Differential induction of ATF3 and HO-1 in myeloid cells and keratinocytes via Dimethylfumarate or Cyclosporine A.

BACKGROUND: Chronic inflammatory skin diseases are characterized by controlled proliferation of keratinocytes. Here, activating transcription factor 3 (ATF3) might play a fundamental role. In these inflammatory diseases, proliferation is controlled and only rarely leads to cancer development which can be supported by an inflammatory microenvironment. ATF3 is a dual function protein as it suppresses pro-inflammatory IL-6 and IL-8, but also acts as a pro-oncogenic factor by the suppression of p53. We therefore analyzed ATF3 expression comparing myeloid cells with keratinocytes.

OBJECTIVE: To dissect the bi-modal role of ATF3 we pharmacologically induced ATF3 and analyzed its influence on cytokine expression and secretion in a cell type specific manner.

METHODS: Since inflammatory skin diseases can be treated systemically with Cyclosporin A or Dimethylfumarate we stimulated myeloid cells and primary human keratinocytes with these drugs and analyzed gene expression by quantitative real-time PCR. Cytokine secretion was measured by ELISA.

RESULTS: In the present study, we could show that ATF3 is induced in PBMCs by DMF and weakly by Ebselen, while CsA is the most prominent inducer of ATF3 in keratinocytes without enhancing HO-1 transcription. Further we could show that induction of stress by LPS treatment elevates IL-1β and IL-6 and weakly ATF3 transcription in PBMCs. While transcription of both cytokines is elevated, LPS treatment mediates IL-6 secretion with only little IL-1β secretion. Treatment with DMF dampens LPS-induced transcription.

CONCLUSIONS: Taken together, our results shed light into the different carcinogenic potential of CsA and DMF, which both target ATF3. Collectively our data demonstrate that CsA strongly induces pro-carcinogenic ATF3 in keratinocytes, whereas ATF3 induction by DMF in myeloid cells acts anti-inflammatory.