Hypoxia-inducible factor 2 (cid:2) (HIF-2 (cid:2) ) promotes colon cancer growth by potentiating Yes-associated protein 1 (YAP1) activity

Colorectal cancer (CRC) is the third-leading cause of cancer mortality in the United States and other industrialized coun-tries. A hypoxic microenvironment is a hallmark for solid tumors. The hypoxia-induced signal transduction is transcrip-tionally mediated by hypoxia-inducible factor (HIF). Three major HIF isoforms, HIF-1 (cid:2) , HIF-2 (cid:2) , and HIF-3 (cid:2) , are present in the intestine. Our previous work demonstrates that HIF-2 (cid:2) is essential for CRC growth and progression. However, the mechanisms mediating cell proliferation after hypoxia or HIF-2 (cid:2) activation in CRC are unclear. Data mining of RNA-Seq experiments with mouse models of intestinal HIF-2 (cid:2) or Yes-associ-ated protein 1 (YAP1) overexpression indicates a significant overlap of genes in these conditions. YAP1 is a transcriptional co-activator in the Hippo signaling pathway, and YAP1-induced transcriptional responses are essential in cancer cell proliferation. Here, we report that HIF-2 (cid:2) robustly increases YAP1 expression and activity in CRC-derived cell lines and in mouse models.ThepotentiationofYAP1activitybyHIF-2 (cid:2)

In response to hypoxia, tumor cells activate genes that are critical in angiogenesis, cell survival, cell proliferation, and glucose metabolism (3). Hypoxia-induced responses are transcriptionally mediated by hypoxia-inducible factor-1␣ (HIF-1␣), HIF-2␣, and HIF-3␣, which are members of the Per-ARNT-Sim family of basic helix-loop-helix transcription factors that bind hypoxia response elements at target gene loci under hypoxic conditions (4,5). Among the HIFs, HIF-2␣ is essential in CRC growth and progression. Our recent work has demonstrated an essential role of HIF-2␣-induced epithelial-elicited inflammation and regulation of intratumoral iron homeostasis as important mechanisms leading to increase in colon cancer (6,7). In cancer-derived cell lines and in renal carcinomas, HIF-2␣ induces proliferation via an increase in c-Myc activity (8,9). However, in CRC, mechanisms of hypoxic and HIF-2␣ potentiation of cancer cell growth are unclear.
Hippo signaling is a central pathway that regulates intestinal growth. Yes-associated protein 1 (YAP1) is a downstream effector in the Hippo signaling pathway, which functions as an essential regulator of proliferation, organ size, and cell differentiation (10,11). In response to a decrease in Hippo signaling, YAP1 translocates into the nucleus and acts as transcriptional co-activator to the transcriptional enhancer factors (TEA) domain (TEAD) family of transcription factors (TEAD1-4) (12,13). YAP1 is dispensable for normal intestinal homeostasis but is absolutely required for intestinal regeneration after tissue injury (14). Recent work has demonstrated that the growthpromoting role of YAP1 is critical for the proliferative response observed during intestinal inflammation as well as in sporadic CRC (15,16). Interestingly, intestine-specific YAP1-overexpressing mice restrict ␤-catenin activation, and under radiation-induced injury, deletion of intestinal YAP1 leads to hyperproliferation (17). This discrepancy is perhaps due to the complex interaction of YAP1 with APC/␤-catenin signaling in intestinal epithelial cells (18 -22). However, several studies have demonstrated an important role of YAP1 as an essential transcription factor in mouse models of colon cancer (14,15,23). The role of HIF-2␣ in YAP1 signaling pathway has not been studied.
In the current study we found a novel role of hypoxia via HIF-2␣ in activating YAP1. Moreover, the increase in YAP1 by HIF-2␣ is essential for cell growth during hypoxic stress in colon-derived cell lines. The study provides a novel HIF-2␣/

HIF-2␣ increases YAP1 transcriptional activity
To investigate the mechanisms of HIF-2␣-induced cell proliferation, RNA-sequencing data were analyzed in a mouse model of HIF-2␣ overexpression specifically in intestinal epithelial cells (HIF2␣ LSL/LSL;VillinCre ) (24). Interestingly, the data showed an increased expression of several well-characterized YAP1 target genes: Cyr61, Ctgf, Tgfbr2, and Tead1. To assess the global regulation of YAP1 target genes by HIF-2␣, data were compared from the HIF2␣ LSL/LSL;VillinCre mice and global gene expression profiles from mice, which have a constitutive activated YAP1 in intestinal epithelial cells (17). 484 genes (279 down-regulated and 205 up-regulated genes) were commonly regulated in YAP1-and HIF-2␣-overexpressing mice (Fig. 1, A  and B). This suggests a potential cross-talk between HIF-2␣ and YAP1 signaling. These commonly regulated genes by HIF-2␣ and YAP1 are involved in not only Hippo signaling pathway but also several other pathways including MAPK, PI3K-AKT, and TGF-␤ signaling pathways (Fig. 1, A and B). To confirm these changes Western blot analyses and quantitative PCR (qPCR) were performed in HIF2␣ LSL/LSL;VillinCre mice for YAP1 signaling and target genes. In the HIF2␣ LSL/LSL;VillinCre mice a significant increase in total-YAP1 protein levels and YAP1 target genes, Cyr61, Ctgf, and Tgfbr2 were observed, whereas HIF-1␣ overexpression decreased YAP1 target genes Tgfbr2 and Tead1 ( Fig. 2A). This indicates that HIF-2␣, but not HIF-1␣, increases YAP1 activity. To demonstrate this was a cell autonomous function of HIF-2␣, an oxygen-stable HIF-2␣ was stably overexpressed in HCT116 colon cancer cell line (Fig. 2B). Consistent with in vivo data, the overexpression of HIF-2␣ in HCT116 significantly increased YAP1 target genes Cyr61 and Ctgf expression (Fig. 2C). Western blotting and immunofluorescent staining showed that HIF-2␣ overexpression significantly increased both total YAP1 (T-YAP1) and phospho-YAP1 (p-YAP1) protein levels and nuclear YAP1 localization without altering the p-YAP/T-YAP1 ratio (Fig. 1, E and F). These data confirm that HIF-2␣ increases the expression of YAP1 target genes.

HIF-2␣ potentiates YAP1 activity
The Cyr61 promoter luciferase (25) and a synthetic promoter consisting of 8ϫ tandem TEAD-binding sites (GTIIC) upstream of a luciferase reporter (26) were assessed for YAP1 activity. Consistent with a previous study, YAP1 significantly increased Cyr61 and GTIIC promoter activities in all of the cell lines assessed (Fig. 3, A and B). Interestingly, HIF-2␣ increased basal Cry61 activity in SW480 and HEK293 cell lines (Fig. 3A) and potentiated the YAP1 transcriptional activity on the Cyr61

The cross-talk between HIF-2␣ and YAP1
and GTIIC promoter in the cell lines assessed (Fig. 3, A and B). These data indicated that HIF-2␣ potentiates YAP1 activity.

Increased YAP1 activity in sporadic tumors is HIF-2␣ independent
YAP1 is important for cell proliferation, adenoma formation, and cancer development (27)(28)(29). In addition, previous reports suggested that YAP is part of ␤-catenin signal cascade and is directly activated after the loss of APC (18, 20 -22). To further understand if HIF-2␣ is essential for YAP1 activation after APC disruption, a novel mouse model was generated using a tamoxifen-inducible colon-specific Cre (CDX2ERT2 cre ) (30). The CDX2ERT2 cre mice were crossed to the Apc floxed mice (CDXERT2 Cre /Apc F/F ) or crossed to the Apc and Hif2␣ floxed mice (CDXERT2 Cre Apc F/F /Hif2␣ F/F ). 7-Days after tamoxifen treatment, Hif2␣ and HIF-2␣ target gene Scl11a2 were significantly decreased in Hif2␣ floxed mice (CDXERT2 Cre Apc F/F / Hif2␣ F/F ) compared with CDXERT2 Cre /Apc F/F group (Fig. 4A). Consistent with previous work (31), Cyr61 and CTGF genes were induced in both CDXERT2 Cre Apc F/F and CDXERT2 Cre Apc F/F /Hif2␣ F/F mice compared with littermate controls Apc F/F mice, but there was no significant change in YAP1 target genes in the CDXERT2 Cre Apc F/F /Hif2␣ F/F mice compared with CDXERT2 Cre Apc F/F mice (Fig. 4A). Moreover, the increase of p-YAP1 and T-YAP1 protein after Apc disruption was not altered in mice with a HIF-2␣ disruption (Fig. 4B). Immunofluorescence staining with Ki67 and H&E staining showed increased cell proliferation in CDXERT2 Cre Apc F/F mice, whereas HIF-2␣ knock-out did not alter the tissue proliferation ( Fig. 4, C and D). These results indicate that HIF-2␣ is not essential for the increase of YAP1 and its target gene expression in sporadic CRC models.

HIF-2␣ potentiation of YAP1 activity is not via kinase signaling or protein-protein interaction
Src family kinases can promote proliferation of intestinal epithelial cells through activation of YAP1, which contributes to the proper regulation of intestinal epithelial cell turnover and intestinal homeostasis (32). To assess if HIF-2␣ potentiates YAP1 through Src, YAP1 reporter assays were performed after treatment with the Src inhibitor, SU6656. Cyr61 promoter luciferase was significantly increased when the cells were transfected with both HIF-2␣ and YAP1, but this potentiation was not inhibited when treated with SU6656 ( Fig. 5A, left panel). The Src activity inhibition with SU6656 was confirmed by Western blot analysis (Fig. 5A, right panel). The PI3K-AKT

Prostaglandin E2 (PGE2) is not essential for HIF-2␣ potentiation of YAP1 activity in CRC
A recent study demonstrates that PGE2 and the synthase needed for PGE2 production, cyclooxygenase-2 (COX2), are essential in YAP1 expression and transcriptional activity in CRC (37). The COX2-PGE2 pathway is critical in CRC progression and non-steroidal inflammatory drugs, which are potent and selective inhibitors of COX2, are effective chemopreventative agents in CRC (38,39). PGE2 production is highly elevated in CRC via a HIF-2␣-dependent activation of COX2 (7). To assess if HIF-2␣ leads to potentiation of YAP1 via COX2-PGE2 signaling loop, a colon tumor model dependent on HIF-2␣ activation was assessed (6,7). Mice with an intestine-specific disruption of Von Hippel Lindau (VHL) tumor suppressor protein led to constitutive HIF activation and an increase in HIF1␣ and HIF2␣ target genes Slc11a2, DcytB, Vegf, PDK1, and Ankrd37 ( Fig. 6A) (40). These mice crossed to a sporadic intestinal tumor model Apc min/ϩ leads to a robust increase in colon tumors that is dependent on HIF-2␣ (but not HIF-1␣) (Vhl ⌬IE ;Apc min/ϩ ) (6). Moreover, our previous study demonstrates that intestinal epithelial cell activation of HIF-1␣ does not increase carcinogenesis or progression of colon cancer (41). Vhl ⌬IE ;Apc min/ϩ and littermate control mice (Vhl F/F ;Apc min/ϩ ) were treated with the COX2 inhibitor nimesulide. Tumor tissue from both Vhl F/F ;Apc min/ϩ and Vhl ⌬IE ;Apc min/ϩ mice had significant increases of Cyr61 and Ctgf gene expression compared with normal tissue (Fig. 6B). Nimesulide treatment significantly suppressed YAP1 target gene expression in Vhl F/F ;Apc min/ϩ tumor tissue (Fig. 6B). HIF-2␣ stabilization led to a further increase in Cyr61 and Ctgf, consistent with our data showing that HIF-2␣ can potentiate YAP1 signaling. Nimesulide did not decrease the HIF-2␣ potentiation of YAP1 target genes in tumor tissues (Fig.  6B). This suggests that the cross-talk between HIF-2␣ and YAP1 was PGE2-independent.

YAP1 is important for HIF-2␣-mediated cancer cell growth
Currently the mechanism by which HIF-2␣ leads to YAP1 potentiation is not clear. However, to understand if YAP1 activation is essential in hypoxia or HIF-2␣-induced colon cancer

The cross-talk between HIF-2␣ and YAP1
cell growth, proliferation assays were assessed. RKO, SW480, and KM12 cell lines were incubated in either normoxia (21% O 2 ) or hypoxia (1% O 2 ) conditions. All three cell lines demonstrated a significant increase in HIF-1␣ and HIF-2␣ protein expression in hypoxia (Fig. 7A). YAP1 was shown to be essential in hypoxic growth, as using our previously validated YAP1 siRNAs (siYAP1) (31) (Fig. 7B). To address the specific role of HIF-2␣ and YAP1 in hypoxic growth, HCT116 cells were assessed that expressed both HIF-1␣ and HIF-2␣ (Fig. 7A). HCT116-overexpressing HIF-2␣ or siYAP1 knockdown did not lead to significant changes in growth in normoxia. Under hypoxic conditions (1% O 2 ), cell growth was significantly inhibited, whereas the growth was increased in cells that overexpressed HIF-2␣. Cells that overexpressed HIF-2␣ but had a knockdown of YAP1 demonstrated a decrease in cell growth under hypoxia (Fig. 7C). Interestingly, knockdown of YAP1 not only reversed the growth-protective effect of HIF-2␣ under hypoxia but also led to a decrease in growth under hypoxia. This result demonstrates that YAP1 is essential for HIF-2␣ and hypoxia-mediated growth.

Discussion
Hypoxia is a hallmark of solid tumors. Hypoxia can activate the expression of numerous genes involved in cell metabolism, cell survival, cell proliferation, and cell apoptosis. In the intestine, activation of HIF-2␣ but not HIF-1␣ promotes the development of CRC (6). However, how HIF-2␣ contributes to proliferation is still not clear. Here, we demonstrate that HIF-2␣ promotes CRC cell growth via regulating YAP1 activity. The hippo pathway can be activated by cell density or mechanical cues (42,43), which leads to phosphorylation and degradation of the YAP/TAZ transcription factors (11). A decrease in Hippo signaling leads to YAP and TAZ nuclear localization and activation of gene expression that promotes cell proliferation and survival. The persistent activation of YAP1 leads to tissue overgrowth and tumor formation in a broad range of tissues (27).

The cross-talk between HIF-2␣ and YAP1
apoptosis (49,50), which is consistent with the decreased cell proliferation in our RKO, SW480, KM12, and HCT116 cells under hypoxia conditions. Here, we showed that HIF-2␣ is an activator for YAP1, which increases colon cancer cell growth. Our data showed that stabilized HIF-2␣ significantly increased YAP1 protein levels and downstream target gene Cyr61, Ctgf, and Tgfbr2 expression in Hif2␣ LSL/LSL;VillinCre mice as well as in the HIF-2␣-overexpressing HCT116 human colon cancer cell line. Indeed, HIF-2␣-induced cancer cell growth was inhibited when YAP1 was knocked down in various colon cancer cell lines, indicating that HIF-2␣-YAP1 cross-talk is important for tumorigenesis. Future work will focus on how the different

The cross-talk between HIF-2␣ and YAP1
HIF-2␣-induced growth proliferative mechanisms integrate and are regulated to increase CRC progression.
The mechanism by which HIF-2␣ induced YAP1 expression and activity was not through direct protein binding or promoter activation of YAP1 target genes but via indirect mechanism. The possible mechanisms that led to potentiation are not clear but are a major future focus. In our sporadic colon cancer mouse model, we demonstrated that the activation of YAP1 is HIF-2␣-independent. This is possibly due to other parallel pathways such as increased ␤-catenin resulting from tumor suppressor APC mutation. The dysregulation of YAP phosphorylation and nuclear localization has been extensively studied (51)(52)(53). However, recent data have revealed that cancer samples have increased total YAP abundance (23,54). Our data support this recently proposed idea and provide a clinically relevant mechanism by which YAP activation by HIF-2␣ is increased in human colon cancers.
Previous studies have demonstrated that the activation of YAP1 leads to hyper-proliferation and spontaneous tumor formation in certain tissues, including the intestine (23,55). Recently, YAP-TEAD complex suppressor Verteporfin has shown potential as a anticancer treatment (31,56). However, YAP1 is a critical component of the ␤-catenin degradation complex, and disruption of YAP1 leads to an enhanced ␤-catenin activation and increased proliferation in the intestine (17). Therefore, better therapeutic methods are necessary. Hypoxia can regulate YAP1 activity in hepatocellular cancerderived cell lines independent of HIF-1␣; however, HIF-2␣ was not assessed (57). Moreover, a complimentary pathway was recently shown in which HIF-1␣ induces the Hippo coactivator TAZ and can serve as transcriptional co-activator (58). Here, we demonstrate a novel mechanism of YAP1 regulation by HIF-2␣, which can serve as a potential therapeutic target for YAP1associated tumorigenesis.
In summary, we discovered that there is novel cross-talk between HIF-2␣ and YAP1, which is one of the mechanisms that HIF-2␣ can increase tumor genesis. Future studies focusing on YAP1 deletion in HIF-2␣-induced cancer models are needed to further understand this cross-talk.

The cross-talk between HIF-2␣ and YAP1
were fed with standard chow and water ad libitum. For the PGE2/COX2 inhibition study, normal tissue and tumor tissue samples were obtained from Vhl ⌬IE /Apc min/ϩ and Vhl F/F / Apc min/ϩ mice fed with powdered laboratory rodent diet 5001 (PMI Nutrition International LLC, Brentwood, MO) with or without 400 mg/kg nimesulide (Sigma) for 8 weeks. All the ani-

The cross-talk between HIF-2␣ and YAP1
mal studies were carried out in accordance with the Institute of Laboratory Animal Resources guidelines and approved by the University Committee on the Use and Care of Animals at the University of Michigan.

Quantitative real-time RT-PCR
RNA extraction, reverse transcription, and qPCR were described previously (61). The primers used in the study are listed in Table 1. Quantification cycle (Cq) values were normalized to ␤-actin and expressed as -fold difference from controls.

Statistical analysis
Results are expressed as the mean Ϯ S.D. Significance between two groups was calculated by an independent t test, and significance among different groups were tested using one-way analysis of variance followed by Dunnett's post-hoc comparisons.
Author contributions-X. M., H. Z., X. X., and Y. M. S. contributed to the conception and design, acquisition of data, and analysis and interpretation of data. X. M. and Y. M. S. drafted the article and revised it critically for important intellectual content. X. M., H. Z., X. X., and Y. M. S. gave final approval of the version to be published. X. M., H. Z., X. X., and Y. M. S. agreed to be accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. The cross-talk between HIF-2␣ and YAP1