We present the structure and nucleotide sequence of a gene encoding the human epidermal keratin. Three genomic clones were isolated from a X Charon 4A human genomic library by hybridization to a specific cDNA probe. vealed that a mammalian species expresses a total of about 20 different keratin subunits, which can be divided into distinct acidic (pI 4.5-5.5) and neutral-basic (pI 6.5-7.5) groups (1). This large number of subunits appears to originate at the gene level, since translation of poly(A)-enriched RNA extracted from these tissues reveals that a discrete mRNA species exists for each subunit (2-5). These protein chemical and mRNA translation experiments, as well as analyses involving monoclonal antibodies (6, 7), show that any one epithelial-cell type usually expresses only a few of these subunits, and these always occur as coordinated doublets consisting of at least one acidic and one neutral-basic keratin. For example, the hyperproliferating basal cells of newborn mouse or human epidermis express four keratin subunits: two acidic subunits (50 and 55 kDa in mouse, 50 and 52 kDa in human) and two neutral-basic subunits (59 and 60 kDa in mouse, 56 and 58 kDa in human). Suprabasal epidermal cells, committed to terminal differentiation, express one acidic subunit (59 kDa in mouse or 56.5 kDa in human) and one basic subunit of 67 kDa. This co-expression phenomenon appears to pertain throughout vertebrate epithelia (8). These findings all imply that two keratin subunits, one of each charge group, are required for keratin IF assembly in vivo, as was found in earlier in vitro assembly experiments (9). Partial or complete amino acid sequences are now available for several of the epidermal keratins (10)(11)(12)(13)(14)38), and the data show that all of them are constructed on a common plan: a central a-helical domain of conserved length and secondary structure, flanked by end-domains of highly variable size (<50 to >150 residues) and amino acid sequence. Comparisons of the sequences of the a-helical domains of different keratin subunits show that they can be divided into two distinct homologous types: the acidic keratins contain type I a-helical sequences, and neutral-basic keratins contain type II a-helical sequences (refs. 11, 13, and 14, and unpublished observations). The reason for the expression of many different subunits and the structural and functional significance of the coincident expression of two types of keratin subunits in one tissue is not known.In an interesting counterpoint to the complexity of the keratins of epithelial tissues, the IF proteins vimentin, desmin, and glial fibrillary acidic protein, normally expressed in mesenchymal, muscle, and astroglial tissues, respectively, each consist of a single protein; that is, they form homopolymer IF (9). Their amino acid sequences are known (15-17) and their secondary structures conform to that found for keratin IF subunits, but their a-helical coiledcoil sequences form a distinctly different type III group (18).In terms of the gene structure...