The potential of cerium oxide nanoparticles (nanoceria) for neurodegenerative disease therapy

Free radicals, molecules with an unpaired electron in the outer orbital shell, are highly reactive and strip electrons from biomolecules such as proteins, lipids and DNA, causing cellular damage. Since they can be formed as normal byproducts of cellular respiration, cells have endogenous antioxidant systems, such as enzymes like catalase and superoxide dismutase (SOD), to help minimize damage due to rogue free radicals. However, under pathological conditions, the production of free radicals can exceed the capacity of endogenous systems to neutralize them, resulting in a condition known as oxidative stress. Oxidative stress has been shown to play a role in aging and in the pathology of stroke and many neurodegenerative diseases [1]. Although we have a good understanding of the role of free radicals and oxidative stress in these disorders, there is still a paucity of antioxidants that have proven effective for treating neurodegenerative diseases.

CeO₂ nanoparticles (nanoceria) have the potential to be developed as a therapeutic for oxidative stress diseases due to their catalytic antioxidant properties. Cerium is a rare earth metal that, when combined with oxygen, can adopt a fluorite crystalline lattice structure that has a highly reactive surface area for neutralization of radicals. In addition, nanoceria can reversibly bind oxygen and shift oxidation states (Ce³⁺/Ce⁴⁺) depending on the conditions. Because of these properties, nanoceria have been studied in biological systems and shown to display antioxidant effects in various models of disease [2]. For example, we have shown the neuroprotective effects of nanoceria using a hippocampal brain slice model of cerebral ischemia [3], as well as a murine model of multiple sclerosis, experimental autoimmune encephalomyelitis [4].

The use of nanoceria for therapeutic purposes provides multiple advantages over other novel antioxidant approaches. For example, the delivery of nanoencapsulated antioxidant enzymes, such as SOD or catalase, has the limitation that only one type of reactive oxygen species can be scavenged by each enzyme, whereas, multiple species are implicated in neurodegenerative diseases [1] and nanoceria have been shown to decrease their levels [3,5]. In addition, there is a limited amount of enzyme that can be delivered within the nanostructure. In contrast, nanoceria exhibits both catalase- and SOD-mimetic activity with SOD-mimetic catalysis exceeding that of the endogenous enzyme [6,7]. One study showed that 5.8 μM nanoceria was equivalent to 527 U of SOD [8]. Alternatively, activation of endogenous antioxidants via the Nrf-2 pathway has also been pursued. Nrf-2 is a transcription factor that mediates the activation of a suite of antioxidant genes, including genes that encode catalase, SOD and several proteins involved in glutathione homeostasis, via the antioxidant response element [9]. Indeed, synthetic triterpanoids that function as activators of the Nrf-2/antioxidant response element pathway have been shown effective in models of neurodegenerative disease [10]. However, this strategy may be limited under pathological conditions in which increased gene expression alone is not sufficient to mitigate the increase in oxidative stress.

Despite these advantages of nanoceria over other therapeutic strategies, there are toxicity issues to consider. In particular, pro-oxidant effects of cerium oxide nanoparticles have been reported in various model systems [11,12], and this would be cause for concern if the ultimate goal is to translate their use to a clinical setting. However, cerium oxide nanoparticles, although comprised of the same core elements, do not all display the same biological effects. We hypothesize that disparate reports – of pro-oxidant toxicity in some cases and antioxidant protective effects in others – could be explained by the different physiochemical parameters of the various nanoparticles tested, such as method of synthesis, particle size and stabilizer. These parameters can then influence the biological identity of the particle due to the adsorption of various proteins on the nanoparticle surface. Differing biological identities could lead to varying interactions with cellular constituents, affecting the biodistribution of the particles and their ultimate effects on the organism. We will explore these ideas further in the sections below.

Role of synthesis, particle size & stabilizer

Different synthesis methods can endow cerium oxide nanoparticles with varying physiochemical and catalytic properties that could contribute to pro-oxidant or antioxidant properties, as Dowding et al. recently demonstrated [13]. Another complication is that the properties of the particle measured in vitro (e.g., size, zeta potential and redox activity) could change under physiological conditions. For example, Kumari et al. showed that the hydrodynamic diameter nanoceria increased dramatically in cell culture media due to the tendency of the particles to agglomerate in physiological buffers [12]. In addition, adsorption of proteins present in biological fluids, such as blood, can also influence the size and distribution of metal oxide nanoparticles [14,15].

Many particles have been synthesized with a coating or stabilizer with the ultimate goal of yielding monodispersed particles that are less prone to aggregation while still retaining the chemistry inherent in the particle. Citric acid is a commonly used stabilizer due to its biocompatibility. However, it is important to note that the stabilizer too can change once immersed into physiological solution. Specifically, citrate-stabilized particles have been shown to precipitate or salt-out in high ionic strength solutions [16].

Biological identity of nanoparticles

Studies looking at the physiochemical characteristics of nanoparticles after interaction with biological fluids provide some insight into the role of protein–nanoparticle interactions in ultimate nanoparticle effects. Walczyk et al. incubated various types of nanoparticles with human blood plasma and showed that plasma proteins can interact with the nanoparticles to form a stable, long-lasting corona on the particle surface [17]. Walkey et al. systematically analyzed the protein corona of 105 different surface-modified gold nanoparticles [18]. Incubation of these nanoparticles with undiluted human serum led to a different set of proteins adsorbed onto the nanoparticle surface, a so-called corona fingerprint for each type of particle, which then either promoted or inhibited the interaction of the particle with human lung carcinoma cells. Taken together, these and other studies suggest that the protein corona endows the nanoparticle with a particular biological identity that subsequently plays an important role in the ultimate interactions of a nanoparticle with target cells. A protein corona could, for example, induce novel surface receptor interactions, influence the intracellular targeting of the particle to specific organelles or trigger immune cell activation [15,19]. Although a systematic analysis of the protein corona fingerprint of cerium oxide nanoparticles has yet to be conducted, one can extrapolate these general conclusions to help interpret the varying effects of nanoceria in whole animals observed in some recent studies [4,11,20].

Sorting out the effects of nanoceria in whole animals

Using the experimental autoimmune encephalomyelitis model of multiple sclerosis, we recently showed that custom-synthesized citrate/ethylenediaminetetraacetic acid (EDTA)-stabilized nanoceria reduced free radical levels in the brain and alleviated the clinical symptoms and motor deficits associated with disease progression. The mice received multiple doses of the particles (up to 30 mg/kg) over a 35-day period. We used the same particles in a G93A mouse model of amyotrophic lateral sclerosis and observed dose-dependent improvement in motor function and survival when particles were delivered at the onset of motor symptoms (Erlichman et al., unpublished observations). However, in a long-term study, Hardas et al. measured significant increases in multiple oxidative stress parameters in rat hippocampus after single intravenous infusion (87 mg/kg) of a 30 nm citrate-stabilized nanoceria, even though the nanoceria did not penetrate the brain [11]. Oxidative stress was elevated for up to 30 days post-infusion, but most parameters returned back to baseline by 90 days.

One possibility to account for these differing reports is the concentration of the nanoparticles used, as toxicological studies typically use higher doses. Another possibility is that in the studies where nanoceria displayed toxicity, the particles agglomerated in the blood if the stabilizer precipitated out [16] or bound different protein coronas, both of which could lead to an enhanced immune response. Indeed, hippocampal oxidative stress was observed with minimal detection of nanoceria in the brain but high accumulation in reticuloendothelial organs, such as the liver and spleen, as well as increases in the levels of interleukin-1 beta, an inflammatory cytokine [11,20]. On the other hand, the particles used in the study by Heckman et al. were stabilized with a custom-synthesized mixture of citrate/EDTA and TEM analysis demonstrated that these particles had an average size of 2.9 nm, were monodispersed even in high salt solutions, and could be detected in the brain following intravenous administration [4].

Conclusion

The core redox chemistry of nanoceria presents multiple advantageous properties for the treatment of oxidative stress diseases. However, an important consideration in their clinical translation is how cerium oxide nanoparticles behave in biological systems. Addressing this is not a simple endeavor. Rational design of surface coatings may mitigate some of the agglomeration issues in vitro, but may not be predictive of the whole animal condition. As the nanoparticles traverse multiple biological compartments, each with varying ionic and protein compositions, some coatings may become destabilized. In addition, different repertoires of proteins may adsorb onto the nanoparticle surface, ultimately driving the biological effects. Since there are currently no unifying principles for predicting these effects, each particle must therefore be individually characterized in a biological setting.