The goal of the major project proposed is to determine a consistent set of 3-dimensional (3-D) structural and assembly data at the molecular level of intermediate filaments (IF) which, ultimately, may lead to a unified model of IF structure and assembly. To achieve this, we will continue using high resolution electron microscopy (EM) and image processing, and combine these with proteinchemical, biophysical and immunological analyses of human epidermal keratin filaments. We propose to proceed as follows: (1) Individual or specific sets of keratins will be isolated by HPLC and other relatively mild procedures and used to reconstitute filaments from defined subunit mixtures. (2) These filaments, intact or partially assembled, will be analyzed structurally (a) by conventional transmission EM under a variety of specimen preparation conditions, (b) by scanning transmission EM, unstained, to measure their masses and native dimensions, and (c) by analytical ultracentrifugation. (3) To probe into their molecular topography, we will employ (a) chemical crosslinking and (b) immuno-EM with subunit-specific antibody including some raised against synthetic peptide immunogens. (4) The formation of paracrystalline arrays of keratin IF constituents will be fostered to allow high resolution 3-D analysis of IF structure to be performed. (5) The relationship between the hierarchic order of IF structure and assembly will be explored by dissecting the in vitro polymerization reaction into the underlying nucleation and condensation steps via a combination of EM and various polymerization assays. (6) We will pursue the question of preferential 'segregation' into different versus random 'integration' into all IFs of specific sets of keratins during filament reconstitution under a number of conditions. Other supramolecular assemblies to be studied will include the nuclear pore complex (NPC) and the nuclear lamina (NL), two major constituents of the nuclear envelope. The subunit topography of the NPC will be explored by immuno-EM using subunit-specific antibody. Attempts will be made to reconstitute the NL meshwork from isolated components and to determine its 3-D molecular organization. Last, but not least, we will continue refining methodologies which should enable us to preserve protein structure in a more native state for subsequent 3-D structural analysis by high resolution EM and digital image processing.
Showing the most recent 10 out of 15 publications
