Search

Secretory signal peptides from individual prokaryotic and eukaryotic species have been analyzed, and the lengths and amino acid compositions of the positively charged amino‐terminal region, the central hydrophobic region, and the carboxy‐terminal cleavage‐region have been compared. We find distinct differences between species in all three regions. Implications for protein secretion in foreign hosts are discussed.

Translocation of proteins through the inner membrane of E. coliis normally catalyzed by the so‐called sec‐machinery. Yet, many integral inner membrane proteins appear not to require a fully functional sec‐machinery for proper insertion, in spite of the fact that sometimes quite sizable domains have to be translocated to the periplasmic side. This review will focus on recent studies of sec‐independent translocation events in an attempt to pin‐point the main differences between sec‐dependent and sec‐independent translocation.

Research on the targeting of proteins to the cytoplasmic membrane of E. colihas mainly focused on the so‐called ‘general secretory pathway’ (GSP) which involves the Sec‐proteins. Recently, evidence has been obtained for an alternative targeting pathway in E. coliwhich involves the signal recognition particle (SRP). The constituents of this SRP pathway in E. coliare homologous to those of the well‐characterized eukaryotic SRP pathway, which is the main targeting pathway for both proteins translocated across and inserted into the endoplasmic reticulum membrane. However, until recently, no clear function could be assigned to the SRP in E. coli. New studies point to an important role of the E. coliSRP in the assembly of inner membrane proteins.

The secmachinery catalyzes the translocation of nascent polypeptide chains across the inner membrane of E. coli, yet some inner membrane proteins depend only weakly or not at all on an intact secfunction for membrane insertion even though they have stretches of chain protruding into the periplasmic space. Earlier work has demonstrated that the length of a periplasmic loop correlates with its degree of sec‐dependence. We now show that the content of positively charged residues in a translocated loop also correlates with the degree of dependence on SecA function, suggesting that arginines and lysines may be inherently difficult to move across the membrane during sec‐dependent translocation.

Chloroplast transit peptides from the green alga Chlamydomonas reinhardtiihave been analyzed and compared with chloroplast transit peptides from higher plants and mitochondrial targeting peptides from yeast, Neurospora and higher eukaryotes. In terms of length and amino acid composition, chloroplast transit peptides from C. reinhardtiiare more similar to mitochondrial targetting peptides than to chloroplast transit peptides from higher plants. They also contain the potential amphiphilic α‐helix characteristic of mitochondrial presequences. However, in similarity with chloroplast transit peptides from higher plants, they contain a C‐terminal region with the potential to form an amphiphilic β‐strand. As in higher plants, transit peptides that route proteins to the thylakoid lumen consist of an N‐tenninal domain similar to stroma‐targeting transit peptides attached to a C‐terminal apolar domain that share many characteristics with secretory signal peptides.

The notion that mitochondrial targeting peptides form amphiphilic α-helices with one apolar and one polar, positively charged face is controversial, since some experimental results seem to imply that non-amphiphilic targeting peptides can also function as import signals. However, the standard methods used to assess the amphiphilicity of a peptide may be misleading, since they do not take the flexibility of the amino acid side chains into account. To demonstrate this, we have developed a new method for calculating the amphiphilicity of helical peptides.

We describe a secretory E. coliprotein with a novel phenotype: signal peptide cleavage is largely unaffected whereas chain translocation is efficiently blocked under conditions where SecA, a central component of the secretory machinery, is rendered non‐functional, and we have traced this phenotype to the presence of a mildly hydrophobic segment located ~30 residues downstream of the signal peptide. When this segment is deleted, normal SecA‐dependent signal peptide cleavage and chain translocation is observed; when its hydrophobicity is increased, it becomes a permanent membrane anchor with cleavage of the signal peptide and membrane insertion both being SecA‐independent. These findings suggest that the initial insertion of the signal peptide across the membrane can be uncoupled from the translocation process proper.

Proline residues are rarely found in the three most C-terminal positions of bacterial signal peptides, and have never been found in position +1 immediately following the cleavage site. It was recently shown that a Pro +1mutation in the E. colimaltose binding protein precursor not only prevents cleavage of the signal peptide but also inhibits the leader peptidase enzyme, resulting in cessation of cell growth (Barkocy-Gallagher, G.A. and Bassford, P.J. (1992) J. Biol. Chem. (in press)). Since maltose binding protein is dependent on the secmachinery for translocation across the inner membrane, it was not clear if this ‘Pro +1’ effect was restricted to sec-dependent proteins, or whether it applies also to proteins that do not require the secfunctions for translocation. We now present data suggesting that the striking phenotypic effects of Pro +1mutations can be elicited also by sec-independent proteins.

The role of the signal sequence in protein export is reviewed, and some difficulties inherent in the conventional picture of how it interacts with other components of the export machinery are pointed out. An alternative model is suggested, which seems to account better for some of the critical experimental findings made so far.

The evolution of the mitochondrial genome towards the compact organization found in the higher eukaryotes is discussed. It is suggested that the machinery for co-translational protein export across the endoplasmic reticulum membrane sets strict limits on the kinds of protein-coding genes that can be successfully transferred from the mitochondrial to the nuclear genome. This hypothesis is in perfect agreement with the pattern of mitochondrially vs nuclearly encoded mitochondrial proteins found in species such as man, mouse, and Xenopus.

Structural analysis of chloroplast transit peptides (cTPs), including secondary structure prediction and analysis of ‘cTP-like’ peptides of known 3D structure, suggests that cTPs are essentially flexible peptides devoid of regular secondary or tertiary structure. It is proposed that cTPs may be designed to interact with a succession of different chaperones on the chloroplast protein-import pathway.

Targeting of the cytoplasmic membrane protein leader peptidase (Lep) and a Lep mutant (Lep-inv) that inserts with an inverted topology compared to the wild-type protein was studied in Escherichia colistrains that are conditional for the expression of either Ffh or 4.5S RNA, the two components of the E. coliSRP. Depletion of either component strongly affected the insertion of both Lep and Lep-inv into the cytoplasmic membrane. This indicates that SRP is required for the assembly of cytoplasmic membrane proteins in E. coli.