In mammalian cells, most membrane proteins are cotranslationally inserted into the membrane of the endoplasmic reticulum (ER) by the Sec61 translocon (1). During the insertion process, hydrophobic segments in the nascent polypeptide are dispelled laterally from the translocon to form transmembrane (TM) ␣-helices with either an N in -C out (i.e., with the N terminus in the cytosol) or N out-C in orientation relative to the membrane (2). From the point of view of the translocon, N out -C in TM helices enter the translocation channel as part of a translocating nascent chain, whereas N in -C out TM helices must gate the channel open and presumably remain in or very near the channel during translocation of the C-terminal parts of the nascent chain (Fig. 1). In principle, this means that the sequence requirements for membrane insertion may be different for N in -C out and N out -C in TM helices.In previous work (3-5), we have carried out a detailed analysis of how different amino acids contribute to the overall efficiency of membrane insertion of TM helices with an N out -C in orientation (also called stop-transfer sequences). Here, we present a similar analysis, but for TM helices of the opposite orientation, i.e., N in -C out . The analysis is done both by using in vitro translation of model constructs in the presence of dog pancreas rough microsomes (RMs) and by expression in the yeast Saccharomyces cerevisiae. We find that individual amino acids affect membrane insertion in much the same way irrespective of the orientation of the TM helix. We also show that the hydrophobicity required for 50% membrane insertion for a N in -C out TM helix can be as much as Ϸ1 kcal/mol less than for a N out -C in TM helix; this difference can be explained in part by the influence from a neighboring N out -C in TM helix on the membrane insertion efficiency of the N in -C out helix.
Results
Model Protein and Membrane Insertion Assay.In our previous studies of N out -C in TM helices, we used a construct derived from the Escherichia coli leader peptidase (Lep) protein shown in Fig. 2A Left. Lep has two transmembrane helices (TM1 and TM2) and a large C-terminal luminal domain (P2). Test segments (H-segments) are inserted into the P2 domain where they are flanked by two engineered acceptor sites for N-linked glycosylation (G1 and G2). The acceptor sites provide a convenient way to measure insertion efficiency into the ER: Lep constructs with H-segments that insert into the membrane are only glycosylated by the luminally disposed