Fig 1.
Identification of CAP molecules as WNT-TCF response blockers.
A) Chemical structure of the CAP1, CAP2 and CAP3 steroidal lactone small molecules. B) Changes in relative fluorescence intensity of CC14-GFP+ cells over time. At 0.15μM only CAP2 is effective whereas at 2.5μM CAP1, CAP2 and CAP3 repress increases in fluorescence (see also panel C in S3 Fig), related here to cell state and mass. P values are given for the end points in relation to DMSO-treated controls. The inset shows a heat map with normalized ratios over control highlighting the repression of AXIN2, LGR5 and cMYC by CAP2 but not by CAP1 at 0.1μM. C) Heat maps of the normalized expression levels of different WNT-TCF targets after treatment of multiple colon cancer cells, melanoma, glioblastoma and lung cancer cells with increasing concentrations of CAP2 as noted. LGR5 levels in CAP2-treated CC36 were undetectable (0*). All values are normalized ratios with DMSO-controls, which were equated to 1. All analyses shown in this figure were performed with the third CAP2 batch. D) Scatter (volcano) plot of the changes in gene expression obtained by microarray of Ls174T cells treated with 5μM CAP2 (blue) or 5μM with Ivermectin (red) for 16h in vitro as 2D cultures. Values are over controls and are per probe with the x-axis giving the fold change in log2 scale and the y-axis the corrected p value in log10. The position of four WNT-TCF targets is highlighted. E) Analyses of WNT-TCF, BMP/TGFß and cytokine/chemokine pathway hits identified for CAP2 and ivermectin from the microarray data (D). Change is given in fold (positive or negative) and only reliable _at or _a_at probes are counted. For probe set ID see panel C in S7 Fig.
Fig 2.
CAP2 and ivermectin regulate a large set of TCF-targets known to be repressed by dnTCF.
A) Venn diagram representation of the overlap of 106 TCF-responsive genes in Ls174T cells [13] with the set repressed by CAP2 (75 genes), ivermectin (50 genes) or both (36 genes). 39 genes are repressed only by CAP2 and not by ivermectin, and 14 by ivermectin and not by CAP2. 17 genes are neither repressed by CAP2 nor ivermectin. B) Color-coded lists of genes in each class shown in (A). C) Heat map of the changes (in fold) observed after CAP2 or ivermectin treatment of Ls174T cells for a core set of 15 genes repressed by dnTCF1/4 in both DLD1 and Ls174T cells from [13]. The upper part is derived from the microarray data and the bottom was directly determined in cells treated as those used for the microarrays by RT-qPCR with specific primers. Whereas both CAP2 and ivermectin repress the entire set each shows gene-specific differences in the extent of the repression. All values are normalized over DMSO-treated controls.
Fig 3.
CAP2 inhibits DLD1 human colon cancer xenograft growth.
A) Graph of tumor growth over time measuring tumor volume in control (cyclodextrin alone) DLD1 xenografts in Nude mice versus those receiving cyclodextrin-conjugated CAP2. The p value is given for the endpoint of the experiment noted by the bracket. n = number of tumors. B) Representative image of subcutaneous DLD1 xenograft tumor (arrows) growth seen in control mice (up) and those receiving CAP2 systemically administered via IP (bottom). C) Histograms of the repression of LGR5 mRNA levels by CAP2 treatment in cells derived from control tumors (left) and those previously treated with CAP2 (right), both after dissociation and in vitro treatment for 16h with different concentrations of CAP2. Note the dose-dependent repression of LGR5 levels in relation to the base of controls (gray bars), which are lower in DMSO-treated cells derived from CAP2 treated tumors in vivo than in control cells from control tumors. The level on DMSO-treated cells from control tumors is set to 1 and the other values (light blue for CAP2-treated cells from cyclodextrin only-treated control tumors or darker blue for CAP2-treated cells from CAP2-treated tumors) are set in reference to this. P values are in relation to the respective DMSO-treated controls (gray bars) unless noted by the bracket that shows the P value between in vivo pretreated with CAP2 and non-pretreated DMSO controls (gray bars).
Fig 4.
Long-lasting repression of stem cell clonogenic growth and of WNT-TCF targets by CAP2.
A) Quantification of 3D clonal spheroids in primary, secondary or tertiary assays from DLD1 (upper panel) or Ls174T (bottom panel) cells previously treated in 2D culture with CAP2 (blue) or ivermectin (red) at 0.5μM (light blue and orange) and 2.5μM (dark blue or red), as compared with control (100%, not shown). Spheroids were not treated with drugs. CAP2 treatment in 2D results in a decrease in primary 3D clonogenic spheroids in both cell types, arguing for a decrease in clonogenic stem cells. Similar results were obtained with ivermectin [6]. However, secondary and tertiary clonal spheroid assays reveal that the repression imposed by CAP2 is long-lasting, in contrast with that imposed by treatment with ivermectin. CAP2 pretreatment of Ls174T cells at 5μM leads to the exhaustion of clonogenic cancer stem cells after three passes. In the case of ivermectin, clonal frequencies recover and become identical to those obtained from DMSO-pretreated control cells, the value of which is here equated to 100 (not shown). p values are in relation to DMSO controls. B) Heat map of showing the repression of normalized AXIN2 and cMYC levels, used here as direct targets of the WNT-TCF pathway, in DLD1 cells treated in vitro and in secondary spheroids derived from these 2D treated cells. Note that CAP2 repression is long-lasting as it is evident in secondary spheroids. In contrast, the repression imposed by treatment with ivermectin is evident in 2D but absent in secondary spheroids.
Fig 5.
Regulation of chromatin components and remodelers by CAP2.
A) Genomic analysis of transcriptomic changes in Ls174T cells imposed by CAP2 but not by ivermectin on an epigenetic gene set revealed large repression or increases in the levels of chromatin remodelers and components. The x-axis denotes the genes (using only reliable _a or _a_at probes) with the associated fold change in relation with DMSO-treated controls, both for CAP2 (blue) and for ivermectin (red) treatments in the y-axis. Changes off scale are written near the respective column. B) Modulation of selected WNT-TCF, BMP, cytokine/chemokine and epigenetic regulator genes by stable dnTCF expression (left) and after 2 or 4h following CAP2 treatment as indicated (right). C) Heat maps of the changes in gene expression after CAP2 treatment (0.5μM batch 4 for 16h) in different human cancer cells as indicated. All values in the heat maps in (B,C) show normalized ratios over DMSO controls. D) Histogram showing CAP2-driven changes in HISTONE 3 modifications. The average of four assays is shown for control DMSO-only treated Ls174T cells and CAP2-treated sibling cells as described in Materials and Methods. The only significative change (p<0.05) was detected in the levels of monomethylated lysine 4.