Novel Porcine Retina Cultivation Techniques Provide Improved Photoreceptor Preservation

Abstract

Age-related macular degeneration (AMD) is the leading cause of blindness in industrialized countries among people over 60 years. It has multiple triggers and risk factors, but despite intense research efforts, its pathomechanisms are currently not completely understood. AMD pathogenesis is characterized by soft drusen in Bruch's membrane and involves the retinal pigment epithelium-Bruch's membrane-choroid complex and adjacent structures, like photoreceptors. This study explores the potential of novel cultivation techniques to preserve photoreceptors in retinal explants to gain better insights in AMD pathology. The porcine retina explants were cultured for 4 and 8 days using three different explantation techniques, namely, control (photoreceptors facing down, touching the filter), filter (photoreceptors facing up, turned sample using a filter), and tweezers (photoreceptors facing up, turned sample using tweezers). Optical coherence tomography revealed that the tweezers method had the best capacity to limit thinning of the retinal explants. Both novel methods displayed advantages in maintaining outer segment thickness. Additionally, immunofluorescence evaluation revealed a better preservation of opsin⁺ cells and rhodopsin signal intensity in both novel methods, especially the tweezers method. Furthermore, RT-qPCR analysis demonstrated an upregulation of OPSIN and RHODOPSIN mRNA expression in tweezers samples at 8 days. Amacrine and bipolar cell numbers were not altered at day 4 of cultivation, while cultivation until 8 days led to reduced bipolar cell numbers. At 4 days, CALRETININ mRNA was upregulated in filter samples, but protein kinase C alpha expression was downregulated. Retinal ganglion cells were diminished in both novel techniques due to a direct physical contact with the insert. Remarkably, no difference in TUBB3 mRNA expression was detected among the techniques. Nevertheless, both novel methods exhibited an improved retention of photoreceptor cells. In conclusion, the tweezers technique was the most promising one. Due to the high homology of the porcine to the human retina, it provides a reasonable alternative to in vivo rodent models. Consequently, an adapted coculture system based on the current findings may serve as an ex vivo model suitable to analyze AMD pathomechanisms and novel therapeutic approaches.

Keywords: age-related macular degeneration; opsin; optical coherence tomography; organotypic retina culture; photoreceptor; porcine; rhodopsin.

FIGURE 1

Scheme of the used explant methods and study design. (A) Scheme of the three different explantation techniques, named control, filter, and tweezers method. The fourth group was a native one consisting of samples gained via the three different methods, which were analyzed at day 0. (B) Timeline of the study to investigate which explantation method best preserves the photoreceptor morphology ex vivo. Three techniques were compared during the cultivation periods of 4 and 8 days using spectral domain optical coherence tomography (SD-OCT), immunofluorescence (IF), and quantitative real-time PCR (RT-qPCR). Native samples were also included in the analysis.

FIGURE 2

Spectral Domain Optical Coherence Tomography (SD-OCT) analysis of porcine retinas. (A) Exemplary pictures of all measured time points and used techniques. Samples of all three explantation methods and the native group were investigated at 0 (= native), 4, and 8 days via SD-OCT. (B) At 4 days, no changes were noted in regard to the retinal thickness between control and native samples. A significantly thinner retinal thickness was revealed in filter (p < 0.001), but not in tweezers samples, compared to native retinas. No alterations were noted in filter and tweezers retinas compared to the controls. The filter group showed a significantly decreased retinal thickness compared to the tweezers group (p = 0.008). The retinal thickness of control and native samples was comparable at 8 days. While a significant thinning of the filter group was observed compared to native retinas (p < 0.001), no changes were noted between tweezers and native samples. Furthermore, the retinal thickness of the filter group was significantly diminished compared to tweezers samples (p = 0.008). OS, photoreceptor outer segments; ONL, outer nuclear layer; OPL, outer plexiform layer; INL, inner nuclear layer; IPL, inner plexiform layer; GCL, ganglion cell layer. Scale bars: 50 μm, values are mean ± SEM. n = 5/group. ***p < 0.001 vs. native group; ^¥¥p < 0.01 vs. filter group.

FIGURE 3

Total and bacillary layer thickness measurement in stained retinas. (A) Representative images of H&E-stained retinas for all three explantation methods and the native group in 400 × (upper panel) and 630 × magnification (lower panel). (B) At 4 days, a significant reduction of the total retinal thickness in control compared to native samples (p = 0.005) was observed, while a preservation was detected in the novel techniques filter and tweezers when compared to native samples. Comparing tweezers to control samples, a significantly thicker retinal thickness was noted (p = 0.009), while a similar thickness was observed between filter and control retinas. Similar results were found at 8 days. A reduction of the total retinal thickness was revealed comparing control to native samples (p = 0.04). A conservation of the total retinal thickness was detected comparing filter and tweezers to native or control retinas. (C) A significant thinning of the bacillary layer was seen in the control group compared to native (p = 0.04) and filter samples (p = 0.046). However, a better-preserved bacillary layer was noted in the tweezers method compared to the native and control samples at 4 days. A reduction in the bacillary layer thickness was measured at day 8 comparing the control and native samples (p = 0.001). The novel methods filter and tweezers maintained the bacillary layer thickness better than did control retinas. OS, photoreceptor outer segments; ONL, outer nuclear layer; OPL, outer plexiform layer; INL, inner nuclear layer; IPL, inner plexiform layer. Scale bars: 20 μm, values are mean ± SEM. n = 9–10/group. *p < 0.05 and **p < 0.01 vs. native group; ^#p < 0.05 and ^##p < 0.01 vs. controls.

FIGURE 4

Analysis of photoreceptors. (A) Exemplary immunofluorescence pictures of opsin (red) staining for L-cones and rhodopsin (green) staining for rods in photoreceptor outer segments. Nuclei were labeled with 4′,6-diamidino-2-phenylindole (DAPI) (blue). (B) Fewer opsin⁺ cells were found in control compared to native retinas (p = 0.02). The filter and tweezers method showed, compared to native retinas, a better preservation over a cultivation period of 4 days. The number of opsin⁺ cells was significantly higher in tweezers samples (p = 0.04) compared to the control ones, while no changes were noted between filter and control retinas. A severe loss of opsin⁺ cells was discovered in the control compared to native samples after 8 days (p < 0.001). In the novel methods filter and tweezers, the number of opsin-labeled cells was comparable to native retinas. When comparing filter and tweezers samples (both: p < 0.001) to the controls, more opsin⁺ cells could be detected. (C) OPSIN expression was not altered at 4 days. OPSIN mRNA expression in tweezers samples was significantly upregulated compared to control retinas at 8 days (p = 0.002). (D) The rhodopsin signal found in control (p < 0.001) and filter samples (p = 0.01) was significantly less intense at 4 days compared to that in native samples. A significantly higher signal intensity was documented in tweezers (p < 0.001) and filter retinas (p = 0.047) compared to the controls. Moreover, a higher rhodopsin signal intensity was observed in tweezers retinas compared to filter samples (p = 0.04) at 4 days. At 8 days of cultivation, all three methods showed a significantly diminished rhodopsin intensity compared to native samples (all: p < 0.001). (E) RHODOPSIN mRNA expression was not altered at 4 days. Quantitative real-time PCR (RT-qPCR) examination of RHODOPSIN demonstrated an upregulation in tweezers samples compared to control retinas at 8 days (p = 0.02). OS, photoreceptor outer segments; ONL, outer nuclear layer. Scale bar: 20 μm, values are mean ± SEM for immunofluorescence (IF) and median ± quartile + min/max for RT-qPCR. IF: n = 9–10/group; RT-qPCR: n = 5/group. *p < 0.05 and ***p < 0.001 vs. native group; ^#p < 0.05, ^##p < 0.01, and ^###p < 0.001 vs. controls; ^¥p < 0.05 vs. filter group.

FIGURE 5

Effect on cells in the inner nuclear layer. (A) Amacrine cells were labeled with calretinin (green) and bipolar cells with protein kinase C alpha (PKCα) (red). Cell nuclei were stained with 4′,6-diamidino-2-phenylindole (DAPI) (blue). (B) No differences were observed regarding the number of calretinin⁺ cells between all groups at 4 and 8 days. (C) An upregulation of the relative CALRETININ mRNA expression was detected in filter retinas compared to control samples at 4 days (p = 0.001). No differences were noted at 8 days. (D) The number of bipolar cells was not altered in any group at 4 days. However, fewer PKCα⁺ cells were discovered in control (p = 0.04), filter (p = 0.02), and tweezers retinas (p = 0.02) compared to those in native samples at 8 days. (E) A significant PKCα mRNA downregulation was observed in filter samples compared to control ones at 8 days (p = 0.02). ONL, outer nuclear layer; OPL, outer plexiform layer; INL, inner nuclear layer; IPL, inner plexiform layer. Scale bar: 20 μm, values are mean ± SEM for immunofluorescence (IF) and median ± quartile + min/max for quantitative real-time PCR (RT-qPCR). IF: n = 9–10/group; RT-qPCR: n = 5/group. *p < 0.05 vs. native group; ^#p < 0.05 and ^##p < 0.01 vs. controls.

FIGURE 6

Evaluation of retinal ganglion cells (RGCs). (A) RGCs were labeled with RNA-binding protein with multiple splicing (RBPMS) (red) and cell nuclei with 4′,6-diamidino-2-phenylindole (DAPI) (blue). (B) The amount of RBPMS⁺ cells in control retinas was comparable to that in native ones. The number of RGCs was significantly reduced in both novel methods, filter (p < 0.001) and tweezers (p < 0.001), compared to the native method at 4 days. A significant loss of RGCs was noted in filter (p = 0.002) and tweezers samples (p = 0.03) compared to control ones. At 8 days, fewer RGCs were visible for all three groups control, filter, and tweezers compared to those in native samples (all: p < 0.001). (C) No differences in the TUBB3 mRNA expression levels were detected between all groups at 4 and 8 days. GCL, ganglion cell layer; IPL, inner plexiform layer; INL, inner nuclear layer. Scale bar: 20 μm, values are mean ± SEM for immunofluorescence (IF) and median ± quartile + min/max for quantitative real-time PCR (RT-qPCR). IF: n = 9–10/group; RT-qPCR: n = 5/group. ***p < 0.001 vs. native group; ^#p < 0.05 and ^##p < 0.01 vs. controls.