Chemical hypoxia was induced in RGC-5 cells using cobalt (II) chloride (CoCl₂). After hypoxia, profound cell loss was observed in the vehicle-treated hypoxic group (Figure 1b,e; p < 0.05 versus normal control). Cells appeared to be more round, with loss of processes (Figure 1b) when compared with the normal control (Figure 1a). However, lutein treatment reversed the cytotoxic effect of CoCl₂ and led to less damage to RGC-5 cells (Figure 1c,d). Cells treated with 20 μM lutein (Figure 1d) showed morphology similar to that of the normal control (Figure 1a). Quantitative analysis by cell counting showed that more RGC-5 cells were observed in the lutein-treated group (Figure 1e; p < 0.05 at 20 μM versus vehicle-treated group).

Hydrogen peroxide (H₂O₂) was used to induce oxidative stress. Exposure to H₂O₂ led to cell death in the vehicle-treated group (Figure 2b,e; p < 0.01 versus normal control). Upon lutein treatment at both at 10 μM and 20 μM, the number of RGC-5 cells was increased (Figure 2c,d,e; p < 0.05 versus vehicle-treated group) to a number similar to that of the normal control (Figure 2e; p > 0.05).

Ischemia and oxidative stress are common causes of many ocular diseases, which lead to irreversible RGC damage. In the present study, we examined the neuroprotective effect of lutein on RGC-5 cells against CoCl₂-induced chemical hypoxia and H₂O₂-induced oxidative stress in vitro. Our data demonstrated that lutein exerted neuroprotection on RGC-5 cells against hypoxia and oxidative stress.

Retinal ischemia is a feature of many ocular pathologies such as glaucoma, DR and retinal vessel occlusion [1]. In experimental studies, CoCl₂ is one of the common agents to induce hypoxia [14–18]. CoCl₂ treatment simulates hypoxia, a key event during ischemic injury, by altering gene and protein expression similarly to ischemia [19]. It induces hypoxia by blocking the degradation of hypoxia-inducible factor-1α (HIF-1α) and subsequent HIF-1α accumulation [20]. Moreover, CoCl₂ also induced apoptosis through activation of caspase-3/8, cleavage of anti-apoptotic protein Mcl-1 and generation of reactive oxygen species (ROS) in a variety of in vitro studies [17,21]. In animal studies, CoCl₂ has been shown to induce apoptosis and retinal photoreceptor degeneration [22]. In addition, CoCl₂-induced hypoxia has been adopted in RGC in vitro [16,23] and in vivo [16]. Accumulation of HIF-1α protein increased expression of heat shock protein-27, and generation of β-amyloid peptide [23] was shown in CoCl₂-treated RGC-5 cells. In our results, we demonstrated that CoCl₂ attributed to hypoxia-induced injury in RGC-5 cells.

Oxidative stress is one of the key factors leading to neuronal injury. The retina is highly susceptible to oxidative stress because of the high content of polyunsaturated fatty acids and high oxygen consumption [24]. Under normal situations, cells possess several intrinsic antioxidant enzymes such as superoxide dismutase, catalase and glutathione peroxidase to cope with oxidative stress resulted from normal metabolism in our body [3]. However, during injuries such as in ischemia/reperfusion, glaucoma and DR, overproduction of ROS and free radicals overwhelms the intrinsic antioxidant mechanisms [1,3]. RGC is sensitive to oxidative stress in pathological situations in vivo [3] and in vitro [25]. In experimental studies, H₂O₂ is widely used to induce oxidative stress [25]. Exogenous H₂O₂ increases intracellular accumulation of ROS [26], apoptosis and leads to loss of cell viability [25] in RGC-5. H₂O₂-induced apoptosis in RGCs has been shown to be caspase-independent and yet involves the activation of poly(ADP-ribose) polymerase and apoptosis-inducing factor [25]. In the present study, H₂O₂ also induced significant cell damage to RGC-5 cells, which was comparable to previously found [25].

In the macula, lutein absorbs high energy blue light and protects the retina from oxidative injury [10]. Low levels of lutein intake have been shown to associate with the prevalence of AMD [27]. Lutein supplementation has been shown to improve vision and retard progression of AMD in clinical trial studies [28,29]. To our knowledge, there is no reported toxic effect of lutein even at a high dose of intake [30]. No significant clinical, hematological, biochemical or histopathological side effects were noted in rats fed with 733 mg/kg per day of purified crystalline lutein [10,30]. More importantly, lutein has been regarded and approved safe to be used as a daily supplement and to be included into certain food and beverage application in USA [10]. However, the use of lutein is still limited in treating AMD, which is an outer retinal disease.

Recently, intensive efforts have been made to explicate the neuroprotective effects of carotenoids in ocular diseases in vivo [12,31,32] and in vitro [5,33]. Lutein treatment in DR mice restored malondialdehyde and glutathione protein levels, glutathione peroxidase activity as well as electroretinogram response to control values [31]. Lutein also reversed the activation of factor-kappa B transcription, which is involved in oxidative stress and inflammation response [31]. In mice with retinal inflammation, lutein reduced inflammatory response and oxidative stress through reversal of STAT3 activation, downstream of inflammatory cytokine signals [32]. In addition, the activation of glial fibrillary acidic protein, an indicator of pathological change of Muller glial cells, was prevented in animal treated with lutein. In our recently reported study, we found that lutein was also protective to inner retinal neurons in ischemia/reperfusion injury in vivo [12]. Reduced immunoreactivity of nitrotyrosine and poly(ADP-ribose) in the inner retina, indicating a reduced oxidative stress, was observed in lutein-treated ischemic retina. Effects of carotenoids on a specific cell population were investigated using an in vitro approach. Zeaxanthin and astaxanthin have been shown to protect RGC-5 cells from oxidative injuries [5,33]. In the present study, our results suggested that lutein treatment protected RGC-5 from CoCl₂-induced chemical hypoxia and H₂O₂-induced oxidative stress. Indeed, it has been proposed that lutein can effectively reduce the intracellular accumulation of H₂O₂ by scavenging H₂O₂ and superoxide as well as inhibit NFκB-regulated inflammatory gene expression in lipopolysaccharide-stimulated macrophages in vivo and in vitro [34]. This indicates that lutein is able to penetrate into cells and scavenge intracellular H₂O₂ to prevent cell damage. Furthermore, lutein-binding protein [35] and retinal tubulin [36] are found in the ganglion cell layer of primate retina and bovine retina. These proteins are suggested to be involved in lutein transport. However, further study is necessary to investigate the expression and localization of lutein-bind protein in rodents and the RGC-5 cell line.

The RGC-5 cell line was previously used as a RGC-specific in vitro model [4,5,18,26,33]. However, several recent reports have questioned the validity of this cell line. It was demonstrated that RGC-5 cells lack critical biochemical and physiological RGC properties [37–39]. In addition, RGC-5 cells do not express RGC-specific markers such as neurofilaments or Thy 1.2 [38]. Moreover, RGC-5 cells are unexcitable, with no voltage-dependent inward Na⁺ or Ca²⁺ currents or action potentials, which are critical properties of cultured postnatal and adult rat RGCs [37]. All these pieces of evidence imply the limitation of using the RGC-5 cell line as an in vitro model of RGCs. Primary RGC culture may be a more appropriate in vitro model of RGC study.