Trigger factor is an abundant cytosolic protein of bacteria, first discovered in Escherichia coli by Wickner and co-workers in 1987 (1). They searched in a biochemical screen for cytosolic components involved in protein secretion of the precursor of OmpA protein, proOmpA, and identified a protein with an apparent molecular weight of 60 kDa that had the ability to form 1-1 stoichiometric complexes with proOmpA; this stabilized the precursor and facilitated its translocation into membrane vesicles. The protein was termed "trigger factor" because of its ability to trigger the folding of proOmpA into a membrane assembly-competent form in vitro.
Recent investigations revealed new interesting features of trigger factor. That from E. coli has peptidyl-prolyl-cis/trans isomerase (PPIase) activity and is capable of catalyzing protein folding in vitro much more efficiently than all other PPIases tested so far (2). In addition, trigger factor binds to the large subunit of ribosomes (3) and associates with cytosolic and secretory nascent polypeptide chains in vitro (4, 5). These findings led to the hypothesis that trigger factor acts as a cotranslational folding catalyst for newly synthesized polypeptides. Trigger factor was also described to cooperate with the chaperonin GroEL in stimulating protein degradation of an unstable protein, apparently by enhancing the interaction of GroEL with the substrate (6). Together, these findings suggest an involvement of trigger factor in processes related to cellular protein folding and metabolism. However, in vivo functions of trigger factor have not yet been demonstrated.
Genes encoding homologous trigger factor proteins were recently identified in other eubacteria, including Bacillus subtilis, Campylobacter jejuni, Haemophilus influenzae, Haemophilus actinomycetemcomitans, Mycoplasma pneumoniae, Mycoplasma genitalium, and Synechocystis PCC6803, whereas no homologous protein has been described so far in archaebacteria and eukaryotic cells. This review will focus mainly on the E. coli trigger factor, since most mechanistic analyses were performed with this homologue.
1. A peptidyl-prolyl-cis/trans Isomerase
Peptidyl-prolyl-cis/trans isomerases (PPIases) are enzymes that catalyze the cis-trans isomerization of peptide bonds preceding proline residues (Xaa-Pro peptide bonds) (see Peptidyl Prolyl Cis/Trans Isomerases). This isomerization can be rate-limiting in the folding of a polypeptide chain into its native structure (see Protein folding). Trigger factor is proposed to represent a new member of the family of FK506 binding (FKBP)-type PPIases on the basis of the following findings. Stoller et al. (2) identified trigger factor as a ribosome-bound PPIase. Using series of tetrapeptides of the sequence Suc-Ala-Xaa-Pro-Phe- p-nitroanilide to monitor PPIase activity, trigger factor resembled in its Xaa-Pro substrate specificity PPIases of the FKBP-family (2). Independently, two groups reported on a significant sequence homology of trigger factor with PPIases of the FKBP-family (5, 7). In particular, hydrophobic and aromatic residues forming the substrate-binding pocket in the well-known structure of the human FKBP12 homologue are found in trigger factor as well. A puzzling finding, however, is that the E. coli trigger factor is not inhibited by FK506 at concentrations of up to 100 mM (2), while the B. subtilis homologue is half maximally inhibited by 0.5 |MFK506 (8). Confirmation of the FKBP-domain as a structural and functional element of trigger factor was achieved by limited proteolysis of the native protein, which generated a stable fragment of 12 kDa constituting the predicted FKBP domain (Fig. 1) When assayed for PPIase activity toward tetrapeptides, this fragment displayed the same specific activity as the full-length protein (9, 10).
Alignment of the sequences of trigger factor proteins from different prokaryotic species revealed that the central parts of the proteins representing the FKBP domain exhibit a strikingly higher degree of conservation than do their N- or C-terminal parts, suggesting that the PPIase activity is a particularly conserved feature of the trigger factor proteins (10).
Figure 1. Schematic representation of the domain structure of E. coli trigger factor. Hatched boxes show the domains for which activities have been assigned. Arrows with numbers indicate the residues at the domain boundaries.
2. Chaperone activity
The ability of trigger factor to form 1-1 stoichiometric complexes with unfolded proOmpA and to stabilize the translocation-competent form of this precursor is reminiscent of molecular chaperones. A chaperone-like activity is further indicated by the trigger factor ability to associate with high affinity (apparent dissociation constant ~0.7 |iM) with the unfolded form of a mutant ribonuclease T1 (11), and this high affinity binding is a prerequisite for trigger factor’s excellent catalysis of refolding of the same protein in vitro (2, 11). Refolding of this protein is rate-limited by the cis-trans isomerization of a single proline peptide bond. Because of the high affinity of trigger factor for the unfolded substrate, it outscores all other PPIases in its efficiency of catalyzing the prolyl isomerization, although the catalytic constant is low (&cat~1.3 |iM) (11). The activity of the isolated central PPIase domain of trigger factor is reduced 800-fold, whereas its PPIase activity toward tetrapeptide substrates is comparable to that of full-length trigger factor (11).
These findings indicate distinct sites on trigger factor for high affinity substrate binding and prolyl-isomerization and suggest that the exceptional catalytic efficiency of trigger factor originates from a cooperation of its PPIase and chaperone-like activities (11). Tight substrate binding may position the stretch of the protein substrate that contains the critical proline residue close to the active site of the PPIase.
3. Association with Nascent Polypeptide Chains
Two research groups independently demonstrated that E. coli trigger factor can associate with nascent polypeptide chains. Luirink and coworkers discovered that trigger factor can be efficiently crosslinked to all nascent chains arrested on the ribosome (4). Trigger factor competes with P48, a component of the E. coli signal recognition particle (SRP), for crosslinking to nascent chains of precursors of secretory proteins. In contrast to P48, however, trigger factor also crosslinks efficiently to nascent chains of cytoplasmic polypeptides (4). Bukau and co-workers identified trigger factor as the main component that associates with ribosomes translating b-galactosidase (5). This association was resistant to high salt but was disrupted by puromycin treatment, which leads to premature termination of translation, suggesting an association of trigger factor with b-galactosidase nascent chains (5). Further evidence for an association of trigger factor with nascent chains was provided by its photocrosslinking to nascent chains of secretory as well as nonsecretory derivatives of preprolactin (5).
The ability of trigger factor to associate with nascent chains emerging from the ribosome suggests that this protein has a binding site on the ribosome that positions it near the exit site for nascent chains. Trigger factor is indeed exclusively associated with the large ribosomal subunit, which harbors this site (3), and trigger factor cannot be crosslinked to nascent polypeptides after their puromycin-mediated release from the ribosome (4, 5). Interestingly, trigger factor also associates with eukaryotic wheat germ ribosomes, suggesting conservation in evolution of its binding site on the ribosome (5).
4. Role with GroEL in Binding and Degradation of Proteins
A further role for trigger factor in protein metabolism is indicated by the finding that trigger factor stimulates the degradation of CRAG, an artificial protein composed of fragments of the l cro repressor, b-galactosidase, and protein A that has been used as a model unfolded substrate for studies of protein degradation in E. coli. Proteolysis of CRAG by the ClpP proteinase depends on the interaction of CRAG with the chaperonin GroEL, and formation of this complex was the rate-limiting step in degradation. Trigger factor accelerates the chaperone-dependent degradation of CRAG by promoting its binding to GroEL. A ternary complex of CRAG, GroEL, and trigger factor was demonstrated by affinity chromatography. Addition of ATP causes dissociation of the complex of GroEL and trigger factor from the substrate (6). Trigger factor interacts directly with GroEL prior to association with CRAG. This interaction differs from that of GroEL with denatured substrates in that addition of ATP and GroES, which causes substrate release from GroEL, does not dissociate the complex (12). Trigger factor also enhances the binding of GroEL to unfolded proteins other than CRAG (12). This indicates that trigger factor is involved in both degradation and folding of denatured substrates by stimulating their binding to the chaperonin (12).
5. Domain Organization
The E. coli trigger factor consists of 432 amino acids with a molecular weight of 48 kDa. Limited proteolysis of the full-length protein with proteinases revealed a compactly folded proteinase-resistant central domain comprising residues 145-247 (Fig. 1) (9, 10). This domain has homology to FKBPs and displays PPIase activity, and thus represents the catalytic core of the protein.
The ^-terminal part of trigger factor forms another structural and functional module. The ^-terminal 144 residues are necessary and sufficient for specific binding of trigger factor to the large ribosomal subunit, both in vitro and in vivo (13). This fragment contains a compactly folded domain comprising the amino-terminal 118 amino acids (Fig. 1). It copurifies with E. coli ribosomes from cell extracts and is capable of associating with isolated ribosomes in vitro; therefore, it represents the ribosome-binding domain of trigger factor (13).
Intriguingly, the N-terminal fragment alone, or with the adjacent PPIase domain, is unable to form complexes with translating ribosomes that are resistant to high salt concentrations. Furthermore, the association of both fragments with ribosomes was not altered by puromycin, suggesting that the sensing of the translational status requires the C-terminus of trigger factor (13). It is conceivable that the C-terminus mediates the tight association with nascent polypeptide chains or with the translating ribosomes, which consequently leads to the formation of high salt-resistant complexes of trigger factor with the translating ribosome.
6. Cellular Function
Despite the considerable progress made in the biochemical dissection of trigger factor in vitro, the role of this protein in E. coli remains unclear. The known properties of trigger factor strongly suggest an important biological function in folding of newly synthesized proteins and raises the attractive hypothesis that trigger factor acts as a cotranslational folding catalyst. Interestingly, expression of the tig gene encoding trigger factor is growth-phase-controlled and coregulated with genes encoding ribosomal components (14). A possible role of trigger factor in GroEL-assisted protein metabolism (12) is not yet established.
Although a knockout mutation of the tig gene is not available yet, the available data indicate that trigger factor is not essential for E. coli growth at 37°C. Depletion of trigger factor to less than 5% of the cellular wild-type levels causes formation of short filaments, indicating cell division defects. Trigger factor-depleted cells remain fully viable at 37°C, however, and have no additional cellular defects (14). For example, although trigger factor enhances translocation of proOmpA into vesicles in vitro (1), trigger factor-depleted cells have no defect in translocation of the precursor, even in a secB mutant background that lacks activity of the secretion-specific chaperone SecB (14). It is unclear whether the activities of other chaperones or PPIases existing in the E. coli cytosol can compensate at 37°C for the missing activity of trigger factor in the depleted cells.
The only detected phenotype of trigger factor-depleted cells was observed at low temperature. Depleted cells stored at 4°C die faster than do normal cells, while cells overproducing trigger factor show enhanced viability at this temperature (15). This suggests an involvement of trigger factor in cell survival at temperatures below the growth temperature limit. Trigger factor may be particularly important at low temperatures for folding of proteins requiring prolyl-bond isomerization. This isomerization is slower at low temperature, so there is an increased requirement for catalysis. Furthermore, the cellular level of trigger factor is about twofold higher at 16°C and 4°C than at 37°C (15).
The importance of trigger factor for metabolism of bacteria is also indicated by the discovery of a tig gene in Mycoplasma genitalium. This bacterium is believed to be free from genetic redundancy and thus contains only the minimal set of genes required for life. Trigger factor appears to be the only PPIase of this organism (16). Identification of the in vivo functions of trigger factor thus seems to be of importance for understanding the fundamental process of protein folding in the bacterial cytosol.
