HIF-1α activates hypoxia-induced BCL-9 expression in human colorectal cancer cells

B-cell CLL/lymphoma 9 protein (BCL-9), a multi-functional co-factor in Wnt signaling, induced carcinogenesis as well as promoting tumor progression, metastasis and chemo-resistance in colorectal cancer (CRC). However, the mechanisms for increased BCL-9 expression in CRC were not well understood. Here, we report that hypoxia, a hallmark of solid tumors, induced BCL-9 mRNA expression in human CRC cells. Analysis of BCL-9 promoter revealed two functional hypoxia-responsive elements (HRE-B and HRE-C) that can be specifically bound with and be transactivated by hypoxia inducible factors (HIF) -1α but not HIF-2α. Consistently, ectopic expression of HIF-1α but not HIF-2α transcriptionally induced BCL-9 expression levels in cells. Knockdown of endogenous HIF-1α but not HIF-2α by siRNA largely abolished the induction of HIF by hypoxia. Furthermore, there was a strong association of HIF-1α expression with BCL-9 expression in human CRC specimens. In summary, results from this study demonstrated that hypoxia induced BCL-9 expression in human CRC cells mainly through HIF-1α, which could be an important underlying mechanism for increased BCL-9 expression in CRC.


INTRODUCTION
Colorectal cancer (CRC) is the third most common cancer and the fourth most major cause of malignancies death, accounting for ~1.2 million new cases and ~600,000 deaths annually worldwide [1]. Several significant efforts have been devoted to identify the etiology of CRC; however, the accurate mechanism of CRC carcinogenesis remains unknown.
Wnt signaling pathway plays critical roles in gene functions during normal to malignant development and transcriptional programs of cell self-renewal [2]. Abnormal activation of Wnt signaling pathway is reported to be associated with a variety of human cancers such as CRC, breast cancer and prostate cancer [3][4][5][6][7]. The great majority (> 90%) of CRC carries mutation of at least one gene in Wnt signaling pathway [8]. β-catenin as a key effector phosphorylates and targets for degradation by Axin complex, which is inhibited in respond to Wnt stimulation. This process flow allows unphosphorylated β-catenin to accumulate and bind with the nuclear TCF/ LEF to stimulate the transcription of Wnt target gene, resulting in recruiting chromatin modifying and remodeling a range of transcriptional complex and co-factors [9][10][11]. Well-established transcriptional complexes and co-factors involving B-cell CLL/lymphoma 9 protein (BCL-9), Pygopus (PYGO) and polymerase-associated factor 1 (PAF1) have been reported to be recruited by β-catenin [9].
BCL-9, one of the crucial counterparts of the nuclear component Legless (Lgs) in Wnt pathway, is highly expressed in human CRC tissues, and the BCL-9/β-catenin complex has been identified as a crucial target for cancer

Research Paper
Oncotarget 25886 www.impactjournals.com/oncotarget therapy [12]. The molecular role of BCL-9 is currently debated. One function of Lgs/BCL-9 proteins could recruit transcriptional co-factors synergized with β-catenin and its co-factors during the transcription of TCF target genes. Likewise, they could capture nuclear β-catenin to facilitate its recruitment to TCF target genes [9,13]. However, there are very little studies working on the mechanisms of regulation on BCL-9 and its co-factors related functional. In this study, we focused on the mechanism of regulation on BCL-9 via hypoxia signaling pathway, and hypothesized that BCL-9 in both Wnt and hypoxia signaling plays a vital role on the carcinogenesis of human CRC.

Association between BCL-9 expression and clinicopathological parameters in CRC patients
Expression of BCL-9 showed nuclear staining scored as −, +, ++ in cases examined. Different staining levels of BCL-9 expressed in the cytoplasm of epithelial cells were show in Figure 1. The positive expression of BCL-9 was observed in all stages of CRC samples (13.9% in stage I, 10 of 72 cases; 47.4% in stage II, 63 of 133 cases; 55.8% in stage III, 43 of 77 cases; and 100% in stage IV, 2 of 2 cases).
As our related work described previously, specimen selection for 284 cases were included. Demographic clinicopathological characteristics including gender, age, invasive depth, lymph node metastasis, distant metastasis, AJCC TNM stage and tumor pathological type were summarized in Table 1. BCL-9 expression was associated with AJCC TNM stage (p = 0.000), invasive depth (p = 0.000) and lymph node metastasis (p = 0.000). No significant association was observed in BCL-9 expression compared with distant metastasis and tumor pathological type (both, p > 0.05).

Hypoxia induces BCL-9 expression in human CRC cells
We have shown that a high percentage of human colorectal cancer specimens display elevated BCL-9 expression levels, but the underlying mechanism for the increased expression of BCL-9 in CRC is unclear. So we introduce the microenvironment Hypoxia, a hallmark of solid tumors including CRC. Interestingly, we found that BCL-9 expression was induced by hypoxia. Human CRC cell lines SW480 and HCT116 cells were treated by hypoxia (0.1% O 2 ). The effective induction of hypoxia in cells was confirmed by the increased expression of VEGF, a hypoxia-responsive gene, at mRNA levels and the increased protein levels of HIF-1α and HIF-2α, the major hypoxia inducible factors. Notably, BCL-9 mRNA levels were significantly increased (p < 0.01) in both SW480 and HCT116 cells cultured under the hypoxic condition as determined by real-time PCR (Figure 2A-2B). The induction of BCL-9 expression by hypoxia was confirmed at the protein levels in these cells by using Western-blot assay ( Figure 2C). These data clearly demonstrated that hypoxia induces BCL-9 expression in human colorectal cancer cells.
Hypoxia transactivates the BCL-9 promoter region containing hypoxia-responsive elements (HREs) The transcriptional response to hypoxia in cells is largely mediated by HIFs. HIF-1α and HIF-2α induce a number of common genes, and at the same time they each have their unique target genes. To investigate whether the transcriptional induction of BCL-9 by hypoxia is activated by HIFs, we searched for the HRE consensus sequence in the promoter region of the BCL-9 gene from 3 kb upstream of transcriptional start site. Three putative HRE sites (HRE-A, HRE-B and HRE-C) were identified in the promoter region of the BCL-9 gene ( Figure 3A). To investigate whether these putative HREs account for the hypoxia-mediated induction of BCL-9, the DNA fragments containing one copy of these putative HRE sites were inserted into a pGL2 luciferase reporter plasmid. VEGF is a well-known hypoxia-inducible gene. HIFs physically bind with the HRE in the VEGF promoter and induce VEGF expression levels under hypoxia. The DNA fragment containing the HRE in the VEGF promoter was inserted into a pGL2 luciferase reporter plasmid to serve as a positive control. SW480 and HCT116 cells were transiently transfected with these reporter constructs, and then exposed to hypoxia for 36 h. TK100 plasmids was co-transfected as an internal standard to normalize transfection efficiency. As shown in Figure 3B, hypoxia clearly enhanced luciferase expression levels of the reporter plasmids containing the HRE from the VEGF promoter, the HRE-B and HRE-C from the BCL-9 promoter, but did not have a clear effect on HRE-A form the BCL-9 promoter in both cells. These results indicate that hypoxia transactivates the BCL-9 promoter containing functional HREs (HRE-B and HRE-C).

HIF-1α instead of HIF-2α binds to HRE-B and HRE-C in the BCL-9 promoter under the hypoxic condition
To further investigate which HIF (HIF-1α or HIF-2α) activates the hypoxia-induced BCL-9 expression, chromatin immunoprecipitation (ChIP) assays were employed to determine whether HIF-1α and HIF-2α physically bind to HREs, especially HRE-B and HRE-C in the BCL-9 promoter. Both SW480 and HCT116 cells were cultured under the normoxic or hypoxic conditions for 36 h, and the ChIP assays were performed using an antibody against HIF-1α or HIF-2α. The degree of pull-down of chromatin www.impactjournals.com/oncotarget fragments was determined by quantitative real-time PCR. As shown in Figure 3C, both HIF-1α and HIF-2α bound with the chromatin fragments containing the HRE from the VEGF promoter under the hypoxic but not normoxic conditions. However, no clear immunoprecipitation of the chromatin fragments containing HRE-A, HRE-B, and HRE-C by the antibody against HIF-2α was observed in both SW480 and HCT116 cells under either the hypoxic or normoxic conditions ( Figure 3C). Interestingly, the chromatin fragments containing HRE-B and HRE-C but not HRE-A were pulled down by the antibody against HIF-1α in both SW480 and HCT116 cells under the hypoxic condition ( Figure 3C). Furthermore, these chromatin fragments were not co-immunoprecipitated with the HIF-1α antibody in cells under the normoxic condition ( Figure 3C). The specific pull-down of chromatin fragment containing HRE-B and HRE-C but not HRE-A by the HIF-1α antibody under the hypoxic condition was also confirmed by conventional PCR followed by agarose gel electrophoresis ( Figure 3D). These results demonstrated that HIF-1α but not HIF-2α binds with HRE-B and HRE-C in the BCL-9 promoter under the hypoxic condition, which may stimulate the hypoxiainduced BCL-9 expression.

HIF-1α transcriptionally stimulates BCL-9 expression
To directly investigate whether HIF-1α but not HIF-2α transcriptionally induces BCL-9 expression levels, SW480 and HCT116 cells were transfected with the plasmids expressing HIF-1α (pcDNA3-Flag-HIF-1α) and HIF-2α (pcDNA3-Flag-HIF-2α) respectively. The ectopic expression of either HIF-1α or HIF-2α clearly induced VEGF mRNA expression ( Figure 4A and 4B). Ectopic HIF-1α expression clearly increased BCL-9 expression at both mRNA and protein levels as determined by Real-time PCR and Western-blot assays respectively ( Figure 4A and 4C). However, ectopic HIF-2α expression had no apparent effect on BCL-9 expression ( Figure 4B). Furthermore, the degree of pull-down of chromatin fragments containing HREs in the BCL-9 promoter by HIF-1α was determined by ChIP assays in HCT116 cells with ectopic HIF-1α expression. As shown in Figure 4D-4E, chromatin fragments containing HRE-B and HRE-C but not HRE-A was co-immunoprecipitated with the Flag antibody in cells with ectopic HIF-1α expression as determined by both realtime PCR and conventional PCR assays. These observations clearly showed that HIF-1α bound with HRE-B and HRE-C in the BCL-9 promoter, which is consistent with the ChIP results obtained in cells under the hypoxic condition. We further investigated whether HIF-1α directly transactivates HREs in the BCL-9 promoter. HCT116 cells were cotransfected with an expression plasmid of HIF-1α and the pGL2 reporter plasmids containing HREs in the BCL-9 promoter or the HRE in the VEGF promoter along with TK100 plasmids. Ectopic HIF-1α expression increased the luciferase activity of the reporter plasmids containing the HRE in the VEGF promoter ( Figure 4F). Notably, ectopic HIF-1α expression clearly transactivated the reporter plasmids containing HRE-B and HRE-C from the BCL-9 promoter, but had no apparent effect on the reporter plasmids containing HRE-A from the BCL-9 promoter ( Figure 4F). Collectively, these results clearly showed that HIF-1α binds with and transactivates HRE-B and HRE-C in the BCL-9 promoter to directly induce BCL-9 expression in cells.

HIF-1α activates the induction of BCL-9 expression by hypoxia
To investigate whether HIF-1α but not HIF-2α activates the induction of BCL-9 expression by hypoxia, the effect of hypoxia on BCL-9 expression levels was determined in HCT116 cells with knock-down of endogenous HIF-1α or HIF-2α by siRNA. As shown in Figure 5A and 5B, knock-down of endogenous HIF-1α largely abolished hypoxia-induced BCL-9 expression in HCT116 cells as determined at both mRNA and protein levels by Real-time PCR and Western-blot assays respectively. In contrary, knockdown of endogenous HIF-2α by siRNA had no apparent effect on hypoxiainduced BCL-9 expression ( Figure 5A and 5B). It has been shown that VEGF is a common target of HIF-1α and HIF-2α. Indeed, knockdown of either HIF-1α or HIF-2α can all inhibit hypoxia-induced VEGF expression ( Figure 5A). The effective knockdown of HIF-1α and HIF-2α was confirmed at the mRNA level by Real-time PCR assays ( Figure 5C). These results clearly demonstrated that hypoxia induces BCL-9 expression mainly through HIF-1α.

HIF-1α expression is associated with BCL-9 expression in human CRC specimens
To investigate whether hypoxia and HIF-1α contribute to the increased expression of BCL-9 in human CRC samples, the expression levels of BCL-9 and HIF-1α were determined by the double immunofluorescence (IF) co-localization and stained on slides comprising 244 human CRC samples. The representative IF images of BCL-9 and HIF-1α were shown in Figure 6A. A significant portion of CRC samples (54%; 132 out of 244 cases) showed positive BCL-9 staining. There is also a significant portion of CRC samples (58%; 142 out of 244 cases) showed positive HIF-1α staining. Notably, HIF-1α expression is strongly associated with BCL-9 expression ( Figure 6B) (p < 0.000, Pearson chi-square test). HIF-1α positive staining was observed in 72% of cases with BCL-9 positive staining (95 of 132 cases), but in 42% of cases with BCL-9 negative staining (47 of 112 cases). These results strongly suggested that the transcriptional induction of BCL-9 by hypoxia and HIF-1α is an important mechanism accounting for the BCL-9 expression in human CRC.

DISCUSSION
Hypoxia-signaling pathway as a classic hallmark of solid tumor was widely concerned. Due to restricted blood flow, tumor cells experienced hypoxic condition that inhibited cell proliferation and altered energy metabolism from oxidative phosphorylation to glycolysis pathway, resulting in modifications of hypoxia-inducible genes expressions [14,15]. These are mediated by the hypoxia inducible factors (HIFs), which contained three HIF-α  Oncotarget 25892 www.impactjournals.com/oncotarget (HIF-1α, HIF-2α and HIF-3α) and two HIF-β (HIF-1β and HIF-2β) [16]. Compared to the Wnt signaling, the hypoxia signaling does not require secreted or transmembranous proteins but does sense oxygen concentrations instead. Once activated under hypoxia, HIF-α are facilitated and dimerized with HIF-1 β; this complex binds to HRE and triggers transcription of target genes such as vascular endothelial growth factor (VEGF), leading to stimulation of carcinogenesis [17,18]. Since there was a crosstalk with different signaling pathways like hypoxia, Hedgehog and Wnt pathway, it was suggested that carcinogenesis is often dependent on at least two pathways [24]. In our study, one of the first evidence for crosstalk between Wnt and hypoxia signaling emerged when hypoxia leaded to upregulation of the key Wnt co-activatior BCL-9 in HIF-1α/2α-correlated manners, suggesting that hypoxia pathway might promote HIF-1α and HIF-2α are often overexpressed in solid tumors and tumor-derived cell lines [19]. Although sometimes both these proteins share similar properties mainly via transcriptional regulation of common target genes, they each target own genes contributing to specific characteristics [20,21]. In addition, the different selection of HIF-1α and HIF-2α target genes can be contributed by interacting with different transcription factors and/or chromatin context [22,23]. Base on this characteristic of HIFs, in our study, we found that BCL-9 promoter contains Oncotarget 25894 www.impactjournals.com/oncotarget 3 putative HRE sites including HRE-A, HRE-B and HRE-C. Meanwhile ChIP assays showed that HIF-1α but not HIF-2α bound with HRE-B and HRE-C but not HRE-A in the BCL-9 promoter under hypoxic condition. Moreover, knockdown of HIF-1α but not HIF-2α largely abolished hypoxia-induced BCL-9 expression. These results clearly demonstrate that BCL-9 is a target gene for HIF-1α but not HIF-2α. Intriguingly, transfection and knockdown of either HIF-1α or HIF-2α can respectively promote and inhibit hypoxia-induced VEGF expression, suggesting that VEGF is a common identical target gene for hypoxia signaling pathway.
In summary, this study demonstrated that BCL-9 expression could be induced by hypoxia in human CRC cells; BCL-9 induction under hypoxic condition is predominantly activated by HIF-1α. Given the critical role of BCL-9 in Wnt signaling pathway promoting carcinogenesis and chemo-resistance, the specific regulation of BCL-9 expression by HIF-1α is considered to be a underlying crosstalk between Wnt signaling and hypoxia signaling pathway, providing foundation for further study to develop new target of potential therapeutic strategies by blocking HIF-1α/BCL-9 signaling related expression in CRC.

Patients and samples
A total of CRC specimens were obtained from patients who underwent colectomy at the Department of Surgery, Changzheng Hospital, Second Millitary Medical University, Shanghai, China, from January 2012 to October 2015. None of the patients in our study received neoadjuvant chemotherapy. Information on patient gender, age, invasive depth, lymph node metastasis, distant metastasis, AJCC TNM stage and tumor pathological type were summarized in Table 1. The tumor pathological types of the samples were assessed by the experienced colorectal pathologists. Informed consent forms were signed from all patients and the research protocol was approved by the Ethics Committee.

IHC and IFC staining assays
Tumor tissues or other samples were fixed with 4% paraformaldehyde and were then dehydrated through a graded series of ethanol, embedded in paraffin and sectioned at 5 μm. Immunohistochemical staining for BCL-9 was carried out using standard histological procedures described in the manual for Histostain-Plus IHC Kit, DAB (Phygene). BCL-9 staining was scored in accordance by two investigators with previous protocols as negative (−), weak positive (+), or strong (++). Scoring was performed blindly with respect to the histologic grade of CRC specimens.
Immunofluorescence staining was performed with primary and secondary antibodies diluted in 10% BSA and nucleus was stained by DAPI (Sigma). All fluorescent secondary antibodies were used at a dilution of 1:200 for 30 min (invitrogen). Quantification of IF was performed using NIH ImageJ.

Real-time PCR
Total RNA was isolated by using RNAiso Plus (TAKARA) following the manufacturer's instruction. RNA was reverse transcribed into cDNA by using the Taqman Reverse Transcription Reagents kit (Applied Biosystems) with random hexamers. Human BCL-9, VEGF, HIF-1α, HIF-2α and actin mRNA levels were determined in 7900HT Fast Real-time PCR system (Life Technology Corporation, USA). All primers were purchased from Applied Biosystems. Real-time PCR was done in triplicate with TaqMan PCR mixture (Applied Biosystems). The expression of genes was normalized to the actin gene.