Evidence for multiple mechanisms for membrane binding and integration via carboxyl-terminal insertion sequences.

Subcellular localization of proteins with carboxyl-terminal insertion sequences requires the molecule be both targeted to and integrated into the correct membrane. The mechanism of membrane integration of cytochrome b5 has been shown to be promiscuous, spontaneous, nonsaturable, and independent of membrane proteins. Thus endoplasmic reticulum localization for cytochrome b5 depends primarily on accurate targeting to the appropriate membrane. Here direct comparison of this mechanism with that of three other proteins integrated into membranes via carboxyl-terminal insertion sequences [vesicle-associated membrane protein 1(Vamp1), polyomavirus middle-T antigen, and Bcl-2] revealed that, unlike cytochrome b5, membrane selectivity for these molecules is conferred at least in part by the mechanisms of membrane integration. Bcl-2 membrane integration was similar to that of cytochrome b5 except that insertion into lipid vesicles was inefficient. Unlike cytochrome b5 and Bcl-2, Vamp1 binding to canine pancreatic microsomes was saturable, ATP-dependent, and abolished by mild trypsin treatment of microsomes. Surprisingly, although the insertion sequence of polyomavirus middle-T antigen was sufficient to mediate electrostatic binding to membranes, binding did not lead to integration into the bilayer. Together these results demonstrate that there are at least two different mechanisms for correct membrane integration of proteins with insertion sequences, one mediated primarily by targeting and one relying on factors in the target membrane to mediate selective integration. Our results also demonstrate that, contrary to expectation, hydrophobicity is not sufficient for insertion sequence-mediated membrane integration. We suggest that the structure of the insertion sequence determines whether or not specific membrane-bound receptor proteins are required for membrane integration.