Retina regeneration in the chick embryo is not induced by spontaneous Mitf downregulation but requires FGF/FGFR/MEK/Erk dependent upregulation of Pax6

Abstract

Purpose: To elucidate the early cellular events that take place during induction of retina regeneration in the embryonic chick, focusing on the relationship between fibroblast growth factor (FGF) signaling and the regulation of Pax6 and Mitf.

Methods: The retina of embryonic day 4 (E4) chicks was removed and a heparin coated bead soaked in fibroblast growth factor 2 (FGF2) was placed into the optic cup. The pharmacological inhibitor PD173074 was used to inhibit FGF receptors, PD98059 was used to inhibit MAP kinase-kinase/extracellular signal-regulated kinase (MEK/Erk) signaling. Retroviral constructs for paired box 6 (Pax6), MEK, and microphthalmia (Mitf) were also used in overexpression studies. Immunohistochemistry was used to examine pErk, Pax6, Mitf, and melanosomal matrix protein 115 (MMP115) immunoreactivity and bromodeoxyuridine (BrdU) incorporation at different time points after removing the retina.

Results: The embryonic chick has the ability to regenerate a new retina by the process of transdifferentiation of the retinal pigment epithelium (RPE). We observed that during the induction of transdifferentiation, downregulation of Mitf was not sufficient to induce transdifferentiation at E4 and that FGF2 was required to drive Pax6 protein expression and cell proliferation, both of which are necessary for transdifferentiation. Furthermore, we show that FGF2 works through the FGFR/MEK/Erk signaling cascade to increase Pax6 expression and proliferation. Ectopic Mitf expression was able to inhibit transdifferentiation by acting downstream of FGFR/MEK/Erk signaling, likely by inhibiting the increase in Pax6 protein in the RPE.

Conclusions: FGF2 stimulates Pax6 expression during induction of transdifferentiation of the RPE through FGFR/MEK/Erk signaling cascade. This Pax6 expression is accompanied by an increase in BrdU incorporation. In addition, we show that Mitf is spontaneously downregulated after removal of the retina even in the absence of FGF2. This Mitf downregulation is not accompanied by Pax6 upregulation, demonstrating that FGF2 stimulated Pax6 upregulation is required for transdifferentiation of the RPE. Furthermore, we show that ectopic Mitf expression is able to protect the RPE from FGF2 induced transdifferentiation by inhibiting Pax6 upregulation.

Figure 1

Stimulation and inhibition of transdifferentiation. A: Schematic of an intact eye at E4. B: Schematic of a transdifferentiating eye three days after retina removal. Boxed area demonstrates orientation of C-H. F represents FGF2 bead. All eyes underwent retinectomy at E4 (C-H). C: Infection of the RPE with Rcas-MEK^DD at E4 induced transdifferentiation by E7. AMV3C2 staining represents the presence of the Rcas virus in the RPE. D: Infection of the RPE with Rcas-Pax6 at E4 induced transdifferentiation by E11. AMV3C2 staining represents the presence of the Rcas virus in the RPE. E: Infection of the RPE with Rcas-Mitf at E4 is able to inhibit transdifferentiation stimulated by FGF at E7. AMV3C2 staining represents the presence of the Rcas virus in the RPE. Only areas of the RPE that were not infected with Rcas-Mitf were able to transdifferentiate. F: Infection of the RPE with Rcas-MEK^DD at E4 induced transdifferentiation by E11. Some of the transdifferentiated RPE started to differentiate and express the photoreceptor marker visinin (green). Tissue was counterstained with DAPI (blue). G: Infection of the RPE with Rcas-Pax6 at E4 induced transdifferentiation by E11. Some of the transdifferentiated RPE starts to differentiate and express the photoreceptor marker visinin (green). Tissue is counterstained with DAPI (blue). H: Infection of the RPE with Rcas-GFP control did not induce transdifferentiation. AMV3C2 staining represents the presence of the Rcas virus in the RPE. Dashed lines outline the RPE and areas of transdifferentiation. Note: For C and D only infected RPE responded by transdifferentiating. L represents lens; RPE represents retina pigmented epithelium; tr represents transdifferentiation; r represents retina. Infected mesenchyme (m) did not transdifferentiate. Scale bar in H represents 50 μm and applies to A-H.

Figure 2

Expression of Pax6, Mitf, and MMP115 in the developing retinal pigment epithelium. A: Expression of Pax6 (top), Mitf (middle), and MMP115 (bottom) at E3.5. B: Expression of Pax6 (top), Mitf (middle), and MMP115 (bottom) at E4. C: Expression of Pax6 (top), Mitf (middle) and MMP115 (bottom) at E5. Scale bar represent 50 μm and applies to all panels.

Figure 3

FGF2 stimulates Pax6 expression in the retinal pigment epithelium when the retina is removed at E4. A: Expression of Pax6 (top), Mitf (middle), and MMP115 (bottom) after retina removal at E4 and in response to 6 h of exposure to FGF2. B: Expression of Pax6 (top), Mitf (middle), and MMP115 (bottom) after retina removal at E4 and in response to 12 h of exposure to FGF2. C: Expression of Pax6 (top), Mitf (middle), and MMP115 (bottom) after retina removal at E4 and in response to 24 h of exposure to FGF2. D: Expression of Pax6 (top), Mitf (middle), and MMP115 (bottom) after retina removal at E4 and in response to 24 h of exposure to a heparin bead. Inset image shows Pax6 expression without DIC overlay. Note that few cells are Pax6 positive. E: Expression of Pax6 (top), Mitf (middle), and MMP115 (bottom) after retina removal at E4 with no bead and visualized 24 h post-retinectomy. Inset image shows Pax6 expression without DIC overlay. Note that Pax6 levels in no-bead and heparin bead controls are significantly lower than those treated with FGF2. Scale bars represent 50 μm. A was taken at a closer magnification than that of B-E (see scale bars).

Figure 4

Pax6 and BrdU incorporation increase in response to FGF2 after retina removal at E4. A: Expression of Pax6 (top), BrdU (second), Pax6 and BrdU, Merge (third), and the DIC image (bottom) of the RPE after retina removal at E4 and 24 h of exposure to FGF2. B: Expression of Pax6 (top), BrdU (second), Pax6 and BrdU Merge (third) and the DIC image (bottom) of the RPE after retina removal at E4 and after 24 h of exposure to a heparin control bead. C: Expression of Pax6 (top), BrdU (second), Pax6 and BrdU Merge (third), and the DIC image (bottom) of the RPE after retina removal at E4 and after 24 h of exposure to FGF2 and FGFR1/FGFR2 blocking antibodies. D: Expression of Pax6 (top), BrdU (second), Pax6 and BrdU Merge (third), and the DIC image (bottom) of the RPE after retina removal at E4 and after 24 h of exposure to FGF2 and the FGFR inhibitor PD173074. E: Expression of Pax6 (top), BrdU (second), Pax6 and BrdU Merge (third) and the DIC image (bottom) of the RPE after retina removal at E4 and after 24 h of exposure to FGF2 and the MEK inhibitor PD98059. F: The number of Pax6-positive cells per μm of RPE in treatments shown in A-E was quantitated. The number of Pax6 positive cells was significantly reduced in all treatments compared to FGF alone. *p<0.01. G: The number of BrdU-positive cells per μm of RPE in treatments shown in A-E was quantitated. The number of BrdU positive cells was significantly reduced in all treatments compared to FGF alone. *p<0.01. H: The number of Pax6/BrdU positive cells per um of RPE in treatments shown in A-E was quantitated. The number of Pax6/BrdU positive cells was significantly reduced in all treatments compared to FGF alone. *p<0.01. I: Erk phosphorylation in the RPE after retina removal at E4 and after 24 h of exposure to FGF2 (top) and DIC image of the RPE (bottom). J: Erk phosphorylation in the RPE after retina removal at E4 and after 24 h of exposure to heparin (top) and DIC image of the RPE (bottom). K: Erk phosphorylation in the RPE after retina removal at E4 and after 24 h of exposure to FGF2 plus FGFR1/FGFR2-blocking antibodies (top) and DIC image of the RPE (bottom). L: Erk phosphorylation in the RPE after retina removal at E4 and after 24 h of exposure to FGF2 and the FGFR inhibitor PD173074 (top) and DIC image of the RPE (bottom). M: Erk phosphorylation in the RPE after retina removal at E4 and after 24 h of exposure to FGF2 plus the MEK inhibitor PD98059 (top) and DIC image of the RPE (bottom). N: Negative control for the immunostaining in panels I-M. O: DIC image of N. Scale bars represent 50 μm; ** represents FGF2 bead in C-E.

Figure 5

Mitf inhibits transdifferentiation by inhibiting Pax6, but not FGF signaling. A: Expression of Pax6 (top), BrdU (second), Pax6 and BrdU merge (third), and the DIC image (bottom) of the RPE after subretinally injecting Rcas-Mitf at E3, removing the retina at E4 and exposing the RPE to FGF2 for 24 h. B: Expression of Pax6 and BrdU in the RPE after 24 h of exposure to FGF2 and retina removal at E4. C: Erk phosphorylation in the RPE after subretinally injecting Rcas-Mitf at E3, removing the retina at E4 and exposing the RPE to FGF2 for 24 h (top), and the corresponding DIC image of the RPE (bottom). D: Mitf protein levels in the RPE after subretinally injecting Rcas-Mitf at E3, removing the retina at E4 and exposing the RPE to FGF2 for 24 h (top), and the corresponding DIC image of the RPE (bottom).