ANTICANCER RESEARCH 28: 1993-2000 (2008)
Evaluating the Function of Matriptase and N-Acetylglucosaminyltransferase V in Prostate Cancer Metastasis KE-HUNG TSUI1,2,3, PHEI-LANG CHANG1,3, TSUI-HSIA FENG4, LI-CHUAN CHUNG5, HSIN-CHING SUNG5 and HORNG-HENG JUANG2,3,5
1Department
of Urology, 2Molecular Image Center and 3Bioinformation Center, Chang Gung Memorial Hospital, Kwei-Shan, Tao-Yuan; 4School of Nursing and 5Department of Anatomy, Chang Gung University, Kwei-Shan, Tao-Yuan, Taiwan, R.O.C.
Abstract. Previous studies have identified a subclone cell line (PC-J) which was isolated from a metastatic human prostate cell line, PC-3. In vitro matrigel invasion assays and xenograft animal studies suggested that matriptase was a putative metastatic gene in human prostate carcinoma cells. Although low metastatic prostate tumor cells, LNCaP, also expressed high levels of matriptase mRNA, gelatin zymography indicated that LNCaP cells had extremely low matriptase activity. Further studies using RT-PCR and lectin blotting assays revealed that the expression of N-acetylglucosaminyltransferase V (MGAT5), a glycoprotein that stabilizes matriptase, was low in LNCaP cells compared to PC-3 and PC-J cells. The transient overexpression of MGAT5 significantly enhanced the activity of matriptase and the invasion ability in the LNCaP cells. Knock-down of MGAT5 in PC-3 cells attenuated the metastatic ability of the cells, as determined by the in vitro invasion assay and the xenograft animal studies. Matriptase and MGAT5 may play important role in the metastasis of prostate cancer.
Metastasis is a complex process involving multiple dysregulation of gene expression that results in the transformation of tumor cells to productive metastasis. Nonetheless, the precise mechanisms of prostate tumor metastasis remain ill-defined. Previous studies using the novel PC-J cell line, a subclone from the PC-3 metastatic human prostate cell line, have demonstrated that the bikunin-matriptase system could be involved in prostate tumor metastasis (1, 2). The novelty of the PC-J cell line, Correspondence to: Dr. Horng-Heng Juang, Department of Anatomy, Chang Gung University, Kwei-Shan, Tao-Yuan, Taiwan, R.O.C. Fax: +886 3 2118112, e-mail:
[email protected]
Key Words: Prostate, metastasis, PC-3, proliferation, tumorigenicity, matriptase, N-acetylglucosaminyltransferase V.
0250-7005/2008 $2.00+.40
making it so suitable for study, is that it has high tumorigenesis, but low metastasis both in vitro and in vivo. Although a previous study has indicated that N-acetylglucosaminyltransferase V (MGAT5) stabilizes matriptase activity by adding β1-6 GlcNac branching to the N-glycan of matriptase (3), no significant difference in the gene expression of MGAT5 was found between PC-3 and PC-J cells (1). Several reports have indicated that MGAT5 might facilitate angiogenesis and malignant transformation (4, 5), and knockout of MGAT5 suppressed tumor growth and metastasis in animal studies (6). Although an earlier study has indicated that increased MGAT5 activity enhanced the metastatic potential of rat prostate carcinoma cells (7), currently, there is no data concerning the function of MGAT5 in the human prostate. β1,6-N-Acetylglucosaminyltransferase V (EC 2.4.1.155) catalyzes the addition of β1,6-GlcNAc to N-glycan intermediates found on newly synthesized glycoproteins transiting the medial Golgi apparatus, is encoded by the MGAT5 gene (8). The cDNA for human MGAT5 has been cloned from a fetal liver cDNA library. This gene encodes a 741-amino acid polypeptide (9). An earlier study has indicated that malignant transformation was accompanied by increased β1,6-GlcNAc branching of N-glycans attached to Asn-X-Ser/Thr sequences in mature glycoproteins, and the level of MGAT5 products was correlated with metastasis progression (10). Studies from MGAT5deficient mice have suggested that inhibitors of MGAT5 may be useful in the treatment of malignancies by targeting their dependency on focal adhesion signaling for proliferation (6, 11). Other studies have also indicated that the expression of MGAT5 could correlate with metastasis of colorectal cancer, non-small cell lung cancer, breast cancer, bladder cancer, gastric cancer, fibrosarcoma, gliomas and hepatocellular carcinoma both in vitro and in vivo (3, 12-17). The objectives of this study were to evaluate the function of matriptase and MGAT5 in human prostate cancer metastasis. 1993
Materials and Methods
ANTICANCER RESEARCH 28: 1993-2000 (2008)
Cell culture and materials. The PZ-HPV-7, CA-HPV-10, LNCaP, PC-3, and DU145 cell lines were obtained from the American Type Culture Collection (ATCC; Manassas, VA, USA) and were maintained as described previously (18). The PC-J cell, a subculture cell line of PC-3 cells, was isolated in our laboratory (1). Fetal calf serum (FCS) was purchased from HyClone (Logan, UT, USA), and RPMI 1640 media were purchased from Life Technologies (Rockville, MD, USA). Matrigel was purchased from BD Biosciences (Bedford, MA, USA).
Reverse transcription-polymerase chain reaction. The total RNA from the different cells was isolated with Trizol reagent and the cDNA was synthesized using Superscript III Preamplification, as described previously (19). The primers for MGAT5 (5’GACCTGCAGTTCCTTCTTCG-3’ and 5’-AGAGCAGGAGGTC ACCTTG-3’) and matriptase (5’-AAACAGTACAGAGGACC CAGGA-3’ and 5’-AGGAGT GGAAGGTCAGGTTGTA-3’) were used for the amplification of sequences specific to human MGAT5 and matriptase mRNAs. The quality of the cDNA was verified by performing controlled reactions using primers derived from β-actin (5’-GAAGATCAAGATCATTGCTCCTCC-3’ and 5’-CTGTGCT CAAGTCAGTGTACA GG-3’). The PCR reaction was performed in a thermal cycler (T3000 Thermocycler, Biometra, Goettingen, Germany) under the following parameters: 30 cycles of 94˚C for 0.5 min, 55˚C for 1 min, and 72˚C for 1 min. The PCR products were separated by 2% agarose gel electrophoresis and visualized with ethidium bromide staining.
Gelatin zymography. The cells were plated in a 24-well plate and grown in serum-containing medium for 24 h and then changed to a serum-free medium for another 24 h. The media were collected and centrifuged at 300 × g for 10 min to remove cell debris. A gelatin (2 mg/ml) substrate was copolymerized with regular SDS-PAGE and 50 μl of medium was then subjected to electrophoresis at a constant current of 15 mA. The gels were washed three times with 50 mM TrisHCl (pH 7.5) containing 2.5% Triton X-100 and then incubated in 20 mM Tris-HCl (pH 7.5) buffer containing 5 mM CaCl2 at 37˚C overnight. The gels were stained with Coomassie Blue.
Immunoblotting of human matriptase and β-actin. Immunoblotting of human matriptase and β-actin was conduced as follows. Equal amounts of protein (30 μg) were loaded onto a 10% SDS-polyacrylamide gel. After electrophoresis, the blot was transferred to a membrane which was probed with 1:3000 diluted polyclonal matriptase antibody (Calbiochem, Dan Diego, CA, USA) or 1:1000 diluted anti-β-actin antibody (C11, Santa Cruz Biotechnology, Santa Cruz, CA, USA). The membrane was analyzed using an electrochemiluminescence (ECL) detection system, as described by the manufacturer (Amersham Biosciences, New Territories, Hong Kong).
Transient overexpression MGAT5 LNCaP cells. Liver Marathonready cDNA was purchased from BD Biosciences (Cat N0. 74071). The PCR reaction was performed with 2 μl of cDNA and with primers (P1 5’-AGAGCCAAGGACAGGTGAAGTTG-3’ and R2 5’-CTATAGGCAGTCTTTGCAGAGAG-3’). The PCR products were ligated into the pGEM-Teasy vector (Promega Biosciences, San Luis Obispo, CA, USA). This cDNA fragment was then cloned into the pcDNA3 expression vector (Invitrogen, Carlsbad,
1994
CA, USA) by ligating at the Eco RI site, resulting in the insertion of the full coding region of MGAT5 cDNA into the polyadenylate region controlled by the CMV promoter (pcDNA3MGAT5). Transient overexpression of MGAT5 by LNCaP cells was achieved by transfecting the human MGAT5 overexpression vector into the LNCaP cells using electroporation. Ten million cells were suspended in 20 μg of pcDNA3MGAT5 and 500 μl of RPMI medium without serum in a 4-mm gap cuvette. The electroporation was conduced using the ECM 830 (BTX, San Diego, CA), set at 170 V, 70-msec pulse length, and one pulse setting. The mock-transfection LNCaP cells (LN-DNA) were transfected with control pcDNA3 expression vector in the same manner as the LN-MGAT5 cells.
Bikunin overexpression and MGAT5 knock-down PC-3 cells. The PC-3 cells were transfected with the pcDNA3-bikunin overexpression vector and selected using geniticin, as described previously (2). The MGAT5 shRNA vector (V2HS_151753) and the control vector, shEGFP vector, were purchased from Open Biosystems (Huntsville, AL, USA). The interference RNA targeted the 4881-4899 site of MGAT5 cDNA (NM_002410). The expression vector was introduced into the PC-3 cells by electroporation. Ten million cells were suspended in 20 μg of shRNA vector and 500 μl of RPMI medium without serum in a 4mm gap cuvette. The electroporation was conduced using the ECM 830 (BTX), set at 180 V, 70-msec pulse length, and one pulse setting. The cells were maintained in RPMI medium with 10% FCS and 2 μg/ml of puromycin for 48 h after electroporation.
Lectin blotting. The cells were homogenized in TNE buffer (10 mM Tris-HCl [pH 7.8], 1% NP-40, 0.15 M NaCl, 1 mM EDTA and protease inhibitors) for 30 min on ice. Next, 20 μg of the protein extracted from the cells was separated on a 10% polyacrylamide gel, and then transferred onto a polyvinylidene difluoride (PVDF) transfer membrane (PerkinElmer, Boston, MA, USA). After blocking, the membrane was incubated with 1 μg/ml of biotinlabeled lectins (EY Laboratories Inc., San Mateo, CA, USA) at room temperature for 1 h. After washing three times for 10 min with Tris-buffered saline containing 0.1% Tween 20, the membrane was incubated with 1/1000 diluted horseradish peroxidase Avidin D (Vector Laboratories Ltd., Burlingame, CA, USA) for 1 h and then washed three times with Tris-buffered saline containing 0.1% Tween 20 for 10 min. The membranes were stained with immunoblot detection reagents (Perkin Elmer).
Matrigel invasion assay. One hundred and fifty μl of cold matrigel (800 μg/ml in serum free cell culture media) was added to the upper chamber of a 24-well transwell plate and incubated at 37˚C for 4 h for gelling. The coated-membrane (8 μm pore size, Millicell-PCF, Millipore, Billerica, MA, USA) was rehydrated with 200 μl of serum free RPMI1640 medium. Five hundred microliters of RPMI 1640 medium with 10% FCS were added to the lower chamber of the 24well plate. The prostate carcinoma cells at a density of 5×105 cells/ml were cultured in 200 μl serum free RPMI1640 medium in the upper well. The plates were incubated in a 37˚C incubator equilibrated with 5% CO2 for 48 h. The cells which migrated to opposite side of the membrane were fixed with 4% paraformaldehyde (0.4% NaOH, 4% paraformaldehyde and 3.8% Borax pH 9.5) then stained with 0.1% crystal violet solution for 30 min. The transmembrane migrated cells were digitally photographed under a microscopy (IX71, Olympus,
Tsui et al: Matriptase and N-Acetylglucosaminyltransferase V
Figure 1. Expression of matriptase and MGAT5 in prostate carcinoma cells. Total RNA was extracted from the cells for the RT-PCR assay (A, C) and the cells were lysed to obtain the corresponding proteins for the immunoblot assay (B). The matriptase activity was determined by gelatin zymography. The cDNA quality and protein levels of β-actin were regarded as the internal controls (PZ, PZ-HPV-7; CA, CA-HPV10; LN, LNCaP; PC, PC-3; DU, DU145).
Tokyo, Japan). The membrane was soaked in 10% acetic acid with agitation at 37˚C for 1 h. The acetic acid solution was then read by spectrophotometer at 635 nm (DU640, Beckman Instruments, Fullerton, CA, USA). The experiments were performed in triplicate and repeated at least three times.
Xenograft animal model. The cells were detached from the cell-flask by treatment with 0.5 μM EDTA in PBS solution, washed with RPMI1640 medium with 10% FCS and resuspended in PBS solution. Male nude mice (BALB/cAnN-Foxn1, 4 weeks of age) were purchased from the animal center of the National Science Council in Taiwan. The PC-3 cells that were stably transfected with the bikunin overexpression vector (PC-BIK), MGAT5 knock-down vector (PCMGAT5si), or the EGFP control vector (PC-EGFPsi) were used for the animal studies. The mice were anesthetized intraperitoneal injection (50 μl/kg of body weight) with ketamine (100 μg/ml)xylazine (10 μg/ml). Ten million cells in 5 μl PBS were injected into the prostate. After 8 weeks the prostate-injected nude mice were dissected to determine prostate tumor size. Tumors found in the pelvic lymph nodes of the mice were considered metastatic. These studies were performed in accordance with the United State National Institutes of Health Guide for the Care and Use of Laboratory Animals and were approved by the Chang Gung University Animal Research Committee. In all the procedures, every effort has been made to minimize the suffering of the laboratory animals and the number of animals used.
Statistical analyses. The results are expressed as means±standard errors (SEs) of at least three independent replications of each experiment. Statistical significance was determined using Student’s paired t-test analysis, using the SigmaStat program for Windows, version 2.03 (SPSS Inc, Chicago, IL, USA).
Results
The RT-PCR and immunoblotting assays showed that the LNCaP cells expressed high levels of matriptase (Figure 1A, B). However, the zymography assays revealed that the PC-J and LNCaP cells had extremely low matriptase enzymatic activity in comparison to the PC-3 cells (Figure 1B). Additional RT-PCR assays revealed that the LNCaP cells expressed low levels of the MGAT5 gene, compared with the PC-J and PC-3 cells (Figure 1C). The expression of MGAT5 in the MGAT5 transfected LNCaP cells (LN-MGAT5) was higher than in the mocktransfected LNCaP cells (LN-DNA) as shown by RT-PCR assay; moreover, zymography assay revealed that the LNMGAT5 cells had very much higher matriptase enzymatic activity in comparison to the LN-DNA cells (Figure 2A). 1995
ANTICANCER RESEARCH 28: 1993-2000 (2008)
Figure 2. LNCaP cells after transient-overexpression of MGAT5. (A). Expression of MGAT5 in LNCP cells after transient transfection with the MGAT5 overexpression vector determined by RT-PCR. The matriptase activity was determined by gelatin zymography assay (B). The invasion ability of mock-transfected LNCaP cells (LN-DNA) and MGAT5transfected LNCaP cells (LN-MGAT5) determined by in vitro matrigel invasion assay. The transmembrane migrated cells were recorded microscopically with a digital camera (top). Experiments were performed in triplicate and repeated at least three times. Mean percentage (±SE) of the OD in relation to that of LN-DNA cell group (bottom). (**p
