Fucoidan from Fucus vesiculosus Protects Retinal Pigment Epithelium from Lipid-Induced Damage Related to AMD

Abstract

Fucoidans are natural compounds that exhibit bioactivity against age-related macular degeneration (AMD), the leading cause of central vision loss in industrialized nations. Pathological factors like oxidative stress and lipid peroxidation play vital roles in AMD pathogenesis. Lipid-induced alterations in the retinal pigment epithelium (RPE) contribute to AMD development. In this study, a commercial fucoidan from Fucus vesiculosus (FVs) was tested for its activity regarding lipid-peroxidation-related effects. The human RPE cell line ARPE-19, primary porcine RPE, and RPE/choroid explants were stimulated with erastin, acting as an inducer of lipid peroxidation, and treated with fucoidan. Effects on cell viability (tetrazolium bromide (MTT) or calcein staining), vascular endothelial growth factor (VEGF) and interleukin 8 (IL8) secretion (ELISA), reactive oxygen species (ROS), protein expression (glutathione peroxidase 4 (GPX4), CD59, and retinoid isomerohydrolase (RPE65), analyzed via Western blot), and gene expression (RT-qPCR) were investigated. FVs showed protective effects against erastin-induced reduction in viability (with a 12.7% increase in viability compared to erastin), RPE65 expression (with a 4.2-fold increase compared to erastin), and GPX4 expression (with a 2.3-fold increase compared to erastin) in primary RPE. Erastin-induced VEGF secretion was attenuated by FVs in ARPE-19 and primary RPE (with an up to 1.7-fold reduction compared to erastin). Elevated IL8 levels were reduced by FV treatment in primary RPE (with a 9.1-fold reduction compared to erastin). Induced VEGF in RPE/choroid explants was reduced by FVs (with an up to 2.9-fold reduction compared to erastin), and this reduction was correlated with slight improvements in viability. In conclusion, FVs exerted protective effects against lipid-induced stress. This study reveals further effects of fucoidans against AMD-related pathologies.

1. Introduction

Age-related macular degeneration (AMD) is the main cause of blindness in older populations in industrialized nations. In addition to its medical significance, it causes sociological and economic problems, with 288 million affected humans forecast for 2040 [1]. It is a multifactorial disease. The main cells affected are the retinal pigment epithelium (RPE) and photoreceptors of the macula lutea, with the latter being responsible for high-acuity central vision. The main pathomechanisms of the development of AMD are oxidative stress, inflammation, lipid-induced damage (contributing to geographic atrophies—dry AMD), and excess vascular endothelial growth factor (VEGF, contributing to vessel growth/retina rupture—wet AMD) [2]. Currently, only wet AMD can be treated with anti-VEGF therapeutics and their derivatives, which are not curative and need to be regularly intravitreally applied [3]. Recently, complement component 3 (C3) and complement component 5 (C5) inhibitors have been developed for the treatment of advanced dry AMD; however, they only decelerate the progression of atrophy. Furthermore, they can lead to a conversion from dry to wet AMD [4], and thus their use in the EU has not been approved by the EMA. Alternatives for treatment and prevention are desirable.

RPE cells have many functions in the retina, including phagocytosis of shed outer segments of photoreceptors, upholding outer retina–blood barrier and immune privilege, and recycling of visual pigments [5]. They protect photoreceptors against reactive oxygen species (ROS); e.g., with glutathione or superoxide dismutase [6]. RPE cells secrete VEGF to uphold vessel support and endothelial fenestration [7]. Disturbance of RPE functions can lead to severe pathological conditions. Pro-inflammatory behaviors in RPE can be activated by danger-associated patterns like lipopolysaccharide (LPS), leading to the secretion of the pro-inflammatory cytokines interleukin 6 (IL6) and interleukin 8 (IL8) [8]. The RPE is involved in complement system regulation. The activation of the complement system is suppressed at the basolateral cell site through negative regulators like CD59 (protectin) [9], which is expressed by the RPE [9]. Lowered CD59 expression in the RPE is associated with the development of AMD [10]. Researchers have demonstrated that the expression of retinoid isomerohydrolase (RPE65), a key enzyme for visual pigment recycling, can be lowered by inflammatory activation with LPS or tumor necrosis factor alpha [11].

The Western diet is known to cause chronic damage, which affects the retina. In contrast, a seafood diet containing foods like seaweed has various positive effects on the retina [12]. Fucoidans from brown algae are often described as antioxidant, anti-inflammatory, and conducive to lipid metabolism [13]. Fucoidans are sulfated polysaccharides from the cell walls of brown algae, contributing to their integrity and defense against environmental damage [14]. They show high variability, which is caused by different algal origins, environmental conditions, and extraction methods, making a proper chemical characterization mandatory for medical research and applications [15]. In this study, a commercially available fucoidan was tested for its possible protective effects against lipid-induced damage in in vitro models. Extensive work has been conducted on fucoidans’ antioxidant, antiangiogenic, and anti-inflammatory activities in the RPE [16]. As shown in published studies, fucoidans are promising substances that address three out of four of the most important pathomechanisms of AMD. The fourth factor, lipid-induced damage, which can lead to degeneration of the RPE, has not been investigated in depth, but a previous study showed that fucoidans had protective effects against iron-induced cell death in ARPE-19 [17]. In this study, erastin was used as an inducer of lipid peroxidation in RPE cell culture models, such as the RPE cell line ARPE-19, primary porcine RPE, and RPE/choroid organ cultures.

A previous study described, for the first time, a commercial fucoidan from Fucus vesiculosus (FV) reducing VEGF secretion and expression in the RPE as well as the angiogenic structures of endothelial cells [18]. In the RPE cell line ARPE-19, fucoidans did not act as ROS scavengers but rather, at the cellular signaling level, had positive effects on iron-induced oxidative stress and increased levels of the key antioxidant molecule GPX4 [17]. Fucoidan from FV reduced VEGF release in RPE/choroid organ cultures [18,19]. Anti-inflammatory activity was demonstrated, as fucoidans inhibited IL6 and IL8 release from the RPE after pro-inflammatory activation [20,21]. The FV fucoidan reduced the gene expression of pro-inflammatory and proangiogenic cytokines in the RPE [22]. Taken together, fucoidans show remarkable potential against factors relevant for AMD, paving the way for further tests regarding lipid-induced damage. To further investigate this area, a universally available commercial FV fucoidan from Sigma-Aldrich was used and investigated regarding lipid-induced pathologies after erastin stimulation. Erastin inhibits the system x_c⁻ cystine/glutamate antiporter, which prevents glutathione formation and thereby hinders glutathione peroxidase 4 (GPX4) activity [23]. This leads to an increase in levels of phospholipid hydroperoxides, which damage the cell membranes via lipid-induced oxidative stress (lipid peroxidation/lipotoxicity) [23].

We hypothesized that pure high-molecular-weight fucoidan from FV would have protective effects against lipid-induced toxicity mimicked by erastin stimulation by reducing pro-inflammatory cytokine secretion (IL8) and proangiogenic chemokine secretion (VEGF), along with protective effects on gene expression level and markers of visual pigment recycling (RPE65) and antioxidative measures (GPX4).

2. Results

2.1. Prelimianry Testing

2.1.1. Cell Viability

Initial experiments were performed to determine the proper erastin treatment concentration for consecutive experiments. Human RPE cell line ARPE-19 and primary porcine RPE were treated with erastin concentrations of between 10 and 35 µM. ARPE-19 cells were stimulated for 1, 2, 3, or 7 days, while RPE cells were treated for 3, 7, or 14 days. Cell viability was assessed using the tetrazolium bromide (MTT) assay. Viability is depicted as a percentage relative to the untreated control, for which the value is set to 100%.

All erastin treatments significantly reduced ARPE-19 viability, which was further reduced by increasing stimulation time (Figure 1A). After 1 day, viability ranged from 70.39 ± 7.79% (30 µM erastin, p < 0.001) to 81.48 ± 5.50% (15 µM erastin, p < 0.001); after 3 days, it ranged from 57.07 ± 4.45% (32.5 µM erastin, p = 0.004) to 75.72 ± 7.92% (15 µM erastin, p = 0.034). From 3 days on, viability was attenuated to a greater extent and ranged from 18.81 ± 4.52% (22.5 µM erastin, p = 0.001) to 32.06 ± 8.66% (10 µM erastin, p = 0.005), while after 7 days, it reached the lowest values: between 12.86 ± 9.17% (15 µM erastin, p = 0.004) and 24.66 ± 17.95% (10 µM erastin, p = 0.018). Erastin significantly reduced cell viability, which was more dependent on time than concentration.

Primary RPE cells were more resistant to erastin than ARPE-19 cells (Figure 1B). After 3 days of stimulation, a significant reduction in viability was achieved with 30 µM of erastin only (84.99 ± 4.36%, p = 0.027). After 7 days, there were no significant findings, although numerically, viability was reduced even further with higher doses. After 14 days of stimulation, every erastin dose tested significantly reduced cell viability, falling from 30.50 ± 12.91% (35 µM erastin, p = 0.011) to 76.71 ± 3.07% (10 µM erastin, p = 0.006).

2.1.2. Interleukin 8 Secretion

Supernatants of the ARPE-19 and RPE cells (treated as mentioned in Section 2.1.1) were applied in ELISA to determine IL8 secretion, which was normalized with the viability data.

For ARPE-19, after 1 day, the control exhibited IL8 secretion amounting to 231.15 ± 118.56 pg/mL, which was significantly reduced by 22.5 µM, 25 µM, 27.5 µM, 30 µM, 32.5 µM, and 35 µM of erastin to 54.71 ± 42.26 pg/mL (p = 0.040), 62.79 ± 35.71 pg/mL (p = 0.050), 51.32 ± 36.07 pg/mL (p = 0.030), 20.97 ± 21.51 pg/mL (p = 0.007), 4.98 ± 4.89 pg/mL (p = 0.003), and 0.00 ± 0.00 pg/mL (p = 0.002), respectively (Figure 2A). After 2 days, IL8 secretion ranged from 83.75 ± 87.87 pg/mL (20 µM erastin) to 333.72 ± 193.69 pg/mL (35 µM erastin) with no significant differences compared to the control (197.71 ± 155.26 pg/mL). After 3 days of stimulation, the data displayed great variability. All applied erastin concentrations increased IL8 levels (from 300.35 ± 14.01 pg/mL to 820.78 ± 447.87 pg/mL compared to the control, with 142.01 ± 82.59 pg/mL), but analysis of variance (ANOVA) showed no significance. After 7 days, 12.5–27.5 µM of erastin halted IL8 secretion (0.00 pg/mL), corresponding to a level that was significantly different from the control, with 256.78 ± 158.01 pg/mL (all p < 0.05).

In primary RPE, most of the erastin concentrations elevated IL8 secretion after 7 and 14 days without reaching significance (Figure 2B). After 3 days of stimulation, 25 µM of erastin increased IL8 secretion to 1500.01 ± 329.56 pg/mL compared to the control, with 2.75 ± 4.76 pg/mL (p = 0.001).

Although IL6 secretion was also assessed, it was generally undetectable.

2.1.3. Vascular Endothelial Growth Factor Secretion

Secreted VEGF was assessed in ARPE-19 cells (Figure 3). After 1 day, each erastin concentration tested increased VEGF secretion (from 1273.97 ± 308.03 pg/mL to 1802.88 ± 418.38 pg/mL) compared to the control, for which the value was 1127.09 ± 235.79 pg/mL, but this difference did not reach significance. Similarly, after 2 days, no significant effects were found, with each erastin dose elevating VEGF levels (from 1115.36 ± 132.27 pg/mL to 1459.09 ± 243.79 pg/mL) compared to the control (with 1056.84 ± 164.18 pg/mL), with the exception of 35 µM erastin, which lowered VEGF secretion. After 3 days, 10–30 µM erastin elevated VEGF secretion (from 1172.33 ± 1119.83 pg/mL to 2827.67 ± 1179.86 pg/mL) compared to the control (with 1070.00 ± 305.02 pg/mL), but this increase did not reach significance. After 7 days, 10 µM, 15 µM, and 27.5 µM erastin markedly decreased VEGF secretion to 963.33 ± 43.84 pg/mL, 103.67 ± 179.56 pg/mL, and 752.67 ± 206.02 pg/mL, respectively, relative to the control, for which the value was 1388.00 ± 55.97 pg/mL.

2.2. Protection Assay for Viability

Based on the preliminary findings, 35 µM erastin was chosen to assess the potential of fucoidan to protect ARPE-19 against loss of cell viability (Figure 4A) caused by lipid peroxidation (reduction to 50% cell viability). Cells were treated for 1 day or 2 days with 50 µg/mL of fucoidan FVs and/or 35 µM erastin. Viability was determined using the MTT assay (Figure 4). FV treatment alone led to a transient and minor viability reduction after 1 day (93.38 ± 2.77%, p < 0.001); this effect was no longer evident after 2 days (98.39 ± 3.09%, p = 0.115). Erastin treatment decreased viability significantly after 1 day (42.88 ± 27.13%, p < 0.001) and 2 days (12.47 ± 12.07%, p < 0.001). Combined treatment of FVs plus erastin resulted in a slight increase in viability after 2 days (20.76 ± 19.19%); this increase was not significant relative to single erastin stimulation.

For primary RPE (Figure 4B), erastin doses and stimulation times were chosen based on the preliminary findings concerning a reproducible and biologically relevant decline in viability. Cells were treated with 50 µg/mL of fucoidan FVs and/or 35 µM erastin (7 days) or 20 µM erastin (14 days). Both erastin applications reduced cell viability efficiently to 64.00 ± 6.90% (p < 0.001, 35 µM erastin) and 66.57 ± 10.58% (p < 0.001, 20 µM erastin), respectively. FVs did not show any significant antiproliferative effects at any of the tested time points. While combined FV-plus-erastin treatment reduced viability (7 days: 73.00 ± 13.04%, p = 0.004; 14 days: 79.29 ± 7.54%, p < 0.001) compared to the untreated controls, the values were higher than those for the pure erastin treatments and reached significance after 14 days (p = 0.025).

2.3. Protection Assay for Interleukin 8 Secretion

To evaluate the potential anti-inflammatory effects of fucoidan after lipid-induced damage, ARPE-19 cells were treated with 50 µg/mL of FVs and/or 12.5 µM erastin for 1 day (Figure 5A). The erastin dose was chosen based on preliminary tests (it induced the highest level of IL8 secretion, with viability above 70%). IL8 was assessed using ELISA. Erastin increased IL8 secretion significantly to 200.92 ± 32.22 pg/mL (p = 0.009) compared to the control (82.00 ± 31.00 pg/mL). FV treatment alone reduced IL8 secretion significantly to 11.24 ± 31.45 pg/mL (p = 0.010. Combined erastin-plus-FV treatment (50.14 ± 44.36 pg/mL) led to a significant reduction compared to treatment with erastin alone (p < 0.001).

Primary RPE cells were treated with 50 µg/mL of FVs and/or 25 µM erastin for 3 days (Figure 5B) or with 10 µM erastin for 7 days (Figure 5C). Erastin doses were chosen based on preliminary tests (highest IL8 secretion, and viability not significantly lowered). Both tested erastin concentrations increased IL8 levels significantly to 975.40 ± 840.68 pg/mL (p < 0.001, 25 µM erastin) and 736.35 ± 614.23 pg/mL (p < 0.001, 10 µM) compared to the control (with 50.00 ± 101.00 pg/mL and 23.80 ± 75.30 pg/mL, respectively). FVs reduced IL8 secretion at both time points to 0.00 ± 0.00 pg/mL. After 3 days, combined erastin-plus-FV treatment led to significantly higher IL8 levels, with 300.25 ± 349.86 pg/mL (p = 0.020), which was lower than treatment with erastin alone. After 7 days, combined FV-plus-erastin treatment led to significantly lower IL-8 secretion than pure erastin stimulation, with 80.70 ± 167.77 pg/mL (p = 0.004). Similarly to ARPE-19, FVs were to exert anti-inflammatory effects during lipid peroxidation.

2.4. Protection Assay for Vascular Endothelial Growth Factor Secretion

To evaluate the potential antiangiogenic effects of fucoidan after lipid-induced damage, ARPE-19 cells were treated with 50 µg/mL of FVs and/or 12.5 µM erastin or 20 µM erastin for 1 day (Figure 6A). An erastin dose of 20 µM was chosen based on preliminary tests (it induced the highest level of VEGF secretion, and viability was not significantly lowered), and 12.5 µM erastin was also tested. Supernatants were applied in ELISA to detect secreted VEGF (Figure 6). We found that 12.5 µM erastin led to a significant increase in VEGF secretion, with 1774.48 ± 167.26 pg/mL (p < 0.001, compared to the control (1323.19 ± 187.00 pg/mL)). FVs significantly reduced VEGF secretion, resulting in 985.51 ± 182.64 pg/mL (p < 0.001, 12.5 µM erastin experiment) or 1038.64 ± 324.51 pg/mL (p = 0.004, 20 µM erastin experiment). Combined treatment with FVs and 12.5 µM erastin (1268.93 ± 213.29, p < 0.001), as well as FVs and 20 µM erastin (1416.22 ± 474.03, p = 0.048), led to significant attenuation of VEGF secretion compared to erastin alone.

Primary RPE cells were treated with FVs and/or 25 µM erastin for 3 days or 10 µM erastin for 7 days (the same supernatants mentioned in the IL8 experiments in Section 2.3 were used, and there were no preliminary VEGF tests) (Figure 6B). Erastin was not observed to have had any significant effects on VEGF secretion. After 3 days, FVs significantly reduced VEGF secretion to 560.01 ± 31.54 pg/mL (p = 0.001) compared to the control (684.59 ± 52.17 pg/mL). In addition, combined FV-plus-erastin treatment had significant anti-VEGF effects (3 days: 550.23 ± 42.70 pg/mL, p < 0.001; 7 days: 331.99 ± 167.56 pg/mL p = 0.005) compared to the control and the erastin-alone treatments (3 days: 648.96 ± 52.41 pg/mL, p = 0.006; 7 days: 563.04 ± 138.86 pg/mL, p = 0.017). The results indicate that FVs have anti-VEGF effects irrespective of lipid-induced stress.

2.5. Reactive Oxygen Species

To detect generation of ROS due to erastin treatment and potential scavenging effects of FVs, ARPE-19 cells were treated with FVs and/or erastin in a concentration range of 10–35 µM for 30 min. The 2′,7′-dichlorofluorescein assay (DCF) was used to detect ROS generation (Figure 7). Increasing the erastin concentration led to a significant dose-dependent increase in ROS formation from 20 to 30 µM (from 2134.83 ± 1187.21 to 2331.06 ± 1153.83 in terms of relative fluorescence intensity compared to the control, for which relative fluorescence intensity was equal to 586.07 ± 319.06). At 30 µM erastin and above, ROS formation decreased, which might be related to cell death, as a decrease in the number of cells leads to a decrease in quantities of ROS. FVs showed fewer ROS in the absence of erastin relative to the untreated control; however, the difference was not statistically significant. Similarly, combined treatment with FVs and 30 µM erastin led to a decrease in ROS, so the increase was no longer significant. Taken together, the ROS-scavenging effects of FVs during erastin stimulation are minor and might not be of biological relevance. This type of assay could not be conducted using primary porcine RPE cells, as they do not grow properly in 96-well plates (which are indispensable for the DCF assay employed).

2.6. Gene Expression

To assess the effects on gene expression after lipid-induced damage and choose optimal parameters for fucoidan assays (and endogenous controls), primary RPE cells were treated with 15, 20, and 25 µM erastin for 4 h, 24 h, and 3 days (Figure 8). RNA was isolated, transcribed to cDNA, and applied in qPCR using porcine gene expression assays [22]. Markers for angiogenesis (ANGPT2), oxidative stress (GPX4), and inflammation (CXCL8–IL8) were tested. Endogenous controls—18S rRNA, ACTB, GAPDH, and GUSB—were employed. Data are depicted as relative gene expression compared to the untreated control (=1.0). After 4 h of stimulation, ANGPT2 showed an erastin-induced dose-dependent reduction in gene expression, which reached significance with 25 µM erastin (RQ = 0.18 ± 0.17, p = 0.003). In addition, after 24 h, 25 µM erastin resulted in the lowest ANGPT2 expression, with RQ = 0.46 ± 0.27, but without reaching statistical significance. After 3 days, 25 µM erastin significantly reduced GPX4 expression (RQ = 0.56 ± 0.15, p = 0.040). GPX4 exhibited the lowest expression after 4 h (RQ = 0.54 ± 0.96), without reaching significance. IL8 gene expression showed no significant reactions to erastin, but it was highest after 24 h of stimulation using 25 µM erastin (RQ = 1.96 ± 1.28).

Further tests with fucoidans were conducted with 25 µM erastin and 4 h of stimulation because ANPGT2 showed the strongest significant gene expression change out of all the genes tested in the preliminary tests (Figure 9). ANGPT2 expression was reduced similarly by erastin (RQ = 0.61 ± 0.28) and FVs (RQ = 0.67 ± 0.14). Combined treatment with FVs and erastin significantly reduced ANGPT2 expression to a greater extent (RQ = 0.50 ± 0.13, p = 0.009). KDR expression was noticeably but not significantly reduced by erastin (RQ = 0.42 ± 0.45) and combined FV-plus-erastin treatment (RQ = 0.39 ± 0.18). FLT1 expression was noticeably lowered by FVs (RQ = 0.63 ± 0.13), whereas erastin (RQ = 3.75 ± 1.67, p = 0.030) significantly increased it. FVs led to significantly lower FLT gene expression relative to erastin treatment (p < 0.001). IL6 expression was similarly reduced by erastin (RQ = 0.13 ± 0.33) or combined FV-plus-erastin treatment (RQ = 0.13 ± 0.02) without reaching significance. CXCL8 expression, on the other hand, was markedly reduced by FVs (RQ = 0.25 ± 0.12). No biological relevant effects were observed for VEGFA, NOS2, CFH, SOD1, GPX4, or GSS.

2.7. Protein Expression

To test the influence of lipid-induced damage on primary RPE, cells were treated with 50 µg/mL of FVs and/or 25 µM erastin for 3 days or 10 µM erastin for 7 days (the same conditions used for the protein secretion assays discussed in Section 2.3 and Section 2.4 were employed). Protein lysates were employed in a Western blot to assess protein expression for markers of rod visual cycle (RPE65, Figure 10A), regulation of glutathione-dependent antioxidative measures (GPX4, Figure 10B), and complement regulation (CD59, Figure 10C). Regarding RPE65 expression, there were no significant findings. After 3 days of stimulation, erastin noticeably reduced RPE65 expression to 0.46 ± 0.49 [arb. unit], whereas FVs increased it to 1.39 ± 0.41 [arb. unit] compared to the control (with 0.64 ± 0.48 [arb. unit]). After 7 days, erastin stimulation noticeably reduced RPE65 expression to 0.74 ± 0.65 [arb. unit] compared to the control (with 2.50 ± 1.72 [arb. unit]). Both pure FVs and combined FV-plus-erastin treatment led to increased expression. The increase for the combined treatment was noticeably higher than treatment with erastin alone, with a value of 3.13 ± 1.69 [arb. unit]. All conditions, especially treatment with FVs, led to higher RPE65 expression at 7 days compared to 3 days.

Concerning GPX4, after 3 days, there were no significant findings, but erastin decreased GPXX4 expression to 0.48 ± 0.23 [arb. unit], whereas FVs increased expression, corresponding to a value of 5.45 ± 6.16 [arb. unit], compared to the control, for which the value was 3.47 ± 2.44 [arb. unit]. After 7 days, erastin treatment significantly reduced GPX4 expression to 0.24 ± 0.09 [arb. Unit] (p = 0.010), and combined FV-plus-erastin treatment reduced it to 0.37 ± 0.07 [arb. Unit] (p = 0.030). The effect of the combined treatment was noticeably greater than that of treatment with erastin alone. In general, GPX4 expression was lower at 7 days than at 3 days.

Regarding CD59 expression, there were no significant findings, as it remained relatively stable. Altogether, FVs were observed to have protective and activating effects on RPE65 and, to a lesser extent, GPX4. CD59 was not influenced by fucoidan or lipid-induced damage.

2.8. Tests Using Organ Cultures

2.8.1. Interleukin 8 Secretion in Organ Cultures

Organ cultures from RPE/choroids cultivated in Ussing chambers were treated with 50 µg/mL of FVs and/or 25 µM erastin for 3 days. Parameters suitable for primary RPE were selected. Supernatants were collected each day. After 3 days, tissue was stained with calcein and assessed with a fluorescence microscope (Figure 11). The confluence of viable cells was estimated visually. Qualitative assessment showed that erastin-treated (38.57% ± 20.35% viable cells per photo) tissue showed fewer viable cells than the untreated control (77.14% ± 9.51% viable cells per photo), the tissue subjected to treatment with FV alone (82.14% ± 6.99% viable cells per photo), and the tissue subjected to combined treatment with FVs plus erastin (47.14% ± 27.52% viable cells per photo).

Supernatants were analyzed for secreted IL8 (Figure 12). Concerning apical secretion (Figure 12A), after 2 days, erastin and combined erastin-plus-FV treatment significantly reduced IL8 secretion to 279.84 ± 288.41 pg/mL (p < 0.001) and 509.72 ± 515.82 pg/mL (p < 0.001) compared to the control (with 2472.98 ± 885.26 pg/mL). After 3 days, erastin and combined erastin-plus-FV treatment significantly reduced IL8 secretion to 35.25 ± 49.17 pg/mL (p < 0.001) and 116.70 ± 141.91 pg/mL (p < 0.001) compared to the control, for which the value was 2233.79 ± 601.32 pg/mL. From days 1 to 3, IL8 secretion was significantly lowered by erastin and combined FV-plus-erastin treatment.

Concerning basolateral secretion (Figure 12B), after 1 day, erastin and combined erastin-plus-FV treatment significantly decreased IL8 secretion to 2023.55 ± 644.86 pg/mL (p = 0.008) and 2169.70 ± 755.80 pg/mL (p = 0.023) compared to the control, for which the value was 3158.98 ± 618.72 pg/mL. After 2 days, erastin and combined FV-plus-erastin treatment significantly reduced IL8 secretion to 663.50 ± 330.99 pg/mL (p < 0.001) and 1029.70 ± 577.99 pg/mL (p < 0.001) compared to the control, for which the value was 3062.27 ± 525.60 pg/mL. After 3 days, erastin and combined FV-plus-erastin treatment significantly reduced IL8 secretion to 262.89 ± 144.14 pg/mL (p < 0.001) and 412.17 ± 319.47 pg/mL (p < 0.001) compared to the control, for which the value was 3180.27 ± 496.51 pg/mL. From days 1 to 2, IL8 secretion was significantly lowered by erastin (p = 0.002) and combined FV-plus-erastin treatment (p = 0.030).

Normalized three-day secretion (Figure 12C) revealed a reduction in IL8 secretion at the basolateral site for erastin treatment, amounting to 753.50 ± 439.52 pg/mL (p < 0.001), and for combined FV-plus-erastin treatment, amounting to 1260.64 ± 1103.55 pg/mL (p < 0.001), compared to the control, for which the value was 4179.62 ± 854.10 pg/mL. The combined stimulation was descriptively higher than that induced by pure erastin treatment without reaching significance. Secretion at the apical site was significantly decreased to 191.90 ± 307.03 pg/mL (p < 0.001) by erastin and to 176.99 ± 191.94 pg/mL (p < 0.001) by the FV-plus-erastin treatment relative to the control, for which the value was 2940.91 ± 883.65 pg/mL. The apical site showed lower IL8 secretion than the basolateral site for the control (p = 0.020). Lipid-induced stress reduction in IL8 induced by erastin in RPE/choroids was demonstrated, with no relevant effects of the fucoidan observed.

2.8.2. Vascular Endothelial Growth Factor Secretion in Organ Cultures

The supernatants discussed in Section 2.8.1 were analyzed for VEGF secretion (Figure 13). Concerning apical VEGF secretion (Figure 13A), there were no significant findings. The combined FV-plus-erastin treatment led to lower VEGF secretion relative to the other conditions. Concerning basolateral secretion (Figure 13B), again, there were no significant findings, but combined stimulation with FVs plus erastin led to less VEGF secretion than in the other conditions. After normalization with viable cells (Figure 13C), at the apical site, erastin noticeably increased VEGF secretion to 486.48 ± 358.34 pg/mL compared to the control, for which the value was 182.35 ± 99.73 pg/mL, while the combined FV-plus-erastin treatment led to the least VEGF secretion, with 170.26 ± 175.22 pg/mL. At the basolateral site, erastin elevated VEGF secretion to 746.03 ± 300.45 pg/mL compared to the control, for which the value was 541.57 ± 93.95 pg/mL, while combined FV-plus-erastin treatment resulted in the least VEGF secretion, with 382.88 ± 320.09 pg/mL, which is significantly less than the value for the erastin treatment (p = 0.049).

3. Discussion

In this study, we aimed to assess the potential protective effects of fucoidans in a lipid-induced stress model using erastin. Fucoidans are well known for their beneficial effects and have been shown to exert anticoagulant, antioxidative, and anti-inflammatory effects, among others [24], rendering them very interesting for potential use in treating or preventing AMD-relevant molecular pathologies: oxidative stress, chronic inflammation, lipid dysregulation, and disturbed angiogenesis (wet AMD) [2]. In particular, commercially available (Sigma-Aldrich, St. Louis, MO, USA) pure fucoidan derived from Fucus vesiculosus has shown remarkable potential in studies concerning anti-angiogenic (anti-VEGF), anti-inflammatory, and antioxidative activity [18,20,22,25]. Fucoidans’ effects on lipid dysregulation and lipid-induced stress have not been investigated in regard to primary RPE, but the beneficial effects of FVs could complete the pathological AMD quartet. In this study, the first step was to establish lipid-induced cellular stress models using erastin. Our focus was on primary porcine RPE models using a fucoidan that is commercially available and has already shown relevant bioactivity (fucoidan from Sigma-Aldrich—Cat.-No.: F8190, Batch: SLBT5471). In other studies, protective effects against erastin were observed for ARPE-19 using fucoidan derived from Laminaria hyperborea [17], while FVs demonstrated slightly protective effects against erastin [17] and H₂O₂ [26]. FVs were chemically characterized in a previous study [17]. They were shown to have a high fucose content (86.2 mol %) and a high degree of sulfation (0.61 [arb. unit]), with no proteins. The extract has a molecular weight of 52 kDa, classifying it as a high-molecular-weight fucoidan [27]. Commercial fucoidan has a purity of ≥ 95%, according to the data sheet of the provider.

Erastin, whose name stands for eradicator of oncogene RAS and small-T-antigen-expressing cells, was first described as an antitumor reagent [28]. Erastin functions as a potent activator of indirect oxidative stress and lipid peroxidation by inhibiting the system x_c⁻ cystine/glutamate antiporter [23]. This action blocks the glutathione supply of cells and thereby reduces GPX4 activity [23]. The accumulating phospholipid hydroperoxides damage the cell membrane through lipid-induced oxidative stress. This model was used to simulate the lipid-based toxic effects induced by lipofuscin and drusen [29]. In this study, we used an indirect approach to generate lipid peroxides using erastin as proxy. Alternatively, direct application of lipid peroxidation products might be employed in the future using prominent dry-AMD-related agents like malondialdehyde, isolevuglandins, or 7-ketocholesterol [30,31,32]. The influence on lipid metabolism should be investigated too [33].

First, we found that erastin could efficiently lower cell metabolism and viability, while primary RPE and the RPE/choroid explants were more robust than ARPE-19. Erastin not only leads to glutathione-based oxidative stress imbalance, but also induces lipotoxicity [34] and activates mitochondrial ROS formation [35]. ARPE-19 cells might be more sensitive to erastin; they are not as differentiated as primary RPE cells and have a weaker cell barrier [36]. Primary RPE cells are described as highly resistant to H₂O₂-induced apoptosis [37]. Protective effects as well as pro-inflammatory and pro-angiogenic protein secretion (specific for the RPE models used) were assessed. In other studies, lower concentrations of erastin and short-term stimulation were usually used (such as 10 µM for 24 h [38,39]) while, in this study, erastin was applied for two weeks and in higher concentrations. This chronic, low-grade pathological environment mimics the chronic inflammatory conditions brought on by AMD. Insights into the long-term effects of erastin were studied. FVs were observed to have moderate protective effects concerning viability in primary RPE and RPE/choroid explants, depending on the stimulation time. Any protective effects were completely absent in ARPE-19 cells after short-term treatments. ARPE-19 cells were more sensitive to an erastin insult and died after only three days of stimulation. The exact way FVs mediate this slight protection in RPE is still unknown, but it might be related to protein secretion and protein expression, which will be discussed later. Fucoidans can elevate nuclear factor erythroid 2-related factor 2 (NRF2) expression in human umbilical vein endothelial cells [40]. They have been observed to protect the expression of HO-1, NRF2, and GPX4 in hepatocytes and mice [41]. We observed a small, basic reduction in ROS quantities brought on by the application of FVs alone, but erastin-elevated ROS formation was not influenced by FVs. This once again demonstrates that high-molecular-weight fucoidans (FVs = 52 kDa [17]) do not exert antioxidative effects via simple ROS scavenging, thereby contrasting with small fucoidans (<10 kDa), which display higher reduction potential and scavenging activity [42]. ROS scavenging correlates with the polyphenol content of the algal extract [43], and the commercial fucoidan used in this study is pure. In addition, concerning the oxidative-stress-related genes examined in this study (GSS, SOD1, and GPX4) no significant influence of erastin or fucoidans was detected, although in a previous study, LPS-induced upregulation of SOD1 was inhibited by FVs [22]. At the protein level, we observed that FVs had a protective effect on GPX4 protein expression after erastin application, which is in accordance with previous studies involving GPX4 expression in ARPE-19 [17]. Fucoidans can enter cells via clathrin-mediated endocytosis [44]. They increase the activity of antioxidant compounds like super oxide dismutase 1 [45] via direct interaction with proteins or transcription factors by, for example, augmenting NRF2 or downstream antioxidative defenses [46].

VEGF secretion in ARPE-19 and the RPE/choroid was increased by erastin, whereas primary RPE cells were not affected. VEGF strongly correlates with cell survival in the RPE and retina [47]. The primary RPE cells were more robust against erastin-based cell death and VEGF induction. VEGF secretion can be induced by oxidative stress [48]. ARPE-19 cells are more sensitive to oxidative stress than primary RPE cells, which have highly developed natural defense mechanisms [49]. This may explain why VEGF secretion can be activated by erastin, an indirect inducer of oxidative stress, in ARPE-19 cells. ANGPT2, which was investigated in this study, maintains vascular integrity [50]. ANGPT2 gene expression was lowered by FVs, even in the presence of erastin. Angiopoietins may play an important role in the biological activities described in this study, and this topic should be investigated in future studies. The anti-VEGF effect of FVs, as previously described [25], was found after a stress insult induced by erastin in this study. In RPE/choroid explants (in which apical and basolateral sites were treated with FVs and erastin), erastin increased VEGF secretion bidirectionally, while the anti-VEGF activity of FVs was significant at the basolateral choroid site. This might lead us to the conclusion that while lipid-induced VEGF secretion occurs at both cell sites, fucoidan inhibits VEGF more effectively at the basolateral site. Interaction with specific basolateral receptors like VEGF receptor 2 (KDR) and VEGF receptor 3 (FLT-4) could occur. It was demonstrated that KDR and FLT-4 are prominently expressed at the choroidal site and in the endothelium of the choriocapillaris facing the RPE, contributing to protection and vessel supply. VEGF receptor 1 (FLT1) is distributed equally on both RPE sites [51]. Erastin increased gene expression of FLT1 in primary RPE cells, and this increase was attenuated by FVs. Typically, FLT1 is antiangiogenic, serving as a decoy receptor, in its soluble form and proangiogenic, but less effective, in its membrane-bound form [52]. It has been shown that the inhibition of FLT1 (via inhibition of PLG) reduces recruitment of macrophages and secretion of pro-inflammatory cytokines [53]. Conversely, VEGF-induced RPE permeability is mediated through FLT1 [54]. Hypoxia-inducible factor HIF-1α is positively correlated with VEGF expression and its receptors. We assume that erastin increases FLT1 gene expression via HIF-1α, which is inducible by ROS [55] and for which the FLT1 promoter contains binding sequences [54]. FVs might decrease HIF-1 α activity, as Li et al. found (in THP-1 cells) an increase in ubiquitination via prolyl hydroxylase [56]. Further investigation is necessary to investigate the involvement of HIF-1α.

FVs could interfere with membrane-bound FLT1, as fucoidans are known to interfere with membrane-bound cell-surface VEGF receptors [57], contributing to fucoidans’ antiangiogenic effects [25]. This interference with VEGF-related receptors might inhibit self-regulated autocrine VEGF loops and gene expression [7]. As erastin increases oxidative stress [58], this might be related to the greater degree of VEGF-related pathway activity [48] and increased FLT1 gene expression. The role and pathways of FLT1 must be investigated further.

As erastin induces cell death in organ cultures, normalizing the VEGF data leads to greater values and significance. While this is a sound method for normalizing cell viability, we are aware that interpretations should be made with caution, as cell death could actually interfere with the cells’ VEGF secretion activity.

In this study, IL8 secretion was elevated by erastin both in ARPE-19 and primary RPE cells. Iron-dependent stress has been described as being linked to neurodegenerative and inflammatory pathologies such as IL8 secretion [59]. Erastin affects IL8 and the signal transducer and activator of transcription 3 axis in terms of metastasis of tumor cells or the cell migration of macrophages [60]. In our study, FVs displayed anti-inflammatory effects by reducing IL8 secretion after erastin stimulation. Again, the mechanism is unknown, but the effects are consistent with anti-IL8’s effects after the activation of toll-like receptor 3 (TLR3) and TLR4 by Poly I:C or LPS, respectively, as observed in other studies [21,22]. In tumor cells, fucoidans can activate TLR4, inducing ROS formation and apoptosis [61]. Why fucoidans acted differently in our non-transformed cells should be further investigated. Fucoidans are known for their anti-tumoral activity and can exhibit different types of activity in non-transformed cells [26,62]. IL8 release in RPE/choroid explants was markedly reduced by erastin treatment. This may be related to cell death, as demonstrated with calcein staining, as there are fewer cells secreting IL8. This reduction was found at both the apical and basolateral sites. At the same time, as expected, basolateral secretion was always greater than apical secretion, irrespective of the treatment employed [63]. FVs, however, did not reduce basolateral RPE/choroid IL8 release, a result that contrasts with the primary RPE. IL8 has many functions, such as ensuring endothelial cell survival, angiogenesis, and neutrophil recruitment [64]. Further detailed studies addressing IL8 during lipotoxicity in RPE cells are warranted. Conversely, IL6 was not inducible by erastin in either the ARPE-19 or RPE cells. Therefore, IL6 does not seem to be involved in RPE lipid-induced damage pathways.

No effects pertaining to CD59 (all complement pathways) or complement factor H (CFH, alternative pathway) expression were found during the lipid-induced damage brought about by erastin. Little is known about the role of the complement system after lipid-induced damage or cross interaction. One study described an interaction of lipid-induced stress with the classical pathway components complement factors C3 and C4 in a myocarditis mouse model [65]. In this study, however, we demonstrated that erastin (and fucoidan) were not involved in bioactivity concerning regulatory complement factors CD59 and CFH, but this does not exclude involvement of complement activation. To the best of our knowledge, we have demonstrated the qualitative downregulation of RPE65—a key enzyme in the rod visual cycle—by erastin for the first time. Remarkably, FVs showed protective effects against this erastin-induced downregulation by elevating RPE65 levels. Similarly, an LPS-induced reduction in RPE65 expression was counteracted by this fucoidan [20]. RPE65 contributes to an iron-depended retinoid visual cycle and uses iron as a cofactor [66]. Fucoidans could increase p-AMPK (phosphorylated 5′ adenosine monophosphate-activated protein kinase) protein levels in podocytes [67] and HepG2 cells [68]. P-AMPK can control protein synthesis via negatively regulating mTORC1 (mammalian target of rapamycin complex 1), promote autophagy [69], and phosphorylate and inactivate ACC (acetyl-CoA-carboxylase), resulting in attenuated biosynthesis of polyunsaturated fatty acids [70]. The lipid-induced damage mediated by erastin and related inflammatory reactions of the cells could interfere with RPE65 too [11], while fucoidan blocks TLR3 or TLR4 to prevent this. These are possible explanations for a general protective effect—resulting in, for example, increased protein expression of RPE65 and GPX4—but further studies are needed to provide more evidence.

The modes of action and mechanistic interactions of fucoidan- and ferroptosis-related pathways might be relevant, because erastin is an inducer of ferroptosis [23]. NRF2 and heme oxygenase-1 (HO-1) are active in both ARPE-19 and RPE cells [71], and they are involved in the regulation of ferroptosis [72]. High expression of NRF2 and HO-1 may lead to a rupture of the tight junctions due to axial elongation of the RPE [73]. Zhao et al. demonstrated a positive correlation between ferroptosis and VEGF expression while applying ferroptosis over choroidal neovascularization induction by laser or solute carrier family 7 member 11 inhibitor in mice and ARPE-19 cells [74]. Proangiogenic angiopoietin-like 4 (ANGPT4) can activate ferroptosis by promoting NADPH oxidase 2 and ROS formation [75] and is involved in lipid metabolism [76]. Hypoxia-inducible factor-1α (HIF-1α) was shown to counteract ferroptosis and could play a role in fucoidan bioactivity [77]. Ferroptosis, an iron-dependent form of cell death, might have a direct connection with respect to RPE65 activity. This finding is supported by previous studies, which showed degeneration of photoreceptors and disturbances in retinal recycling due to ferroptosis [78]. Further studies including ferroptosis markers like SLC7A11, ACSL4, PTGS2, NRF2, and HO-1 as well as ferroptosis-positive controls should be conducted to explore the full potential of these marine compounds [41,79,80].

The main limitations of this work include the lack of vehicle controls and in vivo validation. Future studies should address these aspects to confirm the mechanistic basis of the protection observed.

In this study, erastin-based stress models were used to test a promising fucoidan in regard to the prevention and treatment of AMD. The beneficial effects of a fucoidan derived from Fucus vesiculosus regarding lipid-induced pathological modulation were demonstrated in different RPE models. The fucoidan shows promising activity for lipid-induced proangiogenic and pro-inflammatory protein secretion as well as GPX4 and RPE65 reduction. Studies with fucoidans involving more in-depth analysis of molecular pathways and receptor interactions, lipid metabolism, and in vivo models are planned.

4. Materials and Methods

4.1. Cell Culture, Preparation, and Organ Cultures

The immortalized human ARPE-19 cell line was purchased from the American Type Culture Collection (ATCC, Manassas, VA, USA; RRID: CVCL_0145; #CRL-2302) [36]. Cells were split every week in a 1:4 ratio and seeded into 12-well plates with a density of 100,000 cells per mL (1 mL per well, treated after one week at full confluence) or 96-well plates (100 µL per well, treated after 24 h, subconfluent) (plates from Sarstedt, Nümbrecht, Germany). Cultivation media consisted of phenol red free HyClone DMEM (Cytiva, Marlborough, MA, USA) with 1% penicillin–streptomycin (Pen/Strep, Merck, Darmstadt, Germany), 1% MEM non-essential amino acid solution without L-glutamine (Pan-Biotech, Aidenbach, Germany), and 10% Gibco fetal bovine serum (Thermo Fisher Scientific, Waltham, MA, USA).

Primary RPE cells were isolated from porcine eyes (constituting waste material from food production), as previously described [81]. Eyes were cleaned of attached tissue, disinfected with betaisodonna (Mundipharma, Cambridge, UK), and washed in cold 0.9% NaCl (B. Braun, Melsunen, Germany). The cornea, iris, and vitreous body were removed using enucleation scissors, and the retina was detached with tweezers and micro scissors in phosphate-buffered saline (D-PBS, Pan-Biotech) containing 1% Pen/Strep. The uvea was treated with trypsin (Pan-Biotech) for 10 min at 37 °C. Afterwards, trypsin was removed, and the cells were incubated with trypsin/ethylenediaminetetraacetic acid (EDTA) (Pan-Biotech) for 35 min at 37 °C. The trypsinized RPE cells were detached, collected, and washed two times with 25 mL of media containing Gibco DMEM (Thermo Fisher Scientific), 1% MEM non-essential amino acid solution without L-glutamine, 1 mM sodium pyruvate (Pan-Biotech), and 10% Gibco fetal bovine serum. Cells were seeded with a density of 100,000 cells/mL into 12-well (1 mL per well) or 24-well plates (500 µL per well). Cells were continuously fed every 3–4 days. After 14–28 days of cultivation, the cells were used for experiments.

RPE/choroid explants were prepared as described previously [82]. The initial procedure is the same one used for RPE cell preparation, up until the removal of the vitreous body and iris. With micro scissors and tweezers, the choroid, with the RPE attached, was carefully cut from the connected sclera and put into fixation rings (Minucells & Minutissue, Bad Abbach, Germany) directly after removal of the retina. The explant was transferred to modified Ussing chambers (provided by the Technical Faculty of Kiel University) with separate media chambers for the apical and basolateral sites. In each chamber, 1.1 mL of media was added, consisting of DMEM/F12 (1:1) (Pan-Biotech) with 1% Pen/Strep, 0.2% CaCl₂ (Sigma-Aldrich), 0.04% taurine (Sigma-Aldrich), and 10% Gibco™ fetal bovine serum (Thermo Fisher Scientific).

All cell cultures and organ cultures were kept in a humidified incubator at 37 °C with 5% CO₂.

4.2. Stimulation

Preliminary tests were conducted to determine suitable erastin concentrations and stimulation times. ARPE-19 cells were treated with 10–35 µM erastin (Cayman Chemical, Ann Arbor, MI, USA) for 1–7 days, whereas primary RPE cells were treated with 10–35 µM erastin for 3–14 days. The stimulation time for RPE cells was longer, as they were more resistant to the erastin insult. Cell viability was determined (Section 4.3.), and supernatants were collected and applied in VEGF and IL8 ELISAs (Section 4.5.).

For all further tests, 50 µg/mL of fucoidan (FVs, Sigma-Aldrich, Cat No. F8190, Lot: SLBT5471, ≥95% purity) was applied to the cells 30 min before additional erastin stimulation. The concentration of fucoidan and stimulation procedure were employed as established in previous studies [21,25]. Stimulation was conducted according to the method described in

Table 1. Stimulation overview. Listed are all experimental parameters used in this study concerning assays (research question), biological models, plates, stimulation concentrations, stimulation times, and endpoint analysis. Fucoidan FVs were applied 30 min before erastin at 50 µg/mL. RPE = primary porcine retinal pigment epithelium, ARPE-19 = human RPE cell line, d = days, MTT = tetrazolium bromide viability assay, VEGF = vascular endothelial growth factor, IL = interleukin, DCF = 2′,7′-dichlorofluorescein, ROS = reactive oxygen species.

Assay	Model	Plate	Stimuli	Stim. Time	Analysis
preliminary tests	RPE	12-Well	erastin (10–35 µM)	3, 7, or 14 days	MTT, IL8 ELISA
	ARPE-19	12-Well	erastin (10–35 µM)	1, 2, 3, or 7 days	MTT, IL8/VEGF ELISA
protection	RPE	12-Well	FVs; 35 µM (7 day) or 20 µM (14 days) erastin	7 or 14 days	MTT
	ARPE-19	12-Well	FVs; 35 µM erastin	1 day or 2 days	MTT
ROS formation	ARPE-19	96-Well	FVs; 10–35 µM erastin	30 min	DCF
IL8 secretion	RPE	12-Well	FVs; 25 µM (3 days) or 10 µM (7 days) erastin	3 or 7 days	IL8 ELISA
	ARPE-19	12-Well	FVs; 12.5 µM erastin	1 day	IL8 ELISA
VEGF secretion	RPE	12-Well	FVs; 25 µM (3 days) or 10 µM (7 days) erastin	3 or 7 days	VEGF ELISA
	ARPE-19	12-Well	FVs; 12.5 µM or 20 µM erastin	1 day	VEGF ELISA
gene expression	RPE	12-Well	15, 20, 25 µM erastin	4, 24, or 72 h	qPCR
	RPE	12-Well	FVs; 25 µM erastin	4 h	qPCR
protein expression	RPE	12-Well	FVs; 25 µM (3 days) or 10 µM (7 days) erastin	3 or 7 days	Western blot
organ culture	RPE/choroid	Ussing chamber	FVs; 25 µM Erastin	1, 2, or 3 days	calcein, IL-8/VEGF ELISA

. This fucoidan was shown to be endotoxin-free in a previous study [83]. Fucoidan and erastin were not tested in vehicle controls, as no relevant effects of solvents were expected due to the small amounts.

4.3. Cell Viability—Tetrazolium Bromide and Calcein Assay

To determine the cell viability of ARPE-19 or RPE cells, the cells were washed with PBS after stimulation and treated with 0.5 mg/mL of tetrazolium bromide staining solution (4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide, MTT; Sigma-Aldrich) in HyClone DMEM without phenol red for 2 h according to the method of Riss et al., with modifications [84]. MTT was removed, and cells were dissolved in dimethyl sulfoxide (Carl Roth, Karlsruhe, Germany). Samples were measured at 550 nm using Elx800 BioTek (Bad Friedrichshall, Germany).

To determine the cell viability of the RPE/choroid explants, a calcein assay was conducted [85]. The rings were removed from the Ussing chambers, transferred to 12-well plates, and treated with 10 µM calcein-acetoxymethyl (AnaSpec, Fremont, CA, USA) for 30 min at 37 °C. The rings were washed twice with PBS, and fluorescence visualization was conducted using an AxioVert 160 microscope (Carl Zeiss, Oberkochen, Germany) absorbance: 494 nm, emission: 517 nm).

4.4. Detection of Reactive Oxygen Species

To detect ROS formation in ARPE-19, the DCF assay (Abcam, Cambridge, UK), modified according to the protocol of the provider, was used. Cells were grown as triplicates in 96-well plates. After 24 h of cultivation, the cells were stimulated with 50 µg/mL of FVs for 30 min. The media in all the wells were carefully aspirated, and 50 µL of DCFDA 1X solution was applied. The plates were incubated for 45 min at 37 °C with 5% CO₂. They were then emptied by tapping them on a towel and washed with PBS. Erastin solution was applied in different concentrations in a 100 µL medium. The plates were incubated at 37 °C for 10 min, washed with PBS, and measured in fresh PBS with a Tecan Spark Reader (Tecan, Männedorf, Switzerland) (excitation: 485 nm, emission: 535 nm).

4.5. Enzyme-Linked Immunosorbent Assay

To determine secreted amounts of IL8 and VEGF, supernatants of cells or explants were analyzed with a corresponding ELISA kit after the stimulation times mentioned above. To ensure placement in a linear measurement range, the cell medium with reagents was renewed according to standardized timing. For the VEGF content in RPE, stimuli were renewed 4 h before supernatants were taken; for all other measurements, stimuli were renewed 24 h before supernatants were taken. DuoSet ELISA for IL8 and VEGF were used (R&D Systems, Minneapolis, MN, USA). ELISAs were conducted according to the providers’ instructions. Data were normalized with cell viability data from the MTT assay (cells) or calcein assay (organ cultures).

4.6. Real-Time Polymerase Chain Reaction

To prepare RNA, a NucleoSpin RNA Mini Kit including DNase was used (Macherey-Nagel, Düren, Germany). To measure the concentration and purity of the RNA samples, NanoDrop™ One was utilized (Thermo Fisher Scientific). A High-Capacity cDNA Reverse Transcription Kit was used to prepare 1500 ng of cDNA (Thermo Fisher Scientific). Quantitative qPCR was performed with TaqMan gene expression assays (Thermo Fisher Scientific) and TaqMan Fast Advanced Master Mix (Thermo Fisher Scientific). Gene expression assays for IL6, CXCL8, IL1B, NOS2, CFH, SOD1, GPX4, GSS, ANGPT2, VEGFA, FLT1, and KDR were used. Endogenous controls were preselected from 18S rRNA, ACTB, GAPDH, and GUSB. GAPDH and ACTB were determined appropriate by referring to the M score. All steps were performed as instructed by the manufacturers.

4.7. Western Blotting

RPE65, CD59, and GPX4 expression was assessed using Western blotting, as described elsewhere [20]. After being washed and detached with trypsin/EDTA, the cells were lysed for 45 min in Nonidet^® P-40 lysis buffer (Sigma-Aldrich) on ice and then subjected to centrifugation steps. To determine protein concentrations, a DC Protein Assay Kit II (Bio-Rad, Munich, Germany) was utilized as instructed by the manufacturers, and a Tecan Spark Reader was used for measurement. Fifteen micrograms of protein was applied in SDS-PAGE using 12% acrylamide gels with consecutive wet-tank Western blots. Membranes were blocked with 4% skimming milk (Carl Roth) and treated with 2% skimming milk antibody solutions containing mouse anti-RPE65 (Abcam, Cambridge, UK; 65 kDa, 1:6000), rabbit anti-CD59 (Proteintech Group, Rosemont, IL, USA; 18 kDa, 1:3000), rabbit anti-GPX4 (abcam; 22 kDa, 1:1000), or rabbit anti-β-actin (Cell Signaling Technologies, Denver, CO, USA 37 kDa; 1:1000, #4967). Secondary conjugate solutions in 2% skim milk containing anti-rabbit-IgG-HRP (Cell Signaling) or anti-mouse-IgG-HRP (Cell Signaling) were applied for 1 h. Clarity western ECL substrate (Bio-Rad, Hercules, CA, USA) was applied to detect chemiluminescent signals in a ChemiDoc MP Imaging System (Bio-Rad). Band volumes were calculated with Image Lab 6.1.0 build 7 software (Bio-Rad), and protein expression was normalized with data on β-actin.

4.8. Statistical Analysis

Microsoft Excel was used for basic data management and evaluation (Microsoft, Redmond, WA, USA). GraphPAD Prism 9 (version 9.1.1, San Diego, CA, USA) was used for statistical evaluation and visualization. Tests for normal distribution were conducted using the Shapiro–Wilk test. Parametric data were evaluated with a one-sample t-test or ANOVA with a post hoc Student’s t-test. Non-parametric data were analyzed with a Wilcoxon signed-rank test or a Kruskal–Wallis test with a post hoc Mann–Whitney test. Data from qPCR analysis were evaluated using the ΔΔCT method [86]. Data corresponding to p < 0.05 were considered significant. The sample size for all the fucoidan-related tests was at least n = 6. For preliminary tests, the sample size n was reduced to at least n = 3. These tests were conducted to choose the best parameters for erastin concentration and stimulation time and were not dedicated to stating significant research results.

5. Conclusions

In our study, we revealed further promising biological activity of fucoidan derived from Fucus vesiculosus in regard to age-related macular degeneration pathomechanisms. Lipid peroxidation was induced by erastin in in vitro models of retinal pigment epithelium. We found that fucoidan can attenuate the inflammatory and proangiogenic factors induced by erastin and protect the physiological expression of proteins like retinoid isomerohydrolase. Further studies regarding fucoidans’ effects on lipid metabolism and related pathomechanisms for AMD are highly warranted.