MITF and PAX3 Play Distinct Roles in Melanoma Cell Migration; Outline of a "Genetic Switch" Theory Involving MITF and PAX3 in Proliferative and Invasive Phenotypes of Melanoma

Abstract

Melanoma is a very aggressive neoplasm with a propensity to undergo progression and invasion early in its evolution. The molecular pathways underpinning invasion in melanoma are now just beginning to be elucidated, but a clear understanding of the transition from non-invasive to invasive melanoma cells remains elusive. Microphthalmia-associated transcription factor (MITF), is thought to be a central player in melanoma biology, and it controls many aspects of the phenotypic expression of the melanocytic lineage. However, recently the paired box transcription factor PAX3 was shown to transcriptionally activate POU3F2/BRN2, leading to direct repression of MITF expression. Here we present a theory to explain melanoma phenotype switching and discuss the predictions that this theory makes. One prediction is that independent and opposing roles for MITF and PAX3 in melanoma would be expected, and we present empirical evidence supporting this: in melanoma tissues PAX3 expression occurs independently of MITF, and PAX3 does not play a key role in melanoma cell proliferation. Furthermore, we show that knockdown of PAX3 inhibits cell migration in a group of "lower MITF" melanoma cell lines, while knockdown of MITF promotes cell migration in a complementary "higher MITF" group of melanoma cell lines. Moreover, the morphological effects of knocking down PAX3 versus MITF in melanoma cells were found to differ. While these data support the notion of independent roles for MITF and PAX3, additional experiments are required to provide robust examination of the proposed genetic switch theory. Only upon clear delineation of the mechanisms associated with progression and invasion of melanoma cells will successful treatments for invasive melanoma be developed.

Keywords: melanoma; microphthalmia-associated transcription factor; migration and invasion; paired box transcription factors; pax3; phenotype switching.

Figure 1

A schematic of the “genetic switch” model. The MITF-miR-211 axis is counterbalanced by the PAX3-POU3F2 axis, with high levels of each pathway inhibiting the corresponding other pathway. “Wnt” and “PI3K” represent one of several possible signaling pathways promoting the activity of MITF-miR-211 axis versus the PAX3-POU3F2 axis, respectively. High levels of the PAX3-POU3F2 axis (represented by the thick end of the wedge) are associated with cell migration and invasion of melanoma cells, while high levels of the MITF-miR-211 axis are associated with proliferation and/or differentiation of melanoma cells.

Figure 2

PAX3 and MITF expression vary in their relative intensity in different regions of melanoma. (A) The figure shows photomicrographs of dual immunofluorescent staining of MITF expression (green label), PAX3 expression (red label), DAPI nuclear staining (blue label), and a merged image in normal skin, metastatic melanoma (Primary MM), and Lentigo maligna melanoma. The scale bar, which varies in length on the images, represents 50 μm. (B) Higher magnification photomicrographs taken from the merged image in (A) of the Primary and Lentigo maligna melanomas show in greater detail the difference in the relative immunofluorescence color intensity of MITF labeling versus PAX3 labeling in the tumor cells immediately below the skin surface (“Top dermis”) versus cells located deeper in the tumor (“Deep dermis”). Below the panels is shown a color intensity scale, with one end representing relatively strong MITF intensity, and the opposite end representing relatively strong PAX3 intensity. The scale bar in the bottom right image is for all the panels and represents 25 μm.

Figure 3

Dual immunofluorescent staining of Ki67 and PAX3 in benign and malignant melanocytic lesions. (A) Graph showing the percentage of PAX3 and Ki67 double positive (dp) cells as a percentage of the total number of PAX3-positive cells for superficial spreading melanoma (SSM), lentigo maligna melanoma and nodular melanoma (LMM & NM), and metastatic melanoma (MM), ***p < 0.001. (B) Photomicrographs showing the results of immunofluorescent staining for PAX3, Ki67, and merged images for an intradermal nevus (IN), superficial spreading melanoma (SSM), nodular melanoma (NM), and metastatic melanoma (MM). The scale bar, which varies in length on the images, represents 50 μm.

Figure 4

Knockdown of PAX3 or MITF results in differential effects on the migration of melanoma cell lines in vitro. siRNA-mediated knockdown of PAX3 inhibits the migration of melanoma cells expressing lower levels of MITF (green bars), but not melanoma cells expressing higher levels of MITF (red bars). Conversely, siRNA-mediated knockdown of MITF does not enhance the migration of melanoma cells expressing lower levels of MITF (green bars), but does enhance the migration of melanoma cells expressing high levels of MITF (red bars). **p < 0.001.

Figure 5

Knockdown of PAX3 results in a decrease of POU3F2 and an increase of MITF gene transcript levels in melanoma cell lines in vitro. Transcript levels of PAX3, POU3F2, and MITF mRNA were examined by RT-qPCR in NZM12, and in NZM11 (PAX3 and POU3F2 only) melanoma cell lines following RNAi treatment of the cell lines with either siRNAs against luciferase (siGL2) as a negative control, or against PAX3 (siPAX3). The results were calculated as the fold difference in transcript level relative to the level of the housekeeping gene GNB2L1, normalized to the siGL2 data.

Figure 6

Knockdown of PAX3 or MITF results in differential morphological effects in NZM15 melanoma cells. NZM15 melanoma cells were grown in media without any transfection (MC), or transfected with siRNA to luciferase (siGL2) as a non-targeting control, siRNA to MITF (siMITF), or siRNA to PAX3 (siPAX3) and then stained after 48 h with β-tubulin antibody.