Neuropilin-1 expression in the tumour microenvironment of oral squamous cell carcinoma

Abstract

Background: Neuropilin-1 (Nrp-1) is a transmembrane glycoprotein involved in multiple biologic and pathologic processes including carcinogenesis. It is widely expressed in cancers where it has been correlated with poorer prognosis. Nevertheless its expression profiles and significance in oral squamous cell carcinoma (OSCC) remain elusive. Objective: To investigate the expression of Nrp-1 in OSCC and oral epithelial dysplasia (OED) using immunohistochemistry (IHC), and to examine its co-localisation with markers for angiogenesis and T regulatory cell (Tregs), namely vascular endothelial growth factors 2 (VEGFR2) and forkhead box P3 (FoxP3) respectively, using double-labelling immunofluorescence (DLIF). Method: Formalin-fixed paraffin-embedded (FFPE) blocks were obtained from the archives of the Oral Pathology Centre, University of Otago. These specimens were divided into three groups namely OSCC, OED and non-specifically inflamed mucosa (NIM). The samples from these groups comprised well-differentiated OSCC (n=21) and moderately-differentiated OSCC (n=2), low-risk OED (n=8) and high-risk OED (n= 9), and NIM (n= 14). Immunohistochemistry was performed with an antibody against Nrp-1. Hotspot locations in the epithelium and connective tissue (n=3 each) from each specimen were photographed and analysed semi-quantitatively. Kruskal-Wallis tests were performed with p<0.05 denoting significance. Intensity and proportion scores were used to calculate the immunoreactive score (IRS). DLIF results were qualitatively examined to determine the co-localisation of Nrp-1 with VEGFR2 and FoxP3. Results: The OED epithelium showed significantly higher expression of Nrp-1 than NIM and OSCC (p=0.04). No other significant quantitative differences were observed in the epithelium or connective tissue between the groups. Qualitatively Nrp-1 was more highly expressed on the stromal cells in OSCC and OED tissues than in NIM. Nrp-1+/FoxP3+ co-localisation showed a similar pattern of co-localisation in OED and OSCC where Nrp-1 was present on the surface of FoxP3+ cells. Nrp-1+/FoxP3+ co-localisation was also present in NIM samples but here it showed nuclear co-expression. There was no evidence of Nrp-1+/VEGFR2+ co-localisation in OSCC and NIM samples but it was detected in stromal cells of OED samples. Conclusion: The epithelial overexpression of Nrp-1 in OED suggests that it may be important in the initial dysplastic transformation, but not necessarily in progression once malignant transformation has occured. The high Nrp-1 expression on variable stromal cells in OED and OSCC demonstrates that Nrp-1 is likely to play a significant and diverse role in immune function and regulation via its wide-ranging interactions in the human immune system. The Nrp-1+/FoxP3+ co-localisation on stromal mononuclear cells in OED and OSCC suggests that Nrp-1 may modulate the microenvironment to make it conducive to disease progression, while the variability in the co-expression patterns may indicate the ability of Nrp-1 to differentiate Tregs subsets. VEGFR2 does not appear to be a meaningful co-receptor for Nrp-1 in OSCC and thus Nrp-1-associated angiogenesis may involve alternative mechanisms. Further research may show the potential of Nrp-1 as a novel prognostic and therapeutic marker in OSCC potentially through the mechanism of immune regulation. A more detailed study is required to further elucidate the role of Nrp-1 in cancer angiogenesis.