DOI: 10.1002/chem.201500065
Communication
& Carbohydrates
Automated Synthesis of Arabinoxylan-Oligosaccharides Enables Characterization of Antibodies that Recognize Plant Cell Wall Glycans Deborah Schmidt,[a, b] Frank Schuhmacher,[a, b] Andreas Geissner,[a, b] Peter H. Seeberger,[a, b] and Fabian Pfrengle*[a, b] Abstract: Monoclonal antibodies that recognize plant cell wall glycans are used for high-resolution imaging, providing important information about the structure and function of cell wall polysaccharides. To characterize the binding epitopes of these powerful molecular probes a library of eleven plant arabinoxylan oligosaccharides was produced by automated solid-phase synthesis. Modular assembly of oligoarabinoxylans from few building blocks was enabled by adding (2-naphthyl)methyl (Nap) to the toolbox of orthogonal protecting groups for solid-phase synthesis. Conjugation-ready oligosaccharides were obtained and the binding specificities of xylan-directed antibodies were determined on microarrays.
a common backbone consisting of b-1,4-linked d-xylopyranoses. This backbone structure may be partially acetylated and substituted with l-arabinofuranosyl or d-(4-O-methyl) glucuronyl residues.[2b] High-resolution imaging of cell wall microstructures has provided insights into the structure, function, dynamics, and biosynthesis of plant cell wall polysaccharides.[5] It has been shown that different glycans are present within subdomains of a single cell wall and that the occurrence of specific substructures changes during plant growth.[6] The composition of glycans in the cell wall can be monitored using monoclonal antibodies developed against plant polysaccharides.[7] The specificities of these powerful molecular probes were investigated using carbohydrate microarrays equipped with poly- and oligosaccharides of plant origin.[8] However, the precise epitopes
Plant cells are surrounded by a polysaccharide-rich matrix that constitutes the cell wall of all higher plants.[1] One of the main components of plant cell wall polysaccharides is the hemicellulose xylan, the second most abundant polysaccharide in Nature (Figure 1).[2] Xylans are dietary carbohydrates in everyday food that can provide medicinal benefits including immuFigure 1. The hemicellulose xylan cross-links cellulose microfibrils in plant cell walls. nomodulatory, antitumor, and antimicrobial effects.[3] In addithat are recognized by these antibodies are yet to be detertion, xylans are potential resources for the production of food mined. additives, cosmetics, and biofuels.[4] Although the structure of Despite the potential of xylan fragments as tools for plant xylans varies between plant species, they all possess biology and in the development of various practical applica[a] D. Schmidt, F. Schuhmacher, A. Geissner, Prof. Dr. P. H. Seeberger, tions,[4] access to xylo-oligosaccharides is mostly limited to Dr. F. Pfrengle their isolation from plants.[9] However, structurally defined and Department of Biomolecular Systems highly homogeneous samples are obtained by chemical synMax-Planck-Institute of Colloids and Interfaces thesis.[10] Am Mìhlenberg 1, 14476 Potsdam (Germany) Automated oligosaccharide synthesis[11] provides the means [b] D. Schmidt, F. Schuhmacher, A. Geissner, Prof. Dr. P. H. Seeberger, Dr. F. Pfrengle to construct carbohydrate libraries in a facile and timely Freie Universitt Berlin, Institute of Chemistry and Biochemistry manner, as structurally related oligosaccharides can be proArnimallee 22, 14195 Berlin (Germany) duced from common building blocks. Here, we report the auE-mail: [email protected] tomated synthesis of a collection of arabinoxylan fragments Supporting information for this article is available on the WWW under http://dx.doi.org/10.1002/chem.201500065.
Chem. Eur. J. 2015, 21, 5709 – 5713
5709
Ó 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Communication and their use in determining the binding epitopes of antibodies that are derived from natural xylan polysaccharides. High coupling efficiencies at every step of the automated synthesis mandate carefully designed building blocks (Figure 2). For the construction of the b-1,4-linked xylan backbone two different d-xylopyranose building blocks were employed. Scheme 1. Synthesis of xylose building blocks 6 a and 6 b. Reagents and conditions: a) TrCl, DMAP, NEt3, DMF; b) RBr, NaH, TBAI; c) TsOH, MeOH/Et2O/ H2O (100:10:1), 2 a: 72 %, 2 b: 62 % (3 steps); d) H2O/AcOH, reflux; e) Ac2O, DMAP; f) HSTol, BF3·OEt2, 0 8C, 4 a: 62 %, 4 b: 48 % (3 steps); g) NaOMe, MeOH, CH2Cl2 ; h) Bz2O, 5–10 mol % Yb(OTf)3, dioxane, 5 a: 71 %, 5 b: 61 % (2 steps); j) FmocCl, pyridine, CH2Cl2, a: 79 %, b: 76 %; k) HOP(O)(OBu)2, N-iodosuccinimide, triflic acid, 6 a: 87 %, 6 b: 95 %.
Figure 2. Monosaccharide building blocks for the automated solid-phase synthesis of arabinoxylan fragments.
Each of the C-4 hydroxyls was protected with a fluorenylmethoxycarbonyl (Fmoc) group because it can be selectively removed during the automated synthesis using an amine base. The required b-selectivity in the glycosylation steps was ensured through the installation of benzoate esters on the C-2 hydroxyls. The C-3 hydroxyl was either equipped with a benzyl ether as permanent protecting group or with a (2-naphthyl)methyl (Nap) group as temporary protecting group for substitution of the backbone with arabinose residues. Selective cleavage of Nap ethers in the presence of benzyl ethers has been reported under oxidative conditions using dichlorodicyanobenzochinone (DDQ),[12] and we envisioned that similar reaction conditions might be applicable to solid-phase synthesis. We chose the Nap ether because of its similar electronic properties to benzyl ethers and the relatively straightforward synthesis of the respective xylose building block. Finally, dibutylphosphate was chosen as the anomeric leaving group because glycosyl phosphates gave the best results in the assembly of large oligomeric glycans previously.[11e, f] Installation of the arabinose substituents on the xylan backbone relied either on a perbenzoylated or a 2-Fmoc protected l-arabinofuranosyl building block depending on further side-chain elongation. The synthesis of the desired xylose building blocks 6 a and 6 b was accomplished starting from the commercially available d-xylofuranose derivative 1 (Scheme 1). By following a strategy reported by Paquette et al. we were able to install either a benzyl or a Nap ether in 3-position of thioglycosides 4 a and 4 b, respectively.[13] After removal of the acetyl groups using sodium methoxide selective protection of the 2-position was performed following an ytterbium triflate-catalyzed benzoylation protocol.[14] Protection of the 4-hydroxyl group of thioglycosides 5 a and 5 b with fluorenylmethoxycarbonyl chloride (FmocCl) in the presence of pyridine was followed by the conversion of the thioglycoside leaving group into the corresponding phosphate affording the desired building blocks 6 a Chem. Eur. J. 2015, 21, 5709 – 5713
www.chemeurj.org
and 6 b in good overall yields. The arabinose building blocks were synthesized following literature procedures.[15] With the required building blocks in hand a set of eleven plant arabinoxylan fragments was chosen to be produced using an automated oligosaccharide synthesizer.[16] We focused on specific substructures of naturally occurring arabinoxylans comprised of a linear b-1,4-linked xylan backbone substituted with a-1,3-linked single l-arabinofuranosyl and b-1,2-d-xylopyranosyl-a-1,3-l-arabinofuranosyl disaccharide residues. Linear oligoxylans (9–12) and oligoarabinoxylans of different complexity (13–19) were produced via the iterative addition of building blocks 6 a,b and 7 a,b to linker-functionalized resin 8 (Scheme 2). The photocleavable linker ensures smooth cleavage of the assembled oligosaccharides from the solid support using UV light in a microfluidic photoreactor.[17] The linker is designed to withstand strong acidic and basic conditions and delivers conjugation-ready glycans after cleavage from the solid support. The glycosylation steps were performed in duplicate using five equivalents of glycosyl donor and either equimolar TMSOTf or N-iodosuccinimide together with catalytic amounts of TfOH for activation. Each glycosylation using a xylose building block was followed by the removal of either the temporary Fmoc protecting group at C-4, allowing for elongation of the xylan backbone, or of the Nap protecting group at C-3, for installation of arabinose substituents. After optimization of the reaction conditions on a model disaccharide, we established a reliable protocol for the selective cleavage of Nap groups on solid support.[18] Thus, a new temporary protecting group was added to the toolbox of orthogonal protecting groups for automated oligosaccharide synthesis. The Nap ether differs from the routinely used Fmoc and levulinoyl protecting groups by exerting no disarming effect on the respective building block, facilitating the fine-tuning of glycosyl donor reactivity in solid-phase syntheses.[19] For the synthesis of pentasaccharide 15, which contains a b-1,2-d-xylopyranosyla-1,3-l-arabinofuranosyl substituent, the linear xylan backbone was capped by reaction with acetic anhydride after assembly. After removal of the Nap-group at the central xylose unit and installation of the arabinose residue using building block 7 b, the side chain was selectively elongated by deprotection of
5710
Ó 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Communication
Scheme 2. Automated solid-phase synthesis of arabinoxylan fragments 9–19. The letter code below the structures represents the reaction sequence applied in the respective synthesis. Reagents and conditions: a) or b) 2 Õ 5 equiv of BB 6 a or 6 b, TMSOTf, DCM, ¢35 8C (5 min)!¢15 8C (30 min); c) or d) 2 Õ 5 equiv of BB 7 a or 7 b, NIS, TfOH, DCM/dioxane, ¢40 8C (5 min)!¢20 8C (40 min); e) 3 cycles of 20 % NEt3 in DMF, 25 8C (5 min); f) 7 cycles of 0.1 m DDQ in DCE/ MeOH/H2O (64:16:1), 20 min; g) 3 cycles of Ac2O, pyridine, 25 8C (30 min); h) hn (305 nm); i) NaOMe, THF/MeOH, 12 h; j) H2, Pd/C, EtOAc/MeOH/H2O/HOAc, 12 h. 9: 15 %, 10: 27 %, 11: 20 %, 12: 43 %, 13: 23 %, 14: 19 %, 15: 42 %, 16: 8 %, 17: 15 %, 18: 7 %, 19: 21 % (yields are based on resin loading).
the Fmoc-substituted arabinose residue and a final glycosylation with xylose donor 6 a. Following oligosaccharide assembly, the resin was removed from the synthesizer and the crude oligoarabinoxylosides were cleaved from the solid support.[17] Subsequent removal of the benzoates with sodium methoxide and hydrogenolytic cleavage of the benzyl ethers and the Cbz protecting group afforded the fully deprotected oligoarabinoxylans 9–19 in 7–43 % overall yield, based on the calculated loading of linker-functionalized resin 8. The products were purified by preparative reversed-phase HPLC, either at the stage of the semi-protected oligosaccharides before hydrogenation, or at the final stage yielding the pure products. All products were characterized by Chem. Eur. J. 2015, 21, 5709 – 5713
www.chemeurj.org
analytical HPLC, NMR spectroscopy, and high-resolution mass spectrometry.[18] With arabinoxylan fragments 9–19 in hand, microarray slides[20] were printed and used as tools to probe the binding preferences of 31 anti-xylan monoclonal antibodies.[21] Selected examples are shown in Figure 3. As expected all linear oligoxylans were detected by monoclonal antibody LM10 that was originally raised against a synthetic pentaxyloside–BSA conjugate.[22] LM10 also bound to all arabinose substituted structures containing a terminal xylose residue. Similarly CCRCM150, raised against a natural corn stover xylan, recognized almost all oligoxylans independent of backbone substitution. In some cases the binding epitopes were precisely character-
5711
Ó 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Communication ized. CCRC-M147 recognized selectively all structures containing an unsubstituted xylan disaccharide and CCRC-M154 bound exclusively to oligoxylans substituted with a-1,3-linked arabinose residues. The antibodies used in this study were previously grouped into four clades based on their binding properties towards natural xylan polysaccharides.[21] Interestingly, we observed different binding patterns towards the synthetic oligoarabinoxylans not only for antibodies from the same clade but also for antibodies raised against the same polysaccharide antigen. For example, while CCRC-M140 detected only the linear hexa- and octaxylosides, CCRC-M153 also bound to arabinose substituted xylosides. Both antibodies were originally raised against the same natural oat xylan and bind the same group of natural xylans. None of the antibodies showed either strong or selective binding to pentasaccharide 15 containing the b-1,2-d-xylopyranosyl-a-1,3-l-arabinofuranosyl substituent making this antigen a promising candidate for immunizations and production of monoclonal antibodies towards this otherwise undetectable epitope. Our results demonstrate that libraries of synthetic plant oligosaccharides can be used to determine the binding specificities of cell wall glycan-directed antibodies as exemplified by the identification of antibody CCRCM154 as a selective marker for xylans that are densely substituted with a-1,3-linked arabinose residues.
Employing automated solid-phase synthesis enabled us to rapidly procure a collection of oligosaccharide fragments of plant arabinoxylans. The library of synthetic arabinoxylan fragments was printed as microarray and probed with anti-xylan antibodies, which are valuable probes for immunolabeling studies of plant cell walls. The binding specificities of several antibodies were determined, and currently other oligoxylans such as glucuronoxylans are being prepared for characterization of antibodies that did not bind the structures presented in this report. With the synthetic ability to rapidly create libraries of precision plant carbohydrates, future glycan array studies will provide the essential information required for a detailed interpretation of immunolabeling studies of plant cell walls.
Acknowledgements We gratefully acknowledge financial support from the Max Planck Society, the Fonds der Chemischen Industrie (Liebig Fellowship to F.P.) and an ERC Advanced Grant (AUTOHEPARIN to P.H.S.). We thank Dr. M. Schlegel for helpful discussions and Dr. K. Gilmore for critically editing this manuscript. Generation of the CCRC series of monoclonal antibodies used in this work was supported by a grant from the NSF Plant Genome Program (DBI-0421683). Keywords: arabinoxylan · carbohydrates · microarray · plant cell wall · solid-phase synthesis
Figure 3. Detection of oligoarabinoxylans 9–19 by anti-xylan monoclonal antibodies (mAb): a) printing pattern; b) microarray scans. Representative scans of at least two independent experiments are shown. The intensity of the spots corresponds to the binding affinity of the respective mAb. The structures are drawn according to the Consortium for Functional Glycomics (CFG) nomenclature. Chem. Eur. J. 2015, 21, 5709 – 5713
www.chemeurj.org
[1] P. Albersheim, A. Darvill, K. Roberts, R. Sederoff, A. Staehelin, Plant Cell Walls: From Chemistry to Biology, Garland Science, New York, 2010. [2] a) H. V. Scheller, P. Ulvskov, Annu. Rev. Plant Biol. 2010, 61, 263 – 289; b) A. Ebringerov, T. Heinze, Macromol. Rapid Commun. 2000, 21, 542 – 556. [3] a) A. Ebringerov, A. Kardosˇov, Z. Hromdkov, A. Malovkov, V. Hrˇbalov, Int. J. Biol. Macromol. 2002, 30, 1 – 6; b) A. Proksch, H. Wagner, Phytochemistry 1987, 26, 1989 – 1993; c) R. L. Whistler, A. A. Bushway, P. P. Singh, W. Nakahara, R. Tokuzen, Adv. Carbohydr. Chem. Biochem. 1976, 32, 235 – 275; d) P. Christakopoulos, P. Katapodis, E. Kalogeris, D. Kekos, B. J. Macris, H. Stamatis, H. Skaltsa, Int. J. Biol. Macromol. 2003, 31, 171 – 175. [4] a) A. Moure, P. Gullûn, H. Domnguez, J. C. Parajû, Process Biochem. 2006, 41, 1913 – 1923; b) F. B. Sedlmeyer, Food Hydrocolloids 2011, 25, 1891 – 1898; c) D. Dodd, I. K. O. Cann, Glob. Change Biol. Bioenergy 2009, 1, 2 – 17. [5] K. J. D. Lee, S. E. Marcus, J. P. Knox, Mol. Plant 2011, 4, 212 – 219. [6] G. Freshour, C. P. Bonin, W.-D. Reiter, P. Albersheim, A. G. Darvill, M. G. Hahn, Plant Physiol. 2003, 131, 1602 – 1612. [7] W. G. T. Willats, C. G. Steele-King, L. McCartney, C. Orfila, S. E. Marcus, J. P. Knox, Plant Physiol. Biochem. 2000, 38, 27 – 36. [8] a) H. L. Pedersen, J. U. Fangel, B. McCleary, C. Ruzanski, M. G. Rydahl, M.C. Ralet, V. Farkas, L. von Schantz, S. E. Marcus, M. C. F. Andersen, R. Field, M. Ohlin, J. P. Knox, M. H. Clausen, W. G. T. Willats, J. Biol. Chem. 2012, 287, 39429 – 39438; b) H. L. P. J. U. Fangel, S. Vidal-Melgosa, L. I. Ahl, A. A. Salmean, J. Egelund, M. G. Rydahl, M. H. Clausen, W. G. Willats, Methods Mol. Biol. 2012, 918, 351 – 362; c) C. Herv¦, S. Marcus, J. P. Knox, in The Plant Cell Wall, Vol. 715 (Ed.: Z. A. Popper), Humana, 2011, pp. 103 – 113; d) I. Moller, I. Sørensen, A. J. Bernal, C. Blaukopf, K. Lee, J. Øbro, F. Pettolino, A. Roberts, J. D. Mikkelsen, J. P. Knox, A. Bacic, W. G. T. Willats, Plant J. 2007, 50, 1118 – 1128. [9] Y. Han, V. Agarwal, D. Dodd, J. Kim, B. Bae, R. I. Mackie, S. K. Nair, I. K. O. Cann, J. Biol. Chem. 2012, 287, 34946 – 34960. [10] K. Takeo, Y. Ohguchi, R. Hasegawa, S. Kitamura, Carbohydr. Res. 1995, 278, 301 – 313.
5712
Ó 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Communication [11] a) O. J. Plante, E. R. Palmacci, P. H. Seeberger, Science 2001, 291, 1523 – 1527; b) P. H. Seeberger, Chem. Soc. Rev. 2008, 37, 19 – 28; c) L. Krçck, D. Esposito, B. Castagner, C.-C. Wang, P. Bindschadler, P. H. Seeberger, Chem. Sci. 2012, 3, 1617 – 1622; d) M. T. C. Walvoort, H. van den Elst, O. J. Plante, L. Krçck, P. H. Seeberger, H. S. Overkleeft, G. A. van der Marel, J. D. C. Cod¦e, Angew. Chem. Int. Ed. 2012, 51, 4393 – 4396; Angew. Chem. 2012, 124, 4469 – 4472; e) O. Calin, S. Eller, P. H. Seeberger, Angew. Chem. Int. Ed. 2013, 52, 5862 – 5865; Angew. Chem. 2013, 125, 5974 – 5977; f) M. W. Weishaupt, S. Matthies, P. H. Seeberger, Chem. Eur. J. 2013, 19, 12497 – 12503. [12] a) J. Xia, S. A. Abbas, R. D. Locke, C. F. Piskorz, J. L. Alderfer, K. L. Matta, Tetrahedron Lett. 2000, 41, 169 – 173; b) J. A. Wright, J. Yu, J. B. Spencer, Tetrahedron Lett. 2001, 42, 4033 – 4036; c) J. Xia, J. L. Alderfer, C. F. Piskorz, K. L. Matta, Chem. Eur. J. 2001, 7, 356 – 367. [13] L. A. Paquette, L. Barriault, D. Pissarnitski, J. N. Johnston, J. Am. Chem. Soc. 2000, 122, 619 – 631. [14] T.-Y. Huang, M. M. L. Zulueta, S.-C. Hung, Org. Lett. 2011, 13, 1506 – 1509. [15] a) J. Kandasamy, M. Hurevich, P. H. Seeberger, Chem. Commun. 2013, 49, 4453 – 4455; b) P. I. Abronina, N. M. Podvalnyy, S. L. Sedinkin, K. G. Fedina, A. I. Zinin, A. O. Chizhov, V. I. Torgov, L. O. Kononov, Synthesis 2012, 44, 1219 – 1225.
Chem. Eur. J. 2015, 21, 5709 – 5713
www.chemeurj.org
[16] The automated syntheses were either performed on a home-built synthesizer or on a demonstrator synthesizer that is now commercialized by GlycoUniverse GmbH & CoKG. [17] S. Eller, M. Collot, J. Yin, H. S. Hahm, P. H. Seeberger, Angew. Chem. Int. Ed. 2013, 52, 5858 – 5861; Angew. Chem. 2013, 125, 5970 – 5973. [18] For details, see the Supporting Information. [19] a) D. R. Mootoo, P. Konradsson, U. Udodong, B. Fraser-Reid, J. Am. Chem. Soc. 1988, 110, 5583 – 5584; b) Z. Zhang, I. R. Ollmann, X.-S. Ye, R. Wischnat, T. Baasov, C.-H. Wong, J. Am. Chem. Soc. 1999, 121, 734 – 753. [20] C. D. Rillahan, J. C. Paulson, Annu. Rev. Biochem. 2011, 80, 797 – 823. [21] S. Pattathil, U. Avci, D. Baldwin, A. G. Swennes, J. A. McGill, Z. Popper, T. Bootten, A. Albert, R. H. Davis, C. Chennareddy, R. Dong, B. O’Shea, R. Rossi, C. Leoff, G. Freshour, R. Narra, M. O’Neil, W. S. York, M. G. Hahn, Plant Physiol. 2010, 153, 514 – 525. [22] L. McCartney, S. E. Marcus, J. P. Knox, J. Histochem. Cytochem. 2005, 53, 543 – 546.
Received: January 7, 2015 Published online on February 26, 2015
5713
Ó 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Communication
& Carbohydrates
Automated Synthesis of Arabinoxylan-Oligosaccharides Enables Characterization of Antibodies that Recognize Plant Cell Wall Glycans Deborah Schmidt,[a, b] Frank Schuhmacher,[a, b] Andreas Geissner,[a, b] Peter H. Seeberger,[a, b] and Fabian Pfrengle*[a, b] Abstract: Monoclonal antibodies that recognize plant cell wall glycans are used for high-resolution imaging, providing important information about the structure and function of cell wall polysaccharides. To characterize the binding epitopes of these powerful molecular probes a library of eleven plant arabinoxylan oligosaccharides was produced by automated solid-phase synthesis. Modular assembly of oligoarabinoxylans from few building blocks was enabled by adding (2-naphthyl)methyl (Nap) to the toolbox of orthogonal protecting groups for solid-phase synthesis. Conjugation-ready oligosaccharides were obtained and the binding specificities of xylan-directed antibodies were determined on microarrays.
a common backbone consisting of b-1,4-linked d-xylopyranoses. This backbone structure may be partially acetylated and substituted with l-arabinofuranosyl or d-(4-O-methyl) glucuronyl residues.[2b] High-resolution imaging of cell wall microstructures has provided insights into the structure, function, dynamics, and biosynthesis of plant cell wall polysaccharides.[5] It has been shown that different glycans are present within subdomains of a single cell wall and that the occurrence of specific substructures changes during plant growth.[6] The composition of glycans in the cell wall can be monitored using monoclonal antibodies developed against plant polysaccharides.[7] The specificities of these powerful molecular probes were investigated using carbohydrate microarrays equipped with poly- and oligosaccharides of plant origin.[8] However, the precise epitopes
Plant cells are surrounded by a polysaccharide-rich matrix that constitutes the cell wall of all higher plants.[1] One of the main components of plant cell wall polysaccharides is the hemicellulose xylan, the second most abundant polysaccharide in Nature (Figure 1).[2] Xylans are dietary carbohydrates in everyday food that can provide medicinal benefits including immuFigure 1. The hemicellulose xylan cross-links cellulose microfibrils in plant cell walls. nomodulatory, antitumor, and antimicrobial effects.[3] In addithat are recognized by these antibodies are yet to be detertion, xylans are potential resources for the production of food mined. additives, cosmetics, and biofuels.[4] Although the structure of Despite the potential of xylan fragments as tools for plant xylans varies between plant species, they all possess biology and in the development of various practical applica[a] D. Schmidt, F. Schuhmacher, A. Geissner, Prof. Dr. P. H. Seeberger, tions,[4] access to xylo-oligosaccharides is mostly limited to Dr. F. Pfrengle their isolation from plants.[9] However, structurally defined and Department of Biomolecular Systems highly homogeneous samples are obtained by chemical synMax-Planck-Institute of Colloids and Interfaces thesis.[10] Am Mìhlenberg 1, 14476 Potsdam (Germany) Automated oligosaccharide synthesis[11] provides the means [b] D. Schmidt, F. Schuhmacher, A. Geissner, Prof. Dr. P. H. Seeberger, Dr. F. Pfrengle to construct carbohydrate libraries in a facile and timely Freie Universitt Berlin, Institute of Chemistry and Biochemistry manner, as structurally related oligosaccharides can be proArnimallee 22, 14195 Berlin (Germany) duced from common building blocks. Here, we report the auE-mail: [email protected] tomated synthesis of a collection of arabinoxylan fragments Supporting information for this article is available on the WWW under http://dx.doi.org/10.1002/chem.201500065.
Chem. Eur. J. 2015, 21, 5709 – 5713
5709
Ó 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Communication and their use in determining the binding epitopes of antibodies that are derived from natural xylan polysaccharides. High coupling efficiencies at every step of the automated synthesis mandate carefully designed building blocks (Figure 2). For the construction of the b-1,4-linked xylan backbone two different d-xylopyranose building blocks were employed. Scheme 1. Synthesis of xylose building blocks 6 a and 6 b. Reagents and conditions: a) TrCl, DMAP, NEt3, DMF; b) RBr, NaH, TBAI; c) TsOH, MeOH/Et2O/ H2O (100:10:1), 2 a: 72 %, 2 b: 62 % (3 steps); d) H2O/AcOH, reflux; e) Ac2O, DMAP; f) HSTol, BF3·OEt2, 0 8C, 4 a: 62 %, 4 b: 48 % (3 steps); g) NaOMe, MeOH, CH2Cl2 ; h) Bz2O, 5–10 mol % Yb(OTf)3, dioxane, 5 a: 71 %, 5 b: 61 % (2 steps); j) FmocCl, pyridine, CH2Cl2, a: 79 %, b: 76 %; k) HOP(O)(OBu)2, N-iodosuccinimide, triflic acid, 6 a: 87 %, 6 b: 95 %.
Figure 2. Monosaccharide building blocks for the automated solid-phase synthesis of arabinoxylan fragments.
Each of the C-4 hydroxyls was protected with a fluorenylmethoxycarbonyl (Fmoc) group because it can be selectively removed during the automated synthesis using an amine base. The required b-selectivity in the glycosylation steps was ensured through the installation of benzoate esters on the C-2 hydroxyls. The C-3 hydroxyl was either equipped with a benzyl ether as permanent protecting group or with a (2-naphthyl)methyl (Nap) group as temporary protecting group for substitution of the backbone with arabinose residues. Selective cleavage of Nap ethers in the presence of benzyl ethers has been reported under oxidative conditions using dichlorodicyanobenzochinone (DDQ),[12] and we envisioned that similar reaction conditions might be applicable to solid-phase synthesis. We chose the Nap ether because of its similar electronic properties to benzyl ethers and the relatively straightforward synthesis of the respective xylose building block. Finally, dibutylphosphate was chosen as the anomeric leaving group because glycosyl phosphates gave the best results in the assembly of large oligomeric glycans previously.[11e, f] Installation of the arabinose substituents on the xylan backbone relied either on a perbenzoylated or a 2-Fmoc protected l-arabinofuranosyl building block depending on further side-chain elongation. The synthesis of the desired xylose building blocks 6 a and 6 b was accomplished starting from the commercially available d-xylofuranose derivative 1 (Scheme 1). By following a strategy reported by Paquette et al. we were able to install either a benzyl or a Nap ether in 3-position of thioglycosides 4 a and 4 b, respectively.[13] After removal of the acetyl groups using sodium methoxide selective protection of the 2-position was performed following an ytterbium triflate-catalyzed benzoylation protocol.[14] Protection of the 4-hydroxyl group of thioglycosides 5 a and 5 b with fluorenylmethoxycarbonyl chloride (FmocCl) in the presence of pyridine was followed by the conversion of the thioglycoside leaving group into the corresponding phosphate affording the desired building blocks 6 a Chem. Eur. J. 2015, 21, 5709 – 5713
www.chemeurj.org
and 6 b in good overall yields. The arabinose building blocks were synthesized following literature procedures.[15] With the required building blocks in hand a set of eleven plant arabinoxylan fragments was chosen to be produced using an automated oligosaccharide synthesizer.[16] We focused on specific substructures of naturally occurring arabinoxylans comprised of a linear b-1,4-linked xylan backbone substituted with a-1,3-linked single l-arabinofuranosyl and b-1,2-d-xylopyranosyl-a-1,3-l-arabinofuranosyl disaccharide residues. Linear oligoxylans (9–12) and oligoarabinoxylans of different complexity (13–19) were produced via the iterative addition of building blocks 6 a,b and 7 a,b to linker-functionalized resin 8 (Scheme 2). The photocleavable linker ensures smooth cleavage of the assembled oligosaccharides from the solid support using UV light in a microfluidic photoreactor.[17] The linker is designed to withstand strong acidic and basic conditions and delivers conjugation-ready glycans after cleavage from the solid support. The glycosylation steps were performed in duplicate using five equivalents of glycosyl donor and either equimolar TMSOTf or N-iodosuccinimide together with catalytic amounts of TfOH for activation. Each glycosylation using a xylose building block was followed by the removal of either the temporary Fmoc protecting group at C-4, allowing for elongation of the xylan backbone, or of the Nap protecting group at C-3, for installation of arabinose substituents. After optimization of the reaction conditions on a model disaccharide, we established a reliable protocol for the selective cleavage of Nap groups on solid support.[18] Thus, a new temporary protecting group was added to the toolbox of orthogonal protecting groups for automated oligosaccharide synthesis. The Nap ether differs from the routinely used Fmoc and levulinoyl protecting groups by exerting no disarming effect on the respective building block, facilitating the fine-tuning of glycosyl donor reactivity in solid-phase syntheses.[19] For the synthesis of pentasaccharide 15, which contains a b-1,2-d-xylopyranosyla-1,3-l-arabinofuranosyl substituent, the linear xylan backbone was capped by reaction with acetic anhydride after assembly. After removal of the Nap-group at the central xylose unit and installation of the arabinose residue using building block 7 b, the side chain was selectively elongated by deprotection of
5710
Ó 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Communication
Scheme 2. Automated solid-phase synthesis of arabinoxylan fragments 9–19. The letter code below the structures represents the reaction sequence applied in the respective synthesis. Reagents and conditions: a) or b) 2 Õ 5 equiv of BB 6 a or 6 b, TMSOTf, DCM, ¢35 8C (5 min)!¢15 8C (30 min); c) or d) 2 Õ 5 equiv of BB 7 a or 7 b, NIS, TfOH, DCM/dioxane, ¢40 8C (5 min)!¢20 8C (40 min); e) 3 cycles of 20 % NEt3 in DMF, 25 8C (5 min); f) 7 cycles of 0.1 m DDQ in DCE/ MeOH/H2O (64:16:1), 20 min; g) 3 cycles of Ac2O, pyridine, 25 8C (30 min); h) hn (305 nm); i) NaOMe, THF/MeOH, 12 h; j) H2, Pd/C, EtOAc/MeOH/H2O/HOAc, 12 h. 9: 15 %, 10: 27 %, 11: 20 %, 12: 43 %, 13: 23 %, 14: 19 %, 15: 42 %, 16: 8 %, 17: 15 %, 18: 7 %, 19: 21 % (yields are based on resin loading).
the Fmoc-substituted arabinose residue and a final glycosylation with xylose donor 6 a. Following oligosaccharide assembly, the resin was removed from the synthesizer and the crude oligoarabinoxylosides were cleaved from the solid support.[17] Subsequent removal of the benzoates with sodium methoxide and hydrogenolytic cleavage of the benzyl ethers and the Cbz protecting group afforded the fully deprotected oligoarabinoxylans 9–19 in 7–43 % overall yield, based on the calculated loading of linker-functionalized resin 8. The products were purified by preparative reversed-phase HPLC, either at the stage of the semi-protected oligosaccharides before hydrogenation, or at the final stage yielding the pure products. All products were characterized by Chem. Eur. J. 2015, 21, 5709 – 5713
www.chemeurj.org
analytical HPLC, NMR spectroscopy, and high-resolution mass spectrometry.[18] With arabinoxylan fragments 9–19 in hand, microarray slides[20] were printed and used as tools to probe the binding preferences of 31 anti-xylan monoclonal antibodies.[21] Selected examples are shown in Figure 3. As expected all linear oligoxylans were detected by monoclonal antibody LM10 that was originally raised against a synthetic pentaxyloside–BSA conjugate.[22] LM10 also bound to all arabinose substituted structures containing a terminal xylose residue. Similarly CCRCM150, raised against a natural corn stover xylan, recognized almost all oligoxylans independent of backbone substitution. In some cases the binding epitopes were precisely character-
5711
Ó 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Communication ized. CCRC-M147 recognized selectively all structures containing an unsubstituted xylan disaccharide and CCRC-M154 bound exclusively to oligoxylans substituted with a-1,3-linked arabinose residues. The antibodies used in this study were previously grouped into four clades based on their binding properties towards natural xylan polysaccharides.[21] Interestingly, we observed different binding patterns towards the synthetic oligoarabinoxylans not only for antibodies from the same clade but also for antibodies raised against the same polysaccharide antigen. For example, while CCRC-M140 detected only the linear hexa- and octaxylosides, CCRC-M153 also bound to arabinose substituted xylosides. Both antibodies were originally raised against the same natural oat xylan and bind the same group of natural xylans. None of the antibodies showed either strong or selective binding to pentasaccharide 15 containing the b-1,2-d-xylopyranosyl-a-1,3-l-arabinofuranosyl substituent making this antigen a promising candidate for immunizations and production of monoclonal antibodies towards this otherwise undetectable epitope. Our results demonstrate that libraries of synthetic plant oligosaccharides can be used to determine the binding specificities of cell wall glycan-directed antibodies as exemplified by the identification of antibody CCRCM154 as a selective marker for xylans that are densely substituted with a-1,3-linked arabinose residues.
Employing automated solid-phase synthesis enabled us to rapidly procure a collection of oligosaccharide fragments of plant arabinoxylans. The library of synthetic arabinoxylan fragments was printed as microarray and probed with anti-xylan antibodies, which are valuable probes for immunolabeling studies of plant cell walls. The binding specificities of several antibodies were determined, and currently other oligoxylans such as glucuronoxylans are being prepared for characterization of antibodies that did not bind the structures presented in this report. With the synthetic ability to rapidly create libraries of precision plant carbohydrates, future glycan array studies will provide the essential information required for a detailed interpretation of immunolabeling studies of plant cell walls.
Acknowledgements We gratefully acknowledge financial support from the Max Planck Society, the Fonds der Chemischen Industrie (Liebig Fellowship to F.P.) and an ERC Advanced Grant (AUTOHEPARIN to P.H.S.). We thank Dr. M. Schlegel for helpful discussions and Dr. K. Gilmore for critically editing this manuscript. Generation of the CCRC series of monoclonal antibodies used in this work was supported by a grant from the NSF Plant Genome Program (DBI-0421683). Keywords: arabinoxylan · carbohydrates · microarray · plant cell wall · solid-phase synthesis
Figure 3. Detection of oligoarabinoxylans 9–19 by anti-xylan monoclonal antibodies (mAb): a) printing pattern; b) microarray scans. Representative scans of at least two independent experiments are shown. The intensity of the spots corresponds to the binding affinity of the respective mAb. The structures are drawn according to the Consortium for Functional Glycomics (CFG) nomenclature. Chem. Eur. J. 2015, 21, 5709 – 5713
www.chemeurj.org
[1] P. Albersheim, A. Darvill, K. Roberts, R. Sederoff, A. Staehelin, Plant Cell Walls: From Chemistry to Biology, Garland Science, New York, 2010. [2] a) H. V. Scheller, P. Ulvskov, Annu. Rev. Plant Biol. 2010, 61, 263 – 289; b) A. Ebringerov, T. Heinze, Macromol. Rapid Commun. 2000, 21, 542 – 556. [3] a) A. Ebringerov, A. Kardosˇov, Z. Hromdkov, A. Malovkov, V. Hrˇbalov, Int. J. Biol. Macromol. 2002, 30, 1 – 6; b) A. Proksch, H. Wagner, Phytochemistry 1987, 26, 1989 – 1993; c) R. L. Whistler, A. A. Bushway, P. P. Singh, W. Nakahara, R. Tokuzen, Adv. Carbohydr. Chem. Biochem. 1976, 32, 235 – 275; d) P. Christakopoulos, P. Katapodis, E. Kalogeris, D. Kekos, B. J. Macris, H. Stamatis, H. Skaltsa, Int. J. Biol. Macromol. 2003, 31, 171 – 175. [4] a) A. Moure, P. Gullûn, H. Domnguez, J. C. Parajû, Process Biochem. 2006, 41, 1913 – 1923; b) F. B. Sedlmeyer, Food Hydrocolloids 2011, 25, 1891 – 1898; c) D. Dodd, I. K. O. Cann, Glob. Change Biol. Bioenergy 2009, 1, 2 – 17. [5] K. J. D. Lee, S. E. Marcus, J. P. Knox, Mol. Plant 2011, 4, 212 – 219. [6] G. Freshour, C. P. Bonin, W.-D. Reiter, P. Albersheim, A. G. Darvill, M. G. Hahn, Plant Physiol. 2003, 131, 1602 – 1612. [7] W. G. T. Willats, C. G. Steele-King, L. McCartney, C. Orfila, S. E. Marcus, J. P. Knox, Plant Physiol. Biochem. 2000, 38, 27 – 36. [8] a) H. L. Pedersen, J. U. Fangel, B. McCleary, C. Ruzanski, M. G. Rydahl, M.C. Ralet, V. Farkas, L. von Schantz, S. E. Marcus, M. C. F. Andersen, R. Field, M. Ohlin, J. P. Knox, M. H. Clausen, W. G. T. Willats, J. Biol. Chem. 2012, 287, 39429 – 39438; b) H. L. P. J. U. Fangel, S. Vidal-Melgosa, L. I. Ahl, A. A. Salmean, J. Egelund, M. G. Rydahl, M. H. Clausen, W. G. Willats, Methods Mol. Biol. 2012, 918, 351 – 362; c) C. Herv¦, S. Marcus, J. P. Knox, in The Plant Cell Wall, Vol. 715 (Ed.: Z. A. Popper), Humana, 2011, pp. 103 – 113; d) I. Moller, I. Sørensen, A. J. Bernal, C. Blaukopf, K. Lee, J. Øbro, F. Pettolino, A. Roberts, J. D. Mikkelsen, J. P. Knox, A. Bacic, W. G. T. Willats, Plant J. 2007, 50, 1118 – 1128. [9] Y. Han, V. Agarwal, D. Dodd, J. Kim, B. Bae, R. I. Mackie, S. K. Nair, I. K. O. Cann, J. Biol. Chem. 2012, 287, 34946 – 34960. [10] K. Takeo, Y. Ohguchi, R. Hasegawa, S. Kitamura, Carbohydr. Res. 1995, 278, 301 – 313.
5712
Ó 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Communication [11] a) O. J. Plante, E. R. Palmacci, P. H. Seeberger, Science 2001, 291, 1523 – 1527; b) P. H. Seeberger, Chem. Soc. Rev. 2008, 37, 19 – 28; c) L. Krçck, D. Esposito, B. Castagner, C.-C. Wang, P. Bindschadler, P. H. Seeberger, Chem. Sci. 2012, 3, 1617 – 1622; d) M. T. C. Walvoort, H. van den Elst, O. J. Plante, L. Krçck, P. H. Seeberger, H. S. Overkleeft, G. A. van der Marel, J. D. C. Cod¦e, Angew. Chem. Int. Ed. 2012, 51, 4393 – 4396; Angew. Chem. 2012, 124, 4469 – 4472; e) O. Calin, S. Eller, P. H. Seeberger, Angew. Chem. Int. Ed. 2013, 52, 5862 – 5865; Angew. Chem. 2013, 125, 5974 – 5977; f) M. W. Weishaupt, S. Matthies, P. H. Seeberger, Chem. Eur. J. 2013, 19, 12497 – 12503. [12] a) J. Xia, S. A. Abbas, R. D. Locke, C. F. Piskorz, J. L. Alderfer, K. L. Matta, Tetrahedron Lett. 2000, 41, 169 – 173; b) J. A. Wright, J. Yu, J. B. Spencer, Tetrahedron Lett. 2001, 42, 4033 – 4036; c) J. Xia, J. L. Alderfer, C. F. Piskorz, K. L. Matta, Chem. Eur. J. 2001, 7, 356 – 367. [13] L. A. Paquette, L. Barriault, D. Pissarnitski, J. N. Johnston, J. Am. Chem. Soc. 2000, 122, 619 – 631. [14] T.-Y. Huang, M. M. L. Zulueta, S.-C. Hung, Org. Lett. 2011, 13, 1506 – 1509. [15] a) J. Kandasamy, M. Hurevich, P. H. Seeberger, Chem. Commun. 2013, 49, 4453 – 4455; b) P. I. Abronina, N. M. Podvalnyy, S. L. Sedinkin, K. G. Fedina, A. I. Zinin, A. O. Chizhov, V. I. Torgov, L. O. Kononov, Synthesis 2012, 44, 1219 – 1225.
Chem. Eur. J. 2015, 21, 5709 – 5713
www.chemeurj.org
[16] The automated syntheses were either performed on a home-built synthesizer or on a demonstrator synthesizer that is now commercialized by GlycoUniverse GmbH & CoKG. [17] S. Eller, M. Collot, J. Yin, H. S. Hahm, P. H. Seeberger, Angew. Chem. Int. Ed. 2013, 52, 5858 – 5861; Angew. Chem. 2013, 125, 5970 – 5973. [18] For details, see the Supporting Information. [19] a) D. R. Mootoo, P. Konradsson, U. Udodong, B. Fraser-Reid, J. Am. Chem. Soc. 1988, 110, 5583 – 5584; b) Z. Zhang, I. R. Ollmann, X.-S. Ye, R. Wischnat, T. Baasov, C.-H. Wong, J. Am. Chem. Soc. 1999, 121, 734 – 753. [20] C. D. Rillahan, J. C. Paulson, Annu. Rev. Biochem. 2011, 80, 797 – 823. [21] S. Pattathil, U. Avci, D. Baldwin, A. G. Swennes, J. A. McGill, Z. Popper, T. Bootten, A. Albert, R. H. Davis, C. Chennareddy, R. Dong, B. O’Shea, R. Rossi, C. Leoff, G. Freshour, R. Narra, M. O’Neil, W. S. York, M. G. Hahn, Plant Physiol. 2010, 153, 514 – 525. [22] L. McCartney, S. E. Marcus, J. P. Knox, J. Histochem. Cytochem. 2005, 53, 543 – 546.
Received: January 7, 2015 Published online on February 26, 2015
5713
Ó 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
