Inhibition of thyroid hormone receptor locally in the retina is a therapeutic strategy for retinal degeneration

Abstract

Thyroid hormone (TH) signaling regulates cell proliferation, differentiation, and metabolism. Recent studies have implicated TH signaling in cone photoreceptor viability. Using mouse models of retinal degeneration, we demonstrated that antithyroid drug treatment and targeting iodothyronine deiodinases (DIOs) to suppress cellular tri-iodothyronine (T3) production or increase T3 degradation preserves cones. In this work, we investigated the effectiveness of inhibition of the TH receptor (TR). Two genes, THRA and THRB, encode TRs; THRB2 has been associated with cone viability. Using TR antagonists and Thrb2 deletion, we examined the effects of TR inhibition. Systemic and ocular treatment with the TR antagonists NH-3 and 1-850 increased cone density by 30–40% in the Rpe65^−/− mouse model of Leber congenital amaurosis and reduced the number of TUNEL⁺ cells. Cone survival was significantly improved in Rpe65^−/− and Cpfl1 (a model of achromatopsia with Pde6c defect) mice with Thrb2 deletion. Ventral cone density in Cpfl1/Thrb2^−/− and Rpe65^−/−/Thrb2^−/− mice was increased by 1- to 4-fold, compared with age-matched controls. Moreover, the expression levels of TR were significantly higher in the cone-degeneration retinas, suggesting locally elevated TR signaling. This work shows that the effects of antithyroid treatment or targeting DIOs were likely mediated by TRs and that suppressing TR protects cones. Our findings support the view that inhibition of TR locally in the retina is a therapeutic strategy for retinal degeneration management.—Ma, H., Yang, F., Butler, M. R., Belcher, J., Redmond, T. M., Placzek, A. T., Scanlan, T. S., Ding, X.-Q. Inhibition of thyroid hormone receptor locally in the retina is a therapeutic strategy for retinal degeneration.

Rod and cone photoreceptors degenerate in a variety of pathologic conditions, such as retinitis pigmentosa, Leber congenital amaurosis (LCA), cone–rod dystrophies, and age-related macular degeneration (AMD). Defects in a large number of genes are linked to inherited retinal degenerative disorders (see http://www.sph.uth.tmc.edu/RetNet/disease.htm), including those encoding enzymes involved in the recycling of 11-cis retinal in the retinal pigment epithelium (RPE), retinoid isomerase (RPE65) and lecithin retinol acyltransferase, and the phototransduction-associated proteins (opsins, subunits of transducin, cGMP phosphodiesterase-6, guanylate cyclase, and cyclic nucleotide–gated channel). There are currently no treatments for human retinal dystrophies. Despite a high genetic heterogeneity, the degenerating photoreceptors show common cellular disorder features, including oxidative damage (1, 2), endoplasmic reticulum stress (3, 4), and apoptosis (5, 6). These features offer the possibility of targeting common cell survival and death regulators/pathways to reduce photoreceptor death, regardless of the genetic origins of the diseases.

Thyroid hormone (TH) signaling regulates numerous physiologic functions, including cell growth, differentiation, and metabolic homeostasis (7–9). In the retina, TH signaling is well known for its regulation in cone opsin expression and patterning; it suppresses short-wave-sensitive (S)-opsin 1 expression, induces medium-wave-sensitive (M)-opsin 1 expression (10–12), and controls the dorsal–ventral gradient expression of cone opsin (10). TH signaling has also been linked to cone viability and degeneration. Treatment with tri-iodothyronine (T3) leads to cone death (13). Treatment with antithyroid drug or targeting iodothyronine deiodinases (DIOs) to suppress cellular T3 production or increase T3 degradation improves cone survival in mouse models of retina degeneration (14, 15).

T3 acts through TH receptors (TRs) that belong to the nuclear hormone receptor superfamily and function as ligand-dependent transcription factors (9). Two genes, THRA and THRB, encode related receptors across vertebrate species (9, 16). THRA1 is encoded by the THRA gene, and 2 THRB isoform splice variants, THRB1 and -2, are encoded by the THRB gene. In the retina, THRB2 is expressed in cones, whereas THRA1 is more widely expressed (17–19). THRB2 has been shown to mediate the regulation of TH in cone opsin expression (10, 11, 12) and cone viability (13). Cone death caused by high T3 was completely reversed by Thrb2 deletion (13). To understand the mechanisms of TH suppression effects and determine the potential of targeting TR locally in the retina for cone protection, we examined the effects of TR inhibition. We found that treatment with TR antagonist and Thrb2 deletion significantly improved cone survival in mouse models of retinal degeneration. Moreover, we observed upregulation of TR expression in the diseased retina, suggesting locally elevated TR signaling. Our findings provide insights into how antithyroid treatment or targeting DIOs leads to cone protection and the potential of targeting TR locally in the retina for cone protection.

MATERIALS AND METHODS

Mice, antibodies, and reagents

The Rpe65^−/−, Cngb3^−/−, Rpe65^−/−/Nrl^−/−, and Cngb3^−/−/Nrl^−/− mouse lines were generated as described previously (15, 20–22). Wild-type (C57BL/6J) and Cpfl1 mouse lines were obtained from The Jackson Laboratory (Bar Harbor, ME, USA). The Nrl^−/− mouse line was provided by Dr. Anand Swaroop (National Eye Institute, National Institutes of Health (NIH), Bethesda, MD, USA] (23). The Thrb2^−/− mouse line was provided by Dr. Douglas Forrest (National Institute of Diabetes and Digestive and Kidney Diseases, NIH) (12). The Rpe65^−/−/Thrb2^−/− and Cpfl1/Thrb2^−/− mouse lines were generated by cross-mating. All mice were maintained under cyclic light (12-h light–dark) conditions. Cage illumination was ∼7 foot-candles during the light cycle. All animal maintenance and experiments were approved by the local Institutional Animal Care and Use Committee (University of Oklahoma Health Sciences Center) and conformed to the guidelines on the care and use of animals adopted by the Society for Neuroscience and the Association for Research in Vision and Ophthalmology (Rockville, MD, USA).

Rabbit antibodies against mouse M-opsin and cone arrestin (CAR) were provided by Dr. Cheryl Craft (University of Southern California, Los Angeles, CA, USA). Rabbit anti-S-opsin antibody was provided by Dr. Muna Naash (University of Houston, Texas, TX, USA). Rabbit anti-THRB2 antibody was provided by Dr. Douglas Forrest (18). Biotinylated-peanut agglutinin (PNA) was purchased from Vector Laboratories, Inc. (B-1075; Burlingame, CA, USA). Rabbit anti-TATA-box-binding protein (TBP) antibody was obtained from Thermo Fisher Scientific (Waltham, MA, USA). NH-3 was synthesized and characterized, as described previously (24, 25). 1-850 was procured from Santa Cruz Biotechnology (Santa Cruz, CA, USA). Phospholipon 90G was provided by Dr. Bruce Baretz (Lipoid LLC, Newark, NJ, USA). All other reagents were purchased from Sigma-Aldrich (St. Louis, MO, USA), Bio-Rad (Hercules, CA, USA), and Thermo Fisher Scientific (Carlsbad, CA, USA).

Treatment of TR antagonists

For systemic treatment, mice received NH-3 (10.0 nmol/mouse/d) and vehicle (DMSO) by intraperitoneal injection once a day for 20 d, beginning on postnatal day (P)5. For ocular treatment, mice received the drug via intravitreal injection or topical administration. The intravitreal injections were performed under an Opmi Visu 140 surgical operating microscope (Zeiss, Thornwood, NY, USA). Mice received injections on P5 and were analyzed at P25. In brief, after anesthesia on ice for 2–3 min and complete dilation with 1% cyclopentolate hydrochloride (Akorn, Lake Forest, IL, USA), a drop of 2.5% methylcellulose was added to the corneal surface to visualize the fundus. A 28-gauge beveled hypodermic needle (BD Biosciences, Franklin Lakes, NJ, USA) was used to puncture the cornea carefully. TR inhibitor (1.0 µl) was injected intravitreally into 1 eye of each mouse with a NanoFil microsyringe injector system with a 33-gauge blunt needle (Hamilton Co., Reno, NV, USA). The contralateral eye was injected with 1.0 µl of vehicle. Only eyes with no apparent surgical complications were retained for further evaluation. For topical administration, the NH-3 eye drop (0.3% NH-3 in PBS containing 25% Phospholipon 90G) was formulated as described by Scheppke et al. (26). Mice received NH-3 eye drops or vehicle (25% Phospholipon 90G in PBS; ∼10 μl) 3 times a day for 10 d, beginning on P5. Eyes were collected at the end of the experiments for evaluation of cone death by TUNEL assay. Table 1 shows information for treatment of TR antagonists.

TABLE 1.

Information for TR antagonist treatment

TR antagonist	Dose/volume	Vehicle	Mouse line	Treatment route	Evaluation
NH-3	10.0 nmol/mouse/d	DMSO	Rpe65−/−	Intraperitoneal injection	Mice received daily injection for 20 d, beginning on P5, and were analyzed for cone density and cone death at P25.
NH-3	10 mM/1.0 μl	DMSO	Rpe65−/−	Intravitreal injection	Mice received a single injection at P5, and were analyzed for cone density at P25 (Rpe65−/−) and cone death at P15 (Rpe65−/−/Nrl−/−).
NH-3	0.3%/10.0 μl	25% Phospholipon 90G in PBS	Rpe65−/−/Nrl−/−	Eye drops	Mice received eye drops 3 times a day for 10 d, beginning on P5 and were analyzed for cone death at P15.
			Rpe65−/−/Nrl−/−
1-850	10 mM/1.0 μl	DMSO	Rpe65−/−	Intravitreal injection	Mice received a single injection at P5 and were analyzed for cone density at P25.

Antithyroid drug treatment

Mother Rpe65^−/−/Nrl^−/− mice received methimazole (0.05% wt/vol ) and sodium perchlorate monohydrate (1.0% wt/vol, both from Sigma-Aldrich) in drinking water (14), beginning on the day they delivered the pups. The treatment of both mother and pups continued for 15 d.

Weri-Rb1 cell culture and drug treatment

The retinoblastoma cell line Weri-RB1 [American Type Culture Collection (ATCC), Manassas, VA, USA] was cultured in RPMI-1640 medium (ATCC) with 10% fetal bovine serum (FBS) (15). For T3 treatment, cells were cultured in RPMI-1640 medium with 10% FBS replaced with 10% charcoal-stripped serum (Thermo Fisher Scientific) for 24 h, followed by T3 treatment at various concentrations for another 24 h. For TR antagonist treatment, cells cultured in RPMI1640 medium containing 10% charcoal-stripped serum were treated with the TR antagonist NH-3 or 1-850 at various concentrations for 16 h, followed by T3 treatment for another 24 h.

Scotopic and photopic electroretinography recordings

Full-field electroretinography (ERG) testing was performed (27). In brief, after overnight dark adaptation, animals were anesthetized by intraperitoneal injection of 85 mg/kg ketamine and 14 mg/kg xylazine. ERGs were recorded with an LKC system (Gaithersburg, MD, USA). Potentials were recorded with a platinum wire contacting the corneal surface through a layer of 2.5% methylcellulose. For assessment of scotopic responses, a stimulus intensity of 1.89 log · cd · s · m⁻² was presented to dark-adapted dilated mouse eyes in a Ganzfeld (GS-2000; Nicolet Instruments, Inc., Madison, WI, USA). To evaluate photopic responses, mice were adapted to a 1.46 log · cd · s · m ^–2 light for 5 min, then a light intensity of 1.89 log · cd · s · m^–2 was administered. Responses were differentially amplified, averaged, and stored with a Nicolet Compact-4 signal-averaging system.

Eye preparation, immunofluorescence labeling, and confocal microscopy

Mouse retinal wholemounts or cross sections were prepared for immunofluorescence labeling, as described previously (27, 28). For wholemount preparations, eyes were enucleated, marked at the superior pole with a green dye, and fixed in 4% paraformaldehyde (Polysciences, Inc., Warrington, PA, USA) for 30 min at room temperature. For retinal cross sections, mouse eyes were enucleated and fixed in Prefer (Anatech Ltd., Battle Creek, MI, USA) for 15 min at room temperature. Paraffin-embedded sections (5 µm thickness) passing vertically through the retina (along the vertical meridian passing through the optic nerve head) were prepared on a Leica microtome.

Immunofluorescence labeling of cones using antibodies against cone opsin and CAR was performed, as has been published (27, 28). In brief, retinal wholemounts or sections were blocked in Hanks’ balanced salt solution for 1 h at room temperature. Primary antibody incubation (anti-M-opsin, 1:1000; anti-S-opsin, 1:1000; and anti-CAR, 1:1000) was performed at room temperature for 1 h, followed by incubation with AlexaFluor-488 or -568, or FITC-conjugated secondary antibody. PNA immunohistochemistry was performed by using biotinylated PNA (1:250) and streptavidin-Cy3 (1:500). Fluorescent signals were imaged using an FV1000 confocal laser scanning microscope (Olympus, Melville, NY, USA) with FluoView imaging software (Olympus). Evaluation of cone density on retinal wholemounts and retinal cross sections was performed as published (28).

TUNEL assay

The TUNEL assays were conducted to evaluate photoreceptor apoptotic death (22). We used paraffin-embedded retinal sections and the In Situ Cell Death Fluorescein Detection Kit (Roche Diagnostics, Indianapolis, IN, USA) in this analysis. Immunohistochemical labeling was imaged with an FV1000 (Olympus) confocal laser-scanning microscope. The total TUNEL⁺ cells in the outer nuclear layer that passed through the optical nerve were counted and averaged from 4 sections of 1 eye.

Quantitative RT-PCR

Total RNA preparation and reverse transcription was performed (29). Human and mouse hypoxanthine guanine phosphoribosyl transferase 1 (HPRT1 and Hprt1) were included as an internal control. Supplemental Table 1 shows the primers used. The quantitative RT-PCR (qRT-PCR) assays were performed with a real-time PCR detection system (iCycler; Bio-Rad Laboratories). The relative gene expression value was calculated based on the ^ΔΔC_t method (29).

Retinal protein preparation, SDS-PAGE, and Western blot analysis

Protein SDS-PAGE and Western blot analysis were performed (22). Retinas were homogenized in homogenization buffer containing 20 mM HEPES-NaOH (pH 7.4), 3 mM EDTA, 320 mM sucrose, and protease inhibitor cocktail (4693124001) and phosphatase inhibitor cocktail (04906837001; both from Roche Applied Science). The homogenates were centrifuged at 1000 g for 10 min at 4°C. The resulting supernatant and pellet were subjected to extraction of cytosolic/membrane proteins and nuclear protein, respectively. The nuclear protein was extracted in the same buffer by sonication, followed by incubation on ice for 1 h with gentle agitation.

Protein separation and immunoblot analysis were performed (22). Retinal protein preparations were subjected to SDS-PAGE and transferred onto PVDF membranes. After blocking in 5% nonfat milk at room temperature for 1 h, blots were incubated with primary antibody overnight at 4°C (anti-THRB2, 1:1000; anti-TBP, 1:1000). After rinsing in Tris-buffered saline, blots were incubated with horseradish peroxidase (HRP)-conjugated secondary antibodies (1:20,000) for 1 h at room temperature. SuperSignal West Dura Extended Duration chemiluminescent substrate (Thermo Fisher Scientific) was used to detect binding of the primary antibodies to their cognate antigens. HyBlot CL autoradiography films (Denville Scientific, Inc., Metuchen, NJ, USA) were used to develop the target proteins, and Photoshop CS5 (Adobe, Mountain View, CA, USA) was used to analyze the signal density (30).

Statistical analysis

Results are expressed as means ± sem of number of observations. One-way ANOVA was used to analyze for significance within sets of data, and unpaired, 2-sample Student’s t test was used for differences between 2 groups of data. Differences were considered statistically significant at P < 0.05. Data were analyzed and graphed with Prism software (GraphPad Software, La Jolla, CA, USA).

RESULTS

Intraperitoneal delivery of TR antagonist NH-3 improved cone survival and function in Rpe65^−/− mice

NH-3, a well-characterized TR antagonist (24, 25), was the first drug tested to determine whether blocking TR protects cones. Before performing the animal experiments, we examined the effects of NH-3 (a 5′-phenylethynyl derivative of the GC-1 agonist) on a T3-induced response in a human retinoblastoma-derived photoreceptor-like cell line, Weri-Rb1 cells. These cells express TR and respond to T3 with increased expression of OPN1MW (31, 32). We first established the concentration-dependent response pattern of expression after T3 treatment (Supplemental Fig. 1A). To examine the effects of NH-3, cells were pretreated with NH-3 at various concentrations for 16 h, followed by T3 treatment for another 24 h. Cells were then analyzed for OPN1MW expression using qRT-PCR. The assay results revealed that pretreatment with NH-3 inhibited T3-induced OPN1MW expression in a concentration-dependent manner (Supplemental Fig. 1B).

We then performed animal experiments using Rpe65^−/− mice, a model of LCA. These mice display early onset, rapid loss of cones, with more severe loss of cones in the ventral and central areas than in the dorsal area, and severe impairment of the cone light response (20, 33), and show improved cone survival after antithyroid drug treatment (14) and targeting DIOs to suppress cellular T3 production or increase T3 degradation (15). P5 Rpe65^−/− mice received NH-3 (10.0 nmol/mouse/d, i.p injection) or vehicle (DMSO) daily for 20 d and were then subjected to retinal function and survival evaluations by ERG recordings and cone marker labeling. We found that treatment with NH-3 significantly increased photopic light responses in Rpe65^−/− mice. The photopic a- and b-wave responses in NH-3-treated mice were increased by ∼50%, compared with vehicle-treated controls (Fig. 1A). In contrast, the scotopic responses were not different between NH-3-treated and vehicle-treated mice. Cone survival and death were evaluated by analyzing cone density and TUNEL assays. We found that the treatment increased cone density by ∼30%, as evaluated by PNA labeling on retinal wholemounts (Fig. 1B and Supplemental Fig. 2A shows retinal wholemount images of an entire retina with the indicated regions used for counting PNA-labeled cones), and PNA and M-opsin labeling on retinal sections (Fig. 1C, D), and reduced the number of TUNEL⁺ cells by about 35% (Fig. 1E).

Figure 1.

Intraperitoneal delivery of NH-3 improved cone survival and light response in Rpe65^−/− mice. Mice received NH-3 (10.0 nmol/mouse) or vehicle via daily intraperitoneal injection for 20 d, beginning on P5 and were evaluated for photopic and scotopic light responses with ERG and cone density with immunolabeling. A) Increased photopic light response in Rpe65^−/− mice treated with NH-3. Shown are representative photopic waveforms and a/b-wave amplitude quantifications of photopic and scotopic responses. B–D) Increased cone density in Rpe65^−/− mice treated with NH-3. Shown are representative confocal images of immunofluorescence labeling of PNA on retinal wholemounts (B), and immunofluorescence labeling of PNA (C) and M-opsin (D) and on retinal sections, and the corresponding quantitative analysis. E) Reduced cone death in Rpe65^−/− mice treated with NH-3. Representative confocal images show TUNEL⁺ cells on retinal sections and the corresponding quantitative analysis. OS, outer segment; ONL, outer nuclear layer; INL, inner nuclear layer. Scale bars, 50 μm. Data are means ± sem of assays of eyes from 4 to 9 mice.

Intravitreal delivery of NH-3 improved cone survival in Rpe65^−/− and Rpe65^−/−/Nrl^−/− mice

We also examined the effects of local ocular delivery of NH-3. We treated P5 Rpe65^−/− mice with a single intravitreal injection of NH-3 (10 mM, 1.0 μl) or vehicle (DMSO), and evaluated cone survival/death at P25. As with the systemic treatment experiments, the NH-3-treated mice showed significantly improved cone survival and reduced cone death. Cone density in NH-3-treated mice was increased by ∼30%, compared with vehicle-treated controls (Fig. 2A, B and Supplemental Fig. 2B shows retinal wholemount images of an entire retina with the indicated regions used for counting PNA-labeled cones). The number of TUNEL⁺ cells in NH3-treated Rpe65^−/− mice was reduced by ∼30% (Fig. 2C). In this experiment, we also included a group of Rpe65^−/−/Nrl^−/− mice with RPE65 deficiency on a cone-dominant background to verify that the TUNEL⁺ cells detected in Rpe65^−/− mice reflected the status of cone photoreceptors. Like Rpe65^−/− mice, Rpe65^−/−/Nrl^−/− mice display early-onset, rapid cone degeneration (34, 35). We treated P5 Rpe65^−/−/Nrl^−/− mice with a single intravitreal injection of NH-3 (10 mM, 1.0 μl) or vehicle (DMSO), and performed TUNEL assays at P15. Similar to the findings in Rpe65^−/− mice, the number of TUNEL⁺ cells was significantly reduced in NH-3-treated Rpe65^−/−/Nrl^−/− mice, compared with vehicle-treated controls (Fig. 2D). Hence, the experimental results from both Rpe65^−/− and Rpe65^−/−/Nrl^−/− mice showed that a single intravitreal delivery of NH-3 reduced cone death.

Figure 2.

Intravitreal delivery of NH-3 increased cone density and reduced cone death in Rpe65^−/− and Rpe65^−/−/Nrl^−/− mice. A, B) Intravitreal delivery of NH-3 increased cone density in Rpe65^−/− mice. Mice received a single injection at P5 and were evaluated for cone survival at P25. Shown are representative confocal images of immunofluorescence labeling of PNA on retinal wholemounts (A) and retinal sections (B) and the corresponding quantitative analysis. C, D) Intravitreal delivery of NH-3 reduced the number of TUNEL⁺ cells in Rpe65^−/− (C) and Rpe65^−/−/Nrl^−/− (D) mice. Rpe65^−/− and Rpe65^−/−/Nrl^−/− mice received injections at P5, and were analyzed for TUNEL⁺ cells at P25 and P15, respectively. Representative confocal images show TUNEL⁺ cells on retinal sections, and the corresponding quantitative analysis. ONL, outer nuclear layer; INL, inner nuclear layer. Scale bars, 50 μm. Data are represented as means ± sem of assays of eyes of 5 to 7 mice.

Topical delivery of NH-3 improved cone survival in Rpe65^−/−/Nrl^−/− mice

We topically delivered NH-3 to further study its effects. P5 Rpe65^−/−/Nrl^−/− mice received NH-3 by eye drop (0.3% NH-3 in PBS containing 25% Phospholipon 90G) 3 times a day for 10 d. Mice were then evaluated for cone death by using TUNEL assays. We found that the number of TUNEL⁺ cells in NH-3-treated mice was reduced by ∼40%, compared with the number in vehicle-treated controls (Fig. 3A, B).

Figure 3.

Topical delivery of NH-3 reduced cone death in Rpe65^−/− and Rpe65^−/−/Nrl^−/− mice. A, B) Topical delivery of NH-3 reduced the number of TUNEL⁺ cells in Rpe65^−/−/Nrl^−/− mice. Mice received topical administration of NH-3 at P5, and were analyzed for TUNEL⁺ cells at P15. C, D) Treatment with antithyroid drugs reduced the number of TUNEL⁺ cells in Rpe65^−/−/Nrl^−/− mice. Mice received antithyroid drug treatment for 15 d, beginning on P1, and were analyzed for TUNEL⁺ cells at P15. Shown are representative confocal images of TUNEL⁺ cells on retinal sections (A, C), and the corresponding quantitative analysis (B, D). ONL, outer nuclear layer; INL, inner nuclear layer. Scale bars, 50 μm. Data are represented as means ± sem of assays of eyes from 4 to 11 mice.

We previously showed increased cone density in Rpe65^−/− mice after antithyroid drug treatment. In the present work, we examined the effects of antithyroid drug treatment on cone death. Mother Rpe65^−/−/Nrl^−/− mice received antithyroid drug treatment for 15 d when pups were at P1, as described previously (14). The eyes of pups were then collected for TUNEL assays. The number of TUNEL⁺ cells in the antithyroid drug-treated mice was significantly less than that of untreated controls (Fig. 3C, D).

Intravitreal delivery of 1-850 improved cone survival in Rpe65^−/− mice

We further examined the effects of TR inhibition with a second TR antagonist, the phenyl-hydrazine-carboxamide compound 1-850 (36). Before performing the animal experiment, we examined the effects of 1-850 on T3-induced M-opsin expression in Weri-Rb1 cells. Cells were pretreated with 1-850 at various concentrations for 16 h, followed by T3 treatment for another 24 h. Cells were then analyzed for M-opsin expression by qRT-PCR. The assay results showed that pretreatment with 1-850 inhibited T3-induced M-opsin expression in a concentration-dependent manner (Supplemental Fig. 3).

We next treated P5 Rpe65^−/− mice with 1-850 (10 mM, 1.0 μl) or vehicle (DMSO) via a single intravitreal injection and evaluated cone survival at P25. We found that treatment with 1-850 significantly improved cone survival in Rpe65^−/− mice. Cone density was increased by ∼25% in the drug-treated mice, as evaluated by PNA (Fig. 4A) and CAR labeling (Fig. 4B) of retinal wholemounts and PNA labeling (Fig. 4C) and S-opsin (Fig. 4D) expression of retinal sections.

Figure 4.

Intravitreal delivery of 1-850 improved cone survival in Rpe65^−/− mice. Mice received injections at P5 and were evaluated for cone survival at P30. Representative confocal images show immunofluorescence labeling of PNA (A) and CAR (B) on retinal wholemounts, immunofluorescence labeling of PNA (C) and S-opsin (D) on retinal sections, and the corresponding quantitative analyses. OS, outer segment; ONL, outer nuclear layer; INL, inner nuclear layer. Scale bars, 50 μm. Data are means ± sem of assays of eyes from 4 to 9 mice.

Deletion of Thrb2 improved cone survival in Rpe65^−/− mice

THRB2 is expressed only in cones in the retina; Thrb2 deletion was shown to rescue high T3-induced cone death (13). Therefore, we generated the double-knockout Rpe65^−/−/Thrb2^−/− strain to explore the role of THRB2. We found that the deletion of Thrb2 significantly improved cone survival in Rpe65^−/− mice. As analyzed via PNA labeling on retinal wholemounts, cone density in the more severely degenerated area (ventral part) was increased by about 4-fold in P30 Rpe65^−/−/Thrb2^−/− mice, compared with that in the age-matched Rpe65^−/− mice (Fig. 5A and Supplemental Fig. 4 show retinal wholemount images of an entire retina with the indicated regions used for counting PNA-labeled cones). Similar findings were obtained from analysis on retinal sections (Fig. 5B). Deletion of Thrb2 also significantly increased the number of cones in both ventral and dorsal regions of the retinas labeled by anti-S-opsin antibody (Fig. 5C). This finding is consistent with the previous findings that suppressing TH signaling increases S-opsin expression and diminishes the dorsal-ventral gradient patterning (12). However, despite the increased expression, the S-opsin is mislocalized.

Figure 5.

Improved cone survival in Rpe65^−/−/Thrb2^−/− mice. Cone density was analyzed in P30 Rpe65^−/−/Thrb2^−/− and Rpe65^−/− mice. Representative confocal images show immunofluorescence labeling of PNA on retinal wholemounts (A), immunofluorescence labeling of PNA (B) and S-opsin (C) on retinal sections, along with the corresponding quantitative analysis. OS, outer segment; IS, inner segment; ONL, outer nuclear layer; OPL, outer plexiform layer. Data are represented as means ± sem of assays of eyes from 6 to 11 mice.

Deletion of Thrb2 improved cone survival in Cpfl1 mice

The effects of Thrb2 deletion was also examined in Cpfl1 mice, a naturally occurring achromatopsia/cone dystrophy model with Pde6c defect (37, 38). These mice show improved cone survival after antithyroid drug treatment (14) and targeting DIO to increase T3 degradation (15). The double-knockout Cpfl1/Thrb2^−/− line was generated by cross-breeding. We found that the deletion of Thrb2 significantly improved cone survival in Cpfl1 mice. As analyzed, via PNA labeling on retinal wholemounts, cone density in the ventral retina was increased by ∼40% and 110% in Cpfl1/Thrb2^−/− mice at P30 and P60, respectively, compared with age-matched Cpfl1 mice (Fig. 6A, B and Supplemental Fig. 5 show retinal wholemount images of an entire retina with the indicated regions used for counting PNA-labeled cones). Similar to results from Rpe65^−/−/Thrb2^−/− mice, deletion of Thrb2 also significantly increased the number of cones labeled with anti-S-opsin antibody in Cpfl1/Thrb2^−/− mice (Fig. 6C and Supplemental Fig. 6). However, the S-opsin mislocalization was less significant (Supplemental Fig. 7) than that in Rpe65^−/−/Thrb2^−/− mice (Fig. 5C).

Figure 6.

Improved cone survival in Cpfl1/Thrb2^−/− mice. Cone density was analyzed in P30 and P60 Cpfl1/Thrb2^−/− and Cpfl1 mice. Representative confocal images show immunofluorescence labeling of PNA on retinal wholemounts of P30 (A) and P60 (B) mice and S-opsin labeling on retinal wholemounts of P30 (C) mice, along with the corresponding quantitative analysis. Data are means ± sem of assays of eyes from 4 to 7 mice.

Enhanced expression levels of TR in cone degeneration retinas

The observations that antithyroid drug treatment or TR inhibition protects cones lead to the question of how suppressing TR signaling preserves cones. Excessive TH is known to cause cone death in both healthy and degenerating cones (13, 14). The TR signaling suppression-induced protection suggests that there is elevated TR signaling in degenerating cones/retinas the degenerating cones become more sensitive to TH, or both. Indeed, we observed that the expression of DIO type 2, which converts the prohormone thyroxine (T4) to the bioactive hormone T3, was increased in cone degeneration retinas (15). In this work, we examined TR expression in Rpe65^−/−/Nrl^−/− retinas. The use of the cone-dominant mice was to facilitate the detection of cellular alterations in cones. We first examined retinal expression of Thrb2. Total RNA and proteins extracted from P15 and P30 Rpe65^−/−/Nrl^−/− and Nrl^−/− retinas were used for qRT-PCR and immunoblot analysis. We found that the levels of Thrb2 were increased by ∼4-fold (Fig. 7A), whereas the levels of THRB2 were increased by ∼30%, in Rpe65^−/−/Nrl^−/− mice, compared with Nrl^−/− controls (Fig. 7B). The levels of Thrb1 and Thra1 were also significantly elevated in Rpe65^−/−/Nrl^−/− mice (Fig. 7C, D). In contrast, there was no difference in Gnat2 expression between the 2 genotypes (Fig. 7E). In this experiment, we also examined the expression of the nongenomic TH receptor. Besides the classic TR-mediated responses, TH has been suggested to exert some action through nongenomic (non-TR) actions, which do not include initial nuclear actions of TR or gene transcription, but involve the cell surface receptor integrin αVβ3 subunits (39, 40). We found no elevation of the expression levels of Itgav and Itgb3 mRNA in the Rpe65^−/−/Nrl^−/− retinas (Fig. 7F, G). In addition, we examined retinal expression of Thrb2 in Cngb3^−/−/Nrl^−/− mice with cone cyclic nucleotide–gated (CNG) channel B subunit (CNGB3) deficiency on the Nrl^−/− background, a model of achromatopsia/cone dystrophy (14, 22, 30). These mice showed increased expression of Thrb2 (Fig. 7H), reduced expression of Opn1sw (Fig. 7I), and unchanged expression of Gnat2 (Fig. 7J).

Figure 7.

Enhanced expression of TR in the cone-degeneration retinas. Retinal expression levels of TR, Gnat2, Itgav, and Itgb3 were analyzed in P15 and P30 Rpe65^−/−/Nrl^−/− and Nrl^−/− mice, using qRT-PCR and immunoblot analysis. A) Increased levels of Thrb2 mRNA in Rpe65^−/−/Nrl^−/− retinas. B) Increased levels of THRB2 protein in Rpe65^−/−/Nrl^−/− retinas. Shown are representative images of the immunoblot detections and the corresponding densitometric analyses. The nuclear protein TBP was used as a loading control. C–E) Increased levels of Thrb1 (C) and Thra1 (D) and unchanged levels of Gnat2 (E) in Rpe65^−/−/Nrl^−/− retinas. F, G) No elevation of Itgav (F) and Itgb3 (G) in Rpe65^−/−/Nrl^−/− retinas. H–J) Retinal expression levels of Thrb2 (H), Opn1sw (I), and Gnat2 (J) mRNA in P15 Cngb3^−/−/Nrl^−/− and Nrl^−/− retinas. Data are means ± sem of 3 assays with retinas from 4 to 6 mice for mRNA analysis and from 8 to 9 mice for protein analysis.

DISCUSSION

Based on our previous findings that antithyroid drug treatment and targeting cellular DIOs to suppress cellular T3 production or increase T3 degradation protect cones in retinal degeneration model mice (14, 15), the current work investigated the effectiveness of TR inhibition. We found that treatment with a TR antagonist or deletion of Thrb2 improved cone survival, supporting the idea that cone protection from antithyroid drug treatment and targeting cellular DIOs is mediated by TR. The protective effects of TR antagonist treatment were demonstrated with both systemic and ocular administration. Thus, targeting TR locally in the retina may represent a therapeutic strategy to protect cones in retinal degeneration. We also observed upregulation of TR expression in the cone degeneration retinas. Along with increased expression of DIO2 in the cone degeneration retinas (15), our findings suggest locally elevated TH signaling, and a mechanism through which suppressing TH signaling protects cones.

Inhibition of TR protects cones

The pharmacological evidence was obtained through the use of the 2 different TR antagonists, NH-3 and 1-850. There are several reported TR antagonists (24, 25, 36, 41–43), and NH-3 and 1-850 are best characterized pharmacologically in terms of potency and specificity to TR (24, 25, 36, 42). NH-3 was shown to inhibit T3-induced cofactor binding (24, 25), gene expression (42), and tadpole metamorphosis (44). NH-3 has also been shown to increase thyroid stimulating hormone in rats (43) and suppress TH signaling-induced filial imprinting in an avian model (45). 1-850 was shown to inhibit T3-mediated coactivator recruitment to TR and block T3-dependent gene expression in rats (36).

In the present work, we demonstrated the cone protective effects of NH-3 given through different administration routes: intraperitoneal injection, intravitreal injection, and topical eye drops. After NH-3 treatment, mice showed increased cone density, reduced cone death, and improved cone light responses. The improved cone function is likely a result of improved cone survival. It is worth mentioning that we observed a significant topical treatment effect, which was as effective as or a little better than the other routes, in reducing the number of TUNEL⁺ cells (see Figs. 3A, B and 2C, D for comparison of topical treatment and intravitreal treatment, and Figs. 3A, B and 1E for comparison of topical treatment and intraperitoneal treatment). It is generally thought that topical treatment does not provide adequate delivery of drug to the posterior eye. Our findings might have been affected by the following situations: 1) The intravitreal treatment was only administered once (single injection), and observations were conducted at 10 d (Rpe65^−/−/Nrl^−/− mice) or 20 d (Rpe65^−/− mice) after injection. In contrast, the topical eye drop treatment was performed 3 times a day for 10 d (Rpe65^−/−/Nrl^−/− mice); 2) The systemic treatment effect could be significantly affected by the blood–retinal barrier penetration capability of the drug, about which we know little at this time. In addition, the vehicle Phospholipon 90G (a liposome) obviously worked. Phospholipon 90G has been shown to be a promising vehicle for topical delivering of small molecules to the posterior eye. It has been reported to successfully deliver small-molecule VEGFR2/Src kinase inhibitors to posterior ocular tissues in mice and rabbits, and the topical treatment of the inhibitors suppressed VEGF-induced retinal vascular leakage in mice (26). The effects of 1-850 were demonstrated with a single intravitreal injection. After 1-850 treatment, mice showed increased cone density and reduced cone death. The rescue effects from NH-3 appeared better than those from 1-850, which might be associated with the greater potency of NH-3. We found NH-3 to be ∼100-fold more potent than 1-850 in inhibiting T3-induced responses in Weri-Rb1 cells (see Supplemental Figs.1 and 3). The findings from Rpe65^−/−/Thrb2^−/− and Cpfl1/Thrb2^−/− mice provided genetic evidence supporting the effects of TR inhibition and the involvement of THRB2.

As has been well recognized, Rpe65^−/− mice display early-onset, rapid loss of ventral and central cones (20, 33). Deletion of Thrb2 significantly improved cone survival in Rpe65^−/− mice. Ventral cone density increased by ∼4-fold in Rpe65^−/−/Thrb2^−/− mice at P30, compared with age-matched Rpe65^−/− mice. The improvement in the ventral retina in Rpe65^−/−/Thrb2^−/− mice was more significant than that in Rpe65^−/− mice treated with TR antagonists, which may be associated with the efficacy and bioavailability of the drug administration. The Cpfl1 mouse is a naturally occurring achromatopsia/cone dystrophy model with Pde6c defect, and displays early-onset, rapid cone degeneration (37, 38). The double-knockout Cpfl1/Thrb2^−/− mice show significantly improved cone survival, compared with age-matched Cpfl1 mice, demonstrating cone protection by TR suppression.

It should be pointed out that the rescue achieved by TR inhibition is partial. The drug treatment resulted in 30–40% increase in cone density. This partial rescue could be associated with: 1) the drug treatment effectiveness, which is associated with drug efficacy, bioavailability, and distribution; 2) the possible involvement of nongenomic TH action; and 3) the contributions from non-TH factors. Previous studies have shown that the regulation of TH signaling in cone viability is mediated mainly via TR, as T3-induced cone death was completely rescued in Thrb2^−/− mice (13). Unlike TR, the expression of Itgav and Itgb3 (encoding the nongenomic receptor integrin αVβ3 subunits) was not elevated in the diseased retinas. These findings do not support a major contribution from a nongenomic mechanism; however, we cannot fully exclude a minor contribution. Such a nongenomic mechanism via TH binding to the integrin αVβ3 subunits has been shown to participate in angiogenesis and neovascularization (46, 47). It is also important to emphasize that, although TH signaling plays a role in cone degeneration, it is not the sole factor. Other non-TH factors could contribute to disease progression. In addition, the original pathologic conditions would not be corrected by TR inhibition (i.e., the lack of 11-cis-retinal chromophore due to deficiency of RPE65) (20, 48). Indeed, although S-opsin expression was increased in Rpe65^−/−/Thrb2^−/− mice, cone opsin mislocalization was not corrected.

The effect of TR inhibition appears to be independent of TR regulation of cone opsin expression

TR signaling plays a central role in cone opsin expression and the dorsal–ventral patterning in the mouse retina. It increases expression of M-opsin, suppresses expression of S-opsin, and promotes/maintains their dorsal–ventral gradient distribution (10, 12). Thrb2^−/− mice show upregulation of S-opsin expression and downregulation of M-opsin (12). The expression level of S-opsin has been linked to cone viability. The high expression of S-opsin and its accumulation/aggregation in the ventral retina has been associated with the rapid and severe cone degeneration in Lrat^−/− mice (49), a model with disease pathogenesis similar to that of Rpe65^−/− mice; deletion of Opn1sw was shown to dramatically improve cone survival in Lrat^−/− mice (50). Nevertheless, the TR regulation of cone viability appears to be independent of its regulation of cone opsin expression, because high TH signaling (which inhibits expression of S-opsin) leads to cone death (13), whereas suppressing TH signaling (which stimulates S-opsin expression) protects cones (14). Thus, the effect of TR inhibition is most likely independent of TR regulation of cone opsin expression. This notion is supported by the following findings: 1) the antithyroid drug-induced cone protection was accompanied by increased expression of S-opsin (14); 2) deletion of Thrb2, which upregulates expression of S-opsin (12), resulted in improved cone survival; and 3) T3-induced cone death was accompanied by reduced S-opsin expression (14). The view is also supported by the findings from noncone cells, which do not express cone opsin: 1) stimulating TH signaling induces death of rods; 2) high TH signaling causes auditory deficits and cochlear degeneration (51); and 3) high TH signaling induces death of other cell types, including cancer cells (52–55).

Upregulation of TR expression in the cone-degenerating retinas

The expression levels of TR were elevated in Rpe65^−/−/Nrl^−/− retinas, which contrasts with the reduced expression of cone phototransduction proteins in these mice and Rpe65^−/− mice (33–35), and the unchanged expression of the nongenomic membrane receptor. The elevated expression of TR was also observed in the Cngb3^−/−/Nrl^−/− mouse model of cone dystrophy/achromatopsia. The increased TR expression may represent a common alteration in the degenerating retinas, although the underlying mechanism remains to be determined. Together with our previous observations showing the increased expression of DIO2 in cone degeneration retinas, our findings suggest locally elevated TH signaling. It is worth noting that the upregulation of TR has also been reported in other pathologic conditions. Upregulation of Thrb1 and Thrb2 was detected in the myocardium of dogs with dilated cardiomyopathy or chronic valvular disease (56). This upregulation could be a result of the pathologic conditions, which could in turn further contribute to or worsen the disease progression. The observation that suppressing TR signaling reduces cone death suggests that the elevated TR signaling is harmful for cones. In the present work, we also analyzed retinal expression levels of THRB2 by immunoblot analysis in Rpe65^−/−/Nrl^−/− mice that received intravitreal injection of NH-3 or topical administration of NH-3. We observed no significant differences between NH-3- and vehicle-treated mice (data not shown), suggesting that the TR inhibition with treatment approaches used did not affect the receptor expression levels.

The finding that high TH signaling is detrimental has been documented in several pathologic conditions (52, 54, 55). In sensory/neuronal systems, excessive TH signaling induces death of cone photoreceptors (13, 14), auditory defects/cochlear degeneration (51), and cerebellar abnormalities (57). TH signaling has also been implicated in other neurodegenerative conditions, including Alzheimer’s disease (58, 59). More recently, population-based studies showed that high serum T4 values are associated with an increased risk of AMD (60, 61), suggesting a possible role of thyroid hormone signaling in the pathogenesis of AMD. Nevertheless, the finding that the expression levels of TR were upregulated may explain why suppressing TR, anti-thyroid drug treatment, or targeting DIOs protects cones.

Future perspectives and therapeutic significance

In terms of disease management, the eye has advantages over other organs because of its accessibility to topically delivered therapeutics (i.e., eye drops) (26, 62–64). We demonstrated that topical delivery of TR inhibitor protects cones in retinal-degeneration model mice. These findings may imply a novel strategy for retinal degeneration management. The significance and future perspectives include the following: 1) The development of more potent TR antagonists would benefit cone protection. As THRB2 is expressed in cones only in the retina and Thrb2 deletion completely abolished high T3-induced deterioration of cones, the use of THRB2-specific pharmacological blockers may represent a more specific approach for cone protection; and 2) topical administration of TR inhibitor represents a promising approach for cone protection. However, it is important to note that topical treatment effect is determined by drug delivery efficiency and bioavailability, and we still face challenges to effectively deliver therapeutic drugs to the posterior eye via the topical route. Optimizing the treatment regimen, especially the formulations to enhance bioavailability, represents a significant area for exploration. In considering the size of the human eye compared with the experimental animal eye, the translational efforts for eventual patient benefits based on results from experimental animals present even more challenges. Indeed, testing the effective delivery formulation/regimen is as essential as developing effective drugs/receptor blockers. It has been shown that suppressing TH signaling has no negative effect on cone/rod structure and function or survival during development and in adult mice (12, 65, 66). Thus, the ocular delivery of TR inhibitor is perhaps less likely to result in detrimental effects. However, pharmacokinetic studies are needed to understand the effects of local administration of TH inhibitors on other organ systems. We hope that future retinal degeneration patients can benefit from the use of TR inhibitor eye drops.

In summary, we demonstrate that ocular inhibition of TR using a TR antagonist or by deletion of Thrb2 protects cones in retinal degeneration model mice. Our findings provide insights into how antithyroid treatment or targeting DIOs lead to cone protection and suggest that targeting TR locally in the retina represents a novel strategy for cone protection in retinal degeneration management. Indeed, even a small locus of remaining cones can provide useful vision, maintain the visual circuits, and assure the promise of gene therapy.

Glossary

AMD

age-related macular degeneration

CAR

cone arrestin

CNG

cyclic nucleotide-gated

DIO

iodothyronine deiodinase

ERG

electroretinography

FBS

fetal bovine serum

GNAT2

guanine nucleotide-binding protein G(t) subunit α-2

LCA

Leber congenital amaurosis

M-opsin

medium-wave-sensitive opsin

PNA

peanut agglutinin

RPE

retiinal pigment epithlium

RPE65

retinoid isomerase

qRT-PCR

quantitative RT-PCR

T3

triiodothyronine

S-opsin

short-wave-sensitive opsin

TBP

TATA-box-binding protein

TH

thyroid hormone

TR

thyroid hormone receptor

AUTHOR CONTRIBUTIONS

H. Ma performed ocular treatment and cell culture experiments, performed immunofluorescence labeling experiment and data analysis, and conducted TUNEL assays; F. Yang performed ocular treatment and cell culture experiments, performed immunofluorescence labeling experiment and data analysis; conducted TUNEL assays, and performed expression studies by qRT-PCR and immunoblot analysis; M. R. Butler performed ocular treatment experiments; J. Belcher performed animal experiments and immunoblot experiments; T. M. Redmond provided Rpe65^−/− mice; A. T. Placzel and T. S. Scanlan contributed to NH3 synthesis and characterization; H. Ma, F. Yang, and X.-Q. Ding designed the study and wrote the manuscript; and T. M. Redmond and T. S. Scanlan contributed to the study design and wrote the manuscript.