Figure 1. Subdomain organization of intermediate filament (IF) proteins. The central rod domain consists of -helical segments 1A, 1B, 2A and 2B connected by short nonhelical linker sequences (L). The head and tail domains are divided into homologous (H), variable (V) and end (E) subdomains. Dashed line denotes 42 amino acid stretch of rod 1B specific for type V proteins. The nuclear localization sequence and sites of posttranslational modifications and of frequent mutations in epidermolysis bullosa simplex (EBS) are indicated by arrows. Potential functions of subdomains are shown at the bottom. cdc2, PKA and PKC, protein kinases.

Figure 2. Model of the alignment of intermediate filament (IF) subunits during IF assembly. Homodimers or heterodimers form by arranging their rod domains in a parallel coiled-coil organization, tetramers by antiparallel arrangement of dimers unstaggered or nearly half staggered, leading to overlap of region 2B, or region 1B. Only type V IFs form head-to-tail associated dimers. If these interactions are present in 10-nm filaments consisting of eight tetramers per cross-section, short overlaps of the rod sections 1A and 2B (dark bars) are predicted. The arrangement of nonhelical domains within the 10-nm filament are not known.