Secretion in yeast: preprotein binding to a membrane receptor and ATP- dependent translocation are sequential and separable events in vitro

We have used a cytosol-free assay in which efficient translocation and signal peptide cleavage is achieved when the affinity-purified precursor of OmpA (proOmpA) is diluted out of 8 M urea into a suspension of yeast rough microsomes. This aspect of protein targeting and transport occurs in two discernible steps: (a) in the absence of ATP and cytosolic factors, the precursor binds to the membranes but is not translocated; (b) addition of ATP results in the translocation of the bound precursor and its processing to the mature form. The binding to microsomes of radiolabeled proOmpA is saturable and inhibited by the addition of unlabeled proOmpA but not by mature OmpA or other proteins. The binding of radiolabeled prepro-alpha-factor is also effectively competed by other preproteins, but not by mature ones. Scatchard analysis showed the Kd of proOmpA to be 7.5 X 10(-9) M. Binding is most likely protein mediated as treatment of the microsomes with the protease papain was found to be inhibitory. These results represent the first functional characterization of secretory protein precursor binding to membranes. Alkylation of the microsomes with NEM, washing the membranes with urea or using membranes from the (translocation) mutant ptll at the nonpermissive temperature, did not affect binding, but did eliminate the subsequent ATP-dependent translocation. The ability to subdivide translocation into individual reactions provides a more precise means of determining the membrane components involved in this process.     

ProOmpA spontaneously folds in a membrane assembly competent state which trigger factor stabilizes.

The precursor protein proOmpA can translocate across purified Escherichia coli inner membrane vesicles in the absence of any other soluble proteins. ProOmpA, purified 2000-fold in the presence of 8 M urea, is competent for translocation following rapid renaturation via dilution. ATP, the transmembrane electrochemical potential, and functional secY protein are essential for the translocation of proOmpA renatured by dilution. The kinetics of its translocation and the level of translocation at each concentration of ATP are indistinguishable from that of proOmpA renatured by dialysis with trigger factor. After dilution, the proOmpA rapidly loses its competence for membrane assembly. However, this competence is stabilized by trigger factor. Assembly-competent proOmpA is in a protease-sensitive conformation, whereas proOmpA which has lost this competence is more resistant to degradation. This suggests that the primary role for trigger factor in in vitro protein translocation is to maintain precursor proteins in a translocation-competent conformation. We propose that a properly folded precursor protein and ATP are the only soluble components which are essential for bacterial protein translocation.     

Secretion in yeast: structural features influencing the post-translational translocation of prepro-alpha-factor in vitro.

In vitro, efficient translocation and glycosylation of the precursor of yeast alpha-factor can take place post-translationally. This property of prepro-alpha-factor appears to be unique as it could not be extended to other yeast protein precursors such as preinvertase or preprocarboxypeptidase Y. In order to determine if specific domains of prepro-alpha-factor were involved in post-translational translocation, we carried out a series of experiments in which major domains were either deleted or fused onto reporter proteins. Fusion of various domains of prepro-alpha-factor onto the reporter protein alpha-globin did not allow post-translational translocation to occur in the yeast in vitro system. Prepro-alpha-factor retained its ability to be post-translationally translocated when parts or all of the pro region were deleted. Removal of the C-terminal repeats containing mature alpha-factor had the most profound influence as post-translational translocation decreased in proportion to the number of repeats deleted. Taken together, these results suggest that efficient post-translational translocation requires a signal sequence and the four C-terminal repeats. There does not however, appear to be specific information contained within the C-terminus, as their presence in fusion did not enable the post-translational translocation of reporter proteins. Lastly, the ability to post-translationally translocate radiochemically pure prepro-alpha-factor that had been isolated by immuno-affinity chromatography required the addition of a yeast lysate fraction. Moreover, post-translational translocation is a function of the microsomal membrane of yeast microsomes and not of a factor peculiar to the yeast lysate, as reticulocyte lysate supported this as well.     

Prepro-carboxypeptidase Y and a truncated form of pre-invertase, but not full-length pre-invertase, can be posttranslationally translocated across microsomal vesicle membranes from Saccharomyces cerevisiae

We have determined that prepro-carboxypeptidase Y and a truncated form of pre-invertase can be translocated across the yeast microsomal membrane post-translationally in a homologous in vitro system. The yeast secretory protein prepro-alpha-factor which was previously shown to be an efficient posttranslational translocation substrate is therefore not unique in this regard, but rather the yeast ER protein translocation machinery is generally capable of accepting substrates from a ribosome-free, soluble pool. However, within our detection limits, full-length pre-invertase could not be translocated posttranslationally, but was translocated co-translationally. This indicates that not every fully synthesized pre-protein can use this pathway, presumably because normal or aberrant folding characteristics can interfere with translocation competence.     

Secretion in yeast: translocation and glycosylation of prepro-alpha-factor in vitro can occur via an ATP-dependent post-translational mechanism

In an in vitro system comprising a yeast cell-free translation system, yeast microsomes and mRNA encoding prepro-alpha-factor, the translocation of this protein across the membrane of the microsomal vesicle and its glycosylation could b uncoupled from its translation. Such post-translational processing is dependent upon the presence of ATP in the system. It is not, however, affected by a variety of uncouplers, ionophores or inhibitors, including carbonyl cyanide m-chlorophenyl hydrazone (CCCP), valinomycin, nigericin, dinitrophenol (DNP), potassium cyanide (KCN) or N-ethyl maleimide (NEM). This mechanism of translocation is significant as it indicates that a protein of 18 000 daltons is capable of crossing an endoplasmic reticulum-derived membrane post-translationally. For the moment, this phenomenon seems to be restricted to prepro-alpha-factor in the yeast in vitro system. Neither invertase nor IgG chi light chain could be translocated post-translationally in yeast, nor was such processing observed for prepro-alpha-factor in a wheat germ system supplemented with canine pancreatic microsomes.     

Mutations in the signal sequence of prepro-alpha-factor inhibit both translocation into the endoplasmic reticulum and processing by signal peptidase in yeast cells.

The effects of five single-amino-acid substitution mutations within the signal sequence of yeast prepro-alpha-factor were tested in yeast cells. After short pulse-labelings, virtually all of the alpha-factor precursor proteins from a wild-type gene were glycosylated and processed by signal peptidase. In contrast, the signal sequence mutations resulted in the accumulation of mostly unglycosylated prepro-alpha-factor after a short labeling interval, indicating a defect in translocation of the protein into the endoplasmic reticulum. Confirming this interpretation, unglycosylated mutant prepro-alpha-factor in cell extracts was sensitive to proteinase K and therefore in a cytosolic location. The signal sequence mutations reduced the rate of translocation into the endoplasmic reticulum by as much as 25-fold or more. In at least one case, mutant prepro-alpha-factor molecules were translocated almost entirely posttranslationally. Four of the five mutations also reduced the rate of proteolytic processing by signal peptidase in vivo, even though the signal peptide alterations are not located near the cleavage site. This study demonstrates that a single-amino-acid substitution mutation within a eucaryotic signal peptide can affect both translocation and proteolytic processing in vivo and may indicate that the recognition sequences for translocation and processing overlap within the signal peptide.     

Export of prepro-alpha-factor from Escherichia coli

Yeast prepro-alpha-factor translocates posttranslationally into yeast microsomes in vitro. This process is strongly influenced by the extreme carboxyl-terminal region of the protein. These features contrast with the properties of most eucaryotic proteins which are translocated into the endoplasmic reticulum. We have extended these studies by introducing the gene for the wild-type and several mutant forms of prepro-alpha-factor into Escherichia coli. Prepro-alpha-factor is secreted into the periplasm and processed to pro-alpha-factor. Its translocation across the plasma membrane requires the membrane potential and the secY gene product. Deletion mutant analysis showed that features of the pro-segment were essential for secretion of prepro-alpha-factor in E. coli, while the carboxyl-terminal region, which is required in yeast, is dispensible in E. coli. Neither size nor the presence of a unique topogenic sequence was sufficient to explain the requirement for the pro-segment.     

ProOmpA is stabilized for membrane translocation by either purified E. coli trigger factor or canine signal recognition particle

We have isolated large amounts of E. coli outer-membrane protein A precursor (proOmpA). Purified proOmpA is active in membrane assembly, and this assembly is saturable with respect to the precursor protein. A proOmpA-Sepharose matrix allows affinity isolation of trigger factor, a soluble, 63,000 dalton monomeric protein that stabilizes proOmpA in assembly competent form. Comparison of trigger factor's amino-terminal sequence with those in a computer data bank and with those encoded by sec genes, as well as groEL and heat shock gene dnaK, suggests that trigger factor is encoded by a previously undescribed gene. Trigger factor and proOmpA form a 1:1 complex that can be isolated by gel filtration. Purified canine signal recognition particle (SRP) can also stabilize proOmpA for membrane insertion. This postribosomal activity of SRP suggests a unifying theme in protein translocation mechanisms.     

Secretion in yeast: in vitro analysis of the sec53 mutant.

Of central importance to studying protein translocation via a combined genetic and biochemical approach is the in vitro analysis of yeast conditionally-lethal secretory mutants. Analysis of sec53 presented an opportunity not only to see if mutants could be examined in recently developed yeast in vitro translocation systems, but also to characterize further the nature of this mutant originally postulated to be defective in protein translocation. Membranes from sec53 were capable of translocating and glycosylating nascent prepro-alpha-factor in vitro in both sec53 and wild-type lysates at temperatures that were non-permissive for growth of the mutant cells. These results suggested that the Sec53 protein does not function directly in the translocation and glycosylation of prepro-alpha-factor. To examine this point further, we isolated membranes from sec53 cells that had been grown at the non-permissive temperature prior to disruption. In such cases, regardless of assay temperature, membranes from sec53 cells efficiently translocated but failed to glycosylate prepro-alpha-factor in vitro. The in vitro phenotype of sec53 could be mimicked by isolating rough microsomes from wild-type cells that had been grown for 1 h in the presence of tunicamycin. Together, these results demonstrate that sec53 is not defective in translocation, rather in assembly of the dolichol-oligosaccharide substrate needed for N-linked glycosylation.     

In vitro protein translocation across the yeast endoplasmic reticulum: ATP-dependent posttranslational translocation of the prepro-alpha-factor

The in vitro synthesized precursor of the alpha-factor pheromone, prepro-alpha-factor, of Saccharomyces cerevisiae was translocated across yeast microsomal membranes in either a homologous or a wheat germ cell free system. Translocated prepro-alpha-factor was glycosylated, sedimented with yeast microsomal vesicles, and was protected from digestion by added protease, but was soluble after alkaline sodium carbonate treatment. Thus prepro-alpha-factor was properly sequestered within yeast microsomal vesicles, but was not integrated into the lipid bilayer. In marked contrast to protein translocation across mammalian microsomal membranes, translocation of prepro-alpha-factor across yeast microsomal membranes could occur posttranslationally. This reaction required protein components in the yeast microsomal fraction that could be inactivated by alkylation or proteolysis, was ATP-dependent, and was insensitive to the presence of a variety of uncouplers and ionophores.     

Secretion in yeast. Purification and in vitro translocation of chemical amounts of prepro-alpha-factor

The Saccharomyces cerevisiae mating pheromone precursor, prepro-alpha-factor, can be translocated across yeast endoplasmic reticulum membranes post-translationally in an in vitro system. This characteristic makes prepro-alpha-factor potentially useful as a probe in the biochemical dissection of the mechanism of this basic cellular process. Efforts have been limited by the inability to isolate sufficient quantities of such secretory protein precursors in a translocation-competent form. We report here the one-step purification of chemical amounts of translocation-competent prepro-alpha-factor using nickel ion affinity chromatography on nitrilotriacetate resin. An oligonucleotide encoding 6 histidine residues was inserted into a genomic clone encoding prepro-alpha-factor 5' of the naturally occurring translational stop codon by site-directed mutagenesis. The construct was expressed at high levels in a SecY- strain of Escherichia coli. The produced preprotein was solubilized in 6 M guanidine hydrochloride and bound to nitrilotriacetate resin. Prepro-alpha-factor was recovered at a purity in excess of 95% by elution with 0.25 M imidazole, 8 M urea, which competitively displaced the histidine affinity tag from the nickel column. The chemical amounts of prepro-alpha-factor obtained in this way were determined to be competent for translocation across yeast microsomal membranes and for subsequent modifications such as signal sequence cleavage and N-linked glycosylation.     

SecA protein, a peripheral protein of the Escherichia coli plasma membrane, is essential for the functional binding and translocation of proOmpA.

We have reconstituted protein translocation across plasma membrane vesicles of Escherichia coli using purified proOmpA and trigger factor, a 63 kd soluble protein. Treatment of membrane vesicles with urea inactivates them for translocation unless a factor present in cytoplasmic extracts is added during the translocation reaction. Sedimentation analysis showed that the stimulatory activity is of distinctly higher mol. wt than trigger factor. Cytoplasmic extracts from a strain that greatly overproduces the SecA protein are highly enriched in the stimulatory activity for untreated membranes and restore translocation to urea-treated membranes, suggesting that this protein is the stimulatory factor. This assay was used to monitor the isolation of SecA protein from the overproducing strain. The purified protein is soluble, yet binds peripherally to membranes with high affinity and supports translocation. Using pure proOmpA, SecA protein, trigger factor and urea-treated membranes, the protein export process was resolved into binding and translocation steps. We find that proOmpA binds to membrane vesicles with or without SecA protein, but that translocation only occurs when SecA was bound prior to proOmpA.     

Full-length prepro-alpha-factor can be translocated across the mammalian microsomal membrane only if translation has not terminated

We have previously shown that fully synthesized prepro-alpha-factor (pp alpha F), the precursor for the yeast pheromone alpha-factor, can be translocated posttranslationally across yeast rough microsomal (RM) membranes from a soluble, ribosome-free pool. We show here that this is not the case for translocation of pp alpha F across mammalian RM. Rather we found that a small amount of translocation of full-length pp alpha F is observed, but is solely due to polypeptide chains that were still ribosome bound and covalently attached to tRNA, i.e., not terminated. In addition, both signal recognition particle (SRP) and SRP receptor are required, i.e., the same targeting machinery that is normally responsible for the coupling between protein synthesis and translocation. Thus, the molecular requirements for targeting are distinct from posttranslational translocation across yeast RM. As termination is generally regarded as part of translation, the translocation of full-length pp alpha F across mammalian RM does not occur "posttranslationally," albeit independent of elongation. Most other proteins for which posttranslational translocation across mammalian RM was previously claimed fall into the same category in that ribosome attachment as peptidyl-tRNA is required. To clearly separate these two distinct processes, we suggest that the term posttranslational be reserved for those processes that occur in the complete absence of the translational machinery. We propose the term "ribosome-coupled translocation" for the events described here.     

Translocation of ProOmpA possessing an intramolecular disulfide bridge into membrane vesicles of Escherichia coli. Effect of membrane energization

In the absence of delta mu H+, the in vitro translocation of proOmpA resulted in the stable accumulation of a possible translocation intermediate in addition to a transiently accumulating one. The stable intermediate was detected on a polyacrylamide gel as two proteinase K-resistant bands corresponding to a molecular weight of about 28,000. The appearance of the bands was appreciably enhanced when proOmpA was oxidized with ferricyanide. No mature OmpA appeared. When proOmpA reduced with dithiothreitol was used, on the other hand, the bands did not appear at all. Upon the replacement of Cys302 of OmpA with Gly, the intermediate accumulation was abolished. The proOmpA treated with dithiothreitol was labeled with N-[3H]-ethylmaleimide, whereas that treated with ferricyanide was not. The ferricyanide-treated proOmpA was translocated into membrane vesicles in the presence of delta mu H+. The mature OmpA thus translocated and processed was not labeled with N-[3H]ethylmaleimide. It is concluded that proOmpA possessing the Cys290-Cys302 disulfide bridge can be translocated without cleavage of the bridge, when delta mu H+ is imposed. The accumulation of the disulfide bridge-containing intermediate was ATP-dependent, whereas its conversion to the translocated mature form was not blocked in the presence of adenosine 5'-(beta, gamma-imino)triphosphate. It is concluded that the early and late stages of the translocation reaction require ATP and delta mu H+ differently.     

Dissociation of complexes between 70 kDa stress proteins and presecretory proteins is facilitated by a cytosolic factor.

Members of the 70 kDa stress protein family were shown previously to facilitate the posttranslational translocation of presecretory proteins into the endoplasmic reticulum and protein precursors into mitochondria. To identify proteins that interact with 70 kDa stress proteins during the early steps of posttranslational translocation, polyclonal antibodies were raised against purified yeast cytosolic stress proteins. They were used to immunoprecipitate complexes between 70 kDa stress proteins and a radiolabeled presecretory protein, prepro-alpha-factor, that was translated in vitro. Complexes between prepro-alpha-factor and 70 kDa stress proteins were stable, but could be dissociated in the presence of ATP and crude cytosolic extracts from yeast. These results are consistent with the idea that 70 kDa stress proteins act as molecular chaperones in translocation by binding to precursor proteins before or during their passage across membranes.     

Roles of cytosolic Hsp70 and Hsp40 molecular chaperones in post-translational translocation of presecretory proteins into the endoplasmic reticulum

Hsp70 molecular chaperones and their co-chaperones work together in various cellular compartments to guide the folding of proteins and to aid the translocation of proteins across membranes. Hsp70s stimulate protein folding by binding exposed hydrophobic sequences thereby preventing irreversible aggregation. Hsp40s stimulate the ATPase activity of Hsp70s and target unfolded proteins to Hsp70s. Genetic and biochemical evidence supports a role for cytosolic Hsp70s and Hsp40s in the post-translational translocation of precursor proteins into endoplasmic reticulum and mitochondria. To gain mechanistic insight, we measured the effects of Saccharomyces cerevisiae Ssa1p (Hsp70) and Ydj1p (Hsp40) on the translocation of histidine-tagged prepro-alpha-factor (ppalphaF6H) into microsomes. Radiolabeled ppalphaF6H was affinity purified from wheat germ translation reactions (or Escherichia coli) to remove endogenous chaperones. We demonstrated that either Ssa1p or Ydj1p stimulates post-translational translocation by preventing ppalphaF6H aggregation. The binding and/or hydrolysis of ATP by Ssa1p were required to maintain the translocation competence of ppalphaF6H. To clarify the contributions of membrane-bound and cytosolic Ydj1p, we compared the efficiency of chaperone-dependent translocation into wild-type and Ydj1p-deficient microsomes. Neither soluble nor membrane-bound Ydj1p was essential for post-translational protein translocation. The ability of Ssa1p, Ydj1p, or both chaperones to restore the translocation competence of aggregated ppalphaF6H was negligible.     

Inhibition of protein translocation in permeabilized cells of Schizosaccharomyces pombe by puromycin.

To investigate protein translocation in eukaryotes, we reconstituted a protein translocation system using the permeabilized spheroplasts (P-cells) of the fission yeast Schizosaccharomyces pombe. The precursor of a sex pheromone of Saccharomyces cerevisiae, prepro-alpha-factor, was translocated across the endoplasmic reticulum (ER) of S. pombe posttranslationally, and glycosylated to the same extent as in the ER of S. cerevisiae. This suggested that the size of N-linked core-oligosaccharide in the ER of S. pombe is similar to that in S. cerevisiae. This translocation into the ER of S. pombe was inhibited by puromycin, but the translocation in the P-cells of S. cerevisiae was not inhibited. This difference in sensitivity to puromycin was due to the membrane but not the cytosolic fraction. Our results suggested that the translocation machinery of S. pombe was sensitive to puromycin and different from that of S. cerevisiae.     

Trigger factor depletion or overproduction causes defective cell division but does not block protein export.

Trigger factor is an abundant cytosolic protein of Escherichia coli which can stabilize proOmpA for in vitro translocation across inner membrane vesicles. The gene encoding E. coli trigger factor was isolated and sequenced, allowing construction of strains in which the expression of trigger factor is readily regulated. We found no defect in the in vivo rate of synthesis or secretion of proOmpA in trigger factor-depleted cells. The primary physiological defect in trigger factor-depleted or -overproducing cells is an enrichment of filamented cells. Filamentation of the trigger factor-overproducing strain is suppressed by a multicopy plasmid expressing the essential division gene ftsZ, suggesting that trigger factor has an important role in cell division.     

Translocation of nascent secretory proteins across membranes can occur late in translation.

Signal recognition particle (SRP) causes an arrest in the translation of nascent secretory proteins in a wheat germ cell-free system. In order to examine at what point during the synthesis of a secretory protein its translocation across the endoplasmic reticulum (ER) membrane can occur, SRP was used to arrest nascent chain elongation at various times during a synchronous translation, thus allowing the generation of nascent chains of increasing length. It was found that SRP can still bring about an arrest as late as when an average of two-thirds of nascent IgG light chain was completed. Rough microsomes were added to translations blocked with SRP to determine if such relatively long nascent chains could still be translocated across the membrane. It was found that nascent chains which had been arrested by SRP, regardless of their length, could be translocated into rough microsomes. In the case of IgG light chain, translocation levels of 50% were still observed with nascent chains corresponding to as much as 70-75% of the intact preprotein. Similar results were observed for the nascent bovine prolactin precursor. These results demonstrate that the synthesis of secretory proteins can be uncoupled from their translocation, and that fairly large nascent chains are capable of crossing the membrane of the ER post-translationally.     

Reconstitution of protein translocation from solubilized yeast membranes reveals topologically distinct roles for BiP and cytosolic Hsc70

We reconstituted prepro-alpha-factor translocation and signal peptide processing using a yeast microsomal detergent soluble fraction formed into vesicles with soybean phospholipids. Reconstituted translocation required ATP, and was deficient when sec63 and kar2 (BiP) mutant cells were used as a source of membranes. Normal translocation was observed with vesicles reconstituted from a mixture of pure wild-type yeast BiP and a soluble fraction of kar2 mutant membranes. Two other heat-shock cognate (hsc) 70 homologs, yeast cytosolic hsc70 (Ssalp) and E. coli dnaK protein did not replace BiP. Conversely, BiP was not active under conditions where translocation into native ER vesicles required cytosolic hsc70. We conclude that cytosolic hsc70 and BiP serve noninterchangeable roles in polypeptide translocation, possibly because distinct, asymmetrically oriented membrane proteins are required to recruit each protein to opposing surfaces of the ER membrane.