Reactivating Hippo by drug compounds to suppress gastric cancer and enhance chemotherapy sensitivity

The Hippo signaling pathway plays an essential role in organ size control and tumorigenesis. Loss of Hippo signal and hyper-activation of the downstream oncogenic YAP signaling are commonly observed in various types of cancers. We previously identified STRN3-containing PP2A phosphatase as a negative regulator of MST1/2 kinases (i.e., Hippo) in gastric cancer (GC), opening the possibility of selectively targeting the PP2Aa–STRN3–MST1/2 axis to recover Hippo signaling against cancer. Here, we further discovered 1) disulfiram (DSF), an FDA-approved drug, which can similarly block the binding of STRN3 to PP2A core enzyme and 2) CX-6258 (CX), a chemical inhibitor, that can disrupt the interaction between STRN3 and MST1/2, both allowing reactivation of Hippo activity to inhibit GC. More importantly, we found these two compounds, via an MST1/2 kinase-dependent manner, inhibit DNA repair to sensitize GC towards chemotherapy. In addition, we identified thiram, a structural analog of DSF, can function similarly to inhibit cancer cell proliferation or enhance chemotherapy sensitivity. Interestingly, inclusion of copper ion enhanced such effects of DSF and thiram on GC treatment. Overall, this work demonstrated that pharmacological targeting of the PP2Aa–STRN3–MST1/2 axis by drug compounds can potently recover Hippo signal for tumor treatment.


Supplementary Tables
Primary AlphaScreen results for STRN3-PP2Aa disruption.Table S2.Result of the 2 nd round validation screen for STRN3-PP2Aa disruption.Table S3.Primary AlphaScreen results for STRN3-MST2 disruption.Table S4.Result of the 2 nd round validation screen for STRN3-MST2 disruption.      A. Result of inhibition rate for disrupting the STRN3-MST2 interaction in the 1 st round AlphaScreen.Dash line refers to inhibition efficiency higher than 60% (n=47).B. Result of the 2 nd round AlphaScreen to kick out false positive hits which disrupted interactions between protein and donor/acceptor beads instead of breaking the STRN3-MST2 proteinprotein interaction.Vertical dash line refers to inhibition efficiency higher than 50% while horizontal dash line refers to false positive probability less than 35% (n=10).

Fig. S1 .
Illustration of the AlphaScreen and inhibition of 9 compounds in HGC-27 cells.Fig.S2.DSF binds to PP2Aa to disrupt STRN3-PP2Aa interaction.Fig.S3.DSF reactivates Hippo pathway to inhibit GC growth.Fig.S4.Combined tumor-killing potency of cisplatin and DSF.Fig.S5.TH functions as a structural analog of DSF for GC therapy.Fig.S6.Identification of a STRN3-MST2 disruptor CX for Hippo-based GC therapy.

Fig. S1 .
Fig.S1.Illustration of AlphaScreen and inhibition of 9 compounds in HGC-27 cells A. Graphic illustration of our previous strategy for restoration of Hippo activity for cancer therapy by peptide inhibitor (SHAP) (previous work) or novel chemical compounds via disrupting PP2Aa-STRN3-MST1/2 interaction (this work).B. Illustration of the secondround AlphaScreen exclusion of false positive ones interrupting beads and protein association.C. Cell viability of HGC-27 cells treated with 9 indicated compounds.D. GST pulldown assay to validate the dose-dependent effect of DSF on disrupting the PP2Aa-STRN3 interaction.E. The chemical structure of DSF.

Fig. S2 .
Fig.S2.DSF binds to PP2Aa to disrupt STRN3-PP2Aa interaction.A. Detection of PP2Aa binding with STRN3 by Isothermal titration calorimetry (ITC) assay.B. Detection of DSF binding with PP2Aa by ITC assay.C. Detection the binding ability of DSF with STRN3 by ITC assay.

Fig. S4 .
Fig.S4.Tumor-killing efficiency of combined treatment with cisplatin and DSF. A. Cell viability of several indicated GC cell lines treated with cisplatin.B. Killing efficiencies of MKN-28 cells treated with DSF and cisplatin or in combination.Significance was tested using one-way ANOVA, followed by the Tukey's post-hoc test.***, p<0.001; ****, p<0.0001.

Fig. S5 .
Fig.S5.TH functions as a structural analog of DSF for GC therapy. A. The top 6 chemical structural analogs of DSF with 90% similarity.B. Cell viability assay to assess the 6 DSF structural analogs in GES1 and HGC-27 cells.And their IC50 values were summarized on the right.C. Co-IP analysis of the interaction between PP2Aa and STRN3 in HEK293A cells treated with indicated doses of TH.D. Virtual docking of TH binding to first HEAT repeat of the PP2Aa.PP2Aa is colored in green and shown as cartoon.TH is shown as sticks and colored in yellow.The right panel illustrates the occupied position of TH in STRN3-PP2Aa complex.E. MST assay to show the incapability of PP2Aa single point mutants (V15D or A41D) in binding with TH.F. Volcano plot of altered genes in HGC-27 cells treated with TH.G. Cell viability of HGC-27 cells treated with CuCl2, TH, TH-Copper complex (n = 3).H. Immunoblots of γ-H2AX expression incubated with or without TH.Experimental workflow of TH treatment in HEK293A cells pre-treated with 20 μM etoposide for 2 hr.
C. MBP pulldown assay to assess the disruption effect of STRN3-MST2 association with the indicated 10 compounds treatment.HS: His-Sumo tag.D. IC50 values of the 10 candidate compounds in HGC-27 cells assessed by cell viability assay.E. Cell viability of HGC-27 cells treated with 10 indicated compounds.F. Co-IP analysis of the interaction between PP2Aa and STRN3 in HEK293A cells treated with or without CX.G. Volcano plot of altered genes in HGC-27 cells treated with CX. H. Venn diagram illustration of altered genes in both gene sets treated with DSF and CX.I. Relative mRNA levels of CTGF versus the DMSO control in GES1 and HGC-27 cells treated with CX for 48 hr (n = 3).Data are presented as means ± SD.The data were analyzed using one-way ANOVA, followed by the Tukey's post-hoc test.**, p<0.01; ***, p < 0.01; n.s., no significance.J. Colony number of HGC-27 cells treated with CX corresponding to figure 7H.Significance was tested using one-way ANOVA, followed by the Tukey's post-hoc test.**, p<0.01; ****, p<0.0001.K. Cell viability of HGC-27 cells treated with DSF and CX respectively.L. Immunoblots of γ-H2AX expression incubated with or without CX.Experimental workflow of CX treatment in HEK293A cells pre-treated with 20 μM etoposide for 2 hr.