P21 Tyrosine and serine STAT3 phosphorylation drives cutaneous squamous cell carcinoma in recessive dystrophic epidermolysis bullosa as demonstrated using a 3D in vitro model

Abstract

Abstract The early onset of aggressive cutaneous squamous cell carcinoma (cSCC) and its rapid progression in recessive dystrophic epidermolysis bullosa (RDEB) leads to nearly 100% mortality rates in these patients. Elevated levels of phospho-STAT3 on tyrosine residue (STAT3Tyr705) have been intensively demonstrated in RDEB-derived cSCC cells, demonstrating that the constitutive activation and dysregulation of the Janus kinase (JAK)/STAT pathway plays a role in RDEB cSCC pathogenesis. We efficiently applied ruxolitinib (JAK1/2 inhibitor that blocks phosphorylation on STAT3Tyr705) to a RDEB–cSCC xenograph mouse model, reducing the size of small tumours but found that it was insufficient to reduce bigger tumours completely. Recently, phosphorylation at serine-727 (STAT3Ser727) has been shown to be relevant in the evolution of some cancers, but it has not yet been studied in RDEB cSCC. We hypothesized that phosphorylation at serine-727 of the STAT3 pathway might play a role in RDEB cSCC tumour development and progression, and could not be inhibited by ruxolitinib. To validate our hypothesis, we studied two RDEB cSCC human patient samples with different degrees of dedifferentiation (normal RDEB skin, edge of the tumour and centre of the RDEB cSCC). We demonstrated that STAT3Tyr705 was preferentially upregulated in the normal RDEB skin followed by the tumour edge skin, presenting less phosphorylation on the tumour. When we analysed STAT3 phosphorylation on STAT3Ser727, we could see that it presented the opposite phosphorylation pattern, being more phosphorylated in the tumour and less in normal RDEB skin. Furthermore, we validated the STAT3Ser727residue pattern across other cSCC cell lines and found that serine phosphorylation is present in other SCCs, demonstrating its crucial role in tumour development. To further investigate and analyse cancer progression in more detail, an easy and robust RDEB cSCC model is instrumental. Therefore, we generated RDEB cSCC spheroids in vitro, which has four key advantages over the previous models developed: (i) it is easy to generate; (ii) small amounts of cells are required; (iii) it is suitable for drug testing; and (iv) it can be used to recapitulate different stages of tumour development. We have been able to control their size and use them for drug testing, observing a 20% size decrease after 6 days of ruxolitinib treatment. Taken together, this three-dimensional RDEB cSCC in vitro model, along with analysis of human samples, will allow us to better understand RDEB cSCC progression at different stages, allowing us to investigate new treatments for RDEB cSCC.