Vincristine treatment generated a population of RMS cells with stem-like characteristics
To characterize the cellular characteristics and state of vincristine-resistant RMS cells, we treated a panel of FN and FP RMS cell lines (FN: RD and SMS-CTR; FP: Rh5 and Rh30) with a concentration of vincristine that killed 80–90% of cells (IC80-90) over 7 days. This treatment resulted in a population of slow-growing cells in comparison to the parental lines (Fig. 1A-E; Fig. S1A). These vincristine-resistant cells showed enlarged and irregular cell size with bizarre-shaped nuclei. By cell cycle analysis, vincristine-resistant cells showed a delay in cell cycle with increased number of cells in the quiescent G0/G1 phase and reduced number of cells in the proliferative S phase compared to the parental cells (Fig. 1F).
To assess whether vincristine treatment alters the cell state of the RMS cells, we showed that in comparison with the parental cells, FN RD and SMS-CTR cells following 7-day treatment of vincristine at IC 80–90 were enriched with CD133-positive cells on immunofluorescence (IF) (Fig. 1G-J). These cells also showed statistically significant up-regulation of stem cell markers, CD133, PAX7, OCT4, NANOG and SOX2, by quantitative RT-PCR (Unpaired t-test, P < 0.01, Fig. 1K).
FN RMS spheres enrich for CD133-expressing stem-like cells and show increased resistance to standard-of-care therapeutics
RMS spheres cultured in stem-cell media have previously been shown to be enriched for CD133-positive stem-like cells and are more resistant to the chemotherapeutics, cisplatin and chlorambucil 13. Similarly, we showed that serially-passaged spheres derived from RD showed an increase in CD133 expression as well as CD133-expressing cells compared to the bulk xenograft tumor-derived from the RD line (ANOVA, P < 0.0001, Fig. S1 B-C). The spheres generated from the SMS-CTR cells also showed a similar increase in the CD133-expressing cells (Fig. S1 D). By contrast, the spheres generated from the two FP RMS lines (Rh5 and Rh30) lacked CD133-positive expressing cells (Fig. S1 D), suggesting that a different marker may better highlight the stem cell-like population in FP RMS cells. The spheres generated from both FN RMS cell lines (RD and SMS-CTR) and FP RMS cell lines (Rh5 and Rh30) treated with varying doses of the standard-of-care chemotherapeutic agent, vincristine, all exhibited decreased sensitivity compared to the parental adherent cells, as evidenced by ~ 3 to 9-fold increase in IC50 (Fig. S1 E).
MYC and YBX1-driven transcriptional regulatory networks are up-regulated in CD133-positive stem-like RMS cells
As CD133-positive stem-like cells are enriched in the FN RMS spheres, we used serially passaging of RD sphere cells cultured in stem cell media to generate a sufficient number of CD133-positive cells for gene expression profiling by single cell RNA sequencing. Sphere cells harvested at passages 1 (P1) and 5 (P5) were subjected to processing and analysis by single cell RNA sequencing. As displayed by Uniform Manifold Approximation and Projection (UMAP) (Fig. 2A), both P1 and P5 RD sphere cells showed similar clustering patterns (clustering for P1 and P5 shown in Fig. 2A) with approximately 19 unique clusters identified. CD133 (PROM1)-expressing cells showed relatively diffuse distribution throughout all clusters (Fig. 2B), indicating dynamic transcriptional plasticity of CD133-expressing cells. Through the enrichment analysis against TRRUST (Transcriptional Regulatory Relationships Unraveled by Sentence-based Text mining) database using Metascape, we observed that MYC and YBX1 are among the top-scored master transcription factors in CD133-high expressing RD cells compared to CD133-low expressing cells (Fig. 2C). Protein expression of MYC and YBX1 was also detected in serially passaged spheres by immunohistochemistry (RD P5 spheres shown in Fig. 2D) with MYC showing nuclear localization and YBX1 showing cytoplasmic localization. In addition, we found stronger correlations between MYC and YBX1, MYC and MYC target genes as well as between YBX and YBX1-target genes, in CD133-high expressing sphere cells in comparison with CD133-low expressing cells (Fig. 2E, 2F ; Fig. S2 A-C). Up-regulation of well-known downstream targets genes of MYC and YBX1 was noted in CD133-high expressing RD cells (Fig. 2D-E), which were independently validated on RD sphere cells using quantitative RT-PCR (Fig. S2 D).
Targeted disruption of MYC and YBX1 inhibits tumor RMS tumor cell growth and sphere formation and induces cell death in CD133-positive stem-like cells
To assess the loss-of-function effects of MYC and YBX1 on RMS tumor cell growth and sphere formation, FN RMS (RD, SMS-CTR) and FP RMS (Rh5 and Rh30) Cas9-expressing stable lines were transduced with lentiviral constructs expressing double gRNAs (dgRNA) against MYC and YBX1. CRISPR/Cas9-mediated disruption of MYC and YBX1 (Fig. 3A; Fig. S3 A) inhibited cell growth by cell counts and CellTiter-Glo viability assays (Fig. 3B; Fig. S3 B) as well as reduced sphere formation frequency and size at two cell dilutions (Fig. 3C-G; Fig. S3 C-D). To demonstrate gene targeting specificity, we showed that overexpression of Cas9-resistant MYC and YBX1 alleviated the cell growth defect resulted from CRISPR/Cas9-mediated gene targeting of MYC and YBX1, respectively (Fig. S3 E).
To assess the effect of MYC and YBX1 gene disruption on the viability of CD133-positive sphere cells real-time, we generated a CD133:GFP SMS-CTR reporter line by knocking in the GFP cassette in frame with the CD133 locus via CRISPR/Cas9. By live imaging of dissociated sphere cells derived from the CD133:GFP SMS-CTR stained with the fluorogeneic NucView®530 Caspase-3 substrate 5 days following lentiviral transduction of the CRISPR dgRNA constructs and antibiotic selection, we observed an increased number of cells undergoing apoptosis including the CD133-positive cells (indicated by arrowheads in Fig. 3H; quantitation in Fig. S3 F, ANOVA, P < 0.001). Together, MYC- and YBX1-targeted disruption decreased stem-like characteristics of RMS cells and contributed to increased apoptosis in the CD133-positive stem-like cell population at least in FN RMS.
MYC and YBX1 play an important role in modulating chemosensitivity of the RMS cells
Given that MYC and YBX1 are essential for the viability of the CD133-positive stem-like FN RMS cells, and short-term vincristine-treated FN RMS-resistant cells have shown an increase in CD133-positive stem-like cells, we next asked whether MYC and YBX1 play a role in modulating chemosensitivity of FN RMS cells to vincristine. We first showed by quantitative RT-PCR that there was a relative increase in expression of MYC and YBX1 in the resistant RD and SMS-CTR cells following 7-day treatment of vincristine at IC 80–90 compared to vehicle (DMSO)-treated parental cells (Unpaired t-test, P < 0.01, Fig. 4A). To assess whether YBX1 and MYC are essential for the viability of resistant FN RMS cells following a short-term treatment with vincristine at IC 80–90, we transduced the Cas9-expressing RD cells that survived 7-day treatment of vincristine at IC80-90 with gene-specific double gRNAs (dgRNAs) and re-started vincristine treatment following transduction and antibiotic selection (schematic in Fig. 4B). While the vincristine-treated cells with targeted disruption of a safe-harbor genomic region targeting as a control remained relatively tolerant to the vincristine compared to the parental cells, targeted disruption of MYC and YBX1 reduced cell growth and increased cell death in the vincristine-treated cells (ANOVA, P < 0.05, Fig. 4B-D and Fig. S4 A). In addition, we generated RD and SMS-CTR resistant cells following a long-term treatment with incremental increase in vincristine concentration until a 5-10-fold increase in the IC50 compared to the parental cells (~ 9 months duration). In contrast to short-term treatment of vincristine, which generated resistant cells in predominantly dormant G0 state (Fig. 1F), long-term treatment of vincristine generated RD and SMS-CTR cells that showed cell cycle re-entry from G0 with delay in the G2-M phase (Fig. S4 C). CRISPR/Cas9-mediated targeted disruption of MYC and YBX1 in these resistant cells generated from long-term exposure of vincristine also showed reduced growth compared to safe-harbor region-targeted control resistant cells (data for RD in Fig. 4E, ANOVA, P < 0.05), indicating that MYC and YBX1 play an important role in maintaining the viability of vincristine-resistant FN RMS cells over time.
To assess whether MYC or YBX1 overexpression can confer tolerance to vincristine treatment on RMS cells, we showed that overexpression of MYC but not YBX1 in the Rh30 FP RMS line with low MYC expression conferred increased tolerance to vincristine in a dose-dependent manner (ANOVA, P < 0.0005, Fig. 4F).
Reciprocal regulation of MYC and YBX1 in FN RMS cells
Given the role of MYC in gene transcription and the DNA and RNA-binding capacity of YBX1, we next asked whether MYC and YBX1 showed mutual regulation at the transcriptional or post-transcriptional level. We first observed that upon targeted disruption of MYC in two FN RMS lines, RD and SMS-CTR, there was reduced mRNA expression of YBX1 by quantitative RT-PCR and protein expression by Western blot analysis (Fig. 3A, Fig. S3 A and Fig. 5A-B). Similarly, targeted disruption of YBX1 in RD and SMS-CTR cells also reduced MYC mRNA and protein expression (Fig. 3A and Fig. 5A-B).
To assess whether MYC regulates expression of YBX1 at the level of transcription, we showed by the CUT and RUN assay performed on the RD and SMS-CTR cells treated with vincristine that MYC bound to the promoter containing the consensus MYC binding motif (CACGTG) of YBX1, CD133 and muscle satellite/RMS tumor propagating cell markers, MYF5 and PAX7 (Fig. 5C-D). YBX1 was expressed in the cytoplasm in RMS cells (Fig. 2D), suggesting its functional role at the post-transcriptional level. To assess whether YBX1 binds to the MYC mRNA, we demonstrated the YBX1 protein-MYC mRNA interaction by the RNA immunoprecipitation assay (RIP) performed on RD cells using the antibody against YBX1 in the RNA pull down and primers against the 3’ UTR region of MYC mRNA (Fig. 5E).
Overall, the findings indicate that MYC and YBX show reciprocal regulation with MYC regulating gene expression of YBX1, and YBX1 regulating mRNA stability of MYC.
The combination treatment of the MYC inhibitor, MYCi361, and vincristine significantly inhibits RMS tumor cell growth and depletes stem-like cell subpopulation in vivo.
MYCi361 is a MYC inhibitor previously described by Han et al to decrease MYC protein activity by disrupting MYC/MAX dimer formation and impairing MYC-driven gene expression 14. We showed in cultured RD and SMS-CTR cells that treatment with the combination of MYCi361 and vincristine significantly reduced cell growth compared to treatment with each agent alone (Fig. S5 A). By quantitative RT-PCR, MYCi361-treated RD and SMS-CTR cells showed decreased expression levels of known MYC target genes (Fig. S5 B). To assess whether MYCi361 could be a potential therapeutic agent in combination with vincristine to reduce RMS tumor cell growth in vivo, we tested the two-agent combination in the KRAS(G12D)-induced zebrafish FN RMS model. Primary zebrafish RMS tumor cells expressing a transgenic (rag2:dsRed) fluorescent reporter were transplanted into a pool of syngeneic CG1-strain host zebrafish, engrafted tumor fish were treated with DMSO (vehicle control), MYCi361 (100 mg/kg), vincristine (0.4 mg/kg) or MYCi361 and vincristine in combination with half the dosage for each drug for 7 days. The dosage of MYCi361 and vincristine was titrated to a level that only slightly affected tumor growth as a single agent in vivo. We observed that treatment of zebrafish RMS tumors with the combination of vincristine and MYCi361 significantly reduced tumor growth compared to the tumors treated with vincristine or MYCi361 alone (Fig. 6A-B). This was supported by a significantly greater decrease in the proliferative index and increase in the number of apoptotic cells in tumors treated with the two-drug combination, as highlighted by immunohistochemistry for Ki67 and Caspase-3, respectively (see representative tumor sections from each treatment cohort and quantitative results in Fig. 6C-D).
To assess whether MYCi361 modulates the relative proportions of tumor cell subpopulations based on their differentiation status, we took advantage of the myf5:GFP/mylz2:mCherry transgenic zebrafish line to label distinct cell subpopulations of FN RMS based on their differentiation status in vivo 15. The myf5:GFP+/mylz2:mCherry-negative (G + R-) stem cell-like tumor-propagating-cell (TPC) population has been shown to have self-renewal capacity while late-differentiating myf5:GFP-negative/mylz2:mCherry + tumor cells (G-R+) lack the capacity to self-renew 15,16. Following 7-day drug treatment of fish-bearing engrafted myf5:GFP/mylz2:mCherry expressing RMS tumors, relative proportions of cell subpopulations were assessed by Fluorescence Activated Cell Sorting (FACS) (n = 3 independent tumors). The treatment with MYCi361 resulted in significant depletion in the proportion of the myf5:GFP+/mylz2:mCherry-negative TPC population and a concomitant increase in the proportion of myf5:GFP-negative/mylz2:mCherry + late-differentiating cells (Representative flow cytometric analysis and quantitation of cell subpopulations in Fig. 6E-F; p < 0.05), MYCi361-treated tumor cells also showed decreased expression of stem genes, pax7 and myf5 (Fig. S5 C).
High MYC expression correlates with poor survival outcomes of FN RMS.
To assess whether MYC and YBX1 expression could be detected in recurrent or metastatic RMS samples, we assessed by immunohistochemistry protein expression of MYC and YBX1 in a tissue microarray of tissue cores from 32 RMS patients, including 17 primary tumors (12 FN and 5 FP) and 15 recurrent or metastatic tumors (8 FN and 7 FP) along with 5 muscle controls. FN samples include ERMS (n = 9), FN ARMS (n = 2) and spindle cell/sclerosing variant (n = 1). MYC and YBX1 were expressed in at least a subset of primary RMS samples (MYC: 3 of 9 (33%) in FN RMS, 1 of 5 (20%) in FP RMS; YBX1: 12 of 12 (100%) in FN RMS, 5 of 5 (100%) in FP RMS) as well as recurrent/metastatic RMS samples (MYC: 4 of 8 (50%) in FN RMS, 1 of 7 (14.3%) in FP RMS; YBX1: 8 of 8 (100%) in FN RMS, 7 of 7 (100%) in FP RMS). MYC staining showed a patchy distribution, while YBX1 staining showed a diffuse distribution (Fig. 7A). All normal muscle tissue samples were negative for both MYC and YBX1 (Fig. 7A). In all, MYC showed a trend of increased frequency of positive expression in recurrent/metastatic FN RMS, and YBX1 was expressed in all primary and metastatic FN and FP RMS tumor tissue samples examined (Fig. 7B), suggesting that MYC and YBX1 play a role in both primary and recurrent/metastatic RMS tumors.
To examine potential roles of MYC and YBX as biomarkers for RMS patients, we studied MYC or YBX mRNA expression in 81 RMS cases with survival data (63 fusion-negative and 18 fusion-positive). we first showed that high expression of MYC correlated positively with high expression of YBX1 in both FN and FP patients, separately (Fig. 7C-D). We then asked whether MYC or YBX1 expression in RMS tumors associated with patients’ survival. Kaplan–Meier curves were generated based on gene expression values dichotomized into over- and under-expressed groups using the median expression value within each cohort as a cutoff. However, while expression of YBX did not correlate with overall survival, high expression of MYC correlated with decreased overall survival in the FN RMS patients (Fig. 7E, log-rank test, p = 0.043, HR = 2.08, 95% CI = 1.04–4.51) but not the FP RMS cases (Fig. 7F, p = 0.673, HR = 1.46, 95% CI = 0.25–8.48). Although the small sample size in both subsets compromises the ability to draw firm conclusions, the findings suggest that expression of MYC might be useful as a clinical prognostic biomarker if they are confirmed in a prospective analysis.