Verena Böhrsch

Glycan‐Specific Metabolic Oligosaccharide Engineering of C7‐Substituted Sialic Acids

Intact and integral glycosylation of membrane-associated as well as secreted glycoproteins has been shown to be essential for many aspects of the proper function of biological systems. Recombinantly expressed glycoproteins, such as antibodies, growth factors, hormones, vaccines, and contrast agents are key elements in medical applications. The quality of these therapeutically administered glycoproteins can be efficiently improved by the incorporation of chemically functionalized monosaccharides into their glycan moieties, a process denoted as metabolic oligosaccharide engineering (MOE). In addition to these pharmaceutical applications, MOE has greatly advanced diagnostics by localizing and visualizing glycans even in living animals. To date, a multitude of chemically modified monosaccharides have been designed for MOE applications. Owing to their terminal position at glycan structures of glycoproteins and relevance for cellular recognition, sialic acids and their metabolic precursor N-acetylmannosamine (ManNAc), are the most prominent targets for MOE. Several ManNAc derivatives with N-acetyl side-chain modifications have been synthesized and metabolically incorporated by the sialic acid biosynthetic pathway into a corresponding sialic acid C5 analogue (Figure 1). This approach was beneficial to extending the understanding of the biological role of the N-acyl side chain of sialic acids, for example, in virus infection or neuronal differentiation. Alternatively, C9 modifications of sialosides have also been achieved by directly administering synthetic sialic acid analogues. Additionally, selective cleavage of the glycol moiety led to a truncated sialic acid equipped glycans with an aldehyde for labeling reactions (Figure 1). All of these modifications address sialylation of both, Nand O-glycosylation of glycoproteins, to almost the same extent. Herein we investigate whether the biosynthetic machinery for sialic acids also tolerates other ManNAc derivatives as substrates, which are modified directly at the six-membered carbohydrate ring. The modification of the C4 position appeared most attractive, because it is not enzymatically modified during cellular glycoprotein production and would deliver previously unknown C7-modified sialic acid containing glycoproteins (Figure 1). To probe the biosynthetic promiscuity, we targeted a C4-modified ManNAc derivative, N-acetyl-4-azido-4-deoxymannosamine (4-azido-ManNAc, 1), in our study to enable postglycosylational conjugation and visualization by bioorthogonal reactions. N-acetyl-(1,3,6-O-acetyl)-4-azido-4-deoxy-mannosamine (Ac3-4-azido-ManNAc) was generated by an optimized literature method (Figure S1 in the Supporting InformaFigure 1. Methods for the structural modification of glycan-bound sialic acids by application of chemically modified ManNAc or direct periodate oxidation of glycan-bound sialic acids (left). Specific modification of the C7 position of sialic acids was achieved by C4-modified ManNAc in this study (right; note that to date these methods were carried individually, resulting in only one modification of a single sialic acid molecule).

Suzuki–Miyaura Couplings on Proteins: A Simple and Ready‐to‐use Catalytic System in Water

The palladium-catalyzed Suzuki–Miyaura coupling, which is often referred to as the Suzuki coupling, has become a widely used reaction since its introduction in 1979. 2] The reaction allows an E-selective C C bond formation between sp or sp hybridized carbons, specifically between vinyl or aryl halides and aryl, vinyl, or alkinyl boronic acids. Of the compendium of palladium-catalyzed cross coupling reactions, the Suzuki coupling is particularly attractive due to the stability of the starting materials, their tolerance of functional groups, and their compatibility to a range of solvents, including water. 4] The low toxicity of organoboron compounds in comparison with, for example, organotin compounds used in Stille-couplings, contributes even more to the appeal of the reaction. Consequently, the Suzuki-coupling has already been envisioned as a useful chemical tool for the modification of biologically relevant compounds, especially peptides and proteins. Following the pioneering work by Beletskaya and coworkers in 1989, who coupled water-soluble aryl iodides with boronic acids catalyzed by Pd salts in water, the use of aqueous media for Suzuki couplings on a broader range of substrates has been widely established since 1995. 10] However, mostly mixtures containing 20–30 % of water in acetonitrile, dimethoxyethane, or alcohols were applied, with the water having a positive effect on both, reactivity and yield. More recently, the Suzuki-coupling was performed in pure water on a broader range of substrates using phosphine ligands with a sulfate group for better solubility or N-heterocyclic carbenes. However, these ligands are expensive, not easily accessible, and elevated temperatures and relatively high catalyst loadings (1–4 % in comparison to organic media) are required. Other systems that use heterogenic conditions or surfactants have given excellent results on difficult organic substrates. 13–18] Among the first post-synthetically modified peptides via Suzuki reaction were achieved by the Kotha group for the combinatorial synthesis of biologically active peptide libraries. In addition, the Suzuki-coupling has proven to be effective in solid-phase peptide synthesis and for the heterogeneous modification of peptides. Another interesting application reported by van Vranken and coworkers was the intermolecular crosslinking of two tyrosines in peptides by a Suzuki coupling, although the use of longer peptides inhibited the dimerization reaction. Alternatively, a two-step palladium-catalyzed coupling procedure could be successfully applied to the dimerization of a heptapeptide. Despite these successful transformations of peptides, all of the reactions were performed in organic media and with elevated temperatures, which usually leads to protein denaturation. These challenges were met first by the group of Hamachi who performed the Suzuki coupling on a synthetic protein domain, which maintained structure and function after the catalytic reaction. 27] However, 50 % glycerol in the reaction mixture was needed for complete conversion within several hours. In a second example, an artificial handle on a protein was synthetically modified, again using organic additives. Here, the Suzuki-coupling turned out to be less efficient than a coppercatalyzed reaction. In a recent contribution by Ben Davis and co-workers, a catalytic system previously used for the air-sensitive Sonogashira coupling in organic media was adopted to the modification of proteins in water by use of the Suzuki-coupling. This beneficial ligand system is inexpensive, completely water soluble, and highly active. In their study, they describe an extensive investigation, which is documented by a remarkable data set in the supporting information, of both biological and classical synthetic applications. By reaction of 2-amino-4,6-dihydroxypyrimidine with palladium acetate under basic conditions, a salt was formed, which shows catalytic activity under the reaction conditions (Scheme 1). The catalyst can be used as a stock solution and can couple sterically hindered and functionalized biaryls in excellent yields, in which catalyst loadings of 0.01–1 % Pd can withstand any competition in the aqueous solution. Even loadings as low as 10 4 % Pd required only moderately extended reaction times under microwave conditions for the coupling of aryliodides. The transfer of the pyrimidine catalyst system to the Suzuki coupling of peptides and proteins was first probed by the modification of several, commercially available aryl halide-containing amino acids. Para-bromo and para-iodo phenylalanine could be coupled to phenyl boronic acid in very good yields (92–95 %) under conditions suitable for proteins (37 8C, pH 8, open air). However, the method did not convert the corresponding chloride derivative. Additionally, thioethers at the Cterminus, which occur naturally in methionine, and free thiols in cysteine, appeared to inhibit the catalyst. [a] Dr. V. Bçhrsch , Dr. C. P. R. Hackenberger Institute for Chemistry and Biochemistry, FU Berlin Takustr. 3, 14195 Berlin Fax: (+ 49) 30-838-52551 E-mail : hackenbe@chemie.fu-berlin.de [b] Dr. V. Bçhrsch Beuth Hochschule f r Technik Berlin Fachbereich Life Sciences & Technology Labor f r Biochemie, Seestrasse 64, 13347 Berlin

Site-Specific Modification of Proteins by the Staudinger-Phosphite Reaction

Chemoselective reactions are important tools for the modification of peptides and proteins. Thereby the modification is desired to be site specific and bioorthogonal. Here we describe the site-specific modification of azido-proteins via a Staudinger-type phosphite ligation. The reaction was carried out in aqueous system on proteins containing p-azido-phenylalanine in a single position introduced by the amber codon technique. A selective introduction of branched polyethylene scaffolds can be achieved with the application of the methodology reported herein.

N-Azidoacetylmannosamine and N-Azidoacetylgalactosamine Incorporation into N-Glycans of Recombinantly Expressed Human Lactotransferrin by Metabolic Oligosaccharide Engineering

Over the past years the application of therapeutic and diagnostic proteins such as antibodies or hormones has had a substantial impact on modern medicine. Many of these agents are glycoproteins with conjugated oligosaccharide chains. The introduction of chemically modified monosaccharides into glycan structures of such proteins provides an excellent option to modify their efficacy, affinity, and tolerance. Here we demonstrate the efficient introduction of azido-modified N-acetylmannosamine and N-acetylgalactosamine into glycans of the promising protein therapeutic candidate lactotransferrin (LTF).

Applying Acylated Fucose Analogues to Metabolic Glycoengineering

Manipulations of cell surface glycosylation or glycan decoration of selected proteins hold immense potential for exploring structure-activity relations or increasing glycoprotein quality. Metabolic glycoengineering describes the strategy where exogenously supplied sugar analogues intercept biosynthetic pathways and are incorporated into glycoconjugates. Low membrane permeability, which so far limited the large-scale adaption of this technology, can be addressed by the introduction of acylated monosaccharides. In this work, we investigated tetra-O-acetylated, -propanoylated and -polyethylene glycol (PEG)ylated fucoses. Concentrations of up to 500 µM had no substantial effects on viability and recombinant glycoprotein production of human embryonic kidney (HEK)-293T cells. Analogues applied to an engineered Chinese hamster ovary (CHO) cell line with blocked fucose de novo synthesis revealed an increase in cell surface and recombinant antibody fucosylation as proved by lectin blotting, mass spectrometry and monosaccharide analysis. Significant fucose incorporation was achieved for tetra-O-acetylated and -propanoylated fucoses already at 20 µM. Sequential fucosylation of the recombinant glycoprotein, achieved by the application of increasing concentrations of PEGylated fucose up to 70 µM, correlated with a reduced antibody’s binding activity in a Fcγ receptor IIIa (FcγRIIIa) binding assay. Our results provide further insights to modulate fucosylation by exploiting the salvage pathway via metabolic glycoengineering.