054 Evidence for Protofibrils in Trichocyte Keratin Intermediate Filaments.

PM Steinert, LN Marekov, DAD Parry. Laboratory of Skin Biology, NIAMS, NIH, Bethesda, MD; and Institute of Fundamental Sciences, Massey University, Palmerson North, New Zealand.

Since their first visualization by electron microscopic and X-ray diffraction analyses, there has been much speculation about the organization of subfilamentous units in trichocyte keratin intermediate filaments (TKIF). One popular model has proposed that axially aligned rows of pairs of coiled-coil molecules are bundled into protofilaments, two of which form a protofibril, of which several in turn (four or more) assemble to form a KIF. However, to date only indirect evidence, such as electron microscopic images of unraveling fibers or the existence of mass quanta, has been adduced in support of the existence of protofibrils. We have re-examined this issue in TKIF. First, we have devised methods for the efficient in vitro assembly of mouse Type Ia and IIa chains into TKIF in high yield. Second, crosslinks were induced in reduced and oxidized forms with disulfosuccinimidyl tartrate using improved conditions. Analyses of crosslinks documented that molecules in reduced TKIF are aligned as in cytokeratin KIF, allowing a 1 nm overlap between molecules in the same axial row. However, in oxidized TKIF, molecules in the A11 alignment mode are shifted forward by 2 nm, allowing a 1 nm gap instead. This presumably confers substantial additional stability to the TKIF due to maximal formation of intermolecular disulfide bonds. In addition, several crosslinks were shown to correspond to links between molecules four rows apart on twodimensional surface lattice models. Such data affirm and provide robust support for the concept of an eightchain (four-molecule) protofibril substructure in TKIF. Interestingly, in the reduced TKIF, crosslinks were found between any adjacent four rows of molecules in groupings of six or more. Thus there may be redundancy or promiscuity in protofibril organization since any adjacent grouping of four rows of molecules could bundle together. These data might explain the difficulty of prior reproducible visualization of protofibrils in microscopic images.