Retention and retrieval in the endoplasmic reticulum and the Golgi apparatus.

Resident proteins of the exocytic pathway contain at least two types of information in their primary sequence for determining their subcellular location. The first type of information is found at the carboxyl terminus of soluble proteins of the endoplasmic reticulum (ER) and in the cytoplasmic domain of some ER and Golgi membrane proteins. It acts as a retrieval signal, returning proteins that have left the compartment in which they reside. The second type of information has been found in the membrane-spanning domain of several ER and Golgi proteins and, though the mechanism by which it operates is still unclear, it acts as a retention signal, keeping the protein at a particular location within the organelle. The presence of both a retrieval signal and a retention signal in a trans-Golgi network resident protein suggests that more than one mechanism operates to ensure correct localization of resident proteins along the exocytic pathway.

Control of nucleocytoplasmic transport.

The movement of macromolecules between the nucleus and cytoplasm is tightly controlled. In the past few years it has become increasingly apparent that nuclear traffic is regulated not only by recognition of specific signals on proteins and RNAs, but also by cellular factors that modulate the efficacy with which these signals are recognized.

Anchor sequence-dependent endogenous processing of human immunodeficiency virus 1 envelope glycoprotein gp160 for CD4+ T cell recognition.

Human CD4+ T cell clones and cell lines were shown to lyse recombinant vaccinia virus-infected cells that synthesize the HIV-1 envelope glycoprotein gp160. The processing of endogenously synthesized gp160 for recognition by CD4+ T cells required that the protein, after synthesis on the rough endoplasmic reticulum and during subsequent cellular transport, remain attached to the luminal/extracellular membrane face by a hydrophobic anchor sequence.

Endogenously synthesized peptide with an endoplasmic reticulum signal sequence sensitizes antigen processing mutant cells to class I-restricted cell-mediated lysis.

The HLA-A2-positive human mutant cell line T2 is not lysed by influenza virus-specific HLA-A2-restricted cytotoxic lymphocytes after virus infection. However, lysis does occur when cells are incubated with the antigenic influenza matrix protein-derived peptide M57-68. To examine the nature of this defect, T2 cells were transfected with two different plasmids. One plasmid encoded the peptide M57-68, and the other encoded the same peptide preceded by an endoplasmic reticulum translocation signal sequence. Mutant T2 cells expressing the M57-68 peptide without the signal sequence were not susceptible to lysis by M57-68-specific HLA-A2-restricted cytotoxic T lymphocytes, whereas T2 cells expressing the M57-68 peptide plus signal sequence were lysed effectively. Lysis of parental T1 cells with either plasmid was equally effective. These results suggest that the T2 mutant cells are defective in the transport of antigenic peptides from the cytosol into the secretory pathway.

In vitro translation and assembly of a complete T cell receptor-CD3 complex.

The T cell receptor for antigen (TCR) is a multisubunit complex that consists of at least seven polypeptides: the clonotypic, disulfide-linked alpha/beta heterodimer that is noncovalently associated with the invariant polypeptides of the CD3 complex (CD3-gamma, -delta, -epsilon) and zeta, a disulfide-linked homodimer. We achieved the complete assembly of the human TCR in an in vitro transcription/translation system supplemented with dog pancreas microsomes by simultaneous translation of the messenger RNAs encoding the TCR-alpha, -beta and CD3-gamma, -delta, -epsilon, and -zeta subunits. CD3-epsilon, one of the subunits that initiates the assembly of the TCR in living cells, forms misfolded, disulfide-linked homooligomers when translated alone. However, co-translation of one of its first binding partners in the course of assembly, CD3-gamma or -delta, led to the expression of mainly monomeric and correctly folded epsilon subunits, the only form we could detect as part of a properly assembled TCR complex. In the absence of these subunits, the ER-resident chaperone calnexin interacted with oligomeric, i.e. misfolded, structures of CD3-epsilon in a glycan-independent manner. A glycan-dependent interaction between CD3-epsilon and calnexin was mediated by CD3-gamma and concerned only monomeric CD3-epsilon complexed with CD3-gamma, but was dispensable for proper folding of CD3-epsilon. We suggest that in addition to its signaling function, CD3-epsilon serves as a monitor for proper subunit assembly of the TCR.

Qa-1b binds conserved class I leader peptides derived from several mammalian species.

Qa-1b binds a peptide (AMAPRTLLL), referred to as Qdm (for Qa-1 determinant modifier), derived from the signal sequence of murine class Ia molecules. This peptide binds with high affinity and accounts for almost all of the peptides associated with this molecule. Human histocompatibility leukocyte antigen (HLA)-E, a homologue of Qa-1b, binds similar peptides derived from human class Ia molecules and interacts with CD94/NKG2 receptors on natural killer cells. We used surface plasmon resonance to determine the ability of Qa-1b to bind related ligands representing peptides derived from the leaders of class I molecules from several mammalian species. All of the peptides reported to bind HLA-E bound readily to Qa-1b. In addition, peptides derived from leader segments of different mammals also bound to Qa-1b, indicating a conservation of this "Qdm-like" epitope throughout mammalian evolution. We have attempted to define a minimal peptide on a polyglycine backbone that binds Qa-1b. Our previous findings showed that P2 and P9 are important but not sufficient for binding to Qa-1b. Although a minimum peptide (GMGGGGLLL) bound Qa-1(b), its interaction was relatively weak, as were peptides sharing five or six residues with Qdm, indicating that multiple native residues are required for a strong interaction. This finding is consistent with the observation that this molecule preferentially binds this single ligand.

Downregulation of peptide transporter genes in cell lines transformed with the highly oncogenic adenovirus 12.

The expression of class I major histocompatibility complex antigens on the surface of cells transformed by adenovirus 12 (Ad12) is generally very low, and correlates with the high oncogenicity of this virus. In primary embryonal fibroblasts from transgenic mice that express both endogenous H-2 genes and a miniature swine class I gene (PD1), Ad12-mediated transformation results in suppression of cell surface expression of all class I antigens. Although class I mRNA levels of PD1 and H-2Db are similar to those in nonvirally transformed cells, recognition of newly synthesized class I molecules by a panel of monoclonal antibodies is impaired, presumably as a result of inefficient assembly and transport of the class I molecules. Class I expression can be partially induced by culturing cells at 26 degrees C, or by coculture of cells with class I binding peptides at 37 degrees C. Analysis of steady state mRNA levels of the TAP1 and TAP2 transporter genes for Ad12-transformed cell lines revealed that they both are significantly reduced, TAP2 by about 100-fold and TAP1 by 5-10-fold. Reconstitution of PD1 and H-2Db, but not H-2Kb, expression is achieved in an Ad12-transformed cell line by stable transfection with a TAP2, but not a TAP1, expression construct. From these data it may be concluded that suppressed expression of peptide transporter genes, especially TAP2, in Ad12-transformed cells inhibits cell surface expression of class I molecules. The failure to fully reconstitute H-2Db and H-2Kb expression indicates that additional factors are involved in controlling class I gene expression in Ad12-transformed cells. Nevertheless, these results suggest that suppression of peptide transporter genes might be an important mechanism whereby virus-transformed cells escape immune recognition in vivo.

Chloroplast transit peptides: structure, function and evolution.

It is thought that two to three thousand different proteins are targeted to the chloroplast, and the 'transit peptides' that act as chloroplast targeting sequences are probably the largest class of targeting sequences in plants. At a primary structural level, transit peptide sequences are highly divergent in length, composition and organization. An emerging concept suggests that transit peptides contain multiple domains that provide either distinct or overlapping functions. These functions include direct interaction with envelope lipids, chloroplast receptors and the stromal processing peptidase. The genomic organization of transit peptides suggests that these domains might have originated from distinct exons, which were shuffled and streamlined throughout evolution to yield a modern, multifunctional transit peptide. Although still poorly characterized, this evolutionary process could yield transit peptides with different domain organizations. The plasticity of transit peptide design is consistent with the diverse biological functions of chloroplast proteins.

The ornithine transcarbamylase leader peptide directs mitochondrial import through both its midportion structure and net positive charge.

The cytoplasmically synthesized precursor of the mitochondrial matrix enzyme, ornithine transcarbamylase (OTC), is targeted to mitochondria by its NH2-terminal leader peptide. We previously established through mutational analysis that the midportion of the OTC leader peptide is functionally required. In this article, we report that study of additional OTC precursors, altered in either a site-directed or random manner, reveals that (a) the midportion, but not the NH2-terminal half, is sufficient by itself to direct import, (b) the functional structure in the midportion is unlikely to be an amphiphilic alpha-helix, (c) the four arginines in the leader peptide contribute collectively to import function by conferring net positive charge, and (d) surprisingly, proteolytic processing of the leader peptide does not require the presence of a specific primary structure at the site of cleavage, in order to produce the mature OTC subunit.

The signal sequence interacts with the methionine-rich domain of the 54-kD protein of signal recognition particle.

The signal sequence of nascent preprolactin interacts with the 54-kD protein of the signal recognition particle (SRP54). To identify the domain or site on SRP54 that interacts with the signal sequence we used a photocross-linking approach followed by limited proteolysis and immunoprecipitation using anti-peptide antibodies specific for defined regions of SRP54. We found that the previously identified methionine-rich RNA-binding domain of SRP54 (SRP54M domain) also interacts with the signal sequence. The smallest fragment that was found to be crosslinked to the signal sequence comprised the COOH-terminal 6-kD segment of the SRP54M domain. No cross-link to the putative GTP-binding domain of SRP54 (SRP54G domain) was found. Proteolytic cleavage between the SRP54M domain and SRP54G domain did not impair the subsequent interaction between the signal sequence and the SRP54M domain. Our results show that both the RNA binding and signal sequence binding functions of SRP54 are performed by the SRP54M domain.

A signal sequence domain essential for processing, but not import, of mitochondrial pre-ornithine carbamyl transferase.

Studies using deletion mutagenesis indicate that a processing recognition site lies proximal to the normal cleavage position between gln32 and ser33 of pre-ornithine carbamyl transferase (pOCT). pOCT cDNA was manipulated to delete codons specifying amino acids 22-30 of the signal sequence. The mutant precursor, designated pOCT delta 22-30, was imported to the matrix compartment by purified mitochondria, but remained largely unprocessed; the low level of processing that was observed did not involve the normal cleavage site. Several manipulations performed downstream of the normal pOCT processing site (deletion, substitution, and hybrid protein constructions) affected neither import nor correct processing. Our data suggest that domains specifying import and processing site recognition may be functionally segregated within the signal peptide; that processing is not requisite for import of pOCT; and that a proximal region, not involving the normal signal peptide cleavage site, is required for processing site recognition.

Cotranslational and posttranslational proteolytic processing of preprosomatostatin-I in intact islet tissue.

Preprosomatostatin-I (PPSS-I) is processed in anglerfish islets to release a 14-residue somatostatin (SS-14). However, very little is known regarding other processing events that affect PPSS-I. This is the first study to identify and quantify the levels of nonsomatostatin products generated as a result of processing of this somatostatin precursor in living islet tissue. The products of PPSS-I processing in anglerfish islet tissue were identified in radiolabeling studies using a number of criteria. These criteria included immunoreactivity, specific radiolabeling by selected amino acids, radiolabel sequencing, and chromatographic comparison to isolated, structurally characterized fragments of anglerfish PPSS-I using reverse-phase high performance liquid chromatography. Intact prosomatostatin-I (aPSS-I) was isolated from tissue incubated with [3H]tryptophan and [14C]leucine. Significant 14C radioactivity was observed in the products of 11 of the first 44 sequencer cycles in positions consistent with the generation of a 96-residue prosomatostatin. These results indicate that signal cleavage occurs after the cysteine located 25 residues from the initiator Met of PPSS-I, resulting in a signal peptide 25 amino acids in length. Nonsomatostatin-containing fragments of the precursor were also found in tissue incubated with a mixture of 3H-amino acids. Only a small quantity of the dodecapeptide representing residues 69-80 in the prohormone was found (10 nmol/g tissue). Two other fragments of aPSS-I, also observed to be present in low abundance, were found to correspond to residues 1-27 (16 nmol/g tissue) and to residues 1-67 (7 nmol/g tissue) of aPSS-I. No evidence for the presence of the fragment corresponding to residues 29-67 was found. However, large quantities of SS-14 were observed (287 nmol/g tissue), indicating that the major site of aPSS-I cleavage is at the basic dipeptide immediately preceding SS-14. Recovery of much lower levels of the nonsomatostatin fragments of aPSS-I suggests that prohormone processing at the secondary sites identified in this study occurs at a low rate relative to release of SS-14 from aPSS-I.

Yeast carboxypeptidase Y can be translocated and glycosylated without its amino-terminal signal sequence.

We have constructed a series of mutations in the signal sequence of the yeast vacuolar protein carboxypeptidase Y (CPY), and have used pulse-chase radiolabeling and immunoprecipitation to examine the in vivo effects of these mutations on the entry of the mutant CPY proteins into the secretory pathway. We find that introduction of a negatively charged residue, aspartate, into the hydrophobic core of the signal sequence has no apparent effect on signal sequence function. In contrast, internal in-frame deletions within the signal sequence cause CPY to be synthesized as unglycosylated precursors. These are slowly and inefficiently converted to glycosylated precursors that are indistinguishable from the glycosylated forms produced from the wild-type gene. These precursors are converted to active CPY in a PEP4-dependent manner, indicating that they are correctly localized to the vacuole. Surprisingly, a deletion mutation that removes the entire CPY signal sequence has a similar effect: unglycosylated precursor accumulates in cells carrying this mutant gene, and greater than 10% of it is posttranslationally glycosylated. Thus, the amino-terminal signal sequence of CPY, while important for translocation efficiency, is not absolutely required for the translocation of this protein.

Differential protein import deficiencies in human peroxisome assembly disorders.

Two peroxisome targeting signals (PTSs) for matrix proteins have been well defined to date. PTS1 comprises a COOH-terminal tripeptide, SKL, and has been found in several matrix proteins, whereas PTS2 has been found only in peroxisomal thiolase and is contained within an NH2-terminal cleavable presequence. We have investigated the functional integrity of the import routes for PTS1 and PTS2 in fibroblasts from patients suffering from peroxisome assembly disorders. Three of the five complementation groups tested showed a general loss of PTS1 and PTS2 import. Two complementation groups showed a differential loss of peroxisomal protein import: group I cells were able to import a PTS1- but not a PTS2- containing reporter protein into their peroxisomes, and group IV cells were able to import the PTS2 but not the PTS1 reporter into aberrant, peroxisomal ghostlike structures. The observation that the PTS2 import pathway is intact only in group IV cells is supported by the protection of endogenous thiolase from protease degradation in group IV cells and its sensitivity in the remaining complementation groups, including the partialized disorder of group I. The functionality of the PTS2 import pathway and colocalization of endogenous thiolase with the peroxisomal membranes in group IV cells was substantiated further using immunofluorescence, subcellular fractionation, and immunoelectron microscopy. The phenotypes of group I and IV cells provide the first evidence for differential import deficiencies in higher eukaryotes. These phenotypes are analogous to those found in Saccharomyces cerevisiae peroxisome assembly mutants.

COOH-terminal signals mediate the trafficking of a peptide processing enzyme in endocrine cells.

Peptidylglycine alpha-amidating monooxygenase (PAM) catalyzes the COOH-terminal amidation of bioactive peptides through a two step reaction catalyzed by separate enzymes contained within the PAM precursor. To characterize the trafficking of integral membrane PAM proteins in neuroendocrine cells, we have generated stable AtT-20 cell lines expressing full length and COOH-terminally truncated integral membrane PAM proteins. Full length integral membrane PAM was present on the cell surface in low but detectable amounts and PAM proteins which reached the cell surface were rapidly internalized but not immediately degraded in lysosomes. Internalized PAM complexed with PAM antibody was found in a subcellular compartment which overlapped with internalized transferrin and with structures binding WGA. Thus the punctate juxtanuclear staining of full length PAM represents PAM in endosomes. Endoproteolytic processing of full length PAM-1 and PAM-2 resulted in the secretion of soluble PAM proteins; the secretion of these soluble PAM proteins was stimulus dependent. Although some of the truncated PAM protein was also processed and stored in AtT-20 cells, much of the expressed protein was redistributed to the plasma membrane. Soluble proteins not observed in large amounts in cells expressing full length PAM were released from the surface of cells expressing truncated PAM and little internalization of truncated integral membrane PAM was observed. Thus, the COOH-terminal domain of PAM contains information important for its trafficking within the regulated secretory pathway as well as information necessary for its retrieval from the cell surface.

Interaction between BiP and Sec63p is required for the completion of protein translocation into the ER of Saccharomyces cerevisiae.

To clarify the roles of Kar2p (BiP) and Sec63p in translocation across the ER membrane in Saccharomyces cerevisiae, we have utilized mutant alleles of the essential genes that encode these proteins: kar2-203 and sec63-1. Sanders et al. (Sanders, S. L., K. M. Whitfield, J. P. Vogel, M. D. Rose, and R. W. Schekman. 1992. Cell. 69:353-365) showed that the translocation defect of the kar2-203 mutant lies in the inability of the precursor protein to complete its transit across the membrane, suggesting that the lumenal hsp70 homologue Kar2p (BiP) binds the transiting polypeptide in order to facilitate its passage through the pore. We now show that mutation of a conserved residue (A181-->T) (Nelson, M. K., T. Kurihara, and P. Silver. 1993. Genetics. 134:159-173) in the lumenal DnaJ box of Sec63p (sec63-1) results in an in vitro phenotype that mimics the precursor stalling defect of kar2-203. We demonstrate by several criteria that this phenotype results specifically from a defect in the lumenal interaction between Sec63p and BiP: Neither a sec62-1 mutant nor a mutation in the cytosolically exposed domain of Sec63p causes precursor stalling, and interaction of the sec63-1 mutant with the membranebound components of the translocation apparatus is unimpaired. Additionally, dominant KAR2 suppressors of sec63-1 partially relieve the stalling defect. Thus, proper interaction between BiP and Sec63p is necessary to allow the precursor polypeptide to complete its transit across the membrane.

Posttranslational processing of the prohormone-cleaving Kex2 protease in the Saccharomyces cerevisiae secretory pathway.

The Kex2 protease of the yeast Saccharomyces cerevisiae is a prototypical eukaryotic prohormone-processing enzyme that cleaves precursors of secreted peptides at pairs of basic residues. Here we have established the pathway of posttranslational modification of Kex2 protein using immunoprecipitation of the biosynthetically pulse-labeled protein from a variety of wild-type and mutant yeast strains as the principal methodology. Kex2 protein is initially synthesized as a prepro-enzyme that undergoes cotranslational signal peptide cleavage and addition of Asn-linked core oligosaccharide and Ser/Thr-linked mannose in the ER. The earliest detectable species, I1 (approximately 129 kD), undergoes rapid amino-terminal proteolytic removal of a approximately 9-kD pro-segment yielding species I2 (approximately 120 kD) before arrival at the Golgi complex. Transport to the Golgi complex is marked by extensive elaboration of Ser/Thr-linked chains and minor modification of Asn-linked oligosaccharide. During the latter phase of its lifetime, Kex2 protein undergoes a gradual increase in apparent molecular weight. This final modification serves as a marker for association of Kex2 protease with a late compartment of the yeast Golgi complex in which it is concentrated about 27-fold relative to other secretory proteins.

Fusion of sequence elements from non-anchored proteins to generate a fully functional signal for glycophosphatidylinositol membrane anchor attachment.

Glycophosphatidylinositol (GPI) membrane anchor attachment is directed by a cleavable signal at the COOH terminus of the protein. The complete lack of homology among different GPI-anchored proteins suggests that this signal is of a general nature. Previous analysis of the GPI signal of decay accelerating factor (DAF) suggests that the minimal requirements for GPI attachment are (a) a hydrophobic domain and (b) a cleavage/attachment site consisting of a pair of small residues positioned 10-12 residues NH2-terminal to a hydrophobic domain. As an ultimate test of these rules we constructed four synthetic GPI signals, meeting these requirements but assembled entirely from sequence elements not normally involved in GPI attachment. We show that these synthetic signals are able to direct human growth hormone (hGH), a secreted protein, to the plasma membrane via a GPI anchor. Our results indicate that different hydrophobic sequences, derived from either the prolactin or hGH NH2-terminal signal peptide, can be linked to different cleavage sites via different hydrophilic spacers to produce a functional GPI signal. These data confirm that the only requirements for GPI-anchoring are a pair of small residues positioned 10-12 residues NH2 terminal to a hydrophobic domain, no other structural motifs being necessary.

Import of rat ornithine transcarbamylase precursor into mitochondria: two-step processing of the leader peptide.

The mitochondrial matrix enzyme ornithine transcarbamylase (OTC) is synthesized on cytoplasmic polyribosomes as a precursor (pOTC) with an NH2-terminal extension of 32 amino acids. We report here that rat pOTC synthesized in vitro is internalized and cleaved by isolated rat liver mitochondria in two, temporally separate steps. In the first step, which is dependent upon an intact mitochondrial membrane potential, pOTC is translocated into mitochondria and cleaved by a matrix protease to a product designated iOTC, intermediate in size between pOTC and mature OTC. This product is in a trypsin-protected mitochondrial location. The same intermediate-sized OTC is produced in vivo in frog oocytes injected with in vitro-synthesized pOTC. The proteolytic processing of pOTC to iOTC involves the removal of 24 amino acids from the NH2 terminus of the precursor and utilizes a cleavage site two residues away from a critical arginine residue at position 23. In a second cleavage step, also catalyzed by a matrix protease, iOTC is converted to mature OTC by removal of the remaining eight residues of leader sequence. To define the critical regions in the OTC leader peptide required for these events, we have synthesized OTC precursors with alterations in the leader. Substitution of either an acidic (aspartate) or a "helix-breaking" (glycine) amino acid residue for arginine 23 of the leader inhibits formation of both iOTC and OTC, without affecting translocation. These mutant precursors are cleaved at an otherwise cryptic cleavage site between residues 16 and 17 of the leader. Interestingly, this cleavage occurs at a site two residues away from an arginine at position 15. The data indicate that conversion of pOTC to mature OTC proceeds via the formation of a third discrete species: an intermediate-sized OTC. The data suggest further that, in the rat pOTC leader, the essential elements required for translocation differ from those necessary for correct cleavage to either iOTC or mature OTC.

Direct probing of the interaction between the signal sequence of nascent preprolactin and the signal recognition particle by specific cross-linking.

We have studied the interaction between the signal sequence of nascent preprolactin and the signal recognition particle (SRP) during the initial events in protein translocation across the endoplasmic reticulum membrane. A new method of affinity labeling was used, whereby lysine residues, carrying the photoreactive group 4-(3-trifluoromethyldiazirino) benzoic acid in their side chains, are incorporated into a protein by means of modified lysyl-tRNA, and cross-linking to the interacting component is induced by irradiation. SRP interacts through its Mr 54,000 polypeptide component with the signal sequences of nascent preprolactin chains containing about 70 residues, and with decreasing affinity with longer chains as well; it causes inhibition of elongation. Binding of SRP is reversible and requires the nascent chain to be bound to a functional ribosome. SRP cross-linked to the signal sequence still inhibits elongation but does not prevent it completely. We conclude that SRP does not block the exit site of the polypeptide chain on the ribosome. The SRP receptor of the endoplasmic reticulum membrane displaces the signal sequence from SRP and, even if SRP is cross-linked, releases elongation arrest.