Introduction

Photoreceptor cells have the highest rate of oxygen metabolism of any cells in the body and are continuously exposed to the deleterious effects of oxidative stress and to constant bombardment by photons of light. In all types of blindness, including those caused by inherited retinal degeneration, diabetic retinopathy, macular degeneration and retinal detachment, irrespective of the initiating defect, the intracellular concentration of reactive oxygen intermediates (ROIs) is thought to rise chronically or acutely and to activate a cell death pathway^{1, 2, 3}. The term 'oxidative stress' refers to the damage to cells and cellular components caused by exposure to the highly unstable small molecules called reactive oxygen intermediates⁴. These ROIs, including hydrogen peroxide, hypochlorite ions, hydroxyl radicals, hydroxyl ions and superoxide anions⁵, react with almost any nearby DNA, RNA, lipid, carbohydrate or protein. They are produced primarily by the normal oxygen metabolism that occurs in the mitochondrial respiratory chain, although they are also made by activated leukocytes. Such ROIs are also present in the air (if it contains cigarette smoke, radon or ozone) and are generated by the Sun's ultraviolet rays.

Cerium is a rare earth element of the lanthanide series and cerium oxide (CeO₂) is an inorganic compound that is insoluble in water, which has routinely been used in polishing glass and jewellery, and in catalytic converters for automobile exhaust systems and other commercial applications. Although most of the rare earth elements exist in the trivalent state, cerium also occurs in a +4 state and may flip-flop between the two in a redox reaction^{6, 7, 8}. Cerium oxides are excellent oxygen buffers because of their redox capacity⁹. As a result of alterations in the cerium oxidation state, CeO₂ forms oxygen vacancies or defects in the lattice structure by loss of oxygen and/or its electrons. The valence and defect structure of CeO₂ is dynamic and may change spontaneously or in response to physical parameters such as temperature, the presence of other ions, and oxygen partial pressure^{7, 8, 10}. Earlier studies have shown that, with a decrease in particle size, nanoceria particles demonstrate the formation of more oxygen vacancies^{11, 12}. The increased surface area to volume ratio that exists in a nanoparticle of 5 nm diameter enables CeO₂ to regenerate its activity and thereby act catalytically.

We hypothesized that these engineered nanoceria particles would be able to scavenge ROIs within retinal neurons. To test this, we used an in vitro cell culture system and an in vivo albino rat light-damage model. The photoreceptor cells in the albino rat are very sensitive to light and can be induced to degenerate by varying the exposure time or the intensity of the environmental lighting¹³. This sensitivity requires rhodopsin, the photoreceptor-cell visual pigment¹⁴. The nanoceria particles were effective in both cases and most importantly prevented light-induced loss of functional vision in vivo. In the light-damage model, retinal function, as determined by electroretinography (ERG), is preserved in a dose-dependent manner by the intravitreal injection of nanoceria particles prior to application of bright light, but they are also effective even when injected after light damage has been initiated. Our use of vacancy-engineered nanoceria particles represents a completely new therapeutic strategy, which we think has a very high probability of also being successful in preventing a variety of degenerative diseases. The nanoceria particles will not eliminate the cause, but should inhibit the progressive degenerative response to the ROIs.

Results

Nanoceria particles inhibit the intracellular accumulation of peroxides

As an initial test of the ability of nanoceria particles to protect retinal neurons, primary cell cultures of dissociated rat retinas were grown for seven days, as previously described¹⁵, and incubated with 1, 3, 5, 10 or 20 nM nanoceria particles for 0.5, 12, 24 or 96 h. After incubation, the cells were washed with phosphate buffer saline three times and incubated with 10 M 2,7-dichlorofluorescein diacetate (DCFH-DA) at 37 °C for 30 min. The cells were then washed again and incubated with 1 mM H₂O₂ at 37 °C for 30 min. The cells were harvested and the intensity of fluorescence was determined by flow cytometry. The ROI level was calculated as a ratio (mean intensity of experimental cells/mean intensity of control cells) and plotted as shown in Fig. 1. The data demonstrate that the nanoceria particles, when present for 12, 24 and 96 h, are effective in inhibiting the H₂O₂-induced rise of intracellular ROIs at all concentrations of nanoceria particles tested. The addition of 5 nM nanoceria particles at the time of plating also resulted in fewer apoptotic cells when assayed at 1, 2, 3 and 4 weeks later (data not shown).

Figure 1: Inhibition of ROIs by nanoceria particles.

Pretreatment of cultured retinal neurons with nanoceria particles (1, 3, 5, 10 and 20 nM) inhibits the intracellular accumulation of ROIs in response to exposure to 1 mM H₂O₂ for 30 min. a–d, Incubation with nanoceria particles for 0.5 (a), 12 (b), 24 (c) and 96 h (d). The protection is dose- and time-dependent. (Statistical analysis was done by one-way ANOVA and Newman–Keuls test for post hoc analysis. Data are shown as mean  s.d., n = 3, P < 0.05.).

Full size image (40 KB)

In vivo protection of photoreceptor cells

The cell culture data prompted us to test the efficacy of the nanoceria particles in vivo. To do this we used a light-damage animal model in which albino rats are exposed to 2,700 lux of light for 6 h, conditions that result in the death of about 50–60% of the photoreceptor cells throughout the retina. For the experimental paradigm, rats were injected with 2 l of saline or a 0.1, 0.3 or 1.0 M suspension of the nanoceria particles in saline into the vitreous of both eyes on day 0. Three days later, the animals were exposed to 2,700 lux of light for 6 h and then returned to normal cyclic lighting (12 h dark, 12 h light at 100 lux) for seven days. The rats were then killed, the eyes enucleated, and fixed, processed, paraffin-embedded sections cut and stained with haematoxylin and eosin (H&E). Morphological evaluation of photoreceptor cell protection was performed using bright field microscopy. Exposure to the 2,700 lux of light reduced the thickness of the outer nuclear layer (ONL) of photoreceptor cell nuclei from the 10–13 rows (Fig. 2a, control) in normal rats to 1–2 rows in the most degenerated region of the retinas in light-exposed (LE) rats (Fig. 2b) or vehicle- (0.9% NaCl) injected rats (Fig. 2c). However, in rats treated with nanoceria particles, there was significant dose-dependent protection of photoreceptors, resulting in the ONL having 5–6 rows (Fig. 2d, 0.1 M CeO₂), 7–8 rows (Fig. 2e, 0.3 M CeO₂) or 11–12 rows (Fig. 2f, 1 M CeO₂) of nuclei. In addition, the outer and inner segments of treated retinas were highly preserved compared to the LE and vehicle-injected retinas. The data shown are representative of those obtained in three separate experiments and are from the right eye only as there was no difference between right and left eyes of individual rats (data not shown).

Figure 2: Intravitreal injection of nanoceria particles protects rat retina photoreceptor cells from light-induced degeneration.

Representative images of photomicrographs of H&E stained sections adjacent to the optic nerve are shown. The white bars indicate the thickness of the layer of nuclei of rods and cones (the outer nuclear layer, ONL). Injections were given on day 0, rats were exposed to light on day 3, and the experiment ended on day 10. a, No light exposure (LE), no injection. b–f, Rats were exposed to 6 h of 2,700 lux white light. No injection (b), injections of saline (c), 0.1 M CeO₂ (d), 0.3 M CeO₂ (e) and 1.0 M CeO₂ (f) 3 days before LE. The scale bars in the lower right corner represent 25 m.

Full size image (56 KB)

Quantitative histology

To obtain an overall view of the light damage and protection throughout the retina¹⁶, the thickness of the ONL was measured every 220 m, starting from the optic nerve and proceeding in the superior and inferior directions to where the retina ends. The thickness of the ONL is directly proportional to the number of photoreceptor cell nuclei (5 m diameter) at each point and a plot of the thickness versus each 220 m distance from the optic nerve is presented in Fig. 3a. These data demonstrate that the superior retina (left of the optic nerve) is more sensitive to light damage than the inferior retina (right of optic nerve) and that the nanoceria particles are effective in protecting the photoreceptor cells in both regions. Injections of 0.1 M and 0.3 M suspensions are partially protective, but the data obtained for a 1.0 M suspension demonstrate almost complete protection and are essentially indistinguishable from the control animal, which was not exposed to light. These conclusions are especially obvious when the data from the average thickness of the most sensitive region, the superior retina, are plotted on a bar graph (Fig. 3b).

Figure 3: Nanoceria particles provide pan-retinal protection against light damage.

The thickness of the ONL of the retina was measured along the vertical meridian every 220 m, starting at the optic nerve (ON). a, The ONL thickness of control retinas (no LE, no injection), LE, and LE and vehicle-treated nanoceria-particle-treated retinas are plotted against each of the successive 220 m distances from the optic nerve (ON). The negative x-axis numbers on the left side of the ON refer to the points on the superior hemisphere and those on the right side refer to the inferior hemisphere. Statistical analysis was done by one-way ANOVA and Newman–Keuls test for post hoc analysis. The results are expressed as mean ONL thickness s.e.m. (n = 6 for each point; *P < 0.05; **P < 0.001 versus light exposure (LE) group). b, Summation of the ONL thickness across the superior hemisphere provides a quantitative assessment of the protective effects of the nanoceria particles in the most sensitive hemisphere. Statistical analysis was done by one-way ANOVA and Newman–Keuls test for post hoc analysis. The results are expressed as mean ONL thickness  s.e.m. (n = 6 for each group).

Full size image (36 KB)

Protection against long-term light damage

Exposure to bright light results in the rapid death of numerous photoreceptor cells, but it also damages some cells less extensively and they die over a longer time period¹⁷. To visualize these damaged cells, animals were killed five days after light exposure and their eyes processed for histological analysis using the TUNEL (terminal deoxynucleotidyl transferase dUTP nick end labelling) assay. This fluorescent assay reveals those cells that have been damaged by the light and have progressed along an apoptotic cell death pathway. The results shown in Fig. 4 are representative of such experiments and demonstrate that exposure to bright light results in TUNEL-positive cells in those retinas from eyes that were either uninjected (Fig. 4b, c) or injected with saline (Fig. 4e, f). However, all three concentrations (0.1, 0.3, 1.0 M) of nanoceria particles protected the retinas and prevented the appearance of TUNEL-positive cells (Fig. 4 h, k, n), suggesting that even at lower concentrations, nanoceria particles are effective in inhibiting long-term light-induced injury to photoreceptor cells.

Figure 4: Nanoceria particles prevent the appearance of TUNEL-positive photoreceptor cells, which occurs days after exposure to damaging light.

The eyes were injected at day 0, exposed to light on day 3, and the experiment ended on day 8. a, d, g, j, m, Sections stained with H&E. The black rectangles indicate the sites from which the images in the second and third columns were taken. b, e, h, k, n, Sections stained with anti-digoxigenin-rhodamine for the TUNEL assay. c, f, i, l, o, The same sections visualized using Nomarsky optics. INL, inner nuclear layer where bipolar cell nuclei are located; ONL, outer nuclear layer where photoreceptor cell nuclei are located. The scale bars in a, d, g, j, m represent 100 m; the scale bars in c, f, i, l, o represent 25 m.

Full size image (34 KB)

Nanoceria particles protect retinal function

The morphological data clearly indicate that the nanoceria particles preserved the photoreceptor cells, but the best criterion for showing protection of vision from light damage is to demonstrate the preservation of retinal function. To address this, the same experimental paradigm was used, but prior to ending the experiment, ERG was performed on day 7. Representative ERG data of those experiments are shown in Fig. 5. The control animals, not exposed to light, exhibit a strong ERG (Fig. 5a), whereas those animals that were exposed to light and were either uninjected (Fig. 5b) or injected with saline (Fig. 5c) had almost flat ERGs, indicating the absence of retinal function. The ERG data from the rats that were injected with nanoceria particles but not exposed to the bright light (Fig. 5d) are essentially identical to those obtained with rats that were not injected and not exposed to the bright light, indicating that the nanoceria particles were not simply preventing the passage of light into the eye. However, rats injected with nanoceria particles that were exposed to light showed ERGs that reflected a progressive improvement in retinal function that paralleled the progressive increase in the concentration of nanoceria particles from 0.1 to 1.0 M. The quantification of ERG A- and B-wave amplitudes (Fig. 5h) clearly demonstrates the functional improvement of retinas following treatment with nanoceria particles. This is especially evident in the eyes pretreated with 1 M nanoceria particles, which had 79% and 87% of the control A- and B-wave amplitudes, compared with 22% and 26% for the retinas with 0.9% NaCl vehicle treatment.

Figure 5: Retinal function is protected by the nanoceria particles in a dose-dependent manner.

a–g, Representative ERGs of each group are plotted as voltage (V) against time (ms). h, Mean of the ERG B-wave and A-wave amplitudes are shown for each group of animals. Statistical analysis was done by one-way ANOVA and Newman–Keuls test for post hoc analysis. The results are expressed as mean amplitude  s.d. (n = 6 for each point). Control animals had no injections and were not exposed to bright light. LE animals had no injections and were exposed to bright light

Full size image (47 KB)

Nanoceria particles can rescue retina function after light damage

The data presented show that nanoceria particles were very effective in protecting retina function when administered three days before light exposure. To determine if the nanoceria particles had any ability to rescue photoreceptor cells after they had been exposed to damaging light, rats were subjected to 6 h of 2,700 lux light and were injected intravitreally, 2 h later, with 2 l of a 1 M nanoceria particle suspension. The retinal function of the animals was determined seven days later using ERG and quantitative histology applied to the retinal sections. A summary of the histology data from such an experiment is presented in Fig. 6a, which indicates that, even after light damage, there is protection of photoreceptor cells throughout the retina. Similarly, a summary of the ERG data (Fig. 6b) demonstrates that a significant amount of retinal function is rescued by post-treatment with the nanoceria particles.

Figure 6: Nanoceria particles protect photoreceptor cells even when administered after LE.

Rats were exposed to 2,700 lux for 6 h, and 2 h later were either uninjected or injected intravitreously with 2 l of saline or a 1 M suspension of CeO₂ in saline. a, The thickness of the ONL is plotted against the site of measurement along the vertical meridian starting at the optic nerve (ON) for each group of animals. Control: no injections, no light exposure; LE: light exposure, no injections; LE + 0.9% NaCl: light exposure then saline injection; LE + CeO₂: light exposure then 1 M CeO₂ injection. Statistical analysis was done by one-way ANOVA and Newman–Keuls test for post hoc analysis. The results are expressed as the mean ONL thickness  s.e.m. (n = 4 for each point; *P < 0.05, ^¶P < 0.001 versus the LE group). b, Summary of the functional protection (ERG) provided by post-treatment nanoceria particle injection. These data are from the same animals as in a. Each column represents the mean of the ERG B- and A-wave amplitudes. Statistical analysis was done by one-way ANOVA and Newman–Keuls test for post hoc analysis. The results are expressed as mean amplitude  s.d. (n = 4; P < 0.01 versus LE group). The insets are representative ERG waveforms from four independent experiments.

Full size image (28 KB)

Discussion

Vision is dependent on rod and cone photoreceptors, and human diseases that cause blindness do so by a variety of primary events that ultimately produce two subsequent effects in the photoreceptor cells. The first effect is thought to be the production of either an acute or a chronic rise in the intracellular concentration of ROIs in the photoreceptor cells, which in turn induces the second effect, the activation of an apoptotic signal transduction pathway and the eventual death of the photoreceptor cells. The intracellular ROIs are especially damaging because they can react with almost every type of organic molecule that exists within the cell. Most of the ROIs within the body are generated by respiration, the use of oxygen to produce energy from the metabolism of organic molecules. The retina of the eye is exposed to large amounts of such toxic compounds, generated by normal oxidative reactions as well as those produced by the constant absorption of photons of light. The retina of the albino rat is extremely sensitive to photon-induced damage, which results in the production of excess ROIs, the subsequent degeneration of the photoreceptor cells, and the loss of visual function¹⁸. We have used retinal neurons in culture and the light-damage model in vivo to demonstrate and characterize the ability of a suspension of vacancy-engineered nanoceria particles to prevent ROI-induced injury to the photoreceptor cells and to preserve the visual function of the retina.

Investigations of nanocrystalline CeO₂ have revealed that its lattice constant increases with decreasing nanoparticle size^{11, 12}. This has been attributed to an increase in oxygen vacancies in the crystal structure^{11, 12, 19}. It is implied that the migration enthalpy of the oxygen vacancy in CeO₂ is smaller at the nanoscale^{8, 20}. Additionally, at the nanoscale, the surface area of CeO₂ particles is dramatically enlarged in relation to its volume to increase oxygen exchange and redox reactions. Thus, oxygen vacancies are likely to form more readily at the nanoscale. We hypothesized that nanoceria particles, owing to their chemical and physical structure, protect cells from free-radical-induced damage. The defects in the nanoceria particles can act as chemical spin traps similar to nitrosone compounds, currently used as biological antioxidants. However, one CeO₂ particle may offer many sites of spin-trap activity, where current pharmacological agents offer only a few per molecule. Additionally, the lattice defects in nanoceria particles possess the potential for regeneration and do not require repetitive dosage as seen with the use of vitamins C and E. We propose that nanoceria particles act as free-radical scavengers by switching between the +3 and +4 valence states via various surface chemical reactions. The radical-scavenging mechanism can be given by the following set of chemical reactions.

This regenerative property makes nanoceria particles a very attractive proposition for treating ROI-mediated cellular damage and diseases, as most of the present free-radical scavengers do not have such a property and need a repetitive dosage.

Initially, we tested the ability of nanoceria particles to protect retinal neurons in culture and found that their presence at 5 nM was sufficient to prevent the increase in apoptotic cells seen in cultures grown in the absence of the nanoceria particles. Their protective effects were further demonstrated by showing that the nanoceria particles prevented the ROI-induced apoptosis that occurred in cultured retinal neurons exposed to exogenously added H₂O₂. The addition of the nanoceria particles to such cultures was also shown to prevent the intracellular accumulation of ROIs that was observed in retinal neurons to which H₂O₂ was added. To determine if the nanoceria particles would act in a similar manner in vivo, a light-damage paradigm was established and the number of photoreceptor cells was determined by measuring the thickness of the ONL. The histological evaluation of individual sections supported the conclusion that the nanoceria particles provided protection from light damage, even at very low concentrations (2 l of 0.1, 0.3 and 1.0 M). The light-dependent degeneration, although more severe in the superior retina, occurs throughout the retina, and the protective effect of the nanoceria particles is not limited to the area of the injection but was also shown to occur across the entire retina. The pan-retinal dose-dependent protection was further documented using ERG, to show that pretreatment with the nanoceria particles protected retinal function.

Injection of the nanoceria particles three days prior to light exposure prevented the subsequent appearance of TUNEL-positive cells seen in the retinas of uninjected or saline-injected control animals. We did not observe any dead, damaged or TUNEL-positive retinal pigment epithelium (RPE) cells in our experiments and the ERG data indicate that the RPE are fully functional in the nanoceria-particle-treated cells. This is contrary to the observations made with the ZUR SIV albino rat²¹. Even when injected after light damage, the nanoceria particles were still partially effective, suggesting that, as long as the cells have not died, they may be prevented from either entering or proceeding down an apoptotic pathway. We do not yet know how the nanoceria particles are taken up into photoreceptor cells, nor the process by which they are eliminated. However, because ROIs are formed within the photoreceptor cells in response to light damage, and because ROIs react over very short distances, we think that the nanoceria particles must actually enter the photoreceptor cells. Similarly, inhibition of the appearance of TUNEL-positive cells days after the light insult suggests a relatively long active lifespan in the retina. Because our data demonstrated that the nanoceria particles prevent the peroxide-induced increase in the intracellular concentration of ROIs in cultured retinal cells, we think the nanoceria particles function by the same mechanism in vivo. However, it may be that they regulate the concentration of redox couples such as glutathione disulphide/glutathione²² or induce scavenging enzymes^{23, 24}, although we think it is more likely that they inhibit apoptosis by directly preventing the rise in ROI levels.

Our in vivo results were obtained by direct injection of the nanoceria particles into the vitreous of the eye and it is likely that other routes of administration would also be effective. Neurodegeneration within the retina is not unlike neurodegeneration within other areas of the central nervous system. It is not generally possible to directly analyse human tissue in vivo, but the preponderance of data suggests that Alzheimer's disease²⁵, Parkinson's disease¹ and Huntington's disease²⁶ are caused by different initiating events that result in an increase in ROI levels within the populations of neurons undergoing degeneration. This suggests that the nanoceria particles can also be effective in preventing the ROI-mediated death or loss of function of the neurons uniquely affected in each of these illnesses. Similarly, there are many diverse diseases initiated by ROI, including atherosclerosis²⁷ and stroke²⁸, and inductive reasoning leads to the conclusion that the vacancy-engineered nanoceria particles can be beneficial against any of the diseases in which ROIs are involved.

Material and Methods

Author contributions

J.C. and J.F.M. conceived and designed the experiments, J.C. performed the experiments, J.F.M. and J.C. analysed the data, S.P. and S.S. generated the nanoceria particles, and J.C., S.P., S.S. and J.F.M. cowrote the paper.