Deep-etch views of clathrin assemblies.

Clathrin assemblies were adsorbed to mica and freeze-dried by a new procedure that yields 3-D images with much topological detail. These permitted renewed inquiry into how clathrin trimers (i.e. "triskelions") assemble into polygonal coats or baskets. Freeze-drying revealed unsuspected differences in the relative shapes and dimensions of individual trimer building blocks, as compared with the completed polygonal networks, which indicate that the assembly scheme first proposed by Crowther and Peare (1) requires modification. Specifically, the freeze-etch images display the following new features: (1) Trimer arms possess terminal scroll-shaped hooks that can open or close and thus determine their lengths. (2) When extended, trimer arms are sufficiently long to pass around three sides of the final polygonal facet. Since current views hold that the arms pass around only two sides, the remaining length, including the terminal hook, must point into the basket interior. (3) Freeze-dried trimers display bends in their arms at specific loci that determine their final distribution in the completed baskets. (4) The completed struts of the final assemblies are uniformed in the calibre, cylindrical in profile, and travel directly between the vertices of each polygon, without any sign of the slew or width-variation that is predicted by the Crowther and Pearse model. Based on this direct comparison of promoter vs product, by a single technique that can image both, we offer a modified scheme for clathrin coat assembly, in which we predict that the individual arms in each clathrin triskelion emanate from its center in a slewed manner, but the final assembled struts of the basket need not be slewed. Attempts were made to capture assembly intermediates on mica to obtain support for the scheme, but these unfortunately yielded ambiguous images of incomplete polygons with blunt projections, rather than the expected "halo" of uncommitted trimer arms. These we interpret to be "dead ends" that failed to polymerize further because they included proteolyzed components. Further assembly experiments, avoiding such hazards, are indicated.