ANTICANCER RESEARCH 27: 2325-2330 (2007)
Effect of CDT6 on Factors of Angiogenic Balance in Tumour Cell Lines DIANE R.M.H. BOU´S, WENDY A. DAM, COBY MEIJER, NANNO H. MULDER and GEKE A.P. HOSPERS
Department of Medical Oncology, University Medical Centre Groningen and University of Groningen, Groningen, The Netherlands
Abstract. Background: Cornea-derived transcript 6 (CDT6, also known as AngX) is an angiopoietin-related factor resulting in anti-tumour effect in vivo. However, a recent abstract reported that CDT6 can also induce angiogenesis and promotes tumour growth. In our previous work, CDT6 had failed to show pro- or anti-angiogenic effects. It is unknown if CDT6 expression occurs in human cancer. Materials and Methods: An array of human tumour cell lines and tumour tissues was tested for CDT6-gene expression using RT-PCR. To address the controversial role of CDT6 on angiogenesis in different tumour models, the expression levels of four factors of the angiogenic balance (VEGF, endostatin, TIMP-1 and PAI-1) were determined in CDT6-transfected and control cells of the human and murine melanoma cell lines BLM and B16F10. Endostatin was significantly up-regulated by CDT6 expression in the human model and significantly downregulated in the mouse model. None of 18 cell lines or 23 tumours expressed CDT6. Conclusion: This contradictory effect on endostatin expression in human and mouse models may be an explanation for the conflicting results for the effect of CDT6 expression on angiogenesis. The outgrowth of new blood vessels from pre-existing vasculature, angiogenesis, is an essential process in embryogenesis and wound healing but also plays a major role in several pathological processes such as tumour vascularisation, diabetic retinopathy, psoriasis and rheumatoid arthritis. The angiogenic process is tightly regulated by a multitude of pro- and anti-angiogenic factors, constituting the angiogenic balance. The cornea is a proliferating tissue constantly subjected to external factors such as UV light, yet no tumours of
Correspondence to: G.A.P. Hospers, MD, Ph.D., Department of Medical Oncology, University Medical Centre Groningen, P.O. Box 30.001, 9700 RB Groningen, The Netherlands. Tel: +31 50 3612821, Fax: +31 50 3614862, e-mail:
[email protected] Key Words: CDT6, endostatin, PAI, VEGF, TIMP, angiogenic balance.
0250-7005/2007 $2.00+.40
corneal origin are known. Furthermore, the cornea is normally avascular. It could be hypothesised that a powerful anti-angiogenis stimulus dominates this tissue. This prompted Peek et al. (1) to search for a cornea-specific antiangiogenic factor, a search which led to the description of cornea-derived transcript 6 (CDT6). The gene structure of CDT6 is reminiscent of the angiopoietins (Ang), a family of potent pro- and anti-angiogenic modulators. Investigations of CDT6 as an anti-angiogenic factor has yielded quite controversial results: Peek et al. (2) reported substantial tumour growth inhibition through deposition of large amounts of collagen by the gene product. They could not clearly relate this effect to anti-angiogenesis and concluded that CDT6 was a morphogen for the cornea (2). However, a later study claimed a strongly pro-angiogenic activity of CDT6 in vitro as well as increased tumour growth in vivo (3). We found a neutral effect of the gene and its product on endothelial proliferation in vitro and on tumour growth in an immune-competent murine tumour model in vivo (4). These conflicting results could imply that CDT6 influences angiogenesis not directly, but rather in a way facilitated by other angiogenic factors in the microenvironment. Differences in the concentration of these factors could explain these different results. We therefore tested a panel of human tumour cell lines as well as patient tumour material for the expression of CDT6. Furthermore, the effect of CDT6-expression on the expression of pro- and anti-angiogenic factors was investigated, comparing our own tumour cell line model to the one described by Peek et al. (2).
Materials and Methods Cell lines. The cell lines examined for CDT6-expression were the following: Cervical carcinoma cell lines SiHa (5), CaSki (6), C33A (7) and HeLa (8); Colon carcinoma cell lines Caco-2 (9), Colo320 (10) and SW948 (11); Testis carcinoma cell lines Tera (NT2/D1) (12), Tera-CP (13), 833KE (14) and Scha (15); Ovarian carcinoma cell lines A2780 (16) and its induced doxorubicin-resistant subline A2780-Adr (17); Small cell lung carcinoma cell lines GLC2 (18), GLC3 (18), GLC14 (19) and the induced doxorubicin- and cisplatin-resistant sublines GLC4-Adria (20), GLC4-CDDP (21).
2325
ANTICANCER RESEARCH 27: 2325-2330 (2007) The two tumour cell line models used for detection of factors of the angiogenic balance were: The human melanoma cell line BLM (22) and its transfected sublines BLM-CMV-6C (empty-vector transfected), BLM-CDT6-3E (CDT6-transfected) (2); the murine melanoma cell line B16-F10 (23) and its sublines B16-CMV (empty-vector transfected) and B16-CDT6 (CDT6-transfected) (24). All cell lines were cultured according to their references with media from Invitrogen, (Merelbeke, Belgium) and foetal calf serum (FCS) from Bodinco (Alkmaar, The Netherlands). For passage, the cell lines growing in monolayers were either scraped (Tera(NT2/D1), Tera-CP, Scha, A2780, A2780-Adr and GLC cell lines) or washed with phosphate-buffered saline (PBS; 136.0 mM NaCl, 2.7 mM KCl, 6.4 mM Na2HPO4, 1.5 mM KH2PO4, pH 7.4 all from Merck (Amsterdam, The Netherlands)) and released from culture flasks with 0.125% trypsin (Invitrogen, Merelbeke, Belgium) in PBS, 0.5 mM EDTA (SiHa, CaSki, C33A and HeLa as well as BLM and B16-F10 and their sublines) or 0.005% Protease (Sigma, St. Louis, MO, USA) in PBS for Colo320 and 833KE or 0.01% in PBS for Caco-2 and SW948. Human tumour material. Tumour samples were collected at the time of surgery from patients who had previously given informed consent that their material could be used for research. The entire patient material used in this report had been collected for other studies between 1992 and 1998 in our department and was used in the various publications (25-28). There were 15 lung tumours, of which five were non-small cell lung carcinomas (NSCLC), four planocellular, four adenocarcinomas and two squamous tumours. There were also six colon carcinomas, three cervical tumours, of which two were squamous, and three ovarian cancers as well as two ductal breast cancers, a squamous uterine tumour and a squamous tumour of the endometrium. Tumours were immediately snapfrozen in liquid nitrogen after surgery and stored at –80ÆC until pulverisation and RNA-isolation. RNA-isolation. The tumour tissue was pulverised (MikroDismembrator U, Braun Biotech International, Germany) and dissolved in 500 Ìl guanidine-isothiocyanate (GT)-buffer (4M guanidine-isothiocyanate, 0.5% n-lauroyl sarcosine (both Invitrogen), 25 mM sodium citrate (pH 7 Merck), 0.1 M ‚-mercaptoethanol (Sigma). The cell lines were released from their culture flasks as described above, centrifuged and the resulting cell pellet was dissolved in 500 Ìl GT-buffer as for the tumour samples. The RNA isolation was performed as described by Meersma et al. (29). The resulting RNA-pellet was dissolved in 20 Ìl of diethylpyrocarbonate (DEPC, Sigma, St. Louis, MO, USA)-treated water. RNA concentration was determined spectrophotometrically and its quality was monitored with gel electrophoresis. The RNA samples were stored at –80ÆC until further analysis. RT-PCR (reverse-transcriptase polymerase chain reaction). The isolated RNA was treated with DNAse I (RNAse-free) and RNAse inhibitor (both Invitrogen) followed by cDNA synthesis from 4 Ìg RNA as described by the manufacturers protocol using T11VN oligonucleotides, M-MLV reverse transcriptase, 1st strand buffer and DTT (Invitrogen). The resulting cDNA was stored at 4ÆC and used in a PCR for CDT6 and using the mouse housekeeper gene ‚actin as a control for the murine B16 cell lines and the human
2326
housekeeper gene glyceraldehyde-3-phosphatase dehydrogenase (GAPDH) for the human cell lines and tumour samples. The primers and conditions were as follows: human CDT6: (oligo15.515.6) 5'-ggacttctggctggggaacg-3', 5'-ccagtagccacctttgcgga-3', with 2 mM MgCl2, 63ÆC annealing temperature and 25 cycles; human GAPDH: 5’-caccaccatggagaaggccgg-3’, 5’-ccaaagttgtcatggatgacc-3’ with 1.5 mM MgCl2, 63ÆC annealing temperature and 25 cycles; mouse ‚-actin (conditions and primers kindly provided by Dr. G. de Haan, Department of Stem Cell Biology, Groningen): 5'-ggccca gagcaagagaggtat -3', 5'-ctaggagccagagcagtaatc-3', with 2 mM MgCl2, 63ÆC annealing temperature and 25 cycles. Taq polymerase, 50mM MgCl2 and PCR-buffer were all from Invitrogen. The presence of a PCR-product was monitored using agarose gel electrophoresis. ELISA (Enzyme-linked immunosorbant assay). To test for the expression of factors of the angiogenic balance the cell lines of the two models (BLM-CMV-6C and BLM-CDT6-1E for the human model and B16-CMV and B16-CDT6 for the murine model) were cultured until sub-confluency. They were then washed with PBS and incubated 18-24 h with medium without FCS. The supernatant was harvested and rid of contaminating cells by centrifugation or passage through a 0.2 Ìm filter (Corning Life Sciences, SchipholRijk, The Netherlands) and frozen at –80ÆC until ELISA analysis. Cells were released from the culture flasks and lysed by freezing in PBS and resuspending vigorously. The total protein concentration of the cell lysate was determined according to Bradford (30). Supernatants and cells of four independent passages of each cell line were harvested and tested. The culture supernatant concentrations of human and murine VEGF and endostatin were determined with human and murine ELISA kits (Quantikine, R&D, Abingdon, UK). Human tissue inhibitor of metalloprotease (TIMP) and plasminogen activator inhibitor (PAI-1) were measured using human ELISA kits (Quantikine and Asserachrom, Stago, Parippany, NJ, USA). The measured supernatant concentrations were related to the total protein level of the cell lysates. Statistical analysis. A two-sided Student’s t-test with unequal variances assumed was performed comparing CDT6-expressing to empty vector-transfected control cells.
Results A panel of 18 human tumour cell lines and 31 human tumour samples was tested for CDT6 expression using RTPCR. Of the 31 tumours, eight showed minimal or no expression of the housekeeper gene GAPDH and therefore could not be evaluated. Figure 1 shows that none of the evaluable 23 tumours or cell lines expressed CDT6. The CDT6-transfected and empty vector-transfected sublines of the human melanoma cell line BLM (BLMCDT6 and BLM-CMV) as well as the murine melanoma cell line B16-F10 (B16-CDT6 and B16-CMV) were tested for expression of human or murine VEGF, endostatin, TIMP and PAI (Table I). CDT6 expression did not lead to significant changes in VEGF secretion in either of the two models. Still, a tendency towards reduced VEGF expression was observed
Bouïs et al: Effect of CDT6 on Angiogenic Balance Factors
Figure 1. Agarose-gel analysis of RT-PCR-products of (A) 31 human tumour samples and (B) 18 cell lines as well as CDT6- and empty vector-transfected sublines of the human BLM and the murine B16 cell lines. Top panels: human GAPDH (exception: last two lanes on the cell line panel: murine ‚actin), lower panels: human CDT6. U: uterus, E: endometrium, H: negative control, CDT6: B16-CDT6, CMV: B16-CMV, 3E: BLM-CDT6-3E, 6C: BLM-CMV-6C.
Table I. Culture supernatant concentration of human or murine VEGF, endostatin, TIMP-1 or PAI-1 per mg of total cell lysate protein.
Human cell line BLMC-CMV BLM-CDT6 Murine cell line B16-CMV B16-CDT6
VEGF pg/mg
Endostatin pg/mg
TIMP-1 ng/mg
7041 (±1473) 6184 (±1210) N.S.
47 (±8) 77 (±29) p
