The 'Bayer bridges' confronted with results from improved electron microscopy methods.

In electron micrographs of conventionally prepared thin sections of Escherichia coli one observes (i) a wavy appearance of the two membranes showing frequent appositions (named adhesion sites) and (ii) intermembrane bridges after plasmolysis which, it is claimed, occur at the adhesion sites and are related to intermembrane protein transport (transmigration). When chemical fixation is replaced by cryofixation, the observations are very different. (a) The two membranes are equally spaced and no contacts, adhesions or other sorts of connections are visible. (b) After plasmolysis the protoplast is shrunken, but the typical bridges are no longer produced. (c) In addition, when peptidoglycan is stained on conventionally prepared sections, it is revealed as a 7-nm-thick sacculus which is not interrupted at the sites of apposition. In view of the new observations, the structural concepts derived from conventionally prepared material must be revised. It is proposed that the intermembrane space is entirely filled by a gel, the outer part of which is the 7 nm thick, very stable, chemically resistant peptidoglycan (or murein). The inner part is much less stable and is proposed to undergo rapid autolytic changes upon cell death. The large 'Bayer bridges' might then tentatively be explained as an artificial post-mortem enhancement of either a stream of proteins transmigrating across the periplasm or of a pre-existing, but not yet resolved, structure. This enhancement probably occurs during the 7-10 min between plasmolysis and fixation that are prescribed for the procedure necessary for revealing 'Bayer bridges'.