Deficiency of type 2 iodothyronine deiodinase reduces necroptosis activity and oxidative stress responses in retinas of Leber congenital amaurosis model mice

Abstract

Thyroid hormone (TH) signaling has been shown to regulate cone photoreceptor viability. Suppression of TH signaling with antithyroid drug treatment or by targeting iodothyronine deiodinases and TH receptors preserves cones in mouse models of retinal degeneration, including the Leber congenital amaurosis Rpe65-deficient mice. This work investigates the cellular mechanisms underlying how suppressing TH signaling preserves cones in Rpe65-deficient mice, using mice deficient in type 2 iodothyronine deiodinase (Dio2), the enzyme that converts the prohormone thyroxine to the active hormone triiodothyronine (T3). Deficiency of Dio2 improved cone survival and function in Rpe65^−/− and Rpe65 deficiency on a cone dominant background (Rpe65^−/−/Nrl^−/−) mice. Analysis of cell death pathways revealed that receptor-interacting serine/threonine-protein kinase (RIPK)/necroptosis activity was increased in Rpe65^−/−/Nrl^−/− retinas, and Dio2 deficiency reversed the alterations. Cell-stress analysis showed that the cellular oxidative stress responses were increased in Rpe65^−/−/Nrl^−/− retinas, and Dio2 deficiency abolished the elevations. Similarly, antithyroid drug treatment resulted in reduced RIPK/necroptosis activity and oxidative stress responses in Rpe65^−/−/Nrl^−/− retinas. Moreover, treatment with T3 significantly induced RIPK/necroptosis activity and oxidative stress responses in the retina. This work shows that suppression of TH signaling reduces cellular RIPK/necroptosis activity and oxidative stress responses in degenerating retinas, suggesting a mechanism underlying the observed cone preservation.—Yang, F., Ma, H., Butler, M. R., Ding, X.-Q. Deficiency of type 2 iodothyronine deiodinase reduces necroptosis activity and oxidative stress responses in retinas of Leber congenital amaurosis model mice.

Rod and cone photoreceptors degenerate in hereditary retinal diseases, including retinitis pigmentosa and Leber congenital amaurosis (LCA, a pediatric retinal degenerative disease), as well as in age-related macular degeneration (AMD). Inherited retinal degenerative diseases affect ∼1 in 3000 individuals worldwide (1, 2). LCA accounts for blindness in >20% of children attending schools for the blind, whereas AMD is the leading cause of blindness in the elderly. The progressive death of cone photoreceptors ultimately leads to vision impairment and blindness. Mutations in numerous genes are associated with inherited retinal degeneration, including those involved in phototransduction, such as cGMP phosphodiesterase 6 (3) and guanylate cyclase (4), and those involved in the visual cycle, such as retinal pigment epithelium-specific 65 kDa protein (RPE65) (5) and lecithin retinol acyltransferase (6), with mutations in RPE65 gene accounting for about 16% of all LCA cases (7). Specifically, RPE65 mutation/chromophore deficiency leads to early onset, severe, and rapid cone photoreceptor degeneration, followed by rod degeneration and retinal dystrophies in human patients (5, 7) and mouse models (8–10). Thus, preservation of cone and rod photoreceptors becomes critical in patients with LCA/chromophore-deficient blindness. There is currently no cure for retinal degeneration. Nevertheless, degenerating photoreceptors in conditions of high genetic heterogeneity show common cellular-disorder features, including oxidative stress/damage (11–18) and necroptosis/apoptosis (19–25). Those features offer the possibility of targeting common cell survival and death pathways for photoreceptor preservation.

Thyroid hormone (TH) regulates cell proliferation, differentiation, and metabolism. In the retina, TH regulates retinal development and cone opsin expression (26–28) and is associated with cone photoreceptor viability (10, 29–32). Using cone degeneration mouse models, including the LCA model, Rpe65^−/− and Rpe65 deficiency on a cone-dominant background (Rpe65^−/−/Nrl^−/−) mice, we have shown that suppression of TH signaling with antithyroid drug treatment (31) or by targeting the intracellular TH components iodothyronine deiodinases (10) and thyroid hormone receptor (TR) (32) protects cones. In contrast with those findings, stimulation of TH signaling via triiodothyronine (T3) treatment or inhibition of T3 degradation was shown to cause cone death (30, 31). Excessive TH signaling has also been linked to a variety of neurodegenerative diseases, including AMD (33–35) and Alzheimer’s disease (36, 37).

Intracellular TH homeostasis/signaling is a highly regulated process locally, controlled by the iodothyronine deiodinases and TR (38, 39). In mammals, the thyroid gland predominantly produces the prohormone thyroxine (T4; ∼95%) as well as a small amount of the bioactive hormone T3 (∼5%). The cellular metabolism of T4 and T3 is catalyzed by 2 iodothyronine deiodinases: type 2 iodothyronine deiodinase (Dio2) and type 3 iodothyronine deiodinase (Dio3). T4 is converted to T3 by Dio2, and T3 then binds to TRs, initiating downstream gene expression responses (38). Intracellular T4 and T3 are degraded by Dio3. We have shown (10) that treatment with Dio2 inhibitor or overexpression of Dio3 preserves cones in Rpe65^−/− and Rpe65^−/−/Nrl^−/− mice. Using Dio2 deletion and antithyroid treatment, this work investigated the potential mechanisms underlying how suppression of TH signaling preserves cones in Rpe65-deficient mice. We found that receptor-interacting serine/threonine-protein kinase (RIPK)/necroptosis activity and oxidative stress responses were significantly elevated in Rpe65^−/−/Nrl^−/− retinas, that Dio2 deficiency or antithyroid treatment reversed those alterations, and that treatment with T3 significantly induced RIPK/necroptosis activity and oxidative stress responses in the retina. Our findings suggest that TH signaling suppression-induced cone protection involves the inhibition of RIPK/necroptosis signaling activity and oxidative stress responses in the retina.

MATERIALS AND METHODS

Mice, antibodies, and reagents

The Rpe65^−/− mouse line was provided by T. Michael Redmond [National Institutes of Health/U.S. National Eye Institute (NIH/NEI), Bethesda, MD, USA] (8). The Dio2^−/− mouse line was purchased from The Jackson Laboratory (Bar Harbor, ME, USA) (40). The Nrl^−/− mouse line was provided by Dr. Anand Swaroop (NIH/NEI) (41). Rpe65^−/−, Nrl^−/−, and Dio2^−/− mouse lines were generated on a C57BL/6J background. Rpe65^−/−/Dio2^+/−, Rpe65^−/−/Dio2^−/−, Rpe65^−/−/Nrl^−/−, Rpe65^−/−/Nrl^−/−/Dio2^+/−, and Rpe65^−/−/Nrl^−/−/Dio2^−/− mouse lines were generated by crossmating and on the same genetic background. Littermates were used for experiments. All mice were maintained under cyclic light (12-h light/dark) conditions. In the light cycle, cage illumination was 7 foot-candles (75.35 lux). 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 (Washington, DC, USA) and the Association for Research in Vision and Ophthalmology (Rockville, MD, USA).

The primary antibody information is listed in Table 1. Biotinylated peanut agglutinin (PNA) was purchased from Vector Laboratories (Burlingame, CA, USA). Fluorescent goat anti-rabbit antibody and streptavidin-cyanine 3 (Cy3) were purchased from Thermo Fisher Scientific (Waltham, MA, USA). Horseradish peroxidase–conjugated anti-rabbit or anti-mouse secondary antibodies were purchased from SeraCare (Milford, MA, USA). All other reagents were purchased from MilliporeSigma (Burlington, MA, USA), Bio-Rad Laboratories (Hercules, CA, USA), and Thermo Fisher Scientific.

TABLE 1.

List of primary antibodies

Antibody	Provider	Catalog no.	Dilutions used in IB
M-opsin	MilliporeSigma	AB5405	1:500
S-opsin	Dr. Muna Naash, University of Houston, Houston, TX, USA		1:500
CAR	Dr. Cheryl Craft, University of Southern California, Los Angeles, CA, USA		1:500
Gnat2	Santa Cruz Biotechnology, Dallas, TX, USA	sc-390	1:500
Ripk3	Santa Cruz Biotechnology	sc-135170	1:500
Pepck-M	Santa Cruz Biotechnology	sc-271204	1:500
Gpx4	Cayman Chemicals, Ann Arbor, MI, USA	10005258	1:500
β-actin	Abcam, Cambridge, United Kingdom	ab6276	1:2000

Drug treatment

For antithyroid drug treatment, mother Rpe65^−/−/Nrl^−/− mice were treated with an antithyroid drug [methimazole (0.05% w/v) and sodium perchlorate monohydrate (1.0% w/v)] in drinking water, beginning on the day they delivered pups, and the treatment continued for 15 d. At the end of the treatment, retinas of the pups were collected for biochemical evaluations. For T3 treatment, Nrl^−/− mice received T3 (0.75 μg/g body weight/d, s.c. injection) from postnatal day (P)10 to P15 or from P21 to P32. At the end of the treatment, retinas were collected for biochemical evaluations.

Eye preparation, immunofluorescence labeling, and confocal microscopy

Mouse retinal whole mounts or cross sections were prepared, and immunofluorescence labeling with antibodies against cone markers was performed (10). For whole mount preparations, eyes were enucleated, marked at the superior (dorsal) pole with a green dye for dorsal–ventral orientation, and fixed in 4% paraformaldehyde (Polysciences, Inc., Warrington, PA, USA) for 1 h at room temperature. The cornea and lens were then removed, and the retinas were fixed in 4% paraformaldehyde for another 1 h. For retinal cross sections, mouse eyes were enucleated (the superior portion of the cornea was marked for orientation before enucleation) and fixed in Prefer (Anatech Ltd., Battle Creek, MI, USA) for 25 min at room temperature, and fixed eyes were then stored in 70% ethanol until processing for sections. Paraffin-embedded sections (5 µm thick) passing vertically through the retina (along the vertical meridian passing through the optic nerve head, to allow an examination of the retina in the dorsal and ventral hemisphere) were prepared with a microtome (Leica Camera, Wetzlar, Germany). PNA immunohistochemistry was performed with biotinylated PNA (1:250) and streptavidin-Cy3 (1:500). Fluorescent signals were imaged with an FV1000 confocal laser scanning microscope (Olympus, Tokyo, Japan) with FluoView imaging software (Olympus). The averages of the counts were analyzed and graphed with Prism software (GraphPad Software, La Jolla, CA, USA).

Scotopic and photopic electroretinogram recordings

Full-field electroretinogram (ERG) recordings were performed as previously described by Xu et al. (42). Briefly, after overnight dark adaptation, mice were anesthetized with an i.p. injection of 80 mg/kg ketamine and 16 mg/kg xylazine. ERGs were recorded with an Espion Visual Electrophysiology System with a Ganzfeld ColorDome system (Diagnosys, Lowell, MA, USA). Potentials were recorded using a gold-wire electrode to contact the corneal surface through a layer of 2.5% hypromellose (Gonak; Akorn, Decatur, IL, USA). For assessment of scotopic responses, a stimulus intensity of 157 lux was presented to dark-adapted, dilated mouse eyes. To evaluate photopic responses, mice were adapted to a 30 lux light for 5 min, and then a light intensity of 77 lux was given. Responses were differentially amplified, averaged, and analyzed with Espion 100 software (Diagnosys).

TUNEL assay

TUNEL was performed to evaluate photoreceptor apoptotic death, as previously described by Ma et al. (43). Paraffin-embedded retinal cross sections and the In Situ Cell Death Fluorescein Detection kit (MilliporeSigma) were used. Immunohistochemical images were taken with an FV1000 confocal laser scanning microscope (Olympus). The total TUNEL-positive cells in the outer nuclear layer were counted and averaged from ≥3 sections/eye. Data were analyzed and graphed with Prism software.

PCR array and quantitative RT-PCR

Total RNA preparation and reverse transcription was performed as previously described by Ma et al. (44). The Mouse Oxidative Stress and Antioxidant Defense RT² Profiler PCR Array (Qiagen, Hilden, Germany), which profiles the expression of 84 genes related to oxidative stress, was used following the manufacturer’s instructions. The quantitative RT-PCR (qRT-PCR) assays were performed with a real-time PCR detection system (iCycler; Bio-Rad Laboratories). Mouse hypoxanthine guanine phosphoribosyl transferase 1 (Hprt1) was included as an internal control. Table 2 shows the primers used. The relative gene expression value was calculated based on the cycle threshold ∆∆C_t method (44).

TABLE 2.

List of primers used for qRT-PCR

	Primer, 5′–3′
Gene	Forward	Reverse
Hprt1	GCAAACTTTGCTTTCCCTGGTT	CAAGGGCATATCCAACAACA
Casp3	GACTGATGAGGAGATGGCTTG	TGCAAAGGGACTGGATGAAC
Casp7	CCCACTTATCTGTACCGCATG	GGTTTTGGAAGCACTTGAAGAG
Casp8	AACTTCCTAGACTGCAACCG	TCTCAATTCCAACTCGCTCAC
Ripk1	GGAAGGATAATCGTGGAGGC	AAGGAAGCCACACCAAGATC
Ripk3	TCTTTACTGAGACTCCCGGT	AGTTCCCAATCTGCACTTCAG
Gpx4	GCAATGAGGCAAAACTGACG	CTTGATTACTTCCTGGCTCCTG
Nox4	TCCAAGCTCATTTCCCACAG	CGGAGTTCCATTACATCAGAGG
Ucp2	GCATTGGCCTCTACGACTC	AAGCGGACCTTTACCACATC
Gss	GATCCTGTCCAATAACCCCAG	GCACGCTGGTCAAATATGTTC
Ctsb	AGACCTGCTTACTTGCTGTG	GGAGGGATGGTGTATGGTAAG
Ncf1	TCATCCTTCAGACCTATCGGG	ACCTCGCTTTGTCTTCATCTG
Ehd2	AGCTCAACGACCTAGTGAAAC	TCGCAAAGATGACAGGCAG
Il-22	AGCTTGAGGTGTCCAACTTC	GGTAGCACTGATCTTTAGCACTG
Tnf-α	AAAATTCGAGTGACAAGCCTGTAG	CCCTTGAAGAGAACCTGGGAGTAG
Tnfrsf1a	GCCGGATATGGGCATGAAGC	TGTCTCAGCCCTCACTTGAC
Tnfrsf9	CCTGTGATAACTGTCAGCCTG	TCTTGAACCTGAAATAGCCTGC
Mlkl	ACTGTGAACTTGGAACCCTG	TGCTGATGTTTCTGTGGAGTG
Tradd	ACGAACTCACTAGTCTAGCAGAG	AATACCCCAACAGCCACC
Dio2	ACTCGGTCATTCTGCTCAAG	ACACTGGAATTGGGAGCATC
Dio3	GTGGTCGGAGAAGGTGAAG	TGCACAAGAAATCTAAAAGCCAG
Thrb2	AGTCAGTCCAGCCAGCCTGCACAT	GCTTCCGCTTGGCTAGCCTCTTGCT

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

Retinal protein preparation, SDS-PAGE, and Western blot analysis were performed as previously described by Ma et al. (43). Briefly, retinas were homogenized in homogenization buffer A [0.32 M sucrose, 20 mM 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid (HEPES), pH 7.4, and 3 mM EDTA containing protease and phosphatase inhibitors] (MilliporeSigma), and homogenates were centrifuged at 3000 rpm for 10 min at 4°C. The resulting supernatant was then centrifuged at 13,000 rpm for 35 min at 4°C to separate cytosolic (supernatant) and membrane (pellet) fractions. The resulting membrane pellet was resuspended in homogenization buffer B [0.32 M sucrose, 20 mM 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid (HEPES), pH 7.4, 3 mM EDTA, and 0.1% Triton X-100 containing protease and phosphatase inhibitors, as previously described], sonicated twice for 15 s on ice at a medium speed using an XL2000 Ultrasonic Cell Disruptor (Misonix, Farmingdale, NY, USA), with a 30-s recovery between disruptions, and incubated for 1 h at 4°C with gentle agitation. After incubation, the homogenate was centrifuged at 13,000 rpm for 35 min at 4°C, and the resulting supernatant was used as the membrane fraction. All protein concentrations were determined by a protein-assay kit from Bio-Rad Laboratories. Retinal protein samples were then subjected to SDS-PAGE and transferred to PVDF membranes, which were subsequently blocked in 5% nonfat milk for 1 h at room temperature. Immunoblots were incubated with primary antibody overnight at 4°C (Table 1 for antibody information). After washing in Tris-buffered saline with 0.1% Tween 20, the immunoblots were incubated with horseradish peroxidase–conjugated secondary antibody (1:20,000) for 1 h at room temperature. Chemiluminescent substrate (Thermo Fisher Scientific) was used to detect primary antibodies binding to respective cognate antigens. An Odyssey CLx Imager (Li-Cor Biosciences, Lincoln, NE, USA) and Li-Cor Biosciences software were used for detection and densitometric analysis.

Statistical analysis

Results are expressed as means ± sem of the 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 when P < 0.05. Data were analyzed and graphed with Prism software.

RESULTS

Dio2 deficiency improved cone survival and function in Rpe65-deficient mice

We previously showed that suppression of TH signaling, including inhibition of Dio2 using a Dio2 inhibitor, reduced cone death in Rpe65^−/− and Rpe65^−/−/Nrl^−/− mice. Rpe65^−/−/Nrl^−/− mice show a cone defect phenotype similar to that in Rpe65^−/− mice and have been used to study LCA cone defects (45–47). To investigate how suppressing TH signaling protects cones, we generated Rpe65-knockout mice with a Dio2 deficiency and used Rpe65^−/−/Nrl^−/−/Dio2^+/− and Rpe65^−/−/Nrl^−/−/Dio2^−/− mice as models for analysis of biochemical events in a cone-dominant retina (because cones comprise only 2–3% of the total photoreceptor population in the wild-type mouse retina). Dio2-deficient mice have been used to evaluate the role of thyroid hormone signaling/Dio2 in development and function of various tissues/organs (48–50). Specifically, Dio2-deficient mice have been used in cochlear development studies. Dio2 activity was measured in these mice, showing that the activity was undetectable in Dio2^−/− mice and was reduced by about 50% in Dio2^+/− mice (49). We first evaluated the effects of Dio2 deletion on retinal phenotype by examining cone density, retinal function, and cone death. In agreement with observations obtained from Dio2 inhibitor treatments, Dio2 deficiency significantly improved cone survival in Rpe65^−/− mice. As evaluated by PNA labeling on retinal whole mounts, cone density in the more severely degenerated area (ventral region) was increased by 100 and 40% in P30 Rpe65^−/−/Dio2^+/− and Rpe65^−/−/Dio2^−/− mice, respectively, compared with age-matched Rpe65^−/− mice (Fig. 1A). Cone preservation was also evaluated by examining the expression level of the cone-specific protein cone arrestin (CAR). Expression levels of CAR were increased by about 1-fold in P30 Rpe65^−/−/Dio2^+/− and Rpe65^−/−/Dio2^−/− mice, compared with age-matched Rpe65^−/− mice (Fig. 1B). ERG recordings were used to evaluate retinal function, and showed that the photopic light response was significantly improved in Rpe65^−/− mice with Dio2 heterozygous deficiency. Photopic b-wave response in Rpe65^−/−/Dio2^+/− mice was restored to the wild-type level (Fig. 1C). Interestingly, Rpe65^−/−/Dio2^−/− mice did not show improved photopic response. No significant improvement was observed in scotopic light responses. TUNEL labeling was used to evaluate cone death and revealed a significant reduction of cone death in the Dio2-deficient mice. The numbers of TUNEL-positive cells were reduced by 30 and 50% in P15 Rpe65^−/−/Nrl^−/−/Dio2^+/− and Rpe65^−/−/Nrl^−/−/Dio2^−/− mice, respectively, compared with age-matched Rpe65^−/−/Nrl^−/− controls (Fig. 2A). We also examined expression levels of cone proteins in these mice. Expression levels of S-opsin, M-opsin, CAR, and cone transducin α-subunit [guanine nucleotide-binding protein G(t) subunit α-2 (Gnat2)] were increased by 40–60% in P30 Rpe65^−/−/Nrl^−/−/Dio2^+/− mice, compared with age-matched controls (Fig. 2B). Thus, similar to pharmacologic inhibition of Dio2, Dio2 deficiency preserved cones in Rpe65^−/− and Rpe65^−/−/Nrl^−/− mice.

Figure 1.

Dio2 deficiency improved cone survival and function in Rpe65^−/− mice. Cone survival and function were evaluated in P30 Rpe65^−/− and Rpe65^−/− mice with Dio2 deficiency. A) Dio2 deficiency increased cone density in Rpe65^−/− mice. Shown are representative confocal images of immunofluorescence labeling of PNA on retinal whole mounts. B) Dio2 deficiency increased expression of CAR in Rpe65^−/− mice. Shown are representative immunoblotting images of CAR with corresponding quantitative analysis. C) Dio2 heterozygous deficiency improved retinal function in Rpe65^−/− mice. Full-field ERG recordings were performed. For assessment of scotopic responses, a stimulus intensity of 157 lux was presented to dark-adapted, dilated mouse eyes. To evaluate photopic responses, mice were adapted to a 30 lux light for 5 min, and then a light intensity of 77 lux was given. Shown are the a/b wave amplitude quantification of scotopic and photopic ERG responses. Data are represented as means ± sem (n = 9–16 mice). *P < 0.05, **P < 0.01, ***P < 0.001.

Figure 2.

Dio2 deficiency preserved cones in Rpe65^−/−/Nrl^−/− mice. Cone survival/death was evaluated in P15 and P30 Rpe65^−/−/Nrl^−/− mice and Rpe65^−/−/Nrl^−/− mice with Dio2 deficiency. A) Dio2 deficiency reduced cone death in P15 Rpe65^−/−/Nrl^−/− mice. Shown are representative confocal images of TUNEL-positive cells in retinal cross sections and corresponding quantitative analysis. ONL, outer nuclear layer; INL, inner nuclear layer. B) Dio2 heterozygous deficiency increased expression levels of cone-specific proteins in P30 Rpe65^−/−/Nrl^−/− mice. Shown are representative immunoblotting images of S-opsin, M-opsin, Gnat2, and CAR in the retinas, with corresponding quantitative analysis. Data are represented as means ± sem (n = 8–11 mice). *P < 0.05, **P < 0.01.

Increased expression of the RIPK/necroptosis genes in Rpe65^−/−/Nrl^−/− retinas and abolition of elevation in Rpe65^−/−/Nrl^−/− mice with Dio2 heterozygous deficiency or after antithyroid treatment

We then evaluated cell death pathway activities in Rpe65^−/−/Nrl^−/− mice with Dio2 deficiency. Retinal expression of the genes involved in cellular necroptosis and apoptosis were analyzed by qRT-PCR and immunoblotting. The necroptosis genes examined included Tnf-α (encoding the TNF-α), Tnfra1 (encoding the TNF receptor α1), Tradd (encoding the TNF receptor-associated DEATH domain protein), Ripk1 (encoding receptor-interacting serine/threonine-protein kinase 1), Ripk3 (encoding receptor-interacting serine/threonine-protein kinase 3), and Mlkl (encoding the mixed-lineage kinase domain–like pseudokinase). We found that the mRNA levels of Ripk3, Tnf-α, Tnfra1, Mlkl, and Tradd were increased by 3–5-fold in P15 Rpe65^−/−/Nrl^−/− mice, compared with age-matched Nrl^−/− controls, and the elevated gene expressions were completely abolished in Rpe65^−/−/Nrl^−/−/Dio2^+/− mice (Fig. 3A). In agreement with qRT-PCR results, the protein level of Ripk3 was significantly elevated in Rpe65^−/−/Nrl^−/− mice, and Dio2 heterozygous deficiency abolished that elevation (Fig. 3B). In contrast, the mRNA levels of the canonical apoptosis genes Casp3 (encoding caspase-3), Casp7 (encoding caspase-7), and Casp8 (encoding caspase-8) were not altered in P15 Rpe65^−/−/Nrl^−/− mice, relative to age-matched Nrl^−/− controls (Fig. 3C). The effects of TH signaling suppression were also examined in Rpe65^−/−/Nrl^−/− mice after antithyroid treatment. Similar to findings from Dio2 heterozygous deficiency experiments, the increased expression levels of Ripk3 mRNA and protein in Rpe65^−/−/Nrl^−/− mice were abolished after antithyroid treatment (Fig. 4A, B). Thus, Rpe65-deficient retinas showed increased Ripk3/necroptosis signaling, and the heterozygous deficiency of Dio2 or antithyroid treatment reversed the elevations.

Figure 3.

Increased expression of the Ripk3/necroptosis genes in Rpe65^−/−/Nrl^−/− retinas and decreased expression in Rpe65^−/−/Nrl^−/− mice with Dio2 heterozygous deficiency. Expressions of the necroptosis genes were analyzed in retinas of P15 Rpe65^−/−/Nrl^−/− mice and Rpe65^−/−/Nrl^−/− mice with Dio2 heterozygous deficiency. A) qRT-PCR results showing increased expression of the necroptosis genes in Rpe65^−/−/Nrl^−/− retinas, and decreased expression in Dio2 heterozygous deficiency. B) Increased expression level of Ripk3 protein in Rpe65^−/−/Nrl^−/− retinas and decreased expression in Dio2 heterozygous deficiency are shown as representative immunoblotting images of Ripk3 with corresponding quantitative analysis. C) qRT-PCR results showing no expression alterations of the canonical apoptosis genes examined in retinas of Rpe65^−/−/Nrl^−/− mice. Data are represented as means ± sem (n = 8–10 mice). *P < 0.05, **P < 0.01.

Figure 4.

Antithyroid treatment reduced expression of Ripk3 in Rpe65^−/−/Nrl^−/− retinas. Expressions of the necroptosis genes were analyzed in retinas of P15 Rpe65^−/−/Nrl^−/− mice and in those mice after antithyroid treatment. A) qRT-PCR results showing increased expression of Ripk3 in Rpe65^−/−/Nrl^−/− mice, and decreased expression after antithyroid treatment. B) Increased expression level of Ripk3 protein in Rpe65^−/−/Nrl^−/− mice and decreased expression after antithyroid treatment are shown as representative immunoblotting images of Ripk3 with corresponding quantitative analysis. Data are represented as means ± sem (n = 8–10 mice). *P < 0.05, **P < 0.01.

Treatment with T3 increased expression of the RIPK/necroptosis genes in Nrl^−/− retinas

The regulation of TH signaling on cell death pathway activities was also evaluated in mice after TH signaling stimulation. P10 Nrl^−/− mice received T3 treatment for 5 d and were then analyzed for gene expression by qRT-PCR. We found that treatment with T3 significantly increased expression of necroptosis genes in the mouse retinas. The expression levels of Ripk3, Tnfrsf9 (encoding the TNF receptor superfamily member 9), Tnf-α, Tnfr1, and Mlkl were increased by about 10–15-fold, compared with vehicle-treated controls (Fig. 5A). Those results suggested that T3 treatment exerts a strong effect on necroptosis signaling in the retina. In contrast, the apoptosis genes Casp-3, Casp-7, and Casp-8 were not significantly altered after T3 treatment (Fig. 5B). As expected, treatment with T3 induced expression of Dio3, which may reflect a cellular compensatory mechanism, but the expression levels of Dio2 and Thrb2 (encoding the B2 type of TR) were not altered (Fig. 5C).

Figure 5.

Treatment with T3 increased expression of the necroptosis genes in Nrl^−/− retinas. Expressions of the necroptosis genes, apoptosis genes, and TH component genes were analyzed in retinas of P15 Nrl^−/− mice and in those mice after T3 treatment. qRT-PCR results showing expression levels of the necroptosis genes (A), apoptosis genes (B), and TH component genes (C) in retinas of Nrl^−/− mice after T3 treatment, compared with vehicle-treated controls. Veh, vehicle. Data are represented as means ± sem (n = 6–8 mice). *P < 0.05, **P < 0.01.

Increased expression of the oxidative stress response genes in Rpe65^−/−/Nrl^−/− retinas and abolition of elevation in Rpe65^−/−/Nrl^−/− mice with Dio2 heterozygous deficiency or after antithyroid treatment

Because TH signaling regulates cellular oxidative stress responses/mitochondrial functions (51), which have a critical role in photoreceptor death and retinal degeneration (52, 53), we examined whether oxidative stress responses were involved in TH-signaling suppression-induced cone protection. We first examined cellular oxidative stress responses in Rpe65^−/−/Nrl^−/− retinas using The Mouse Oxidative Stress and Antioxidant Defense RT² Profiler PCR Array. The array profiles expression of 84 genes related to oxidative stress, including peroxidases, oxidative stress responsive genes, and genes involved in superoxide metabolism. We found that 28 of 84 genes were up-regulated in Rpe65^−/−/Nrl^−/− mice, compared with age-matched Nrl^−/− mice (data not shown). Among the 28 genes, 8 were chosen as representatives for further expression analysis by qRT-PCR and immunoblotting. The 8 representative genes were Gpx4 (encoding glutathione peroxidase 4), Nox4 (encoding NADPH oxidase 4), Ucp2 (encoding uncoupling protein 2), Gss (encoding glutathione synthetase), Ctsb (encoding cathepsin B), Ncf1 (neutrophil cytosolic factor 1), Ehd2 (encoding EH-domain containing 2), and Il22 (encoding IL-22). The expression levels of Gpx4, Ucp2, Gss, Ctsb, Ncf1, and Ehd2 were 2–5-fold greater in Rpe65^−/−/Nrl^−/− mice than in Nrl^−/− controls, and those elevations were completely reversed in Rpe65^−/−/Nrl^−/−/Dio2^+/− mice (Fig. 6A). Similar results were obtained in Rpe65^−/−/Nrl^−/−/Dio2^−/− mice (data not shown). Consistent with qRT-PCR results, the protein level of Gpx4 was significantly elevated in Rpe65^−/−/Nrl^−/− mice, and Dio2 heterozygous deficiency abolished the alteration (Fig. 6B). We also examined the effects of antithyroid treatment. Similar to findings from Dio2 heterozygous deficient mice, antithyroid treatment significantly reduced Gpx4 expression in Rpe65^−/−/Nrl^−/− mice (Fig. 6C). Hence, Rpe65-deficient retinas showed elevated cellular oxidative stress responses, and heterozygous deficiency of Dio2 or antithyroid treatment reversed the elevations.

Figure 6.

Increased expression of the oxidative stress genes in Rpe65^−/−/Nrl^−/− retinas and decreased expression in Rpe65^−/−/Nrl^−/− mice with Dio2 heterozygous deficiency or after antithyroid treatment. Expressions of the oxidative stress genes were analyzed in retinas of P15 Rpe65^−/−/Nrl^−/− mice, Rpe65^−/−/Nrl^−/− mice with Dio2 heterozygous deficiency, and Rpe65^−/−/Nrl^−/− mice after antithyroid treatment. A) qRT-PCR results showing increased mRNA levels of the oxidative stress responsive genes in Rpe65^−/−/Nrl^−/− mice, and Dio2 heterozygous deficiency reversed the elevations. B) Increased Gpx4 protein level in Rpe65^−/−/Nrl^−/− mice, and Dio2 heterozygous deficiency reversed the elevation. Shown are representative immunoblotting images of Gpx4 with corresponding quantitative analysis. C) qRT-PCR results showing increased expression of Gpx4 and Ucp2 in Rpe65^−/−/Nrl^−/− retinas and decreased expression after antithyroid treatment. Data are represented as means ± sem (n = 8–10 mice). *P < 0.05, **P < 0.01.

Treatment with T3 increased expression of the oxidative stress response genes in Nrl^−/− retinas

The regulation of TH signaling on cellular oxidative stress responses was also evaluated in mice after TH signaling stimulation. P10 Nrl^−/− mice received T3 treatment for 5 d and were then analyzed for expression of the oxidative stress genes using the Mouse Oxidative Stress and Antioxidant Defense RT² Profiler PCR Array. We found that T3 treatment significantly increased expression levels of 36 genes (40% of the genes examined), including Gpx, Gpx6, Gpx5, Gpx4, Gpx2, and Gpx1, Gss, Nox4, and Ucp2 in the mouse retina, compared with vehicle-treated controls. Table 3 shows genes with change >2.0-fold. In a separate experiment, we examined the effects of T3 treatment in adult mice. P21 Nrl^−/− mice received T3 treatment for 10 d and were then analyzed for gene expression. The PCR array showed that treatment with T3 significantly increased expression levels of 29 genes (35% of the genes examined) in the mouse retina, compared with vehicle-treated controls (Table 3).

TABLE 3.

RT²-PCR array results showing increased expression of the oxidative stress response genes in retinas of Nrl^−/− mice after T3 treatment (genes with fold change >2)

Description	Gene	GenBank accession no.	Fold change Nrl−/−+T3 (P10-P15)	Fold change Nrl−/−+T3 (P20-P32)
Recombination activating gene 2	Rag2	NM_009020	4.49	4
Cathepsin B	Ctsb	NM_007798	4.49	6.06
IL-22	Il-22	NM_016971	3.85	4.29
Glutathione peroxidase 6	Gpx6	NM_145451	3.7	4.92
Neutrophil cytosolic factor 1	Ncf1	NM_010876	3.41	4.59
Glutathione peroxidase 5	Gpx5	NM_010343	4.25	6.5
Heme oxygenase (decycling) 1	Hmox1	NM_010442	3.59	4.92
Intraflagellar transport 172 homolog (Chlamydomonas)	Ift172	NM_026298	3.7	4.59
Glutathione synthetase	Gss	NM_008180	3.99	5.23
Excision repair cross-complementing rodent repair deficiency, complementation group 2	Ercc2	NM_007949	4.02	5.28
Heat shock protein 1A	Hspa1a	NM_010479	3.7	4.29
EH-domain containing 2	Ehd2	NM_153068	3.7	4.29
Uncoupling protein 2 (mitochondrial, proton carrier)	Ucp2	NM_011671	3.7	4.59
RecQ protein-like 4	Recql4	NM_058214	3.26	3.73
Prostaglandin-endoperoxide synthetase 2	Ptgs2	NM_011198	2.84	3.73
Glutathione peroxidase 1	Gpx1	NM_008160	3.45	3.48
Neuroglobin	Ngb	NM_022414	3.31	4.59
Thioredoxin 1	Txn1	NM_011660	3.18	2.3
Glutathione peroxidase 4	Gpx4	NM_008162	5.16	2.64
Neutrophil cytosolic factor 2	Ncf2	NM_010877	2.89	3.48
Ferritin heavy chain 1	Fth1	NM_010239	2.51	4
NADPH oxidase 4	Nox4	NM_015760	2.89	3.73
Lactoperoxidase	Lpo	NM_080420	2.59	2.64
Serine (or cysteine) peptidase inhibitor, clade B, member 1b	Serpinb1b	NM_173052	2.92	3.48
Xeroderma pigmentosum, complementation group A	Xpa	NM_011728	2.61	2.46
Myeloperoxidase	Mpo	NM_010824	2.12	3.48
Amyotrophic lateral sclerosis 2 (juvenile) homolog (human)	Als2	NM_028717	3	2.14
Eosinophil peroxidase	Epx	NM_007946	2.33	3.48
Prostaglandin-endoperoxide synthetase 1	Ptgs1	NM_008969	2.42	2.3
Parkinson disease (autosomal recessive, early onset 7)	Park7	NM_020569	2.12
Stearoyl-coenzyme A desaturase 1	Scd1	NM_009127	2.25
Glutathione peroxidase 2	Gpx2	NM_030677	2.58
Sequestosome 1	Sqstm1	NM_011018	2.41

GenBank, National Center for Biotechnology Information, Bethesda, MD, USA; https://www.ncbi.nlm.nih.gov/genbank/)

Increased expression of the mitochondrial metabolic enzymes in Rpe65^−/−/Nrl^−/− retinas and abolition of elevation in Rpe65^−/−/Nrl^−/− mice with Dio2 heterozygous deficiency or after antithyroid treatment

TH has a profound effect on mitochondrial function and metabolic homeostasis. We examined effects of Dio2 heterozygous deficiency on the expression of the mitochondrial metabolic enzymes. Phosphoenolpyruvate carboxykinase (Pepck) and pyruvate dehydrogenase kinase 4 (Pdk4) have important roles in mitochondrial lipid and glucose metabolism, and expression of those 2 enzymes are transcriptionally regulated by TH (54–59). We found that the expression levels of Pdk4 and Pck2 were increased by about 5- and 15-fold in Rpe65^−/−/Nrl^−/− retinas, respectively, relative to the control Nrl^−/− retinas. Heterozygous deficiency of Dio2 completely reversed expression elevation of those 2 genes (Fig. 7A). Consistent with qRT-PCR results, the Pepck-M protein level was about 30-fold higher in Rpe65^−/−/Nrl^−/− retinas than in control Nrl^−/− retinas, and Dio2 heterozygous deficiency significantly reduced the expression (Fig. 7B). Similarly, antithyroid treatment also greatly reduced expression level of Pepck-M in Rpe65^−/−/Nrl^−/− retinas (Fig. 7B). Hence, the mitochondrial metabolic enzymes were dramatically up-regulated in Rpe65^−/−/Nrl^−/− retinas, and Dio2 heterozygous deficiency or antithyroid treatment suppressed those elevations. As expected, the mRNA levels of Pdk4 and Pck2 were greatly increased in Nrl^−/− retinas after T3 treatment (Fig. 7C).

Figure 7.

Increased expression of the mitochondrial metabolic enzymes in Rpe65^−/−/Nrl^−/− retinas and decreased expression in Rpe65^−/−/Nrl^−/− mice with Dio2 heterozygous deficiency or after antithyroid treatment. Expressions of the mitochondrial metabolic enzymes Pepck and Pdk4 were analyzed in retinas of P15 Rpe65^−/−/Nrl^−/− mice, Rpe65^−/−/Nrl^−/− mice with Dio2 heterozygous deficiency, or Rpe65^−/−/Nrl^−/− mice after antithyroid treatment. A) qRT-PCR results showing increased expression of Pdk4 and Pck2 in Rpe65^−/−/Nrl^−/− retinas, and Dio2 heterozygous deficiency reversed the increases. B) Increased Pepck-M protein level in Rpe65^−/−/Nrl^−/− mice, and Dio2 heterozygous deficiency or antithyroid treatment suppressed the elevation. Shown are representative Western blotting images of Pepck-M with corresponding quantitative analysis. C) Increased mRNA levels of Pdk4 and Pck2 in Nrl^−/− retinas after T3 treatment. Veh, vehicle. Data are represented as means ± sem (n = 8–10 mice). *P < 0.05, **P < 0.01, ***P < 0.001.

DISCUSSION

We previously showed that antithyroid treatment or targeting iodothyronine deiodinases/TR protects cones in Rpe65^−/− and Rpe65^−/−/Nrl^−/− mice. This work investigated how suppression of TH signaling preserves cones. We found that the cellular RIPK/necroptosis activity and oxidative stress responses were increased in Rpe65-deficient retinas, and Dio2 heterozygous deficiency or antithyroid treatment reversed those elevations. Consistent with findings from Rpe65^−/−/Nrl^−/− mice, treatment with T3 significantly induced expression of the RIPK/necroptosis genes and oxidative stress response genes in Nrl^−/− retinas. Hence, suppression of TH signaling-induced cone protection might be associated with inhibition of the cellular necroptosis activity and oxidative stress responses in the diseased retinas.

Dio2 deficiency improves cone survival and function in Rpe65-deficient mice

Dio2 inhibitor treatment has been shown to preserve cones in Rpe65-deficient mice (10). This work demonstrated the effects of Dio2 deficiency. Cone density was significantly increased in Rpe65^−/−/Dio2^+/− and Rpe65^−/−/Dio2^−/− mice, relative to Rpe65^−/− controls. Dio2 deficiency reduced cone death in Rpe65^−/−/Nrl^−/− mice, accompanied by increased expression of cone specific proteins. Along with improved cone survival, Rpe65^−/−/Dio2^+/− mice showed greatly improved cone responses to light. Together with findings from pharmacologic inhibition of Dio2, the findings from Dio2-deficient mice demonstrate that Dio2 inhibition preserves cones in retinal degeneration, and support the view that Dio2 inhibition locally in the retinas may represent a therapeutic strategy for cone protection. The beneficial effects of Dio2 deficiency have also been shown in other models of photoreceptor degeneration. For example, Dio2 inhibitor treatment has been shown to accelerate reinnervation of the optic tectum after optic nerve crush in zebrafish (60). Recently, deletion of Dio2 was shown to rescue cones in Dio3^−/− mice (29). Deletion of Dio3 induces elevation of cellular T3 level and subsequent cone death (30), and that effect is reversed by deletion of TR (30).

Similar to the effects from other strategies used to inhibit TH signaling, including antithyroid treatment or overexpression of Dio3, the rescue achieved by deficiency of Dio2 is partial. This partial rescue might be associated with: 1) the requirement to target multiple TH components, 2) the potential involvement/compensatory responses of nongenomic actions of TH, and 3) the contributions from non-TH factors. Although TH signaling has a role in cell death, it is not the sole factor. Other non-TH factors could contribute to disease progression. In addition, the initial pathologic conditions will not be corrected by suppression of TH signaling. Critical among those is the requirement of cone photoreceptors for the 11-cis-retinal chromophore absent in the Rpe65-deficient mice (8).

Rpe65^−/−/Dio2^+/−, but not Rpe65^−/−/Dio2^−/−, mice showed significant improvement in cone ERG responses. That observation might be associated with the essential role of Dio2 in retinal development and function. Dio2^−/− mice have been shown to have impaired photoreceptor function. Sawant et al. (61) reported reduced rod and cone ERG responses in Dio2^−/− mice. Ng et al. (29) showed reduced rod ERG response in Dio2^−/− mice. We observed a similar phenotype showing impaired cone and rod ERG responses in Dio2^−/−, but not Dio2^+/−, mice (data not shown). Moreover, knockdown of Dio2 has been shown to affect zebrafish eye development at the level of gene expression, morphology, and function (62). For those reasons, we used retinas of Rpe65^−/−/Nrl^−/−/Dio2^+/− mice in our biochemical evaluations. We think TH signaling is likely decreased in the retinas of mice with Dio2 heterozygous deficiency because of reduced Dio2 activity [Dio2 activity was reduced by about 50% in Dio2^+/− mice (49)]. This presumption is also supported by our findings that showed reduced expression of the TH receptor-target genes in Rpe65^−/−/Nrl^−/−/Dio2^+/− retinas, relative to that in Rpe65^−/−/Nrl^−/− retinas. The genes Pepck, Pdk4, Cd44 (encoding CD44), and Hes5 (encoding Hes family bHLH transcription factor 5) are the known TH receptor-target genes (54, 63–65). Expression of these genes was significantly reduced in Rpe65^−/−/Nrl^−/−/Dio2^+/− retinas, relative to that in Rpe65^−/−/Nrl^−/− retinas (Fig. 7A, B for expression of Pepck and Pdk4; our unpublished observations showed reduced expression of Cd44 and Hes5 in Rpe65^−/−/Nrl^−/−/Dio2^+/− retinas, relative to that in Rpe65^−/−/Nrl^−/− retinas).

TH signaling suppression-induced cone protection involves the inhibition of the RIPK/necroptosis activity

To investigate how suppression of TH signaling preserves cones in Rpe65-deficient mice, we examined the death pathway activities. We found that many genes involved in cellular necroptosis were elevated in Rpe65-deficient mice. Necroptosis is a regulated caspase-independent cell death mechanism that resembles necrosis (66, 67) and is characterized by activation of the death receptors including the TNF receptor superfamily, the TNF receptor 1–associated death domain protein, and the RIPK necrosome (Fig. 8). Expression of those necroptosis signaling components was significantly increased in Rpe65^−/−/Nrl^−/− retinas, and that elevation was completely abolished in Rpe65^−/−/Nrl^−/−/Dio2^+/− retinas and by antithyroid drug treatment. Moreover, treatment with T3 significantly increased expression of those genes. These findings support the view that TH signaling suppression-induced cone protection in Rpe65-deficient mice involves the inhibition of RIPK/necroptosis activity (Fig. 8). Elevated RIPK/necroptosis signaling and its role in photoreceptor death have been implicated in other types of pathologic conditions. RIPK-mediated necroptosis was linked to the death of various cell types (68–71), including photoreceptors. RIPK was shown to mediate necroptotic cone death in mouse (24, 72–75) and zebrafish models (76) of inherited retinal degeneration, and in mouse models of experimental retinal detachment (73, 77) and ischemia–reperfusion injury (78).

Figure 8.

Schematic diagram showing how suppression of TH signaling leads to cone protection. The RIPK3/necroptosis signaling is initiated by activation of a death receptor (such as TNFR1), stimulated by a death ligand (such as TNF-α), and recruits TRADD, which allows the formation of the RIPK1/RIPK3/mixed lineage kinase domain-like pseudokinase (MLKL) necrosome, initiating necroptosis. The mitochondrion is the site of ROS production and is associated with oxidative stress responses. These two cellular events, RIPK3/necroptosis signaling and oxidative stress responses, are elevated in Rpe65-deficicient cones. The elevated RIPK3/necroptosis signaling may also induce oxidative stress responses. In addition to the contributions of the original pathologic conditions, TH signaling, which is likely activated in the diseased retina, may contribute to the elevated RIPK/necroptosis activity and oxidative stress responses. Suppression of TH signaling reduces cellular RIPK/necroptosis activity and oxidative stress in degenerating retinas, thereby leading to reduced cell death/cone preservation.

TH signaling suppression-induced cone protection involves the inhibition of the oxidative stress responses

Oxidative stress is caused by an imbalance between the antioxidant defense system and the production of reactive oxygen species (ROS), which are mostly generated in the mitochondrial respiratory chain. Photoreceptors are packed with mitochondria and have extremely high metabolic activity and oxygen consumption. We examined potential involvement of the oxidative stress responses in TH signaling suppression-induced cone protection for the following 2 reasons. First, oxidative stress is a well-established mechanism that contributes to cellular dysfunction/cell death, including photoreceptor death. Oxidative stress has been reported in development and acceleration of many retinal diseases, including retinitis pigmentosa (11, 15, 52, 79), AMD (80–82), glaucoma (83–85), and diabetic retinopathy (86, 87). Second, TH has a profound effect on mitochondrial function/metabolic homeostasis. TH has a central role in oxidative phosphorylation, thus regulating metabolic efficiency, energy expenditure, thermogenesis, and production of ROS. TH also stimulates mitochondriogenesis, thereby augmenting cellular oxidative capacity. High TH signaling has been shown to accelerate the basal metabolic rate, which is associated with increased production of ROS and results in mitochondrial dysfunction and increased oxidative damage in rat liver (88, 89), heart (90), and muscle (91). TH has also been found to affect antioxidant defense (92). We found that many genes involved in oxidative stress responses were up-regulated in Rpe65-deficient retinas, and those elevations were completely abolished in Rpe65^−/−/Nrl^−/−/Dio2^+/− retinas and by antithyroid treatment. Moreover, treatment with T3 significantly increased expression of those genes. These findings support the view that TH signaling suppression-induced cone protection in Rpe65-deficient mice involves inhibition of oxidative stress responses (Fig. 8).

The elevated oxidative stress responses/mitochondrial function in the diseased retina were also reflected by the increased expression of the 2 mitochondrial glucose metabolism–regulating enzymes Pdk4 and Pepck-M. We found that Pdk4 and Pck2 were up-regulated in Rpe65^−/−/Nrl^−/− retina, and those elevations were completely abolished in Rpe65^−/−/Nrl^−/−/Dio2^+/− retinas and by antithyroid treatment. Moreover, treatment with T3 significantly increased expression of Pdk4 and Pck2. Those findings are in agreement with the regulation of TH signaling on mitochondrial function. Increased expression of Pdk4 has been associated with many diseases, such as diabetes (93), neurologic disorders (94), and several cancers (95). Because Pdk4 and Pck2 are known to be transcriptionally regulated by TH (54), increased expression of these genes may also reflect an elevated TH transcriptional activity in the diseased retina. Along with our previous findings showing increased expression of TR and Dio2 in the diseased retinas, increased expression of Pdk4 and Pck2 supports the view that TH signaling is up-regulated in the diseased retinas.

Potential contribution of TH signaling to elevations of RIPK/necroptosis activity and oxidative stress responses in Rpe65-deficient retinas

It is not currently clear how the RIPK/necroptosis activity and oxidative stress responses are elevated in Rpe65-deficient mice. They are likely associated with the initial pathologic conditions, including the lack of 11-cis-retinal and subsequent impaired protein trafficking/endoplasmic reticulum stress (96). Additionally, TH signaling may contribute, at least in part, to the elevated RIPK/necroptosis activity and oxidative stress responses. That theory is supported by the following findings: 1) suppression of TH signaling reduced RIPK/necroptosis activity and oxidative stress responses, 2) treatment with T3 increased the signaling activity and oxidative stress responses, and 3) TH signaling activity is likely elevated in the Rpe65^−/−/Nrl^−/− retina. The finding that suppressing TH signaling protects cones in mice with normal serum TH levels suggests that there is locally elevated TH signaling in degenerating cones/retinas, the degenerating cones become more sensitive to TH, or both. We have previously shown that the expression levels of TR and Dio2 were increased in Rpe65-deficient mice (10, 32). The present work showed elevated expression of the TR target genes Pdk4 and Pck2 in Rpe65^−/−/Nrl^−/− mice and reversal of those elevations by Dio2 heterozygous deficiency or after antithyroid treatment, supporting elevated TR transcriptional activity. Together, these findings support the view that TH signaling is elevated in the diseased retina. That elevation may reflect a stress response, which may sequentially promote or accelerate cell death. TH has been shown to induce necrosis/necroptosis in the rat liver (97) and cancer cells (98) and to increase expression of the TNF-related apoptosis-inducing ligand in humans (99).

In addition to inducing cell necroptosis directly, the elevated RIPK/necroptosis signaling in Rpe65-deficient mice may affect mitochondrial function/oxidative stress responses (Fig. 8). RIPK/necroptosis signaling has been shown to regulate ROS production/oxidative stress responses. The necroptotic signaling-induced cell death is mediated, at least in part, by oxidative stress responses/ROS production. The evidence includes mitochondrial translocation of the RIPK/mixed lineage kinase domain-like pseudokinase necrosome, RIPK-dependent ROS production, stabilization of the necrosome complex by ROS (100), and the subsequent execution of the necroptotic process (68, 101–104). In addition, inhibition of RIPK-dependent ROS production was shown to reduce cell death (103). Thus, the effects of TH signaling on oxidative stress responses in Rpe65-deficient retinas are likely mediated via its direct action on the mitochondrion and the indirect action through RIPK/necroptosis signaling (Fig. 8).

In summary, the present work demonstrated that deficiency of Dio2 protects cones in the LCA model Rpe65-deficient mice. Retinas of these mice showed significantly elevated RIPK/necroptosis activity and oxidative stress responses, and TH signaling suppression, via heterozygous deficiency of Dio2 or antithyroid treatment, reversed those elevations. Moreover, treatment with T3 significantly increased RIPK/necroptosis activity and oxidative stress responses in the retinas, demonstrating the regulation of TH signaling in those cellular events. Findings from this work support the view that TH signaling suppression-induced cone protection involves the inhibition of the RIPK/necroptosis activity and oxidative stress responses, and suppression of TH signaling locally in the retina may represent a strategy for cone protection.

Glossary

AMD

age-related macular degeneration

CAR

cone arrestin

Cy3

cyanine 3

Dio2/3

type 2/3 iodothyronine deiodinase

ERG

electroretinography

Gnat2

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

LCA

Leber congenital amaurosis

Pdk4

pyruvate dehydrogenase kinase 4

Pepck

phosphoenolpyruvate carboxykinase

PNA

peanut agglutinin

qRT-PCR

quantitative RT-PCR

RIPK

receptor-interacting serine/threonine-protein kinase

ROS

reactive oxygen species

Rpe65

retinal pigment epithelium-specific 65 kDa protein

Rpe65^−/−/Nrl^−/−

Rpe65 deficiency on a cone dominant background

T3

triiodothyronine

T4

thyroxine

TH

thyroid hormone

TR

thyroid hormone receptor

AUTHOR CONTRIBUTIONS

F. Yang, H. Ma, and M. R. Butler performed research and analyzed data; and F. Yang, H. Ma, M. R. Butler, and X.-Q. Ding designed research and wrote the manuscript.