Vascular Endothelial Growth Factor Expression in Nasal Polyps of Aspirin-Intolerant Patients FREE

Chronic rhinosinusitis (CRS) is an inflammatory disease of the upper respiratory tract and one of the most common chronic diseases worldwide with high socioeconomic costs.¹ Nasal polyp formation in CRS is characteristic and frequently associated with an intolerance reaction after aspirin intake, such as rhinorrhea, bronchospasm, and shock symptoms related to a non–IgE-mediated pharmacological hypersensitivity reaction and bronchial asthma, termed the Samter triad. The development of the complete triad can take years.² Nasal polyps related to the Samter triad are additionally characterized by high recurrence rates after sinus surgery. The etiology of nasal polyps is still under controversial debate. Staphylococcal superantigens, airborne fungi, and the dysregulation of arachidonic acid metabolism as the crucial part of polyp formation within the Samter triad are under discussion.^3- 4

Vascular endothelial growth factor (VEGF) is known to play an important role during the angiogenesis and modulation of capillary permeability.⁵ Nasal polyps are characterized by remarkable edema that could be explained by an upregulated VEGF release. Several studies^6- 11 have reported high levels of VEGF in the epithelium and endothelium of nasal polyps. However, other basic mechanisms of tissue homeostasis, such as proliferation and apoptosis, might be involved in nasal polyp formation and potentially regulated by VEGF, as demonstrated by Coste et al¹² and Lee et al.¹³ These authors found evidence of a VEGF-dependent inhibition of apoptosis in nasal mucosa and observed an increased cell growth rate in nasal polyps compared with control mucosa.

The expression of VEGF receptors (VEGFRs) is an essential precondition to allow any effect of VEGF on its target cells. Several types of VEGFRs, including Flk for VEGFR2, Flt for VEGFR1, and neuropilin for NRP1, have been identified. Receptors 1 and 2 for VEGF are the major receptors important for proper embryonic vasculature. Knockouts for VEGFR2 are nonviable, with solely minimal developed vasculature, whereas knockouts for VEGFR1 show a disorganized vasculature. Neuropilin 1 acts as a VEGF coreceptor to enhance VEGFR2 binding and bioactivity.^14- 17

Until now, differences between VEGF and VEGFR expression in nasal polyps from patients with aspirin intolerance compared with those without remained speculative. Identification of specificities at the molecular level could lead to a better understanding of the different phenotypic growth patterns with a potential therapeutic impact. Therefore, in the present immunohistochemical study, we analyzed the expression levels of VEGF and VEGFRs in healthy nasal tissue samples and in nasal polyps from aspirin-intolerant and aspirin-tolerant patients. To gain insight into potentially regulated downstream mechanisms, we also analyzed markers of proliferation (Ki67) and apoptotic activity (caspase 3). Finally, we outlined and discussed differences at the molecular level of nasal polyps from aspirin-tolerant and aspirin-intolerant patients.

We collected a total of 67 specimens. Forty-three specimens of nasal polyps, including 22 from clinically determined aspirin-intolerant patients and 21 from aspirin-tolerant patients, were obtained during sinus surgery (30 male and 13 female patients; mean age, 48.8 years; range, 16-59 years). Twenty-four nasal mucosa samples serving as control specimens were taken from the inferior turbinates of patients undergoing surgery for nasal obstruction without any endoscopic and radiologic signs or symptoms of CRS (13 male and 11 female patients; mean age, 37.1 years; range, 16-59 years).

Chronic rhinosinusitis with nasal polyps was diagnosed by means of nasal endoscopy and computed tomography according to the guidelines of the 2007 European position paper on rhinosinusitis and nasal polyps.¹⁸ Aspirin intolerance was diagnosed using the patients' history and symptoms.

Patients with recurrent nasal polyps after sinus surgery combined with bronchial asthma and an intolerant reaction such as bronchospasm after aspirin intake were identified as aspirin intolerant. Patients who did not have signs of aspirin intolerance underwent primary surgery. We analyzed the expression of VEGF, VEGFRs (Flk for VEGFR2, Flt for VEGFR1, and neuropilin for NRP1), Ki67 (proliferation), caspase 3 (apoptosis), and the spatial distribution (endothelium vs epithelium) of these proteins by means of immunohistochemistry.

This study was approved by the institutional review board and performed in accordance with the actual version of the Declaration of Helsinki.

Biopsy specimens were snap frozen in liquid nitrogen immediately after resection and stored at −80°C until needed for procedures. Specimens were embedded in optimal cutting temperature medium (Cryomatrix; Shandon). We prepared 6-μm sections using a cryostat (Leica) and mounted them on glass slides (Superfrost Plus; Thermo Scientific). Sections and cells were fixed using ice-cold acetone (Sigma-Aldrich Corp), and endogenous peroxidases were blocked by 2.5% methanol/hydrogen peroxide (Roth). An avidin/biotin blocking kit (Zymed Laboratories) was used to reduce background staining. After 30 minutes of preincubation with 10% normal goat or normal rabbit serum (Dako), the primary antibodies consisting of polyclonal rabbit anti-VEGF (1:150 dilution; Santa Cruz Biotechnology), polyclonal rabbit anti-Flt (1:100 dilution; Santa Cruz Biotechnology), monoclonal mouse anti-Flk (1:150 dilution; Santa Cruz Biotechnology), polyclonal goat anti-NRP1 (1:150 dilution; Santa Cruz Biotechnology), monoclonal mouse anti-Ki67 (1:100 dilution; Dako), and polyclonal rabbit anti–caspase 3 (1:100 dilution; Cell Signaling Technology) were overlaid without washing and incubated overnight at 4°C. Slides were consecutively washed twice with Tris-buffered saline (TBS)/polysorbate (Tween) wash buffer (400mM sodium chloride); incubated with the secondary antibodies (Dako) consisting of biotinylated goat antirabbit (1:250 dilution), goat antimouse (1:250 dilution), or rabbit antigoat (1:600 dilution) for 30 minutes; washed 2 times for 5 minutes with TBS/Tween wash buffer (147mM sodium chloride); incubated with horseradish peroxidase–labeled streptavidin (1:200 dilution; Dako) for 30 minutes; and washed 2 times for 5 minutes with TBS/Tween wash buffer. Finally, the slides were incubated with diaminobenzidine tetrahydrochloride (Dako) for 1 minute and washed with TBS/Tween wash buffer. The slides were counterstained with hematoxylin (1:5 dilution; Merck KGaA) for 5 minutes and washed with tap water for 10 minutes. All slides were then treated with an increasing isopropanol series (70%, 80%, 90%, and 100%; Hedinger GmbH & Co), treated 2 times for 5 minutes with xylol (Merck), and mounted with a coverslip. Sections incubated with TBS/Tween wash buffer instead of the primary antibodies served as negative controls, and normal kidney and placenta tissue served as positive controls (data not shown).

We applied a computer-based analysis method to quantify the immunohistochemical staining results as described before.^6,19 In brief, 3 representative areas (original magnification ×400, corresponding to 25 × 25 μm) were documented in jpg format using an inverted microscope (Axiovert 200; Zeiss). For quantification, the jpg documents were analyzed using commercially available software (Photoshop; Adobe Systems Inc). All images of specimens from a given experimental group (control, aspirin tolerant, and aspirin intolerant) were imported in the same master file, and the staining results were analyzed in parallel. This method allows measuring of the stained area and the labeling intensity and hence the quantification and relative comparison of expression levels. Results were expressed as arbitrary units.

To calculate the proliferative and apoptotic activities of a given sample, we counted Ki67- and caspase 3–positive cells in 3 randomly selected grids (25 × 25 μm) at original magnification ×400 and then expressed the results as a percentage of all cells within the 3 grids (approximately 300 cells/grid).

Differences among the groups (control and nasal polyps with and without aspirin intolerance) concerning continuous variables (VEGF, Flk, Flt, neuropilin, and Ki67) were analyzed by means of the nonparametric Kruskal-Wallis test. In the case of P < .05, multiple comparisons were performed using pairwise comparisons (Mann-Whitney test).

All P values are 2 sided. Analyses are not adjusted for multiple testing, and P values serve for description only. P < .05 demonstrates a difference between the observed groups.

We observed a strongly increased level of VEGF expression in epithelial cells of nasal polyps from aspirin-tolerant patients compared with the controls (P = .01). When analyzing tissue samples from aspirin-intolerant patients, we found higher VEGF levels compared with those from aspirin-tolerant patients (and controls) (P < .001). In the endothelium, the level of VEGF expression in nasal polyps from aspirin-intolerant patients was remarkably higher compared with the controls (P = .004) and compared with nasal polyps from aspirin-tolerant patients (P = .049) (Figure 1 and Figure 2).

A precondition for a stimulatory activity of VEGF is the expression of VEGFR on the potential target cells. Therefore, we analyzed the expression of Flk, Flt, and NRP1. Expression of Flk was comparable in the epithelium and endothelium in all analyzed tissue samples (data not shown). Staining for Flt yielded lower levels in the epithelium of nasal polyps from aspirin-intolerant patients compared with the controls (P = .007) and lower levels in the endothelium of nasal polyps from aspirin-intolerant compared with those from aspirin-tolerant patients (P = .046), but the difference did not reach statistical significance compared with the controls (P = .30) (Figure 3 and Figure 4). In the endothelium and epithelium of nasal polyps, NRP1 levels were increased compared with those in the controls. No differences were observed between nasal polyps of aspirin-intolerant and aspirin-tolerant patients (Figure 5 and Figure 6).

The percentage of Ki67-positive cells, indicating proliferating cells, was clearly higher in the epithelium of nasal polyps from aspirin-intolerant (P = .01) and aspirin-tolerant patients (P = .001) compared with the controls (Figure 7 and Figure 8) but did not differ between aspirin-intolerant and aspirin-tolerant patients. The staining of caspase 3 as an indicator of apoptotic cells demonstrated a generally low level and did not reveal any differences between the groups (data not shown).

Increasing evidence supports the concept of nasal polyp growth related to VEGF tissue levels.^6- 13 However, no differentiation concerning aspirin intolerance has yet been reported. In accordance with these reports, we found high levels of VEGF in epithelial and endothelial cells of nasal polyps compared with healthy mucosa. In nasal polyps from aspirin-intolerant patients, VEGF levels were also heightened compared with those of aspirin-tolerant patients. This key observation of our study demonstrates molecular differences between nasal polyps from aspirin-tolerant and aspirin-intolerant patients.

Furthermore, it is known that leukotrienes can transcriptionally activate VEGF production.²⁰ One might speculate that the imbalance of leukotrienes and prostaglandins with a predominance of leukotrienes in aspirin-intolerant patients might lead to the observed additionally increased VEGF levels. Vascular endothelial growth factor is a strong mitogen of endothelial cells and a prosurvival factor for VEGFR-expressing cells in general. The VEGF homodimers bind Flk and Flt, causing receptor dimerization and therefore signal transduction. Simultaneously, VEGF can bind neuropilin, thereby enhancing signal transduction through Flk and Flt. Vascular endothelial growth factor receptors transmit the VEGF growth signal from outside the cell via tyrosine kinases to the nucleus, resulting in the upregulated transcription of a multitude of growth-promoting genes and in the direct upregulation of mitosis.^14,17

High levels of VEGF in polyps might activate VEGFRs and downstream pathways, resulting in increased proliferation, thereby promoting polyp formation. Therefore, we analyzed the expression of all 3 VEGFRs. We found the major VEGFR Flk expressed at comparable levels in the analyzed tissue samples, potentially indicating a basic susceptibility for VEGF stimulation. Levels of the coreceptor NRP1 were found to be increased in polyps, and levels of Flt were lowered. Neuropilin 1 is known to enhance bioactivity and binding affinity of VEGF.¹⁷ One might suggest that high VEGF levels in combination with high levels of expressed NRP1 receptors as found in nasal polyps, especially in those from aspirin-intolerant patients, are part of the pathophysiology of nasal polyp formation. These data are in accordance with the functional data reported by Lee et al,¹³ who demonstrated by means of in vitro inhibitor experiments that proliferation in polyps is largely regulated by neuropilin but less so by Flt. Consequently, we observed increased proliferation in nasal polyps. Our observation of a low level of Flt expression strengthens the assumption of a minor impact of Flt regarding cell proliferation in nasal polyps.¹³

Furthermore, our findings suggest an impact of VEGF on both epithelial and endothelial cells. One might assume that the impact of VEGF on epithelial cells mainly results in cell proliferation, indicated by high levels of Ki67. In endothelial cells, VEGF might lead to increased vessel permeability and finally to the well-known characteristic tissue edema of nasal polyps according to the initially described function of VEGF as a vascular permeability factor.⁵ Surprisingly, heightened VEGF levels in aspirin-intolerant patients were not associated with additionally increased proliferation compared with the aspirin-tolerant patients. The enhanced VEGF expression in aspirin-intolerant patients obviously has no additional effect on epithelial cell growth but might stimulate tissue edema by enhanced vessel permeability of the endothelium.

Nasal polyps are characterized by high proliferation, VEGF expression, and neuropilin expression. Furthermore, nasal polyps from aspirin-intolerant patients show remarkably increased levels of VEGF that might be related to increased vessel permeability and the observed tissue edema. Tissue samples of the inferior turbinates from patients without CRS are widely used as controls; however, biological specificities compared with nonpolypoid osteomeatal complex tissue cannot be definitely excluded.

In addition to sinus surgery, current options for treatment of CRS include the application of systemic and topical corticosteroids to reduce inflammation, aspirin desensitization in patients with the Samter triad, nasal saline washings, and others. As a clinical consequence of our results, we should establish novel therapeutic attempts that modulate cell proliferation and VEGF expression locally and analyze their benefit within clinical trials with emphasis on polyp growth and quality of life.

In summary, our data suggest that nasal polyp growth of aspirin-tolerant and aspirin-intolerant patients differs with respect to VEGF expression. The definite clinical relevance of this observation needs further investigation.