Developing rods transplanted into the degenerating retina of Crx-knockout mice exhibit neural activity similar to native photoreceptors

Abstract

Replacement of dysfunctional or dying photoreceptors offers a promising approach for retinal neurodegenerative diseases, including age-related macular degeneration and retinitis pigmentosa. Several studies have demonstrated the integration and differentiation of developing rod photoreceptors when transplanted in wild-type or degenerating retina; however, the physiology and function of the donor cells are not adequately defined. Here, we describe the physiological properties of developing rod photoreceptors that are tagged with green fluorescent protein (GFP) driven by the promoter of rod differentiation factor, Nrl. GFP-tagged developing rods show Ca(2 +) responses and rectifier outward currents that are smaller than those observed in fully developed photoreceptors, suggesting their immature developmental state. These immature rods also exhibit hyperpolarization-activated current (Ih ) induced by the activation of hyperpolarization-activated cyclic nucleotide-gated (HCN) channels. When transplanted into the subretinal space of wild-type or retinal degeneration mice, GFP-tagged developing rods can integrate into the photoreceptor outer nuclear layer in wild-type mouse retina and exhibit Ca(2 +) responses and membrane current comparable to native rod photoreceptors. A proportion of grafted rods develop rhodopsin-positive outer segment-like structures within 2 weeks after transplantation into the retina of Crx-knockout mice and produce rectifier outward current and Ih upon membrane depolarization and hyperpolarization. GFP-positive rods derived from induced pluripotent stem (iPS) cells also display similar membrane current Ih as native developing rod photoreceptors, express rod-specific phototransduction genes, and HCN-1 channels. We conclude that Nrl-promoter-driven GFP-tagged donor photoreceptors exhibit physiological characteristics of rods and that iPS cell-derived rods in vitro may provide a renewable source for cell-replacement therapy.

Figure 1. Physiology of wild type adult photoreceptors

(A) Calcium imaging of six week-old mouse retinal slices. (A-a) Slice preparation were loaded with Fura-2 and (A-b-d) stimulated with 100 μM glutamate (Glu) at 0:48 min and 80 mM KCl at 2:13 min. (A-b) Representative ratio images and (A-c) quantitation of [Ca²⁺]_i in GCL, red; INL, orange, yellow, green; OPL, blue; ONL, purple. Resting [Ca²⁺]_i and response to Glu stimulation were lower in the ONL compared to other layers. (A-d) Amplitude of Ca²⁺ response to Glu and KCl in the GCL (n = 9) and INL (n = 25) were compared with those in the ONL (n = 24). Data represent mean ± S.D. *p < 0.05. (A-e) Expression by microarray of voltage-dependent calcium channel subunits in isolated developing rod photoreceptors. (B) Patch clamp recording of rod photoreceptors in six week-old wild type retina. (B-a) DIC and Alexa⁵⁹⁴ fluorescence image of sliced retina showing the recorded cell. (B-b) Membrane voltage-clamp recording from −100 mV (red line) to 10 mV, with 10 mV increments. (B-c) Postnatal day (P)8 and adult retinal sections immunostained with anti-HCN-1 antibody. HCN-1 channel protein was expressed in the cell membrane of rod photoreceptors and in the INL at P8, and mostly localized to photoreceptor IS in the adult retina (white frame). (B-d) HCN-1 mRNA expression increased during postnatal retinal development. Data represent mean ± S.D of N = 3. *p < 0.05. (B-e) Expression by microarray of voltage-dependent potassium channel subunits in isolated developing rod photoreceptors. Scale bars, 50 μm in (A-a), 20 μm in (B-c). GCL, ganglion cell layer; INL, inner nuclear layer; OPL, outer plexiform layer; ONL, outer nuclear layer; OS, outer segment.

Figure 2. Cellular physiology of developing Nrlp-eGFP-positive rods

(A) Nrlp-eGFP-positive (blue, dark blue, and purple) and –negative (red, orange, yellow, and green) cells dissociated from postnatal day (P)3 Nrlp-eGFP mouse retina were cultured for 5 days on glass-coated dishes (B) loaded with Fura-2, and (C-E) stimulated with 100 μM Glu (at 1:00 min) and 80 mM KCl (at 5:40 min). (C-D) Absolute [Ca²⁺]_i was calculated from F340/F380 ratio images and plotted. (E) Average Ca²⁺ responses to Glu were comparable, whereas, Ca²⁺ responses to KCl in Nrlp-eGFP-positive cells (n = 7) were smaller than those in Nrlp-eGFP-negative cells (n = 8). (F-G) Membrane current in Nrlp-eGFP-positive developing rods was recorded by patch clamp technique from −100 mV (red line in G) to 10 mV, at 10 mV increments using Alexa⁵⁹⁴ (red) in the electrode solution for visualization (F). (H) I-V curve plotted from the data in (G) representing one single cell measurement. Data represent mean ± S.D. *p < 0.05. Scale bars, 20 μm in (A) and (F).

Figure 3. Ca²⁺ responses of Nrlp-eGFP-positive cells grafted in adult wild type retina

(A) Some Nrlp-eGFP-positive cells (red) transplanted in the sub-retinal space of wild type mouse, integrated in the ONL (arrows) and (B) showed IS-like structures (arrows). OS were stained with rhodopsin antibody (green) and nuclei with DAPI (blue). GFP does not localize to the OS. (C, D) Retinal slices were labeled with Fura-2 and [Ca²⁺]_i was calculated from ratio images. Retinal cells in the INL (red) showed higher baseline level of [Ca²⁺]_i than donor cells (orange, yellow, blue) or endogenous photoreceptors (green). Scale bars, 200 μm in (A); 20 μm in (B); 10 μm in (C). INL, inner nuclear layer; ONL, outer nuclear layer; OS, outer segment.

Figure 4. Membrane properties of Nrlp-eGFP-positive cells grafted in Crx-KO mouse retina

(A) Retinal flat mount image of developing Nrlp-eGFP-positive rods transplanted into the sub-retinal space of Crx-KO mice and expressing rhodopsin. (B) Some Nrlp-eGFP-positive donor cells extended rhodopsin-positive protrusions. (C) Patch clamp recording of membrane current in Nrlp-eGFP-positive donor cells identified by including Alexa⁵⁹⁴ in the electrode solution. (D, E) Membrane current in (D) control or (E) ZD7288-treated Nrlp-eGFP-positive donor cells recorded by membrane voltage-clamp from −100 mV (red line) to 10 mV, with 10 mV increments. (F) I-V curves were plotted using data from 4 trials before and after ZD7288 treatment. The voltage-dependent inward current recorded at membrane hyperpolarization was suppressed by treatment with ZD7288, an HCN channel blocker. Data represent mean ± S.D. *p < 0.05. (G) The transplanted retinas were sectioned and stained with anti-GFP (green), anti-HCN-1 antibodies (red) and DAPI (blue). Scale bars, 100 μm in (A), 20 μm in (B), (C) and (G).

Figure 5. Mouse iPS cell-derived Nrlp-eGFP-positive cells express rod photoreceptor markers

Mouse iPS cells generated from Nrlp-eGFP mouse fibroblasts were differentiated in vitro for 25 days to express Nrlp-eGFP. (A) Nrlp-eGFP-positive cells express rhodopsin and recoverin. (B) RT-PCR analysis for rod and phototransduction genes in mouse ES cell-derived Rx-GFP-positive cells (mES Rx+), mouse iPS cell-derived Nrlp-eGFP-positive cells (miPS Nrl+), Nrlp-eGFP-positive cells from P5 mouse retina (P5 Nrl+), retinal tissue from P8 mouse (P8 retina), and retinal tissue from adult (6w) mouse. (C) Confocal images of iPS-derived Nrlp-eGFP-positive cells expressing HCN-1 channel. Scale bars, 50 μm in (A) and (C).

Figure 6. Mouse iPS cell-derived Nrlp-eGFP-positive cells show rod-like properties in vitro

(A) Nrlp-eGFP-positive cells are visible in the differentiation culture. (B) Blight field image of the culture. (C) Patch clamp of mouse iPS cell-derived Nrlp-eGFP-positive cells visualized with Alexa594 in the electrode solution. (D) Membrane current in Nrlp-eGFP-positive developing rods was recorded from −100 mV (red line) to 10 mV, at 10 mV increments. (E) Outward current and I_h in developing rods (P8) were compared with those in mouse iPS cell-derived Nrlp-eGFP positive cells, grafted cells or wild-type adult rods. (F) Mouse iPS cell-derived Nrlp-eGFP positive cells were stained with Fura-2 and the Ca²⁺ imaging was performed. Arrows indicate Nrlp-eGFP positive cells. (G) Mouse iPS cell-derived Nrlp-eGFP positive cells showed Ca²⁺ responses to high-K⁺ stimulation. (H) Ca²⁺ responses in developing rods (P8) were compared with those in mouse iPS cell-derived Nrlp-eGFP positive cells, grafted cells or wild-type adult rods. Scale bars, 200 μm (A, B); 20 μm (C). Data represent mean ± S.D. *p < 0.05.