Representative photomicrographs of staining for vascular endothelial growth factor receptors (VEGFRs) in head and neck squamous cell carcinoma tissue: A, VEGFR-1 in tumor cells and fibroblasts (arrow); B, VEGFR-2 in tumor cells; C, VEGFR-3 in tumor cells and vascular endothelial cells (arrow); and D, negative control for VEGFR-1 demonstrates lack of staining (immunohistochemical stain, original magnification ×40).
Mean staining intensity for the vascular endothelial growth factor receptors (VEGFRs) for different cell populations in head and neck squamous cell carcinoma tissue. Data were analyzed for 2 groups: (1) for all tumor samples studied for each receptor and (2) for only the 10 common tumors that were studied for all 3 receptors. An asterisk indicates statistically significant difference compared with the staining intensity of vascular endothelial cells (VECs) for the same receptor in the same group. TC indicates tumor cell; MAC, macrophages; and FIB, fibroblasts.
Representative photomicrographs of staining for vascular endothelial growth factor receptors (VEGFRs) in a cultured oral squamous cell carcinoma cell line (SCC-25): A, VEGFR-1; B, VEGFR-2; C, VEGFR-3; and D, negative control for VEGFR-1 demonstrates lack of staining (immunohistochemical stain, original magnification ×40).
Lalla RV, Boisoneau DS, Spiro JD, Kreutzer DL. Expression of Vascular Endothelial Growth Factor Receptors on Tumor Cells in Head and Neck Squamous Cell Carcinoma. Arch Otolaryngol Head Neck Surg. 2003;129(8):882-888. doi:10.1001/archotol.129.8.882
Angiogenesis is essential for the growth of solid tumors, including head and neck squamous cell carcinoma (HNSCC). Angiogenesis is regulated by angiogenic factors such as vascular endothelial growth factor (VEGF) and VEGF receptors (VEGFRs) 1, 2, and 3 known to be located on vascular endothelial cells (VECs). We hypothesize that VEGFRs are also expressed on HNSCC tumor cells in vitro and in vivo and likely control tumor function in vivo.
Immunohistochemical analysis for VEGFR-1 (n = 13), VEGFR-2 (n = 21), and VEGFR-3 (n = 16) was performed on human HNSCC tumor samples. Specimens were analyzed for receptor expression and staining intensity. A cultured oral SCC cell line (SCC-25) and a pharyngeal SCC cell line (FADU) were also studied for receptor expression.
The HNSCC tumor cells expressed VEGFR-1, VEGFR-2, and VEGFR-3 in all specimens evaluated. Staining for all 3 receptors was also found on tumor-associated macrophages and fibrobasts, except that VEGFR-2 was not present on fibroblasts. Staining intensity for VEGFR-1 and VEGFR-2 was significantly higher in tumor cells and macrophages than in VECs stained for the same receptor. Both cultured HNSCC cell lines demonstrated expression of all 3 receptors.
This represents the first report of all 3 VEGFRs being expressed by HNSCC cells. These findings indicate that VEGF may be an autocrine regulator of tumor cell activity in addition to its known angiogenic effects on VECs. The presence of VEGFRs on tumor-associated macrophages and fibroblasts contributes to the complexity of the VEGF/VEGFR system in human cancer.
TUMOR ANGIOGENESIS has been described as the formation of new blood vessels in the tumor microenvironment. It is now well accepted that angiogenesis is vital for continued growth and metastasis of solid tumors. Tumors that are not vascularized are limited in size and do not metastasize.1 Additional studies have also indicated that increased tumor angiogenesis (microvessel density) correlates with poor outcome in several human tumors.2 In head and neck squamous cell carcinoma (HNSCC), there have been conflicting studies on the relationship between tumor angiogenesis and prognosis. However, a recent review of the studies in this field concluded that most studies in HNSCC including large numbers or focusing on specific tumor groups have found positive associations of microvessel density with nodal metastasis or recurrence after treatment.3
Previous work in our laboratory has demonstrated an elevated expression of vascular endothelial growth factor (VEGF), an important angiogenic factor, in the ex vivo HNSCC tumor microenvironment compared with normal controls.4 Other investigators have addressed the pivotal role of VEGF in HNSCC angiogenesis,5 and indeed in the growth of human cancers in general.6 The VEGF family consists of several growth factors known to act directly on vascular endothelial cells (VECs), inducing endothelial cell proliferation, migration, and chemotaxis.7 In HNSCC, as well as other tumor systems, VEGF has been shown to be expressed by tumor cells and to induce proliferation of adjacent VECs via a paracrine mechanism.5,8 Recent evidence has demonstrated that VEGF may also have a direct effect on tumor cell activity.9 These data suggest a possible autocrine mechanism for VEGF-regulated vasculogenesis and tumorigenesis.
Vascular endothelial growth factor interacts with a family of class III tyrosine-kinase receptors, kinase insert domain–containing receptor (KDR or VEGF receptor [VEGFR] 2), and Fms-like tyrosine kinases 1 and 4 (Flt-1 or VEGFR-1 and Flt-4 or VEGFR-3).10 Previously considered to be localized primarily to VECs, VEGFR-1 and VEGFR-2 specifically bind VEGF and other VEGF family members. In contrast, VEGFR-3 is considered a specific marker for lymphatic vessels in most normal tissues, including skin11 and lymph nodes.12 Some investigations have also localized VEGFR-1 and VEGFR-2 to specific tumorigenic cell types, such as melanoma and ovarian and pancreatic carcinoma.9 Recently, VEGFR-3 expression has been demonstrated on AIDS-associated Kaposi sarcoma spindle cells and associated lymphatic endothelial cells.12
Since VEGF regulation of angiogenesis appears to be critical for tumor growth and invasion, it is important to begin to fill the gaps in our knowledge of the distribution and role of angiogenic factors and their receptors. To date, however, there are no published reports examining the expression of all 3 known VEGFRs directly on HNSCC tumor cells. Vascular endothelial growth factor and its family members are well described as potent growth factors for VECs, but relatively little is known regarding their effects on other cell types. Evaluating the distribution and level of expression of these VEGFRs on tumor cells seems to be a logical step in further understanding the complexity of VEGF-regulated tumor cell growth, proliferation, and metastasis. Since our laboratory has demonstrated that VEGF levels are elevated in the HNSCC tumor microenvironment,4 we hypothesized that the specific receptors for this ligand are also distributed throughout this microenvironment. We found that VEGFR-1, VEGFR-2, and VEGFR-3 are expressed by HNSCC tumor cells in vivo and in vitro. This may represent an autocrine mechanism for tumor growth in response to VEGF that complements its proliferative effects on endothelial cells. Other cellular elements such as macrophages and fibroblasts demonstrate varying levels of expression, adding to the complexity of this system.
A total of 23 archival, formalin-fixed, paraffin-embedded HNSCC specimens were obtained from the Department of Pathology, University of Connecticut School of Medicine. Specimens were selected after initial review of hematoxylin-eosin–stained slides to contain tumor from primary oral, pharyngeal, and laryngeal sites. These patients had undergone diagnostic biopsy or primary surgical resection at the University of Connecticut Health Center during the years 1989 through 1997. Table 1 lists the primary tumor site, stage, and which VEGFR(s) each tumor was studied for. Owing to limitations in specimen availability, total numbers of specimens were not equal in each receptor group. Thus, 13 tumors were studied for VEGFR-1 expression, 21 for VEGFR-2, and 16 for VEGFR-3. Of these, 10 common tumors were studied for all 3 receptors.
Immunohistochemical analysis of the tissue specimens was performed by indirect immunoperoxidase staining. For receptor studies, the 5-µm paraffin-embedded tumor sections were deparaffinized in xylene and rehydrated in sequential graded alcohol (100%, 95%, 70%, and 50%). To inhibit endogenous peroxidase, the sections were immersed in 100% methanol containing 0.01% hydrogen peroxide for 20 minutes at room temperature. In addition, a target unmasking step was performed prior to the initial blocking step by immersing the slides in 0.01M sodium citrate at 90° C for 5 minutes and rinsing in phosphate-buffered saline (PBS). This procedure unmasks the epitopes from the formalin bonds to improve antibody recognition.
The specimens were then incubated with 5% goat or rabbit blocking serum (Vector, Burlington, Calif) for 1 hour at room temperature. After washing with PBS, each slide was reacted with primary antibody overnight at 4°C. The primary antibodies, goat polyclonal antihuman VEGFR-1, mouse monoclonal antihuman VEGFR-2, and rabbit polyclonal antihuman VEGFR-3 were obtained from Santa Cruz Biotechnologies (Santa Cruz, Calif) and used at 1:200, 1:100, and 1:300 dilutions, respectively. These concentrations had been determined in preliminary experiments to be optimal concentrations for each antibody. A biotinylated rabbit antigoat, rabbit antimouse, or goat antirabbit secondary antibody (Vector) at 1:200 was added after the overnight incubation for 1 hour at room temperature.
Following a PBS wash, horseradish peroxidase streptavidin (Zymed, San Francisico, Calif) at 1:100 was added to the slides for 45 minutes at room temperature. The slides were washed in PBS, and the reactions were developed in a 3-amino-9-ethylcarbazole solution, counterstained with hematoxylin, rinsed with distilled water, and coated in crystal mounting solution for microscopic analysis.
Negative controls were prepared on a matched set of slides as follows: For VEGFR-1 and VEGFR-3, the primary antibody was preincubated with a 5-fold excess (by weight) of its specific antigen before being applied to the slides. The antigen was supplied as a blocking peptide that included epitopes that the antibody is known to react with. For VEGFR-2, a blocking peptide specific for the primary antibody used was not available. Therefore, control slides were treated with nonimmune serum (mouse IgG) instead of the primary antibody. Further steps and reagents for the controls were identical to those described above.
Human oral SCC (SCC-25) and human hypopharyngeal SCC (FADU) cells were purchased from American Type Culture Collection, Rockville, Md. All cells were grown on polystyrene culture flasks (Becton, Dickinson and Company, Lincoln Park, NJ). The cell culture medium for SCC-25 cells was composed of 90% 1:1 mixture of Ham F12 and Dulbecco modified Eagle medium (Invitrogen, Carlsbad, Calif) with 0.4 µg/mL of hydrocortisone and 10% fetal bovine serum (Hyclone, Logan, Utah). The cell culture medium for FADU cells consisted of 90% minimum essential medium (Eagle) with 0.1mM nonessential amino acids, 1.0mM sodium pyruvate, and Earle balanced salt solution (Invitrogen), and 10% fetal bovine serum (Hyclone). All cells were grown under 5% carbon dioxide and humidified conditions at 37°C. Routine subcultures were prepared at 1:3 split ratios with 0.05% trypsin/0.02% edetic acid (Invitrogen).
For immunocytochemical analysis, cells were grown directly on Lab-Tek tissue culture chamber slides (Miles Scientific, Naperville, Ill) and fixed in 10% formalin at appropriate subconfluent cell density. Immunocytochemical analysis was performed using the reagents and indirect immunoperoxidase staining technique described above, with the exception of the deparaffinization, rehydration, and target unmasking steps. Negative controls were prepared as described above.
The stained slides were evaluated in a blinded fashion by 2 observers, one of whom is highly experienced in evaluation of immunohistochemical results. Expression of the receptors was evaluated in HNSCC tumor cells, VECs, infiltrating macrophages, and fibroblasts. Intensity of staining was evaluated for each cell group on a 0 to 4 grading system, with 0 signifying no staining and 4 signifying intense staining.
Data were analyzed using the JMP statistical package (SAS Institute Inc, Cary, NC). Data for each receptor were expressed as mean staining intensity (SI) ± SD for each cell type. Since VECs were previously known to express all 3 VEGFRs, we compared staining for each receptor in tumor cells, fibroblasts, and macrophages against staining for the same receptor in VECs. These comparisons were carried out for all tumor samples studied and also for the group of 10 common tumors that were studied for all 3 receptors. Group means were compared using the t test. Differences were considered statistically significant if the P values were less than or equal to .05.
Data for each receptor are expressed as mean SI ± SD for each cell type. We initially evaluated HNSCC specimens for the high-affinity receptor VEGFR-1 (Flt-1). Tumor cells in 100% of specimens evaluated for VEGFR-1 stained positively (Figure 1A) with moderate intensity of staining (2.15 ± 0.69). Vascular endothelial cells demonstrated less consistent (66%) and less intense (0.66 ± 0.65) staining. Interestingly, tumor-associated macrophages and fibroblasts (Figure 1A, arrow) stained positive for VEGFR-1 in almost all specimens (91% and 100%, respectively). All control slides were negative for expression of VEGFR-1 (Figure 1D). Staining intensity for VEGFR-1 in tumor cells, macrophages, and fibroblasts was significantly higher than that in VECs (P<.001 for all 3) (Figure 2).
We next evaluated the HNSCC specimens for VEGFR-2 (KDR). Expression of VEGFR-2 (Figure 1B) was evident on tumor cells in 100% of specimens with moderate staining intensity (1.95 ± 0.67). Staining for VEGFR-2 on VECs was also present (71%) but with weaker intensity (0.48 ± 0.46). Expression on macrophages was uniformly present and very intense (4.00 ± 0.00) in all specimens assayed. Interestingly, expression of VEGFR-2 was not found on tumor-associated fibroblasts in any specimens. Control specimens were consistently negative for VEGFR-2 expression (not shown). Staining intensity for VEGFR-2 in tumor cells and macrophages was significantly higher than that in VECs (P<.001 for both) (Figure 2).
Staining for VEGFR-3 (Flt-4) by tumor cells in all HNSCC specimens was found to be intense (3.25 ± 0.68). A high-intensity staining (2.87 ± 1.10) was also seen on VECs in all specimens (Figure 1C, arrow). Strong intensity of staining was also seen for macrophages (3.15 ± 0.99) and fibroblasts (2.86 ± 1.12) in all specimens. All control slides were negative for VEGFR-3 expression (not shown). No significant differences were found when VEGFR-3 expression in VECs was compared with that in tumor cells, macrophages, or fibroblasts (Figure 2).
Immunocytochemical analysis using anti-VEGFR-1, anti-VEGFR-2, and anti-VEGFR-3 antibodies was performed on pharyngeal (FADU) and oral (SCC-25) HNSCC lines. Strong expression was found for all 3 receptors on the oral SCC cells (SCC-25) (Figure 3A-C), with no staining on control slides for VEGFR-1 (Figure 3D). Controls for the remaining receptors were also consistently negative for staining (not shown). Positive staining for VEGFR-1, VEGFR-2, and VEGFR-3 was also noted on the pharyngeal SCC (FADU) cell line, with no staining of controls (not shown).
Angiogenesis, the process of formation of new blood vessels, has been shown to be important for the growth and metastasis of tumors in general and specifically for HNSCC.1,3 The importance of angiogenesis in tumor growth and metastasis is based on the concept that solid tumors cannot grow beyond a few millimeters in diameter without new blood vessel development. Angiogenesis is regulated by a network of chemical signals (angiogenic factors) that interact with angiogenic factor receptors, which control the growth, differentiation, and maturation of new capillaries. Angiogenic factors such as interleukin 8 and basic fibroblast growth factor are produced in elevated levels by HNSCC tumor cells.13 Receptors for these cytokines are also present on tumor cells themselves, in addition to VECs, implying that tumor cells have the ability to regulate their own growth. These complex paracrine/autocrine pathways likely interact to promote angiogenesis and uncontrolled tumor proliferation and to regulate the metastatic processes. Understanding the distribution and role of angiogenic factors and their receptors in the progression of HNSCC will be essential to the development and design of new therapeutic strategies.
The primary member of the VEGF family,VEGF-A, is a 45-kDa heparin-binding protein considered to be the central regulator of angiogenesis in normal growth states (such as embryogenesis) and in the growth and metastasis of solid tumors.10 Additional growth factors from the VEGF family (VEGF-B, VEGF-C, VEGF-D and VEGF-E) and placental growth factors (PlGF1 and PlGF2) share common VEGFRs. Receptor specificities and functions of these various growth factors are summarized in Table 2. Our laboratory has demonstrated that elevated levels of VEGF (ie, VEGF-A) are present in the HNSCC tumor microenvironment and may be related to tumor aggressiveness.4 Specifically, VEGF was expressed by HNSCC tumor cells and the endothelial cells of microvessels located in and around areas of tumor. Benefield et al5 demonstrated that VEGF in supernatants from HNSCC cells mediates endothelial cell proliferation. This response was attenuated by treating the supernatants with anti-VEGF, supporting the concept that VEGF is an important angiogenic factor in HNSCC.
The VEGFR family consists of 3 receptor-type tyrosine kinases, denoted VEGFR-1 (Flt-1), VEGFR-2 (KDR/Flk-1), and VEGFR-3 (Flt-4). The expression of VEGFR-1 and VEGFR-2 has been described mainly on VECs.10 In addition, nonendothelial localization of VEGFR-1 and VEGFR-2 has been described on various tumorigenic cell types9 and in human peripheral blood monocytes.14 Expression of VEGFR-3 has been described on lymphatic endothelium in several normal tissues11 and on Kaposi sarcoma spindle cells.12 Signaling via these receptors involves their dimerization, autophosphorylation, and coupling to signal transduction pathways, which results in various biologic responses. Elaboration of metalloproteases and specific integrins and the initiation of cell proliferation and migration results from their activation.10 Further evidence to support the role of these receptors in tumorigenesis arises from studies in which inhibition of normal receptor function results in inhibition of tumor progression in glioblastoma.15
We performed immunohistochemical analysis for VEGFR-1, VEGFR-2, and VEGFR-3 on ex vivo HNSCC tissue from surgical specimens. Essentially, all 3 VEGFRs were consistently expressed on tumor cells and VECs in our specimens. In addition, all 3 receptors demonstrated specific staining on cultured oral and pharyngeal SCC cell lines. The presence of VEGFRs on VECs is consistent with previous studies indicating that the VEGF family is an important paracrine mediator for angiogenesis. Furthermore, VECs themselves produce VEGF, indicating an autocrine mechanism for VEGF-induced angiogenesis.16 Importantly, the presence of VEGFRs on HNSCC tumor cells suggests an autocrine regulatory function for VEGF in tumor growth. This concept is supported by several studies in other tumor systems. Masood et al9 found that melanoma, ovarian cancer, and pancreatic cancer cell lines express VEGFR-1 and VEGFR-2. Inhibition of VEGF using antisense and inhibition of VEGFRs using antibodies each resulted in reduced tumor cell proliferation in all these tumor lines. Vascular endothelial growth factor is also an autocrine growth factor for human gastric adenocarcinoma17 and malignant mesothelioma.18 Our finding of VEGFR expression by HNSCC tumor cells supports a similar autocrine regulatory role for VEGF in HNSCC.
Strong expression of VEGFR-3 by tumor cells and VECs was evident in our study. Additionally, VEGFR-3 has been described on lymphatic endothelium in areas adjacent to lymphoma and breast carcinoma,12 suggesting a paracrine signaling network between tumor cells and the lymphatic endothelium. Vascular endothelial growth factor C, which binds VEGFR-3, strongly induces the proliferation of lymphatic endothelial cells and development of new lymphatic sinuses.19 We found that VEGFR-3 is strongly expressed on VECs in HNSCC. This finding is consistent with a recent study by Valtola et al,20 who demonstrated a significant up-regulation of VEGFR-3 expression by VECs in invasive breast cancer compared with normal breast tissue. Vascular endothelial growth factor C was located in the cytoplasm of the tumor cells. Since both ligands for VEGFR-3, VEGF-C and VEGF-D, are capable of stimulating endothelial cell proliferation, it is likely that VEGFR-3 has a role in angiogenesis in HNSCC. In addition, our finding of strong expression of VEGFR-3 on tumor cells indicates a possible role for this receptor in autocrine regulation of cancer growth.
An interesting additional finding in our study was that of significant expression of all 3 VEGFRs on tumor-associated macrophages. Macrophages are known to produce VEGF; additionally, increased macrophage infiltration is associated with increased VEGF expression, increased angiogenesis, and worsened prognosis in breast cancer.21 Vascular endothelial growth factor receptor 1 mediates human peripheral blood monocyte chemotaxis in vitro when treated with VEGF derived from tumor cell supernatants.22 Thus, the presence of VEGFRs on macrophages supports the concept that VEGF produced by tumor cells may play a role in attraction of macrophages into the HNSCC environment. The presence of VEGFR-1 and VEGFR-2 on stromal fibroblasts associated with tumor has recently been described in non–small cell lung carcinomas.23 Interestingly, we found expression of VEGFR-1 and VEGFR-3, but not VEGFR-2, on stromal fibroblasts in HNSCC. Vascular endothelial growth factor D, which binds VEGFR-3, stimulates mitogenic activity and morphologic alterations in fibroblasts.24 Thus, our finding that VEGFR-3 is present on fibroblasts in HNSCC suggests that the VEGF family also affects fibroblast function in the tumor environment. Vascular endothelial growth factor acting on fibroblasts may be derived from fibroblasts themselves in addition to tumor cells.25 Further investigation is needed regarding the effects of cytokine activation on varying cell populations in the tumor microenvironment. However, it is likely that recruitment and activation of macrophages and stromal fibroblasts play a role in the complex regulation of tumor growth and metastasis via the VEGF/VEGFR system.
In conclusion, interactions of the various VEGF/VEGFR systems in the tumor microenvironment are clearly complex. Our finding that the VEGFRs are expressed by HNSCC tumor cells may indicate that VEGF is an autocrine regulator of tumor cell activity in addition to its known angiogenic effects on VECs. Tumor-associated macrophages and stromal fibroblasts also express these receptors to various degrees and contribute to the complexity of the VEGF/VEGFR system in human cancers. Anti-VEGF therapies designed to block angiogenesis may also have a direct effect on tumor cell activity. Further understanding of the VEGF/VEGFR system will be useful in designing future therapeutic agents.
Corresponding author and reprints: Donald L. Kreutzer, PhD, Department of Pathology MC 3105, University of Connecticut Health Center, 263 Farmington Ave, Farmington, CT 06030-3105 (e-mail: firstname.lastname@example.org).
Submitted for publication June 21, 2001; final revision received August 23, 2002; accepted December 13, 2002.
This research was supported in part by funds from the Division of Otolaryngology of the Department of Surgery and the Department of Oral Diagnosis, University of Connecticut Health Center, Schools of Medicine and Dental Medicine, Farmington.
This study was presented at the Fifth International Conference on Head and Neck Cancer; July 29-August 2, 2000; San Francisco, Calif.