Molecular Vision 2006; 12:915-930 <http://www.molvis.org/molvis/v12/a104/> Received 19 October 2005 | Accepted 21 July 2006 | Published 11 August 2006

Download Reprint

Effects of induced systemic hypothyroidism upon the retina: Regulation of thyroid hormone receptor alpha and photoreceptor production

Michelle M. Mader, David A. Cameron
 
 

Department of Neuroscience and Physiology, SUNY Upstate Medical University, Syracuse, NY

Correspondence to: David A. Cameron, Department of Neuroscience and Physiology, SUNY Upstate Medical University, 750 East Adams Street; Syracuse, NY, 13210; Phone: (315) 464-8149; FAX: (315) 464-7712; email: camerond@upstate.edu

Abstract

Purpose: Investigate the effects of systemic hypothyroidism upon the differentiated, growing, and regenerating retina of postmetamorphic winter flounder, a vertebrate that experiences a thyroid hormone (TH) induced metamorphosis during development.

Methods: A loss-of-signal strategy was utilized in which TH signaling was disrupted by inhibiting TH synthesis. Induced hypothyroidism was confirmed by radioimmunoassay. Reverse transcriptase PCR (RT-PCR), real-time quantitative PCR (qPCR), molecular cloning, non-isotopic in situ hybridization, western blot analysis, and indirect immunohistochemistry techniques were performed to analyze retinal thyroid hormone receptors (TR), photoreceptor production, and the phenotypic repertoire of differentiated retinal cells as a function of TH signaling status.

Results: Molecular bases for TH signaling were supported by retinal expression of TH receptors α and β. TH-dependent transcriptional regulation of TRα but not TRβ was indicated, with induced hypothyroidism producing an increase in TRα expression. Evidence for post-transcriptional regulation of retinal TRα was observed. The repertoire of inner retinal cell types in premetamorphic fish (a naturally low TH condition) matched that observed in the central retinas of both normal postmetamorphic fish (a naturally elevated TH condition) and postmetamorphic fish rendered hypothyroidic. In differentiated postmetamorphic retina there was no evidence for significant differences in opsin expression between normal and hypothyroidic animals. Induced hypothyroidism did, however, significantly affect the types of photoreceptors that were produced in postmetamorphic retina: as a hypothyroidic postmetamorphic retina grew or regenerated following injury, the phenotypic repertoire of newly-produced photoreceptors matched that observed for premetamorphic retina, in which rods, SWS2-expressing "blue" cones, and LWS-expressing "red" cones are absent, and only the RH2-expressing "green" cone type is present. The effects of induced hypothyroidism upon photoreceptor specification (manifestation of the rod lineage) and differentiation (expression of a particular opsin by specified cones) were apparently reversible.

Conclusions: The results suggest a TH-dependent regulation of retinal TRα, a lack of TH-dependent regulation of the phenotypic identity of differentiated retinal cells, and the operation of similar cytogenic mechanisms during retinal growth and regeneration. The principal conclusion is that TH signaling significantly affects, in a targeted manner, the production of both rod and cone photoreceptors during retinal growth and regeneration.

Introduction

Thyroid hormone (TH) is a general hormone that regulates many cellular functions [1]. Two TH receptor subtype genes, termed α and β (TRα, TRβ), have been identified, each of which encodes a member of the DNA-binding nuclear receptor superfamily. TRα and TRβ exhibit DNA binding activity in both the ligand-bound and unbound states [2], and thus cellular effects of TH signaling can be achieved, in association with recruited co-activators or co-repressors, at the level of gene transcription [2-5].

TH signaling is a critical determinant of vertebrate development. TH is known to trigger vertebrate metamorphosis [6-14] and its importance for proper assembly of the human central nervous system (CNS) has been established [15,16]. Along with other components of the developing CNS the neural retina is a known target of TH signaling [17]. The presence of deiodinase enzymes, TRα, and TRβ in the developing retina has been reported [18-20], and previous investigations have provided biochemical evidence for complex TH signaling networks within the retina [18,21-24], a structure with known phenotypic lability [25]. Photoreceptor development in both anamniotes and amniotes is particularly sensitive to manipulations of TH signaling [26-33]. Many aspects of how TH affects retinal cells, however, including differential affects upon phenotypic maintenance on the one hand, and control of developmental/regenerative events on the other, remain unresolved.

Because of its importance to CNS development in general-and retinal photoreceptor development in particular-we investigated potential roles of TH upon TRs, photoreceptor production, and cellular phenotypic maintenance in the growing and regenerating vertebrate retina. A metamorphic vertebrate, the winter flounder, served as the model system, for which there are several empirical advantages. First, like other fish the flounder adds new cells to the retina throughout its life, and it can regenerate retinal cells following an injury [32], thus providing a convenient and accessible substrate for evaluating the mechanisms of cellular production during these two aspects of retinal assembly. Second, normal development of the flounder involves a TH-induced metamorphosis during which a substantial reorganization of the body plan occurs [8-10]. Within the population of metamorphic changes is a significant re-organization of the neural retina [34,35], including a complex set of changes in opsin expression that, in an earlier report, suggested targeted effects of TH upon photoreceptor development [32]. Third, loss-of-signal experiments can be performed in flounder, with systemic hypothyroidism induced effectively via systemic drug exposure [36,37], thus permitting direct evaluation of how altered TH signaling affects molecular and cellular attributes of the retina.

We report that retinal TRα is a target of TH signaling at both the transcriptional and post-transcriptional levels. In the growing and regenerating postmetamorphic flounder retina TH signaling is apparently required for producing the full phenotypic repertoire of photoreceptors, but has little if any effect upon the production of other retinal cell classes. Although TH has no evident role in the maintenance of differentiated cellular phenotypes in the retina, including those of photoreceptors, it appears to influence photoreceptor production at the level of both specification (rod production) and differentiation (expression of opsin by specified cones). These results indicate that TH signaling is a controlling element of photoreceptor development that affects multiple stages of the photoreceptor lineage.

Methods

Postmetamorphic winter flounder (Pleuronectes americanus) were purchased from the Marine Biological Laboratory (Woods Hole, MA) and housed in standard fish tanks. Premetamorphic (larval) flounder less than or equal to 21 days after hatching were acquired from the Coastal Marine Laboratory (University of New Hampshire, New Castle, NH) and used immediately for experiments. For all experimental analyses there were no evident differences associated with eye (left or right) or retinal hemifield.

Many of the methodologies employed in this study were identical to those reported previously, including: introduction of focal lesion to the dorsal retina of one eye in each fish [38,39]; intraocular injection of the thymidine analog 5-bromo-2-deoxyuridine (BrdU; Sigma, St. Louis, MO) at the time of initial exposure to thiourea [40]; euthanasia 30-47 d subsequent to the beginning of thiourea exposure [41]; cryosectioning and indirect fluorescence immunohistochemical (IHC) analysis of retinal cells [41]; non-isotopic in situ hybridization analysis using sense and anti-sense digoxin-labeled cRNA probes [32,42]; measuring and quantifying distances of IHC- or in situ hybridization-labeled photoreceptors to each other and to sites of retinal cytogenesis [32]; reverse transcriptase polymerase chain reaction (RT-PCR) [43]. Methodological details unique to this report follow.

Thiourea administration and thyrois hormone measurement

A loss-of-signal strategy was utilized in which TH signaling was disrupted by inhibiting TH synthesis. Systemic hypothyroidism was induced in postmetamorphic flounder (n=7) via constitutive exposure to thioruea (TU). TU was dissolved in salt water and added directly to fish tanks to a final concentration of 40 μM [36,37]. Charcoal/activated carbon was removed from all fish tank filters to prevent chemical removal of the TU, necessitating quarter-volume tank water changes every three days with fresh TU-containing salt water. Blood was collected from experimental flounder at the time of euthanasia and spun down to allow removal of the serum fraction, which was immediately stored at -20 °C. Serum samples were analyzed with a T₃ Solid Phase Component System RIA kit to determine bound and unbound T₃ concentration in control and TU-exposed samples (MP Biomedicals). The RIA analyses indicated that in all TU-exposed flounder the serum T₃ concentration was at or below the detection limit of the assay, corresponding to a >95% reduction in circulating TH level (Figure 1A). Because serum is likely a major source of retinal T₃ [23], TU exposure was inferred to significantly reduce retinal TH.

Immunohistochemistry

The primary antibodies utilized to detect cells in pre- and postmetamorphic flounder retina were: zpr1 and zpr3 (University of Oregon Monoclonal Antibody Facility, Eugene, OR); rabbit anti-rhodopsin (courtesy B. Knox, SUNY Upstate); mouse anti-proliferating cell nuclear antigen (PCNA; Catalog number M0879; Dako Cytomation; Glostrup, Denmark); mouse anti-BrdU (Catalog number RPN202; Amersham; Little Chalfont, England); rabbit anti-zebrafish glutamine synthetase (courtesy Paul Linser, University of Florida, Gainesville, FL); anti-calretinin (Catalog number AB149; Chemicon; Temecula, CA); mouse anti-tyrosine hydroxylase (Catalog number MAB318; Chemicon); rabbit anti-neuropeptide Y (Catalog number 22940; DiaSorin; Stillwater, MN); mouse anti-parvalbumin (Catalog number P3088; Sigma); rabbit anti-GABA (Catalog number A2052; Sigma); rabbit anti-PKC (Catalog number AB1610; Chemicon; this antibody is no longer available commercially); and anti-TRα-1 (Catalog number PA1-211A; Affinity BioReagents; this anti-TRα recognizes a peptide sequence that matches exactly a sequence inferred by conceptual translation of the cloned flounder TRα). Antibodies that recognize flounder TRβ are currently unavailable. Antibodies were utilized at 1/100 dilution, except for anti-rhodopsin (1/500) and anti-GABA (1/200), in PBS/0.3% Triton-X (pH=7.4). Detection of primary antibodies was achieved with cyanine 2 (Cy2)-, cyanine 3 (Cy3)-, or cyanine 5 (Cy5)-conjugated secondary antibodies (Jackson ImmunoResearch; West Grove, PA). Images of reacted sections were captured with standard epifluorescence digital microscopy, and images of triple-label retinal cryosections were acquired with a laser scanning confocal microscope (SUNY Upstate Center for Biomedical Imaging).

Molecular cloning and gene expression analyses

TH receptors (TR) were cloned from cDNA derived from postmetamorphic neural retina and premetamorphic whole bodies using techniques reported previously [32]. Initial PCR-mediated amplification of partial-length sequences encoding TR subtypes α and β were achieved with degenerate primers designed from the corresponding TR sequences of two other flatfish, the Atlantic halibut (Hippoglossus hippoglossus; GenBank accession number AF143296) and the Japanese flounder (Paralichthys olivaceus; Genbank accession numbers D16461, D16462, and D45245). All nucleotide and inferred amino acid sequences of the winter flounder TR have been deposited in the GenBank database (TRα, AY794223; TRβ, AY794222).

The coding sequences of the P. americanus opsins (RH1, RH2, SWS2, and LWS) [32] and the 60S ribosomal protein L8 (Genbank accession number AW013168) [44] have also been reported. For RT-PCR and qPCR analyses of these genes-retinal expression the following primer sets were utilized (5'-3'): TRα forward, CGA GAC GCT GAC GCT GAA C; TRα reverse, GTC ACC TTC ATC AGC AGC; TRβ forward, TCA TCC GAT CGG CCA GGC C; TRβ reverse, AAT CCA TTG GTC AGT CCT C; RH2 ("green" cone) opsin forward, ATG CAG TGC TCA TGC GGA CC; RH2 opsin reverse, CTG AAC GTC ACT CAG TCC; RH1 opsin (rhodopsin) forward, GTG GCC TGG TAT ATC TTC C; RH1 opsin reverse, ACA TTA GCC TTC AGT CGT TTC C; LWS ("red" cone) opsin forward, GTC CAA TTA CCA VAT TGC TCC; LWS opsin reverse, GTA GCA CAA GAT GAT AAC ACC C; SWS2 ("blue" cone) opsin forward, TGC ACT TAT TGC CTC AGC TC; SWS2 opsin reverse, TTC TTC ATC ATC ACC TCC ACC C; 60S ribosomal protein L8 forward, ACA GTT CAT CTA CTG CGG C; 60S ribosomal protein L8 reverse, ACT TGT GGT AGG CAC GAC C. The primer sequences used to amplify PCR-derived templates for DIG-labeled cRNA synthesis were as reported previously for opsins [32], and for TRs were: TRα forward, ATT CGA AGC GCG TGG CCA AGC; TRα reverse, CCA AGC TTT TGC CCA AAT C; TRβ forward, TCC AGA AGA ACC TCA ACC C; TRβ reverse, CGA CAT GAT CTC CAT GCA AC. The identity of each PCR-amplified product was qualitatively assessed with standard gel electrophoresis and confirmed with sequencing (Biotechnology Resource Center, Cornell University; Ithaca, NY).

Real-time quantitative PCR (qPCR) analysis was performed using a Smart Cycler system (Cepheid) and a SYBR green detection method [43]. Briefly, qPCR reaction mixtures used the primer pairs listed above at 0.1 nmol/μl each, with cDNA sample at approximately 0.1 μg/μl per 25 μl reaction tube. Neural retina cDNA derived from control and TU-exposed fish were analyzed, and produced equivalent intra-condition results. The qPCR conditions (i.e., primer annealing temperatures) were optimized for each primer pair, as judged by the amplification of a single product with a melt temperature that was significantly different from that obtained with the same reaction mixture and protocol in the absence of cDNA. For each qPCR analysis each product (i.e., primer pair) was run in triplicate, and a common "master mix" was used to ensure equivalent cDNA concentration in each tube. Threshold cycle values (C_T) were defined for the fluorescence growth curve of each reaction tube as the qPCR cycle at which the second derivative of the growth curve was maximal, and the relative C_T differences for each product were determined relative to the C_T values for RH1, a high-abundance transcript in all retinal samples.

Western blot analysis

Neural retina for western blot analysis was removed from dark-adapted flounder and immediately frozen on dry ice; for the TU-exposure condition retinas were extracted following 14 d constitutive exposure to TU. The procedures for tissue processing and western blot analysis were as described previously [45]. Protein concentrations were determined with a Bradford assay. The anti-TRα antibody used in the IHC analyses was also used in the western blot analysis. After developing, the blots were stripped and stained with amido black (45% methanol, 10% acetic acid, 0.1% amido black). Because tested antibodies against actin failed to label protein material from winter flounder, the TRα band "intensities" were referenced to apparently TU-independent, amido black-labeled bands, which served as loading controls.

Results

Thyroid hormone receptors in flounder retina

RT-PCR analysis revealed expression of TRα and TRβ in winter flounder, both in premetamorphic animals and postmetamorphic retina (Figure 1B). The coding sequence of each TR message was determined, and conceptual translation revealed the corresponding amino acid sequences (Figure 1C). The deduced amino acid sequences for TRα and TRβ contained the expected DNA-binding (c4 zinc finger) and ligand-binding domains, and were 96% and 93% identical, respectively, to the corresponding TR of Japanese flounder (P. olivaceous) [46,47]. The winter flounder TRβ included a contiguous sequence of 20 amino acids in the hinge domain, starting at amino acid 172, that was not evident in the TRβ of other fish (GenBank accession number AY794222).

The expression patterns of TRα and TRβ in pre- and postmetamorphic flounder retina were determined with non-isotopic in situ hybridization. In pre- and postmetamorphic retina TRα expression was evident in most, if not all, cells of the outer nuclear layer (ONL), inner nuclear layer (INL) and retinal ganglion cell (RGC) layers (Figure 2). In premetamorphic retina there was a particularly strong field of TRα expression at the retinal margin (Figure 2, blue arrows), a region that corresponds to the circumferential germinal zone (CGZ), the site of substantial cytogenesis in the growing flounder retina [32]. The expression pattern of TRβ, like that of TRα, was fairly broad throughout the premetamorphic retina, although qualitatively at a lower overall level than TRα. In postmetamorphic retina the distribution of TRα expression was similar to that of premetamorphic retina, whereas TRβ expression was qualitatively stronger than premetamorphic retina (Figure 1B) and characterized by occasional strong "patches" at the ONL (Figure 2, black arrows). TRβ expression was also evident at the postmetamorphic CGZ. The RT-PCR and in situ hybridization analyses thus indicated that TRα and TRβ are expressed in the retinas of pre- and postmetamorphic flounder.

Distribution of TRα protein in the retina

Western blot analysis was performed to confirm the specificity of the anti-TRα antibody. The antibody labeled a product of the expected size for TRα (48 kda; Figure 3A). Based upon this confirmation of antibody recognition, immunohistochemical labeling was performed to determine the distribution of TRα protein in the postmetamorphic retina. Positive labeling for TRα was observed in the ONL, and double labeling with zpr1 indicated that TRα is located proximal to, but not coincident with, cone pedicles at the outer plexiform layer (Figure 3B). TRα label was also evident at the pigmented epithelium somata, RGC fiber layer, INL, and CGZ (Figure 3B), the latter two regions consistent with the expression patterns observed with in situ hybridization analysis (Figure 2). Double label analysis with an antibody against PCNA suggested that cells expressing TRα at the CGZ partially overlap with (presumably) undifferentiated cells that are mitotically active (Figure 4). These results were consistent with those of the RT-PCR, in situ hybridization, and western blot analyses, and indicated that TRα is expressed and its protein is present in the retina, including a region associated with cytogenesis.

Regulation of thyroid hormone receptors by thyroid hormone

Both RT-PCR and qPCR analyses indicated that TU exposure leads to an increase in retinal mRNA encoding TRα. This outcome was judged qualitatively by a TU-dependent increase in gel intensity for the TRα product in the RT-PCR analysis and by a left-ward shift of the TRα growth curves relative to those for RH1 in the TU condition compared to control (Figure 5A). The decrease in the TRα C_T corresponds to an estimated five-fold increase in TRα expression in the TU condition. The qPCR analyses revealed no TU-dependent changes in expression were observed for 60S ribosomal protein L8, nor for TRβ, consistent with the RT-PCR tests (gel images and corresponding growth curves of Figure 5A). Because the mRNA derived from whole retina displayed a significant change in TRα expression in the TU condition, it was further concluded that TU exposure rendered the entire retina hypothyroidic.

To determine effects of induced hypothyroidism upon levels of TRα protein in the retina, densitometry analysis of western blots was performed with retinal material collected from control and TU-exposed flounder. As a loading control the "intensity" of the TRα bands was referred to a non-target, amido black-stained band in the blots. No significant difference in TRα protein level was observed between control and TU-exposed retinas (Figure 3A). Consistent with the western blot results IHC analysis of TRα in control and TU-exposed postmetamorphic retinas indicated no evidence for TU-dependent changes in TRα protein level or distribution (Figure 5B). Combined with the RT-PCR and qPCR tests, the protein analyses suggest that TH-dependent regulation of TRα expression in the postmetamorphic retina is accompanied by a "counteracting" post-transcriptional regulation, the apparent combined effects of which are a maintenance of the control level and distribution of TRα protein.

TH signaling and phenotypic maintenance of retinal cells

IHC, RT-PCR, qPCR, and in situ hybridization techniques were used to assess the role of TH signaling in the phenotypic maintenance of differentiated retinal cells. Retinas from premetamorphic (a natural low TH condition), normal postmetamorphic (a normal high TH condition), and hypothyroidic postmetamorphic fish (a thiourea-induced low TH condition) were screened with an array of antibodies intended to label a broad spectrum of distinct retinal cell types [48]. Most major retinal cell classes could be labeled in all three conditions, including cone photoreceptors (zpr1), horizontal cells (anti-calretinin; anti-GABA), bipolar cells (anti-PKC), Müller glia (anti-zebrafish glutamine synthetase), interplexiform cells (anti-tyrosine hydroxylase), amacrine cells (anti-calretinin, anti-GAGA, anti-neuropeptide Y, anti-parvalbumin), ganglion cells (anti-GABA), and perhaps displaced amacrine cells within the ganglion cell layer (anti-parvalbumin; Figure 6). Consistent with an earlier report the only observed differential labeling involved an antibody that labels rod photoreceptors (zpr3), a cell type that is absent from premetamorphic retina (Figure 6, top row) [32]. The lack of evident differences in labeling patterns for the majority of these immunohistochemical markers suggested that neither a difference in TH level associated with normal development (i.e., the metamorphic transition), nor experimental diminution of TH level in postmetamorphic fish, affects the phenotypic characteristics of pre-existing, differentiated retinal cells, including a cell type (rod photoreceptors) that is absent during the hypothyroidic period of normal development.

Retinal opsin expression was evaluated in an effort to further investigate potential roles of TH in the phenotypic maintenance of retinal cells. RT-PCR analysis of opsin expression in normal and hypothyroidic postmetamorphic retinas qualitatively indicated that expression of the RH1, RH2, SWS2, and LWS opsins were not significantly affected by induced hypothyroidism, nor was the expression of the "housekeeping" gene L8 (Figure 7A, left). These results were supported by qPCR analysis, in which, compared to rhodopsin mRNA, no significant hypothyroidism-induced changes in the relative amounts of cone opsin or L8 mRNA were observed (p>0.05 for all cases, Student's independent t-test; Figure 7A, right). Because the total amount of new retinal area added from the CGZ during the period of TU exposure was estimated to be <2% of the total retinal area, it was not expected that TH-dependent effects upon photoreceptor production would significantly impact the RT-PCR or qPCR analyses of opsin expression. Lastly, in situ hybridization analysis of differentiated (i.e., central) retina of hypothyroidic postmetamorphic fish revealed no apparent, hypothyroidism-induced effects upon the patterns of RH1, RH2, SWS2, or LWS opsin expression (Figure 7B, top row) compared to control retina (see figure 4 of Flamant et al. [31]). The RT-PCR, qPCR, and in situ hybridization analyses were thus consistent with the IHC experiments in suggesting no apparent role for TH in the phenotypic maintenance of differentiated retinal cells, including photoreceptors.

TH effects upon photoreceptor specification and differentiation

Because TH signaling has been implicated as a controlling element of vertebrate photoreceptor development, the effects of induced hypothyroidism upon photoreceptor production in the growing and regenerating flounder retina were investigated. IHC techniques were used to assess potential impacts of TH manipulation upon normal retinal growth. The presence of PCNA-positive cells at, and the central displacement of BrdU-positive cells from, the CGZ of both normal and hypothyroidic postmetamorphic retinas indicated that induced hypothyroidism does not halt retinal cytogenesis per se (Figure 8).

Both IHC and in situ hybridization techniques were used to evaluate the production of photoreceptors at the CGZ and regenerative lesion sites. For TH-normal retinas, coincident screening with an antibody that selectively labels rod photoreceptors (anti-rhodopsin) and another antibody that selectively labels cone photoreceptors (zpr1) indicated a qualitative differential distribution of rods and cones proximal to regions of retinal cytogenesis (Figure 8, bottom). Inter-photoreceptor measurements indicated that the distance between the rods and cones nearest to the CGZ and/or regeneration site was 6.5±4.1 μm (mean±standard deviation, measurements taken from n=30 sections derived from five fish). This differential distribution of newly-born rods and cones was consistent with an earlier report [32], and was further supported by measurements taken from retinas processed for in situ hybridization analysis of opsin expression: in normal postmetamorphic retinas the expression of the RH2, SWS2, and LWS cone opsins is always spatially closer to the regions of cytogenesis than the RH1 opsin (Figure 8, Figure 9A,B) [32]. These results indicated that during retinal assembly cones are produced prior to rods, similar to the sequence of photoreceptor specification observed during normal retinal development; that is, across the metamorphic transition.

At regions of cytogenesis in hypothyroidic retinas the patterns of photoreceptor-specific antibody labeling and opsin expression were significantly different from those of normal retinas. Compared to normal retinas there was a significantly greater distance between the rods (rhodopsin-positive) and cones (zpr1-positive) nearest to regions of cytogenesis (17.3±6.9 μm, measurements taken from n=28 sections derived from four fish; p<0.001, Student's t-test analysis of the 28 sections; compare red bars in the bottom row of Figure 8). Furthermore, a differential pattern of cone opsin expression was observed in these regions, with the RH2 opsin expressed closer to the cytogenic regions than the SWS2 and LWS opsins (CGZ: Figure 7B, bottom row). This effect was particularly evident with in situ hybridization analysis of retinal regeneration sites (Figure 10). Quantification of the in situ hybridization patterns revealed two statistically significant differences in hypothyroidic retinas compared to control: (a) the nearest points of SWS2 and LWS opsin expression relative to the cytogenic regions were coincident with that of RH1 expression (Figure 9D,E), and (b) a greater differential between the location of the RH2 and RH1 opsin-expressing cells (Figure 9A-E). The differential effects of TH upon the patterns of opsin expression at regions of cytogenesis are schematized in Figure 9C (normal) and Figure 9F (hypothyroidic). These observations are consistent with a substantial deficiency, or lack, of rod production following the onset of TU exposure, and suggest that like the rods but unlike the "green" cones, "blue" and "red" cones are not produced in hypothyroidic postmetamorphic retinas.

The reversibility of the induced hypothyroidism effects was evaluated by returning TU-exposed fish to normal seawater for 14 days. In situ hybridization analysis suggested that the resumption of the normal TH condition converts the premetamorphic-like retina produced during the period of TU exposure to a phenotypic profile similar to that of postmetamorphic retina. Specifically, an atypical, "spotty" pattern of RH1, SWS2, and LWS expression was observed in those regions where, in the hypothyroidic condition, only RH2 opsin was expressed (Figure 10, Figure 11). The re-emergence of postmetamorphic opsins in these regions provided additional support for the hypothesis that TH signaling directly affects the patterns of photoreceptor development in flounder retina.

Discussion

The goals of the current study were to investigate TH signaling mechanisms in the flounder retina, and to identify specific roles of TH upon the production and phenotypic maintenance of retinal cells. The loss-of signal experiments indicated that although not apparently involved in the phenotypic maintenance of differentiated cells, TH signaling regulates one of its receptors at the transcriptional and post-transcriptional levels, and controls specific aspects of photoreceptor development. We here present interpretations of the results within the contexts of TH signaling mechanisms, retinal growth, and retinal regeneration.

TH regulation of TH signaling in the flounder retina

Many vertebrates experience a TH-dependent metamorphosis during development [7,10,12,14]. In the winter flounder the TH-triggered metamorphic transition includes substantial cellular and molecular reorganization of the retina [32,34,35]. The current investigation suggests that TH-mediated effects upon the growing flounder retina include a differential regulation of TRs. Flounder rendered systemically hypothyroidic via TU exposure exhibited higher levels of TRα mRNA compared to controls, as indicated by real-time and RT-PCR analyses (Figure 5A). No corresponding, TH-dependent change in TRβ expression was observed. The functional implication of this selective, TH-dependent regulation of TRα is not clear, but might be indicative of a complex feedback mechanism between TH signaling and its receptors [49]. Interestingly this hypothesized regulatory mechanism was not manifest post-transcriptionally: neither the western blot nor immunohistochemical analyses indicated any TU-dependent change in the level of retinal TRα protein. This result admits at least two alternative interpretations: (a) TH affects TRα expression transiently, and the subsequent effect upon TRα protein is delayed beyond the two week post-TU point at which TRα protein was analyzed in this investigation; (b) there is an active, negative-feedback, post-transcriptional regulation of TRα by TH, perhaps at the level of translation or protein turnover. The latter interpretation could involve mechanisms that regulate protein degradation or translation of pre-existing mRNA, such that endogenous TH levels might control a translational suppressor mechanism upon TRα. Our observed "mismatch" between the levels of TRα mRNA and protein is reminiscent of an earlier investigation in which 5' UTR variants of human TRβ were shown to directly inhibit translation [50]. Mechanistically this type of post-transcriptional regulation could involve ribosome access, translation efficiency, initiation codon selection, and/or recruitment of mRNA binding proteins. Although the functional role(s) of this inferred regulation is unclear, it suggests potential independence between the effects of TH upon transcriptional and post-transcriptional phenomena for a single gene product in the flounder retina. Additional investigation is required to identify the precise regulatory mechanism(s) that operate in this model system.

TH signaling and retinal cytogenesis

In the current investigation, with the exception of rod photoreceptors there was no overt difference in the phenotypic repertoire of retinal cells between pre- and postmetamorphic flounder (Figure 6). This suggests that the substantial change in TH status associated with the flounder's metamorphic transition affects neither the major phenotypic characteristics of differentiated retinal cells, nor the development of most retinal cell types. This interpretation was supported by the apparent lack of effect of induced hypothyroidism upon differentiated retinal cells in the postmetamorphic retina, including differentiated photoreceptors (Figure 10, Figure 11). Additionally, the phenotypic signature and production of an inner retinal neuron commonly associated with the rod signaling pathway, the PKC-positive bipolar cell [51], was apparently not influenced by TH status, even though the production of rods is apparently TH-dependent (Figure 10, Figure 11). These results indicate that if TH signaling influences retinal cell phenotype, it is likely to do so for a relatively limited number of retinal cell types, and perhaps only prior to terminal differentiation.

We hypothesize that TH signaling mediates the cellular and molecular reorganization of the flounder retina following metamorphosis, particularly at the level of photoreceptors. How the specific TRs affect these distinct aspects of photoreceptor development are unclear, but there is ample precedent from several model systems that photoreceptor development is a target of TH signaling. For example, TH exposure has been reported to expand the field of rhodopsin expression in developing Xenopus [20], affect the development of UV-sensitive cones in trout [21], and drive a production of cones at the expense of rods in a cell culture model of rat retinal development [22]. The importance of specific TRs in photoreceptor development is also becoming clear, with TH signaling via TRβ being implicated in several model systems [19,28,52]. Furthermore, a human form of monochromatism is associated with mutation of the C-terminal region of TRβ [28,53]. Because of its contribution to photoreceptor development, and because it is expressed at the ONL in flounder retina (Figure 4), TH signaling via TRβ cannot be ruled out as a candidate mechanism for inducing the changes in photoreceptor phenotype (morphology and opsin expression) that accompany flounder metamorphosis. As stated earlier, however, there is independent indication that TRα is a particularly important signaling element during fish metamorphosis [49]. Additionally, because it is expressed at the ONL and CGZ, and is transcriptionally regulated by TH (Figure 4, Figure 8), TRα presents itself as a candidate for regulating important cellular and molecular events in the flounder retina. Additional investigation is required to identify and characterize the specific contributions of each TR to specific cellular and molecular aspects of retinal development in this system, and particularly the changes in retinal organization that accompany metamorphosis.

TH signaling and photoreceptor development

TH signaling is known to affect photoreceptor development in the vertebrate retina, and the underlying molecular mechanisms are the subject of ongoing study [17,22]. Although many interpretations are possible, our data suggest that TH is a key determinant of both photoreceptor specification (the production of rod photoreceptors) and differentiation (cone photoreceptor morphologies and the repertoire of cone opsin expression). Specification, defined as a flexible, alterable selection of cell fate that can be influenced by "environmental" factors [54] is here applied to the manifestation of a particular photoreceptor lineage (rod or cone) during retinal cytogenesis. Differentiation, on the other hand, refers to the acquisition of "mature" phenotypic characteristics by cells that have already been specified, a process that is here defined to include the expression of a particular opsin (SWS2, RH2, or LWS) by a specified cone photoreceptor. Specification and differentiation are thus distinct events along a cell's developmental trajectory, and our data supports a model in which TH influences both the specification and differentiation of photoreceptors.

With respect to specification our evidence suggests that TH is required for manifestation of the rod photoreceptor lineage. This requirement was evidenced by both the lack of rod photoreceptors in premetamorphic retina (i.e., a developmental state with low TH), and the lack of rod photoreceptor production in postmetamorphic retinas that were rendered hypothyroidic. As assessed by IHC, RT-PCR, qPCR, and in situ hybridization analyses, however, TH did not affect the phenotypic status of differentiated rods, suggesting a selective role for TH in rod photoreceptor development and an apparent lack of TH effect upon RH1 expression per se. We hypothesize that rod precursor cells [55,56] are a cellular target of this TH signaling, and that the re-emergence of rod photoreceptor production in postmetamorphic retinas following the resumption of the normal TH level is due to a currently undefined effect of TH upon latent rod precursor function.

The data further suggest that TH influences the differentiation, but perhaps not specification, of cone photoreceptors. As the developing, growing, or regenerating flounder retina produces photoreceptors there is an apparent TH-independent lineage toward RH2-expressing cones, reminiscent of the "default" photoreceptor fate reported for Drosophila [57]. Evidence in support of this interpretation was provided by the presence and production of RH2-expressing cones in premetamorphic retina, normal postmetamorphic retina, and postmetamorphic retina rendered hypothyroidic. Although a role for TH in the specification of cone photoreceptors cannot be formally ruled out, the results suggest that cone specification in flounder retina may occur independently of TH status. The production of SWS2- ("blue") and LWS- ("red") expressing cones, in contrast, was observed only in retinas exposed to the normal, postmetamorphic level of TH. This suggests that the choice of cone opsin expression by specified cones and thus the determination of a major function-defining characteristic of a mature cone is directly controlled by TH signaling.

TH receptors, including both the α and β isoforms, are expressed in flounder retina at regions that correspond to cytogenesis (Figure 4). In particular, TRα is apparently expressed robustly at the CGZ, which is the site of cone photoreceptor production. This observation indirectly suggests a role for TRα in the TH-dependent cone differentiation events reported here. Within this context we note that previous investigations have implicated TH signaling via TRα as a significant contributor to metamorphosis [45,58]. The downstream targets of TH signaling that affect photoreceptor specification and/or differentiation in the flounder retina, however, are unknown. For example, it is not yet clear if TH-dependent aspects of cone opsin expression are regulated in this system via direct effects upon opsin promoters [59], indirectly via regulation of, or interaction with, transcription factors that have known binding sites in opsin promoters [60-63], or a combination of both mechanisms. Molecular dissection of the transcriptional mechanisms through which TH affects photoreceptor development in flounder retina is a subject of ongoing investigation.

Adult retinal growth and regeneration recapitulate development

Premetamorphic winter flounder are pelagic visual day feeders [64] that express only a single visual pigment based upon a RH2 opsin [32]. Following TH-driven metamorphosis the flounder inhabit benthic regions, acquire appropriate feeding strategies, and manifest a complex capacity for body camouflage, all of which are likely to be supported by the new repertoire of retinal photoreceptor subtypes (rods, single cones, and double cones) and expressed opsins (RH1, SWS2, RH2, and LWS). Consistent with our earlier report [32], the current study provides evidence that ongoing growth of the postmetamorphic retina involves a recapitulation of the mechanisms that drive retinal assembly during normal development. Specifically, the phenotypic organization of the premetamorphic retina, which is produced during low TH conditions, is consistent with the premetamorphic-like retina produced by the growing postmetamorphic retina during induced hypothyroidic conditions. Additionally, because a similar effect of TH upon photoreceptor production was observed for regenerating postmetamorphic retina, our results support the hypothesis that regeneration of the adult vertebrate retina involves a recapitulation of the mechanisms that drive and direct cytogenesis during normal development and growth [65]. This observation indicates that retinal regeneration in adult fish, which proceeds at a substantially faster rate than normal retinal growth [32], is a convenient and advantageous model system for evaluating hypothesized mechanisms of retinal cytogenesis.

Acknowledgements

The authors thank Vera McIlvain, Melinda Tyler, and Patrick Yurco for helpful discussions and comments on the manuscript, Hunt Howell (Coastal Marine Lab, University of New Hampshire, New Castle, NH) for supplying biological materials, Barry Knox (SUNY Upstate) for supplying the anti-rhodopsin antibody, Paul Linser (University of Florida, Gainesville, FL) for supplying the anti-zebrafish glutamine synthetase antibody, and Beth Pritts (Lemoyne College) and Julie Siegenthaler (SUNY Upstate) for technical advice. Supported in part by the Edward F. MacNichol Memorial Fund, SUNY Upstate.

References

1. Bayliss RIS. Thyroid Disease: the facts. New York: Oxford University Press; 1982.

2. McNabb FMA. Thyroid hormones. New York: Prentice Hall; 1992.

3. Tsai MJ, O'Malley BW. Molecular mechanisms of action of steroid/thyroid receptor superfamily members. Annu Rev Biochem 1994; 63:451-86.

4. Brent GA. Tissue-specific actions of thyroid hormone: insights from animal models. Rev Endocr Metab Disord 2000; 1:27-33.

5. Zhang J, Lazar MA. The mechanism of action of thyroid hormones. Annu Rev Physiol 2000; 62:439-66.

6. Gudernatsch J. Feeding experiments on tadpoles I. The influence of specific organs given as food on growth and differentiation. A contribution to the knowledge of organs with internal secretion. Wilhelm Roux Arch Entwicklungsmech Org 1912; 35:457-83.

7. Dodd M, Dodd J. The biology of metamorphosis. In: Moore JA, Lofts B, editors. Physiology of the amphibia. New York: Academic Press; 1976. p. 467-598.

8. Inui Y, Miwa S. Thyroid hormone induces metamorphosis of flounder larvae. Gen Comp Endocrinol 1985; 60:450-4.

9. Graf W, Baker R. Neuronal adaptation accompanying metamorphosis in the flatfish. J Neurobiol 1990; 21:1136-52.

10. Hoskins SG. Metamorphosis of the amphibian eye. J Neurobiol 1990; 21:970-89.

11. Denver RJ, Pavgi S, Shi YB. Thyroid hormone-dependent gene expression program for Xenopus neural development. J Biol Chem 1997; 272:8179-88.

12. Denver RJ. The molecular basis of thyroid hormone-dependent central nervous system remodeling during amphibian metamorphosis. Comp Biochem Physiol C Pharmacol Toxicol Endocrinol 1998; 119:219-28.

13. Tata JR. Amphibian metamorphosis as a model for studying the developmental actions of thyroid hormone. Biochimie 1999; 81:359-66.

14. Power DM, Llewellyn L, Faustino M, Nowell MA, Bjornsson BT, Einarsdottir IE, Canario AV, Sweeney GE. Thyroid hormones in growth and development of fish. Comp Biochem Physiol C Toxicol Pharmacol 2001; 130:447-59.

15. Dussault JH, Ruel J. Thyroid hormones and brain development. Annu Rev Physiol 1987; 49:321-34.

16. Shagam JY. Thyroid disease: an overview. Radiol Technol 2001; 73:25-40.

17. Harpavat S, Cepko CL. Thyroid hormone and retinal development: an emerging field. Thyroid 2003; 13:1013-9.

18. Marsh-Armstrong N, Huang H, Remo BF, Liu TT, Brown DD. Asymmetric growth and development of the Xenopus laevis retina during metamorphosis is controlled by type III deiodinase. Neuron 1999; 24:871-8.

19. Azadi S, Zhang Y, Caffe AR, Holmqvist B, van Veen T. Thyroid-beta2 and the retinoid RAR-alpha, RXR-gamma and ROR-beta2 receptor mRNAs; expression profiles in mouse retina, retinal explants and neocortex. Neuroreport 2002; 13:745-50.

20. Cossette SM, Drysdale TA. Early expression of thyroid hormone receptor beta and retinoid X receptor gamma in the Xenopus embryo. Differentiation 2004; 72:239-49.

21. Orozco A, Linser P, Valverde C. Kinetic characterization of outer-ring deiodinase activity (ORD) in the liver, gill and retina of the killifish Fundulus heteroclitus. Comp Biochem Physiol B Biochem Mol Biol 2000; 126:283-90.

22. Forrest D, Reh TA, Rusch A. Neurodevelopmental control by thyroid hormone receptors. Curr Opin Neurobiol 2002; 12:49-56.

23. Plate EM, Adams BA, Allison WT, Martens G, Hawryshyn CW, Eales JG. The effects of thyroxine or a GnRH analogue on thyroid hormone deiodination in the olfactory epithelium and retina of rainbow trout, Oncorhynchus mykiss, and sockeye salmon, Oncorhynchus nerka. Gen Comp Endocrinol 2002; 127:59-65.

24. Brown DD. The role of deiodinases in amphibian metamorphosis. Thyroid 2005; 15:815-21.

25. Beatty DD. Visual pigments and the labile scotopic visual system of fish. Vision Res 1984; 24:1563-73.

26. Browman H, Hawryshyn C. Retinoic acid modulates retinal development in the juveniles of a teleost fish. J Exp Biol 1994; 193:191-207.

27. Kelley MW, Turner JK, Reh TA. Ligands of steroid/thyroid receptors induce cone photoreceptors in vertebrate retina. Development 1995; 121:3777-85.

28. Ng L, Hurley JB, Dierks B, Srinivas M, Salto C, Vennstrom B, Reh TA, Forrest D. A thyroid hormone receptor that is required for the development of green cone photoreceptors. Nat Genet 2001; 27:94-8.

29. Forrest D. Twists in the tail-change-of-function mutations in thyroid hormone receptors. Endocrinology 2002; 143:2466-8.

30. Yanagi Y, Takezawa S, Kato S. Distinct functions of photoreceptor cell-specific nuclear receptor, thyroid hormone receptor beta2 and CRX in one photoreceptor development. Invest Ophthalmol Vis Sci 2002; 43:3489-94.

31. Flamant F, Samarut J. Thyroid hormone receptors: lessons from knockout and knock-in mutant mice. Trends Endocrinol Metab 2003; 14:85-90.

32. Mader MM, Cameron DA. Photoreceptor differentiation during retinal development, growth, and regeneration in a metamorphic vertebrate. J Neurosci 2004; 24:11463-72.

33. Roberts MR, Hendrickson A, McGuire CR, Reh TA. Retinoid X receptor (gamma) is necessary to establish the S-opsin gradient in cone photoreceptors of the developing mouse retina. Invest Ophthalmol Vis Sci 2005; 46:2897-904.

34. Evans BI, Fernald RD. Retinal transformation at metamorphosis in the winter flounder (Pseudopleuronectes americanus). Vis Neurosci 1993; 10:1055-64.

35. Evans BI, Harosi FI, Fernald RD. Photoreceptor spectral absorbance in larval and adult winter flounder (Pseudopleuronectes americanus). Vis Neurosci 1993; 10:1065-71.

36. Miwa S, Inui Y. Effects of various doses of thyroxine and triiodothyronine on the metamorphosis of flounder (Paralichthys olivaceus). Gen Comp Endocrinol 1987; 67:356-63.

37. Schreiber AM, Specker JL. Metamorphosis in the summer flounder (Paralichthys dentatus): stage-specific developmental response to altered thyroid status. Gen Comp Endocrinol 1998; 111:156-66.

38. Hitchcock PF, Lindsey Myhr KJ, Easter SS Jr, Mangione-Smith R, Jones DD. Local regeneration in the retina of the goldfish. J Neurobiol 1992; 23:187-203.

39. Cameron DA, Easter SS Jr. Cone photoreceptor regeneration in adult fish retina: phenotypic determination and mosaic pattern formation. J Neurosci 1995; 15:2255-71.

40. Cameron DA. Asymmetric retinal growth in the adult teleost green sunfish (Lepomis cyanellus). Vis Neurosci 1995; 12:95-102.

41. Cameron DA, Carney LH. Cell mosaic patterns in the native and regenerated inner retina of zebrafish: implications for retinal assembly. J Comp Neurol 2000; 416:356-67.

42. Yurco P, Cameron DA. Responses of Muller glia to retinal injury in adult zebrafish. Vision Res 2005; 45:991-1002.

43. Cameron DA, Gentile KL, Middleton FA, Yurco P. Gene expression profiles of intact and regenerating zebrafish retina. Mol Vis 2005; 11:775-91 <http://www.molvis.org/molvis/v11/a93/>.

44. Douglas SE, Gallant JW, Bullerwell CE, Wolff C, Munholland J, Reith ME. Winter Flounder Expressed Sequence Tags: Establishment of an EST Database and Identification of Novel Fish Genes. Mar Biotechnol (NY) 1999; 1:458-64.

45. Siegenthaler JA, Miller MW. Transforming growth factor beta1 modulates cell migration in rat cortex: effects of ethanol. Cereb Cortex 2004; 14:791-802.

46. Yamano K, Araki K, Sekikawa K, Inui Y. Cloning of thyroid hormone receptor genes expressed in metamorphosing flounder. Dev Genet 1994; 15:378-82.

47. Yamano K, Inui Y. cDNA cloning of thyroid hormone receptor beta for the Japanese flounder. Gen Comp Endocrinol 1995; 99:197-203.

48. Yazulla S, Studholme KM. Neurochemical anatomy of the zebrafish retina as determined by immunocytochemistry. J Neurocytol 2001; 30:551-92.

49. Marchand O, Duffraisse M, Triqueneaux G, Safi R, Laudet V. Molecular cloning and developmental expression patterns of thyroid hormone receptors and T3 target genes in the turbot (Scophtalmus maximus) during post-embryonic development. Gen Comp Endocrinol 2004; 135:345-57.

50. Frankton S, Harvey CB, Gleason LM, Fadel A, Williams GR. Multiple messenger ribonucleic acid variants regulate cell-specific expression of human thyroid hormone receptor beta1. Mol Endocrinol 2004; 18:1631-42.

51. Suzuki S, Kaneko A. Identification of bipolar cell subtypes by protein kinase C-like immunoreactivity in the goldfish retina. Vis Neurosci 1990; 5:223-30.

52. Sjoberg M, Vennstrom B, Forrest D. Thyroid hormone receptors in chick retinal development: differential expression of mRNAs for alpha and N-terminal variant beta receptors. Development 1992; 114:39-47.

53. Chin WW. Molecular mechanisms of thyroid hormone action. Thyroid 1994; 4:389-93.

54. Gilbert SF. Developmental Biology. 5th ed. Sunderland (MA): Sinauer Associates; 1997.

55. Johns PR, Fernald RD. Genesis of rods in teleost fish retina. Nature 1981; 293:141-2.

56. Raymond PA, Rivlin PK. Germinal cells in the goldfish retina that produce rod photoreceptors. Dev Biol 1987; 122:120-38.

57. Cook T, Pichaud F, Sonneville R, Papatsenko D, Desplan C. Distinction between color photoreceptor cell fates is controlled by Prospero in Drosophila. Dev Cell 2003; 4:853-64.

58. Schreiber AM, Das B, Huang H, Marsh-Armstrong N, Brown DD. Diverse developmental programs of Xenopus laevis metamorphosis are inhibited by a dominant negative thyroid hormone receptor. Proc Natl Acad Sci U S A 2001; 98:10739-44.

59. Luo W, Williams J, Smallwood PM, Touchman JW, Roman LM, Nathans J. Proximal and distal sequences control UV cone pigment gene expression in transgenic zebrafish. J Biol Chem 2004; 279:19286-93.

60. Mani SS, Batni S, Whitaker L, Chen S, Engbretson G, Knox BE. Xenopus rhodopsin promoter. Identification of immediate upstream sequences necessary for high level, rod-specific transcription. J Biol Chem 2001; 276:36557-65.

61. Whitaker SL, Knox BE. Conserved transcriptional activators of the Xenopus rhodopsin gene. J Biol Chem 2004; 279:49010-8.

62. Peng GH, Ahmad O, Ahmad F, Liu J, Chen S. The photoreceptor-specific nuclear receptor Nr2e3 interacts with Crx and exerts opposing effects on the transcription of rod versus cone genes. Hum Mol Genet 2005; 14:747-64.

63. Dann SG, Ted Allison W, Veldhoen K, Johnson T, Hawryshyn CW. Chromatin immunoprecipitation assay on the rainbow trout opsin proximal promoters illustrates binding of NF-kappaB and c-jun to the SWS1 promoter in the retina. Exp Eye Res 2004; 78:1015-24.

64. Chambers RC, Leggett WC. Size and age at metamorphosis in marine fishes: an analysis of laboratory-reared winter flounder (Pseudopleuronectes americanus) with a review of variation in other species. Canadian journal of fisheries and aquatic sciences. 1987; 44:1936-47.

65. Otteson DC, Hitchcock PF. Stem cells in the teleost retina: persistent neurogenesis and injury-induced regeneration. Vision Res 2003; 43:927-36.