What can be learned from intermediate filament gene regulation in the mouse embryo.

In recent years, intermediate filaments (IFs) have attracted much interest, largely because their constitutive polypeptide units are specifically expressed in various cell types and thus represent excellent differentiation markers. Data obtained through biochemical studies and molecular cloning have allowed the classification of IFs into five types according to their protein structure. The expression of most IF types is characteristic of a given cell type: cytokeratins (IF types I and II) are produced in epithelia, neurofilaments and alpha-internexin (type IV) in neurons and nestin (type IV) in neuroblast and myoblast. On the other hand the four type III IFs are highly related proteins which are expressed in different cell types. Thus the study of type III IF gene regulation provides an excellent approach towards the analysis of cell-specific transcription. This review focuses on type III IF gene regulation during mouse embryogenesis and describes the latest data obtained through the combination of both in vitro (in cell lines) and in vivo (in transgenic mice) approaches. It appears that, while intragenic sequences play a major role in the regulation of the expression of the genes encoding other types of IFs, a major contribution to the transcriptional regulation of type III IF genes is brought by 5' upstream sequences. However, recent evidence obtained through the use of transgenic mice indicate that upstream sequences must cooperate with intragenic elements to establish the complex and dynamic expression pattern characteristic of type III IF genes. The very high similarity between the coding sequences of type III IF genes raises the question of the significance of the occurrence of four members of this class. We propose a model for the amplification of this small gene family based on the increasing complexity of expression patterns in higher organisms. This could have led first to the requirement for a highly sophisticated control region in an ancestral type III IF gene, followed by two successive gene duplications, thus leading to the appearance of four different regulatory regions directing the cell-specific transcription of nearly identical genes in different cell types.