Role of glycosylation on the secretion and biological activity of erythropoietin.

The erythropoietin (EPO) molecule contains four carbohydrate chains. Three contain N-linkages to asparagines at positions 24, 38, and 83, and one contains an O-linkage to a serine at position 126. We constructed human EPO variants that eliminated the three N-glycosylation sites by replacing the asparagines with glutamines singly or in combination. The O-linked carbohydrate chain was removed by replacing the serine with glutamine, valine, histidine, or alanine. A variant with a double mutation (Gln38,83) and another with a triple mutation (Gln24,38,83) were secreted poorly from COS1 and CHO cells even though RNA encoding these variants was present. All other variants with mutations in N-linked glycosylation sites were secreted normally. Removal of any of the N-glycosylation sites reduced the in vivo but not the in vitro biological activity of the EPO molecule. All the mutations at Ser126, the O-glycosylation site, were secreted normally. In vitro activity was also unaffected except for Ala126 which had a 50-fold decrease. The Val126 variant was tested in vivo, and its specific activity was only slightly less than that of the native EPO, which indicates that the O-linked carbohydrate is not essential for activity.     

Sequence differences between glycosylated and non-glycosylated Asn-X-Thr/Ser acceptor sites: implications for protein engineering

In N-glycosylated glycoproteins, carbohydrate is attached to Asn in the sequence Asn-X-Ser/Thr, where X denotes any amino acid. However, the presence of this consensus peptide does not always lead to glycosylation. We have compiled an extensive collection of glycosylated and non-glycosylated Asn-X-Thr/Ser sites and present a statistical study based on this data set. Our results indicate that non-glycosylated sites tend to be found more frequently towards the C termini of glycoproteins, and that proline residues in positions X and Y in the consensus Asn-X-Thr/Ser-Y strongly reduce the likelihood of N-linked glycosylation. Beyond this, there are no obvious local sequence features that seem to correlate with the absence or presence of N-linked glycosylation. These findings are discussed in terms of the prediction and engineering of glycosylation sites in secretory proteins.     

Amino acid sequence of guinea pig prostate kallikrein

The primary structure of the major arginine esteropeptidase from guinea pig prostate has been deduced from automated Edman degradation of peptides generated by clostripain, cyanogen bromide, endoproteinase Lys-C, and Staphylococcus aureus V8 protease digestion of the protein. The esteropeptidase is a single polypeptide chain comprised of 239 amino acids and contains 2 apparent sites of carbohydrate attachment, Asn-78 and Asn-169. Both occur in consensus sequences for N-linked glycosylation sites. The esteropeptidase exhibits approximately 35% homology with trypsin including conservation of the catalytic residues and the aspartic acid which confers specificity toward basic amino acids. The sequence identity, however, extends to greater than 60% with the kallikrein family of serine proteases. In addition to the overall homology, the guinea pig enzyme displays a number of features characteristic of kallikreins including 10 conserved half-cystine residues, a C-terminal proline, and the "kallikrein loop". On the basis of this structural relatedness, the enzyme has been designed as guinea pig prostate kallikrein. In contrast to many of the kallikreins of other species and tissues, this enzyme does not contain any sites within the kallikrein loop sensitive to proteases that result in internal breaks in the polypeptide chain.     

Introduction of N-linked glycans in the lectin domain of surfactant protein D: impact on interactions with influenza A viruses

Porcine surfactant protein D (pSP-D) displays distinctively strong, broad-range inhibitory activity against influenza A virus (IAV). N-Linked glycosylation of the carbohydrate recognition domain (CRD) of pSP-D contributes to the high affinity of this collectin for IAV. To investigate the role of the N-linked glycan further, HEK293E protein expression was used to produce recombinant pSP-D (RpSP-D) that has similar structural and antiviral properties as NpSP-D. We introduced an additional N-linked glycan in the CRD of RpSP-D but this modification did not alter the antiviral activity. Human SP-D is unglycosylated in its CRD and less active against IAV compared with pSP-D. In an attempt to modify its antiviral properties, several recombinant human SP-D (RhSP-D) mutants were constructed with N-linked glycans introduced at various locations within its CRD. To retain lectin activity, necessary for the primary interactions between SP-D and IAV, N-linked glycosylation of RhSP-D was shown to be restricted to the corresponding position in the CRD of either pSP-D or surfactant protein A (SP-A). These N-glycosylated RhSP-D mutants, however, did not show increased neutralization activity against IAV. By developing RhSP-D mutants that also have the pSP-D-specific Ser-Gly-Ala loop inserted in the CRD, we could demonstrate that the N-linked glycan-mediated interactions between pSP-D and IAV involves additional structural prerequisites of the pSP-D CRD. Ultimately, these studies will help to develop highly effective SP-D-based therapeutic and prophylactic drugs against IAV.     

Analysis of post-translational modification of mycobacterial proteins using a cassette expression system

A recombinant expression system was developed to analyse sequence determinants involved in O-glycosylation of proteins in mycobacteria. By expressing peptide sequences corresponding to known glycosylation sites within a chimeric lipoprotein construct, amino acids flanking modified threonine residues were found to have an important influence on glycosylation. The expression system was used to screen mycobacterial sequences selected using a neural network (NetOglyc) trained on eukaryotic O-glycoproteins. Evidence of glycosylation was obtained for eight of 11 proteins tested. The results suggest that sites involved in O-glycosylation of mycobacterial and eukaryotic proteins share similar structural features.     

Enhancement of therapeutic protein in vivo activities through glycoengineering

Delivery of protein therapeutics often requires frequent injections because of low activity or rapid clearance, thereby placing a burden on patients and caregivers. Using glycoengineering, we have increased and prolonged the activity of proteins, thus allowing reduced frequency of administration. Glycosylation analogs with new N-linked glycosylation consensus sequences introduced into the protein were screened for the presence of additional N-linked carbohydrates and retention of in vitro activity. Suitable consensus sequences were combined in one molecule, resulting in glycosylation analogs of rHuEPO, leptin, and Mpl ligand. All three molecules had substantially increased in vivo activity and prolonged duration of action. Because these proteins were of three different classes (rHuEPO is an N-linked glycoprotein, Mpl ligand an O-linked glycoprotein, and leptin contains no carbohydrate), glycoengineering may be generally applicable as a strategy for increasing the in vivo activity and duration of action of proteins. This strategy has been validated clinically for glycoengineered rHuEPO (darbopoetin alfa).     

Definition of the bacterial N-glycosylation site consensus sequence

The Campylobacter jejuni pgl locus encodes an N-linked protein glycosylation machinery that can be functionally transferred into Escherichia coli. In this system, we analyzed the elements in the C. jejuni N-glycoprotein AcrA required for accepting an N-glycan. We found that the eukaryotic primary consensus sequence for N-glycosylation is N terminally extended to D/E-Y-N-X-S/T (Y, X not equalP) for recognition by the bacterial oligosaccharyltransferase (OST) PglB. However, not all consensus sequences were N-glycosylated when they were either artificially introduced or when they were present in non-C. jejuni proteins. We were able to produce recombinant glycoproteins with engineered N-glycosylation sites and confirmed the requirement for a negatively charged side chain at position -2 in C. jejuni N-glycoproteins. N-glycosylation of AcrA by the eukaryotic OST in Saccharomyces cerevisiae occurred independent of the acidic residue at the -2 position. Thus, bacterial N-glycosylation site selection is more specific than the eukaryotic equivalent with respect to the polypeptide acceptor sequence.     

Evaluation of human N-linked glycosylation sites in murine granulocyte-macrophage colony-stimulating factor.

Nonglycosylated murine and human granulocyte-macrophage colony-stimulating factor have a molecular mass of approximately 14.5 kDa predicted from the primary amino acid sequence. The expression of both proteins in COS cells leads to a heterogeneous population of molecules that differ in the degree of glycosylation. Both human and murine molecules contain two N-linked glycosylation sites that are situated in nonhomologous locations along the linear sequence. Despite this difference both proteins show a similar size distribution among the glycosylation variants. These studies analyze the effects of introducing in the murine protein novel N-linked glycosylation sites corresponding to those sites found in the human molecule. A panel of molecules composed of various combinations of human N-linked glycosylation sites in either the presence or the absence of murine N-linked glycosylation was compared. Substitution of a proper human N-linked glycosylation consensus sequence at Asn 24 did not result in N-linked glycosylation, nor was there any considerable effect on bioactivity. Replacement of the N-linked glycosylation consensus sequence at Asn 34 results in glycosylation similar to that found in the human molecule and causes a significant decrease in bioactivity. These data suggest that the position of N-linked glycosylation is critical for maximal bioactivity in a particular species and that the changes in position of these sites in different species probably occurred during evolution in response to changes in their receptors.     

Cytoplasmic N-glycosyltransferase of Actinobacillus pleuropneumoniae is an inverting enzyme and recognizes the NX(S/T) consensus sequence

N-Linked glycosylation is a frequent protein modification that occurs in all three domains of life. This process involves the transfer of a preassembled oligosaccharide from a lipid donor to asparagine side chains of polypeptides and is catalyzed by the membrane-bound oligosaccharyltransferase (OST). We characterized an alternative bacterial pathway wherein a cytoplasmic N-glycosyltransferase uses nucleotide-activated monosaccharides as donors to modify asparagine residues of peptides and proteins. N-Glycosyltransferase is an inverting glycosyltransferase and recognizes the NX(S/T) consensus sequence. It therefore exhibits similar acceptor site specificity as eukaryotic OST, despite the unrelated predicted structural architecture and the apparently different catalytic mechanism. The identification of an enzyme that integrates some of the features of OST in a cytoplasmic pathway defines a novel class of N-linked protein glycosylation found in pathogenic bacteria.     

Glycosylation at specific sites of erythropoietin is essential for biosynthesis, secretion, and biological function

The glycoprotein hormone erythropoietin (Ep), the primary regulator of erythropoiesis, is synthesized by the kidney and secreted as the mature protein with three N-linked and one O-linked oligosaccharide chains. To investigate the role(s) of each carbohydrate moiety in the biosynthesis and function of Ep, we have used oligonucleotide-directed mutagenesis of a cDNA for human Ep to alter the amino acids at each of the carbohydrate attachment sites. Each mutated cDNA construct was expressed in stably transfected sublines of a kidney cell line, baby hamster kidney. We show, by preventing attachment of N-linked carbohydrate at asparagines 38 or 83, or preventing O-linked glycosylation at serine 126, that glycosylation of each of these specific sites is critical for proper biosynthesis and secretion of Ep. Fractionation of cellular extracts demonstrated that the mutant proteins lacking glycosylation at each of these three sites, (38, 83, and 126) were associated mainly with membrane components or were degraded rapidly. Less than 10% of these three mutant proteins were processed properly and secreted from the cells. The Ep protein lacking N-linked glycosylation at asparagine 24 is synthesized and secreted as efficiently as native Ep. The carbohydrates at positions 24 and 38 may be involved in the biological activity of Ep, since the absence of either of the oligosaccharide side chains at these positions reduced the hormone's biological activity.     

From peptide to protein: comparative analysis of the substrate specificity of N-linked glycosylation in C. jejuni

The gram-negative bacterium Campylobacter jejuni was recently discovered to contain a general N-linked protein glycosylation pathway. Central to this pathway is PglB, a homologue of the Stt3p subunit of the eukaryotic oligosaccharyl transferase (OT), which is involved in the transfer of an oligosaccharide from a polyisoprenyl pyrophosphate carrier to the asparagine side chain of proteins within the conserved glycosylation sites D/E-X1-N-X2-S/T, where X1 and X2 can be any amino acids except proline. Using a library of peptide substrates and a quantitative radioactivity-based in vitro assay, we assessed the amino acids at each position of the consensus glycosylation sequence for their impact on glycosylation efficiency, whereby the sequence DQNAT was found to be the optimal acceptor substrate. In the context of a full-length folded protein, the differences between variations of the glycosylation sequences were found to be consistent with the trends observed from their peptidyl counterparts, though less dramatic because of additional influences. In addition to characterizing the acceptor preferences of PglB, we also assessed the selectivity toward the glycan donor. Interestingly, despite recent reports of relaxed selectivity toward the glycan donor, PglB was not found to be capable of utilizing glycosyl donors such as dolichyl-pyrophosphate-chitobiose, which is the minimum substrate for the eukaryotic OT process.     

Secretion of glycosylation site mutants can be rescued by the signal/pro sequence of tissue plasminogen activator.

Strategies that prevent the attachment of N-linked carbohydrates to nascent glycoproteins often impair intracellular transport and secretion. In the present study, we describe a method to rescue the intracellular transport and secretion of glycoproteins mutagenized to delete N-linked glycosylation sites. Site-directed mutagenesis was used to delete N-linked glycosylation sites from a chimeric protein, TNFR-IgG1. Deletion of any of the three glycosylation sites in the TNFR portion of the molecule, alone or in combination, resulted in a moderate or near total blockade of TNFR-IgG1 intracellular transport and secretion. Pulse chase experiments suggested that the glycosylation site mutants accumulated in the endoplasmic reticulum (ER) and were inefficiently exported to the Golgi apparatus (GA). Replacement of the TNFR signal sequence with the signal/pro sequence of human tissue plasminogen activator (tPA) overcame the blockade to intracellular transport, and restored secretion to levels comparable to those achieved with the fully glycosylated molecule. Ligand binding studies suggested that the secreted glycosylation variants possessed binding characteristics similar to the fully glycosylated protein. This study demonstrates that N-terminal sequences of tPA are unexpectedly efficient in facilitating transport from the ER to the GA and suggests that these sequences contain a previously unrecognized structural element that promotes intracellular transport.     

The role of constant region carbohydrate in the assembly and secretion of human IgD and IgA1

Immunoglobulins are glycoproteins, containing N- linked carbohydrates in the heavy chain constant regions of all isotypes and O-linked carbohydrates in the hinge regions of human IgA1 and IgD. A previous study showed that IgD synthesized in the presence of tunicamycin and lacking the three N-linked glycans on the heavy chain was not secreted (Shin, S. U., Wei, D. F., Amin, A. R., Thorbecke, G. J., and Morrison, S. L. (1992) Hum. Antibodies 3, 65-74). The contribution of each of the carbohydrates in the Fc of IgD to assembly and secretion was now analyzed by eliminating the carbohydrate addition sequence, Asn-X-Ser/Thr, through site-directed mutagenesis. Only the carbohydrate nearest the sole disulfide bond between heavy chains, which remained high mannose and appeared to be buried within the folded molecule, was found to be essential for secretion. When IgD lacked that glycan, assembly reached only the heavy/light chain half-molecule stage, and heavy chains were held inside the endoplasmic reticulum. Using benzyl 2-acetamido-2-deoxy-alpha-d-galactopyranoside (BADG) to inhibit complete O-linked glycosylation, we found that IgA1 and IgD with incomplete hinge carbohydrates were assembled and secreted from cells. Thus, one N-linked glycan plays a structural role in IgD and is required for proper assembly and secretion, but the O-linked carbohydrates in the hinge of IgD and IgA1 are not required for folding and export.     

Different roles of individual N-linked oligosaccharide chains in folding, assembly, and transport of the simian virus 5 hemagglutinin-neuraminidase.

The role of N-linked glycosylation in protein maturation and transport has been studied by using the simian virus 5 hemagglutinin-neuraminidase (HN) protein, a model class II integral membrane glycoprotein. The sites of N-linked glycosylation on HN were identified by eliminating each of the potential sites for N-linked glycosylation by oligonucleotide-directed mutagenesis on a cDNA clone. Expression of the mutant HN proteins in eucaryotic cells indicated that four sites are used in the HN glycoprotein for the addition of N-linked oligosaccharide chains. These functional glycosylation sites were systematically eliminated in various combinations from HN to form a panel of mutants in which the roles of individual carbohydrate chains and groups of carbohydrate chains could be analyzed. Alterations in the normal glycosylation pattern resulted in the impairment of HN protein folding and assembly which, in turn, affected the intracellular transport of HN. The severity of the consequences on HN maturation depended on both the number of deleted carbohydrate sites and their position in the HN molecule. Analysis of the reactivity pattern of HN conformation-specific monoclonal antibodies with the mutant HN proteins indicated that one specific carbohydrate chain plays a major role in promoting the correct folding of HN. Another carbohydrate chain, which is not essential for the initial folding of HN was found to play a role in preventing the aggregation of HN oligomers. The HN molecules which were misfolded, owing to their altered glycosylation pattern, were retained in the endoplasmic reticulum. Double-label immunofluorescence experiments indicate that misfolded HN and folded HN are segregated in the same cell. Misfolded HN forms disulfide-linked aggregates and is stably associated with the resident endoplasmic reticulum protein, GRP78-BiP, whereas wild-type HN forms a specific and transient complex with GRP78-BiP during its folding process.     

Nucleotide sequence of the Saccharomyces cerevisiae PUT4 proline-permease-encoding gene: similarities between CAN1, HIP1 and PUT4 permeases

The complete nucleotide (nt) sequence of the PUT4 gene, whose product is required for high-affinity proline active transport in the yeast Saccharomyces cerevisiae, is presented. The sequence contains a single long open reading frame of 1881 nt, encoding a polypeptide with a calculated Mr of 68,795. The predicted protein is strongly hydrophobic and exhibits six potential glycosylation sites. Its hydropathy profile suggests the presence of twelve membrane-spanning regions flanked by hydrophilic N- and C-terminal domains. The N terminus does not resemble signal sequences found in secreted proteins. These features are characteristic of integral membrane proteins catalyzing translocation of ligands across cellular membranes. Protein sequence comparisons indicate strong resemblance to the arginine and histidine permeases of S. cerevisiae, but no marked sequence similarity to the proline permease of Escherichia coli or to other known prokaryotic or eukaryotic transport proteins. The strong similarity between the three yeast amino acid permeases suggests a common ancestor for the three proteins.     

Cloning, sequencing and site of origin of the rat sperm receptor protein, ZP3

The ZP3 gene encodes for a zona glycoprotein that serves as both a cell-specific binding site for capacitated spermatozoa and an inducer of acrosomal exocytosis during fertilisation. In this study we have determined the nucleotide sequence of rat ZP3 (accession no. Y10823), predicted primary amino acid structure and determined the cellular origin of this molecule within the ovary. Rat ZP3 was found to have an open reading frame of 1272 nucleotides encoding a polypeptide chain of 424 amino acids that was expressed exclusively by the actively growing oocyte population. Rat ZP3 exhibited 91%, 78% and 66% identity with the mouse, hamster and human homologues, respectively. Key features of mouse ZP3, including the number and location of cysteine and proline residues and N-linked glycosylation sites, were also conserved in the rat homologue. The putative O-linked glycosylation sites, a series of serine residues at ZP3(329-334), were also conserved in rat and mouse ZP3, although immediately downstream of this site the amino acid sequences deviated over a short stretch of amino acids. The hydropathicity profile revealed two hydrophobic domains. The first was associated with a putative N-terminal signal sequence which is unusual in the rat in possessing a proline residue at the -1 position relative to the signal cleavage site, a feature it shares with human and marmoset ZP3 but not mouse. The second hydrophobic domain was observed at the C-terminus downstream of a TGF-beta type III receptor domain that appears to be common to all ZP3 sequences examined to date.     

Secretion of glycosylated alpha-lactalbumin in yeast Pichia pastoris

The secretion of N-linked glycosylated alpha-lactalbumin was much higher in the expression system of yeast Pichia pastoris carrying goat alpha-lactalbumin cDNA than in mammalian milk. This is possibly because of the presence of N-linked glycosylation signal sequences, Asn(45)-Asp(46)-Ser(47) and Asn(74)-Ile(75)-Ser(76), in wild-type alpha-lactalbumin. Attempts to elucidate the mechanism of the higher secretion of glycosylated alpha-lactalbumin in P. pastoris were made. Mutant N45D that deleted the N-linked glycosylation signal sequence at position 45 predominantly secreted nonglycosylated protein. On the other hand, mutant D46N with another N-glycosylation signal site at position 46 only secreted N-linked glycosylated alpha-lactalbumin, i.e. not the nonglycosylated protein. The total secreted amount of mutant N45D was greatly enhanced, while the secreted amounts of the wild-type and mutant D46N were very low, suggesting that the increase in the number of glycosylation sites greatly reduced the secretion of alpha-lactalbumin. It seems likely that the glycosylated alpha-lactalbumin may be degraded by the quality control system.     

HLA-D region antigen-associated invariant polypeptides as revealed by two-dimensional gel analysis. Glycosylation and structural inter-relationships

Two-dimensional polyacrylamide gel analyses of immunoprecipitates of HLA-D region antigens prepared from [35S]methionine-labeled B lymphoblastoid cells revealed a number of invariant polypeptides (Ii and theta) that co-precipitate with the alpha and beta polypeptides of the class II (Ia) antigens. The invariant polypeptides comprised at least three Ii spots of Mr = 31,000 (Ii1-Ii3) and a series of six theta spots of Mr = 34,000 (theta 1-theta 6). The structural inter-relationships of these polypeptides have been investigated. Tryptic peptide fingerprints showed that Ii and theta have closely related amino acid sequences. In contrast, the fingerprints of the HLA-DR alpha and beta polypeptides clearly differed from those of theta and Ii as well as from each other. Analyses of immunoprecipitates prepared from cells cultured in the presence of tunicamycin revealed the presence of two N-linked oligosaccharides on each invariant polypeptide and suggested that the more acidic theta polypeptides (theta 1 and theta 2) differed from the other invariant polypeptides by the presence of sialic acid on one or both N-linked oligosaccharides. Removal of sialic acid by neuraminidase simplified the pattern of theta spots into three distinct Ii-related polypeptides. Endo-beta-N-acetylglycosaminidase H digestion indicated that the individual theta polypeptides represent stages in carbohydrate processing whereby Ii with two N-linked immature oligosaccharides are converted initially to theta 6-theta 3 with one immature and one complex, but nonsialylated, oligosaccharide and finally to theta 2-theta 1 with two complex oligosaccharides. Digestion of the theta polypeptides with N-acetylgalactosamine oligosaccharidase indicated that the theta spots are also derived by O-glycosylation from the Ii polypeptides. This assignment is supported by results obtained using monensin to block glycosylation within the Golgi. At least three spots persisted after complete removal of the N- and O-linked oligosaccharides, suggesting the presence of a family of invariant polypeptides differing in amino acid sequence.     

Post-translational modification of bone morphogenetic protein-1 is required for secretion and stability of the protein

Bone morphogenetic protein (BMP)-1 is a glycosylated metalloproteinase that is fundamental to the synthesis of a normal extracellular matrix because it cleaves type I procollagen, as well as other precursor proteins. Sequence analysis suggests that BMP-1 has six potential N-linked glycosylation sites (i.e. NXS/T) namely: Asn(91) (prodomain), Asn(142) (metalloproteinase domain), Asn(332) and Asn(363) (CUB1 domain), Asn(599) (CUB3 domain), and Asn(726) in the C-terminal-specific domain. In this study we showed that all these sites are N-glycosylated with complex-type oligosaccharides containing sialic acid, except Asn(726) presumably because proline occurs immediately C-terminal of threonine in the consensus sequence. Recombinant BMP-1 molecules lacking all glycosylation sites or the three CUB-specific sites were not secreted. BMP-1 lacking CUB glycosylation was translocated to the proteasome for degradation. BMP-1 molecules lacking individual glycosylation sites were efficiently secreted and exhibited full procollagen C-proteinase activity, but N332Q and N599Q exhibited a slower rate of cleavage. BMP-1 molecules lacking any one of the CUB-specific glycosylation sites were sensitive to thermal denaturation. The study showed that the glycosylation sites in the CUB domains of BMP-1 are important for secretion and stability of the molecule.     

Enzymatic conversion of proteins to glycoproteins by lipid-linked saccharides: A study of potential exogenous acceptor proteins

Previous studies have shown that a membrane preparation from hen oviduct catalyzes transfer of oligosaccharide from oligosaccharide-P-P-dolichol to denatured RNase and α-lactalbumin. To gain further insight into the structural requirements of a protein that allow it to serve as a substrate for glycosylation, the acceptor ability of a variety of other modified proteins containing the tripeptide sequence -ASN-X-(SER/THR)- has been investigated. Of 7 proteins tested, 2 (ovine prolactin and rabbit muscle triosephosphate isomerase) could be enzymatically glycosylated by a particulate preparation from hen oviduct. The remaining 5 proteins, assayed as either S-carboxy-methylated or S-aminoethylated derivatives, were inactive as carbohydrate acceptors. However, cyanogen bromide treatment of 2 of the inactive proteins, bovine catalase and concanavalin A from jack bean, yielded peptide fragments which served as substrates for glycosylation. These results suggests that for some proteins, disruption of the tertiary structure is sufficient to allow attachment of carbohydrate. Other denatured proteins may possess additional restrictions imposed by their secondary structure. In certain cases, these restrictions are removed when the polypeptide chain is fragmented.