Enoxacin has a cancer-specific growth-inhibitory effect on ESCC cells
To analyze the effects of enoxacin on ESCC cells, we first asked if enoxacin could reduce ESCC cell viability. The initial step was to determine the optimized enoxacin concentration to treat the cells. To this end, we cultured ESCC cells in 96-well plates, treated them for 5 days with different concentrations of enoxacin (0, 10, 25, 50, 75, 100, 125, 150, 200, 250, 300, and 350 µM), and measured ESCC cell viability using MTS assays. We observed that enoxacin dose-dependently suppressed the growth of ESCC cells, and determined its median inhibitory concentration (IC50) to be 125 µM (Supplementary Fig. 1a). Since several other studies have determined the IC50 of enoxacin to be 124 µM15,17, we decided to perform all our subsequent treatments with 124 µM which is almost identical to the IC50 that we determined in the current study. Notably, this inhibitory effect was observed to be cancer-specific, as enoxacin did not adversely affect the growth of non-cancerous cells i.e. human foreskin fibroblasts (HFFs), judging from crystal violet staining, MTS viability assessments, and live/dead staining (Supplementary Fig. S1b-d). We subsequently treated the ESCC cells with 124 µM of enoxacin continuously for 5 days for various experiments, as the 5-day treatment duration exerted a more potent inhibition than 3- or 4-day treatments on cell viability (Supplementary Fig. 1e). We also noticed that it was not necessary to treat the cells daily with fresh enoxacin as there was not a significant difference in cell viability when changing the media either every 24 h or every 48 h (Supplementary Fig. 1f). Therefore, we exposed the ESCC cells to enoxacin for 5 continuous days and refreshed the media of the cells every 48 h for the subsequent assays.
Next, the effect of enoxacin on growth and survival of the ESCC cell line, KYSE-30, was investigated. The cells were exposed to enoxacin for 5 days, and their survival was measured on day 5. We found that the number and viability of KYSE-30 cells were considerably decreased by enoxacin treatment compared to the untreated group, as evidenced by phase-contrast imaging (Fig. 1a), crystal violet staining (Fig. 1b), and live/dead assays where enoxacin decreased the number of live (green) cells and simultaneously increased the number of dead (red) and dying (yellow) cells (Fig. 1c-d). This result was further confirmed using MTS viability assays (Fig. 1e) and the analysis of cell growth rate on the basis of cell counting following enoxacin treatment (Fig. 1f). We next repeated the same experiments in a second ESCC cell line, YM-1, to examine how reproducible the growth-inhibitory effects of enoxacin are. We obtained a highly-similar set of results for YM-1 cells and indicated that enoxacin potently inhibited the growth of YM-1 cells judging from phase-contrast imaging (Fig. 2a), crystal violet staining (Fig. 2b), live/dead staining (Fig. 2c-d), MTS viability assays (Fig. 2e), and cell counting (Fig. 2f). These results indicate that enoxacin reproducibly inhibits the survival of ESCC cells whereas it has no inhibitory effects on the viability of normal cells.
ESCC cells undergo cell cycle arrest and apoptosis upon treatment with enoxacin
After noticing a clear decline in survival of ESCC cells while being exposed to enoxacin, we looked at the cell cycle and apoptosis rate of the cells to examine if these two phenomena contribute to the decreased cell number observed upon enoxacin treatment. To determine the cell cycle profile, ESCC cells were seeded in 35-mm culture dishes, and were synchronized in G1 phase by being maintained under FBS-free conditions before the treatment. We found that 5 days treatment with enoxacin resulted in cell cycle arrest in the S phase of the KYSE-30 cell cycle (Fig. 3a-b) and in the G1 phase of YM-1 cell cycle (Fig. 3c-d), indicating that enoxacin impairs the cell cycling of the cancer cells.
Next, we sought to determine the apoptotic rate of the cells following exposure to enoxacin. Four days post-treatment, the KYSE-30 cells in the treatment and control groups were harvested and prepared for annexin-PI staining. Flow cytometry analysis of the cells stained with annexin-PI revealed that in the enoxacin-treated cells compared to the control group, (i) the number of live (non-apoptotic) cells was lower and (ii) the percentage of cells in the late phase of apoptosis was significantly higher (Fig. 3e-f). We also observed a similar increase in apoptosis rate when exposing YM-1 cells to enoxacin (Fig. 3g-h). Overall, enoxacin treatment induces cell cycle arrest as well as apoptosis in ESCC cells, thereby explaining the observed reductions in the number of ESCC cells upon treatment with the chemical.
Enoxacin inhibits capacity of ESCC cells to form colonies and migrate
Next, we sought to determine whether enoxacin affects the clonogenicity of ESCC cells. To this end, the cells were treated with enoxacin for 5 days, and on day 5, the cells in both control and treatment groups were trypsinized, from which 200 cells were transferred to 35-mm cell culture dishes containing complete ESCC medium for colony formation. Within 3 weeks, several colonies emerged from the seeded cells which were then stained on day 21 with crystal violet solution and counted. The results indicated that enoxacin efficiently inhibited colony formation of KYSE-30 cells (Fig. 4a). The number and type of colonies were then analyzed and it was revealed that the colonies of the control group cells were mostly holoclones with a few colonies being meroclones or paraclones (Fig. 4b-c). In contrast, the colonies derived from the enoxacin-treated cells were more similar to cell clusters than cell colonies with few of them being paraclones (i.e. no holoclones and meroclones observed) (Fig. 4b-c). We also studied the YM-1 cell line in terms of clonogenicity in response to enoxacin exposure, and found a clear reduction in the number of colonies formed in the treated group compared to the control group (Fig. 4d-f), confirming that enoxacin is reproducibly detrimental to colony formation by ESCC cells.
Next, we asked if enoxacin inhibits the ability of cancer cells to migrate. Toward this goal, we carried out wound healing (i.e. scratch) assays post-treatment and examined the scratch area over 24 h. As shown in Fig. 5a-b, enoxacin considerably inhibited the closure of scratch wounds of the cultured cancer cells compared to the untreated cells. Since cancer cell migration is known to be dependent on cancer stem cells, we then analyzed the expression of several stemness genes e.g. OCT4, SOX2, NANOG, KLF4, c-MYC, and LIN28 upon treatment with enoxacin. Our qRT-PCR analysis revealed that enoxacin suppressed the expression of almost all these genes 48 h post-treatment compared to the untreated control group (Fig. 5c). Moreover, it significantly decreased the expression of typical cancer-associated genes ALDH1, CD44, and CD133 (Fig. 5d). In conclusion, enoxacin blocks ESCC cell migration by downregulating key stemness and cancer genes.
TARBP2 mediates the anti-tumor effects of enoxacin in ESCC cells
Previous research has demonstrated that enoxacin induces cancer cell suppression by speeding up the maturation of miRNAs through binding and stimulating the TARBP2 protein, which is the physical partner of DICER1. To examine if enoxacin similarly promotes TARBP2-mediated processing of cellular miRNAs in ESCC cells, we treated the cancer cells with enoxacin and analyzed the expression of a selected set of miRNAs 48 h after exposure. It was observed using qRT-PCR assays that all the tested miRNAs were upregulated in response to enoxacin treatment (Fig. 6a), which was consistent with what reported previously15.
Next, we attempted to confirm whether TARBP2 is the mediator of enoxacin actions in ESCC cells. To this end, we first transfected the cells with a FITC-conjugated oligonucleotide to determine the efficiency of our transfection-based approach for the delivery of the TARBP2-targeting ASOs. We observed that our lipid-based transient transfection method delivers the FITC-labelled oligonucleotides to the cultured ESCC cells in a highly efficient manner (~ 70% of the cells received the oligonucleotide), as evidenced by fluorescence microscopy and flow cytometry (Supplementary Fig. 2a-b). Next, we suppressed the expression of TARBP2 transcript using a pool of two different TARBP2-targeting ASOs with two different concentrations to determine the optimal ASO concentration for TARBP2 silencing, and found that a significant downregulation of TARBP2 transcripts could be similarly achieved upon treatment with either of the ASO concentrations (Fig. 6b), therefore we chose the lower concentration (100 nM) for transfection. Then, we treated the ESCC cells according to the procedure shown in Fig. 6c to examine if TARBP2 knockdown abolishes the growth-inhibitory effects of enoxacin. As shown in Fig. 6d-i, (i) enoxacin alone sharply reduced cell viability; (ii) TARBP2-targeting ASOs had no inhibitory effects on ESCC cell growth; and (iii) enoxacin-ASO combination group did not reduce cell viability as potent as enoxacin alone, judging from phase-contrast imaging (Fig. 6d), crystal violet staining (Fig. 6e), live/dead staining (Fig. 6f-g), cell counting (Fig. 6h), and MTS viability assays (Fig. 6i). Taken together, these results confirm that enoxacin suppresses the growth of ESCC cells by TARBP2-mediated enhancement of miRNA maturation.