Freeze-Dried Extracts of Eruca sativa Selectively Inhibit Nuclear Factor-κB and Vascular Endothelial Growth Factor Levels in Retinal Pigment Epithelial Cells Authors: Minseo Cho, Jisu Lee, Jiwoo Lee, Caglar Doguer, Jeong Ho Hwang, Jung-Heun Ha Published: 2026 ABSTRACT AbstractDiabetic retinopathy and age-related macular degeneration are driven by retinal pigment epithelium (RPE) dysfunction accompanied by chronic inflammation and an upregulation of vascular endothelial growth factor (VEGF). This study investigated the protective effects of Eruca sativa freeze-dried extract against metabolic and inflammatory stress in the ARPE-19 retinal epithelial cell line. Gene expression levels and protein contents were assessed via quantitative real-time polymerase chain reaction and Western blots. VEGF secretion and cell calcium levels were quantified using enzyme-linked immunosorbent assay and Fluo-4 NW calcium assay, respectively. Freeze-dried E. sativa extract (ESF) exposure did not induce any cytotoxicity across the concentration range tested. Under thapsigargin (Tg) stimulation, ESF at 1 mg/mL significantly reduced VEGF expression and protein secretion by 32% and 34%, respectively. Tg-induced calcium mobilization within cells and the expression of unfolded protein response markers, including CCAAT/enhancer-binding protein homologous protein, binding immunoglobulin protein, and X-box binding protein-1, remained unchanged following ESF exposure. Under lipopolysaccharide stimulation, in a dose-dependent manner, ESF suppressed nuclear factor-κB (NF-κB) phosphorylation, tumor necrosis factor-α, and interleukin-1β expression; however, c-Jun N-terminal kinase signaling was not affected. Hot-air-dried E. sativa extract demonstrated contrasting regulatory profiles, which exhibited a broader modulation of endoplasmic reticulum stress and mitogen-activated protein kinase signaling. Collectively, the findings demonstrate that ESF selectively attenuates inflammatory and angiogenic responses by specifically inhibiting the NF-κB-VEGF axis. Dehydration, therefore, is a critical method determining ESF bioactivity, preserving native phytochemicals that confer focused anti-inflammatory and anti-angiogenic efficacy in RPE cells.Keywords: Eruca sativa, inflammation, NF-kappa B, retinal pigment epithelium, vascular endothelial growth factors INTRODUCTION Diabetic retinopathy (DR) and age-related macular degeneration (AMD) constitute the leading causes of irreversible visual impairment worldwide; their prevalence is increasing steadily due to an aging population and a global rise in metabolic disorders (Yongpeng et al., 2022; Hwang et al., 2023). DR affects >100 million adults, while AMD impacts ∼200 million individuals, with projections estimating an increase of up to 288 million cases by 2040 (Chaudhuri et al., 2023). Beyond clinical severity, these retinal diseases impose substantial socioeconomic burdens, as vision loss markedly compromises independence, productivity, and quality of life (Hariprasad et al., 2025). Despite distinct pathogenic origins-hyperglycemia-driven microvascular injury in DR and oxidative degenerative processes in AMD-the two disorders converge on retinal pigment epithelium (RPE) dysfunction (Ferrington et al., 2020; Grigoryan and Markitantova, 2025). The RPE monolayer plays an indispensable role in photoreceptor survival, regulation of the visual cycle, and maintenance of the outer blood-retina barrier (Taylor et al., 2021). Persistent metabolic stressors, including chronic hyperglycemia, lipid overload, and redox imbalance, disrupt RPE homeostasis and initiate inflammatory and angiogenic signaling cascades (Miller et al., 2021; He et al., 2025). Activation of nuclear factor-κB (NF-κB) and mitogen-activated protein kinase (MAPK) pathways promotes the transcriptional upregulation of tumor necrosis factor-α (TNF-α), interleukin (IL)-1β, and IL-6 (Mao et al., 2025). These proinflammatory mediators compromise tight junction integrity and facilitate leukocyte infiltration of retinal tissues (Ferreira et al., 2023). In parallel, sustained inflammation-related signaling elevates vascular endothelial growth factor (VEGF) expression, thereby enhancing vascular permeability and driving pathological neovascularization during advanced disease stages (Lee et al., 2025a). Current therapeutic strategies primarily target inflammatory and angiogenic mediators (Wang et al., 2021). However, prolonged pathway activation frequently exceeds endogenous cell defense capacity, culminating in irreversible barrier disruption and progressive retinal degeneration. Identification of therapeutic approaches capable of suppressing these pathogenic cascades while restoring RPE homeostasis, therefore, remains a critical unmet need (Bian et al., 2012). Conventional clinical management of DR and AMD relies predominantly on anti-VEGF agents and corticosteroids (Wang et al., 2021). Their long-term application is limited by procedural invasiveness, high economic burden, and cumulative adverse effects (Kim et al., 2024). Moreover, such interventions primarily address late-stage anatomical manifestations rather than metabolic and inflammatory disturbances, which precede overt pathology (Lavalle et al., 2025). These drawbacks have redirected investigative efforts toward nutritional and phytochemical strategies designed to modulate disease-associated pathways during initial, subclinical stages (Jang et al., 2025). Among emerging approaches, plant-derived bioactive compounds have attracted increasing attention due to demonstrated efficacy in regulating cell stress responses through pleiotropic mechanisms (Lee et al., 2025c). In this context, cruciferous vegetables are particularly notable as dietary sources rich in secondary metabolites; exhibiting potent anti-inflammatory and antiangiogenic activities, thereby offering a sustainable avenue for retinal neuroprotection (Ağagündüz et al., 2022). Eruca sativa Mill., commonly referred to as rocket or arugula, is a cruciferous vegetable common in Mediterranean diets and characterized by a diverse phytochemical composition, including glucosinolates, polyphenols, and essential vitamins (Bell and Wagstaff, 2019). Flavonoids such as quercetin and kaempferol found in E. sativa inhibit NF-κB and MAPK signaling pathway, reducing proinflammatory cytokine production and VEGF expression (Cavaiuolo and Ferrante, 2014). In addition, glucosinolate-derived isothiocyanates, including erucin and sulforaphane analogs, activate the nuclear factor erythroid 2-related factor 2/heme oxygenase-1 axis, thereby enhancing cellular resistance to metabolic and oxidative stress (Kim and Lee, 2023; Mthembu et al., 2023). E. sativa possesses integrated bioactivities capable of attenuating inflammation- and VEGF-associated retinal injury (Ger et al., 2026). Preservation of these functions, however, is highly dependent on processing conditions, as several phytochemicals are sensitive to thermal and oxidative degradation (Frlin et al., 2025). The phytochemical profile and bioefficacy of E. sativa are robustly influenced by dehydration methodologies (Hassan Algarni, 2025). Freeze-drying effectively preserves the native structures of glucosinolates, flavonoids, and vitamins, yielding extracts with superior antioxidant potential (Coşkun et al., 2024). In contrast, hot-air drying partially degrades heat-labile compounds and simultaneously promotes the Maillard reaction, forming products with distinct bioactivities (Pan-utai et al., 2025; Yin et al., 2025). Hot-air-dried E. sativa extract (ESH) attenuates inflammation and endoplasmic reticulum (ER) stress in ARPE-19 cells (Cho et al., 2026). However, the biological effects of freeze-dried E. sativa extract (ESF), which may retain native phytochemicals more comprehensively, remain insufficiently characterized and have not been evaluated in retinal epithelial models. The present study was designed to establish the influence of ESF on stress-induced VEGF dysregulation in ARPE-19 cells and to determine whether these effects are mediated through the modulation of inflammatory or ER stress pathways. Our hypothesis proposes that ESF attenuates inflammation-driven VEGF overexpression by restoring RPE cell homeostasis and preserving barrier integrity under metabolic stress. Through comprehensive molecular analyses, the investigation aims to clarify the potential utility of ESF as a functional nutritional strategy for preventing progressive retinal degeneration. A comparative framework addressing processing-dependent bioactivity further provides mechanistic insight into the extent to which dehydration methods dictate the therapeutic relevance of cruciferous vegetable extracts in ocular disease models. MATERIALS AND METHODS Preparation of E. sativa Mill. extract Fresh leaves of E. sativa Mill. were obtained from Greemfarm, in May 2018 and processed following a method described previously (Son et al., 2020). The pooled extracts were passed through a Whatman No. 2 filter paper (GE Healthcare), concentrated under reduced pressure employing a rotary evaporator (N-1110S-W, EYELA), and lyophilized to obtain the ESF powder, which was stored at −70°C. Cell culture and treatment Human retinal pigment epithelial cell line (ARPE-19, ATCC) were cultured in Dulbecco’s Modified Eagle Medium/Nutrient Mixture F-12 (Gibco) supplemented with 10% fetal bovine serum, 1% penicillin-streptomycin (Gibco) and 100 µg/mL gentamicin (Thermo Fisher Scientific). Cells were maintained at 37°C under a humidified atmosphere containing 5% CO2. ARPE-19 cells were seeded in six-well plates at 5×105 cells/well and cultured until reaching 80% confluence. Cells were treated with ESF at 100-1,000 µg/mL or vehicle control (0.1% dimethyl sulfoxide, Sigma-Aldrich) prior to stimulation with 1 µg/mL lipopolysaccharide (LPS, Sigma-Aldrich) or 5 µM thapsigargin (Tg, Sigma-Aldrich). Tg and LPS were applied under separate experimental settings to model distinct pathological conditions; Tg was used to induce ER stress-associated signaling, whereas LPS was employed to activate Toll-like receptor-mediated inflammatory pathways. Cell viability assay Cell viability was determined utilizing the EZ-CYTOX assay kit (DoGenBio) following a described method (Jang et al., 2025). ARPE-19 cells were seeded into 96-well plates at 2×104 cells/well via incubation for 24 h, exposed to ESF for 24 h, and assayed using colorimetry, according to the manufacturer’s protocol. The OD450 was measured with a microplate reader (SpectraMax iD3, Molecular Devices). Measurement of secreted VEGF and intracellular Ca2+ Secreted VEGF was quantified using a human VEGF enzyme-linked immunosorbent assay kit (Abcam) per the manufacturer’s protocol (Lee et al., 2025b). Intracellular Ca2+ concentration was determined employing the Fluo-4 NW Calcium Assay Kit (Thermo Fisher Scientific) following a previously described methodology (Jeong et al., 2025). ARPE-19 cells were seeded at 5.0×104 cells/well in Corning black-walled 96-well plates (Corning) and cultured overnight for adherence. ARPE-19 cells were seeded in 6-well plates at 5×105 cells/well and cultured until 80% confluence. Cells were treated with ESF at 100-1,000 µg/mL or vehicle, loaded with Fluo-4 NW dye, incubated at 37°C for 30 min, and then stimulated with 5 µM Tg. Fluorescence was measured using a SpectraMax iD3 microplate reader. Quantitative real-time polymerase chain reaction (qRT-PCR) Total RNA was extracted using the NucleoZOL reagent (Macherey-Nagel), and genomic DNA was degraded with the DNA-FreeTM Kit (Thermo Fisher Scientific) per a published protocol (Jang et al., 2024). RNA concentration and purity were evaluated spectrophotometrically with a SpectraDropTM Micro-Volume Microplate reader (Molecular Devices). Complementary DNA (cDNA) was synthesized from 1 µg of total RNA using the iScriptTM cDNA Synthesis Kit (Bio-Rad). Quantitative PCR was performed on a CFX Connect Real-Time System (Bio-Rad) utilizing SsoAdvancedTM Universal SYBRⓇ Green Supermix. The mRNA levels were normalized to glyceraldehyde-3-phosphate dehydrogenase (GAPDH) and relatively quantified using the 2−ΔΔCt method. The primer sequences used for qRT-PCR are listed in Table 1. Table 1Primer sequences used for qRT-PCRGeneDirectionPrimer sequence (5’→3’) TNF-α ForwardGGC AGT CAG ATC ATC TTC TCGReverseGGT TTG CTA CAA CAT GGG CTA IL-1β ForwardCTC GCC AGT GAA ATG ATG GCTReverseGTC GGA GAT TCG TAG CTG GAT IL-6 ForwardCTT CTC CAC AAG CGC CTT CReverseCAG GCA ACA CCA GGA GCA VEGF ForwardCTA CCT CCA CCA TGC CAA GTReverseGCA GTA GCT GCG CTG ATA GA GAPDH ForwardGGC CTC CAA GGA GTA AGA CCReverseAGG GGA GAT TCA GTG TGG TGqRT-PCR, quantitative real-time polymerase chain reaction; TNF-α, tumor necrosis factor α; IL-1β, interleukin 1β; IL-6, interleukin 6; VEGF, vascular endothelial growth factor; GAPDH, glyceraldehyde 3-phosphate dehydrogenase. Primer sequences used for qRT-PCR qRT-PCR, quantitative real-time polymerase chain reaction; TNF-α, tumor necrosis factor α; IL-1β, interleukin 1β; IL-6, interleukin 6; VEGF, vascular endothelial growth factor; GAPDH, glyceraldehyde 3-phosphate dehydrogenase. Western blot analysis ARPE-19 cells were lysed in RIPA buffer containing protease and phosphatase inhibitors (Thermo Fisher Scientific). Lysates were sonicated, centrifuged at 12,000 g for 20 min at 4°C, and the supernatants were collected for protein quantification using the bicinchoninic acid assay. Equal amounts of protein-30 µg-were separated by 10%-15% sodium dodecyl sulfate–polyacrylamide gel electrophoresis and transferred to polyvinylidene difluoride membranes. These were blocked with 5% skim milk and incubated overnight at 4°C with the primary antibodies listed in Table 2. Following washing, the membranes were exposed to horseradish peroxidase-conjugated secondary antibodies. Immunoreaction bands were visualized using SuperSignalTM West Pico PLUS (Thermo Fisher Scientific) and imaged with an Azure Biosystems system (Dublin). Densitometry was performed with the ImageJ software (National Institutes of Health), and α-tubulin served as the loading control. Table 2Antibodies used for Western blotsTypeAntibodyCatalog numberDilutionSupplierPrimary antibodyp-NF-κB30331:1,000Cell Signalingp-IκBα28591:1,000Cell Signalingp-JNK92511:500Cell Signalingα-tubulinsc-80351:1,000Santa Cruz BiotechnologyCHOP28951:2,000Cell SignalingBiP31831:1,000Cell SignalingXBP-1834181:1,000Cell SignalingCleaved caspase-396611:1,000Cell SignalingCleaved caspase-995071:1,000Cell SignalingBax27721:1,000Cell SignalingBcl-2sc-73821:1,000Santa Cruz BiotechnologySecondary antibodyAnti-rabbit IgG70741:3,000Cell SignalingAnti-mouse IgG70761:1,000Cell Signalingp-NF-κB, phosphorylated nuclear factor-κB; p-IκBα, phosphorylated inhibitor of kappa B alpha; p-JNK, phosphorylated c-Jun N-terminal kinase; CHOP, CCAAT/enhancer-binding protein homologous protein; BiP, binding immunoglobulin protein; XBP-1, X-box binding protein 1; Bax, bcl-2-associated X protein; Bcl-2, b-cell lymphoma 2. Antibodies used for Western blots p-NF-κB, phosphorylated nuclear factor-κB; p-IκBα, phosphorylated inhibitor of kappa B alpha; p-JNK, phosphorylated c-Jun N-terminal kinase; CHOP, CCAAT/enhancer-binding protein homologous protein; BiP, binding immunoglobulin protein; XBP-1, X-box binding protein 1; Bax, bcl-2-associated X protein; Bcl-2, b-cell lymphoma 2. Statistical analysis All data are presented as mean±standard deviation from three independent experiments. Differences among groups were analyzed for statistical significance using one-way analysis of variance followed by Tukey’s multiple-comparison test. P<0.05 indicated statistical significance. RESULTS Impact of ESF on cell viability in ARPE-19 A noncytotoxic concentration range was established to ensure that the molecular effects observed during mechanistic investigations were not confounded by cell toxicity. ARPE-19 cells were treated with ESF at 0-1,000 µg/mL for 24 h; none of the concentrations tested induced any statistically significant reduction in cell viability (Fig. 1). Maintenance of metabolic activity at all concentrations indicated that ESF did not exert any acute cytotoxicity in RPE cells. Based on these results, 100, 500, and 1,000 µg/mL were selected for subsequent experiments to encompass a broad, biologically tolerable exposure range. Effect of freeze-dried Eruca sativa extract (ESF) on ARPE-19 cell viability. ARPE-19 cells were treated with 0-1,000 µg/mL ESF for 24 h, and viability was assessed using the EZ-CYTOX assay. Data are presented as mean±standard deviation (n=3) relative to untreated controls. Statistical significance was evaluated using one-way analysis of variance followed by Tukey’s post-hoc test. NS, not significant. Modulation of Tg-induced VEGF expression and angiogenic responses ESF-mediated regulation of angiogenic signaling was evaluated under Tg-induced ER stress to model pathological stimuli relevant to retinal disease. Tg stimulation markedly elevated VEGF expression to ∼483% of control levels (Fig. 2A). Pretreatment with ESF significantly attenuated this induction, with respective reductions of 26%, 38%, and 32% observed at 100, 500, and 1,000 µg/mL. Influence of freeze-dried Eruca sativa extract (ESF) on vascular endothelial growth factor (VEGF) expression and intracellular calcium levels. Cells were pretreated with 0-1,000 µg/mL ESF for 3 h and stimulated with 5 µM thapsigargin (Tg) for 24 h. (A) VEGF levels were determined by quantitative real-time polymerase chain reaction and normalized to those of glyceraldehyde 3-phosphate dehydrogenase (GAPDH). (B) Secreted VEGF proteins were quantified by enzyme-linked immunosorbent assay. (C) Intracellular calcium levels were assessed using the Fluo-4 NW assay. Data are presented as mean±standard deviation (n=3). Various letters indicate statistically significant differences (P<0.05). Consistent with gene-transcription findings, VEGF secretion increased by 88% following Tg exposure (Fig. 2B). Pretreatment with ESF at 1,000 µg/mL remarkably reduced VEGF secretion by 34%. Similarly, intracellular Ca2+ was elevated by 39% following Tg stimulation; however, ESF did not induce any significant alterations in Ca2+ mobilization (Fig. 2C). Selective suppression of VEGF in the absence of Ca2+-modulation suggests that ESF regulates angiogenic signaling downstream or independently of Ca2+ flux. In contrast, ESH suppresses VEGF synthesis and Ca2+ elevation concurrently (Cho et al., 2026). Suppression of LPS-induced NF-κB activation and proinflammatory cytokine expression Inflammatory modulation was assessed in LPS-stimulated ARPE-19 cells. LPS exposure significantly enhanced the phosphorylation of NF-κB, inhibitor of κBα (IκBα), and c-Jun N-terminal kinase (JNK), by 116.7%, 2,145.7%, and 307.7%, respectively (Fig. 3A-3D), confirming the activation of canonical inflammatory cascades. Transcriptional analysis demonstrated a marked upregulation of TNF-α, IL-1β, and IL-6 (Fig. 3E-3G) following LPS stimulation. Effect of freeze-dried Eruca sativa extract (ESF) on nuclear factor-κB (NF-κB)/mitogen-activated protein kinase signaling and inflammatory cytokine expression. Cells were pretreated with 0-1,000 µg/mL ESF for 3 h and stimulated with 1 µg/mL lipopolysaccharide (LPS) for 2 h. (A) Representative Western blots. The levels of (B) phosphorylated nuclear factor-κB (p-NF-κB), (C) phosphorylated inhibitor of kappa B alpha (p-IκBα), and (D) phosphorylated c-Jun N-terminal kinase (p-JNK) proteins were normalized to those of α-tubulin. (E) tumor necrosis factor-α (TNF-α), (F) interleukin (IL)-1β, and (G) IL-6 mRNA levels were normalized to those of glyceraldehyde 3-phosphate dehydrogenase (GAPDH). Data are presented as mean±standard deviation (n=3). Various letters indicate statistically significant differences (P<0.05). Pretreatment with ESF induced a dose-dependent reduction in NF-κB phosphorylation by 34.9%, 51.4%, and 74.3% at 100, 500, and 1,000 µg/mL, respectively. The phosphorylation status of IκBα and JNK remained largely unchanged, indicating selective inhibition of NF-κB activation rather than broad suppression of upstream MAPK signaling. Correspondingly, ESF significantly inhibited TNF-α and IL-1β expression, whereas that of IL-6 was not altered markedly. Such a selective profile contrasts with the broader inhibitory effects of ESH, which suppressed both NF-κB and MAPK pathways (Cho et al., 2026). Evaluation of Tg-induced ER stress and apoptotic signaling The unfolded protein response (UPR) was investigated to identify potential cytoprotective effects on ER homeostasis. Tg exposure markedly increased the expression of CCAAT/enhancer-binding protein homologous protein (Fig. 4A and 4B), binding immunoglobulin protein (Fig. 4C), and X-box binding protein 1 (Fig. 4D). Pretreatment with ESF at ≥1,000 µg/mL did not remarkably alter the expression of these markers, indicating a limited involvement of ESF in direct UPR modulation. Effect of freeze-dried Eruca sativa extract (ESF) on the expression of endoplasmic reticulum stress- and apoptosis-related proteins. Cells were pretreated with 0-1,000 µg/mL ESF for 3 h and stimulated with 5 µM thapsigargin (Tg) for 24 h. (A) Representative Western blots. The levels of (B) CCAAT/enhancer-binding protein homologous protein (CHOP), (C) binding immunoglobulin protein (BiP), and (D) X-box binding protein-1 (XBP-1) proteins were normalized to α-tubulin. The levels of (E) cleaved caspase-3, (F) cleaved caspase-9, (G) Bcl-2-associated X protein (Bax), (H) B-cell lymphoma 2 (Bcl-2) proteins and (I) the Bax/Bcl-2 ratio was quantified. Data represent mean±standard deviation (n=3). Various letters indicate statistically significant differences (P<0.05). NS, not significant. An analysis of apoptosis-related signaling did not demonstrate any marked changes in the expression of cleaved caspase-3, cleaved caspase-9, Bax, or Bcl-2 following ESF pretreatment (Fig. 4E-4I). The Bax/Bcl-2 ratio also remained unchanged, suggesting a lack of anti-apoptotic modulation. Such a lack of ER stress and apoptosis regulation contrasts prior findings on ESH (Cho et al., 2026), which attenuated UPR activation and apoptotic pathways. These observations underscore dehydration as a critical determinant of the bioactivity of E. sativa extracts. DISCUSSION DR and AMD constitute major causes of irreversible visual impairment and share a pathological foundation characterized by progressive dysfunction and degeneration of RPE cells (Bowes Rickman et al., 2013; Zając-Pytrus et al., 2015; Kim et al., 2023). Such deterioration initiates a cascade of pathological events, including chronic inflammation (Park et al., 2024), ER stress (Zhang et al., 2014), and oxidative imbalance (Böhm et al., 2023), collectively promoting VEGF overexpression and pathological neovascularization (Klettner and Roider, 2012). Limitations of current clinical interventions, including procedural invasiveness of repeated intraocular injections and cumulative adverse effects of prolonged corticosteroid administration, have increased interest in plant-derived bioactive compounds as complementary strategies for long-term ocular protection (Formica et al., 2021; Dhallu et al., 2025). E. sativa extract represents a rich dietary source of phytochemicals with documented anti-inflammatory and cytoprotective properties (Testai et al., 2022). However, many constituents are sensitive to thermal degradation and oxidative modification (Son et al., 2020). The selection of an optimal drying method, therefore, exerts a decisive influence on chemical composition and biological efficacy. Hot-air-dried E. sativa promotes the oxidative modification of polyphenols and facilitates Maillard-type reactions, structurally transforming native phytochemicals and generating heat-derived metabolites (Cho et al., 2026). Thermal reaction products contribute substantially to anti-inflammatory and cytoprotective activities by enhancing antioxidant capacity and modulating cell-stress signaling pathways (Kuehu et al., 2024). In contrast, freeze-drying employs low-temperature and low-pressure conditions that minimize degradation of thermolabile constituents, thereby preserving native polyphenols, glucosinolates, and vitamins (Pasten et al., 2024; El-Saadony et al., 2025). ESH effectively suppressed inflammatory signaling and attenuated ER stress in ARPE-19 cells (Cho et al., 2026). The translation of these findings through an evaluation of ESF enabled a direct comparison of drying-dependent physicochemical alterations and corresponding functional outcomes in RPE cells. Induction of ER stress is commonly associated with disruption of Ca2+ homeostasis within the ER lumen, leading to cytosolic Ca2+ release and a subsequent activation of UPR pathways that contribute to VEGF upregulation (Yu et al., 2025). ESF did not alter Tg-induced elevation in intracellular Ca2+ contents or the expression of canonical UPR markers, while VEGF transcription and secretion were significantly suppressed. Divergence between VEGF regulation and ER stress signaling suggests the involvement of Ca2+-independent regulatory mechanisms (Noren et al., 2016). VEGF expression is modulated through transcriptional repression mediated by inhibition of the NF-κB-hypoxia-inducible factor-1α signaling axis (Park et al., 2014; Liu et al., 2022). NF-κB functions as the primary transcriptional regulator of VEGF under inflammatory stress (Kiriakidis et al., 2003; Shakir et al., 2026). Suppression of NF-κB activation by ESF therefore provides a mechanistically plausible explanation for its antiangiogenic effects. Preservation of native polyphenols and flavonoids through freeze-drying likely underlies a selective regulation of the transcription of inflammatory and angiogenic genes rather than a direct modulation of ER stress pathways. Pretreatment with ESF markedly suppressed NF-κB activation and the transcription of TNF-α and IL-1β (proinflammatory cytokines) in LPS-stimulated ARPE-19 cells. Notably, although IL-1β expression exhibited a modest rebound at higher ESF concentrations, the transcription levels remained substantially lower than those observed in the LPS-only group. Such an effect may reflect stress-responsive transcriptional modulation of genes rather than genuine proinflammatory activation, which was supported by the concomitant suppression of NF-κB signaling and TNF-α expression. Comparable anti-inflammatory activity has been reported for ESH; it inhibited both NF-κB and MAPK signaling pathways. Despite compositional differences arising from divergent drying processes, both extracts robustly suppressed inflammatory mediators. Conservation of core phytochemicals, including quercetin, kaempferol, and glucosinolate-derived metabolites, likely accounts for shared anti-inflammatory efficacy (De Rossi et al., 2025; Ochar et al., 2025). These compounds are intrinsic to E. sativa and exhibit a well-documented inhibition of NF-κB-mediated transcriptional activity and oxidative stress amplification (Kim et al., 2014; Neugart et al., 2018). Quercetin and kaempferol attenuate inflammatory cascades through combined antioxidant and transcription-based regulatory mechanisms (Kim et al., 2014). The anti-inflammatory activity of ESF is therefore primarily attributable to preserved native phytochemicals rather than heat-induced derivatives. Functional divergence between ESF and ESH was most evident in the regulation of ER stress signaling. ESF did not significantly modulate the expression of UPR-related genes or protein expression, whereas ESH markedly attenuated ER stress markers (Cho et al., 2026). Such divergence reflects unique phytochemical signatures generated by different drying methodologies. Freeze-drying preserves thermolabile native constituents such as vitamin C and specific flavonoids, whereas hot-air drying promotes amino acid degradation and Maillard reactions that generate bioactive thermal derivatives (Son et al., 2020; Kittibunchakul et al., 2023). Several heat-induced compounds enhance protein-folding capacity and activate endogenous antioxidant defense systems, thereby mitigating ER stress responses (Lee et al., 2020; Özcan et al., 2021). Relative to ESF, elevated levels of ER-stabilizing components, including specific nonessential amino acids and vitamin E derivatives, have been identified in ESH (Son et al., 2020). Therefore, ESF primarily exhibits anti-inflammatory and VEGF-suppressive activity, whereas ESH demonstrates broader cytoprotection through a concurrent modulation of inflammation, angiogenesis, and ER stress. The experimental framework employed was an in vitro RPE cell model, limiting direct extrapolation to complex cellular interactions and pharmacokinetic dynamics inherent to the intact retina. The lack of a comprehensive characterization of phenolic subclasses and Maillard reaction intermediates further restricts the definitive attribution of bioeffects to specific compounds. Future investigations integrating quantitative metabolomics and lipidomics with in vivo retinal disease models and diverse processing conditions may delineate precise structure-function relationships. In addition, conclusions regarding Ca2+-independent mechanisms warrant cautious interpretation, as direct modulation of Ca2+ signaling or downstream regulators of gene transcription was not experimentally interrogated. Despite these limitations, the current findings provide the first integrated comparison demonstrating that drying-induced physicochemical modifications in E. sativa extracts produce distinct functional responses in RPE cells. Selection of processing methodology should therefore be guided by intended therapeutic objectives, with freeze-drying favoring anti-inflammatory and anti-angiogenic applications and hot-air drying offering additional benefits for ER stress modulation. Footnotes FUNDING None. AUTHOR DISCLOSURE STATEMENT The authors declare no conflict of interest. AUTHOR CONTRIBUTIONS Concept and design: MC, Jisu L, Jiwoo L, CD, Jeong Ho H, Jung-Heun H. Analysis and interpretation: MC. Data collection: Jisu L, Jiwoo L, CD, Jung-Heun H. Writing the article: MC, Jisu L, CD, Jung-Heun H. Critical revision of the article: MC, Jisu L, Jiwoo L, CD, Jeong Ho H, Jung-Heun H. Final approval of the article: All authors. Statistical analysis: MC. Overall responsibility: Jung-Heun H. References