Direct Effect of Sodium Iodate on Neurosensory Retina

Abstract

Purpose.: To systematically characterize the effects of NaIO₃ on retinal morphology and function.

Methods.: NaIO₃ at 10, 20, or 30 mg/kg was administered by retro-orbital injection into adult C57BL/6J mice. Phenotypic and functional changes of the retina were assessed at 1, 3, 5, and 8 days postinjection by fundus imaging, optical coherence tomography (OCT), ERG, and histology. Direct NaIO₃ cytotoxicity on ARPE-19 and 661W cells was quantified using lactate dehydrogenase (LDH) apoptosis assay. Effect of NaIO₃ on RPE and photoreceptor gene expression was assessed in vitro and in vivo by quantitative PCR.

Results.: While little to no change was observed in the 10 mg/kg NaIO₃-injected group, significant retinal anomalies, such as RPE atrophy and retinal thinning, were observed in both 20 and 30 mg/kg NaIO₃-injected groups. Gene expression analysis showed rapid downregulation of RPE-specific genes, increase in heme oxygenase 1 expression, and induction of the ratio of Bax to Bcl-2. Electroretinographic response loss and photoreceptor gene repression preceded gross morphological changes. High NaIO₃ toxicity on 661W cells was observed in vitro along with reactive oxygen species (ROS) induction. NaIO₃ treatment also disrupted oxidative stress, phototransduction, and apoptosis gene expression in 661W cells. Exposure of ARPE-19 cells to NaIO₃ increased expression of neurotrophins and protected photoreceptors from direct NaIO₃ cytotoxicity.

Conclusions.: Systematic characterization of changes associated with NaIO₃ injection revealed a large variability in the severity of toxicity induced. Treatment with >20 mg/kg NaIO₃ induced visual dysfunction associated with rapid suppression of phototransduction genes and induced oxidative stress in photoreceptors. These results suggest that NaIO₃ can directly alter photoreceptor function and survival.

Introduction

Retinal degeneration, a characteristic of neurodegenerative diseases such as age-related macular degeneration (AMD) or retinitis pigmentosa, can lead to severe visual impairment and eventually blindness. Millions of people worldwide suffer varying degrees of irreversible vision loss because of these untreatable, degenerative eye disorders. To understand the molecular mechanisms of disease induction and progression, and to develop therapeutic strategies for vision preservation in these patients, extensive research has been done using a large number of both inherited and induced animal models. ^1,2 However, the mechanisms of formation/development of these diseases remain poorly understood. Systematic delivery of sodium iodate (NaIO₃), a stable oxidizing agent, has been proven to be an effective way to induce retinal degeneration associated with regional loss of retinal pigment epithelium (RPE) recapitulating some of the morphological features of geographic atrophy. ^3–5 NaIO₃ retinal toxicity has been demonstrated in many different mammalian species, including sheep, ⁶ rabbit, ^7,8 rat, ^9–11 and mouse. ^12–14 In the retina, NaIO₃ is thought to target primarily the RPE cells, ^6,15 inducing their necrosis, followed by choriocapillaris atrophy ¹⁶ and panretinal degeneration. ^10,15 Despite the fact that the NaIO₃-dependent model of retinal degeneration has been previously investigated ^9,17–19 and used for the evaluation of neuroprotective treatments, ^11,20 the acute effect of NaIO₃ toxicity remains poorly characterized. Most studies used relatively high doses of NaIO₃ (from 50 to 100 mg/kg) and reported rapid RPE damage characterized by defragmentation and loss of RPE cell nuclei at ∼2 to 12 hours, ^15,21 followed by disorganization of the rod outer segment discs at ∼6 hours ¹⁰ —results consistent with the concept that RPE cells are the primary target of NaIO₃ toxicity. However, this mechanism was recently challenged by a study showing that high doses of NaIO₃ are able to directly induce intraretinal neuron injury. ²² Because the model of NaIO₃ ocular toxicity is widely used to assess the efficacy of cell replacement therapy through transplantation, ^13,23,24 it is important to clarify whether and/or how NaIO₃ affects different retinal cells, whether or not this effect is direct or indirect, and, specifically, how NaIO₃ administration alters the gene expression profiles of photoreceptor (PR) and RPE cells. To date no report has characterized the early retinal function changes together with a precise, sequential, morphological observation in order to more precisely define this particular model of retinal degeneration. In the present study, we conducted a systematic characterization of the early effects of low NaIO₃ concentration (<30 mg/kg) on the morphology, function, and gene expression of the murine retina, and demonstrated evidence of direct NaIO₃ toxicity on PR cells both in vivo and in vitro. 

Materials and Methods

Animals

Six- to eight-week-old male C57BL/6J mice (Jackson Laboratory, Bar Harbour, ME) were used in this study. The mice were housed in a standard laboratory environment and maintained on a 12-hour light–dark cycle at 21°C. All experimental procedures involving animals were performed in accordance with the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research and were approved by the local institutional animal care and use committee. 

NaIO₃ Injection

A sterile 1% NaIO₃ solution was freshly prepared from solid NaIO₃ (S4007; Sigma St. Louis, MO) diluted in 0.9% sterile sodium chloride. Mice were anesthetized with isoflurane/oxygen and injected via the orbital venous plexus with three different doses of NaIO₃: 10, 20, and 30 mg/kg body weight. Control mice were injected with similar volumes of physiological saline. 

Electroretinography (ERG)

Following 6 hours of dark adaptation, mice were anesthetized and prepared under red light. Eyes were dilated with 1% tropicamide solution (Bausch and Lomb, Tampa, FL). Ground and reference electrodes were placed subdermally near the tail and in the forehead, respectively. Scotopic and phototopic ERGs were recorded from gold loop electrodes placed on the cornea using the Espion E ³ machine (Diagnosys, Boxborough, MA). Scotopic responses were elicited using a graded series of flashes from −4 to 2 log cd*s/m², with increasing interflash intervals for each step. Before photopic ERG recordings, mice were light adapted using a series of bright white flashes. Photopic responses were elicited using a graded series of flashes from −1 to 3 log cd*s/m² presented on a constant 30 cd*s/m² amber background, with increasing interflash intervals for each step. All ERG data were collected at the same time of day. The a-wave amplitude was measured from the baseline to the trough of the first negative wave; the b-wave amplitude was measured from the trough of the a-wave to the peak of the first positive wave or, if the a-wave was absent, from baseline to the peak of the first positive wave. The scotopic a-wave implicit time and the ratio of b- to a-wave were calculated at 500 cd*s/m². 

Spectral Domain Optical Coherence Tomography (SD-OCT) and Fundus Photography

Ultrahigh-resolution SD-OCT images centered on the optic nerve head (ONH) were acquired on live anesthetized animals using an SD-OCT system (Bioptigen, Morrisville, NC). Fundus images were obtained using the Micron III retinal imaging system (Phoenix Research Laboratories, Pleasanton, CA). 

Histology and Immunofluorescence

For histology, whole eyes were enucleated, fixed in 10% formalin, and prepared for paraffin embedding; then 3-μm-thick slides were sectioned. The thickness of the retinal layers and nuclei/row numbers were measured on hematoxylin- and eosin-stained slides. For each section, photographs were taken at a distance of 1000 to 1500 μm from the optic nerve (midperiphery). Three measurements were made per photograph and averaged. For immunohistochemistry of the posterior eye cups, the eyes were fixed in 4% paraformaldehyde overnight, and the neural retina was dissected out. The remaining RPE cell–containing eye cups were flattened by making long radial cuts and incubated in a blocking buffer containing 1% BSA, 3% goat serum in 1× PBS (vol/vol) and 0.1% Triton X-100 for 2 hours at room temperature (RT). Retinal pigment epithelium–choroid flat mounts were incubated for 2 hours at RT with rabbit anti-RPE65 primary antibody (1:300; a gift from Michael Redmond, National Eye Institute, Bethesda, MD). The secondary antibody Alexa Fluor 488-conjugated anti-rabbit antibody (A11034, 1:500; Invitrogen, Carlsbad, CA) was added with Alexa 586-phalloidin (1:100; Invitrogen) to stain F-actin. After mounting, images were taken with an Axioscope Mot 2 microscope (Carl Zeiss Meditec, Inc., Dublin, CA). 

661W and ARPE-19 Cell Cultures In Vitro

Murine 661W PR-derived cells were plated at 15,000 cells/cm² in 12-well plates and cultured in regular Dulbecco's modified Eagle's medium (DMEM) medium containing 10% fetal bovine serum (FBS; Atlanta Biologicals, Lawrenceville, GA) for 36 hours before the experiments were performed. The human RPE cell line ARPE-19 was expanded in DMEM plus 10% FBS in plastic culture dishes. Once cells reached confluence, medium was replaced with RPE differentiation medium (DMEM plus 1% FBS) for another 5 days before the experiments were performed. For conditioned medium collection, ARPE-19 cells were cultured in serum-free, phenol-free medium with or without 250 μg/mL NaIO₃ for 24 hours. 

Lactate Dehydrogenase Assay

Cytoplasmic enzyme lactate dehydrogenase (LDH) was measured in cell culture supernatants by a colorimetric cell death assay (Roche Molecular Diagnostics, Pleasanton, CA). The LDH assays were performed using 661W or ARPE-19 cell cultures grown to confluence in 48-well plates, and LDH was measured after 250, 500, 750, 1000, or 3000 μg/mL NaIO₃ for 24 hours in serum-free, phenol-free medium. After collection, the supernatants were centrifuged at 250g for 10 minutes to remove suspended cells. The supernatants were then assayed at a 1:4 dilution for a final volume of 100 μL/well. Absorbance was measured at 490 nm using a microplate reader (Synergy2; BioTek, Winooski, VT). For each experiment, each condition was replicated six times; and background levels, determined using media not exposed to cells, were subtracted from absorbance values obtained for each condition. Lactate dehydrogenase values were normalized to the mean maximal value (defined as 100%) in parallel cultures exposed to 2% Triton X-100 for 3 hours—conditions that resulted in the death of all cells. 

Measurement of Reactive Oxygen Species (ROS) Production

The levels of intracellular ROS were measured using the fluorescent probe 5-(and-6)-chloromethyl-2′,7′-dichlorodihydrofluorescein diacetate (CM-H₂DCFDA) following the manufacturer's instructions (C6827; Invitrogen). In brief, 661W cells were incubated with 10 μM CM-H₂DCFDA for 30 minutes at 37°C, then washed in serum-free, phenol red–free medium, before the addition of 250 μg/mL NaIO₃. Twenty-four hours later, DCF fluorescence was measured and corrected for cell number. 

RNA Isolation and Quantitative PCR (qPCR)

Total RNA from RPE cells was isolated using the simultaneous RPE cell isolation and RNA stabilization (SRIRS) method. ²⁵ Total RNA from the neuroretina was extracted using the RNeasy Mini Kit (Qiagen, Valencia, CA). Residual DNA was removed by treatment with DNase I (Qiagen) during the extraction procedures. Ribonucleic acid (1 μg) was reverse transcribed using the iScript cDNA synthesis kit (Bio-Rad, Mississauga, ON). Real-time qPCR was performed using the SYBR green Master Mix and the Lightcycler 480 (Roche Applied Science, Indianapolis, IN) according to the manufacturer's instructions. To ensure accurate gene quantification, the PCR data were normalized to the mean of multiple internal control genes (Hprt1 and B2m), ²⁶ and gene expression was calculated according to the ΔΔC_T method. Each sample was run in triplicate, and each experiment included three nontemplate control wells. The genes investigated by qPCR are listed in Table 1, and the respective primers used are listed in Table 2. 

Table 1

 

List of Genes Studied

Symbol	Name	Function
Hprt1	Hypoxanthine phosphoribosyl-transferase 1	Housekeeping
B2m	Beta-2-microglobulin	Housekeeping
Rpe65	Retinal pigment epithelium-specific protein 65kDa	Visual cycle
Opn1mw	Opsin 1 (cone pigments), medium wave sensitive	Color vision
Pde6b	Phosphodiesterase 6B, cGMP-specific, rod, beta	Phototransduction
Rho	Rhodopsin	Phototransduction
Prph2	Peripherin 2	Phototransduction
Arr3	Arrestin 3	Phototransduction
Gnat2	Guanine nucleotide binding protein, alpha transducing activity polypeptide 2	Phototransduction
Cnga1	Cyclic nucleotide gated channel alpha 1	Phototransduction
Cnga3	Cyclic nucleotide gated channel alpha 3	Phototransduction
Cngb3	Cyclic nucleotide gated channel beta 3	Phototransduction
Rora	RAR-related orphan receptor A	Transcription factor for photoreceptor development and functions
Rorb	RAR-related orphan receptor B	Transcription factor for photoreceptor development and functions
Bax	Bcl2-asociated X protein	Proapoptotic
Bcl2	B-cell CLL/lymphoma 2	Antiapoptotic
Gpx1	Glutathione peroxidase 1	Antioxidant
Gpx4	Glutathione peroxidase 4	Antioxidant
Sod2	Superoxide dismutase 2, mitochondrial	Antioxidant
Hmox1	Heme oxygenase 1	Heme catabolism (inducible)
Hmox2	Heme oxygenase 2	Heme catabolism (constitutive)
Txn2	Thioredoxin 2	Antioxidant
Pspn	Persephin	Neurotrophic factor
Ctf1	Cardiotrophin 1	Neurotrophic factor
Bdnf	Brain-derived neurotrophic factor	Neurotrophic factor
Ngf	Nerve growth factor	Neurotrophic factor
Vegf	Vascular endothelial growth factor	Neurotrophic factor
Pedf	Pigment epithelium-derived factor	Neurotrophic factor
NT4	Neurotrophin-4	Neurotrophic factor
Gdnf	Glial cell-derived neurotrophic factor	Neurotrophic factor
Tgfb1	Transforming growth factor beta 1	Neurotrophic factor
Igf1	Insulin-like growth factor 1	Neurotrophic factor
Lif	Leukemia inhibitory factor	Neurotrophic factor

Table 2

 

Primer Sequences for qPCR Analysis

Symbol	Forward Primer (5′–3′)	Reverse Primer (5′–3′)
Hprt1	GTTAAGCAGTACAGCCCCAAA	AGGGCATATCCAACAACAAACTT
B2m	GTGGCCCTTAGCTGTGCTCG	ACCTGAATGCTGGATAGCCTC
Opn1mw	TGCAAGCCCTTTGGCAATG	GCCGTCCATATAGCAGCCC
Pde6b	TGAAGATGAAGATGTTTTCACG	CTCTGCGTGTCTCACAGTTG
Rho	GCCACACTTGGAGGTGAAATC	AAGCGGAAGTTGCTCATCG
Prph2	CGGGACTGGTTCGAGATTC	ATCCACGTTGCTCTTGATGC
Arr3-2	CCATTGATGGAGTCGTCCTT	CTTTGCGAAATGTCAGACCA
Gnat2	CTGTCAAGCTGCTGTTGCTC	CTAGGCACTCTTCGGGTGAG
Cnga1	ACTCGTACAAAAGGCGAGGAC	CTTTGTTGCTGCTGTTGTTGAC
Cnga3	CCTTCGATAAGCAGGCATTG	GGTGTTCACCTTTGCCATCT
Cngb3	GACTGGGAAAGAAAGGCCAG	AGGTTGTGTAGGTGGGCATC
Rora	AGGCAGAGCTATGCGAGC	TTCCTGACGATTTGTCTCCA
Rorb	ATGCCAGCTGATGGAGTTCT	TAGCTCCCGGGATAACAATG
Bax	GATCAGCTCGGGCACTTTAG	TTGCTGATGGCAACTTCAAC
Bcl2	CCGGGAGAACAGGGTATGATAA	CCCACTCGTAGCCCCTCG
Gpx1	GTCTGGGACCTCGTGGACT	CAGGTCGGACGTACTTGAGG
Gpx4	CCGTCTGAGCCGCTTACTTA	GCTGAGAATTCGTGCATGG
Sod2	CAGACCTGCCTTACGACTATGG	CTCGGTGGCGTTGAGATTGTT
Hmox1	GAGCCTGAATCGAGCAGAAC	CCTTCAAGGCCTCAGACAAA
Hmox2	AGCACATGACCGAGCAGAAAA	GCTCCGTGGGGAAATATAAGGG
Txn2	CACACAGACCTTGCCATTGA	ATCCCCACAAACTTGTCCAC
PSPN	GTCTGAACAGGTGGCAAAGG	ACAGGGTCAGGCTCCACAG
CTF1	CGGAGGGAGGGAAGTCTG	AGGCTGTGTGTCTGACGGAT
VEGFA	GGGCAGAATCATCACGAAGTG	ATTGGATGGCAGTAGCTGCG
PEDF	TATCACCTTAACCAGCCTTTCATC	GGGTCCAGAATCTTGCAATG
NT4	CCTCCGCCAGTACTTCTTTGA	CCGGGCCACCTTCCTC
GDNF	GACTCAAATATGCCAGAGGTTATCC	GGTGGCTTGAATAAAATCCATGAC

 Rpe65, BDNF, and NGF primers were ordered from SABiosciences (Valencia, CA).

Statistical Methods

Statistical comparisons were made by Student's two-tailed t-test. Significance was set at P < 0.05, and the data are presented as the mean ± SEM unless otherwise specified (*P < 0.05, **P < 0.01, ***P < 0.001). 

Results

Morphological Changes in Retinal Tissue After NaIO₃ Administration

In order to evaluate the ocular effects of a low-dose NaIO₃ injection (i.e., 10 mg/kg), fundus photographs at days 1, 3, 5, and 8 were collected. No obvious changes were found for the 10 mg/kg group at any time point evaluated (Fig. 1, top) when compared to saline-injected controls (see Supplementary Fig. S1A). Higher doses (i.e., 20 and 30 mg/kg) were associated with progressive retinal degeneration characterized by significant pigmentary changes of the RPE layer at day 3 (Fig. 1, middle and bottom). The RPE pigmentary disruption and atrophy quickly became more widespread in the following days. Retinal vasculature and optic nerve tissue appeared to remain normal at all examined time points and concentrations. 

Figure 1

 

Color fundus photographs of C57BL/6J mice treated with increasing doses of NaIO₃. Top, middle, bottom: 10, 20, 30 mg/kg NaIO₃ groups, respectively. Left to right: fundus at days 1, 3, 5, and 8, respectively. Estimated scale bar based on mouse optic nerve diameter: 220 μm.

Spectral domain OCT evaluation confirmed a lack of obvious anatomical changes in mice treated with the lower doses of NaIO₃ (Fig. 2) when compared with control animals (see Supplementary Fig. S1B). However, higher concentrations of NaIO₃ injected were associated with loss of retinal lamination at day 3 and significant outer retina damage by day 5. Interestingly, hyperreflective opacities were observed in both the vitreous and neuroretina for both the 20 and 30 mg/kg groups as early as 1 day postinjection. Inside retinal tissue, these hyperreflective foci are likely to represent glia cell migration or the presence of phagocytic retinal cells. The origin of the vitreal opacities is unknown but could represent migratory inflammatory cells. However, no sign of acute ocular inflammation was observed by fundus photography (Fig. 1). Loss of lamination was also accompanied by significant total retinal thinning, with prominent loss of the outer nuclear layer (ONL) and the outer segment (OS) layers by day 8 (Fig. 2). 

Figure 2

 

OCT of C57BL/6J mice treated with increasing doses of NaIO₃. Top, middle, bottom: 10, 20, 30 mg/kg NaIO₃ groups, respectively. Left to right: days 1, 3, 5, and 8, respectively. Arrows point to hyperreflective spots in both the vitreous (green) and the retina (yellow). Red double-headed arrows mark total retinal thickness. Scale bar: 100 μm.

Results from fundus and OCT studies were confirmed by histological examination of ocular specimens. Retinal pigment epithelium swelling and migration of pigmented cells into the OS layer were observed for both the 20 and 30 mg/kg-injected groups as early as day 3 (Fig. 3). Cystic spaces in the inner nuclear layer were also observed; however, they were not present in either the control retinas (see Supplementary Fig. S1C) or the lower NaIO₃-dose groups. By day 5, disorganization of PRs and significant thinning of the entire retina, especially the ONL, were observed along with large regions of complete RPE loss, which had expanded over time (Fig. 3). 

Figure 3

 

Hematoxylin and eosin staining of paraffin-embedded retinal cross sections of NaIO₃-treated C57BL/6J mice at different time points. Top, middle, bottom: 10, 20, 30 mg/kg NaIO₃-injected group, respectively. Left to right: days 1, 3, 5, and 8, respectively. Starting from day 3, swelling and bundling of the RPE cells (arrowheads) associated with macrophage migration and photoreceptor disorganization (arrows) were observed. Significant loss of RPE cells and ONL thinning were detected at days 5 and 8 (double-headed arrows). Scale bar: 100 μm.

To better characterize changes of the RPE monolayer, we used immunostaining for F-actin and RPE65 on flat-mounted posterior eye cups. There were no observable changes in the organization or RPE65 expression of the RPE layer of mice treated with 10 mg/kg NaIO₃ (Figs. 4A–D). Retinal pigment epithelium cells presented their normal cuboidal shape and even distribution, with robust expression of RPE65 in both the central and peripheral retina—similar to the appearance of RPE cells in normal saline–injected mice. Day 8 following the 20 mg/kg NaIO₃ injection was associated with dramatic RPE atrophy in large areas of the retina. In the peripheral region, many of the remaining cells appeared significantly enlarged and disorganized, and had lost RPE65 expression (Figs. 4E, 4F). Groups treated with 30 mg/kg NaIO₃ demonstrated similar results (data not shown). 

Figure 4

 

Immunofluorescence of RPE/choroid flat mounts 8 days post-NaIO₃ injection. F-actin was stained with phalloidin (red) and RPE65 (green). (A, B) Saline-injected control eyes displayed normal RPE morphology with evenly distributed RPE65 expression in the central (A) and peripheral retina (B). (C, D) 10 mg/kg NaIO₃-treated RPE flat mounts were indistinguishable from saline control (E, F). In 20 mg/kg NaIO₃-treated animals, RPE cells appeared highly disorganized in the central retina (E). Significantly enlarged RPE cells characterized by strongly reduced RPE65 expression were observed in the peripheral retina (F). OD, optic disc. Scale bar: 100 μm.

Characterization of PR Degeneration and Functional Defects Following NaIO₃ Treatment

Quantification of ONL, inner and outer segment layer thicknesses (IS/OS), and ONL nuclei/row demonstrated significant and progressive PR degeneration by day 5 in mice treated with either 20 or 30 mg/kg NaIO₃; this led to over ∼50% thinning of the ONL by day 8 (Fig. 5). Retinal function changes were examined by scotopic and photopic ERG 1 week before the NaIO₃ injections at baseline and at 1, 3, 5, and 8 days postinjection. As we observed through our anatomical examination, the 10 mg/kg injection group maintained a normal ERG response (Fig. 6). Surprisingly, however, an attenuated ERG response, characterized by a reduction of both scotopic and photopic b-wave amplitudes, was observed as early as 1 day post-NaIO₃ injection in both the 20 and 30 mg/kg groups, preceding any morphological indication of PR damage (Figs. 6B, 6C). These rapid ERG changes were also characterized by a reduction of the ratio of b to a amplitude (Fig. 6D) and a delayed a-wave implicit time (Fig. 6E). Moreover, visual function decreased over the experimental time course and, by day 8, both scotopic (a- and b-wave) and photopic (b-wave) responses were near zero (Fig. 6). 

Figure 5

 

Altered outer retina thickness following NaIO₃ administration. Outer nuclear layer ([A], ONL), photoreceptor inner and outer segment layer thicknesses ([B], IS/OS), and ONL nuclei/row numbers (C) were quantified on histological sections. Asterisks indicate the statistically significant deviation of thickness from the control. **P < 0.01.

Figure 6

 

Longitudinal analysis of visual function of mice treated with NaIO₃. (A) Scotopic a-wave recorded at 500 cd*s/m². (B) Scotopic b-wave recorded at 0.1076 cd*s/m². (C) Saturating photopic b-wave recorded at 10³ cd*s/m². Significant reduction of all ERG responses was observed at day 1 in both the 20 and 30 mg/kg NaIO₃ groups. Both a- and b-waves continued to decrease over time, and by day 8, all ERG responses were abolished. No functional change was detected for the 10 mg/kg NaIO₃ group at any time point. At day 1, a significant reduction of the ratio of b to a amplitude (D) and increase of the a-wave implicit time (E) in animals treated with 20 and 30 mg/kg NaIO₃ were also observed. The results are presented as mean ± SEM. ‡,¶ or *P < 0.05, ‡‡,¶¶ or **P < 0.01, ‡‡‡,¶¶¶ or ***P < 0.001.

Induction of Oxidative Stress and Phototransduction Gene Dysregulation by NaIO₃

The acute effect of NaIO₃ on the expression profile of oxidative stress, cell death, and phototransduction-relative genes was measured using qPCR on isolated RPE and neuroretina (free of RPE) samples 1 and 3 days postinjection of 10 or 20 mg/kg NaIO₃. Treatment with 20 mg/kg NaIO₃ was associated with a dramatic loss of Rpe65 expression within 24 hours, followed by repression of both Mitf and Otx2 expression at day 3 (Fig. 7A). This acute NaIO₃ effect on RPE65 gene expression was also accompanied by the induction of Hmox1 (iron-generated ROS antioxidant) expression and increase of ratio of Bax to Bcl2 (indicative of increased apoptosis) at day 1 (Figs. 7C, 7D), a time point at which no major RPE morphological changes were observed (Figs. 1, 3). Two other markers of oxidative stress, Txn2 and Gpx4, were also found at slightly decreased levels at day 3 in the RPE cells, while other ROS markers, such as Sod2, showed no significant changes (data not shown). Gene analysis of the neuroretina revealed a rapid dysregulation of many phototransduction genes, such as Rho, Opn1mw, and Pde6b, following treatment with 20 mg/kg NaIO₃ (Fig. 7B). Similarly, as observed in the RPE cells, both Hmox1 expression and the ratio of Bax to Bcl2 were strongly induced in neuroretina treated with 20 mg/kg NaIO₃. However, no changes in Gpx4 and Sod2 expression were detected (data not shown). Intriguingly, treatment with 10 mg/kg NaIO₃ was associated with induction of Rpe65, Mitf, and Otx2 expression at day 3 (Fig. 7A). The same induction trend was observed in the neuroretina with the phototransduction genes Rho, Pde6b (Fig. 7B), and Cnga1 (data not shown). 

Figure 7

 

Gene expression changes in mouse RPE and neuroretina following NaIO₃ treatment. RPE cells and neuroretina were isolated from mice treated with 10 or 20 mg/kg NaIO₃ for 1 or 3 days. (A) Expression of Rpe65, Mitf, and Otx2 from mouse RPE cells. (B) Expression of phototransduction-related genes from mouse neuroretinas. (C, D) Expression of Hmox1 (C) and the apoptosis marker ratio of Bax to Bcl2 (D) in mouse RPE cells and neuroretinas, at days 1 and 3, respectively. The results are presented as mean ± SEM. *P < 0.05, **P < 0.01, ***P < 0.001. RE, relative expression.

Characterization of a Direct Toxic Effect of NaIO₃ on PRs

The observation of a major alteration of the phototransduction pathway in the neuroretina of animals treated with >20 mg/kg NaIO₃, together with the identification of ERG anomalies, suggests that NaIO₃ could exert a direct toxic effect on PRs. This is supported by our observations that these changes occurred at days 1 to 3 (Figs. 6, 7), preceding a measurable PR loss at day 5 (Fig. 5). To test this hypothesis, we utilized the murine cone-like 661W cell line and characterized the effect of experimentally relevant doses of NaIO₃. In a serum-containing culture system, NaIO₃ treatment led to dose-dependent 661W cell death, as measured by LDH release, and was detectable only after high doses of NaIO₃ (above 750 μg/mL). However, significant cytotoxicity was detected in the serum-free culture system at doses as low as 250 μg/mL (Fig. 8A). This NaIO₃-dependent cell death was also associated with a robust increase in ROS production (Fig. 8B). Similar to our observations in vivo, gene expression analysis revealed a rapid increase of the oxidative stress–related genes Hmox1 and Gpx1 along with a strong induction of the ratio of Bax to Bcl2 (Fig. 8E). Furthermore, most of the cone phototransduction-related genes tested were significantly downregulated (Fig. 8D), which is also consistent with our in vivo data. 

Figure 8

 

Effect of NaIO₃ on 661W cells in vitro. (A) 661W cell death secondary to NaIO₃ treatment was quantified by assessing LDH activity in serum-containing and serum-free culture media. NaIO₃ toxicity appears strongly dependent on serum level, as serum-free media condition elicited a much higher LDH release. (B) 24-hour treatment of 661W cells with 250 μg/mL NaIO₃ in a serum-free condition was also associated with increased ROS production. (C, D) Gene expression analysis on 661W cells treated with 250 μg/mL NaIO₃ showed upregulation of oxidative stress (C) and downregulation of phototransduction-related genes (D). (E) Expression ratio of Bax to Bcl2 was strongly induced by NaIO₃ treatment. The results are presented as mean ± SEM. *P < 0.05, **P < 0.01, ***P < 0.001 compared to controls. RE, relative expression.

In contrast, results of similar experiments using the human RPE cell line ARPE-19 indicated that RPE cells were less vulnerable to NaIO₃-induced cell death than were 661W cells. Moreover, only high concentrations of NaIO₃ (>750 μg/mL) induced a significant, but low, cytotoxicity, although an excessive concentration (3000 μg/mL) led to 80% cell death (Fig. 9A). To identify if the paracrine interaction between RPE cells and PRs could affect PR sensitivity to NaIO₃ in a context similar to in vivo conditions, we exposed 661W cells to conditioned media (CM) collected from ARPE-19 cells precultured with or without 250 μg/mL NaIO₃. Interestingly, conditioning the NaIO₃-containing media on ARPE-19 cells significantly decreased (by 50%) NaIO₃ toxicity on 661W cells (Fig. 9B). Through gene expression analysis, we observed that only a few known neurotrophic factors released from RPE cells—cardiotrophin 1 (CTF1), persephin (PSPN), and brain-derived neurotrophic factor (BDNF)—were significantly increased after NaIO₃ treatment (Fig. 9C). Expression of other neurotrophic factors analyzed, which included NGF, VEGF, PEDF, N4, GDNF (Fig. 9C), TGFβ1, IGF1, and LIF (data not shown), were not affected by NaIO₃ treatment. 

Figure 9

 

Effect of NaIO₃ on RPE cell neurotrophic function. (A) ARPE-19 cells are resistant to direct NaIO₃ cytotoxicity, as only concentrations over 750 μg/mL are associated with increased cell death. (B) Conditioned medium was transferred from either control ARPE-19 cells or 24-hour 250 μg/mL NaIO₃-treated ARPE-19 cells onto 661W cells for 24-hour culture, and comparison was made between 661W cells cultured with regular serum-free medium, medium with NaIO₃, and conditioned medium from either ARPE-19 cells or NaIO₃-treated ARPE-19 cells. CM, conditioned media collected from 24-hour ARPE-19 culture; CM+NaIO₃, conditioned media collected from 24-hour ARPE-19 culture treated with 250 μg/mL NaIO₃. (C) Expression of neurotrophic factors was significantly upregulated in ARPE-19 cells after 250 μg/mL NaIO₃ treatment at 24 hours. Asterisks indicate the statistically significant deviation of amplitude from the control. The results are presented as mean ± SEM. *P < 0.05, **P < 0.01, ***P < 0.001. RE, relative expression.

Discussion

In this study, we have reassessed the mechanisms of NaIO₃-induced retinal damage model in pigmented C57BL/6J mice. To this end, we examined the effects of different doses and postinjection time points on several aspects of both retinal structure and function. Our data showed that even low doses of NaIO₃ (20 or 30 mg/kg) led to significant loss of retinal function at day 1, followed by local morphological changes in RPE cells and PRs at day 3, and significant ONL thinning at day 5. Thus, we were able to detect functional deficits in the absence of overt morphological changes. 

Although it is difficult to compare results of the current study to previously published reports, due to differences in the source of NaIO₃, methods of administration, animals (type, strain, and age), and time points analyzed after injection, trends in changes can still be meaningfully compared. The retinal toxicity following administration of high doses of NaIO₃ (30–100 mg/kg) was characterized by whole retina thinning associated with degeneration of RPE, rods, and cones, leading to the complete ablation of the outer nuclear and RPE layers. ²⁷ Under these conditions, cell toxicity was too rapid to allow unequivocal identification of the mechanisms leading to PR damage. 

NaIO₃ is traditionally considered to be highly specific to RPE cells. However, this concept was challenged by a recent study demonstrating that NaIO₃ injection in the rat leads to rapid and direct intraretinal neuron injury. ²² In our study, we observed acute PR dysfunction before notable changes to the RPE and ONL layers. Such function loss could be due to several factors. First, NaIO₃ treatment could cause synaptic damage, as a defect in synaptic transmission between the PRs and depolarizing bipolar cells can lead to a diminished ERG b-wave. ²⁸ This idea is supported by a recent study using SD-OCT to characterize early changes associated with injection of high doses of NaIO₃ (40 mg/kg) in the rat and demonstrating a rapid swelling of PR outer segments due to hydropic change, ⁹ although no significant change of the inner plexiform layer (IPL), where the synapses are located, was observed. ⁹ In our study, we were unable to accurately measure IPL thickness due to the ONL disorganization and loss of uniformity at the IPL, but swelling in the inner retina was clearly observed in histological sections after NaIO₃ treatment (Fig. 3). However, our observations of a rapid reduction of the ratio of b to a amplitude following NaOI₃ injection is in agreement with a defective synaptic transmission to the second-order neurons. Secondly, NaIO₃ could directly alter PR function and survival. We tested this hypothesis by studying both RPE cells and PRs in vivo and in vitro, with a special emphasis on oxidative stress and phototransduction processes. We observed that rhodopsin, opsin, Cnga1, and Pde6b were strongly downregulated secondary to NaIO₃ treatment both in vivo and in vitro. As these genes are critically involved in the visual function of the PR, such dysregulation might be sufficient to cause the early ERG response loss with the lengthened a-wave implicit time observed in NaIO₃-treated mice. 

We also observed an induction of the oxidative stress–related gene Hmox1 in PRs by 7-fold in vivo (Fig. 7C) and 2-fold in vitro (Fig. 8C). This result was further supported by the accumulation of ROS in 661W cells after NaIO₃ exposure. Involvement of excessive generation of ROS and induction of oxidative stress has been considered to be a major factor in PR apoptosis. ^29,30 Indeed, we measured a significant, directly toxic effect of NaIO₃ on 661W cells at a concentration (250 μg/mL) similar to that used in vivo (20 mg/kg or 342 μg/mL in blood). Interestingly, serum content in the culture media strongly affected the sensitivity of PR to NaIO₃, as serum-free conditions were associated with much higher percentages of cell death. It has been previously shown that FBS-containing culture medium can lessen the oxidative injury caused by the transition metal ion in pheochromocytoma cells, ³¹ likely due to the presence of many growth factors able to alter cell antioxidant responses. Together, our findings suggest that NaIO₃-induced oxidative stress in the neuroretina is able to dysregulate phototransduction gene expression and lead to subsequent PR cell death. 

In our study, the 10 mg/kg NaIO₃-injected group did not display any morphological or functional changes; however, gene expression analysis revealed that this low dose led to some modulation of RPE-specific (Rpe65, Mitf) and antioxidant (Hmox1) genes. In vivo, RPE cells are particularly exposed to oxidative stress due to their location between the choriocapillaris and PRs, high levels of light irradiation, and abundance of photosensitizers such as lipofuscin. Retinal pigment epithelium cells' endogenous antioxidant capability as a defense system allows them to withstand such oxidative conditions. ³² Moreover, induction of antioxidant enzymes secondary to oxidative stress in RPE has been well documented. ^33,34 However, the effect of oxidative stress on RPE-specific genes, especially RPE65, is less well understood. Treatment of ARPE-19 with cytotoxic doses of the oxidant t-butyl hydroperoxide (tBH) has been shown to inhibit RPE65 expression, ³⁵ however, not to the extent observed in our study. Controversially, RPE65 has been shown to be upregulated in RPE cells after short-term exposure to intense light, as a cellular oxidizer. ³⁶ Thus, RPE65 expression appears to be tightly controlled by the oxidative status of RPE cells. It is possible that noncytotoxic oxidative stress levels, such as those triggered by administration of 10 mg/kg NaIO₃, may be associated not only with induction of antioxidant enzymes but also with induction of other genes critical for RPE function such as Rpe65 and Mitf. However, excess oxidative stress associated with activation of apoptotic pathways could lead to major disruption of RPE gene expression including loss of Rpe65. Indeed, decreased Rpe65 expression secondary to intense photic injury was recently reported. ³⁷  

Our in vitro study clearly demonstrated that neuronal cells are more sensitive to NaIO₃ toxicity than are RPE cells. It is well established not only that RPE cells are highly resistant to oxidative stress, ³⁸ but also that one of their main functions is to support PR survival under homeostatic or pathologic oxidative stress conditions. Indeed, we observed that conditioning the NaIO₃-containing medium on ARPE-19 cells significantly reduced PR cell death, suggesting that RPE cells are able to process NaIO₃ and/or secrete neurotrophic factors in response. Further analysis confirmed that known neurotrophic factors, including CTF1, PSPN, and BDNF, were induced in RPE cells after NaIO₃ exposure. Retinal pigment epithelium cells are known for their ability to secrete a wide variety of growth factors ³⁹ that are able to support surrounding cells by paracrine and autocrine signaling and hence promote PR survival and retinal homeostasis. ^40,41 In addition to their survival effect on PRs, many RPE-secreted neurotrophins, including BDNF and PSPN, can promote RPE cell survival due in part to their synthesis of neuroprotectin D1, which counteracts the cytotoxic effect of various proapoptotic stressors such as the accumulation of A2E. ⁴² The neurotrophic function of CTF1 in the eye, unlike that of BDNF and PSPN, has not been evaluated. Although our data suggest that CTF1 is involved in RPE cells' protective role against direct NaIO₃ toxicity on PRs, further research is required to better characterize RPE-derived CTF1 functions. 

Taken together, our results indicate that intravenous administration of low-dose NaIO₃ triggers a direct effect on PRs, leading to rapid retinal dysfunction before the appearance of measurable RPE atrophy. This direct and early PR damage likely plays a causative role in the pathological events associated with NaIO₃-induced retinal degeneration. The identification of this new mechanism of retinal NaIO₃ toxicity is also critical for the accurate interpretation of results from experiments testing approaches such as RPE replacement strategies using this model. 

Supplementary Materials

pdf S1 

Acknowledgments

Supported by the National Institutes of Health through the NIH Director's New Innovator Award Program, 1-DP2-OD006649 (MS-G) and the award number T32EY007145 from the National Eye Institute (JI). The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Eye Institute or the National Institutes of Health. 

Disclosure: J. Wang, None; J. Iacovelli, None; C. Spencer, None; M. Saint-Geniez, None