Glycosylation Improves the Proteolytic Stability of Exenatide.

Exenatide was the first marketed GLP-1 receptor agonist for the treatment of type 2 diabetes. Modification to the chemical structure or the formulation has the potential to increase the stability of exenatide. We introduced human complex-type sialyloligosaccharide to exenatide at the native Asn28 position. The synthesis was achieved using both solid phase peptide synthesis (SPPS) and Omniligase-1-mediated chemoenzymatic ligation. The results demonstrate that glycosylation increases the proteolytic stability of exenatide while retaining its full biological activity.

Regulating Antifreeze Activity through Water: Latent Functions of the Sugars of Antifreeze Glycoprotein Revealed by Total Chemical Synthesis.

Antifreeze glycoprotein (AFGP), which inhibits the freezing of water, is highly O-glycosylated with a disaccharide, d-Galß1-3-d-GalNAcα (GalGalNAc). To elucidate the function of the sugar residues for antifreeze activity at the molecular level, we conducted a total chemical synthesis of partially sugar deleted AFGP derivatives, and unnatural forms of AFGPs incorporating glucose (Glc)-type sugars instead of galactose (Gal)-type sugars. These elaborated AFGP derivatives demonstrated that the stereochemistry of each sugar residue on AFGPs precisely correlates with the antifreeze activity. A hydrogen-deuterium exchange experiment using synthetic AFGPs revealed a different dynamic behavior of water around sugar residues depending on the sugar structures. These results indicate that sugar residues on AFGP form a unique dynamic water phase that disturbs the absorbance of water molecules onto the ice surface, thereby inhibiting freezing.

Rapid Chemical Synthesis of Serine Protease Inhibitor Kazal-type 13 (SPINK13) Glycoform by a Combined Method with Glycan Insertion Strategy and Fast-Flow Fmoc SPPS.

Serine protease inhibitor Kazal type 13 (SPINK13) is a secreted protein that has been recently studied as a therapeutic drug and an interesting biomarker for cancer cells. Although SPINK13 has a consensus sequence (Pro-Asn-Val-Thr) for N-glycosylation, the existence of N-glycosylation and its functions are still unclear. In addition to this, the preparation of glycosylated SPINK 13 has not been examined by both the cell expression method and chemical synthesis. Herein we report the chemical synthesis of the scarce N-glycosylated form of SPINK13 by a rapid synthetic method combined with the chemical glycan insertion strategy and a fast-flow SPPS method. Glycosylated asparagine thioacid was designed to chemoselectively be inserted between two peptide segments where is the sterically bulky Pro-Asn(N-glycan)-Val junction by two coupling reactions which consist of diacyl disulfide coupling (DDC) and thioacid capture ligation (TCL). This insertion strategy successfully afforded the full-length polypeptide of SPINK13 within two steps from glycosylated asparagine thioacid. Because the two peptides used for this synthesis were prepared by a fast-flow SPPS, the total synthetic time of glycoprotein was considerably shortened. This synthetic concept enables us to repetitively synthesize a target glycoprotein easily. Folding experiments afforded well-folded structure confirmed by CD and disulfide bond map. Invasion assays of glycosylated SPINK13 and non-glycosylated SPINK13 with pancreatic cancer cells showed that non-glycosylated SPINK-13 was more potent than that of glycosylated SPINK13.

Design and Synthesis of Glycosylated Cholera Toxin B Subunit as a Tracer of Glycoprotein Trafficking in Organelles of Living Cells.

Glycosylation of proteins is known to be essential for changing biological activity and stability of glycoproteins on the cell surfaces and in body fluids. Delivering of homogeneous glycoproteins into the endoplasmic reticulum (ER) and the Golgi apparatus would enable us to investigate the function of asparagine-linked (N-) glycans in the organelles. In this work, we designed and synthesized an intentionally glycosylated cholera toxin B-subunit (CTB) to be transported to the organelles of mammalian cells. The heptasaccharide, the intermediate structure of various complex-type N-glycans, was introduced to the CTB. The synthesized monomeric glycosyl-CTB successfully entered mammalian cells and was transported to the Golgi and the ER, suggesting the potential use of synthetic CTB to deliver and investigate the functions of homogeneous N-glycans in specific organelles of living cells.

Semisynthesis of a Homogeneous Glycoprotein Using Chemical Transformation of Peptides to Thioester Surrogates.

Semisynthesis using recombinant polypeptides as building blocks is a powerful approach for the preparation of proteins with a variety of modifications such as glycosylation. The activation of the C terminus of recombinant peptides is a key step for coupling peptide building blocks and preparing a full-length polypeptide of a target protein. This article reports two chemical approaches for transformation of the C terminus of recombinant polypeptides to thioester surrogates. The first approach relies on efficient substitution of the C-terminal Cys residue with bis(2-sulfanylethyl)amine (SEA) to yield peptide-thioester surrogates. The second approach employs a native tripeptide, cysteinyl-glycyl-cysteine (CGC), to yield peptide-thioesters via a process mediated by a thioester surrogate. Both chemical transformation methods employ native peptide sequences and were thereby successfully applied to recombinant polypeptides. As a consequence, we succeeded in the semisynthesis of a glycosylated form of inducible T cell costimulator (ICOS) for the first time.

Optimizing the Semisynthesis towards glycosylated interferon-ß-polypeptide by utilizing bacterial protein expression and chemical modification.

The synthesis of a sufficient amount of homogeneous glycoprotein is of great interest because natural glycoproteins show considerable heterogeneity in oligosaccharide structures, making the studies on glycan structure-function relationship difficult. Herein, we report optimized methods that can accelerate the semisynthesis of homogeneous glycoproteins based on recombinant expression and chemical conversion. Peptide thioesters and peptides with Cys residues at their N-terminals are necessary intermediates to perform native chemical ligation. We successfully performed thioesterification for a peptide prepared in E. coli via Cys-cyanylation at its C-terminal followed by hydrazinolysis and acidic thiolysis. These optimized conditions could tolerate an acid labile Thz protected Cys at the N-terminal of a peptide-hydrazide and specific cyanylation of the C-terminal Cys to yield a peptide thioester. To reduce the amount of precious oligosaccharide that is required in the conventional SPPS method, an improved liquid phase glycopeptide coupling was also optimized in a good yield (46% over four steps). Lastly, chemoselective protection of the internal cysteines and activation of the N-terminal cysteine were optimized toward a long peptide prepared in E. coli. By using these strategies, a full-length interferon-ß glycosyl polypeptide as a model was successfully obtained.

Glycoprotein Semisynthesis by Chemical Insertion of Glycosyl Asparagine Using a Bifunctional Thioacid-Mediated Strategy.

Glycosylation is a major modification of secreted and cell surface proteins, and the resultant glycans show considerable heterogeneity in their structures. To understand the biological processes arising from each glycoform, the preparation of homogeneous glycoproteins is essential for extensive biological experiments. To establish a more robust and rapid synthetic route for the synthesis of homogeneous glycoproteins, we studied several key reactions based on amino thioacids. We found that diacyl disulfide coupling (DDC) formed with glycosyl asparagine thioacid and peptide thioacid yielded glycopeptides. This efficient coupling reaction enabled us to develop a new glycoprotein synthesis method, such as the bifunctional thioacid-mediated strategy, which can couple two peptides with the N- and C-termini of glycosyl asparagine thioacid. Previous glycoprotein synthesis methods required valuable glycosyl asparagine in the early stage and subsequent multiple glycoprotein synthesis routes, whereas the developed concept can generate glycoproteins within a few steps from peptide and glycosyl asparagine thioacid. Herein, we report the characterization of the DDC of amino thioacids and the efficient ability of glycosyl asparagine thioacid to be used for robust glycoprotein semisynthesis.

Chemical Synthesis and Characterization of a Nonfibrillating Glycoglucagon.

The current commercially available glucagon formulations for the treatment of severe hypoglycemia must be reconstituted immediately prior to use, owing to the susceptibility of glucagon to fibrillation and aggregation in an aqueous solution. This results in the inconvenience of handling, misuse, and wastage of this drug. To address these issues, we synthesized a glycosylated glucagon analogue in which the 25th residue (Trp) was replaced with a cysteine (Cys) and a Br-disialyloligosaccharide was conjugated at the Cys thiol moiety. The resulting analogue, glycoglucagon, is a highly potent full agonist at the glucagon receptor. Importantly, glycoglucagon exhibits markedly reduced propensity for fibrillation and enhanced thermal and metabolic stability. This novel analogue is thus a valuable lead for producing stable liquid glucagon formulations that will improve patient compliance and minimize wastage.

Chemical Synthesis of an Erythropoietin Glycoform Having a Triantennary N-Glycan: Significant Change of Biological Activity of Glycoprotein by Addition of a Small Molecular Weight Trisaccharide.

The glycosylation of proteins contributes to the modulation of the structure and biological activity of glycoproteins. Asparagine-linked glycans (N-glycans) of glycoproteins naturally exhibit diverse antennary patterns, such as bi-, tri-, and tetra-antennary forms. However, there are no chemical or biological methods to obtain homogeneous glycoproteins via the intentional alteration of the antennary form of N-glycans. Thus, the functions of the individual antennary form of N-glycan at a molecular level remain unclear. Herein, we report the chemical synthesis of an erythropoietin (EPO) glycoform having a triantennary sialylglycan at position 83, as well as two biantennary sialylglycans at both positions 24 and 38. We demonstrated efficient liquid-phase condensation reactions to prepare a sialylglycopeptide having a triantennary N-glycan prepared by the addition of a Neu5Ac-α-2,6-Gal-ß-1,4-GlcNAc element to the biantennary glycan under semisynthetic conditions. The molecular weight of the newly added antennary element was ∼3% of the EPO glycoform, and the introduced position was the most distant from the bioactive protein. However, in vivo assays using mice revealed that the additional antennary element at position 83 dramatically increased the hematopoietic activity compared to a commercially available native EPO. These unprecedented data clearly indicate that the antennary pattern of N-glycans inherently plays a critical role in the modulation of protein functions.

Total Chemical Synthesis of a Nonfibrillating Human Glycoinsulin.

Glycosylation is an accepted strategy to improve the therapeutic value of peptide and protein drugs. Insulin and its analogues are life-saving drugs for all type I and 30% of type II diabetic patients. However, they can readily form fibrils which is a significant problem especially for their use in insulin pumps. Because of the solubilizing and hydration effects of sugars, it was thought that glycosylation of insulin could inhibit fibril formation and lead to a more stable formulation. Since enzymatic glycosylation results in heterogeneous products, we developed a novel chemical strategy to produce a homogeneous glycoinsulin (disialo-glycoinsulin) in excellent yield (∼60%). It showed a near-native binding affinity for insulin receptors A and B in vitro and high glucose-lowering effects in vivo, irrespective of the route of administration (s.c. vs i.p.). The glycoinsulin retained insulin-like helical structure and exhibited improved stability in human serum. Importantly, our disialo-glycoinsulin analogue does not form fibrils at both high concentration and temperature. Therefore, it is an excellent candidate for clinical use in insulin pumps.

Chemical-Synthesis-Based Approach to Glycoprotein Functions in the Endoplasmic Reticulum.

The introduction of Asn-linked glycans to nascent polypeptides occurs in the lumen of the endoplasmic reticulum of eukaryotic cells. After the removal of specific sugar residues, glycoproteins acquire signals in the glycoprotein quality control (GPQC) system and enter the folding cycle composed of lectin-chaperones calnexin (CNX) and calreticulin (CRT), glucosidase II (G-II), and UDP-Glc:glycoprotein glucosyltransferase (UGGT). G-II initiates glycoproteins' entry and exit from the cycle, and UGGT serves as the "folding sensor". This account summarizes our effort to analyze the properties of enzymes and lectins that play important roles in GPQC, especially those involved in the CNX/CRT cycle. To commence our study, general methods for the synthesis of high-mannose-type glycans and glycoproteins were established. Based on these, various substrates to analyze components of the GPQC were created, and properties of CRT, G-II, and UGGT have been clarified.

Identification of the epitope of 10-7G glycan antibody to recognize cancer-associated haptoglobin.

We previously identified fucosylated haptoglobin (Fuc-Hpt) as a clinical serum biomarker of pancreatic cancer and established the novel glycan monoclonal antibody (mAb) 10-7G. This antibody recognizes cancer-associated haptoglobin including Fuc-Hpt and the precursor of haptoglobin. Interestingly, Western blot analysis showed that the 10-7G mAb reacts with the haptoglobin α chain, which has no N-glycan potential sites; haptoglobin ß chain has four N-glycan sites. In this study, we identified the epitope for the 10-7G mAb using haptoglobin deletion mutants, as well as inhibition ELISA with recombinant peptides. We illustrated molecular graphics to show a relationship between the epitope and the ß chain. Furthermore, we hypothesized that the 10-7G mAb minimally recognizes normal haptoglobin, but aberrant glycosylation on the ß chain causes conformational changes, enabling the 10-7G mAb to easily access the epitope within the α chain. Because 10-7G values, an enzyme-linked immunosorbent assay-immobilized 10-7G mAb, in patients with pancreatic cancer varied by haptoglobin phenotype, the amount of aberrant glycosylation needed to affect haptoglobin conformation probably depends on haptoglobin phenotype. Taken together, the 10-7G mAb recognized characteristic peptides on the haptoglobin α chain as a result of conformational changes and is a promising tool for diagnosing pancreatic cancer by haptoglobin phenotype.

Chemical Synthesis of Ubiquitinated High-Mannose-Type N-Glycoprotein CCL1 in Different Folding States.

Degradation of misfolded glycoproteins by the ubiquitin-proteasome system (UPS) is a very important process for protein homeostasis. To demonstrate the accessibility toward a ubiquitinated glycoprotein probe for the study of glycoprotein degradation by UPS, we synthesized ubiquitinated glycoprotein CC motif chemokine 1 (CCL1) bearing a high-mannose-type N-glycan, starting from six peptide segments. A native isopeptide linkage was constructed using δ-thiolysine (thioLys)-mediated chemical ligation. CCL1 glycopeptide with a high-mannose-type N-glycan as well as a δ-thioLys residue was synthesized chemically. The chemical ligation between δ-thioLys-containing glycopeptide and ubiquitin-α-thioester successfully yielded a ubiquitinated glycopeptide with a native isopeptide bond after desulfurization, even in the presence of a large N-glycan. In vitro folding experiments under reduced and redox conditions gave the desired two types of ubiquitinated glycosylated CCL1s, consisting of unfolded CCL1 and folded ubiquitin, and the folded form of both CCL1 as well as ubiquitin. We achieved the chemical synthesis of a complex protein molecule that contains not only the two major post-translational modifications, ubiquitination and glycosylation, but also controlled folding states of ubiquitin and CCL1. These chemical probes could have useful applications in the study of complex ubiquitin biology and glycobiology.

Acceleration and Deceleration Factors on the Hydrolysis Reaction of 4,6-O-Benzylidene Acetal Group.

The benzylidene acetal group is one of the most important protecting groups not only in carbohydrate chemistry but also in general organic chemistry. In the case of 4,6-O-benzylidene glycosides, we previously found that the stereochemistry at 4-position altered the reaction rate constant for hydrolysis of benzylidene acetal group. However, a detail of the acceleration or deceleration factor was still unclear. In this work, the hydrolysis reaction of benzylidene acetal group was analyzed using the Arrhenius and Eyring plot to obtain individual parameters for glucosides (Glc), mannosides (Man), and galactosides (Gal). The Arrhenius and Eyring plot indicated that the pre-exponential factor (A) and ΔS⧧ were critical for the smallest reaction rate constant of Gal among nonacetylated substrates. On the other hand, both Ea/ΔH⧧ and A/ΔS⧧ were influential for the smallest reaction rate constant of Gal among diacetylated substrates. All parameters obtained suggested that the rate constant for hydrolysis reaction was regulated by protonation and hydration steps along with solvation. The obtained parameters support wide use of benzylidene acetal group as orthogonal protection of cis- and trans-fused bicyclic systems through the fast hydrolysis of the trans-fused benzylidene acetal group.

Regioselective α-Peptide Bond Formation Through the Oxidation of Amino Thioacids.

Biological systems, including ribosomes and enzymes, produce peptides with an extraordinary high speed and accuracy. On the other hand, a rational and regioselective α-peptide bond formation, without involving protecting groups, is difficult to achieve in chemical synthesis. In this study, α-amino thioacids were utilized for the generation of polypeptides without using any protecting groups. We found that an α-amino thioacid could oxidatively form a diaminoacyl-disulfide moiety and undergo a subsequent intramolecular S- to N-acyl transfer to form an α-peptide bond. Even the thioacid form of lysine, which has a free ε-amino group, generated a regioselective α-peptide bond. The oxidation of amino thioacids generated the oligomers of amino acids. Interestingly, this oligomerization reaction proceeded even in the presence of iron ore, a prebiotic element, thus suggesting a plausible prebiotic peptide bond forming reaction.

Chemical Modification of the N Termini of Unprotected Peptides for Semisynthesis of Modified Proteins by Utilizing a Hydrophilic Protecting Group.

A simple and efficient strategy for the selective modification of the peptide N terminus with an unnatural amino acid is described. A peptide having a SUMO-HisTag-TEV sequence (SUMO: small ubiquitin-related modifier, TEV: tobacco etch virus) preceding the N terminus of the target peptide was designed. Recombinant expression in E. coli and subsequent SUMO protease cleavage yielded the HisTag-TEV-target peptide. Partial protection of the lysine side chains of this peptide with d-glucopyranosyloxycarbonyl and removal of the HisTag-TEV sequence by TEV protease yielded the partially protected peptide with a free N-terminal amine. Coupling of selenocysteine selectively at the N terminus and subsequent acidic deprotection of the carbohydrate protecting groups yielded a modified peptide that can be used for native chemical ligation (NCL). As a proof of concept, the modification of a longer recombinant peptide with selenocysteinylserine (GalNAc) at the N terminus was demonstrated.

Monitoring of Glycoprotein Quality Control System with a Series of Chemically Synthesized Homogeneous Native and Misfolded Glycoproteins.

The glycoprotein quality control (GQC) system in the endoplasmic reticulum (ER) effectively uses chaperone-type enzymes and lectins such as UDP-glucose:glycoprotein glucosyltransferase (UGGT), calnexin (CNX), calreticulin (CRT), protein disulfide bond isomerases (ERp57 or PDIs), and glucosidases to generate native-folded glycoproteins from nascent glycopolypeptides. However, the individual processes of the GQC system at the molecular level are still unclear. We chemically synthesized a series of several homogeneous glycoproteins bearing M9-high-mannose type oligosaccharides (M9-glycan), such as erythropoietin (EPO), interferon-ß (IFN-ß), and interleukin 8 (IL8) and their misfolded counterparts, and used these glycoprotein probes to better understand the GQC process. The analyses by high performance liquid chromatography and mass spectrometer clearly showed refolding processes from synthetic misfolded glycoproteins to native form through folding intermediates, allowing for the relationship between the amount of glucosylation and the refolding of the glycoprotein to be estimated. The experiment using these probes demonstrated that GQC system isolated from rat liver acts in a catalytic cycle regulated by the fast crosstalk of glucosylation/deglucosylation in order to accelerate refolding of misfolded glycoproteins.

Semisynthesis of Complex-Type Biantennary Oligosaccharides Containing Lactosamine Repeating Units from a Biantennary Oligosaccharide Isolated from a Natural Source.

Poly-N-acetyllactosamine (poly-LacNAc) structures on glycoproteins play important roles in essential biological events such as cell-cell adhesion. Here, we report a new strategy for the semisynthesis of LacNAc-extended complex-type biantennary oligosaccharides. We found an efficient isopropylidenation reaction that selectively protects the terminal Gal-3,4-OH of a biantennary complex-type nonasaccharide isolated from a natural source. This finding enabled the conversion of the nonasaccharide into the two types of oligosaccharides containing di-LacNAc units at one or two antennae via ten-step chemical sequences.

Effects of N-Glycans on Glycoprotein Folding and Protein Dynamics.

This chapter describes the folding of synthetic homogeneous glycosylpolypeptides into glycoproteins depending on the position and number of glycosylation sites and oligosaccharide structures. To evaluate the role of oligosaccharides in protein folding, we synthesized small glycoprotein models, homogeneous misfolded glycoproteins, and erythropoietins. In addition to these chemical syntheses, this chapter introduces a unique method for 15N-labeling of synthetic glycoproteins to enable structural analysis. Based on experimental results, it can be suggested that N-glycans stabilize the structure of glycoproteins.

Functional analysis of endoplasmic reticulum glucosyltransferase (UGGT): Synthetic chemistry's initiative in glycobiology.

UGGT1 is called as a folding sensor protein that recognizes misfolded glycoproteins and selectively glucosylates high-mannose-type glycans on the proteins. However, conventional approaches using naturally occurring glycoproteins is not optimum in performing precise analysis of the unique properties of UGGT1. We have demonstrated that high-mannose-type glycans, in which various hydrophobic aglycons were introduced, act as good substrates for UGGT1 and are useful analytical tools for its characterization. Moreover, we found that UGGT2, an isoform UGGT1, is also capable of glucosylating these synthetic substrates. Our strategy stemmed on synthetic chemistry has been further strengthened by total synthesis of homogeneous glycoproteins in correctly folded as well as in intentionally misfolded forms.