The long noncoding RNA HOTAIR has tissue and cell type-dependent effects on HOX gene expression and phenotype of urothelial cancer cells.

Background: Urothelial carcinoma (UC) is the fifth most common cancer in the developed world. Delineation of differentiation subtypes in UC highlighted the importance of aberrant differentiation. Understanding underlying mechanisms may facilitate diagnosis and development of efficient therapy strategies. It is well accepted that epigenetic mechanisms are involved. Long noncoding RNAs (lncRNAs), a new class of epigenetic factors, are thought to mediate molecular differences between cell types to control cellular identity. The present study focuses on the lncRNA HOTAIR, originating from the HOXC locus. Its overexpression induces an aggressive phenotype in many cancers and aberrant expression of homeotic HOX transcription factors, especially HOXD10, that regulate differentiation and tissue homeostasis. The aim of the present study was to determine the functional role of HOTAIR in UC with regard to aggressive phenotype, regulation of aberrant differentiation and altered HOX gene expression.

Methods: We determined RNA expression levels of HOTAIR and HOX genes in UC tissues and cell lines. Knockdown of HOTAIR and ectopic overexpression was performed to determine the effect on reported target genes in UC. Cell lines were stably transfected with HOTAIR to investigate changes in phenotype and HOX gene expression.

Results: HOTAIR was overexpressed in approximately half of UC tissues and cell lines. Effects of HOTAIR overexpression differed between cell lines. Whereas VM-CUB1 cells acquired the expected phenotype with increased proliferation, clonogenicity, anchorage independent growth, migratory activity and epithelial-to-mesenchymal transition, 5637 cells grew more slowly displaying induction of senescence and related immune response genes. Other UC lines showed intermediate effects. Expression profiling revealed divergent effects on HOX genes, cell cycle regulators and differentiation according with the phenotypic differences between HOTAIR-overexpressing VM-CUB1 and 5637 cells.

Conclusions: Our data indicate that HOTAIR overexpression may affect differentiation state and aggressiveness of UC cells, but in a cell-type dependent manner. Our functional studies and the comparison of our expression data sets with those from other cancer cell types, which revealed minimal overlaps, indicate that effects of HOTAIR are strongly tissue-dependent and can even differ within one cancer type. Thus, HOTAIR functions and target genes cannot simply be transferred from one cancer type to the other.