Functional significance of Hippo/YAP signaling for drug resistance in colorectal cancer

Colorectal cancer is a leading cause of cancer‐related death worldwide. While early stage colorectal cancer can be removed by surgery, patients with advanced disease are treated by chemotherapy, with 5‐Fluorouracil (5‐FU) as a main ingredient. However, most patients with advanced colorectal cancer eventually succumb to the disease despite some responded initially. Thus, identifying molecular mechanisms responsible for drug resistance will help design novel strategies to treat colorectal cancer. In this study, we analyzed an acquired 5‐FU resistant cell line, LoVo‐R, and determined that elevated expression of YAP target genes is a major alteration in the 5‐FU resistant cells. Hippo/YAP signaling, a pathway essential for cell polarity, is an important regulator for tissue homeostasis, organ size, and stem cells. We demonstrated that knockdown of YAP1 sensitized LoVo‐R cells to 5‐FU treatment in cultured cells and in mice. The relevance of our studies to colorectal cancer patients is reflected by our discovery that high expression of YAP target genes in the tumor was associated with an increased risk of cancer relapse and poor survival in a larger cohort of colorectal cancer patients who underwent 5‐FU‐related chemotherapy. Taken together, we demonstrate a critical role of YAP signaling for drug resistance in colorectal cancer.

Colorectal cancer is a leading cause of cancer-related death worldwide. While early stage colorectal cancer can be removed by surgery, patients with advanced disease are treated by chemotherapy, with 5-Fluorouracil (5-FU) as a main ingredient. However, most patients with advanced colorectal cancer eventually succumb to the disease despite some responded initially. Thus, identifying molecular mechanisms responsible for drug resistance will help design novel strategies to treat colorectal cancer. In this study, we analyzed an acquired 5-FU resistant cell line, LoVo-R, and determined that elevated expression of YAP target genes is a major alteration in the 5-FU resistant cells.
Hippo/YAP signaling, a pathway essential for cell polarity, is an important regulator for tissue homeostasis, organ size, and stem cells. We demonstrated that knockdown of YAP1 sensitized LoVo-R cells to 5-FU treatment in cultured cells and in mice. The relevance of our studies to colorectal cancer patients is reflected by our discovery that high expression of YAP target genes in the tumor was associated with an increased risk of cancer relapse and poor survival in a larger cohort of colorectal cancer patients who underwent 5-FU-related chemotherapy. Taken together, we demonstrate a critical role of YAP signaling for drug resistance in colorectal cancer. Chemotherapy, particularly 5-Fluorouracil (5-FU)-based adjuvant chemotherapy, has been widely used to treat CRC since early 1990s.
To elucidate the underlying mechanism for 5-FU resistance in CRC, we used an acquired resistant cell line, LoVo-R, to identify gene expression profiles in comparison with that of the parental LoVo cells.
We discovered elevated expression of YAP1 target genes in LoVo-R cells. We demonstrated both in cultured cells and in mice that YAP1 is significant for 5-FU resistance. We also investigate the relevance of our results to colorectal cancer patients by examining a large cohort of patients with colorectal cancer who underwent 5-FU-based chemotherapy (n = 633) for the rate of cancer relapse and patient survival in patients with high and low expression of YAP1 target genes. protein levels by Western blotting with available specific antibodies ( Figure 1C).
The Hippo/YAP signaling pathway involves a kinase cascade including Mst1/2 and Lats1/2. Lats1/2 phosphorylate YAP1, leading to degradation and inactivation of YAP1. 30,31 In the active form, YAP1 is a co-activator to promote DNA binding of TEAD1-4, causing elevated expression of YAP1 target genes. To confirm the data from target gene expression, we detect protein levels of YAP1 in 5-FU resistant LoVo-R and the parental LoVo cells. Nuclear YAP1 is an indicator for activated YAP1 signaling, and we detected YAP1 protein level by two approaches: immunofluorescent staining, cell fractionation followed by Western blotting. By immunofluorescent staining, we detected more nuclear staining of YAP1 in the 5-FU resistant LoVo-R cells, but not in the parental LoVo cells (Figure 2A). By cell fractionation analysis, we detected a high ratio of nuclear YAP1/Lamin C in 5-FU resistant LoVo-R cells ( Figure 2B). These results indicate that more YAP1 protein is accumulated in the nucleus in the 5-FU resistant cells, suggesting elevated YAP signaling in the 5-FU resistant cells.

| Molecular mechanisms by which YAP signaling is elevated in 5-FU resistant LoVo-R cells
To determine the molecular mechanisms responsible for elevated Although RTKs and EMT regulate YAP1 signaling, knocking down YAP1 seems to be more effective in suppression of YAP1 signaling because up-regulation of YAP1 and TEAD2 is one important mechanism by which YAP1 target genes are induced. We knocked down YAP1 by specific shRNAs in LoVo-R cells. We examined YAP1 target genes expression. As shown in Figure 4A, expression of ANKRD1 and CTGF were significantly suppressed whereas NRG1 and TEAD2 were not affected.
We further tested the IC50 of LoVo-R shYAP cells ( Figure 4A). The IC50 value of LoVo-R shYAP is around 0.4 mM which is lower than 13.803 mM, the value observed for LoVo-R control cells ( Figure 4B).
Next, we determined whether 5-FU sensitivity is altered by YAP1 knocking down in LoVo formed tumors in mice. After injection of LoVo-R-shYAP or the LoVo-R control cells into immune deficient NSG mice and tumor formation, we treated mice with 5-FU (50 mg/kg) once a week. We measured the tumor size at different time points. As shown in Figure 5, we found that the tumors derived from the LoVo-R control cells were not significantly different from the tumors from 5-FU treated mice ( Figure 5A). In contrast, tumors from LoVo-R-shYAP1 FIGURE 2 Nuclear localization of YAP1 protein. We detected YAP1 protein in LoVo and LoVo-R cells by immunofluorescent staining and cell fractionation analysis. A, Shows typical YAP1 staining. Please note the different cell morphology between LoVo-R cells and LoVo cells. YAP1 was indicated by red staining, counter stained with DAPI. B, Shows detection of YAP1 in nuclear fractionated proteins by Western blotting, with LaminC as the internal control. Left is the Western blotting results and the right is the quantitative analysis of the nuclear YAP1/LaminC ratios. Cells treated with trypsin or without trypsin treatment had no significant difference. Significant difference was indicated by *** (P < 0.0005)  Through RNA-seq analyses, we discovered elevated YAP1 target gene expression as the major alteration in the 5-FU resistant cells. By knocking down YAP1, we demonstrated that YAP1 is required for 5-FU resistance. We also showed the relevance of our results to human colorectal cancer patients through gene expression association studies. We showed that patients with high YAP1 target gene Previous studies have linked YAP1 signaling to colorectal cancer development. In one study, a tyrosine kinase YES1 was shown to be up-regulated in the 5-FU resistant cells, and YES1 regulates drug resistance through regulation of YAP1. 36 In other studies, YAP1 signaling is shown to be associated with cell proliferation in colorectal cancer cells as well as metastasis of colon cancer. 31 In our studies, we itself. It thus appears that there are different mechanisms by which YAP1 signaling is activated in drug resistance in colorectal cancer.
Using a large cohort of colorectal cancer patients (n = 633), we were able to show that high expression of YAP1, TEAD2, and YAP1 target genes CYR61 and ANKRD1 is associated with a high risk of cancer relapse (>33% vs ∼10%) and poor survival (Figure 6), suggesting that our findings are relevant to colorectal cancer patients. As complicated as drug resistance, it is likely that multiple mechanisms are involved in regulation of 5-FU resistance in colorectal cancer.
It remains unclear how YAP1 signaling regulates 5-FU drug resistance. It is known that YAP1 signaling is an important pathway for regulation of tissue polarity. Like other cell polarity regulators, including hedgehog and Wnt signaling, YAP1 is known to regulate residual cancer cells or cancer stem cells. Efficiency of tumor sphere formation is one feasible biology assay for cancer stem cells. However, LoVo-R or its parental LoVo cells do not form tumor spheres, which prevents further analyses. Additional 5-FU resistant cell lines with activated YAP1 signaling will be sought to further the mechanism study. 5-FU is widely used in cancer treatment for chemotherapy or neoadjuvant chemotherapy. In addition to colorectal cancer, 5-FU has FIGURE 4 The effect of YAP1 knockdown on YAP1 target genes expression and on 5-FU response in culture. A, Real-time PCR detection of gene expression. B, The effect of YAP1-ShRNAs (shown as shYAP) on the IC50 of 5-FU. Significant difference was indicated by * (P < 0.05); ** (P < 0.005); or *** (P < 0.0005) been used for treatment with pancreatic cancer, head and neck cancers, skin and pediatric cancers. We predict that similar mechanisms may also exist for 5-FU resistance in other cancer types. In addition to be a drug target, elevated expression of YAP1 and its target genes may be also used to predict cancer relapse in chemotherapy.
Proteins were separated by SDS-PAGE, and then transferred onto

| Immunofluorescence staining
To test the expression of YAP between LoVo cells and LoVo-R cells.
Cells were seeded in six-well plates with coverslips at the bottom when cell confluence reaches 50-80%, cells were fixed with 4% of PFA/PBS for 15 min at room temperature. There were two groups: one group with the first antibody (anti-YAP antibody 1:400 dilution) and the second antibody (1:300); another group with the second antibody (1:300) only. DAPI was used for nuclear staining.

| Statistical analysis
All the data are presented as mean ± SD and from three independent experiments. IC50 values of 5-FU were calculated with the GraphPad Prism 7.0 software. Student's test was done for statistical analysis with two groups and P-value was calculated by two-tail unpaired t-test.
Bands of Western blotting and pictures of immunofluorescence staining were quantified by ImageJ software. Analyses of a large cohort of colorectal cancer patients was through a public CTGA database (Cbioportal) with Kaplan-Meier survival generated automatically.