Advertisement

Additional N-Glycosylation and Its Impact on the Folding of Intestinal Lactase-phlorizin Hydrolase*

  • Ralf Jacob
    Affiliations
    Department of Physiological Chemistry, School of Veterinary Medicine Hannover, D-30559 Hannover, Germany
    Search for articles by this author
  • Jocelyn R. Weiner
    Affiliations
    Department of Physiological Chemistry, School of Veterinary Medicine Hannover, D-30559 Hannover, Germany
    Search for articles by this author
  • Stephanie Stadge
    Affiliations
    Department of Physiological Chemistry, School of Veterinary Medicine Hannover, D-30559 Hannover, Germany
    Search for articles by this author
  • Hassan Y. Naim
    Correspondence
    To whom correspondence should be addressed: Dept. of Physiological Chemistry, School of Veterinary Medicine, Hannover, Bünteweg 17, D-30559 Hannover, Germany. Tel.: 49-511-9538780; Fax: 49-511-9538585
    Affiliations
    Department of Physiological Chemistry, School of Veterinary Medicine Hannover, D-30559 Hannover, Germany
    Search for articles by this author
  • Author Footnotes
    * This work was supported by Deutsche Forschungsgemeinschaft (DFG), Bonn, Germany, Grant Na 331/1-2 (to H. Y. N.) and Sonderforschungsbereich (SFB) 280.The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
      Lactase-phlorizin hydrolase (LPH) is a membrane bound intestinal hydrolase, with an extracellular domain comprising 4 homologous regions. LPH is synthesized as a large polypeptide precursor, pro-LPH, that undergoes several intra- and extracellular proteolytic steps to generate the final brush-border membrane form LPHβfinal. Pro-LPH is associated through homologous domain IV with the membrane through a transmembrane domain. A truncation of 236 amino acids at the COOH terminus of domain IV (denoted LAC236) does not significantly influence the transport competence of the generated mutant LPH1646MACT (Panzer, P., Preuss, U., Joberty, G., and Naim, H. Y. (1998) J. Biol. Chem. 273, 13861–13869), strongly suggesting that LAC236 is an autonomously folded domain that links the ectodomain with the transmembrane region. Here, we examine this hypothesis by engineering several N-linked glycosylation sites into LAC236. Transient expression of the cDNA constructs in COS-1 cells confirm glycosylation of the introduced sites. The N-glycosyl pro-LPH mutants are transported to the Golgi apparatus at substantially reduced rates as compared with wild-type pro-LPH. Alterations in LAC236 appear to sterically hinder the generation of stable dimeric trypsin-resistant pro-LPH forms. Individual expression of chimeras containing LAC236, the transmembrane domain and cytoplasmic tail of pro-LPH and GFP as a reporter gene (denoted LAC236-GFP) lends strong support to this view: while LAC236-GFP is capable of forming dimersper se, its N-glycosyl variants are not. The data strongly suggest that the LAC236 is implicated in the dimerization process of pro-LPH, most likely by nucleating the association of the ectodomains of the enzyme.
      ER
      endoplasmic reticulum
      LPH
      lactase-phlorizin hydrolase (all forms)
      pro-LPH
      uncleaved precursor of LPH
      mAb
      monoclonal antibody
      pro-LPHh
      mannose-rich precursor
      pro-LPHc
      complex glycosylated precursor
      LAC236
      a stretch of 236 amino acids (Arg1647-Thr1882) located juxtapose the membrane anchoring domain of pro-LPH
      GFP
      green fluorescent protein
      PAGE
      polyacrylamide gel electrophoresis
      endo
      endoglycosidase
      TEMED
      N,N,N′,N′-tetramethylethylenediamine
      The pathways by which membrane and secretory proteins attain their three-dimensional structure, in particular the implication of glycosylation in these events has been the target of extensive investigation in the past few years. N-Glycosylation is necessary for efficient folding in the ER1 (
      • Fan H.
      • Meng W.
      • Kilian C.
      • Grams S.
      • Reutter W.
      ,
      • Letourneur O.
      • Sechi S.
      • Willette-Brown J.
      • Robertson M.W.
      • Kinet J.P.
      ,
      • Riederer M.A.
      • Hinnen A.
      ,
      • Zhang Y.
      • Dahms N.M.
      ,
      • Roberts P.C.
      • Garten W.
      • Klenk H.D.
      ), interaction with calnexin and calreticulin (
      • Cannon K.S.
      • Hebert D.N.
      • Helenius A.
      ), receptor-ligand binding (
      • Zhang R.
      • Cai H.
      • Fatima N.
      • Buczko E.
      • Dufau M.L.
      ), transport to lysosomes (
      • Chapman R.L.
      • Kane S.E.
      • Erickson A.H.
      ), polarized sorting to the apical plasma membrane (
      • Scheiffele P.
      • Peranen J.
      • Simons K.
      ), and also for optimal expression of some proteins (
      • Shen F.
      • Wang H.
      • Zheng X.
      • Ratnam M.
      ). The addition ofN-linked core oligosaccharides to membrane and secretory glycoproteins occurs co-translationally at asparagine residues in the tripeptide sequon Asn-Xaa-Ser/Thr soon after translocation of the nascent polypeptide into the lumen of the endoplasmic reticulum (ER). However, the presence of the sequon does not ensure core glycosylation, as many proteins contain sequons that remain either unglycosylated or glycosylated to a variable extent (
      • Allen S.
      • Naim H.Y.
      • Bulleid N.J.
      ). N-Glycosyl sugar chains are core-glycosylated in the ER and become complex glycosylated by passing through the Golgi apparatus. AlthoughN-glycosylation in the ER constitutes the critical step in the initial folding of proteins, the complex glycosylated chains in some glycoproteins, acquired in the Golgi, are also required for the acquisition of a correct conformation (
      • Loo T.W.
      • Clarke D.M.
      ). Several roles forN-linked oligosaccharides in protein folding have been described so far. Their presence assists folding by facilitating disulfide bond formation (
      • Allen S.
      • Naim H.Y.
      • Bulleid N.J.
      ,
      • Feng W.
      • Matzuk M.M.
      • Mountjoy K.
      • Bedows E.
      • Ruddon R.W.
      • Boime I.
      ,
      • Rickert K.W.
      • Imperiali B.
      ) or through a chaperone-mediated glucose trimming and reglycosylation cycle in the ER (
      • Hebert D.N.
      • Foellmer B.
      • Helenius A.
      ). The absence of N-glycosyl chains influences the quaternary structure of proteins as has been demonstrated for the dimerization of MUC2 mucin (
      • Asker N.
      • Axelsson M.A.B.
      • Olofsson S.-O.
      • Hansson G.C.
      ). Also highly glycosylated protein domains, like the carboxyl-terminal peptide (CTP) of the human placental hormone CG-β subunit, can participate in the folding of the whole subunit (
      • Muyan M.
      • Boime I.
      ).
      We used human lactase-phlorizin hydrolase (EC 3.2.1.23–3.2.1.62) (LPH) as the appropriate model protein to study the influence ofN-glycosyl sugar chains on the folding of single protein domains. This enzyme is synthesized in intestinal cells as a pro-LPH that undergoes one intracellular proteolytic cleavage in the Golgi apparatus to generate a mature LPHβinitial that is targeted with high fidelity to the apical membrane. A final cleavage step takes place in the intestinal lumen by trypsin to generate LPHβfinal that exerts its biological function in hydrolyzing lactose. Pro-LPH consists of four homologous regions with 38–55% homology and is associated with the plasma-membrane via a membrane anchor at the COOH terminus of the LPHβfinaldomain (
      • Mantei N.
      • Villa M.
      • Enzler T.
      • Wacker H.
      • Boll W.
      • James P.
      • Hunziker W.
      • Semenza G.
      ). The different protein domains are likely involved together or independently in the folding of pro-LPH. Most notably is the critical role of the profragment, LPHα, in the maturation of LPHβinitial (
      • Wuthrich M.
      • Grunberg J.
      • Hahn D.
      • Jacob R.
      • Radebach I.
      • Naim H.Y.
      • Sterchi E.E.
      ,
      • Jacob R.
      • Radebach I.
      • Wuthrich M.
      • Grunberg J.
      • Sterchi E.E.
      • Naim H.Y.
      ). Deletion of this NH2-terminal domain leads to a drastic reduction in the proportion of correctly folded and transport-competent molecules (
      • Naim H.Y.
      • Jacob R.
      • Naim H.
      • Sambrook J.F.
      • Gething M.J.
      ,
      • Oberholzer T.
      • Mantei N.
      • Semenza G.
      ). Apart from the LPHα domain, the transmembrane and perhaps also the cytoplasmic domains of pro-LPH play a key role in acquisition of a correct dimeric protein structure (
      • Naim H.Y.
      • Naim H.
      ) and transport competence of the enzyme (
      • Panzer P.
      • Preuss U.
      • Joberty G.
      • Naim H.Y.
      ). The homologous region IV contains a subdomain with structural characteristics of unfolded, highlyO-glycosylated stalk regions found in a variety of membrane proteins such as low density lipoprotein receptor, sucrase isomaltase, and aminopeptidase N (
      • Hunziker W.
      • Spiess M.
      • Semenza G.
      • Lodish H.F.
      ,
      • Olsen J.
      • Cowell G.M.
      • Konigshofer E.
      • Danielsen E.M.
      • Moller J.
      • Laustsen L.
      • Hansen O.C.
      • Welinder K.G.
      • Engberg J.
      • Hunziker W.
      ). This domain harbors the catalytic site of the lactose hydrolytic activity. We could show that deletion of 236 amino acids located in the immediate vicinity of the membrane in homologous region IV had almost no influence on the quaternary structure, transport, and sorting of LPH to the plasma membrane (
      • Panzer P.
      • Preuss U.
      • Joberty G.
      • Naim H.Y.
      ). By contrast, a further deletion of 87 amino acids upstream of the 236-amino acid stretch have led to an inhibition of the dimerization event and to a substantial reduction in the transport capacity of the deletion mutant. This suggested that the last 236 amino acids of domain IV fold independently of the extracellular domain and might function as a link between the globular protein and the membrane by forming a stalk region. Structural predictions of the first 20 amino acids of the deleted region reveal an α-helical structure similar to the stalk of aminopeptidase-N (
      • Vogel L.K.
      • Noren O.
      • Sjostrom H.
      ). Additionally, according to data derived from a survey of large number of O-glycosylation sites (
      • Wilson I.B.
      • Gavel Y.
      • von Heijne G.
      ), the first 50 amino acids of the deleted region contain several potentialO-glycosylation sites, which are presumably glycosylated in the mature form of LPH (amino acids 869–1927) (
      • Naim H.Y.
      • Lentze M.J.
      ). In this paper we addressed the role of this domain in the context of the overall folding of pro-LPH by introducing potential N-glycosylation sites into this domain. We demonstrate that the mutated constructs are at least partially misfolded and transported to the cell surface at markedly lower rates than the wild-type species. The effect of the additional N-glycosylation sites is in all likelihood limited to the folding of a region including the 236 amino acids stretch, but nevertheless implicates the dimerization event and subsequent trafficking of the mutants.

      EXPERIMENTAL PROCEDURES

       Materials and Reagents

      Tissue culture reagents, streptomycin, penicillin, Dulbecco's modified Eagle's medium (DMEM), and methionine-free DMEM were purchased from Life Technologies, Inc. Pepstatin, leupeptin, aprotinin, and molecular weight standards for SDS-PAGE were purchased from Sigma.l-[35S]Methionine (>800 Ci/mmol) and protein A-Sepharose were obtained from Amersham Pharmacia Biotech. Acrylamide and N,N′-methylenebisacrylamide were obtained from Carl Roth GmbH & Co., Karlsruhe, Germany. SDS, TEMED, ammonium persulfate, dithiothreitol, and Triton X-100 were obtained from Merck, Darmstadt, Germany. Endo-β-N-acetylglucosaminidase H (endo H), endo-β-N-acetylglucosaminidase F (containing N-glycosidase F (or N-glycanase) (endo F/GF), and phenylmethanesulfonyl fluoride were purchased from Roche Diagnostics (Mannheim). All other reagents were of superior analytical grade.

       Reagents for Recombinant DNA Techniques

      Restriction enzymes, T4 polymerase, and ligase were obtained from New England Biolabs. Oligonucleotide-directed mutagenesis was performed with the “Altered SitesTM in Vitro mutagenesis System” from Promega Corp., Madison, WI. The following oligonucleotides were used in this context: 5′-CAAAAGAAGAGTTGGCAGTGGCAT-3′ for the LPHI1697N, 5′-GCCACGAGCGATTTGCGATGGAAGC-3′ for the LPHD1711N, 5′-CATAAATTGGAGAGTCATTGTATTC-3′ for the LPHP1743S.

       Addition of Potential N-Glycosylation Sites in the LPHcDNA

      The plasmid pLPH containing the full-length cDNA of LPH inserted into the unique EcoRI site of vector pGEM4Z (
      • Naim H.Y.
      • Lacey S.W.
      • Sambrook J.F.
      • Gething M.J.
      ) was used for subcloning of the distalEcoRI/PstI fragment of LPH into the vector pSelect (Promega). This construct was the template for all three mutagenesis reactions with the Altered SitesTM in Vitro Mutagenesis System. Mutagenized plasmids were sequenced to confirm the mutagenesis reaction and cloned back into the full-length LPH cDNA. For expression in COS-1 cells, the pSG8-LPHI1697N, pSG8-LPHD1711N, and pSG8-LPHP1743S expression vectors were constructed by ligating the full-length LPH cDNA clones into the uniqueEcoRI site of the pSG8 vector.

       Generation of GFP Constructs

      TheEcoRI/ScaI fragment from pLPH (
      • Naim H.Y.
      • Lacey S.W.
      • Sambrook J.F.
      • Gething M.J.
      ) containing full-length LPH was cloned in-frame into the polycloning site of peGFP-N1 (CLONTECH) to create pLPH-GFP, encoding the cDNA for a pro-LPH molecule with COOH-terminal fused GFP (pro-LPH-GFP). For the generation of the LAC236-GFP series the signal sequence of LPH, LPHsignal, was synthesized at first by polymerase chain reaction using full-length LPH cDNA as template as described (
      • Naim H.Y.
      • Jacob R.
      • Naim H.
      • Sambrook J.F.
      • Gething M.J.
      ). Another cDNA comprising the last 20 nucleotides of LPHsignal (nucleotides 52–71) and nucleotides 4955 to 4974 of pro-LPH were synthesized by polymerase chain reaction using pLPH, pSG8-LPHI1697N, pSG8-LPHD1711N, or pSG8-LPHP1743S as template and the following oligonucleotides: LPHsig/1646, GTTTTTCATGCTGGGGGTCACGTGACAGGAGCTTGGCTGC, and cLPH6869, CTCTAACGGTGCAGCAGGAC. The resulting DNA molecules were fused to LPHsignal in four different assembly polymerase chain reactions. The reaction products were each digested withEcoRI/ScaI and cloned in-frame into the polycloning site of peGFP-N1 (CLONTECH) to create pLAC236wt-GFP, pLAC236I1697N-GFP, pLAC236D1711N-GFP, and pLAC236P1743S-GFP. The sequence of the constructs was confirmed by DNA sequencing.

       Transfection, Biosynthetic Labeling of Transfected Cells, and Immunoprecipitation of Cell Extracts

      COS-1-cells were cultured in DMEM supplemented with 10% fetal calf serum, 50 units/ml penicillin, and 50 mg/ml streptomycin (denoted complete medium). They were transfected without DNA (mock) or 2 μg of the appropriate recombinant DNA using DEAE-dextran essentially as described before (
      • Naim H.Y.
      • Lacey S.W.
      • Sambrook J.F.
      • Gething M.J.
      ). Transiently transfected COS-1 cells were labeled 48–60 h post-transfection with 80 μCi of [35S]methionine in methionine-free DMEM containing 2% fetal calf serum, 50 units/ml penicillin, and 50 mg/ml streptomycin (denoted Met-free medium). In pulse-chase experiments labeling was performed for 1 h at 37 °C followed by a chase with non-labeled methionine for different periods of time. The labeled cells were rinsed two times with phosphate-buffered saline. Cells were solubilized with 1 ml/dish of cold lysis buffer essentially as described before (
      • Panzer P.
      • Preuss U.
      • Joberty G.
      • Naim H.Y.
      ). The cell extracts were centrifuged to remove nuclei and debris. Thereafter the supernatants were incubated with mouse anti-human lactase-phlorizin hydrolase monoclonal antibody HBB 1/909/34/74 (
      • Hauri H.P.
      • Sterchi E.E.
      • Bienz D.
      • Fransen J.A.
      • Marxer A.
      ) from Dr. H.-P. Hauri (Biocenter, Basel, Switzerland) or with the mouse anti-GFP antibody B34 (Babco) and precipitated with protein A-Sepharose. Following immunoprecipitation, the protein A-Sepharose beads were washed 3 times with washing buffer A (0.5% Triton X-100, 0.05% sodium deoxycholate in phosphate-buffered saline) and 3 times with washing buffer B (500 mm NaCl, 10 mm EDTA, 0.5% Triton X-100 in 125 mm Tris/HCl, pH 8.0) prior to analysis of the samples by SDS-PAGE and fluorography. Cells transfected with constructs containing GFP were also fixed with 3% paraformaldehyde 48 h post-transfection and the recombinant proteins were visualized by confocal laser microscopy (Leica TCS SP).

       Enzymatic Assays

      Lactase activity was measured according to Dahlqvist (
      • Dahlqvist A.
      ) using lactose (Fluka) as substrate. Detergent extracts of transfected COS-1 cells (10 dishes of confluent cells) were used for immunoprecipitation of LPH. The immunoprecipitates were subsequently assayed for lactase activity essentially as described by Naim et al. (
      • Naim H.Y.
      • Lacey S.W.
      • Sambrook J.F.
      • Gething M.J.
      ).

       Trypsin Treatment of Immunoprecipitates

      Immunoprecipitates from biosynthetically labeled COS-1 cells were treated with 0.5 mg of the pancreatic protease trypsin at 37 °C for 0, 5, or 15 min, and the reaction was stopped by the addition of 2 mg of soybean trypsin inhibitor (Roche Molecular Biochemicals) at 4 °C followed by boiling in SDS-PAGE sample buffer.

       Cell Lysate Fractionation on Sucrose Density Gradients

      According to Ref.
      • Panzer P.
      • Preuss U.
      • Joberty G.
      • Naim H.Y.
      , transfected COS-1 cells were labeled 48 h post-transfection for 6 h with [35S]methionine, washed two times with phosphate-buffered saline, and solubilized in 500 μl of 6 mm n-dodecyl-β-d-maltoside, 50 mmTris/HCl, 100 mm NaCl, pH 7.4, and a mixture of protease inhibitors. To remove nuclei and debris the cell extracts were centrifuged for 20 min at 17,000 × g at 4 °C. The supernatants were layered onto a continuous gradient from 10 to 30% or 5 to 25% (w/v) sucrose. After centrifugation for 22 h at 100,000 × g at 4 °C, different fractions were collected and immunoprecipitated with mAb anti-LPH or mAb anti-GFP and analyzed on 6% SDS-PAGE.

       Other Procedures

      Digestion of 35S-labeled immunoprecipitates with endo-β-N-acetylglucosaminidase H (endo H) and endo-β-N-acetylglucosaminidase F/glycopeptidase F (endo F/GF) was performed as described previously (
      • Naim H.Y.
      • Sterchi E.E.
      • Lentze M.J.
      ).

      RESULTS

       Amino Acid Exchange at Three Different Positions in the Primary Sequence of LPH Yields Additional N-Glycosyl Sugar Chains

      Truncation of 236 amino acids immediately upstream of the membrane anchor of LPH did not affect the quaternary structure, transport competence, or polarized sorting of the enzyme (
      • Panzer P.
      • Preuss U.
      • Joberty G.
      • Naim H.Y.
      ). These data suggested that this stretch (LAC236), which lies in the homologous region IV of LPH, is independently folded and is probably a linker between the transmembrane of LPH and the remainder of its ectodomain. To investigate this hypothesis and explore further structural features and the role of this region within the ectodomain of LPH, we introduced mutations that result in potential N-glycosylation sites (sequon Asn-X-Ser/Thr). The mutants contained one of the following amino acid substitutions: Ile1697 to Asn1697 (LPHI1697N), Asp1711 to Asn1711 (LPHD1711N), and Pro1743 to Ser1743 (LPHP1743S) (Fig.1). The influence of these mutations on the biosynthesis, folding, and transport of pro-LPH was investigated in COS cells transfected with expression vectors containing the mutated LPH cDNAs. Fig. 2 shows an SDS-PAGE analysis of the immunoprecipitated mutants. Within 6 h of biosynthetic labeling each mutant revealed two polypeptides. Treatment of these forms with endo H demonstrated the type of glycosylation of these forms. The higher molecular mass species (∼230 kDa) were resistant to endo H indicating their complex type of glycosylation, while the smaller polypeptides (∼215 kDa) were sensitive and shifted to approximately 200-kDa band (Fig. 2 B). In analogy with wild-type pro-LPH, these forms correspond to the mannose-rich (∼215 kDa) and complex glycosylated (∼230 kDa) species and could both be deglycosylated by treatment with endo F/GF. The introduced potentialN-glycosylation sites in pro-LPHI1697N, pro-LPHD1711N, and pro-LPHP1743S could be electrophoretically demonstrated to be indeed glycosylated when compared with wild-type pro-LPH. The mannose-rich polypeptides as well as their complex glycosylated forms revealed slight shifts in their apparent molecular masses as compared with their wild-type counterparts (indicated by the asterisk). In contrast to the complex glycosylated counterparts, the endo H-treated deglycosylated mannose-rich forms of the mutants and wild-type pro-LPH shifted to identical apparent molecular weights. Variations in the apparent molecular weights of the complex glycosylated pro-LPH glycoforms could be detected, whereby wild-type pro-LPH was smaller than each of the mutated proteins (pro-LPHc versuspro-LPHc*). Since the N-deglycosylated mannose-rich pro-LPH molecules are all of the same size, an increase in molecular weight is the consequence of additionalN-glycosylation at the added sites. In addition to size shifts, variations in the proportions of the complex glycosylated forms of the mutants as compared with wild-type pro-LPH could be detected. We therefore compared the transport kinetics of pro-LPHI1697N, pro-LPHD1711N, and pro-LPHP1743S with those of wild-type pro-LPH by applying a pulse-chase protocol. The results are shown in Fig. 3. Mutant forms and wild-type pro-LPH were detected exclusively as mannose-rich species after 1 h of pulse. Wild-type pro-LPH was processed in the Golgi apparatus much faster than the mutants. Within 2 h of chase complex glycosylated pro-LPH appeared, while 4 to 8 h were required to see the complex glycosylated species of pro-LPHI1697N, pro-LPHD1711N, and pro-LPHP1743S. While approximately 50% of wild-type pro-LPH was converted to a complex glycosylated species within 12 h of chase, 20% of pro-LPHI1697N, 25% of pro-LPHD1711N, and only 18% of pro-LPHP1743S were found as complex glycosylated species within an identical chase period (see Fig. 3 B). Taken together, the results demonstrate that the glycosylation mutants were transported to the Golgi apparatus, albeit at a much slower rate than wild-type pro-LPH.
      Figure thumbnail gr1
      Figure 1Schematic representation of the structure of pro-LPH in human small intestinal cells. Some important structural features of pro-LPH in human small intestinal cells compiled from data which employed biosynthetic studies in human small intestinal explants (
      • Naim H.Y.
      • Lentze M.J.
      ,
      • Hauri H.P.
      • Sterchi E.E.
      • Bienz D.
      • Fransen J.A.
      • Marxer A.
      ,
      • Naim H.Y.
      • Sterchi E.E.
      • Lentze M.J.
      ), cDNA cloning (
      • Mantei N.
      • Villa M.
      • Enzler T.
      • Wacker H.
      • Boll W.
      • James P.
      • Hunziker W.
      • Semenza G.
      ), and recombinant expression in COS-1 (
      • Wuthrich M.
      • Grunberg J.
      • Hahn D.
      • Jacob R.
      • Radebach I.
      • Naim H.Y.
      • Sterchi E.E.
      ,
      • Jacob R.
      • Radebach I.
      • Wuthrich M.
      • Grunberg J.
      • Sterchi E.E.
      • Naim H.Y.
      ,
      • Naim H.Y.
      • Lacey S.W.
      • Sambrook J.F.
      • Gething M.J.
      ) or Madin-Darby canine kidney (
      • Boll W.
      • Wagner P.
      • Mantei N.
      ,
      • Grunberg J.
      • Sterchi E.E.
      ) cells. The NH2 terminus starts with a cleavable signal sequence (Met1-Gly19) for co-translational translocation into the ER. The ectodomain extends from Ser20 to Thr1882 and can be divided into four homologous domains as indicated. Two proteolytic cleavage steps of pro-LPH take place: one intracellularly between Arg734 and Leu735 to generate LPHβinitial and another one in the intestinal lumen between Arg868 and Ala869 to generate the brush-border mature enzyme, LPHβfinal (
      • Wuthrich M.
      • Grunberg J.
      • Hahn D.
      • Jacob R.
      • Radebach I.
      • Naim H.Y.
      • Sterchi E.E.
      ,
      • Jacob R.
      • Radebach I.
      • Wuthrich M.
      • Grunberg J.
      • Sterchi E.E.
      • Naim H.Y.
      ). Deletion of 236 amino acids at the COOH terminus of domain IV (Arg1647-Thr1882) yields a deletion mutant, denoted LPH1646MACT (
      • Panzer P.
      • Preuss U.
      • Joberty G.
      • Naim H.Y.
      ), a transport competent molecule. The sites of amino acid exchange, I1697N, D1711N, and P1743S, are indicated byarrows; MA, membrane anchoring; CT, cytoplasmic tail.
      Figure thumbnail gr2
      Figure 2Expression of wild-type pro-LPH , pro-LPHI1697N, pro-LPHD1711N, and pro-LPHP1743S in COS-1 cells. Transfected COS-1 cells were labeled for 6 h (pro-LPHwt) or 10 h (pro-LPHI1697N, pro-LPHD1711N, and pro-LPHP1743S) with [35S]methionine. After immunoprecipitation with mAb anti-LPH (HBB) the samples were divided into three aliquots and treated with endo H, endo F/PNGase F, or not treated. The samples were analyzed by SDS-PAGE on 6% slab gels.
      Figure thumbnail gr3
      Figure 3Transport kinetics of wild-type pro-LPH and mutant proteins. A, COS-1 cells were transfected with the cDNA of wild-type pro-LPH and the mutants, pulse labeled for 1 h with [35S]methionine and chased for the indicated periods of time with cold methionine. The immunoprecipitates were analyzed by SDS-PAGE on 6% slab gels. B, densitometric scanning of the fluorograms shown in A.

       Assessment of the Folding of the N-Glycosylation Mutants as Compared with Wild-type Pro-LPH

      The reduced transport rates of the pro-LPH glycosylation mutants is most likely the consequence of an altered three-dimensional structure as has been shown previously for other polypeptides (
      • Fan H.
      • Meng W.
      • Kilian C.
      • Grams S.
      • Reutter W.
      ,
      • Letourneur O.
      • Sechi S.
      • Willette-Brown J.
      • Robertson M.W.
      • Kinet J.P.
      ,
      • Riederer M.A.
      • Hinnen A.
      ,
      • Zhang Y.
      • Dahms N.M.
      ,
      • Roberts P.C.
      • Garten W.
      • Klenk H.D.
      ). We investigated therefore the folding properties of the mutants as compared with wild-type pro-LPH by utilizing (i) enzymatic activity assays and (ii) protease sensitivity.

       Enzymatic Activity Measurements of LPH

      COS-1 cells were transiently transfected with the cDNAs of pro-LPHI1697N, pro-LPHD1711N, pro-LPHP1743S, and wild-type pro-LPH and were labeled for 6 h with [35S]methionine. The cell lysates were immunoprecipitated with mAb anti-LPH (HBB). Part of each immunoprecipitate was separated by SDS-PAGE to confirm the expression of the expected species, while the remainder was assayed for enzymatic activity using lactose as a substrate. The electrophoretic analysis revealed the expected molecular forms (not shown, but refer to Fig. 2). The enzymatic activities of the three N-glycosylation mutants reached only background levels indicating that the mutants are not active. The wild-type pro-LPH control was as expected enzymatically active with a specific activity levels similar to that reported previously (5.6–5.8 × 109 IU/mol) (
      • Naim H.Y.
      • Lacey S.W.
      • Sambrook J.F.
      • Gething M.J.
      ).

       Protease Sensitivity

      The activity site of lactase within pro-LPH has been assigned to glutamic acid residues 1271 and 1747 in rabbit LPH (
      • Wacker H.
      • Keller P.
      • Falchetto R.
      • Legler G.
      • Semenza G.
      ). One explanation for the lack of enzymatic activity in the three mutants are misfolded structures around glutamic acid 1747, which lies in a close proximity to the mutations made in pro-LPHI1697N, pro-LPHD1711N, and pro-LPHP1743S. We probed the folding pattern of the mutants by using a trypsin sensitivity assay. Trypsin cleaves correctly folded pro-LPH to a 160-kDa LPHβfinal mature enzyme which is resistant toward further protease treatment (
      • Naim H.Y.
      • Lacey S.W.
      • Sambrook J.F.
      • Gething M.J.
      ). A change in the trypsin cleavage pattern is compatible with altered protein folding, since novel trypsin cleavage sites become exposed which are normally shielded in the core of the correctly folded native protein (
      • Naim H.Y.
      • Lacey S.W.
      • Sambrook J.F.
      • Gething M.J.
      ,
      • Jacob R.
      • Bulleid N.J.
      • Naim H.Y.
      ).
      Pro-LPH was immunoprecipitated from transfected COS-1 cells labeled with [35S]methionine for 10 h and the immunoprecipitates were treated with trypsin for 5 or 15 min at 37 °C. As shown in Fig. 4 the 160-kDa LPHβfinal appeared within 5 min of trypsin treatment and persisted at the same size and intensity after 15 min of digestion. By contrast, a similar band did not appear when the mutants were treated with trypsin. Faint bands corresponding to the pro-LPH glycoforms were still observed in the 5-min time point, but these forms were completely degraded within 15 min of trypsin treatment. Taken together, the lack of enzymatic activity and the accessibility to degradation by trypsin strongly suggest that pro-LPHI1697N, pro-LPHD1711N, and pro-LPHP1743S have altered protein folding patterns as compared with wild type pro-LPH.
      Figure thumbnail gr4
      Figure 4Trypsin sensitivity of wild-type and mutant pro-LPH proteins. COS-1 cells were transfected with DNA encoding wild-type pro-LPH, pro-LPHI1697N, pro-LPHD1711N, and pro-LPHP1743S. The cells were biosynthetically labeled with [35S]methionine for 10 h and solubilized on ice in the absence of protease inhibitors. Each cell extract was divided into equal aliquots, which were treated with trypsin at 37 °C as indicated. Thereafter, the protease was inhibited using soybean trypsin inhibitor and a mixture of protease inhibitors. Immmunoprecipitation was performed with mAb anti-LPH and the immunoprecipitates were subjected to SDS-PAGE using 6% slab gels.

       Analysis of the Quaternary Structure of the N-Glycosylation Mutants

      The acquisition of pro-LPH to transport competence occurs in the ER and includes dimerization of the mannose-rich polypeptides. This event implicates the membrane anchoring domain of pro-LPH, the deletion of which results in an accumulation of monomeric forms of pro-LPH in the ER and an ultimate degradation. A correctly folded ectodomain alone is not sufficient for dimerization to ensue (
      • Naim H.Y.
      • Naim H.
      ). The marked decrease in the amount of complex glycosylated forms of the N-glycan mutants relative to the total synthesized and processed glycoforms is indicative of a slow transport rate out of the ER. We wanted therefore to examine whether changes in the quaternary structure of the mutants have also occurred. For this the dimerization event of the mutants was analyzed in sucrose gradient centrifugation. The results are depicted in Fig. 5. Along previous data (
      • Mantei N.
      • Villa M.
      • Enzler T.
      • Wacker H.
      • Boll W.
      • James P.
      • Hunziker W.
      • Semenza G.
      ,
      • Naim H.Y.
      • Naim H.
      ), monomeric mannose-rich wild-type pro-LPH appeared in the light sucrose phase peaking in fraction 7 + 8 and the dimeric counterpart in the heavy sucrose phase (fraction 5). The complex glycosylated dimeric wild-type pro-LPH appeared exclusively in the heavy fraction. Likewise, the mutants revealed monomeric mannose-rich pro-LPH in the light fractions (
      • Zhang R.
      • Cai H.
      • Fatima N.
      • Buczko E.
      • Dufau M.L.
      ,
      • Chapman R.L.
      • Kane S.E.
      • Erickson A.H.
      ,
      • Scheiffele P.
      • Peranen J.
      • Simons K.
      ) as well as the dimeric forms in the heavier fractions (4 + 5). It is obvious therefore that the mutants dimerize prior to acquisition of complex type of glycans as has been demonstrated for wild-type pro-LPH. However, the substantial reduction in the labeling intensity of the dimeric forms of the mutants relative to wild-type pro-LPH as well as the recovery of these forms in less fractions indicate a lower rate of dimerization of the mutants as compared with wild-type pro-LPH. We conclude that the reduced dimerization efficiency of the mutants is responsible for their lower rate of processing to complex glycosylated forms. Since the membrane anchoring domain and the cytosolic tails have not been changed in the mutants and the only apparent structural differences are restricted to subregions in the extracellular portion of pro-LPH, we conclude that alterations in the ectodomain of pro-LPH have markedly affected or hampered the efficiency of dimerization of the mutants.
      Figure thumbnail gr5
      Figure 5Separation of assembled proteins on sucrose density gradients. COS-1 cells were transfected with DNA encoding wild-type pro-LPH, pro-LPHI1697N, pro-LPHD1711N, and pro-LPHP1743S and biosynthetically labeled for 6 h with [35S]methionine. Cell lysates were layered on a sucrose density gradient. After centrifugation for 22 h at 100,000 ×g, fractions were collected, immunoprecipitated with mAb anti-LPH, and analyzed on SDS-PAGE. The protein markers used (β-amylase (200 kDa), apoferritin (450 kDa), and thyroglobulin (660 kDa)) were recovered in fractions 7, 4, and 3, respectively, of the gradient. The localization of the marker proteins is indicated witharrows.

       Analysis of the LAC236 Domain and Its N-Glycosylation Mutants

      In light of the observations that N-glycosyl mutants in the LAC236 domain are associated with reduced dimerization and subsequent retarded transport of pro-LPH a direct role of the LAC236 domain in these events could be hypothesized. To delineate directly a possible role of LAC236 in the dimerization event we expressed this domain in conjunction with the transmembrane domain of pro-LPH in COS cells. Also the N-glycosyl variants described above that are present within this domain were individually expressed. Due to the lack of antibodies specific for the LAC236 domain we used GFP as a reporter gene that could be also utilized to trace the trafficking of these domains within the cell. The GFP protein was fused to the COOH-terminal end of a construct comprising LAC236 or its variants, the transmembrane domain, and the cytoplasmic tail (Fig.6).
      Figure thumbnail gr6
      Figure 6Schematic representation of LAC236wt-GFP and its N-glycosylation variants. The orientation of the signal sequence (ss), LAC236, membrane anchor (MA), cytosolic tail (CT), and GFP are indicated. The different positions ofN-glycosyl sugar chains, i.e. I1697N, D1711N, and P1743S, are marked with a “Y”.
      Cells transfected with these constructs were continuously labeled for 6 h with [35S]methionine and the detergent extracts were immunoprecipitated with anti-GFP antibodies. The wild-type LAC236-GFP (LAC236wt-GFP) chimera revealed a double band pattern consisting of a major 60-kDa polypeptide and a slightly smaller faint species (Fig. 7, lane 1). Endo H shifted these forms to one single band indicating that the two forms are different glycosylation intermediates (Fig. 7,lane 2). The size of the N-deglycosylated protein corresponds to that calculated for the LAC236-GFP chimera. The complete sensitivity of the LAC236wt-GFP toward endo H indicates that this chimera was co-translationally mannose-rich glycosylated in the ER, but was not further processed in the Golgi apparatus to a mature complex glycosylated polypeptide. In a fashion similar to LAC236wt-GFP, the other three constructs, LAC236I1697N-GFP, LAC236D1711N-GFP, and LAC236P1743S-GFP, persisted also as mannose-rich polypeptides during the same biosynthetic labeling period as determined by their sensitivities toward endo H (Fig. 7). The major biosynthetic forms of these mutants, however, revealed slightly increased apparent molecular weights due to the additional N-glycosylation sites. Subcellular localization of the chimeras was visualized by tracking the GFP fluorescence within the cell. As shown in Fig.8 LAC236wt-GFP revealed fluorescence images typical of an ER localization and the same pattern could be detected for the three other GFP-constructs (data not shown). These results clearly support the biochemical analyses that the chimeras are transport-incompetent proteins and do not exit the ER. Fusing the GFP to wild-type full-length pro-LPH did not affect its transport competence as demonstrated by the acquisition of the pro-LPH-GFP chimera to a complex glycosylated and endo H-resistant species (Fig. 7). This is further supported by the fluorescence images in which pro-LPH-GFP was localized at the cell surface, in the Golgi apparatus and in vesicular structures. Since it had no marked effects on the transport of pro-LPH it is unlikely that GFP may have been the cause of transport-incompetence of the LAC236-GFP chimeras.
      Figure thumbnail gr7
      Figure 7Expression of pro-LPH-GFP, LAC236wt-GFP, LAC236I1697N-GFP, LAC236D1711N-GFP, and LAC236P1743S-GFP in COS-1 cells. Transfected COS-1 cells were labeled for 6 h with [35S]methionine. After immunoprecipitation with mAb anti-GFP the samples were divided into three aliquots and treated with endo H, endo F/PNGase F, or not treated. The samples were analyzed by SDS-PAGE on 6% (pro-LPH) or 8% slab gels (LAC236wt-GFP, LAC236I1697N-GFP, LAC236D1711N-GFP, and LAC236P1743S-GFP).
      Figure thumbnail gr8
      Figure 8Confocal analysis pro-LPH-GFP and LAC236wt-GFP in COS-1 cells. COS-1 cells were transiently transfected with expression vectors encoding pro-LPH-GFP or LAC236wt-GFP. The cells were fixed with 3% paraformaldehyde 48 h post-transfection and examined with a confocal laser microscope (Leica TCS SP).
      Next we examined the quaternary structure of the LAC236wt-GFP chimera and its N-glycosyl variants by employing sucrose gradients. LAC236wt-GFP appeared in two main areas of the gradient that peaked at different sucrose densities (Fig. 9). LAC236wt-GFP contained in the light fractions corresponds to a monomeric form of the protein, while the peak representing the heavy fractions contained a dimeric form of LAC236wt-GFP as suggested by protein markers run on similar gradients. It is clear therefore that the LAC236wt-GFP protein forms mannose-rich dimers. The results obtained above with wild-type pro-LPH and its glycosyl variants suggested that additional N-glycosylation in the LAC236 domain reduces markedly the dimerization of pro-LPH. Having shown that LAC236wt-GFP is a dimeric molecule we wanted to determine directly the effect of the novelN-glycosylation sites on the dimerization of this form. Fig.9 demonstrates that the extent of dimerization of all three LAC236 mutants has been drastically reduced. In fact, the proportion of the mutants in the heavy gradient fractions containing dimers was many fold less than that of wild-type LAC236 compatible with reduced dimerization of the mutants. Obviously the introduction of N-glycosyl sites has markedly affected the dimerization event. The data support therefore the notion that LAC236 is critical in the dimerization event of pro-LPH and structural alterations in this domain, such as those induced by novel N-glycosylation sites, are sufficient to drastically affect the dimerization of the full-length molecule. It is interesting, however, that dimerization of LAC236 was not sufficient for this form to exit the ER or to reach the Golgi apparatus.
      Figure thumbnail gr9
      Figure 9Separation of LAC236wt-GFP, LAC236I1697N-GFP, LAC236D1711N-GFP, and LAC236P1743S-GFP on sucrose density gradients. COS-1 cells were transfected with expression vectors encoding LAC236wt-GFP, LAC236I1697N-GFP, LAC236D1711N-GFP, and LAC236P1743S-GFP. The cells were biosynthetically labeled for 6 h with [35S]methionine and cell lysates were layered on a sucrose density gradient from 5 to 25% sucrose. After centrifugation for 22 h at 100,000 × g, 12 fractions of each gradient were collected, immunoprecipitated with mAb anti-GFP, and analyzed on SDS-PAGE. The protein markers used were bovine serum albumin (66 kDa), β-galactosidase (116 kDa), and myosin (205 kDa) and were recovered in fractions 10, 5, and 3, respectively, of the gradient. The localization of the marker proteins is indicated witharrows.

      DISCUSSION

      The acquisition of a protein to its final native configuration implicates a cascade of complex events occurring mainly in the ER (
      • Hurtley S.M.
      • Helenius A.
      ,
      • Doms R.W.
      • Lamb R.A.
      • Rose J.K.
      • Helenius A.
      ) and, in some cases, in the Golgi apparatus (
      • Bonfanti L.
      • Mironov A.A.
      • Martinez-Menarguez J.A.
      • Martella O.
      • Fusella A.
      • Baldassarre M.
      • Buccione R.
      • Geuze H.J.
      • Mironov A.A.
      • Luini A.
      ). Along current concepts and views, protein folding ensues by rapid interaction of hydrophobic residues in the polypeptide chain, formation of secondary structures such as α-helices and β-sheets, and finally formation of disulfide bonds or other covalent interactions that stabilize particular regions of the protein (for a review, see Ref.
      • Gething M.J.
      • Sambrook J.
      ). The folding of a large multidomain protein, such as pro-LPH, is expected to be a complex process that implicates inter- and intramolecular protein-protein interactions. With the lack of three-dimensional structural analysis of pro-LPH no information is available yet on the organization of the four homologous regions (
      • Mantei N.
      • Villa M.
      • Enzler T.
      • Wacker H.
      • Boll W.
      • James P.
      • Hunziker W.
      • Semenza G.
      ) or the individual exons (
      • Boll W.
      • Wagner P.
      • Mantei N.
      ) in the native LPH confirmation. However, structural and functional analysis of deletion mutants of LPH have provided some clues on the significance of specific polypeptide stretches in the folding, intracellular transport, and sorting. The large profragment, LPHα, for example, exerts its function predominantly as an intramolecular chaperone and is essential in the correct folding of the membrane-associated COOH-terminal located LPHβ. This precedes homodimerization of pro-LPH (
      • Naim H.Y.
      • Naim H.
      ,
      • Grunberg J.
      • Sterchi E.E.
      ), an event that involves the transmembrane domain and is required for exit of LPH from the ER (
      • Naim H.Y.
      • Naim H.
      ). The main focus of the present investigation is a region that is located immediately upstream of the membrane in the homologous region IV and comprises 236 amino acids (LAC236). Deletion of this region from pro-LPH remains without substantial influence on the quaternary structure and folding of the resulting mutant as well as on its transport competence and polarized sorting (
      • Panzer P.
      • Preuss U.
      • Joberty G.
      • Naim H.Y.
      ) thus proposing that this region folds autonomously and is not a part of critical structures that are implicated in trafficking events of LPH. Here, we examine this possibility by introducing several potential N-linked glycosylation sites to regions which are predicted to form turns or helices. The creation of N-linked glycosylation sites within these turns hardly produces global disruption of the overall structure of the protein. However, the addition of a bulky oligosaccharide group serves as an umbrella that shields an area equivalent to 10–12 amino acids and effectively blocks access to the site. The addition of supernummary oligosaccharides can therefore be used to identify particular regions on the surface of pro-LPH that are functionally important and others that are probably involved in protein transport and sorting. Each of the mutants generated here has one additionalN-glycosylation site. This modification is sufficientper se to substantially affect many of the biosynthetic and structural features of pro-LPH. None of the mutants is as transport competent as wild-type pro-LPH; although all the mutants acquire complex type of carbohydrates and hence have been transported through the Golgi apparatus, the proportion of the mature form within total pro-LPH is substantially reduced as compared with the wild-type counterpart. This is the consequence of a marked reduction in the dimerization event of the mutants and dimerization is absolutely required for pro-LPH to acquire transport competence and exit the ER (
      • Naim H.Y.
      • Naim H.
      ,
      • Grunberg J.
      • Sterchi E.E.
      ). It is obvious that engineering an additional carbohydrate chain in LAC236 generates substantial effects on the transport kinetics of full-length pro-LPH while deletion of the same region does not significantly influence the folding, transport, and sorting of the truncated pro-LPH mutant, pro-LPH1646MACT (
      • Panzer P.
      • Preuss U.
      • Joberty G.
      • Naim H.Y.
      ).
      What putative role could be assigned to the LAC236 stretch in the context of the pro-LPH molecule? Individual expression of the LAC236 domain fused to the transmembrane domain of pro-LPH and the reporter gene GFP (LAC236wt-GFP) demonstrates that this domain is capable of dimerizing. This event is drastically reduced or even blocked when N-glycosyl sites are introduced. Nevertheless, the fact that it persists as a mannose-rich polypeptide is indicative of transport incompetence per se. In this context dimerization is not a sufficient requirement for a protein to leave the ER and exit of LAC236 from the ER requires therefore additional domains of the pro-LPH molecule. It has been shown that deletion of the LAC236 domain generates a transport-competent LPH1646MACT deletion mutant (
      • Panzer P.
      • Preuss U.
      • Joberty G.
      • Naim H.Y.
      ). However, the marked delay in the transport of this mutant from the ER to the Golgi apparatus relative to the full-length pro-LPH molecule points to a direct role of LAC236 in the initial phases of assembly of the ectodomain of pro-LPH. In fact, alterations in the structural features of LAC236 due to additionalN-glycosylation are not restricted to the domain itself, but have rather subsequent implications on the folding of the remainder of pro-LPH, the pro-LPH1646MACT (
      • Panzer P.
      • Preuss U.
      • Joberty G.
      • Naim H.Y.
      ). This is shown in protease sensitivity assays in which N-glycosyl mutants are readily degraded by trypsin in contrast to wild-type pro-LPH suggestive of changes occurring in the overall structure of pro-LPH glycoforms. Individual expression of LAC236 and its N-glycosyl mutants indicate that the additional N-glycosyl chains sterically hinder the association of LAC236 of one monomer with that of a neighboring one. It is conceivable therefore that these structural changes induce subtle changes in the folding patterns of the pro-LPHN-glycosyl mutants. While such changes do not completely prevent dimerization of the mutants, the generated dimers are not stable toward trypsin suggestive of a loose association between the monomeric units. Obviously, strict folding criteria of the LAC236 domain are absolutely required in the context of pro-LPH, since slight alterations in this domain cannot be tolerated by the remainder of the pro-LPH molecule. Presumably LAC236 nucleates the association of the ectodomain subunits of neighboring pro-LPH molecules and hence a native configuration of this region is necessary and sufficient for this process to ensue. By virtue of the potential role of LAC236, it is worth shedding light on some structural features of this domain. The LAC236 stretch is rich in hydrophobic amino acids (
      • Mantei N.
      • Villa M.
      • Enzler T.
      • Wacker H.
      • Boll W.
      • James P.
      • Hunziker W.
      • Semenza G.
      ) and algorithmic predictions reveal that it is composed of two major α-helices. The potential N-glycosylation sites in the wild-type species as well as the introduced ones are embedded in the predicted helices. Since the active site of pro-LPH is located at residue Glu1747 in this domain, it is possible to directly assess folding alterations in the mutants by measuring the enzymatic activity of pro-LPH. That all the mutants are enzymatically inactive is a clear indication of changes in the folding of the 236-amino acid domain. Two possibilities could be offered to explain these alterations. In the first, the amino acid substitutions per se may have affected the folding of a microdomain in a close proximity to the active site. This is likely for the mutant pro-LPHP1743S, in which one of two consecutive proline residues has been replaced by a serine leading to a more relaxed conformation. The mutants pro-LPHI1697N and pro-LPHD1711N with mutations located farther away from the active site may not follow this hypothesis.
      The second possibility is that additional glycans in LAC236 lead to conformational distortion of this region. This assigns carbohydrates present in this domain a critical role in modulating the activity or influencing the folding pattern and the resulting transport kinetics of pro-LPH. Pro-LPH activity is known to vary with differences inO-glycosylation (
      • Naim H.Y.
      • Lentze M.J.
      ), which has been proposed to take place in the 236-amino acid domain. A similar role could be also proposed forN-linked glycans. Several potentialN-glycosylation sites are found in this domain (
      • Mantei N.
      • Villa M.
      • Enzler T.
      • Wacker H.
      • Boll W.
      • James P.
      • Hunziker W.
      • Semenza G.
      ) and it is possible that not all of these sites are always glycosylated. In line with the data presented here, one would expect that addition of anN-linked carbohydrate chain leads to alterations in the folding of pro-LPH. Along this model the LAC236 domain may play a regulatory role in the context of structure and function of pro-LPH.

      Acknowledgments

      We thank Dr. Hans-Peter Hauri, Biozentrum, University of Basel, and Dr. Erwin Sterchi, Institute of Biochemistry and Molecular Biology, University of Bern, Switzerland, for generous gifts of the monoclonal anti-LPH antibody.

      REFERENCES

        • Fan H.
        • Meng W.
        • Kilian C.
        • Grams S.
        • Reutter W.
        Eur. J. Biochem. 1997; 246: 243-251
        • Letourneur O.
        • Sechi S.
        • Willette-Brown J.
        • Robertson M.W.
        • Kinet J.P.
        J. Biol. Chem. 1995; 270: 8249-8256
        • Riederer M.A.
        • Hinnen A.
        J. Bacteriol. 1991; 173: 3539-3546
        • Zhang Y.
        • Dahms N.M.
        Biochem. J. 1993; 295: 841-848
        • Roberts P.C.
        • Garten W.
        • Klenk H.D.
        J. Virol. 1993; 67: 3048-3060
        • Cannon K.S.
        • Hebert D.N.
        • Helenius A.
        J. Biol. Chem. 1996; 271: 14280-14284
        • Zhang R.
        • Cai H.
        • Fatima N.
        • Buczko E.
        • Dufau M.L.
        J. Biol. Chem. 1995; 270: 21722-21728
        • Chapman R.L.
        • Kane S.E.
        • Erickson A.H.
        J. Biol. Chem. 1997; 272: 8808-8816
        • Scheiffele P.
        • Peranen J.
        • Simons K.
        Nature. 1995; 378: 96-98
        • Shen F.
        • Wang H.
        • Zheng X.
        • Ratnam M.
        Biochem. J. 1997; 327: 759-764
        • Allen S.
        • Naim H.Y.
        • Bulleid N.J.
        J. Biol. Chem. 1995; 270: 4797-4804
        • Loo T.W.
        • Clarke D.M.
        J. Biol. Chem. 1998; 273: 14671-14674
        • Feng W.
        • Matzuk M.M.
        • Mountjoy K.
        • Bedows E.
        • Ruddon R.W.
        • Boime I.
        J. Biol. Chem. 1995; 270: 11851-11859
        • Rickert K.W.
        • Imperiali B.
        Chem. Biol. 1995; 2: 751-759
        • Hebert D.N.
        • Foellmer B.
        • Helenius A.
        Cell. 1995; 81: 425-433
        • Asker N.
        • Axelsson M.A.B.
        • Olofsson S.-O.
        • Hansson G.C.
        J. Biol. Chem. 1998; 273: 18857-18863
        • Muyan M.
        • Boime I.
        Mol. Endocrinol. 1998; 12: 766-772
        • Mantei N.
        • Villa M.
        • Enzler T.
        • Wacker H.
        • Boll W.
        • James P.
        • Hunziker W.
        • Semenza G.
        EMBO J. 1988; 7: 2705-2713
        • Wuthrich M.
        • Grunberg J.
        • Hahn D.
        • Jacob R.
        • Radebach I.
        • Naim H.Y.
        • Sterchi E.E.
        Arch. Biochem. Biophys. 1996; 336: 27-34
        • Jacob R.
        • Radebach I.
        • Wuthrich M.
        • Grunberg J.
        • Sterchi E.E.
        • Naim H.Y.
        Eur. J. Biochem. 1996; 236: 789-795
        • Naim H.Y.
        • Jacob R.
        • Naim H.
        • Sambrook J.F.
        • Gething M.J.
        J. Biol. Chem. 1994; 269: 26933-26943
        • Oberholzer T.
        • Mantei N.
        • Semenza G.
        FEBS Lett. 1993; 333: 127-131
        • Naim H.Y.
        • Naim H.
        Eur. J. Cell Biol. 1996; 70: 198-208
        • Panzer P.
        • Preuss U.
        • Joberty G.
        • Naim H.Y.
        J. Biol. Chem. 1998; 273: 13861-13869
        • Hunziker W.
        • Spiess M.
        • Semenza G.
        • Lodish H.F.
        Cell. 1986; 46: 227-234
        • Olsen J.
        • Cowell G.M.
        • Konigshofer E.
        • Danielsen E.M.
        • Moller J.
        • Laustsen L.
        • Hansen O.C.
        • Welinder K.G.
        • Engberg J.
        • Hunziker W.
        FEBS Lett. 1988; 238: 307-314
        • Vogel L.K.
        • Noren O.
        • Sjostrom H.
        FEBS Lett. 1992; 308: 14-17
        • Wilson I.B.
        • Gavel Y.
        • von Heijne G.
        Biochem. J. 1991; 275: 529-534
        • Naim H.Y.
        • Lentze M.J.
        J. Biol. Chem. 1992; 267: 25494-25504
        • Naim H.Y.
        • Lacey S.W.
        • Sambrook J.F.
        • Gething M.J.
        J. Biol. Chem. 1991; 266: 12313-12320
        • Hauri H.P.
        • Sterchi E.E.
        • Bienz D.
        • Fransen J.A.
        • Marxer A.
        J. Cell Biol. 1985; 101: 838-851
        • Wacker H.
        • Keller P.
        • Falchetto R.
        • Legler G.
        • Semenza G.
        J. Biol. Chem. 1992; 267: 18744-18752
        • Dahlqvist A.
        Anal. Biochem. 1968; 22: 99-107
        • Naim H.Y.
        • Sterchi E.E.
        • Lentze M.J.
        Biochem. J. 1987; 241: 427-434
        • Jacob R.
        • Bulleid N.J.
        • Naim H.Y.
        J. Biol. Chem. 1995; 270: 18678-18684
        • Hurtley S.M.
        • Helenius A.
        Annu. Rev. Cell Biol. 1989; 5: 277-307
        • Doms R.W.
        • Lamb R.A.
        • Rose J.K.
        • Helenius A.
        Virology. 1993; 193: 545-562
        • Bonfanti L.
        • Mironov A.A.
        • Martinez-Menarguez J.A.
        • Martella O.
        • Fusella A.
        • Baldassarre M.
        • Buccione R.
        • Geuze H.J.
        • Mironov A.A.
        • Luini A.
        Cell. 1998; 95: 993-1003
        • Gething M.J.
        • Sambrook J.
        Nature. 1992; 355: 33-45
        • Boll W.
        • Wagner P.
        • Mantei N.
        Am. J. Hum. Genet. 1991; 48: 889-902
        • Grunberg J.
        • Sterchi E.E.
        Arch. Biochem. Biophys. 1995; 323: 367-372