RSL3 induces apoptosis in MDS cells.
To determine whether RSL3 has some roles in MDS cells, MDS-L and SKM-1 cells were treated with different concentrations of RSL3 at specific times, and cell viability was analyzed by manual counting and soft agar clonogenic assays. The results showed that RSL3 significantly inhibited MDS cell expansion (Fig. 1A). And soft agar clonogenic assay also demonstrated that RSL3 inhibited colony formation of MDS-L and SKM-1 cells compared with control groups (Fig. 1B). Furthermore, we explored the mechanism by which RSL3 suppresses MDS cell proliferation. It was found that RSL3 treatment led to a series of morphological changes, including reduced volume, ruptured cytomembrane and decreased cell density, in both MDS-L and SKM-1 cells (Fig. 1C). Therefore, we speculated that RSL3 induces MDS cell death to inhibit cell proliferation. Ferroptosis, apoptosis, autophagy, and necroptosis are common types of programmed cell death26. To identify which kind of cell death was involved, MDS cells were treated with RSL3 alone or in combination with liproxstatin-1 (Lip-1; a ferroptosis inhibitor targeting GPX4), ferrostatin-1 (Fer-1; a ferroptosis inhibitor targeting SCL7A11), deferoxamine (DFO; a chelator), Z-VAD-FMK (an apoptosis inhibitor targeting caspase), chloroquine (CQ; an autophagy inhibitor), or necrostatin-1 (Nec-1; a necroptosis inhibitor). The results showed that only Z-VAD-FMK notably reversed the inhibition of proliferation induced by RSL3 in a concentration-dependent manner (Fig. 1D, E). Western blot results showed that the levels of cleaved PARP, cleaved caspase3 and γ-H2AX were increased and Bcl-2 was decreased with increasing concentrations of RSL3 in both MDS-L and SKM-1 cells (Fig. 1F), further confirming RSL3 induces apoptosis in MDS cells. Flow cytometric analysis also showed that RSL3 significantly induced MDS cell apoptosis and that Z-VAD-FMK restored cell viability (Fig. 1G). Taken together, these data showed that RSL3 suppressed the malignant proliferation of MDS cells through inducing apoptosis.
RSL3-mediated apoptosis is dependent on ROS.
ROS accumulation occurs during apoptosis27. Based on the above results, we measured ROS levels in RSL3-treated MDS cells. MDS cells were treated for with RSL3 alone or in combination with NAC (a ROS inhibitor) and subsequently analyzed by flow cytometry. RSL3 increased ROS levels in MDS cells in a dose-dependent manner (Fig. 2A). And NAC partially reversed the RSL3-induced ROS accumulation (Fig. 2B). Furthermore, fluorescence staining also revealed that RSL3 treatment increased ROS fluorescence in MDS cells and NAC could attenuate this effect (Fig. 2C). These results showed the accumulation of ROS in RSL3-treated MDS cells. At the meanwhile, cell proliferation assay showed NAC can rescue the proliferation inhibition induced by RSL3 treatment in MDS cells in a dose-dependent manner, indicating that ROS production plays an important role in the RSL3-induced MDS cell proliferation inhibition (Fig. 2D). Importantly, the levels of cleaved PARP, Bcl-2 and cleaved caspase 3, were greatly reduced after NAC treatment (Fig. 2E), and NAC reduced the RSL3-mediated apoptosis rates in MDS-L and SKM-1 cells (Fig. 2F), indicating that inhibition of ROS prevented RSL3-induced apoptosis. Thus, RSL3-triggered cell apoptosis is dependent on ROS.
MYB is an efficacious target of RSL3-mediated apoptosis.
To explore the detailed molecular mechanism of RSL3-induced cell death, RNA-seq was performed in MDS-L cells treated with RSL3 or DMSO. Principal component analysis (PCA) showed clustering of triplicated samples in each group (Fig. 3A). Volcano plot showed the differential expression of all genes. Compared with control group, there were 3099 genes significantly up-regulated and 2265 genes significantly down-regulated in RSL3 treatment group (Fig. 3B). Importantly, GSEA analysis showed that apoptosis pathway was significantly enriched in RSL3 treated cells, which was consistent with RSL3’s roles (Fig. 3C). We further performed Venn diagram analysis to compare downregulated genes identified by RNA-seq with upregulated genes in MDS patients identified in a GEO dataset (GSE114869). The Venn diagram contained 26 overlapping molecules, which were visualized in a heatmap (Fig. 3D). These molecules could promote the initiation and development of MDS and be potential targets of RSL3. Our previous study found that MYB is overexpressed in clinical MDS samples compared with normal samples and that MYB regulates MDS-L and SKM-1 cell viability via PI3K-AKT pathway. Interestingly, RNA-seq results showed that RSL3 treatment decreased MYB expression (Fig. 3D), which was further confirmed by quantitative PCR (Fig. 3E). RNA seq analysis showed that MYB was downregulated and ferroptosis-associated genes like EGLN1 and KEAP1 had no significant changes after RSL3 treatment (Supplemental Fig. 1A).
To further explore whether RSL3-mediated MDS cell apoptosis was related with decreased MYB levels. We did rescue assays by transfected MDS-L and SKM-1 cells with MYB overexpression plasmids. Western blot result confirmed the over-expression efficiency (Fig. 3F). And cell proliferation assay showed that MYB overexpression partially rescued the decrease in MDS cell viability caused by RSL3 (Fig. 3G). Flow cytometric analysis showed that RSL3 increased ROS production in MDS-L and SKM-1 cells and overexpression of MYB reverted this effect (Fig. 3H). More importantly, western blot results showed that MYB overexpression elevated the Bcl-2 protein level, decreased the cleaved caspase 3 and γ-H2AX and reversed the effects of apoptotic protein expression caused by RSL3 (Fig. 3I). And the apoptosis rate in the RSL3-treated group was significantly decreased upon MYB overexpression (Fig. 3J). Taken together, RSL3 induced apoptosis in MDS cells at least partially via decreasing MYB expression.
RSL3 reduces MYB expression via binding MYB.
We further explored the mechanism of RSL3 regulating MYB expression. Western blot results showed that RSL3 treatment decreased MYB protein in dose and time-dependent manners (Fig. 4A-B). And in both MDS-L and SKM-1 cells, RSL3 treatment dramatically decreased MYB protein levels in as short as 6 hours (Fig. 4B). This fast reduction of protein levels usually involves stabilization and degradation pathway. Indeed, proteasome inhibitor MG132 treatment can rescue decreased MYB protein levels led by RSL3 treatment (Fig. 4C). Furthermore, cellular thermal shift assay (CETSA) result showed that RSL3 protected MYB from thermal degradation, indicating that RSL3 can directly bind to MYB protein (Fig. 4D). In summary, these results demonstrated that RSL3 decreased MYB expression via directly binding to MYB protein and mediating its degradation via proteasome-ubiquitin pathway.
RSL3 inhibits tumor growth in vivo.
To further explore the antitumor role of RSL3, SKM-1 cells were inoculated subcutaneously into the flanks of female BALB/c mice (Fig. 5A). After the tumor volume increased to 50–100 mm3, the mice were divided into the RSL3-treated group and the control group. RSL3 (20 mg/kg) or vehicle was administered intraperitoneally into mice once every two days. It was shown that tumor volume and weight were markedly smaller in RSL3-treated mice than control mice, while the body weight did not differ significantly between the RSL3-treated group and control group, indicating the safety of RSL3 administration (Fig. 5B-E). Importantly, immunofluorescence staining results showed that RSL3 treatment decreased MYB and Bcl-2 expression in the tumor tissues, which was further confirmed in western blot results (Fig. 5F-H). In conclusion, RSL3 inhibits tumor growth in vivo.
RSL3 treatment had therapeutic efficiency in MDS patients’ derived samples.
To determine the potential clinical significance of RSL3 in MDS, we first treated peripheral blood mononuclear cells (PBMCs) from MDS patients with RSL3. Our results showed that RSL3 had an inhibitory effect in mononuclear cells from MDS patients but almost no effect in cells from the healthy counterparts (Fig. 6A-D). In clinical, resistance to hypomethylating agents (HMA) such as azacitidine (AZA) and decitabine (DAC) leads to treatment failure in MDS patients. Given the advantage of RSL3 in decreasing the MDS cells, we tested that whether RSL3 can synergize with HMA via combining RSL3 and DAC to treat MDS cells. The results showed that RSL3 had an obvious synergistic effect with DAC in MDS cell lines (Fig. 6E-F). Taken together, these results indicated that RSL3 inhibited the cell viability of MDS patients and was promising in combination with HMA agents to treat MDS cells.