SemOpenAlex |

SemOpenAlex

Matches in SemOpenAlex for { <https://semopenalex.org/work/W2000504183> ?p ?o ?g. }

Showing items 1 to 87 of 87 with 100 items per page.

W2000504183 endingPage "2472" @default.
W2000504183 startingPage "2467" @default.
W2000504183 abstract "Retention of soluble proteins in the endoplasmic reticulum is dependent on their interaction with the KDEL (Lys-Asp-Glu-Leu) receptor in the Golgi apparatus and their subsequent retrieval back to the endoplasmic reticulum. We have studied the three-dimensional organization of the human KDEL receptor using site-directed mutagenesis and sulfhydryl-specific labeling. We have identified four amino acid residues, Arg-5, Asp-50, Tyr-162, and Asn-165, which we suggest participate in the formation of the ligand binding pocket. Arg-5 and Asp-50 are shown to be located on the lumenal side of the membrane and are inaccessible from the cytoplasm. In addition, our results strongly support a topology of the KDEL receptor similar to the family of G-protein-coupled receptors with seven transmembrane domains. Furthermore, Asp-50 plays a crucial role in the binding of His/Lys-Asp-Glu-Leu ligands, but is not required for Asp-Asp-Glu-Leu binding, suggesting that this residue forms an ion pair with the positively charged amino acid residue positioned 4 residues from the carboxyl terminus of the ligand. Retention of soluble proteins in the endoplasmic reticulum is dependent on their interaction with the KDEL (Lys-Asp-Glu-Leu) receptor in the Golgi apparatus and their subsequent retrieval back to the endoplasmic reticulum. We have studied the three-dimensional organization of the human KDEL receptor using site-directed mutagenesis and sulfhydryl-specific labeling. We have identified four amino acid residues, Arg-5, Asp-50, Tyr-162, and Asn-165, which we suggest participate in the formation of the ligand binding pocket. Arg-5 and Asp-50 are shown to be located on the lumenal side of the membrane and are inaccessible from the cytoplasm. In addition, our results strongly support a topology of the KDEL receptor similar to the family of G-protein-coupled receptors with seven transmembrane domains. Furthermore, Asp-50 plays a crucial role in the binding of His/Lys-Asp-Glu-Leu ligands, but is not required for Asp-Asp-Glu-Leu binding, suggesting that this residue forms an ion pair with the positively charged amino acid residue positioned 4 residues from the carboxyl terminus of the ligand. The precise sorting of proteins along the secretory pathway is crucial for the maintenance of organelle integrity. The localization of chaperones and other soluble proteins to the ER 1The abbreviations used are: ER, endoplasmic reticulum; AMS, 4-acetamido-4′-maleimidylstilbene-2,2′-disulfonic acid; BCM, 3-(N-maleimidylpropionyl)biocytin; NEM,N-ethylmaleimide.1The abbreviations used are: ER, endoplasmic reticulum; AMS, 4-acetamido-4′-maleimidylstilbene-2,2′-disulfonic acid; BCM, 3-(N-maleimidylpropionyl)biocytin; NEM,N-ethylmaleimide. is achieved by their continuous retrieval from post-ER compartments by the KDEL receptor, a membrane protein localized in the Golgi apparatus. Binding of soluble ER proteins to the receptor is dependent on a conserved tetrapeptide sequence at their carboxyl terminus, which is usually KDEL in animal cells and HDEL in Saccharomyces cerevisiae (see Ref. 1Pelham H.R.B. Trends Biochem. Sci. 1990; 15: 483-486Abstract Full Text PDF PubMed Scopus (432) Google Scholar for review). The HDEL receptor was originally isolated from yeast; it is the product of the ERD2 gene and has subsequently been cloned from other organisms (2Semenza J.C. Hardwick K.G. Dean N. Pelham H.R.B. Cell. 1990; 61: 1349-1357Abstract Full Text PDF PubMed Scopus (388) Google Scholar, 3Lewis M.J. Pelham H.R.B. Nature. 1990; 348: 162-163Crossref PubMed Scopus (243) Google Scholar, 4Tang B.L. Wong S.H. Qi X.L. Low S.H. Hong W. J. Cell Biol. 1993; 120: 325-338Crossref PubMed Scopus (115) Google Scholar, 5Elmendorf H.G. Haldar K. EMBO J. 1993; 12: 4763-4773Crossref PubMed Scopus (138) Google Scholar, 6Lee H.I. Gal S. Newman T.C. Raikhel N.V. Proc. Natl. Acad. Sci. U. S. A. 1993; 90: 11433-11437Crossref PubMed Scopus (91) Google Scholar). Erd2 determines both capacity and specificity of the H/KDEL retrieval system (7Lewis M.J. Sweet D.J. Pelham H.R.B. Cell. 1990; 61: 1359-1363Abstract Full Text PDF PubMed Scopus (163) Google Scholar), and a direct interaction between KDEL ligands and the human receptor has recently been demonstrated (8Wilson D.W. Lewis M.J. Pelham H.R.B. J. Biol. Chem. 1993; 268: 7465-7468Abstract Full Text PDF PubMed Google Scholar, 9Scheel A.A. Pelham H.R.B. Biochemistry. 1996; 35: 10203-10209Crossref PubMed Scopus (44) Google Scholar). Hydrophobicity analyses and the “positive inside rule” (10Sipos L. von Heijne G. Eur. J. Biochem. 1993; 213: 1333-1340Crossref PubMed Scopus (253) Google Scholar) suggest seven primarily hydrophobic membrane-spanning domains with the amino terminus in the lumen and the carboxyl terminus in the cytoplasm (11Semenza J.C. Pelham H.R.B. J. Mol. Biol. 1992; 224: 1-5Crossref PubMed Scopus (23) Google Scholar, 12Townsley F.M. Wilson D.W. Pelham H.R.B. EMBO J. 1993; 12: 2821-2829Crossref PubMed Scopus (110) Google Scholar). The topology of the KDEL receptor would thus be similar to the organization of G-protein-coupled receptors, although no amino acid sequence homology is observed. Based on studies using fusions to N-linked glycosylation sites, a different topology has been proposed with six transmembrane domains and the amino terminus located in the cytoplasm (13Singh P. Tang B.L. Wong S.H. Hong W. Mol. Cell. Biol. 1993; 13: 6435-6441Crossref PubMed Scopus (21) Google Scholar). Although the KDEL recycling system is well established, several significant questions have remained unanswered. For example, the sorting of KDEL proteins is believed to be controlled by the pH difference between the Golgi apparatus and the ER (8Wilson D.W. Lewis M.J. Pelham H.R.B. J. Biol. Chem. 1993; 268: 7465-7468Abstract Full Text PDF PubMed Google Scholar). However, it is unclear whether the pH difference between these two compartments, which is estimated to be approximately 0.5 pH units (8Wilson D.W. Lewis M.J. Pelham H.R.B. J. Biol. Chem. 1993; 268: 7465-7468Abstract Full Text PDF PubMed Google Scholar), is sufficient to ensure efficient binding of KDEL proteins in the Golgi apparatus and their release in the ER. In addition, ligand binding was suggested to induce retrograde transport of the receptor-ligand complexes, involving a signal transduction mechanism across the membrane (14Lewis M.J. Pelham H.R.B. Cell. 1992; 68: 353-364Abstract Full Text PDF PubMed Scopus (306) Google Scholar). Neither the mechanism of signal transduction nor any components of the machinery apart from the KDEL receptor have been identified. It has been shown that the KDEL receptor, presumably together with bound ligand, is selectively recruited into COPI-coated vesicles from Golgi membranes, but this process is not well characterized (15Sönnichsen B. Watson R. Clausen H. Misteli T. Warren G. J. Cell Biol. 1996; 134: 1411-1425Crossref PubMed Scopus (88) Google Scholar, 16Orci L. Stamnes M. Ravazzola M. Amherdt M. Perrelet A. Söllner T.H. Rothman J.E. Cell. 1997; 90: 335-349Abstract Full Text Full Text PDF PubMed Scopus (351) Google Scholar). As a better understanding of these events requires detailed analyses of the receptor structure, we set out to identify amino acid residues involved in ligand binding and to investigate the topology of the human Erd2.1 protein, one of two related KDEL receptors identified in humans (17Lewis M.J. Pelham H.R.B. J. Mol. Biol. 1992; 226: 913-916Crossref PubMed Scopus (60) Google Scholar). We undertook a sulfhydryl-specific labeling approach, creating site-directed mutants with unique cysteine residues engineered at different positions in the receptor, followed by chemical labeling with thiol-specific reagents. We have investigated the effect of derivatization of single cysteine residues on ligand binding and the sidedness of these residues by labeling of receptor mutants with membrane-permeant and -impermeant sulfhydryl-specific compounds. The baculovirus transfer vector pVL1393 and linearized baculovirus DNA were from PharMingen, Ni-NTA resin was from Qiagen, NEM and 3-(N-maleimidylpropionyl)biocytin (biotin maleimide) were from Sigma, 4-acetamido-4′-maleimidylstilbene-2,2′-disulfonic acid (AMS) was from Molecular Probes, and a streptavidin-horseradish peroxidase conjugate was from Pierce. Site-directed mutagenesis was performed according to Kunkel (18Kunkel T.A. Roberts J.D. Zakour R.A. Methods Enzymol. 1987; 154: 367-382Crossref PubMed Scopus (4558) Google Scholar) in a pBluescript vector containing the human Erd2.1 cDNA modified at the 3′-end to add a Myc and a hexahistidine tag (hErd2-Myc-His, see Ref. 9Scheel A.A. Pelham H.R.B. Biochemistry. 1996; 35: 10203-10209Crossref PubMed Scopus (44) Google Scholar). All sequences were confirmed by dye terminator cycle sequencing using an Applied Biosystems model 377 sequencer. The cDNAs encoding receptor mutants were cloned into aSmaI/BglII-digested baculovirus transfer vector pVL1393. Mutants are named as (wild-type residue and number) → (mutant residue). Recombinant baculoviruses containing mutant Erd2.1 cDNAs under the control of a polyhedrin promoter were created using the Baculogold kit from PharMingen according to the information provided by the manufacturer. Isolation of virus clones and infection of Sf9 cells were carried out as described (19Summers M.D. Smith G.E. A Manual of Methods for Baculovirus Vectors and Insect Cell Culture Procedures. Texas Agricultural Experiment Station, College Station, TX1987Google Scholar). Preparation of a crude membrane fraction from infected cells and ligand binding assays were carried out as described previously (9Scheel A.A. Pelham H.R.B. Biochemistry. 1996; 35: 10203-10209Crossref PubMed Scopus (44) Google Scholar). Briefly, standard binding assays contained membranes derived from infected insect cells expressing wild-type or mutant KDEL receptors, approximately 180,000 cpm of 125I-YTSEKDEL, 100 ng of YTSEKDELGL (nonfunctional ligand) in binding assay buffer (20 mm NaCl, 250 μg/ml bovine serum albumin, and 50 mm sodium cacodylate pH 5.0). Unless otherwise mentioned, all binding data are shown without background subtraction. Competition assays contained 180,000 cpm125I-YTFEHDEL and variable amounts of competitor peptide. All binding assays were done under nonsaturating conditions. To assess the effect of alkylation of individual cysteine residues on the binding activity of receptor mutants, 1–10 μg of crude membranes were incubated with NEM or biotin maleimide in 10 mm Hepes, pH 7.0, for 10 min on ice and were then directly used in the peptide binding assay. For analysis by Western blotting, 100 μg of crude membranes were incubated with 100 μm biotin maleimide for 10 min on ice in phosphate buffer (50 mm sodium phosphate, pH 7.0), quenched with 10 mm β-mercaptoethanol and solubilized with 0.5% SDS at room temperature. The soluate was spun at 14,000 × g for 20 min, and the receptor was bound to Ni-NTA resin by incubation of the soluble detergent extract with 40 μl of resin under rotation for 1 h at room temperature. Unbound proteins were washed off and bound receptor protein eluted with 1m imidazole, pH 7.0. All buffers during the purification contained 0.5% SDS. Biotin-labeled KDEL receptor was visualized by immunoblotting using a streptavidin-horseradish peroxidase conjugate (Pierce, 1 μg/ml in 1% bovine serum albumin) and the ECL chemoluminescence detection system (Amersham International, United Kingdom). To determine the sidedness of amino acid residues relative to the membrane, 100 μg of crude membranes were incubated in the presence or absence of 100 μm AMS for 10 min on ice and the membranes were pelleted in a microcentrifuge at 14,000 × g. After a washing step, the membranes were resuspended in phosphate buffer and incubated with 100 μm biotin maleimide for 10 min on ice before quenching the reaction with 10 mmβ-mercaptoethanol. Biotinylated receptor was purified and visualized as described above. The human KDEL receptor Erd2.1 contains four cysteine residues, none of which is strictly conserved in all receptor homologues. There are no cysteines in theS. cerevisiae receptor, indicating that cysteine residues are not strictly required for receptor function (2Semenza J.C. Hardwick K.G. Dean N. Pelham H.R.B. Cell. 1990; 61: 1349-1357Abstract Full Text PDF PubMed Scopus (388) Google Scholar). However, Cys-29 in the human sequence (Fig. 1) is conserved in six out of nine receptor homologues, and a change to alanine at this position reduces the binding activity of the human KDEL receptor approximately 2–3-fold, implying that a mutation at this position has indirect affects on the receptor structure (12Townsley F.M. Wilson D.W. Pelham H.R.B. EMBO J. 1993; 12: 2821-2829Crossref PubMed Scopus (110) Google Scholar). The first objective was to test whether any individual cysteine residue was strictly required for structure or function of the human KDEL receptor. Therefore, site-directed mutagenesis was used to change all four cysteines to serines, and Cys-less and wild-type receptors were expressed in insect cells. We used the baculovirus expression system, as we have recently demonstrated that it can be used to produce large amounts of functional human KDEL receptor (9Scheel A.A. Pelham H.R.B. Biochemistry. 1996; 35: 10203-10209Crossref PubMed Scopus (44) Google Scholar). We expressed a both Myc- and hexahistidine-tagged version of the human Erd2.1 protein (hErd2-Myc-His, see “Experimental Procedures”), since the addition of these tags to the carboxyl terminus of the receptor does not alter the affinity of the receptor for KDEL ligand and Myc-His-tagged receptor can be easily purified in a single step by nickel affinity chromatography (9Scheel A.A. Pelham H.R.B. Biochemistry. 1996; 35: 10203-10209Crossref PubMed Scopus (44) Google Scholar). Crude membranes containing equal amounts of wild-type receptor and a cysteine-less receptor mutant were analyzed for KDEL peptide binding. The binding activity of Cys-less human KDEL receptor is significantly higher than the wild-type receptor (Fig.2 A). This is not due to a change in ligand affinity and thus suggests an increase in the amount of functional receptor, which was estimated to be approximately 2.3-fold as judged by Scatchard analysis (data not shown). Together with earlier data, which showed around 40% of the human KDEL receptor to be functional when expressed in insect cells (9Scheel A.A. Pelham H.R.B. Biochemistry. 1996; 35: 10203-10209Crossref PubMed Scopus (44) Google Scholar), this suggests that more than 90% of the Cys-less KDEL receptor in these preparations is functional. To determine which cysteine residue is responsible for this effect, we mutated three cysteines at a time to serine, leaving one of the endogenous cysteine residues. Fig. 2 A shows that a receptor containing only Cys-29, which is predicted to be in the first cytoplasmic loop, shows a binding activity similar to wild-type receptor. Thus, it is evidently the presence of Cys-29 that causes some of the overexpressed wild-type receptor in insect cells to accumulate in an inactive state. Immunoblotting showed an increase in the amount of receptor dimers for the Cys-29-containing mutant compared with Cys-less receptor (data not shown), suggesting that oxidation of Cys-29 may play a role in receptor inactivation.Figure 2Effect of cysteine replacements on the KDEL binding activity of the human KDEL receptor. A, binding of125I-YTSEKDEL to membranes derived from insect cells containing equal amounts of wild-type and mutant KDEL receptors was determined. Wild-type, human Erd2-Myc-His; ΔC, human Erd2-Myc-His with all four endogenous cysteines (Cys-29, Cys-62, Cys-70, Cys-192) changed to Ser; the numbering of the other mutants refers to the only cysteine residue still present in the receptor sequence, e.g. Cys-29, residues Cys-62, Cys-70, and Cys-192 changed to Ser. B, Western blot analysis of the samples used in A probed with the 9E10 anti-Myc antibody.View Large Image Figure ViewerDownload Hi-res image Download (PPT) We reintroduced individual cysteine residues into the Cys-less KDEL receptor sequence at 25 different positions by site-directed mutagenesis to probe their function and location (Fig. 1). In the choice of the residues, we were guided by an earlier study, which used mutational analysis to identify amino acid residues crucial for ligand binding (12Townsley F.M. Wilson D.W. Pelham H.R.B. EMBO J. 1993; 12: 2821-2829Crossref PubMed Scopus (110) Google Scholar). Although this approach could not be used to distinguish between residues directly contacting the ligand or ones that merely altered the receptor structure, thus affecting ligand binding indirectly, it allowed us to predict which residues might be involved in forming the binding pocket. We concentrated on charged residues or residues predicted to be on the same side of a potential α-helix as charged or polar residues (Fig.1). Recombinant baculoviruses encoding receptor mutants were used to infect insect cells, and crude membrane preparations were subsequently analyzed for their ability to bind ligands in vitro (TableI). The replacement of charged residues by cysteines had a significant effect on KDEL binding activity, consistent with earlier findings (12Townsley F.M. Wilson D.W. Pelham H.R.B. EMBO J. 1993; 12: 2821-2829Crossref PubMed Scopus (110) Google Scholar). The first half of transmembrane domain 1 was particularly sensitive to mutations, even a replacement of Gly by Cys (Gly-8 → Cys) strongly affected ligand binding, whereas less conserved residues in the same region (Leu-10, Ser-11, and Leu-13) could be changed without significant effect (Table I). In general, mutation of less conserved residues or nonpolar residues had only minor effects on KDEL binding. Surprisingly, a minor change in the fifth transmembrane domain (Ser-123 → Cys) greatly reduced KDEL binding, whereas an identical amino acid mutation in the same transmembrane domain (Ser-128 → Cys) has no effect on the binding activity (see below).Table IEffects of mutations on KDEL binding activityPositionReceptor mutantKDEL binding activity1-a+, less than 30%; ++, 30–60%, +++, greater than 60% of KDEL binding activity compared to the Cys-less receptor.ΔC+++TMD 1Arg-5 → Cys+Gly-8 → Cys+Asp-9 → Cys+Leu-10 → Cys++Ser-11 → Cys+++His-12 → Cys+Leu-13 → Cys+++Loop 1Cys-291-bThese cysteine residues are present in the wild-type receptor and were introduced into a Cys-less receptor mutant. TMD, transmembrane domain.++TMD 2Asp-50 → Cys+Leu-51 → Cys+++Loop 2Asn-54 → C++TMD 3Cys-621-bThese cysteine residues are present in the wild-type receptor and were introduced into a Cys-less receptor mutant. TMD, transmembrane domain.+++Tyr-67 → C+Cys-701-bThese cysteine residues are present in the wild-type receptor and were introduced into a Cys-less receptor mutant. TMD, transmembrane domain.+++Loop 3Thr-85 → Cys+++TMD 4Thr-102 → Cys+++Loop 4Thr-114 → Cys+++TMD 5Ser-123 → Cys+Ser-128 → Cys+++Ala-130 → Cys+++Gln-134 → Cys+Loop 5Thr-146 → Cys+++TMD 6Tyr-162 → Cys+Asn-165 → Cys+Loop 7Gly-174 → Cys+++TMD 7Gln-187 → Cys++Thr-188 → Cys+++Cys-1921-bThese cysteine residues are present in the wild-type receptor and were introduced into a Cys-less receptor mutant. TMD, transmembrane domain.+++Asp-193 → Cys+++1-a +, less than 30%; ++, 30–60%, +++, greater than 60% of KDEL binding activity compared to the Cys-less receptor.1-b These cysteine residues are present in the wild-type receptor and were introduced into a Cys-less receptor mutant. TMD, transmembrane domain. Open table in a new tab Crude membranes containing wild-type or mutant receptors were treated with NEM or biotin maleimide (BCM), and then KDEL binding activity was determined (see Fig.3 for examples). Biotin maleimide is a bulky maleimide-derivative, which was shown to be membrane-permeant at higher concentrations (20Loo T.W. Clarke D.M. J. Biol. Chem. 1995; 270: 843-848Abstract Full Text Full Text PDF PubMed Scopus (261) Google Scholar). This reagent was used to determine whether the size of the modifying reagent is of importance, and it also facilitated detection of labeled receptor. We have identified four receptor mutants (Arg-5 → Cys, Asp-50 → Cys, Tyr-162 → Cys, and Asn-165 → Cys), which upon derivatization with NEM or BCM show significantly reduced binding activity (Fig. 3 and TableII). Binding of KDEL peptides to wild-type or Cys-less KDEL receptor was not affected by derivatization (Fig. 3). Arg-5 → Cys, Asp-50 → Cys, Tyr-162 → Cys, and Asn-165 → Cys already show a significant defect in binding (Table I), which was enhanced when the cysteine residue was alkylated; note that the binding experiments in Fig. 3 used a larger amount of Arg-5 → Cys mutant than wild-type receptor to compensate for its lower binding activity. We consider it to be unlikely that these mutations have a severe effect on folding of the receptor protein, the insertion into the bilayer or protein stability, as their expression levels are comparable with that of wild-type or Cys-less Erd2 (data not shown). Labeling with biotin maleimide was visualized by purification of the KDEL receptor using nickel affinity chromatography followed by immunoblotting and probing with a streptavidin-peroxidase conjugate. Biotin maleimide is specific for cysteine residues under our conditions, since we were unable to detect reaction of BCM with the Cys-less receptor (Fig. 4 A). Arg-5 → Cys, Asp-50 → Cys, Tyr-162 → Cys, and Asn-165 → Cys could be labeled with both BCM and NEM, where labeling with NEM was visualized by its ability to prevent biotinylation (Fig.4 A).Table IIEffects on KDEL binding of receptor mutants by BCM and NEMMutantRelative YTSEKDEL binding after treatment of membranes with500 μm BCM2-aFigures represent the mean of three independent experiments performed in duplicates using constant amounts of125I-YTSEKDEL under nonsaturating conditions (the standard deviation was less than 10%). 100% binding corresponds to KDEL binding in the absence of modifying reagent.500 μm NEM2-aFigures represent the mean of three independent experiments performed in duplicates using constant amounts of125I-YTSEKDEL under nonsaturating conditions (the standard deviation was less than 10%). 100% binding corresponds to KDEL binding in the absence of modifying reagent.%%Cys-less100100Arg-5 → Cys4033Asp-50 → Cys50100Thr-85 → Cys100100Thr-114 → Cys100100Ser-123 → Cys100100Tyr-162 → Cys8044Asn-165 → Cys75352-a Figures represent the mean of three independent experiments performed in duplicates using constant amounts of125I-YTSEKDEL under nonsaturating conditions (the standard deviation was less than 10%). 100% binding corresponds to KDEL binding in the absence of modifying reagent. Open table in a new tab Figure 4Labeling of receptor mutants with BCM and NEM. Crude membranes (100 μg) were incubated in the absence (−) or presence (+) of 100 μm NEM for 10 min on ice. The membranes were washed with phosphate buffer, and incubated with 100 μm BCM for 10 min on ice before quenching with 10 mm β-mercaptoethanol. The KDEL receptor was solubilized and purified as described under “Experimental Procedures.”A, biotinylated KDEL receptor was visualized by immunoblotting and probing with a streptavidin-peroxidase conjugate. The positions of the KDEL receptor and of molecular size markers are indicated. An unrelated protein of approximately 30 kDa was occasionally found to copurify. B, total amounts of KDEL receptor were detected using anti-Myc antiserum.View Large Image Figure ViewerDownload Hi-res image Download (PPT) Binding of KDEL ligands to Arg-5 → Cys was similarly affected by NEM and biotin maleimide. Interestingly, while Asp-50 → Cys was labeled with both compounds (Fig. 4 A), only treatment with biotin maleimide inhibited KDEL binding (Table II), suggesting that a large group added to the cysteine side chain is required to inhibit binding. Inactivation of Tyr-162 → Cys and Asn-165 → Cys was more efficient with NEM, presumably because the labeling efficiency with biotin maleimide was low compared with Arg-5 → Cys or Asp-50 → Cys (Fig.4 A). KDEL binding to receptor mutant Ser-123 → Cys was not further affected by treatment with either reagent (Table II). Thus, the drastic effect of Ser-123 → Cys on KDEL binding is presumably indirect, and even minor changes of the amino acid side chain at this position seem to have severe effects on the receptor conformation. All other receptor mutants could be treated with NEM or biotin maleimide without any effect on KDEL binding (see Table II for examples). We were interested to see whether ligand binding could protect Arg-5 → Cys, Asp-50 → Cys, Tyr-162 → Cys, and Asn-165 → Cys from chemical modification, supporting the localization of these residues to the ligand binding site. However, ligand binding is strongly pH-dependent and is maximal at pH 5 (8Wilson D.W. Lewis M.J. Pelham H.R.B. J. Biol. Chem. 1993; 268: 7465-7468Abstract Full Text PDF PubMed Google Scholar), where we were unable to obtain sufficient biotinylation. We investigated the location of the amino acid residues identified above relative to the membrane by testing the ability of a membrane-impermeant maleimide, stilbenemaleimide disulfonic acid (AMS), to prevent derivatization with biotin maleimide. These experiments crucially depended on intact membrane vesicles with the receptor being oriented in its natural way with the ligand binding side facing the lumen of the vesicles. To test the orientation of the receptor, binding assays were performed in the presence or absence of low amounts of detergent. The addition of 0.3% CHAPS stimulated the binding of 125I-YTSEKDEL to wild-type receptor and all mutants examined at least 4-fold (data not shown). This suggests that 80% of the KDEL binding sites are inaccessible to the hydrophilic peptide without permeabilization of the membrane. We assume that a small fraction of the membrane vesicles are either inverted or slightly leaky, allowing access of the peptide to the ligand binding side in the absence of detergent. AMS has been reported to be a membrane-impermeant thiol-specific compound (20Loo T.W. Clarke D.M. J. Biol. Chem. 1995; 270: 843-848Abstract Full Text Full Text PDF PubMed Scopus (261) Google Scholar) and to confirm its physical properties we chose two mutants, Thr-85 → Cys and Thr-114 → Cys, with changes in positions that are predicted to be on the cytoplasmic and lumenal side of the membrane, respectively. The treatment of single cysteine mutants with AMS prior to incubation with biotin maleimide should prevent the biotinylation of cysteine residues located in the cytoplasmic loops, in contrast to cysteines located on the lumenal side of the membrane. The reaction of biotin maleimide with Thr-85 → Cys was completely prevented by AMS, indicating that residue 85 is indeed located on the cytoplasmic side of the membrane (Fig.5 A). In contrast, the binding of biotin maleimide to Thr-114 → Cys was unaffected by pretreatment with AMS, which locates Thr-114 to the lumenal compartment. This experiment confirms the proposed orientation of Thr-85 and Thr-114 (12Townsley F.M. Wilson D.W. Pelham H.R.B. EMBO J. 1993; 12: 2821-2829Crossref PubMed Scopus (110) Google Scholar,13Singh P. Tang B.L. Wong S.H. Hong W. Mol. Cell. Biol. 1993; 13: 6435-6441Crossref PubMed Scopus (21) Google Scholar). Furthermore, it shows that AMS is indeed sufficiently membrane-impermeant under our conditions to be used to determine the sidedness of individual cysteine residues relative to the membrane. We focused on a mutant containing the endogenous cysteine residue Cys-29 and the mutants Arg-5 → Cys and Asp-50 → Cys, since the topology of the amino-terminal part of the KDEL receptor has remained controversial (12Townsley F.M. Wilson D.W. Pelham H.R.B. EMBO J. 1993; 12: 2821-2829Crossref PubMed Scopus (110) Google Scholar, 13Singh P. Tang B.L. Wong S.H. Hong W. Mol. Cell. Biol. 1993; 13: 6435-6441Crossref PubMed Scopus (21) Google Scholar). As shown in Fig. 5 B, biotinylation of Cys-29 is almost completely blocked by preincubation of the membranes with AMS, strongly suggesting that this residue is located in the cytoplasm. Both Arg-5 → Cys and Asp-50 → Cys behave like Thr-114 → Cys and are not accessible to AMS under these conditions. The biotinylation of all residues studied could be blocked at higher concentrations of AMS and after longer incubation periods, which indicates that these residues are in principle accessible to AMS. Thus, the amino acids Arg-5 and Asp-50, which we suggest to be directly involved in ligand binding are located on the lumenal side of the membrane and are not accessible from the cytoplasm. The tetrapeptide signal used for the retention of soluble ER proteins is conserved among different species, although some variations can be found. On most major ER proteins in higher organisms, KDEL acts as the signal (1Pelham H.R.B. Trends Biochem. Sci. 1990; 15: 483-486Abstract Full Text PDF PubMed Scopus (432) Google Scholar), whereas S. cerevisiae uses HDEL and Kluyveromyces lactis uses both HDEL and DDEL (7Lewis M.J. Sweet D.J. Pelham H.R.B. Cell. 1990; 61: 1359-1363Abstract Full Text PDF PubMed Scopus (163) Google Scholar). Intriguingly, one of the residues described above, Asp-50, which we propose to be involved in ligand binding is located in a region which was shown earlier to be important in determining ligand specificity in the K. lactis and S. cerevisiaereceptors (11Semenza J.C. Pelham H.R.B. J. Mol. Biol. 1992; 224: 1-5Crossref PubMed Scopus (23) Google Scholar). The same region in the human receptor (amino acids 50–56) was shown to be crucial for binding of KDEL ligands (8Wilson D.W. Lewis M.J. Pelham H.R.B. J. Biol. Chem. 1993; 268: 7465-7468Abstract Full Text PDF PubMed Google Scholar, 14Lewis M.J. Pelham H.R.B. Cell. 1992; 68: 353-364Abstract Full Text PDF PubMed Scopus (306) Google Scholar). To determine whether a change of Asp-50 affects ligand specificity, we tested the binding activity of KDEL and DDEL peptides to the Asp-50 → Cys mutant. Octapeptides were shown to bind with higher affinity than tetrapeptides, 2A. Scheel, unpublished observation. presumably because sequences upstream of the tetrapeptide signal contribute to the interaction with the receptor. To eliminate these additional effects, we used tetrapeptides as competitors for binding of radioactively labeled YTFEHDEL. Binding of 125I-YTFEHDEL to the Cys-less receptor was competed efficiently with the peptide KDEL, whereas the peptide DDEL competed poorly (Fig.6 A), consistent with earlier findings (8Wilson D.W. Lewis M.J. Pelham H.R.B. J. Biol. Chem. 1993; 268: 7465-7468Abstract Full Text PDF PubMed Google Scholar). Binding of DDEL peptides to Asp-50 → Cys was comparable with binding to the Cys-less receptor; a concentration of approximately 250 μm DDEL was required to compete 50% of125I-YTFEHDEL bound to the Cys-less mutant or Asp-50 → Cys (Fig. 6). Strikingly, KDEL was a substantially worse competitor with the Asp-50 → Cys mutant than with the Cys-less receptor (Fig.6 B). The same effect, which was specific for Asp-50 → Cys, was observed for HDEL and RDEL peptides. These results suggest that the efficient binding of peptides with a positive charge at position −4 from the carboxyl terminus of the ligand (HDEL, KDEL, RDEL) requires the aspartic acid at position 50, presumably involving a direct ionic interaction. This paper describes an investigation of the three-dimensional organization of the human KDEL receptor Erd2.1, using a site-directed sulfhydryl labeling approach. We show that all four endogenous Cys residues of the human KDEL receptor can be removed without affecting the ligand binding activity. Furthermore, the removal of cysteine 29 leads to the expression of almost homogeneously active receptor protein in Sf9 cells. Together with an earlier study, in which we describe the high level expression and the purification of functional receptor protein (9Scheel A.A. Pelham H.R.B. Biochemistry. 1996; 35: 10203-10209Crossref PubMed Scopus (44) Google Scholar), this will prove to be a significant step toward the determination of the receptor structure by two-dimensional crystallization and cryo-electron microscopy. We have investigated the topology of the KDEL receptor by testing the sidedness of amino acids relative to the lipid bilayer using membrane-permeant and -impermeant thiol-specific reagents. This was of particular interest since the topology of the amino-terminal half of the KDEL receptor has been controversial (12Townsley F.M. Wilson D.W. Pelham H.R.B. EMBO J. 1993; 12: 2821-2829Crossref PubMed Scopus (110) Google Scholar, 13Singh P. Tang B.L. Wong S.H. Hong W. Mol. Cell. Biol. 1993; 13: 6435-6441Crossref PubMed Scopus (21) Google Scholar). Using this method, we have demonstrated that both Arg-5 → Cys and Asp-50 → Cys are not accessible to membrane-impermeant reagents, indicating that these residues are located in the lumen of the membrane compartments and contribute to the formation of a hydrophilic binding site for KDEL ligands, which is not accessible from the cytoplasmic side of the membrane. Furthermore, Cys-29 was shown to be located on the cytoplasmic side of the membrane. This suggests that the amino terminus is located in the lumen as proposed previously (12Townsley F.M. Wilson D.W. Pelham H.R.B. EMBO J. 1993; 12: 2821-2829Crossref PubMed Scopus (110) Google Scholar). Together with the location of Thr-85 → Cys and Thr-114 → Cys, which were shown to be located on the cytoplasmic and lumenal side of the membrane, respectively, these findings strongly point toward a two-dimensional organization of the KDEL receptor analogous to the family of G-protein-coupled receptors with seven membrane-spanning segments. By determining the binding activity of single cysteine mutants of the human KDEL receptor after reaction with thiol-specific reagents, we have identified four residues that we believe contribute to the formation of a hydrophilic binding pocket for KDEL ligands (Arg-5, Asp-50, Tyr-162, and Asn-165). The KDEL binding activity of these mutants is considerably decreased compared with the Cys-less receptor and alkylation of these cysteines further inhibits KDEL binding. This suggests that these four residues play a critical role in KDEL binding and that the alkylation of the cysteine replacing the wild-type residue leads to steric hindrance of ligand binding. We do not believe that these mutations cause a significant change in receptor conformation, as these mutants retain substantial KDEL binding activity and no differences in expression levels could be observed compared with the wild-type receptor. Furthermore, mutations of these residues described earlier (Arg-5 → Gln, Asp-50 → Asn, and Asn-165 → Ala) do not interfere with the proper localization of these mutants to the Golgi apparatus and no accumulation of the mutants in the ER, indicative of misfolded proteins, was observed (12Townsley F.M. Wilson D.W. Pelham H.R.B. EMBO J. 1993; 12: 2821-2829Crossref PubMed Scopus (110) Google Scholar). The reaction of Tyr-162 → Cys and Asn-165 → Cys with biotin maleimide was less efficient compared with Arg-5 → Cys and Asp-50 → Cys, presumably because they are located closer to the inside of the hydrophilic binding pocket, making their side chains less accessible to a bulkier reagent such as BCM, whereas a smaller reagent, NEM, was still efficient in inhibiting binding of YTSEKDEL. Arg-5 → Cys was equally well labeled by both NEM and biotin maleimide, and residual KDEL binding to Arg-5 → Cys was efficiently inhibited by both reagents. Strikingly, all four residues are predicted to be located at the lumenal end of the transmembrane domains based on a model with seven transmembrane domains (see Fig. 1). Asp-50 → Cys is a particularly interesting mutant. KDEL binding, which is substantially reduced compared with the Cys-less receptor, could only be inhibited by reaction with biotin maleimide, although labeling of this mutant with both NEM and biotin maleimide was efficient. This suggests that this residue is located closer to the opening of the binding pocket than Arg-5 → Cys, and a large group added to the cysteine side chain is required to perturb peptide binding. Furthermore, we have studied the effects of this mutation on the specificity of ligand binding. As mentioned above, the yeastK. lactis uses both HDEL and DDEL as retention signals, whereas the S. cerevisiae receptor only recognizes HDEL. Studies in yeast suggest that amino acids 51–57 in the K. lactis receptor determine ligand specificity in vivo(11Semenza J.C. Pelham H.R.B. J. Mol. Biol. 1992; 224: 1-5Crossref PubMed Scopus (23) Google Scholar). Similar results were obtained when the corresponding residues (amino acids 50–56) in the human receptor were replaced with theK. lactis sequence, which significantly decreased the ability of the human receptor to bind KDEL ligands but not DDEL ligands, both in vitro and in vivo (8Wilson D.W. Lewis M.J. Pelham H.R.B. J. Biol. Chem. 1993; 268: 7465-7468Abstract Full Text PDF PubMed Google Scholar, 14Lewis M.J. Pelham H.R.B. Cell. 1992; 68: 353-364Abstract Full Text PDF PubMed Scopus (306) Google Scholar). We have demonstrated that Asp-50 in the human receptor sequence is required for the efficient binding of KDEL ligands; the binding of KDEL (and other peptides with a positive charge at the position −4 such as HDEL and RDEL) to Asp-50 → Cys was strongly reduced, whereas the binding of DDEL peptides was unaffected. This suggests that DDEL, at least partly, interacts with different amino acids than H/KDEL as proposed previously (11Semenza J.C. Pelham H.R.B. J. Mol. Biol. 1992; 224: 1-5Crossref PubMed Scopus (23) Google Scholar). The fact that binding of peptides with a positive charge (His, Lys, or Arg) is specifically weakened by a mutation in the receptor that changes a negatively charged residue to a polar residue implicates a direct interaction. Close proximity between this residue and the ligand is also supported by the finding that ligand binding to Asp-50 → Cys was inhibited by alkylation of the cysteine side chain (see above). Strikingly, Asp-50 is conserved in all Erd2 sequences known except for the K. lactis receptor. The importance of this residue is further demonstrated by the fact that one of the receptor mutants originally isolated in S. cerevisiaewas localized to the equivalent residue in the yeast receptor, changing the aspartate to an asparagine (2Semenza J.C. Hardwick K.G. Dean N. Pelham H.R.B. Cell. 1990; 61: 1349-1357Abstract Full Text PDF PubMed Scopus (388) Google Scholar). However, there must be other interactions significantly contributing to ligand binding, in particular to the binding of HDEL peptides; the K. lactisreceptor, which has an asparagine residue at the corresponding location, binds HDEL ligands in vivo, but not KDEL ligands. In addition, the binding of HDEL was substantially worsened by the Asp-50 → Cys mutation, but it still binds considerably tighter to Asp-50 → Cys than all other peptides tested, suggesting additional interactions of the histidine residue both in the human and theK. lactis receptor. Furthermore, our competition experiments were performed with tetrapeptides. They bind with lower affinity to the receptor than the corresponding octapeptides, indicating that interactions with residues upstream of the conserved tetrapeptide contribute to ligand binding, although no sequence conservation among different ER proteins can be observed (1Pelham H.R.B. Trends Biochem. Sci. 1990; 15: 483-486Abstract Full Text PDF PubMed Scopus (432) Google Scholar). Additionally, soluble ER proteins have been identified without a positively charged residue at position −4 (1Pelham H.R.B. Trends Biochem. Sci. 1990; 15: 483-486Abstract Full Text PDF PubMed Scopus (432) Google Scholar). In conclusion, our study provides insight into the three-dimensional organization of the KDEL receptor; we present evidence for the existence of a hydrophilic ligand binding pocket formed by conserved amino acid residues of the transmembrane domains. Together with earlier findings (12Townsley F.M. Wilson D.W. Pelham H.R.B. EMBO J. 1993; 12: 2821-2829Crossref PubMed Scopus (110) Google Scholar, 13Singh P. Tang B.L. Wong S.H. Hong W. Mol. Cell. Biol. 1993; 13: 6435-6441Crossref PubMed Scopus (21) Google Scholar), our data suggest a secondary structure of the KDEL receptor with seven membrane-spanning domains, similar to the family of G-protein-coupled receptors. Furthermore, we have shown that Asp-50 is involved in determining the ligand specificity, presumably by directly pairing with a positively charged amino acid at position −4 from the COOH terminus of the ligand." @default.
W2000504183 created "2016-06-24" @default.
W2000504183 creator A5053567351 @default.
W2000504183 creator A5079274133 @default.
W2000504183 date "1998-01-01" @default.
W2000504183 modified "2023-09-26" @default.
W2000504183 title "Identification of Amino Acids in the Binding Pocket of the Human KDEL Receptor" @default.
W2000504183 cites W1482546821 @default.
W2000504183 cites W1536827165 @default.
W2000504183 cites W1963912518 @default.
W2000504183 cites W1972341173 @default.
W2000504183 cites W1983464528 @default.
W2000504183 cites W1983986068 @default.
W2000504183 cites W1987955769 @default.
W2000504183 cites W2003144507 @default.
W2000504183 cites W2019113077 @default.
W2000504183 cites W2030463413 @default.
W2000504183 cites W2040966208 @default.
W2000504183 cites W2062876904 @default.
W2000504183 cites W2069116829 @default.
W2000504183 cites W2095429868 @default.
W2000504183 cites W2133110610 @default.
W2000504183 cites W2166247822 @default.
W2000504183 cites W221247672 @default.
W2000504183 cites W2888511473 @default.
W2000504183 doi "https://doi.org/10.1074/jbc.273.4.2467" @default.
W2000504183 hasPubMedId "https://pubmed.ncbi.nlm.nih.gov/9442098" @default.
W2000504183 hasPublicationYear "1998" @default.
W2000504183 type Work @default.
W2000504183 sameAs 2000504183 @default.
W2000504183 citedByCount "47" @default.
W2000504183 countsByYear W20005041832012 @default.
W2000504183 countsByYear W20005041832013 @default.
W2000504183 countsByYear W20005041832015 @default.
W2000504183 countsByYear W20005041832016 @default.
W2000504183 countsByYear W20005041832017 @default.
W2000504183 countsByYear W20005041832018 @default.
W2000504183 countsByYear W20005041832019 @default.
W2000504183 countsByYear W20005041832020 @default.
W2000504183 countsByYear W20005041832021 @default.
W2000504183 countsByYear W20005041832022 @default.
W2000504183 countsByYear W20005041832023 @default.
W2000504183 crossrefType "journal-article" @default.
W2000504183 hasAuthorship W2000504183A5053567351 @default.
W2000504183 hasAuthorship W2000504183A5079274133 @default.
W2000504183 hasBestOaLocation W20005041831 @default.
W2000504183 hasConcept C104317684 @default.
W2000504183 hasConcept C116834253 @default.
W2000504183 hasConcept C170493617 @default.
W2000504183 hasConcept C185592680 @default.
W2000504183 hasConcept C2779377555 @default.
W2000504183 hasConcept C40767141 @default.
W2000504183 hasConcept C515207424 @default.
W2000504183 hasConcept C55493867 @default.
W2000504183 hasConcept C59822182 @default.
W2000504183 hasConcept C86803240 @default.
W2000504183 hasConceptScore W2000504183C104317684 @default.
W2000504183 hasConceptScore W2000504183C116834253 @default.
W2000504183 hasConceptScore W2000504183C170493617 @default.
W2000504183 hasConceptScore W2000504183C185592680 @default.
W2000504183 hasConceptScore W2000504183C2779377555 @default.
W2000504183 hasConceptScore W2000504183C40767141 @default.
W2000504183 hasConceptScore W2000504183C515207424 @default.
W2000504183 hasConceptScore W2000504183C55493867 @default.
W2000504183 hasConceptScore W2000504183C59822182 @default.
W2000504183 hasConceptScore W2000504183C86803240 @default.
W2000504183 hasIssue "4" @default.
W2000504183 hasLocation W20005041831 @default.
W2000504183 hasOpenAccess W2000504183 @default.
W2000504183 hasPrimaryLocation W20005041831 @default.
W2000504183 hasRelatedWork W1927591950 @default.
W2000504183 hasRelatedWork W1987955769 @default.
W2000504183 hasRelatedWork W1994546225 @default.
W2000504183 hasRelatedWork W2044628742 @default.
W2000504183 hasRelatedWork W2125533158 @default.
W2000504183 hasRelatedWork W2148960370 @default.
W2000504183 hasRelatedWork W2170754052 @default.
W2000504183 hasRelatedWork W4205670493 @default.
W2000504183 hasRelatedWork W4323248528 @default.
W2000504183 hasRelatedWork W4379618501 @default.
W2000504183 hasVolume "273" @default.
W2000504183 isParatext "false" @default.
W2000504183 isRetracted "false" @default.
W2000504183 magId "2000504183" @default.
W2000504183 workType "article" @default.