One of the possibilities is that the cells actually melanize (e.g. due to interactions with some components of the medium, e.g. Eagle's MEM, DMEM), and the terminally differentiated cells get lost over serial passaging. The phenomenon is pretty well described in early papers by prof. Slominski, in particular, e.g. Anticancer Res 1985; 5:403, and a latter  review: Pigment Cell Melanoma Res 2011; 25:14; DOI: 10.1111/j.1755-148X.2011.00898.x

I particularly recommend the early papers from Biosci Rep cited in these reviews.(if you wish, I can send privately)

Best regards!

Dear Olivier and Przemyslaw,

Thanks a lot for your informative insights.

As you guys mentioned, many genes involved in melanogenesis might have altered in melanoma cells. And this is exactly the point where I think should be studied with more care! This is because, many of the genes involved in melanogenesis are also involved in melanocyte's differentiation and development such as Kit and MITF (as mentiioned). And this could make many other changes to the cells in culture which would make these cells very different to the actual tumor cells in vivo! Then the question is that how reliable these cells are in studying Melanoma, specially when people look at the differentially gene expressions in NGS and omics analysis in general.

Also, in agreement with Przemyslaw, looking at the early papers would be of great help. I would appreciate if you can send me the mentioned papers @Przemyslaw.

Thanks