Abstract
Although not complete, the available sequence data on smooth muscle desmin, a prototype of 10 nm filaments present in living vertebrate cells, and two wool alpha-keratin components indicate a common structural motif . A similarly sized rod-like middle domain based mainly on alpha-helices probably able to form coiled-coils is flanked by differently sized terminal domains of non-alpha-helical nature. Within the middle domain there seem to be at least two regions where wool keratins and 10 nm filament proteins show a noticeable degree of sequence homology. In general, however, the proteins have diverged to an astonishing degree. Although the analysis seems to support, in general terms, a separation of the rod into two nearly equally long coiled-coils it raises doubts about additional aspects of current models of 10 nm filament organization. We propose that the terminal domains are directly involved in filament assembly making this process permanent in wool alpha-keratins because of the many disulfide bonds present in these regions. The 10 nm filaments of most living cells seem to avoid this frozen state and lack a similar wealth of cysteine residues.