• We are sorry, but NCBI web applications do not support your browser and may not function properly. More information
Logo of pnasPNASInfo for AuthorsSubscriptionsAboutThis Article
Proc Natl Acad Sci U S A. May 17, 2005; 102(20): 7121–7126.
Published online May 10, 2005. doi:  10.1073/pnas.0502669102
PMCID: PMC1129144

Studies of yeast oligosaccharyl transferase subunits using the split-ubiquitin system: Topological features and in vivo interactions


Oligosaccharyl transferase (OT) catalyzes the cotranslational N-glycosylation of nascent polypeptides in the endoplasmic reticulum in all eukaryotic systems. Due to the inherent difficulty in characterizing this membrane protein complex, the mode of enzymatic action has not been resolved. Here, we used a membrane protein two-hybrid approach, the split-ubiquitin system, to address two aspects of the enzyme complex in yeast: the topological features, as well as the in vivo interactions of all of the components. We investigated the N- and C-terminal orientation of these proteins and the presence or the absence of a cleavable signal sequence at their N termini. We found that Ost2p and Stt3p have only their N terminus located in the cytosol, whereas Ost3p and Swp1p have only their C terminus oriented in the cytosol. In the case of Ost5p and Ost6p, both their N and C termini are present in the cytosol. These findings also suggested that Ost2p, Stt3p, Ost5p, and Ost6p do not have a cleavable N-terminal signal sequence. The pairwise analysis of in vivo interactions among all of the OT subunits demonstrated that OT subunits display specific interactions with each other in a functional complex. By comparing this interaction pattern with that detected in vitro in a nonfunctional complex, we proposed that a distinct conformation rearrangement takes place when the enzyme complex changes from the nonfunctional state to the activated functional state. This finding is consistent with earlier work by others indicating that OT exhibits allosteric properties.

Keywords: conformation rearrangement, membrane topology, signal sequence

Oligosaccharyl transferase (OT) is an enzyme complex that catalyzes the transfer of an oligosaccharyl moiety from a dolichol-linked pyrophosphate to the side chain of asparagine as the nascent polypeptides are translocated into the endoplasmic reticulum (ER). It has been shown in both yeast and higher eukaryotic organisms that a large number of protein subunits are involved in the catalytic action. These components assemble together to form a heteromeric, multisubunit complex in the ER membrane. In the yeast Saccharomyces cerevisiae, at least nine genes encoding OT subunit proteins have been cloned and identified: OST1, OST2, OST3, OST4, OST5, OST6, WBP1, SWP1, and STT3 (for reviews, see refs. 1-4).

Although OT plays an essential role in an early step of the secretory pathway, its catalytic mechanism and the role of each of these gene products remains unknown, partially because of the inherent difficulty in isolating bioactive membrane protein complexes and the fact that catalytic N-glycosylation takes place in a highly dynamic manner during the protein translation and translocation. Thus far, only limited approaches have been used to study this unusual enzyme complex. The yeast two-hybrid assay is not applicable to membrane proteins because it is required that the proteins or the domains of the proteins of interest be soluble and express well in the yeast nucleus (5). Indeed, a two-hybrid screening to investigate the interaction partners of yeast OT subunits using the luminal domains of these proteins as baits has failed to reveal any of the interactions among them (6). Recently, a modified two-hybrid approach that specifically applies to membrane proteins, the split-ubiquitin system, has emerged as a powerful tool to detect protein-protein interaction in live cells (7-9). The system utilizes complementation between separable domains of ubiquitin: the N terminus of ubiquitin (Nub, amino acids 1-34) and the C terminus of ubiquitin (Cub, amino acids 35-76), which is followed by a reporter protein (Rep) (7). Wild-type Nub (NubI, with I being isoleucine at position 13) spontaneously assembles with Cub-Rep, resulting in proteolytic cleavage at the C terminus of Cub by a ubiquitin-specific protease(s) and subsequent release of the reporter fragment. However a mutant of Nub (NubG) in which Ile-13 is changed to Gly-13 is unable to assemble with Cub-Rep unless two proteins X and Y that interact with each other are fused to the Cub-R and to the NubG so that this interaction can force the reassociation of the two halves of ubiquitin. As a result, the interaction between two proteins X and Y can be monitored by the cleavage and subsequent activation of the reporter genes. Appealing features of this approach are the diversity of the reporter format to be selected and the ability to detect transient interactions that take place in a dynamic pathway. Indeed, this system has been used to examine a wide spectrum of protein-protein interactions in a variety of organisms, including charactering the interacting partners in sucrose transporters, ion channels, and G protein-coupled receptor signaling pathway, as well as transcriptional regulation processes in plants, yeast, and mammals (10-13).

In this article, we used this in vivo approach in the study of yeast OT subunits, i.e., the topology features and the interactions among them in a functional complex. In a previous exploration to study the structural organization of the yeast OT complex, we carried out a complementary in vitro cross-linking study on all of the subunits in microsomes (14), and, from this study, a model to demonstrate the specific interactions of OT subunit proteins was developed (4). Because the biochemical cross-linking study was carried out in a system of isolated microsomal membranes without the presence of acceptor nascent polypeptides, we speculate that the model probably merely represents the organization of the OT complex in a nonfunctional mode. Here, we undertake an in vivo interaction study on OT and ask whether there is any variation of the interaction pattern compared with that in the nonworking state. This alteration probably represents a conformational rearrangement after the enzyme binds one or both of the substrates, and therefore would be expected to provide useful information on the catalytic mechanism of OT in the long term.

To detect protein interactions by means of the split-ubiquitin system, the protein of interest that is fused to Cub-Rep must be membrane-associated to prevent the reporter protein from entering the nucleus and causing self activation. In a previous split-ubiquitin study, Wbp1p was used as the protein to be attached by Cub-Rep at its C terminus, and a strain YG0673, in which the WBP1 allele was replaced by the nucleotides encoding Wbp1-Cub-PLV, was generated [where PLV is a format of reporter protein composed of protein A and transcriptional activation domain of the bacterial LexA and the herpes simplex VP16 (8)]. Upon cleavage, PLV is liberated from Cub and transported into nucleus to activate LacZ and HIS3 reporter genes, which, when expressed, generate a blue color in the presence of X-Gal and allow growth on medium lacking the amino acid histidine. Because our previous chemical cross-linking study showed that Wbp1p could be cross-linked to most of the other OT subunits (14), we chose to initiate our study by using Wbp1p as the protein to be fused with Cub-Rep and other OT components to be fused with Nub.

A second prerequisite of the split-ubiquitin system is that Nub and Cub moieties must be attached to parts of the membrane protein that localize to the cytosol because the necessary ubiquitin-specific protease(s) is present in the cytosol and not in the ER lumen (8). This fact promoted us to ask whether the split-ubiquitin assay can also be used to characterize the topology features of the membrane proteins. As a matter of fact, in the case of yeast OT subunits [except for the three single transmembrane proteins Ost1p (15), Wbp1p (16) and Ost4p (17)], the definitive membrane topology of the others (Ost2p, Ost3p, Ost5p, Ost6p, Stt3p, and Swp1p) has not been described (see Note). To investigate the orientation of the N and C termini of these proteins, a strategy was developed that two fusion forms of a single protein (termed X, where X represents Ost2p, Ost3p, Ost5p, Ost6p, Stt3p, or Swp1p) were prepared: NubG-Xp (where NubG is fused to the N terminus of the protein) and X-NubGp (where the Nub part is fused to the C terminus of X). We then coexpressed either one of them with Wbp1-Cub-PLVp in the parental strain of YG0673 to investigate which form of the NubG fusion is capable of reassociation with the Cub moiety. This reassociation, in turn, will indicate which end of the protein is located in the cytosol. As illustrated in Fig. 1, if the C terminus of protein X is oriented toward the cytoplasm and it interacts with Wbp1p, cleavage of the reporter gene can be detected only when the Wbp1-Cub-PLVp is coexpressed with X-NubGp (Fig. 1 A) but not with NubG-Xp (Fig. 1B). If neither form of the NubG fusion protein displays interaction with Wbp1-Cub-PLVp, another bait protein fused by Cub-PLV can be used for the analysis, given that the C terminus of this protein has been shown to be located in the cytosol. It is noteworthy that the result of split-ubiquitin assay using NubG-Xp to coexpress with Wbp1-Cub-PLVp also provides potential information on another important topology feature of membrane protein X: the presence or absence of an N-terminal signal sequence. Because membrane proteins are translocated through the signal sequence-dependent pathway (19-21), if the NubG-X fusion protein can be successfully incorporated into the OT complex and interact with Wbp1-Cub-PLVp, the N-terminal Nub moiety followed by an N-terminal signal sequence must not have been cleaved.

Fig. 1.
Using the split-ubiquitin system to investigate which terminus of protein X is located in the cytosol (X represents Ost2p, Ost3p, Ost5p, Ost6p, Stt3p, or Swp1p). Two fusion forms of protein X, X-NubGp (A) and NubG-Xp (B), were generated and transformed ...

Using this approach, we established that Ost2p and Stt3p have only their N terminus in the cytosol, whereas Ost3p and Swp1p have only their C terminus orientated in the cytosol. In the case of Ost5p and Ost6p, both of their N and C termini are present in the cytosol. We also found that Ost2p, Stt3p, Ost5p, and Ost6p do not have a cleavable signal sequence at their N terminus. The pairwise in vivo interaction analysis on all of the OT subunits demonstrated that OT subunit proteins display specific interactions with each other in a working complex. A model to describe their interrelationship in live cells was developed. By comparing this model with that developed based on the cross-linking study of OT in a nonfunctioning mode in microsomes, we found that a structural rearrangement does indeed take place after substrate(s) binding, and the change from the nonworking state to the working state results in variations in interaction patterns among several subunits.

Materials and Methods

Construction of Plasmids. All plasmids were constructed by the strategy of homologous recombination. Plasmids pRS314 (NubG-AlG5), pRS314 (OST1-NubG), and pRS305 (Δwbp1-Cub-PLV) are parental plasmids (8); the plasmids and strains used in this study are summarized in Table 2, which is published as supporting information on the PNAS web site.

Construction of pRS314 (NubG-X) series. pRS314 (NubG-AlG5) was treated with BglII to be linearized. The 5′ end homologous recombination region was selected to complement the sequence on the vector that precedes the ATG codon of ALG5 gene, and the 3′ end of the homologous region was chosen to complement the sequences on the vector that follows the stop codon of ALG5 gene. Each homologous region contained the following 50 nucleotides: 5′-CGACAACGTTAAGTCGAAAATTCAAGACAAGGAAGGGATCCCTGGTGGGT-3′ and 5′-GCTTATGTAATAATAATAATAATGATAATAATAACATAAAAATAATTACT-3′. A gene of interest was amplified by PCR using genomic DNA as a template and two primers, a 5′ end primer complementing 18 nucleotides following the start codon of the gene with the homologous region sequence and a 3′ end primer complementing 18 nucleotides preceding the stop codon of the gene with the homologous segment of that region. Genomic DNA was isolated from W303-1a strain (MATa ade2 can1 his3 leu2 trp1 ura3). A yeast strain L40 (MATa trp1 leu2 his3 LYS2::lexA-HIS3 URA3::lexA-lacZ) was transformed with 100 ng of linearized pRS314 (NubG-AlG5) vector and 300 ng of PCR product carrying the gene of interest flanked by the homologous region sequences. Transformants were selected on SD-Trp. Plasmids were isolated by using an RPM yeast plasmid isolation kit (Q-Biogene, Irvine, CA) and were transformed into Escherichia coli JBE181 [ΔlacX74 hsr- rpsL pyrF::TN5(kan) leuB600 trpC 9830 gal E galK] cells by means of electroporation. Plasmids were subsequently isolated by using a High Pure plasmid isolation kit (Roche,Gipf-Oberfrick, Switzerland) and verified by PCR analysis and DNA sequencing.

Construction of pRS314 (X-NubG) series. The method was the same as described above. pRS314 (OST1-NubG) was selected as the parental vector, linearized by enzyme digestion of EcoRI. The 50-nucleotide homologous sequences were as follows: 5′-CTTCAGCCAACTTGGAGACGAATCTAGCTTTGACGATAACTGGAACATTTG-3′ and 5′-CCGGTCAAAGTCT TGACGAAAATCTGCATGGTCGACCCACCAGACTCGAG-3′.

Construction of pRS315 (OT-Cub-PLV). The method was the same as described above. The parental vector pRS315 was linearized by means of the enzyme digestion of XbaI. The homologous sequence of 50 nucleotides was as follows: 5′-GTCGACGGTATCGATA AGCT TGATATCGA AT TCCTGCAGCCCGGGGGATC-3′ and 5′-ATTGTAATACGACTCACTATAGGGCGAATTGGAGCTCCACCGCGGTGGCG-3′. The DNA fragment encoding Cub-ProteinA-LexA-VP16 was amplified by using the plasmid pRS305 (Δwbp1-Cub-PLV) as template and two primers: the 5′ end primer complementing the linker between the WBP1 gene and the start codon of the Cub gene with the homologous sequence region and a 3′ end primer complementing 18 nucleotides of VP16 gene preceding the stop codon with the homologous sequence region. The resulting plasmid served as the parental plasmid to construct a series of pRS315 (OT-Cub-PLV) constructs. It was linearized by means of the enzyme digestion of HindIII. The homologous sequences of 50 nucleotides used to construct various Cub-PLV fused OT subunit genes were as follows: 5′CTAAAGGGAACAAAAGCTGGGTACCGGGCCCCCCCTCGAGGTCGACGGTAT-3′ and 5′CT TACCGGCA A AGATCA ATCT T TGT TGATCTGGAGGGATCCCCCCCGACAT.

β-Gal Activity Assay. Transformants carrying two series of plasmids, pRS314 (X-NubG) or pRS314 (NubG-X) and pRS315 (OT subunit-Cub-PLV), were grown for 2 days at 30°C on dropout agar plates lacking tryptophan and leucine and were transferred to sterile Whatman filter paper. The cells were permeabilized by dipping the filters into liquid nitrogen for 30 seconds. After thawing, the filters were overlaid with a piece of Whatman filter paper soaked by 3 ml of Z buffer (60 mM Na2HPO4/40 mM NaH2PO4/10 mM KCl/1 mM MgSO4) containing 9 μl of 2-mercaptoethanol and 0.3 mg/ml X-Gal, and incubated at 30°C for 1-23 h.

Growth Assay. Equal numbers of cells were collected after the strains were grown to early log phase in liquid media lacking tryptophan and leucine. Seven microliters of 1:10 serial dilutions of the cells was spotted on SD plates lacking tryptophan and leucine as well as SD plates lacking tryptophan, leucine, and histidine and incubated at 30°C for 2 days.


Because it had been shown that the proteins being investigated could be cross-linked to Wbp1p (14), we first chose Wbp1-Cub-PLV as the bait and investigated its interaction with X-NubG and with NubG-X.

Interaction of NubG-X with Wbp1-Cub-PLV. A series of plasmids in which the Nub moiety was fused to the N termini of X (i.e., X represents Ost2p, Ost3p, Ost5p, Ost6p, Swp1p or Stt3p) were prepared. They were subsequently transformed into the YG0673 strain in which Wbp1-Cub-PLVp was expressed. The resulting transformants were analyzed for β-gal activity as well as for growth on plates lacking histidine (-His plates). As shown in Fig. 2, among the transformants examined, coexpression of Wbp1-Cub-PLVp with NubG-Ost2p, NubG-Ost6p, and NubG-Stt3p resulted in positive β-gal activity (Fig. 2 A) and growth on -His plates (Fig. 2B), indicating that the N termini of Ost2p, Ost6p, and Stt3p are oriented in the cytosol. This observation also indicated that the NubG-fused membrane proteins were successfully incorporated into the OT complex, suggesting that these three proteins do not have a cleavable signal sequence at their N terminus, because, if they had contained a cleavable signal sequence at their N termini, the Nub moiety would be cleaved off along with the signal sequence when the fusion protein was translocated before its incorporation into the OT complex.

Fig. 2.
Split-ubiquitin analysis between Wbp1-Cub-PLVp and NubG-Xp (X represents Ost2p, Ost3p, Ost5p, Ost6p, Stt3p, or Swp1p). (A) β-Gal activity of the transformants expressing Wbp1-Cub-PLVp together with the NubG-Xp fusion proteins. Cells were grown ...

On the other hand, coexpression of Wbp1-Cub-PLVp with NubG-Ost3p, NubG-Ost5p and NubG-Swp1p yielded transformants that showed no β-gal activity (Fig. 2 A) and no growth on -His plates (Fig. 2B). This result, however, was not sufficient to indicate that the N termini of these proteins face the ER lumen rather than the cytoplasm, because there are other possibilities that could yield negative results. One possibility is that these proteins do not associate with Wbp1-Cub-PLVp in vivo and, as a consequence, there is no binding force allowing the reassociation of the two halves of ubiquitin. This probability is not unlikely, because two proteins shown to be within 12 Å of close proximity detected by in vitro cross-linking experiments might not necessarily interact with each other in a working complex in vivo, especially under circumstances in which a conformational change may take place when the enzyme binds one or both of the substrates. The other possibility is that, if the protein being tested has a cleavable signal sequence at its N terminus, the Nub moiety could be cleaved off along with the N-terminal signal sequence before its incorporation into the OT complex. Under this condition, the split-ubiquitin approach is unable to yield a definitive result on the N-terminal orientation of a membrane protein. To investigate whether the negative results with these three proteins are due to the lack of in vivo interactions of them with Wbp1p, we carried out the split-ubiquitin analysis between X-NubGp and Wbp1-Cub-PLVp.

Interaction of X-NubG with Wbp1-Cub-PLV. We prepared another class of plasmids in which the NubG was fused to the C terminus of these proteins, and transformed them individually into the YG0673 strain. As shown in Fig. 3, coexpression of Wbp1-Cub-PLVp with Ost3-Nubp, Ost6-Nubp, and Swp1-Nubp resulted in positive β-gal activity (Fig. 3A) and growth on -His plates (Fig. 3B), suggesting that the C termini of all three of these proteins are located in the cytosol. In the case of Ost3p and Swp1p, these results ruled out the possibility that the absence of β-gal activity resulting from the coexpression of NubG-Ost3p or NubG-Swp1p with Wbp1-Cub-PLVp was due to a lack of in vivo interaction between these proteins with Wbp1p. Rather, it was either because the N termini of these proteins are located in the ER lumen or probably the Nub part was cleaved along with the signal sequences when the protein was translocated into the ER. On the other hand, in the case of Ost2p and Stt3p, because only coexpression of NubG-Ost2p or NubG-Stt3p, but not the Ost2-NubGp or Stt3-NubGp with Wbp1-Cub-PLpV, led to positive β-gal activity and growth capability on -His, these two proteins must have their N termini located in the cytosol and their C termini located in the ER lumen. In the case of Ost6p, the findings in Figs. Figs.22 and and33 indicate that it contains an even number of transmembrane segments and that both its N and C termini face the cytosol.

Fig. 3.
Split-ubiquitin analysis between Wbp1-Cub-PLVp and X-NubGp (X represents Ost2p, Ost3p, Ost5p, Ost6p, Stt3p, or Swp1p). Shown are β-Gal activity (A) and growth of cells coexpressing Wbp1-Cub-PLVp together with X-NubGp fusion proteins on agar plates ...

The N-Terminal and C-Terminal Orientation of Ost5p. It is noteworthy that, in the case of Ost5p, neither coexpression of NubG-Ost5p with Wbp1-Cub-PLVp nor Ost5-NubGp with Wbp1-Cub-PLV resulted in positive β-gal activity. To test whether this result is because both the N and C termini of Ost5p are located in the ER lumen, or because Ost5p does not interact with Wbp1p in vivo, we chose another protein, Ost1p, as the bait protein to be fused by Cub-PLV. We cotransformed the plasmid expressing Ost1-Cub-PLVp with the plasmid expressing NubG-Ost5p or the plasmid expressing Ost5-NubGp into L40 strain, and examined the β-gal activity of the transformants. As shown in Fig. 4, the positive β-gal activity (Fig. 4A) and positive growth behavior on -His plates (Fig. 4B) of both two transformants indicate that, as in the case of Ost6p, both the N-terminal and C-terminal of Ost5p are oriented into the cytosol with an even number of transmembrane segments. It is also revealed that Ost5p does not contain a cleavable signal sequence at its N terminus. Thus, the negative β-gal activity observed with Ost5p and Wbp1p discussed above is due to the absence of a direct interaction between these two proteins in vivo.

Fig. 4.
Split-ubiquitin analysis between Ost1-Cub-PLVp and NubG-Ost5p or Ost5-NubGp. (A) β-Gal activity of cells coexpressing Ost1-Cub-PLVp together with NubG-Ost5p or Ost5-NubGp fusion proteins, respectively. (B) Growth of the cells coexpressing Ost1-Cub-PLVp ...

Mapping the in Vivo Interaction of OT Subunits. Having determined which parts of the OT subunit proteins are located in the cytosol, we subsequently prepared a collection of Cub-PLV and NubG fused constructs as indicated in Table 1 and carried out a pairwise in vivo interaction analysis for almost all of them. The only exception was the pair of Ost2p and Stt3p, because both of them have only their N termini located in the cytosol, whereas the split-ubiquitin assay requires that at least one of a pair of proteins being studied has its C terminus located in the cytosol so that the moiety of Cub-PLV can be accessed by ubiquitin-specific protease(s). As can be seen from Table 1, each of the Cub-PLV fusion proteins displays distinct interactions with Nub fused other OT subunits, indicating they are functionally incorporated into the OT complex. It also suggests that the N and C termini orientation of these proteins determined earlier is correct. Table 1 also indicates that, although an unusually large number of subunits are involved in the catalytic action of OT, each of them displays specific interactions with other components in a working enzyme complex. Consistent with the previous cross-linking study, the essential gene products Ost1p, Wbp1p, and Stt3p display extensive contacts with other components, indicating their essential roles in the catalytic event. The nonessential gene products Ost3p, Ost5p, and Ost6p were found to associate with only a limited number of other components. Notably, Ost2p and Ost4p, which showed limited interactions with other OT subunits when probed in vitro by using a cross-linking reagent with a 12-Å spacer arm, displayed contacts with most of the other components in this study. The recently established 3D structure of Ost4p might help to interpret its extensive contacts with other OT subunits (22). NMR structural analysis revealed that this unusually small protein (36 aa) folds into a well formed, kinked α-helix. The kink is formed between residues Phe-14 and Gly-15 with an angle of 37°, and the residue of Gly-15 that induces this kink structure is strictly conserved through Ost4p homologues in all eukaryotes. Kinks, which represent changes in direction of the helical axis, have been proven a functionally important feature of many TM helices (23, 24). An example is the γ subunit of SecY complex; it plays a critical role in clamping the two halves of the α-subunit together by means of its contacts with TM1 and TM5 in one half (composed of TM1 to -5) and contacts with TM6 and TM10 in the other half (composed of TM6 to -10) (25). Thus, the typical kink structure of Ost4p is likely to be used to apply extensive contacts with other OT subunit proteins in a working complex.

Table 1.
In vivo interactions detected between OT subunits by the split-ubiquitin system


In this study, we determined the complementary in vivo interactions of OT subunits in a functional complex in live cells. A model was developed to illustrate these interactions (Fig. 5B). We also compared it with the one developed based on the in vitro cross-linking study. As can be seen from Table 1 and Fig. 5, substantial changes on the interaction pattern were observed on Ost2p and Ost4p (Fig. 5B), the extensively membrane-embedded proteins, whereas other components displayed only minor alterations when compared with the interaction pattern detected in vitro by the cross-linker molecule with a spacer arm of 12 Å (Fig. 5A). Ost3p and Ost6p display exactly the same interaction pattern with all other components, and this pattern is virtually unchanged in both models. It seems that the plane formed by Ost2p and Ost4p bends toward the center of the complex, resulting in a deviation of Ost1p from the core position and from Stt3p. Meanwhile, Ost5p moves toward Stt3p, and, as a result, it is close to Ost2p, Ost4p, and Stt3p. Consequently, Ost5p loses contact with Wbp1p. This conformational change is strongly supported by previous kinetic data on yeast OT, which showed substantial deviations from a typical Michaelis-Menten equation, and it was proposed to be due to the involvement of allosteric regulation at the binding sites of donor substrate (26).

Fig. 5.
Comparison of the structural organization of OT complex in a nonworking state (in vitro)(A) and a working state (in vivo)(B). The nonworking model was developed based on previous cross-linking studies in isolated microsomal membranes, and the working ...

In addition to the kinetic and structural studies, there is genetic evidence supporting our proposal. An STT3 mutation strain, stt3-3, is specifically defective in the transfer of incomplete lipid-linked oligosaccharides to the peptide substrate, whereas the transfer of full-length substrate is hardly affected, suggesting that Stt3p might function in discriminating the full-length oligosaccharide chain from various intermediate species (27). The OST5 gene was isolated in a synthetic lethal screen in combination with the alg5 deletion, which accumulates nonglucosylated lipid-linked oligosaccharide; therefore, Ost5p is highly likely to be involved in regulation of donor substrate specificity (28). Furthermore, although neither Δost5 nor stt3-3 causes any growth phenotype, a combination of Δost5 and stt3-3 is lethal under various of conditions (28), suggesting the specific functional interaction between these two proteins in vivo. These observations are in agreement with our proposal that Ost5p and Stt3p are involved in the conformational change of the OT complex.

It is of interest that enzymes that adopt an allosteric regulation mechanism tend to be larger and more complex than nonallosteric enzymes. An example is aspartate transcarbamoylase (29), which catalyzes an early reaction in the biosynthesis of pyrimidine nucleotides. It has 12 polypeptide chains organized into two stacked catalytic clusters, each with three catalytic polypeptide chains and three regulatory polypeptide chains (29). Another example is the non-enzyme oxygen-carrying protein hemoglobin, which contains two αβ dimers (α1β1 and α2β2). Upon oxygen binding, one pair of αβ subunits shifts with respect to the other by a rotation of 15° to convert the T (tense) state to the R (relaxed) state (30, 31). These multisubunit systems usually employ an allosteric regulation mechanism to accomplish the sequential reactions involved in a pathway or allow the fine-tuning changes in activity. Important examples are found in a variety of physiological processes such as digestion, blood clotting, hormone action, and vision. Enzymatic protein N-glycosylation is obviously a highly complex process because it governs the subsequent protein folding, stability, and processing. Thus, it is not surprising that the enzyme that catalyzes this process utilizes a fine-tuning system to ensure the selective N-glycosylation.

In this study, we also used the split-ubiquitin system to investigate the topology features of several yeast OT subunits that contain multiple transmembrane domains. Our study has resulted in definitive information on the orientation of the C termini of OT subunit proteins and indication on the presence or the absence of a cleavable N-terminal signal sequence. The findings are consistent with previous studies in the case of Ost2p (32) and Stt3p (18) by other groups and the predictions on Ost5p (27). In the case of Ost6p, it was previously predicted to have a cleavable signal sequence at its N terminus (33). However, our observations strongly suggest that it does not have an N-terminal cleavable signal sequence. As a matter of fact, current prediction programs often mis-predict N-terminal transmembrane segments as cleaved signal peptides (H. Kim and G. Van Heijne, personal communication). Our findings also corrected the previous prediction on the topology of Ost3p and Ost6p developed based on hydropathy analysis according to the algorithm of Kyte and Doolittle (34), which concluded that both of their N and C termini are located in the ER lumen with an even number of transmembrane segments. However, we clearly established that both N and C termini of Ost6p as well as the C terminus of Ost3p are located in the cytosol.

It has been shown that the flanking domain adjacent to the signal sequence, especially the charged residues in the flanking fragment, might affect the orientation of the signal sequence (19, 21). Therefore, it was important to ask whether the Nub moiety in the form of NubG-X disturbed the N-terminal orientation of the native protein and therefore yielded an artifactual result in our study. This result seems unlikely for three reasons. (i) The polypeptide encompassing the 1-34 aa of ubiquitin contains an equal number of positive (the α-amino group of the polypeptide and five lysine residues) and negative charges (three aspartic acid and three glutamic acid residues) and the net charge of this polypeptide should be close to neutral. (ii) The N-terminal orientation of the proteins being studied was shown not to be identical. If this flanking fragment of Nub strongly defined the orientation of the signal sequence, it would be expected that these proteins would display the same orientation pattern. (iii) Study of the N-terminal orientation of Ost4p using the same approach (data not shown) revealed exactly the same result as that established previously (17), indicating that Nub moiety does not affect its orientation. In summary, as an alternative tool for topology mapping, the split-ubiqutin assay can also be used to characterize the topology features of membrane proteins. The drawback of this approach is that it is not applicable to an unknown protein, unless it has a known interaction partner with its C terminus located in the cytosol.

Despite this potential drawback, the split-ubiquitin system has been proven to be an emerging powerful tool to characterize membrane proteins by us and others. Our findings have provided a plausible interpretation to several previous biochemical and genetic observations and partially explained why an unusually large number of components are involved in the cotranslational protein N-glycosylation. A hypothetical proposal is that binding of the lipid-linked oligosaccharide donor to the Stt3p subunit triggers the concerted motion of other subunits that are involved in the conformational rearrangement. However, at present, there is only indirect evidence to support this speculation. Clearly, extensive genetic, as well as biochemical, studies and x-ray crystallographic analysis are needed to test this idea.

Note. During the preparation of this manuscript, the topology of Stt3p was reported by Kim et al. (18).

Supplementary Material

Supporting Table:


We thank Dr. Hong Fang (Vanderbilt University) and Dr. Hyun Kim (Stockholm University) for helpful discussions with us on the topology features of membrane proteins. We appreciate Dr. Robert Noiva (University of South Dakota) and all members of the W.J.L. laboratory for their critical reading of the manuscript. We are grateful for the financial support of National Institutes of Health Grant GM33185 (to W.J.L.).


Author contributions: A.Y. designed research; A.Y. and E.W. performed research; A.Y. analyzed data; W.J.L. was the principal investigator on the paper; and A.Y. wrote the paper.

Abbreviations: OT, oligosaccharyl transferase; ER, endoplasmic reticulum; -His plates, plates lacking histidine.


1. Breuer, W. & Bause, E. (1995) Eur. J. Biochem. 228, 689-696. [PubMed]
2. Karaoglu, D., Silberstein, S., Kelleher, D. J. & Gilmore, R. (1995) Cold Spring Harbor Symp. Quant. Biol. 60, 83-92. [PubMed]
3. Knauer, R. & Lehle, L. (1999) Biochim. Biophys. Acta 1426, 259-273. [PubMed]
4. Yan, A. & Lennarz, W. J. (2005) J. Biol. Chem. 280, 3121-3124. [PubMed]
5. Van Criekinge, W. & Beyaert, R. (1999) Biol. Proced. Online 2, 1-38. [PMC free article] [PubMed]
6. Park, H. & Lennarz, W. J. (2000) Glycobiology 10, 737-744. [PubMed]
7. Johnsson, N. & Varshavsky, A. (1994) Proc. Natl. Acad. Sci. USA 91, 10340-10344. [PMC free article] [PubMed]
8. Stagljar, I., Korostensky, C., Johnsson, N. & te Heesen, S. (1998) Proc. Natl. Acad. Sci. USA 95, 5187-5192. [PMC free article] [PubMed]
9. Stagljar, I. (December 16, 2003) Sci. STKE, .10.1126/stke.2132003pe56 [PubMed] [Cross Ref]
10. Reinders, A., Schulze, W., Kuhn, C., Barker, L., Schulz, A., Ward, J. M. & Frommer, W. B. (2002) Plant Cell 14, 1567-1577. [PMC free article] [PubMed]
11. Reinders, A., Schulze, W., Thaminy, S., Stagljar, I., Frommer, W. B. & Ward, J. M. (2002) Structure (Camb.) 10, 763-772. [PubMed]
12. Obrdlik, P., El-Bakkoury, M., Hamacher, T., Cappellaro, C., Vilarino, C., Fleischer, C., Ellerbrok, H., Kamuzinzi, R., Ledent, V., Blaudez, D., et al. (2004) Proc. Natl. Acad. Sci. USA 101, 12242-12247. [PMC free article] [PubMed]
13. Pandey, S. & Assmann, S. M. (2004) Plant Cell 16, 1616-1632. [PMC free article] [PubMed]
14. Yan, A., Ahmed, E., Yan, Q. & Lennarz, W. J. (2003) J. Biol. Chem. 278, 33078-33087. [PubMed]
15. Silberstein, S., Collins, P. G., Kelleher, D. J., Rapiejko, P. J. & Gilmore, R. (1995) J. Cell Biol. 128, 525-536. [PMC free article] [PubMed]
16. te Heesen, S. & Aebi, M. (1992) Yeast 8, 1101-1103. [PubMed]
17. Kim, H., Yan, Q., Von Heijne, G., Caputo, G. A. & Lennarz, W. J. (2003) Proc. Natl. Acad. Sci. USA 100, 7460-7464. [PMC free article] [PubMed]
18. Kim, H., von Heijne, G. & Nilsson, I. (March 21, 2005) J. Biol. Chem., .10.1074/jbc.M412213200 [PubMed] [Cross Ref]
19. Goder, V. & Spiess, M. (2003) EMBO J. 22, 3645-3653. [PMC free article] [PubMed]
20. Ng, D. T., Brown, J. D. & Walter, P. (1996) J. Cell Biol. 134, 269-278. [PMC free article] [PubMed]
21. Spiess, M. (1995) FEBS Lett. 369, 76-79. [PubMed]
22. Zubkov, S., Lennarz, W. J. & Mohanty, S. (2004) Proc. Natl. Acad. Sci. USA 101, 3821-3826. [PMC free article] [PubMed]
23. Riek, R. P., Rigoutsos, I., Novotny, J. & Graham, R. M. (2001) J. Mol. Biol. 306, 349-362. [PubMed]
24. Cordes, F. S., Bright, J. N. & Sansom, M. S. (2002) J. Mol. Biol. 323, 951-960. [PubMed]
25. Van den Berg, B., Clemons, W. M., Jr., Collinson, I., Modis, Y., Hartmann, E., Harrison, S. C. & Rapoport, T. A. (2004) Nature 427, 36-44. [PubMed]
26. Karaoglu, D., Kelleher, D. J. & Gilmore, R. (2001) Biochemistry 40, 12193-12206. [PubMed]
27. Zufferey, R., Knauer, R., Burda, P., Stagljar, I., te Heesen, S., Lehle, L. & Aebi, M. (1995) EMBO J. 14, 4949-4960. [PMC free article] [PubMed]
28. Reiss, G., te Heesen, S., Gilmore, R., Zufferey, R. & Aebi, M. (1997) EMBO J. 16, 1164-1172. [PMC free article] [PubMed]
29. Villeret, V., Clantin, B., Tricot, C., Legrain, C., Roovers, M., Stalon, V., Glansdorff, N. & Van Beeumen, J. (1998) Proc. Natl. Acad. Sci. USA 95, 2801-2806. [PMC free article] [PubMed]
30. Lukin, J. A., Kontaxis, G., Simplaceanu, V., Yuan, Y., Bax, A. & Ho, C. (2003) Proc. Natl. Acad. Sci. USA 100, 517-520. [PMC free article] [PubMed]
31. Bolton, W. & Perutz, M. F. (1970) Nature 228, 551-552. [PubMed]
32. Silberstein, S., Collins, P. G., Kelleher, D. J. & Gilmore, R. (1995) J. Cell Biol. 131, 371-383. [PMC free article] [PubMed]
33. Knauer, R. & Lehle, L. (1999) J. Biol. Chem. 274, 17249-17256. [PubMed]
34. Kyte, J. & Doolittle, R. F. (1982) J. Mol. Biol. 157, 105-132. [PubMed]

Articles from Proceedings of the National Academy of Sciences of the United States of America are provided here courtesy of National Academy of Sciences
PubReader format: click here to try


Related citations in PubMed

See reviews...See all...

Cited by other articles in PMC

See all...


Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...