Ru(II)-based complexes containing 2-thiouracil derivatives suppress liver cancer stem cells by targeting NF-B and Akt … – Nature.com

Ru(II) complexes containing 2-thiouracil derivatives display potent cytotoxicity in cancer cell lines

The cytotoxic effects of complexes 1 and 2 were evaluated on a panel of 23 cancer cell lines (HepG2, HCT116, NB4, THP-1, Jurkat, K-562, HL-60, KG-1a, MDA-MB-231, MCF-7, 4T1, HSC-3, CAL 27, SCC-25, SCC4, SCC-9, A549, PANC-1, OVCAR-3, DU 145, U-87 MG, A-375, and B16-F10) and three noncancerous cells (PBMC, MRC-5, and BJ) using the Alamar Blue assay after 72h of treatment (Fig. 1B and Table S1). Both complexes displayed potent cytotoxicity against all cancer cell lines, with half-maximal inhibitory concentration (IC50) values ranging from 2.4M in OVCAR-3 ovarian cancer cells to 17.5M in A549 lung cancer cells for complex 1 and from 1.6M in OVCAR-3 ovarian cancer cells to 10.5M in A549 lung cancer cells for complex 2. Doxorubicin was used as a positive control and showed cytotoxicity in all cell lines.

In noncancerous cells, complex 1 had an IC50 of 17.7M in MRC-5 pulmonary fibroblasts, 7.0M in BJ foreskin fibroblasts and 14.2M in PBMCs. In comparison, complex 2 presented IC50 values of 15.2M in MRC-5 pulmonary fibroblasts, 7.8M in BJ foreskin fibroblasts and 11.7M in PBMCs. The selectivity indices (SIs) were calculated by the following formula: SI=IC50 ([noncancerous cells]/IC50 [cancerous cells]). Figure 1C and Table S2 present the calculated SI. Curiously, both complexes showed an SI>2 for many of the cancer cells investigated.

To study the anti-HCC potential of these complexes, the HCC cell line HepG2 was used in further experiments. Therefore, the viability of HepG2 cells treated with complex 1 at concentrations of 2, 4, and 8M and complex 2 at concentrations of 1.5, 3, and 6M was determined by trypan blue assay after 12, 24, 48, and 72h of incubation. Both complexes reduced HepG2 cell viability in a concentration- and time-dependent manner (Fig. S1A1D). After 72h of incubation, complex 1 reduced cell viability by 46.5, 72.2, and 95.8%, while complex 2 inhibited cell viability by 39.4, 61.4, and 93.5%, respectively.

To determine whether complexes 1 and 2 can act against liver CSCs, we first performed a long-term colony formation assay to determine whether these complexes affect the clonogenic ability of HCC HepG2 cells. Clonogenic assays are well-known methods for evaluating the stemness of CSCs since a single CSC can form clonogenic colonies [16, 17]. Interestingly, treatment with both complexes significantly decreased the clonogenic viability of HepG2 cells in a concentration- and time-dependent manner (Fig. 2A, B).

A Representative images and (B) quantification of the number of colonies formed from HepG2 cells after treatment with complexes 1 and 2. (C and D) Quantification of CD133 expression on HepG2 cells after 24h of incubation with 8M complex 1 or 6M complex 2, as determined by flow cytometric analysis. (E and F) Quantification of CD44high in HepG2 cells after 24h of incubation with 8M complex 1 or 6M complex 2, as determined by flow cytometric analysis. The vehicle (0.2% DMSO) was used as a negative control (CTL). The data are expressed as the meanS.E.M. of three biological replicates carried out in duplicate. *P<0.05 compared to CTL by one-way analysis of variance (ANOVA) followed by Dunnetts multiple comparisons test.

Next, we quantified the expression of two biomarkers of liver CSCs, CD133 [18] and CD44 [19], in HepG2 cells treated with complexes 1 and 2. Likewise, both complexes reduced the percentage of HepG2 CD133-positive cells (Fig. 2C, D), while complex 2 reduced the percentage of HepG2 CD44high cells (Fig. 2E, F).

In a new set of experiments, we measured the effects of complexes 1 and 2 on three-dimensional (3D) tumorspheres formed from HepG2 cells since multicellular 3D tumor spheroids are well-known cell culture systems that can enrich cells with CSC characteristics [20, 21]. Both complexes reduced HepG2 tumorsphere growth (Fig. S2 and S3) and caused cell death (Fig. S4), corroborating that these molecules may inhibit CSCs in HCC HepG2 cells.

A series of cellular and molecular analyses were performed to examine the mechanism of cell death in HepG2 cells treated with complexes 1 and 2. HepG2 cells that were treated with complexes 1 and 2 for 24, 48, and 72h showed cell morphology changes that were associated with apoptosis, including a reduction in cell volume, chromatin condensation, and fragmentation of the nuclei, as observed in May-Grunwald-Giemsa-stained cells (Fig. S5).

Light scattering characteristics measured by flow cytometry were used to analyze cellular parameters such as size and complexity/granularity in HepG2 cells treated with complexes 1 and 2 (Fig. S6AS6F). Forward light scattering (FSC) was employed as a cell size metric in this experiment, while side scattering (SSC) was used to determine cell complexity/granularity. Treatment with these complexes caused cell shrinkage, as indicated by a decrease in the FSC, accompanied by an increase in the SSC, probably due to nuclear condensation. Both morphological changes are associated with cellular apoptosis, corroborating the findings observed in cells stained with May-Grunwald-Giemsa.

Internucleosomal DNA fragmentation and cell cycle distribution were evaluated in HepG2 cells after 24, 48, and 72h of incubation with complexes 1 and 2 via a DNA content-based flow cytometry assay (Fig. 3AG). All DNA of subdiploid size (sub-G0/G1) was considered fragmented. Both complexes induced DNA fragmentation in a time- and concentration-dependent manner. After 72h of incubation, complex 1, at concentrations of 2, 4, and 8M, caused DNA fragmentation by 12.3, 26.7, and 43.1%, respectively, while complex 2, at concentrations of 1.5, 3, and 6M, induced DNA fragmentation by 18.6, 39.1, and 72.9%, respectively (against the 5.6% detected in the control). The cell cycle phases G0/G1, S and G2/M decreased proportionally in HepG2 cells treated with complexes 1 and 2. Doxorubicin, used as a positive control, also caused DNA fragmentation.

Representative flow cytometric histograms of the cell cycle distribution of HepG2 cells after treatment with complexes 1 and 2 after 24 (A), 48 (B) and 72 (C) h of incubation. The percentages of cells in the sub-G0/G1 (D), G0/G1 (E), S (F) and G2/M (G) phases were quantified via flow cytometric analysis. The vehicle (0.2% DMSO) was used as a negative control (CTL), and doxorubicin (DOX, 1M) was used as a positive control. The data are expressed as the meanS.E.M. of three biological replicates carried out in duplicate. *P<0.05 compared to CTL by one-way analysis of variance (ANOVA) followed by Dunnetts multiple comparisons test.

Annexin V-FITC/propidium iodide (PI) double staining was also applied to HepG2 cells treated with complexes 1 and 2 for 24, 48 and 72h to quantify phosphatidylserine exposure and cell membrane integrity, which are markers of apoptosis and necrosis, respectively. Both complexes induced a significant increase in the percentage of apoptotic cells in a time- and concentration-dependent manner, and no significant increase in the percentage of necrotic cells was detected (Fig. 4AF). After 72h of incubation, complex 1, at concentrations of 2, 4, and 8M, increased apoptosis by 9.0, 48.9, and 76.1%, respectively, while complex 2, at concentrations of 1.5, 3, and 6M, increased apoptosis by 9.1, 40.7, and 76.9%, respectively (against 4.8% found in the control). Treatment with doxorubicin, which was used as a positive control, also led to apoptosis.

Representative flow cytometric dot plots of HepG2 cells stained with annexin V-FITC/PI after treatment with complexes 1 and 2 after 24 (A), 48 (B) and 72 (C) h of incubation. The percentages of viable (annexin V-FITC-/PI- cells) (D), apoptotic (early apoptotic [annexin V-FITC+/PI- cells] plus late apoptotic [annexin V-FITC+/PI+ cells]) (E) and necrotic (annexin V-FITC-/PI+ cells) (F) cells were quantified. The vehicle (0.2% DMSO) was used as a negative control (CTL), and doxorubicin (DOX, 1M) was used as a positive control. The data are expressed as the meanS.E.M. of three biological replicates carried out in duplicate. *P<0.05 compared to CTL by one-way analysis of variance (ANOVA) followed by Dunnetts multiple comparisons test.

As mitochondrial dysfunction and PARP cleavage are well-known events in apoptotic cell death, mitochondrial transmembrane potential and PARP (Asp214) cleavage were also determined by flow cytometry. Significant mitochondrial depolarization (Fig. 5A) and increased levels of PARP (Asp214) cleavage (Fig. 5B, C) were found in HepG2 cells treated with complexes 1 and 2, corroborating that these complexes can cause cell death via apoptosis. Moreover, the BAD KO SV40 MEF cell line, as well as its parental cell line, WT SV40 MEF, were used to assess the involvement of the proapoptotic protein BAD in the cell death caused by complexes 1 and 2 (Fig. 5D, E). On the other hand, these complexes cause cell death independent of the protein BAD.

A Quantification of mitochondrial membrane depolarization in HepG2 cells after 24h of incubation with complex 1 or 2, as determined by flow cytometry. B and C Quantification of PARP (Asp214) cleavage in HepG2 cells after 24h of incubation with complex 1 (8M) or 2 (6M), as determined by flow cytometric analysis. MFI: Mean fluorescence intensity. D Survival curves of WT SV40 MEFs and BAD KO SV40 MEFs upon treatment with complexes 1 and 2 and 5-fluorouracil (5-FU, used as a positive control). The curves were obtained from at least three biological replicates carried out in duplicate using the Alamar Blue assay after 72h of incubation. E DNA fragmentation (sub-G0/G1 cells) and cell cycle distribution (G0/G1, S and G2/M phases) of WT SV40 MEFs and BAD KO SV40 MEFs after 48h of incubation with complexes 1 (10M) and 2 (10M) or 5-FU (40M). The vehicle (0.2% DMSO) was used as a negative control (CTL). The data are expressed as the meanS.E.M. of three biological replicates carried out in duplicate. *P<0.05 compared to CTL by one-way analysis of variance (ANOVA) followed by Dunnetts multiple comparisons test.

To investigate the molecular mechanism of action of complexes 1 and 2, we analyzed the transcripts of 82 target genes using a qPCR array (Fig. 6A, B and Table S3). Among the altered gene transcripts, genes related to NF-B (the NFKB1 gene with RQ=0.45 for complex 1), PI3K/Akt/mTOR (the PIK3CA gene with RQ=0.44 for complex 2; the MTOR gene with RQ=0.39 for complex 1) and oxidative stress (the GSTP1 gene with RQ=0.49 for complex 1 and RQ=0.27 for complex 2; the TXN gene with RQ=0.37 for complex 2; and the TXNRD1 gene with RQ=0.35 for complex 2) were downregulated in HepG2 cells treated with complexes 1 and 2.

A, B Genes up- and downregulated in HepG2 cells after 12h of treatment with complexes 1 (8M) and 2 (6M). The vehicle (0.2% DMSO) was used as a negative control (CTL). The data are expressed as the relative quantification (RQ) compared to the CTL data. The genes were upregulated if RQ2 (red bars) and downregulated if RQ0.5 (green bars). Quantification of the levels of phospho-NF-B p65 (S529) (C, D), Akt1 (E, F), phospho-Akt (S473) (G, H), phospho-mTOR (S2448) (I, J), and phospho-S6 (S235/S236) (K, L) in HepG2 cells after 24h of incubation with complexes 1 (8M) and 2 (6M), as determined by flow cytometry. The vehicle (0.2% DMSO) was used as a negative control (CTL). The data are expressed as the meanS.E.M. of three biological replicates carried out in duplicate. *P<0.05 compared to CTL by one-way analysis of variance (ANOVA) followed by Dunnetts multiple comparisons test. MFI: Mean fluorescence intensity.

Next, the protein levels of several elements of the NF-B and Akt/mTOR signaling pathways were quantified. The levels of phospho-NF-B p65 (S529) (Fig. 6C, D), Akt1 (Fig. 6E, F), phospho-Akt (S473) (Fig. 6G, H), phospho-mTOR (S2448) (Fig. 6I, J), and phospho-S6 (S235/S236) (Fig. 6K, L) were reduced in complex 1-treated HepG2 cells. In contrast, the levels of Akt1 (Fig. 6E, F), phospho-Akt (S473) (Fig. 6G, H), and phospho-mTOR (S2448) (Fig. 6I, J) were reduced after treatment with complex 2, indicating that these complexes interfere with NF-B and Akt/mTOR signaling. The levels of phospho-PI3K p85/p55 (T458/T199) (Fig. S7A and S7B), phospho-Akt (T308) (Fig. S7C and S7D), phospho-4EBP1 (T36/T45) (Fig. S7E and S7F), and phospho-elF4E (S209) (Fig. S7G and S7H) were not affected by treatment with these complexes.

As complexes 1 and 2 downregulate the level of phospho-mTOR (S2448), a negative regulator of autophagy [22], the effect of these complexes on the induction of autophagy was investigated. On the other hand, none of them caused autophagy, as assessed by quantification of p62/SQSTM1 expression levels in HepG2 cells treated with complexes 1 and 2 (Fig. S8AS8C).

Since both complexes reduced the proportion of liver CSC markers in HepG2 cells and liver CSCs are directly associated with cell migration and invasion [23], we hypothesized that these complexes could reduce HepG2 cell motility. Initially, noncytotoxic concentrations of complexes 1 and 2 were selected (Fig. S9) and tested in the wound healing assay. Both complexes reduced HepG2 cell migration after 72h of incubation at noncytotoxic concentrations (0.5M for complex 1 and 0.3M for complex 2) (Fig. 7A, B). Similarly, both complexes, at the same concentrations, also reduced motility in a transwell cell migration assay (Fig. 7C, D) using HepG2 cells.

A Representative images and (B) quantification of HepG2 cell migration in the wound healing assay after 72h of incubation with complexes 1 and 2. C Representative images and (D) quantification of HCT116 cell migration in the transwell migration assay after 24h of incubation with complexes 1 (8M) and 2 (6M). Quantification of vimentin (E, F) and E-cadherin (G, H) expression in HepG2 cells after 24h of incubation with complexes 1 (8M) and 2 (6M), as determined by flow cytometry. The vehicle (0.2% DMSO) was used as a negative control (CTL). The data are expressed as the meanS.E.M. of three biological replicates carried out in duplicate. *P<0.05 compared to CTL by one-way analysis of variance (ANOVA) followed by Dunnetts multiple comparisons test. MFI: Mean fluorescence intensity.

Next, the epithelialmesenchymal transition (EMT) markers vimentin and E-cadherin were evaluated in HepG2 cells treated with complexes 1 and 2 after 24h of incubation. Vimentin (Fig. 7E, F) was reduced, and E-cadherin (Fig. 7G, H) was increased by treatment with complex 2, indicating that this molecule can modulate EMT.

The in vivo antitumor activity of complexes 1 and 2 was investigated in C.B-17 SCID mice grafted with HepG2 cells. The animals were treated with 2 or 4mg/kg of both complexes intraperitoneally once a day for 21 consecutive days. Both complexes inhibited the growth of HepG2 cells in mice (Fig. 8A, B). At the end of treatment, the mean tumor weight in the negative control group was 981mg, while it was 665mg in the doxorubicin-treated group. In complex 1-treated animals, the mean tumor weights were 635 and 455mg, corresponding to 35.3 and 53.6% tumor inhibition, respectively. In complex 2-treated animals, the mean tumor weights were 358 and 340mg, corresponding to 63.6 and 65.4% tumor inhibition, respectively. Doxorubicin reduced the tumor weight by 32.2%.

A, B In vivo antitumor activity of complexes 1 and 2 on C.B-17 SCID mice inoculated with HepG2 cells. The animals were treated with complexes 1 and 2 at doses of 2 or 4mg/kg intraperitoneally once a day for 3weeks. C Representative photomicrographs of HepG2 tumors from animals treated with complexes 1 and 2. Histological sections were stained with hematoxylin-eosin and analyzed by light microscopy. The asterisks indicate areas of tissue necrosis. Scale bar=50m. The vehicle (5% DMSO) was used as a negative control (CTL), and doxorubicin (DOX, 1mg/kg) was used as a positive control. The data are expressed as the meanS.E.M. from 8 animals. *P<0.05 compared to CTL by one-way analysis of variance (ANOVA) followed by Dunnetts multiple comparisons test.

The tumors presented histological characteristics compatible with hepatocellular carcinoma, such as intense cellular and nuclear pleomorphism, hyperchromatism, atypical mitotic figures, and hepatocyte-like cells (Fig. 8C). The histological grading of the tumors varied from poorly to moderately differentiated in all the experimental groups. The tumor cells were organized in nodules or cords surrounded by a poorly vascularized collagen matrix. Areas of coagulative necrosis were frequent, especially in more central tumor regions. In addition, an infiltrate of inflammatory cells, predominantly mononuclear, was observed mainly adjacent to the necrotic areas. Areas of dystrophic calcification were observed in some of the tumors in the negative control, doxorubicin and complex 2 (4mg/kg) groups. Furthermore, invasion fronts in the muscular tissue were observed in the control groups.

The toxicity parameters of animals treated with complexes 1 and 2 were also examined. No significant changes in body weight or organs (liver, kidney, lung, or heart) were detected in the animals treated with these complexes (P>0.05) (Fig. S10AS10F). Histopathological analysis of the kidneys (Fig. S11), livers (Fig. S12), and lungs (Fig. S13) of mice treated with complexes 1 and 2 revealed some alterations that were minor and/or reversible, indicating little damage to normal tissues. No significant changes were observed in the hearts of the animals treated with complexes 1 and 2 (data not shown).

Excerpt from:
Ru(II)-based complexes containing 2-thiouracil derivatives suppress liver cancer stem cells by targeting NF-B and Akt ... - Nature.com