RESUMO
Metastatic melanoma is a very aggressive form of cancer in both humans and dogs. Dogs primarily develop oral melanoma of mucosal origin. Although oral melanoma in humans is rare, both diseases are highly aggressive with frequent metastases. This disease represents a "One Health" opportunity to improve molecular and mechanistic understanding of melanoma progression. Accumulating evidence suggests that cyclooxygenase-2 (COX-2) may play a critical role in the malignant behaviour of melanoma. In this study we analysed 85 histologically confirmed melanomas from canine patients and showed that COX-2 is overexpressed in both oral and cutaneous melanomas and that COX-2 expression correlates with established markers of poor prognosis. To determine the role of COX-2 in melanoma we developed two melanoma cell lines with stable integration of an inducible doxycycline-regulated expression vector containing a COX-2 targeted micro-RNA (miRNA). Using this system, we showed that cellular proliferation, migration and invasion are COX-2 dependent, establishing a direct relationship between COX-2 expression and malignant behaviour in canine melanoma. We have also developed a powerful molecular tool to aid further dissection of the mechanisms by which COX-2 regulates melanoma progression.
RESUMO
Melanoma is a fast-growing tumour in dogs and represents 7% of the total malignant neoplasms from the skin and is the most common tumour found in the oral cavity. In these tumours, high expression of cyclooxygenase-2 (COX-2) is associated with a poor prognosis. The aim of this study was to verify if the overexpression of COX-2 is related to the modulation of lymphocytes and if it is associated with the angiogenic and proliferative capacity of the melanoma. Canine melanoma samples (n = 85) were analysed by immunohistochemistry to detect the expression of S-100, Melan-A, PNL-2, COX-2, Factor VIII, Ki-67 and immune cells markers (CD3, CD4, FOXP3 and MAC387); and expression levels of MAC387, NOS and CD206 were determined by immunofluorescence. Our study showed a concurrent difference between the expression of COX-2 and inflammatory cell infiltration: Oral melanomas showed positivity for COX-2 in 34% of the cases and this expression was associated with CD3 positivity in the inflammatory infiltrate and angiogenesis; whereas cutaneous melanomas presented positivity for COX-2 in 42% of the cases and this expression was associated with positive staining for CD3, CD4, FOXP3 and MAC387. These markers are associated with inflammatory cells, angiogenesis and proliferation. Interestingly, melanomas were highly infiltrated by FOXP3+ cells, this is related to angiogenesis, whereas CD3, CD4 and MAC387 expression was only associated with cutaneous melanomas. The macrophage profile analysis showed that both oral and cutaneous melanomas with low COX-2 expression have an M1 phenoptype, whereas the cases with high COX-2 expression demonstrate a hybrid M1/M2 profile pattern. We concluded that the COX-2 is overexpressed in 42% of cutaneous melanomas and in 34% of oral melanomas, with a direct association with angiogenesis, proliferation, and intratumoral lymphocyte infiltration. We propose that COX-2 is a key regulator of immune cell infiltration and may drive tumour associated macrophage activation.