Differential Antioxidant Capacities of Human Endometriotic and Endometrial Cell Models Under H2O2 Exposure

All reagents were obtained from Sigma-Aldrich and used without further purification. NAC and H 2 O 2 solutions were prepared with ultrapure MilliQ water (MIliQ, Burlington, MA, USA), and peroxide concentration was determined before each experiment by UV-Vis (UV-1800; Shimadzu Corp., Kyoto, Japan) spectroscopy as reported in the literature [ 89 ]. Immortalized Ishikawa endometrial epithelial adenocarcinoma cells (ECACC 99040201) and human endometriotic epithelial 12Z cells (SCC443) were purchased from Sigma-Aldrich (St Louis, MO, USA) and maintained at 37 °C in a humidified atmosphere with 5% CO 2 . Ishikawa cells were cultured in MEM supplemented with 5% FBS, while 12Z cells were cultured in High-Glucose DMEM supplemented with 10% FBS. Both media were supplemented with penicillin (100 U/mL), streptomycin (10 μg/mL), non-essential amino acids, and sodium pyruvate. For 3D spheroid formation, cells were seeded in 96-well Nunclon TM Sphera TM U-bottom low-adhesion plates (Thermo Scientific, Waltham, MA, USA) and incubated for 48 h. Based on literature guidance [ 41 ], seeding densities of 0.5, 1.0, and 2.0 × 10 4 cells/well were evaluated to obtain spheroids with minimal baseline cell death (negative control) and diameters of 400–600 μm [ 41 ]. Spheroid morphology was assessed by inverted microscopy prior to the experiments. Cell viability in 2D cultures was assessed by the MTT assay [ 90 ]. Ishikawa and 12Z cells were seeded in 96-well plates (4 × 10 4 cells/cm 2 ) and allowed to attach for 24 h. Cells were then exposed to H 2 O 2 (50–3000 μM) for 24 h. After treatment, MTT solution (5 mg/mL; 30 μL/well) was added, and plates were incubated for 2 h at 37 °C and 5% CO 2 , protected from light. Cell viability was calculated relative to untreated controls, and EC 50 values were obtained by non-linear regression in GraphPad Prism 10.0. All experiments were performed in triplicate as independent assays. NAC cytotoxicity was evaluated using the same protocol by treating cells with NAC (1, 5 and 10 mM) for 24 h, with results expressed as percentage viability relative to untreated controls. Based on these data, the protective effect of NAC against H 2 O 2 was determined by pre-treating cells with NAC (5 mM) for 10 min, followed by incubation with H 2 O 2 at the previously defined concentrations for 24 h. Untreated cells were used as negative controls and H 2 O 2 -treated cells as positive controls. Viability was measured by MTT as previously described. The concentration of H 2 O 2 and NAC selected in 2D was subsequently used to guide treatments in spheroids. Viability in 3D cultures was measured using the LIVE/DEADTM Cell Imaging Kit (Calcein-AM/BOBO-3 iodide; Invitrogen TM , Carlsbad, CA, USA) diluted 1:1 in HBSS (pH 7.5) and added at 50 μL/well. Fluorescence images were acquired on a Leica AF6000 microscope (Wetzlar, Germany) (λ Ex/Em 488/515 nm for Calcein-AM; 570/602 nm for BOBO-3 iodide). Images were analyzed in ImageJ V1.54 by calculating corrected total cell fluorescence (CTCF = integrated density − [area × mean background]). For each fluorophore, CTCF values from all fields were summed to obtain total fluorescence (CTCF_total), and viability was expressed as [CTCF_individual/CTCF_total] × 100. EC 50 values were determined by non-linear regression in GraphPad Prism 10.0. All experiments were performed in triplicate as independent assays. ROS production was evaluated by DCFH-DA staining followed by fluorescence microscopy [ 91 , 92 , 93 ]. DMNQ (100 μM) was used as a positive control for ROS generation, and NAC as an antioxidant control. H 2 O 2 (EC 50 ) and NAC (MEC) concentrations were defined beforehand using the LIVE/DEAD TM assay. Untreated spheroids with and without staining served as negative controls. Spheroids were generated as previously described and then exposed to treatments for 3 h. Sequentially, spheroids were washed with PBS (2×) and incubated with DCFH-DA (50 μM; 200 μL/well) for 45 min at 37 °C protected from light. After two additional PBS washes, fluorescence was recorded on a Leica AF6000 microscope (λ Ex/Em 498/522 nm). Images were quantified in ImageJ, and results are reported as mean gray value. All experiments were performed in triplicate as independent assays. 2D cells were seeded (4 × 10 4 cells/cm 2 , 25 cm 2 flasks, 24 h) and spheroids were generated as described above (48 h), followed by 24 h treatments with H 2 O 2 , NAC or H 2 O 2 + NAC; untreated samples were used as controls. 2D cells were collected by trypsinization, whereas spheroids were lysed by sonication (40% amplitude; cycle 1; 10 s). Lysates were centrifuged (1300 rpm, 5 min, 4 °C) and washed with ice-cold PBS (1500× g , 3 min, 4 °C, 3×). Pellets were resuspended in 240 μL of cold ultrapure water and subjected to a freeze–thaw step in liquid nitrogen (−196 °C, 1 min). After thawing, 40 μL were stored at −80 °C for protein quantification (Lowry method). The remaining 200 μL were deproteinized with 50 μL of 10% w/v sulfosalicylic acid and centrifuged (4000× g , 5 min, 4 °C); GSH and GSSG were then quantified in the supernatant. GSH was measured by monitoring the increase in absorbance at 412 nm (DTBN/GR system), and GSSG by monitoring the decrease in NADPH at 340 nm. Concentrations were calculated by the Lambert-Beer law, and redox status was expressed as the GSH/GSSG ratio. Total glutathione was reported as ×10 −7 mol/mg of protein, with protein levels determined by the Lowry method using BSA as a standard [ 42 , 43 , 45 , 93 , 94 , 95 , 96 , 97 ]. Experiments were performed in triplicate as independent assays. DNA damage after H 2 O 2 exposure was assessed by alkaline comet assay with SYBR Gold staining [ 98 , 99 ]. Cells were seeded in 24-well plates (1 × 10 5 cells/well) for 24 h and then treated for 24 h with H 2 O 2 , NAC or H 2 O 2 + NAC; untreated cells were used as controls. Cells were collected, washed with PBS, embedded in 1% w/v low-melting agarose, and layered onto slides pre-coated with 1.5% w/v agarose. After solidification at 4 °C, slides were lysed for 1 h at 4 °C (2.5 M NaCl, 100 mM EDTA, 10 mM Tris, 10% v/v DMSO, 1% v/v Triton X-100). Slides were then incubated in alkaline electrophoresis buffer (300 mM NaOH, 100 mM EDTA, pH > 13) for 20 min and electrophoresed at 25 V and 300 mA for 15 min. After electrophoresis, slides were neutralized with Tris buffer (pH 7.5), washed twice with dd-H 2 O, dried at 37 °C for 2 h, and fixed in cold ethanol for 5 min. DMSO at 20% ( v / v ) was used as an internal reference for robust DNA damage induction, reflecting cytotoxicity-associated genomic instability rather than a classical genotoxic mechanism. For staining, slides were incubated with 1× SYBR Gold in TE Buffer for 30 min protected from light, dried at 37 °C for 2 h, and imaged by fluorescence microscopy (λ Ex/Em 496/539 nm). Comet parameters were quantified using the OpenComet plugin in ImageJ. Experiments were performed in triplicate as independent assays. To assess lipid peroxidation, TBARS were measured as described in the literature [ 100 ]. Cells were seeded in 25 cm 2 flasks (4 × 10 4 cells/cm 2 ) for 24 h, treated as described above, and incubated for an additional 24 h at 37 °C and 5% CO 2 . Cells were washed twice with saline, resuspended in 1 mL of 50 mM sodium phosphate buffer, and stored at −80 °C until analysis. Samples were scraped and transferred to 15 mL tubes; an aliquot (50 μL) was reserved for protein quantification (Lowry method), and the remaining suspension was mixed with 1 mL TBA-TCA reagent. After heating (45 min, boiling water bath) and cooling to room temperature, samples were centrifuged (300× g , 5 min) and absorbance was read at 535 nm. MDA equivalents were calculated using ε = 1.49 × 10 5 L/mol.cm and expressed as nmol MDA/mg of protein. Experiments were performed in triplicate as independent assays. Protein carbonylation was quantified by DNPH derivatization as previously described [ 100 ]. Cells were seeded in 25 cm 2 flasks (4 × 10 4 cells/cm 2 ) for 24 h, treated as previously described, then collected by trypsinization, washed three times with PBS, and lysed in RIPA buffer (100 μL). Lysates were kept on ice for 30 min and centrifuged (14,000 rpm, 20 min, 4 °C). An aliquot (10 μL) was reserved for protein quantification (Lowry method). Volumes corresponding to 1 mg of protein were reacted with DNPH (10 mM in 2.0 M HCl; 500 μL) for 1 h at 37 °C. Proteins were precipitated with 20% v/v TCA (500 μL), centrifuged (11,000 rpm, 5 min, 4 °C), and the pellet was washed three times with EtOH/EtOAc (1:1, 3 × 100 μL). Pellets were then solubilized in 6 M guanidine in 20 mM phosphate buffer (15 min, 37 °C. Carbonyl content was determined at 370 nm (ε = 22 × 10 3 L/mol.cm) and expressed as nmol of carbonyl/mg of protein. Experiments were performed in triplicate as independent assays. All experiments were repeated at least three times as independent replicates, and results from cellular assays are presented as mean ± standard deviation. Statistical analyses were performed using ANOVA followed by Bonferroni’s multiple comparisons test. For each comparison, mean differences were reported together with their corresponding 95% confidence intervals and multiplicity-adjusted p -value.

Endometriosis is a chronic inflammatory condition that affects about 10% of women of reproductive age and is characterized by the presence of endometrial-like tissues outside the uterus, usually found in the peritoneum, ovaries and rectovaginal septum [ 1 , 2 , 3 , 4 ]. Besides the chronic pelvic pain, about 40% of diagnosed women present infertility, and current treatments focus on symptom management with surgical intervention, through the combination of hormones and laparoscopy, resulting in several side effects and a high rate of recurrence of endometriotic lesions [ 5 , 6 , 7 , 8 ]. Since 1920, the main theory regarding the origin of endometriosis has been John A. Sampson’s hypothesis of retrograde menstruation. Over the decades, new hypotheses have emerged, including the lymphovascular dissemination of stem cells, the transdifferentiation of peritoneal coelomic epithelial cells, and embryonic remnants in response to estrogen [ 9 , 10 , 11 , 12 , 13 , 14 , 15 , 16 ]. However, regardless of the proposed theory, all result in activated macrophages and apoptosis of endometrial cells and erythrocytes, leading to the presence of free iron derived from hemoglobin, promoting the cyclic overproduction of H 2 O 2 and other ROS, and subsequent inflammatory signaling [ 17 ]. Clinical studies analyzing fluids and tissues from women diagnosed with endometriosis identified oxidative stress biomarkers, such as 8-hydroxy-2′-deoxyguanosine (8-OHdG) adducts resulting from DNA damage, lipid peroxidation products such as Malondialdehyde (MDA), protein products of advanced oxidation, pro-inflammatory cytokines, and others [ 18 , 19 , 20 , 21 ]. Thus, the overproduction of mediators resulting from oxidative stress can be attributed as one of the sources of endometrial dysfunction contributing to disease progression [ 22 ]. Given the correlation between disease occurrence and pro-inflammatory markers, it is interesting to study these pathways as targets for alternative therapies [ 23 ]. Traditional antioxidants, such as N-acetylcysteine (NAC), can serve as inspiration for such discoveries [ 24 , 25 , 26 , 27 ], as it participates in several relevant cellular mechanisms, such as Glutathione (GSH) replenishment, H 2 O 2 scavenging, activation of the Nuclear Factor E2-related factor 2 (Nrf2) antioxidant defense pathway, prevention of apoptosis by inhibition of Nuclear Factor Kappa-light-chain-enhancer of activated B cells (NF-κB), and increased cytotoxic capacity of immune cells [ 24 , 28 , 29 ]. Based on that, this work aimed to investigate oxidative stress responses relevant to endometriosis-associated inflammatory and hemolysis-related processes, using H 2 O 2 exposure as a controlled pro-oxidant stimulus, and to evaluate its effects on biomolecular oxidation, the protective effect of NAC, and the biochemical profiles of two widely used models of immortalized endometrial cell lines—Ishikawa cells, which were immortalized from human endometrial epithelial adenocarcinoma and retain several characteristics of endometrial epithelium, allowing their use in studying normal endometrial behavior and functions [ 30 , 31 ], and the 12Z cell line, immortalized in the 2000s from peritoneal endometrial biopsies, and widely used in constructing in vitro endometriosis models, as it expresses the same gene patterns found in women diagnosed with the disease [ 32 , 33 , 34 , 35 ]. The pathogenesis of endometriosis is also deeply driven by genetic predispositions that establish an altered cellular microenvironment [ 33 , 36 , 37 ]. At the regulatory level, microRNA dysfunction acts as a key genetic determinant by modulating gene expression related to cell survival, proliferation, and immune signaling; their systemic regulation in affected women suggests an epigenetic scenery that favors pro-inflammatory signaling and impaired apoptosis. This susceptibility is further evidenced by a constitutive imbalance in estrogen metabolism enzymes, suggesting a genetically programmed disturbance in metabolic clearing [ 2 ]. This enzymatic imbalance leads to the accumulation of reactive hydroxy-estrogens and altered ROS production, acting as a biochemical trigger for the proliferation of ectopic endometrial tissue rather than being a mere byproduct of established lesions. However, despite extensive evidence linking oxidative stress to endometriosis pathophysiology, comparative studies evaluating the biochemical oxidative responses of cellular models relevant to endometrial and endometriosis-associated biology under controlled oxidative challenge remain limited. Thus, the hypothesis of this work was that, compared to Ishikawa cells, 12Z cells would exhibit greater oxidative vulnerability, including increased sensitivity to H 2 O 2 , elevated production of Reactive Oxygen Species (ROS), depletion of the glutathione antioxidant system, and greater damage to the evaluated biomolecules.

The use of H 2 O 2 as an oxidative stress inducer for in vitro studies is well described in the literature [ 38 , 39 ]. Before evaluating the effects caused in biomolecules by H 2 O 2 , the cellular density of cells for spheroids construction to achieve recommended size (400 µm) ( Figure 1 ), as well as the toxicity of H 2 O 2 in bi- and tri-dimensional cell cultures for 12Z and Ishikawa cell lines, were determined by MTT assay and by LIVE/DEAD™ Cell Imaging Kit (Invitrogen), for each cell culture model, respectively. For a 3D tissue-like model that recapitulates key architectural features relevant to endometrial and endometriosis-related in vitro studies, the ideal condition for spheroids establishment is required for both endometrial cell lines [ 16 , 40 , 41 ] was a 48 h incubation of 0.5 × 10 4 cells per well in Nunclon™ Sphera™ ( Figure 1 ). In 2D cultures, with concentrations ranging from 50 to 3000 µM, it was possible to observe that Ishikawa cells exhibited higher tolerance to H 2 O 2 , showing an Effective Concentration 50% (EC 50 ) of 1300 ± 148.5 µM, while 12Z cells were more sensitive, presenting an EC 50 of 296 ± 35.5 µM ( Figure 2 A,C). These EC 50 results reveal a differential sensitivity greater than fourfold between the two cell lines, supporting their use for comparative oxidative stress analysis. The cell viability results from 2D cultures guided the selection of the concentrations subsequently assessed in the spheroids. The following H 2 O 2 concentrations were assessed for Ishikawa and 12Z spheroids, respectively: 112.5, 225, 450 and 1300 µM, and 37, 75, 150 and 300 µM. Similar results for H 2 O 2 EC 50 were observed across all concentrations (LIVE/DEAD™ of approximately 30/70%), so we used the lowest peroxide concentration (112.5 µM for Ishikawa and 37 µM for 12Z) ( Figure 2 B,D). In our hypothesis regarding H 2 O 2 -induced oxidative stress, we selected NAC as a model antioxidant. NAC showed no cytotoxic effects at any of the tested concentrations in either cell line, and when its ability to protect cells from H 2 O 2 toxicity was evaluated, treatment with 5 mM in 2D cultures and 1.25 mM in 3D cultures, followed by H 2 O 2 exposure, resulted in cell viability comparable to the negative control, suggesting a strong protective effect of NAC ( Figure 2 E,G). The following NAC concentrations were assessed for Ishikawa and 12Z spheroids: 1.25, 2.5, and 5 mM. Equivalent results for NAC’s Minimum Effective Concentration (MEC) were observed across all concentrations (LIVE/DEAD™ of approximately 60/40%), so the lowest effective concentration (1.25 mM) was selected ( Figure 2 F,H). Once the treatment concentrations were determined, experiments were conducted to verify the generation of ROS and the redox status of the cells. ROS production in 3D cultures was assessed by a sensitive cell-permeable redox probe, 2′,7′-Dichlorodihydrofluorescein diacetate (DCFH-DA), with the positive control 2,3-Dimethoxy-1,4-naphthoquinone (DMNQ), and the evaluation of the Redox Status in 2D and 3D cultures was performed through GSH/GSSG ratio and quantification of total glutathione (×10 −7 mol/mg protein), indirectly reflecting the activity of the glutathione antioxidant enzyme system. Results from the DCFH-DA assay are expressed as mean gray value, an average gray value within the Region of Interest (ROI) of the spheroids, representing the sum of the gray values of all the pixels in the selection divided by the number of pixels. In the negative control, 12Z cells produced approximately four times more ROS than Ishikawa cells ( Figure 3 D). When observing the different responses between the negative control and the induced oxidative stress, Ishikawa spheroids showed a three-fold increase in mean gray value ( Figure 3 B), and 12Z spheroids showed an increase of approximately two times ( Figure 3 C). Regarding the incubation of the H 2 O 2 stressor with the antioxidant, NAC showed an inhibitory effect, apparently reducing the presence of ROS in both cell lines, although this reduction occurred more considerably in Ishikawa cells since the statistical difference between the negative control and the treatment with both H 2 O 2 and NAC was more significant than for the same condition in 12Z ( Figure 3 B,C). The activity of the glutathione antioxidant system was assessed by measuring the classic redox index GSH/GSSG [ 42 , 43 ] and total glutathione levels, quantified and expressed as a normalized value (×10 −7 mol/mg protein) [ 44 , 45 ]. When comparing the different cell lines in 2D culture, even though the GSH/GSSG ratio was lower for 12Z cells regardless of the treatment, demonstrating a naturally more fragile redox status, 12Z cells showed the same overall trend as Ishikawa cells: lower activity of the antioxidant system in the negative control when compared to the H 2 O 2 + NAC treatment (negative control: 6.02 in Ishikawa vs. 4.06 in 12Z; H 2 O 2 + NAC: 5.48 in Ishikawa vs. 3.28 in 12Z), and higher activity when treated only with H 2 O 2 (4.75 in Ishikawa vs. 3.1 in 12Z) ( Figure 4 A). Total glutathione values reinforce the negative control and H 2 O 2 + NAC treatment results (negative control: 2.12 × 10 −7 mol/mg protein in Ishikawa vs. 2.69 × 10 −7 mol/mg protein in 12Z; H 2 O 2 + NAC: 5.46 × 10 −7 mol/mg protein in Ishikawa vs. 5.92 ×10 −7 mol/mg protein in 12Z), but showed a more notable difference under H 2 O 2 treatment. When exposed to oxidative stress, 12Z cells exhibited a drastic increase in total glutathione compared to Ishikawa cells (1.77 × 10 −7 mol/mg protein in Ishikawa vs. 12.05 × 10 −7 mol/mg protein in 12Z) ( Figure 4 B). The GSH/GSSG ratio of the 3D culture showed similarity when comparing the different treatments, suggesting that the redox balance was maintained at negative control levels ( Figure 4 C,E,G). The total glutathione measurement in the negative control was slightly lower in 12Z spheroids; however, when treated with H 2 O 2 , Ishikawa maintained the amount of total glutathione at negative control levels (similar to 2D culture), as also observed when exposed to H 2 O 2 + NAC ( Figure 4 F). For 12Z, the opposite occurs to what happens in its 2D culture: a sharp decrease in total glutathione levels with both H 2 O 2 and H 2 O 2 + NAC ( Figure 4 H). Levels of lipid peroxidation in 2D cultures were assessed through the Thiobarbituric Acid Reactive Substances (TBARS) assay. Our results indicate no significant difference in the negative control when comparing cell lines. However, when cells were exposed to H 2 O 2 , the concentration of MDA produced in 12Z cells increased threefold (974.9 nmol of MDA/mg of protein), indicating that it is more sensitive to induced oxidative stress when compared to Ishikawa cells (619.5 nmol of MDA/mg of protein). Our results also corroborate the protective effect of NAC against H 2 O 2 -induced damage, since cells incubated previously with NAC presented MDA concentrations similar to those of untreated cells for both cell lines ( Figure 5 A–C). The levels of carbonylated proteins in 2D cultures were determined by the reaction of carbonyl groups with 2,4-dinitrophenylhydrazine (DNPH). Our results indicate that in basal conditions, 12Z cells present a higher concentration of carbonylated proteins in comparison to Ishikawa cells. When exposed to oxidative stress induced by H 2 O 2 , Ishikawa cells did not present changes in levels of carbonylated proteins, while 12Z cells showed a four-fold increase in carbonyl content. It was also possible to observe that NAC inhibits the production of carbonylated proteins in cells previously protected ( Figure 5 E,F). The Comet Assay was used to assess DNA damage in 2D cultures. Results indicated that Ishikawa cells exhibited higher resistance to oxidative challenge, as observed for the positive control (DMSO 20% v / v ) and H 2 O 2 treatment, in comparison to 12Z cells, presenting a Tail Length (pixels) of 750 and 600, for each treatment, respectively, while 12Z cells showed values of 2250 and 1150 under the same conditions. Despite NAC being able to preserve cell viability under H 2 O 2 exposure, it did not reduce DNA damage ( Figure 5 G–I). All the results presented here are summarized in Figure 6 , and the numerical values are available in Supplementary Table S1, Figure S1 .

Since 1920, several hypotheses about the origins of endometriosis have been discussed [ 46 , 47 , 48 , 49 ], and more recent studies have identified inflammation as a hallmark of the disease, particularly mechanisms linked to oxidative stress [ 50 , 51 , 52 , 53 ]. Despite remaining incompletely characterized, the imbalance between oxidants and antioxidants is strongly implicated in endometriosis-associated pain [ 54 , 55 , 56 ]. Based on this framework, the present study evaluated and compared the effects of H 2 O 2 -induced oxidative stress on the biochemical profile of two widely used in gynecological epithelial cell lines in vitro: Ishikawa, an endometrial adenocarcinoma cell line that retains several characteristics of endometrial epithelium and was therefore used here as a comparator; and 12Z cell line, an N -cadherin-positive human endometrial cell line widely used in endometriosis-related in vitro studies, including investigations of invasion, cellular behavior, and therapeutic responses [ 40 , 57 , 58 ]. Consistent with our hypothesis, 12Z cells proved to be more sensitive to H 2 O 2 . The EC 50 values obtained for Ishikawa cells are comparable to previously reported data [ 37 ]. NAC, a widely employed antioxidant, acts through several mechanisms, including GSH replenishment, H 2 O 2 scavenging, and Nrf2 pathway modulation [ 28 , 29 ]. Consistent with its known properties, NAC served as an effective positive control in our models, restoring cell viability after H 2 O 2 exposure to levels comparable to the negative control, and increasing viability in both cell lines. The LIVE/DEAD™ assay for 3D cultures corroborated MTT results. However, the limited discrimination between H 2 O 2 concentrations suggests reduced assay sensitivity, due to limited dye penetration into the spheroid core and the heterogeneous diffusion of reagents [ 59 , 60 ]. Thus, it did not demonstrate a dose-dependent response to H 2 O 2 treatment, but the assay remained adequate for defining experimental conditions. Baseline fluorescence observed in negative controls is consistent with the intrinsic oxidation dynamics of the DCFH-DA probe, which is further enhanced under oxidative conditions [ 61 ]. The variation in 12Z when exposed to H 2 O 2 is low compared to its high basal level of oxidative stress [ 2 , 18 , 62 ], may reflect adaptive mechanisms that limit the additional impact of exogenous oxidative challenge. In contrast, Ishikawa cells displayed lower ROS levels across all conditions, highlighting intrinsic differences in redox status between the two cell lines. When treated simultaneously with H 2 O 2 and NAC, the antioxidant showed an inhibitory effect on H 2 O 2 -induced oxidative stress. These results are consistent with reports of enhanced ROS generation in biological samples from patients with endometriosis [ 14 , 15 , 17 , 63 ]. Regarding 2D cultures, 12Z cells exhibited a lower GSH/GSSG ratio than the Ishikawa cells, though both followed the same tendencies, with NAC restoring redox balance and protecting the GSH pool [ 28 , 29 ]. Notably, H 2 O 2 exposure triggered a pronounced increase in total glutathione exclusively in 12Z cells, suggesting an exacerbated compensatory response [ 64 ]. This likely reflects the immortalized nature of the model [ 62 , 65 , 66 ], potentially corresponding to an early adaptive response rather than a sustained physiological state [ 62 ]. The 3D microenvironment provides a more constrained and physiologically relevant context for redox regulation, as cells prioritize the GSH/GSSG ratio for survival over the total glutathione pool. Although negative control spheroids maintained higher total glutathione than 2D cultures due to increased antioxidant demand [ 67 ], their response to H 2 O 2 differed significantly: while Ishikawa spheroids preserved their levels, 12Z spheroids showed a prominent decrease. This indicates a collapse of the antioxidant system in 12Z spheroids under oxidative stress conditions; while GR activity sustains the GSH/GSSG ratio via recycling, it incurs a high metabolic cost, evidenced by a 72% depletion of total glutathione [ 64 , 68 , 69 ]. This underscores the robust but energetically costly antioxidant response of 12Z cells, achieved through GSH pool exhaustion or compensatory mechanisms like CAT and SOD [ 2 , 21 ]. Lipid peroxidation is expected to be higher in cellular systems associated with endometriosis-related oxidative stress [ 70 , 71 , 72 , 73 , 74 , 75 ], a process largely mitigated by GPx4 through the reduction in lipid hydroperoxides to nontoxic alcohols using the glutathione system [ 62 , 64 , 76 , 77 , 78 , 79 , 80 , 81 ]. TBARS assays revealed comparable and lower lipid damage in the negative control and H 2 O 2 + NAC groups for both cell lines, consistent with a high GSH/GSSG ratio, lower GPx4 demand, and lower total GSH. However, isolated H 2 O 2 exposure tripled MDA levels in 12Z cells and doubled them in Ishikawa cells compared to controls, confirming the greater sensitivity of 12Z to induced oxidative stress. This aligns with observed GSH oxidation and the drastic increase in total glutathione in 12Z cells, indicating compensatory redox hyperactivation. Consistent with lipid peroxidation patterns, H 2 O 2 exposure induced a moderate but significant increase in PCO levels in Ishikawa cells, only partially reversed by NAC—potentially due to limited uptake or alternative oxidative pathways [ 29 , 37 ]. In contrast, 12Z cells exhibited a significantly more pronounced PCO increase upon H 2 O 2 treatment. Notably, NAC co-treatment effectively restored PCO levels to near-control levels. This suggests that 12Z cells possess a higher susceptibility to oxidative damage but also a greater reliance on redox-sensitive mechanisms that are highly responsive to NAC supplementation [ 50 , 82 ]. DNA damage analysis revealed minimal DNA damage in untreated controls of both cell lines. However, upon exposure to H 2 O 2 , 12Z cells exhibited greater DNA fragmentation compared to Ishikawa cells, indicating heightened susceptibility to oxidative stress-induced damage. Despite NAC’s efficacy in other biochemical assays, it failed to attenuate H 2 O 2 -induced DNA damage. This outcome may reflect the contribution of iron-dependent oxidative processes described in endometriosis, which are not directly modeled in the present system, as well as NAC’s limited capacity for iron chelation and •OH neutralization [ 83 ]. Recent studies highlight that iron-driven oxidative reactions contribute to genomic instability and ferroptosis processes in endometriosis [ 84 , 85 ]. These findings highlight the limitations of conventional antioxidant strategies in iron-enriched oxidative environments [ 86 , 87 , 88 ]. A major limitation of this study lies in the use of immortalized cell lines, which lack the multicellular and inflammatory complexity of endometriosis lesions in vivo. Additionally, the adenocarcinoma origin of Ishikawa cells may contribute to increased oxidative stress resistance due to cancer-associated redox adaptations, which should be considered when interpreting the differences observed between cell lines. Therefore, future studies employing primary cultures, organoids, or ex vivo tissues will be essential for translational validation. This study was designed as a controlled in vitro investigation of oxidative stress mechanisms and did not aim to directly model the full pathophysiology of endometriosis or to provide clinical or therapeutic conclusions. Despite these limitations, from a mechanistic perspective, these findings reinforce oxidative stress as a vital component of endometriosis pathophysiology. 12Z cells displayed intrinsic redox fragility, exaggerated biomolecular damage, and incomplete antioxidant rescue under oxidative challenge. Consistent with this, the 3D spheroid model revealed a pronounced depletion of total glutathione in 12Z cells under oxidative stress, indicative of redox exhaustion and supporting its application in controlled mechanistic studies.

Our findings confirm that 12Z cells exhibit reduced antioxidant capacity and increased susceptibility to oxidative damage compared to Ishikawa cells. This is evidenced by lower EC 50 values, elevated basal ROS, and significant damage to DNA, lipids, and proteins, patterns that are consistent with oxidative features associated with endometriosis. 12Z cells maintained fragile redox homeostasis, preserving GSH/GSSG ratio under stress despite total glutathione depletion in 3D cultures, a pattern consistent with redox exhaustion under sustained oxidative challenge. Furthermore, while NAC effectively mitigated ROS production and restored cell viability, its inability to prevent DNA damage suggests that oxidative genomic injury in this cellular system involves pathways not directly modulated by NAC’s antioxidant effects. These data indicate that antioxidant approaches relevant to endometriosis-associated oxidative stress may need to go beyond generic ROS scavenging. The inability of NAC to prevent genomic injury, despite maintaining cell viability, underscores the importance of investigating additional mechanisms contributing to oxidative DNA damage in more complex experimental systems. The antioxidant exhaustion observed in 3D models suggests that metabolic interventions aimed at preserving glutathione homeostasis and genomic stability deserve further investigation. Ultimately, these results highlight oxidative stress as a key component of the mechanisms associated with endometriosis. Our data support the use of 3D models to better capture cellular metabolic stress under constrained conditions, evidencing glutathione dysregulation as a relevant feature of this microenvironment. These findings, together with iron-mediated ROS pathways well-described in endometriosis but not directly modeled here, represent promising directions for future investigation in endometriosis-related contexts.