Sequestration revisited: integrating traditional electron microscopy, de novo assembly and new results.

Electron microscopy analysis of the autophagic sequestration membrane (SM) in various metazoan cell types after different fixation methods shows that: (1) the growing SM cannot derive from preformed rough surfaced endoplasmic reticulum (RER) membranes by transformation; (2) the empty cleft between the two layers of the SM after aldehyde fixation is an artifact of sample preparation; (3) the SM emerges from and grows de novo in cytoplasmic areas where membranous precursors cannot be identified by traditional electron microscopy; (4) the growing SM consists of two tightly packed membrane layers with a sharp bend at the edge; (5) changes in the environment of the growing SM participate in the determination of the size and shape of the autophagosome. We suggest that expansion as well as regression takes place at the edge of the growing SM. Stabilization and irreversibility of formation of the SM is achieved by closure. The immediate source of lipids for the SM must be in the cytoplasmic matrix, supposedly in the form of special phospholipid carrying vesicles that might involve the transmembrane Atg9 protein. To explain the apparent lack of such vesicles by electron microscopy we suggest that they are too small, have a similar density to other frequently occurring structures, or are destroyed during sample preparation.