in vivo retinas and retinas examined 90 min after enucleation (Fig. 1H; inner retina 0.35 ± 0.67; outer retina 0.67 ± 3.41). Retinas examined 240 min after enucleation displayed a significant increase in labeled cells in the inner retina compared with the in vivo, as well as 90-min counterparts (Fig. 1H; inner retina 0.68 ± 1.52, p < 0.01 and p < 0.05). No significant difference in labeled cells was seen in the outer retina (0.23 ± 0.51).

RBPMS immunohistochemistry revealed well-labeled, large cell bodies in the GCL in the in vivo retinas (Fig. 1I). A highly significant decline in labeled retinal ganglion cells was detected in retinas examined 90 and 240 min after enucleation compared to in vivo controls (Fig. 1I–L; in vivo 35.42 ± 16.70; 90 min 17.86 ± 12.09; 240 min 11.90 ± 7.31, p < 0.0001). In addition, the number of labeled cells was significantly lower in retinas examined 240 min after enucleation compared with 90-min counterparts (p < 0.0001).

Retinal tissue explanted after 90 or 240 min and then cultured for 48 h, displayed a varying degree of disorganized lamination, pyknotic cells, as well as vacuoles in all 3 cellular layers in H&E-stained sections (Fig. 2A, B). A statistical analysis of the tissue damage score revealed no difference between the 2 groups (Fig. 2C; 90 min + culture: inner retina 0.77 ± 0.42, outer retina 0.96 ± 0.62; 240 min + culture: inner retina 0.80 ± 0.31, outer retina 0.91 ± 0.45).

TUNEL staining indicated apoptotic cells in all 3 nuclear layers in both culture groups, especially in the inner retina (Fig. 2D, E). Retinas cultured 240 min after enucleation displayed a significant increase of TUNEL-positive cells in the inner and outer retina compared with 90-min counterparts (Fig. 2F; 90 min + culture: inner retina 86.83 ± 67.74, outer retina 56.99 ± 47.06; 240 min + culture: inner retina 171.78 ± 60.86, outer retina 67.47 ± 38.54, p < 0.0001 (inner), p < 0.05 (outer)).

A disparity between the 2 time points was also evident in RBPMS-labeled sections where explants cultured 240 min after enucleation displayed a strong significant decrease of ganglion cells when compared with 90-min counterparts (Fig. 2G–I; 90 min + culture: 9.51 ± 4.73; 240 min + culture: 4.62 ± 5.12, p < 0.0001).

The level of the intracellular enzyme LDH was measured at 2, 24, and 48 h in the culture medium to assess cellular viability in the explants. LDH analysis revealed significantly higher activity (absorbance level) in culture medium derived from explants cultured 240 min after enucleation compared with 90-min counterparts at all time points (Fig. 3; 90 min + culture 2 h: 0.33 ± 0.12; 240 min + culture 2 h: 0.86 ± 0.52, p < 0.001; 90 min + culture 24 h: 1.12 ± 0.49; 240 min + culture 24 h: 2.04 ± 0.52, p < 0.0001; 90 min + culture 48 h: 2.20 ± 0.86; 240 min + culture 48 h: 2.53 ± 0.76, p < 0.05).

In this paper, we have explored the impact of time after enucleation in relation to tissue damage in the adult porcine retina. We focused our exploration on overall retinal morphological damage and apoptosis in the inner and outer retina and used immunohistochemistry to explore ganglion cell loss specifically. The results show that with the exception of ganglion cells, the retina is relatively well preserved 90 min after enucleation, whereas profound changes in the inner and outer retina are seen after 240 min. Enucleation of the eye induces global ischemia, and the increase in tissue damage seen 240 min after enucleation corresponds well to that reported in previous in vivo experiments where 90 min of ischemia have been reported to be well tolerated whereas 240 min are detrimental [Hayreh et al., 1980, 2004].

Even though retinas examined 90 min after enucleation displayed only minimal morphological deterioration and apoptosis, RBPMS labeling of ganglion cells was significantly decreased when compared to in vivo controls. In addition to global ischemia, the enucleation paradigm also includes transection of the optic nerve, which is well-known to produce ganglion cell death. However, this phenomenon is usually not detectable until after longer time periods [Watanabe and Fukuda, 2002]. When isolating the tissue prior to fixation, additional axonal trauma is induced within the retina, more central to the ganglion cell body. Therefore, we cannot exclude the possibility that the loss of RBPMS labeling in the ganglion cells 90 min after enucleation is primarily the result of axonal trauma and not ischemia.

Our results are at least in part in contrast to earlier findings concerning time-dependency after enucleation and may, therefore, be relevant to experimental research involving the large animal adult retina. Reinhard et al. [2016] reported that ganglion cell activity in the mini-pig retina was maintained for at least 50 h after enucleation when the tissue was kept under hypothermic conditions at 4°C. Similarly, Schultheiss et al. [2016] found that hypothermia prolonged ganglion cell function and preserved the retinal structure for up to 340 min in bovine eyes after enucleation. To minimize tissue damage, we ensured that our eyes were kept cold at 4°C after enucleation until the time of dissection. In accordance with the previous studies, our retinas examined 240 min after enucleation displayed only a minimal increase in TUNEL detectable apoptosis compared with 90-minute counterparts, however, overall morphology and ganglion cell survival were profoundly affected, as evident in H&E-stained sections and RBPMS immunohistochemistry, suggesting that at least in our experimental setup, hypothermia failed to protect the retina 240 min after enucleation.

To further elucidate any possible extended time-dependent effects in vitro, we examined retinas harvested 90 or 240 min after enucleation with subsequent culturing for 48 h. The in vitro retinal explant setup is increasingly used for a wide array of research activities, including disease mechanism exploration, pharmacological intervention as well as transplantation [Engelsberg and Ghosh, 2007; Rettinger and Wang, 2018]. Immature retinal porcine tissue can be cultured for extended time periods and even show much of the normal development, however, the adult retina degenerates rapidly within the in vitro environment, which has been attributed to axotomy and the loss of retinal pigment epithelium-photoreceptor contact [Winkler et al., 2002; Engelsberg et al., 2005]. We have previously shown that cell survival in adult retinal porcine explants, including ganglion cells and photoreceptors, can be significantly enhanced by placing the explant with the inner limiting membrane facing the culture membrane [Taylor et al., 2014]. To limit the culture-induced effect, we thus used this method for the culture experiment.

The cultured retinas of both groups displayed a similar degree of overall tissue damage in H&E staining. However, a strong significant increase in TUNEL-labeled cells in the inner retina combined with profound loss of ganglion cells in the 240-min group indicates that the initial discrepancy in tissue damage discussed above is indeed extended in vitro. The increased tissue damage state of retinas 240 min after enucleation was further confirmed in the culture medium analysis in which the 240-min explants already after 2 h of culture displayed a highly significant increase in LDH activity. The implication of these findings is relevant to a range of research involving large animal adult retinal tissue experiments and may be of particular interest for explorations in vitro.

To summarize, we here show that time after enucleation is a crucial factor for retinal tissue survival in a large animal model. The adult porcine retina displays profound tissue damage 240 min after enucleation, whereas retinas examined after 90 min remain relatively intact, and this discrepancy is extended when the tissue is placed under culture conditions. These time points are in accordance with previous in vivo experiments and should be considered in experiments involving the ex vivo adult porcine retina.

The authors extend their gratitude to Hodan Abdshill for excellent technical assistance and knowledge regarding all aspects of the study, Linnéa Taylor for advice on technical aspects of the culture experiment, and Jesper Bergwik and Amanda Kristiansson at the Section for Infection Medicine, Dept. of Clinical Sciences, Lund University, for guidance regarding the LDH analysis.

All procedures and handling of animals were in accordance with the guidelines and requirements of the Government Committee on Animal Experimentation at Lund University and also complied with the ARVO guidelines for animal experimentation.

The authors have no conflicts of interest to declare.

The study was supported by the Swedish Research Council (grant number 2015-02,772), The King Gustaf V and Queen Victoria Freemason Foundation.

F.S.: Conception, experimentation, data collection and analysis, manuscript preparation, correspondence. B.Å.: Conception, data analysis (LDH), proof reading. F.G.: Conception, data collection and analysis, manuscript preparation.