Decreased, Triiodothyronine (T3) leads to Altered, retinal layer structure

THs, TH receptors, and deiodinase (DIO) enzymes are important for eye and retinal development in vertebrates. Dio enzymes activate and inactivate THs, consequently playing a central role in regulating TH levels in target tissues. In zebrafish, TH receptors and dio enzymes have been localized in the retina from 24 hpf onwards, probably regulating the differentiation of retinal structures and photoreceptors (Gan et al. (2010), Duval, M. G., & Allison, W. T. (2018)). It is known from amphibians that when TH levels start to rise at the beginning of metamorphosis, the morphology of the eyes starts changing (Fini et al. 2012). In chicken, the developing eye shows a dynamic expression pattern of Deiodinase 2 (DIO2) and Deiodinase 3 (DIO3), probably regulating photoreceptor differentiation and cornea development (reviewed by Darras 2015).

There is ample evidence that reduced THs have an influence on development of the retinal layer structure in fish and other vertebrates. Although the assumed site of T3 decrease is assumed to be in the retinal layers itself, most fish early life stage studies only quantify whole body T3 levels which does not allow for making the distinction between systemic and local T3 levels.

Evidence from exposure to PTU, 6-n-propylthouracil, a classic positive control for inhibition of TPO responsible for TH synthesis:

• Reduced whole body T4 and T3 levels at 14, 21 and 32 days post fertilization (dpf) were observed after exposure of zebrafish to 111 mg/L PTU (Stinckens et al., 2020). Exposure to 37 mg/L PTU reduced T4 levels at 14, 21 and 32 dpf and significantly reduced T3 levels at 32 dpf, while the more limited decrease of T3 levels at 14 and 21 dpf was not statistically significant (Stinckens et al., 2020). Schmidt and Braunbeck (2011) also showed reduced T4 levels in juvenile zebrafish exposed to PTU for 5 weeks. PTU was also shown to reduce T4 levels already at 72 and 120 hours post fertilization (Walter et al., 2019). T3 levels tended to decrease at 72 and 120 hpf but these changes were not significant. Exposures were always continuous and started immediately after fertilization.
• Baumann et al. (2016) described alterations in retinal structure, pigmentation and eye size in 5 day old zebrafish embryos after exposure to PTU. Exposures to 100 and 250 mg/L PTU reduced retinal pigment epithelial diameter and exposure to 250 mg/L increased the grey value of the pigment layer which is a measure of decreased pigmentation.
• Gan et al. (2010) showed that thyroid hormones accelerate opsin expression in differentiating cones and induce the opsin switch, a shift from expression of UV opsin to blue opsin, in differentiated single cones in salmonids. Using in situ hybridization, they characterized the spatiotemporal dynamics of opsin expression and switching in embryos treated with exogenous TH or PTU. The results show that PTU repressed the opsin switch. THs are required for opsin switching in the retina of salmonid fishes.

Evidence from exposure to methimazole, a model thyroperoxidase inhibitor:

• Methimazole was shown to reduce whole body T4 and T3 levels at 14, 21 and 32 days post fertilization after exposure of zebrafish to 50 and 100 mg/L (Stinckens et al., 2020). Exposures were always continuous and started immediately after fertilization.
• Komoike et al. (2013) exposed zebrafish embryos to 10 mM methimazole and observed moderately disrupted retinal structure with apoptosis of retinal      cells already at 48 hpf and more severely disrupted retinal structure at 72 hpf. Major gaps and malformations of the retinal structure occurred at 72 hpf. The observed retinal anomalous morphologies have a direct analogy to the congenital anomalies observed in children exposed to methimazole in utero.
• Reider and Connaughton (2014) exposed zebrafish embryos to methimazole until 66, 70 or 72 hpf and analysed the retina at 72 hpf. The thickness of the ganglion cell layer (GCL) was decreased in embryos exposed to MMI until 66 hpf compared to controls. An increase in GCL thickness was observed in embryos exposed until 70 hpf, and normal thickness was observed in embryos exposed until 72 hpf. Although the impact of the exposure windows cannot be entirely explained, this confirms the relation between reduced T3 and altered retinal structure.

Evidence from other chemical exposures:

• Baumann et al. (2016): After exposure to 200 and 300 μg/L TBBPA, a compound with several mechanisms including a direct interaction with the TH     receptor and binding to the TH binding protein transthyretin, grey values were increased at 5 dpf indicative of reduced pigmentation in the eyes. There were no significant effects on the retinal pigment epithelium diameter. Experiments from Zhu et al. (2018) and Yu et al. (2021) confirm a reduction in T3 levels in both the larvae and embryos (whole body) after exposure to 300 μg/L TBBPA and the locomotor activity of larval offspring was significantly reduced.
• Besson et al. (2020) used treatment with NH3 to highlight the role that THs play in retinal development in metamorphosing convict surgeonfish. They analysed different cell segments, types, and layers of the retina, such as (i) the densities of photoreceptor external segments (perceiving light signals), (ii) photoreceptor nuclei, (iii) bipolar cells (which integrate the synaptic signals originating from the photoreceptors), and (iv) ganglion cells (which integrate signals from bipolar cells and create action potential toward the optic nerve). They investigated the role of THs in the development of these sensory structures by injecting fish daily from d0 to d5 with NH3 (10−6 M), a TH antagonist, to achieve TH signal disruption. NH3 prevents the binding of TH such as T3 to TR, therefore impairing the binding of transcriptional coactivators to TR, which therefore remain in an inactive and repressive conformation.The NH3 treatment was thus applied to repress TH signaling by disrupting the TH pathway leading to an adverse outcome on retinal layer level. Repressed retinal development at both d2 and d5 with a 10- 25 % decrease of bipolar cell density was detected.
• Besson et al. (2020) further showed that treatment with chlorpyrifos reduced T3 levels and reduced bipolar cell density by 10%.
• Bhumika et al. (2014) found that lowering T3 signaling through exposure to different chemicals accelerates optic tectum reinnervation following optic      nerve crush in zebrafish and that this is accompanied by a more rapid resolution of the inflammatory response. Unlike in mammals, full recovery of the damaged CNS is possible in adult fish and amphibians and, for instance, the optic nerve of fish can regenerate completely after injury. Adult zebrafish were exposed to 10 μM of iopanoic acid (IOP), which lowered intracellular T3 availability, or to 7 μM of the TH receptor β antagonist methylsulfonylnitrobenzoate (C1). Both treatments accelerated optic tectum (OT) reinnervation. At 7 days post injury (7 dpi) there was a clear increase in the biocytin labeled area in the OT following anterograde tracing as well as an increased immunostaining of Gap43, a protein expressed in outgrowing axons. This effect was attenuated by T3 supplementation to IOP-treated fish. ON crush induced limited cell death and proliferation at the level of the retina in control, IOP- and C1-treated fish.

Evidence from genetic knockdown and knockout studies:

• Houbrechts (2016) performed deiodinase (DIO) knockdown in zebrafish embryos and observed reduced eye size, disturbed retinal lamination and strong reduction in rods and all four cone types. DIO 1 and 2 are both responsible for converting T4 to the more active T3. Combined knockdown of DIO 1 and 2, leading to reduced T3 levels, altered the structure of the ganglion cell layer (GCL), making it wider and less dense. DIO3 deactivates T3 and defects were more prominent and persistent in D3-deficient fish with observations of marked disorganization across all retinal layers.
• Using genetic zebrafish experiments Duval and Allison (2018) investigated the role of the thyroid hormone receptor thrb in cone differentiation at different time points. Disrupting thrb activity via expression of a dominant negative thrb (dnthrb) at either early or late retinal development had differential outcomes on red cones (reduced abundance), versus UV and blue cones (increased abundance). The effects of thrb change through photoreceptor development, first promoting red cones and restricting UV cones, and later restricting UV and blue cones. Knockdown of thrb causes near-complete absence of red cones and an increase in UV cone abundance (by approximately 35%), whereas expression of dnthrb via heat shock at 52 hpf leads to increased UV (by 27%) and blue cone abundance (by 36%) relative to heat shocked nontransgenic siblings. Inducing dnthrb expression at other time points, including 24 hpf, 30 hpf, and 36 hpf, did not alter cone abundances as dramatically relative to controls (<20% change). This revealed an effect of thrb that is limited to later photoreceptor development: the endogenous receptor negatively regulates blue cone determination. In contrast, disrupting Thrb activity either early (with morpholino knock down) or late leads to more UV cones.
• Ng et al. (2010) showed in mice that knockout of the thyroid receptor, THRb2, results in important changes in the numbers of specific cone types in the retina and M opsins do not even appear at all. Knockout of a thyroid hormone receptor conceptually corresponds to decreased activation of the thyroid hormone receptor due to decreased T3 levels.

Other models of hypothyroidism:

• Gamborino (2000) analysed eye development in a rat model of congenital-neonatal hypothyroidism (HG), induced by combined chemical-surgical thyroidectomy. Histopathological analyses of the eyes of TH-deficient animals revealed decrease in photoreceptor and ganglion cell layer thickness, a delay in photoreceptor outer segment morphogenesis and significantly lower values for ganglion cell nuclear volumes and nuclear pore density.

indirect evidence:

Trimarchi et al. (2008) observed three waves of expression of components of the HPT-axis in specific locations in the retina in progenitor cells and photoreceptor cells during development of the chicken, indicating that thyroid hormones are required for normal retinal development and photoreceptor differentiation

Several studies have shown molecular responses to hypothyroidism that are related to eye development (Bagci et al., 2015; Houbrechts et al., 2016; Baumann et al., 2019) but the exact molecular processes linking lower TH level to disturbances of the layers in the retina is not yet fully understood. 

Both decreased as well as increased TH action has been shown to impact retinal development.

• For example, Ng et al. (2010) showed altered cone appearance in the retina following both DIO3 knockout (leading to hyperthyroidism) and THRb2 knockout (corresponding to hypothyroidism).
• Besson et al. (2020) used pharmacological treatments (T3 + iopanoic acid (IOP), NH3) to not only disrupt but also activate the TH signaling pathway. They used 10−6M T3 + (iopanoic acid) (T3 treatment) to achieve TH signal activation. Here, IOP was used as an inhibitor of deiodinase enzymes, following comparable work in mammals and amphibians, and as routinely used in fish to prevent the immediate degradation of injected T3. The combined treatment thus causes elevated T3 levels. Detected effects on retinal layers were elevated densities of bipolar cells at day 2 in surgeonfish.
• Suppressing TH signaling in retina dystrophy mouse models (a mouse model of retinal degeneration) seems to protect cone viability (Ma et al., 2014; 2016). The authors suggested that the impact of TH on cone survival is independent of its impact on cone opsin expression. The mechanism underlying the effect on cone viability has not been elucidated. 
• Bhumika et al. (2014) showed accelerated reinnervation of the optic tectum after optic nerve crush in zebrafish that had been treated with IOP or a TR antagonist. B     oth treatments cause hypothyroidism. Supplementation of T3 reduced the rate of reinnervation.

Another uncertainty lies in the systemic versus local changes in T3 levels. Although the assumed site of T3 decrease is assumed to be in the retinal layers itself, most fish early life stage studies only quantify whole body T3 levels which does not allow for making the distinction between systemic and local T3 levels.

Most knowledge comes from effects observed in developing organisms. There are some gaps in our knowledge about how TH levels affect the eyes of already fully developed organisms and/or whether they have similarly serious effects on the retinal layers. It can be assumed that the effects, if any, are weaker. Studies (Reider et al. 2014) found that layer thickness varied across ages suggesting that these retinal layers are differentially sensitive to for example MMI and/or that there are different critical periods of sensitivity of the retinal tissue.

• One feedback loop mechanism could be triggered by iodine deficiency or inhibition of iodine uptake. It appears probably that the inhibition increases the secretion of Thyroid stimulating hormone, which could stimulate the expression of the NIS-transporter. This increase in TSH could shift the ratio in favour of T3.