*2.7. SpD Protected against Doxorubicin-Induced Mitochondrial Damage in AC16 Cells*

Using tetramethylrhodamine (TMRE) and rhodamine-2 (rhod-2) staining, we compared mitochondrial membrane potential and mitochondrial Ca2+ in doxorubicin and SpD-treated AC16 cells (Figure 9). Doxorubicin treatment for 24 h decreased the TMRE intensity of AC16 cells in a dose-dependent manner. The doxorubicin-treated cells also showed increased cytosolic diffusion of rhod-2. We treated AC16 cells with 10 μM of SpD in the presence of 0.1–1.0 μM doxorubicin. SpD treatment attenuated the loss of mitochondrial membrane potential that is induced by doxorubicin (Figure 9A,B). SpD treatment reduced the cytosolic overload of mitochondrial Ca2+ in doxorubicin-treated cells (Figure 9C,D).

**Figure 9.** SpD attenuated doxorubicin-induced mitochondrial membrane potential and mitochondrial calcium changes in AC16 cells. (**A**) Mitochondrial membrane potential (Δψm) in AC16 cells was indicated by TMRE fluorescence. The cells were treated with doxorubicin (0.1–1.0 μM) with/without SpD (10 μM); (**B**) The intensity of tetramethylrhodamine (TMRE) staining was measured using fluorometry at 550Ex/590Em nm. Doxorubicin decreased mitochondrial membrane potential in a dose-dependent manner. Co-treatment with SpD attenuated the membrane potential loss; (**C**) Mitochondrial calcium was localized using rhod-2, a selective indicator for mitochondrial Ca2+; and, (**D**) The intensity of rhod-2 was measured at 552Ex/581Em nm. Doxorubicin induced diffusion of mitochondrial Ca2+ to the cytosolic space but SpD co-treatment attenuated the Ca2+ diffusion. \* *p* < 0.05 compared with the doxorubicin-treated group.

#### *2.8. SpD Did Not Interfere with the Anticancer Effects of Doxorubicin in MCF-7 Cells*

SpD did not inhibit the cytotoxic activity of doxorubicin in MCF-7 cells (Figure 10). Human breast cancer MCF-7 cells and human cervical cancer HeLa cells showed reduced cell viability after treatment with SpD above 100 μM (Figure 10A and Figure S7A). When the SpD/doxorubicin co-treated cells were compared with the cells that were treated with doxorubicin alone, there were no significant differences in ROS level (Figure 10B), loss of mitochondrial membrane potential (Figure 10C,D), or Ca2+ overload (Figure 10E,F). Similar experiments using the HeLa human cervical cancer cell line showed the same lack of inhibition of cytotoxicity (Figure S7). Interestingly, SpD treatment inhibited cell migration in a wound healing test (Figure S8).

**Figure 10.** SpD did not inhibit the cytotoxicity of doxorubicin. (**A**) SpD induced MCF-7 cell death at 100–200 μM. \* *p* < 0.05 compared with the untreated control; (**B**) The co-treatment of SpD and doxorubicin showed similar levels of ROS compared with the doxorubicin alone group; (**C**) Mitochondrial membrane potential was visualized using TMRE staining in SpD/doxorubicin-treated MCF-7 cells; (**D**) The TMRE intensity measured by fluorometry is shown; (**E**) Mitochondrial Ca2+ level is indicated using rhod-2 dye in SpD/doxorubicin-treated MCF-7 cells; (**F**) The rhod-2 intensity was measured by fluorometry. \* *p* < 0.05 compared with doxorubicin treated groups without SpD. ROS = reactive oxygen species.

#### **3. Discussion**

There have been many efforts to overcome the cardiotoxicity of doxorubicin in cancer treatment. The IC50 (drug concentration required to inhibit cell growth by 50%) of doxorubicin for breast cancer cell lines has been reported to be between 1–4 μM after 24 h treatment (IC50 = 1 μM for MCF-10F; 4 μM for MCF-7; and, 1 μM for MDA-MB-231 cells) [19]. We used 10 μM of SpD, which induced no significant viability changes in either cardiomyocytes or cancer cells. From animal models using Histochrome® (echinochrome A), 1–10 mg/kg of doses have been reported to act as antioxidant in cardiomyocytes, which approximately correspond to 3–30 μM [20]. In our study, echinochrome A and SpD showed cardioprotective activity when treated with 0.1 μM doxorubicin. In equimolar treatment,

SpD showed better antioxidant activity and ATP production than echinochrome A. It might be reasonable to assume that cell viability decreases with increased doxorubicin incubation time. However, lower concentrations (<0.25 μM) of doxorubicin often show a low dose–time response relationship in cancer cells [21]. Cardiomyocytes and cancer cells have different mechanisms of doxorubicin-induced apoptosis. In cardiomyocytes, doxorubicin induces apoptosis by a H2O2-mediated mechanism, which is largely independent of p53 activation. In contrast, the p53 tumor suppressor plays an important role in doxorubicin-induced apoptosis in cancer cells [22–24].

Univariate analyses, such as fold change comparison, *t*-test, and volcano plot suggest overall shapes of measured data and multivariate analyses, including PCA, PLS-DA, and OPLS-DA, often reveal the latent structure of the data. When quantitatively analyzing multi-parametric metabolite responses, it is critical to specify all of the independent and dependent variables to be included [25]. In biological systems, metabolites are the end product of enzymatic and other protein activity, and therefore they are not independent from biological interactions. In our study, correlation network analysis and pathway enrichment analysis of VIP > 1.0 metabolites (e.g., GPC, acetate) showed that glycerolipid metabolism, glutathione metabolism, and pyruvate metabolism were significantly affected by SpD treatment in cardiomyocyte [26]. In addition to production during ethanol metabolism, acetate is transported into cells by proteins of the monocarboxylate transporter family or it is generated intracellularly by protein deacetylases and acetyl-CoA hydrolases [27,28]. In the cytosol, CoA is acetylated by acetyl-CoA synthetase to produce acetyl-CoA. In contrast, in mitochondria, acetyl-CoA is produced through the pyruvate dehydrogenase complex reactions. Acetyl-CoA participates in the citric acid cycle and β-oxidation of fatty acids to produce cellular energy (e.g., ATP). In addition to increased acetate consumption, SpD might increase cytosolic glycolysis and entrance of glutamate into the citric acid cycle, which could be shown by an increased lactate concentration and increased ratio of glutamate to glutamine (from 20.9 to 35.3; calculated as shown in Table 1). SpD treatment increased the accumulation of cytosolic osmolytes such as GPC, myo-inositol, and free amino acids (e.g., taurine and glycine), which are critical for the viability of cells. The integrated analysis with mass spectrometry based proteomics indicated that the glutathione metabolism of AC16 cells was most affected by SpD treatment. Since we measured the reduced form of glutathione (GSH) using Chenomx NMR Suit 7.1, the increased concentration of glutathione represents the increase of GSH reduced glutathione/GSSG oxidized glutathione ratio. Using the luciferase mediated method, we confirmed the increase of GSH/GSSG ratio in a dose-dependent manner (Figure S9).

Based on the acquired results, we hypothesized that SpD mainly exerted its function as antioxidant in the process of protection of cardiomyocytes against the cytotoxicity of doxorubicin. Since the cytotoxicity of doxorubicin on cardiomyocytes is known to be based on ROS increase, we tested SpD activity in ROS generating environments. We tested 1 mM H2O2 concentration, which might be increased by constitutively active NADPH oxidase 4 (NOX4) [29]. In addition, we assessed the antioxidant ability of SpD in cobalt chloride and hyperglycemic condition. The hypoxia mimetic cobalt chloride and hyperglycemic concentrations of glucose (33.3 mM) are known to increase intracellular ROS [30–34]. Hyperglycemia induces hypoxia-induced cell death via the influx of calcium in diabetic cardiomyopathy [35–37]. Oxidative stress might cause cardiac mitochondrial dysfunction, leading to cell death [38,39].

Since the integrated analysis that is located the mitochondrial proteins clustered together at SpD treatment, we had to focus the mitochondrial ATP production by SpD treatment. To differentiate ATP production from the cytosol versus mitochondria, we added D-galactose (10 mM) to the culture media. By competing with glucose in the cytosol, D-galactose reduces cytosolic glycolysis, resulting in decreased cytosolic ATP production. Galactokinase produces galactose 1-phosphate from galactose, utilizing ATP. Uridine diphosphate (UDP)-galactose 4-epimerase converts UDP-glucose and galactose 1-phosphate into UDP-galactose and glucose 1-phosphate, respectively. Galactose participates in glycolysis by consuming ATP and reducing cytosolic glycolysis rates [40–42]. SpD treatment increased

ATP production, even in 10 mM galactose media, which suggested the enhancement of mitochondrial ATP production with increased OCR.

In cardiomyocytes, doxorubicin has been known to induce oxidized state in mitochondrial redox potential to trigger mitochondrial depolarization and elevated calcium levels, which suppresses ATP production via oxidative phosphorylation.

As the mitochondrial dysfunction occurs, the cells are subjected to ATP depletion and become more dependent on ADP metabolism to compensate the ATP/ADP ratio [43]. In our study, the SpD treatment did not inhibit the anticancer activity of doxorubicin while protecting cardiomyocytes at identical concentration via increasing ATP production. Our approaches might provide some clues for the potential cardioprotective mechanisms of SpD in a combination therapy with doxorubicin. Nevertheless, further studies are still needed for evaluation of the drug.
