Selective upregulation of vascular endothelial growth factor receptors neuropilin-1 and -2 in human neuroblastoma
Abstract
BACKGROUND
Recent studies show that vascular endothelial growth factor (VEGF) and its receptors Flt-1 and KDR, and a series of other angiogenic molecules, are upregulated in advanced but not low stage human neuroblastoma. Neuropilin-1 and 2 (NRP) are novel specific receptors of VEGF165, whose role is unknown in human neuroblastoma.
METHODS
Tissue biopsies of 37 children with Stage I-IV neuroblastoma were obtained, as well as biopsies of 7 normal adrenals as controls. The mRNA expression of VEGF165 and its receptors Flt-1, KDR, NRP1, and NRP2 was evaluated by real-time reverse transcripton polymerase chain reaction. NRP protein expression was detected by immunocytochemistry and Western blotting.
RESULTS
VEGF165 mRNA was upregulated in Stage III and IV and Flt-1 and KDR gene expression was increased in Stage III, while NRP1 and 2 mRNA and protein levels were higher in Stages I-IV vs. controls (P < 0.05). NRP was expressed in vascular endothelial but not tumor cells.
CONCLUSIONS
These results show for the first time that human neuroblastoma expresses NRP, and that NRP co-regulates VEGF angiogenic effect in human neuroblastoma. NRP might be a sensitive angiogenic measure of VEGF systems in neuroblastoma, particularly in its early stages. Cancer 2002;94:258–63. © 2002 American Cancer Society.
Neuroblastoma, the most common extracranial solid malignancy of childhood, is thought to develop from pluripotent cells of the embryonic neural crest. Poor prognosis is expected in patients with advanced Stages III and IV, of whom about 30% and almost 80%, respectively, die from the disease.1 The pathogenetic changes determining poor outcome are largely unknown. The invasive, metastatic, and hypervascular nature of high-stage neuroblastoma may be one of the key obstacles to the cure of this disease. Evidence is accumulating that angiogenic molecules may be prognostic factors for a variety of solid tumors,2-6 and recent studies show that gene expression of vascular endothelial growth factor (VEGF) is upregulated in advanced but not early stage human neuroblastoma.7
VEGF165 is an endothelial cell mitogen in vitro and angiogenesis factor in vivo whose activities are mediated by the VEGF tyrosine kinase receptors Flt-1 and Flk-1/KDR.4-6, 8, 9 Neuropilin-1 (NRP1), a transmembrane glycoprotein first described in the developing nervous system,10 has been shown to be an isoform-specific VEGF receptor that binds VEGF165 but not VEGF121 to the surface of endothelial and tumor cells.11 Expression of NRP1 in endothelial cells enhances the binding of VEGF165 to KDR, as well as enhancing VEGF165-induced endothelial cell chemotaxis.11 Thus, it appears that, in endothelial cells, NRP1 acts as a co-receptor that enhances KDR activity, and subsequently enhances angiogenesis. NRP1 is a part of a receptor system that also includes NRP2, a receptor that displays highly similar structural features, sharing 44% of its amino acids with NRP1.12 Since neuroblastoma is hypervascular7 and originates from the neural crest,1 and because NRP1 is both neurogenic and angiogenic,10, 11 we hypothesized that human neuroblastoma co-expresses NRP/VEGF165/Flt-1/KDR. We found upregulated VEGF165 mRNA in Stages III and IV, and increased Flt-1 and KDR gene expression in Stage III. Further, we show, to our knowledge for the first time, that NRP is expressed in vascular endothelial cells of neuroblastoma and that NRP1 and 2 mRNA levels are upregulated in Stages I–IV neuroblastomas vs. controls.
MATERIALS AND METHODS
Patients
This study was approved by the Institutional Committee for Human Studies of the Medical Faculty at the University of Vienna. Parents gave informed consent for their children to be involved in the study and for biopsies taken during surgery to be used in the study. A total of 37 patients diagnosed with neuroblastoma were investigated. Patients were scheduled for surgical intervention and had routine radiologic and medical programs prior to surgery. The control group was made up of seven children with normal adrenals, admitted for nephrectomy. The patients' case histories, the results of radiologic tests (computed tomography scans, magnetic resonance imaging), and the histopathologies of the excised tumors were documented. Each patient was given a code and, except for one member of our group who had access to patients' data, the authors were unaware of patients' case histories until the study was finished and the key was broken. Biopsies of all tumors and control organs were taken intraoperatively and examined using real-time polymerase chain reaction (PCR), immunocytochemistry, and Western blotting as described next.
Real-time PCR
Total RNA was isolated from the homogenized tumor biopsies by a standard guanidinium thiocyanate-phenol-chloroform extraction.13 cDNA was synthesized using avian myeloblastosis virus reverse transcriptase and 2 μg of total RNA primed with oligo dT-primer.14 After reverse transcription of RNA into cDNA, real-time PCR was used to monitor gene expression using a Light Cycler Instrument (Roche, Mannheim, Germany) according to the standard procedure. Primers for real-time PCR were designed using Prime software (Genetics Computer Group Package, WI). For sequences of oligonucleotides used, see Table 1. PCR was performed using “Hot Start” reaction mix (Light Cycler – FastStart DNA Master SYBR Green I; Roche, Mannheim, Germany). cDNA (1 μL of the reverse transcripton reaction) was added to a 20-μL PCR reaction mixture according to the manufacturer's protocol. The temperature profile included an initial denaturation for 10 minutes at 95 °C, followed by 37 cycles of denaturation at 95 °C for 15 seconds, annealing at different temperatures (Table 1) for 5 seconds, elongation at 72 °C (elongation time depended on the size of the fragment; the number of base pairs divided by 25 yielded the time in seconds), and fluorescence monitoring at 85 °C. Light Cycler Data Analysis Software (Roche) Version 3.1.102 was used for PCR data analysis. The specificity of the amplification reaction was determined by performing a melting curve analysis. Standard curves for expression of each gene were generated by serial dilution of known quantities of the respective cDNA gene template. Relative quantification of the signals was done by normalizing the signals of the different genes with the β-actin signal.
| cDNA and primer | Primer sequence | Annealing temperature |
|---|---|---|
| VEGF165 | ||
| sense | 5′-AGCCTTGCCTTGCTGCTCTA-3′ | 60 °C |
| antisense | 5′-GTGCTGGCCTTGGTGAGG-3′ | |
| Flt-1 | ||
| sense | 5′-GTCACAGAAGAGGATGAAGGTGTC-3′ | 60 °C |
| antisense | 5′-CACAGTCCGGCACGTAGGTGATT-3′ | |
| KDR | ||
| sense | 5′-GCATCTCATCTGTTACAGC-3′ | 64 °C |
| antisense | 5′-CTTCATCAATCTTTACCCC-3′ | |
| Neuropilin-1 | ||
| sense | 5′-AAAAGCCCACGGTCATAG-3′ | 60 °C |
| antisense | 5′-TGTCATCCACAGCAATCC-3′ | |
| Neuropilin-2 | ||
| sense | 5′-CAAGTTGCTGTGGGTCATC-3′ | 60 °C |
| antisense | 5′-AATTGCTCCAGTCCACCTC-3′ | |
| β-Actin | ||
| sense | 5′-TGCCATCCTAAAAGCCAC-3′ | 58 °C |
| antisense | 5′-TCAACTGGTCTCAAGTCAGTG-3′ |
- VEGF: vascular endothelial growth factor.
Neuropilin Immunostaining
Cryosections 7 μm thick were prepared from all tumor and control biopsies. These sections were incubated with a primary antibody (90 minutes at room temperature) against NRP (rabbit polyclonal, 1:400 dilution, Santa Cruz Biotechnologies, Santa Cruz, CA). A biotinylated secondary antibody was used to detect positive staining (anti-rabbit: 1:100 dilution; Amersham Pharmacia Biotech, Buckinghamshire, UK). Texas-Red conjugated avidin (1:100 dilution; Amersham Pharmacia Biotech) was used to visualize the antibody reaction. Negative controls were generated by omitting the respective first antibody. Sections were washed twice in phosphate-buffered saline and mounted in Aquatex mounting medium (Merck, Darmstadt, Germany) containing 0.01% 4',6-Diamidino-2-Phenylindole (DAPI; Sigma, St. Louis, MO). Labeled sections were analyzed and digitized using a confocal laser scanning microscope (Olympus Fluoview IX70; UPlanFl 60x objective, Tokyo, Japan) and a digital camera (Olympus DP50).
Western Blotting
Biopsies were lyzed in solubilization buffer (10 mM Tris-Cl, 50 mM NaCl, 1% Triton X-100, 30 mM sodium pyrophosphate, 100 μM Na3VO4, 1 mM phenylmethylsulfonyl fluoride, 1× Complete™ - EDTA-free Protease Inhibitor Cocktail [Roche, Mannheim, Germany]). Insoluble material was removed by centrifugation (15,000 rpm, 10 minutes, 4 °C). Tissue lysates (50 μg/lane) were separated by 7% sodium dodecyl sulfate polyacrylamide gel electrophoresis prior to electrophoretic transfer onto Hybond C super (Amersham Pharmacia Biotech). The blots were probed with a polyclonal antibody against NRP (Santa Cruz Biotechnology) prior to incubation with horseradish peroxidase-conjugated secondary antibody (Amersham Pharmacia Biotech). Proteins were immunodetected on the membrane using chemiluminescence (ECL, Amersham Pharmacia Biotech) and specific protein bands were quantified using Easy plus Win 32 software (Herolab, Wiesloch, Germany).
Data Analysis
The mRNA expression levels of all molecules were compared between control and neuroblastoma tissue biopsies by using analysis of variance. Data are expressed as mean values and statistical significance is set at P < 0.05.
RESULTS
Table 2 summarizes patient demographic data, as well as tumor staging and location. Most of the tumors were located in the adrenal with clinical Stages I-IV. VEGF165 gene expression was significantly upregulated in Stage III and IV tumor biopsies vs. controls (P < 0.05; Fig. 1). There was no statistically significant difference between the VEGF165 mRNA expression in biopsies of Stage I-II neuroblastomas vs. control biopsies. VEGF receptor 1, Flt-1, had a significantly higher gene expression in tumor biopsies of Stage III, but not Stage I, II, and IV, as compared to control tissues (Fig. 1). VEGF receptor 2, KDR, showed a similar gene expression pattern, and its mRNA levels were upregulated in Stage III tumor biopsies, but not Stage I, II, and IV, as compared to control biopsies (Fig. 1). NRP1 mRNA levels were significantly increased in all tumor stages vs. controls, and Stage I neuroblastoma had a higher NRP1 gene expression in comparison to Stages II-IV (P < 0.05; Fig. 1). NRP2 gene expression was upregulated in Stages I-IV tumor biopsies vs. controls, with Stages III and IV having higher NRP2 mRNA levels vs. Stages I and II (P < 0.05; Fig. 1). Having shown that VEGF receptor NRP might be related to neuroblastoma staging at mRNA level, we examined NRP protein expression in tissue sections. We found that vascular endothelial cells surrounded by neuroblastoma cells, but not the neuroblastoma cells themselves, express NRP (Fig. 2). Endothelial cell nuclei with specific NRP positive staining could be clearly identified in DAPI staining, and hence vessels showing positive NRP immunostaining could be clearly localized by digitized comparison (overlay) of NRP and DAPI immunostaining (Fig. 2). Western blotting and negative controls in immunocytochemistry verified the signals in tissue sections as NRP specific. NRP immunostaining was localized in capillaries as well as in post capillary venules in the early stages of neuroblastoma, while in advanced stages of the tumor, NRP immunostaining was largely localized in mid-sized and larger vessels (Fig. 2).
| Group | Age (months) | Gender (M/F) | Tumor location |
|---|---|---|---|
| Control | 21 ± 17 | 3/4 | — |
| Stage Ia | 7 ± 4 | 5/3 | Adrenal |
| Stage II | 9 ± 3 | 7/3 | 9 Adrenal/1 vagus nerve |
| Stage III | 10 ± 4 | 4/5 | 8 Adrenal/1 sympathetic trunk |
| Stage IV | 14 ± 11 | 6/4 | Adrenal |
- a Staging according to Evans.
Vascular endothelial growth factor165(VEGF), Flt-1, KDR, and neuropilin (NRP) mRNA expression in neuroblastoma and control biopsies. VEGF165 gene expression is significantly elevated in tumor Stages III and IV vs. controls (A, asterisks). Flt-1 gene expression is upregulated in neuroblastoma Stage III vs. controls (B, asterisk). Similarly, KDR gene expression is increased in tumor Stage III vs. controls (C, asterisk). Both NRP1 (D) and NRP2 (E) gene expression are significantly increased in neuroblastoma Stages I-IV vs. control biopsies (asterisks). Asterisks mark significance vs. controls, (†) in D marks significance vs. Stages II-IV, (†) in E marks significance vs. Stage II, and (‡) in E marks significance vs. Stages I and II. P < 0.05.
Representative double staining immunocytochemistry (neuropilin [NRP] and DAPI staining) and Western blot images of NRP in neuroblastoma. In Stage I neuroblastoma (A) many capillaries and postcapillary venules are positively stained for NRP (arrows). Bar = 50 μm. B is an overlay negative control image (omitting the first antibody against NRP) and visualizing cell nuclei with DAPI staining. No specific NRP immunostaining is visible; only neuroblastoma cancer cells are visualized (arrows). Bar = 50 μm. C is a Western blot image showing the specific 135 kDa band for NRP. In Stage II neuroblastoma (D), NRP positive immunostaining is visible in endothelial cells of capillaries (arrow) as well as postcapillary venules (arrowheads). Bar = 50 μm. E is the corresponding double staining overlay image to D, showing NRP red immunostaining and endothelial cell nuclei with DAPI staining (arrowheads and arrow). Bar = 50 μm. NRP immunostaining (F) and the corresponding DAPI double staining overlay image (G) of a neuroblastoma Stage IV section. A mid-sized vein with positive NRP (F) and DAPI (G) staining is shown (arrows). Bar = 25 μm. Staging according to Evans.
CONCLUSIONS
To our knowledge, this study reports for the first time that human neuroblastoma expresses the novel VEGF receptors NRP1 and NRP2, and that the expression of these VEGF receptors is upregulated not only in advanced but also in early stage neuroblastoma. A recent study has shown that overexpression of NRP1 in an experimental tumor model in rats resulted in larger tumor volumes and enhanced angiogenesis.15 Besides an apparent increase in blood vessel density, the vessels in NRP1-overexpressing tumors were more dilated, suggesting possible increased blood flow in the tumors. NRP1 expression also decreased tumor cell and tumor vascular endothelial cell apoptosis, indicating a survival function for NRP1.15 Previous results had shown that endothelial cell NRP1 acts as a co-receptor enhancing VEGF receptor-2 mediated chemotactic activities.15 Expression of NRP1 in tumor cells enhances binding of VEGF165 to these cells, and greatly enhances VEGF protein levels.16 Alternatively, VEGF165 could stimulate tumor cells directly via NRP1 and NRP2.15 NRP1 overexpression might also induce downstream genes that are responsible for enhancing tumor vascularization. The current NRP protein data show that in human neuroblastoma, NRP is only expressed in vascular endothelial cells, indicating that this protein is mainly involved in mediating or enhancing the neuroblastoma-induced angiogenesis,15 rather than mediating (or enhancing) the interaction of VEGF with neuroblastoma cells.16 In other words, the likelihood of existence for an autocrine regulatory loop for tumor cell progression and/or invasion in human neuroblastoma, co-mediated by VEGF and its receptor NRP, is very low. In any case, the current results support the co-expression/co-regulation of VEGF165 receptors, KDR and NRP, in human neuroblastoma, explain the highly hypervascular nature of this malignancy, and add evidence to the signaling/regulation pathway of VEGF system in neuroblastoma.
NRP1 enhances the binding of VEGF165 to KDR.
The chemotaxis of endothelial cells co-expressing NRP1 and KDR toward a gradient of VEGF165 is enhanced compared to those cells expressing KDR alone.
Blocking VEGF165 access to NRP1 substantially inhibits VEGF165 binding to KDR and its mitogenic activity. The current study provides evidence that human neuroblastoma expresses VEGF165, KDR, NRP1, and NRP2. The fact that NRP1 and 2 but not KDR mRNA expression was significantly upregulated in the early stages of neuroblastoma, together with strong NRP immunostaining in vascular endothelial cells, suggests that early steps of angiogenesis in this tumor are mainly mediated by VEGF165 receptors NRP1 and 2, and that angiogenesis in advanced stages of neuroblastoma is co-mediated/co-regulated by both NRP and KDR.7, 11, 12 The shift of NRP expression pattern toward advanced stages of the tumor both at mRNA and protein levels supports our assumptions with regard to co-regulation of neuroblastoma-induced angiogenesis by VEGF and its receptor family, although another, yet undefined role of NRP in neuroblastoma might exist. The reasons for upregulated KDR expression levels in Stage III but not Stage IV remain unclear and deserve further study.
Recent studies suggest that several factors have a biologic role in neuroblastoma angiogenesis, contributing synergistically to a more aggressive unfavorable tumor biology.7, 19 The ubiquitous expression of many angiogenesis stimulators in neuroblastoma (i.e., VEGF-A, B, C, angiopoetin 1 and 2, platelet derived growth factor,7, 17-20 and NRP) suggests that neuroblastoma-induced angiogenesis is a multi-step process. This may also explain different expression levels of VEGF isoforms or its receptors Flt-1 and KDR7, 18 in lower vs. advanced-stages of neuroblastoma. Therefore, these molecular mechanisms of neuroblastoma-induced angiogenesis7, 15-20 might be considered for designing future angiostatic therapeutic strategies. Finally, the finding that NRP1 and 2 gene expression is significantly upregulated in early and advanced stage neuroblastoma, while VEGF165 and KDR mRNA levels are elevated in advanced stage neuroblastoma, might constitute a novel diagnostic and prognostic tool. The present NRP protein expression studies support the applicability of these novel VEGF receptors for diagnosis of human neuroblastoma.
