SCAND1 Reverses Epithelial-to-Mesenchymal Transition (EMT) and Suppresses Prostate Cancer Growth and Migration

Abstract

Epithelial–mesenchymal transition (EMT) is a reversible cellular program that transiently places epithelial (E) cells into pseudo-mesenchymal (M) cell states. The malignant progression and resistance of many carcinomas depend on EMT activation, partial EMT, or hybrid E/M status in neoplastic cells. EMT is activated by tumor microenvironmental TGFβ signal and EMT-inducing transcription factors, such as ZEB1/2, in tumor cells. However, reverse EMT factors are less studied. We demonstrate that prostate epithelial transcription factor SCAND1 can reverse the cancer cell mesenchymal and hybrid E/M phenotypes to a more epithelial, less invasive status and inhibit their proliferation and migration in DU-145 prostate cancer cells. SCAND1 is a SCAN domain-containing protein and hetero-oligomerizes with SCAN-zinc finger transcription factors, such as MZF1, for accessing DNA and the transcriptional co-repression of target genes. We found that SCAND1 expression correlated with maintaining epithelial features, whereas the loss of SCAND1 was associated with mesenchymal phenotypes of tumor cells. SCAND1 and MZF1 were mutually inducible and coordinately included in chromatin with hetero-chromatin protein HP1γ. The overexpression of SCAND1 reversed hybrid E/M status into an epithelial phenotype with E-cadherin and β-catenin relocation. Consistently, the co-expression analysis in TCGA PanCancer Atlas revealed that SCAND1 and MZF1 expression was negatively correlated with EMT driver genes, including CTNNB1, ZEB1, ZEB2 and TGFBRs, in prostate adenocarcinoma specimens. In addition, SCAND1 overexpression suppressed tumor cell proliferation by reducing the MAP3K-MEK-ERK signaling pathway. Of note, in a mouse tumor xenograft model, SCAND1 overexpression significantly reduced Ki-67(+) and Vimentin(+) tumor cells and inhibited migration and lymph node metastasis of prostate cancer. Kaplan–Meier analysis showed high expression of SCAND1 and MZF1 to correlate with better prognoses in pancreatic cancer and head and neck cancers, although with poorer prognosis in kidney cancer. Overall, these data suggest that SCAND1 induces expression and coordinated heterochromatin-binding of MZF1 to reverse the hybrid E/M status into an epithelial phenotype and, inhibits tumor cell proliferation, migration, and metastasis, potentially by repressing the gene expression of EMT drivers and the MAP3K-MEK-ERK signaling pathway.