Introduction
FHL2 is a LIM-only protein that belongs to the four and a half LIM-only protein family. It consists of four LIM domains and one N-terminal half LIM domain. The abbreviation LIM represents the names of three transcription factors (Lin-11, Isl-1 and Mec-3) in which such a domain was first identified (4). It is strongly expressed in cardiac and skeletal muscle cells but a much lower level was observed in other tissues and cell types (1–4). The LIM domains are double zinc finger motifs that play multiple roles in protein-protein interaction (5). FHL2 can interact with over 50 different proteins with diverse functions such as receptors, enzymes, transcription factors, cofactors or splicing factors (6). Thus, FHL2 is involved in the regulation of various cellular processes including proliferation, differentiation, migration, adhesion, motility and contraction.
The role of FHL2 in cancer is particularly intriguing since FHL2 binds to different proteins and can function in a cell-type dependent fashion as transcriptional co-activators of several transcription factors, including androgen receptor, AP-1, CREB, BRCA1, WT-1 and NF-κB in various transformed cell types (7–11), or as transcriptional co-repressors of ERK2, PLZF, SRF and FOXO1 (12–15). FHL2 was first identified as being downregulated in human rhabdomyosarcoma cells, suggesting a suppressor role in tumor development (16). Many recent studies have reported the differential expression of FHL2 in tumor tissues. Interestingly, FHL2 is overexpressed in breast cancer (17), prostate cancer (18), ovarian cancer (4), gastrointestinal cancer (19) and glioma (20) but downregulated in liver cancer (21), making the role of FHL2 in cancer development elusive. It is also notable that FHL2 triggers apoptosis in human RD, monkey kidney COS-1 and normal mouse fibroblast NIH 3T3 cell lines (22) but plays an anti-apoptotic role in glioblastoma (20). The intriguing aspects of FHL2 being as oncoprotein or tumor suppressor may be related to its interaction with different partner proteins even in different cell types (23). Suppression of FHL2 was detected to induce cell differentiation and inhibit tumorigenesis in FHL2 high expression colon cancer cells, while in HT29 cells with bare FHL2 expression, attenuated FHL2 suppressed cell growth and differentiation (19,23). We and other groups have reported that FHL2 was also a potent EMT inducer by stimulating vimentin and MMP-9 expressions and causing a loss of E-cadherin in colon DLD1 cells (24) and FHL2 negatively regulated the transcription of E-cadherin through interaction with Snail1 (25). For the tissue and cellular function specificity, FHL2 might be the target in cancer biological therapy.
In our study, we have demonstrated that a novel and effective way to knockdown FHL2, the rAAV-FHL2-shRNA can induce apoptosis significantly, inhibit tumorigenesis and tumor growth, strongly enhance the antitumor activity of 5-FU and markedly prolonged the survival time of animals bearing tumor xenografts in vivo. Our results document that rAAV-shRNA-FHL2 is a promising candidate for gene therapy of colon cancer.
Materials and methods
Cell culture
Culture reagents were purchased from Invitrogen (Carlsbad, CA, USA). Colon cancer cell line LoVo was obtained from American Type Culture Collection (Rockville, MD, USA) and cultured as described (19). Cell was maintained at 37°C in a 5% CO2 humidified incubator and subcultured using 0.25% trypsin every 2–3 days before confluence was reached.
Constructs
As we have described previously (20), the enhanced green fluorescence protein (EGFP) was constructed by inserting the EGFP gene between the XhoI and EcoRI sites of the AAV2 vector. Short hairpin RNA (shRNA) targeting FHL2 (sense: 5′-TCGACGCGAATCTCTCTTTGGCAAGTTCAAGAG ACTTGCCAAAGAGAGATTCGTTTTTTGGAAT-3′; anti-sense: 5′-CTAGATTCCAAAAAACGAATCTCTCTTTGG CAAGTCTCTTGAACTTGCCAAAGAGAGATTCGCG-3′) or luciferase (sense: 5′-TCGACGCGTACGCGGAATACT TCGATTCAAGAGATCGAAGTATTCCGCGTACGTTTT TTGGAAT-3′; antisense: 5′-CTAGATTCCAAAAAACGT ACGCGGAATACTTCGATCTCTTGAATCGAAGTATTC CGCGTACGCG-3′) was initially inserted into the SalI and XbaI sites of pAVU6+7 plasmid and then the construct was sub-cloned into the AAV2 expression plasmid in order to substitute the CMV promoter with the U6 promoter.
Western blot analysis
The whole cell lysates were prepared with lysis buffer (20 mM Tris-HCl, 1 mM EDTA, 1 mM EGTA, 1 mM sodium vanadate, 0.2 mM phenylmethylsulfonyl fluoride, 0.5% NP-40, 1 μg/ml leupeptin, 1 μg/ml aprotinin, and 1 μg/ml pepstatin A). In total, 10 or 30 μg of cell lysate was subjected to SDS-PAGE, transferred to PVDF membranes, and probed with first antibodies against CDK6, cyclin D, procaspase 3, cleaved caspase 3, procaspase 8, cleaved caspase 8 (Santa Cruz Biotechnology, Santa Cruz, CA, USA), procaspase 9 (Alexis Biochemical, San Diego, CA, USA), cleaved caspase 9 (Imgenex, San Diego, CA, USA) or FHL2 (MBL International Incorporation, Woburn, Japan) followed by the HRP-conjugated secondary antibody. Goat antihuman actin antibody (I-19, Santa Cruz Biotechnology) was used as internal control. Antigen-antibody complexes were visualized by the enhanced chemiluminescence (ECL) system (Amersham Biosciences, Little Chalfont, UK).
Preparation of rAAV
The rAAV particles were produced using a helper virus free system as previously described in HEK 293 cells (20). The viruses were purified by HiTrap Heparin column chromatography (Sigma Chemical Co., St. Louis, MO, USA) and viral titer was determined by real-time PCR using the SYBR-Green I kit (Applied Biosystems, Foster City, CA, USA) with a forward primer (5′-CGGCTGTTGGGCACTGA-3′) and a reverse primer (5′-CCGAAGGGACGAAGCAGAAG-3′). Aliquot of viral stocks (1.5×1012 viral genomes/ml) were stored at −80°C before use.
Gene transfection in vitro
LoVo cells were cultured in 6-well plastic plates and transfected with rAAV-EGFP or rAAV-FHL2-shRNA at different multiplicity of infection (MOI = 1×104, 1×105, 5×105). From then on, the enhanced green fluorescent protein (EGFP) was observed under a fluorescent microscope and the expression of rAAV-FHL2-shRNA was detected by western blot analysis.
Flow cytometry scan
To analyze cell cycle, cells were collected and fixed with ice-cold 70% ethanol in PBS and stored at −4°C until use. After resuspension, cells were incubated with 100 μl of RNase I (1 μg/ml) and 100 μl of propidium iodide (PI) (400 μg/ml) at 37°C and analyzed by flow cytometry (Becton-Dickinson, Franklin Lakes, NJ, USA). The cell cycle phase distribution was calculated from the resultant DNA histogram using Multicycle AV software (Phoenix Flow Systems, San Diego, CA, USA). Apoptosis was detected using the Annexin V-FITC kit according to the manufacturer’s instructions (Trevigen Inc., Gaithersburg, MD, USA). Briefly, cells with various treatments were collected and stained with Annexin V-FITC and PI in the dark for 15 min at room temperature. After addition of binding buffer, cells were analyzed by flow cytometry and analyzed using Winmdi 2.8.
Cell viability assay
The colorimetric WST-1 assay was performed to assess the effect of rAAV-FHL2-shRNA on cell proliferation (26). The measurement was based on the ability of viable cells to cleave the sulfonated tetrazolium salt WST-1 (4-(3-(4-iodophenyl)-2-(4-nitrophenyl)-2H-5-tetrazolio)-1,3-benzene disulfonate) by mitochondrial dehydrogenases. LoVo cells (5,000 cells/well) were plated in a 96-well plate in regular growth medium, and after 16 h the medium was replaced with 2% FBS containing medium. After 72 h, 10 μl WST-1 reagent was added in each well followed by additional incubation for 2 h. The absorbance at 450 nm was measured using a microplate reader.
Morphological detection of apoptosis
Morphological evaluation of apoptotic cell death was performed as previously described with some modification (27). Cells were fixed for 5 min in 3% paraformaldehyde in phosphate-buffered saline (PBS). After air-drying, cells were stained for 10 min in Hoechst 33258 (10 μg/ml), mounted in 50% glycerol containing 20 mM citric acid and 50 mM orthophosphate, and stored at −20°C before analysis, nuclear morphology was evaluated using a Zeiss IM 35 fluorescent.
Caspase 3, 8, 9 activity assay
Caspase 3, 8, 9 activity was determined using the ApoAlert caspase colorimetric assay kit according to the manufacturer’s instructions (Clontech, Moutain View, CA, USA) (28). Briefly, assays were performed on 96-well microtiter plates by incubating 10 μl protein of cell lysate per sample in 80 μl reaction buffer [1% NP-40, 20 mmol/l Tris-HCl (pH 7.5), 137 mmol/l Nad, and 10% glycerol] containing 10 μl caspase 3, 8, 9 substrate (Ac-DEVDpNA) (2 mmol/l). Lysates were incubated at 37°C for 4 h. Samples were measured with an enzyme-linked immunosorbent assay reader at an absorbance of 405 nm. All of experiments were performed at least 4 times.
Experimental animal model and tumorigenesis assay
Five to six-weeks-old female BALB/c nude mice were bred under pathogen-free conditions at the Southern Medical University (Guangzhou, China). All animal studies were approved by the Southern Medical University Animal Care and Use Committee. LoVo cells in exponential growth phase were harvested and washed twice in PBS. The cells were resus-pended in PBS at a density of 5×107 cells/ml and 0.1 ml (5×106 cells) of the cell suspension was then injected subcutaneously into the right flank of each nude mouse (29). Seven days after subcutaneous tumor cell injection, mice were anesthetized and 1.5×1011 v.g. of rAAV-Luc-shRNA, rAAV-FHL2-shRNA, rAAV-Luc-shRNA + 5-FU or rAAV-FHL2-shRNA + 5-FU was injected directly into the tumors at three different locations by a 10-ml micro-syringe (Hamilton, Reno, NV, USA). Four mice (4 injections) were included in each group (each transduced cell line). The volumes of tumor were calculated as follows: V=(4/3) R12R2, where R1 is radius 1 and R2 is radius 2 and R1<R2. Mice were sacrificed and tumors were dissected and weighed on day 35 after inoculation.
Statistical analysis
Data are presented as the mean ± standard error of mean (SEM). The significance of the difference between groups was evaluated with the Student’s t-test or one-way ANOVA test. P<0.05 was considered significant difference.
Results
Concentration and time kinetics of rAAV transduction in LoVo cell
LoVo cells were infected with rAAV-EGFP at different multiplicity of infection (MOIs) ranging from 1×104 to 5×105. Forty-eight hours after infection, expression of EGFP was analyzed for by flow cytometry (30). At a MOI of 1×104, an average of 13.6% of cells (SEM= 0.2%) was EGFP positive. At a MOI of 1×105, a significant increase of positive cells up to 40.1% (SEM=1.2%) was observed. At a MOI of 5×105, a slight but insignificant increase of EGPF cells was detected (53.7%; SEM=4.0%; Fig. 1A, one-way ANOVA analysis). As shown in Fig. 1B, FHL2 expressed at high level in LoVo cells, while with different dosage rAAVFHL2-shRNA transduction, FHL2 expression decreased dose-dependently.
With this efficient gene interference by rAAV, FHL2 was specifically and highly inhibited. No significance decrease in 5×105 titer, suggested 1×105 as the most appropriate MOI in the future experiments.
rAAV-FHL2-shRNA induces G0/G1 cell cycle arrest
To further address effect of FHL2 on the cell cycle, we employed rAAV vectors expressing either FHL2-shRNA or Luc-shRNA to infect LoVo cells at a dose of 1×105 MOI. Flow cytometry revealed that rAAV-FHL2-shRNA transduction led to significant G0/G1 phase accumulation in LoVo cells (Fig. 2A).
The cyclin-dependent kinases together with the cyclin D proteins are specifically involved in the progression of cells through the G1 phase and the entry into the S phase. Here, the protein expression of cyclin D1 and CDK6 was analyzed in total protein lysates after transduced with FHL2-shRNA. Cyclin D1 and CDK6 were decreased in rAAV-FHL2-shRNA compared with rAAV-Luc-shRNA (Fig. 2B).
Ectopic expression of FHL2-shRNA suppresses cell growth
WST-1 assay is commonly used to estimate cell survival, growth, and differentiation (26). The cells with rAAV-Luc-shRNA was considered as control, cultured in 0.4, 2.0 and 10% FBS for 72 h. The OD 450 values of control group were 0.672±0.031, 1.284±0.162 and 2.236±0.091; whereas those of rAAV-FHL2-shRNA were 0.482±0.083, 0.662±0.028 and 0.826±0.209, respectively (Fig. 2C). Ectopic expression of rAAV-FHL2-shRNA suppressed cell growth in a serum-dependent manner (P<0.05).
We assessed the proliferation of rAAV-FHL2-shRNA cells cultured in complete medium (5% FBS) for various time points. The OD 450 of rAAV-Luc-shRNA were 0.522±0.022, 0.986±0.062 and 1.456±0.152% with 24, 48 and 72 h culture individually, whereas those of rAAV-FHL2-shRNA cells were inhibited as 0.496±0.124, 0.562±0.041 and 0.62±0.084%, respectively (Fig. 2D, P<0.05).
rAAV-FHL2-shRNA activates intrinsic and extrinsic apoptotic pathways
To investigate the mechanism of rAAV-FHL2-shRNA induced growth suppression, apoptosis was assayed by flow cytometry. As shown in Fig. 3A, rAAV-FHL2-shRNA induced more apoptosis than rAAV-Luc-shRNA in LoVo cells, indicating that rAAV-FHL2-shRNA inhibits cell growth by inducing apoptosis.
Caspases are essential in cells for apoptosis and have been termed ‘executioner’ proteins for their roles (28,31–33). rAAV-FHL2-shRNA activated caspases 3, 8 and 9 as evident by the increasing protein level of cleaved caspases (Fig. 3B and C). Moreover, the activity of caspases 3, 8 and 9 was analyzed after rAAV-shRNA-FHL2 transduction. The ratio between OD 405 nm of transfected cells and OD 405 nm of parental cells was calculated. The activity of caspases 3, 8 and 9 was significantly increased in rAAV-FHL2-shRNA compared with rAAV-Luc-shRNA (Fig. 3D).
Suppression of rAAV-FHL2-shRNA increases cell susceptibility to apoptotic stimuli
To evaluate the role of rAAV-FHL2-shRNA in chemotherapy-induced apoptosis, the cells were transduced with rAAV, treated or not with 5-FU (50 μg/ml in NS) (34), apoptotic morphological changes was analyzed by staining with Hoechst 33258 blue fluorescence; a brightly blue-fluorescent condensed nuclei and chromatin fragmentation was considered as apoptosis by fluorescence microscopy (Fig. 4A). Similarly, the ratios of condensed nuclei positive cells were higher in rAAV-FHL2-shRNA + 5-FU of LoVo cells comparing with the rAAV-Luc-shRNA and rAAV-Luc-shRNA + 5-FU controls (P<0.05 comparing with rAAV-Luc-shRNA control) (Fig. 4B). Moreover, rAAV-Luc-shRNA and rAAV-FHL2-shRNA cells were treated with 5-FU for 48 h (NS was used as vehicle), double stained with Annexin V-FITC and PI, followed by flow cytometry analysis to determine the apoptosis. As shown in Fig. 4C, the apoptotic index of rAAV-Luc-shRNA + 5-FU and rAAV-FHL2-shRNA + 5-FU was significantly increased relative to rAAV-Luc-shRNA controls.
These findings suggested that rAAV-FHL2-shRNA enhanced the susceptibility of cancer cells to apoptotic triggers induced by 5-FU.
In vitro transduction of rAAV-FHL2-shRNA inhibits tumor formation in vivo
Moreover, the antitumor effect of rAAV-FHL2-shRNA and/or 5-FU in vivo was demonstrated by a xenograft model in nude mice. LoVo cells were subcutaneously injected in the right flanks of the nude mice. When the tumor nodules became visible (about 3–5 mm in diameter), rAAV-Luc-shRNA or rAAV-FHL2-shRNA was injected directly into the tumor and 5-FU was intraperitoneally injected as a co-treatment. The tumor sizes were continuously monitored on a weekly basis. As shown in Fig. 5, the tumor volumes of the rAAV-FHL2-shRNA and rAAV-FHL2-shRNA + 5-FU treated mice were markedly smaller than those of the rAAV-Luc-shRNA treated mice 4 weeks after rAAV injection. These data suggest that rAAV-shRNA-FHL2-transduction with 5-FU treatment is sufficient to suppress tumorigenesis.
Discussion
Identification of the differences in the genetics of cancer cells with normal cells and exploration of cancer-associated genes are important for the development of targeted therapies in cancer treatment. Nude mice have been used extensively in cancer research because they do not reject allografts and often do not reject xenografts, making them an invaluable animal model to assess the therapeutic effect of novel molecules before human clinical trials. Here, we have demonstrated the anti-tumor efficacy of FHL2 inhibition by rAAV-shRNA in cells and nude mouse xenograft models through inducing cell cycle arrest and apoptosis, yielding that rAAV-shRNA-FHL2 might be a novel and potent therapeutic or 5-FU co-therapeutic agent for colon cancer.
As shown in our previous study, FHL2 was an oncogene in gastrointestinal cancers and inhibited cell differentiation and induced tumorigenesis (19), methods targeting FHL2 might be a promising strategy for the treatment of GI cancers. AAV is a non-enveloped, single-strand DNA virus, which belongs to the family Parvoviridae. AAV was used as our gene delivery vector because it offered several advantages over other delivery methods. First, it can infect a wide range of host cells irrespective of their cell cycle stages. Second, it mediates long-term gene expression. Third, when coupled to a strong promoter like the hybrid CMV enhancer/chicken β-actin (CAG) promoter used in this study, it is capable of delivering high levels of transgene expression in a wide variety of cell types. For these reasons, AAV carrying therapeutic genes like VEGF-Trap, AAV-hTERT-TRAIL and endostatin have been employed in the treatment of malignant glioblastoma, hepatic carcinoma and pancreatic cancer with varying degree of success.
Before in vivo study, the antitumor effect of rAAV-FHL2-shRNA was evaluated in vitro. Two key classes of regulatory molecules, cyclins and cyclin-dependent kinases (CDKs), determine a cell’s progress through the cell cycle. Consistent with studies from other groups (17,23), we found that rAAV-FHL2-shRNA inhibited cell cycle progression and caused significant G0/G1 arrest by inhibiting cyclin D1 and CDK6 in colon cancer cells (Fig. 2A and B); rAAV-FHL2-shRNA induced G0/G1 arrest was also contributed to the inhibition of cell proliferation time- and serum-dependently (Fig. 2C and D). Next, rAAV-FHL2-shRNA induced apoptosis was revealed by western blot analysis and caspase activity assay (Fig. 3). In general, there are two well-characteristic pathways of apoptosis that are termed ‘extrinsic pathway’ and ‘intrinsic pathway’. The former is triggered by the interaction between the membrane death receptors such as Fas, TNF-RI, DR3, DR4, DR5 and caspase 8 and their respective ligands. Regarding the intrinsic pathway, also known as the mitochondrial pathway, involves the increase of the mitochondrial permeability and the release of apoptogenic molecules such as cytochrome c and Smac/DIABLO from mitochondria into cytosol, resulting in the activation of pro-caspase 9 and the downstream caspases. Activity of caspases 3, 8 and 9 were significantly higher in rAAV-FHL2-shRNA group, indicating that apoptosis induced by FHL2 inhibition was through both intrinsic and extrinsic pathway (31–33).
The effect of rAAV-FHL2-shRNA with or without chemo-therapy 5-FU treatment on colon cancer was investigated in vitro and in vivo (Figs. 4 and 5) (29,34). Although few studies have shown the antitumor effect of FHL2 inhibition in different cancers, to the best of our knowledge there are no reports regarding therapeutic value of rAAV-FHL2-shRNA in colon cancer. Inhibition of FHL2 by rAAV-FHL2-shRNA, as a long-term effective and specific targeting method, inhibited colon tumorigenesis and enhanced 5-FU-induced apoptosis in vitro. In vivo, it inhibited xenograft tumorigenesis and resulted in near eradication of established colon cancer xeno-graft when combined with 5-FU.
In conclusion, we have demonstrated that suppression of FHL2 by rAAV-FHL2-shRNA decreases colon cancer cell growth and cell cycle in vitro, and hinder tumor progression and outright regression in combination with 5-FU in vivo. Therefore, rAAV-FHL2-shRNA is potentially an important molecular target for the design of novel anti-colon cancer therapy.