CHEMBIOCHEM FULL PAPERS DOI: 10.1002/cbic.201300764

Biological Evaluation of Multivalent Lewis X–MGL-1 Interactions Magdalena Eriksson,[a, b] Sonia Serna,[c] Maha Maglinao,[a, b] Mark K. Schlegel,[a, b] Peter H. Seeberger,[a, b] Niels-Christian Reichardt,*[c, d] and Bernd Lepenies*[a, b] Myeloid C-type lectin receptors (CLRs) expressed by antigenpresenting cells are pattern-recognition receptors involved in the recognition of pathogens as well as of self-antigens. The interaction of carbohydrate ligands with a CLR can trigger immune responses. Although several CLR ligands are known, there is limited insight into CLR targeting by carbohydrate ligands. The weak affinity of lectin–carbohydrate interactions often renders multivalent carbohydrate presentation necessary. Here, we have analyzed the impact of multivalent presentation of the trisaccharide Lewis X (LeX) epitope on its interaction with the CLR macrophage galactose-type lectin-1 (MGL-1). Glycan arrays, including N-glycan structures with terminal LeX, were prepared by enzymatic extension of immobilized synthet-

ic core structures with two recombinant glycosyltransferases. Incubation of arrays with an MGL-1-hFc fusion protein showed up to tenfold increased binding to multiantennary N-glycans displaying LeX structures, compared to monovalent LeX trisaccharide. Multivalent presentation of LeX on the model antigen ovalbumin (OVA) led to increased cytokine production in a dendritic cell /T cell coculture system. Furthermore, immunization of mice with LeX-OVA conjugates modulated cytokine production and the humoral response, compared to OVA alone. This study provides insights into how multivalent carbohydrate– lectin interactions can be exploited to modulate immune responses.

Introduction Carbohydrate–protein interactions play an important role in various biological phenomena, such as cell adhesion,[1] cell-cell interaction, and signaling.[2] These interactions are crucial for the recognition of glycan structures on pathogens as well as for maintaining tolerance towards self-antigens.[3] Interactions between lectins and their carbohydrate ligands are usually relatively weak, and therefore depend on a multivalent ligand display.[4] One family of immunologically relevant lectins is C-type lectin receptors (CLRs). Myeloid CLRs are expressed by subsets of antigen-presenting cells, such as dendritic cells (DCs) and macrophages.[5] CLRs recognize glycan struc[a] Dr. M. Eriksson,+ Dr. M. Maglinao, Dr. M. K. Schlegel, Prof. Dr. P. H. Seeberger, Dr. B. Lepenies Department of Biomolecular Systems Max Planck Institute of Colloids and Interfaces Am Mhlenberg 1, 14476 Potsdam (Germany) E-mail: [email protected] [b] Dr. M. Eriksson,+ Dr. M. Maglinao, Dr. M. K. Schlegel, Prof. Dr. P. H. Seeberger, Dr. B. Lepenies Institute for Chemistry and Biochemistry, Freie Universitt Berlin Arnimallee 22, 14195 Berlin (Germany) [c] Dr. S. Serna,+ Dr. N.-C. Reichardt Biofunctional Nanomaterials Unit, CICbiomaGUNE Paseo Miramon 182, 20009 San Sebastian (Spain) E-mail: [email protected] [d] Dr. N.-C. Reichardt CIBER-BBN Paseo Miramon 182, 20009 San Sebastian (Spain) [+] These authors contributed equally to this work. Supporting information for this article is available on the WWW under http://dx.doi.org/10.1002/cbic.201300764.

 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

tures expressed by pathogens and self-antigens; this can lead to a cascade of events, such as phagocytosis of antigens,[6] release of cytokines, and modulation of an adaptive immune response.[7] CLR targeting has proven useful for cell-specific drug delivery[8] and for modulating T cell responses.[9] Because of the cell-specific expression of particular CLRs, antigen delivery to specific DC subtypes can be achieved, and thus enhance intracellular antigen processing or antigen presentation by MHC-I or MHC-II molecules. Numerous studies have focused on antibody-mediated CLR targeting. For instance, directing antigens to DEC-205 (expressed by CD8 + DCs) induced antigen uptake and intracellular routing to MHC-I presentation, thus promoting CD8 + T cell activation.[10] In contrast, antibody-mediated targeting of antigens to the CLR DC immunoreceptor resulted in antigen presentation by MHC-II molecules.[10] Similarly, antigens captured by DC-SIGN were routed to endosomes/lysosomes followed by presentation on MHC-II molecules.[11] Although the specificity of antibody-mediated CLR targeting is well proven, antibodies can elicit unwanted immune responses against the antibody and can induce antibody-dependent cellmediated cytotoxicity. Therefore, glycan-based targeting has gained more interest in the last couple of years.[12] Glycanbased CLR targeting has the advantage that the glycan presentation can be controlled and fine-tuned as needed. Numerous different architectures have been employed for the multivalent presentation of carbohydrates, including nanoparticles, dendrimers, polymers, fullerenes, and liposomes. Gold nanoparticles displaying oligomannosides were shown to bind to the CLR DC-SIGN (DC-specific intercellular adhesion molecule-3ChemBioChem 2014, 15, 844 – 851

844

CHEMBIOCHEM FULL PAPERS grabbing non-integrin) and to inhibit binding of the HIV glycoprotein gp120 to the receptor.[13] Nanoparticles functionalized with DC-SIGN carbohydrate ligands induced enhanced antigen presentation along with increased CD4 + and CD8 + T cell proliferation.[12c] In addition, sugar-capped quantum dots were used to study carbohydrate–lectin interactions in vivo.[14] Nanoparticles functionalized with sialyl Lewis X (sLex) were used to target the endothelial inflammation markers E-/P-selectins to allow early detection of brain inflammation.[15] Besides nanoparticles, glycoliposomes were employed for cell-specific delivery and immune modulation.[16] Finally, polymers have served as scaffolds for multivalent carbohydrate presentation. DCSIGN targeting by Lewis B–polyamidoamine polymers led to specific antigen delivery to DCs, thus promoting cytokine production and T cell proliferation.[17] A promising strategy to enhance specific targeting of antigens to DCs is glycan modification of the respective antigen. Dectin-1-mediated targeting by antigen crosslinking with the b-glucan polysaccharide zymosan routed it to lysosomes followed by efficient MHC-II presentation.[18] This effect was also observed in mice immunized with b-glucan-modified antigens, and resulted in robust antifungal responses accompanied by increased activity by Th1 and Th17 cells.[19] CLRs that have been successfully targeted for immune modulation are the macrophage galactose-type lectin (MGL) and its murine homologues MGL-1 and MGL-2. MGL and its homologues are expressed by immature DCs and macrophages,[20] and are involved in the recognition of pathogens[21] and tumor antigens.[22] Chemoenzymatic glycosylation of the mucin MUC1 with N-acetylgalactosamine (GalNAc) residues induced MGLmediated endocytosis by monocyte-derived DCs.[22a] The murine MGL homologues are involved in the recognition of carbohydrate tumor antigens[22b] and were reported to play a role in the immune response against lymph node metastasis.[23] In line with these observations, adding GalNAc-modified antigens to DCs induced MGL2-dependent antigen uptake and led to enhanced CD8 + T cell proliferation.[24] MGL-1 binds strongly to the trisaccharide LeX (also termed stage-specific embryonic antigen-1, SSEA-1). This antigen is expressed in various healthy organs; however, it is highly abundant in tumor cells as well as in embryonic stages.[25] LeX is also present in helminths such as Schistosoma mansoni[26] and has been demonstrated to be involved in Schistosoma-induced immune modulation.[21c, 26a, 27] One useful method to detect multivalent carbohydrate– lectin interactions is the glycan array platform, which allows high-throughput screening of carbohydrate interactions with lectins or antibodies.[28] Carbohydrates are presented in high density on a chip, thus mimicking the multivalent glycan presentation on cells and tissues. Glycan arrays can be used to identify novel ligands for lectins[29] and to screen for inhibitors of such interactions.[30] A number of different array platform types are available: covalently attached chemically synthesized carbohydrates to activated glass slides,[29b] noncovalently immobilized neoglycolipids and glycosphingolipids,[31] and attached extracted glycans or glycoproteins.[32]

 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.chembiochem.org To address the difficulties in rapidly obtaining complex glycan structures by chemical synthesis or isolation from natural sources, we recently explored the use of recombinant glycosyltransferases for the on-chip derivatization of synthetic glycan core structures in nanodroplets.[29c] A series of N-glycan core structures (with variable terminal sugars and branching) were immobilized and enzymatically derivatized with a series of glycosyltransferases, to generate a medium-sized array of complex N-glycan structures within a few days; this employed only microgram amounts of the valuable enzymes and glycan ligands. Enzymatic conversion on the immobilized glycans was assessed by incubating the slides with pairs of lectins specific for detecting starting and end products. The results were later confirmed by on-chip MALDI-MS with a slightly different chip format.[33] Glycosyltransferases catalyze the formation of new glycosidic bonds with complete regio- and stereoselectivity, and constitute versatile tools for the synthesis of glycoconjugates without the need for protecting groups. Our collection of multiantennary N-glycans is particularly interesting for studying the effects of multiple and branch-differentiated epitope presentation on glycan–lectin interactions.[34] Multiantennary compounds could be seen solely as miniature dendrimers that present multiple copies of a terminal epitope in a very confined space. But branch specificity in lectin binding and glycosidase and glycosyltransferase activity highlights the importance of the overall glycan conformation of the pseudosymmetric presentation of terminal glycans.[35] On the other hand, the effect of asymmetric epitope presentation in naturally occurring but only partially extended multiantennary N-glycans has been shown to have a profound effect on lectin binding.[36] This is significant, as enzymatic on-chip elongation of crowded structures (such as multiantennary N-glycans) can be incomplete. Whereas on-chip galactosylation with bovine beta 1–4 GalT is a robust and high-yielding process (in part, because of the relatively shallow binding site), Lewis-type fucosylation with a recombinant a-1,3-fucosyltransferase from Caenorhabditis elegans is likely to produce discrete mixtures of structures with varying degrees of terminal fucosylation. In this study, we analyzed the impact of multivalent LeX display on its interaction with MGL-1. A glycan array containing multiantennary N-glycan structures was prepared by stepwise incubation of synthetic core structures with glycosyltransferases. Binding analysis with MGL-1-hFc (see below) demonstrated that multiantennary LeX presentation led to a tenfold increase of fluorescence intensity when detection was performed with a fluorescently labeled secondary antibody. The binding of MGL-1-hFc to LeX was confirmed by surface plasmon resonance (SPR) measurements. Finally, we employed a T cell receptor transgenic mouse model to determine whether the LeX modification of the model antigen (ovalbumin, OVA) impacted T cell responses in vitro and in vivo. Our results highlight the potential of CLR targeting by multivalent glycan display to modulate immune responses.

ChemBioChem 2014, 15, 844 – 851

845

CHEMBIOCHEM FULL PAPERS Results and Discussion The goal of this study was to analyze the impact of multivalent LeX presentation on the LeX–MGL-1 interaction, and to explore whether efficient CLR targeting in vitro and in vivo could be mediated by multivalent LeX display. To study the interaction of MGL-1 with LeX, a fusion protein (MGL-1-hFc) consisting of the extracellular part of MGL-1 and the Fc part of human IgG1 was produced in a mammalian expression system.[37] SPR measurements were performed to obtain quantitative analysis of the interaction between MGL-1 and LeX. To this end, neutravidin was immobilized on a polycarboxylated CM5 sensor chip, followed by injection of LeX-PAA-biotin. Specific binding of 100 mm MGL-1-hFc to this immobilized LeX construct was detected and compared with that of a control (100 mm hFc, Figure 1 A). Kinetic measurements were then performed with vari-

Figure 1. A) SPR sensorgram of LeX-PAA-biotin binding to MGL-1-hFc and hFc at 100 mm. B) Binding between LeX and MGL-1-hFc (“MGL-1”; 0–50 mm).

ous concentrations of MGL-1-hFc (0–50 mm, Figure 1 B) to determine the dissociation constant (Kd) based on a 1:1 Langmuir interaction model, which is the simplest model describing the interaction between an immobilized ligand (LeX) and an analyte (MGL-1). It assumes that the association and disassociation phases are monoexponential; it has the advantage of allowing slight deviation from the raw data (e.g., baseline drift) and takes into account mass transport limitations. The calculated Kd of 3.79 mm by the 1:1 Langmuir model is in agreement with  2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.chembiochem.org a previously reported Kd value of about 1.5 mm for the MGL-1– LeX interaction.[38] Next, we determined whether multivalent presentation of LeX increases the avidity of its interaction with MGL-1. To this end, we performed a binding study on an array of N-glycans with broad variability in the number and presentation of the terminal LeX epitopes, by comprising specific features not simulated in previous synthetic dendrimer conjugates.[39] LeX epitopes were introduced in the antennae of complex and hybrid type N-glycans by sequential incubation of the array with two glycosyltransferases. N-glycan core structures comprising high-mannose, complex, and hybrid-type structures (1–14, all with an alkylamine linker) were robotically printed onto N-hydroxysuccinimide (NHS)-activated glass slides. Monovalent LeX was included as a positive control; high-mannose N-glycans (1, 2), N,N’-diacetyllactosamine (LacdiNAc, 13), and fucosylated LacdiNAc (LDNF, 14) were included as negative controls. Printed structures presenting terminal GlcNAc (3–12) were first galactosylated by incubation with a bovine milk b-1,4-galactosyltransferase and the corresponding nucleotide sugar donor (UDP-galactose). On-chip galactosylation was followed by incubation with a Caenorhabditis elegans a-1,3-fucosyltransferase, which has known in vitro Lewis-type fucosylation activity,[40] in the presence of GDP-fucose (Figure 2). The introduction of new sugar moieties onto the array was monitored by incubating the slide with commercially available fluorescently labeled lectins. Ricinus communis agglutinin was used to detect terminal galactose residues, and Aleuria aurantia lectin was used to detect fucose (Figure S1 in the Supporting Information). Although analysis with specific lectins could not provide quantitative measurement of the conversion at the surface, high local concentrations of both enzymes and glycosidic donors were employed (as well as extended reaction times) to ensure maximum conversion. In this way, a series of mono- and polyvalent LeX structures were obtained from mono-, di-, tri-, and tetra-antennary glycan scaffolds (Figure 3). The glycan array was incubated with different concentrations of MGL-1-hFc. Binding to glycan structures displaying terminal GlcNAc, Gal, or LeX was analyzed. In the array containing glycans 1–14 and in the galactosylated array, only binding to monovalent LeX was detected (data not shown), in accordance with previous reports.[20, 22c] After galactosylation and fucosylation, however, all N-glycans containing LeX epitopes in their antennae showed binding to MGL-1-hFc. MGL-1-hFc exhibited markedly increased affinity towards two of the extended N-glycans: tri- and tetra-antennary structures 6-LeX and 7-LeX. The fluorescence intensities for these interactions were almost ten times higher than for the monovalent LeX ligand (Figure 3 B), thus showing the importance of multivalent ligands in lectin binding. Interestingly, this effect was not seen with 5-LeX (Figure 3 A) which can present up to three LeX epitopes (as for 6-Lex). This indicates the importance of epitope presentation within the molecular context for effective lectin binding. Whether partially elongated structures or the fully extended glycans are the strongest binders in this series will be clarified in future studies, for example, by employing purified singleChemBioChem 2014, 15, 844 – 851

846

CHEMBIOCHEM FULL PAPERS

www.chembiochem.org

Figure 2. Enzymatic generation of LeX branched N-glycans. Printed N-glycan structures, after enzymatic reaction with bovine milk b-1,4-galactosyltransferase (GalT), and after enzymatic reaction with C. elegans a-1,3-fucosyltransferase (FucT) to generate LeX-type epitopes. &: N-Acetylglucosamine (GlcNAc), &: N-Acetylgalactosamine (GalNAc), *: mannose (Man), *: galactose (Gal), 3 : fucose (Fuc).

end, LeX was conjugated to the model antigen OVA, and the neoglycoconjugate was characterized by gel electrophoresis and MALDI-TOF MS (average 4.8 LeX residues per OVA molecule; Figure S3). To determine the ability of LeX-modified OVA to target DCs and modulate antigen presentation and T cell effector functions, splenic CD11c + DCs from C57BL/6 mice were purified by magnetic-activated cell separation (MACS, Figure 4 A) and either OVA alone or the OVALeX neoglycoconjugate was added. These DCs were then cocultured for 72 h with purified OVA-specific CD4 + T cells from OT-II transgenic mice (Figure 4 A). In this model system, OVA protein is internalized by the CD11c + DCs, processed, degraded, and then OVA323–339 peptide is presented by MHC-II I-Ab molecules on the cell surface.[6] These peptide/MHC-II complexes are specifically recognized by the OVA323– Figure 3. A) N-glycan microarray images after incubation with mMGL-1-hFc (10 mg mL 1) [43] followed by an Alexa Fluor 488-labeled mouse anti-human IgG1 antibody. B) Fluores339 peptide-specific T cell receptor of OT-II T cells. cence intensities after incubation with mMGL-1-hFc followed by anti-IgG-488. Each Thus, the measured T cell response indirectly reflects histogram represents the average RFU values for five replicates (mean  SD). antigen uptake, processing, and presentation by DCs. Interestingly, OT-II T cells cocultured with DCs to which OVA-LeX has been added produced significantligand entities. No binding was detected when the glycan ly higher amounts of IL-2 and IFN-g, compared with stimulaarray studies were done with other CLR-Fc fusion proteins, tion with OVA alone (Figure 4 B). IL-2 is a cytokine produced such as Clec9a-Fc or fluorescently labeled secondary antibody mainly by activated T cells and induces autocrine stimulation only (Figure S2). and T cell proliferation.[44] Increased IFN-g levels indicate inPrevious studies have shown that adding glycosylated creased CD4 + T cell differentiation into Th1. In contrast, the [41] model antigens to DCs can induce DC maturation, enhanced levels of IL-4 (a classical Th2 cytokine in cell supernatant) were below the detection limit (data not shown). These data indiantigen presentation, and T cell activation.[12c, 19, 42] For instance, cate that the in vitro CD4 + T cell response was enhanced by pulsing DCs with OVA conjugated to sulfated LeX or tri-GlcNAc enhanced cross-presentation of antigens, thus inducing inLeX-modified OVA. The significant increase of IFN-g is in agree+ [24] creased CD8 T cell proliferation. As the glycan array studies ment with a report in which both CD4 + and CD8 + T cell reindicated enhanced binding of MGL-1-hFc to multivalently presponses were enhanced by targeting MGL-2 with GalNAcsented LeX, we aimed at evaluating the potential of multivalent modified antigens.[24] X Le display to modulate OVA-specific T cell responses. To this  2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

ChemBioChem 2014, 15, 844 – 851

847

CHEMBIOCHEM FULL PAPERS

www.chembiochem.org geting and immune modulation. Glycan arrays were used to determine MGL-1 binding to multiantennary LeX structures, prepared by stepwise chemoenzymatic elongation of synthetic core glycan structures. Multiantennary LeX presentation enhanced binding of the recombinantly expressed MGL-1-hFc fusion protein, as indicated by a tenfold increased fluorescence intensity. Finally, we found in a T cell receptor transgenic system that LeX-modified OVA led to augmented cytokine production in vitro and to modulation of the humoral response in vivo. In conclusion, this study highlights the potential of CLR targeting by multivalent glycan display to modulate immune responses.

Experimental Section

Figure 4. A) Representative dot plots of MACS-purified splenic CD11c + cells from wild-type C57BL/6 mice (upper panel) and splenic T cells from transgenic OT-II mice (lower panel). Cells were stained with rat anti-mouse CD11c-APC or rat anti-mouse CD3-APC-eFluor780 antibody and analyzed by flow cytometry. FSC = forward scatter, SSC = side scatter. B) Cytokine production by OT-II T cells with CD11c + DCs to which OVA and OVA-LeX had been added. Cytokines in the supernatant were analyzed after 72 h of incubation. Statistical analysis was performed with a Student’s t-test. * p < 0.05, ** p < 0.01.

Finally, to investigate the immune modulatory properties of the neoglycoconjugates in vivo, an adoptive transfer model was employed, with OT-II cells intravenously injected into wildtype C57BL/6 mice. One day after cell transfer, mice were immunized with the OVA-LeX conjugates, then boosted with the same conjugates on day 21. Anti-OVA antibody serum titers were measured on day 30 (Figure 5 A). Interestingly, a decreased OVA-specific IgG response was detected in mice immunized with LeX-modified OVA, compared with mice immunized with OVA alone (Figure 5 B). In line with the results of the in vitro cell stimulation, a higher percentage of spleen cells from mice immunized with LeX-modified OVA produced IL-2 and IFN-g (Figure 5 C). Thus, this study highlights the potential of glycanbased DC targeting to modulate immune responses.

Conclusions We analyzed the impact of multivalent LeX display on interactions with MGL-1, and we demonstrated its impact on DC tar 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

General microarray methods: Microarrays were printed on glass slides by employing a sciFLEXARRAYER S11 robotic noncontact spotter (Scienion, Berlin, Germany). Nexterion H NHS-activated glass slides were purchased from Schott (Mainz, Germany). A. aurantia lectin (L-1390) and R. communis agglutinin I, RCA120 (L-1080) were purchased from Vector Laboratories (Burlingame, CA). Lectins were fluorescently labeled with the AnaTag HiLyte Fluor 555 protein labeling kit from AnaSpec (Freemont, CA). Enzymatic reactions were performed in SureHyb hybridization chambers (Agilent Technologies). Lectin incubations were performed in Whatman FAST slide incubation chambers (GE Healthcare). Fluorescence measurements were performed in an Agilent G265BA microarray scanner system (Agilent Technologies), and quantification was achieved with ProScanArray Express software (PerkinElmer) with an adaptive circle quantitation method (spot diameter: 50–300 mm). Average RFU values (mean  SD from five spots, local background subtracted) are reported as histograms. N-glycan array printing: Ligand solutions were prepared from stock solutions (1 mm in water) by dilution (200 mm) in sodium phosphate (300 mm, pH 8.5) containing, Tween-20 (0.005 %), and arrayed (1.25 nL each, 500 mm apart) onto NHS-functionalized glass slides. Each glycan was spotted as five replicates, and the entire glycan array (14 glycans, 5 copies per glycan) was printed 14 times on each slide. After printing, the slides were placed in a 75 % humidity chamber (saturated NaCl) at 25 8C for 18 h. The remaining NHS groups were quenched by placing the slide in ethanolamine (50 mm) in sodium borate (50 mm, pH 9.0), at RT for 1 h. The slides were washed with PBST (PBS containing Tween-20 (0.5 %)), then with PBS and with water. The slides were dried in a slide spinner. N-glycan array reaction with GalT and FucT: Bovine milk b-1,4galactosyltransferase (85 mU; E.C. 2.4.1.22, Sigma–Aldrich) in HEPES buffer (50 mm, pH 7.4; 500 mL) containing UDP-galactose (1 mm) and MnCl2 (20 mm) was incubated with the N-glycan array in a SureHyb microarray hybridization chamber at 37 8C for 72 h. The slide was washed with PBST, PBS, and water, then dried. C. elegans core type a-1,3-fucosyltransferase (CeFUT6, 200 mg) was expressed and purified as previously reported[40]) in 2-(N-morpholino)ethanesulfonic acid (MES; 80 mm, pH 6.5; 500 mL) containing GDP-fucose (GDP-Fuc 2 mm) and MnCl2 (20 mm) was incubated with the Nglycan array in the SureHyb chamber at RT for 72 h. The slide was then washed and dried (as above). Glycan array binding: The microarray slides were compartmentalized with a 16-well gasket (Fast Frame, Whatman). The glycan arrays were incubated for 1 h in the dark at room temperature with the following lectins: R. communis agglutinin (10 mg mL 1; ChemBioChem 2014, 15, 844 – 851

848

CHEMBIOCHEM FULL PAPERS

www.chembiochem.org by injecting neutravidin (100 mg mL 1) in acetate buffer (pH 4.0) until the immobilization level reached 2090 RU. The remaining activated carboxyl groups were then capped by injection of ethanolamine (1 m) for 10 min. Control flow cells were treated with EDC/NHS followed by ethanolamine as described. For binding analysis, LeXPAA-biotin (1 mg mL 1, Lectinity, Moscow, Russia) in HBSN (GE Healthcare) was injected, followed by either MGL1-hFc or hFc (100 mm). For kinetics analysis, MGL-1-hFc (0.1, 0.5, 1.0, 5, 10, or 50 mm) was injected over the LeXPAA-biotin-functionalized surfaces at 10 mL min 1 for 5 min for contact and for dissociation times, followed by regeneration with phosphoric acid (10 mm) at 30 mL min 1 for 30 s. Experimental data were obtained and analyzed with a Biacore T100 system (Biacore/GE Healthcare) and its accompanying software. Kinetic analysis (based on a 1:1 interaction model) was performed with Scrubber2 (BioLogic Software, Campbell, Australia).

Figure 5. A) Immunization protocol. First, adoptive transfer of splenic OT-II transgenic mouse cells into wild-type C57BL/6 mice was performed. One day later, mice were prime-immunized with either OVA or the OVA-LeX neoglycoconjugate. On day 21, boost immunization was performed with OVA or OVA-LeX. On day 30, mice were sacrificed and cellular and humoral immune responses were analyzed. B) OVA-specific antibody titers in mice immunized with OVA or OVA-LeX. C) The numbers of IL-2- and IFN-g-producing spleen cells from mice immunized with OVA or OVA-LeX were analyzed by ELISpot. sfu = spot-forming unit.

after GalT reaction) or A. aurantia lectin (AAL-555, 50 mg mL 1; after FucT reaction). Lectin solutions were prepared in PBS supplemented with CaCl2 (5 mm), MgCl2 (5 mm), and Tween-20 (0.05 %). The slide was washed and dried as above. Fluorescence was analyzed with the G265BA microarray scanner. Production of MGL-1-hFc: A recombinant fusion protein consisting of the extracellular part of MGL-1 and the Fc part of human IgG1 was stably expressed and produced in CHO cells as previously described.[37] The fusion protein was purified from the cell supernatant over a protein G column (GE Healthcare). The functionality of the recombinant fusion protein was verified by binding to two known ligands (Lewis X, a-GalNAc) in a glycan array as previously reported.[37] Binding studies on N-glycan microarray slides: The microarray slides were compartmentalized with a 16-well gasket (Fast Frame incubation chambers, Whatman). MGL-1-hFc (5 or 10 mg mL 1) in lectin binding buffer (HEPES (50 mm, pH 7.4), MgCl2 (5 mm), CaCl2 (5 mm)) with Tween-20 (0.01 %) was added to the microarrays, and these were incubated at 4 8C for 16 h, before washing in lectin binding buffer for 10 min. Microarrays were incubated in the dark with Alexa Fluor 488 Mouse Anti-Human IgG1 antibody (100 mg mL 1, Invitrogen/Life Technologies) in lectin buffer supplemented with Tween-20 (0.01 %) and BSA (0.5 %) at RT for 1 h. The slides were washed with lectin binding buffer for 10 min then with nanopure water. Fluorescence was detected with a microarray scanner. Surface plasmon resonance (SPR): For the preparation of neutravidin-coated surfaces, neutravidin (Thermo Scientific) was immobilized at a flow rate of 10 mL min 1. The CM5 chip (GE Healthcare) was activated by injection of a mixture of N-ethyl-N’-(diethylaminopropyl)carbodiimide (EDC) and NHS for 10 min, and functionalized

 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Conjugation of Lewis X to ovalbumin: LeX trisaccharide was synthesized by automated oligosaccharide synthesis as previously described.[37] LeX was conjugated to OVA (Hyglos, Bernried am Starnberger See, Germany) with a di-N-succinimidyl adipate (DSAP) linker as previously described.[45] For the conjugation, 1 mg of OVA-carbohydrate conjugate, was used with 2 mg of LeX. Samples were concentrated in a 10 kDa Amicon centrifugation filter unit (Merck Millipore) and dissolved in sterile PBS.

MALDI-TOF-MS: To characterize the average LeX conjugated to each OVA molecule, MALDI-TOF-MS analysis was performed on an Autoflex Speed mass spectrometer (Bruker Daltonics) equipped with smartbeam II laser optics (1 kHz). The instrument was controlled by FlexControl 3.3 software. Equal amounts of sample (1 mg mL 1 in water) and matrix 2’,4’,6’-trihydroxyacetophenone monohydrate (THAP; 50 nmol mL 1 in 50 % acetonitrile) were mixed, and 1 mL was spotted onto a 384-spot MALDI target (600 mm polished steel) and allowed to dry at RT. Internal calibration was performed by using a standard calibration mix. MS spectra were acquired in linear positive ion mode (m/z 25 000–65 000). Acquired spectra were baseline corrected and smoothed by using a Gaussian algorithm with FlexControl software (m/z 0.2 widths; one cycle). Cell stimulation assay: CD11c + cells were isolated from spleen of C57BL/6 mice, and T cells were isolated from the spleens of transgenic OT-II mice by magnetic-activated cell separation (MACS) by using CD11c MicroBeads and a Pan T cell Isolation Kit II, respectively (Miltenyi Biotec, Bergisch Gladbach, Germany). The purity of the cell subsets was analyzed by flow cytometry using an APC-labeled rat anti-mouse CD11c antibody (eBioscience, for purified CD11c+ cells) or an APC-eFluor780-labeled rat anti-mouse CD3 antibody (eBioscience, for purified T cells). Purified CD11c + cells (2  104 cells per well) were seeded on a round-bottomed 96-well cell culture plate in complete Iscove’s modified Dulbecco’s medium (FCS (10 %), l-glutamine (2 mm), penicillin-streptomycin (5 mm; PAN Biotech, Aidenbach, Germany), and gentamicin (50 mg mL 1)) and OVA or OVA-LeX (final concentration: 30 mg mL 1) were added. After 1 h, purified OT-II T cells (1  104 cells per well) were added. Cytokines in the supernatant were analyzed after a further 72 h. ELISA: The concentration of cytokines IL-2, IFN-g, and IL-4 in cell culture supernatant were determined by ELISA (Murine IL-2/IFN-y/ IL-4 ELISA Development Kits, PeproTech, Hamburg, Germany) as described by the manufacturer. Briefly, after blocking, supernatants ChemBioChem 2014, 15, 844 – 851

849

CHEMBIOCHEM FULL PAPERS

www.chembiochem.org

and standards were incubated on 96-well plates (Greiner, Solingen, Germany) coated with capture antibody. After incubation with biotinylated detection antibody, avidin-HRP (BD Bioscience, Heidelberg, Germany) was added. The ELISA was developed by adding freshly made ABTS substrate (2,2’-azino-bis(3-ethylbenzothiazoline6-sulfonic acid); AppliChem, Darmstadt, Germany) and color development was measured. The concentrations of cytokines were calculated from standards provided by the manufacturer. Repeated washing with PBS supplemented with Tween-20 (0.01 %) was performed between each step.

H2O2 (30 %, 5 mL) was added to the substrate solution (10 mL) directly before use. The reaction was stopped by washing with deionized water. The plates were air dried overnight, and the numbers of spots were detected in a Bioreader 5000 (Biosys, Karben, Germany).

Serum titers of OVA-specific IgGs were determined by direct ELISA. Briefly, OVA was coated onto a 96-well plate, and this was incubated overnight followed by blocking. Serial dilutions of serum were incubated at RT for 2 h before incubation with AP-conjugated goat anti-mouse IgG (H + L, Jackson Immunoresearch, West Grove, PA) for 2 h. Development was performed with freshly made p-nitrophenyl-phosphate (PNPP; Thermo Scientific), and color development was measured at 405 nm.

Acknowledgements

C57BL/6 and OT-II mice were housed and bred in the animal facility of the Federal Institute for Risk Assessment (BfR) in a temperatureand humidity-controlled room under specific pathogen-free conditions. Food and water were provided ad libitum. Animal experiments were performed in strict accordance with the German regulations of the Society for Laboratory Animal Science (GV-SOLAS) and the European Health Law of the Federation of Laboratory Animal Science Associations (FELASA). The protocol was approved by the Landesamt fr Gesundheit und Soziales (LAGeSo) Berlin. Immunization: Spleen cells were obtained from ten-week-old female OT-II transgenic mice. Cells (1.5  107 in PBS (100 mL)) were injected intravenously into female C57BL/6 mice. One day later, preimmunization sera were obtained from the recipient mice. These mice were then prime-immunized by intraperitoneal injection of PBS, 22 mg of OVA (in PBS), or an equivalent amount of OVA-LeX (in PBS). On day 21, boost immunization was performed with the same amounts of OVA or OVA-LeX conjugate. Sera were obtained on days 7, 13, 21, and 30 postimmunization. ELISpot: To determine the number of IL-2, IL-4 or IFN-g secreting cells in spleens from mice immunized with OVA-LeX conjugates, PVDF membranes in 96-well ELISpot plates (Millipore) were treated with ethanol (35 %) and washed three times with PBS. Capture antibody (diluted to 5 mg mL 1; BD) was added, and the plates were incubated at RT for 6 h. Subsequently, wells were washed twice with RPMI 1640 medium (Life Technologies), then nonspecific binding was blocked by adding RPMI 1640 (200 mL) supplemented with l-glutamine (2 mm), penicillin-streptomycin (5 mm), and gentamicin (50 mg mL 1) to each well and incubation at RT for 2 h. After the blocking medium had been decanted, OVA323–339 (100 mL of a 50 mg mL 1 solution; Innovagen, Lund, Sweden) or anti-mouse CD3/CD28 antibody (10 mg mL 1, eBioscience, cat. no. 16-0031-86 and 16-0281-83) was added to the wells. Subsequently, splenic cells (4  105) were added to each well, and the plates were incubated at 37 8C for 18 h. Wells were washed with PBS containing Tween-20 (0.05 %), followed by the addition of biotinylated detection antibodies (50 mL, 2 mg mL 1; BD) and incubation for 2 h. After the plates had been washed with PBS containing Tween-20 (0.05 %), avidin-HRP (50 mL, diluted 1:1000 in PBS with FCS (10 %); BD) was added, and the plates were incubated at RT for 1 h. After four washes with PBS containing Tween-20 (0.05 %) and two with PBS, the substrate solution was added. The substrate solution was prepared by diluting a stock solution of 3-amino-9-ethylcarbazole (10 mg mL 1 in DMF) in acetate buffer (0.1 m, pH 5.0; 1:30 dilution).  2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Statistical analysis: Statistical analyses were performed with the unpaired Student’s t-test in Prism (GraphPad Software, La Jolla, CA; p < 0.05 considered statistically significant).

M.E., M.M., M.K.S., P.H.S., and B.L. thank the Max Planck Society for generous support. Funding by the German Federal Ministry of Education and Research (to B.L.) and the SFB 765 (to P.H.S. and B.L.) is also gratefully acknowledged. We also thank Susanne Eisenschmidt for technical assistance with animal husbandry. N.C.R. and S.S. acknowledge generous funding from the Spanish Ministerio de Ciencia e Innovacin, (CTQ2008-04444/BQA, grant), the Government of the Basque Country (Etortek grants), the European Union (ITN-Euroglycoarrays grant), and from the Guipuzcoan Regional Government (Guipuzcoan program of Science, Technology and Innovation), and thank Prof. Iain Wilson (BOKU, Vienna) for supplying a clone for the FUT6 expression. Keywords: antigens · carbohydrates · C-type lectins · dendritic cells · multivalency · targeting [1] K. Ley, G. S. Kansas, Nat. Rev. Immunol. 2004, 4, 325 – 335. [2] M. Sperandio, C. A. Gleissner, K. Ley, Immunol. Rev. 2009, 230, 97 – 113. [3] G. A. Rabinovich, Y. van Kooyk, B. A. Cobb, Ann. N.Y. Acad. Sci. 2012, 1253, 1 – 15. [4] a) B. E. Collins, J. C. Paulson, Curr. Opin. Chem. Biol. 2004, 8, 617 – 625; b) J. Huskens, Curr. Opin. Chem. Biol. 2006, 10, 537 – 543; c) I. Papp, J. Dernedde, S. Enders, S. B. Riese, T. C. Shiao, R. Roy, R. Haag, ChemBioChem 2011, 12, 1075 – 1083; d) C. Fasting, C. A. Schalley, M. Weber, O. Seitz, S. Hecht, B. Koksch, J. Dernedde, C. Graf, E.-W. Knapp, R. Haag, Angew. Chem. Int. Ed. 2012, 51, 10472 – 10498; Angew. Chem. 2012, 124, 10622 – 10650. [5] a) N. Kanazawa, J. Dermatol. Sci. 2007, 45, 77 – 86; b) Y. van Kooyk, Biochem. Soc. Trans. 2008, 36, 1478 – 1481. [6] S. Burgdorf, A. Kautz, V. Bçhnert, P. A. Knolle, C. Kurts, Science 2007, 316, 612 – 616. [7] a) D. Sancho, C. Reis e Sousa, Annu. Rev. Immunol. 2012, 30, 491 – 529; b) M. Maglinao, R. Klopfleisch, P. H. Seeberger, B. Lepenies, J. Immunol. 2013, 191, 2551 – 2559. [8] P. J. Tacken, C. G. Figdor, Semin. Immunol. 2011, 23, 12 – 20. [9] a) L. C. Bonifaz, D. P. Bonnyay, A. Charalambous, D. I. Darguste, S.-I. Fujii, H. Soares, M. K. Brimnes, B. Moltedo, T. M. Moran, R. M. Steinman, J. Exp. Med. 2004, 199, 815 – 824; b) S. B. Boscardin, J. C. R. Hafalla, R. F. Masilamani, A. O. Kamphorst, H. A. Zebroski, U. Rai, A. Morrot, F. Zavala, R. M. Steinman, R. S. Nussenzweig, M. C. Nussenzweig, J. Exp. Med. 2006, 203, 599 – 606. [10] D. Dudziak, A. O. Kamphorst, G. F. Heidkamp, V. R. Buchholz, C. Trumpfheller, S. Yamazaki, C. Cheong, K. Liu, H.-W. Lee, C. G. Park, R. M. Steinman, M. C. Nussenzweig, Science 2007, 315, 107 – 111. [11] A. Engering, T. B. H. Geijtenbeek, S. J. van Vliet, M. Wijers, E. van Liempt, N. Demaurex, A. Lanzavecchia, J. Fransen, C. G. Figdor, V. Piguet, Y. van Kooyk, J. Immunol. 2002, 168, 2118 – 2126. [12] a) B. Lepenies, J. Lee, S. Sonkaria, Adv. Drug Delivery Rev. 2013, 65, 1271 – 1281; b) Y. van Kooyk, W. W. J. Unger, C. M. Fehres, H. Kalay, J. J. Garca-Vallejo, Mol. Immunol. 2013, 55, 143 – 145; c) W. W. J. Unger, A. J. van Beelen, S. C. Bruijns, M. Joshi, C. M. Fehres, L. van Bloois, M. I. Verstege, M. Ambrosini, H. Kalay, K. Nazmi, J. G. Bolscher, E. Hooijberg, T. D.

ChemBioChem 2014, 15, 844 – 851

850

CHEMBIOCHEM FULL PAPERS

[13] [14] [15] [16]

[17]

[18]

[19] [20]

[21]

[22]

[23] [24] [25] [26]

[27] [28] [29]

de Gruijl, G. Storm, Y. van Kooyk, J. Controlled Release 2012, 160, 88 – 95; d) J. J. Garca-Vallejo, W. W. J. Unger, H. Kalay, Y. van Kooyk, Oncoimmunology 2013, 2, e23040. O. Martnez-vila, K. Hijazi, M. Marradi, C. Clavel, C. Campion, C. Kelly, S. Penads, Chem. Eur. J. 2009, 15, 9874 – 9888. R. Kikkeri, B. Lepenies, A. Adibekian, P. Laurino, P. H. Seeberger, J. Am. Chem. Soc. 2009, 131, 2110 – 2112. S. I. van Kasteren, S. J. Campbell, S. Serres, D. C. Anthony, N. R. Sibson, B. G. Davis, Proc. Natl. Acad. Sci. USA 2009, 106, 18 – 23. a) N. Jayaraman, K. Maiti, K. Naresh, Chem. Soc. Rev. 2013, 42, 4640 – 4656; b) M. S. Macauley, F. Pfrengle, C. Rademacher, C. M. Nycholat, A. J. Gale, A. von Drygalski, J. C. Paulson, J. Clin. Invest. 2013, 123, 3074 – 3083. J. J. Garca-Vallejo, M. Ambrosini, A. Overbeek, W. E. van Riel, K. Bloem, W. W. J. Unger, F. Chiodo, J. G. Bolscher, K. Nazmi, H. Kalay, Y. van Kooyk, Mol. Immunol. 2013, 53, 387 – 397. J. Herre, A. S. J. Marshall, E. Caron, A. D. Edwards, D. L. Williams, E. Schweighoffer, V. Tybulewicz, C. Reis e Sousa, S. Gordon, G. D. Brown, Blood 2004, 104, 4038 – 4045. H. Huang, G. R. Ostroff, C. K. Lee, C. A. Specht, S. M. Levitz, mBio 2010, 1, e00164-10. a) M. Sato, K. Kawakami, T. Osawa, S. Toyoshima, J. Biochem. 1992, 111, 331 – 336; b) M. Tsuiji, M. Fujimori, Y. Ohashi, N. Higashi, T. M. Onami, S. M. Hedrick, T. Irimura, J. Biol. Chem. 2002, 277, 28892 – 28901. a) K. Saba, K. Denda-Nagai, T. Irimura, Am. J. Pathol. 2009, 174, 144 – 152; b) J. P. Upham, D. Pickett, T. Irimura, E. M. Anders, P. C. Reading, J. Virol. 2010, 84, 3730 – 3737; c) C. H. Hokke, M. Yazdanbakhsh, Parasite Immunol. 2005, 27, 257 – 264. a) C. Napoletano, A. Rughetti, M. P. Agervig Tarp, J. Coleman, E. P. Bennett, G. Picco, P. Sale, K. Denda-Nagai, T. Irimura, U. Mandel, H. Clausen, L. Frati, J. Taylor-Papadimitriou, J. Burchell, M. Nuti, Cancer Res. 2007, 67, 8358 – 8367; b) S. Ichii, Y. Imai, T. Irimura, J. Leukocyte Biol. 1997, 62, 761 – 770; c) S. K. Singh, I. Streng-Ouwehand, M. Litjens, D. R. Weelij, J. J. Garca-Vallejo, S. J. van Vliet, E. Saeland, Y. van Kooyk, Mol. Immunol. 2009, 46, 1240 – 1249. S. Ichii, Y. Imai, T. Irimura, Cancer Immunol. Immunother. 2000, 49, 1 – 9. S. K. Singh, I. Streng-Ouwehand, M. Litjens, H. Kalay, E. Saeland, Y. van Kooyk, Int. J. Cancer 2011, 128, 1371 – 1383. J. Heimburg-Molinaro, M. Lum, G. Vijay, M. Jain, A. Almogren, K. Rittenhouse-Olson, Vaccine 2011, 29, 8802 – 8826. a) A. I. Ko, U. C. Drger, D. A. Harn, Proc. Natl. Acad. Sci. USA 1990, 87, 4159 – 4163; b) M. Wuhrer, C. A. M. Koeleman, J. M. Fitzpatrick, K. F. Hoffmann, A. M. Deelder, C. H. Hokke, Glycobiology 2006, 16, 991 – 1006. P. Velupillai, E. A. dos Reis, M. G. dos Reis, D. A. Harn, Hum. Immunol. 2000, 61, 225 – 232. B. Lepenies, P. H. Seeberger, Immunopharmacol. Immunotoxicol. 2010, 32, 196 – 207. a) S. Boonyarattanakalin, X. Liu, M. Michieletti, B. Lepenies, P. H. Seeberger, J. Am. Chem. Soc. 2008, 130, 16791 – 16799; b) T. Horlacher, M. A.

 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.chembiochem.org

[30] [31] [32]

[33]

[34] [35]

[36] [37]

[38]

[39]

[40] [41] [42] [43]

[44] [45]

Oberli, D. B. Werz, L. Krçck, S. Bufali, R. Mishra, J. Sobek, K. Simons, M. Hirashima, T. Niki, P. H. Seeberger, ChemBioChem 2010, 11, 1563 – 1573; c) S. Serna, J. Etxebarria, N. Ruiz, M. Martin-Lomas, N.-C. Reichardt, Chem. Eur. J. 2010, 16, 13163 – 13175. Y. Zhang, Q. Li, L. G. Rodriguez, J. C. Gildersleeve, J. Am. Chem. Soc. 2010, 132, 9653 – 9662. Y. Liu, A. S. Palma, T. Feizi, Biol. Chem. 2009, 390, 647 – 656. a) A. R. de Boer, C. H. Hokke, A. M. Deelder, M. Wuhrer, Anal. Chem. 2007, 79, 8107 – 8113; b) E. Lonardi, C. I. A. Balog, A. M. Deelder, M. Wuhrer, Expert Rev. Proteomics 2010, 7, 761 – 774. a) A. Sanchez-Ruiz, S. Serna, N. Ruiz, M. Martin-Lomas, N.-C. Reichardt, Angew. Chem. Int. Ed. 2011, 50, 1801 – 1804; Angew. Chem. 2011, 123, 1841 – 1844; b) A. Beloqui, J. Calvo, S. Serna, S. Yan, I. B. H. Wilson, M. Martin-Lomas, N. C. Reichardt, Angew. Chem. Int. Ed. 2013, 52, 7477 – 7481; Angew. Chem. 2013, 125, 7625 – 7629. S. Serna, S. Yan, M. Martin-Lomas, I. B. H. Wilson, N.-C. Reichardt, J. Am. Chem. Soc. 2011, 133, 16495 – 16502. A. Canales, A. Mallagaray, J. Prez-Castells, I. Boos, C. Unverzagt, S. Andr, H.-J. Gabius, F. J. CaÇada, J. Jimnez-Barbero, Angew. Chem. Int. Ed. 2013, 52, 13789 – 13793; Angew. Chem. 2013, 125, 14034 – 14038. Z. Wang, Z. S. Chinoy, S. G. Ambre, W. Peng, R. McBride, R. P. de Vries, J. Glushka, J. C. Paulson, G.-J. Boons, Science 2013, 341, 379 – 383. a) M. K. Schlegel, J. Htter, M. Eriksson, B. Lepenies, P. H. Seeberger, ChemBioChem 2011, 12, 2791 – 2800; b) M. Maglinao, M. Eriksson, M. K. Schlegel, S. Zimmermann, T. Johannssen, S. Gçtze, P. H. Seeberger, B. Lepenies, J. Controlled Release 2014, 175, 36 – 42. S. Oo-puthinan, K. Maenuma, M. Sakakura, K. Denda-Nagai, M. Tsuiji, I. Shimada, S. Nakamura-Tsuruta, J. Hirabayashi, N. V. Bovin, T. Irimura, Biochim. Biophys. Acta Gen. Subj. 2008, 1780, 89 – 100. N. Parera Pera, H. M. Branderhorst, R. Kooij, C. Maierhofer, M. van der Kaaden, R. M. J. Liskamp, V. Wittmann, R. Ruijtenbeek, R. J. Pieters, ChemBioChem 2010, 11, 1896 – 1904. S. Yan, S. Serna, N.-C. Reichardt, K. Paschinger, I. B. H. Wilson, J. Biol. Chem. 2013, 288, 21015 – 21028. K.-C. Sheng, D. S. Pouniotis, M. D. Wright, C. K. Tang, E. Lazoura, G. A. Pietersz, V. Apostolopoulos, Immunology 2006, 118, 372 – 383. E. W. Adams, D. M. Ratner, P. H. Seeberger, N. Hacohen, ChemBioChem 2008, 9, 294 – 303. a) M. J. Barnden, J. Allison, W. R. Heath, F. R. Carbone, Immunol. Cell Biol. 1998, 76, 34 – 40; b) J. M. Robertson, P. E. Jensen, B. D. Evavold, J. Immunol. 2000, 164, 4706 – 4712. J. Gordon, L. D. MacLean, Nature 1965, 208, 795 – 796. C. E. Martin, F. Broecker, M. A. Oberli, J. Komor, J. Mattner, C. Anish, P. H. Seeberger, J. Am. Chem. Soc. 2013, 135, 9713 – 9722.

Received: December 5, 2013 Published online on March 11, 2014

ChemBioChem 2014, 15, 844 – 851

851