Design and Characterization of Galectin-3 Fusion Proteins and Novel Multivalent Galectin-3 Ligands
Von der Fakultät für Mathematik, Informatik und Naturwissenschaften der RWTH Aachen University zur Erlangung des akademischen Grades einer Doktorin der Naturwissenschaften genehmigte Dissertation
vorgelegt von
Master of Science (M.Sc.) Biotechnologie
Sophia Böcker
aus
Eisenhüttenstadt
Berichter: Universitätsprofessor Dr. rer. nat. Lothar Elling
Universitätsprofessor Dr. rer. nat. Wilhelm Jahnen-Dechent
Tag der mündlichen Prüfung: 24.04.2018
Diese Dissertation ist auf den Internetseiten der Universitätsbibliothek verfügbar.
Acknowledgment
My first thanks go to Univ.-Prof. Dr. rer. nat. Lothar Elling for his supervision, for letting me work on this interesting topic and dive into the world of galectins and sugars. Thank you for being a member of the laboratory for biomaterials.
I also gratefully thank Univ.-Prof. Dr. rer. nat. Wilhelm Jahnen-Dechent for kindly agreeing to be my second reviewer.
Furthermore, I thank Univ.-Prof. Dr. rer. nat. Marc Spehr to be my third examiner.
I enjoyed all the inspiring collaborations and want to thank: Dr. rer. nat. Sebastian Lowins, each experiment with you was a lot of fun; Anne Rix, you enabled cell experiments with my proteins, thank you especially for performing FACS-analysis with HUVECs.
A special thanks to Prof. Dr. rer. nat. Andreas Walther and Prof. Dr. rer. nat. Martin Möller (DWI Leibniz Institute for Interactive Materials, Aachen) for the opportunity to perform measurements using their SPR device.
Of course, I have to thank for the financial support. Thanks for being a scholarship holder of the Research Training Group “GRK 1035: Biointerface – Detection and Control of Interface- induced Biomolecular and Cellular Functions” by DFG. Other thanks go to the excellence initiative of the German federal and state governments through ERS@RWTH Aachen University.
I thank all the students I supervised for their participation in my topic and for giving me the chance to practice leadership. I learned a lot of new perceptions.
I want to cordially thank all present and former members of the laboratory for biomaterials, especially Anna Eisele, Bastian Lange, Dr. rer. nat. Christiane Kupper, Dennis Hirtz, Dominic Laaf, Manja Henze, Dr. rer. nat. Ruben R. Rosencrantz and Thomas Fischöder. Thank you for the nice working atmosphere. An extra thanks to Manja for not breaking-down the contact and all the motivating talks.
The final thanks are the biggest thanks and go to my family and to the very best, Ruben. Without you this thesis would never have been finished. Thank you for your help, support and motivation.
Publications derived from this work
S. Böcker, L. Elling, Biotinylated N-Acetyllactosamine and N,N-Diacetyllactosamine based Oligosaccharides as Novel Ligands for Human Galectin-3. Bioengineering 2017, 4(2), 31.
S. Böcker, L. Elling, Binding Characteristics of Galectin-3 Fusion Proteins. Glycobiology 2017, 27, 457-468.
S. Böcker, D. Laaf, L. Elling, Galectin binding to neo-glycoproteins: LacDiNAc conjugated BSA as ligand for human galectin-3. Biomolecules 2015, 5, 1671-1696.
A. Šimonová, C. E. Kupper, S. Böcker, A. Müller, K. Hofbauerová, H. Pelantová, L. Elling, V. Křen, P. Bojarová, Chemo-enzymatic synthesis of LacdiNAc dimers of varying length as novel galectin ligands. Journal of Molecular Catalysis B: Enzymatic 2014, 101, 47-55.
C. E. Kupper, S. Böcker, H. L. Liu, C. Adamzyk, J. van de Kamp, T. Recker, B. Lethaus, W. Jahnen-Dechent, S. Neuss, G. Müller-Newen, L. Elling, Fluorescent SNAP-tag galectin fusion proteins as novel tools in glycobiology. Current Pharmaceutical Design 2013, 19, 5457-5467.
Presentations
S. Böcker, D. Laaf, L. Elling (2015), lecture/poster: “LacDiNAc Conjugated BSA: a Neo- glycoprotein as Multivalent and Selective Ligand for Galectin-3”, COST Action CM1102 MultiGlycoNano Spring Training School, Bangor, UK
S. Böcker (2013), lecture: “Binding Studies of Truncated Galectin-3“, 24th Joint Glycobiology Meeting, Halle-Wittenberg, Germany
S. Böcker, C. E. Kupper, L. Elling (2013), lecture/poster: “Fluorescent SNAP-Tag Galectin-3 Fusion Protein as a Novel Tool in Glycobiology”, 10th Carbohydrate Bioengineering Meeting, Prague, Czech Republic
S. Böcker, C. E. Kupper, L. Elling (2012), poster: “Fluorescent SNAP-Tag Galectin-3 Fusion Protein as a Novel Tool in Glycobiology”, 23rd Joint Glycobiology Meeting, Wageningen, The Netherlands
Table of contents
Abbreviations ...... I Summary ...... III Zusammenfassung ...... IV 1. Introduction ...... 1 1.1 Carbohydrates – The sweet world ...... 1 1.2 Galectins – Sweet and sour effects ...... 10 1.3 Design of galectin-3 ligands – Turning sweet into affine ...... 17 1.4 The aim ...... 26 1.4.1 Design, cloning, expression, and characterization of truncated galectin-3 ...... 26 1.4.2 Design of novel multivalent neo-glycoproteins as promising ligands for galectin-3 ...... 26 1.5 References ...... 28 2. Binding characteristics of galectin-3 fusion proteins – Influence of truncation and fusion . 39 Abstract ...... 39 2.1 Introduction ...... 40 2.2 Materials and methods ...... 41 2.2.1 Cloning of galectin constructs ...... 41 2.2.2 Expression and purification ...... 42 2.2.3 SDS-PAGE and western blot ...... 43 2.2.4 Size exclusion chromatography ...... 43 2.2.5 Immobilization of recombinant galectin to Sepharose beads ...... 43 2.2.6 Self-association/crosslinking assays of galectin-3 ...... 44 2.2.7 Galectin binding assay on asialofetuin ...... 45 2.2.8 Inhibition of galectin binding with (Di-)LacNAc-linker-tBoc ...... 45 2.2.9 Surface plasmon resonance spectroscopy ...... 45 2.2.10 Flow cytometry with human umbilical vein endothelial cells ...... 46 2.3 Results and discussion ...... 47 2.3.1 Production and characterization of full-length and truncated galectin-3 fusion constructs 47 2.3.2 Self-association/crosslinking potential of galectin-3 fusion proteins ...... 49 2.3.3 Binding of galectin-3 fusion proteins to asialofetuin in a solid-phase assay ...... 52 2.3.4 Surface plasmon resonance spectroscopy of galectin-3 fusion proteins on immobilized asialofetuin ...... 57 2.3.5 Binding of galectins to human umbilical vein endothelial cells ...... 61 2.4 Conclusion ...... 62 2.5 Contributions ...... 63 2.6 References ...... 63
3. Neo-glycoproteins as novel ligands for human galectin-3 –Albumin as carrier for multivalent presentation of non-biotinylated and 6-biotinylated tetrasaccharides to gain high- affinity ligands ...... 67 Abstract ...... 67 3.1 Introduction ...... 68 3.2 Materials and methods ...... 70 3.2.1 Production of recombinant enzymes ...... 70 3.2.2 Chemo-enzymatic synthesis of glycans (4 and 5) ...... 71 3.2.3 Synthesis of biotinylated glycans (9 and 10) ...... 71 3.2.4 Synthesis of squaric acid monoamide esters of non-biotinylated glycans (16 and 17) ...... 72 3.2.5 Synthesis of squaric acid monoamide esters of biotinylated glycans (18 and 19) ...... 73 3.2.6 Conjugation of glycans to bovine serum albumin ...... 73 3.2.7 TNBSA-assay ...... 73 3.2.8 SDS-PAGE and streptavidin blot ...... 74 3.2.9 Expression and purification of recombinant galectins ...... 74 3.2.10 Galectin binding assays on immobilized glycans and neo-glycoproteins ...... 74 3.2.11 Inhibition of galectin binding with neo-glycoproteins ...... 75 3.2.12 Surface plasmon resonance spectroscopy ...... 75 3.3 Results and discussion ...... 76 3.3.1 Chemo-enzymatic synthesis of LacNAc-LacNAc and LacdiNAc-LacNAc ...... 76 3.3.2 Biotinylation of LacNAc-LacNAc and LacdiNAc-LacNAc ...... 77 3.3.3 Synthesis and analysis of neo-glycoproteins ...... 78 3.3.4 Binding of galectin-3 to immobilized tetrasaccharides ...... 85 3.3.5 Binding of galectin-3 and galectin-1 to neo-glycoproteins ...... 86 3.3.6 Galectin-3 binding to neo-glycoproteins at different galectin concentrations ...... 89 3.3.7 Neo-glycoproteins as inhibitors for galectin-3 ...... 98 3.3.8 Neo-glycoproteins as galectin-3 ligands in surface plasmon resonance spectroscopy ...... 103 3.4 Conclusions ...... 105 3.5 Contributions ...... 106 3.6 References ...... 106 Appendix ...... 115 Supporting Information for Chapter 2...... 117 Supporting Information for Chapter 3...... 123
Abbreviations
APTS 8-Aminopyrene-1,3,6-Trisulfonic Acid ASF Asialofetuin BG Benzylguanine BSA Bovine serum albumin CAZy Carbohydrate active enzyme database CRD Carbohydrate recognition domain DMF Dimethyl-formamide DNA Desoxyribonucleic acid DTT Dithiothreitol ECM Extracellular matrix EDC 1-Ethyl-3-(3-dimethylaminopropyl)-carbodiimide EDTA Ethylenediamine tetraacetic acid ELISA Enzyme linked immunosorbent assay EPBS EDTA-containing PBS ER Endoplasmatic reticulum ESI-MS Electrosprayionization mass spectrometry FACS Fluorescence activated cell sorting Fuc Fucose GAG Glycosaminoglycan Gal Galactose Gal-1 Galectin-1 Gal-3 Galectin-3 GalNAc N-acetylgalactosamine GalOx Galactose oxidase GH Glycoside hydrolase Glc Glucose GlcA Glucuronic acid GlcNAc N-acetylglucosamine GnT N-acetylglucosaminyltransferase GT Glycosyltransferase HEPES 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid HPLC High performance liquid chromatography IC50 Half-maximal inhibitory concentration IL Interleukin IMAC Immobilized metal ion affinity chromatography Kd Concentration at half-maximum binding KD Dissociation constant koff Desorption rate kon Adsorption rate LacdiNAc N,N-diacetyllactosamine LacNAc N-acetyllactosamine LB-Medium Lysogeny broth medium LC Liquid chromatography LDS Lithium dodecyl sulfate LNT Lacto-N-tetraose
I
LNnT Lacto-N-neo-tetraose Man Mannose MeOH Methanol MES 2-(N-Morpholino)ethansulfonsäure MMP Matrix-metalloproteinases MOPS 3-(N-Morpholino)propansulfonsäure MS Mass spectrometry MW Molecular weight NeuAc N-acetylneuraminic acid NHS N-hydroxysuccinimide OPD o-phenylenediamine OST Oligosaccharyltransferase PBS Phosphate-buffered saline pp Pyrophoshate ppGalNAcT Polypeptide GalNAc transferase Rf Retardation factor RPM Rounds per minute SDS-PAGE Sodium dodecyl sulfate polyacrylamide gel electrophoresis SEC Size exclusion chromatography SPR Surface plasmon resonance spectroscopy Sulfo-NHS Sulfo-N-hydroxysuccinimide TB-Medium Terrific broth medium tBoc tert-butyloxycarbonyl TNBSA Trinitrobenzene sulfonic acid UDP Uridine-diphosphate VEGF Vascular endothelial growth factor VEGFR VEGF receptor Xyl Xylose YFP Yellow fluorescent protein α3GalT α3-galactosyltransferase β3GlcNAcT β3-N-acetylglucosaminyltransferase β4GalT-1 β4-galactosyltransferase
II
Summary
Galectin-3 (Gal-3) is a key player in a variety of biological processes including cancer progression and immune response driven by specific binding and cross-linking of galactose based carbohydrate ligands. In tumor regions, Gal-3, composed of an N-terminal domain and a carbohydrate recognition domain, can be N-terminally cleaved by metalloproteinases. Truncation of Gal-3 plays an important role in regulation of affinity and self-association. In this work, we aimed for a detailed analysis of Gal-3 from two points of view: Firstly, binding of Gal-3 was analyzed after modification of the protein. Twelve Gal-3 constructs with N-terminal truncation (∆1-62 and ∆1-116) and fusions with SNAP-tag and/or yellow fluorescent protein (YFP) were designed and recombinantly produced. We investigated the influence of the truncation and the fusion partners on Gal-3 binding. Altered binding to asialofetuin (ASF) in ELISA-type and surface plasmon resonance (SPR) binding assays was observed. Highest affinity was proved for Gal-3(Δ1-62) and native Gal-3, respectively, whereas Gal-3(Δ1-116) shows only weak binding. Moreover, we demonstrate here for the first time that SNAP-tag and YFP fusions of Gal-3 and truncated Gal-3 modulated and improved binding affinity to ASF. Fusion of truncated Gal-3 with YFP reconstituted binding properties similar to native Gal-3. In combination with a SNAP-tag even improved binding characteristics were obtained. These results emphasize that the N-terminal domain is important for ligand binding. The self-association potential of Gal-3 seemed not be affected by the modifications leading us to the conclusion that Gal-3 interaction involves the C-terminus. Secondly, we analyzed Gal-3 binding from the ligand’s point of view. As Gal-3 is strongly upregulated in many tumor cells it is a potential target for anti-cancer therapy and cancer diagnosis. In search of high-affinity ligands for Gal-3, we established a method to efficiently synthesize multivalent neo-glycoproteins. By conjugating N-acetyllactosamine (LacNAc) based tetrasaacharides to albumin via the homobifunctional linker squaric acid diethyl ester we obtained glycan-protein conjugates with tunable multivalency with up to 29 binding sites. For the first, time the influence of defined glycan density of neo-glycoproteins on Gal-3 binding was investigated. We observed higher affinity with increasing glycosylation density and found multivalent effects at serum level concentrations of Gal-3. As modifications of glycans can lead to promising ligands, we synthesized novel biotin modified glycans and proved high affinity and selectivity for Gal-3 over galectin-1. After conjugation to albumin, biotinylated neo-glycoproteins achieved high binding levels of Gal-3 at lower glycosylation density compared to non-biotinylated neo-glycoproteins. The efficient and selective binding of our tailor-made neo-glycoproteins make them suitable candidates for targeting Gal-3 in cancer related biomedical research. The two points of view of this work complete the circle of Gal-3 binding to carbohydrate ligands paving the path to future applications in tumor therapy and diagnostics. On the one hand, it gives important insights into the binding characteristics of truncated Gal-3 and how galectins may be tailored by fusions and beneficially tuned for higher binding affinity. On the other hand, it broadens the recent design approaches for producing high-affinity ligands.
III
Zusammenfassung
Galectin-3 (Gal-3) spielt eine Schlüsselrolle in vielen biologischen Prozessen, wie der Tumorprogression und Immunantwort. Dies wird durch spezifische Bindung und Vernetzung von Galaktose-basierenden Glykanen moduliert. In Tumorregionen kann Gal-3, das aus einer N-terminalen Domäne und einer Glykan-Erkennungsdomäne besteht, N-terminal durch Metalloproteinasen gespaltet werden. Die Trunkierung von Gal-3 spielt eine wichtige Rolle bei der Regulation von Affinität und Selbstassoziation. Ziel dieser Arbeit war die detaillierte Analyse von Gal-3 aus zwei Blickwinkeln: Erstens wurde die Bindung von Gal-3 nach Modifikation des Proteins analysiert. Zwölf Gal-3-Konstrukte mit N- terminaler Trunkierung (Δ1-62 und Δ1-116) und Fusionen mit SNAP-Tag und/oder gelb fluoreszierendem Protein (YFP) wurden designt und rekombinant hergestellt. Anschließend wurde der Einfluss der Trunkierung und der Fusionspartner auf die Gal-3-Bindung untersucht. Veränderte Bindung zu Asialofetuin (ASF) wurde in ELISA und Oberflächenplasmonenresonanz (SPR) Bindungsassays beobachtet. Die höchste Affinität wurde für Gal-3(Δ1-62) bzw. natives Gal-3 nachgewiesen, während Gal-3(Δ1-116) nur schwach bindet. Zum ersten Mal wird gezeigt, dass SNAP- Tag und YFP-Fusionen von Gal-3 und trunkiertem Gal-3 die Bindungsaffinität zu ASF modulieren und verbessern. Fusionen von trunkiertem Gal-3 mit YFP stellten die Bindungseigenschaften ähnlich zu nativem Gal-3 wieder her. In Kombination mit einem SNAP-Tag wurden sogar verbesserte Bindungseigenschaften erhalten. Die Ergebnisse zeigen, dass die N-terminale Domäne für die Ligandenbindung wichtig ist. Das Selbstassoziationspotential von Gal-3 schien von den Modifikationen nicht beeinflusst zu sein, was auf die Wechselwirkung von Gal-3 über den C- Terminus hindeutet. Zweitens wurde die Gal-3-Bindung aus Ligandensicht analysiert. Da Gal-3 in vielen Krebszellen hochreguliert wird, ist es ein potenzielles Target für die Tumortherapie und -diagnostik. Auf der Suche nach hochaffinen Gal-3-Liganden wurde eine Methode zur effizienten Synthese von multivalenten Neo-Glykoproteinen etabliert. Durch Konjugation von N-Acetyllactosamin (LacNAc) basierten Tetrasacchariden an Albumin mit Hilfe des homobifunktionellen Linkers Squaratsäurediethylester erhielten wir Glykan-Protein-Konjugate mit einstellbarer Multivalenz mit bis zu 29 Bindungsstellen. Zum ersten Mal wurde der Einfluss der definierten Glykandichte von Neo-Glykoproteinen auf die Gal-3-Bindung untersucht. Die Affinität nahm mit zunehmender Glykosylierungsdichte zu und multivalente Effekte traten bei Gal-3 Konzentrationen, wie sie im Serum vorkommen, auf. Des Weiteren synthetisierten wir neuartige biotin-modifizierte Glykane und wiesen eine hohe Affinität und Selektivität für Gal-3 gegenüber Galectin-1 nach. Nach deren Konjugation an Albumin erreichten biotinylierte Neo-Glykoproteine hohe Gal-3-Bindungsniveaus bei geringerer Glykosylierungsdichte als Nicht-biotinylierte. Die effiziente und selektive Bindung unserer maßgeschneiderten Neo- Glykoproteine macht sie zu geeigneten Kandidaten für das Targeting von Gal-3 in der biomedizinischen Forschung. Die beiden Aspekte dieser Arbeit ebnen den Weg für zukünftige Anwendungen in der Tumortherapie und -diagnostik. Es wurden wichtige Erkenntnisse über die Bindungseigenschaften von trunkiertem Gal-3 geliefert und darüber hinaus, wie Galektine durch Fusionen maßgeschneidert und auf eine höhere Bindungsstärke gebracht werden können. Die Ergebnisse erweitern zudem die neusten Ansätze zum Design hochaffiner Liganden.
IV
1. Introduction
1.1 Carbohydrates – The sweet world Carbohydrates are the building blocks of organism specific glycosylation allowing information-coding as a basis of numerous biochemical signals [1-3]. High variety biocoding is realized by the structural complexity of oligosaccharides. Beyond stability and energy metabolism, the glycan mediated information transfer between biological entities is the most striking feature of carbohydrates in nature [4,5]. Glycans are found on every living higher cell and may be described as antennas reaching into the extracellular space. This makes them ideal structures for allowing communication between a cell and its surroundings. The interaction partners may be other cells or extracellular matrix (ECM) proteins in adhesion processes. Glycans are attached to the cell surface in various ways [6]. They appear linked to lipids of the cell-membrane as glycolipids or as glycoproteins linked to integral membrane proteins. The structures found on glycolipids are biochemically characterized as glycosphingolipids, whereas structures on glycoproteins are mostly divided into the class of N- or O-linked glycans. Additionally, glycosaminoglycans (GAGs), mainly as part of proteoglycans, are found in the ECM. All ‘glycan-antennas’ together form the so called glycocalyx [7]. GAGs are a large class of polymeric glycans which consist of a huge number of linked oligosaccharides and play important roles in biophysical properties of the ECM, hormone interactions and various other physiological processes [8,9]. The most known GAGs are heparin/heparan sulfate, chondroitin/dermatan sulfate and hyaluronic acid. GAGs represent huge polymeric saccharides consisting of repeating units and achieving their function by complex ionic and 3D-structure-driven interactions with their binding partners [10,11]. In contrast, glycans linked to proteins and lipids are better described as core-structure based oligosaccharides with a very defined composition of saccharides [12,13]. The saccharide building-blocks of these glycans are derived from a pool of nine monosaccharides. To explain why sugars are biomolecules with an extremely high coding capacity it helps to take a look at the common coding biomolecules like DNA and proteins: Two DNA bases may be connected to each other in two different ways, the same stands for two amino acids. Two monosaccharides, however, may be connected theoretically in 11 different ways. Taking not only the regio-dependent variation in connecting saccharides into account, but also the type- dependent variation (as there are nine commonly used monosaccharides), the extraordinary high possibilities of building up a glycan becomes obvious. Besides the molecular-based coding abilities of oligosaccharides, the presentation type of glycan-ligands influence the read-out by specific glycan recognizing proteins, called lectins [14-16]. The avidity of lectins
1 towards a certain glycan is modulated by the actual molecule one the one hand, but also by the number of binding sites, which may be altered by glycan branching or e.g. lipid raft formation, on the other [17]. The affinity of a lectin to a single glycan-ligand is rather low and binding constants in the millimolar range are observed. However, it is stated that a dense glycan presentation enhances avidity of lectins in orders of magnitude reaching nanomolar ranged binding constants [18,19]. This is described as “cluster glycoside effect” and the basis of biomaterial research for constructing multivalent glycan presenting materials, which should show strong lectin binding. There are several screws to alter the coding capacity of glycans or the avidity of lectins towards them. But tuning multivalency is for sure the most sensitive one, as it allows the precise tuning of binding efficiency by altering the glycan branching or changing the microenvironment. Other possibilities like changing the glycan sequence and composition are less subtle and better described as on-off-switches for lectin recognition, whereas changes of multivalency allows nature to precisely tune the attracting forces of the glycan towards lectins.
Figure 1.1: Monosaccharides main building-blocks in vertebrates The nine major building-blocks are D-glucose (1), D-N-acetylglucosamine (2), D-glucoronic acid (3), D-galactose (4), D-N-acetylgalactosamine (5), D-mannose (6), L-fucose (7), D-xylose (8) and D-N- acetylneuraminic acid (9).
So far, we focused on glycans important for information transfer located at the border between cells and their surrounding environment. However, the machinery of producing carbohydrate structures relies within the cell and is a multi-compartment process. Besides glycolipids and glycoproteins, there are numerous examples for intracellular glycosylated proteins. In fact, over 50% of all mammalian proteins are glycosylated [12,20,21]. Like their extracellular counterparts, also intracellular glycosylation (O-GlcNAc) is an information coding process, not for cell-cell or cell-ECM interactions, but for proteolytic degradation,
2 transcription control or signal transduction [22]. However, within this work, we will concentrate on the glycans situated in the glycocalyx and their interactions with lectins for extracellular communication processes. As mentioned above, the biosynthesis of glycans is a multi-compartment process and controlled by a vast of different enzymes. In total about 2% of all mammalian genes code solely for enzymes that are related to the glycan biosynthesis machinery [20]. To understand this complex process and also its translation to synthetic approaches, it is worthwhile to take a deeper look at the principles and properties underlying the enzymatic biosynthesis of carbohydrates.
In higher organisms, namely vertebrates, nine different monosaccharides are the main building-blocks of the glycans: glucose (Glc), galactose (Gal), N-acetylglucosamine (GlcNAc), N-acetylgalactosamine (GalNAc), fucose (Fuc), xylose (Xyl), N-acetylneuraminic acid (NeuAc), mannose (Man), and glucuronic acid (GlcA) (Figure 1.1) [23]. Free glycans are rather rare, mostly they are connected to an aglycon forming so called glycoconjugates. Besides lipids, glycoproteins are the major class of glycoconjugates and as well the key- player in numerous biological processes. The glycans on the protein are involved in expression quality control, folding, targeting and receptor-ligand interactions at the interface between cell and ECM. Protein glycosylation may be distinguished by the type of connection between glycan and protein. Links involving the δ-amine of asparagine in the consensus sequence Asn-X-Ser/Thr are named N-glycosidic bonds, whereas links via the terminal hydroxyl group to Ser or Thr are called O-glycosidic bonds [24,25].
N-glycosylation, the most abundant form of glycosylation, may be characterized by defined yet branched glycan structures. The biosynthesis is a multi-compartment route involving various enzymes residing in rough Endoplasmic Reticulum (rER) and Golgi apparatus (Figure 1.2) [26]. Starting point is a lipid-linked precursor which is formed by concerted action of various glycosyltransferases and activated monosaccharides as glycosyldonors, which are synthesized in the cytoplasm. This dolichol linked structure Glc3Man9GlcNAc2-P-P-dolichol, the typical precursor of N-glycans, is transferred en bloc by the key enzyme oligosaccharyltransferase (OST) to an Asn within the consensus sequence in a growing polypeptide chain [25]. Trimming by glucosidase-I/II and quality control by calnexin/calreticulin takes place in the rER before mannosidase-III gives the signal for protein folding and vesicularization of the premature glycoprotein with transport to the cis-Golgi apparatus [27]. Trimming to Man3GlcNAc2 takes place in the Golgi apparatus and involves enzymes like mannosidase-I and mannosidase-II. This minimal structure is the intermediate point where sophisticated biosynthesis steps take place to evolve the complex N-glycan structure by concerted action of at least six different N-acetylglucosaminyltransferases (GnTs)
3
Figure 1.2: Biosynthesis of secreted N-glycoproteins Starting from the production of dolichol-1-p and the addition of N-acetylglucosamine-1-phosphate (GlcNAc-1-p) to form dolichol-p-p-GlcNAc, another GlcNAc and several mannose (Man) are added. After a flip into the lumen of the Endoplasmic Reticulum further Man units as well as glucose (Glc) are transferred. The glycan chain is transferred to a protein backbone via Asparagine by the oligosaccharyltransferase (OST). Glucosidase-I/II (Glc-I/II) split Glc units off again, so that the structure passes the calnexin/calreticulin quality control. Further trimming of Man by mannosidase-III (Man-III) initiates protein folding and transport to the Golgi apparatus, where trimming goes on before processing starts. Man-I/II cleaves and N-acetylglucosaminyltransferases (GnTs) elongates the glycan chain, while fucosyltransferase (FucT), galactosyltransferase and sialyltransferase add fucose (Fuc), galactose (Gal) and N-acetylneuraminic acid (NeuAc). The glycosyltransferases need activated nucleotide sugar as donor substrates (UDP-GlcNAc, GDP-Man, UDP-Glc, GDP-Fuc, UDP-Gal, CMP- NeuAc). Afterwards the finished glycoprotein leaves the cell via secretory vesicles.
[26,28]. The GnTs differ in their substrate preference and therefore address specific mannose residues of the core structure Man3GlcNAc2. However, GnT-I and GnT-III may also act on the premature N-glycan before trimming to Man3GlcNAc2 is completed. This activates or inhibits follow-up glycosylation and increases the complexity and the dynamics of Golgi- resident N-glycosylation. The final outcome is mainly determined by the GnT substrate
4 specificity, the action of preceding GnTs and finally the expression level of the biocatalysts responsible for modifications like fucosylation, galactosylation or sialylation. In this way, di- to penta-antennary complex or hybrid-type N-glycans are synthesized. After reaching the trans-Golgi the matured glycoprotein is further processed and finally enters the secretory pathway.
This complex process leads to the understanding that glycosylation is not directly template- coded within the genetic code, but a highly dynamic process, which makes it difficult to decipher the mechanisms responsible for the formation of certain glycans.
O-glycosylated proteins in mammals are mostly mucin-type glycans [24,29-31]. The typical minimal core structure is the GalNAcα1-OR pattern, which is linked to a Ser or Thr of the polypeptide [32]. This minimalistic consensus sequence on protein level explains the large variety of O-glycosylation patterns on the one hand, but is also crucial for the dense mucin- type glycosylation. In GAGs (Figure 1.4B) like chondroitin sulfate, dermatan sulfate or heparin sulfate as well as derivatives of that the carbohydrate chains are linked to the backbone via xylose. O-glycosylation is involved in diverse biological processes from embryonic development over cell adhesion processes to diseases like cardiovascular risk or cancer, but also plays important roles in maintaining the viscoelasticity of the cell-clusters and tissue [33-36]. The glycosylation patterns are very tissue specific and differ strongly with function. Although in both cases hydrogel-forming mucins for protection of underlying cells
Figure 1.3: Mucin-type O-glycosylation of proteins and the eight core structures In the Golgi apparatus polypeptide-N-acetylgalactosaminyltransferases (ppGalNAcT) transfer N- acetylgalactosamin (GalNAc) from activated UDP-GalNAc to the Serin/Threonin residues of a protein. This starting unit is elongated with further GalNAc, N-acetylglucosamin (GlcNAc) and/or galactose (Gal) via different glycosidic bonds. Eight core structures are available for O-glycans that are not as complex as for N-glycans.
5 and for maintaining smoothness are involved, the glycan structure in mouth, eye and intestine are very different [37,38]. The main enzymes initiating O-glycosylation are polypeptide GalNAc transferases (ppGalNAcTs) [39,40]. In human over 20 different ppGalNacTs are reported with distinct yet overlapping substrate specificity. In contrast to N-glycosylation, eight core structures have been identified for O-glycosylation (Figure 1.3), whereas core 1 and core 2 are the most common, core 3 and core 4 are less abundant and the remaining four core structures are considered as rare [24,41]. While the core structure determines the follow-up glycosylation and is one key-point of the large variety of oligosaccharide structures in O- glycosylation, the influence of modifications of the saccharide backbone, like fucosylation, sialylation, sulfation, methylation and acetylation, cannot be underestimated. Similar to N- glycosylation, also the biosynthetic pathways of O-glycosylation take place at the ER and Golgi-apparatus [42]. An exception is proteoglycan O-glycosylation which takes place only in the Golgi. The regulating enzymes are the aforementioned ppGalNAcTs, which represents a highly conserved enzyme family. The hierarchical action of these enzymes leads finally to the cell- and tissue specific glycosylation of a polypeptide and is the guiding event for the actual glycosylated site, the glycosylation density, and the oligosaccharide modification or extension leading to a matured glycoprotein. Because of the large diversity it is not surprising that mucin-type O-glycosylation is involved on various biological tasks, ranging from cell growth control and proliferation to trafficking of glycoproteins and immunological or signaling pathways. The understanding of the synthesis pathways and the functions of O-glycans will lead to new forms of vaccination, therapy and diagnosis [43].
Both, minimal N- and O-glycans serve as carrier for various terminal glycan structures called epitopes (Figure 1.4A). Important epitopes are LacNAc, Lewis antigens or blood group antigens, which serve as main information carrier for interaction with lectins.
At this point it makes sense to take a deeper look at the biocatalysts responsible for the build- up and trimming of glycans. Two important groups, the glycosyltransferases (GTs) and the glycosylhydrolases (GHs) will be introduced as they are not only the important classes of enzymes in glycosylation modulation, but are also important for in vitro production of glycans.
GTs may be characterized as enzymes that utilize a donor sugar substrate, where the glycan- part is transferred to a nucleophilic group [44,45]. The activated donor contains a leaving group, which is in most cases a phosphate. Commonly, the donor substrates are nucleotide sugars like e.g. UDP-Gal, UDP-Glc or GDP-Man. However, monophosphate nucleotide sugars like CMP-NeuAc, phospho-lipids, sugar-phosphates or even sucrose may as well be
6
Figure 1.4: Glycan epitopes Various epitopes like N-acetyllactosamine (LacNAc), Lewis antigens or blood group antigens can terminate glycans (A). Glycosaminoglycans like heparan, dermatan and chondroitin sulfate are long- chain polysaccharides with highly negative charge due to the acidic residues and sulfations (B). utilized by certain GTs. The most abundant class of GTs is nucleotide-sugar dependent, known as Leloir-GTs [46]. Although the donor substrate promiscuity of GTs is rather low, a large variety of acceptor substrates ranging from saccharides to peptides or lipids may be accepted by the same enzyme. The attacked nucleophilic atom is often oxygen within an alcohol-group, but it can also be nitrogen, sulfur or even carbon, which underlines the broad range of substrates addressed by GTs. About 200,000 enzymes from all kind of organisms are divided into 97 families so far. This is the second largest group within the carbohydrate active enzyme database (CAZy-database, http://www.cazy.org). In contrast to the very diverse structural properties of GHs, GTs show only two main folding motifs, named GT-A and GT- B fold [47]. The GT-A fold consist of two Rossmann domains, which are in very close proximity to each other. These enzymes often use coordinative binding to an enzyme bound metal-ion to interact with their donor substrate. Consensus sequence for this is the metal- coordinating DxD motif. The GT-B fold consists as well of two Rossman folds, but they are linked with a flexible spacious linker peptide. This linker peptide, rich in cationic amino- acids, is responsible for binding of the anionic donor substrate and acts as general catalytic cleft. Besides these evident two types of fold, a predicted GT-C fold is discussed, which should be abundant in transmembrane GTs [44]. However, so far, there is a lack of experimental evidence for this folding type. Besides their structural classification, a
7 mechanistic classification of GTs into inverting and retaining GTs is useful. Inverting GTs catalyze a group transfer under inversion of the stereochemistry at the anomeric atom of the donor substrates. The reaction mechanism of inverting GTs is well understood and may be described as a direct SN2-reaction. As intermediate state a single oxocarbenium ion-like transition state is found between donor substrate, acceptor substrate and the catalytic base within the catalytic center of the GT. Contrary, a retaining GT catalyzes the transfer under retention of the stereochemistry at the anomeric position. Based on the mechanism of retaining GH a double-displacement mechanism is proposed involving a covalent intermediate between donor substrate and catalytic base of the enzyme, this is also known as
Koshland-type mechanism. Another proposed mechanism is a SNi-type reaction, where the nucleophile attacks from the same side as the leaving group evades. Here, the other side of the molecule is sterically blocked by the enzyme [47,48].
In general, GTs from mammalian are hard to express recombinantly, as they show distinct posttranslations modifications (PTM), which hampers their use in large scale applications as the expression must take place in insect cells or cell lines. An alternative are bacterial GTs, which do not show PTM and should be expressed on bacterial or yeast systems with sufficient activity. Examples for biocatalytic applications of GT is the synthesis of oligo-LacNAc with human and bacterial GTs, with a library of soluble and immobilized acceptor substrates, as well as the synthesis of keratan sulfate mimicking glycans which involves the cascade reaction of GTs and subsequent chemical sulfation [49-55].
An alternative to GTs for in vitro synthesis of glycans are GHs, which comprise the largest group within carbohydrate modifying enzymes [56].From a sequence point of view, they are a very diverse group which is divided into more than 45 different families [57]. Their structural appearance is more conserved, but a differentiation in only very few classes like for GTs is not possible. Maybe this is related to the fact that GHs are the older enzyme class compared to GTs and therefore evolved into a broader range of different sequences and structures. Interestingly, even the type of active-site is different between various GHs, as its topology ranges from cleft-like to tunnel-like or crater-like [56]. However, as with GTs, GHs may be distinguished by their reaction mechanism into inverting and retaining. The retaining mechanism follows a Koshland-type mechanism with a covalent glycosyl-enzyme intermediate, the inverting mechanism is once more a classical SN2-type reaction. To maintain their activity, all GHs share a catalytic acid and base within their active site. Commonly, GHs hydrolyze glycosidic bonds by substituting one glycan by water. Under certain conditions, mainly an unnatural high substrate concentration or non-natural substrates, the inverted reaction, called transglycosylation can take place [58]. The phenomenon is of interest for the
8 in vitro production of glycans as GHs do not require expensive donor substrates or metal cofactors like GTs. Moreover, as the more ancient enzyme group they are also readily expressed on bacteria or yeast. By mutation of the active site with removal of the catalytic acid, GHs are created that are not able to hydrolyze glycosidic bonds anymore and are solely able to perform transglycosylation [59,60]. These artificial enzymes are named glycosynthases. The main drawback of usage of glycosynthases is the necessity of labile substrates like glycosyl fluorides or glycosyl azides with good leaving groups, which are often laboriously in production. However, the advantage is the facile synthesis of glycoconjugates. Multiple reports are published for the utilization and characterization of GHs and derived glycosynthases. The focus lies on finding new enzymes capable of efficient transglycosylation, the creation of glycosynthases or finding unconventional reaction parameters for increasing the yield and turn-over. Interesting examples include the synthesis of nucleotide oligosaccharides and the application of microwave irradiation to speed-up the reaction with thermophilic enzymes as well as minimize product hydrolysis [61-64].
Due to better controllable product distribution, GTs are the favorable biocatalysts for this work. However, in future it may become possible to have tightly controlled reactions also with GHs, which would lead to a more cost-efficient synthesis of carbohydrate ligands.
Besides the synthesis of complex glycan-structures via sophisticated biocatalytic routes, the interaction between glycans and proteins is of high interest. If the enzymes may be described as a carefully tuned machinery to wrap-up complex biological information into the glycan- code, lectins are the decoding “enigmas”, capable of deciphering the highly dynamic code written in carbohydrates. Lectins are abundant in all organisms, from bacteria to plants and mammals. Especially the function of lectins in the latter is of immense interest in biomedical research as it is more and more revealed how prominent lectin-glycan interactions are in various diseases. After this summary of the biosynthesis and in vitro synthesis of glycans, we will now focus on certain lectins, which are responsible for translating the glycan-code into biological action, and on their unique properties, functions and ligand requirements. Only the understanding of these basics makes it possible to create novel materials optimized for carbohydrate-mediated interactions, making the diverse functions of glycans accessible for biomedical research.
9
1.2 Galectins – Sweet and sour effects Contrary to numerous membrane-anchored lectins, galectins are small soluble molecules that show affinity for the β-galactosides. Up to now, about 15 galectins have been found in mammals sharing a conserved carbohydrate recognition domain (CRD) but differing in their fine specificity [65-67]. At least, twelve of them are found in humans. Galectins contain one or two CRDs and are classified into three major groups: proto-type, chimeric, and tandem- repeat (Figure 1.5) [65,68]. Galectins with one CRD and possible homodimer formation are proto-typical (e.g. galectin-1). Tandem-repeat galectins (e.g. galectin-8) are characterized by two CRDs bridged by a small peptide domain. The only known mammalian member of the chimeric type is galectin-3 (Gal-3).
Figure 1.5: The galectin-family Galectins are divided into three groups: A prototype galectin, B chimeric galectin and C tandem-repeat galectin. Mostly, they consist of two carbohydrate recognition domains, present as homo- or heterodimers.
Gal-3 contains a single CRD and long N-terminal domain. The non-lectin domain of approximately 120 amino acids is rich in proline, tyrosine, and glycine [69-71] and sensitive for collagenases and matrix-metalloproteinases (MMP), especially MMP-2 and MMP-9 [72]. Cleavage of Gal-3 yields the N-terminally truncated form of Gal-3, still bearing the CRD (Figure 1.6G). This process takes place in-vivo particularly in tumor regions where MMPs are upregulated [73,74]. The main cleavage site is at Ala62-Tyr63 resulting in 22 kDa fragment but also truncated Gal-3 with larger or shorter N-terminal sequences are possible [75-77]. Showing enhanced affinity for carbohydrate ligands and competing with uncleaved Gal-3 truncated Gal-3 plays a part in tumor regulation [78,79]. The importance of Gal-3 cleavage increases the interest in the N-terminal Gal-3 domain. The non-lectin tail promotes self- association of multiple Gal-3 molecules forming pentamers (Figure 1.6A) and also probably assists the binding process for tighter ligand interaction (Figure 1.6F) [80,81]. While Gal-3 is monomeric in solution it oligomerizes in the presence of multivalent ligands. The N-terminal fragment also regulates cellular targeting of Gal-3 [77,82,83] and stimulates capillary-like
10 morphogenesis and migration of endothelial cells by receptor clustering [74]. The N-terminal functions are diminished after Gal-3 cleavage. Not only the N-term is responsible for Gal-3 self-interactions; the C-term is also involved. C-terminal self-association occurs between two CRDs facing each other if no ligand is bound (Figure 1.6D) [84] and within multiple CRD interactions described as ligand induced binding of one CRD to the back of another (Figure 1.6C) [85]. Also N-C domain interactions are possible (Figure 1.6E) [86]. Contrary to N- terminal self-associations, CRD interactions block the Gal-3 binding sites for glycan ligands.
Figure 1.6: Self-association and cleavage of galectin-3 In the presence of (multivalent) ligands galectin-3 self-associates in variable manners: A N-term mediated pentamer formation, B interaction of both N- and C-term, C C-term mediated self- association after one ligand is bound, D interaction of two binding sites and E binding of one C-term another N-term. The N-term of galectin-3 possibly stabilizes the ligand binding through conformational change (F) and contains a collagenase cleavage site (G).
The crystal structure of the CRD of galectins shows a beta-sandwich formed by six-stranded (S1–S6) and five-stranded (F1–F5) antiparallel β-sheets [87,88]. Carbohydrates are bound in the concave side (S1-S6). The binding site is optimal for small ligands, but long enough for linear tetrasaccharides [89]. Gal-3 shows an extended binding site leading to increased affinity for longer oligosaccharides [88,90]. Interaction of galectin and carbohydrate ligand is based on hydrogen bonds that are built between conserved amino acids and galactose, glucose, or GlcNAc [88,91,92]. The interaction with galactose is mostly conserved. The different fine specificity of galectin members depends mainly on the variation in interplay with other monosaccharide units. Among all galectins, the affinity towards galactose as single monosaccharide is rather weak [93,94]. In fact, galactose based oligosaccharides, e.g. poly-
11
LacNAc, or multivalent glycan presentation increase binding affinity. For understanding the ligand binding and different affinities, the CRD can be divided in five subsites (Figure 1.7) [89,95,96]. In site C the galactose residue is bound, supported by site D for disaccharide partner of the β-glycosidic linked galactose. Site D demands no strict structure requirements and gives room for varying specificity between galectins. Different galectins prefer different saccharide units as well as linkages. Units A and B allow interactions with elongated ligands such as saccharides or other groups mainly at C3 position of galactose. Extensions at the reducing end of the saccharide in D are possible owing to site E. Thus, longer saccharides, glycoproteins or glycolipids can be bound and are even bound with higher affinity than sole galactose due to multiplexed interactions. The different specificity of different galectin CRDs is also present within tandem-repeat type galectins. The N-terminal domain of galectin-8 (Gal-8) shows high affinity towards lactose, whereas the C-terminal domain bounds preferably LacNAc [97]. Moreover, both Gal-8 CRDs tolerate different sialylations and sulfations [98]. Galectin-1 (Gal-1) binds preferably terminal LacNAc units at the non- reducing end and shows no to weak binding to internal galactose [99-102]. However, type I and type II LacNAc are recognized equally by Gal-1, while Gal-3 prefers type II LacNAc over type I [66,103]. Longer chains of type II LacNAc are Gal-3 ligands with increasing affinity due to recognition of internal LacNAc units [89,103,104]. Moreover, Gal-3 and some more galectins tolerate additional sulfate group or N-acetylneuraminic acid at C3-position of galactose and modifications with fucose, galactose, or GalNAc, but do not bind C6-modified galactose [89,105]. If a glycan carries terminal α2-6 sialylation Gal-3 binding is blocked which is used to regulate its biological function [106].
Figure 1.7: Subunits of the carbohydrate recognition domain of galectin-3 The carbohydrate recognition domain can be divided into five subsites. Each site plays an important part in ligand binding. The broad distribution of galectins in human body indicates their involvement in various biological processes. Galectins are found in the extracellular matrix (ECM), cytoplasm, outer plasma membrane, the nucleus, and on cell surfaces [107]. They are expressed by many cells like activated T- and B-cells, dendritic cells, mast cells, or neutrophils. These ancient glycan
12 binding proteins regulate inflammation, immune responses, angiogenesis, and tumor progression. Most of their functions are due to their multimeric character like bridging of molecules, cell-cell and cell-matrix interactions, matrix network assembly, receptor clustering, and cell migration and routing. Galectin interactions with binding partners trigger signal cascades leading to the induction or suppression of effector release, cell growth control, or induction of apoptosis. Every galectin can trigger multiple biological reactions, often influenced by their cellular localization. Gal-3 is a pro-inflammatory mediator that is associated with the activation of T-cells by binding to N-glycans presented on the T-cell receptor, as well as mast cells, neutrophils, and monocytes. Gal-3 is overexpressed and its cellular localization is changed in various inflammatory cells. Increased or suppressed expression of Gal-3 by certain cell types modulates their proliferation and differentiation. Whether cell growth is promoted or reduced depends on the cell type. In contrast, Gal-1 generally has an anti-inflammatory role by inducing anti-inflammatory cytokines and inhibiting pro-inflammatory cytokine production.
Human Gal-3 is expressed in different cell types and tissues. It is existent in the nucleus or cytoplasm [108], or secreted by non-classical pathways [69,71]. Thus, it is also located on plasma membranes and in the extracellular space [69,109]. In Figure 1.8 the main functions of Gal-3 are illustrated. Extracellular Gal-3 acts as an adhesion modulator due to its ability to multimerize and bind glycoconjugates on the cell surface or of the ECM [110,111]. Binding partners in the ECM are laminin, fibronectin, collagen IV, or elastin [78,112-114]. Cell surface ligands are integrins and growth factor receptors. Gal-3 promotes adhesion of cells to ECM glycoproteins or other cells, e.g. human neutrophils to laminin [115], but also breaks cell interactions [113]. Gal-3 induced cross-linking is an important step of signal transduction cascades and leads to the activation of various cellular reactions. In that way, Gal-3 induces the production of interleukin-1 (IL-1) in the present of lipopolysaccharides, IL-2 by human Jurkat T cells, IgE in B-lymphocytes [116-118]. On the other hand, IL-5 production and B- lymphocyte differentiation are inhibited [119-121]. Gal-3 ligand complexes also affect angiogenesis by the induction of endothelial cell motility, capillaries and multicellular network formation. Intracellular Gal-3 has cell-cycle regulating properties like the control of cell proliferation by protecting from apoptosis, cell death and differentiation [122]. It acts as a pre-RNA splicing factor and associates with ribonucleoprotein complexes [123]. Binding to B-cell lymphoma 2 (Bcl-2), activated K-Ras, β-catenin, and synexin triggers some of Gal-3 intracellular functions [124-128].
While Gal-1 is known for inducing apoptosis, especially in T-cells, Gal-3 displays pro- as well as anti-apoptotic properties. It is observed that extracellular Gal-3 mainly induces
13 apoptosis, whereas intracellular Gal-3 prevents cell death. If Gal-3 translocates from the cytosol or nucleus to the perinuclear mitochondrial membrane due to apoptotic stimuli, caspase activation is reduced by inhibited cytochrome c release after binding to Bcl-2 [129,130]. The cell is protected against mitochondrial damage and therefore from apoptosis. For different cell types it was shown that expression of Gal-3 prevents cell apoptosis [131,132] while cells in Gal-3 deficient species undergo apoptosis more rapidly [133,134].
For the anti-apoptotic property the phosphorylation of Gal-3 at Ser6 is required, because the non-phosphorylated mutant failed to protect the cell [135]. In contrast, the apoptotic signaling pathway is triggered by Gal-3 binding to molecules involved in the regulation of apoptosis like CD95 or Nucling [136,137].
The influences of Gal-3 on cellular functions are exploited by different cancer cell lines revealing the involvement of Gal-3 in tumor progression [138]. Influence of Gal-3 expression on tumor growth, migration, and invasion is present in many tumor types [139], such as lung cancer, melanoma, gastric cancer, breast cancer, ovarian cancer, and prostate cancer. Cancer cells express Gal-3 for defending themselves against the immune system and pushing their own growth [140] [141,142]. Intracellular Gal-3 protects the cancer cell from apoptosis and triggers cell proliferation, while extracellular Gal-3 induces monocyte and T-cell apoptosis, helps the cancer cell to adhere and spread, and promotes angiogenesis. Thus, overexpression of Gal-3 and its translocation from nucleus to the cytoplasm are predominant features of many tumor types. Tumor progression is supported by regulation of specific gene expression after Gal-3 interaction with transcription factors as β-catenin. Gal-3 mediated adhesion processes by binding ECM proteins and receptors are important for tumor cell migration and invasion [140]. Multimerization of Gal-3 helps cancer cells to aggregate allowing invasion in blood vessels as well as adhesion to endothelial cells to reach distant organs using the blood stream. Homotypic and heterotypic adhesions of melanoma and colon cancer cells are increased by Gal-3 binding to Thomsen-Friedenreich antigen (Galβ1,3GalNAcα1) on the transmembrane mucin protein MUC1 [143]. In breast tumor cells Gal-3 favors tumor-stromal interactions increasing endothelial cell adhesion [144]. Elevated Gal-3 levels cause secretion of cytokines like IL-6 that upregulates E-selectin, ICAM-1, or integrins on endothelial cells for enhanced adhesion [145]. Angiogenesis is an important process for tumor growth, where secreted Gal-3 is involved in. Capillary tube formation is promoted by endothelial cell behavior influenced by Gal-3 binding to αvβ3 integrins presented on endothelial cells [146]. The induced integrin clustering leads to signaling and triggers angiogenetic activity involving vascular endothelial growth factor (VEGF). Similar signaling effects are observed by binding of Gal-3 to VEGF
14 receptor-2 (VEGFR-2) [147]. For melanoma cells expressing Gal-3 a distinct VEGF secretion was demonstrated.
Figure 1.8: Functions of galectin-3 Galectin-3 appears intra- as well as extracellularly and plays important roles in adhesion, apoptosis, signaling, inflammation, cell differentiation, and tumor progression.
In metastatic processes MMPs contribute decisively as they break down physical barriers like the ECM for invasion, intravasation and extravasation of cancer cells [148]. MMPs are therefore key regulators of tumor growth and angiogenesis. These enzymes are either expressed by tumor cells in the later stage of tumor progression or are produced by a host response to the tumor. Appearance of MMP-2 (gelatinase A) and MMP-9 (gelatinase B) is closely related to tumor metastasis [149] and synthesized as a latent proenzyme that has to be activated by removal of the propeptide [148,150]. Both MMPs are found in colon, breast, and lung cancer [151,152]. With the presence of MMPs cleavage of Gal-3 is upregulated. Because
15
MMPs appear extracellularly, secreted Gal-3 is cleaved while intracellular Gal-3 stays intact [76]. Cleavage of Gal-3 takes place at most epithelial cells and invasive cell clusters in carcinomas [74]. In breast cancer, truncated Gal-3 interacts with endothelial cells promoting rapid attachment of cells and inducing morphogenesis. Due to 20-fold higher affinity of truncated Gal-3 to human umbilical vein endothelial cells (HUVECs) and tighter interaction with laminin than full-length Gal-3 the cleaved form is important for stabilization of epithelial-endothelial interactions for promoting invasion and angiogenesis [74,76-78]. Cells transfected with cleavage-resistant Gal-3 show significantly lower invasion [74]. However, truncated Gal-3 is also able to reduce tumor progression. VEGF-mediated angiogenesis that is facilitated by Gal-3 is inhibited by truncated Gal-3 (Gal-3C) [146]. The treatment of human breast cancer in mouse model with exogenous Gal-3C decreases tumor volumes and weights while Gal-3C as molecule shows a low toxicity profile [153]. The same observation is made in murine model of human multiple myeloma where Gal-3C treatment inhibits tumor growth [79,154,155]. Moreover, in vitro Gal-3C exhibits modest anti-proliferative effects and inhibits chemotaxis and invasion. In combination with proteasome inhibitor bortezomib multiple myeloma induced angiogenesis activity is almost completely inhibited due to supported anticancer activity [79]. Full-length Gal-3 contains the NWGR domain of the Bcl-2 family [129] that inhibits apoptosis [156] and increases chemo-resistance in cancer [157].The absence of this domain may be responsible for the contraire properties of truncated Gal-3.
16
1.3 Design of galectin-3 ligands – Turning sweet into affine The broad field of biological functions makes Gal-3 an important target in biomedical applications. Especially in cancer therapy, Gal-3 is a point of attack for reducing tumor progression and metastasis. In this context, the design of competitive inhibitor is an emerging technology. High affinity ligands have the potential to inhibit undesired Gal-3 binding processes and prevent responses associated with them.
Ligand binding can be explained by characterization of the crystallographic structure of the Gal-3 CRD. Modelling of the binding groove in complex with LacNAc shows the orientation and interactions with amino acids (Figure 1.9) [88,158,159] Tryptophan (Trp181) is conserved for binding throughout the galectin family and its side chain forms π-stacking interactions with the carbohydrate residue galactose. Another seven conserved amino acids are responsible for carbohydrate ligand binding: Arg144, His158, Asn160, Arg162, Asn174, Glu184 and Arg186. During Gal-3 binding of lactose or LacNAc, hydrogen bonds are formed between C4 and C6 of galactose as well as C3 of glucose/GlcNAc and His158, Asn160, Arg162, Asn174 and Glu184, and van der Waals contacts exist between the ligand and Trp181 and Arg186 [104]. The binding groove of Gal-3 exhibits some individual amino acids that are flexible in conformation. The side chain of arginine (Arg144), in particular, is able to adopt different conformations by directing itself away from the protein surface and opens the possibility for ligand modifications. An aromatic substituent at C3 of the galactose unit of LacNAc derivatives leads to increased affinity due to the arginine-arene interaction as a cation-π stacking interaction [160,161]. This extended binding groove near the C3 OH group of the galactose led to further ligand developments based on C3 modifications.
Figure 1.9: Binding model of the carbohydrate recognition domain of galectin-3 LacNAc is bound through direct and water mediated hydrogen bonds or through van der Waals contacts (A) (modified from Seetharaman et al. 1998). The model of galectin-3 binding to Lacto-N- tetraose and Lacto-N-neo-tetraose gives hints for possible recognition after further elongation in C2 and C3 direction of β1-3 linked galactose (B) but limited extensions of β1-4 linked galactose due to close proximity (C) (modified from Collins et al. 2014).
17
C1-symmetrical thiodigalactosides consisting of two galactose units are bound identically like lactose or LacNAc with similar affinity [162,163]. By modifying the C3-postitions potent
Gal-3 inhibitors can be designed. Affinity values in the nanomolar range (Kd ~ 0.05 µM) were achieved with C3-modifications at both units with an aryl group allowing two simultaneous arginine-arene interactions (Figure 1.10A) [162,164,165]. An emerging C3-modification is the 1,2,3-triazolylation of galactose. 3’-triazole-LacNAc (Figure 1.10B) and di-triazole- thiodigalactoside (Figure 1.10C) derivatives showed Kd values of 0.66 µM and 0.03 µM, respectively, in a fluorescence polarization assay with Gal-3 [94]. Triazolyl-LacNAc derivatives with an additional aglyconic modification at the reducing end (Figure 1.10D) showed four-fold inhibitory potency compared to lactose [166]. Even monosaccharide derivatives have the potential to highly inhibit Gal-3 binding. Triazole-galactosides (Figure 1.10E) synthesized by Cu(I)-assisted 1,3-dipolar azide-alkyne cycloaddition (CuAAC) exhibit up to 40 times higher inhibition than galactose [167]. Further O-3 triazole-galactose analogues with aromatic groups at C3 of galactose (Figure 1.10F) are designed exploiting the positive effect of interactions with arginine of Gal-3 CRD [164,165,168]. Besides the effective cation- π interactions with arginine, additional C2 substituents with phosphate or sulfate (Figure 1.10G) allow polar interactions around the arginine guanidinium hydrogens and strengthen the Gal-3 affinity [93,169]. For the substitution of larger aromatic groups, a positive effect is reported due to the establishment of favorable aryl-arginine interactions with Gal-3 [165]. Galactosyl oximes with C3-triazole fragments (Figure 1.10H) showed promising Gal-3 binding (Kd ~ 11 µM) due to strong interaction of the indole aldoxime fragment and Arg144 of Gal-3 [170]. Besides the effective interactions with the Gal-3 CRD, hydrophobic C3- modifications of galactose-derived monosaccharides have additional advantages: they improve half-life and membrane transport [171]. The anomeric C-atom can also be functionalized to develop new Gal-3 inhibitors. C3-modification as well as anomeric functionalization of galactose with oxime ether (Figure 1.10I) showed 24-fold increased affinity (Kd 180 µM) when compared with free galactose [172]. Three times higher inhibitory potency in a hemagglutination inhibition assay than lactose was achieved with an aromatic S- lactosides with a naphthalene fragment at the reducing end (Figure 1.10J) [173]. Selenoglycosides consisting of two symmetrical galactose units connected via the anomeric C-atom (Figure 1.10K) show better Gal-3 inhibition than lactose [174].
18
Figure 1.10: Design of galectin-3 ligands by C3-modifications and/or anomeric functionalization of saccharides A: symmetrical diamido-thio-digalactoside, B: naphtyl-3-triazol-N-acetyllactosamine, C: butylamine- di-triazol-thio-digalactoside, D: 1-triazol-N-acetyllactosamine E: phenyl-3-acetoxyltriazol-thio- galactoside , F: 1,2,3-triazole lysine-derived-3-O-galactoside, G: Methyl-3-methylbenzamido-2-sulfo- thio-galactopyranoside, H: 3C-triazol-1-yl-O-galactopyranosyl aldoximes , I: indole-3-carbaldoxime- galactoside, J: β-napthylsulfonyl lactoside , K: symmetrical (di)selenodigalactoside
Modifications at C6-position are not tolerated during Gal-3 binding because forming of a hydrogen bond at this site is obligatory for the binding process [88,104,175]. In contrast and as seen for the C3-position, the C2 OH group as well does not establish a hydrogen bond [90,92] and C2-modifications of galactosides with the objective of high affinity Gal-3 inhibitors are possible. N-acetylation at the C2-position of galactose in LacNAc forming LacdiNAc (Figure 1.11A) inhibits Gal-3 binding to ASF 1.5-times better than LacNAc [54]. The additional N-acetyl group of LacdiNAc can engage in a hydrogen bond with Arg144, whereas LacNAc does not [176]. Synthetic octyl lacto-N-biose (Galβ1-3GlcNAc) with a naphthalene residue at the C2-position (Figure 1.11B) of GlcNAc is bound by Gal-3 seven- fold better than sole lacto-N-biose [177].
19
Figure 1.11: Design of galectin-3 ligands by C2-modifications of disaccharides A: LacdiNAc, B: N-naphthoyl octyl-lacto-N-biose
Besides galactoside ligands, further monosaccharides, like mannose or talose, have the potential to serve as base for high affinity Gal-3 ligands. Similar tools are used to modify these carbohydrates. β-1,2-oligomannosides on macrophages are bound by Gal-3 [178]. The affinity of Gal-3 towards mannose monosaccharides is considered weak but can be increased by C1-triazolylation (Figure 1.12A) to a moderate Kd value of 1.5 mM [179]. C1-triazole- mannose does not optimally mimic galactose-based Gal-3 inhibitors. Talose is seen as stable candidate as it does not appear naturally in mammalians. It is a potential Gal-3 ligand because the inverted C2-configuration of talose compared to galactose enables interaction with polar amino acids, different from those accessed by galactosides. C2 and C3 substituted talosides with aromatic groups and amides, sulfate or acetyl groups (Figure 1.12B, C) shows up to 17 times higher affinity than methyl β-galactoside [180-182]. The lowest Kd was 0.25 mM.
Figure 1.12: Design of galectin-3 ligands on the basis of mannose or talose A: propylaminocarbonyl-triazol-mannosid, B: 2-sulfate-3-toluoyl-methyl-taloside, C: 3-p- nitrobenzamido-2-toluoyl-methyl-taloside
Crystallographic structures of Gal-3 and lacto-N-tetraose as well as lacto-N-neotetraose (Galβ1-3/4GlcNAcβ1-3Galβ1-4Glc) indicate that Gal-3 is able to bind extended oligosaccharide like poly-LacNAc (Figure 1.9B and C) [90]. The lactose at the reducing end forms identical hydrogen bonding interactions with the protein as sole lactose. The GlcNAc residue that is β1-3 linked to the lactose portion of the tetrasaccharides extends along the Gal- 3 binding site forming direct hydrogen bonds with the protein via OH at C6 to side chains of Arg144 and Asp148. Additionally, van der Waals contacts are made between GlcNAc and Asp148 and His158. Thus, favorable interactions between Gal-3 and larger oligosaccharides are shown and can explain why poly-LacNAc consisting of at least two LacNAc units was proven to be bound by Gal-3 with higher affinity than one LacNAc unit [183].
20
Peptide ligands mimic carbohydrates, especially in immune responses [184]. Pentapeptides based on Tyr-X-Tyr motif of glycomimetic peptides show inhibitory potentials in millimolar range [185]. Longer peptides (15 amino acids) inhibit Gal-3 binding to Thomsen-Friedenreich antigen [186]. In phage-display studies, sequences were found with affinity and specificity for Gal-3 in nanomolar range. Moreover, inhibition of metastasis-associated cell adhesion is reported.
Since interactions of lectins with carbohydrate ligands are rather weak, presentation of multiple glycan ligands can drastically increases the binding strength in so-called glyco clusters [18,187-189]. Carbohydrates are multivalently presented with the help of different scaffolds that define arrangement and orientation of the saccharides [28,190-200]. As scaffolds azido-sugar cores [201], dendritic systems [202-206], glyconanoparticles [207,208], calixarenes [199,209] or cyclodextrins [210] are used to attain possible cluster glycoside effects. The reason for this binding enhancement may be related to chelation, which is characterized by improved ligand binding when another binding site is already bound to a closely neighbored ligand of the multivalent system.
Often, the reported modifications of carbohydrates are combined with multivalent presentation. Lactoside triazoles arranged multivalently on carbohydrate scaffolds yield up to tetravalent ligand (Figure 1.13A) with four times higher inhibitory potency by competitive fluorescent polarization assays (Kd 16 µM) than reference with one lactose residue [201]. The backbone is an azido-sugar core consisting of glucose, maltose or maltotriose. A trivalent triazole-lactoside based on polyfunctional unnatural amino acids (Figure 1.13B) shows in competitive fluorescent polarization assays 13-times higher affinity (Kd 17 µM) than the methyl β-lactoside [211]. The 1,2,3-triazole di-lactose-derived glycoconjugate (Figure 1.13C) shows strong interaction with Gal-3 and can bridge two independent Gal-3 CRDs [168]. Dimeric thio-lactosides linked via carbon-carbon bonds (Figure 1.13D) are active Gal-3 inhibitors in hemagglutination assays [212]. A trimeric triazole galactohybrid with an azidogalactofuranose core (Figure 1.13E) is bound by Gal-3 with a Kd of 50 µM [213].
21
Figure 1.13: Design of galectin-3 ligands by multivalent presentation of modified saccharides A: tetravalent lactoside triazol, B: trivalent lactosyl methylester carbamate, C: triazole di-lactose- derived glycoconjugate, D: dimeric thio-lactoside, E: trimeric triazole-linked galactohybrid
Lactose-functionalized poly(amidoamine) dendrimers are another possibility for multivalent ligand presentation [214,215]. Rigid multivalent lactose ligands with an alkyne-branched tetraamino backbone (Figure 1.14A) showed strong multivalent effects with more than 4000- fold enhanced inhibition (IC50 70 nM) than free lactose in a solid phase inhibition assay [216]. This inhibitory potency was 300-times higher compared to the monovalent reference and 75- times higher in relation to a single lactose-containing arm. In a fluorescence titration assay 44- fold higher potency than lactose was obtained for the dendrimeric ligands (Kd 14 µM). Gal-3 binding to monovalent lactose dendrimer was already about 10 times higher than to free lactose indicating interactions with the dendrimer backbone [203,216].
22
Figure 1.14: Design of galectin-3 ligands by synthesis of lactose dendrimers and calixarenes A: tetravalent lactose, B: cone lactosylcalix[4]arene, C: 2/3' substituted LacNAc calix[4/6]arene, D: isomeric lactosylcalix[4]arene
Phenol-based and cup-shaped calixarenes with four or more aromatic units can be decorated with carbohydrates at each subunit. Multivalent lactose conjugated calixarenes (Figure 1.14B) increased the inhibitory potency towards Gal-3 by a factor of 12.5 reaching IC50 of 200 µM
[209]. Calix[4/6]arenes bearing 2/3' substituted LacNAc (Figure 1.14C) showed IC50 values up to 0.15 µM for Gal-3 binding in a solid-phase assay, which were about 1500 times lower compared to monovalent species [217]. Therefore, Gal-3 binds calixarenes bearing LacNAc more strongly than presenting lactose. Galactosyl- and lactosylcalix[4]arenes in its cone and isomeric form (Figure 1.14B, D) were proved as Gal-3 ligands in surface plasmon resonance (SPR) spectroscopy [218]. Gal-3 preferred lactosylcalixarenes over galactosylcalixarenes as well as the cone form compared to the isomeric 1,3-alternate structure.
23
Quantum dots can also be used as carrier of ligands. With multivalent LacNAc modified quantum-dots presenting 108 glycans the affinity towards Gal-3 (KD 57 nM) was 92-fold higher compared to soluble LacNAc in SPR measurements [219].
Hepta-antennated β-cyclodextrins carrying galactose, lactose or LacNAc (Figure 1.15A) reached relatively strong binding with Gal-3; the lowest inhibitory potential is reported for galactosyl cyclodextrin [220]. Some multivalency based binding enhancement occurred, especially for lactosyl cyclodextrin, depending on the competing ligand. In another study, novel gold nanoparticle species containing multiple long, flexible linkers capped with lactose or a mixture of lactose and β-cyclodextrin (Figure 1.15B) are designed that interacted efficiently with Gal-3 [221].However, only the lactose units are the targeting ligands for Gal- 3.
Figure 1.15: Design of galectin-3 ligands by synthesis of functionalized cyclodextrins and gold nanoparticles A: perglycosylated β-cyclodextrins, B: gold nanoparticles containing lactose or a mixture of lactose and β-cyclodextrin
24
Due to their polyvalency, neo-glycoproteins are of high interest for the investigation of carbohydrate-protein interactions [5,222]. Albumin as non-glycosylated protein is often used as backbone. About 15 biantennary N-glycans were conjugated to human serum albumin (HSA) (Figure 1.16A) by reaction with an activated ester and CuAAC [223]. These neo- glycoproteins showed high affinity towards Gal-3 in SPR measurements (EC50 4.91 nM) with clear multivalent effects. The affinity related to one glycan was with EC50 of 74 nM about
450-fold higher than to the monovalent reference (EC50 33 µM). Moreover, the synthetic multivalent ligands could efficiently inhibit the attachment of Gal-3 to human prostate cancer and lung cancer cells in an ELISA-type assay indicating the usefulness in tumor therapy. Another method to synthesize neo-glycoproteins is the usage of diethyl squarate for the coupling of oligosaccharide amines to carrier proteins, e.g. bovine serum albumin (BSA) [224]. With this method, LacNAc and LacdiNAc were coupled to BSA (Figure 1.16B) and use these neo-glycoproteins as ligands for Gal-3 to prove Gal-3 binding to both epitopes [225].
Figure 1.16: Design of galectin-3 ligands by glycan coupling to proteins yielding multivalent neo- glycoproteins A: N-glycans conjugated human serum albumin, B: LacdiNAc conjugated bovine serum albumin
In summary, there are numerous glycan modifications available to promote effective binding of Gal-3. In this way, low affinity ligands can be turned into potent inhibitors.
25
1.4 The aim The aim of this work is to reveal with well-designed in vitro assays new insights into molecular recognition and the influence of sequential changes of Gal-3 on its binding characteristics and ultimately on its potential as drug or target in biomedicine. As multiple aspects of Gal-3 properties will be explored, this work divides into two major parts:
1.4.1 Design, cloning, expression, and characterization of truncated galectin-3 In nature, truncation of Gal-3 plays an important role in regulation of affinity, self-association and -organization potential, and signal modulation. As the detailed characterization of truncated Gal-3 in terms of binding behavior was yet not explored in detail, we go now deeply into this topic to reveal the actual influence of the truncation. This part involves the design of truncated Gal-3 variants as well as multiple fusion-proteins including YFP and SNAP-tag to investigate the influence of the truncation and the fusion partner on Gal-3 binding. This will give a new alternative in the design of Gal-3 variants that may be used as protein-based drugs. Moreover, the fusions enable imaging or coupling of effector molecules like active components. Mostly, insertion and deletion of sequences accompany with alteration of protein function. Therefore, this work deals with the question about the influence of the N-terminal domain on the molecular properties of human Gal-3 – the affinity and specificity of carbohydrate ligands as well as the self-association potential. An important step is the design and cloning of Gal-3 variants as well as the recombinant expression in small scale processes in E. coli. The Gal-3 variants will be purified and characterized by classical methods in protein chemistry like SDS-PAGE, Western-Blot and SEC. The main aim is to monitor the differences in affinity towards glycan ligands. This will be accomplished in ELISA-type assays as well as SPR and FACS experiments. This part will give new insight into modulating the binding behavior of Gal-3 by creating novel versions of this protein, which will become important in general galectin research.
1.4.2 Design of novel multivalent neo-glycoproteins as promising ligands for galectin-3 The second part of this work addresses not the binder but the ligand. In nature, the key point of glycan ligands is their multivalent organization in glyco clusters. This mode of presentation has high impact on the avidity in a way that lectin binding potency is increased in orders of magnitude due to a high glycosylation density. Various scaffolds for reaching a multivalent effect have been investigated and published in the last years. Mostly chemical, artificial scaffolds have been employed for galectin binding. In this work, we follow a path of nature- mimicking presentation of glycans utilizing a naturally non-glycosylated serum protein. Albumin is highly abundant and non-toxic, and contains multiple lysine residues which are easily addressable for chemical conjugation methods. The aim is to establish a novel process
26 for the synthesis of potent Gal-3 ligands. We will develop a simple, variable synthesis method for attaching linear oligosaccharides via the homobifunctional linker squaric acid diethyl ester to the lysines of bovine serum albumin (BSA). Prior to this chemical conjugation, defined tetrasaccharides will be synthesized under optimized conditions with recombinant bacterial and mutated human glycosyltransferases. The saccharides carry a primary amino-group at the reducing end that will be used for amine-dependent conjugation. The synthesis of neo- glycoproteins presenting LacNAc-derived tetrasaccharides will be established and the progress of the reaction will be monitored by chromogenic assays and SDS-PAGE as well as Western Blot. By changing the ratio of substrates different glycosylation densities are accessible. This work enters into the question how the glycan density presented by the synthesized neo-glycoproteins influences the binding affinity and inhibition potency towards Gal-3. Investigation of Gal-3 binding by solid phase assays and SPR measurements will give the answers. Additionally, chemically modified tetrasaccharides step into the game. The question is how C6-biotinylation affects the binding properties of Gal-3. Therefore, the synthesis of C6-biotinylated tetrasaccharides will be developed as well as their presentation as neo-glycoprotein with various glycan densities. The process of biotinylation will be performed in an optimized manner utilizing a chemo-enzymatic reaction cascade. The potential of modified tetrasaccharides on protein scaffolds to act as Gal-3 ligands will also be investigated in ELISA-type assays and via SPR. The gathered interesting results will enable the future design of novel Gal-3 ligands based on nature-mimicking multivalent scaffolds. Additionally, the toolbox of modified glycan-structures as ligands for Gal-3 will be enlarged by the so far not addressed C6-modified structures.
This work combines different aspects of biomedical related research on Gal-3. It answers questions from the sequence-function interplay as well as from the ligand-display point of view, to give a detailed insight in Gal-3 targeting or Gal-3 based methods valuable for biomedicine. Gal-3 mediated interactions are a highly complex, dynamic field with multiple proposed applications in future. In this work, human Gal-3, as all-round player in numerous biological processes, will be characterized in depth from different angles. This will offer new methodic possibilities in glycobiology and glycobiomedicine, two strongly emerging technologies with growing impact on research and industrial application.
27
1.5 References 1. Gabius, H.-J.; Siebert, H.-C.; André, S.; Jiménez-Barbero, J.; Rüdiger, H. Chemical biology of the sugar code. ChemBioChem 2004, 5, 740-764. 2. Šebestík, J.; Reiniš, M.; Ježek, J. Biomedical applications of peptide-, glyco- and glycopeptide dendrimers, and analogous dendrimeric structures. In Sugar code (glycocode), Springer Vienna: 2012; pp 23-27. 3. Solís, D.; Bovin, N.V.; Davis, A.P.; Jiménez-Barbero, J.; Romero, A.; Roy, R.; Smetana, K.; Gabius, H.-J. A guide into glycosciences: How chemistry, biochemistry and biology cooperate to crack the sugar code. BBA Gen. Subj. 2015, 1850, 186-235. 4. Gabius, H.-J. Cell surface glycans: The why and how of their functionality as biochemical signals in lectin-mediated information transfer. Crit. Rev. Immunol. 2006, 26, 43-79. 5. Varki, A. Biological roles of oligosaccharides: All of the theories are correct. Glycobiology 1993, 3, 97-130. 6. Collins, B.E.; Paulson, J.C. Cell surface biology mediated by low affinity multivalent protein– glycan interactions. Curr. Opin. Chem. Biol. 2004, 8, 617-625. 7. Reitsma, S.; Slaaf, D.W.; Vink, H.; van Zandvoort, M.A.M.J.; oude Egbrink, M.G.A. The endothelial glycocalyx: Composition, functions, and visualization. Pflugers Arch. 2007, 454, 345-359. 8. Lamberg, S.I.; Stoolmiller, A.C. Glycosaminoglycans. A biochemical and clinical review††from the departments of medicine, pediatrics, and biochemistry, the joseph p. Kennedy jr. Mental retardation center, and the larabida-university of chicago institute, pritzer school of medicine, university of chicago, chicago, illinois, 60633. J. Invest. Dermatol. 1974, 63, 433-449. 9. Jackson, R.L.; Busch, S.J.; Cardin, A.D. Glycosaminoglycans: Molecular properties, protein interactions, and role in physiological processes. Physiol. Rev. 1991, 71, 481-539. 10. Sattelle, B.M.; Almond, A. Less is more when simulating unsulfated glycosaminoglycan 3d- structure: Comparison of glycam06/tip3p, pm3-carb1/tip3p, and scc-dftb-d/tip3p predictions with experiment. J. Comput. Chem. 2010, 31, 2932-2947. 11. Imberty, A.; Lortat-Jacob, H.; Pérez, S. Structural view of glycosaminoglycan–protein interactions. Carbohydr. Res. 2007, 342, 430-439. 12. Spiro, R.G. Protein glycosylation: Nature, distribution, enzymatic formation, and disease implications of glycopeptide bonds. Glycobiology 2002, 12, 43R-56R. 13. Walsh, C. Posttranslational modification of proteins: Expanding nature's inventory. Roberts and Company Publishers: 2006. 14. Lis, H.; Sharon, N. Lectins: Carbohydrate-specific proteins that mediate cellular recognition. Chem. Rev. 1998, 98, 637-674. 15. Sharon, N.; Lis, H. Lectins as cell recognition molecules. Science 1989, 246, 227-234. 16. Sharon, N.; Lis, H. History of lectins: From hemagglutinins to biological recognition molecules. Glycobiology 2004, 14, 53R-62R. 17. Garner, Omai B.; Baum, Linda G. Galectin–glycan lattices regulate cell-surface glycoprotein organization and signalling. Biochem. Soc. Trans. 2008, 36, 1472-1477. 18. Lundquist, J.J.; Toone, E.J. The cluster glycoside effect. Chem. Rev. 2002, 102, 555-578. 19. Monsigny, M.; Mayer, R.; Roche, A.-C. Sugar-lectin interactions: Sugar clusters, lectin multivalency and avidity. Carbohydr. Lett. 2000, 4, 35-52. 20. Yarema, K.J.; Bertozzi, C.R. Characterizing glycosylation pathways. Genome Biol. 2001, 2, REVIEWS0004. 21. Hart, G.W.; Copeland, R.J. Glycomics hits the big time. Cell 2010, 143, 672-676. 22. Bektas, M.; Rubenstein, D.S. The role of intracellular protein o-glycosylation in cell adhesion and disease. J. Biomed. Res. 2011, 25, 227-236. 23. Moremen, K.W.; Tiemeyer, M.; Nairn, A.V. Vertebrate protein glycosylation: Diversity, synthesis and function. Nat. Rev. Mol. Cell. Biol. 2012, 13, 448-462. 24. Hounsell, E.; Davies, M.; Renouf, D. O-linked protein glycosylation structure and function. Glycoconj. J. 1996, 13, 19-26. 25. Schwarz, F.; Aebi, M. Mechanisms and principles of n-linked protein glycosylation. Curr. Opin. Struct. Biol. 2011, 21, 576-582. 26. Schachter, H. The joys of hexnac. The synthesis and function of n-ando-glycan branches. Glycoconj. J. 2000, 17, 465-483.
28
27. Hammond, C.; Braakman, I.; Helenius, A. Role of n-linked oligosaccharide recognition, glucose trimming, and calnexin in glycoprotein folding and quality control. Proc. Natl. Acad. Sci. 1994, 91, 913-917. 28. Bojarová, P.; Rosencrantz, R.R.; Elling, L.; Křen, V. Enzymatic glycosylation of multivalent scaffolds. Chem. Soc. Rev. 2013, 42, 4774-4797. 29. Carraway, K.L.; Hull, S.R. O‐glycosylation pathway for mucin‐type glycoproteins. Bioessays 1989, 10, 117-121. 30. Hanisch, F.-G. O-glycosylation of the mucin type. In Biol. Chem., 2001; Vol. 382, p 143. 31. Jensen, P.H.; Kolarich, D.; Packer, N.H. Mucin-type o-glycosylation--putting the pieces together. FEBS J. 2010, 277, 81-94. 32. Paulson, J.C. Glycoproteins: What are the sugar chains for? Trends Biochem. Sci. 1989, 14, 272-276. 33. Feizi, T.; Childs, R.A. Carbohydrates as antigenic determinants of glycoproteins. Biochem. J. 1987, 245, 1-11. 34. Kobata, A. Structures and functions of the sugar chains of glycoproteins. Eur. J. Biochem. 1992, 209, 483-501. 35. Tarp, M.A.; Clausen, H. Mucin-type o-glycosylation and its potential use in drug and vaccine development. BBA Gen. Subj. 2008, 1780, 546-563. 36. Tian, E.; Ten Hagen, K. Recent insights into the biological roles of mucin-type o- glycosylation. Glycoconj. J. 2009, 26, 325-334. 37. Argüeso, P.; Gipson, I.K. Epithelial mucins of the ocular surface: Structure, biosynthesis and function. Exp. Eye Res. 2001, 73, 281-289. 38. Robbe, C.; Capon, C.; Coddeville, B.; Michalski, J.-C. Structural diversity and specific distribution of o-glycans in normal human mucins along the intestinal tract. Biochem. J. 2004, 384, 307-316. 39. Gerken, T.A.; Jamison, O.; Perrine, C.L.; Collette, J.C.; Moinova, H.; Ravi, L.; Markowitz, S.D.; Shen, W.; Patel, H.; Tabak, L.A. Emerging paradigms for the initiation of mucin-type protein o-glycosylation by the polypeptide galnac transferase family of glycosyltransferases. J. Biol. Chem. 2011, 286, 14493-14507. 40. Ten Hagen, K.G.; Fritz, T.A.; Tabak, L.A. All in the family: The udp-galnac:Polypeptide n- acetylgalactosaminyltransferases. Glycobiology 2003, 13, 1R-16R. 41. Steen, P.V.d.; Rudd, P.M.; Dwek, R.A.; Opdenakker, G. Concepts and principles of o-linked glycosylation. Critical Reviews in Biochemistry and Molecular Biology 1998, 33, 151-208. 42. Bennett, E.P.; Mandel, U.; Clausen, H.; Gerken, T.A.; Fritz, T.A.; Tabak, L.A. Control of mucin-type o-glycosylation: A classification of the polypeptide galnac-transferase gene family. Glycobiology 2012, 22, 736-756. 43. Bansil, R.; Turner, B.S. Mucin structure, aggregation, physiological functions and biomedical applications. Curr. Opin. Colloid Interface Sci. 2006, 11, 164-170. 44. Breton, C.; Šnajdrová, L.; Jeanneau, C.; Koča, J.; Imberty, A. Structures and mechanisms of glycosyltransferases. Glycobiology 2006, 16, 29R-37R. 45. Coutinho, P.M.; Deleury, E.; Davies, G.J.; Henrissat, B. An evolving hierarchical family classification for glycosyltransferases. J. Mol. Biol. 2003, 328, 307-317. 46. Unligil, U.M.; Rini, J.M. Glycosyltransferase structure and mechanism. Curr. Opin. Struct. Biol. 2000, 10, 510-517. 47. Lairson, L.L.; Henrissat, B.; Davies, G.J.; Withers, S.G. Glycosyltransferases: Structures, functions, and mechanisms. Annu. Rev. Biochem. 2008, 77, 521-555. 48. Lee, S.S.; Hong, S.Y.; Errey, J.C.; Izumi, A.; Davies, G.J.; Davis, B.G. Mechanistic evidence for a front-side, sni-type reaction in a retaining glycosyltransferase. Nat. Chem. Biol. 2011, 7, 631-638. 49. Adamiak, K.; Anders, T.; Henze, M.; Keul, H.; Möller, M.; Elling, L. Chemo-enzymatic synthesis of functionalized oligomers of n-acetyllactosamine glycan derivatives and their immobilization on biomaterial surfaces. J. Mol. Catal. B Enzym. 2012, 84, 108-114. 50. Bojarová, P.; Křenek, K.; Wetjen, K.; Adamiak, K.; Pelantová, H.; Bezouška, K.; Elling, L.; Křen, V. Synthesis of lacdinac-terminated glycoconjugates by mutant galactosyltransferase – a way to new glycodrugs and materials. Glycobiology 2009, 19, 509-517. 51. Kupper, C.E.; Rosencrantz, R.R.; Henssen, B.; Pelantová, H.; Thönes, S.; Drozdova, A.; Křen, V.; Elling, L. Chemo-enzymatic modification of poly-n-acetyllactosamine (lacnac) oligomers
29
and n,n-diacetyllactosamine (lacdinac) based on galactose oxidase treatment. Beilstein J. Org. Chem. 2012, 8, 712-725. 52. Rech, C.; Rosencrantz, R.R.; Křenek, K.; Pelantová, H.; Bojarová, P.; Römer, C.E.; Hanisch, F.-G.; Křen, V.; Elling, L. Combinatorial one-pot synthesis of poly-n-acetyllactosamine oligosaccharides with leloir-glycosyltransferases. Adv. Synth. Catal. 2011, 353, 2492-2500. 53. Sauerzapfe, B.; Krenek, K.; Schmiedel, J.; Wakarchuk, W.W.; Pelantova, H.; Kren, V.; Elling, L. Chemo-enzymatic synthesis of poly-n-acetyllactosamine (poly-lacnac) structures and their characterization for cgl2-galectin-mediated binding of ecm glycoproteins to biomaterial surfaces. Glycoconj. J. 2009, 26, 141-159. 54. Šimonová, A.; Kupper, C.E.; Böcker, S.; Müller, A.; Hofbauerová, K.; Pelantová, H.; Elling, L.; Křen, V.; Bojarová, P. Chemo-enzymatic synthesis of lacdinac dimers of varying length as novel galectin ligands. J. Mol. Catal. B Enzym. 2014, 101, 47-55. 55. Lange, B.; Šimonová, A.; Fischöder, T.; Pelantová, H.; Křen, V.; Elling, L. Towards keratan sulfate – chemoenzymatic cascade synthesis of sulfated n-acetyllactosamine (lacnac) glycan oligomers. Adv. Synth. Catal. 2016, 358, 584-596. 56. Davies, G.; Henrissat, B. Structures and mechanisms of glycosyl hydrolases. Structure 1995, 3, 853-859. 57. Henrissat, B. A classification of glycosyl hydrolases based on amino acid sequence similarities. Biochem. J. 1991, 280, 309-316. 58. Cote, G.L.; Tao, B.Y. Oligosaccharide synthesis by enzymatic transglycosylation. Glycoconj. J. 1990, 7, 145-162. 59. Perugino, G.; Cobucci-Ponzano, B.; Rossi, M.; Moracci, M. Recent advances in the oligosaccharide synthesis promoted by catalytically engineered glycosidases. Adv. Synth. Catal. 2005, 347, 941-950. 60. Mackenzie, L.F.; Wang, Q.; Warren, R.A.J.; Withers, S.G. Glycosynthases: Mutant glycosidases for oligosaccharide synthesis. J. Am. Chem. Soc. 1998, 120, 5583-5584. 61. Henze, M.; You, D.-J.; Kamerke, C.; Hoffmann, N.; Angkawidjaja, C.; Ernst, S.; Pietruszka, J.; Kanaya, S.; Elling, L. Rational design of a glycosynthase by the crystal structure of β- galactosidase from bacillus circulans (bgac) and its use for the synthesis of n- acetyllactosamine type 1 glycan structures. J. Biotechnol. 2014, 191, 78-85. 62. Kamerke, C.; Pattky, M.; Huhn, C.; Elling, L. Synthesis of udp-activated oligosaccharides with commercial β-galactosidase from bacillus circulans under microwave irradiation. J. Mol. Catal. B Enzym. 2012, 79, 27-34. 63. Henze, M.; Schmidtke, S.; Hoffmann, N.; Steffens, H.; Pietruszka, J.; Elling, L. Combination of glycosyltransferases and a glycosynthase in sequential and one-pot reactions for the synthesis of type 1 and type 2 n-acetyllactosamine oligomers. ChemCatChem 2015, 7, 3131- 3139. 64. Henze, M.; Merker, D.; Elling, L. Microwave-assisted synthesis of glycoconjugates by transgalactosylation with recombinant thermostable β-glycosidase from pyrococcus. Int. J. Mol. Sci. 2016, 17, 210. 65. Leffler, H.; Carlsson, S.; Hedlund, M.; Qian, Y.; Poirier, F. Introduction to galectins. Glycoconj. J. 2004, 19, 433-440. 66. Rabinovich, G.A.; Toscano, M.A. Turning 'sweet' on immunity: Galectin-glycan interactions in immune tolerance and inflammation. Nat. Rev. Immunol. 2009, 9, 338-352. 67. Rapoport, E.M.; Bovin, N.V. Specificity of human galectins on cell surfaces. Biochem. (Mosc) 2015, 80, 846-856. 68. Hirabayashi, J.; Kasai, K.-i. The family of metazoan metal-independent β-galactoside-binding lectins: Structure, function and molecular evolution. Glycobiology 1993, 3, 297-304. 69. Dumic, J.; Dabelic, S.; Flögel, M. Galectin-3: An open-ended story. BBA-Gen. Subjects 2006, 1760, 616-635. 70. Krześlak, A.; Lipińska, A. Galectin-3 as a multifunctional protein. Cell. Mol. Biol. Lett. 2004, 9, 305-328. 71. Barondes, S.H.; Cooper, D.N.; Gitt, M.A.; Leffler, H. Galectins. Structure and function of a large family of animal lectins. J. Biol. Chem. 1994, 269, 20807-20810. 72. Ochieng, J.; Fridman, R.; Nangiamakker, P.; Kleiner, D.E.; Liotta, L.A.; Stetlerstevenson, W.G.; Raz, A. Galectin-3 is a novel substrate for human matrix metalloproteinases-2 and -9. Biochem. 1994, 33, 14109-14114.
30
73. Wang, Y.; Nangia-Makker, P.; Tait, L.; Balan, V.; Hogan, V.; Pienta, K.J.; Raz, A. Regulation of prostate cancer progression by galectin-3. Am. J. Pathol. 2009, 174, 1515-1523. 74. Nangia-Makker, P.; Wang, Y.; Raz, T.; Tait, L.; Balan, V.; Hogan, V.; Raz, A. Cleavage of galectin-3 by matrix metalloproteases induces angiogenesis in breast cancer. Int. J. Cancer 2010, 127, 2530-2541. 75. Guévremont, M.; Martel-Pelletier, J.; Boileau, C.; Liu, F.T.; Richard, M.; Fernandes, J.C.; Pelletier, J.P.; Reboul, P. Galectin-3 surface expression on human adult chondrocytes: A potential substrate for collagenase-3. Ann. Rheum. Dis. 2004, 63, 636-643. 76. Shekhar, M.P.V.; Nangia-Makker, P.; Tait, L.; Miller, F.; Raz, A. Alterations in galectin-3 expression and distribution correlate with breast cancer progression - functional analysis of galectin-3 in breast epithelial-endothefial interactions. Am. J. Pathol. 2004, 165, 1931-1941. 77. Nangia-Makker, P.; Raz, T.; Tait, L.; Hogan, V.; Fridman, R.; Raz, A. Galectin-3 cleavage: A novel surrogate marker for matrix metalloproteinase activity in growing breast cancers. Cancer Res. 2007, 67, 11760-11768. 78. Ochieng, J.; Green, B.; Evans, S.; James, O.; Warfield, P. Modulation of the biological functions of galectin-3 by matrix metalloproteinases. BBA-Gen. Subjects 1998, 1379, 97-106. 79. Mirandola, L.; Yu, Y.F.; Chui, K.; Jenkins, M.R.; Cobos, E.; John, C.M.; Chiriva-Internati, M. Galectin-3c inhibits tumor growth and increases the anticancer activity of bortezomib in a murine model of human multiple myeloma. PLoS One 2011, 6. 80. Ahmad, N.; Gabius, H.J.; André, S.; Kaltner, H.; Sabesan, S.; Roy, R.; Liu, B.C.; Macaluso, F.; Brewer, C.F. Galectin-3 precipitates as a pentamer with synthetic multivalent carbohydrates and forms heterogeneous cross-linked complexes. J. Biol. Chem. 2004, 279, 10841-10847. 81. Barboni, E.A.M.; Bawumia, S.; Henrick, K.; Hughes, R.C. Molecular modeling and mutagenesis studies of the n-terminal domains of galectin-3: Evidence for participation with the c-terminal carbohydrate recognition domain in oligosaccharide binding. Glycobiology 2000, 10, 1201-1208. 82. Gong, H.C.; Honjo, Y.; Nangia-Makker, P.; Hogan, V.; Mazurak, N.; Bresalier, R.S.; Raz, A. The nh2 terminus of galectin-3 governs cellular compartmentalization and functions in cancer cells. Cancer Res. 1999, 59, 6239-6245. 83. Nakahara, S.; Oka, N.; Wang, Y.; Hogan, V.; Inohara, H.; Raz, A. Characterization of the nuclear import pathways of galectin-3. Cancer Res. 2006, 66, 9995-10006. 84. Yang, R.Y.; Hill, P.N.; Hsu, D.K.; Liu, F.T. Role of the carboxyl-terminal lectin domain in self-association of galectin-3. Biochem. 1998, 37, 4086-4092. 85. Lepur, A.; Salomonsson, E.; Nilsson, U.J.; Leffler, H. Ligand induced galectin-3 protein self- association. J. Biol. Chem. 2012, 287, 21751-21756. 86. Kuklinski, S.; Probstmeier, R. Homophilic binding properties of galectin-3: Involvement of the carbohydrate recognition domain. J. Neurochem. 1998, 70, 814-823. 87. Rini, J.M.; Lobsanov, Y.D. New animal lectin structures. Curr. Opin. Struct. Biol. 1999, 9, 578-584. 88. Seetharaman, J.; Kanigsberg, A.; Slaaby, R.; Leffler, H.; Barondes, S.H.; Rini, J.M. X-ray crystal structure of the human galectin-3 carbohydrate recognition domain at 2.1-angstrom resolution. J. Biol. Chem. 1998, 273, 13047-13052. 89. Hirabayashi, J.; Hashidate, T.; Arata, Y.; Nishi, N.; Nakamura, T.; Hirashima, M.; Urashima, T.; Oka, T.; Futai, M.; Muller, W.E.G., et al. Oligosaccharide specificity of galectins: A search by frontal affinity chromatography. BBA-Gen. Subjects 2002, 1572, 232-254. 90. Collins, P.M.; Bum-Erdene, K.; Yu, X.; Blanchard, H. Galectin-3 interactions with glycosphingolipids. J. Mol. Biol. 2014, 426, 1439-1451. 91. Di Lella, S.; Ma, L.; Díaz Ricci, J.C.; Rabinovich, G.A.; Asher, S.A.; Álvarez, R.M.S. Critical role of the solvent environment in galectin-1 binding to the disaccharide lactose. Biochem. 2009, 48, 786-791. 92. Diehl, C.; Engstrom, O.; Delaine, T.; Hakansson, M.; Genheden, S.; Modig, K.; Leffler, H.; Ryde, U.; Nilsson, U.J.; Akke, M. Protein flexibility and conformational entropy in ligand design targeting the carbohydrate recognition domain of galectin-3. J. Am. Chem. Soc. 2010, 132, 14577-14589. 93. Öberg, C.T.; Leffler, H.; Nilsson, U.J. Inhibition of galectins with small molecules. Chimia 2011, 65, 18-23.
31
94. Salameh, B.A.; Cumpstey, I.; Sundin, A.; Leffler, H.; Nilsson, U.J. 1h-1,2,3-triazol-1-yl thiodigalactoside derivatives as high affinity galectin-3 inhibitors. Bioorg. Med. Chem. 2010, 18, 5367-5378. 95. Knibbs, R.N.; Agrwal, N.; Wang, J.L.; Goldstein, I.J. Carbohydrate-binding protein 35. Ii. Analysis of the interaction of the recombinant polypeptide with saccharides. J. Biol. Chem. 1993, 268, 14940-14947. 96. Meynier, C.; Feracci, M.; Espeli, M.; Chaspoul, F.; Gallice, P.; Schiff, C.; Guerlesquin, F.; Roche, P. Nmr and md investigations of human galectin-1/oligosaccharide complexes. Biophys. J. 2009, 97, 3168-3177. 97. Carlsson, S.; Oberg, C.T.; Carlsson, M.C.; Sundin, A.; Niisson, U.J.; Smith, D.; Cummings, R.D.; Almkvist, J.; Karlsson, A.; Leffler, H. Affinity of galectin-8 and its carbohydrate recognition domains for ligands in solution and at the cell surface. Glycobiology 2007, 17, 663-676. 98. Ideo, H.; Matsuzaka, T.; Nonaka, T.; Seko, A.; Yamashita, K. Galectin-8-n-domain recognition mechanism for sialylated and sulfated glycans. J. Biol. Chem. 2011, 286, 11346- 11355. 99. Leppänen, A.; Stowell, S.; Blixt, O.; Cummings, R.D. Dimeric galectin-1 binds with high affinity to alpha 2,3-sialylated and non-sialylated terminal n-acetyllactosamine units on surface-bound extended glycans. J. Biol. Chem. 2005, 280, 5549-5562. 100. Stowell, S.R.; Dias-Baruffi, M.; Penttila, L.; Renkonen, O.; Nyame, A.K.; Cummings, R.D. Human galectin-1 recognition of poly-n-acetyllactosamine and chimeric polysaccharides. Glycobiology 2004, 14, 157-167. 101. Di Virgilio, S.; Glushka, J.; Moremen, K.; Pierce, M. Enzymatic synthesis of natural and c-13 enriched linear poly-n-acetyllactosamines as ligands for galectin-1. Glycobiology 1999, 9, 353-364. 102. Qun, Z.; Cummings, R.D. L-14 lectin recognition of laminin and its promotion of in vitro cell adhesion. Arch. Biochem. Biophys. 1993, 300, 6-17. 103. Salomonsson, E.; Carlsson, M.C.; Osla, V.; Hendus-Altenburger, R.; Kahl-Knutson, B.; Oberg, C.T.; Sundin, A.; Nilsson, R.; Nordberg-Karlsson, E.; Nilsson, U.J., et al. Mutational tuning of galectin-3 specificity and biological function. J. Biol. Chem. 2010, 285, 35079- 35091. 104. Rapoport, E.M.; Andre, S.; Kurmyshkina, O.V.; Pochechueva, T.V.; Severov, V.V.; Pazynina, G.V.; Gabius, H.J.; Bovin, N.V. Galectin-loaded cells as a platform for the profiling of lectin specificity by fluorescent neoglycoconjugates: A case study on galectins-1 and-3 and the impact of assay setting. Glycobiology 2008, 18, 315-324. 105. Stowell, S.R.; Arthur, C.M.; Mehta, P.; Slanina, K.A.; Blixt, O.; Leffler, H.; Smith, D.F.; Cummings, R.D. Galectin-1,-2, and-3 exhibit differential recognition of sialylated glycans and blood group antigens. J. Biol. Chem. 2008, 283, 10109-10123. 106. Zhuo, Y.; Bellis, S.L. Emerging role of α2,6-sialic acid as a negative regulator of galectin binding and function. J. Biol. Chem. 2011, 286, 5935-5941. 107. Liu, F.T.; Rabinovich, G.A. Galectins: Regulators of acute and chronic inflammation. In Year in immunology 2, 2010; Vol. 1183, pp 158-182. 108. Davidson, P.J.; Davis, M.J.; Patterson, R.J.; Ripoche, M.-A.; Poirier, F.; Wang, J.L. Shuttling of galectin-3 between the nucleus and cytoplasm. Glycobiology 2002, 12, 329-337. 109. Probstmeier, R.; Montag, D.; Schachner, M. Galectin-3, a β-galactoside-binding animal lectin, binds to neural recognition molecules. J. Neurochem. 1995, 64, 2465-2472. 110. Hsu, D.K.; Zuberi, R.I.; Liu, F.T. Biochemical and biophysical characterization of human recombinant ige-binding protein, an s-type animal lectin. J. Biol. Chem. 1992, 267, 14167- 14174. 111. Colin Hughes, R. Galectins as modulators of cell adhesion. Biochimie 2001, 83, 667-676. 112. Massa, S.M.; Cooper, D.N.W.; Leffler, H.; Barondes, S.H. L-29, an endogenous lectin, binds to glycoconjugate ligands with positive cooperativity. Biochem. 1993, 32, 260-267. 113. Sato, S.; Hughes, R.C. Binding specificity of a baby hamster kidney lectin for h type i and ii chains, polylactosamine glycans, and appropriately glycosylated forms of laminin and fibronectin. J. Biol. Chem. 1992, 267, 6983-6990. 114. Ochieng, J.; Warfield, P.; Green-Jarvis, B.; Fentie, I. Galectin-3 regulates the adhesive interaction between breast carcinoma cells and elastin. J. Cell. Biochem. 1999, 75, 505-514.
32
115. Kuwabara, I.; Liu, F.T. Galectin-3 promotes adhesion of human neutrophils to laminin. J. Immunol. 1996, 156, 3939-3944. 116. Jeng, K.C.G.; Frigeri, L.G.; Liu, F.-T. An endogenous lectin, galectin-3 (ϵbp / mac-2), potentiates il-1 production by human monocytes. Immunol. Lett. 1994, 42, 113-116. 117. Hsu, D.K.; Hammes, S.R.; Kuwabara, I.; Greene, W.C.; Liu, F.T. Human t lymphotropic virus-i infection of human t lymphocytes induces expression of the beta-galactoside-binding lectin, galectin-3. Am. J. Pathol. 1996, 148, 1661-1670. 118. Kimata, H. Enhancement of ige production in b cells by neutrophils via galectin-3 in ige- associated atopic eczema/dermatitis syndrome. International Archives of Allergy and Immunology 2002, 128, 168-170. 119. Cortegano, I.; del Pozo, V.; Cárdaba, B.; de Andrés, B.; Gallardo, S.; del Amo, A.; Arrieta, I.; Jurado, A.; Palomino, P.; Liu, F.-T., et al. Galectin-3 down-regulates il-5 gene expression on different cell types. J. Immunol. 1998, 161, 385-389. 120. Acosta-Rodríguez, E.V.; Montes, C.L.; Motrán, C.C.; Zuniga, E.I.; Liu, F.-T.; Rabinovich, G.A.; Gruppi, A. Galectin-3 mediates il-4-induced survival and differentiation of b cells: Functional cross-talk and implications during trypanosoma cruzi infection. J. Immunol. 2004, 172, 493-502. 121. Oliveira, F.L.; Chammas, R.; Ricon, L.; Fermino, M.L.; Bernardes, E.S.; Hsu, D.K.; Liu, F.- T.; Borojevic, R.; El-Cheikh, M.C. Galectin-3 regulates peritoneal b1-cell differentiation into plasma cells. Glycobiology 2009, 19, 1248-1258. 122. Kim, H.-R.C.; Lin, H.-M.; Biliran, H.; Raz, A. Cell cycle arrest and inhibition of anoikis by galectin-3 in human breast epithelial cells. Cancer Res. 1999, 59, 4148-4154. 123. Dagher, S.F.; Wang, J.L.; Patterson, R.J. Identification of galectin-3 as a factor in pre-mrna splicing. Proc. Natl. Acad. Sci. 1995, 92, 1213-1217. 124. Yang, R.Y.; Hsu, D.K.; Liu, F.T. Expression of galectin-3 modulates t-cell growth and apoptosis. Proc. Natl. Acad. Sci. 1996, 93, 6737-6742. 125. Elad-Sfadia, G.; Haklai, R.; Balan, E.; Kloog, Y. Galectin-3 augments k-ras activation and triggers a ras signal that attenuates erk but not phosphoinositide 3-kinase activity. J. Biol. Chem. 2004, 279, 34922-34930. 126. Shalom-Feuerstein, R.; Cooks, T.; Raz, A.; Kloog, Y. Galectin-3 regulates a molecular switch from n-ras to k-ras usage in human breast carcinoma cells. Cancer Res. 2005, 65, 7292-7300. 127. Shimura, T.; Takenaka, Y.; Tsutsumi, S.; Hogan, V.; Kikuchi, A.; Raz, A. Galectin-3, a novel binding partner of β-catenin. Cancer Res. 2004, 64, 6363-6367. 128. Yu, F.; Finley, R.L.; Raz, A.; Kim, H.-R.C. Galectin-3 translocates to the perinuclear membranes and inhibits cytochrome c release from the mitochondria: A role for synexin in galectin-3 translocation. J. Biol. Chem. 2002, 277, 15819-15827. 129. Akahani, S.; Nangia-Makker, P.; Inohara, H.; Kim, H.-R.C.; Raz, A. Galectin-3: A novel antiapoptotic molecule with a functional bh1 (nwgr) domain of bcl-2 family. Cancer Res. 1997, 57, 5272-5276. 130. Zörnig, M.; Hueber, A.-O.; Baum, W.; Evan, G. Apoptosis regulators and their role in tumorigenesis. Biochimica et Biophysica Acta (BBA) - Reviews on Cancer 2001, 1551, F1- F37. 131. Chen, H.Y.; Liu, F.-T.; Yang, R.-Y. Roles of galectin-3 in immune responses. Arch. Immunol. Ther. Exp. 2005, 53, 497-504. 132. Nakahara, S.; Oka, N.; Raz, A. On the role of galectin-3 in cancer apoptosis. Apoptosis 2005, 10, 267-275. 133. Hsu, D.K.; Yang, R.-Y.; Pan, Z.; Yu, L.; Salomon, D.R.; Fung-Leung, W.-P.; Liu, F.-T. Targeted disruption of the galectin-3 gene results in attenuated peritoneal inflammatory responses. Am. J. Pathol. 2000, 156, 1073-1083. 134. Jiang, H.-R.; Al Rasebi, Z.; Mensah-Brown, E.; Shahin, A.; Xu, D.; Goodyear, C.S.; Fukada, S.Y.; Liu, F.-T.; Liew, F.Y.; Lukic, M.L. Galectin-3 deficiency reduces the severity of experimental autoimmune encephalomyelitis. J. Immunol. 2009, 182, 1167-1173. 135. Yoshii, T.; Inohara, H.; Takenaka, Y.; Honjo, Y.; Akahani, S.; Nomura, T.; Raz, A.; Kubo, T. Galectin-3 maintains the transformed phenotype of thyroid papillary carcinoma cells. Int. J. Oncol. 2001, 18, 787-792.
33
136. Fukumori, T.; Takenaka, Y.; Oka, N.; Yoshii, T.; Hogan, V.; Inohara, H.; Kanayama, H.-o.; Kim, H.-R.C.; Raz, A. Endogenous galectin-3 determines the routing of cd95 apoptotic signaling pathways. Cancer Res. 2004, 64, 3376-3379. 137. LIU, L.; SAKAI, T.; SANO, N.; FUKUI, K. Nucling mediates apoptosis by inhibiting expression of galectin-3 through interference with nuclear factor kappab signalling. Biochem. J. 2004, 380, 31-41. 138. Nangia-Makker, P.; Balan, V.; Raz, A. Galectin-3 binding and metastasis. In Metastasis research protocols, Dwek, M.; Brooks, S.A.; Schumacher, U., Eds. Humana Press: Totowa, NJ, 2012; pp 251-266. 139. Newlaczyl, A.U.; Yu, L.G. Galectin-3-a jack-of-all-trades in cancer. Cancer Letters 2011, 313, 123-128. 140. Liu, F.T.; Rabinovich, G.A. Galectins as modulators of tumour progression. Nat. Rev. Cancer 2005, 5, 29-41. 141. Califice, S.; Castronovo, V.; Bracke, M.; van den Brule, F. Dual activities of galectin-3 in human prostate cancer: Tumor suppression of nuclear galectin-3 vs tumor promotion of cytoplasmic galectin-3. Oncogene 2004, 23, 7527-7536. 142. Sun, W.; Li, L.; Yang, Q.; Shan, W.; Zhang, Z.; Huang, Y. G3-c12 peptide reverses galectin-3 from foe to friend for active targeting cancer treatment. Mol. Pharm. 2015, 12, 4124-4136. 143. Yu, L.-G.; Andrews, N.; Zhao, Q.; McKean, D.; Williams, J.F.; Connor, L.J.; Gerasimenko, O.V.; Hilkens, J.; Hirabayashi, J.; Kasai, K., et al. Galectin-3 interaction with thomsen- friedenreich disaccharide on cancer-associated muc1 causes increased cancer cell endothelial adhesion. J. Biol. Chem. 2007, 282, 773-781. 144. Fortuna-Costa, A.; Gomes, A.M.; Kozlowski, E.O.; Stelling, M.P.; Pavão, M.S.G. Extracellular galectin-3 in tumor progression and metastasis. Front. Oncol. 2014, 4, 138. 145. Yang, R.-Y.; Rabinovich, G.A.; Liu, F.-T. Galectins: Structure, function and therapeutic potential. Expert Rev. Mol. Med. 2008, 10. 146. Markowska, A.I.; Liu, F.T.; Panjwani, N. Galectin-3 is an important mediator of vegf- and bfgf-mediated angiogenic response. J. Exp. Med. 2010, 207, 1981-1993. 147. Markowska, A.I.; Jefferies, K.C.; Panjwani, N. Galectin-3 protein modulates cell surface expression and activation of vascular endothelial growth factor receptor 2 in human endothelial cells. J. Biol. Chem. 2011, 286, 29913-29921. 148. Wojtowicz-Praga, S.M.; Dickson, R.B.; Hawkins, M.J. Matrix metalloproteinase inhibitors. Invest. New Drug. 1997, 15, 61-75. 149. Sato, H.; Seiki, M. Membrane-type matrix metalloproteinases (mt-mmps) in tumor metastasis1. The Journal of Biochemistry 1996, 119, 209-215. 150. Atkinson, S.J.; Crabbe, T.; Cowell, S.; Ward, R.V.; Butler, M.J.; Sato, H.; Seiki, M.; Reynolds, J.J.; Murphy, G. Intermolecular autolytic cleavage can contribute to the activation of progelatinase a by cell membranes. J. Biol. Chem. 1995, 270, 30479-30485. 151. Zucker, S.; Lysik, R.M.; Zarrabi, M.H.; Moll, U. mr 92,000 type iv collagenase is increased in plasma of patients with colon cancer and breast cancer. Cancer Res. 1993, 53, 140-146. 152. Brown, P.D.; Bloxidge, R.E.; Stuart, N.S.A.; Galler, K.C.; Carmichael, J. Association between expression of activated 72-kilodalton gelatinase and tumor spread in non-small-cell lung carcinoma. J. Natl. Cancer Inst. 1993, 85, 574-578. 153. John, C.M.; Leffler, H.; Kahl-Knutsson, B.; Svensson, I.; Jarvis, G.A. Truncated galectin-3 inhibits tumor growth and metastasis in orthotopic nude mouse model of human breast cancer. Clin. Cancer Res. 2003, 9, 2374-2383. 154. Mirandola, L.; Nguyen, D.D.; Rahman, R.L.; Grizzi, F.; Yuefei, Y.; Figueroa, J.A.; Jenkins, M.R.; Cobos, E.; Chiriva-Internati, M. Anti-galectin-3 therapy: A new chance for multiple myeloma and ovarian cancer? Int. Rev. Immunol. 2014, 33, 417-427. 155. Mirandola, L.; Yu, Y.; Cannon, M.J.; Jenkins, M.R.; Rahman, R.L.; Nguyen, D.D.; Grizzi, F.; Cobos, E.; Figueroa, J.A.; Chiriva-Internati, M. Galectin-3 inhibition suppresses drug resistance, motility, invasion and angiogenic potential in ovarian cancer. Gynecol. Oncol. 2014, 135, 573-579. 156. Moon, B.-K.; Lee, Y.J.; Battle, P.; Jessup, J.M.; Raz, A.; Kim, H.-R.C. Galectin-3 protects human breast carcinoma cells against nitric oxide-induced apoptosis: Implication of galectin-3 function during metastasis. Am. J. Pathol. 2001, 159, 1055-1060.
34
157. Fukumori, T.; Kanayama, H.-o.; Raz, A. The role of galectin-3 in cancer drug resistance. Drug Resis. Updat. 2007, 10, 101-108. 158. Diehl, C.; Genheden, S.; Modig, K.; Ryde, U.; Akke, M. Conformational entropy changes upon lactose binding to the carbohydrate recognition domain of galectin-3. J. Biomol. NMR 2009, 45, 157-169. 159. Saraboji, K.; Hakansson, M.; Genheden, S.; Diehl, C.; Qvist, J.; Weininger, U.; Nilsson, U.J.; Leffler, H.; Ryde, U.; Akke, M., et al. The carbohydrate-binding site in galectin-3 is preorganized to recognize a sugarlike framework of oxygens: Ultra-high-resolution structures and water dynamics. Biochem. 2012, 51, 296-306. 160. Sörme, P.; Kahl-Knutsson, B.; Wellmar, U.; Magnusson, B.-G.; Leffler, H.; Nilsson, U.J. Design and synthesis of galectin inhibitors. Methods Enzymol. 2003, 363, 157-169. 161. Sörme, P.; Arnoux, P.; Kahl-Knutsson, B.; Leffler, H.; Rini, J.M.; Nilsson, U.J. Structural and thermodynamic studies on cation-π interactions in lectin-ligand complexes: High-affinity galectin-3 inhibitors through fine-tuning of an arginine-arene interaction. J. Am. Chem. Soc. 2005, 127, 1737-1743. 162. Cumpstey, I.; Sundin, A.; Leffler, H.; Nilsson, U.J. C2‐symmetrical thiodigalactoside bis‐ benzamido derivatives as high‐affinity inhibitors of galectin‐3: Efficient lectin inhibition through double arginine–arene interactions. Angew. Chem. 2005, 117, 5240-5242. 163. Bianchet, M.A.; Ahmed, H.; Vasta, G.R.; Amzel, L.M. Soluble beta-galactosyl-binding lectin (galectin) from toad ovary: Crystallographic studies of two protein-sugar complexes. Proteins 2000, 40, 378-388. 164. Cumpstey, I.; Salomonsson, E.; Sundin, A.; Leffler, H.; Nilsson, U.J. Double affinity amplification of galectin–ligand interactions through arginine–arene interactions: Synthetic, thermodynamic, and computational studies with aromatic diamido thiodigalactosides. Chem. Eur. J. 2008, 14, 4233-4245. 165. van Hattum, H.; Branderhorst, H.M.; Moret, E.E.; Nilsson, U.J.; Leffler, H.; Pieters, R.J. Tuning the preference of thiodigalactoside- and lactosamine-based ligands to galectin-3 over galectin-1. J. Med. Chem. 2013, 56, 1350-1354. 166. Rauthu, S.R.; Shiao, T.C.; André, S.; Miller, M.C.; Madej, É.; Mayo, K.H.; Gabius, H.-J.; Roy, R. Defining the potential of aglycone modifications for affinity/selectivity enhancement against medically relevant lectins: Synthesis, activity screening, and hsqc-based nmr analysis. ChemBioChem 2015, 16, 126-139. 167. Giguère, D.; Patnam, R.; Bellefleur, M.-A.; St-Pierre, C.; Sato, S.; Roy, R. Carbohydrate triazoles and isoxazoles as inhibitors of galectins-1 and -3. Chem. Commun. 2006, 2379-2381. 168. Marchiori, M.F.; Souto, D.E.; Bortot, L.O.; Pereira, J.F.; Kubota, L.T.; Cummings, R.D.; Dias-Baruffi, M.; Carvalho, I.; Campo, V.L. Synthetic 1,2,3-triazole-linked glycoconjugates bind with high affinity to human galectin-3. Bioorg. Med. Chem. 2015, 23, 3414-3425. 169. Öberg, C.T.; Leffler, H.; Nilsson, U.J. Arginine binding motifs: Design and synthesis of galactose-derived arginine tweezers as galectin-3 inhibitors. J. Med. Chem. 2008, 51, 2297- 2301. 170. Tejler, J.; Salameh, B.; Leffler, H.; Nilsson, U.J. Fragment-based development of triazole- substituted o-galactosyl aldoximes with fragment-induced affinity and selectivity for galectin- 3. Org. Biomol. Chem. 2009, 7, 3982-3990. 171. Salameh, B.A.; Leffler, H.; Nilsson, U.J. 3-(1, 2, 3-triazol-1-yl)-1-thio-galactosides as small, efficient, and hydrolytically stable inhibitors of galectin-3. Bioorg. Med. Chem. Lett. 2005, 15, 3344-3346. 172. Tejler, J.; Leffler, H.; Nilsson, U.J. Synthesis of o-galactosyl aldoximes as potent lacnac- mimetic galectin-3 inhibitors. Bioorg. Med. Chem. Lett. 2005, 15, 2343-2345. 173. Giguère, D.; Sato, S.; St-Pierre, C.; Sirois, S.; Roy, R. Aryl o- and s-galactosides and lactosides as specific inhibitors of human galectins-1 and -3: Role of electrostatic potential at o-3. Bioorg. Med. Chem. Lett. 2006, 16, 1668-1672. 174. André, S.; Kövér, K.E.; Gabius, H.-J.; Szilágyi, L. Thio- and selenoglycosides as ligands for biomedically relevant lectins: Valency–activity correlations for benzene-based dithiogalactoside clusters and first assessment for (di)selenodigalactosides. Bioorg. Med. Chem. Lett. 2015, 25, 931-935.
35
175. Lobsanov, Y.D.; Gitt, M.A.; Leffler, H.; Barondes, S.H.; Rini, J.M. X-ray crystal structure of the human dimeric s-lac lectin, l-14-ii, in complex with lactose at 2.9-a resolution. J. Biol. Chem. 1993, 268, 27034-27038. 176. Krzeminski, M.; Singh, T.; Andre, S.; Lensch, M.; Wu, A.M.; Bonvin, A.; Gabius, H.J. Human galectin-3 (mac-2 antigen): Defining molecular switches of affinity to natural glycoproteins, structural and dynamic aspects of glycan binding by flexible ligand docking and putative regulatory sequences in the proximal promoter region. BBA-Gen. Subjects 2011, 1810, 150-161. 177. Fort, S.; Kim, H.-S.; Hindsgaul, O. Screening for galectin-3 inhibitors from synthetic lacto-n- biose libraries using microscale affinity chromatography coupled to mass spectrometry. J. Org. Chem. 2006, 71, 7146-7154. 178. Fradin, C.; Poulain, D.; Jouault, T. Β-1,2-linked oligomannosides from candida albicans bind to a 32-kilodalton macrophage membrane protein homologous to the mammalian lectin galectin-3. Infect. Immun. 2000, 68, 4391-4398. 179. Tejler, J.; Skogman, F.; Leffler, H.; Nilsson, U.J. Synthesis of galactose-mimicking 1h-(1,2,3- triazol-1-yl)-mannosides as selective galectin-3 and 9n inhibitors. Carbohydr. Res. 2007, 342, 1869-1875. 180. Öberg, C.T.; Blanchard, H.; Leffler, H.; Nilsson, U.J. Protein subtype-targeting through ligand epimerization: Talose-selectivity of galectin-4 and galectin-8. Bioorg. Med. Chem. Lett. 2008, 18, 3691-3694. 181. Öberg, C.T.; Noresson, A.-L.; Leffler, H.; Nilsson, U.J. Synthesis of 3-amido-3-deoxy-β-d- talopyranosides: All-cis-substituted pyranosides as lectin inhibitors. Tetrahedron 2011, 67, 9164-9172. 182. Collins, P.M.; Öberg, C.T.; Leffler, H.; Nilsson, U.J.; Blanchard, H. Taloside inhibitors of galectin-1 and galectin-3. Chem. Biol. Drug Des. 2012, 79, 339-346. 183. Kupper, C.E.; Böcker, S.; Liu, H.L.; Adamzyk, C.; van de Kamp, J.; Recker, T.; Lethaus, B.; Jahnen-Dechent, W.; Neuss, S.; Müller-Newen, G., et al. Fluorescent snap-tag galectin fusion proteins as novel tools in glycobiology. Curr. Pharm. Des. 2013, 19, 5457-5467. 184. Monzavi-Karbassi, B.; Cunto-Amesty, G.; Luo, P.; Kieber-Emmons, T. Peptide mimotopes as surrogate antigens of carbohydrates in vaccine discovery. Trends Biotechnol. 2002, 20, 207- 214. 185. Arnusch, C.J.; André, S.; Valentini, P.; Lensch, M.; Russwurm, R.; Siebert, H.-C.; Fischer, M.J.E.; Gabius, H.-J.; Pieters, R.J. Interference of the galactose-dependent binding of lectins by novel pentapeptide ligands. Bioorg. Med. Chem. Lett. 2004, 14, 1437-1440. 186. Zou, J.; Glinsky, V.V.; Landon, L.A.; Matthews, L.; Deutscher, S.L. Peptides specific to the galectin-3 carbohydrate recognition domain inhibit metastasis-associated cancer cell adhesion. Carcinogenesis 2005, 26, 309-318. 187. Kiessling, L.L.; Young, T.; Gruber, T.D.; Mortell, K.H. Multivalency in protein–carbohydrate recognition. In Glycoscience, Fraser-Reid, B.; Tatsuta, K.; Thiem, J., Eds. Springer: Berlin, Heidelberg, Germany: 2008; pp 2483-2523. 188. Pieters, R.J. Maximising multivalency effects in protein-carbohydrate interactions. Org. Biomol. Chem. 2009, 7, 2013-2025. 189. Pieters, R.J.; Arnusch, C.J.; Breukink, E. Membrane permeabilization by multivalent anti- microbial peptides. Protein and peptide letters 2009, 16, 736-742. 190. Bernardi, A.; Jimenez-Barbero, J.; Casnati, A.; de Castro, C.; Darbre, T.; Fieschi, F.; Finne, J.; Funken, H.; Jaeger, K.-E.; Lahmann, M., et al. Multivalent glycoconjugates as anti-pathogenic agents. Chem. Soc. Rev. 2013, 42, 4709-4727. 191. Peri, F. Clustered carbohydrates in synthetic vaccines. Chem. Soc. Rev. 2013, 42, 4543-4556. 192. Branson, T.R.; Turnbull, W.B. Bacterial toxin inhibitors based on multivalent scaffolds. Chem. Soc. Rev. 2013, 42, 4613-4622. 193. Chabre, Y.M.; Roy, R. Multivalent glycoconjugate syntheses and applications using aromatic scaffolds. Chem. Soc. Rev. 2013, 42, 4657-4708. 194. Marradi, M.; Chiodo, F.; Garcia, I.; Penades, S. Glyconanoparticles as multifunctional and multimodal carbohydrate systems. Chem. Soc. Rev. 2013, 42, 4728-4745. 195. Martinez, A.; Ortiz Mellet, C.; Garcia Fernandez, J.M. Cyclodextrin-based multivalent glycodisplays: Covalent and supramolecular conjugates to assess carbohydrate-protein interactions. Chem. Soc. Rev. 2013, 42, 4746-4773.
36
196. Hatano, K.; Matsuoka, K.; Terunuma, D. Carbosilane glycodendrimers. Chem. Soc. Rev. 2013, 42, 4574-4598. 197. Chen, Y.; Star, A.; Vidal, S. Sweet carbon nanostructures: Carbohydrate conjugates with carbon nanotubes and graphene and their applications. Chem. Soc. Rev. 2013, 42, 4532-4542. 198. Kennedy, D.C.; Grünstein, D.; Lai, C.-H.; Seeberger, P.H. Glycosylated nanoscale surfaces: Preparation and applications in medicine and molecular biology. Chem. Eur. J. 2013, 19, 3794-3800. 199. Sansone, F.; Casnati, A. Multivalent glycocalixarenes for recognition of biological macromolecules: Glycocalyx mimics capable of multitasking. Chem. Soc. Rev. 2013, 42, 4623-4639. 200. Cecioni, S.; Imberty, A.; Vidal, S. Glycomimetics versus multivalent glycoconjugates for the design of high affinity lectin ligands. Chem. Rev. 2014, 115, 525-561. 201. Gouin, S.G.; García Fernández, J.M.; Vanquelef, E.; Dupradeau, F.-Y.; Salomonsson, E.; Leffler, H.; Ortega-Muñoz, M.; Nilsson, U.J.; Kovensky, J. Multimeric lactoside “click clusters” as tools to investigate the effect of linker length in specific interactions with peanut lectin, galectin-1, and -3. ChemBioChem 2010, 11, 1430-1442. 202. André, S.; Pieters, R.J.; Vrasidas, I.; Kaltner, H.; Kuwabara, I.; Liu, F.T.; Liskamp, R.M.; Gabius, H.J. Wedgelike glycodendrimers as inhibitors of binding of mammalian galectins to glycoproteins, lactose maxiclusters, and cell surface glycoconjugates. ChemBioChem 2001, 2, 822-830. 203. Parera Pera, N.; Branderhorst, H.M.; Kooij, R.; Maierhofer, C.; van der Kaaden, M.; Liskamp, R.M.; Wittmann, V.; Ruijtenbeek, R.; Pieters, R.J. Rapid screening of lectins for multivalency effects with a glycodendrimer microarray. ChemBioChem 2010, 11, 1896-1904. 204. Li, Y.; Cheng, Y.; Xu, T. Design, synthesis and potent pharmaceutical applications of glycodendrimers: A mini review. Curr. Drug. Discov. Technol. 2007, 4, 246-254. 205. Wolfenden, M.; Cousin, J.; Nangia-Makker, P.; Raz, A.; Cloninger, M. Glycodendrimers and modified elisas: Tools to elucidate multivalent interactions of galectins 1 and 3. Molecules 2015, 20, 7059-7096. 206. Turnbull, W.B.; Stoddart, J.F. Design and synthesis of glycodendrimers. Rev. Mol. Biotechnol. 2002, 90, 231-255. 207. Penadés, S.; de la Fuente, J.M.; Barrientos, Á.G.; Clavel, C.; Martínez-Ávila, O.; Alcántara, D. Multifunctional glyconanoparticles : Applications in biology and biomedicine. In Nanomaterials for application in medicine and biology, Giersig, M.; Khomutov, G.B., Eds. Springer Netherlands: Dordrecht, 2008; pp 93-101. 208. de la Fuente, J.M.; Barrientos, A.G.; Rojas, T.C.; Rojo, J.; Cañada, J.; Fernández, A.; Penadés, S. Gold glyconanoparticles as water-soluble polyvalent models to study carbohydrate interactions. Angew. Chem. Int. Ed. 2001, 40, 2257-2261. 209. André, S.; Sansone, F.; Kaltner, H.; Casnati, A.; Kopitz, J.; Gabius, H.J.; Ungaro, R. Calix[n]arene-based glycoclusters: Bioactivity of thiourea-linked galactose/lactose moieties as inhibitors of binding of medically relevant lectins to a glycoprotein and cell-surface glycoconjugates and selectivity among human adhesion/growth-regulatory galectins. ChemBioChem 2008, 9, 1649-1661. 210. Nelson, A.; Belitsky, J.M.; Vidal, S.; Joiner, C.S.; Baum, L.G.; Stoddart, J.F. A self- assembled multivalent pseudopolyrotaxane for binding galectin-1. J. Am. Chem. Soc. 2004, 126, 11914-11922. 211. Tejler, J.; Tullberg, E.; Frejd, T.; Leffler, H.; Nilsson, U.J. Synthesis of multivalent lactose derivatives by 1,3-dipolar cycloadditions: Selective galectin-1 inhibition. Carbohydr. Res. 2006, 341, 1353-1362. 212. Giguère, D.; Bonin, M.-A.; Cloutier, P.; Patnam, R.; St-Pierre, C.; Sato, S.; Roy, R. Synthesis of stable and selective inhibitors of human galectins-1 and -3. Bioorg. Med. Chem. 2008, 16, 7811-7823. 213. Mackeviča, J.; Ostrovskis, P.; Leffler, H.; Nilsson, U.J.; Rudovica, V.; Viksna, A.; Belyakov, S.; Turks, M. Synthesis of 1, 2, 3-triazole-linked galactohybrids and their inhibitory activities on galectins. Arkivoc 2014. 214. André, S.; Cejas Ortega, P.J.; Perez, M.A.; Roy, R.; Gabius, H.-J. Lactose-containing starburst dendrimers: Influence of dendrimer generation and binding-site orientation of receptors
37
(plant/animal lectins and immunoglobulins) on binding properties. Glycobiology 1999, 9, 1253-1261. 215. Michel, A.K.; Nangia-Makker, P.; Raz, A.; Cloninger, M.J. Lactose-functionalized dendrimers arbitrate the interaction of galectin-3/muc1 mediated cancer cellular aggregation. ChemBioChem 2014, 15, 2106-2112. 216. Vrasidas, I.; Andre, S.; Valentini, P.; Bock, C.; Lensch, M.; Kaltner, H.; Liskamp, R.M.; Gabius, H.J.; Pieters, R.J. Rigidified multivalent lactose molecules and their interactions with mammalian galectins: A route to selective inhibitors. Org. Biomol. Chem. 2003, 1, 803-810. 217. André, S.; Grandjean, C.; Gautier, F.M.; Bernardi, S.; Sansone, F.; Gabius, H.J.; Ungaro, R. Combining carbohydrate substitutions at bioinspired positions with multivalent presentation towards optimising lectin inhibitors: Case study with calixarenes. Chem. Commun. 2011, 47, 6126-6128. 218. Bernardi, S.; Fezzardi, P.; Rispoli, G.; Sestito, S.E.; Peri, F.; Sansone, F.; Casnati, A. Clicked and long spaced galactosyl- and lactosylcalix[4]arenes: New multivalent galectin-3 ligands. Beilstein J. Org. Chem. 2014, 10, 1672-1680. 219. Yang, Y.; Xue, X.C.; Jin, X.F.; Wang, L.J.; Sha, Y.L.; Li, Z.J. Synthesis of multivalent n- acetyl lactosamine modified quantum dots for the study of carbohydrate and galectin-3 interactions. Tetrahedron 2012, 68, 7148-7154. 220. André, S.; Kaltner, H.; Furuike, T.; Nishimura, S.-I.; Gabius, H.-J. Persubstituted cyclodextrin-based glycoclusters as inhibitors of protein−carbohydrate recognition using purified plant and mammalian lectins and wild-type and lectin-gene-transfected tumor cells as targets. Bioconjug. Chem. 2004, 15, 87-98. 221. Aykaç, A.; Martos-Maldonado, M.C.; Casas-Solvas, J.M.; Quesada-Soriano, I.; García- Maroto, F.; García-Fuentes, L.; Vargas-Berenguel, A. Β-cyclodextrin-bearing gold glyconanoparticles for the development of site specific drug delivery systems. Langmuir 2014, 30, 234-242. 222. Stowell, C.P.; Lee, Y.C. Neoglycoproteins the preparation and application of synthetic glycoproteins. In Adv. Carbohydr. Chem. Biochem., Tipson, R.S.; Derek, H., Eds. Academic Press: Waltham, MA, USA, 1980; Vol. 37, pp 225-281. 223. Wang, H.; Huang, W.; Orwenyo, J.; Banerjee, A.; Vasta, G.R.; Wang, L.X. Design and synthesis of glycoprotein-based multivalent glyco-ligands for influenza hemagglutinin and human galectin-3. Bioorg. Med. Chem. 2013, 21, 2037-2044. 224. Kamath, V.P.; Diedrich, P.; Hindsgaul, O. Use of diethyl squarate for the coupling of oligosaccharide amines to carrier proteins and characterization of the resulting neoglycoproteins by maldi-tof mass spectrometry. Glycoconj. J. 1996, 13, 315-319. 225. Van den Berg, T.K.; Honing, H.; Franke, N.; van Remoortere, A.; Schiphorst, W.; Liu, F.T.; Deelder, A.M.; Cummings, R.D.; Hokke, C.H.; van Die, I. Lacdinac-glycans constitute a parasite pattern for galectin-3-mediated immune recognition. J. Immunol. 2004, 173, 1902- 1907.
38
2. Binding characteristics of galectin-3 fusion proteins – Influence of truncation and fusion Parts of this chapter were published in: S. Böcker, L. Elling, Glycobiology 2017, 27, 457-468.
Abstract Galectin-3 is a key player in cell adhesion and signaling events driven by specific binding and cross-linking of galactose based carbohydrate ligands. By cleavage caused by metalloproteinases N-terminally truncated galectin-3 still bearing the carbohydrate recognition domain is produced in vivo. Galectin-3 as truncated version has been shown to act in vivo as a negative inhibitor of full-length galectin-3 due to different affinity for carbohydrate ligands. We present studies on a series of twelve human galectin-3 protein constructs. With truncated galectin-3 (∆1-62 and ∆1-116) and fusions with SNAP-tag and/or yellow fluorescent protein (YFP), we show altered binding affinities to asialofetuin (ASF) in ELISA-type and surface plasmon resonance (SPR) binding assays. Galectin-3(Δ1-62) and native galectin-3 show highest affinity to ASF in both ELISA and SPR experiments, whereas galectin-3(Δ1-116) shows only weak binding. It is demonstrated here for the first time that SNAP-tag and YFP fusions of galectin-3 and truncated galectin-3 proteins modulate and improve binding affinity to ASF. SNAP-tagged galectin-3, galectin-3(Δ1-62), and galectin- 3(Δ1-116) give a three- to six-fold increased binding affinity in SPR compared to native galectin-3. Fusion of truncated galectin-3 with YFP reconstitutes binding properties similar to native galectin-3. In combination with a SNAP-tag even improved binding characteristics are obtained. These results emphasize the impact of the N-terminal domain of human galectin-3 on ligand affinity. Most importantly, by carefully selected fusion proteins it may become possible to tailor diagnostic and therapeutic tools with beneficially tuned binding properties. The resulting novel protein toolbox will be advantageous for the biomedical application of galectin-3 in cancer diagnosis and therapy.
39
2.1 Introduction Galectins are a class of lectins which share the common affinity towards β-galactoside glycan structures [1,2]. Galectins cover a broad range of biological functions such as mediation of cell adhesion, inflammation, signaling, and even tumor progression [3-8]. Galectin-1 (Gal-1) and galectin-3 (Gal-3) trigger tumor angiogenesis by crosslinking vascular endothelial growth factor receptor 2 (VEGFR2) and expanding its cell-surface retention [9-12]. This is initiated by binding of the galectins to N-acetyllactosamine (LacNAc), which is presented by VEGFR2. This makes both galectins potential targets for cancer therapy. Furthermore, labeled galectins are useful analysis-tools to detect cell-surface presented poly-LacNAc [13]. In the context of tumor angiogenesis, it could be possible to utilize fusion proteins of galectin-3 for imaging and/or drug delivery purposes. However, it is crucial to understand the binding characteristics of galectin fusion proteins to gather evidence for their potential as novel ‘theranostic’ biomolecules.
Gal-3 contains three distinct domains and is the only member of the ancient chimera family of galectins. The three domains are: a ~20 amino acids short N-terminal tail, a ~100 amino acids collagen-like domain, that is cleavable by metalloproteinases and the highly conserved ~130 amino acids carbohydrate recognition domain (CRD) [2,14]. The N-terminal domain of Gal-3 is crucial for the oligomerization of the protein [15-20]. However, the CRD itself is as discussed to be also involved in the self-association process [16-18,21]. Matrix- metalloproteinases MMP-2 and -9 cleave Gal-3 at position Ala62-Tyr63 which results in a 22 kDa truncated galectin, containing a truncated ~50 amino acids collagen-like domain and the intact CRD [22,23]. It should be noted that in vivo, this process takes place during tumor progression [24]. It has been reported that collagenase truncated Gal-3(Δ1-107) (Gal-3C), and MMP truncated Gal-3(Δ1-62) act as negative inhibitors of Gal-3. Interestingly, the truncated versions display higher binding affinity towards carbohydrate ligands [25,26]. Additionally, the self-association capability is hampered after MMP treatment [26,27]. It is thought that truncated Gal-3 may therefore act as inhibitor of tumor angiogenesis preventing tumor growth. Moreover, it may support the effect of anti-tumor drugs [25,28,29]. The analysis of the properties of truncated Gal-3 is an important step to understand Gal-3 mediated biological processes in more detail. However, to the best of our knowledge a detailed characterization of truncated Gal-3’s binding properties has not been performed, so far.
Fluorescently labeled galectins are an interesting tool for the facile evaluation of their binding characteristics [30-32]. Labeling may be realized by randomized chemical modification of galectins with activated dyes [31-35], which may affect residues involved in binding. Alternatively, fusions with fluorescent proteins like eGFP have been used for detection [36-
40
39]. However, the influence of the fusion partner on the actual functionality in comparison with native galectin has been rarely performed. Another fusion protein enabling specific modification of the target protein is reached by the SNAP-tag technology. The SNAP-tag introduces a mutated DNA-repair protein O6-alkylguanine-DNA alkyltransferase. This enzyme reacts with benzylguanine (BG) to form a covalent bond. Derivatives of BG can be used to enable e.g. directed immobilization of functional proteins [40-43] or directed conjugation of fluorescent dyes [44-47].
Recently, we have introduced fusion proteins of Gal-1 and Gal-3 with both, N-terminal yellow fluorescent protein (YFP) and SNAP-tag [13]. The protein constructs proved similar binding characteristics to the glycoprotein standard asialofetuin (ASF) in comparison to non- fused galectins. In flow cytometry experiments we could demonstrate specific binding of fluorescent SNAP-tagged Gal-3 to mesenchymal stem cells. We also immobilized Gal-3 fusion protein on BG-activated Sepharose beads and achieved high binding capacity for glycoproteins. We concluded so far that the N-terminal SNAP-YFP fusion does not alter the binding characteristics of Gal-3.
Here, we report on the creation and analysis of a toolbox of full-length and truncated Gal-3 fusion proteins. N-terminal His6-tagged full-length Gal-3 and two truncated versions, Gal- 3(Δ1-62) and Gal-3(Δ1-116), were fused to a SNAP-tag and/or YFP, respectively. For evaluating the binding properties of all twelve Gal-3 fusion proteins (Figure 2.1) to ASF, static and flow conditions using solid phase ELISA (enzyme-linked immunosorbent assay) and SPR (surface plasmon resonance), respectively, were performed. Our results prove protein truncation and fusion protein partners dependent binding properties of the Gal-3 fusion proteins.
2.2 Materials and methods 2.2.1 Cloning of galectin constructs All full-length and truncated Gal-3 constructs for this study have been designed based on the pETDuet-1-vector encoding H6Gal-3 and the pET17b-vector encoding H6SYGal-3 [13]. For the truncations of Gal-3 we followed on the one hand the main cleavage site of the metalloproteinases MMP-2 and -9 between Ala62-Tyr63 (Gal-3(Δ1-62)) and on the other hand the truncation of the entire N-terminal and collagen-like domain until Pro117 (Gal-3(Δ1-116)).
The genes for Gal-3(Δ1-62) and Gal-3(Δ1-116) were amplified from pETDuet-H6Gal-3 and the restriction sites for EcoRI and NotI were simultaneously introduced using the primers 5´- AAGAA TTCAA TGTAC CCTGG AGCAC CTGG-3´ and 5´-TTTGC GGCCG CTTAT
41
ATCAT GGTAT ATGAA GCACT GGTG-3´ for H6Gal-3(Δ1-62) and 5´-AAGAA TTCAA TGCCT TATAA CCTGC CTTTG CC-3´ and 5´-TTTGC GGCCG CTTAT ATCAT GGTAT
ATGAA GCACT GGTG-3´ for H6Gal-3(Δ1-116). After cutting the pETDuet-H6Gal-3 and the PCR products with EcoRI and NotI the final vectors pETDuet-H6Gal-3(Δ1-62) and -
H6Gal-3(Δ1-116) were ligated.
The cloning of the Gal-3 fusion proteins with SNAP-tag and YFP was very similar. Amplification and insertion of the restriction sites for BsrGI and NotI were done using the primers 5´-ACATG TACAA AATGT ACCCT GGAGC ACCTG G-3´ and 5´-TTTGC
GGCCG CTTAT ATCAT GGTAT ATGAA GCACT GGTG-3´ for H6SYGal-3(Δ1-62) and 5´-ACATG TACAA AATGC CTTAT AACCT GCCTT TGCC-3´ and 5´-TTTGC GGCCG
CTTAT ATCAT GGTAT ATGAA GCACT GGTG-3´ for H6SYGal-3(Δ1-116). Afterwards, the restricted pET17b-H6SYGal-3 and the particular PCR product were ligated to pET17b-
H6SYGal-3(Δ1-62) and -H6SYGal-3(Δ1-116).
The SNAP-tag fusion constructs H6SGal-3, H6SGal-3(Δ1-62), and H6SGal-3(Δ1-116) were cloned via restriction of pET17b-H6SYGal-3, -H6SYGal-3(Δ1-62), and -H6SYGal-3(Δ1-116), respectively, with AgeI und BsrGI to remove YFP-gene, Klenow fill-in of the sticky ends and subsequent blunt-end ligation of the vectors.
To generate YFP fusion constructs the genes for YFP-Gal-3, YFP-Gal-3(Δ1-162), and YFP- Gal-3(Δ1-116) with restriction sites for NdeI and NotI were amplified from pET17b-
H6SYGal-3, pET17b-H6SYGal-3(Δ1-62), and pET17b-H6SYGal-3(Δ1-116), respectively, using the primers 5´-AAAAA ACATA TGGTG AGCAA GGGC-3´and 5´-TTTGC GGCCG CTTAT ATCAT GGTAT ATGAA GCACT GGTG-3´. After restriction of pET28a-vector and PCR-products both were ligated resulting in pET28a-H6YGal-3, -H6YGal-3(Δ1-62), and -
H6YGal-3(Δ1-116). All ligation products were transformed in competent cells either E.coli NovaBlue (Novagen/Merck, Darmstadt, Germany) or E.coli NEB Turbo (NEB, Frankfurt/Main, Germany) for plasmid selection and isolation. Successful cloning was confirmed by sequencing.
2.2.2 Expression and purification All Gal-3 fusion proteins were expressed in E. coli Rosetta (DE3) pLysS (Novagen/Merck, Darmstadt, Germany). TB medium with 0.5 mM IPTG was used for induction of expression. After 24 h, cells were harvested by centrifugation and appropriate cell mass was sonicated two times for 30 s for follow-up affinity chromatography. Recombinant Gal-3 constructs were purified by immobilized metal-ion affinity chromatography (IMAC) using HisTrapTM HP
42
5 mL column (GE Healthcare, Munich, Germany). Purification was performed according to manufacturer’s instructions. After column equilibration bacterial crude extracts were loaded onto the column at a flow rate of 1 mL/min and washed until baseline signal was reached.
Target proteins were eluted with buffer containing 500 mM imidazole. His6-tagged and SNAP-tagged Gal-3 constructs were stored in phosphate buffered saline containing 2 mM EDTA (EPBS, pH 7.5) at 4°C, while YFP as well as SNAP-YFP Gal-3 constructs were frozen in EPBS at -20 °C. Buffer exchange was achieved by ultrafiltration (Amicon® Ultra-15, Merck Millipore, Darmstadt, Germany). Protein concentrations were determined by Bradford assay (Roti®-Quant, Carl Roth, Karlsruhe, Germany) using bovine serum albumin for calibration.
Affinity chromatography purification with α-Lactose-Agarose (Sigma-Aldrich, Taufkirchen, Germany) was performed using EPBS as running buffer. 5 to 10 mg IMAC-purified protein in EPBS were concentrated by ultrafiltration (Amicon® Ultra-15) and loaded onto Lactose- Agarose, washed, and eluted with 200 mM lactose.
2.2.3 SDS-PAGE and western blot The purification and the size of the purified proteins were checked by SDS-PAGE followed by western blot. Here, 0.4 µg for Coomassie staining and 0.08 µg for western blot were applied. The proteins transferred to the PVDF-membrane were detected by incubation with CRD-specific anti-Gal- 3 antibody (Gal379, Biolegend, Fell, Germany) as primary antibody followed by incubation with anti- mouse-peroxidase (Sigma-Aldrich).
2.2.4 Size exclusion chromatography
Double purified proteins were analyzed by SEC using TSK-GEL® G3000SWXL HPLC column (5 µm, 7.8 mm ID x 30.0 cm L, 10-500 kDa (globular proteins), Tosoh bioscience, Stuttgart, Germany). 0.2 nmol protein in 20 µL EPBS was applied onto the column with a flow rate of 0.5 mL/min and detected at 280 nm. The buffer consisting of 0.05 M phosphate, 0.15 M NaCl, pH 7 was used for protein elution. For molecular mass calculation different calibration standards (Serva, Heidelberg, Germany and Bio-Rad, Munich, Germany) were used (Figure S1).
2.2.5 Immobilization of recombinant galectin to Sepharose beads NHS-activated Sepharose (GE Healthcare) was used for immobilizing selected Gal-3 constructs. The coupling was generally done as described previously [13]. For non-directed
43 coupling of His-tagged proteins via amino groups, the material was prepared according to manufacturer’s instructions. The reaction was performed in coupling buffer pH 8.3 and residual NHS was blocked with 500 mM ethanolamine for 4 h followed by final inactivation of NHS by alternating washing with basic and acidic buffers. Non-reacted protein was removed by washing with PBS.
In case of fluorescent SNAP-tagged Gal-3 variants, coupling to NHS-activated Sepharose was done in a directed way applying the SNAP-tag technology with prior binding of amino- benzylguanine. The material was prepared for subsequent reaction of aminobenzylguanine (NEB) with NHS-Sepharose in dimethyl-formamide (DMF) over night at 4°C. Residual NHS groups were blocked and inactivated as described. SNAP reaction of the protein construct and the benzylguanine modified Sepharose was performed in PBS with 1 mM DTT (dithiothreitol) 1 h at room temperature and followed by overnight incubation at 4°C. Non- reacted protein was removed by washing with PBS.
2.2.6 Self-association/crosslinking assays of galectin-3 25 µL Sepharose beads with immobilized Gal-3 were incubated with 200 µL soluble Gal-3 (200 µg/mL in PBS) for one hour in presence or absence of 150 mM lactose. After washing three times with PBS by centrifugation the beads were treated with LDS-sample buffer (NuPAGE™,Thermo Fisher Scientific). The supernatant was analyzed by SDS-PAGE with subsequent western blot as described before.
The ability of Gal-3 constructs to crosslink laminin and asialofetuin (ASF) was performed in a ELISA-type assay, generally as described in literature [13]. After immobilization of ASF (Sigma-Aldrich, 5 µg/mL in sodium carbonate buffer pH 9.6) and blocking with bovine serum albumin (BSA, 2% in PBS) 10 µM of His-tagged Gal-3 variants or Gal-3 fusion proteins with SNAP-tag and YFP were incubated for one hour in PBS. After washing three times with PBS containing 0.05% Tween® 20 (AppliChem, Darmstadt, Germany) 5 µg/mL laminin (from mouse Engelbreth-Holm-Swarm sarcoma, Sigma-Aldrich) were incubated for one hour and excess was removed by washing with PBS-Tween. Laminin was detected by primary anti- laminin antibody (Sigma-Aldrich, 1:1000 in PBS) followed by washing and subsequent incubation with peroxidase labeled anti-rabbit antibody (Sigma-Aldrich, 1:1000 in PBS). Signal determination was done by conversion of o-phenylenediamine (OPD, Dako, Hamburg, Germany) and read-out at 492 nm.
44
2.2.7 Galectin binding assay on asialofetuin Binding of full-length and truncated Gal-3 fusion proteins to ASF was analyzed in 96 well microtiter plate format as previously described [13]. Briefly, ASF (200 µL of 5 µg/mL bovine ASF in sodium carbonate buffer pH 9.6) was immobilized in microtiter plates (MaxiSorp, Nunc, Wiesbaden, Germany) over night. After blocking of residual binding sites with BSA (2% in PBS) different amounts of galectin were incubated for 1 h in EPBS. Three times washing was done with 250 µL PBS-Tween between the incubation steps. Bound galectin was detected by incubation with anti-His6-peroxidase (Roche, Basel, Switzerland, 1:1000 in PBS) and subsequent conversion of OPD-substrate with read-out at 492 nm. The presented data are the results of at least three independent measurements. Kd and the maximal binding response were calculated for each galectin-3 fusion construct by non-linear fitting (ligand binding
퐵푚푎푥∙x model: 푦 = , SigmaPlot (Systat, Erkrath, Germany)). Kd resembles therefore the 퐾푑+푥 concentration at half-maximum binding signal.
2.2.8 Inhibition of galectin binding with (Di-)LacNAc-linker-tBoc Specific binding of the full-length and truncated Gal-3 constructs was proved by two oligosaccharides used for competitive inhibition studies of galectin-ASF binding. LacNAc- linker-tBoc (LacNAc) and Di-LacNAc-linker-tBoc (Di-LacNAc) were synthesized and purified as described (Chapter 3.2.2) [48,49]. After immobilization of ASF and blocking as described before, different concentrations of saccharide were simultaneously incubated with galectin for 1 h in EPBS. Controls without glycan and without galectin were performed to indicate minimal and maximal binding, respectively. Residual bound galectin was detected as described by anti-His6-peroxidase and OPD conversion. All assays were reproduced in at least three independent measurements.
2.2.9 Surface plasmon resonance spectroscopy Surface plasmon resonance (SPR) spectroscopy was performed with Reichert SR7500DC System (XanTec, Düsseldorf, Germany; provided by Dr. Andreas Walther and Prof. Dr. Martin Möller (DWI Leibniz Institute for Interactive Materials, Aachen)) using carboxymethyldextran hydrogel sensor chips (200M, XanTec). ASF was immobilized on a chip via EDC-Sulfo-NHS coupling. The immobilization was carried out with a flowrate of 10 µL/min. After activating the surface with 40 µL Sulfo-NHS/EDC in 100 mM MES 10 µL ASF (5 µg/mL in 10 mM acetate buffer, pH 4.5) was applied on the sample channel. To get similar surface condition for sample and reference both flow cells were treated with 2% BSA
45 for two minutes. Remaining NHS esters were blocked by injecting 1 M ethanolamine (60 µL, pH 8.5).
The binding experiments were carried out with a flow rate of 20 µL/min by successively injecting seven different galectin concentrations in EPBS from low to high concentration (0.08 to 40 µM). The dissociation time was three minutes. Between the measurements of different Gal-3 fusion proteins the surface was regenerated several times with 1 M NaCl/20 mM HCl and 500 mM lactose.
The measured data were subtracted by reference and blank values using Scrubber2 (BioLogic Software, Campbell, Australia). The averaged binding response values for each concentration at the equilibrium binding (average responses from 60 s to 140 s) of two different measurements were plotted against the galectin concentration and Kd and the maximal binding response were calculated for each Gal-3 fusion construct by non-linear fitting (ligand 퐵 ∙x binding model: 푦 = 푚푎푥 , SigmaPlot (Systat, Erkrath, Germany)). The maximal binding 퐾푑+푥 signals were related to protein size by division by corresponding molecular masses.
2.2.10 Flow cytometry with human umbilical vein endothelial cells Fluorescence-activated cell sorting (FACS) was performed in collaboration with Anne Rix from the Department of Experimental Molecular Imaging (Prof. Dr. med. Fabian Kiessling, Helmholtz Institute for Biomedical Engineering, RWTH Aachen University). Human umbilical vein endothelial cells (HUVECs, PromoCell, Heidelberg, Germany) were cultivated in VascuLife VEGF medium (Lifeline Cell Technologies, Frederick, MD, US) including 1% penicillin/streptomycin (Invitrogen/Thermo Fisher Scientific, Dreieich, Germany) until reaching passage 5 or 6. Cells were trypsinized with trypsin-EDTA solution (0.25% trypsin, 0.05% EDTA, in PBS, Invitrogen/Thermo Fisher Scientific) and 106 cells (cell counting with Cedex XS, Roche, Mannheim, Germany) for one measurement were centrifuged at 1000 rpm for 5 min. After 30 min incubation with 135 nM fusion construct (H6SYGal-3, H6SYGal-
3(Δ1-62) and H6SYGal-3(Δ1-116)), cells were washed with PBS including 10% fetal bovine serum and by centrifugation. The cell pellet was resuspended in 500 µL FACSFlow (BD Bioscience, Heidelberg, Germany) and analyzed by FACS (FACSCalibur (BD Bioscience)) at 488 nm excitation and detection with a 530/30 BP filter. For inhibition analyses, the cells were first incubated with inhibitor (75 mM lactose, 75 mM sucrose, 1.25 mM LacNAc) for 15 min before incubation with galectin. As control, pure cells were measured. The data were evaluated by the geometric mean result and by using Flowing Software2 (version 2.5.1, by Perttu Terho, Turku Centre of Biotechnology, University of Turku, Finland).
46
2.3 Results and discussion 2.3.1 Production and characterization of full-length and truncated galectin-3 fusion constructs
Twelve different His6-tagged Gal-3 constructs with alternated fusion protein partners and different truncation were designed (Figure 2.1), produced by expression in E.coli Rosetta (DE3) pLysS and purified by IMAC. The received constructs were named as shown in Figure 2.1. Protein yields ranging from 2.5 mg to 19 mg protein per 1 g cells were reached depending on the fusion partners and the truncations (Table 2.1). H6Gal-3 showed the lowest expression yield, whereas SNAP-tag as well as truncation increased the expression yield.
Figure 2.1: Protein fusion constructs of human Gal-3 Twelve different Gal-3 fusion proteins were designed. The constructs differ in the length of the N- terminus and/or type of fusion protein partners. N-terminal His6-tagged Gal-3 proteins with full-length Gal-3, MMP-derived Δ1-62 truncation of Gal-3 (Gal-3(Δ1-62)), and N-terminal Δ1-116 truncation of Gal-3 (Gal-3(Δ1-116)) were fused to SNAP-tag (S) and/or YFP (Y), respectively.
Table 2.1: Yields of Gal-3 fusion proteins after IMAC purification Amount of purified Gal-3 fusion proteins is given in mg protein per g cells.
Protein Yield [mg/g] Protein Yield [mg/g] H6Gal-3 2.5 H6YGal-3 3.5 H6Gal-3(Δ1-62) 20.0 H6YGal-3(Δ1-62) 10.0 H6Gal-3(Δ1-116) 17.0 H6YGal-3(Δ1-116) 10.0 H6SGal-3 9.0 H6SYGal-3 12.0 H6SGal-3(Δ1-62) 19.0 H6SYGal-3(Δ1-62) 17.5 H6SGal-3(Δ1-116) 16.0 H6SYGal-3(Δ1-116) 15.5 (reprinted with permission; Copyright © 2017, Oxford University Press)
To obtain highly purified proteins as shown in SDS-PAGE (Figure 2.2A, B) and western blot (Figure 2.2C, D) a second purification step by lactose agarose affinity chromatography was applied. The detected protein bands fit well to the theoretical molecular mass of the corresponding fusion protein (Figure 2.2 and Table S2.1). The CRD was present in all detected proteins on the stained gel as proven by western blot with a CRD-specific anti-Gal-3
47 antibody. Hence, expression and purification of the constructs results in pure and intact proteins. H6Gal-3 shows an additional faint band at 29 kDa. An additional reported potential cleavage site Gly32-Ala33 for collagenases producing a 27 kDa product could be the reason [27,50]. This implies collagenase impurity during the purification. On the other hand, it could also be an artefact of the SDS-PAGE.
(reprinted with permission; Copyright © 2017, Oxford University Press) Figure 2.2: SDS-PAGE and western blot analysis of purified Gal-3 fusion proteins The purified Gal-3 fusion proteins were analyzed by SDS-PAGE with Coomassie-staining (A, B) and by western blot (C, D) using an anti-Gal-3 antibody.
M – Marker, a – full-length Gal-3, b – Gal-3(Δ1-62), c – Gal-3(Δ1-116), 1 – His6-tag (H6) fusion, 2 –
H6-SNAP-tag fusion, 3 – H6-YFP fusion, 4 – H6-SNAP-tag-YFP fusion
Multiple functions of Gal-3 are triggered by its oligomerization [4,51]. In previous studies, it was shown that truncated Gal-3 (Δ1-62 and Δ1-107) does not self-associate readily [18,26]. To investigate the oligomerization potential of our constructs in solution we performed size exclusion chromatography (SEC) defining the influence of the SNAP-tag and YFP as well as the truncation (Figure S2.1 and Figure S2.2 of supporting information).
48
In Figure 2.3 it becomes obvious, that the SEC-derived molecular masses fit very well to the theoretical values for all protein constructs. Moreover, all proteins eluted from the column in one single peak (see Figure S2.2). We conclude that the presence of monomeric Gal-3 in solution, as described formerly for native human Gal-3 [19,52-54], is neither influenced by YFP nor SNAP-tag.
Figure 2.3: Comparison of molecular masses of Gal-3 fusion proteins obtained by SDS-PAGE and SEC Molecular masses of Gal-3 fusion proteins determined by SEC (light-grey) and by SDS-PAGE (grey) as well as theoretical molecular masses (black) are given. See Table S2.1 in Supplementary data for molecular mass values.
2.3.2 Self-association/crosslinking potential of galectin-3 fusion proteins The self-association potential of selected Gal-3 constructs was determined using beads for immobilization according to a similar experiment [17]. The idea if this experiment was the retention of Gal-3 by the Gal-3 conjugated beads due to the association with each other. Thus, possible interaction of Gal-3 in solution with immobilized Gal-3 was detected after elution under appropriate conditions. After incubation of Gal-3 in PBS with Gal-3 immobilized on beads and washing with PBS, bound proteins were eluted with LDS-sample buffer and analyzed by western blot. There was no elution possible with 150 mM lactose or 100 mM DTT.
Gal-3 constructs carrying only a His-tag were analyzed using random immobilization, whereas Gal-3 fusion proteins with His-tag, SNAP-tag and YFP were immobilized orientated by the SNAP-tag technology [40-43].
49
The immobilized full-length H6Gal-3 was incubated with itself (H6Gal-3) and with the two truncated versions (H6Gal-3(Δ1-62), H6Gal-3(Δ1-116)). The western blot of the eluted fractions indicates interactions of the involved Gal-3 constructs because corresponding bands were detected by the antibody (Figure 2.4A). Both immobilized truncated constructs also interacted with themselves as bands at proper molecular weight are visible (Figure 2.4A). Gal- 3 conjugated beads incubated with PBS showed no or very low signal at the height of the immobilized protein due to minimal release. These interactions also occurred in presence of 150 mM lactose (Figure 2.4B) indicating that the interactions did not involve the carbohydrate binding pocket. These results show that Gal-3 self-associates even if it is N-terminally truncated and are contradictory to published data showing that the self-association of Gal-3 is diminished after N-terminal truncation [26]. C-term dependent interactions of Gal-3 were also observed in a very similar experiment [17], but the described inhibition with lactose is in disagreement with the present study.
Figure 2.4: Western blot analyses of Gal-3 after interaction experiments Gal-3 in solution was incubated with immobilized Gal-3 in the presence and absence of 150 mM lactose and eluted with LDS-sample buffer. The eluates were analyzed by SDS-PAGE and following western blot. The CRD of Gal-3 was detected with a specific antibody.
M – Marker, a – full-length Gal-3, b – Gal-3(Δ1-62), c – Gal-3(Δ1-116), 1 – His6-tag (H6) fusion, 4 –
H6-SNAP-tag-YFP fusion
50
The same observations were made with the fusion constructs. In the absence and presence of 150 mM lactose, full-length and truncated Gal-3 interacted with immobilized full-length and truncated Gal-3, respectively (Figure 2.4C). There is no visible influence of the fusion partners, although an enhancement of the oligomerization potential by the additional amino acid sequences of the fusion proteins cannot be excluded.
In summary, interactions of soluble Gal-3 with immobilized Gal-3 that involve neither the N- term nor the carbohydrate binding site were observed. In consequence, the association has to be C-term mediated but takes place outside the carbohydrate binding pocket. These results prove C-terminal self-association but do not exclude further possible interaction sites. Recently published results implicating the exclusion of C-terminal interactions or lactose inhibitable self-association [17,26] are controversial to the results in this work.
Another assay to analyze the oligomerization potential of the Gal-3 constructs was done in ELISA-format. In this assay, the ability of Gal-3 to crosslink the ECM glycoprotein laminin which carries poly-LacNAc chains [55,56] with immobilized ASF was determined. Crosslinked laminin to an ASF coated microtiter plate was detected by specific antibodies. As laminin itself can recognize glycans on ASF [57,58] the signal of the control experiment without galectin was subtracted (Figure 2.5).
Figure 2.5: Crosslinked laminin to an ASF coated microtiter plate by the different Gal-3 constructs The ability of selected Gal-3 constructs to crosslink laminin with immobilized ASF was determined in a solid-phase ELISA assay.
51
H6Gal-3 showed good crosslinking of laminin with ASF as we could also prove in an earlier study [13]. For H6Gal-3(Δ1-62) a slightly lower crosslink ability was determined, whereas the crosslinking failed for H6Gal-3(Δ1-116) (Figure 2.5). As seen in the next chapter, H6Gal- 3(Δ1-116) binds to ASF only in a weak manner. Thus, there were possibly too less molecules to form oligomers after ASF binding. As aforesaid, H6Gal-3(Δ1-116) showed the ability to self-associate in a ligand-independent way. However, the binding to immobilized ASF may lead to conformational change making oligomerization impossible. Another reason could be a very low binding strength to laminin that was not tested in this work. The amount of laminin crosslinked by H6Gal-3(Δ1-62) was lower than by H6Gal-3, although the absolute binding to ASF was higher (see next chapter). This suggests that either the affinity of laminin is lower or the arrangement of the oligomers is less efficient for binding other ligands. It is possible that more H6Gal-3(Δ1-62) molecules are necessary to lead to full laminin binding. The fusion proteins of full-length Gal-3 with SNAP-tag or SNAP-tag and YFP crosslinked nearly the same amount of laminin compared to H6Gal-3 (Figure 2.5). Moreover, truncation of Gal-3 in fluorescent SNAP-tagged fusion proteins showed no influence on the crosslinking ability. The N-terminal fusion partners may mimic the lacking Gal-3 sequence. This phenomenon is also seen for ASF binding behavior as explained in the following chapter.
2.3.3 Binding of galectin-3 fusion proteins to asialofetuin in a solid-phase assay For analyzing and characterizing the binding properties of Gal-3, the glycoprotein ASF is reported as a suitable ligand [59]. In Figure 2.6 the binding curves for each galectin construct is depicted. Absolute binding signals of galectin constructs purified solely by IMAC as well as by IMAC followed by lactose affinity chromatography are very similar (Figure S2.3). We concluded that the additional purification step can be avoided and IMAC purification is sufficient for comparative binding assays of these constructs.
His6-tagged Gal-3 proteins show significantly different binding characteristics depending on the degree of truncation (Figure 2.6A). With H6Gal-3 as benchmark, H6Gal-3(Δ1-62) binds most efficient whereas the H6Gal-3(Δ1-116) shows weakest binding in this trial (Figure 2.6A and Table S2.2). The calculated apparent Kd values reflect these differences (Figure 2.7 and
Table 2.2). Within the His6-tagged constructs, H6Gal-3(Δ1-62) reaches the highest maximum binding and has the highest affinity to ASF with an apparent Kd value of 4 µM (Figure 2.7 and Table 2.2). This confirms previously published data on tight binding of truncated Gal- 3(Δ1-62) to immobilized laminin utilizing a similar solid-phase assay [26]. Most interestingly,
H6Gal-3(Δ1-116) lacking the complete N-terminal domain, yet presenting the CRD gives a
52 five-fold higher Kd compared to H6Gal-3 indicating poor binding (Table 2.2). Our results emphasize the functional importance of N-terminal sequences of human Gal-3 for ligand binding. A synthetic peptide consisting of amino acids 98 to 112 of human Gal-3, a part of the N-terminal domain, showed association with the CRD [60] and is probably involved in tight ligand interaction.
Figure 2.6: Binding of Gal-3 protein constructs to ASF in a solid-phase ELISA assay
Absolute binding signals are compared for the different fusion types: His6-tag (H6, A); H6-SNAP-tag
(H6S, B); H6-YFP (H6Y, C); H6-SNAP-tag-YFP (H6SY, D), and truncation types: full-length Gal-3 (○); Gal-3(Δ1-62) (▽); Gal-3(Δ1-116) (□). In B binding curves are additionally zoomed for better comparison in the low concentration range.
Binding of all Gal-3 constructs to ASF is significantly improved by N-terminal fusion with a SNAP-tag (Figure 2.6 and Table 2.2). Differences among the differently truncated Gal-3 constructs occur only in subtle manner at low protein concentrations (Figure 2.6B, insert). Protein concentrations below 10 µM are sufficient to achieve maximum binding with best binding performance of H6SGal-3(Δ1-62) followed by H6SGal-3 and H6SGal-3(Δ1-116).
When compared to the His6-tagged Gal-3 constructs, it is striking that significantly lower apparent Kd values (up to 35-fold enhanced affinity) are reached (Figure 2.7B and Table 2.2). Most importantly, the substitution of the complete N-terminal domain by the SNAP-tag
53 reconstitutes the function of Gal-3 CRD to a fully functional lectin with a 6-fold improved binding affinity towards ASF. We summarize that a SNAP-tag fusion is generally beneficial for Gal-3 binding to ASF in a solid-phase ELISA. The individual binding properties of the Gal-3 constructs are as well altered by fusion with YFP (Figure 2.6C). Binding curves similar to those of H6Gal-3 (Figure 2.6A) are obtained. Importantly, also YFP fusion reconstitutes function of Gal-3(Δ1-116) reaching a similar maximum binding and similar Kd value as calculated for full-length Gal-3 (Figure 2.7 and Table S2.2). However, Figure 2.7 indicates that YFP fusion has no pronounced effect on the binding of full length Gal-3 (H6Gal-3 vs.
H6YGal-3). We conclude that N-terminal fusion of YFP or SNAP-tag to truncated Gal-3 may take over the role of the natural N-terminal domain (Figure 2.1).
Table 2.2: Calculated apparent Kd values and relative improvement of binding affinities for Gal-3 fusion proteins in ELISA and SPR binding experiments to ASF
ELISA SPR
Protein Apparent Kd [µM] Factor Apparent Kd [µM] Factor H6Gal-3 5.56 ± 1.35 1.0 32.46 ± 7.40 1.0 H6Gal-3(Δ1-62) 4.02 ± 0.51 1.4 40.37 ± 9.58 0.8 H6Gal-3(Δ1-116) 26.81 ± 5.79 0.2 60.02 ± 28.4 0.5 H6SGal-3 0.70 ± 0.16 7.9 8.56 ± 0.65 3.8 H6SGal-3(Δ1-62) 0.16 ± 0.04 34.8 10.40 ± 2.13 3.1 H6SGal-3(Δ1-116) 0.96 ± 0.14 5.8 5.55 ± 1.60 5.8 H6YGal-3 4.19 ± 1.52 1.3 12.79 ± 0.43 2.5 H6YGal-3(Δ1-62) 5.28 ± 1.34 1.1 27.01 ± 2.08 1.2 H6YGal-3(Δ1-116) 5.18 ± 0.87 1.1 37.22 ± 5.42 0.9 H6SYGal-3 3.03 ± 0.87 1.8 9.89 ± 2.33 3.3 H6SYGal-3(Δ1-62) 1.61 ± 0.19 3.5 12.92 ± 1.45 2.5 H6SYGal-3(Δ1-116) 2.26 ± 0.34 2.5 20.29 ± 3.38 1.6
By fusing SNAP-tag and YFP to Gal-3 the effects of each fusion protein may be combined (Figure 2.6D). Full-length and truncated constructs show similar binding curves despite enhanced binding to ASF. The concerted action of both fusions yield overall improved binding values: The SNAP-tag increases the maximum binding signal as well as the binding affinity and YFP diminishes the effect of truncation (Figure 2.7 and Table 2.2). To exclude that SNAP-tag and YFP were directly involved in binding, a suitable negative control, H6SY lacking complete Gal-3, was used and did not show any binding to ASF (Figure S2.4 of supporting information). However, we cannot exclude an influence of the SNAP-tag on protein oligomerization leading to higher binding signals as complexed SNAP-tagged Gal-3 is detected. It is well known that oligomerization of Gal-3 occurs upon ligand binding [15,19,54] and this may influence binding assays based on direct protein detection as performed here. Additionally, washing steps may remove weakly bound monomeric Gal-3
54 constructs whereas oligomerized proteins will bind more tightly due to chelating effects. As the binding capacity of a fixed ASF amount is considered constant, the required washing procedure between the incubation steps is sufficient to identify binding differences between the fusion proteins.
Figure 2.7: Calculated maximum binding and apparent Kd values for Gal-3 fusion protein binding to ASF in a solid-phase ELISA assay For all fusion and truncation variants of Gal-3 values for maximal binding to ASF (A) as well as apparent Kd (B) are compared. Values were calculated by non-linear fit of binding data (see Figure 2.6).
Sugar-mediated binding of all Gal-3 constructs to ASF was verified by inhibition experiments with competing soluble saccharides. The binding of all tested constructs is inhibited by LacNAc and Di-LacNAc as shown in Figure 2.8. As inhibition by the free glycan can be considered a competitive inhibition, larger inhibition potency is directly related to higher affinity towards the soluble inhibitor. Since the affinity of Gal-3 is higher to Di-LacNAc than to the disaccharide LacNAc [13], it is not surprising that the inhibition curves for Di-LacNAc are shifted to lower inhibitor concentrations. As this is the case for all constructs, it becomes clear that the fusion partners have no influence on the ligand specificity of Gal-3. Recently, we could prove that, compared to H6Gal-3, H6SYGal-3 shows very similar binding specificity even to longer oligosaccharides with up to four LacNAc units [13]. Our results confirm a previous study on a fusion protein of alkaline phosphatase and Gal-3 that also did not show any alteration in binding specificity [36].
55
Figure 2.8: Competitive inhibition of binding of galectin-3 fusion proteins to ASF by (Di-)LacNAc All twelve galectin-3 constructs show inhibition of their binding with LacNAc (●) and Di-LacNAc (●). The inhibition curves shifted to lower inhibitor concentrations if Di-LAcNAc is applied instead of LacNAc. The data reflect the different binding affinity of galectin-3 fusion proteins toward ASF (see Table 2.2). For galectin-3 constructs with high binding affinity to ASF higher competitive inhibitor concentrations were needed to inhibit binding to ASF. Standard deviation of at least three independent measurements is represented as errors bars.
A – H6Gal-3, B – H6SGal-3, C – H6YGal-3, D – H6SYGal-3, E – H6Gal-3(Δ1-62),
F – H6SGal-3(Δ1-62), G – H6YGal-3(Δ1-62), H – H6SYGal-3(Δ1-62), I – H6Gal-3(Δ1-116), J –
H6SGal-3(Δ1-116), K – H6YGal-3(Δ1-116), L – H6SYGal-3(Δ1-116)
Stronger inhibition of H6Gal-3 than for H6Gal-3(Δ1-62) was observed, and for H6Gal-3(Δ1-116) even less inhibitor was required (A, E, I). The inhibition curves for SNAP-tagged galectin-3 fusion proteins (3 µM) are clearly shifted to higher inhibitor concentrations (B, F, J) although lower galectin concentrations were used when compared to His6-tagged galectin-3 constructs (15 µM). The fluorescent galectin-3 fusion proteins (5 µM) show similar inhibition behavior for all truncation types
(C, G, K), which is similar to the inhibition of H6Gal-3 at three-fold higher galectin concentrations. For fluorescent SNAP-tagged galectin-3 fusion proteins (3 µM) inhibition curves similar to these for fluorescent constructs were recorded with slight tendency to higher inhibitor concentrations (D, H, L).
-5 -7 Our determined apparent Kd values in the area of 10 to 10 M fit well with a reported Kd value (2 µM) for Gal-3 binding to ASF, which was obtained by a titration experiment [61]. However, it is clear that truncations as well as fusion protein partners have a significant influence on the Kd values (Figure 2.7B). The strongest binding to ASF was recorded for
H6SGal-3(Δ1-62) shown by an apparent Kd in the nanomolar range, which is about 35-fold
56 lower than the value for H6Gal-3. Once more, the SNAP-tag may influence the CRD and increase the self-oligomerization. Previous studies on differences in the binding characteristics of full-length and truncated Gal-3 are confirmed by our findings. It was shown that MMP-2 cleaved Gal-3 displays 20-fold higher affinity to HUVECs in vivo [50]. In this study, we found a 40% higher binding signal of H6Gal-3(Δ1-62) compared to H6Gal-3 (Table 2.2). Although the crystal structure of human Gal-3(Δ1-112) reveals optimal binding of lactose indicating that the amino acids 1-112 are not involved in ligand binding[62], we observe only low binding of H6Gal-3(Δ1-116). It is possible that certain N-terminally missing amino acids indeed contribute to ligand binding. The four additionally truncated amino acids may contribute to structural stability of Gal-3 despite the fact that they are just outside the canonical galectin beta-sandwich. Another hint for the involvement of the N-terminal domain in ligand binding is a previous study with hamster Gal-3 binding to the glycoprotein laminin [63]: Truncated Gal-3(Δ1-93) binds stronger to laminin than Gal-3(Δ1-103). Here, the contribution of Tyr102 and adjacent amino acids to oligosaccharide binding becomes clear. It was proposed that the small N-terminal region induces a change in CRD structure altering carbohydrate binding specificity.
Not only binding properties but also other biological functions are altered by N-terminal truncation of Gal-3. Substitution of a lacking N-terminal region could also be observed in nuclear localization experiments with hamster and murine Gal-3 [37,64]. Certain sequence motifs within the N-terminal domain are important for nuclear localization, which is hampered after deletion. The relevant motifs could be substituted by a completely unrelated sequence, e.g. the fusion with green fluorescent protein and maltose binding protein, leading to similar localization of truncated and full-length Gal-3.
2.3.4 Surface plasmon resonance spectroscopy of galectin-3 fusion proteins on immobilized asialofetuin To monitor the binding characteristics of the Gal-3 constructs to ASF under flow conditions, SPR experiments were performed. SPR sensor chip with a surface mounted 3D matrix for ligand multilayer immobilization were utilized. In detail, ASF was covalently immobilized via its lysine amino groups on an EDC/Sulfo-NHS activated carboxymethyldextran hydrogel chip. Additionally, BSA was subsequently immobilized on both, sample and reference flow cell. BSA immobilization turned out to be beneficial due to equal protein loading on sample and reference channel as well as blocking unspecific adhesion of galectin.
57
Figure 2.9: Binding of Gal-3 fusion proteins to ASF in SPR measurements Galectin binding to immobilized ASF was monitored under flow (20 µL/min). Binding responses at equilibrium of seven concentrations (0.08 to 40 µM) of Gal-3 fusion proteins were plotted and fitted by non-linear regression. Binding curves are shown for the different fusion types: His6-tag (H6, A); H6-
SNAP-tag (H6S, B); H6-YFP (H6Y, C); H6-SNAP-tag-YFP (H6SY, D), and truncation types: full- length Gal-3 (○); Gal-3(Δ1-62) (▽); Gal-3(Δ1-116) (□).
During binding experiments a constant flow rate of 20 µL/min was maintained. Binding curves are presented in Figure 2.9. They are calculated from the averaged response at equilibrium of association for seven different Gal-3 concentrations (Figure S2.6 of supporting information). In general, the kon and koff signals suggest fast association and dissociation, and dissociation did not reach baseline completely, especially at high protein concentrations (see Figure S2.6). This behavior of galectin-3 was previously shown in other SPR experiments as well [65]. Fusion proteins with SNAP-tag and YFP at high protein concentrations resulted in incomplete dissociation (Figure S2.6). Figure 2.10 summarizes the calculated maximum binding response and apparent Kd values of the Gal-3 constructs. SPR experiments were performed because of two main reasons: First, to get a qualitative idea about binding kinetics and second, to investigate the binding behavior of galectin-constructs under flow conditions.
58
It is crucial to ensure highest comparability between ELISA and SPR analysis. Therefore, we calculated Kd of SPR experiments not by the kinetical constants kon and koff, but in the same way as mentioned for the ELISA approach: Kd is the concentration at which 50% saturation of the binding signal is reached. This methodology was previously used for SPR-based galectin studies by other groups [65].
In contrast to the ELISA assay, SPR reveals highest binding for full-length Gal-3 when compared to Gal-3(Δ1-62) (Figure 2.9A-C). H6Gal-3(Δ1-116) however, binds weakest to ASF which was also found in the ELISA assay. SNAP-tag and YFP fusion proteins of Gal-
3(Δ1-62) and Gal-3(Δ1-116) show very similar binding curves (Figure 2.9B and C). For His6-
YFP and the His6-SNAP-tag-YFP fusion, differences can be observed: The response of Gal- 3(Δ1-116) reaches the highest values, followed by Gal-3(Δ1-62) and Gal-3 (Figure 2.10A). It is possible that constructs with a truncation and therefore lower molecular mass face less steric hindrance and reach higher saturation levels. SNAP-tag fusion proteins, in contrast, show equal maximum binding signals (Figure 2.10A). To diminish the effect of different molecular weight the signals were normalized by the respective molecular mass of the protein. We summarize that, contrary to the ELISA assay, Gal-3(Δ1-62) shows not the best binding to ASF in SPR. However, in both flow and solid-phase assays the N-terminal SNAP- tag and/or YFP fusions are functional substitutes of the lacking N-terminal domain and reconstitute a fully functional lectin.
Figure 2.10: Calculated values for maximum binding and apparent Kd of Gal-3 protein constructs for binding to ASF in SPR measurements For all fusion and truncation variants of Gal-3 values for maximal binding to ASF (A) as well as apparent Kd (B) are compared. Values were calculated by non-linear fit of binding data (see Figure 2.9). Normalized maximal binding signals were obtained after division by molecular mass.
59
In contrast to the solid-phase assay, the binding differences of the His6-tagged Gal-3 constructs are not as distinct, when derived from the apparent Kd values, yet significant
(Figure 2.9B and Table 2.2). Indicated by lower apparent Kd, the SNAP-tag and SNAP-tag- YFP fusions show higher affinity to ASF than the YFP-fused variants (Table 2.2). The order of affinity constants for the different fusion types is the same as the data derived from ELISA assay. However, due to flow conditions apparent Kd values are overall higher than those observed in solid-phase assay (Table 2.2). Flow makes binding to a ligand more difficult due to shear forces and the fast dissociation processes with a generally weak binding. Proteins do not always show the same behavior under flow conditions compared to static conditions in solid-phase assays [66,67]. The procedures in both experiments are very different. With ELISA assay binding efficiency is calculated at equilibrium conditions reached after one hour incubation time and several fixed washing or incubation steps. The washing procedure may influence the maximum signal. Therefore, it is not surprising that under flow conditions like in SPR higher Kd values compared to static ELISA conditions are rendered. The very different time-course of both experiments leads to different times for enabling the establishment of equilibrium. Despite differences in set-up and aim of research it has been reported elsewhere that measured values of static assays and flow-assays may differ. An inhibition assay to detect paralytic shellfish poisoning toxins by ELISA and SPR biosensor technology gave lower IC50 values under static ELISA conditions [66]. Similar results, weaker binding under flow- conditions, were reported for a dengue virus immunoassay [67].
The lowest affinity is detected with His6-tagged galectins, whereas the lowest apparent Kd values (highest affinity) are reached by SNAP-tag fused galectins (Figure 2.10B and Table
2.2). Remarkably, in both assays H6SGal-3(Δ1-116) gives a 6-fold higher affinity for ASF in comparison to H6Gal-3 (Table 2.2). The reported Kd of 21.5 µM for human Gal-3 in a similar experimental setup is in good agreement with our apparent Kd for H6Gal-3 of about 32 µM (Table 2.2) [65]. Interestingly, it was not possible to verify the higher affinity of Gal-3(Δ1-62) compared to Gal-3 in SPR measurements. Although several studies proved better binding of truncated Gal-3 [26,50], isothermal titration calorimetry revealed very similar affinity constants for full-length and truncated Gal-3 [68,69]. Other studies even showed reversed results where a two-fold higher association constant was measured for Gal-3 compared to truncated Gal-3 [70].
In summary, similar conclusions as for ELISA assay can be drawn from our SPR experiments. Binding to ASF is improved by N-terminal fusions of SNAP-tag to full-length and truncated Gal-3 (Δ1-62 and Δ1-116) as well as to YFP-fusion proteins thereof. In addition, the binding affinity of Gal-3(Δ1-116) is enhanced by SNAP-tag and/or YFP fusions.
60
Our results show that SNAP-tagged Gal-3 fusion proteins with their enhanced affinity could be candidates for cancer therapy and may also serve as novel tools in molecular imaging. The SNAP-tag enables conjugation of drugs, fluorescent dyes, and other reporter groups, which eases the localization of Gal-3 in vivo. Endogenous tailored Gal-3 could lead to reduced tumor angiogenesis, metastasis, and progression and enhance the effect of truncated Gal-3 which is present in tumor [50,71]. Exogenous truncated Gal-3 was shown to successfully decrease tumor volumes and metastases in mouse models of breast cancer, multiple myeloma, and ovarian cancer [25,28,29,72].
2.3.5 Binding of galectins to human umbilical vein endothelial cells Gal-3 binds to cell surface ligands like integrins and growth factor receptors. In this way, Gal- 3 enables cell-cell interaction and adhesion to other compounds, e.g. of the ECM. In FACS- analysis we proved the binding of the fluorescent SNAP-tagged Gal-3 constructs to HUVECs. The YFP fusion enabled the direct detection of the proteins.
Figure 2.11: FACS-analysis of SNAP-tag-YFP fusion Gal-3 binding to HUVECs The fluorescent signal of Gal-3 constructs was measured in flow cytometry with HUVECs (A) and in presence of 1.25 mM LacNAc (B). In (C) varying concentrations of LacNAc were added to H6SYGal- 3 cell mixture showing increasing inhibition. 75 mM lactose inhibited the binding of all constructs completely (D) and 75 mM sucrose showed no influence (E).
61
Figure 2.11A shows clear labeling of HUVECs with full-length and truncated Gal-3 fusion proteins. The binding of Gal-3 to cell surface VEGFR2 and αvβ3-integrins, as they are present on HUVECs, is already known [12,73]. The fusion partner has no influence on the cell binding ability of Gal-3. In a recent study, we already proved cell binding of the full length fusion construct to mesenchymal stem cells [13].
Compared to truncated Gal-3, the binding of full-length Gal-3 to HUVECs was less affine. The affinity increased with increasing Gal-3 truncation. Gal-3(Δ1-116) showed slightly better binding than Gal-3(Δ1-62) whereas both binding signals were clearly higher than for full- length Gal-3 (Figure 2.11A). It is also possible that more truncated Gal-3 molecules bind on the cell surface than full-length Gal-3. This observation coincides with reported results of higher HUVECs binding of MMP-2 cleaved Gal-3 compared to full-length Gal-3, both completely inhibited by lactose [50]. The binding of the fusion proteins to HUVECs in this study was also completely inhibited in the presence of 75 mM lactose (Figure 2.11D). Sucrose had no influence on the cell binding as seen in Figure 2.11E. LacNAc inhibited the binding at lower concentration. In Figure 2.11C, different LacNAc concentration were tested showing about half complete inhibition of full-length Gal-3 binding with 2.5 mM LacNAc. In the presence of 1.25 mM LacNAc the three constructs were compared (Figure 2.11B). The range is identical as seen in Figure 2.11A but due to inhibitory influence a shift to the left is observed.The cell-assay supports the results of ELISA and SPR measurements and ensures the cell binding ability of the fusion constructs.
2.4 Conclusion We demonstrate for the first time the influence of truncations and fusion protein partners on the binding characteristics of human Gal-3. We showed a higher affinity of Gal-3(Δ1-62) toward ASF compared to full-length Gal-3 in solid-phase assay, which is in agreement with literature. In contrast, flow conditions alter the behavior of Gal-3 in that way that here full- length Gal-3 showed highest ASF affinity. SNAP-tag fusion of Gal-3 enables increased binding affinity towards ASF, even for Gal-3(Δ1-116). YFP as fusion protein partner substitutes successfully the N-terminal region of Gal-3 and levels out the influence of truncations. We emphasize that the N-terminal domain is important for ligand binding in a structural manner but can be replaced by an unrelated sequence, here provided by the corresponding fusion proteins. However, the self-association potential of Gal-3 seemed not be affected by the modifications. As even highly truncated Gal-3 could self-associate, we conclude that interaction of Gal-3 has to involve the C-terminus. Additionally, we found that binding characteristics of Gal-3 fusion proteins are strongly dependent on the experimental
62 set-up with flow or with static conditions. This is important for galectin-based ‘theranostics’ and their mode of application. It is not surprising that the galectin binding is highly dynamic and adjustable by binding conditions as Gal-3 takes part in a vast of biological actions. This study may give important insights on the binding characteristics of Gal-3 and especially Gal- 3(Δ1-62) as well as Gal-3(Δ1-116) and how galectins may be tailored by fusions and beneficially tuned for higher binding affinity, paving the path to future applications in tumor therapy and diagnostics.
2.5 Contributions S. Böcker cloned the galectin constructs, synthesized the glycans and performed ELISA and SPR binding assays, A. Rix and S. Böcker did FACS analysis, S. Böcker evaluated the data.
2.6 References 1. Cooper, D.N. Galectinomics: Finding themes in complexity. Biochim. Biophys. Acta 2002, 1572, 209-231. 2. Barondes, S.H.; Cooper, D.N.; Gitt, M.A.; Leffler, H. Galectins. Structure and function of a large family of animal lectins. J. Biol. Chem. 1994, 269, 20807-20810. 3. Almkvist, J.; Karlsson, A. Galectins as inflammatory mediators. Glycoconj. J. 2004, 19, 575- 581. 4. Elola, M.T.; Wolfenstein-Todel, C.; Troncoso, M.F.; Vasta, G.R.; Rabinovich, G.A. Galectins: Matricellular glycan-binding proteins linking cell adhesion, migration, and survival. Cell. Mol. Life Sci. 2007, 64, 1679-1700. 5. Hernandez, J.D.; Baum, L.G. Ah, sweet mystery of death! Galectins and control of cell fate. Glycobiology 2002, 12, 127R-136R. 6. Liu, F.T.; Patterson, R.J.; Wang, J.L. Intracellular functions of galectins. BBA-Gen. Subjects 2002, 1572, 263-273. 7. Liu, F.T.; Rabinovich, G.A. Galectins as modulators of tumour progression. Nat. Rev. Cancer 2005, 5, 29-41. 8. Rabinovich, G.A.; Toscano, M.A. Turning 'sweet' on immunity: Galectin-glycan interactions in immune tolerance and inflammation. Nat. Rev. Immunol. 2009, 9, 338-352. 9. Funasaka, T.; Raz, A.; Nangia-Makker, P. Galectin-3 in angiogenesis and metastasis. Glycobiology 2014, 24, 886-891. 10. Croci, Diego O.; Cerliani, Juan P.; Dalotto-Moreno, T.; Méndez-Huergo, Santiago P.; Mascanfroni, Ivan D.; Dergan-Dylon, S.; Toscano, Marta A.; Caramelo, Julio J.; García- Vallejo, Juan J.; Ouyang, J., et al. Glycosylation-dependent lectin-receptor interactions preserve angiogenesis in anti-vegf refractory tumors. Cell 2014, 156, 744-758. 11. Stanley, P. Galectin-1 pulls the strings on vegfr2. Cell 2014, 156, 625-626. 12. Markowska, A.I.; Liu, F.T.; Panjwani, N. Galectin-3 is an important mediator of vegf- and bfgf-mediated angiogenic response. J. Exp. Med. 2010, 207, 1981-1993. 13. Kupper, C.E.; Böcker, S.; Liu, H.L.; Adamzyk, C.; van de Kamp, J.; Recker, T.; Lethaus, B.; Jahnen-Dechent, W.; Neuss, S.; Müller-Newen, G., et al. Fluorescent snap-tag galectin fusion proteins as novel tools in glycobiology. Curr. Pharm. Des. 2013, 19, 5457-5467. 14. Dumic, J.; Dabelic, S.; Flögel, M. Galectin-3: An open-ended story. BBA-Gen. Subjects 2006, 1760, 616-635. 15. Ahmad, N.; Gabius, H.J.; André, S.; Kaltner, H.; Sabesan, S.; Roy, R.; Liu, B.C.; Macaluso, F.; Brewer, C.F. Galectin-3 precipitates as a pentamer with synthetic multivalent carbohydrates and forms heterogeneous cross-linked complexes. J. Biol. Chem. 2004, 279, 10841-10847.
63
16. Lepur, A.; Salomonsson, E.; Nilsson, U.J.; Leffler, H. Ligand induced galectin-3 protein self- association. J. Biol. Chem. 2012, 287, 21751-21756. 17. Yang, R.Y.; Hill, P.N.; Hsu, D.K.; Liu, F.T. Role of the carboxyl-terminal lectin domain in self-association of galectin-3. Biochem. 1998, 37, 4086-4092. 18. Kuklinski, S.; Probstmeier, R. Homophilic binding properties of galectin-3: Involvement of the carbohydrate recognition domain. J. Neurochem. 1998, 70, 814-823. 19. Hsu, D.K.; Zuberi, R.I.; Liu, F.T. Biochemical and biophysical characterization of human recombinant ige-binding protein, an s-type animal lectin. J. Biol. Chem. 1992, 267, 14167- 14174. 20. Fermino, M.L.; Polli, C.D.; Toledo, K.A.; Liu, F.T.; Hsu, D.K.; Roque-Barreira, M.C.; Pereira-da-Silva, G.; Bernardes, E.S.; Halbwachs-Mecarelli, L. Lps-induced galectin-3 oligomerization results in enhancement of neutrophil activation. PLoS One 2011, 6, e26004. 21. Vijayakumar, S.; Peng, H.; Schwartz, G.J. Galectin-3 mediates oligomerization of secreted hensin using its carbohydrate-recognition domain. Am. J. Physiol.-Renal. 2013, 305, F90-F99. 22. Ochieng, J.; Fridman, R.; Nangiamakker, P.; Kleiner, D.E.; Liotta, L.A.; Stetlerstevenson, W.G.; Raz, A. Galectin-3 is a novel substrate for human matrix metalloproteinases-2 and -9. Biochem. 1994, 33, 14109-14114. 23. Guévremont, M.; Martel-Pelletier, J.; Boileau, C.; Liu, F.T.; Richard, M.; Fernandes, J.C.; Pelletier, J.P.; Reboul, P. Galectin-3 surface expression on human adult chondrocytes: A potential substrate for collagenase-3. Ann. Rheum. Dis. 2004, 63, 636-643. 24. Wang, Y.; Nangia-Makker, P.; Tait, L.; Balan, V.; Hogan, V.; Pienta, K.J.; Raz, A. Regulation of prostate cancer progression by galectin-3. Am. J. Pathol. 2009, 174, 1515-1523. 25. Mirandola, L.; Yu, Y.F.; Chui, K.; Jenkins, M.R.; Cobos, E.; John, C.M.; Chiriva-Internati, M. Galectin-3c inhibits tumor growth and increases the anticancer activity of bortezomib in a murine model of human multiple myeloma. PLoS One 2011, 6. 26. Ochieng, J.; Green, B.; Evans, S.; James, O.; Warfield, P. Modulation of the biological functions of galectin-3 by matrix metalloproteinases. BBA-Gen. Subjects 1998, 1379, 97-106. 27. Nangia-Makker, P.; Raz, T.; Tait, L.; Hogan, V.; Fridman, R.; Raz, A. Galectin-3 cleavage: A novel surrogate marker for matrix metalloproteinase activity in growing breast cancers. Cancer Res. 2007, 67, 11760-11768. 28. Mirandola, L.; Nguyen, D.D.; Rahman, R.L.; Grizzi, F.; Yuefei, Y.; Figueroa, J.A.; Jenkins, M.R.; Cobos, E.; Chiriva-Internati, M. Anti-galectin-3 therapy: A new chance for multiple myeloma and ovarian cancer? Int. Rev. Immunol. 2014, 33, 417-427. 29. John, C.M.; Leffler, H.; Kahl-Knutsson, B.; Svensson, I.; Jarvis, G.A. Truncated galectin-3 inhibits tumor growth and metastasis in orthotopic nude mouse model of human breast cancer. Clin. Cancer Res. 2003, 9, 2374-2383. 30. Rapoport, E.M.; Andre, S.; Kurmyshkina, O.V.; Pochechueva, T.V.; Severov, V.V.; Pazynina, G.V.; Gabius, H.J.; Bovin, N.V. Galectin-loaded cells as a platform for the profiling of lectin specificity by fluorescent neoglycoconjugates: A case study on galectins-1 and-3 and the impact of assay setting. Glycobiology 2008, 18, 315-324. 31. Salomonsson, E.; Carlsson, M.C.; Osla, V.; Hendus-Altenburger, R.; Kahl-Knutson, B.; Oberg, C.T.; Sundin, A.; Nilsson, R.; Nordberg-Karlsson, E.; Nilsson, U.J., et al. Mutational tuning of galectin-3 specificity and biological function. J. Biol. Chem. 2010, 285, 35079- 35091. 32. Song, X.Z.; Xia, B.Y.; Stowell, S.R.; Lasanajak, Y.; Smith, D.F.; Cummings, R.D. Novel fluorescent glycan microarray strategy reveals ligands for galectins. Chem. Biol. 2009, 16, 36- 47. 33. Carlsson, S.; Oberg, C.T.; Carlsson, M.C.; Sundin, A.; Niisson, U.J.; Smith, D.; Cummings, R.D.; Almkvist, J.; Karlsson, A.; Leffler, H. Affinity of galectin-8 and its carbohydrate recognition domains for ligands in solution and at the cell surface. Glycobiology 2007, 17, 663-676. 34. Stowell, S.R.; Arthur, C.M.; Mehta, P.; Slanina, K.A.; Blixt, O.; Leffler, H.; Smith, D.F.; Cummings, R.D. Galectin-1,-2, and-3 exhibit differential recognition of sialylated glycans and blood group antigens. J. Biol. Chem. 2008, 283, 10109-10123. 35. Patnaik, S.K.; Potvin, B.; Carlsson, S.; Sturm, D.; Leffler, H.; Stanley, P. Complex n-glycans are the major ligands for galectin-1, -3, and -8 on chinese hamster ovary cells. Glycobiology 2006, 16, 305-317.
64
36. de Melo, F.H.M.; Butera, D.; Medeiros, R.S.; Andrade, L.N.D.; Nonogaki, S.; Soares, F.A.; Alvarez, R.A.; Moura da Silva, A.M.; Chammas, R. Biological applications of a chimeric probe for the assessment of galectin-3 ligands. J. Histochem. Cytochem. 2007, 55, 1015-1026. 37. Davidson, P.J.; Li, S.Y.; Lohse, A.G.; Vandergaast, R.; Verde, E.; Pearson, A.; Patterson, R.J.; Wang, J.L.; Arnoys, E.J. Transport of galectin-3 between the nucleus and cytoplasm. I. Conditions and signals for nuclear import. Glycobiology 2006, 16, 602-611. 38. Nakahara, S.; Oka, N.; Wang, Y.; Hogan, V.; Inohara, H.; Raz, A. Characterization of the nuclear import pathways of galectin-3. Cancer Res. 2006, 66, 9995-10006. 39. Delacour, D.; Cramm-Behrens, C.I.; Drobecq, H.; Le Bivic, A.; Naim, H.Y.; Jacob, R. Requirement for galectin-3 in apical protein sorting. Curr. Biol. 2006, 16, 408-414. 40. Hussain, A.F.; Kampmeier, F.; von Felbert, V.; Merk, H.F.; Tur, M.K.; Barth, S. Snap-tag technology mediates site specific conjugation of antibody fragments with a photosensitizer and improves target specific phototoxicity in tumor cells. Bioconjug. Chem. 2011, 22, 2487- 2495. 41. Engin, S.; Trouillet, V.; Franz, C.M.; Welle, A.; Bruns, M.; Wedlich, D. Benzylguanine thiol self-assembled mono layers for the immobilization of snap-tag proteins on microcontact- printed surface structures. Langmuir 2010, 26, 6097-6101. 42. Kindermann, M.; George, N.; Johnsson, N.; Johnsson, K. Covalent and selective immobilization of fusion proteins. J. Am. Chem. Soc. 2003, 125, 7810-7811. 43. Recker, T.; Haamann, D.; Schmitt, A.; Kuster, A.; Klee, D.; Barth, S.; Müller-Newen, G. Directed covalent immobilization of fluorescently labeled cytokines. Bioconjug. Chem. 2011, 22, 1210-1220. 44. Gronemeyer, T.; Godin, G.; Johnsson, K. Adding value to fusion proteins through covalent labelling. Curr. Opin. Biotechnol. 2005, 16, 453-458. 45. Juillerat, A.; Heinis, C.; Sielaff, I.; Barnikow, J.; Jaccard, H.; Kunz, B.; Terskikh, A.; Johnsson, K. Engineering substrate specificity of o6-alkylguanine-DNA alkyltransferase for specific protein labeling in living cells. ChemBioChem 2005, 6, 1263-1269. 46. Keppler, A.; Pick, H.; Arrivoli, C.; Vogel, H.; Johnsson, K. Labeling of fusion proteins with synthetic fluorophores in live cells. Proc. Natl. Acad. Sci. U. S. A. 2004, 101, 9955-9959. 47. Lukinavicius, G.; Reymond, L.; Johnsson, K. Fluorescent labeling of snap-tagged proteins in cells. Methods Mol. Biol. 2015, 1266, 107-118. 48. Rech, C.; Rosencrantz, R.R.; Křenek, K.; Pelantová, H.; Bojarová, P.; Römer, C.E.; Hanisch, F.-G.; Křen, V.; Elling, L. Combinatorial one-pot synthesis of poly-n-acetyllactosamine oligosaccharides with leloir-glycosyltransferases. Adv. Synth. Catal. 2011, 353, 2492-2500. 49. Sauerzapfe, B.; Krenek, K.; Schmiedel, J.; Wakarchuk, W.W.; Pelantova, H.; Kren, V.; Elling, L. Chemo-enzymatic synthesis of poly-n-acetyllactosamine (poly-lacnac) structures and their characterization for cgl2-galectin-mediated binding of ecm glycoproteins to biomaterial surfaces. Glycoconj. J. 2009, 26, 141-159. 50. Shekhar, M.P.V.; Nangia-Makker, P.; Tait, L.; Miller, F.; Raz, A. Alterations in galectin-3 expression and distribution correlate with breast cancer progression - functional analysis of galectin-3 in breast epithelial-endothefial interactions. Am. J. Pathol. 2004, 165, 1931-1941. 51. Yamaoka, A.; Kuwabara, I.; Frigeri, L.G.; Liu, F.T. A human lectin, galectin-3 (epsilon bp/mac-2), stimulates superoxide production by neutrophils. J. Immunol. 1995, 154, 3479- 3487. 52. Morris, S.; Ahmad, N.; Andre, S.; Kaltner, H.; Gabius, H.J.; Brenowitz, M.; Brewer, F. Quaternary solution structures of galectins-1, -3, and -7. Glycobiology 2004, 14, 293-300. 53. Massa, S.M.; Cooper, D.N.W.; Leffler, H.; Barondes, S.H. L-29, an endogenous lectin, binds to glycoconjugate ligands with positive cooperativity. Biochem. 1993, 32, 260-267. 54. Ochieng, J.; Platt, D.; Tait, L.; Hogan, V.; Raz, T.; Carmi, P.; Raz, A. Structure-function relationship of a recombinant human galactoside-binding protein. Biochem. 1993, 32, 4455- 4460. 55. Arumugham, R.G.; Hsieh, T.C.Y.; Tanzer, M.L.; Laine, R.A. Structures of the asparagine- linked sugar chains of laminin. BBA Gen. Subj. 1986, 883, 112-126. 56. Knibbs, R.N.; Perini, F.; Goldstein, I.J. Structure of the major concanavalin a reactive oligosaccharides of the extracellular matrix component laminin. Biochem. 1989, 28, 6379- 6392.
65
57. Ido, H.; Harada, K.; Futaki, S.; Hayashi, Y.; Nishiuchi, R.; Natsuka, Y.; Li, S.; Wada, Y.; Combs, A.C.; Ervasti, J.M., et al. Molecular dissection of the alpha-dystroglycan- and integrin-binding sites within the globular domain of human laminin-10. J. Biol. Chem. 2004, 279, 10946-10954. 58. Martin, P.T. Dystroglycan glycosylation and its role in matrix binding in skeletal muscle. Glycobiology 2003, 13, 55r-66r. 59. Sato, S.; Hughes, R.C. Binding specificity of a baby hamster kidney lectin for h type i and ii chains, polylactosamine glycans, and appropriately glycosylated forms of laminin and fibronectin. J. Biol. Chem. 1992, 267, 6983-6990. 60. Berbís, M.A.; André, S.; Canada, F.J.; Pipkorn, R.; Ippel, H.; Mayo, K.H.; Kubler, D.; Gabius, H.J.; Jimenez-Barbero, J. Peptides derived from human galectin-3 n-terminal tail interact with its carbohydrate recognition domain in a phosphorylation-dependent manner. Biochem. Biophys. Res. Commun. 2014, 443, 126-131. 61. von Mach, T.; Carlsson, M.C.; Straube, T.; Nilsson, U.; Leffler, H.; Jacob, R. Ligand binding and complex formation of galectin-3 is modulated by ph variations. Biochem. J. 2014, 457, 107-115. 62. Saraboji, K.; Hakansson, M.; Genheden, S.; Diehl, C.; Qvist, J.; Weininger, U.; Nilsson, U.J.; Leffler, H.; Ryde, U.; Akke, M., et al. The carbohydrate-binding site in galectin-3 is preorganized to recognize a sugarlike framework of oxygens: Ultra-high-resolution structures and water dynamics. Biochem. 2012, 51, 296-306. 63. Barboni, E.A.M.; Bawumia, S.; Henrick, K.; Hughes, R.C. Molecular modeling and mutagenesis studies of the n-terminal domains of galectin-3: Evidence for participation with the c-terminal carbohydrate recognition domain in oligosaccharide binding. Glycobiology 2000, 10, 1201-1208. 64. Gaudin, J.C.; Mehul, B.; Hughes, R.C. Nuclear localisation of wild type and mutant galectin-3 in transfected cells. Biol. Cell 2000, 92, 49-58. 65. Maljaars, C.E.P.; André, S.; Halkes, K.M.; Gabius, H.J.; Kamerling, J.P. Assessing the inhibitory potency of galectin ligands identified from combinatorial (glyco)peptide libraries using surface plasmon resonance spectroscopy. Anal. Biochem. 2008, 378, 190-196. 66. Campbell, K.; Huet, A.C.; Charlier, C.; Higgins, C.; Delahaut, P.; Elliott, C.T. Comparison of elisa and spr biosensor technology for the detection of paralytic shellfish poisoning toxins. J. Chromatogr. B Analyt. Technol. Biomed. Life Sci. 2009, 877, 4079-4089. 67. Hu, D.; Fry, S.R.; Huang, J.X.; Ding, X.; Qiu, L.; Pan, Y.; Chen, Y.; Jin, J.; McElnea, C.; Buechler, J., et al. Comparison of surface plasmon resonance, resonant waveguide grating biosensing and enzyme linked immunosorbent assay (elisa) in the evaluation of a dengue virus immunoassay. Biosensors 2013, 3, 297-311. 68. Yegorova, S.; Chavaroche, A.E.; Rodriguez, M.C.; Minond, D.; Cudic, M. Development of an alphascreen assay for discovery of inhibitors of low-affinity glycan-lectin interactions. Anal. Biochem. 2013, 439, 123-131. 69. Ahmad, N.; Gabius, H.J.; Sabesan, S.; Oscarson, S.; Brewer, C.F. Thermodynamic binding studies of bivalent oligosaccharides to galectin-1, galectin-3, and the carbohydrate recognition domain of galectin-3. Glycobiology 2004, 14, 817-825. 70. Dam, T.K.; Gabius, H.-J.; André, S.; Kaltner, H.; Lensch, M.; Brewer, C.F. Galectins bind to the multivalent glycoprotein asialofetuin with enhanced affinities and a gradient of decreasing binding constants. Biochem. 2005, 44, 12564-12571. 71. Jarvis, G.A.; Mirandola, L.; Yuefei, Y.; Cobos, E.; Chiriva-Internati, M.; John, C.M. Galectin- 3c: Human lectin for treatment of cancer. In Galectins and disease implications for targeted therapeutics, Klyosov, A.A.; Traber, P.G., Eds. American Chemical Society: 2012; Vol. 1115, pp 195-232. 72. Mirandola, L.; Yu, Y.; Cannon, M.J.; Jenkins, M.R.; Rahman, R.L.; Nguyen, D.D.; Grizzi, F.; Cobos, E.; Figueroa, J.A.; Chiriva-Internati, M. Galectin-3 inhibition suppresses drug resistance, motility, invasion and angiogenic potential in ovarian cancer. Gynecol. Oncol. 2014, 135, 573-579. 73. Nangia-Makker, P.; Balan, V.; Raz, A. Regulation of tumor progression by extracellular galectin-3. Cancer Microenviron. 2008, 1, 43-51.
66
3. Neo-glycoproteins as novel ligands for human galectin-3 – Albumin as carrier for multivalent presentation of non- biotinylated and 6-biotinylated tetrasaccharides to gain high- affinity ligands Parts of this chapter were published in: S. Böcker, D. Laaf, L. Elling, Biomolecules 2015, 5, 1671-1696. S. Böcker, L. Elling, Bioengineering 2017, 4(2), 31. Abstract Galectin inhibitor design is an emerging research field due to the involvement of galectins in cancer. Galectin-3, in particular, plays an important role in tumor progression. As carbohydrate-lectin interactions are relatively weak, multivalent glycan ligands are designed to enhance binding affinity and inhibitory potency. In the present study, we generated novel multivalent neo-glycoproteins based on albumin as scaffold. We synthesized N- acetyllactosamine and N,N-diacetyllactosamine based tetrasaccharides by multi-step chemo- enzymatic synthesis utilizing recombinant glycosyltransferases The tetrasaccharides were conjugated to lysine groups of bovine serum albumin via squaric acid diethyl ester yielded different neo-glycoproteins with tuned ligand density. Binding strength of human galectin-3 is closely related to modification density and shows enhancement by multivalent ligand presentation. At galectin-3 concentrations comparable to serum levels of cancer patients, we detect the highest avidities. Multivalent ligand presentation of neo-glycoproteins also significantly increased the inhibitory potency towards galectin-3 binding to asialofetuin when compared to free monovalent glycans. Besides multivalency, modifications of the glycan structure can be introduced to find novel high-affinity ligands. Conjugation of hydrophobic compounds to saccharides has proved to be promising due to increased binding of galectin-3. Therefore, we select neo-glycans modified with biotin. We modified the tetrasaccharides at the C6-position of the terminal saccharide unit using selective enzymatic oxidation and subsequent chemical conjugation of biotinamidohexanoic acid hydrazide. These neo-glycans were much better bound by galectin-3 than the unmodified counterparts. High selectivity for galectin-3 over galectin-1 was also proven. The modified glycans were also conjugated to albumin. Compared to non-biotinylated neo-glycoproteins, biotinylated structures achieved high binding levels of galectin-3 with fewer amount of conjugated neo-glycans. The novel neo-glycoproteins may therefore serve as selective and strong galectin-3 ligands in cancer related biomedical research. The positive impact of 6-biotinylation of tetrasaccharides on galectin-3 binding broadens the recent design approaches for producing high-affinity ligands.
67
3.1 Introduction Carbohydrates play an important role in many biological processes as they act as recognition motifs triggering cellular signals and signal cascades [1-8]. Cell-surface bound glycans mediate the interactions and communication between cells and their microenvironment [9-12]. Specialized proteins called lectins mediate the recognition and decoding of the glycan code presented by the cells [13]. Although the binding affinity of lectins towards their glycan ligand is rather low it may be dramatically increased by multivalent presentation of glycans in close proximity to each other as in so-called glyco clusters, like nature does it on glycolipids or glycoproteins [14-18]. These scaffolds may bear multiple glycosylation sites or branched glycans. Moreover, arrangement in lipid rafts increases the multivalency as well [19].
Lectins play various roles in disease associated processes, for example virus or toxin binding, mediation of tumor angiogenesis, tumor progression, cell cycle arrest or metastasis [20-22]. The most ancient group of lectins are the β-galactoside binding galectins, formerly described as S-type lectins, as it was assumed that sulfhydryl groups are crucial for binding activity [23]. So far, at least 15 different human galectins are known. Galectins mediate most of their functions by galectin–glycan interactions [4,24,25]. They play an important role in a variety of biological processes including cancer progression and immune response [26-32]. The only member of the chimera-type galectin family is galectins-3 (Gal-3), which acts cell-type specifically pro- or anti-apoptotic [33,34]. Moreover, it is reported that Gal-3 is strongly upregulated in many tumor cells extra- and intracellularly [28,35-37]. The role of Gal-3 has been investigated extensively and found to be crucial in tumor progression, metastasis and angiogenesis [32,38-40]. This makes Gal-3 a potential target for anti-cancer therapy and cancer diagnosis. [41-43]. Recent reports suggest multiple glycan epitopes and low molecular weight molecules as potential inhibitor molecules [44-48]. The structural features of Gal-3 are an N-terminal non-lectin domain and a highly conserved C-terminal carbohydrate recognition domain (CRD) [25]. It is reported that in the presence of natural ligand environment Gal-3 forms pentamers or even higher oligomers upon binding [49-52]. Therefore, inhibitors with multivalent presentation are prone to show the highest potency [53-56]. Multivalent carbohydrate presentation is mostly achieved by chemical modification of different scaffolds that define arrangement and orientation of the saccharides [57-68]. The reason for binding enhancement is discussed as chelate cooperativity [69,70], which describes interplay of binding sites for enhanced binding to their closely neighbored ligands presented in a multivalent system. These effects are found in many carbohydrate and non-carbohydrate related systems [71-75].
68
However, a crucial point for usage of glycan-based inhibitors is selectivity as all galectins may recognize β-galactoside containing structures. Recently, we confirmed that N,N- diacetyllactosamine (LacdiNAc, GalNAcβ1,4GlcNAc) epitope acts as a selective ligand for Gal-3 compared to galectin-1 (Gal-1) [76]. This was surprising as LacdiNAc has a rather low abundancy in mammalian cells [77-79], but is overexpressed in parasites [80,81]. Nevertheless, Gal-3 was found to bind to LacDiNAc and mediate immune recognition [80]. In human cancer progression, cancer type dependent up- or downregulation of LacdiNAc expression was observed [82-85]. Additionally, it was shown that the LacdiNAc epitope is present on the gastric mucin and recognized by an adhesin from Helicobacter pylori [78,86] which may play a role for bacterial colonization in the gastric mucosa.
In our previous studies, the inhibitory effect of divalent LacdiNAc for Gal-3 binding to asialofetuin glycoprotein, a model ligand for galectins, was only weakly dependent on the linker length [76]. However, a multivalent effect was not observed. Additionally, we found that the poly-LacNAc ([3Galβ1,4GlcNAcβ1]n = 2-4) oligomers are preferentially bound by Gal- 3 [87]. We concluded that multivalent presentation of a tetrasaccharide glycan structure with a LacdiNAc epitope on a suitable scaffold should lead to increased inhibition of Gal-3 binding to glycoproteins.
To enable a multivalent presentation of various glycan ligands in a natural way, we use here serum albumin as scaffold for designing galectin ligands. Albumin is a highly abundant non- glycosylated serum protein and the protein of choice because it contains multiple lysine residues which are easily addressable for chemical conjugation methods [88-90]. Neo- glycoproteins are in general synthesized by chemical conjugation of glycans to native proteins [91,92]. They can be specifically designed for lectin binding studies by attachment of various types of glycan epitopes and by their glycosylation density [93-95].
Here, we developed controlled glycosylation of bovine serum albumin (BSA) to yield neo- glycoproteins with tunable multivalency. First, the glycans carrying an amino terminated linker at their reducing end were synthesized by multi-step chemo-enzymatic synthesis as previously reported [76,96]. Chemical conjugation to lysine residues of BSA was accomplished by homobifunctional amino-reactive linker squaric acid diethyl ester enabling crosslinking via primary amino groups [97-104]. Variation of the molar ratios of glycan with respect to lysine residues resulted in the synthesis of a set of neo-glycoproteins with different degrees of glycan modification. These scaffolds showed very high affinity towards galectins and could confirm the selectivity of LacdiNAc towards Gal-3.
69
In the second part, the search of novel glycan-based inhibitors of galectins led to the synthesis of neo-glycoconjugates based on N-acetyllactosamine (LacNAc) [105] and biotin at the C6- position of terminal galactose. Various approaches to design galectin inhibitors suggest the incorporation of hydrophobic residues into natural ligands [44,45,106,107]. However, these syntheses suffer often from complicated chemical routes, which are time-consuming and give rather low yields. Our reported biotinylated neo-glycoconjugates combine a high affinity leading structure with a hydrophobic modification at the non-reducing end, synthesized by a chemo-enzymatic reaction [105]. The carrier tetrasaccharides were synthesized by a cascade reaction with three recombinant glycosyltransferases and modified at the C6-position via enzymatic oxidation and reductive amination. Subsequently, the neo-glycans were likewise conjugated to BSA via squaric acid diethyl ester coupling with various ligand densities.
We here report the evaluation of non-biotinylated and 6-biotinylated LacNAc and LacdiNAc terminated tetrasaccharides as Gal-3 inhibitors. The binding affinity of Gal-3 towards neo- glycoproteins and free neo-glycans was investigated using solid phase ELISA-type assays and surface plasmon resonance spectroscopy (SPR). In this way, we could identify the impact of biotinylation at C6-position with soluble ligands and immobilized ligands on proteins. The results indicate that the biotinylated glycans are better bound by Gal-3 when they are immobilized. This encourages not only the assumption that multimerization of Gal-3 takes place in a multivalent ligand environment, but also that this effect may be amplified by modification of the glycan. Therefore, this study leads to new insight into the behavior of Gal- 3 on multivalent scaffolds and to alternative approaches for designing potent inhibitors.
3.2 Materials and methods 3.2.1 Production of recombinant enzymes Enzymes were expressed and purified as described previously [76,96,108,109]. Briefly, the fusion protein of human β1,4-galactosyltransferase-1 (His6-Propeptide-catβ4GalT-1, β4GalT) and the Y284L mutant (β4GalTY284L) were expressed in E. coli Shuffle T7 Express (DE3) (NEB, Frankfurt/Main, Germany). The β1,3-N-acetylglucosaminyltransferase from Helicobacter pylori (β3GlcNAcT) and UDP-Glc 4′-epimerase from Campylobacter jejuni, both fused to maltose binding protein, were expressed in E. coli BL21 (DE3) (Novagen/Merck, Darmstadt, Germany). TB medium with 0.5 mM IPTG was used for induction of expression. After 24 h, cells were harvested by centrifugation and appropriate cell mass was sonicated two times for 30 s for follow-up affinity chromatography. Recombinant MBP-tagged proteins were purified by amylose affinity chromatography using TM MBPTrap HP 5 mL column (GE Healthcare, Munich, Germany) and His6-tagged proteins
70 by immobilized metal-ion affinity chromatography (IMAC) using HisTrapTM HP 5 mL column (GE Healthcare). Purification was carried out according to manufacturer’s instructions. After column equilibration bacterial crude extracts were loaded onto the column at a flow rate of 1 mL/min and washed until baseline signal was reached. Target proteins were eluted by application of elution buffer containing 10 mM maltose or 500 mM imidazole. The buffer for β4GalTY284L was subsequently changed to 0.1 M MOPS buffer (pH 6.8) including 20% (v/v) glycerol.
3.2.2 Chemo-enzymatic synthesis of glycans (4 and 5)
Compound 1 (GlcNAc-linker-NH2-tBoc) was synthesized as described [109] and served as starting material for synthesis of 4 (Galβ1,4GlcNAcβ1,3Galβ1,4GlcNAcβ1-linker-NH2-tBoc,
LacNAc-LacNAc-linker-NH2-tBoc) and 5 (GalNAcβ1,4GlcNAcβ1,3Galβ1,4GlcNAcβ1- linker-NH2-tBoc, LacdiNAc-LacNAc-linker-NH2-tBoc). 1 was elongated sequentially using
β4GalT adding galactose resulting in 2 (Galβ1,4GlcNAcβ1-linker-NH2-tBoc) [96] and
β3GlcNAcT producing 3 (GlcNAcβ1,3Galβ1,4GlcNAcβ1-linker-NH2-tBoc) [109]. Either galactose using β4GalT or GalNAc using β4GalTY284L with in-situ production of UDP- GalNAc from UDP-GlcNAc using UDP-Glc 4'-epimerase were added to 3 yielding 4 and 5, respectively. The enzymatic synthesis of compound 4 was carried out as described for compound 2. To synthesize 5 compound 3 (5 mM) was mixed with 15 mM UDP-GlcNAc (Roche, Mannheim, Germany) in reaction buffer (100 mM MOPS-NaOH, pH 6, with 25 mM
KCl, 2 mM MnCl2, 10 mU/mL alkaline phosphatase), 75 mU/mL β4GalTY284L and 7 U/mL U UDP-Glc 4'-epimerase. Reaction was monitored via HPLC (LiChrospher 100 RP 18-5µ (Merck), 15% (v/v) acetonitrile, 0.1% (v/v) formic acid or Multospher APS-HP 5µ (CS- Chromatographie, Langerwehe, Germany), 5 mM ammonium acetate (pH 4.3), 80% (v/v) acetonitrile). To stop the reaction, enzymes were removed via filtration (VivaSpin®20, Sartorius, Göttingen, Germany) and products were purified by solid phase extraction with Sep-Pak® Vac C18 1cc columns (Waters GmbH, Eschborn, Germany) [109] followed by mass analysis by LC-ESI-MS.
3.2.3 Synthesis of biotinylated glycans (9 and 10)
The tetrasaccharides LacNAc-LacNAc-linker-NH2-tBoc (4) and LacdiNAc-LacNAc-linker-
NH2-tBoc (5) were biotinylated at C6-position of the terminal galactose and GalNAc, respectively, after oxidation using galactose oxidase. The overall procedure is described previously [105] with some synthesis optimizations. For oxidation of 15 µmol oligosaccharide
71
48 U galactose oxidase (Worthington, Lakewood, USA) and 480 U peroxidase (Merck, Darmstadt, Germany) were applied in 25 mM MES-NaOH (pH 6.0) saturated with oxygen and containing 5 mM CuCl2. The reaction took place over night under 1 bar oxygen pressure at room temperature. The reaction was stopped by ultrafiltration (10 kDa cut off, VivaSpin®500, Sartorius, Göttingen, Germany) as it is known that α,β-unsaturated aldehydes are formed by heat [105]. Analysis of oxidation reactions were performed by HPLC (LiChrospher 100 RP 18-5μ, 15% (v/v) acetonitrile, 0.1% (v/v) formic acid). The 6-aldehydes 6 and 7 were biotinylated with biotinamidohexanoic acid hydrazide (BACH, 8, Sigma- Aldrich, Taufkirchen, Germany). Compounds 6 and 7 were incubated with 3.5 eq. of 8 for two days followed by the addition of 10-15 eq. of NaBH3CN (Sigma-Aldrich). Reaction mixture was stirred until no further reduction of biotinylated LacNAc-LacNAc (9) and LacdiNAc-LacNAc (10) occurred. The progress was monitored via HPLC (Multokrom 100-5 C18, 250 × 4 mm (CS-Chromatographie), gradient separation using 11-50% (v/v) acetonitrile over a time course of 50 min at a flow rate of 0.5 mL/min, detection at 254 nm). 9 and 10 were purified by preparative HPLC (Eurospher 100-10 C18, 250 x 20 mm, Knauer, Berlin, Germany) using the same method and analyzed by LC-ESI-MS.
3.2.4 Synthesis of squaric acid monoamide esters of non-biotinylated glycans (16 and 17) Linker deprotection of 4 and 5 was carried out under acidic conditions (1 M HCl) resulting in 11 and 12 followed by neutralization using Dowex® 66 free base (Sigma-Aldrich, Taufkirchen, Germany). Compounds 11 and 12 were dissolved in 50% aqueous ethanol (35 mM HEPES, pH 7.0) containing 4 equivalents of squaric acid diethyl ester (15, Sigma- Aldrich) and of triethylamine [101]. Reaction mixture was incubated at room temperature overnight while gently shaking. When thin-layer chromatography (Alugram® Xtra SIL G/UV254, Macherey-Nagel, Düren, Germany, n-butanol/ethanol/water/15% aqueous ammonia, 5:10:8:4) showed disappearance of oligosaccharide starting material, the solvent was evaporated and the products, squaric acid monoamide esters of LacNAc-LacNAc (16) and LacdiNAc-LacNAc (17), respectively, were purified by preparative HPLC (Eurospher 100-10 C18, 15% (v/v) acetonitrile). The correct masses were proven by LC-ESI-MS and concentration was determined via HPLC (LiChrospher 100 RP 18-5µ, 15% (v/v) acetonitrile, 0.1% (v/v) formic acid) using calibration curve of GlcNAc conjugated to 15.
72
3.2.5 Synthesis of squaric acid monoamide esters of biotinylated glycans (18 and 19) Deprotection under acidic conditions (1 M HCl) of 9 and 10 yielded 13 and 14. The present amino group was used to couple squaric acid diethyl ester (15) to the tetrasaccharides as described above with optimized conditions for this application. Glycans 13 and 14, respectively, were incubated with 4-fold molar excess of 15 and triethylamine in 50% aqueous ethanol including 20 mM HEPES (pH 8.0) for one hour at 4 °C and 450 rpm. The products, squaric acid monoamide esters of 6-biotin LacNAc-LacNAc (18) and 6-biotin LacdiNAc-LacNAc (19), were directly purified by preparative HPLC using the gradient separation method described above. Mass analysis of the products was done by LC-ESI-MS and concentration determination by HPLC and absorbance measurement at 254 nm with calibration for GlcNAc conjugated to 15.
3.2.6 Conjugation of glycans to bovine serum albumin Prior to BSA modification, active charcoal delipidation was carried out for removal of lipid contaminations as described [110], because commercial preparations of BSA may contain fatty compounds. BSA (Carl Roth, Karlsruhe, Germany) and appropriate amount of squaric acid monoamide esters 16, 17, 18 or 19 were mixed in sodium tetraborate buffer (40-50 mM, pH 9.0) and incubated at room temperature and 450 rpm for up to ten days. Compounds 16 and 17 were applied in varying molar ratios compared to lysine residues of BSA at one time point (from 0.025:1 to 1.5:1). In contrast, for biotinylated glycans fed-batch mode was chosen beginning with 1.5 molar excess compared to lysine residues of BSA to reduce the consumption of 18 and 19. To gain the desired modification degrees an increasing amount of 18 and 19, respectively, were added at different time points (end molar ratios from 0.03:1 to 1.2:1). Purification of the products 20a-k and 21a-k (non-biotinylated neo-glycoproteins) as well as 22a-f and 23a-f (biotinylated neo-glycoproteins) was performed by ultrafiltration to remove unbound glycans and protein concentrations were determined by Bradford assay (Roti®-Quant, Carl Roth) calibrated with BSA.
3.2.7 TNBSA-assay Determination of free lysine residues (ε-amino groups) in non-biotinylated neo-glycoproteins (20a-k and 21a-k) was carried out using 2,4,6-trinitrobenzene sulfonic acid (TNBSA, Sigma- Aldrich) as detecting agent. The TNBSA assay was performed as described [111] but optimized for our application. Protein samples were diluted in 50 mM sodium tetraborate buffer (pH 9.0) to yield a concentration of 12.5 µM and 50 µL were mixed with equal volume
73 of 7.5 mM TNBSA. After incubation at room temperature for 15 min, absorbance at 420 nm was measured in microplate reader (SPECTRAmax Plus, Molecular Devices, Ismaning, Germany). Standard curves were generated using lysine hydrochloride and crude BSA as standard. Protein amounts were measured by Bradford assay using calibration with unmodified BSA.
3.2.8 SDS-PAGE and streptavidin blot Molecular weights of the neo-glycoproteins were checked by SDS-PAGE followed by streptavidin blot for biotinylated products. Here, protein amounts of 1 µg for Coomassie staining and 0.5 µg for blotting were applied using 8% gels and constant current of 25 mA. The proteins transferred to the PVDF-membrane were detected by incubation with peroxidase conjugated streptavidin (Roche, Mannheim, Germany).
3.2.9 Expression and purification of recombinant galectins
Human His6Gal-1C2S and human His6Gal-3 were expressed as described previously [87,112]. For the expression of the proteins E. coli Rosetta (DE3) pLysS (Novagen/Merck, Darmstadt, Germany) was used.
Disruption of cultivated cells and followed galectins purification by immobilized metal affinity chromatography (IMAC, Ni2+-NTA) was performed as described above. Galectins were stored in phosphate buffered saline containing 2 mM EDTA (EPBS, pH 7.5) at 4°C, including 10% (v/v) glycerol for His6Gal-1C2S. Protein concentrations were measured by Bradford assay using BSA for calibration.
3.2.10 Galectin binding assays on immobilized glycans and neo-glycoproteins The synthesized glycans and neo-glycoproteins were analyzed for binding of Gal-3 and Gal-1 in 96 well microtiter plate formats [87]. LacNAc-LacNAc-linker-NH2 11 and LacdiNAc-
LacNAc-linker-NH2 12 (deprotected forms of 4 and 5) as well as the corresponding biotinylated glycans 18 and 19 were immobilized via the amino group in aminoreactive microtiter plates (Immobilizer Amino, Nunc, Wiesbaden, Germany). The immobilization of 5 nmol glycan in sodium carbonate buffer (100 mM, pH 9.6) was done overnight. For immobilizing neo-glycoproteins, 5 pmol protein were incubated in PBS (pH 7.5) in MaxiSorp microtiter plates (Nunc) overnight. Wells were washed three times with PBS-Tween (0.05% (v/v)) and residual binding sites were blocked with 2% BSA in PBS followed by incubation for one hour with galectins diluted in EPBS. Anti-His6-peroxidase (Roche, 1:2000 in PBS)
74 was used to detect bound galectin that subsequently converted OPD substrate (o- phenylenediamine, Dako, Hamburg, Germany) with read-out at 492 nm. Measured data were analyzed using Sigma Plot (Systat software GmbH, Erkrath, Germany).
3.2.11 Inhibition of galectin binding with neo-glycoproteins Neo-glycoproteins 20j, 20k, 20g; 21j, 21k, 21g; 22b, 22d, 22e and 23b, 23d, 23e as well as glycans 4, 5, 9 and 10 were used as inhibitors of Gal-3 to asialofetuin (ASF) binding in competitive inhibition assays performed as described [87]. ASF (Sigma-Aldrich, 5 pmol in PBS) as standard glycoprotein was immobilized overnight in microtiter plates (MaxiSorp, Nunc). After blocking as described before, varying concentration of inhibitor and 5 µM Gal-3 were simultaneously incubated for one hour in PBS. Wells were washed and residual bound galectin was detected as described above. Evaluation of the measured data was done using Sigma Plot. All assays were reproduced in at least three independent measurements.
3.2.12 Surface plasmon resonance spectroscopy SPR spectroscopy was done with Reichert SR7500DC System (XanTec bioanalytics, Düsseldorf, Germany) and NTA derivatized carboxymethyldextran hydrogel sensor chips (200M, XanTec bioanalytics). These surfaces can be complexed with Ni2+ ions and allow reversible immobilization of His6-tagged proteins. Human His6Gal-3 (5 pmol in EPBS) was immobilized on the sample channel with a flowrate of 10 µL/min after 5 mM NiCl2 solution was injected. The reference channel stayed untreated.
The binding experiments were carried out with a flow rate of 20 µL/min by injecting neo- glycoproteins 20a,j,e,k,g,i; 21a,j,e,k,g,i; 22a-f and 23a-f (2 pmol in PBS) as well as BSA and ASF as control. The dissociation time was three minutes. Between the measurements the surface was regenerated with 0.5 M Na EDTA (pH 8.5) to remove Ni2+ together with Gal-3.
Each cycle started with NiCl2 followed by Gal-3 application.
The measured data were subtracted by reference values and analyzed using Integrated SPRAutolink (Reichert technologies, Depew, USA) and Scrubber2 (BioLogic Software, Campbell, Australia).
75
3.3 Results and discussion The synthesis of novel glycans and neo-glycoproteins took place with the background of designing new ligands for galectins, especially Gal-3. Here, we present neo-glycoproteins with varying glycosylation density based on BSA and their application in galectin binding studies. For our purposes, the oligomers LacNAc-LacNAc and LacdiNAc-LacNAc were synthesized de novo chemo-enzymatically. In addition, we extended the saccharide part by hydrophobic biotin coupling. Decoration of BSA was accomplished by a two-step conjugation reaction using squaric acid diethyl ester as a linker. Irrespective of the accessibility, BSA can be decorated with up to 60 glycans per molecule due to the presence of 60 lysine residues. The synthesized neo-glycoproteins are tested as ligands for human Gal-3 and Gal-1. Moreover, the influence of conjugated biotin on galectin binding to tetrasaccharides was investigated.
3.3.1 Chemo-enzymatic synthesis of LacNAc-LacNAc and LacdiNAc-LacNAc Glycosyltransferases and activated nucleotide sugars as donor substrate were applied in a consecutive synthesis for attachment of monosaccharide residues to GlcNAc-linker-tBoc (1) (Scheme 3.1). By sequential elongation of 1 using recombinantly expressed and purified β4GalT [108] and β3GlcNAcT [96,113], compounds 2 and 3 (both 100% conversion) were obtained. Compound 3 was further elongated using either β4GalT or β4GalTY284L [76,105] for the synthesis of Gal- or GalNAc-terminated tetrasaccharides 4 and 5, respectively. In contrast to our previously published method [76], a three-fold molar excess of UDP-GalNAc and appropriate amount of β4GalTY284L (75 mU/mL) turned out to be crucial to obtaining a quantitative yield for 5. After purification yields of 4 (90.6%) and 5 (98.3%) were determined via HPLC analysis. Integrity and purity of 4 and 5 were confirmed via LC-MS (Figure S3.1A and B). For conjugation of these glycans, deprotection of 4 and 5 carrying tert- butyloxycarbonyl (tBoc) protecting group was carried out under acidic conditions yielding 11 (85.0%) and 12 (96.1%).
In conclusion, multi-step chemo-enzymatic synthesis and deprotection provided two pure tetrasaccharides (LacNAc-LacNAc and LacdiNAc-LacNAc) in reasonable yields.
76
Scheme 3.1: Chemo-enzymatic synthesis of tetrasaccharides 4 and 5 using recombinant glycosyltransferases for complete conversions
3.3.2 Biotinylation of LacNAc-LacNAc and LacdiNAc-LacNAc The biotinylation of LacNAc oligomers is a two-step synthesis (Scheme 3.2) as published before [105]. In the first step, the terminal galactose or N-acetylgalactosamine (GalNAc) was oxidized at C6-position by galactose oxidase yielding C6-aldehydes. The second step describes the reaction with BACH (8) and subsequent reduction.
The conversion of LacNAc-LacNAc-linker-NH2-tBoc (4) and LacdiNAc-LacNAc-linker-
NH2-tBoc (5) to the corresponding C6-aldehydes 6 and 7 by galactose oxidase obtained about 80-90% product. Stopping the reaction by ultrafiltration avoided the formation to α,β- unsaturated aldehydes as unwanted side-products. Further purification was not necessary. The reaction of 8 with 6 and 7, respectively, resulted in 6-biotin LacNAc-LacNAc-linker-NH2- tBoc (9) and 6-biotin LacdiNAc-LacNAc-linker-NH2-tBoc (10). After preparative HPLC, yields of 41% (9) and 27% (10) were obtained. Product loss was probably due to product instability caused by incomplete reduction after oxidation and biotinylation. The isolated products were confirmed by LC-MS (Figure S3.1C and D). Further product characterization was previously described [105]. The tBoc-linker structures of 9 and 10 were deprotected as described above for the non-biotinylated structures to obtain amino-linker functionalized
77 products 13 and 14. MS-spectra of these compounds are shown in supporting information (Figure S3.1E and F).
Scheme 3.2: Two-step synthesis of 6-biotin tetrasaccharides 9 and 10
3.3.3 Synthesis and analysis of neo-glycoproteins Consecutive attachment of two different ligands with the same functional group is the most attractive feature of squaric acid diethyl ester (15) [101]. Thus, 15 offered ideal chemical properties for BSA decoration with our synthesized and modified tetrasaccharides LacNAc- LacNAc, LacdiNAc-LacNAc, 6-biotin LacNAc-LacNAc and 6-biotin LacdiNAc-LacNAc.
Amino group providing oligomers 11 and 12 were conjugated to 15 forming squaric acid monoamide esters 16 and 17 (Scheme 3.3). Formation of squaric acid bisamide was suppressed by constant pH 7 and the presence of triethylamine (Et3N) as external base. Complete conversion of 11 and 12 could only be achieved by applying a four-fold molar excess of 15 and equal amounts of Et3N. The purity of the products 16 and 17 had highest priority as trace impurities of 15 would lead to protein cross-linking reactions during the second conjugation step. Thus, removal of residual 15 was achieved using preparative HPLC. Additionally, LC-MS results (Figure S3.1G and H) confirmed integrity and purity of compounds 16 and 17. Molecular masses of 972.3 m/z and 1013.6 m/z, respectively, were found for 16 and for 17.Overall yields of 77.8% for 16 and 86.3% for 17 were determined after chemo-enzymatic synthesis, deprotection and conjugation to 15.
78
Since compound 15 reacts also with secondary amines [98], which are present in the biotin moiety, reactions of biotinylated tetrasaccharides 13 and 14 were optimized. It turned out that reaction at pH 8.0 and 4 °C for one hour was suitable to produce high amounts of 18 and 19 (Scheme 3.3). The products were separated from by-products by preparative HPLC and yields of 42% for 18 and 74% for 19 were obtained. Starting from tetrasaccharides 4 and 5 overall yields of approximately 15% were obtained. LC-MS results (Figure S3.1I and J) confirmed integrity and purity of compounds 18 and 19.
Scheme 3.3: Two-step neo-glycoprotein synthesis
Aqueous solution with increased pH was utilized as reaction buffer for squaric acid bisamide production, as recommended in literature [98,101,114]. In order to generate neo-glycoproteins presenting variable numbers of sugar moieties, different concentrations of 16 or 17 were applied for the coupling process (Scheme 3.3). Active charcoal delipidated BSA was dissolved in borate buffer (pH 9.0) and ratios of 16 or 17 ranging between 0.025 and 1.5 with respect to BSA lysine residues were adjusted. Thus, neo-glycoproteins 20a–k and 21a–k with variable numbers of LacNAc-LacNAc and LacdiNAc-LacNAc, respectively, were obtained after an incubation period of six days (Table 3.1).
In a modified version, a fed-batch mode, biotinylated products 18 and 19, respectively, were coupled to BSA. The glycans were added consecutively to the BSA containing reaction mixtures at several time points to overcome squaric acid monoamide ester hydrolysis at pH 9 [101,114] and to reduce consumption of the modified tetrasaccharides. Finally, different
79
molar ratios of glycans and BSA lysine residues from 0.03:1 up to 1.2:1 were applied to reach different degrees of glycosylation (22a-f and 23a-f) (Scheme 3.3, Table 3.3).
Neo-glycoproteins were purified by ultrafiltration to remove unbound glycans and analyzed by trinitrobenzene sulfonic acid (TNBSA) assay and/or reducing polyacrylamide gel electrophoresis (SDS-PAGE). TNBSA was utilized for quantification of free lysine residues of BSA. Nucleophilic character of TNBSA as nitrobenzene derivative is well known for precise labeling of primary amines. The photometric measurement of the resulting N-trinitrophenylamine at 420 nm was demonstrated as the suitable wavelength for colorimetric quantification of free lysine residues [111]. In order to minimize consumption of valuable neo-glycoproteins experimental setup of the TNBSA assay was optimized. Lysine hydrochloride and BSA were utilized for highly accurate calibration (Figure S3.2) of the TNBSA assay.
Table 3.1: Glycosylation densities and resulting molecular weights of non-biotinylated neo- glycoproteins 20a–k and 21a–k The molecular weights (MW) of BSA based neo-glycoproteins were calculated according to the number of attached LacNAc-LacNAc (20a–k) and LacdiNAc-LacNAc (21a–k) glycans using TNBSA assay.
Unmodified Modified Unmodified Modified MW MW Compound lysine residues lysine residues Compound lysine residues lysine residues [kDa] [kDa] [mol/mol BSA] [mol/mol BSA] [mol/mol BSA] [mol/mol BSA] BSA 60.0 ± 0.027 0.0 66.4 BSA 60.0 ± 0.027 0.0 66.4 20a 58.4 ± 0.016 1.6 67.9 21a 58.3 ± 0.012 1.7 68.1 20b 56.2 ± 0.005 3.8 70.0 21b 57.8 ± 0.005 2.2 68.6 20c 53.9 ± 0.013 6.1 72.2 21c 54.1 ± 0.012 5.9 72.2 20d 50.0 ± 0.040 10.0 75.9 21d 51.2 ± 0.012 8.8 75.1 20e 45.6 ± 0.015 14.4 80.0 21e 45.9 ± 0.046 14.1 80.3 20f 40.6 ± 0.027 19.4 84.7 21f 38.7 ± 0.091 21.3 87.4 20g 35.8 ± 0.079 24.2 89.3 21g 35.6 ± 0.115 24.4 90.4 20h 34.0 ± 0.009 26.0 90.9 21h 35.0 ± 0.066 25.0 91.0 20i 31.0 ± 0.130 29.0 93.8 21i 32.5 ± 0.097 27.5 93.5 20j 52.5 ± 0.012 7.5 73.5 21j 52.6 ± 0.018 7.4 73.7 20k 42.2 ± 0.084 17.8 83.2 21k 42.0 ± 0.060 18.0 84.1
Finally, the number of unmodified (free) lysine residues of the non-biotinylated neo- glycoproteins 20a–k and 21a–k was determined by the optimized TNBSA assay as shown in Table 3.1. Molecular weights of the glycoproteins were then calculated according to the number of attached LacNAc-LacNAc or LacdiNAc-LacNAc glycans. Increased ratios of 16 or 17 with respect to lysine residues of BSA led to an increasing modification density as more lysine residues were amidated. For 20a–d and 21a–d the coupling efficiencies of squaric acid monoamide esters 16 and 17 were between 73.0% and 94.4%. However, at higher molar
80 ratios coupling efficiency was only 32.2% for 16 and 30.6% for 17 yielding BSA based neo- glycoproteins 20i and 21i, respectively. Possible reasons for this observed limitation may be low accessibility of some of the lysine residues and hydrolysis of the squaric acid monoamide esters [101,114]. Neo-glycoproteins 20j,k and 21j,k share a very similar glycosylation density of 7.5 or 18 glycans per molecule, respectively, and were used for inhibition assays. In summary, different amounts of 16 or 17 were attached to BSA reaching numbers for modified lysine residues between 2 and 29 per BSA molecule. Our results are in accordance with published data, where lactose conjugated BSA with maximum 30 glycation sites was obtained at a lactose to BSA-lysine ratio of 20:1 [104]. In contrast, we reach this grade of modification already with a 1.5-fold molar excess of squaric acid monoamide esters (16 or 17) despite their approximately doubled molecular masses with respect to lactose.
Figure 3.1: SDS-PAGE and streptavidin blot analysis of neo-glycoproteins Electrophoretic mobility of BSA-based neo-glycoproteins presenting (A) LacNAc-LacNAc (20a–k), (B) LacdiNAc-LacNAc (21a–k) and (C) 6-biotin LacNAc-LacNAc (22a–f) and 6-biotin LacdiNAc- LacNAc (23a–f) is compared to unmodified BSA (lane C). The biotinylated neo-glycoproteins were also analyzed by a streptavidin blot (D). M – Protein size standard, C – unmodified BSA, 20a–k – LacNAc-LacNAc conjugated BSA, 21a–k – LacdiNAc-LacNAc conjugated BSA, 22a-f – 6-biotin LacNAc-LacNAc conjugated BSA, 23a-f – 6-biotin LacdiNAc-LacNAc conjugated BSA
81
TNBSA assay results (Table 3.1) indicated consistent increase of theoretical molecular weights of neo-glycoproteins 20a–k and 21a–k. For further evaluation, apparent molecular weights of neo-glycoproteins were determined by SDS-PAGE shown in Figure 3.1. As molecular weight of BSA (66.4 kDa) could be confirmed via SDS-PAGE, gentle active charcoal treatment for removal of lipids does obviously not influence protein integrity. Moreover, Figure 3.1 provides direct evidence for molecular weight shift of neo-glycoproteins towards higher values. As expected, the mass differences of 20a–k or 21a–k compared to unmodified control (lane C) increase with higher initial concentrations of substances 16 or 17. Molecular weights of neo-glycoproteins are evaluated. For this purpose, retardation factors
(Rf) are determined for each sample (20a–k and 21a–k). Molecular weight differences with regard to unmodified BSA were calculated and are summarized in Table 3.2. Coupling reactions of squaric acid monoamide esters of LacNAc-LacNAc 16 or LacdiNAc-LacNAc 17 lead to mass increase of 0.943 kDa or of 0.984 kDa per modified lysine residue. Results from biochemical TNBSA assay and SDS-PAGE provide direct evidence for molecular weight shift of neo-glycoproteins 20a–k and 21a–k. Both approaches are suitable analysis methods as approximately equal differences of molecular weight are identified. Most molecular weights indicated by SDS-PAGE are slightly smaller compared to those resulted from
TNBSA assay. The analysis of diffuse bands leads to certain inaccuracy of Rf values.
Table 3.2: Molecular weight differences (∆MW) of non-biotinylated neo-glycoproteins 20a–k and 21a–k compared to unmodified BSA
Values for retardation factor (Rf) were calculated based on electrophoretic mobility of samples resulting from SDS-PAGE analysis (Figure 3.1). Measured Rf values were transformed to molecular weights by linear regression (y = mx + b; y—log MW; m—slope; x—Rf; b—y-intercept). Differences between the calculated molecular weights of neo-glycoproteins and those of unmodified BSA are compared with data determined by TNBSA assay.
TNBSA SDS-PAGE TNBSA SDS-PAGE Compound Compound ΔMW [kDa] ΔMW [kDa] ΔMW [kDa] ΔMW [kDa] BSA 0.0 0.0 BSA 0.0 0.0 20a 1.5 1.3 21a 1.7 1.5 20b 3.6 3.0 21b 2.2 2.5 20c 5.8 5.0 21c 5.8 4.6 20d 9.4 8.2 21d 8.7 7.3 20e 13.6 12.1 21e 13.9 11.7 20f 18.3 17.4 21f 21.0 19.0 20g 22.8 22.4 21g 24.0 22.2 20h 24.5 24.3 21h 24.6 25.5 20i 27.3 27.6 21i 27.1 28.2 20j 7.1 7.0 21j 7.3 7.1 20k 16.8 15.0 21k 17.7 15.6
82
Neo-glycoproteins and do not appear as distinct bands while smearing effect increases with size. Smearing appearance of glycosylated BSA can be due to the statistical character of the coupling reaction and thus presence of BSA molecules with different degrees of modification. Moreover, hydrophilic glycans are known to interact irregularly with detergents like SDS [115]. Consequently, detergent induced development of linearized proteins is prevented proportionally to the amount of glycan moieties.
The achieved coupling degrees of the neo-glycoproteins presenting biotinylated glycans (22a- f and 23a-f) were estimated only by molecular mass shifts in SDS-PAGE. TNBSA assay produces incorrect results for biotinylated glycans due to high hydrophobicity and bulkiness, which leads to overestimation of the coupling degree. This phenomenon was interestingly also found in other published reports on modification of BSA via its lysine residues (Torres et al. 2014; Adamczyk et al. 1994; Bullock et al. 1997).
Table 3.3: Glycosylation densities of biotinylated neo-glycoproteins 22a-f and 23a-f The differences in molecular weights (ΔMW) compared to BSA determined by SDS-PAGE were used to calculate the number of attached 6-biotin LacNAc-LacNAc (22a-f) and 6-biotin LacdiNAc-LacNAc (23a-f). Consequential molecular weights (MW) are given as well.
Modified Modified MW ΔMW MW ΔMW Compound lysine residues Compound lysine residues [kDa] [kDa] [kDa] [kDa] [mol/mol BSA] [mol/mol BSA] BSA 66.4 0.0 0.0 BSA 66.4 0.0 0.0 22a 67.0 0.6 0.5 23a 66.8 0.4 0.3 22b 70.6 4.2 3.3 23b 69.3 2.9 2.3 22c 74.6 8.2 6.4 23c 71.9 5.5 4.3 22d 78.4 12.0 9.4 23d 74.3 7.9 6.2 22e 83.0 16.6 13.0 23e 78.1 11.7 9.2 22f 84.6 18.2 14.2 23f 80.6 14.2 11.1
Coupling of squaric acid monoamide esters of 6-biotin LacNAc-LacNAc 18 or 6-biotin LacdiNAc-LacNAc 19 to BSA led to a theoretical increase of BSA molecular mass by 1281.5 or 1322.5 g/mol per attached glycan, respectively. Shifts towards higher molecular weights are clearly visible in Figure 3.1C for neo-glycoproteins 22b-f and 23b-f accompanied by increased broadening of the protein bands. Neo-glycoproteins 22b-f and 23b-f show significant increase of molecular masses with increasing glycosylation densities. Neo- glycoproteins 22a and 23a were only modified to a minor degree. In addition, a streptavidin blot analysis proved biotinylation of the neo-glycoproteins 22a-f and 23a-f while unmodified
BSA was not detectable (Figure 3.1D). We determined Rf values of protein bands and calculated the attached number of glycans per BSA by molecular weight differences of biotinylated neo-glycoproteins compared to unmodified BSA (Table 3.3). The variation of
83 molar excess of glycans with respect to BSA lysine residues gave different degrees of glycan modification of BSA. In general, a slightly lower modification of BSA with 6-biotin- LacDiNAc-LacNAc was observed (23a-f) when compared to 6-biotin-LacNAc-LacNAc (22a- f). With the highest applied molar excess (1.2 fold) 14.2 (22f) and 11.1 (23f) glycans per BSA molecule were reached. Coupling efficiencies of 13% to 24%, respectively, as ratio of amount of attached glycan to amount of applied glycan are obtained with higher values for 6-biotin- LacNAc-LacNAc. A maximum number of 30 addressable sites of BSA for modification with lactose were previously reported which is half of the lysine residue amount per BSA molecule [104]. We could confirm that 29 and 28 lysine residues per BSA were modified by LacNAc- LacNAc (20i) and LacdiNAc-LacNAc (21i), respectively. Only half of the maximum glycosylation density was achieved for the biotinylated glycans (Table 3.4) although coupling reaction was performed in fed-batch mode. We conclude that limited coupling efficiency is dependent on the molecular mass and steric demands of the conjugated compound. Biotinylation of the tetrasaccharides increases the molecular masses by 40% and consequently reduces the coupling density by about 50%. This was also recently demonstrated for higher molecular weight biantennary N-glycans where a maximal coupling degree of 15 glycans per albumin was reached by click-chemistry addressing lysine residues, despite over 70-fold molar excess of glycan [116].
Table 3.4: Overview of glycosylation densities of all neo-glycoproteins The amount of attached glycans to BSA is given as [mol/mol BSA].
LacNAc-LacNAc LacdiNAc-LacNAc 6-biotin LacNAc-LacNAc 6-biotin LacdiNAc-LacNAc 20a 1.6 21a 1.7 22a 0.5 23a 0.3 20b 3.8 21b 2.2 20c 6.1 21c 5.9 22b 3.3 23b 2.3 20d 10.0 21d 8.8 20e 14.4 21e 14.1 22c 6.4 23c 4.3 20f 19.4 21f 21.3 20g 24.2 21g 24.4 22d 9.4 23d 6.2 20h 26.0 21h 25.0 20i 29.0 21i 27.5 22e 13.0 23e 9.2 20j 7.5 21j 7.4 22f 14.2 23f 11.1 20k 17.8 21k 18.0
In conclusion, novel neo-glycoproteins based on BSA were produced using biotinylated tetrasaccharides as glycans. Clean BSA based neo-glycoproteins varying in the degree of presented oligosaccharide moieties were synthesized. Maximum 29 of 60 lysine residues were accessible squaric acid monoamide esters by a tuned molar excess (1.5-fold) of those. For biotinylated glycans, glycosylation densities between one and a maximum of 14 lysine
84 residues per BSA molecule were obtained. The biotin-labelled glycans were detected by streptavidin. The impact of multivalent ligand presentation on BSA as well as biotin modified oligosaccharides on galectin binding was evaluated in the following binding assays.
3.3.4 Binding of galectin-3 to immobilized tetrasaccharides The biotinylated tetrasaccharides 13 and 14 as well as the corresponding non-biotinylated and deprotected compounds 4 and 5 (11, 12) were immobilized in amino reactive microplates to test galectin binding in an ELISA-type assay (Figure 3.2). Whereas binding to immobilized non-biotinylated tetrasaccharides 11 and 12 was weak, Gal-3 showed high binding signals with immobilized biotinylated compounds 13 and 14 (Figure 3.2A and B). Most interestingly, highest binding was detected for 6-biotin LacdiNAc-LacNAc (14) and essentially no binding for LacdiNAc-LacNAc (12). Similarly, Gal-3 binding to LacNAc-LacNAc (11) was rather weak but increased significantly by C6-biotinylation (13). We conclude that binding of Gal-3 to surface immobilized tetrasaccharides is limited by glycan length. However, biotin obviously promotes binding of Gal-3 to surface immobilized tetrasaccharides and could be helpful for boosting binding efficiencies of glycan ligands to Gal-3.
Comparison of Gal-1 and Gal-3 revealed a high selectivity of immobilized 6-biotinylated compounds 13 and 14 for Gal-3 binding (Figure 3.2D). This is in contrast to previously published data showing that Gal-3 binding is blocked by 6-sialylation of LacNAc glycans [47,117].
We further investigated competitive inhibition of Gal-3 binding to ASF by 6-biotinylated tetrasaccharides. Figure 3.2C shows similar IC50 values for biotinylated (9, 10) and non- biotinylated (4, 5) tetrasaccharides. IC50 values for 4 and 9 were 14 µM and for 5 and 10 around 8 µM demonstrating that N-acetylation increased inhibition potency by over 40%. The corresponding inhibition curves are shown in Figure 3.7. Hence, we conclude that 6- biotinylated Lac(di)NAc-LacNAc tetrasaccharides are novel ligands for human galectin-3. Glycan immobilization plays an important role for ligand recognition because inhibitory potency of soluble glycans was not altered by biotinylation.
85
Figure 3.2: Influence of 6-biotinylation of tetrasaccharides LacNAc-LacNAc and LacdiNAc-LacNAc on galectin binding and inhibition Binding curves (A) and maximal binding signals (B) of Gal-3 to immobilized glycans shows differences between biotinylated and non-biotinylated ligands. In competitive inhibition assays the inhibitory potency was not affected by C6-biotinylation of the tetrasaccharides (C). Comparison of Gal-3 and Gal-1 at 1 µM protein concentration indicates high specificity of biotinylated tetrasaccharides towards Gal-3 (D).
3.3.5 Binding of galectin-3 and galectin-1 to neo-glycoproteins Coupling of glycans to proteins is a form of immobilization to achieve multivalent ligand presentation. We tested BSA-based neo-glycoproteins for evaluation of selective binding of human Gal-1 and Gal-3 in a solid-phase assay using galectin concentration of 1 µM (Figure 3.3). Comparison was made with Gal-1 as member of the large prototype family. In terms of specificity and selectivity we would expect similar results for other of the prototype galectins because glycan ligand specificity is similar among the members of this family.
Non-biotinylated neo-glycoproteins
Both galectins show no binding to unmodified BSA. Whereas binding signals of both galectins to the standard glycoprotein ASF are similar, Gal-3 shows higher binding to neo- glycoproteins compared to Gal-1. At low modification degrees of LacNAc-LacNAc conjugated BSA (20a–d), Gal-3 binds up to 10 times better, whereas at higher glycation degrees (20e–i), Gal-3 reaches two-fold higher binding signals than Gal-1 (Figure 3.3 and Table S1). Our data confirm previous studies that galactose terminated tetrasaccharides and oligosaccharides have higher selectivity for binding of Gal-3 compared to Gal-1 [47,87,118,119].
86
Most importantly, the difference between Gal-3 and Gal-1 binding is more distinct regarding LacdiNAc-LacNAc conjugated BSA (Figure 3.3 and Table S3.1). The binding of Gal-3 to neo-glycoproteins 21a–d is up to 60-fold higher, and to BSA with higher glycan densities (21e–i) still seven-fold higher when compared to Gal-1. The smaller difference between the binding potencies of both galectins to highly modified neo-glycoproteins is probably caused by reaching the maximal binding density of Gal-3 to neo-glycoproteins as well as increased binding of Gal-1 if multiple ligands are presented. Comparing neo-glycoproteins of similar glycosylation density, higher Gal-3 binding on the one hand, and lower Gal-1 binding on the other hand, was observed for LacdiNAc-LacNAc compared to LacNAc-LacNAc at conjugation degrees of approx. six (20c, 21c) and nine (20d, 21d) glycans per molecule (Figure 3.4). The affinity and selectivity for Gal-3 is enhanced compared to Gal-1.
In addition, the impact of GalNAc as terminal sugar is clearly visible in Table 3.5. Only half the amount of conjugated LacdiNAc-LacNAc glycans was required to reach a Gal-3 binding level of 75% compared to LacNAc-LacNAc glycans.
Figure 3.3: Comparison of Gal-1 and Gal-3 binding to immobilized neo-glycoproteins In ELISA-type assay, binding of 1 µM Gal-1 (■) and 1 µM Gal-3 (■) to neo-glycoproteins 20a-i, 21a- i, 22a-f and 23a-f as well as to ASF and unmodified BSA as controls was measured. All Gal-3 binding signals towards neo-glycoproteins are significantly higher than those of Gal-1.
Biotinylated neo-glycoproteins
For all neo-glycoproteins carrying either 6-biotin LacNAc-LacNAc (22a-f) or 6-biotin LacdiNAc-LacNAc (23a-f), Gal-3 binding signals were more than four-fold higher than Gal-1 binding signals (see also Table S1). The binding differences were more pronounced with LacdiNAc glycans (23a-f). Moreover, selectivity for Gal-3 was more distinct at lower degrees of glycan modification with less than three glycans per BSA molecule (22a, 23a,b). Gal-3
87 binds to 22a and 23b 15 times better and to 23a even 80 times better than Gal-1. In contrast, at higher modification degrees Gal-3 binding signals were in average five-fold higher than those for Gal-1.
Gal-1 should not be able to bind internal galactose [120,121] but in the present and earlier studies, weak binding to internal galactose occurred [87,122,123]. Gal-1 bound weakly to LacdiNAc-LacNAc and 6-biotinylated LacNAc-LacNAc as well LacdiNAc-LacNAc (Figure 3.3). Since Gal-1 does not bind to LacdiNAc [76,80,119,120] and modifications at C6- position like sulfation or sialylation and at C2-position like acetylation are not tolerated by Gal-1 [47,117,119,124], binding signals of Gal-1 were probably caused by weak recognition of internal galactose [87,122,123] when multivalent ligands are presented.
Table 3.5: Required glycan number attached to neo-glycoproteins to reach 75% of maximum Gal-3 binding To reach 75% of maximum Gal-3 binding twice of LacNAc-LacNAc and six-fold more LacdiNAc- LacNAc glycans have to be conjugated to BSA compared to the 6-biotinylated counterparts. LacdiNAc-LacNAc glycans showed higher affinity than LacNAc-LacNAc. Only half the amount of glycans was required to reach same Gal-3 binding level; for biotinylated LacdiNAc-LacNAc even one fifth was necessary compared to 6-biotin LacNAc-LacNAc.
LacNAc-LacNAc LacdiNAc-LacNAc 6-biotin LacNAc- 6-biotin LacdiNAc- LacNAc LacNAc 24.2 14.1 13.0 2.3
The galectin binding signals for neo-glycoproteins presenting biotinylated glycans were at first view similar to those obtained for non-biotinylated tetrasaccharides conjugated BSA. However, at least two times higher Gal-3 binding signals are already reached for neo- glycoproteins presenting at least two biotinylated glycans (Figure 3.3). For the non- biotinylated counterparts more than six glycans per BSA molecule were needed. To reach 75% of the maximum binding signal twice the number of LacNAc-LacNAc glycans compared to 6-biotin LacNAc-LacNAc glycans and six-fold more of LacdiNAc-LacNAc glycans compared to 6-biotin LacdiNAc-LacNAc should be presented on one BSA molecule (Table 3.5). Besides the biotinylation, the N-acetylation of the terminal sugar unit made a positive contribution to Gal-3 binding. A six-fold lower glycosylation degree of 6-biotin LacdiNAc- LacNAc was necessary compared to 6-biotin LacNAc-LacNAc to reach a similar Gal-3 binding level. Considering neo-glycoproteins with similar amount of glycans (Figure 3.4), the binding signals of Gal-3 increased with biotinylation and acetylation. For nine glycans per molecule, Gal-3 binding to LacdiNAc-LacNAc (21d) and 6-biotin LacNAc-LacNAc (22d) conjugated BSA was nearly the same. With N-acetylation of the terminal sugar the selectivity
88 for Gal-3 is enhanced, while biotinylation alone did not increase the selectivity for Gal-3 over Gal-1.
In conclusion, neo-glycoproteins modified either with LacNAc-LacNAc or LacdiNAc- LacNAc, 6-biotin LacNAc-LacNAc or 6-biotin LacdiNAc-LacNAc show higher selectivity for Gal-3 compared to Gal-1. LacdiNAc-LacNAc conjugated BSA exhibits highly distinct selectivity for Gal-3, especially at low modification degrees (21a–d). For putative application, e.g., anti-cancer therapy or imaging, Gal-3 could solely be addressed using low modified LacdiNAc-LacNAc conjugated BSA. In case of 6-biotinylated tetrasaccharides preferably few glycans presented on BSA trigger high Gal-3 binding. These signals were higher compared to the non-biotinylated neo-glycoproteins. Moreover, statistically one or two 6-biotinylated LacdiNAc-LacNAc glycans per BSA molecule (23a,b) are sufficient to achieve almost absolute selectivity for Gal-3 binding over Gal-1. Since Gal-3 shows already exceptional binding to neo-glycoproteins 23a-c presenting up to four glycans, multivalency of biotinylated glycans on neo-glycoproteins seems to play a minor role and does not amplify galectin binding.
Figure 3.4: Comparison of Gal-1 and Gal-3 binding to immobilized neo-glycoproteins In ELISA-type assay, binding of 1 µM Gal-1 (■) and 1 µM Gal-3 (■) to neo-glycoproteins with similar glycosylation density (20c, 21c, 22c, 23d with approx. 6 glycans and 20d, 21d, 22d, 23e with approx. 9 glycans) as well as to ASF is shown. The impact of the different ligands is clearly visible.
3.3.6 Galectin-3 binding to neo-glycoproteins at different galectin concentrations LacNAc and LacdiNAc epitopes are both recognized by human Gal-3 with preferential binding to LacdiNAc [76]. Moreover, we identified in previous studies the LacNAc-LacNAc tetrasaccharide as the preferable Gal-3 ligand [87] pointing out that the glycans 4 and 5 are
89 suitable candidates for developing multivalent neo-glycoproteins. Moreover, we could prove positive impact of biotin modification of glycans 9 and 10 on Gal-3 binding, conjugating these glycans likewise to the protein scaffold. BSA with varying numbers of 4, 5, 9 and 10 were analyzed for their binding characteristics as ligands for human Gal-3.
Binding curves
In Figure 3.5, binding signals of Gal-3 on immobilized neo-glycoproteins increase with higher Gal-3 concentration as well as higher modification densities. The results confirm Gal-3 binding to all four types of BSA based neo-glycoproteins and no binding to unmodified BSA. A significant difference in Gal-3 binding was monitored for those neo-glycoproteins with minor modification degrees. For non-biotinylated neo-glycoproteins 20a-g and 21a-f, respectively, significant increases of saturated binding signals are detected up to a modification degree of approximately 20 glycans per BSA. Binding of Gal-3 hardly improves with increasing numbers of glycans above these modification degrees which may be due to approaching the maximal Gal-3 steric occupancy per molecule. The maximal binding signal of Gal-3 is for 20a-c and 21a-c in the same range but is slightly higher for 21d-i compared to 20d-i which may hint at better binding of Gal-3 to the LacdiNAc epitope (see also Figure 3.6A and B). An even more enhanced Gal-3 binding was achieved for biotinylated neo- glycoproteins. Already with very low glycosylation densities of three glycans per BSA molecule (22b and 23b) high binding signals of about 0.8 were measured (Figure 3.5), that were reached not until nine binding sites of non-biotinylated neo-glycoproteins (20d and 21d).With more than two conjugated glycans, only a minor increase of saturated galectin binding was observed being more pronounced for 6-biotin LacdiNAc-LacNAc conjugated BSA (23a-f). With 23a-f overall higher binding signals of Gal-3 were reached compared to 22a-f (see also Figure 3.6C and D). It is known that the hydroxyl group at C2-position is not involved in interacting with the CRD of Gal-3 [125]. Thus, C2-modification is tolerated by Gal-3 and even enhances galectin binding to the LacdiNAc epitope as reported earlier [76,80,119]. Since modifications at C6-position are not tolerated in the case of Gal-3 binding because hydrogen bond formation is obligatory for the binding process [119,126,127], the role of 6-biotinylation remains unclear. In our study we demonstrate the positive effect of BACH conjugated to tetrasaccharides. The spacer at C6-position used in this study allows even the formation of two hydrogen bonds. Potentially, the disadvantage of C6-modification is therefore circumvented and Gal-3 binding even strengthened. Thus, it is possible to design Gal-3 ligands that carry functionalities that do not interact directly with the binding sites but with more distant residues. The conjugation of the tetrasaccharides with biotin could also enhance the Gal-3 binding to the internal LacNAc unit by optimizing the ligand protein
90 complex regarding orientation or hydrophobicity. Hydrophobic modifications at C3-position were already shown as effective Gal-3 binders [128-131]. Therefore, the C6-biotinylated non- reducing terminal LacNAc and LacdiNAc structures may be considered sophisticated glycan- derived hydrophobic C3-modifications of LacNAc. The precise role of the C6-biotin modification and its interaction with the binding pocket from a structural point of view remains a matter of investigation.
Figure 3.5: Binding of human Gal-3 to neo-glycoproteins presenting different numbers of glycans In an ELISA-type assay Gal-3 at different concentrations was incubated on immobilized neo- glycoproteins (5 pmol per well) presenting (A) LacNAc-LacNAc (20a-i), (B) LacdiNAc-LacNAc (21a-i), (C) 6-biotin LacNAc-LacNAc (22a-f) and (D) 6-biotin LacdiNAc-LacNAc (23a-f). Binding signals were plotted against galectin concentration and fitted.
91
Figure 3.6: Binding of human Gal-3 to neo-glycoproteins presenting different numbers of glycans The diagrams in Figure 3.5 are zoomed. The x-axis is scaled from 0 to 1.4 µM to better see the binding differences.
Differences in Gal-3 binding to neo-glycoproteins are elucidated when the apparent Kd data are compared (Figure 3.7, Table S3.2). The Kd value is the concentration at which 50% saturation of the binding signal is reached and a measure for binding affinity. In general, LacdiNAc-LacNAc conjugated BSA shows better binding of Gal-3 compared to LacNAc-
LacNAc conjugated BSA resulting in general lower Kd values for all modification degrees 21a-i and 23a-f. Significant differences between the types of neo-glycoproteins are obvious reflecting different selectivity of Gal-3 for the glycans. Interestingly, ASF carrying nine LacNAc units (three triantennary N-glycans) is not the preferred glycoprotein (highest apparent Kd).
Kd values for Non-biotinylated neo-glycoproteins
For binding of Gal-3 to 20a-i Kd values decrease more gradually reaching the lowest apparent
Kd value of 0.1 µM for 20i (Table S3.2). A significant drop of the Kd value is observed for the neo-glycoproteins 20f and 20i with about 19 and 29 LacNAc-LacNAc glycans, respectively.
In comparison, Kd values for LacdiNAc-LacNAc conjugated BSA are already significantly lowered for 21d-f with about nine, 14 and 21 LacdiNAc-LacNAc glycans, respectively,
92
−8 reaching the lowest Kd value for Gal-3 in the nanomolar range (3 × 10 M for 21f). Comparing neo-glycoproteins with similar conjugation level (Figure 3.8), BSA with six LacdiNAc-LacNAc glycans (21c) showed slightly higher Gal-3 binding potency than BSA with six LacNAc-LacNAc glycans (20c). This difference in binding potency became more distinct when nine glycans are presented (20d vs. 21d), where 21d showed almost three times higher binding potency than 20d and over five-fold higher Gal-3 binding than for ASF with also nine possible ligands. These data emphasize high selectivity of Gal-3 for LacdiNAc- LacNAc glycans over LacNAc-LacNAc-glycans.
Figure 3.7: Apparent Kd values of human Gal-3 for binding to different immobilized neo- glycoproteins
The Gal-3 concentration [µM] at half-maximal binding (Kd) is shown for the neo-glycoproteins modified with LacNAc-LacNAc 20a-i (■), LacdiNAc-LacNAc 21a-i (■), 6-biotin LacNAc-LacNAc
22a-f (■) and 6-biotin LacdiNAc-LacNAc 23a-f (■) as well as ASF (■). Kd values decreased with increasing modification degree indicating higher affinity of human Gal-3
Kd values for biotinylated neo-glycoproteins
Regarding biotinylation, the values for 23a-f are overall lower compared to 22a-f with a significant drop between two (23b) and four (23c) glycans per molecule. For binding of Gal-3 to 22a-f Kd values decrease gradually reaching the lowest apparent Kd value of 0.22 µM for 22f (Table S3.2). For neo-glycoprotein 23f with 10.5 glycans per BSA molecule Gal-3 shows highest affinity (Kd 0.05 µM). Despite the lower glycosylation density of 23f compared to
22f, the Kd value was more than four-fold lower emphasizing once more the positive influence of the N-acetylated galactosamine. Gal-3 affinity for 6-biotin LacNAc-LacNAc and 6-biotin LacdiNAc-LacNAc conjugated BSA 22a-f and 23a-f is higher than for the corresponding non-biotinylated neo-glycoproteins with similar glycosylation level. With respect to the minimum number of glycans to obtain high Gal-3 affinity, biotinylated glycan
93 ligands are superior over non-biotinylated glycans. A Kd value below 0.1 µM was already reached for 6-biotin LacdiNAc-LacNAc with a glycosylation density of six glycans per BSA molecule (23d) compared to 14 LacdiNAc-LacNAc glycans per BSA molecule (21e), whereas neo-glycoproteins presenting LacNAc-LacNAc did not achieve this Kd value. The direct comparison between neo-glycoproteins of similar glycosylation degree (Figure 3.8) shows the high impact of the N-acetylation of the terminal 6-biotinylated galactose unit. For six and nine conjugated glycans per BSA molecule, over five-fold higher binding potency was reached with 6-biotin LacdiNAc-LacNAc glycans (23d,e) compared to 6-biotin LacNAc- LacNAc glycans (22c,d) showing exceptional Gal-3 binding properties of 6-biotin LacdiNAc- LacNAc. Regarding nine conjugated glycans, BSA presenting LacdiNAc-LacNAc (21d) was even superior to BSA presenting 6-biotin LacNAc-LacNAc (22d) concerning Gal-3 binding. The highest binding potency by far was reached when LacdiNAc-LacNAc glycans were biotinylated (23e). 23e was more than 20 times more potent than ASF. We conclude that terminal GalNAc improved the affinity of Gal-3, and additional 6-biotinylation gained a further high affinity increase.
Figure 3.8: Relative binding potency of Gal-3 towards neo-glycoproteins The potency of Gal-3 binding to neo-glycoproteins with similar glycosylation density (20c, 21c, 22c, 23d with approx. 6 glycans and 20d, 21d, 22d, 23e with approx. 9 glycans) in relation to ASF is compared. Biotinylated LacdiNAc-LacNAc conjugated to BSA showed clearly the highest affinity.
Influence of multivalency
As Gal-3 forms pentamers or oligomers during binding events [49,52] and multivalent carbohydrate presentation plays an important role in binding enhancement [53-56], the binding signal per glycan ligand was calculated for all applied Gal-3 concentrations and was
94 related to 20d, 21d, 22d and 23d, respectively (Figure 3.9 and Table S3.3). For 20a-h and 21a-c, binding signals are only detectable for Gal-3 concentrations from 0.05 µM; 22a-f and 23a-c gave values for Gal-3 concentrations from 0.25 µM (Figure S3.3). In contrast, for LacdiNAc-LacNAc conjugated BSA as well as 6-biotin LacdiNAc-LacNAc conjugated BSA modification densities of at least nine glycans (21d) and six glycans (23d) per BSA are sufficient to reach binding signals of Gal-3 at concentrations even below 0.025 µM. Thus, binding signals per glycan for 20d, 21d, 22d and 23d are used as the benchmark for calculating binding potencies.
Non-biotinlyated neo-glycoproteins Enhancement of binding by multivalent presentation is not observed for LacNAc-LacNAc conjugated BSA (Figure 3.9A). Binding signals of Gal-3 to 20a-i in relation to one binding site reach nearly the same level, in the used concentration range. In contrast, enhanced avidity is shown for neo-glycoproteins 21e-i carrying LadiNAc-LacNAc at concentrations below 0.05 µM with highest multivalent effects for 0.005 µM (Figure 3.9B). Up to 100 times higher binding signal in relation to one binding site is observed for modification densities above 20 glycans per BSA. However, at Gal-3 concentrations above 0.1 µM, the binding signals in relation to one glycan are for higher modified BSA even lower than for weak glycosylated BSA. This is possibly caused by hindered accessibility of all binding sites in the presence of Gal-3 excess. Biotinylated neo-glycoproteins For 6-biotin LacNAc-LacNAc conjugated BSA, the results are similar to the non-biotinylated counterparts (Figure 3.9C). A certain Gal-3 concentration (0.025 µM) was necessary to detect binding signals. In contrast, 6-biotin LacdiNAc-LacNAc conjugated BSA gave binding signals of Gal-3 at concentrations below 0.05 µM but multivalent effects were not observed (Figure 3.9D), contrary to neo-glycoproteins carrying LacdiNAc-LacNAc . Even neo- glycoproteins with low number of glycans were high affinity ligands for Gal-3, especially 6- biotin LacdiNAc-LacNAc conjugated BSA, so that more glycans on one BSA molecule had hardly an influence. Moreover, the maximal achieved glycan density of eleven glycans per molecule is probably too low to gain multivalently enhanced Gal-3 binding. Eight conjugated glycans to BSA-based neo-glycoprotein were reported to be too distant to reach enhanced avidity for Gal-1 [120].
95
Figure 3.9: Binding signals of human Gal-3 for neo-glycoproteins in relation to one binding site Relative signals of Gal-3 binding to neo-glycoproteins with (A) LacNAc-LacNAc 20a-i and (B) LacdiNAc-LacNAc 21a-i as well as (C) 6-biotinylated LacNAc-LacNAc 22a-f and LacdiNAc- LacNAc 23a-f at Gal-3 concentration ranging from 0.005-10 µM are shown. Values of binding enhancement result from binding signals per glycan in relation to the Gal-3 binding signal per glycan for 20d, 21d, 22d and 23d, respectively.
Consequently, multivalent ligand presentation on BSA can induce cluster glycoside effects depending on the ligand, the valency and the galectin concentration. Here, only LacdiNAc-
96
LacNAc conjugated BSA with modification degrees above eight glycans per BSA show enhanced avidity towards Gal-3 at concentrations below 0.05 µM. Multivalent effects were also found for Gal-3 using N-glycan modified human serum albumin synthesized by click- chemistry [116]. The multivalent neo-glycoprotein carrying eight biantennary N-glycans showed a 30-fold increase in affinity per glycan unit in comparison to the monovalent reference.
Mostly, Gal-3 binding was investigated for multivalent ligands based on chemical scaffolds like calixarenes, dendrimers or quantum dots [55,56,132-134]. Comparison of our results with findings for these synthetic scaffolds is difficult but can be drawn with regard to the glycan type and number. With multivalent LacNAc modified quantum-dots presenting 108 glycans, a 92-fold higher affinity towards Gal-3 in comparison to soluble LacNAc was shown using SPR [55]. With respect to one ligand, we gain higher enhancement of Gal-3 binding to BSA based neo-glycoproteins modified with a much smaller number of glycans, namely at least 14 LacdiNAc-LacNAc glycans (21f).
Neo-glycoproteins enhancing the binding potency of low concentrated Gal-3 by a factor of 100 may be promising diagnostic tools for Gal-3 in sera of cancer patients. The expression of Gal-3 is upregulated in different tumor cells and induces tumor progression and metastasis [26-32,135]. The serum level of Gal-3 in patients with cancer is likewise significantly increased compared with healthy individuals [136-138]. Therefore, interfering with Gal-3 by applying effective inhibitors can reduce the angiogenic and metastatic potential of tumor cells [41,139-141]. Multivalent systems, e.g., based on dendrimer scaffolds, are of high interest as they induce cluster formation and exhibit enhanced inhibitory potencies [48,134,142]. In the present study, we have generated neo-glycoproteins which are effective for multivalent binding at high modification degrees (21f-i) with the Gal-3 selective epitope LacdiNAc- LacNAc. The highest multivalent effects occur at Gal-3 concentrations of 5–10 nM (130–260 ng/mL) which is in the range of Gal-3 level in human serum of cancer patients (mean value = 320 ng/mL). This is about five times higher than in healthy individuals (mean value = 62 ng/mL) [138]. Hence, in roughly that concentration range, Gal-3 inhibition should reduce tumor aggressiveness making these LacdiNAc-LacNAc conjugated BSA promising candidates for anti-cancer therapy. In vivo applications are possible, because BSA based neo- glycoproteins were already applied to mice, rats and rabbits [143,144].
In conclusion, BSA-based neo-glycoproteins are high-affinity ligands for Gal-3 with higher binding for increasing modification densities. At Gal-3 concentrations below 0.05 µM, up to 100-fold enhanced avidity is shown for neo-glycoproteins presenting more than 14 LacdiNAc-LacNAc glycans.
97
3.3.7 Neo-glycoproteins as inhibitors for galectin-3 High-affinity glycan ligands for Gal-3 are of interest as inhibitors in biomedical applications [46-48]. In this context, we performed inhibition assays to investigate Gal-3 binding to immobilized ASF in the presence of neo-glycoproteins of three different modification degrees (20j,k,g; 21j,k,g; 22b,d,e; 23b,d,e). To compare inhibitory potencies with soluble oligosaccharides, 4, 5, 9 and 10 were also used in this assay. Plotting residual Gal-3 binding signals against inhibitor concentration provides the sigmoidal curves depicted in Figure 3.10 and are used for IC50 calculation (Figure 3.11 and Table 3.6).
Figure 3.10: Competitive inhibition of Gal-3 binding by free tetrasaccharides and neo-glycoproteins Residual binding of 5 µM Gal-3 to ASF in the presence of variable concentrations of (A) LacNAc- LacNAc conjugated BSA (20j, 20k and 20g), (B) and LacdiNAc-LacNAc conjugated BSA (21j, 21k and 21g), (C) 6-biotin LacNAc-LacNAc conjugated BSA (22b,d,e) and (D) 6-biotin LacdiNAc- LacNAc conjugated BSA (23b,d,e) as well as monovalent controls (4, 5, 9 and 10) is shown. With increasing inhibitor concentrations, Gal-3 binding signals on ASF decreased. The curves shifted to lower inhibitor concentrations with rising glycosylation density. Unmodified BSA has no influence on the Gal-3 binding.
Inhibition curves in Figure 3.10 indicate decrease of Gal-3 signal with increasing inhibitor concentrations for neo-glycoproteins and free glycans. BSA as control showed no inhibitory effect. As stronger inhibition is shown by a curve shift to lower inhibitor concentrations, neo- glycoproteins showed higher inhibition potencies than free tetrasaccharides. With increasing
98 glycosylation densities of the neo-glycoproteins stronger inhibition of Gal-3 binding to ASF is observed. For non-biotinylated neo-glycoproteins (Figure 3.10A and B), BSA modified with 18 and 24 binding sites (20k, 20g and 21k, 21g) are obviously stronger inhibitors than BSA presenting 7.5 glycans (20j and 21j) though only small differences exist between both higher modification degrees. Increased modification of BSA results in higher inhibition, but does not strengthen the inhibition proportionally. When comparing the tetrasaccharides LacdiNAc- LacNAc (5) and LacNAc-LAcNAc (4), 5 reveals a 1.5-fold higher inhibition of Gal-3 binding. This is based on better binding to LacdiNAc terminated glycan as already demonstrated [76]. For biotinylated neo-glycoproteins (Figure 3.10C and D), a significant shift already occurs for BSA-based neo-glycoproteins decorated with at least two glycans (22b, 23b). The effect of increased glycan density was less pronounced, except for the difference of 22b and 22d.
IC50 values were calculated and showed clear differences between the inhibitory potencies of neo-glycoproteins and corresponding free glycans (Figure 3.11 and Table 3.6). IC50 values of 4, 5, 9 and 10 are significantly higher than those reached by the neo-glycoproteins. It is also visible that biotinylation of the free tetrasaccharides gave no improvement of inhibiton due to overall similar IC50 values. Only after immobilization to BSA clear inhibitory differences exist for non-biotinylated and biotinylated glycans. Moreover, for LacNAc-LacNAc and 6- biotin LacNAc-LacNAc conjugated BSA, it could be observed that with increasing glycan number the inhibitory potency is drastically enhanced. 20j with 7.5 LacNAc-LacNAc glycans per BSA improved the inhibition of Gal-3 binding to ASF by a factor of 15 compared to 4. A successive improvement with increasing glycosylation density up to 370-fold higher inhibitory potency than 4 was reached with 24 glycans per BSA. In contrast for LacdiNAc- LacNAc conjugated BSA, above a certain glycan number the increase in glycosylation gained scarcely any further improvement. That applies for LacdiNAc-LacNAc above 18 glycans per BSA molecule (21k). For 21j with 7.5 glycans per BSA a 6.5-fold higher inhibitory potency was reached compared to the free tetrasaccharide 5. The presentation of more glycans on BSA (21k,g) improved the inhibition potential by a factor of 20 compared to 21j.
With BSA presenting three 6-biotin LacNAc-LacNAc glycans (22b) a 35-fold increase of inhibition potency was reached compared to the free biotinylated tetrasaccharide 9. Presenting additional six glycans (22d) the IC50 value is lowered by a factor of 13 compared to 22b reaching low nanomolar concentration. A further addition of four glycans doubled the already high inhibitory potency up to 850-fold higher potency than 9. 6-biotin LacdiNAc-LacNAc with two glycans per BSA (23b) won the highest drop of the IC50 value with the lowest glycosylation density; a 50-fold higher inhibitory potency was gained compared to the free
99 tetrasaccharide 10, respectively. The presentation of more biotinylated LacdiNAc-LacNAc glycans on BSA (23d,e) improved the inhibition potential by a factor merely of four. In summary, all neo-glycoprotein species reached IC50 values in the two-digit nanomolar range. For non-biotinylated neo-glycoproteins, high conjugation levels of at least 20 glycans per molecular were required, while for biotinylated neo-glycoproteins six to nine glycans per molecule were sufficient to get highly efficient inhibitors.
Figure 3.11: IC50 values of monovalent tetrasaccharides and neo-glycoproteins for Gal-3 inhibition
IC50 values of selected neo-glycoproteins (■) and the corresponding free saccharides (■) are compared. Lower values indicate higher inhibitory potencies. Neo-glycoproteins showed higher inhibitory potential compared to monovalent glycans as well as with increasing number of glycans per molecule.
Calculated IC50 values are depicted in Table 3.6.
Multivalent presentation of the glycans had a relevant impact on Gal-3 inhibition by neo- glycoproteins in solution. Relative potencies calculated per glycan reveal multivalency effects (Table 3.6). The improvement of the inhibitory potency by multivalent presentation is overall more pronounced for (6-biotin) LacNAc-LacNAc conjugated BSA than for (6-biotin)
LacdiNAc-LacNAc conjugated BSA since the IC50 value for 4/9 is higher than for 5/10. Thereby, the inhibitory potency can be more enhanced. The general presentation of non- biotinylated glycans on BSA has little influence on the inhibitory potency indicated by only up to two times improved relative potency in relation to one binding site compared to 4 and 5.
100
Here, we can conclude that multivalent effects are not relevant for low glycosylation degrees of about 7.5 glycans per BSA (20j, 21j). In relation to one binding site, inhibitory potency is improved by 20k four times and by 20g 7.5 times in comparison to 20j, and almost twice by 20g compared to 20k. Regarding LacdiNAc-LacNAc conjugation, inhibition potency is enhanced by 21k by a factor of 7 per glycan compared to 21j, whereas with 21g scarcely any further improvement is gained. Higher inhibitory potency of LacdiNAc-LacNAc conjugated BSA over LacNAc-LacNAc conjugated BSA is not observed. This can possible explained by the higher Gal-3 concentration (5 µM) required by the inhibition assay due to weaker binding to ASF than to neo-glycoproteins (Chapter 3.3.6). Therefore, equal inhibition potencies of both types of modifications are obtained.
Table 3.6: Calculated IC50 values and relative inhibitory potencies of soluble tetrasaccharides and neo- glycoproteins Values of inhibitory potentials are stated for tested tetrasaccharides and neo-glycoproteins. Relative potencies are related to the corresponding tetrasaccharide and additionally to one binding site. Neo- glycoproteins show up to 60 times enhanced potencies per glycan compared to monovalent tetrasaccharides. Glycosylation is given as [mol/mol BSA].
Compound Glycosylation IC50 [µM] Relative potency Relative potency per glycan 4 1 13.04 ± 3.43 1.0 ± 0.26 1.0 ± 0.26 20j 7.5 0.85 ± 0.21 15.3 ± 3.14 2.0 ± 0.42 20k 17.8 0.09 ± 0.02 145.2 ± 2.45 8.2 ± 0.14 20g 24.2 0.04 ± 0.01 367.8 ± 5.12 15.2 ± 0.21 5 1 7.23 ± 0.87 1.0 ± 0.12 1.0 ± 0.12 21j 7.4 1.10 ± 0.22 6.6 ± 1.42 0.9 ± 0.19 21k 18.0 0.06 ± 0.02 127.1 ± 2.56 7.1 ± 0.15 21g 24.4 0.04 ± 0.02 182.0 ± 4.06 7.5 ± 0.17 9 1 14.8 ± 2.06 1.0 ± 0.10 1.0 ± 0.10 22b 3.3 0.42 ± 0.18 35.2 ± 15.30 10.7 ± 4.60 22d 9.4 0.03 ± 0.01 464.1 ± 83.70 49.4 ± 8.90 22e 13.0 0.02 ± 0.01 856.0 ± 287.50 65.8 ± 22.10 10 1 7.66 ± 1.24 1.0 ± 0.20 1.0 ± 0.20 23b 2.3 0.15 ± 0.06 50.3 ± 19.10 21.9 ± 8.30 23d 6.2 0.04 ± 0.02 197.6 ± 84.80 31.9 ± 13.70 23e 9.2 0.03 ± 0.00 297.9 ± 52.40 32.4 ± 5.70
Regarding 6-biotin LacNAc-LacNAc, neo-glycoprotein 22b (three glycans) showed eleven- fold, 22d (nine glycans) 50-fold, and 22e (13 glycans) almost 70-fold inhibitory potency in relation to one binding site than 9. The multivalent influence for 6-biotin LacdiNAc-LacNAc conjugated BSA was less pronounced. Relative inhibitory potencies of 22 to 33 per glycan were observed for 23b, 23d and 23e compared to 10. The overall lower potencies are due to the already higher inhibition potential of 10 compared to 9. However, 23b showed higher
101 potency of one binding site than 22b. Consequently, higher modification degrees above six glycans per BSA (23d, 23e) did not achieve elevated potencies because the contribution of each further added glycan chain got lower with increased affinity of the glycan ligand. This applies also for non-biotinylated LacdiNAc-LacNAc. However, presentation of 6-biotinylated Lac(di)NAc terminated tetrasaccharides as neo-glycoprotein affected Gal-3 binding positively when compared to non-biotinylated counterparts by showing higher inhibitory potencies. As C6-biotinylation of the tetrasaccharides caused no improvement using them as soluble ligands in an inhibition assay (Figure 3.2C) the application of neo-glycoproteins is beneficial for binding characterization as well as design of high-affinity ligands.
We determined IC50 values in the low nanomolar range for non-biotinylated BSA based neo- glycoproteins modified with at least 18 glycans and for biotinylated glycan types with BSA modification degrees above six. Reaching such high inhibitory potencies is caused by binding enhancement due to protein chelation in the presence of multivalent ligands [16]. As Gal-3 forms oligomers involving multivalent ligands [49,52] it is prone to be a candidate for binding enhancement by multivalent effects. In contrast, Gal-3 binding to ASF was characterized by high affinity to the first LacNAc epitope of the triantennary N-glycan with increasing negative cooperativity for subsequent binding events [53]. However, lactose based dendrimers showed strong multivalent effects in a solid phase inhibition assay with ASF [145]. Gal-3 binding to monovalent lactose dendrimer was already about 10 times higher than to free lactose indicating interactions with the dendrimer backbone [145,146], contrary to the neo- glycoproteins in the present study. Moreover, calixarenes presenting lactose multivalently increased the inhibitory potency towards Gal-3 binding to ASF by a factor of 12.5 reaching an
IC50 of 200 µM [132]. Multivalent LacNAc bearing calix[4/6]arenes with 2/3' substitutions in the LacNAc core showed IC50 values up to 0.15 µM for Gal-3 binding to ASF, which were about 1500 times lower compared to monovalent species [133]. Hence, multivalent presented LacNAc increases Gal-3 binding more than multivalent lactose. In the present study, multivalent presentation of tetrasaccharides on BSA yields lower IC50 values up to 0.02 µM indicating a large benefit using tetrasaccharides instead of disaccharides. For 6-biotinylated tetrasaccharides even conjugation densities below ten glycans on BSA were necessary to obtain IC50 values in the low nanomolar range In conclusion, we could prove inhibitory potency for Gal-3 binding to ASF applying the presented neo-glycoproteins. For non-biotinylated neo-glycoproteins with at least 18 glycans per BSA, more than seven times higher potencies in relation to one binding site are recorded. For 6-biotinylated tetrasaccharides, even up to 50-fold potency increase in relation to one binding site is reached at low glycan densities of about 10 glycans per molecule. We suggest
102 that IC50 values can even be reduced in assays using lower Gal-3 concentration, e.g., in vivo assays, as we could show up to 100 times increased Gal-3 binding at concentrations below 0.05 µM (Figure 3.9).
3.3.8 Neo-glycoproteins as galectin-3 ligands in surface plasmon resonance spectroscopy Albumin modified with suitable glycans is of high interest for the design of galectin ligands as the present and recent studies show [116,120]. We could already prove Gal-3 binding to neo- glycoproteins in a solid phase assay. SPR spectroscopy was additionally performed to compare neo-glycoproteins as Gal-3 ligands in a flow set-up. Gal-3 was immobilized via 2+ His6-tag on a Ni -dextran surface and binding of neo-glycoproteins as well as ASF were monitored. This set-up allowed complete removal of Gal-3 by releasing of Ni2+ by EDTA solution. Thus, strong interactions of Gal-3 and ligand had no influence on the following measurement because fresh Gal-3 solution was used for each ligand.
Table 3.7: Neo-glycoproteins used in SPR measurements Biotinylated and non-biotinylated LacNAc and LacdiNAc terminated tetrasaccharides conjugated to BSA were used as Gal-3 ligands in SPR measurements. The numbering and modification degree of each neo-glycoprotein are shown.
LacNAc-LacNAc LacdiNAc-LacNAc 6-biotin LacNAc-LacNAc 6-biotin LacdiNAc-LacNAc 20a 1.6 21a 1.7 22a 0.5 23a 0.3 20j 7.5 21j 7.4 22b 3.3 23b 2.3 20e 14.4 21e 14.1 22c 6.4 23c 4.3 20k 17.8 21k 18.0 22d 9.4 23d 6.2 20g 24.2 21g 24.4 22e 13.0 23e 9.2 20i 29.0 21i 27.5 22f 14.2 23f 11.1
Four different tetrasaccharides, LacNAc-LacNAc (20a,j,e,k,g,i), LacdiNAc-LacNAc (21a,j,e,k,g,i), 6-biotin LacNAc-LacNAc (22a-f) and 6-biotin LacdiNAc-LacNAc (23a-f), presented on BSA with different modification degrees (Table 3.7) were compared as Gal-3 ligands in SPR measurements.
Figure 3.12 reveals that the dissociation constants (KD) fall with increasing modification densities for all types of neo-glycoproteins indicating rising Gal-3 affinity. KD is defined as ratio of koff and kon and therefore calculated by fitting dissociation and association curves. We observed typical sensorgrams (Figure S3.4) with clear association and dissociation phase as previously published for a similar set-up [55,116]. The SPR sensorgrams show faster association with increasing modification degree of neo-glycoproteins. The dissociation was
103 very slow with a relative high response level indicating that glycoproteins remained bound to Gal-3 for the tested dissociation time.
Figure 3.12: KD values resulting from binding of neo-glycoproteins on immobilized Gal-3 in SPR measurements Gal-3 was immobilized on a sensorchip and binding to neo-glycoproteins as well as ASF was monitored in flow (20 µL/min). Apparent KD values were calculated for one concentration by fitting association and dissociation using the software Scrubber2. (* KD was not detectable)
All neo-glycoproteins, except 20a as well as 22a and 21a with no calculable values, showed higher binding affinity towards Gal-3 than ASF. Moreover, apparent KD values decreased with increasing glycosylation density but were mostly in the range of 1-10 nM (see Table 3.8 and Table S3.4). No significant increase in affinity caused by increasing multivalency was observed. At similar modification degrees between seven and nine (20j, 21j, 22d, 23e) BSA presenting biotinylated tetrasaccharides gained lower apparent KD values that could be again reduced for the GalNAc containing glycans (23e). In numbers, 20j and 21j showed almost 8- fold, 22d 24-fold and 23e 37-fold better binding to immobilized Gal-3 than ASF that has nine possible ligands. Therefore, the biotinylated neo-glycoproteins were four to five times more affine than the non-biotinylated neo-glycoproteins indicting a minor role of LacdiNAc in this assay-mode. However, the highest binding potency was observed for LacdiNAc-LacNAc conjugated BSA with 27.5 glycans (21i) and was almost 60-fold higher compared to ASF. This result keeps up with LacNAc-quantum dots presenting 108 LacNAc epitopes and showing 92-fold better binding to Gal-3 compared to free LacNAc [55]. Exactly the same affinity enhancement compared to ASF was described for neo-glycoproteins carrying 15 biantennary N-glycans [116]. In summary, with increasing glycosylation density, Gal-3 binds stronger to non-biotinylated neo-glycoproteins. At low to medium modifications degrees,
104 biotinylated conjugated tetrasaccharides led to higher Gal-3 affinity compared to non- biotinylated glycans.
Table 3.8: Values of KD in SPR measurements with neo-glycoproteins and immobilized Gal-3
Apparent KD values determined by SPR are compared for all designed neo-glycoproteins. Neo- glycoproteins and ASF were flowed over the surfaced immobilized with Gal-3. Values were calculated by fitting association and dissociation using Scrubber2. (n.d. – not detectable)
Ligand Apparent KD (nM) Ligand Apparent KD (nM) 20a 103 ± 3 21a n.d. 20j 9.80 ± 0.10 21j 9.50 ± 0.10 20e 4.77 ± 0.07 21e 2.11 ± 0.02 20k 3.18 ± 0.06 21k 2.86 ± 0.02 20g 2.35 ± 0.05 21g 1.50 ± 0.20 20i 1.72 ± 0.04 21i 1.28 ± 0.02 22a n.d. 23a 22 ± 1 22b 9.40 ± 0.10 23b 4.17 ± 0.05 22c 4.84 ± 0.05 23c 2.74 ± 0.03 22d 3.12 ± 0.03 23d 1.95 ± 0.02 22e 2.64 ± 0.03 23e 2.01 ± 0.02 22f 2.57 ± 0.02 23f 1.90 ± 0.20 ASF 75 ± 2
In conclusion, results of SPR measurements show dissociation constants (KD) of (biotinylated) Lac(di)NAc conjugated BSA-based neo-glycoproteins in the low nanomolar range which are in agreement with calculated apparent Kd and IC50 values determined by solid-phase ELISA assays. The biotinylated glycans gave lower dissociation constants (KD) at lower modification densities than non-biotinylated neo-glycoproteins
3.4 Conclusions Here, we report for the first time on biotin modification of glycans for the design of selective high-affinity Gal-3 ligands. Tetrasaccharides with C6-biotinylation of the terminal Galactose/GalNAc sugar unit were synthesized and characterized as Gal-3 ligands. Gal-3 binding to immobilized biotinylated glycans was improved in comparison to non-biotinylated counterparts and even enabled high selectivity for Gal-3 when compared to Gal-1.
With the design of neo-glycoproteins based on BSA, we demonstrate the efficient synthesis of high-affinity ligands for human Gal-3. We show that conjugation of LacNAc-LacNAc or LacdiNAc-LacNAc as well as their biotinylated counterparts to BSA is tunable using squaric acid diethyl ester as linker. For the first time the selective Gal-3 binding to multivalent neo- glycoproteins revealing and quantifying the influence of defined multivalency is investigated. We could show a high impact of presenting ligands multivalently on a protein scaffold. With
105 increasing glycosylation density high Gal-3 avidity, Kd and IC50 values below 50 nM, are observed. The biotinylated ligands show high affinity and high inhibitory potency even at low glycosylation density.
Enhancement of binding by a factor of about 100 was observed with LacdiNAc-LacNAc conjugated BSA modified with at least 21 glycans at a Gal-3 concentration of 5 nM. The efficient and selective binding of such neo-glycoproteins at serum level concentrations of Gal- 3 may have high impact in anti-cancer therapy. In addition, the neo-glycoproteins can be further loaded with cytotoxic compounds and labeled with fluorescent dyes to yield tailor- made theranostics. Further application could be the capture of Gal-3 from serum of cancer patients on suitable surfaces of biosensors.
In conclusion, our tailor-made neo-glycoproteins are suitable candidates for targeting Gal-3 in cancer related biomedical research. Moreover, our findings open new possibilities for C6- modifications of carbohydrate structures in galectin inhibitor design.
3.5 Contributions S. Böcker planned the experiments, D. Laaf performed syntheses of non-biotinylated neo- glycoproteins, S. Böcker synthesized biotinylated glycans and neo-glycoproteins, S. Böcker and D. Laaf performed neo-glycoprotein analyses and galectin binding assays for non- biotinylated structures, S. Böcker performed neo-glycoprotein analyses and galectin binding assays for biotinylated structures, S. Böcker analyzed Gal-3 binding by SPR, S. Böcker evaluated the binding data.
3.6 References 1. Varki, A. Biological roles of oligosaccharides: All of the theories are correct. Glycobiology 1993, 3, 97-130. 2. Zhao, Y.-Y.; Takahashi, M.; Gu, J.-G.; Miyoshi, E.; Matsumoto, A.; Kitazume, S.; Taniguchi, N. Functional roles of n-glycans in cell signaling and cell adhesion in cancer. Cancer Sci. 2008, 99, 1304-1310. 3. Boscher, C.; Dennis, J.W.; Nabi, I.R. Glycosylation, galectins and cellular signaling. Curr. Opin. Cell. Biol. 2011, 23, 383-392. 4. Fred Brewer, C. Binding and cross-linking properties of galectins. BBA Gen. Subj. 2002, 1572, 255-262. 5. Funasaka, T.; Raz, A.; Nangia-Makker, P. Galectin-3 in angiogenesis and metastasis. Glycobiology 2014, 24, 886-891. 6. Lau, K.S.; Dennis, J.W. N-glycans in cancer progression. Glycobiology 2008, 18, 750-760. 7. Noorjahan, P. Role of galectins in wound healing. In Galectins and disease implications for targeted therapeutics, ACS: Washington, DC, USA, 2012; Vol. 1115, pp 415-432. 8. Schattner, M. Platelets and galectins. Ann. Transl. Med. 2014, 2. 9. Collins, B.E.; Paulson, J.C. Cell surface biology mediated by low affinity multivalent protein– glycan interactions. Curr. Opin. Chem. Biol. 2004, 8, 617-625.
106
10. Gabius, H.-J. Cell surface glycans: The why and how of their functionality as biochemical signals in lectin-mediated information transfer. Crit. Rev. Immunol. 2006, 26, 43-79. 11. Gabius, H.-J. The magic of the sugar code. Trends Biochem. Sci. 2015, 40, 341. 12. Sasisekharan, R.; Myette, J.R. The sweet science of glycobiology: Complex carbohydrates, molecules that are particularly important for communication among cells, are coming under systematic study. Am. Sci. 2003, 432-441. 13. Lis, H.; Sharon, N. Lectins: Carbohydrate-specific proteins that mediate cellular recognition. Chem. Rev. 1998, 98, 637-674. 14. Lundquist, J.J.; Toone, E.J. The cluster glycoside effect. Chem. Rev. 2002, 102, 555-578. 15. Kiessling, L.L.; Young, T.; Gruber, T.D.; Mortell, K.H. Multivalency in protein–carbohydrate recognition. In Glycoscience, Fraser-Reid, B.; Tatsuta, K.; Thiem, J., Eds. Springer: Berlin, Heidelberg, Germany: 2008; pp 2483-2523. 16. Pieters, R.J. Maximising multivalency effects in protein-carbohydrate interactions. Org. Biomol. Chem. 2009, 7, 2013-2025. 17. Pieters, R.J.; Arnusch, C.J.; Breukink, E. Membrane permeabilization by multivalent anti- microbial peptides. Protein and peptide letters 2009, 16, 736-742. 18. Monsigny, M.; Mayer, R.; Roche, A.-C. Sugar-lectin interactions: Sugar clusters, lectin multivalency and avidity. Carbohydr. Lett. 2000, 4, 35-52. 19. Simons, K.; Ikonen, E. Functional rafts in cell membranes. Nature 1997, 387, 569-572. 20. Sharon, N. Bacterial lectins, cell-cell recognition and infectious disease. FEBS Lett. 1987, 217, 145-157. 21. Alvarez, C.P.; Lasala, F.; Carrillo, J.; Muñiz, O.; Corbí, A.L.; Delgado, R. C-type lectins dc- sign and l-sign mediate cellular entry by ebola virus in cis and in trans. J. Virol. 2002, 76, 6841-6844. 22. Danguy, A.; Camby, I.; Kiss, R. Galectins and cancer. BBA Gen. Subj. 2002, 1572, 285-293. 23. Rabinovich, G.A. Galectins: An evolutionarily conserved family of animal lectins with multifunctional properties; a trip from the gene to clinical therapy. Cell Death Differ. 1999, 6, 711-721. 24. Barondes, S.H.; Cooper, D.N.; Gitt, M.A.; Leffler, H. Galectins. Structure and function of a large family of animal lectins. J. Biol. Chem. 1994, 269, 20807-20810. 25. Leffler, H.; Carlsson, S.; Hedlund, M.; Qian, Y.; Poirier, F. Introduction to galectins. Glycoconj. J. 2004, 19, 433-440. 26. Griffioen, A.W.; Thijssen, V.L. Galectins in tumor angiogenesis. Ann. Transl. Med. 2014, 2. 27. Rabinovich, G.A.; van Kooyk, Y.; Cobb, B.A. Glycobiology of immune responses. Ann. NY. Acad. Sci. 2012, 1253, 1-15. 28. Liu, F.T.; Rabinovich, G.A. Galectins as modulators of tumour progression. Nat. Rev. Cancer 2005, 5, 29-41. 29. D’Haene, N.; Maris, C.; Rorive, S.; Decaestecker, C.; le Mercier, M.; Salmon, I. Galectins and neovascularization in central nervous system tumors. Glycobiology 2014, 24, 892-898. 30. Compagno, D.; Gentilini, L.D.; Jaworski, F.M.; Pérez, I.G.; Contrufo, G.; Laderach, D.J. Glycans and galectins in prostate cancer biology, angiogenesis and metastasis. Glycobiology 2014, 24, 899-906. 31. Le Mercier, M.; Fortin, S.; Mathieu, V.; Kiss, R.; Lefranc, F. Galectins and gliomas. Brain Pathol. 2010, 20, 17-27. 32. Nakahara, S.; Oka, N.; Raz, A. On the role of galectin-3 in cancer apoptosis. Apoptosis 2005, 10, 267-275. 33. Nangia-Makker, P.; Honjo, Y.; Sarvis, R.; Akahani, S.; Hogan, V.; Pienta, K.J.; Raz, A. Galectin-3 induces endothelial cell morphogenesis and angiogenesis. Am. J. Pathol. 2000, 156, 899-909. 34. Dumic, J.; Dabelic, S.; Flögel, M. Galectin-3: An open-ended story. BBA-Gen. Subjects 2006, 1760, 616-635. 35. Hittelet, A.; Legendre, H.; Nagy, N.; Bronckart, Y.; Pector, J.C.; Salmon, I.; Yeaton, P.; Gabius, H.J.; Kiss, R.; Camby, I. Upregulation of galectins‐1 and‐3 in human colon cancer and their role in regulating cell migration. Int. J. Cancer 2003, 103, 370-379. 36. Hsu, D.K.; Dowling, C.A.; Jeng, K.C.G.; Chen, J.T.; Yang, R.Y.; Liu, F.T. Galectin‐3 expression is induced in cirrhotic liver and hepatocellular carcinoma. Int. J. Cancer 1999, 81, 519-526.
107
37. Pugliese, G.; Iacobini, C.; Pesce, C.M.; Menini, S. Galectin-3: An emerging all-out player in metabolic disorders and their complications. Glycobiology 2015, 25, 136-150. 38. Krześlak, A.; Lipińska, A. Galectin-3 as a multifunctional protein. Cell. Mol. Biol. Lett. 2004, 9, 305-328. 39. Takenaka, Y.; Fukumori, T.; Raz, A. Galectin-3 and metastasis. Glycoconj. J. 2002, 19, 543- 549. 40. Moon, B.-K.; Lee, Y.J.; Battle, P.; Jessup, J.M.; Raz, A.; Kim, H.-R.C. Galectin-3 protects human breast carcinoma cells against nitric oxide-induced apoptosis: Implication of galectin-3 function during metastasis. Am. J. Pathol. 2001, 159, 1055-1060. 41. Johnson, K.D.; Glinskii, O.V.; Mossine, V.V.; Turk, J.R.; Mawhinney, T.P.; Anthony, D.C.; Henry, C.J.; Huxley, V.H.; Glinsky, G.V.; Pienta, K.J. Galectin-3 as a potential therapeutic target in tumors arising from malignant endothelia. Neoplasia 2007, 9, 662-670. 42. Nangia-Makker, P.; Nakahara, S.; Hogan, V.; Raz, A. Galectin-3 in apoptosis, a novel therapeutic target. J. bioenerg. biomembr. 2007, 39, 79-84. 43. Mirandola, L.; Nguyen, D.D.; Rahman, R.L.; Grizzi, F.; Yuefei, Y.; Figueroa, J.A.; Jenkins, M.R.; Cobos, E.; Chiriva-Internati, M. Anti-galectin-3 therapy: A new chance for multiple myeloma and ovarian cancer? Int. Rev. Immunol. 2014, 33, 417-427. 44. Salameh, B.A.; Leffler, H.; Nilsson, U.J. 3-(1, 2, 3-triazol-1-yl)-1-thio-galactosides as small, efficient, and hydrolytically stable inhibitors of galectin-3. Bioorg. Med. Chem. Lett. 2005, 15, 3344-3346. 45. Sörme, P.; Qian, Y.; Nyholm, P.G.; Leffler, H.; Nilsson, U.J. Low micromolar inhibitors of galectin‐3 based on 3′‐derivatization of n‐acetyllactosamine. ChemBioChem 2002, 3, 183-189. 46. Hakon, L.; Ulf, J.N. Low-molecular weight inhibitors of galectins. In Galectins and disease implications for targeted therapeutics, ACS: Washington, DC, USA, 2012; Vol. 1115, pp 47- 59. 47. Stowell, S.R.; Arthur, C.M.; Mehta, P.; Slanina, K.A.; Blixt, O.; Leffler, H.; Smith, D.F.; Cummings, R.D. Galectin-1,-2, and-3 exhibit differential recognition of sialylated glycans and blood group antigens. J. Biol. Chem. 2008, 283, 10109-10123. 48. Téllez-Sanz, R.; Garcia-Fuentes, L.; Vargas-Berenguel, A. Human galectin-3 selective and high affinity inhibitors. Present state and future perspectives. Curr. Med. Chem. 2013, 20, 2979-2990. 49. Ahmad, N.; Gabius, H.J.; André, S.; Kaltner, H.; Sabesan, S.; Roy, R.; Liu, B.C.; Macaluso, F.; Brewer, C.F. Galectin-3 precipitates as a pentamer with synthetic multivalent carbohydrates and forms heterogeneous cross-linked complexes. J. Biol. Chem. 2004, 279, 10841-10847. 50. Nieminen, J.; Kuno, A.; Hirabayashi, J.; Sato, S. Visualization of galectin-3 oligomerization on the surface of neutrophils and endothelial cells using fluorescence resonance energy transfer. J. Biol. Chem. 2007, 282, 1374-1383. 51. Ochieng, J.; Green, B.; Evans, S.; James, O.; Warfield, P. Modulation of the biological functions of galectin-3 by matrix metalloproteinases. BBA-Gen. Subjects 1998, 1379, 97-106. 52. Lepur, A.; Salomonsson, E.; Nilsson, U.J.; Leffler, H. Ligand induced galectin-3 protein self- association. J. Biol. Chem. 2012, 287, 21751-21756. 53. Dam, T.K.; Gabius, H.-J.; André, S.; Kaltner, H.; Lensch, M.; Brewer, C.F. Galectins bind to the multivalent glycoprotein asialofetuin with enhanced affinities and a gradient of decreasing binding constants. Biochem. 2005, 44, 12564-12571. 54. Goodman, C.K.; Wolfenden, M.L.; Nangia-Makker, P.; Michel, A.K.; Raz, A.; Cloninger, M.J. Multivalent scaffolds induce galectin-3 aggregation into nanoparticles. Beilstein J. Org. Chem. 2014, 10, 1570-1577. 55. Yang, Y.; Xue, X.C.; Jin, X.F.; Wang, L.J.; Sha, Y.L.; Li, Z.J. Synthesis of multivalent n- acetyl lactosamine modified quantum dots for the study of carbohydrate and galectin-3 interactions. Tetrahedron 2012, 68, 7148-7154. 56. Wolfenden, M.; Cousin, J.; Nangia-Makker, P.; Raz, A.; Cloninger, M. Glycodendrimers and modified elisas: Tools to elucidate multivalent interactions of galectins 1 and 3. Molecules 2015, 20, 7059-7096. 57. Bernardi, A.; Jimenez-Barbero, J.; Casnati, A.; de Castro, C.; Darbre, T.; Fieschi, F.; Finne, J.; Funken, H.; Jaeger, K.-E.; Lahmann, M., et al. Multivalent glycoconjugates as anti-pathogenic agents. Chem. Soc. Rev. 2013, 42, 4709-4727.
108
58. Peri, F. Clustered carbohydrates in synthetic vaccines. Chem. Soc. Rev. 2013, 42, 4543-4556. 59. Branson, T.R.; Turnbull, W.B. Bacterial toxin inhibitors based on multivalent scaffolds. Chem. Soc. Rev. 2013, 42, 4613-4622. 60. Chabre, Y.M.; Roy, R. Multivalent glycoconjugate syntheses and applications using aromatic scaffolds. Chem. Soc. Rev. 2013, 42, 4657-4708. 61. Marradi, M.; Chiodo, F.; Garcia, I.; Penades, S. Glyconanoparticles as multifunctional and multimodal carbohydrate systems. Chem. Soc. Rev. 2013, 42, 4728-4745. 62. Martinez, A.; Ortiz Mellet, C.; Garcia Fernandez, J.M. Cyclodextrin-based multivalent glycodisplays: Covalent and supramolecular conjugates to assess carbohydrate-protein interactions. Chem. Soc. Rev. 2013, 42, 4746-4773. 63. Hatano, K.; Matsuoka, K.; Terunuma, D. Carbosilane glycodendrimers. Chem. Soc. Rev. 2013, 42, 4574-4598. 64. Chen, Y.; Star, A.; Vidal, S. Sweet carbon nanostructures: Carbohydrate conjugates with carbon nanotubes and graphene and their applications. Chem. Soc. Rev. 2013, 42, 4532-4542. 65. Kennedy, D.C.; Grünstein, D.; Lai, C.-H.; Seeberger, P.H. Glycosylated nanoscale surfaces: Preparation and applications in medicine and molecular biology. Chem. Eur. J. 2013, 19, 3794-3800. 66. Sansone, F.; Casnati, A. Multivalent glycocalixarenes for recognition of biological macromolecules: Glycocalyx mimics capable of multitasking. Chem. Soc. Rev. 2013, 42, 4623-4639. 67. Cecioni, S.; Imberty, A.; Vidal, S. Glycomimetics versus multivalent glycoconjugates for the design of high affinity lectin ligands. Chem. Rev. 2014, 115, 525-561. 68. Bojarová, P.; Rosencrantz, R.R.; Elling, L.; Křen, V. Enzymatic glycosylation of multivalent scaffolds. Chem. Soc. Rev. 2013, 42, 4774-4797. 69. Ercolani, G.; Schiaffino, L. Allosteric, chelate, and interannular cooperativity: A mise au point. Angew. Chem. Int. Ed. 2011, 50, 1762-1768. 70. Hunter, C.A.; Anderson, H.L. What is cooperativity? Angew. Chem. Int. Ed. 2009, 48, 7488- 7499. 71. Maierhofer, C.; Rohmer, K.; Wittmann, V. Probing multivalent carbohydrate-lectin interactions by an enzyme-linked lectin assay employing covalently immobilized carbohydrates. Bioorg. Med. Chem. 2007, 15, 7661-7676. 72. Lee, R.; Lee, Y. Affinity enhancement by multivalent lectin–carbohydrate interaction. Glycoconj. J. 2000, 17, 543-551. 73. Profit, A.A.; Lee, T.R.; Lawrence, D.S. Bivalent inhibitors of protein tyrosine kinases. J. Am. Chem. Soc. 1999, 121, 280-283. 74. Schaschke, N.; Matschiner, G.; Zettl, F.; Marquardt, U.; Bergner, A.; Bode, W.; Sommerhoff, C.P.; Moroder, L. Bivalent inhibition of human β-tryptase. Chem. Biol. 2001, 8, 313-327. 75. Rao, J.; Lahiri, J.; Isaacs, L.; Weis, R.M.; Whitesides, G.M. A trivalent system from vancomycin·d-ala-d-ala with higher affinity than avidin·biotin. Science 1998, 280, 708-711. 76. Šimonová, A.; Kupper, C.E.; Böcker, S.; Müller, A.; Hofbauerová, K.; Pelantová, H.; Elling, L.; Křen, V.; Bojarová, P. Chemo-enzymatic synthesis of lacdinac dimers of varying length as novel galectin ligands. J. Mol. Catal. B Enzym. 2014, 101, 47-55. 77. Breloy, I.; Söte, S.; Ottis, P.; Bonar, D.; Grahn, A.; Hanisch, F.-G. O-linked lacdinac-modified glycans in extracellular matrix glycoproteins are specifically phosphorylated at the subterminal glcnac. J. Biol. Chem. 2012, 287, 18275-18286. 78. Kenny, D.T.; Skoog, E.C.; Lindén, S.K.; Struwe, W.B.; Rudd, P.M.; Karlsson, N.G. Presence of terminal n-acetylgalactosamineβ1-4n-acetylglucosamine residues on o-linked oligosaccharides from gastric muc5ac: Involvement in helicobacter pylori colonization? Glycobiology 2012, 22, 1077–1085. 79. Sasaki, N.; Shinomi, M.; Hirano, K.; Ui-Tei, K.; Nishihara, S. Lacdinac (galnacβ1-4glcnac) contributes to self-renewal of mouse embryonic stem cells by regulating leukemia inhibitory factor/stat3 signaling. Stemm Cells 2011, 29, 641-650. 80. Van den Berg, T.K.; Honing, H.; Franke, N.; van Remoortere, A.; Schiphorst, W.; Liu, F.T.; Deelder, A.M.; Cummings, R.D.; Hokke, C.H.; van Die, I. Lacdinac-glycans constitute a parasite pattern for galectin-3-mediated immune recognition. J. Immunol. 2004, 173, 1902- 1907.
109
81. Wuhrer, M.; Koeleman, C.A.M.; Deelder, A.M.; Hokke, C.H. Repeats of lacdinac and fucosylated lacdinac on n-glycans of the human parasite schistosoma mansoni. FEBS J. 2006, 273, 347-361. 82. Hirano, K.; Matsuda, A.; Shirai, T.; Furukawa, K. Expression of lacdinac groups on n-glycans among human tumors is complex. Biomed. Res. Int. 2014, 2014, 1-7. 83. Do, K.-Y.; Do, S.; Cummings, R.D. Differential expression of lacdinac sequences (galnacβ1- 4glcnac-r) in glycoproteins synthesized by chinese hamster ovary and human 293 cells. Glycobiology 1997, 7, 183-194. 84. Fukushima, K.; Satoh, T.; Baba, S.; Yamashita, K. Α1, 2-fucosylated and β-n- acetylgalactosaminylated prostate-specific antigen as an efficient marker of prostatic cancer. Glycobiology 2010, 20, 452-460. 85. Kaya, T.; Kaneko, T.; Kojima, S.; Nakamura, Y.; Ide, Y.; Ishida, K.; Suda, Y.; Yamashita, K. High-sensitivity immunoassay with surface plasmon field-enhanced fluorescence spectroscopy using a plastic sensor chip: Application to quantitative analysis of total prostate-specific antigen and galnacβ1–4glcnac-linked prostate-specific antigen for prostate cancer diagnosis. Anal. Chem. 2015, 87, 1797-1803. 86. Rossez, Y.; Gosset, P.; Boneca, I.G.; Magalhaes, A.; Ecobichon, C.; Reis, C.A.; Cieniewski- Bernard, C.; Joncquel Chevalier Curt, M.; Leonard, R.; Maes, E., et al. The lacdinac-specific adhesin laba mediates adhesion of helicobacter pylori to human gastric mucosa. The Journal of infectious diseases 2014, 210, 1286-1295. 87. Kupper, C.E.; Böcker, S.; Liu, H.L.; Adamzyk, C.; van de Kamp, J.; Recker, T.; Lethaus, B.; Jahnen-Dechent, W.; Neuss, S.; Müller-Newen, G., et al. Fluorescent snap-tag galectin fusion proteins as novel tools in glycobiology. Curr. Pharm. Des. 2013, 19, 5457-5467. 88. Blackmore, P.F.; Eisoldt, S. The neoglycoprotein mannose-bovine serum albumin, but not progesterone, activates t-type calcium channels in human spermatozoa. Mol. Hum. Reprod. 1999, 5, 498-506. 89. Oh, Y.S.; Ahn, H.S.; Gye, M.C. Fucosyl neoglycoprotein binds to mouse epididymal spermatozoa and inhibits sperm binding to the egg zona pellucida. Andrologia 2013, 45, 363- 368. 90. Huang, B.X.; Kim, H.-Y.; Dass, C. Probing three-dimensional structure of bovine serum albumin by chemical cross-linking and mass spectrometry. J. Am. Soc. Mass Spectrom. 2004, 15, 1237-1247. 91. Luyai, A.; Lasanajak, Y.; Smith, D.F.; Cummings, R.D.; Song, X. Facile preparation of fluorescent neoglycoproteins using p-nitrophenyl anthranilate as a heterobifunctional linker. Bioconjug. Chem. 2009, 20, 1618-1624. 92. Unverzagt, C.; André, S.; Seifert, J.; Kojima, S.; Fink, C.; Srikrishna, G.; Freeze, H.; Kayser, K.; Gabius, H.-J. Structure-activity profiles of complex biantennary glycans with core fucosylation and with/without additional α2,3/α2,6 sialylation: Synthesis of neoglycoproteins and their properties in lectin assays, cell binding, and organ uptake. J. Med. Chem. 2002, 45, 478-491. 93. Gabius, H.J.; Bardosi, A. Neoglycoproteins as tools in glycohistochemistry. Prog. Histochem. Cytochem. 1991, 22, 1-63. 94. Stowell, C.P.; Lee, Y.C. Neoglycoproteins the preparation and application of synthetic glycoproteins. In Adv. Carbohydr. Chem. Biochem., Tipson, R.S.; Derek, H., Eds. Academic Press: Waltham, MA, USA, 1980; Vol. 37, pp 225-281. 95. Kerekgyarto, M.; Fekete, A.; Szurmai, Z.; Kerekgyarto, J.; Takacs, L.; Kurucz, I.; Guttman, A. Neoglycoproteins as carbohydrate antigens: Synthesis, analysis, and polyclonal antibody response. Electrophoresis 2013, 34, 2379-2386. 96. Rech, C.; Rosencrantz, R.R.; Křenek, K.; Pelantová, H.; Bojarová, P.; Römer, C.E.; Hanisch, F.-G.; Křen, V.; Elling, L. Combinatorial one-pot synthesis of poly-n-acetyllactosamine oligosaccharides with leloir-glycosyltransferases. Adv. Synth. Catal. 2011, 353, 2492-2500. 97. Kamath, V.P.; Diedrich, P.; Hindsgaul, O. Use of diethyl squarate for the coupling of oligosaccharide amines to carrier proteins and characterization of the resulting neoglycoproteins by maldi-tof mass spectrometry. Glycoconj. J. 1996, 13, 315-319. 98. Tietze, L.F.; Arlt, M.; Beller, M.; Glüsenkamp, K.-H.; Jähde, E.; Rajewsky, M.F. Anticancer agents, 15. Squaric acid diethyl ester: A new coupling reagent for the formation of drug
110
biopolymer conjugates. Synthesis of squaric acid ester amides and diamides. Chem. Ber. 1991, 124, 1215-1221. 99. Tietze, L.F.; Schroeter, C.; Gabius, S.; Brinck, U.; Goerlach-Graw, A.; Gabius, H.J. Conjugation of p-aminophenyl glycosides with squaric acid diester to a carrier protein and the use of the neoglycoprotein in the histochemical detection of lectins. Bioconjug. Chem. 1991, 2, 148-153. 100. Wang, J.W.; Asnani, A.; Auzanneau, F.I. Synthesis of a bsa-lex glycoconjugate and recognition of lex analogues by the anti-lex monoclonal antibody sh1: The identification of a non-cross reactive analogue. Bioorg. Med. Chem. 2010, 18, 7174-7185. 101. Wurm, F.R.; Klok, H.A. Be squared: Expanding the horizon of squaric acid-mediated conjugations. Chem. Soc. Rev. 2013, 42, 8220-8236. 102. Komarova, B.S.; Orekhova, M.V.; Tsvetkov, Y.E.; Beau, R.; Aimanianda, V.; Latge, J.P.; Nifantiev, N.E. Synthesis of a pentasaccharide and neoglycoconjugates related to fungal alpha-(1->3)-glucan and their use in the generation of antibodies to trace aspergillus fumigatus cell wall. Chemistry 2015, 21, 1029-1035. 103. Jahouh, F.; Hou, S.J.; Kovac, P.; Banoub, J.H. Determination of glycation sites by tandem mass spectrometry in a synthetic lactose-bovine serum albumin conjugate, a vaccine model prepared by dialkyl squarate chemistry. Rapid Commun. Mass Spectrom. 2012, 26, 749-758. 104. Jahouh, F.; Xu, P.; Vann, W.F.; Kovac, P.; Banoub, J.H. Mapping the glycation sites in the neoglycoconjugate from hexasaccharide antigen of vibrio cholerae, serotype ogawa and the recombinant tetanus toxin c-fragment carrier. J. Mass Spectrom. 2013, 48, 1083-1090. 105. Kupper, C.E.; Rosencrantz, R.R.; Henssen, B.; Pelantová, H.; Thönes, S.; Drozdova, A.; Křen, V.; Elling, L. Chemo-enzymatic modification of poly-n-acetyllactosamine (lacnac) oligomers and n,n-diacetyllactosamine (lacdinac) based on galactose oxidase treatment. Beilstein J. Org. Chem. 2012, 8, 712-725. 106. Sörme, P.; Kahl-Knutsson, B.; Wellmar, U.; Magnusson, B.-G.; Leffler, H.; Nilsson, U.J. Design and synthesis of galectin inhibitors. Methods Enzymol. 2003, 363, 157-169. 107. Cumpstey, I.; Sundin, A.; Leffler, H.; Nilsson, U.J. C2‐symmetrical thiodigalactoside bis‐ benzamido derivatives as high‐affinity inhibitors of galectin‐3: Efficient lectin inhibition through double arginine–arene interactions. Angew. Chem. 2005, 117, 5240-5242. 108. Sauerzapfe, B.; Namdjou, D.J.; Schumacher, T.; Linden, N.; Křenek, K.; Křen, V.; Elling, L. Characterization of recombinant fusion constructs of human β1,4-galactosyltransferase 1 and the lipase pre-propeptide from staphylococcus hyicus. J. Mol. Catal. B Enzym. 2008, 50, 128- 140. 109. Sauerzapfe, B.; Krenek, K.; Schmiedel, J.; Wakarchuk, W.W.; Pelantova, H.; Kren, V.; Elling, L. Chemo-enzymatic synthesis of poly-n-acetyllactosamine (poly-lacnac) structures and their characterization for cgl2-galectin-mediated binding of ecm glycoproteins to biomaterial surfaces. Glycoconj. J. 2009, 26, 141-159. 110. Chen, R.F. Removal of fatty acids from serum albumin by charcoal treatment. J. Biol. Chem. 1967, 242, 173-181. 111. Cayot, P.; Tainturier, G. The quantification of protein amino groups by the trinitrobenzenesulfonic acid method: A reexamination. Anal. Biochem. 1997, 249, 184-200. 112. Witten, K.G.; Rech, C.; Eckert, T.; Charrak, S.; Richtering, W.; Elling, L.; Simon, U. Glyco- DNA–gold nanoparticles: Lectin-mediated assembly and dual-stimuli response. Small 2011, 7, 1954-1960. 113. Logan, S.M.; Altman, E.; Mykytczuk, O.; Brisson, J.R.; Chandan, V.; Schur, M.J.; St Michael, F.; Masson, A.; Leclerc, S.; Hiratsuka, K., et al. Novel biosynthetic functions of lipopolysaccharide rfaj homologs from helicobacter pylori. Glycobiology 2005, 15, 721-733. 114. Hou, S.J.; Saksena, R.; Kovac, P. Preparation of glycoconjugates by dialkyl squarate chemistry revisited. Carbohydr. Res. 2008, 343, 196-210. 115. Roth, Z.; Yehezkel, G.; Khalaila, I. Identification and quantification of protein glycosylation. Int. J. Carbohydr. Chem. 2012, 2012, 1-10. 116. Wang, H.; Huang, W.; Orwenyo, J.; Banerjee, A.; Vasta, G.R.; Wang, L.-X. Design and synthesis of glycoprotein-based multivalent glyco-ligands for influenza hemagglutinin and human galectin-3. Bioorg. Med. Chem. 2013, 21, 2037-2044.
111
117. Song, X.Z.; Xia, B.Y.; Stowell, S.R.; Lasanajak, Y.; Smith, D.F.; Cummings, R.D. Novel fluorescent glycan microarray strategy reveals ligands for galectins. Chem. Biol. 2009, 16, 36- 47. 118. Ahmad, N.; Gabius, H.J.; Kaltner, H.; Andre, S.; Kuwabara, I.; Liu, F.T.; Oscarson, S.; Norberg, T.; Brewer, C.F. Thermodynamic binding studies of cell surface carbohydrate epitopes to galectins-1,-3, and-7: Evidence for differential binding specificities. Can. J. Chem. 2002, 80, 1096-1104. 119. Rapoport, E.M.; Andre, S.; Kurmyshkina, O.V.; Pochechueva, T.V.; Severov, V.V.; Pazynina, G.V.; Gabius, H.J.; Bovin, N.V. Galectin-loaded cells as a platform for the profiling of lectin specificity by fluorescent neoglycoconjugates: A case study on galectins-1 and-3 and the impact of assay setting. Glycobiology 2008, 18, 315-324. 120. Stowell, S.R.; Dias-Baruffi, M.; Penttila, L.; Renkonen, O.; Nyame, A.K.; Cummings, R.D. Human galectin-1 recognition of poly-n-acetyllactosamine and chimeric polysaccharides. Glycobiology 2004, 14, 157-167. 121. Leppänen, A.; Stowell, S.; Blixt, O.; Cummings, R.D. Dimeric galectin-1 binds with high affinity to alpha 2,3-sialylated and non-sialylated terminal n-acetyllactosamine units on surface-bound extended glycans. J. Biol. Chem. 2005, 280, 5549-5562. 122. Di Virgilio, S.; Glushka, J.; Moremen, K.; Pierce, M. Enzymatic synthesis of natural and c-13 enriched linear poly-n-acetyllactosamines as ligands for galectin-1. Glycobiology 1999, 9, 353-364. 123. Qun, Z.; Cummings, R.D. L-14 lectin recognition of laminin and its promotion of in vitro cell adhesion. Arch. Biochem. Biophys. 1993, 300, 6-17. 124. Lopez-Lucendo, M.F.; Solis, D.; Andre, S.; Hirabayashi, J.; Kasai, K.; Kaltner, H.; Gabius, H.J.; Romero, A. Growth-regulatory human galectin-1: Crystallographic characterisation of the structural changes induced by single-site mutations and their impact on the thermodynamics of ligand binding. J. Mol. Biol. 2004, 343, 957-970. 125. Diehl, C.; Engstrom, O.; Delaine, T.; Hakansson, M.; Genheden, S.; Modig, K.; Leffler, H.; Ryde, U.; Nilsson, U.J.; Akke, M. Protein flexibility and conformational entropy in ligand design targeting the carbohydrate recognition domain of galectin-3. J. Am. Chem. Soc. 2010, 132, 14577-14589. 126. Seetharaman, J.; Kanigsberg, A.; Slaaby, R.; Leffler, H.; Barondes, S.H.; Rini, J.M. X-ray crystal structure of the human galectin-3 carbohydrate recognition domain at 2.1-angstrom resolution. J. Biol. Chem. 1998, 273, 13047-13052. 127. Lobsanov, Y.D.; Gitt, M.A.; Leffler, H.; Barondes, S.H.; Rini, J.M. X-ray crystal structure of the human dimeric s-lac lectin, l-14-ii, in complex with lactose at 2.9-a resolution. J. Biol. Chem. 1993, 268, 27034-27038. 128. Öberg, C.T.; Noresson, A.-L.; Leffler, H.; Nilsson, U.J. Synthesis of 3-amido-3-deoxy-β-d- talopyranosides: All-cis-substituted pyranosides as lectin inhibitors. Tetrahedron 2011, 67, 9164-9172. 129. Pieters, R.J. Inhibition and detection of galectins. ChemBioChem 2006, 7, 721-728. 130. van Hattum, H.; Branderhorst, H.M.; Moret, E.E.; Nilsson, U.J.; Leffler, H.; Pieters, R.J. Tuning the preference of thiodigalactoside- and lactosamine-based ligands to galectin-3 over galectin-1. J. Med. Chem. 2013, 56, 1350-1354. 131. Marchiori, M.F.; Pires Souto, D.E.; Oliveira Bortot, L.; Francisco Pereira, J.; Kubota, L.T.; Cummings, R.D.; Dias-Baruffi, M.; Carvalho, I.; Campo, V.L. Synthetic 1,2,3-triazole-linked glycoconjugates bind with high affinity to human galectin-3. Bioorg. Med. Chem. 2015, 23, 3414-3425. 132. André, S.; Sansone, F.; Kaltner, H.; Casnati, A.; Kopitz, J.; Gabius, H.J.; Ungaro, R. Calix[n]arene-based glycoclusters: Bioactivity of thiourea-linked galactose/lactose moieties as inhibitors of binding of medically relevant lectins to a glycoprotein and cell-surface glycoconjugates and selectivity among human adhesion/growth-regulatory galectins. ChemBioChem 2008, 9, 1649-1661. 133. André, S.; Grandjean, C.; Gautier, F.M.; Bernardi, S.; Sansone, F.; Gabius, H.J.; Ungaro, R. Combining carbohydrate substitutions at bioinspired positions with multivalent presentation towards optimising lectin inhibitors: Case study with calixarenes. Chem. Commun. 2011, 47, 6126-6128.
112
134. André, S.; Pieters, R.J.; Vrasidas, I.; Kaltner, H.; Kuwabara, I.; Liu, F.T.; Liskamp, R.M.; Gabius, H.J. Wedgelike glycodendrimers as inhibitors of binding of mammalian galectins to glycoproteins, lactose maxiclusters, and cell surface glycoconjugates. ChemBioChem 2001, 2, 822-830. 135. Song, L.; Tang, J.W.; Owusu, L.; Sun, M.Z.; Wu, J.; Zhang, J. Galectin-3 in cancer. Clin. Chim. Acta. 2014, 431, 185-191. 136. Saussez, S.; Lorfevre, F.; Lequeux, T.; Laurent, G.; Chantrain, G.; Vertongen, F.; Toubeau, G.; Decaestecker, C.; Kiss, R. The determination of the levels of circulating galectin-1 and -3 in hnscc patients could be used to monitor tumor progression and/or responses to therapy. Oral Oncol. 2008, 44, 86-93. 137. Sakaki, M.; Oka, N.; Nakanishi, R.; Yamaguchi, K.; Fukumori, T.; Kanayama, H.O. Serum level of galectin-3 in human bladder cancer. J. Med. Invest. 2008, 55, 127-132. 138. Iurisci, I.; Tinari, N.; Natoli, C.; Angelucci, D.; Cianchetti, E.; Iacobelli, S. Concentrations of galectin-3 in the sera of normal controls and cancer patients. Clin. Cancer Res. 2000, 6, 1389- 1393. 139. Sörme, P.; Arnoux, P.; Kahl-Knutsson, B.; Leffler, H.; Rini, J.M.; Nilsson, U.J. Structural and thermodynamic studies on cation-π interactions in lectin-ligand complexes: High-affinity galectin-3 inhibitors through fine-tuning of an arginine-arene interaction. J. Am. Chem. Soc. 2005, 127, 1737-1743. 140. Nangia-Makker, P.; Hogan, V.; Honjo, Y.; Baccarini, S.; Tait, L.; Bresalier, R.; Raz, A. Inhibition of human cancer cell growth and metastasis in nude mice by oral intake of modified citrus pectin. J. Natl. Cancer Inst. 2002, 94, 1854-1862. 141. Iurisci, I.; Cumashi, A.; Sherman, A.A.; Tsvetkov, Y.E.; Tinari, N.; Piccolo, E.; D’Egidio, M.; Adamo, V.; Natoli, C.; Rabinovich, G.A., et al. Synthetic inhibitors of galectin-1 and -3 selectively modulate homotypic cell aggregation and tumor cell apoptosis. Anticancer Res. 2009, 29, 403-410. 142. Michel, A.K.; Nangia-Makker, P.; Raz, A.; Cloninger, M.J. Lactose-functionalized dendrimers arbitrate the interaction of galectin-3/muc1 mediated cancer cellular aggregation. ChemBioChem 2014, 15, 2106-2112. 143. Prasanphanich, N.S.; Song, X.; Heimburg-Molinaro, J.; Luyai, A.E.; Lasanajak, Y.; Cutler, C.E.; Smith, D.F.; Cummings, R.D. Intact reducing glycan promotes the specific immune response to lacto-n-neotetraose-bsa neoglycoconjugates. Bioconjug. Chem. 2015. 144. Mencke, A.J.; Wold, F. Neoglycoproteins. Preparation and in vivo clearance of serum albumin derivatives containing ovalbumin oligosaccharides. J. Biol. Chem. 1982, 257, 14799-14805. 145. Vrasidas, I.; Andre, S.; Valentini, P.; Bock, C.; Lensch, M.; Kaltner, H.; Liskamp, R.M.; Gabius, H.J.; Pieters, R.J. Rigidified multivalent lactose molecules and their interactions with mammalian galectins: A route to selective inhibitors. Org. Biomol. Chem. 2003, 1, 803-810. 146. Parera Pera, N.; Branderhorst, H.M.; Kooij, R.; Maierhofer, C.; van der Kaaden, M.; Liskamp, R.M.; Wittmann, V.; Ruijtenbeek, R.; Pieters, R.J. Rapid screening of lectins for multivalency effects with a glycodendrimer microarray. ChemBioChem 2010, 11, 1896-1904.
113
114
Appendix
Supporting Information for Chapter 2: Binding characteristics of galectin-3 fusion proteins – Influence of truncation and fusion
Supporting Information for Chapter 3: Neo-glycoproteins as novel ligands for human galectin-3 –Albumin as carrier for multivalent presentation of non-biotinylated and 6- biotinylated tetrasaccharides to gain high-affinity ligands
115
116
Supporting Information for Chapter 2: Binding characteristics of galectin-3 fusion proteins – Influence of truncation and fusion Parts of this chapter were published in: S. Böcker, L. Elling, Glycobiology 2017, 27, 457-468.
Size exclusion chromatography and molecular mass determination of galectin-3 protein constructs
Figure S2.1: Calibration of size exclusion chromatography column
A: Calibration of the size exclusion chromatography column TSKgel® G3000SWXL (Tosoh bioscience, Stuttgart, Germany) for determination of molecular masses was performed with different protein standards: mixture 1 (light blue line): horse ferritin (440 kDa), rabbit aldolase (161 kDa), chymotrypsinogen A (25.6 kDa); mixture 2 (green line): bovine thyroglobulin (670 kDa), bovine γ- globulin (158k kDa), egg albumin (44.3 kDa), horse/equine myoglobin (17 kDa), vitamin B12 (1.4 kDa); mixture 3 (red line): bovine catalase (240 kDa), bovine albumin (66.4 kDa), horse/equine myoglobin (17 kDa); mixture 4 (dark blue line): egg albumin (44.3 kDa), cytochrome C (12.4 kDa). The column was run with the elution buffer (0.05 M Phosphate, 0.15 M NaCl, pH 7) at a flow rate of 0.5 mL/min. 2 B: Calibration plot for Kav over logM and fit with y = -0.4894x + 1.4482 (R =0.9941) for protein standards horse ferritin (440 kDa), rabbit aldolase (161 kDa), bovine albumin (66.4 kDa), egg albumin
117
(44.3 kDa), horse/equine myoglobin (17 kDa) and cytochrome C (12.4 kDa). The void volume Vo and the total volume Vt were determined using blue dextran and EDTA, respectively. 푉푒 − 푉표 퐾푎푣 = 푉푒 = elution volume; 푉표 = void volume (6.2 mL); 푉푡 = total volume (12.45 mL) 푉푡 − 푉표 C: Chromatogram of sample buffer showed a clear peak. The containing EDTA eluted at about 12.45 mL.
Figure S2.2: Elution profiles of Gal-3 fusion proteins in size exclusion chromatography
Size exclusion chromatograms are shown for Gal-3 constructs fused to His6-tag (H6, A), H6-SNAP-tag
(B), H6-YFP (C) and H6-SNAP-tag-YFP (D). The peak at 12.5 mL is caused by EDTA in sample buffer. Full-length Gal-3 (blue line), Gal-3(Δ1-62) (green line) as well as Gal-3(Δ1-116) (cyan line) show similar elution profiles. All protein constructs eluted as a main peak before the buffer peak containing EDTA (Figure S1C) which served as an internal standard. In addition, small shoulders at lower elution volumes occurred in the elution profile of SNAP-tag fused Gal-3 proteins (B and D). However, the calculated molecular masses of these proteins do not fit to the corresponding dimers. Since signal intensities are far below the main peaks we considered these as not significant.
118
Table S2.1: Calculated molecular masses (M) of Gal-3 fusion proteins determined by size exclusion chromatography (SEC) and SDS-PAGE vs. theoretical molecular masses
Theoretical SEC SDS-PAGE Protein M [kDa] M [kDa] M [kDa]
H6Gal-3 28.02 25.5 ± 0.56 31.8 ± 0.17
H6Gal-3(Δ1-62) 21.77 20.6 ± 1.00 26.9 ± 0.03
H6Gal-3(Δ1-116) 16.84 14.4 ± 0.95 18.2 ± 0.14
H6SGal-3 46.85 43.6 ± 1.93 47.9 ± 0.15
H6SGal-3(Δ1-62) 41.00 39.7 ± 2,81 42.5 ± 0.08
H6SGal-3(Δ1-116) 36.09 32.2 ± 2.99 36.8 ± 0.22
H6YGal-3 55.29 48,5 ± 3.00 56.8 ± 0.45
H6YGal-3(Δ1-62) 49.43 43.6 ± 3.47 52.4 ± 0.08
H6YGal-3(Δ1-116) 44.53 37.0 ± 3.11 48.0 ± 0.00
H6SYGal-3 74.20 79.0 ± 4.21 76.0 ± 0.43
H6SYGal-3(Δ1-62) 68.34 72.4 ± 5.77 72.3 ± 0.52
H6SYGal-3(Δ1-116) 63.43 63.9 ± 5.64 68.3 ± 1.13
Binding of galectin-3 fusion proteins to asialofetuin
Figure S2.3: Comparison of IMAC and IMAC/LAC purified H6SYGal-3 Galectins were purified by IMAC (●) and IMAC followed by lactose affinity chromatography (LAC) (●), respectively. Binding to ASF (A) and inhibition of binding to ASF by LacNAc (B) is shown for
H6SYGal-3. Similar absolute binding signals were measured.
119
Figure S2.4: Fusion protein lacking Gal-3 in ELISA assay
H6SY (▼), a negative control for Gal-3 fusion proteins, did not bind to immobilized ASF. In contrast,
H6SYGal-3 (●) showed full ASF binding.
Table S2.2: Calculated values for maximum binding and apparent Kd for Gal-3 fusion proteins binding to ASF in ELISA and SPR experiments
The dissociation constant Kd indicates the galectin concentration at half maximal saturation (i.e., at half maximal Gal-3 binding to ASF). Low Kd values signify a strong binding and are measures for high affinity. In case of SPR, maximal binding signals were calculated per molarity.
ELISA SPR Maximal binding Protein Maximal binding Apparent Apparent signal [µRIU], signal Kd [µM] Kd [µM] normalized
H6Gal-3 0.69 ± 0.05 5.56 ± 1.35 1.43 ± 0.18 32.46 ± 7.40
H6Gal-3(Δ1-62) 0.95 ± 0.03 4.02 ± 0.51 1.27 ± 0.18 40.37 ± 9.58
H6Gal-3(Δ1-116) 0.42 ± 0.05 26.81 ± 5.79 1.01 ± 0.31 60.02 ± 28.4
H6SGal-3 1.13 ± 0.05 0.70 ± 0.16 0.64 ± 0.02 8.56 ± 0.65
H6SGal-3(Δ1-62) 1.08 ± 0.03 0.16 ± 0.04 0.62 ± 0.05 10.40 ± 2.13
H6SGal-3(Δ1-116) 1.11 ± 0.03 0.96 ± 0.14 0.58 ± 0.05 5.55 ± 1.60
H6YGal-3 0.58 ± 0.06 4.19 ± 1.52 0.79 ± 0.01 12.79 ± 0.43
H6YGal-3(Δ1-62) 0.59 ± 0.04 5.28 ± 1.34 0.88 ± 0.03 27.01 ± 2.08
H6YGal-3(Δ1-116) 0.63 ± 0.03 5.18 ± 0.87 1.02 ± 0.08 37.22 ± 5.42
H6SYGal-3 1.13 ± 0.08 3.03 ± 0.87 0.41 ± 0.03 9.89 ± 2.33
H6SYGal-3(Δ1-62) 1.01 ± 0.03 1.61 ± 0.19 0.50 ± 0.02 12.92 ± 1.45
H6SYGal-3(Δ1-116) 1.01 ± 0.04 2.26 ± 0.34 0.78 ± 0.06 20.29 ± 3.38
120
association dissociation association dissociation A B H6Gal-3 H6Gal-3(Δ1-62)
association dissociation association dissociation C D H6Gal-3(Δ1-116) H6SGal-3
association dissociation association dissociation E F H6SGal-3(Δ1-62) H6SGal-3(Δ1-116)
association dissociation association dissociation G H H6YGal-3 H6YGal-3(Δ1-62)
121
association dissociation association dissociation I J H6YGal-3(Δ1-116) H6SYGal-3
K L association dissociation association dissociation
H6SYGal-3(Δ1-62) H6SYGal-3(Δ1-116)
Figure S2.5: SPR-sensorgrams for the binding of Gal-3 protein constructs to immobilized ASF Seven different concentrations (top to bottom: 40, 20, 6.67, 2.22, 0.74, 0.25 and 0.08 µM) of Gal-3 protein constructs were applied in SPR measurements. Response values at equilibrium at about 60 to 140 s (defined using software Scrubber2) were used for non-linear regression (ligand binding model, SigmaPlot). Association phase was between 0 and 150 s, after 150 s the dissociation phase started, as indicated by the vertical line. The gathered sensorgrams are analogous to previously published Gal-3 SPR data and may therefore be considered valuable [65].
122
Peak #1 RT: 4.04 min NL: 6.43E6 TR: 5.0% T: {0;0}... Supporting109 Information for Chapter 3: Neo-glycoproteins as novel ligands% 255.2 for human galectin-3 –Albumin as carrier for multivalent presentation of non-biotinylated and 6-biotinylated tetrasaccharides75 to gain high-affinity ligands Parts of this chapter were published in: S. Böcker, D. Laaf, L. Elling, Biomolecules 2015, 5, 1671-1696. 50 S. Böcker,948.3 L. Elling, Bioengineering 2017, 4(2), 31.
Rel. Intensity [%] Intensity Rel. 473.7 628.2 25 Product characterization
0 Peak #1 RT: 5.23 min NL: 1.51E7 TR: 5.0% T: {0;0}... A 109 % 255.2 948.3 (a) Peak #1 RT: 10.48 min NL: 2.34E4 TR: 5.0% T: {0;0}... 109 % 296.9 75 502.7 780.1
7550 Rel. Intensity [%] Intensity Rel. 5025
473.7 628.2 Rel. [%] Intensity
250 Peak #1 RT: 5.25 min NL: 1.34E7 TR: 5.0% T: {0;0}... 109 383.0 % 255.2 0 Peak #1 RT: 11.97 min NL: 3.10E4 TR: 5.0% T: {0;0}... B 109 % 988.7 75
238.3 75 50
948.3 Rel. Intensity [%] Intensity Rel. 50 25 302.8
473.6 628.1 Rel. [%] Intensity
25 m/z 0 100145.4 500 750 1.000 1.250 1.500 1.750 2.000 m/z 0 Peak #1 RT: 11.99 min NL: 1.68E4 TR: 5.0% T: {0;0}... 109 % 424.2
75 157.1
50 Rel. [%] Intensity 25
604.2 m/z 0 100 500 750 1.000 1.250 1.500 1.750 2.000 m/z 123
C Peak #1 RT: 29.62 min NL: 6.02E6 T: {0;0} - c ESI corona sid=100.00 det=1506.00 Full ms [ 100.00-2000.00] 107 % 650.6
80
60
40
20 301.2 1301.8 227.0 355.3 981.7 1954.4 574.2 723.4 879.6 m/z 0 100 400 600 800 1.000 1.200 1.400 1.600 1.800 2.000 D Peak #1 RT: 29.72 min NL: 8.58E6 T: {0;0} - c ESI corona sid=100.00 det=1506.00 Full ms [ 100.00-2000.00] 107 % 671.0
80
60
40
20 301.2 1342.7 227.0 355.1 920.5 1022.6 615.4 1811.7 m/z 0 100 400 600 800 1.000 1.200 1.400 1.600 1.800 2.000 E Peak #1 RT: 5.63 min NL: 3.97E7 T: {0;0} + c ESI corona sid=100.00 det=1506.00 Full ms [ 100.00-2000.00] 107 % 602.3
80
60
40
195.8 372.0 20 485.0 719.2
881.1 1203.4 m/z 0 100 400 600 800 1.000 1.200 1.400 1.600 1.800 2.000
124
F Peak #1 RT: 5.53 min NL: 4.74E7 T: {0;0} + c ESI corona sid=100.00 det=1506.00 Full ms [ 100.00-2000.00] 107 % 622.8
80 Peak #1 RT: 4.23 min NL: 1.78E5 TR: 5.0% T: {0;0}... 109 % 485.9 Peak #1 RT: 3.78 min NL: 9.88E4 TR: 5.0% T: {0;0}... 60 109 % 255.3 75 40
75 50 20 372.1 485.0 972.6 195.9 Rel. [%] Intensity 414.0 760.2 1244.4 50 359.5 907.2 25 250.2 529.8 1366.2 m/z 0 145.3 628.4643.7 1438.4 100 400 600 800 1.000 1.200 1.400 1588.71.600 1.800 2.000 Rel. [%] Intensity 810.3 1164.0 1331.9 1817.4 25 0 816.3 Peak #1 RT: 5.02 min NL: 4.88E5 TR: 5.0% T: {0;0}... G 109 1032.8 % 972.3 0 Peak #1 RT: 3.91 min NL: 3.12E5 TR: 5.0% T: {0;0}... 109 % 506.4 75
5075 Rel. [%] Intensity 325.1 1013.4 2550 255.4 485.7 613.4 1489.8 1672.2 Rel. [%] Intensity 283.3 0 25 325.2 Peak #1 RT: 5.04 min NL: 6.71E4 TR: 5.0% T: {0;0}... 109 627.2 % 972.3 1150.2 787.9 906.4 1573.9 0 325.2 H Peak #1 RT: 3.93 min NL: 5.76E5 TR: 5.0% 1547.7 T: {0;0}... 109 221.3 75 % 506.3
50 75 1476.9
Rel. [%] Intensity 1608.8 25 691.1 1185.0 1836.9 50 466.5 1665.7 646.2 1294.1 1013.6 m/z Rel. [%] Intensity 0 25100 500 750 1.000 1.250 1.500 1.750 2.000 283.4 m/z 1219.4 1684.0 m/z 0 100 500 750 1.000 1.250 1.500 1.750 2.000 m/z
125
I Peak #1 RT: 4.35 min NL: 1.64E6 T: {0;0} - c ESI corona sid=100.00 det=1506.00 Full ms [ 100.00-2000.00] 107 % 662.4
80
60
40
20 325.0
1325.5 255.2 417.0 972.4 1479.1 1677.1 1802.9 574.2 1140.3 m/z 0 100 400 600 800 1.000 1.200 1.400 1.600 1.800 2.000 J Peak #1 RT: 4.90 min NL: 2.60E6 T: {0;0} - c ESI corona sid=100.00 det=1506.00 Full ms [ 100.00-2000.00] 107 % 683.0
80
60
325.0
40
249.1 20 1366.6 980.5 424.0 183.1 555.1 921.4 1811.6 m/z 0 100 400 600 800 1.000 1.200 1.400 1.600 1.800 2.000 Figure S3.1: MS spectra of products 4 (A), 5 (B), 9 (C), 10 (D), 13 (E), 14 (F), 16 (G) and 17 (H), 18 (I) and 19 (J) Via LC-MS using electrospray ionization [M-H]- ions were analyzed using quadrupole mass analyzer. 4 [M-H]- = 949.97 m/z 5 [M-H]- = 990.01 m/z 9 [M-H]- = 1301.8 m/z ([M-2H]2- = 650.6 m/z); calculated m/z: 1302.6 10 [M-H]- = 1342.7 m/z ([M-2H]2- = 671.0 m/z); calculated m/z: 1343.6 13 [M+H]+ = 1204.4 m/z ([M+2H]2+ = 602.3 m/z); calculated m/z: 1202.5 14 [M+H]+ = 1244.4 m/z ([M+2H]2+ = 622.8 m/z); calculated m/z: 1243.5 16 [M-H]- = 973.95 m/z 17 [M-H]- = 1013.99 m/z ([M-2H]2- = 506.3 m/z) 18 [M-H]- = 1325.5 m/z ([M-2H]2- = 662.4 m/z); calculated m/z: 1326.5 19 [M-H]- = 1366.6 m/z ([M-2H]2- = 683.0 m/z); calculated m/z: 1367.6
126
A B
Figure S3.2: TNBSA-assay calibration curves with lysine hydrochloride (A) and bovine serum albumin (B) Chromogenic products (N-trinitrophenylamine) resulting from reaction of trinitrobenzene sulfonic acid (TNBSA) with primary amine (ε-amino group of lysine) were quantified in microplate reader at 420 nm. For calculating the modification degrees of the neo-glycoproteins calibration with bovine serum albumin (B) was used.
Galectin binding to neo-glycoproteins
Table S3.1: Binding signals of Gal-1 and Gal-3 to immobilized neo-glycoproteins 20a-i, 21a-i, 22a-f and 23a-f as well as asialofetuin (ASF) and bovine serum albumin (BSA) Comparison of Gal-1 and Gal-3 binding is shown with standard deviation of at least 9 measured values. Significant higher binding of Gal-3 compared to Gal-1 is observed for neo-glycoproteins presenting LacNAc-LacNAc, LacDiNAc-LacNAc, 6-biotin LacNAc-LacNAc and 6-biotin LacdiNAc- LacNAc, respectively.
Binding signal Binding signal Ligand Ligand Galectin-1 Galectin-3 Galectin-1 Galectin-3 BSA 0.00 ± 0.01 0.00 ± 0.01 ASF 0.36 ± 0.06 0.33 ± 0.11 20a 0.02 ± 0.02 0.23 ± 0.04 21a 0.01 ± 0.01 0.21 ± 0.05 20b 0.02 ± 0.03 0.20 ± 0.04 21b 0.01 ± 0.01 0.39 ± 0.09 20c 0.06 ± 0.02 0.49 ± 0.04 21c 0.01 ± 0.02 0.54 ± 0.08 20d 0.13 ± 0.04 0.57 ± 0.05 21d 0.02 ± 0.02 0.68 ± 0.05 20e 0.23 ± 0.06 0.62 ± 0.05 21e 0.05 ± 0.03 0.74 ± 0.07 20f 0.31 ± 0.07 0.68 ± 0.04 21f 0.13 ± 0.05 0.85 ± 0.05 20g 0.34 ± 0.07 0.75 ± 0.06 21g 0.14 ± 0.06 0.95 ± 0.04 20h 0.34 ± 0.08 0.79 ± 0.05 21h 0.11 ± 0.06 0.90 ± 0.05 20i 0.40 ± 0.08 0.89 ± 0.05 21i 0.15 ± 0.06 0.95 ± 0.05 22a 0.01 ± 0.02 0.20 ± 0.08 23a 0.01 ± 0.01 0.48 ± 0.11 22b 0.11 ± 0.05 0.65 ± 0.07 23b 0.05 ± 0.03 0.75 ± 0.08 22c 0.14 ± 0.06 0.70 ± 0.07 23c 0.10 ± 0.05 0.79 ± 0.07 22d 0.18 ± 0.06 0.72 ± 0.07 23d 0.14 ± 0.06 0.84 ± 0.05 22e 0.19 ± 0.06 0.74 ± 0.08 23e 0.15 ± 0.06 0.91 ± 0.05 22f 0.22 ± 0.06 0.78 ± 0.09 23f 0.15 ± 0.05 0.91 ± 0.06
127
Table S3.2: Kd values and relative potencies of Gal-3 bound to neo-glycoproteins 20a-i, 21a-i, 22a-f and 23a-f and asialofetuin (ASF)
Apparent Kd in [µM] Gal-3 in ELISA-type binding assay to immobilized neo-glycoproteins (5 pmol) and respective standard deviations of at least 9 measured data are shown. Values were calculated by B ∙x data fitting using equation for one site saturation (y = max ). Potencies were calculated in relation to Kd+x ASF and additionally per glycan. Binding affinity of Gal-3 increases with increasing modification densities of neo-glycoproteins, more pronounced for (6-biotin) LacdiNAc-LacNAc conjugated BSA (21a-i, 23a-f).
Ligand Apparent Kd (µM) Relative potency Relative potency per glycan 20a 0.86 ± 0.20 1.46 ± 0.34 0,91 ± 0,21 20b 0.97 ± 0.18 1.30 ± 0.24 0,34 ± 0,06 20c 0.69 ± 0.17 1.82 ± 0.45 0,30 ± 0,07 20d 0.63 ± 0.13 2.01 ± 0.41 0,20 ± 0,04 20e 0.52 ± 0.11 2.43 ± 0.50 0,17 ± 0,03 20f 0.30 ± 0.06 4.16 ± 0.87 0,21 ± 0,04 20g 0.21 ± 0.06 5.98 ± 1.73 0,25 ± 0,07 20h 0.18 ± 0.05 7.06 ± 2.02 0,27 ± 0,08 20i 0.11 ± 0.03 11.05 ± 3.06 0,38 ± 0,11 21a 0.33 ± 0.11 3.86 ± 1.32 2,27 ± 0,78 21b 0.46 ± 0.13 2.73 ± 0.75 1,24 ± 0,34 21c 0.45 ± 0.09 2.79 ± 0.59 0,47 ± 0,10 21d 0.23 ± 0.06 5.45 ± 1.35 0,62 ± 0,15 21e 0.08 ± 0.02 15.35 ± 3.61 1,09 ± 0,26 21f 0.03 ± 0.00 42.00 ± 5.60 1,97 ± 0,26 21g 0.04 ± 0.01 30.51 ± 4.43 1,25 ± 0,18 21h 0.03 ± 0.00 46.49 ± 6.35 1,86 ± 0,25 21i 0.03 ± 0.00 47.55 ± 3.41 1,73 ± 0,12 22a 0.63 ± 0.16 2.00 ± 0.52 4.14 ± 1.08 22b 0.42 ± 0.09 2.98 ± 0.64 0.91 ± 0.20 22c 0.36 ± 0.07 3.51 ± 0.68 0.55 ± 0.11 22d 0.32 ± 0.07 3.96 ± 0.86 0.42 ± 0.09 22e 0.27 ± 0.06 4.67 ± 0.97 0.36 ± 0.07 22f 0.22 ± 0.04 5.66 ± 1.07 0.40 ± 0.08 23a 0.30 ± 0.13 4.16 ± 1.82 14.37 ± 6.28 23b 0.25 ± 0.05 5.11 ± 1.08 2.26 ± 0.48 23c 0.12 ± 0.03 10.83 ± 2.33 2.52 ± 0.54 23d 0.07 ± 0.02 17.94 ± 4.10 2.90 ± 0.66 23e 0.05 ± 0.01 23.26 ± 5.17 2.54 ± 0.56 23f 0.05 ± 0.01 27.30 ± 6.53 2.46 ± 0.59 ASF 1.26 ± 0.25 1.00 ± 0.20 0.11 ± 0.02
128
Figure S3.3: Binding signals of Gal-3 for concentrations of 0 to 0.1 µM to (A) LacNAc-LacNAc conjungated BSA (20a-i), (B) LacDiNAc-LacNAc conjugated BSA (21a-i), (C) 6-biotin LacNAc- LacNAc conjungated BSA (22a-i), (D) 6-biotin LacDiNAc-LacNAc conjugated BSA (23a-i) Data were obtained for all neo-glycoproteins in the Gal-3 concentration range of 0.05 to 10 µM. Moreover, binding signals of Gal-3 at concentrations below 0.05 µM were detectable for 20i and 21d- i, for 22a-f and 23a-f until 0.025 µM and for 23d-f below 0.025 µM.
129
Table S3.3: Binding signals of Gal-3 per glycan of immobilized neo-glycoproteins 20a-i (A), 21a-i (B), 22a-f (C) and 23a-f (D) in relation to 20d, 21d, 22d and 23d, respectively Binding signals at different Gal-3 concentrations are related to one binding site of the neo-glycoproteins and the relative potencies to 20d, 21d, 22d or 23d are given with standard deviations. For Gal-3 concentrations below 0.05 µM binding to 21a-c and 23a-c was not detectable (n.d.), setting binding signal per ligand for 21d and 23d (and also 20d and 22d) to 1.00.
Galectin-3 Relative binding signal per glycan A [µM] 20a 20b 20c 20d 20e 20f 20g 20h 20i 10 2.76 ± 0.35 1.13 ± 0.13 1.37 ± 0.11 1.00 ± 0.08 0.73 ± 0.04 0.57 ± 0.03 0.47 ± 0.02 0.46 ± 0.02 0.43 ± 0.02 5 2.80 ± 0.39 1.15 ± 0.09 1.38 ± 0.10 1.00 ± 0.08 0.74 ± 0.04 0.56 ± 0.03 0.47 ± 0.02 0.46 ± 0.02 0.42 ± 0.02 1 2.53 ± 0.42 0.95 ± 0.18 1.41 ± 0.12 1.00 ± 0.10 0.76 ± 0.06 0.62 ± 0.04 0.55 ± 0.04 0.53 ± 0.03 0.54 ± 0.03 0.5 2.34 ± 0.39 0.88 ± 0.16 1.27 ± 0.19 1.00 ± 0.12 0.82 ± 0.08 0.77 ± 0.05 0.72 ± 0.05 0.73 ± 0.05 0.72 ± 0.04 0.25 1.60 ± 0.70 0.67 ± 0.25 1.10 ± 0.47 1.00 ± 0.31 0.96 ± 0.17 1.27 ± 0.09 1.42 ± 0.11 1.44 ± 0.11 1.56 ± 0.07 0.1 1.62 ± 0.96 0.94 ± 0.42 1.03 ± 0.62 1.00 ± 0.58 0.83 ± 0.36 1.32 ± 0.59 1.53 ± 0.98 2.26 ± 1.24 3.70 ± 0.64 0.05 3.58 ± 2.87 1.70 ± 1.78 0.94 ± 1.49 1.00 ± 0.97 0.76 ± 0.46 1.48 ± 0.72 1.43 ± 0.70 1.09 ± 0.81 3.98 ± 2.86
B Galectin-3 Relative binding signal per glycan [µM] 21a 21b 21c 21d 21e 21f 21g 21h 21i 10 1.35 ± 0.21 2.14 ± 0.28 1.18 ± 0.06 1.00 ± 0.06 0.68 ± 0.04 0.48 ± 0.02 0.45 ± 0.01 0.44 ± 0.02 0.40 ± 0.01 5 1.41 ± 0.26 2.18 ± 0.37 1.16 ± 0.10 1.00 ± 0.06 0.66 ± 0.04 0.47 ± 0.03 0.45 ± 0.02 0.42 ± 0.02 0.40 ± 0.01 1 1.62 ± 0.36 2.27 ± 0.52 1.18 ± 0.10 1.00 ± 0.07 0.68 ± 0.07 0.52 ± 0.03 0.51 ± 0.02 0.47 ± 0.03 0.45 ± 0.02 0.5 1.33 ± 0.40 1.62 ± 0.49 0.86 ± 0.10 1.00 ± 0.11 0.70 ± 0.09 0.55 ± 0.03 0.52 ± 0.02 0.48 ± 0.02 0.47 ± 0.01 0.25 0.75 ± 0.33 0.99 ± 0.36 0.59 ± 0.08 1.00 ± 0.15 0.86 ± 0.09 0.65 ± 0.04 0.63 ± 0.02 0.58 ± 0.04 0.56 ± 0.02 0.1 1.09 ± 0.62 1.05 ± 0.69 0.75 ± 0.23 1.00 ± 0.48 2.34 ± 0.25 2.07 ± 0.15 1.89 ± 0.15 1.76 ± 0.11 1.67 ± 0.09 0.05 0.77 ± 0.75 0.20 ± 0.93 0.35 ± 0.18 1.00 ± 0.58 3.26 ± 0.92 4.92 ± 0.85 3.96 ± 0.61 4.89 ± 0.55 4.14 ± 0.32 0.025 n.d. n.d. n.d. 1.00 ± 0.70 2.24 ± 0.90 6.16 ± 0.43 4.43 ± 1.48 6.50 ± 0.74 6.21 ± 2.31 0.01 n.d. n.d. n.d. 1.00 ± 5.58 10.01 ± 14.78 58.74 ± 15.82 42.18 ± 20.69 46.97 ± 15.05 66.87 ± 18.31 0.005 n.d. n.d. n.d. 1.00 ± 4.58 24.34 ± 13.11 115.27 ± 8.94 63.84 ± 7.09 109.12 ± 15.02 104.96 ± 11.61
130
C Galectin-3 Relative binding signal per glycan [µM] 22a 22b 22c 22d 22e 22f 10 5.57 ± 0.66 2.71 ± 0.60 1.47 ± 0.69 1.00 ± 0.31 0.74 ± 0.48 0.69 ± 0.41 5 5.64 ± 0.64 2.68 ± 0.59 1.45 ± 0.66 1.00 ± 0.45 0.74 ± 0.43 0.68 ± 0.40 1 5.38 ± 1.04 2.60 ± 0.87 1.43 ± 0.95 1.00 ± 0.88 0.75 ± 1.06 0.72 ± 1.17 0.5 3.45 ± 0.75 2.33 ± 1.16 1.35 ± 1.56 1.00 ± 1.58 0.78 ± 1.91 0.75 ± 1.92 0.1 0.94 ± 0.51 1.69 ± 2.33 1.33 ± 3.77 1.00 ± 4.21 0.93 ± 5.74 1.11 ± 7.58 0.05 0.65 ± 1.17 1.02 ± 4.66 1.27 ± 7.39 1.00 ± 7.14 1.18 ± 10.61 1.67 ± 14.64 0.025 0.35 ± 1.92 1.24 ± 5.56 1.29 ± 9.56 1.00 ± 11.57 1.18 ± 17.50 1.76 ± 26.77
D Galectin-3 Relative binding signal per glycan [µM] 23a 23b 23c 23d 23e 23f 10 10.21 ± 0.55 2.51 ± 0.23 1.37 ± 0.24 1.00 ± 0.29 0.74 ± 0.34 0.60 ± 0.29 5 11.57 ± 0.50 2.61 ± 0.23 1.37 ± 0.31 1.00 ± 0.48 0.73 ± 0.41 0.59 ± 0.30 1 12.23 ± 0.76 2.45 ± 0.53 1.36 ± 0.52 1.00 ± 0.37 0.73 ± 0.35 0.60 ± 0.40 0.5 9.06 ± 0.60 2.07 ± 0.72 1.29 ± 0.83 1.00 ± 0.47 0.74 ± 0.35 0.67 ± 0.31 0.1 1.51 ± 0.24 1.06 ± 1.07 1.08 ± 1.96 1.00 ± 1.71 0.78 ± 1.26 0.67 ± 0.83 0.05 0.31 ± 0.18 0.43 ± 0.67 0.80 ± 1.99 1.00 ± 3.14 0.94 ± 3.35 0.89 ± 2.98 0.025 0.09 ± 0.09 0.42 ± 1.03 0.76 ± 3.18 1.00 ± 5.06 0.94 ± 6.41 0.92 ± 7.61 0.01 n.d. n.d. n.d. 1.00 ± 8.80 1.86 ± 20.05 1.47 ± 23.15 0.005 n.d. n.d. n.d. 1.00 ± 11.89 3.87 ± 21.39 0.96 ± 2.51
131
Table S3.4: Values of KD in SPR measurements with neo-glycoproteins and immobilized Gal-3
Apparent KD values determined by SPR are compared for all designed neo-glycoproteins. Neo- glycoproteins and ASF were applied over the surfaced immobilized with Gal-3. Values were calculated by fitting association and dissociation using Scrubber2. (n.d. – not detectable)
Attached Apparent K Attached Apparent K Ligand D Ligand D glycans (nM) glycans (nM) 20a 1.6 103 ± 3 21a 1.7 n.d. 20j 7.5 9.80 ± 0.10 21j 7.5 9.50 ± 0.10 20e 14.4 4.77 ± 0.07 21e 14.1 2.11 ± 0.02 20k 17.8 3.18 ± 0.06 21k 18.0 2.86 ± 0.02 20g 24.2 2.35 ± 0.05 21g 24.4 1.50 ± 0.20 20i 29.0 1.72 ± 0.04 21i 27.5 1.28 ± 0.02 22a 0.5 n.d. 23a 0.3 22 ± 1 22b 3.3 9.40 ± 0.10 23b 2.3 4.17 ± 0.05 22c 6.4 4.84 ± 0.05 23c 4.3 2.74 ± 0.03 22d 9.4 3.12 ± 0.03 23d 6.2 1.95 ± 0.02 22e 13.0 2.64 ± 0.03 23e 9.2 2.01 ± 0.02 22f 14.2 2.57 ± 0.02 23f 11.1 1.90 ± 0.20 ASF 75 ± 2
132
A association dissociation B association dissociation
20e,k,g,i 21j,e,k
20j 21g,i
20a 21a
C association dissociation D association dissociation
22c-f 23b-f 22b
23a 22a
E association dissociation
ASF
Figure S3.4: SPR sensorgrams of neo-glycoproteins bound by immobilized Gal-3 Neo-glycoproteins carrying (A) LacNAc-LacNAc (20a,j,e,k,g,i), (B) LacdiNAc-LacNAc (21a,j,e,k,g,i), (C) 6-biotin LacNAc-LacNAc (22a-f) and (D) 6-biotin LacdiNAc-LacNAc (23a-f) as well as ASF (E) were applied in flow on the surface immobilized with Gal-3. Responses of different ligands at a concentration of 0.2 µM were plotted against the time. With increasing glycan number per BSA steeper slopes are observed.
133