FHL family members suppress vascular ... - Wiley Online Library

Download PDF
2 downloads 25 Views 776KB Size Report
Comment
negatively regulated by the LIM-only protein FHL2 in cardiomyocytes. Mol. Cell Biol. 24, 1081–1095. 27. McGrath, M. J., Cottle, D. L., Nguyen, M. A., Dyson, J. M., ...
IUBMB

Life, 64(11): 921–930, November 2012

Research Communication FHL Family Members Suppress Vascular Endothelial Growth Factor Expression Through Blockade of Dimerization of HIF1a and HIF1b Jing Lin1,2*, Xi Qin1*, Ziman Zhu3*, Jinsong Mu4*, Lingling Zhu5, Kuiwu Wu5, Huabo Jiao3, Xiaojie Xu1, and Qinong Ye1 1

Department of Medical Molecular Biology, Beijing Institute of Biotechnology, Beijing, People’s Republic of China Department of Clinical Laboratory, The First Affiliated Hospital, Chinese PLA General Hospital, Beijing, People’s Republic of China 3 Department of Hepatobiliary Surgery, The First Affiliated Hospital, Chinese PLA General Hospital, Beijing, People’s Republic of China 4 Department of Intensive Care Center, 302 Military Hospital, Beijing, People’s Republic of China 5 Department of Brain Protection and Plasticity, Institute of Basic Medical Sciences, Beijing, People’s Republic of China 2

Keywords Summary Four and a half LIM domain (FHL) proteins belong to a family of LIM-only proteins that have been implicated in the development and progression of various types of cancers. However, the role of FHL proteins in tumor angiogenesis remains to be elucidated. Herein, we demonstrate that FHL1-3 decrease the promoter activity and expression of vascular endothelial growth factor (VEGF), the key regulator of angiogenesis in cancer growth and progression as well as an important target gene of the transcription factor hypoxia-inducible factor 1 (HIF1a/ HIF1b). FHL1-3 interacted with HIF1a both in vitro and in vivo. A single LIM domain of FHL1 was sufficient for its interaction with HIF1a. FHL1 interacted with the HIF1a region containing basic helix-loop-helix (bHLH) motif and PER-ARNT-SIM domain, both of which aid in dimerization with HIF1b and DNA binding. FHL1-3 inhibited HIF1 transcriptional activity and HIF1-mediated VEGF expression in a hypoxia-independent manner. Moreover, FHL1 blocked HIF1a-HIF1b heterodimerization and HIF1a recruitment to the VEGF promoter. These data suggest that FHL proteins are involved in negative regulation of VEGF possibly by interfering with the dimerization and DNA binding of HIF1 subunits and may play an important role in tumor angiogenesis. Ó 2012 IUBMB IUBMB Life, 64(11): 921–930, 2012

Additional Supporting Information may be found in the online version of this article. Address correspondence to: Qinong Ye, Department of Medical Molecular Biology, Beijing Institute of Biotechnology, 27 Tai-Ping Lu Rd, Beijing 100850, People’s Republic of China. Tel: 18610-6818-0809. Fax: 18610-6824-8045. E-mail: [email protected] Received 25 March 2012; accepted 12 August 2012 *J.L., X.Q., Z.Z., and J.M. contributed equally to this work. ISSN 1521-6543 print/ISSN 1521-6551 online DOI: 10.1002/iub.1089

four and a half LIM domain; vascular endothelial growth factor; hypoxia-inducible factor 1; protein–protein interaction; transcriptional activity.

INTRODUCTION Angiogenesis is essential for tumor growth and metastasis. It is strictly regulated by a highly coordinated process that is modulated by many molecules (1). Among them is vascular endothelial growth factor (VEGF), the key regulator of angiogenesis in cancer growth and progression (2). VEGF is a secreted dimeric glycoprotein and can be induced in cells that are not receiving enough oxygen. Under hypoxia, the expression and function of VEGF are upregulated by the transcription factor hypoxia-inducible factor 1 (HIF1a/HIF1b) (3). Under aerobic conditions, HIF1a is hydroxylated at the level of conserved proline and asparagine residues, and then ubiquitinated via binding of the von Hippel-Lindau (VHL) tumor suppressor protein (4– 6). VHL is the recognition component of an E3 ubiquitin-protein ligase that targets HIF1a for proteasomal degradation. In response to hypoxia, the hydroxylation and ubiquitination of HIF1a is reduced. HIF1a becomes stabilized, accumulates in the nucleus, and binds to HIF1b/aryl hydrocarbon receptor nuclear translocator (ARNT), thus forming HIF1a-HIF1b heterodimers. The heterodimeric complex binds to hypoxia response elements upstream of hypoxic-regulated genes, regulating expression of a variety of hypoxia-inducible target genes, including VEGF. The four and a half LIM (FHL) proteins belong to the LIMonly protein family, characterized by four complete LIM domains, preceded by an N-terminal half LIM domain (7). LIM

922

LIN ET AL.

domains are cysteine-rich zinc finger motifs responsible for a wide range of protein–protein interactions. The FHL protein family consists of the members FHL1/SLIM1, FHL2/DRAL/ SLIM3, FHL3/SLIM2, FHL4, and FHL5/ACT (activator of CREM in the testis). There is no human homolog of mouse FHL4. The FHL family members are expressed in a cell- and tissue-specific manner and are involved in various cellular processes, including modulation of cell proliferation, cell differentiation, apoptosis, cell adhesion, migration, transcription, and signal transduction. Recently, some FHL family members have been reported to play important roles in cancer development and progression. FHL1 expression is downregulated in various types of malignancies including breast cancer (8, 9), liver cancer (10), gastric cancer (11), kidney cancer (8), prostate cancer (8), and lung cancer (12). FHL1 acts downstream of the oncogene SRC to specifically block cancer cell growth and migration (13). The role of FHL2 in tumorigenesis is complex, because human FHL2 is downregulated in malignant rhabdomyosarcoma (14) and liver cancer (10), and upregulated in ovarian (15) and gastrointestinal (16) cancers. FHL3 has been shown to be downregulated in breast cancer (17) and liver cancer (10) and suppress breast and liver cancer cell growth. Recently, we demonstrated that human FHL1, FHL2, and FHL3 proteins physically and functionally interact with Smad2, Smad3, and Smad4, important regulators of cancer development and progression, and inhibit human hepatoma cell growth in vitro and in vivo (10). To further elucidate the molecular mechanisms by which FHL proteins regulate tumor growth and progression, we tested whether human FHL1-3 can modulate VEGF expression in cancer cells. We demonstrate that FHL1-3 inhibit HIF1 transcriptional activity and HIF1-mediated VEGF expression possibly by blocking HIF1a-HIF1b heterodimerization and HIF1a recruitment to the VEGF promoter.

MATERIALS AND METHODS Plasmids and Cell Culture VEGF promoter-containing luciferase reporter (VEGF-LUC) and expression vectors for FHL1, FHL2, FHL3, FHL1 siRNA, FHL2 siRNA, and FHL3 siRNA have been described previously (10, 18). Myc-tagged HIF1a and FLAG-tagged HIF1b were prepared by cloning the polymerase chain reaction (PCR)-generated full-length HIF1a cDNA or HIF1b cDNA from a human mammalian cDNA library (Clontech, Mountain View, CA) into the pcDNA3 vector harboring Myc or FLAG epitope sequence. Deletion mutants of FHL1 and HIF1a were constructed by inserting PCR-generated fragments from the corresponding cDNAs into the pcDNA3 vector harboring Myc or FLAG epitope sequence. Cancer cell lines were routinely cultured in Dulbecco’s modified Eagle’s medium (Invitrogen, Carlsbad, CA) containing 10% fetal bovine serum at 37 8C in a humidified atmosphere of 5% CO2 in air. For hypoxia experiments, cells

were grown in a hypoxia chamber in 1% O2, 5% CO2, and 94% N2 for 24 h.

Luciferase Reporter Assay Mammalian cells were transfected using Lipofectamine 2000 (Invitrogen) with 0.2 lg of VEGF-LUC reporter plasmid, 50 ng of HIF1a expression vector, 0.1 lg of b-galactosidase reporter, and various amounts of over-expression vectors for FHL1, FHL2, or FHL3, or small-interfering RNA (siRNA) vectors targeting FHL1, FHL2, or FHL3, and the respective empty vector was used to adjust the total amount of DNA. Twenty-four hours after transfection, cells were either exposed to normoxia or hypoxia for another 24 h before harvested. Luciferase and b-galactosidase activities were determined as described previously (19). Quantitative Real-time Reverse Transcription-PCR (qRT-PCR) Total RNA was isolated using TRIzol Reagent (Invitrogen) and reverse transcribed using SuperScript II Reverse Transcriptase (Invitrogen). Real-time PCR was performed with VEGF- and GAPDH-specific primers. The sense primer for VEGF was 50 TTCTGGGCTGTTCTCGCTTCG-30 , and the antisense primer was 50 -CCCCTCTCCTCTTCCTTCTCT-30 . The sense primer for GAPDH was 50 -ACCACAGTCCATGCCATCAC-30 , and the antisense primer was 50 -TCCACCACCCTGTTGCTG TA-30 . The fold change in expression of VEGF was calculated using the 22DDCt method, with GAPDH as an internal control (20). Enzyme-linked Immunosorbent Assay Cells were transiently transfected with over-expression vectors for FHL1, FHL2, or FHL3, or small-interfering RNA (siRNA) vectors targeting FHL1, FHL2, or FHL3. At 48 h posttransfection, medium were collected for VEGF secretion assay. Human VEGF-A protein concentrations was determined by enzyme-linked immunosorbent assay (ELISA) analysis according to the manufacturer’s protocol (R&D Systems, Minneapolis, MN). The values obtained were normalized to the total protein concentration in the total cell extracts prepared from each dish. GST Pull-down Assay The GST- and His-fusion proteins were expressed and purified according to the manufacturers’ protocols (Amersham Pharmacia, Piscataway, NJ and Qiagen, Valencia, CA). The purified GST fusion proteins bound to glutathione-Sepharose beads were incubated with purified His-fusion proteins, and the adsorbed proteins were analyzed as previously described (10). Co-immunoprecipitation For transfection-based co-immunoprecipitation assays, 293T cells were transfected with the indicated plasmids using Lipo-

FHL FAMILY MEMBERS SUPPRESS VEGF

fectamine 2000 (Invitrogen). Cells were harvested and lysed in lysis buffer. Co-immunoprecipitation was performed with antiFLAG (Sigma-Aldrich, St. Louis, MO) or anti-Myc (SigmaAldrich) as previously described (10). For detecting interaction of endogenous HIF1a with FHL1, FHL2, or FHL3, HepG2 cells were lysed and immunoprecipitated with anti-HIF1a antibody (Sigma-Aldrich) or control serum (Santa Cruz Biotech, Santa Cruz, CA). Immunoprecipitated proteins were run on sodium dodecyl sulfate polyacrylamide gel electropheresis (SDSPAGE), and standard Western blot were performed using anti-HIF1a, anti-FHL1 (Proteintech, Chicago, IL), anti-FHL2 (Santa Cruz Biotech), or anti-FHL3 (Proteintech).

Chromatin Immunoprecipitation Assay Chromatin immunoprecipitation (ChIP) assays were performed as described previously (10). The following primers were used for ChIP PCR analysis (20): VEGF promoter sense, 50 -ACAGACGTTCCTTAGTGCTGG-30 ; VEGF promoter antisense, 50 AGCTGAGAACGGGAAGCTGTG-30 ; VEGF upstream sense, 50 -GAATTCTGTGCCCTCACTCC-30 ; VEGF upstream antisense, 50 -GTAGACATCTTGGGGCAGGA-30 . Statistical Analysis All data shown are means 6 SD of triplicates of one representative experiment and have been repeated three times with similar results. Statistical analysis was performed using SPSS version 11.0 software. Statistical significance in the luciferase activity assays among multiple constructs was determined by one-way analysis of variance. Statistical differences between two groups of data were assessed using the unpaired t-test. Statistical significance was taken as P \ 0.05. RESULTS FHL1-3 Inhibit VEGF Promoter Activity To determine whether FHL1-3 affect VEGF promoter activity, luciferase reporter assays were performed. Over-expression of FHL1, FHL2, and FHL3 in human HepG2 hepatoma cells significantly reduced VEGF reporter activity in a dose-dependent manner (Fig. 1A), whereas knockdown of endogenous FHL1, FHL2, or FHL3 with their respective siRNA effectively enhanced the VEGF promoter activity in a dose-dependent manner (Fig. 1B). Similar results were observed in different cell lines, including human SMMC7721 hepatoma cells, MCF7 and MDA-MB-468 breast cancer cells, and 293T embryonic kidney cells (Supporting Information Fig. S1). Similar to the results obtained under normoxia, over-expression of FHL1-3 decreased VEGF promoter activity under hypoxia (Fig. 1C), whereas reduction of FHL1-3 increased VEGF promoter activity (Fig. 1D). FHL1-3 Repress VEGF Expression To verify the finding that VEGF promoter activity is inhibited by FHL1-3, qRT-PCR and ELISA analysis were performed

923

to determine VEGF mRNA expression and VEGF secretion levels, respectively. In agreement with the results of the luciferase reporter analysis, both under normoxia and under hypoxia, over-expression of FHL1-3 reduced VEGF mRNA expression (Fig. 2A) and VEGF secretion levels (Fig. 2B), whereas reducing endogenous FHL1-3 levels by siRNA constructs increased the levels of VEGF mRNA (Fig. 2C) and secretion of endogenous VEGF protein (Fig. 2D).

FHL1-3 Interact with HIF1a In Vitro and In Mammalian Cells To investigate molecular mechanisms by which FHL1-3 inhibit the promoter activity and expression of VEGF, we tested whether FHL1, FHL2, and FHL3 interact with HIF1a, which has been reported to regulate VEGF expression through binding to the VEGF promoter (3). GST pull-down experiments indicated that GST-FHL1, GST-FHL2, and GST-FHL3, but not GST alone, bound to HIF1a, with FHL1 showing relatively high affinity (Fig. 3A). To confirm the interaction between HIF1a and FHL1-3 in mammalian cells, immunoprecipitation experiments were conducted in 293T cells co-transfected with Myc-tagged HIF1a and FLAG-tagged FHL1, FHL2, or FHL3. Cell lysates were immunoprecipitated with anti-FLAG. FLAG-tagged FHL1, FHL2, and FHL3, but not FLAG, interacted with Myc-tagged HIF1a (Fig. 3B). Importantly, endogenous FHL1, FHL2, or FHL3 proteins interacted with HIF1a in HepG2 cells (Figs. 3C–3E). Mapping of Interaction Domain of HIF1a and FHL1 To define which region of HIF1a interacts with FHL1, various Myc-tagged HIF1a deletion mutants were used for co-immunoprecipitation. Except HIF1a (482–826) containing the trans-activation domain (TAD), the mutants of HIF1a (1–166) and HIF1a (167–481), containing the basic helix-loophelix motif (bHLH) and the PER-ARNT-SIM (PAS) domain, respectively, bound to FHL1 (Fig. 4A). To delineate the domains in the FHL1 that mediate the protein–protein interaction with HIF1a, a series of deletion mutants of FHL1 were used for co-immunoprecipitation. All FHL1 deletion mutants with more than one LIM domain were observed to co-immunoprecipitate with HIF1a (Fig. 4B). To further test whether a half LIM or a single LIM domain of FHL1 is responsible for mediating the interaction with HIF1a, GST pull down assay was performed. A single LIM domain, but not a half LIM domain, of FHL1 interacted with HIF1a (Fig. 4C). Similarly, a single LIM domain of FHL2 or FHL3 is sufficient for binding HIF1a (Figs. 4D and 4E). FHL1-3 Inhibit HIF1a-depedent VEGF Promoter Activity and VEGF Expression To test whether the interaction of HIF1a and FHL1, FHL2, or FHL3 affects VEGF promoter activity, HepG2 cells were cotransfected with the VEGF promoter reporter, Myc-HIF1a, and FLAG-tagged FHL1, FHL2, or FHL3 and were incubated under

924

LIN ET AL.

Figure 1. FHL1-3 inhibit VEGF promoter activity. (A) Effects of over-expression of FHL1-3 on VEGF promoter activity. HepG2 cells were co-transfected with reporter VEGF-LUC and increasing amounts of plasmid expressing FLAG-tagged FHL1, FHL2, or FHL3. Cells were harvested and analyzed for luciferase activity. Western blotting showed expression levels of FLAG-tagged FHL1-3 with antibodies against FLAG and GAPDH (lower panel). GAPDH was used as a loading control. (B) Effects of knockdown of endogenous FHL1-3 on VEGF promoter activity. HepG2 cells were co-transfected with reporter VEGF-LUC and increasing amounts of FHL1 siRNA, FHL2 siRNA, or FHL3 siRNA. Cells were analyzed as in (A). Western blotting showed expression levels of FHL1-3 with antibodies against FHL1-3 and GAPDH (lower panel). (C) Effects of overexpression of FHL1-3 on VEGF promoter activity under hypoxia condition. HepG2 cells were co-transfected with reporter VEGF-LUC and FLAG-tagged FHL1, FHL2, or FHL3. At 24 h post-transfection, cells were exposed to either normoxic or hypoxic condition for another 24 h. Cells were analyzed as in (A). (D) Effects of knockdown of FHL1-3 on VEGF promoter activity under hypoxia condition. HepG2 cells were co-transfected with reporter VEGF-LUC and FHL1 siRNA, FHL2 siRNA, or FHL3 siRNA. Cells were treated and analyzed as in (C). All data shown are means 6 SD of triplicates of one representative experiment and have been repeated three times with similar results. *P \ 0.05, **P \ 0.01 versus corresponding empty vector or control siRNA. normoxic or hypoxic conditions. As previously reported (3–6), overexpression of HIF1a augmented VEGF promoter activity under both normoxic and hypoxic conditions, with relatively high activity under hypoxia. Interestingly, over-expression of FHL1-3

decreased HIF1a transcriptional activity in a hypoxia-independent manner (Fig. 5A). To evaluate the role of endogenous HIF1a in FHL1-3 repression of VEGF promoter activity, HepG2 cells were transfected

FHL FAMILY MEMBERS SUPPRESS VEGF

925

Figure 2. FHL1-3 repress VEGF expression under normoxic or hypoxic conditions. (A and B) HepG2 cells were stably transfected with FLAG-tagged FHL1, FHL2, or FHL3. Total RNA was isolated from the transfected cells, and VEGF mRNA expression was analyzed by quantitative real-time RT-PCR (A). ELISA was performed to determine VEGF protein levels in supernatants (B). Western blotting showed expression levels of FLAG-tagged FHL1-3 with antibodies against FLAG and GAPDH (A, lower panel). (C and D) HepG2 cells were stably transfected with FHL1 siRNA, FHL2 siRNA, or FHL3 siRNA. Expression levels of VEGF mRNA (C) and protein (D) were determined as in (A) and (B). Western blot showed expression levels of FHL1-3 with antibodies against FHL1-3 and GAPDH (C, lower panel). All data shown are means 6 SD of triplicates of one representative experiment and have been repeated three times with similar results. *P \ 0.01 versus empty vector or control siRNA. with HIF1a siRNA or control siRNA. Compared with the control, HIF1a siRNA effectively lowered the protein expression of HIF1a (Fig. 5B). Importantly, knockdown of HIF1a dramatically reduced the ability of FHL1-3 to inhibit VEGF promoter activity under normoxic or hypoxic conditions (Fig. 5C). To test the roles of endogenous FHL1-3 in HIF1a induction of VEGF promoter activity, HepG2 cells were co-transfected with

the VEGF promoter reporter and siRNA vectors targeting FHL1, FHL2, or FHL3. Reduction of FHL1, FHL2, or FHL3 enhanced VEGF promoter activity under normoxic or hypoxic conditions (Fig. 5D). Similar results were observed in SMMC7721, MCF7, and 293T cells (Supporting Information Fig. S2). To corroborate the results of the luciferase reporter assay, the effects of FHL1-3 on HIF1a-mediated VEGF expression

926

LIN ET AL.

of HIF1a and HIF1b. Therefore, to elucidate the mechanism by which FHL1 modulates HIF1 transcriptional activity, the effect of FHL1 on interaction of HIF1a with HIF1b was determined by co-immunoprecipitation assays. Indeed, both under normoxia and under hypoxia, FHL1 over-expression inhibited the interaction of HIF1a with HIF1b in a dose-dependent manner (Fig. 6A and Supporting Information Fig. S3A), whereas reduction of endogenous FHL1 with FHL1 siRNA enhanced the interaction between HIF1a and HIF1b (Fig. 6B and Supporting Information Fig. S3B). These results suggest that FHL1 inhibition of HIF1 transcriptional activity might be due to the repression of dimerization between HIF1a and HIF1b.

Figure 3. FHL1-3 interact with HIF1a in vitro and in vivo. (A) Interaction of HIF1a and FHL1-3 in vitro. Glutathione-sepharose beads bound with GST, GST-FHL1, GST-FHL2, or GST-FHL3 were incubated with purified His-HIF1a protein. After washing the beads, the bound proteins were eluted and subjected to Western blot with anti-His. (B) Interaction of HIF1a and FHL1-3 in mammalian cells by co-immunoprecipitation (Co-IP). 293T cells were transfected with Myc-tagged HIF1a and Flag-tagged FHL1, FHL2, or FHL3. Cell lysates were immunoprecipitated with antiFLAG, followed by immunoblot (IB) with anti-FLAG or antiMyc. (C–E) Interaction of endogenous FHL1-3 with HIF1a in vivo. Co-IP was performed on the HepG2 lysates using antiHIF1a or an IgG control antibody, followed by immunoblot with anti-FHL1, anti-FHL2, or anti-FHL3. were examined. Consistent with previous report (3–6), overexpression of HIF1a increased VEGF mRNA expression (Fig. 5E). Intriguingly, over-expression of FHL1-3 inhibited HIF1amediated VEGF mRNA expression (Fig. 5E), while knockdown of endogenous FHL1, FHL2, or FHL3 promoted HIF1a-induced VEGF mRNA expression (Fig. 5F).

FHL1 Inhibits HIF1a and HIF1b Dimerization HIF1 is a heterodimeric transcription factor consisting of HIF1a and HIF1b subunits and subunit dimerization is an obligatory step for HIF1 activity (3–6). We hypothesized that FHL1 might regulate HIF1 activity by inhibiting dimerization

FHL1 Inhibits the Binding of HIF1a Protein to the VEGF Promoter On dimerization, transcription factors bind to promoter sequences of target genes, thus regulating target gene expression. As FHL1 alters dimerization between HIF1a and HIF1b, we investigated whether FHL1 affects binding of HIF1a to the VEGF promoter by ChIP assay. Consistent with the published results (21), HIF1a was recruited to the VEGF promoter (Fig. 6C). Importantly, under normoxic or hypoxic conditions, FHL1 over-expression decreased recruitment of HIF1a to the VEGF promoter, but not to a region  1.2-kb upstream of the VEGF promoter (Fig. 6C and Supporting Information Figs. S3C and S3E), while knockdown of FHL1 with FHL1 siRNA increased recruitment of HIF1a to the VEGF promoter (Fig. 6D and Supporting Information Figs. S3D and S3F). These results suggest that FHL1 modulation of HIF1 transcription activity may be through repression of HIF1a binding to the VEGF promoter. DISCUSSION In this study, we identified FHL proteins as important regulators of VEGF expression. FHL proteins modulate VEGF expression through interaction with HIF-1a. While we were preparing our article, Hubbi et al. (22) reported that FHL1, FHL2, and FHL3 inhibit HIF1 transcriptional activity in an oxygen-independent manner. They found that FHL1 disrupts binding of HIF-1a to the general co-activators p300/CBP. We identified a new mechanism by which FHL1 regulates HIF1 transcriptional activity, that is, FHL1 blocks HIF1a-HIF1b heterodimerization and subsequent HIF1a recruitment to the VEGF promoter. Hubbi et al. determined the effects of FHL1-3 on VEGF mRNA expression in Hela cervical cancer cell line. We demonstrated that, in liver cancer cells, FHL1-3 inhibit not only the expression of VEGF mRNA but also that of VEGF protein. Our results are relevant to the reported roles of FHL proteins in liver cancer development and progression. Moreover, we defined the minimal interaction region of HIF1a in FHL1, namely, a single LIM domain of FHL1 is sufficient for binding to HIF1a. Although our major conclusions are the same as those of Hubbi et al., there is discrepancy between our study and their study. Hubbi et al. showed that only FHL2 physically interact with

FHL FAMILY MEMBERS SUPPRESS VEGF

927

Figure 4. Mapping of interacting regions of FHL1-3 with HIF1a. (A) Mapping of FHL1 interaction regions in HIF1a. 293T cells were transfected with FLAG-tagged FHL1 and Myc-tagged HIF1a or its deletion mutants. Co-IP was performed using anti-FLAG, followed by immunoblotting with anti-Myc or anti-FLAG. (B) Mapping of HIF1a interaction regions in FHL1. 293T cells were cotransfected with Myc-tagged HIF1a and FLAG-tagged FHL1 or its deletion mutants. Co-IP was performed as in (A). (C–E) Mapping of minimal HIF1a interaction region in FHL1-3. GST or GST-fusion proteins containing the N-terminal half LIM (1/2 LIM) or the first LIM domain (LIM1) of FHL1 (C), FHL2 (D), or FHL3 (E) was incubated with purified His-HIF1a, captured with glutathione-sepharose beads, and then subjected to immunoblot with anti-His.

HIF1a although FHL1, FHL2, and FHL3 repress HIF1 transcriptional activity. In our study, however, all the FHL proteins interact with HIF1a. This notion is supported by several experiments including GST pull-down, co-immunoprecipitation with over-expressed proteins, and co-immunoprecipitation with endogenous proteins. The discrepancy between our study and that reported by Hubbi et al. may be that we used different tags to express FHL and HIF1a proteins, and we determined interaction of three endogenous FHL proteins with HIF1a in HepG2 hepatoma cells. We also noticed that there was severe degradation of purified GST-HIF1a mutant proteins used by Hubbi et al. In our study, we used purified GST-FHL1, GST-FHL2, GSTFHL3, or GST-FHL1 mutant proteins, which had negligible degradation, to perform GST pull-down assays. FHL proteins are characterized by four complete LIM domains, preceded by an N-terminal half LIM domain. LIM domains are cysteine-rich zinc finger motifs involved in a wide range of protein–protein interactions with transcription factors, cell-signaling molecules, and cytoskeleton-associated proteins (23). Increasing amounts of evidence show that different combinations of LIM domains are responsible for interaction with dif-

ferent proteins. For example, all LIM domains of FHL1 are required to interact with RIP140 (24). Similarly, deletion of any complete LIM domain or the N terminal half domain of FHL2 ablates the interaction with TNF receptor-associated factor 6 (TRAF6) (25). At least two LIM domains of FHL2 are required for efficient interaction with extracellular signal-regulated kinase 2 (ERK2) (26). The N-terminal two and a half LIM domains of FHL1 are sufficient for binding to myosin binding protein C (MyBP-C) (27). In contrast, removal of the first two and a half N-terminal LIM domains does not significantly affect the ability of FHL2 to interact with the E1A-targeted transcription factor E4F1 (28). The E4F1-binding region of FHL2 lies within the third LIM domain. Our results indicate that a single LIM domain of FHL1 is sufficient for interaction with HIF1a. Recently, the tumor suppressor protein LIMD1 (LIM domaincontaining protein 1) has been shown to bind prolyl hydroxylases (PHDs), the enzymes involved in regulation of HIF1 hydroxylation, and the VHL tumor suppressor, the recognition component of a ubiquitin-ligase complex (29). Such interaction creates an enzymatic niche that enables efficient degradation of HIF1. The LIM domains of LIMD1 are responsible for

928

LIN ET AL.

Figure 5. FHL1-3 inhibit HIF1a-depedent VEGF promoter activity and expression. (A) Overexpression of FHL1-3 reduces HIF1amediated VEGF promoter activity. HepG2 cells were co-transfected with reporter VEGF-LUC, HIF1a, and FLAG-tagged FHL1, FHL2, or FHL3. At 24 h post-transfection, cells were exposed to either normoxic or hypoxic condition for another 24 h. Cell lysates were subjected to luciferase activity analysis. (B) Western blot showing the specific knockdown effect of HIF1a siRNA on the endogenous HIF1a protein level. HepG2 cells were transfected with expression vector for HIF1a siRNA or scramble siRNA (control) plasmid. Forty-eight hours after transfection, whole-cell extracts were prepared and probed with anti-HIF1a or GAPDH. (C) Effect of HIF1a reduction on VEGF promoter activity regulated by FHL1-3. HepG2 cells were co-transfected with reporter VEGF-LUC, HIF1a siRNA, and FLAG-tagged FHL1, FHL2, or FHL3. Cells were treated and analyzed as in (A). (D) Effects of knockdown of FHL1-3 on HIF1a-depedent VEGF promoter activity. HepG2 cells were co-transfected with reporter VEGF-LUC, HIF1a, and FHL1 siRNA, FHL2 siRNA, or FHL3 siRNA. Cells were treated and analyzed as in (A). (E) Effects of overexpression of FHL1-3 on HIF1a-depedent VEGF mRNA expression. HepG2 cells were co-transfected with HIF1a and FLAG-tagged FHL1, FHL2, or FHL3. Total RNA was isolated from the transfected cells and VEGF mRNA expression was analyzed by quantitative RT-PCR. (F) Effects of reduction of FHL1-3 on HIF1a-depedent VEGF mRNA expression. HepG2 cells were co-transfected with HIF1a and FHL1 siRNA, FHL2 siRNA, or FHL3 siRNA. VEGF mRNA expression was analyzed as in (E). All data shown are means 6 SD of triplicates of one representative experiment and have been repeated three times with similar results. *P \ 0.05, **P \ 0.01 versus corresponding empty vector or control siRNA. #P \ 0.05, ##P \ 0.01 versus HIF1a or HIF1a siRNA.

FHL FAMILY MEMBERS SUPPRESS VEGF

929

Figure 6. FHL1 inhibits HIF1a-HIF1b dimerization and recruitment of HIF1a to the VEGF promoter. (A and B) 293T cells were co-transfected with Myc-tagged HIF1a, FLAG-tagged HIF1b, and increasing amounts of HA-tagged FHL1 (A) or FHL1 siRNA (B), as indicated. Co-IP was performed using anti-Myc, followed by immunoblot with the indicated antibodies. (C and D) HepG2 cells were stably transfected with FLAG-tagged FHL1 (C) or FHL1 siRNA (D), and subjected to ChIP assays with anti-HIF1a or normal IgG. Immunoprecipitated DNA was PCR amplified with primers that annealed to the proximal region of the VEGF promoter (Promoter) or the region  1.2-kb upstream of the VEGF promoter (Upstream).

its interaction with PHDs and VHL. Because FHL1-3 are also LIM-containing proteins, we cannot exclude the possibility that FHL1-3 can facilitate VHL-mediated degradation of HIF1. Heterodimerization between HIF1a and HIF1b is an essential step toward forming active, DNA binding complexes (30–32). HIF1 contains a highly conserved N-terminal bHLH motif, adjacent PAS domain, and loosely conserved C-terminal transactivation and trans-repression regions. The bHLH domain is a well-characterized DNA binding and dimerization domain. The PAS domain contributes to dimerization. Strong dimer formation often requires collaboration between bHLH and PAS domains. The fact that FHL1 interacts with the bHLH and PAS domains of HIF1a and blocks HIF1a-HIF1b dimerization indicates that FHL1 competes with HIF1b for binding HIF1a. Because a single LIM domain of FHL1 is sufficient for association with HIF1a, the presence of multiple interacting regions may allow efficient interaction of FHL1 with HIF1a and subsequent inhibition of HIF1a-HIF1b dimerization. Consistent with FHL1 inhibition of HIF1a-HIF1b dimerization, FHL1 represses HIF1a binding to DNA and subsequent transcriptional activation.

Angiogenesis is fundamental to tumor growth, invasion, and metastasis. The discovery of VEGF as the key regulator of tumor angiogenesis has enabled the specific and successful repression of angiogenesis in experimental and clinical studies (33–37). Different agents including antibodies, aptamers, peptides, and small molecules have been shown to inhibit VEGF and its proangiogenic functions. Some of them were approved to clinical applications. Although anti-VEGF therapy has improved the treatment of several types of cancer during recent years, some cancer patients develop resistance to anti-VEGF therapy following initially successful therapy, while others never show a response. Thus, elucidating the molecular mechanisms underlying VEGF regulation is of great significance. Whether FHL proteins are involved in anti-VEGF therapy remains to be elucidated.

ACKNOWLEDGEMENTS This work was supported by Major State Basic Research Development Program (2012CB945100 and 2011CB504202), National Natural Science Foundation (81071954, 81072173, and 31071174), and National Key Technologies R&D Program for New Drugs (2009ZX09301-002).

930

LIN ET AL.

REFERENCES 1. Bergers, G. and Benjamin, L. E. (2003) Tumorigenesis and the angiogenic switch. Nat. Rev. Cancer 3, 401–410. 2. Cao, Y. and Liu, Q. (2007) Therapeutic targets of multiple angiogenic factors for the treatment of cancer and metastasis. Adv. Cancer Res. 97, 203–224. 3. Majmundar, A. J., Wong, W. J., and Simon, M. C. (2010) Hypoxia-inducible factors and the response to hypoxic stress. Mol. Cell 40, 294– 309. 4. Semenza, G. L. (2010) Defining the role of hypoxia-inducible factor 1 in cancer biology and therapeutics. Oncogene 29, 625–634. 5. Jokilehto, T. and Jaakkola, P. M. (2010) The role of HIF prolyl hydroxylases in tumour growth. J. Cell Mol. Med. 14, 758–770. 6. Onnis, B., Rapisarda, A., and Melillo, G. (2009) Development of HIF-1 inhibitors for cancer therapy. J. Cell Mol. Med. 13, 2780–2786. 7. Johannessen, M., Møller, S., Hansen, T., Moens, U., and Van Ghelue, M. (2006) The multifunctional roles of the four-and-a-half-LIM only protein FHL2. Cell Mol. Life Sci. 63, 268–284. 8. Ding, L., Niu, C., Zheng, Y., Xiong, Z., Liu, Y., et al. (2011) FHL1 interacts with oestrogen receptors and regulates breast cancer cell growth. J. Cell Mol. Med. 15, 72–85. 9. Li, X., Jia, Z., Shen, Y., Ichikawa, H., Jarvik, J., et al. (2008) Coordinate suppression of Sdpr and Fhl1 expression in tumors of the breast, kidney, and prostate. Cancer Sci. 99, 1326–1333. 10. Ding, L., Wang, Z., Yan, J., Yang, X., Liu, A., et al. (2009) Human four-and-a-half LIM family members suppress tumor cell growth through a TGF-b-like signaling pathway. J. Clin. Invest. 119, 349–361. 11. Sakashita, K., Mimori, K., Tanaka, F., Kamohara, Y., Inoue, H., et al. (2008) Clinical significance of loss of Fhl1 expression in human gastric cancer. Ann. Surg. Oncol. 15, 2293–2300. 12. Niu, C., Liang, C., Guo, J., Cheng, L., Zhang, H., et al. (2012) Downregulation and growth inhibitory role of FHL1 in lung cancer. Int. J. Cancer. 130, 2549–2556. 13. Shen, Y., Jia, Z., Nagele, R. G., Ichikawa, H., and Goldberg, G. S. (2006) SRC uses Cas to suppress Fhl1 in order to promote nonanchored growth and migration of tumor cells. Cancer Res. 66, 1543–1552. 14. Scholl, F. A., McLoughlin, P., Ehler, E., de Giovanni, C., and Scha¨fer, B. W. (2000) DRAL is a p53-responsive gene whose four and a half LIM domain protein product induces apoptosis. J. Cell Biol. 151, 495– 506. 15. Gabriel, B., Mildenberger, S., Weisser, C. W., Metzger, E., Gitsch, G., et al. (2004) Focal adhesion kinase interacts with the transcriptional coactivator FHL2 and both are overexpressed in epithelial ovarian cancer. Anticancer Res. 24, 921–927. 16. Wang, J., Yang, Y., Xia, H. H., Gu, Q., Lin, M. C., et al. (2007) Suppression of FHL2 expression induces cell differentiation and inhibits gastric and colon carcinogenesis. Gastroenterology 132, 1066–1076. 17. Niu, C., Yan, Z., Cheng, L., Zhu, J., Zhang, H., et al. (2011) Downregulation and antiproliferative role of fhl3 in breast cancer. IUBMB Life 63, 764–771. 18. Xiong, Z. H., Ding, L. H., Yang, S. X., Sun, Q. H., Yang, X., et al. (2007) The cloning of the VEGF promoter reporter and the effect of estrogen receptor on the reporter activity. Lett. Biotech. 18, 915–917. 19. Ding, L., Yan, J., Zhu, J., Zhong, H., Lu, Q., et al. (2003) Ligand-independent activation of estrogen receptor alpha by XBP-1. Nucleic Acids Res. 31, 5266–5274.

20. Cheng, L., Li, J., Han, Y., Lin, J., Niu, C., et al. (2012) PES1 promotes breast cancer by differentially regulating ERa and ERb. J. Clin. Invest. 122, 2857–2870. 21. Kang, H. J., Kim, H. J., Rih, J. K., Mattson, T. L., Kim, K. W., et al. (2006) BRCA1 plays a role in the hypoxic response by regulating HIF1a stability and by modulating vascular endothelial growth factor expression. J. Biol. Chem. 281, 13047–13056. 22. Hubbi, M. E., Gilkes, D. M., Baek, J. H., and Semenza, G. L. (2012) Four-and-a-half LIM domain proteins inhibit transactivation by hypoxiainducible factor 1. J. Biol. Chem. 287, 6139–6149. 23. Kleiber, K., Strebhardt, K., and Martin, B. T. (2007) The biological relevance of FHL2 in tumour cells and its role as a putative cancer target. Anticancer Res. 27, 55–61. 24. Lin, J., Ding, L., Jin, R., Zhang, H., Cheng, L., et al. (2009) Four and a half LIM domains 1 (FHL1) and receptor interacting protein of 140 kDa (RIP140) interact and cooperate in estrogen signaling. Int. J. Biochem. Cell Biol. 41, 1613–1618. 25. Bai, S., Kitaura, H., Zhao, H., Chen, J., Mu¨ller, J. M., et al. (2005) FHL2 inhibits the activated osteoclast in a TRAF6-dependent manner. J. Clin. Invest. 115, 2742–2751. 26. Purcell, N. H., Darwis, D., Bueno, O. F., Mu¨ller, J. M., Schu¨le, R., et al. (2004) Extracellular signal-regulated kinase 2 interacts with and is negatively regulated by the LIM-only protein FHL2 in cardiomyocytes. Mol. Cell Biol. 24, 1081–1095. 27. McGrath, M. J., Cottle, D. L., Nguyen, M. A., Dyson, J. M., Coghill, I. D., et al. (2006) Four and a half LIM protein 1 binds myosin-binding protein C and regulates myosin filament formation and sarcomere assembly. J. Biol. Chem. 281, 7666–7683. 28. Paul, C., Lacroix, M., Iankova, I., Julien, E., Scha¨fer, B. W., et al. (2006) The LIM-only protein FHL2 is a negative regulator of E4F1. Oncogene 25, 5475–5484. 29. Foxler, D. E., Bridge, K. S., James, V., Webb, T. M., Mee, M., et al. (2012) The LIMD1 protein bridges an association between the prolyl hydroxylases and VHL to repress HIF-1 activity. Nat. Cell Biol. 14, 201–208. 30. Hao, N., Whitelaw, M. L., Shearwin, K. E., Dodd, I. B., and ChapmanSmith, A. (2011) Identification of residues in the N-terminal PAS domains important for dimerization of Arnt and AhR. Nucleic Acids Res. 39, 3695–3709. 31. Kewley, R. J., Whitelaw, M. L., and Chapman-Smith, A. (2004) The mammalian basic helix-loop-helix/PAS family of transcriptional regulators. Int. J. Biochem. Cell Biol. 36, 189–204. 32. Yang, J., Zhang, L., Erbel, P. J., Gardner, K. H., Ding, K., et al. (2005) Functions of the Per/ARNT/Sim domains of the hypoxia-inducible factor. J. Biol. Chem. 280, 36047–36054. 33. Niu, G. and Chen, X. (2010) Vascular endothelial growth factor as an anti-angiogenic target for cancer therapy. Curr. Drug Targets 11, 1000– 1017. 34. Ellis, L. M. and Hicklin, D. J. (2008) VEGF-targeted therapy: mechanisms of antitumour activity. Nat. Rev. Cancer 8, 579–591. 35. Chen, H. X. and Cleck, J. N. (2009) Adverse effects of anticancer agents that target the VEGF pathway. Nat. Rev. Clin. Oncol. 6, 465–477. 36. Duda, D. G., Batchelor, T. T., Willett, C. G., and Jain, R. K. (2007) VEGF-targeted cancer therapy strategies: current progress, hurdles and future prospects. Trends Mol. Med. 13, 223–230. 37. Cardones, A. R. and Banez, L. L. (2006) VEGF inhibitors in cancer therapy. Curr. Pharm. Des. 12, 387–394.