Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 7-Day Trial for You or Your Team.

Learn More →

VEGF and VEGF type C play an important role in angiogenesis and lymphangiogenesis in human malignant mesothelioma tumours

VEGF and VEGF type C play an important role in angiogenesis and lymphangiogenesis in human... British Journal of Cancer (1999) 81(1), 54–61 © 1999 Cancer Research Campaign Article no. bjoc.1999.0650 VEGF and VEGF type C play an important role in angiogenesis and lymphangiogenesis in human malignant mesothelioma tumours 1,4 1 1 1 2 3 4 1 Y Ohta , V Shridhar , RK Bright , GP Kalemkerian , W Du , M Carbone , Y Watanabe and HI Pass 1 2 Aerodigestive Program and Hematology-Oncology, Barbara Ann Karmanos Cancer Institute, Wayne State University School of Medicine, Detroit, MI, USA; 3 4 Loyola Cancer Center, Maywood, IL, USA; First Department of Surgery, School of Medicine, Kanazawa University, Kanazawa, Japan Summary The vascular endothelial growth factor (VEGF) family is a novel regulator of endothelial cell proliferation. We assessed the mRNA expression of VEGF, VEGF type C (VEGF-C) and their receptors together with the microvessel density (VD) and microlymphatic vessel density (LVD) in pursuit of their connection and prognostic value in malignant pleural mesothelioma (MPM). We used four human MPM cell lines, 54 MPM tumours and five normal pleural tissues. Expression levels for receptors and ligands were assessed by semiquantitative reverse transcriptase polymerase chain reaction analysis. Microvessels were highlighted by immunohistochemical staining for factor VIII. The discrimination of lymphatics was performed by enzyme-histochemistry for 5¢-nucleotidase after adequate inhibition of non-specific activity. The expression levels of VEGF, VEGF-C and VEGFRs were high in all MPM cell lines. The percentages of tumours with higher expression compared to the mean values of normal pleural tissues were 31.5% (17/54) for VEGF, 66.7% (36/54) for VEGF-C, 20.4% (11/54) for fms-like tyrosine kinase (flt)-1, 42.6% (23/54) for kinase insert domain-containing recepter (KDR) and 59.3% (32/54) for flt-4. Significant positive correlations were found between VEGF-C and flt-4, VEGF and KDR, VEGF and flt-1 in tumour tissues. The association between LVD and VEGF-C expression level was especially strong (P < 0.0001, r = 0.63). There were also significant correlations between LVD and flt-4, and VD and VEGF. No correlation, however, was found between LVD and nodal metastasis. VD was a negative prognostic indicator in this study. The associations between VEGF/VEGF-C and vessel density suggest that these factors play an important role in angiogenesis and lymphangiogenesis in this tumour, and assessment of vascularity may be a useful prognostic indicator for MPM patients. Keywords: vascular endothelial growth factor; VEGF; VEGF-C; angiogenesis; lymphangiogenesis; malignant mesothelioma Asbestos had been widely used for centuries before a causal rela- 1996; Lee et al, 1996; Olofsson et al, 1996; Yamada et al, 1997; tionship between malignant pleural mesothelioma (MPM) and Meyer et al, 1999). Interestingly, the function of VEGF-C appears exposure to asbestos was confirmed. Although the use of asbestos to extend to the lymphatic system where it serves as a ligand for has recently been restricted, the incidence of MPM remains high fms-like tyrosine kinase 4 (flt-4) (Kukk et al, 1996; Jeltsch et al, and is steadily increasing due to the long latent period for MPM 1997). However, lymphangiogenesis within tumours has not been after the exposure to asbestos (Mossman et al, 1996). Advanced documented. MPM remains an aggressive and highly lethal disease due to its For MPM, little information on tumour angiogenesis is avail- marked resistance to conventional treatments, including surgical able. In this study, we assessed the expression of VEGF, VEGF-C resection, chemotherapy and radiotherapy. and VEGFRs (KDR, flt-1 and flt-4) in human MPM cell lines, Angiogenesis is crucial for the proliferation of tumour cells, and MPM tumour samples and normal pleural tissue. We used the anti-angiogenic therapy is a promising strategy aimed at inhibiting reverse transcriptase polymerase chain reaction technique (RT- tumour growth, invasion and metastasis (Folkman et al, 1995). PCR) to assess the expression of each factor, and we also assessed Among the many reported angiogenic factors, vascular endothelial the VD and microlymphatic vessel density (LVD) within tumours growth factor (VEGF) is the most powerful endothelial cell- using immunohistochemistry for factor VIII and enzyme-histo- specific mitogen associated with tumour neovascularization. chemistry for 5¢-nucleotidase (5¢-NA) respectively. The expres- While a number of investigators have reported the direct relation- sion of the associated angiogenic factors and receptors was also ship between microvessel density (VD) and VEGF expression determined. Clinical correlations of angiogenesis with survival within a variety of tumours (Toi et al, 1994; Mattern et al, 1995; from treatments for the patients was also performed. Samoto et al, 1995; Takahashi et al, 1995), recent studies also suggest that VEGF type B, C, D and E are novel regulators of MATERIALS AND METHODS endothelial cell proliferation (Grimmond et al, 1996; Joukov et al, Cell lines and tissue samples The human malignant mesothelioma cell lines (H-meso, H2818, Received 28 October 1998 H2591 and H2595) used in this study were all developed from Accepted 16 February 1999 tumours diagnosed using a well-defined panel of immunohisto- chemical markers (Pass et al, 1995). H-meso is commercially Correspondence to: Y Ohta, First Department of Surgery, Kanazawa University, School of Medicine, Kanazawa 920-8641, Japan available (Biomeasure, Hopkinton, MA, USA), while H2818, 54 Angiogenesis in malignant mesothelioma 55 Table 1 Nucleotide sequences of the primers used and their PCR conditions. Angiogenic factors Primer sets PCR cycle Product size (bp) VEGF 5¢-GAAGTGGTGAAGTTCATGGATGTC-3¢ (sense) 30 408 (VEGF121) 5¢-CGATCGTTCTGTATCAGTCTTTCC-3¢ (antisense) 541 (VEGF165) 613 (VEGF185) VEGF-C 5¢-CATGTACGAACCGCCAG-3¢ (sense) 25 320 5¢-TTGGCTGTTTGGTCATTGGC-3¢ (antisense) Flt-1 5¢-GAGAATTCACTATGGAAGATCTGATTTCTTACAGT-3¢ (sense) 30 1098 5¢-GAGCATGCGGTAAAATACACATGTGCTTCTAG-3¢ (antisense) Flt-4 5¢-AGCCATTCATCAACAAGCCT-3¢ (sense) 25 298 5¢-GGCAACAGCTGGATGTCATA-3¢ (antisense) KDR 5¢-TATAGATGGTGTAACCCGGA-3¢ (sense) 30 555 5¢-TTTGTCACTGAGACAGCTTGG-3¢ (antisense) H2591 and H2595 were developed from resected MPM tumours at amplification. For quantitative evaluation of the amplified product, the National Cancer Institute (Bethesda, MD, USA) (Pass et al, PCR encompassing 20Ð40 cycles was preliminarily performed to 1995). A human fibrosarcoma cell line (HT1080) was used as a determine the most suitable number of amplifications for each positive control for VEGF, VEGF-C, flt-1 and flt-4, while factor. PCR was performed after adding 80 ml of PCR mixture (50 HUVEC was used for kinase insert domain-containing receptor mM TrisÐHCl, pH 8.3; 40 mM KCl; 8 mM MgCl ; 0.5 mM each (KDR). Cell lines were maintained in RPMI-1640 media (Gibco- dNTP; 50 pmol of each the sense and the antisense primer; 2.5 units BRL, Gaithersburg, MD, USA) supplemented with 10% fetal of Taq polymerase (Takara, Kyoto, Japan)). Each cycle consisted of: bovine serum (FBS), 2 mML-glutamine, 10 mM HEPES buffer, 1 min at 94°C, 1 min at 58°C and 2 min at 72°C for VEGF, flt-1 and Ð1 Ð1 500 units ml penicillin and 500 units ml streptomycin at 37°C KDR; and 1 min at 94°C, 1 min at 52°C and 1.5 min at 72°C for under a humidified atmosphere of 5% carbon dioxide. Five normal VEGF type C, flt-4 and b-actin. PCR primers used and PCR condi- pleural tissue samples were obtained from patients with non- tions are shown in Table 1. The PCR products were electrophoresed cancerous disease (either for repair of hiatus hernia or diagnosis of through a 1.0% agarose gel (4-mm thick). The gel was fluorescently pulmonary infiltrates). In all these cases, the resected pleural tissue stained in SYBR Green I stain Solution (FMC Bio-Products, was confirmed to be free of malignancy by pathological examina- Rockland, ME, USA) at a 1:10 000 dilution in 1 ´ TAE buffer tion. And 54 malignant mesothelioma tumours were randomly (pH 7.2) for 30 min. The intensity of the PCR product bands was chosen from a bank of snap-frozen samples collected by one of us measured by Stormª image analyser (Molecular Dynamics, Inc., (HIP). For tumour samples, consecutive 6-mm cryosections were Sunnyvale, CA, USA). Each expression was standardized using the cut from the invasive tumour margin for immunohistochemical b-actin signal as an internal control and a densitometry index and enzyme-histochemical studies. The pathologic classification defined. For negative control of each factor, PCR was done of each sample was confirmed by review of haematoxylin and excepting the template cDNA. eosin (H&E)-stained sections. The pathological MPM subtypes identified were: epithelial (E), 44; sarcomatous (S), 4; and mixed Enzyme-histochemistry (M), 6. According to the new international TNM staging system An enzyme-histochemical reaction for 5¢-NA was done according (Rusch, 1996), 50 cases could be classified based on pathological to the leading method of Wachstein et al (1954). Briefly, cryostat findings [stage II in nine (E, 7; S, 2), III in 36 (E, 28; S, 2; M, 6) sections were fixed with 5% paraformaldehyde for 10Ð60 min (10, and IV in five (E, 4; M, 1)]. 20, 30, 40, 50 and 60 min for each sample). This step-wise fixation was carried out in order to discover the appropriate conditions to RT-PCR analysis maximally exhibit 5¢-NA activity in microlymphatic vessels compared to that in blood vessels. Next, after washing with water Total RNA was extracted from the human mesothelioma cell lines, for 5 min, sections were treated with an enzyme reaction for 2 h control cell lines (HT1080 and HUVEC), normal pleural tissues, at 37°C using a substrate mixture (20 ml of 0.2 M TrisÐmaleate and each resected MPM tumour using TRIzol (Life Technologies, buffer (pH 7.2), 25 mg adenosine-5-monophosphate (Sigma, Inc., Grand Island, NY, USA) according to the standard acidÐguani- St Louis, MO, USA), 20 mg tetramisole (Sigma, St Louis, MO, diumÐphenolÐchloroform method. Total RNA (1 mg) was denatured USA), 3.5 mg sucrose, 5 ml 0.1 M magnesium sulphate, 3 ml of together with oligo-dT primer (10 pmol) for 15 min at 68°C. After 2% Pb(NO ) , 20 ml water). After the sections were washed with 3 2 this was chilled on ice for 5 min, poly-adenosine (poly-A) RNA was water for 15 min, they were reacted with a 1% ammonium reverse-transcribed at 42°C for 90 min in RT solution (50 mM sulphide solution (Sigma, St Louis, MO, USA) for 2 min. The TrisÐHCl (hychochloric acid), pH 8.3; 40 mM potassium chloride sections were counterstained with methyl green and mounted (KCl); 8 mM magnesium chloride (MgCl ); 0.5 mM each dNTP; using glycerol (Sigma, St Louis, MO, USA). Ð1 225 mg ml bovine serum albumin; 5 mM dithiothreitol; 20 units RNasin (Life Technologies, Inc., Grand Island, NY, USA) and Immunohistochemistry 4 units AMV reverse transcriptase (Gibco-BRL, Gaithersburg, MD, USA)). The cDNA was incubated at 95°C for 5 min to inactivate Consecutive cryostat sections were used for the immunohistochem- the reverse transcriptase, and served as the template for PCR ical assessment of VD. The primary antibodies used in this study © 1999 Cancer Research Campaign British Journal of Cancer (1999) 81(1), 54–61 56 Y Ohta et al Human mesothelioma cell lines Human mesothelioma tissues 320 bp VEGF-C 320 bp VEGF-C 541 bp VEGF 408 bp VEGF 541 bp 408 bp 298 bp flt-4 298 bp flt-4 1098 bp flt-1 1098 bp flt-1 555 bp KDR 555 bp KDR 884 bp 884 bp Actin Actin Figure 1 RT-PCR analysis of VEGF, VEGF-C and their receptors (KDR, flt-1 and flt-4) in human mesothelioma cell lines (A), normal pleural tissue (B) and resected human malignant mesothelioma tumour samples (C). High Normal pleural tissue expression of VEGF, VEGF-C and VEGFRs mRNA was recognized in human mesothelioma cell lines. Compared with the expression in the five normal pleural tissues, VEGF-C and flt-4 mRNA expression levels were greater in tumour tissues 320 bp VEGF-C were reacted with primary antibody for 2 h, then washed with PBSÐ and reacted with biotin-labelled anti-rabbit immunoglobulin 541 bp VEGF 408 bp (Vector Laboratories, Burlingame, CA, USA) for 30 min at room temperature. After sections were washed with PBSÐ, avidinÐbiotinÐperoxidase complex was added and developed by 3-3¢ diaminobenzidine (Sigma, St Louis, MO, USA) with 0.03% 298 bp flt-4 hydrogen peroxide. Counterstaining was done with methyl green. The negative control used all of the reagents except for the primary antibody. As for the flt-4, specificity of the staining was tested 1098 bp using blocking peptide (Santa Cruz Biotechnology, Inc., Santa flt-1 Cruz, CA, USA). 555 bp Assessment of vessel density KDR LVD was determined by enzyme-histochemistry for 5¢-NA, and VD was determined by immunohistochemistry for factor VIII. For 884 bp Actin the assessment of LVD, we initially identified candidate lymphatic vessels and blood vessels in consecutive H&E-stained and immunohistochemically stained sections for each sample. Then, were an anti-factor VIII rabbit polyclonal antibody at a 1:200 dilu- we chose the most appropriate enzyme-histochemical stained tion (Dako Corp., Carpinteria, CA, USA) and an anti-flt-4 rabbit section among the six sections with different fixation times for polyclonal antibody at a 1:200 dilution (Santa Cruz Biotechnology, each sample, i.e. the slides with maximal 5¢-NA activity in Inc., Santa Cruz, CA, USA). The staining was done by the lymphatic vessels compared to in blood vessels. The vessel densi- immunoperoxidase technique. After air-drying for 10 min sections ties were assessed blindly by two investigators according to the were fixed with acetone for 10 min at Ð20°C. Then the cryostat method previously described by Weidner et al (1991). After the sections were washed in DulbeccoÕs phosphate-buffered saline areas of the highest vascularization were chosen under low power (pH 7.2) without calcium or magnesium (PBSÐ) and endogenous (100 ´ magnification), vessel counts within tumours were done in peroxidase was blocked by treatment with 0.3% hydrogen peroxide the three fields at 400 ´ magnification for VD and at 200 ´ magni- in methanol for 30 min. After washing with PBSÐ, the sections fication for LVD. The average counts of the three fields were were blocked with normal goat serum diluted tenfold with PBSÐ for recorded and the mean of the two investigators findings was used 20 min at room temperature. After washing with PBSÐ, the sections for the final VD. British Journal of Cancer (1999) 81(1), 54–61 © 1999 Cancer Research Campaign H-meso H 2591 H 2818 H 2595 t t t t tt t t t t t t t -+ t tt t Angiogenesis in malignant mesothelioma 57 Table 2 Expression of VEGF, VEGF-C and VEGFRs mRNA in malignant Statistics mesothelioma tumours Differences in densitometry indices were analysed using the Percentage of positive Percentage of MannÐWhitney U-test. The SpearmanÕs rank correlation coeffi- tumours (cases) overexpression (cases) cient test was used to examine the associations between different variables. To assess the prognostic effect on overall survival, VEGF 75.9 (41/54) 31.5 (17/54) VEGF, VEGF-C, VEGFRs, VD and LVD were classified as high VEGF-C 85.2 (46/54) 66.7 (36/54) flt-1 74.1 (40/54) 20.4 (11/54) or low expressing group relative to the mean values of densito- KDR 90.7 (49/54) 42.6 (23/54) metry indices in resected tumours. Two combination factors, flt-4 96.3 (52/54) 59.3 (32/54) COM-VEGF and COM-VEGF-C, were defined to reduce the dimensionality of the independent variables list. The COM-VEGF > each mean value of normal pleural tissue. factor was defined as ÔpositiveÕ if two or more of the three associ- ated variables (VEGF, flt-1, KDR) were strongly expressed, ÔnegativeÕ if otherwise. The COM-VEGF-C factor was defined as ÔpositiveÕ if both VEGF-C and flt-4 were strongly expressed, Table 3 The mean densitometry index of each factor in MPM tumours and ÔnegativeÕ if otherwise. Survival curves were obtained by the normal pleural tissue samples KaplanÐMeier method, and compared univariately by log-rank MPM tumours (n = 54) Normal pleural tissue (n = 5) tests. The effects of age (over vs under 50 years), gender, patho- logical stage (I, II, vs III, IV), pathological subtype (epithelial VEGF-C 0.87 ± 0.15 0.16 ± 0.03 vs others), nodal status (positive vs negative), VEGF, VEGF-C, flt-4 1.24 ± 0.25 0.34 ± 0.07 VEGFRs, VD and LVD on overall survival were assessed multi- KDR 1.73 ± 0.38 1.08 ± 0.23 VEGF 1.80 ± 0.41 1.80 ± 0.60 variately using the Cox proportional hazard regression along with flt-1 0.50 ± 0.11 0.95 ± 0.46 a step-wise procedure for variable selection. Factors with P-value < 0.10 were included in the final model. r = 0.67 r = 0.64 1 P = 0.0018 P = 0.0164 0 123456789 10 VEGF-type C flt-4 r = 0.63 r = 0.33 P < 0.0001 P = 0.0188 0 0 02468 10 12 14 16 18 20 VEGF-type C VEGF121 Figure 2 Correlation (non-parametric) between VEGF-C and flt-4 mRNA expression (A), VEGF-C mRNA expression and LVD (B), flt-4 mRNA expression and LVD (C) and VEGF mRNA expression and VD (D) within MPM tumours © 1999 Cancer Research Campaign British Journal of Cancer (1999) 81(1), 54–61 LVD flt-4 VD LVD 58 Y Ohta et al Figure 3 (A) Staining of flt-4 using polyclonal antibody in malignant pleural Figure 4 5¢-nucleotidase reaction of microvessels within tumours by mesothelioma tissue. Cytoplasmic staining was positive in tumour cells. enzyme-histochemistry for 5¢-NA (A) and immunohistochemical staining for (B) Weakly staining of flt-4 was partly identified in vascular endothelial cells factor VIII antigen (B) of the consecutive cryostat sections. Theoretically, closed arrows show lymphatic vessels and open arrows show blood vessels. Closed arrows in A were counted for LVD and both closed and open arrows in B were counted for VD in this study. Scale bar, 40 mm RESULTS Expression of VEGF, VEGF-C and VEGFRs in normal Immunohistochemically, flt-4 antigen was identified in the pleural tissues, human mesothelioma cell lines and cytoplasm of malignant pleural mesothelioma cells (Figure 3A) resected mesothelioma tumours and partly in vascular endothelial cells (Figure 3B). High VEGF, VEGF-C and VEGFRs (KDR, flt-1 and flt-4) mRNA expression was noted in all human MPM cell lines (Figure 1A). In LVD and VD within MPM tumour tissues MPM tumours, the percentages of positive tumours were, VEGF, 75.9% (41/54 cases); VEGF-C, 85.2% (46/54); flt-1, 74.1% (40/54); The 5¢-NA staining was limited to vessels in 31 of the 54 speci- KDR, 90.7% (49/54); flt-4, 96.3% (52/54). The percentages of mens (57.4%). In these samples, fixation for 30Ð40 min was tumours with higher expression compared to the mean values of optimal for highlighting 5¢-NA activity in lymphatic vessels while normal pleural tissues were 31.5% (17/54) for VEGF, 66.7% (36/54) suppressing 5¢-NA activity in blood vessels (Figure 4). In 23 cases for VEGF-C, 20.4% (11/54) for flt-1, 42.6% (23/54) for KDR and (42.6%), 5¢-NA staining was seen both in vessels and in tumour 59.3% (32/54) for flt-4 (Table 2). Compared to the mean values in cells or stromal elements. In 20 of these 23 cases, the intensity of resected tumours, higher densitometry index of VEGF was found in this non-specific staining was very low compared to that of the two (case nos 2 and 5 in Figure 1B) and higher index of VEGF-C lymphatics, and assessment of LVD was easily achieved. was also recognized in two out of five normal pleural samples (case However, in three cases, the intensity of non-specific staining nos 3 and 5 in Figure 1B). The sample data of resected tumour remained very high despite 60-min of fixation. In these three specimens are shown in Figure 1C. Compared with normal pleural cases, LVD was assessed by avoiding areas with high non-specific tissues, the mean densitometry indexes of VEGF-C (0.87 ± 0.15 vs activity. The mean VD and LVD were 10.9 ± 0.8 (range 0Ð31) and 0.16 ± 0.03) (± s.e.m.), flt-4 (1.24 ± 0.25 vs 0.34 ± 0.07) and KDR 9.1 ± 0.9 (range 1Ð24) respectively. LVD and VEGF-C mRNA (1.73 ± 0.38 vs 1.08 ± 0.23) tended to be greater in MPM tumours expression were strongly associated (P < 0.0001, r = 0.63, Figure (Table 3). Positive correlations were observed between VEGF-C 2B). There were also significant positive relationships between and flt-4 (P = 0.0018, r = 0.67, Figure 2A), VEGF and KDR LVD and flt-4 expression (P = 0.0164, r = 0.64, Figure 2C), and (P = 0.0002, r = 0.39), and VEGF and flt-1 (P = 0.0089, r = 0.27). VD and VEGF (P = 0.0188, r = 0.33, Figure 2D). British Journal of Cancer (1999) 81(1), 54–61 © 1999 Cancer Research Campaign Angiogenesis in malignant mesothelioma 59 Table 4 Cox proportional hazard regression analysis using 46 malignant pleural mesothelioma patients with standardized resections Hazard ratio P 95% CI Low VD (n = 13) Gender 3.437 0.0067 1.409 – 8.384 Stage (I, II vs III, IV) 3.729 0.0091 1.387 – 10.025 VD 2.247 0.0326 1.070 – 4.719 *COM-VEGF 0.400 0.0731 0.147 – 1.089 High VD (n = 15) A combination factor of VEGF, flt-1, and KDR. It was defined as ‘positive’ if two or more of the three associated variables are strongly expressed. Years after operation (Kumar-Singh et al, 1997). In this study, higher densitometry Figure 5 Kaplan–Meier survival plots in stage III patients with epithelial type MPM who underwent standardized resections (n = 28). A tumour was index of VEGF compared to the mean value of tumours was found included in the high VD group if VD was greater than 11. High VD tended to in two out of five normal pleural tissues. If pleura is a tissue with be associated with poor survival (P = 0.0866) relatively high VEGF, this particular condition may suit malignant development. Among VEGFRs, the KDR and flt-4 expression levels tended to be higher in the MPM tumours compared to Association with clinicopathological findings normal pleural tissues in this study. Although the percentage of Histopathological information on lymph node metastasis was avail- VEGF overexpression was not so high, our data confirmed that able in 47 cases. The mean LVD and VD were 8.3 ± 1.0 and VEGF expression was significantly associated with VD and the 10.9 ± 1.1 in node-positive cases (n = 30), and 10.2 ± 1.4 and expression of VEGFRs (KDR and flt-1) in MPM tumours. 11.0 ± 1.5 in node-negative cases (n = 17) respectively. Nodal status Furthermore, among the 46 MPM patients undergoing standard- had no relationship with LVD (P = 0.1351) or VD ized resections, high VD was significantly associated with poor (P = 0.8856). VEGF and VEGF-C mRNA expression levels were survival. A combination factor with VEGF and its receptors (flt-1 1.9 ± 0.6 and 0.8 ± 0.2, respectively, in node-positive cases, and and KDR), COM-VEGF, also had a weaker relationship with 1.3 ± 0.2 and 0.9 ± 0.2, respectively, in node-negative cases. The outcome, yet the value only trends toward significance due to the differences were not statistically significant. In stage III patients small sample size. with epithelial type who underwent standardized resections Most conventional immunohistochemistry methods for the detec- (n = 28), the median survival of high and low VD groups were 11 tion of endothelial cells do not distinguish lymphatics from blood and 17 months respectively. The 1-year survival rates for high and vessels. In this study, we tried to assess the LVD within MPM low VD were 40.0% and 61.5% respectively. In this analysis, high tumours using the enzyme-histochemistry assay for 5¢-NA based on VD tended to correlate with poor survival (P = 0.0866, Figure 5). the leading method (Wachstein et al, 1954). 5¢-NA expression in Using all patients with standardized resections (n = 46), age, nodal vascular endothelial cells varies according to vessel types (Turner et status, histological subtype, VEGF-C, flt-4, the COM-VEGF-C and al, 1987; Airas et al, 1997). That is, the activity of 5¢-NA is very high LVD had no impact on survival. As a result of multivariate analysis, in lymphatic endothelial cells, while it is very low or absent in blood male, advanced stage (III and IV) and high VD were independent capillary endothelial cells (Turner et al, 1987). Therefore, this assay negative prognostic indicators. The positive COM-VEGF also has a great potential to discriminate neo-lymphatics from showed weaker correlation with poor survival (Table 4). neo-blood vessels after the adequate inhibition of its activity by paraformaldehyde (Vetter et al, 1970; Ji et al, 1997). In normal lymphatics, this enzyme activity appears to be necessary for the DISCUSSION growth and development of vessels (Ji et al, 1997), and has an Neovascularization has been shown to be necessary for tumour important role in the control of interactions between lymphocytes growth and metastasis (Folkman, 1990). Although some contraries and vascular endothelial cells (Airas et al, 1997). If this enzyme exist, many studies have confirmed the negative impact of tumour activity is essential for the proliferation of lymphatic vessels, the vascularization on prognosis (Chodak et al, 1980; Weidner et al, same condition may be expected in tumour-induced neo-vessels. 1991; Macchiarini et al, 1992). Among the many reported angio- Although the significance of 5¢-NA activity in cancerous lesions is genic factors, VEGF is the most powerful endothelial cell-specific not clear, this enzyme has been reported in some tumour tissues, mitogen associated with tumour neovascularization. A number of including seminoma, malignant fibrous histiocytoma, breast cancer investigators have reported a significant relationship between VD and renal cancer (Wachstein et al, 1954; Wood et al, 1986; Canbolat and VEGF expression in a variety of tumours (Toi et al, 1994; et al, 1996; Durak et al, 1997). In this study, we have used this Mattern et al, 1995, Samoto et al, 1995, Takahashi et al, 1995), and method, for the first time, to study the microlymphatic vessel overexpression of VEGF has been associated with poor prognosis density in MPM tumours and found that 5¢-NA staining was limited in some neoplasms (Toi et al, 1994; Takahashi et al, 1995). to vessels in 31 of the 54 (57.4%) MPM tumour specimens. In However, little information is available on tumour angiogenesis in these samples, lymphatic vessel specificity was optimized by malignant mesothelioma. Recently, Kumar-Singh et al (1997) paraformaldehyde fixation for 30Ð40 min. In 23 cases (42.6%), reported that VD in MPM tumours was significantly higher than 5¢-NA staining was also noted in MPM tumour cells and within that in non-neoplastic mesothelium. They also reported that the stromal elements. In 20 out of these 23 cases, the intensity of the patients with highly vascularized tumours had a significantly non-vessel staining was less than that of the vessels, and LVD could shorter survival than patients with poorly vascularized ones be assessed. In three cases with the intensity of non-specific © 1999 Cancer Research Campaign British Journal of Cancer (1999) 81(1), 54–61 Survival rate (%) 60 Y Ohta et al Durak I, Beduk Y, Kavutcu M, Suzer O, Yaman O, Ozturk HS, Canbolat O and staining, LVD was assessed by avoiding areas with high non- Ulutepe S (1997) Activity of the enzymes participating in purine metabolism of specific activity. Under the condition that very little information is cancerous and non-cancerous human kidney tissues. Cancer Invest 5: 212Ð216 available about the mechanisms of lymphangiogenesis, some Folkman J (1990) What is the evidence that tumors are angiogenesis dependent? reports have recently suggested that the specific pathways involved J Natl Cancer Inst 82: 4Ð6 in lymphangiogenesis are different from those in hemangiogenesis Folkman J (1995) Angiogenesis in cancer, vascular, rheumatoid and other disease. Nat Med 1: 27Ð31 (Wilting et al, 1996; Oh et al, 1997). Among the VEGF family Grimmond S, Lagercrantz J, Drinkwater C, Silins G, Townson S, Pollock P, Gotley members, the function of VEGF-C appears to extend to the D, Carson E, Rakar S, Nordenskjold M, Ward L, Hayward N and Weber G lymphatic system as a ligand for flt-4, and VEGF-C is suspected (1996) Cloning and characterization of a novel human gene related to vascular to play an important role in lymphangiogenesis (Kukk et al, 1996; endothelial growth factor. Genome Res 6: 124Ð131 Hewett PW and Murray JC (1996) Coexpression of flt-1, flt-4 amd KDR in freshly Jeltsch et al, 1997; Oh et al, 1997). In this study, we identified a isolated and cultured human endothelial cells. Biochem Biophys Res Commun strong positive relationship between VEGF-C mRNA expression 221: 697Ð702 and LVD, and flt-4 expression and LVD in MPM, as well as a close Jeltsch M, Kaipainen A, Joukov V, Meng X, Lakso M, Rauvala H, Swartz M, association between flt-4 and VEGF-C expression. Further, the Fukumura D, Jain RK and Alitalo K (1997) Hyperplasia of lymphatic vessels mean vessel count for VD that includes both blood and lymphatic in VEGF-C transgenic mice. Science 276: 1423Ð1425 Ji RC and Kato S (1997) Enzyme-histochemical study on postnatal development of vessels was much greater than that for LVD. VD itself had no rat stomach lymphatic vessels. Microvasc Res 54: 1Ð12 relationship with VEGF-C and flt-4 expression levels in this study. Joukov V, Pajusola K, Kaipainen A, Chilov D, Lahtinen I, Kukk E, Saksela O, These results support the use of the enzyme-histochemistry method Kalkkinen N and Alitalo K (1996) A novel vascular endothelial growth factor, based on 5¢-NA activity for assessment of lymphatics, and also VEGF-C, is a ligand for the flt-4 (VEGFR-3) and KDR (VEGFR-2) receptor tyrosine kinases. EMBO J 15: 290Ð298 suggest that VEGF-C plays a key role in lymphangiogenesis in Kaipainen A, Korhonen J, Mustonen T, van Hinsbergh VWM, Fang G, Dumont D, MPM. In addtion, flt-4 expression, which has originally been found Breitman M and Alitalo K (1995) Expression of the fms-like tyrosine kinase 4 in lymphatic endothelium (Kaipainen et al, 1995; Joukov et al, gene becomes restricted to lymphatic endothelium during development. Proc 1996), was high in both MPM cell lines used and resected tumour Natl Acad Sci USA 92: 3566Ð3570 tissues. Immunohistochemical staining revealed flt-4 expression in Kukk E, Lymboussaki A, Taira S, Kaipainen A, Jeltsch M, Joukov V and Alitalo K (1996) VEGF-C receptor binding and pattern of expression with VEGFR-3 MPM tumour cells except for some vascular endothelial cells. In suggests a role in lymphatic vascular development. Development 122: this study, it was difficult to clarify the correspondence between 3829Ð3837 endothelial cells with flt-4 expression and those with 5¢-NA activity Kumar-Singh S, Vermeulen PB, Weyler J, Segers K, Weyn B, van Daele A, Dirix LY, partly owing to the situation that mesothelioma cells themselves van Oosterom AT and van Marck E (1997) Evaluation of tumour angiogenesis as a prognostic marker in malignant mesothelioma. J Pathol 182: 211Ð216 expressed both flt-4 and 5¢-NA. Although the role of flt-4 in tumour Lee J, Gray A, Yuan J, Luoh S, Avraham H and Wood WI (1996) Vascular cells is not well known, flt-4 expression was confirmed in meso- endothelial growth factor-related protein: a ligand and specific activator of the thelial cells (Hewett et al, 1996). This suggests a possibility that tyrosine kinase receptor flt-4. Proc Natl Acad Sci USA 93: 1988Ð1992 mesothelioma cells express flt-4, and that VEGF-C may have some Macchiarini P, Fontanini G, Hardin MJ, Squartini F and Angeletti CA (1992) role for the autocrine growth or proliferation of tumour cells. Relation of neovascularization to metastasis of non-small-cell lung cancer. Lancet 340: 145Ð146 In conclusion, our study suggests that VEGF and VEGF-C play Mattern J, Koomagi R and Volm M (1995) Vascular endothelial growth factor an important role in angiogenesis and lymphangiogenesis in MPM expression and angiogenesis in non-small-cell lung carcinomas. Int J Oncol 6: tumours, and that VD has a significant impact on the overall 1059Ð1062 survival of MPM patients. The LVD assessed by enzyme-histo- Meyer M, Clauss M, Lepple-Wienhues A, Waltenberger J, Augustin HG, Ziche M, Lanz C, Bittner M, Rziha HJ and Dehio C (1999) A novel vascular endothelial chemistry for 5¢-NA has no impact on nodal metastasis. However, growth factor encoded by Orf virus, VEGF-E, mediates angiogenesis via nodal status does not have a relation with outcome in our samples. signalling through VEGFR-2 (KDR) but not VEGFR-1 (Flt-1) receptor Therefore, a question remains whether VEGF and VEGF-C have a tyrosine kinase. EMBO J 18: 363Ð374 relation with prognosis in cases where nodal status has a great Mossman BT, Kamp DW and Weitzman SA (1996) Mechanisms of carcinogenesis impact on survival. The potential roles of VEGF and VEGF-C in and clinical features of asbestos-associated cancers. Cancer Invest 14: 466Ð480 Oh SJ, Jeltsch MM, Birkenhager R, McCarthy JE, Weich HA, Christ B, Alitalo K lymphatic metastasis warrant further study. and Wilting J (1997) VEGF and VEGF-C: specific induction of angiogenesis and lymphangiogenesis in the differentiated avian chorioallantoic membrane. Dev Biol 188: 96Ð109 ACKNOWLEDGEMENTS Olofsson B, Pajusola K, Kaipainen A, von Euler G, Joukov V, Saksela O, Orpana A, Pettersson RF, Alitalo K and Eriksson U (1996) Vascular endothelial growth We would like to thank Mr Larry Tait for technical assistance. We factor B, a novel growth factor for endothelial cells. Proc Natl Acad Sci USA also thank Professor H Yamamoto of Kanazawa University for his 93: 2576Ð2581 kind gift of the primers for flt-1, flt-4 and KDR. Pass HI, Stevens EJ, Oie H, Tsokos MG, Abati AD, Fetsch PA, Mew DJY, Pogrebniak HW and Matthews WJ (1995) Characteristics of nine newly derived mesothelioma cell lines. Ann Thorac Surg 59: 835Ð8344 REFERENCES Rusch VW (1996) A proposed new International TNM staging system for malignant pleural mesothelioma from the International Mesothelioma Interest Group. Airas L, Niemela J, Salmi M, Puurunen T, Smith DJ and Jalkanen S (1997) Lung Cancer 14: 1Ð12 Differential regulation and function of CD73, a glycosyl-phosphatidylinositol- Samoto K, Ikezaki K, Ono M, Shono T, Kohno K, Kuwano M and Fukui M (1995) linked 70-kD adhesion molecule, on lymphocytes and endothelial cells. J Cell Expression of vascular endothelial growth factor and its possible relation with Biol 136: 421Ð431 neovascularization in human brain tumor. Cancer Res 55: 1189Ð1193 Canbolat O, Durak I, Cetin R, Kavutcu M, Demirci S and Ozturk S (1996) Activities Takahashi Y, Kitadai Y, Bucana CD, Cleary KR and Ellis LM (1995) Expression of of adenosine deaminase, 5¢-nucleotidase, guanase, and cystidine deaminase vascular endothelial growth factor and its receptor, KDR, correlates with enzymes in cancerous and non-cancerous human breast tissues. Breast Cancer vascularity, metastasis, and proliferation of human colon cancer. Cancer Res Res Treat 37: 189Ð193 55: 3964Ð3968 Chodak GW, Haudenschild C, Gittes RF and Folkman J (1980) Angiogenic activity Toi M, Hoshina S, Takayanagi T and Tominaga T (1994) Association of vascular as a marker of neoplasia and preneoplasia in lesions of the human bladder. Ann endothelial growth factor expression with tumor angiogenesis and with early Surg 192: 762Ð771 relapse in primary breast cancer. Jpn J Cancer Res 85: 1045Ð1049 British Journal of Cancer (1999) 81(1), 54–61 © 1999 Cancer Research Campaign Angiogenesis in malignant mesothelioma 61 Turner RR, Beckstead JH, Warnke RA and Wood GS (1987) Endothelial cell Weidner N, Semple JP, Welch WR and Folkman J (1991) Tumor angiogenesis and phenotypic diversity. In situ demonstration of immunologic and enzymatic metastasis Ð correlation in invasive breast carcinoma. N Engl J Med 324: 1Ð8 heterogeneity that correlates with specific morphologic subtypes. Am J Clin Wilting J, Birkenhager R, Eichmann A, Kurz H, Martiny-Baron G, Marme D, Pathol 87: 569Ð575 McCarthy JE, Christ B and Weich HA (1996) VEGF121 induces proliferation Vetter W (1970) Alkaline phosphatasen in mastzellen, blut- und lymphagefŠben der of vascular endothelial cells and expression of flk-1 without affecting rattenzunge. 5¢-nucleotidase-, unspezifische alkalische phosphatase- und lymphatic vessels of chorioallantoic membrane. Dev Biol 176: 76Ð85 polyphosphatase- (ATP¢ase) aktivitŠt unter besondere berŸckscichtigung des Wood GS, Beckstead JH, Turner RR, Hendrickson MR, Kempson RL and Warnke pH. Z Anat Entwicklungsgesch 130: 153Ð176 RA (1986) Malignant fibrous histiocytoma tumor cells resemble fibroblasts. Wachstein M and Meisel E (1954) The histochemical distribution of 5-nucleotidase Am J Surg Pathol 110: 323Ð335 and unspecific alkaline phosphatase in the testicle of various species and in two Yamada Y, Nezu J, Shimane M and Hirata Y (1997) Molecular cloning of a novel human seminomas. J Histochem 2: 137Ð148 vascular endothelial growth factor, VEGF-D. Genomics 42: 483Ð488 © 1999 Cancer Research Campaign British Journal of Cancer (1999) 81(1), 54–61 http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png British Journal of Cancer Springer Journals

VEGF and VEGF type C play an important role in angiogenesis and lymphangiogenesis in human malignant mesothelioma tumours

Download PDF
8 pages

Loading next page...
 
/lp/springer-journals/vegf-and-vegf-type-c-play-an-important-role-in-angiogenesis-and-Ofi86p0yDO

References (37)

Publisher
Springer Journals
Copyright
Copyright © 1999 by The Author(s)
Subject
Biomedicine; Biomedicine, general; Cancer Research; Epidemiology; Molecular Medicine; Oncology; Drug Resistance
ISSN
0007-0920
eISSN
1532-1827
DOI
10.1038/sj.bjc.6690650
Publisher site
See Article on Publisher Site

Abstract

British Journal of Cancer (1999) 81(1), 54–61 © 1999 Cancer Research Campaign Article no. bjoc.1999.0650 VEGF and VEGF type C play an important role in angiogenesis and lymphangiogenesis in human malignant mesothelioma tumours 1,4 1 1 1 2 3 4 1 Y Ohta , V Shridhar , RK Bright , GP Kalemkerian , W Du , M Carbone , Y Watanabe and HI Pass 1 2 Aerodigestive Program and Hematology-Oncology, Barbara Ann Karmanos Cancer Institute, Wayne State University School of Medicine, Detroit, MI, USA; 3 4 Loyola Cancer Center, Maywood, IL, USA; First Department of Surgery, School of Medicine, Kanazawa University, Kanazawa, Japan Summary The vascular endothelial growth factor (VEGF) family is a novel regulator of endothelial cell proliferation. We assessed the mRNA expression of VEGF, VEGF type C (VEGF-C) and their receptors together with the microvessel density (VD) and microlymphatic vessel density (LVD) in pursuit of their connection and prognostic value in malignant pleural mesothelioma (MPM). We used four human MPM cell lines, 54 MPM tumours and five normal pleural tissues. Expression levels for receptors and ligands were assessed by semiquantitative reverse transcriptase polymerase chain reaction analysis. Microvessels were highlighted by immunohistochemical staining for factor VIII. The discrimination of lymphatics was performed by enzyme-histochemistry for 5¢-nucleotidase after adequate inhibition of non-specific activity. The expression levels of VEGF, VEGF-C and VEGFRs were high in all MPM cell lines. The percentages of tumours with higher expression compared to the mean values of normal pleural tissues were 31.5% (17/54) for VEGF, 66.7% (36/54) for VEGF-C, 20.4% (11/54) for fms-like tyrosine kinase (flt)-1, 42.6% (23/54) for kinase insert domain-containing recepter (KDR) and 59.3% (32/54) for flt-4. Significant positive correlations were found between VEGF-C and flt-4, VEGF and KDR, VEGF and flt-1 in tumour tissues. The association between LVD and VEGF-C expression level was especially strong (P < 0.0001, r = 0.63). There were also significant correlations between LVD and flt-4, and VD and VEGF. No correlation, however, was found between LVD and nodal metastasis. VD was a negative prognostic indicator in this study. The associations between VEGF/VEGF-C and vessel density suggest that these factors play an important role in angiogenesis and lymphangiogenesis in this tumour, and assessment of vascularity may be a useful prognostic indicator for MPM patients. Keywords: vascular endothelial growth factor; VEGF; VEGF-C; angiogenesis; lymphangiogenesis; malignant mesothelioma Asbestos had been widely used for centuries before a causal rela- 1996; Lee et al, 1996; Olofsson et al, 1996; Yamada et al, 1997; tionship between malignant pleural mesothelioma (MPM) and Meyer et al, 1999). Interestingly, the function of VEGF-C appears exposure to asbestos was confirmed. Although the use of asbestos to extend to the lymphatic system where it serves as a ligand for has recently been restricted, the incidence of MPM remains high fms-like tyrosine kinase 4 (flt-4) (Kukk et al, 1996; Jeltsch et al, and is steadily increasing due to the long latent period for MPM 1997). However, lymphangiogenesis within tumours has not been after the exposure to asbestos (Mossman et al, 1996). Advanced documented. MPM remains an aggressive and highly lethal disease due to its For MPM, little information on tumour angiogenesis is avail- marked resistance to conventional treatments, including surgical able. In this study, we assessed the expression of VEGF, VEGF-C resection, chemotherapy and radiotherapy. and VEGFRs (KDR, flt-1 and flt-4) in human MPM cell lines, Angiogenesis is crucial for the proliferation of tumour cells, and MPM tumour samples and normal pleural tissue. We used the anti-angiogenic therapy is a promising strategy aimed at inhibiting reverse transcriptase polymerase chain reaction technique (RT- tumour growth, invasion and metastasis (Folkman et al, 1995). PCR) to assess the expression of each factor, and we also assessed Among the many reported angiogenic factors, vascular endothelial the VD and microlymphatic vessel density (LVD) within tumours growth factor (VEGF) is the most powerful endothelial cell- using immunohistochemistry for factor VIII and enzyme-histo- specific mitogen associated with tumour neovascularization. chemistry for 5¢-nucleotidase (5¢-NA) respectively. The expres- While a number of investigators have reported the direct relation- sion of the associated angiogenic factors and receptors was also ship between microvessel density (VD) and VEGF expression determined. Clinical correlations of angiogenesis with survival within a variety of tumours (Toi et al, 1994; Mattern et al, 1995; from treatments for the patients was also performed. Samoto et al, 1995; Takahashi et al, 1995), recent studies also suggest that VEGF type B, C, D and E are novel regulators of MATERIALS AND METHODS endothelial cell proliferation (Grimmond et al, 1996; Joukov et al, Cell lines and tissue samples The human malignant mesothelioma cell lines (H-meso, H2818, Received 28 October 1998 H2591 and H2595) used in this study were all developed from Accepted 16 February 1999 tumours diagnosed using a well-defined panel of immunohisto- chemical markers (Pass et al, 1995). H-meso is commercially Correspondence to: Y Ohta, First Department of Surgery, Kanazawa University, School of Medicine, Kanazawa 920-8641, Japan available (Biomeasure, Hopkinton, MA, USA), while H2818, 54 Angiogenesis in malignant mesothelioma 55 Table 1 Nucleotide sequences of the primers used and their PCR conditions. Angiogenic factors Primer sets PCR cycle Product size (bp) VEGF 5¢-GAAGTGGTGAAGTTCATGGATGTC-3¢ (sense) 30 408 (VEGF121) 5¢-CGATCGTTCTGTATCAGTCTTTCC-3¢ (antisense) 541 (VEGF165) 613 (VEGF185) VEGF-C 5¢-CATGTACGAACCGCCAG-3¢ (sense) 25 320 5¢-TTGGCTGTTTGGTCATTGGC-3¢ (antisense) Flt-1 5¢-GAGAATTCACTATGGAAGATCTGATTTCTTACAGT-3¢ (sense) 30 1098 5¢-GAGCATGCGGTAAAATACACATGTGCTTCTAG-3¢ (antisense) Flt-4 5¢-AGCCATTCATCAACAAGCCT-3¢ (sense) 25 298 5¢-GGCAACAGCTGGATGTCATA-3¢ (antisense) KDR 5¢-TATAGATGGTGTAACCCGGA-3¢ (sense) 30 555 5¢-TTTGTCACTGAGACAGCTTGG-3¢ (antisense) H2591 and H2595 were developed from resected MPM tumours at amplification. For quantitative evaluation of the amplified product, the National Cancer Institute (Bethesda, MD, USA) (Pass et al, PCR encompassing 20Ð40 cycles was preliminarily performed to 1995). A human fibrosarcoma cell line (HT1080) was used as a determine the most suitable number of amplifications for each positive control for VEGF, VEGF-C, flt-1 and flt-4, while factor. PCR was performed after adding 80 ml of PCR mixture (50 HUVEC was used for kinase insert domain-containing receptor mM TrisÐHCl, pH 8.3; 40 mM KCl; 8 mM MgCl ; 0.5 mM each (KDR). Cell lines were maintained in RPMI-1640 media (Gibco- dNTP; 50 pmol of each the sense and the antisense primer; 2.5 units BRL, Gaithersburg, MD, USA) supplemented with 10% fetal of Taq polymerase (Takara, Kyoto, Japan)). Each cycle consisted of: bovine serum (FBS), 2 mML-glutamine, 10 mM HEPES buffer, 1 min at 94°C, 1 min at 58°C and 2 min at 72°C for VEGF, flt-1 and Ð1 Ð1 500 units ml penicillin and 500 units ml streptomycin at 37°C KDR; and 1 min at 94°C, 1 min at 52°C and 1.5 min at 72°C for under a humidified atmosphere of 5% carbon dioxide. Five normal VEGF type C, flt-4 and b-actin. PCR primers used and PCR condi- pleural tissue samples were obtained from patients with non- tions are shown in Table 1. The PCR products were electrophoresed cancerous disease (either for repair of hiatus hernia or diagnosis of through a 1.0% agarose gel (4-mm thick). The gel was fluorescently pulmonary infiltrates). In all these cases, the resected pleural tissue stained in SYBR Green I stain Solution (FMC Bio-Products, was confirmed to be free of malignancy by pathological examina- Rockland, ME, USA) at a 1:10 000 dilution in 1 ´ TAE buffer tion. And 54 malignant mesothelioma tumours were randomly (pH 7.2) for 30 min. The intensity of the PCR product bands was chosen from a bank of snap-frozen samples collected by one of us measured by Stormª image analyser (Molecular Dynamics, Inc., (HIP). For tumour samples, consecutive 6-mm cryosections were Sunnyvale, CA, USA). Each expression was standardized using the cut from the invasive tumour margin for immunohistochemical b-actin signal as an internal control and a densitometry index and enzyme-histochemical studies. The pathologic classification defined. For negative control of each factor, PCR was done of each sample was confirmed by review of haematoxylin and excepting the template cDNA. eosin (H&E)-stained sections. The pathological MPM subtypes identified were: epithelial (E), 44; sarcomatous (S), 4; and mixed Enzyme-histochemistry (M), 6. According to the new international TNM staging system An enzyme-histochemical reaction for 5¢-NA was done according (Rusch, 1996), 50 cases could be classified based on pathological to the leading method of Wachstein et al (1954). Briefly, cryostat findings [stage II in nine (E, 7; S, 2), III in 36 (E, 28; S, 2; M, 6) sections were fixed with 5% paraformaldehyde for 10Ð60 min (10, and IV in five (E, 4; M, 1)]. 20, 30, 40, 50 and 60 min for each sample). This step-wise fixation was carried out in order to discover the appropriate conditions to RT-PCR analysis maximally exhibit 5¢-NA activity in microlymphatic vessels compared to that in blood vessels. Next, after washing with water Total RNA was extracted from the human mesothelioma cell lines, for 5 min, sections were treated with an enzyme reaction for 2 h control cell lines (HT1080 and HUVEC), normal pleural tissues, at 37°C using a substrate mixture (20 ml of 0.2 M TrisÐmaleate and each resected MPM tumour using TRIzol (Life Technologies, buffer (pH 7.2), 25 mg adenosine-5-monophosphate (Sigma, Inc., Grand Island, NY, USA) according to the standard acidÐguani- St Louis, MO, USA), 20 mg tetramisole (Sigma, St Louis, MO, diumÐphenolÐchloroform method. Total RNA (1 mg) was denatured USA), 3.5 mg sucrose, 5 ml 0.1 M magnesium sulphate, 3 ml of together with oligo-dT primer (10 pmol) for 15 min at 68°C. After 2% Pb(NO ) , 20 ml water). After the sections were washed with 3 2 this was chilled on ice for 5 min, poly-adenosine (poly-A) RNA was water for 15 min, they were reacted with a 1% ammonium reverse-transcribed at 42°C for 90 min in RT solution (50 mM sulphide solution (Sigma, St Louis, MO, USA) for 2 min. The TrisÐHCl (hychochloric acid), pH 8.3; 40 mM potassium chloride sections were counterstained with methyl green and mounted (KCl); 8 mM magnesium chloride (MgCl ); 0.5 mM each dNTP; using glycerol (Sigma, St Louis, MO, USA). Ð1 225 mg ml bovine serum albumin; 5 mM dithiothreitol; 20 units RNasin (Life Technologies, Inc., Grand Island, NY, USA) and Immunohistochemistry 4 units AMV reverse transcriptase (Gibco-BRL, Gaithersburg, MD, USA)). The cDNA was incubated at 95°C for 5 min to inactivate Consecutive cryostat sections were used for the immunohistochem- the reverse transcriptase, and served as the template for PCR ical assessment of VD. The primary antibodies used in this study © 1999 Cancer Research Campaign British Journal of Cancer (1999) 81(1), 54–61 56 Y Ohta et al Human mesothelioma cell lines Human mesothelioma tissues 320 bp VEGF-C 320 bp VEGF-C 541 bp VEGF 408 bp VEGF 541 bp 408 bp 298 bp flt-4 298 bp flt-4 1098 bp flt-1 1098 bp flt-1 555 bp KDR 555 bp KDR 884 bp 884 bp Actin Actin Figure 1 RT-PCR analysis of VEGF, VEGF-C and their receptors (KDR, flt-1 and flt-4) in human mesothelioma cell lines (A), normal pleural tissue (B) and resected human malignant mesothelioma tumour samples (C). High Normal pleural tissue expression of VEGF, VEGF-C and VEGFRs mRNA was recognized in human mesothelioma cell lines. Compared with the expression in the five normal pleural tissues, VEGF-C and flt-4 mRNA expression levels were greater in tumour tissues 320 bp VEGF-C were reacted with primary antibody for 2 h, then washed with PBSÐ and reacted with biotin-labelled anti-rabbit immunoglobulin 541 bp VEGF 408 bp (Vector Laboratories, Burlingame, CA, USA) for 30 min at room temperature. After sections were washed with PBSÐ, avidinÐbiotinÐperoxidase complex was added and developed by 3-3¢ diaminobenzidine (Sigma, St Louis, MO, USA) with 0.03% 298 bp flt-4 hydrogen peroxide. Counterstaining was done with methyl green. The negative control used all of the reagents except for the primary antibody. As for the flt-4, specificity of the staining was tested 1098 bp using blocking peptide (Santa Cruz Biotechnology, Inc., Santa flt-1 Cruz, CA, USA). 555 bp Assessment of vessel density KDR LVD was determined by enzyme-histochemistry for 5¢-NA, and VD was determined by immunohistochemistry for factor VIII. For 884 bp Actin the assessment of LVD, we initially identified candidate lymphatic vessels and blood vessels in consecutive H&E-stained and immunohistochemically stained sections for each sample. Then, were an anti-factor VIII rabbit polyclonal antibody at a 1:200 dilu- we chose the most appropriate enzyme-histochemical stained tion (Dako Corp., Carpinteria, CA, USA) and an anti-flt-4 rabbit section among the six sections with different fixation times for polyclonal antibody at a 1:200 dilution (Santa Cruz Biotechnology, each sample, i.e. the slides with maximal 5¢-NA activity in Inc., Santa Cruz, CA, USA). The staining was done by the lymphatic vessels compared to in blood vessels. The vessel densi- immunoperoxidase technique. After air-drying for 10 min sections ties were assessed blindly by two investigators according to the were fixed with acetone for 10 min at Ð20°C. Then the cryostat method previously described by Weidner et al (1991). After the sections were washed in DulbeccoÕs phosphate-buffered saline areas of the highest vascularization were chosen under low power (pH 7.2) without calcium or magnesium (PBSÐ) and endogenous (100 ´ magnification), vessel counts within tumours were done in peroxidase was blocked by treatment with 0.3% hydrogen peroxide the three fields at 400 ´ magnification for VD and at 200 ´ magni- in methanol for 30 min. After washing with PBSÐ, the sections fication for LVD. The average counts of the three fields were were blocked with normal goat serum diluted tenfold with PBSÐ for recorded and the mean of the two investigators findings was used 20 min at room temperature. After washing with PBSÐ, the sections for the final VD. British Journal of Cancer (1999) 81(1), 54–61 © 1999 Cancer Research Campaign H-meso H 2591 H 2818 H 2595 t t t t tt t t t t t t t -+ t tt t Angiogenesis in malignant mesothelioma 57 Table 2 Expression of VEGF, VEGF-C and VEGFRs mRNA in malignant Statistics mesothelioma tumours Differences in densitometry indices were analysed using the Percentage of positive Percentage of MannÐWhitney U-test. The SpearmanÕs rank correlation coeffi- tumours (cases) overexpression (cases) cient test was used to examine the associations between different variables. To assess the prognostic effect on overall survival, VEGF 75.9 (41/54) 31.5 (17/54) VEGF, VEGF-C, VEGFRs, VD and LVD were classified as high VEGF-C 85.2 (46/54) 66.7 (36/54) flt-1 74.1 (40/54) 20.4 (11/54) or low expressing group relative to the mean values of densito- KDR 90.7 (49/54) 42.6 (23/54) metry indices in resected tumours. Two combination factors, flt-4 96.3 (52/54) 59.3 (32/54) COM-VEGF and COM-VEGF-C, were defined to reduce the dimensionality of the independent variables list. The COM-VEGF > each mean value of normal pleural tissue. factor was defined as ÔpositiveÕ if two or more of the three associ- ated variables (VEGF, flt-1, KDR) were strongly expressed, ÔnegativeÕ if otherwise. The COM-VEGF-C factor was defined as ÔpositiveÕ if both VEGF-C and flt-4 were strongly expressed, Table 3 The mean densitometry index of each factor in MPM tumours and ÔnegativeÕ if otherwise. Survival curves were obtained by the normal pleural tissue samples KaplanÐMeier method, and compared univariately by log-rank MPM tumours (n = 54) Normal pleural tissue (n = 5) tests. The effects of age (over vs under 50 years), gender, patho- logical stage (I, II, vs III, IV), pathological subtype (epithelial VEGF-C 0.87 ± 0.15 0.16 ± 0.03 vs others), nodal status (positive vs negative), VEGF, VEGF-C, flt-4 1.24 ± 0.25 0.34 ± 0.07 VEGFRs, VD and LVD on overall survival were assessed multi- KDR 1.73 ± 0.38 1.08 ± 0.23 VEGF 1.80 ± 0.41 1.80 ± 0.60 variately using the Cox proportional hazard regression along with flt-1 0.50 ± 0.11 0.95 ± 0.46 a step-wise procedure for variable selection. Factors with P-value < 0.10 were included in the final model. r = 0.67 r = 0.64 1 P = 0.0018 P = 0.0164 0 123456789 10 VEGF-type C flt-4 r = 0.63 r = 0.33 P < 0.0001 P = 0.0188 0 0 02468 10 12 14 16 18 20 VEGF-type C VEGF121 Figure 2 Correlation (non-parametric) between VEGF-C and flt-4 mRNA expression (A), VEGF-C mRNA expression and LVD (B), flt-4 mRNA expression and LVD (C) and VEGF mRNA expression and VD (D) within MPM tumours © 1999 Cancer Research Campaign British Journal of Cancer (1999) 81(1), 54–61 LVD flt-4 VD LVD 58 Y Ohta et al Figure 3 (A) Staining of flt-4 using polyclonal antibody in malignant pleural Figure 4 5¢-nucleotidase reaction of microvessels within tumours by mesothelioma tissue. Cytoplasmic staining was positive in tumour cells. enzyme-histochemistry for 5¢-NA (A) and immunohistochemical staining for (B) Weakly staining of flt-4 was partly identified in vascular endothelial cells factor VIII antigen (B) of the consecutive cryostat sections. Theoretically, closed arrows show lymphatic vessels and open arrows show blood vessels. Closed arrows in A were counted for LVD and both closed and open arrows in B were counted for VD in this study. Scale bar, 40 mm RESULTS Expression of VEGF, VEGF-C and VEGFRs in normal Immunohistochemically, flt-4 antigen was identified in the pleural tissues, human mesothelioma cell lines and cytoplasm of malignant pleural mesothelioma cells (Figure 3A) resected mesothelioma tumours and partly in vascular endothelial cells (Figure 3B). High VEGF, VEGF-C and VEGFRs (KDR, flt-1 and flt-4) mRNA expression was noted in all human MPM cell lines (Figure 1A). In LVD and VD within MPM tumour tissues MPM tumours, the percentages of positive tumours were, VEGF, 75.9% (41/54 cases); VEGF-C, 85.2% (46/54); flt-1, 74.1% (40/54); The 5¢-NA staining was limited to vessels in 31 of the 54 speci- KDR, 90.7% (49/54); flt-4, 96.3% (52/54). The percentages of mens (57.4%). In these samples, fixation for 30Ð40 min was tumours with higher expression compared to the mean values of optimal for highlighting 5¢-NA activity in lymphatic vessels while normal pleural tissues were 31.5% (17/54) for VEGF, 66.7% (36/54) suppressing 5¢-NA activity in blood vessels (Figure 4). In 23 cases for VEGF-C, 20.4% (11/54) for flt-1, 42.6% (23/54) for KDR and (42.6%), 5¢-NA staining was seen both in vessels and in tumour 59.3% (32/54) for flt-4 (Table 2). Compared to the mean values in cells or stromal elements. In 20 of these 23 cases, the intensity of resected tumours, higher densitometry index of VEGF was found in this non-specific staining was very low compared to that of the two (case nos 2 and 5 in Figure 1B) and higher index of VEGF-C lymphatics, and assessment of LVD was easily achieved. was also recognized in two out of five normal pleural samples (case However, in three cases, the intensity of non-specific staining nos 3 and 5 in Figure 1B). The sample data of resected tumour remained very high despite 60-min of fixation. In these three specimens are shown in Figure 1C. Compared with normal pleural cases, LVD was assessed by avoiding areas with high non-specific tissues, the mean densitometry indexes of VEGF-C (0.87 ± 0.15 vs activity. The mean VD and LVD were 10.9 ± 0.8 (range 0Ð31) and 0.16 ± 0.03) (± s.e.m.), flt-4 (1.24 ± 0.25 vs 0.34 ± 0.07) and KDR 9.1 ± 0.9 (range 1Ð24) respectively. LVD and VEGF-C mRNA (1.73 ± 0.38 vs 1.08 ± 0.23) tended to be greater in MPM tumours expression were strongly associated (P < 0.0001, r = 0.63, Figure (Table 3). Positive correlations were observed between VEGF-C 2B). There were also significant positive relationships between and flt-4 (P = 0.0018, r = 0.67, Figure 2A), VEGF and KDR LVD and flt-4 expression (P = 0.0164, r = 0.64, Figure 2C), and (P = 0.0002, r = 0.39), and VEGF and flt-1 (P = 0.0089, r = 0.27). VD and VEGF (P = 0.0188, r = 0.33, Figure 2D). British Journal of Cancer (1999) 81(1), 54–61 © 1999 Cancer Research Campaign Angiogenesis in malignant mesothelioma 59 Table 4 Cox proportional hazard regression analysis using 46 malignant pleural mesothelioma patients with standardized resections Hazard ratio P 95% CI Low VD (n = 13) Gender 3.437 0.0067 1.409 – 8.384 Stage (I, II vs III, IV) 3.729 0.0091 1.387 – 10.025 VD 2.247 0.0326 1.070 – 4.719 *COM-VEGF 0.400 0.0731 0.147 – 1.089 High VD (n = 15) A combination factor of VEGF, flt-1, and KDR. It was defined as ‘positive’ if two or more of the three associated variables are strongly expressed. Years after operation (Kumar-Singh et al, 1997). In this study, higher densitometry Figure 5 Kaplan–Meier survival plots in stage III patients with epithelial type MPM who underwent standardized resections (n = 28). A tumour was index of VEGF compared to the mean value of tumours was found included in the high VD group if VD was greater than 11. High VD tended to in two out of five normal pleural tissues. If pleura is a tissue with be associated with poor survival (P = 0.0866) relatively high VEGF, this particular condition may suit malignant development. Among VEGFRs, the KDR and flt-4 expression levels tended to be higher in the MPM tumours compared to Association with clinicopathological findings normal pleural tissues in this study. Although the percentage of Histopathological information on lymph node metastasis was avail- VEGF overexpression was not so high, our data confirmed that able in 47 cases. The mean LVD and VD were 8.3 ± 1.0 and VEGF expression was significantly associated with VD and the 10.9 ± 1.1 in node-positive cases (n = 30), and 10.2 ± 1.4 and expression of VEGFRs (KDR and flt-1) in MPM tumours. 11.0 ± 1.5 in node-negative cases (n = 17) respectively. Nodal status Furthermore, among the 46 MPM patients undergoing standard- had no relationship with LVD (P = 0.1351) or VD ized resections, high VD was significantly associated with poor (P = 0.8856). VEGF and VEGF-C mRNA expression levels were survival. A combination factor with VEGF and its receptors (flt-1 1.9 ± 0.6 and 0.8 ± 0.2, respectively, in node-positive cases, and and KDR), COM-VEGF, also had a weaker relationship with 1.3 ± 0.2 and 0.9 ± 0.2, respectively, in node-negative cases. The outcome, yet the value only trends toward significance due to the differences were not statistically significant. In stage III patients small sample size. with epithelial type who underwent standardized resections Most conventional immunohistochemistry methods for the detec- (n = 28), the median survival of high and low VD groups were 11 tion of endothelial cells do not distinguish lymphatics from blood and 17 months respectively. The 1-year survival rates for high and vessels. In this study, we tried to assess the LVD within MPM low VD were 40.0% and 61.5% respectively. In this analysis, high tumours using the enzyme-histochemistry assay for 5¢-NA based on VD tended to correlate with poor survival (P = 0.0866, Figure 5). the leading method (Wachstein et al, 1954). 5¢-NA expression in Using all patients with standardized resections (n = 46), age, nodal vascular endothelial cells varies according to vessel types (Turner et status, histological subtype, VEGF-C, flt-4, the COM-VEGF-C and al, 1987; Airas et al, 1997). That is, the activity of 5¢-NA is very high LVD had no impact on survival. As a result of multivariate analysis, in lymphatic endothelial cells, while it is very low or absent in blood male, advanced stage (III and IV) and high VD were independent capillary endothelial cells (Turner et al, 1987). Therefore, this assay negative prognostic indicators. The positive COM-VEGF also has a great potential to discriminate neo-lymphatics from showed weaker correlation with poor survival (Table 4). neo-blood vessels after the adequate inhibition of its activity by paraformaldehyde (Vetter et al, 1970; Ji et al, 1997). In normal lymphatics, this enzyme activity appears to be necessary for the DISCUSSION growth and development of vessels (Ji et al, 1997), and has an Neovascularization has been shown to be necessary for tumour important role in the control of interactions between lymphocytes growth and metastasis (Folkman, 1990). Although some contraries and vascular endothelial cells (Airas et al, 1997). If this enzyme exist, many studies have confirmed the negative impact of tumour activity is essential for the proliferation of lymphatic vessels, the vascularization on prognosis (Chodak et al, 1980; Weidner et al, same condition may be expected in tumour-induced neo-vessels. 1991; Macchiarini et al, 1992). Among the many reported angio- Although the significance of 5¢-NA activity in cancerous lesions is genic factors, VEGF is the most powerful endothelial cell-specific not clear, this enzyme has been reported in some tumour tissues, mitogen associated with tumour neovascularization. A number of including seminoma, malignant fibrous histiocytoma, breast cancer investigators have reported a significant relationship between VD and renal cancer (Wachstein et al, 1954; Wood et al, 1986; Canbolat and VEGF expression in a variety of tumours (Toi et al, 1994; et al, 1996; Durak et al, 1997). In this study, we have used this Mattern et al, 1995, Samoto et al, 1995, Takahashi et al, 1995), and method, for the first time, to study the microlymphatic vessel overexpression of VEGF has been associated with poor prognosis density in MPM tumours and found that 5¢-NA staining was limited in some neoplasms (Toi et al, 1994; Takahashi et al, 1995). to vessels in 31 of the 54 (57.4%) MPM tumour specimens. In However, little information is available on tumour angiogenesis in these samples, lymphatic vessel specificity was optimized by malignant mesothelioma. Recently, Kumar-Singh et al (1997) paraformaldehyde fixation for 30Ð40 min. In 23 cases (42.6%), reported that VD in MPM tumours was significantly higher than 5¢-NA staining was also noted in MPM tumour cells and within that in non-neoplastic mesothelium. They also reported that the stromal elements. In 20 out of these 23 cases, the intensity of the patients with highly vascularized tumours had a significantly non-vessel staining was less than that of the vessels, and LVD could shorter survival than patients with poorly vascularized ones be assessed. In three cases with the intensity of non-specific © 1999 Cancer Research Campaign British Journal of Cancer (1999) 81(1), 54–61 Survival rate (%) 60 Y Ohta et al Durak I, Beduk Y, Kavutcu M, Suzer O, Yaman O, Ozturk HS, Canbolat O and staining, LVD was assessed by avoiding areas with high non- Ulutepe S (1997) Activity of the enzymes participating in purine metabolism of specific activity. Under the condition that very little information is cancerous and non-cancerous human kidney tissues. Cancer Invest 5: 212Ð216 available about the mechanisms of lymphangiogenesis, some Folkman J (1990) What is the evidence that tumors are angiogenesis dependent? reports have recently suggested that the specific pathways involved J Natl Cancer Inst 82: 4Ð6 in lymphangiogenesis are different from those in hemangiogenesis Folkman J (1995) Angiogenesis in cancer, vascular, rheumatoid and other disease. Nat Med 1: 27Ð31 (Wilting et al, 1996; Oh et al, 1997). Among the VEGF family Grimmond S, Lagercrantz J, Drinkwater C, Silins G, Townson S, Pollock P, Gotley members, the function of VEGF-C appears to extend to the D, Carson E, Rakar S, Nordenskjold M, Ward L, Hayward N and Weber G lymphatic system as a ligand for flt-4, and VEGF-C is suspected (1996) Cloning and characterization of a novel human gene related to vascular to play an important role in lymphangiogenesis (Kukk et al, 1996; endothelial growth factor. Genome Res 6: 124Ð131 Hewett PW and Murray JC (1996) Coexpression of flt-1, flt-4 amd KDR in freshly Jeltsch et al, 1997; Oh et al, 1997). In this study, we identified a isolated and cultured human endothelial cells. Biochem Biophys Res Commun strong positive relationship between VEGF-C mRNA expression 221: 697Ð702 and LVD, and flt-4 expression and LVD in MPM, as well as a close Jeltsch M, Kaipainen A, Joukov V, Meng X, Lakso M, Rauvala H, Swartz M, association between flt-4 and VEGF-C expression. Further, the Fukumura D, Jain RK and Alitalo K (1997) Hyperplasia of lymphatic vessels mean vessel count for VD that includes both blood and lymphatic in VEGF-C transgenic mice. Science 276: 1423Ð1425 Ji RC and Kato S (1997) Enzyme-histochemical study on postnatal development of vessels was much greater than that for LVD. VD itself had no rat stomach lymphatic vessels. Microvasc Res 54: 1Ð12 relationship with VEGF-C and flt-4 expression levels in this study. Joukov V, Pajusola K, Kaipainen A, Chilov D, Lahtinen I, Kukk E, Saksela O, These results support the use of the enzyme-histochemistry method Kalkkinen N and Alitalo K (1996) A novel vascular endothelial growth factor, based on 5¢-NA activity for assessment of lymphatics, and also VEGF-C, is a ligand for the flt-4 (VEGFR-3) and KDR (VEGFR-2) receptor tyrosine kinases. EMBO J 15: 290Ð298 suggest that VEGF-C plays a key role in lymphangiogenesis in Kaipainen A, Korhonen J, Mustonen T, van Hinsbergh VWM, Fang G, Dumont D, MPM. In addtion, flt-4 expression, which has originally been found Breitman M and Alitalo K (1995) Expression of the fms-like tyrosine kinase 4 in lymphatic endothelium (Kaipainen et al, 1995; Joukov et al, gene becomes restricted to lymphatic endothelium during development. Proc 1996), was high in both MPM cell lines used and resected tumour Natl Acad Sci USA 92: 3566Ð3570 tissues. Immunohistochemical staining revealed flt-4 expression in Kukk E, Lymboussaki A, Taira S, Kaipainen A, Jeltsch M, Joukov V and Alitalo K (1996) VEGF-C receptor binding and pattern of expression with VEGFR-3 MPM tumour cells except for some vascular endothelial cells. In suggests a role in lymphatic vascular development. Development 122: this study, it was difficult to clarify the correspondence between 3829Ð3837 endothelial cells with flt-4 expression and those with 5¢-NA activity Kumar-Singh S, Vermeulen PB, Weyler J, Segers K, Weyn B, van Daele A, Dirix LY, partly owing to the situation that mesothelioma cells themselves van Oosterom AT and van Marck E (1997) Evaluation of tumour angiogenesis as a prognostic marker in malignant mesothelioma. J Pathol 182: 211Ð216 expressed both flt-4 and 5¢-NA. Although the role of flt-4 in tumour Lee J, Gray A, Yuan J, Luoh S, Avraham H and Wood WI (1996) Vascular cells is not well known, flt-4 expression was confirmed in meso- endothelial growth factor-related protein: a ligand and specific activator of the thelial cells (Hewett et al, 1996). This suggests a possibility that tyrosine kinase receptor flt-4. Proc Natl Acad Sci USA 93: 1988Ð1992 mesothelioma cells express flt-4, and that VEGF-C may have some Macchiarini P, Fontanini G, Hardin MJ, Squartini F and Angeletti CA (1992) role for the autocrine growth or proliferation of tumour cells. Relation of neovascularization to metastasis of non-small-cell lung cancer. Lancet 340: 145Ð146 In conclusion, our study suggests that VEGF and VEGF-C play Mattern J, Koomagi R and Volm M (1995) Vascular endothelial growth factor an important role in angiogenesis and lymphangiogenesis in MPM expression and angiogenesis in non-small-cell lung carcinomas. Int J Oncol 6: tumours, and that VD has a significant impact on the overall 1059Ð1062 survival of MPM patients. The LVD assessed by enzyme-histo- Meyer M, Clauss M, Lepple-Wienhues A, Waltenberger J, Augustin HG, Ziche M, Lanz C, Bittner M, Rziha HJ and Dehio C (1999) A novel vascular endothelial chemistry for 5¢-NA has no impact on nodal metastasis. However, growth factor encoded by Orf virus, VEGF-E, mediates angiogenesis via nodal status does not have a relation with outcome in our samples. signalling through VEGFR-2 (KDR) but not VEGFR-1 (Flt-1) receptor Therefore, a question remains whether VEGF and VEGF-C have a tyrosine kinase. EMBO J 18: 363Ð374 relation with prognosis in cases where nodal status has a great Mossman BT, Kamp DW and Weitzman SA (1996) Mechanisms of carcinogenesis impact on survival. The potential roles of VEGF and VEGF-C in and clinical features of asbestos-associated cancers. Cancer Invest 14: 466Ð480 Oh SJ, Jeltsch MM, Birkenhager R, McCarthy JE, Weich HA, Christ B, Alitalo K lymphatic metastasis warrant further study. and Wilting J (1997) VEGF and VEGF-C: specific induction of angiogenesis and lymphangiogenesis in the differentiated avian chorioallantoic membrane. Dev Biol 188: 96Ð109 ACKNOWLEDGEMENTS Olofsson B, Pajusola K, Kaipainen A, von Euler G, Joukov V, Saksela O, Orpana A, Pettersson RF, Alitalo K and Eriksson U (1996) Vascular endothelial growth We would like to thank Mr Larry Tait for technical assistance. We factor B, a novel growth factor for endothelial cells. Proc Natl Acad Sci USA also thank Professor H Yamamoto of Kanazawa University for his 93: 2576Ð2581 kind gift of the primers for flt-1, flt-4 and KDR. Pass HI, Stevens EJ, Oie H, Tsokos MG, Abati AD, Fetsch PA, Mew DJY, Pogrebniak HW and Matthews WJ (1995) Characteristics of nine newly derived mesothelioma cell lines. Ann Thorac Surg 59: 835Ð8344 REFERENCES Rusch VW (1996) A proposed new International TNM staging system for malignant pleural mesothelioma from the International Mesothelioma Interest Group. Airas L, Niemela J, Salmi M, Puurunen T, Smith DJ and Jalkanen S (1997) Lung Cancer 14: 1Ð12 Differential regulation and function of CD73, a glycosyl-phosphatidylinositol- Samoto K, Ikezaki K, Ono M, Shono T, Kohno K, Kuwano M and Fukui M (1995) linked 70-kD adhesion molecule, on lymphocytes and endothelial cells. J Cell Expression of vascular endothelial growth factor and its possible relation with Biol 136: 421Ð431 neovascularization in human brain tumor. Cancer Res 55: 1189Ð1193 Canbolat O, Durak I, Cetin R, Kavutcu M, Demirci S and Ozturk S (1996) Activities Takahashi Y, Kitadai Y, Bucana CD, Cleary KR and Ellis LM (1995) Expression of of adenosine deaminase, 5¢-nucleotidase, guanase, and cystidine deaminase vascular endothelial growth factor and its receptor, KDR, correlates with enzymes in cancerous and non-cancerous human breast tissues. Breast Cancer vascularity, metastasis, and proliferation of human colon cancer. Cancer Res Res Treat 37: 189Ð193 55: 3964Ð3968 Chodak GW, Haudenschild C, Gittes RF and Folkman J (1980) Angiogenic activity Toi M, Hoshina S, Takayanagi T and Tominaga T (1994) Association of vascular as a marker of neoplasia and preneoplasia in lesions of the human bladder. Ann endothelial growth factor expression with tumor angiogenesis and with early Surg 192: 762Ð771 relapse in primary breast cancer. Jpn J Cancer Res 85: 1045Ð1049 British Journal of Cancer (1999) 81(1), 54–61 © 1999 Cancer Research Campaign Angiogenesis in malignant mesothelioma 61 Turner RR, Beckstead JH, Warnke RA and Wood GS (1987) Endothelial cell Weidner N, Semple JP, Welch WR and Folkman J (1991) Tumor angiogenesis and phenotypic diversity. In situ demonstration of immunologic and enzymatic metastasis Ð correlation in invasive breast carcinoma. N Engl J Med 324: 1Ð8 heterogeneity that correlates with specific morphologic subtypes. Am J Clin Wilting J, Birkenhager R, Eichmann A, Kurz H, Martiny-Baron G, Marme D, Pathol 87: 569Ð575 McCarthy JE, Christ B and Weich HA (1996) VEGF121 induces proliferation Vetter W (1970) Alkaline phosphatasen in mastzellen, blut- und lymphagefŠben der of vascular endothelial cells and expression of flk-1 without affecting rattenzunge. 5¢-nucleotidase-, unspezifische alkalische phosphatase- und lymphatic vessels of chorioallantoic membrane. Dev Biol 176: 76Ð85 polyphosphatase- (ATP¢ase) aktivitŠt unter besondere berŸckscichtigung des Wood GS, Beckstead JH, Turner RR, Hendrickson MR, Kempson RL and Warnke pH. Z Anat Entwicklungsgesch 130: 153Ð176 RA (1986) Malignant fibrous histiocytoma tumor cells resemble fibroblasts. Wachstein M and Meisel E (1954) The histochemical distribution of 5-nucleotidase Am J Surg Pathol 110: 323Ð335 and unspecific alkaline phosphatase in the testicle of various species and in two Yamada Y, Nezu J, Shimane M and Hirata Y (1997) Molecular cloning of a novel human seminomas. J Histochem 2: 137Ð148 vascular endothelial growth factor, VEGF-D. Genomics 42: 483Ð488 © 1999 Cancer Research Campaign British Journal of Cancer (1999) 81(1), 54–61

Journal

British Journal of Cancer – Springer Journals

Published: Aug 13, 1999

There are no references for this article.