The Ost1p Subunit of Yeast Oligosaccharyl Transferase Recognizes the Peptide Glycosylation Site Sequence, -Asn- X -Ser/Thr-*

Other laboratories have established that oligosaccharyl transferase (OST) from Saccharomyces cerevisiae can be purified as a protein complex containing eight different subunits. To identify the OST subunit that recognizes the peptide sites that can be glycosylated, we developed photoaffinity probes containing a photoreactive benzophenone derivative, p -benzoylphenylalanine (Bpa), as part of an 125 I-labeled peptide that could be expected to be glycosylated. We found that Asn-Bpa-Thr peptides served as substrates for OST and that photoactivation of these probes in the presence of microsomes abolished the OST activity. Photoactivation of 125 I-la-beled Asn-Bpa-Thr in the presence of microsomes re-sulted in specific covalent labeling of a protein doublet of molecular mass 62 and 64 kDa. By carrying out the photoactivation of the probe using microsomes containing epitope-tagged Ost1p, we demonstrated that the 125 I-labeled protein was Ost1p. Radiolabeling of this protein was dependent on irradiation at 350 nm. No labeling was detected using a probe containing Ala instead of Thr as the third amino acid residue. We conclude that Ost1p is the subunit of the OST complex that recognizes the peptide sites in the nascent chains that are destined to be glycosylated. The

Other laboratories have established that oligosaccharyl transferase (OST) from Saccharomyces cerevisiae can be purified as a protein complex containing eight different subunits. To identify the OST subunit that recognizes the peptide sites that can be glycosylated, we developed photoaffinity probes containing a photoreactive benzophenone derivative, p-benzoylphenylalanine (Bpa), as part of an 125 I-labeled peptide that could be expected to be glycosylated. We found that Asn-Bpa-Thr peptides served as substrates for OST and that photoactivation of these probes in the presence of microsomes abolished the OST activity. Photoactivation of 125 I-labeled Asn-Bpa-Thr in the presence of microsomes resulted in specific covalent labeling of a protein doublet of molecular mass 62 and 64 kDa. By carrying out the photoactivation of the probe using microsomes containing epitope-tagged Ost1p, we demonstrated that the 125 Ilabeled protein was Ost1p. Radiolabeling of this protein was dependent on irradiation at 350 nm. No labeling was detected using a probe containing Ala instead of Thr as the third amino acid residue. We conclude that Ost1p is the subunit of the OST complex that recognizes the peptide sites in the nascent chains that are destined to be glycosylated.
The key enzyme in the process of N-glycosylation of proteins, oligosaccharyl transferase (OST), 1 catalyzes the transfer of a preassembled high-mannose oligosaccharide from a lipidlinked oligosaccharide donor (Dol-PP-GlcNAc 2 Man 9 Glc 3 ) onto asparagine acceptor sites on nascent polypeptide chains being translocated into the lumen of the rough ER (1,2). The consensus sequence of the acceptor, -Asn-X-Thr/Ser-, where X is any amino acid other than Pro (3,4), was confirmed using a variety of synthetic peptides as substrates for in vitro glycosylation in microsomes (5)(6)(7)(8).
OST in Saccharomyces cerevisiae was initially purified as a complex consisting of six polypeptides (␣, ␤, ␥, ␦, ⑀, subunits) with enzymatic activity (9). Subsequently, two other groups (10, 11) have described a tetrameric OST complex that lacks the ⑀ and subunits. So far in yeast eight subunits have been cloned that may be components of this enzyme complex. Five of them are essential genes (OST1, OST2, WBP1, SWP1, and STT3). The components of yeast OST complex show significant sequence similarity to the components of the complex purified from higher eukaryotes (12).
Much work has been done to characterize the substrates of OST (13)(14)(15)(16), but little is known about the function of the subunits of the enzyme complex. Following the observation that modification of a cysteine residue on the OST complex by methyl methanethiosulfonate caused time-and concentrationdependent inactivation of enzyme activity, a biotin-tagged form of this reagent was shown to inactivate the enzyme and to label Wbp1p. Based on the finding that the substrate, Dol-PP-Glc-NAc 2 , protected the enzyme from inactivation, it was proposed that Wbp1p may contain a site for the binding of the lipidlinked oligosaccharide (11). Following up on an earlier observation (16) on inactivation of pig liver OST by a hexapeptide in which Thr was replaced by epoxyethylglycine in the -Asn-X-Thr-consensus sequence, Bause et al. (17) used a N-dinitrobenzoylated form of this hexapeptide with the objective being to identify an OST subunit(s) that might become covalently linked to the epoxy inhibitor. Following incubation of the pig liver OST with Dol-PP-14 C-oligosaccharides and the N-dinitrobenzoylated hexapeptide, two polypeptides, proposed to be ribophorin I and Ost48p, were found to be immunolabeled and radioactive.
Our approach has been to identify one or more subunits of the yeast OST complex that would be expected to be involved in recognition of -Asn-X-Ser/Thr-glycosylation sites in the growing polypeptide chain. To accomplish this, we took advantage of the fact that small glycosylatable peptides can mimic nascent chains, because they can enter the ER and become glycosylated (5,6,8). Based on this fact, we developed photoreactive glycosylatable peptides containing a benzophenone moiety. Such probes have several advantages: 1) they are chemically more stable than other photoreactive moieties such as diazo esters, aryl azides, and diazirines; 2) can be manipulated in ambient light and can be activated at 350 nm, conditions that minimize damage to proteins; and 3) react preferentially with normally unreactive C-H bonds, even in the presence of aqueous solvent and bulk nucleophiles. These three properties combine to allow for highly efficient covalent modifications of macromolecules, frequently with high site specificity (18,19). We used photoprobes containing the -Asn-X-Thr-consensus sequence in which the X amino acid was p-benzoylphenylalanine (Bpa). Using this probe we were able to show that Ost1p is the subunit in the OST complex that recognizes the N-glycosylation site sequence.

EXPERIMENTAL PROCEDURES
Strains-W303-1a (MAT a ade2 can1 his3 leu2 trp1 ura3) was used as the parental strain to generate the Ost1p hemagglutinin (HA) and c-Myc epitopes integrated into the chromosome. Polymerase chain reaction was carried out using the ME-3 plasmid (which contains a triple HA tag and the HIS5 gene from Schizosaccharomyces pombe) as the template. The primers were 5Ј GCACTAATGGGAGTTTTTGTCTTAA-AAACTTTGAACATGAACGTAACTAACTACCCATACGATGTTCCT 3Ј and 5Ј TAAAACTTATAAAAAATTAAATACTGAATTTATTGCTCTGC-ATCATAAGGTGTCGACGGTATCGATAAG 3Ј. A linear polymerase chain reaction product was used to transform W303-1a by homologous recombination. The resulting transformant was designated as QYY101. QYY102 was generated in a similar way, but instead of containing the HA epitope, it contained nine copies of a c-Myc tag. QYY103 was generated by crossing QYY102 with RGY330. RGY330 (MAT ␣ leu2-⌬1 lys2-801 ade2-101 trp1-⌬1 his3-⌬200 with HA incorporated into the COOH terminus of Ost3p) was obtained from R. Gilmore (Department of Biochemistry and Molecular Biology, University of Massachusetts). QYY104 was generated by crossing W303-1a with RGY330.
Conditions for Photoactivation and Immunoprecipitation-Irradiation of crude yeast microsomes was conducted at 4°C in a borosilicate glass tube (12 ϫ 75 mm, VWR Scientific) whose bottom and sides were located 2 cm from the surface of four 350 nm photochemical reactor lamps (RMR 3500A, Southern New England Ultraviolet Co., Hamden, CT). In a standard assay, 200 l of crude microsomes (protein concentration around 5 g/l) was irradiated for 10 min with 0.3 Ci of 125 I-bh-Asn-Bpa-Thr-Am or 125 I-bh-Asn-Bpa-Ala-Am in B88 buffer (20 mM Hepes-KOH, pH 6.8, 150 mM KOAc, 250 mM sorbitol, 5 mM MgOAc) containing 10 mM MnCl 2 . After irradiation, immunoprecipitation was performed as described by Karaoglu et al. (25) except that after photoactivation the reaction was adjusted to 1.5% digitonin, 0.5 M NaCl, 20 mM Tris⅐Cl, pH 7.4, 3.5 mM MgCl 2 . The solution was centrifuged for 20 min at 55,000 rpm in a TLA 100.3 rotor (Beckman), and the supernatant fraction was used for immunoprecipitation.
Preparation of Yeast Lysate-Cells were collected by centrifugation from a 5-ml log phase culture (A 600 /ml ϭ 1.0), washed with 5 ml of distilled H 2 O, and resuspended in 200 l of lysis buffer (20 mM Tris⅐HCl, pH 7.5, 100 mM KCl, 10% glycerol, 1 mM EDTA, 1 mM phenylmethylsulfonyl fluoride). After addition of 75 l of acid-washed glass beads, sonication using a Branson sonicator at level 3 was performed on ice for three 15-s intervals with a 15-s wait on ice between sonications. The cell lysate was recovered by centrifugation for 15 min at 16,000 ϫ g; it was subsequently used for analysis by SDS-PAGE with 10% gels followed by Western blot analysis.

RESULTS
Asn-Bpa-Thr Serves as a Substrate for OST-To search for the peptide recognition subunit(s) of OST, 125 I-bh-Asn-Bpa-Thr-Am and [ 3 H]Ac-Asn-Bpa-Thr-Am, two tripeptides in which the amino acid in the X position consisted of photoreactive Bpa, were synthesized (Fig. 1). The assumption was made that binding of these two tripeptides to one or more subunits of OST would occur because these peptides were expected to be substrates for OST. To test this assumption the tripeptides were incubated with a yeast cell lysate without photoactivation at 350 nm. Labeled glycopeptide formed from the labeled peptide was quantitated by binding of the former to ConA-agarose beads as described (24). The results shown in Fig. 2 indicate that formation of a 3 H-labeled glycopeptide that binds to ConA-agarose was dependent on time ( Fig. 2A), the amount of cell lysate (Fig. 2B), and the amount of [ 3 H]Ac-Asn-Bpa-Thr-Am (Fig. 2C). Furthermore, addition of a 100-fold molar excess of the unlabeled peptide, Ac-Asn-Phe-Thr-Am, which is also a substrate for N-glycosylation, markedly inhibited formation of 3 H-glycopeptide (data not shown). These results established that [ 3 H]Ac-Asn-Bpa-Thr-Am was a substrate for OST. Similar experiments with the 125 I-bh-Asn-Bpa-Thr-Am probe gave similar results (data not shown). Since this work was completed a report on the synthesis and glycosylation of Bpa-containing peptides has appeared (26).
Photoactivation of Asn-Bpa-Thr in the Presence of Microsomes Abolishes OST Activity-OST activity is essential for viability of yeast. If covalent attachment of the peptide to an OST subunit occurs, it would be expected that the OST complex would lose its catalytic activity, since the probe would be irreversibly attached to the peptide binding site. Of course this might only be expected if cross-linking occurred to one of the subunits of OST known to be essential for viability, namely, Ost1p, Ost2p, Wbp1p, Swp1p, and Stt3p. To test the effect of cross-linking by irradiation on the OST activity of microsomes, we used unlabeled photoprobe bh-Asn-Bpa-Thr-Am. Although the conditions used in this experiment differed somewhat than those ultimately used to link labeled photoprobe to a polypeptide, the results in Fig. 3 clearly showed that irradiation in the presence of bh-Asn-Bpa-Thr-Am abolished OST activity in a time-dependent manner. In a control experiment, microsomes that were irradiated without probe were found to lose little of their OST activity. Yeast cells were converted to spheroplasts, lysed, and then assayed for OST activity as described previously (24). The lysate was incubated with [ 3 H]Ac-Asn-Bpa-Thr-Am for 20 min or the indicated time at 25°C. Then the reaction was stopped by adding Nonidet P-40 to a final concentration of 1%, and 3 H-glycopeptide formation was measured by binding to ConA-agarose beads. A, formation of 3 H-labeled glycoprobe is dependent on the incubation time. B, formation of glycopeptide is dependent on the amount of cell lysate. An A 600 value of 1 is equivalent to lysate prepared from 2 ϫ 10 7 cells. C, formation of glycopeptide is dependent on the amount of probe added.

Ost1p Recognizes the Peptide Glycosylation Sequence
Irradiation of Microsomes and Labeled Asn-Bpa-Thr Yields an 125 I-Labeled Protein That Is Immunoprecipitated by Antibody to HA-tagged Ost3p-Having established that 125 I-bh-Asn-Bpa-Thr-Am and [ 3 H]Ac-Asn-Bpa-Thr-Am are substrates of OST and that irradiation of the unlabeled photoprobe peptide inactivated OST, we investigated the target of its binding. It has been shown that incorporation of the HA epitope into the COOH terminus of Ost3p, followed by cell rupture and then immunoprecipitation under nondenaturing conditions, results in precipitation of all eight subunits of the OST complex (25). After confirming this observation, we carried out experiments using microsomes prepared from a yeast strain in which an HA-tagged allele of Ost3p was integrated into the chromosome. Photoactivation of these microsomes in the presence of 125 I-bh-Asn-Bpa-Thr-Am was followed by solubilization, and then immunoprecipitation of Ost3-HAp, in order to detect any radiolabeled OST subunits. The results shown in Fig. 4, lane 2, established that irradiation of microsomes prepared from the yeast strain carrying Ost3-HAp in the presence of 125 I-bh-Asn-Bpa-Thr-Am produced a doublet of radiolabeled bands of apparent molecular masses of 62 and 64 kDa (Fig. 4, lane 2). As shown in Fig. 4, lane 1, when the microsomes were incubated with 125 I-bh-Asn-Bpa-Thr-Am, in the absence of irradiation, followed by immunoprecipitation and subsequent SDS-PAGE, no labeled protein bands were detected. Therefore, the radiolabeling of these proteins was irradiation-dependent. In other experiments (data not shown) we have found that reduction of the photoreactive ketone group in the Bpa abolished its ability to become covalently linked to protein. In addition, when the photoactivation was carried out using microsomes that had been inactivated by either heating, or by adding 4-fold molar excess of EDTA (to deplete the divalent cations known to be necessary for OST activity), radiolabeling of these two bands did not occur (lanes 7 and 8, respectively).
To further confirm that photoactivation led to specific labeling of a polypeptide, a competition experiment was performed using a labeled probe and either an unlabeled peptide that is a known substrate for OST or a peptide that has been shown to not be a substrate. As demonstrated in Fig. 4, lane 4, when microsomes were incubated with 125 I-bh-Asn-Bpa-Thr-Am with the presence of a large excess of a competing acceptor peptide, Ac-Asn-Phe-Thr-Am, labeling of the 62/64-kDa doublet did not occur. However, when microsomes were incubated with the labeled probe and a large excess of a peptide, Ac-Asp(NHCH 3 )-Leu-Thr-Am, which is known to be inactive as a substrate for OST (27), labeling of the 62/64-kDa band was unaffected (Fig. 4, lane 5). Furthermore, when the labeled acceptor photoprobe was replaced with another labeled photoprobe that is not a substrate for OST, 125 I-bh-Asn-Bpa-Ala-Am, no labeling was observed (lane 6). These results clearly showed that the photolabeling of the 62/64-kDa protein requires a probe that is a substrate for OST.
The Radiolabeled Protein Is Ost1p-Based on the specificity of anti-HA in precipitating only the OST subunits (25) and the apparent molecular mass of the photolabeled protein we presumed that the subunit being labeled was Ost1p. It has been reported Ost1p migrates as a triplet of protein bands at 60/ 62/64 kDa due to different glycosylation states (26), although as shown in Fig. 4, lane 2, we detected only a doublet at 62 and 64 kDa. To provide more direct proof that the labeled protein was Ost1p, photoactivation was carried out using microsomes prepared from a strain bearing chromosome-integrated HA tag located at the COOH terminus of Ost1p. Immunoprecipitation with anti-HA, followed by SDS-PAGE, revealed two labeled protein bands migrating at 67 and 65 kDa (Fig. 4, lane 3). The HA tag had a mass about 3 kDa, which explains why the doublet of labeled Ost1p now had a slower mobility than the Ost1p doublet observed when the epitope was attached to Radiolabeled photoprobe was added to the microsomes, and photoactivation was carried out at 4°C. After irradiation, membranes were solubilized by mild detergent and the pellet discarded after ultracentrifugation. The supernatant was used for immunoprecipitation using anti-HA antibody. After immunoprecipitation, samples were resolved on SDS-PAGE and detected by autoradiography. In all lanes microsomes prepared from yeast strain containing Ost3-HAp were used, except lane 3, in which the microsomes contained Ost1-HAp. Lane 1, microsomes and photoprobe were incubated in the dark. Lane 2, photoactivation under normal conditions. Lane 3, labeling using microsomes made from a strain containing Ost1-HAp. In other experiments, it was found that the level of labeling was essentially the same as that in lane 2 when an equal amount of microsomes was analyzed. Lane 4, unlabeled acceptor peptide or nonacceptor peptide (lane 5) were added. Because of the limited amount of labeled probe the concentration used was 0.5 nM, which is far below the estimated K m value of 5-10 M. Consequently, in the competition experiments using unlabeled acceptor, Ac-Asn-Phe-Thr-Am (lane 4) or nonacceptor, Ac-Asp(NHCH 2 )-Leu-Thr-Am (lane 5), the compounds were added at 50 M, approximately 5-10 times the K m values. Lane 6, 125 I-bh-Asn-Bpa-Ala-Am, which is not an OST substrate, was tested as a photoprobe. Lane 7, microsomes were heated at 100°C for 5 min before photoactivation. Lane 8, EDTA was added to the reaction mix before photoactivation.
To further confirm that the labeled protein was Ost1p, we took advantage of the molecular weight shift expected upon synthesis of Myc-tagged Ost1p. A strain was generated with nine copies of Myc linked to the COOH terminus of Ost1p (QYY102). Then a diploid strain (QYY103) was made by crossing this Ost1-Mycp strain with strain RGY 330, which contains Ost3-HAp. A control diploid strain containing an untagged form of Ost1p was also made by crossing strain W303-1a with strain RGY330 (QYY104). Having confirmed that these diploid strains have the expected properties by Western blot analysis (data not shown), microsomes were prepared from them and then irradiation and immunoprecipitation were carried out as described above. As shown in Fig. 5, lane 2, two radiolabeled proteins at 62 and 76 kDa were observed. As expected this diploid strain (QYY103) exhibited two different radiolabeled forms of Ost1p; the original Ost1p and Ost1-Mycp, with the increased mass expected from addition of the Myc epitope. As a control, in Fig. 5, lane 1, only one form of radiolabeled Ost1p was observed in microsomes made using the control diploid strain (QYY104), which contained only Ost1p. DISCUSSION In an earlier effort to identify the peptide recognition element of OST by photoactivation of an acceptor, we utilized a glycosylatable peptide containing a photoreactive benzoylazido group attached to the ⑀-NH 2 group of Lys located in the -Xposition (22). We found that this probe did not label a membrane protein, but instead the lumenal protein, protein disulfide isomerase (PDI) (28). Subsequently we excluded the possibility that labeling of a membrane protein failed because of the high abundance of PDI; even when PDI was depleted from the ER, little labeling of membrane proteins occurred. 2 Using considerations described earlier (22), in the current study, we introduced hydrophobicity into the peptide by 1) attachment of a Bolton-Hunter group on the NH 2 terminus, 2) blocking of the carboxy group at the COOH terminus, and 3) introduction of the Bpa group in the X position. These hydrophobic modifications were made because in vitro studies had demonstrated that hydrophobic tripeptides were much better substrates than polar, water-soluble peptides (8). Second, photoreactive Bpa was introduced into the peptide as the middle amino acid, because it is known that peptides containing any amino acid except proline in the X position are substrates (7), whereas those with modifications on either the Asn or the Thr residues are not recognized by OST (6, 8, 22, 29 -31). Based on the fact that in the current study no labeling of PDI was observed, 3 it is clear that the reason for the different results in labeling of polypeptides in the microsomes in this and the earlier study must be due to the nature of the photoreactive moiety itself, rather than the overall design of the peptide.
Using this photoprobe we have established that specific labeling of one of the essential subunits of OST, Ost1p, occurs. This specificity is documented by the finding that only this protein becomes photocovalently modified, even though the antibody directed against Ost3p has been shown to immunoprecipitate the entire OST complex. Previously, it has been directly shown that the OST1 gene is essential for growth and Ost1p is essential for OST activity (32). Our data clearly indicated that the photolabeling of Ost1p is an active site-directed reaction of the tripeptide and the OST subunit. A prerequisite for this active site-directed photolabeling is that the photoreactive peptide should be recognized by and specifically bind to the enzyme. This is the case, since the photoprobe can be glycosylated. Furthermore, photoactivation with microsomes abolishes OST activity and, as expected, photolabeling was dependent upon divalent cations. Additional evidence for the specificity of labeling was afforded by the observation that labeling did not occur when the Thr in the third position was converted to Ala.
It is of interest to consider several aspects of the finding that Ost1p is the subunit that recognizes -Asn-X-Ser/Thr-. The sequence of this protein coupled with its glycosylated state indicates that it is a Type I membrane protein containing a short cytosolic COOH-terminal tail and a long NH 2 -terminal lumenal domain. Given the requirement that Ost1p recognize a small structural element, namely, two of three amino acids in a sequence within a growing polypeptide chain, the extended lumenal domain of Ost1p (427 amino acid residues) could have a sufficiently high degree of order to generate a very structured, specific binding site that can distinguish between Asn and Gln, and between Ser/Thr and Ala. Another aspect of interest is the relationship between this finding in yeast and work in mammalian systems, since Ost1p is homologous to ribophorin I. As mentioned above, Bause et al. (17) have proposed that ribophorin I and Ost48p are involved in recognition of the acceptor peptide and Dol-PP-oligosaccharide, and the work of Pathak et al. (11) in yeast has implicated Wbp1p, the yeast homolog of Ost48p, in recognition of Dol-PP-oligosaccharide. Clearly a next step in extending our findings on the participation of Ost1p in recognition of the peptide binding site will be to use a bifunctional cross-linking probe to determine the possible interaction of Ost1p with other subunits of the yeast OST complex. FIG. 5. Photoactivation of microsomes prepared from diploid strain QYY103 yields two 125 I-labeled proteins, Ost1p and Ost1-Mycp. Microsomes were prepared from diploid cells as indicated in the text. Photoactivation and immunoprecipitation using anti-HA antibody was carried out as described and then radiolabeled proteins were detected by autoradiography. Lane 1, microsomes were prepared from the diploid strain containing only endogenous Ost1p. Lane 2, microsomes were prepared from the diploid strain containing Ost1-mycp and Ost1p.