Mouse vimentin: structural relationship to fos, jun, CREB and tpr.

We have isolated and characterized mouse cDNA clones representing the entire coding region of vimentin. RNA blot analysis of different stages during development has revealed differential control in the expression of vimentin mRNA in the different tissues studied. The nucleotide sequence extends 1800 base pairs and contains the 466 amino acid mouse vimentin polypeptide chain, flanked by 90 base pairs 5' and 312 base pairs 3' untranslated region. Conformational analysis of the deduced amino acid sequence was used to localize the three known structural domains: a non-alpha-helical N-terminal head of 81 residues, a rod-like domain of 330 residues arising from three alpha-helices, and a non-alpha-helical C-terminal domain of 55 residues. Amino acid sequence comparisons with other species revealed high sequence conservation of mouse vimentin to hamster (98.7%), human (96%), and chicken (88%) protein. Computer sequence analysis also revealed domains of significant homology between different alpha helical regions of vimentin and the DNA binding-leucine zipper domain of several proto-oncogenes and transcription regulators. Specifically, 50-70% structural similarity was observed between the basic domain of the DNA binding region of the nuclear proto-oncogene products c-fos and its related antigen fra-1, c-jun and the cAMP-responsive DNA binding protein CREB, with part of the N-terminal half region of helix 1b of vimentin. When the leucine zipper domains of all these proteins were compared to vimentin, at least two different regions of similarity in the vimentin molecule were found reaching up to 53% for jun, 60% for fos, and 76% for CREB. Further analysis revealed several domains of significant similarity (50%) between all alpha-helices of the rod domain of vimentin and the N-terminal (approximately 210 residues) activation domain tpr of the oncogenic raf.