First, we established optimal concentrations of XJB and TSM that could exert maximum anti-ferroptotic effects in H9c2 cells. The cells were exposed to a ferroptotic stimulus by treatment with RSL3 (0.5 µM) for 3 h in the absence and presence of XJB or TSM in a concentration range from 0.1 to 1 µM. Results showed that XJB reached a maximum protective effect at 0.1 µM (Figure 1A), whereas TSM demonstrated the highest protection starting at 0.6 µM (Figure 1B). Next, based on the LDH activity in the cell culture medium to assess cell death, we determined an optimal exposure time in which H9c2 cardiomyocytes developed ferroptosis once exposed to RSL3. The cells were incubated with 0.5 µM RSL3 for 1, 2, and 3 h in the absence and presence of Fer-1, XJB, or TSM (Figure 2). Incubation of the cells with RSL3 for 1 and 2 h had no significant effects on cell survival, as evidenced by no significant differences in LDH activity between control and RSL3-treated groups. However, RSL3 induced approximately a 2.3-fold increase (p < 0.001) in LDH activity after 3 h of incubation compared to the control. Fer-1, XJB, and TSM protected the cells against RSL3-induced cell death and prevented LDH release to the culture medium. The antioxidants alone (no RSL3-treatment) had no detrimental effects on the cells (Figure S1).

To identify and quantify cellular metabolites, cell lysates obtained from H9c2 cardioblasts after incubation with 0.5 µM RSL3 for 3 h in the absence and presence of Fer-1, XJB, and TSM were processed through GC-MS. A total of 52 metabolites were identified and classified in the following groups: amino acids (44%), TCA metabolites (11%), saturated FA (9%), unsaturated FA (6%), pyrroline carboxylic acids (6%), and 2% of pyrimidines, secondary, and sugar alcohols, short-chain acids, and cholesterol, among others (Figure 3A). The characterization of each identified metabolite, including chemical names, fragment ions, retention time, and classification, is detailed in Table S1, whereas information regarding the statistics that include the p.values and fold of change can be found in Table S2. The heatmap shows the results obtained from univariate analysis of metabolite concentration samples, considering the intensity of the color, as discrete variables (−2 to 2), by assorting them in bins (Figure 3B). The conditions (in the abscise axis) and the bins/metabolites (in the ordinate axis) have been arranged according to the Ward-clustering method. The fold change is color-coded according to the bar legend: a red/green value in the heatmap indicates an increase/decrease in average concentrations of the metabolites. The heatmap shows data clusters based on group similarities in which two superior clusters were identified. The first group contains only metabolites in the presence of RSL3, and the second group (control, TSM, RSL3 + XJB, Fer-1, RSL3 + TSM, RSL3 + Fer-1, and XJB) was clustered based on their data value proximity. Data shows the similarities of metabolites between the control and TSM; together with RSL3 + XJB, and Fer-1 with RSL3 + TSM. The latter altogether compares to a separate cluster containing RSL3 + Fer-1 and XJB. These results show how the RSL3-challenged group separates from the rest, indicating significant up- or downregulation of metabolites during ferroptosis. Interestingly, the control group shares a similar metabolite concentration with the TSM-treated group, while RSL3 + Fer-1 and XJB groups differentiate from RSL3 with the most distance, indicating opposing metabolite regulation compared to RSL3 alone. A one-way ANOVA followed by a post hoc test analysis revealed a total of 22 metabolites that had statistically significant differences in their concentrations between the treatment groups and the control. After performing Bonferroni correction, 12 metabolites were identified with a new p-value being p = 0.0022: stearic acid, glycolic acid, succinate, proline, palmitic acid, azelaic acid, beta-alanine, aspartic acid, glycine, alanine, (2R)-pyrrolidine 1,2-dicarboxilic acid [shown as (2R)-pyrrolidine in Figure 3B and Supplementary Materials], and lysine (Table S3). Particularly, in comparison with the control group, RSL3-treated metabolites’ concentration was increased in azelaic, glycolic, palmitic, and stearic acid, and decreased the levels of other metabolites. No changes were found in groups treated with RSL3 in combination with Fer-1 and TSM compared to the control. RSL3, in combination with TSM, decreased the concentrations of azelaic, glycolic, palmitic, and stearic acid slightly; and the same effect was observed in Fer-1 alone. XJB reduced the levels of all metabolites except azelaic, glycolic, palmitic, and stearic acid. In contrast, RSL3 supplemented with XJB reduced proline, (2R)-pyrrolidine 1,2-dicarboxilic acid, alanine, beta-alanine, glycine, and aspartate.

In this section, we examined the effect of glutaminolysis in mitochondria isolated from H9c2 cardiomyocytes exposed to ferroptosis, and the activity of LDH in the medium was measured to estimate cell death at 1 and 3 h. First, the expression of the mitochondrial glutaminase GAC in the cells was silenced with a GLS1 siRNA. Control (NC, non-coding siRNA-treated) and GLS1 knockdown cells were incubated with 0.5 µM RSL3 to induce ferroptosis, in the absence or presence of 1 µM Fer-1. Results showed a 51% (p < 0.05) reduction of GAC protein levels in the cells treated with GLS1 siRNA compared to NC cells (Figure S4). Analysis of LDH release in NC showed no changes in LDH activity compared to the control at 1 h. However, in GLS1 knockdown cells, LDH activity was reduced in the RSL3 group, compared to NC. (Figure 7A). In contrast, the activity of LDH from RSL3-challenged cells was significantly increased compared to the control in both NC and GLS1 knockdown cells after 3 h of incubation (Figure 7B). Likewise, GSH levels were higher in GLS1 knockdown cells in response to RSL3 at 1 h; however, the effects were lost at 3 h of incubation (Figure 7C,D). These data suggest that glutaminolysis is involved only at the early stage (1 h) of ferroptotic signaling.