Early effects of the antineoplastic agent salinomycin on mitochondrial function

Abstract

Salinomycin, isolated from Streptomyces albus, displays antimicrobial activity. Recently, a large-scale screening approach identified salinomycin and nigericin as selective apoptosis inducers of cancer stem cells. Growing evidence suggests that salinomycin is able to kill different types of non-stem tumor cells that usually display resistance to common therapeutic approaches, but the mechanism of action of this molecule is still poorly understood. Since salinomycin has been suggested to act as a K(+) ionophore, we explored its impact on mitochondrial bioenergetic performance at an early time point following drug application. In contrast to the K(+) ionophore valinomycin, salinomycin induced a rapid hyperpolarization. In addition, mitochondrial matrix acidification and a significant decrease of respiration were observed in intact mouse embryonic fibroblasts (MEFs) and in cancer stem cell-like HMLE cells within tens of minutes, while increased production of reactive oxygen species was not detected. By comparing the chemical structures and cellular effects of this drug with those of valinomycin (K(+) ionophore) and nigericin (K(+)/H(+) exchanger), we conclude that salinomycin mediates K(+)/H(+) exchange across the inner mitochondrial membrane. Compatible with its direct modulation of mitochondrial function, salinomycin was able to induce cell death also in Bax/Bak-less double-knockout MEF cells. Since at the concentration range used in most studies (around 10 μM) salinomycin exerts its effect at the level of mitochondria and alters bioenergetic performance, the specificity of its action on pathologic B cells isolated from patients with chronic lymphocytic leukemia (CLL) versus B cells from healthy subjects was investigated. Mesenchymal stromal cells (MSCs), proposed to mimic the tumor environment, attenuated the apoptotic effect of salinomycin on B-CLL cells. Apoptosis occurred to a significant extent in healthy B cells as well as in MSCs and human primary fibroblasts. The results indicate that salinomycin, when used above μM concentrations, exerts direct, mitochondrial effects, thus compromising cell survival.

Figure 1

Effect of salinomycin, nigericin and valinomycin on lymphocytes. Jurkat leukemic T cells (a and b) and human primary B-CLL cells (c) were incubated for 24 h with different concentrations of the compounds, as indicated. Cell survival (a) was measured by MTT assay; values are expressed as the average percentage of cell survival compared with untreated cells ± S.E.M. (n=5). Cell death (b and c) was tested using FACS by the staining with FITC-Annexin V and propidium iodide. Staurosporine (Stauro or ST) was used as positive control. Indicated values refer to the percentage of dead cells ± S.E.M. (n=14). Statistically significant differences (P<0.05) are indicated by asterisks

Figure 2

Salinomycin kills Bax/Bak-less mouse embryonic fibroblasts and alters mitochondrial membrane potential. (a) Cell survival in WT and Bax/Bak DKO MEFs was measured by MTT assay (n=4). Staurosporine was used as the classical apoptosis inducer. No significant differences were observed between WT and DKO cells. (b) Mitochondrial membrane potential determined on rat liver isolated mitochondria (RLM) by measuring the distribution of the TPP⁺ ions across the mitochondrial membrane with a selective TPP⁺ ion-sensitive electrode. Values are presented as variation of Δψ following addition of the indicated compounds. (c and d) Mitosox (c) and TMRM (d) fluorescence on B-CLL cells, measured by FACS analysis after treatment with salinomycin, nigericin and valinomycin at the indicated concentrations; CCCP, antimycin and staurosporine (ST) were used as positive controls. Average values ± S.D. are showed (n=3 independent experiments)

Figure 3

Salinomycin induces an instantaneous acidification of matrix pH in intact cells. (a and b) Measurement of transient variation of mitochondrial matrix pH in MEF WT cells (a) and in HMLE-Twist cells (b) expressing mito-SypHer; changes in pH correspond to variations in the 535-nm fluorescence emission after alternative excitation at 405 and 488 nm. Results are expressed as mean 500/430 nm ratios ± S.E.M. of four different experiments. Addition of drugs is indicated by red arrow, while Na-acetate (NaAc) was used as positive control (blue arrows). (c) Isolated rat liver mitochondria (RLM) swelling was measured as reduction of mitochondrial absorbance over time at 540 nm. CaCl₂ (140 μM) was used to induce PTP opening and swelling. (d) MEF WT cells expressing mito-YFP were treated with SAL at the indicated concentrations and live images were acquired over time to monitor changes in mitochondria morphology. Bars correspond to 25 μm. Magnified images are shown in the lower row. Results shown in (a)–(d) are representative of three independent experiments

Figure 4

Salinomycin and nigericin affect respiration in a similar manner. (a–d) Oxygen consumption rate (OCR) of MEF WT cells (a, c and d) and of HMLE-Twist cells (b) was measured in the presence of 1 μM salinomycin, nigericin and valinomycin (compounds). Representative experiments are shown. The compounds were added to the cells either before (a and b) or after (c and d) inhibition of ATP synthase activity with oligomycin. The effect of 10 μM SAL in MEF cells was comparable to that of 1 μM, indicating that already the lower concentration exerts the maximal effect

Figure 5

Salinomycin, nigericin and valinomycin affect survival of leukemic B cells, mesenchymal stromal cells and fibroblasts. (a) Comparison of apoptosis in pathologic B-CLL cells and B cells from healthy subjects (cells were incubated for 24 h in the presence of the indicated compounds and cell death was determined by FACS by staining with FITC-Annexin V and propidium iodide). Quantification of the cell death (all annexin-positive cells) ± S.E.M. (n=14 for B-CLL cells and n=6 for healthy B cells). (b) Apoptosis of B-CLL (n=14) cells compared with death occurring in the pathologic cells co-cultured with MSC (n=4). Apoptotic cells were identified by FACS as in (a). (c) Mesenchymal stromal cells were photographed after 24 h treatment with SAL and removal of co-cultured B-CLL cells by washing. Bars correspond to 75 μm. At higher concentrations of SAL the number of cells decreased and the cells showed significant morphological alterations. (d) Human primary fibroblasts treated with the indicated substances for 24 h. Annexin-FITC binding is shown. Results shown in (c) and (d) are representative of three independent experiments. Bars correspond to 50 μm