Multivalent glycopeptide dendrimers for the targeted delivery of antigens to dendritic cells

Molecular Immunology 53 (2013) 387–397

Contents lists available at SciVerse ScienceDirect

Molecular Immunology journal homepage: www.elsevier.com/locate/molimm

Multivalent glycopeptide dendrimers for the targeted delivery of antigens to dendritic cells Juan J. García-Vallejo a,∗ , Martino Ambrosini a , Annemieke Overbeek a , Wilhelmina E. van Riel a,1 , Karien Bloem a , Wendy W.J. Unger a , Fabrizio Chiodo b , Jan G. Bolscher c , Kamran Nazmi c , Hakan Kalay a , Yvette van Kooyk a a

Department of Molecular Cell Biology & Immunology, Faculty of Medicine, VU University Medical Center, Amsterdam, The Netherlands Laboratory of GlycoNanotechnology, Biofunctional Nanomaterials Unit, CIC biomaGUNE, San Sebastián, Spain c Department of Oral Biochemistry, Academic Centre for Dentistry Amsterdam, University of Amsterdam and VU Amsterdam, Amsterdam, The Netherlands b

a r t i c l e

i n f o

Article history: Received 7 August 2012 Received in revised form 8 September 2012 Accepted 23 September 2012 Available online 24 October 2012 Keywords: Human Rodent Dendritic cells Cell surface molecules

a b s t r a c t Dendritic cells are the most powerful type of antigen presenting cells. Current immunotherapies targeting dendritic cells have shown a relative degree of success but still require further improvement. One of the most important issues to solve is the efficiency of antigen delivery to dendritic cells in order to achieve an appropriate uptake, processing, and presentation to Ag-specific T cells. C-type lectins have shown to be ideal receptors for the targeting of antigens to dendritic cells and allow the use of their natural ligands – glycans – instead of antibodies. Amongst them, dendritic cell-specific ICAM-3-grabbing non-integrin (DC-SIGN) is an interesting candidate due to its biological properties and the availability of its natural carbohydrate ligands. Using Leb -conjugated poly(amido amine) (PAMAM) dendrimers we aimed to characterize the optimal level of multivalency necessary to achieve the desired internalization, lysosomal delivery, Ag-specific T cell proliferation, and cytokine response. Increasing DC-SIGN ligand multivalency directly translated in an enhanced binding, which might also be interesting for blocking purposes. Internalization, routing to lysosomal compartments, antigen presentation and cytokine response could be optimally achieved with glycopeptide dendrimers carrying 16–32 glycan units. This report provides the basis for the design of efficient targeting of peptide antigens for the immunotherapy of cancer, autoimmunity and infectious diseases. © 2012 Elsevier Ltd. All rights reserved.

1. Introduction Dendritic cells (DCs) are the most powerful antigen-presenting cells in the induction and modulation of antigen-specific immune responses, which makes them ideal candidates for the development of immunotherapies against cancer, autoimmunity, and infectious

Abbreviations: CLR, C-type lectin receptor; DC, dendritic cell; DC-SIGN, DCspecific ICAM-3 grabbing non-integrin; LeX, Y, a, or b , LewisX, Y, a, or b ; MR, mannose receptor; PAMAM, poly(amido amine). ∗ Corresponding author at: Department of Molecular Cell Biology & Immunology, VU University Medical Center, FdG, B242, PO Box 7057, 1007 MB Amsterdam, The Netherlands. Tel.: +31 20 4448529/48080; fax: +31 20 4448081. E-mail addresses: [email protected] (J.J. García-Vallejo), [email protected] (M. Ambrosini), [email protected] (A. Overbeek), [email protected] (W.E. van Riel), [email protected] (K. Bloem), [email protected] (W.W.J. Unger), [email protected] (F. Chiodo), [email protected] (J.G. Bolscher), [email protected] (K. Nazmi), [email protected] (H. Kalay), [email protected] (Y. van Kooyk). 1 Current address: Department of Cell Biology, Faculty of Science, Utrecht University Kruytgebouw, Padualaan 8, Utrecht, The Netherlands. 0161-5890/$ – see front matter © 2012 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.molimm.2012.09.012

diseases. DCs reside in peripheral tissues in an immature state. This state is characterized by an enhanced capacity to sample the environment for pathogens. In the steady state DCs migrate continuously to the primary lymphoid tissue and, upon pathogen recognition, their migration rate increases (Carbone et al., 2004). Another effect of pathogen recognition is the activation of receptors such as TLRs, which initiate signaling events that results in the maturation of DCs (Joffre et al., 2009). The maturation process ensures that DCs up-regulate the expression of co-stimulatory molecules (signal 2) and cytokines (signal 3) that are required for T-cell activation (Steinman and Banchereau, 2007). In parallel with this process, the captured antigen is taken through a maze of endosomes until it reaches acidic lysosomal compartments were it can be conveniently processed for presentation on MHC-I and MHC-II. Antigen processing for presentation on MHC-II is a relatively wellcharacterized process (Steinman and Banchereau, 2007; Neefjes et al., 2011; Paul et al., 2011). Besides classical MHC-II antigen presentation, DCs also process antigens for presentation on MHCI, a phenomenon known as cross-presentation. This phenomenon remains poorly defined and attracts a great interest, since it has

388

J.J. García-Vallejo et al. / Molecular Immunology 53 (2013) 387–397

important implication in the understanding but also the manipulation of cytotoxic anti-tumor T cell responses (Kurts et al., 2010; Heath and Carbone, 2001). Preferred strategies for the targeting of the antigens to DCs in immunotherapy are based on the direct delivery of the antigen to DCs in vivo. This approach requires that the antigen vehicle recognize a specific marker on DCs that is able to induce internalization of the antigenic cargo. Previous research has focused on C-type lectin receptors (CLRs), such as DEC-205 (Hawiger et al., 2001; Bonifaz et al., 2002; Boscardin et al., 2006; Kretschmer et al., 2005), the DC-specific ICAM-3-grabbing non-integrin (DC-SIGN) (Singh et al., 2009; Unger et al., 2012), mannose receptor (MR) (Burgdorf et al., 2007), CLEC9A (Sancho et al., 2008; Caminschi et al., 2008) and Langerin (Idoyaga et al., 2011). Since these receptors often show different expression profiles amongst DC subsets (Cao et al., 2007; Dudziak et al., 2007; Schreibelt et al., 2012; Robinson et al., 2006; Kanazawa, 2007), targeting CLRs provides the opportunity to specifically target a single or multiple DC subsets (Dudziak et al., 2007). Additionally, multiple CLRs have been shown to mediate efficient internalization and routing of their ligands into acidic compartments involved in antigen processing. Probably the most complete set of studies have been performed on DEC-205 using the monoclonal antibody NLDC145 (Kraal et al., 1986). Injection of mice with NLDC-145/antigen complexes resulted in draining lymph nodes filled with antigen-loaded DCs, which efficiently presented to antigen-specific T cells leading to the development of regulatory T cells (eight days post-challenge) and the elimination of effector T cells (three weeks post-challenge) (Mahnke et al., 2003; Bonifaz et al., 2002; Hawiger et al., 2001). A disadvantage of the use of monoclonal antibodies for DC targeting is that, even when humanized, they still elicit adverse immune reactions that can decrease the efficiency of the immunotherapy but also induce severe autoimmune side effects (Wang et al., 2012). An alternative to the use of CLR-specific antibodies is the use of CLR ligands, which lack the problem of immunogenicity and can be synthesized or purified in large amounts at a relatively low cost. A CLR with known glycan specificity and capable of mediating T cell responses is DC-SIGN (Unger and van Kooyk, 2011). The natural ligands of DC-SIGN comprise high-mannose oligosaccharides and Lewis-type epitopes, such as LewisX (LeX ), LewisY (LeY ), Lewisa (Lea ) and Lewisb (Leb ) (Appelmelk et al., 2003; van Liempt et al., 2006). In humans, DC-SIGN is predominantly expressed on DCs at mucosal sites, as well as in skin and lymph nodes (Engering et al., 2004). Glycan recognition by DC-SIGN occurs through its carbohydrate recognition domain, it is a Ca2+ -dependent event, and results in fast and efficient receptor internalization and trafficking to the lysosomes (Engering et al., 2002; Geijtenbeek et al., 2000b). Previous work has demonstrated that the preparation of DCSIGN ligands in multivalent systems increases affinity for the receptor and competes for the recognition of several DC-SIGNspecific pathogens (Berzi et al., 2012; Sattin et al., 2010; Luczkowiak et al., 2011). These data, besides a promising strategy in the prevention of HIV and other DC-SIGN-specific pathogens infections, raises the question of whether multimeric glycan-based compounds that target DC-SIGN provide efficient platforms for the delivery of antigen to DCs. One such type of multivalent platform for antigen targeting is PAMAM dendrimers. PAMAM dendrimers are symmetric highly branched monodisperse polymers with a compact spherical structure and are commercially supplied with their functional groups in an activated form, allowing the design and development of complex multivalent structures by simple chemical reactions (Biricova and Laznickova, 2009). Using DC-ligand conjugated PAMAM dendrimers we aimed to characterize the optimal level of multivalency to achieve the desired internalization, lysosomal delivery, Ag-specific T cell proliferation and cytokine response.

2. Materials and methods 2.1. Synthesis of glycopeptide dendrimers tert-Butoxycarbonyl tert-butyl carbonate, triethylamine, cystamine dihydrochloride, NaH2 PO4 , NaOH, succinic anhydride, dimethylaminopyridine, trifluoroacetic acid, acetic acid, phenol, triisopropylsilane, picoline borane, PAMAMgeneration 3.0 amino dendrimer, Tris(2-carboxyethyl)phosphine hydrochloride, 7-mercapto-4-methylcoumarine, Alexa 488-NHS ester were purchased from Sigma–Aldrich. Sulfosuccinimidyl 4-[N-maleimidomethyl]cyclohexane-1-carboxylate and 4-(4N-maleimidophenyl)butyric acid hydrazide hydrochloride, were purchased from Pierce (Thermo Fisher Scientific). L602 (lacto-N-difucohexaose) and maltohexaose were purchased from Dextralab. Methanol, diethyl ether, ethyl acetate, dichloromethane, dimethylformamide, 2-propanol, acetonitrile, DMSO and the FmocProtected amino acids were purchased from Biosolve. HATU was purchased from IRIS Biotech. 4 -[2-[ 2-(tert-Butoxycarbonylamino)ethyldisulfanyl]ethylamino]-4-oxo-butanoic acid (compound 1) was prepared in a gram scale as previously described (Li and Zeng, 2007). The compound structure and purity was confirmed by NMR and ESI-MS analysis. The peptide sequences SIINFEKL (Barnden et al., 1998) (compound 2 or CKOT1) and ISQAVHAAHAEINEAGR (Hogquist et al., 1994) (compound 2bis or CKOT2) were prepared on an automated peptide synthesizer (Applied Biosystems) in a 50 ␮mol scale, using a double-coupling Fmoc-strategy of solid-phase peptide synthesis. An additional Cys, at the start of the sequence and Lys, flanking the epitopes, were added to facilitate conjugation to the dendrimer and release of the antigenic epitope during peptide processing, respectively. A 4-fold excess of Fmoc-protected amino acids were coupled to the chemmatrix resin using HATU as the activating agents. After the final coupling and the subsequent removal of the Fmoc group the resin was washed with DMF, isopropyl alcohol and DCM and then dried under a flux of nitrogen. The dried resin containing the protected amino acid sequence was used in the coupling reaction with (1). Peptidyl resin (50 ␮mol) was swelled in DMF for 10 min, and then washed with DMF (2 × 5 ml). To the peptidyl resin were added 3 ml of a solution containing 200 ␮mol of (1), 1.2 mmol of DIPEA, 1.2 mmol of HATU previously stirred for 30 min. The reaction mixture was shaken during 90 min at room temperature and then filtered to give a dark yellow resin. The resin was washed to remove unreacted compounds, first with DMF until the filtrate was colorless, then with isopropanol and DCM, before being dried under a flux of nitrogen for 10 min. Cleavage and deprotection was carried out by treatment of the dried resin with 3 ml of a mixture of TFA/phenol/TIS/H2 O, 92/3/2/3 (w/w) at room temperature for 3 h. The solution was collected, the resin was washed with TFA (2 × 2 ml), and the combined filtrates, after partial concentration under a flux of nitrogen to approx. 4–5 ml, were added to ice-cold diethyl ether (35 ml). The white-yellow precipitate was washed with diethyl ether (3 × 20 ml) and dried under a flux of nitrogen. Water was added (7 ml) and then the solution was lyophilized. The purification of the compounds 3 and 3bis (CKOT1/2-Linker) was achieved by reverse-phase HPLC on a Vydac HPLC C18 Mass Spec column (10 mm × 250 mm, 5 ␮m) (Grace Division Discovery Sciences) using a solvent system of water/acetonitrile both containing 0.1% TFA, with a linear gradient from 5% to 60% in 60 min. The fraction containing the peptide was collected and lyophilized (70% yield for compound 3 and 55% for compound 3bis). A solution of 4-(4-N-maleimidophenyl)butyric acid hydrazide hydrochloride (3 eq.) and PicolineBorane (10 eq.), dissolved in DMSO/AcOH (8:2; 100 ␮l), was added to Lewisb (L602: lacto-Ndifucohexaose; 1 mg, 1 eq., Dextra Laboratories), or maltohexaose (G602, 1 mg, 1 eq., Dextra Laboratories). The resulting mixture was

J.J. García-Vallejo et al. / Molecular Immunology 53 (2013) 387–397

stirred at 65 ◦ C for 2 h. After cooling to room temperature, ice-cold iso-propanol (1.4 ml) was added and the mixture was cooled at −20 ◦ C for 30 min. After deposition of the precipitate through centrifugation at 14,000 rpm for 10 min, the solution was removed and more iso-propanol was added. The washing procedure with iso-propanol was repeated two more times, removing the solvent, mixing and centrifuging. After removal of the last portion of solvent, compound 4 (MPBH-Glycan) was recovered as white solid and used immediately in the final step. Sulfosuccinimidyl 4-[N-maleimidomethyl]cyclohexane-1carboxylate (0.337 mg, 35 eq., 1.1 eq./arm) was dissolved in 150 ml of DMSO and this solution was added to the PAMAM amino dendrimer (0.2 mg, 28.9 nmol, Sigma). The mixture was heated at 35 ◦ C for 2 h, then the modified peptide (compounds 3 and 3b) (CKOT1/2-Linker, 96 eq.) was added and the mixture was left reacting overnight at room temperature. The mixture was diluted with 0.85 ml of water and 0.1 ml of TCEP was added. The purification was performed by size-exclusion chromatography over a GE Healthcare PD10 column, elution with PBS buffer, giving compounds 5 and 5bis (Dendrimer-CKOT1/2-Linker) dissolved in 2 ml of PBS. Compounds 5 and 5bis were added to the activated glycan (4) and the mixture was left stirring at room temperature for 2.5 h. The purification was performed by size exclusion chromatography over a GE Healthcare PD10 column, eluting with PBS buffer, giving the final compounds (6 and 6bis, OT-I and OT-II glycopeptide dendrimer) dissolved in 3 ml of PBS. Mass spectrometry was used throughout the synthesis process to verify all the intermediate compounds. Mass characterization of the final product was done by MALDI-TOF mass spectrometry carried out on in linear mode using 7-mercapto-4-methylcoumarine as matrix. Samples were dissolved in PBS at a concentration of 2 mg/ml. The matrix was dissolved in MeOH:CH3 CN (1:1,5) to achieve a saturated solution (20 mg/ml). Sample (2 ␮l) and matrix (2 ␮l) were mixed and then 1 ␮l was spotted on the MALDI plate. 2.2. Cells 2.2.1. Human monocyte-derived dendritic cells Monocytes were isolated from the blood of healthy donors (Sanquin, The Netherlands) through a sequential Ficoll/Percoll gradient centrifugation. Isolated monocytes (purity, >85%) were cultured in RPMI 1640 (Invitrogen, USA) supplemented with 10% FCS (BioWhittaker, USA), 1000 U/ml penicillin (BioWhittaker, USA), 1000 U/ml streptomycin (BioWhittaker, USA), and 2 mM glutamine (BioWhittaker, USA) in the presence of interleukin-4 (IL-4) (500 U/ml; BioSource, Belgium) and granulocyte-macrophage colony-stimulating factor (GM-CSF) (800 U/ml; BioSource, Belgium) for 7 days (Sallusto and Lanzavecchia, 1994). Dendritic cells differentiation was confirmed by flow cytometric analysis (FACScan, BD biosciences, USA) of the expression of DC-SIGN using the monoclonal antibody AZN-D1 (Geijtenbeek et al., 2000b) followed by staining with a secondary FITC-labeled anti-mouse antibody (Zymed, San Francisco, CA). 2.2.2. K562/DC-SIGN and K562/DC-SIGNLL/Y Stable K562/DC-SIGN and K562/DC-SIGNLL/Y transfectants (Engering et al., 2002) were maintained in RPMI 1640 medium containing 10% Fetal Calf Serum 1000 U/ml penicillin (BioWhittaker, USA), 1000 U/ml streptomycin (BioWhittaker, USA), and 2 mM glutamine (BioWhittaker, USA). K562/DC-SIGNLL/Y cells carry mutations in the intracellular domain (L14A, L15A, Y31A) that render DC-SIGN unable to mediate internalization, while K562/DC-SIGN carry the wild-type cDNA for DC-SIGN and are able to mediate internalization. DC-SIGN expression was regularly selected using 1 mg/ml Geneticin (Invitrogen, USA). To check for

389

DC-SIGN expression, cells were incubated with primary antibody (AZN-D1 (Geijtenbeek et al., 2000b)) followed by staining with a secondary FITC-labeled anti-mouse antibody (Zymed, USA) and analyzed by flow cytometry on a FACScan (BD Biosciences, USA). 2.3. Soluble DC-SIGN/Fc binding assay DC-SIGN/Fc consists of the extracellular portion of DC-SIGN (residues 64–404) fused at the C-terminus to a human IgG1 /Fc fragment into the Sig-pIgG1 -Fc vector (Fawcett et al., 1992). DC-SIGN-Fc was produced in Chinese hamster ovary K1 cells by co-transfection of DC-SIGN-Sig-pIgG1 Fc (20 ␮g) and pEE14 (5 ␮g) vector. The soluble DC-SIGN/Fc adhesion assay was performed as follows. Dendrimers were coated at the indicated concentrations onto NUNC maxisorb 96-well plates (NUNC, Denmark) (100 ␮l per well) in PBS for 18 h at room temperature, followed by blocking with 1% bovine serum albumin (BSA) for 30 min at 37 ◦ C. Soluble DCSIGN/Fc supernatant was added for 2 h and incubated under mild shaking at room temperature. Unbound DC-SIGN/Fc was washed away and binding was determined by anti-IgG1 ELISA (horseradish peroxidase-labeled goat anti-human IgG, Fc␥ fragment specific, from Jackson ImmunoResearch, USA). 2.4. Enzyme-linked immuno sorbent assay (ELISA) Dendrimers were coated to NUNC maxisorb 96-well plates (NUNC, Denmark) overnight (100 ␮l per well) in PBS for 18 h at room temperature, followed by blocking with 1% BSA for 30 min at 37 ◦ C. After extensive washing anti-Lewisb antibodies (Calbiochem, Germany) were added and incubation proceeded for 2 h at room temperature while shaking. Peroxidase-labeled F(ab )2 fragment goat anti-mouse IgG, Fc␥ fragment specific antibody (Jackson ImmunoResearch, USA) was used as secondary step antibody. Signal detection was achieved by incubation with 1.3 mM H2 O2 in the presence of TMB (3,3 ,5,5 -tetramethylbenzidine) in 0.1 M sodium acetate–citrate buffer until the development of the reaction. The reaction is then stopped by addition of 1 M H2 SO4 and the absorbance at 450 nm is then measured with the help of a colorimeter (BioRad, The Netherlands). 2.5. Cellular DC-SIGN binding assay Dendrimer binding to cellular DC-SIGN was evaluated using atto488-conjugated dendrimers and K562/DC-SIGNLL/Y . Cells were incubated at 37 ◦ C for 1 h with a fixed concentration of dendrimer in the presence of a titration of Lewisb glycans. Cells were then washed and the amount of atto488 fluorescence per cell measured by flow cytometry (FACS Calibur, BD Biosciences, USA). The concentration of Lewisb glycans able to inhibit 50% binding (IC50 ) of the dendrimers was used to compare the avidity of the different dendrimer generations. The IC50 was calculated with the help of the GraphPAD v4.0 software (GraphPAD Software Inc., USA). 2.6. Imaging flow cytometry K562/DC-SIGN and K562/DC-SIGNLL/Y cells were incubated in the presence of dendrimers at 37 ◦ C for 60 min. Cells were then washed, fixated and prepared for acquisition on the ImageStreamX (Amnis Corp., Seattle) imaging flow cytometer. For the intracellular routing assay, dendritic cells were incubated in the presence of dendrimers at either 4 ◦ C or 37 ◦ C for 60 min. Cells were then washed, fixated, permeabilized and stained with monoclonal antibodies against the early endosomal marker EEA1 (Mu et al., 1995) and the lysosomal marker LAMP1 (Viitala et al., 1988). Both antibodies were from Abcam (UK). An AF594-labeled goat anti-mouse antibody (Invitrogen, USA) was used as a secondary staining. The

390

J.J. García-Vallejo et al. / Molecular Immunology 53 (2013) 387–397

following laser powers were used for the internalization assay: 488 nm at 20 mW and 785 nm at 4.5 mW. The laser powers for the intracellular routing assay were the following: 488 nm at 20 mW, 561 nm at 100 mW, 785 nm at 4.5 mW. Brightfield illumination was set at 800 mW before the acquisition of each sample. Brightfield images were collected in channels 1 and 9. Channels 2 (Atto488) and 6 (granularity) were habilitated for the internalization assay, while channels 2 (Atto488), 4 (AF594), and 6 (granularity) were habilitated for the intracellular routing assay. Cells were acquired at 40× magnification and on the basis of their area (area = the number of pixels in an image reported in square microns). Minimum area for acquisition was set to 50 pixels. A minimum of 15,000 cells was acquired per sample at a flow rate ranging between 50 and 100 cells/s. At least 2000 cells were acquired from single stained samples to allow for compensation. Compensation samples were acquired with all channels habilitated and with the brightfield illumination and the 785 nm laser switched off. A minimum of 5000 cells from the single stained samples were acquired with the same settings as experimental samples to control for over/under compensation. Analysis was performed using the IDEAS v5.0 software (Amnis Corp., Seattle). A compensation table was generated using the compensation macro built in the software. Single stained samples were manually gated for accurate calculation of spectral overlap coefficients (Ortyn et al., 2006). Once the compensation table was calculated, it was applied to the single staining samples that were acquired using the same settings as experimental samples. Proper compensation was then verified by visualizing samples in bivariate fluorescence intensity plots (data not shown). A template analysis file was generated then that include an area vs aspect ratio intensity plot and a gradient root mean square histogram of one of the brightfield channels (channels 1 and 9). Area is the number of squared microns of the cells, while the aspect ratio intensity index is the result of dividing the minor axis (intensity-weighted) by the major axis (intensity weighted) and describes how round or oblong an object is, but also indicates if there are doublets in a population of normally circular cells. The gradient RMS feature measures the sharpness quality of an image by detecting large changes of pixel values in the image and is useful for the selection of focused objects. The gradient RMS feature is computed using the average gradient of a pixel normalized for variations in intensity levels. Using these features a population of focused single cells (SC/F) was gated. This template, together with the corresponding compensation table was applied to all the experimental samples acquired. Each of the data files generated was opened and the SC/F population gated to a new compensated image file. Compensated image files were then merged into the final analysis file. This file allows for the direct comparison of features amongst the different dendrimers. To calculate the internalization of the dendrimers, a mask was designed that characterizes only the intracellular space of the cells. This mask was based on the use of the morphology feature applied to the brightfield image on channel 1, and then eroded until the membrane was left out of the mask. Since cells are gated on a certain level of focusing it is possible to assume that the image acquired represents, in all cells, a 4 ␮m cross-section of the major circumference (Ortyn et al., 2007). At this location, the thickness of the membrane is similar in all cells and allows us to design a mask based exclusively on brightfield images. This is a major advantage over the use of an extracellular fluorescent marker, which introduces new challenges in the experiment: the use of an additional channel complicates the compensation process and the selection of a marker that is exclusively located in the extracellular membrane during the internalization process of the dendrimers is a difficult task. The intracellular mask was then used to calculate the feature internalization applied to channel

2 (Atto488). The internalization score is a log-scaled ratio of the intensity inside the cell (intracellular mask) respect the intensity of the entire cell. Cells that have internalized antigen typically have positive scores while cells that show the antigen still on the membrane have negative scores. Cells with scores around 0 have similar amounts of antigen on the membrane and in intracellular compartments. Colocalization is calculated as the logarithmic transformation of Pearson’s correlation coefficient of the localized bright spots with a radius of 3 pixels or less within the whole cell area in the two input images (bright detail similarity R3). Since the bright spots in the two images are either correlated (in the same spatial location) or uncorrelated (in different spatial locations), the correlation coefficient varies between 0 (uncorrelated) and 1 (perfect correlation). The logarithmic transformation of the correlation coefficient allows the use of a wider range for the colocalization score. In general, cells with a low degree of colocalization or no colocalization at all between two probes show scores below 1. Since only molecules that have internalized are able to show colocalization with intracellular compartment markers, bivariate plots depicting the internalization score (Y axis) and the colocalization score (X axis) provide the best representation of data. In this scatter plot, a gate representing the cells that have internalized the probe and show colocalization was calculated and the percentage of cells within that plot calculated.

2.7. Antigen presentation to murine CD4+ and CD8+ T cells OVA-specific CD4+ and CD8+ T cells were isolated from spleen and lymph node cell suspensions from OT-II (Barnden et al., 1998) and OT-I (Hogquist et al., 1994) mice respectively. Lymph nodes and spleen were mechanically dissociated to single-cell suspensions and pelleted. Erythrocytes were lysed and the cells were passed through a cell strainer. CD4+ and CD8+ T cells were isolated from the suspensions using the Dynal mouse CD4/CD8 negative isolation kit (Invitrogen, USA) according to the manufacturer’s protocol. Bone marrow-derived DCs from human DC-SIGN transgenic mice (Schaefer et al., 2008) were incubated with the indicated concentrations of glycopeptide dendrimers in round-bottom 96 wells plates for 4 h. After extensive washing, either CD4+ T cells from OT-II mice or CD8+ T-cells from OT-I mice were added to each well. Forty-eight hours later, [3 H]-thymidine (1 ␮Ci/well; Amersham Biosciences, USA) was added for 16 h to detect incorporation into DNA of proliferating T cells. Cells were harvested onto filters and [3 H]-thymidine incorporation was assessed using a beta counter.

2.8. DC maturation assay Human DC maturation was assessed by the cell surface levels of the maturation marker CD83 (anti-CD83-PE, Beckman-Coulter, Denmark), and the co-stimulatory molecules CD80 (anti-CD80-PE, BD Biosciences, USA), and CD86 (anti-CD86-PE, BD Biosciences, USA). LPS-treated DCs were used as positive controls, since LPS is known to induce a full maturation of DC.

2.9. IL-10 ELISA For the detection of IL-10, culture supernatants were harvested 24 h after DC stimulation and frozen at −80 ◦ C until analysis. IL-10 was measured by ELISA using antibody pairs from eBioscience (The Netherlands) and according to manufacturer’s protocol.

J.J. García-Vallejo et al. / Molecular Immunology 53 (2013) 387–397

391

Fig. 1. Multivalency enhances avidity of DC-SIGN for its ligands. (A) Glycopeptide dendrimers are synthesized sequentially over PAMAM dendrimers functionalized with primary amines. A maleimide linker is used to attach the C-terminal Cys of the CKOTI/II peptide to the dendrimer. At the other end a glycan is attached to the N-terminus via reductive amination to the side chain of a Lys. (B) OT-II dendrimers were coated at 100 nM on to ELISA plates and binding to DC-SIGN/Fc tested by a soluble DC-SIGN/Fc binding assay. (C) OT-II dendrimers were coated at 4 nM on to ELISA plates and binding to DC-SIGN/Fc tested by a soluble DC-SIGN/Fc binding assay. (D) Mannose IC50 of the different atto488-labeled generation 3 OT-II Leb glycopeptide dendrimers binding to K562/DC-SIGNLL/EEE cells. Data represent the mean + S.D. of triplicate wells. This experiment is representative of at least three independent experiments.

3. Results 3.1. Ligand multivalency increases the avidity of DC-SIGN for its ligands We have previously demonstrated that glycan modification of OVA with DC-SIGN targeting glycans, such as LewisX or Lewisb , results in an efficient delivery of antigen to bone marrow-derived and splenic DCs (Singh et al., 2009). It has been proposed that innate immunity lectins on antigen-presenting cells bind to pathogens by a density-dependent recognition mechanism of surface glycans (Dam and Brewer, 2010). For this mechanism to be able to discriminate between host and foreign glycans the concept of multivalency is crucial. Thus, unique “weak” glycan epitopes on foreign pathogens appear to be strong epitopes when presented in highdensity, such as polysaccharides and lipopolysaccharides (Dam and Brewer, 2010). Dendrimers provide a unique platform for the design of multivalent compounds with a defined geometric orientation. PAMAM dendrimers are polymers with a high degree of molecular uniformity, narrow molecular weight distribution, specific size and shape characteristics, and a highly-functionalized terminal surface. The manufacturing process is a series of repetitive chemical steps starting with a central initiator core. Each subsequent growth step represents a new generation of polymer with twice the number of reactive surface sites. We conjugated Lewisb (Leb ) glycopeptides to PAMAM dendrimer generations 0–7 (4–512 reactive groups) in order to prepare compounds with the same chemical structure but increasing ligand density (Fig. 1A). To use as negative controls, maltohexose (MH) glycopeptides and nonglycosylated (NG) peptides were generated. The mass of each of the resulting 48 compounds was resolved by MALDI-TOF mass spectrometry analysis. The obtained mass corresponded in all cases to

approximately 50–60% of the expected mass, indicating that the site occupancy was consistently around 50%. Additionally, a certain level of polydispersity (approximately 10% of average mass) was observed in each measurement. Beyond generation 6 it was not possible to determine the mass, since it was above the detection limit of the mass spectrometer. Yet, we assumed that the chemistry had worked evenly, irrespective of the number of reactive groups, since the reactions were scaled accordingly. As a final control for the synthesis and functionality of the dendrimers, we assessed reactivity against an anti-Lewisb antibody by ELISA. As shown in Supplementary Fig. 1A and 1B, all Leb glycopeptide dendrimers were recognized by the Lewisb -specific antibody, while MH and NG dendrimers gave absorbances below the detection limit. We then tested the binding of a DC-SIGN/Fc chimera to the glycopeptide dendrimers by ELISA. Glycopeptide dendrimers were coated overnight at 4 ◦ C at a concentration of 100 nM. Detection was achieved using a DC-SIGN/Fc chimera followed by a horseradish peroxidase-labeled goat anti-human Fc antibody. The detection reaction was allowed until maximal color development was achieved. Results (Fig. 1B) indicate that, at this coating concentration, all generations of Leb glycopeptide dendrimers were efficiently recognized by DC-SIGN/Fc. MH or NG dendrimers were not recognized by DC-SIGN/Fc. Results were almost identical for both OT-I and OT-II glycopeptide dendrimers. The DC-SIGN/Fc binding assay was then repeated on dendrimers coated at a lower concentration (4 nM). At this concentration, binding of DC-SIGN to the dendrimers was directly proportional to the density of glycan per dendrimer, as shown in Fig. 1C. Again, results were similar for both OT-I and OT-II glycopeptide dendrimers. The DC-SIGN/Fc chimera consists of a DC-SIGN carbohydrate recognition domain fused to the Fc domain of a human Ig. The construct therefore displays two DC-SIGN carbohydrate recognition

392

J.J. García-Vallejo et al. / Molecular Immunology 53 (2013) 387–397

domains instead of the Ig variable domains. The crystal structure of DC-SIGN, however, indicates that DC-SIGN is expressed as a tetramer (Mitchell et al., 2001) on the surface of dendritic cells. To investigate how multivalency affected interaction with cell-expressed tetrameric DC-SIGN, we set up binding assays to DC-SIGN-expressing K562 cells. Because DC-SIGN is an efficient internalization receptor (Engering et al., 2002) and this could obscure the results obtained, we used a mutant form of DC-SIGN that lacks the di-leucine and the triple acidic motifs in its intracellular domain (Engering et al., 2002). Glycopeptide dendrimers and control dendrimers were labeled with a fluorochrome (Atto488) to allow detection by flow cytometry. After purification, labeling efficiency was assessed by relating the fluorescence emission at 520 nm to the absorbance at 280 nm. All dendrimers were efficiently labeled with Atto-488, however the labeling was not homogeneous for all the compounds, forbidding direct comparison of the compounds. Therefore, we decided to assess the avidity of the interaction by an inhibition assay. Mutant DC-SIGN-expressing cells were incubated with equimolar concentrations of the different glycopeptide dendrimers in the presence of increasing concentrations of mannose, ranging from 0.05 to 100 mM, at 37 ◦ C for 1 h. Cells were then washed in ice-cold TSM and measured by flow cytometry. The decay in the fluorescence signal followed a typical sigmoid curve that allowed curve fitting for the calculation of the IC50 (Supplementary Fig. 2). The IC50 for each of the Leb glycopeptide dendrimers show, as expected, that multivalency increases the avidity of the interaction (Fig. 1D). This increase in avidity follows a sigmoid pattern: a small increase in multivalency results in a discrete increase in avidity, then there is a range of dendrimer that increase dramatically the avidity until a certain level where saturation is achieved. Incubation of K562/DC-SIGN cells with MH and NG dendrimers did not increase the mean fluorescence intensity of the cells (data not shown), indicating that the non-targeting compounds did not bind cellular DC-SIGN. 3.2. DC-SIGN-ligands multivalency facilitates internalization and targeting to the lysosomes Since Leb glycopeptide dendrimers bound both soluble and cellular DC-SIGN and DC-SIGN is an efficient internalization receptor, we speculated that Leb glycopeptide dendrimers would also induce receptor internalization and that receptor internalization would also increase parallel to the increase in multivalency. To test this hypothesis we set up an imaging flow cytometry assay using the K562/DC-SIGNLL/Y cell line and a K562 cell line expressing DC-SIGN with its intracellular domain intact. Imaging flow cytometry is an attractive new technology that allows the study of different aspects of cellular morphology and fluorescent probe localization, such as internalization and colocalization (Basiji et al., 2007). As expected, Leb glycopeptide dendrimers remained membrane-bound in K562LL/Y cells while were efficiently internalized in K562/DC-SIGN cells. The internalization of generation 3 Leb glycopeptide dendrimers is shown as a representative example (Fig. 2A). Since hardly any binding of MH and NG dendrimers was observed, internalization could not be assessed for these compounds. To study whether Leb glycopeptide dendrimers were indeed internalized by a DC-SIGN-mediated mechanism, K562/DC-SIGN cells were incubated with Leb glycopeptide dendrimers in the presence of the blocking monoclonal antibody AZN-D1 (Geijtenbeek et al., 2000a). Incubation with the blocking antibody, indeed, resulted in an inhibition of the binding to DC-SIGN (data not shown). Therefore, internalization could not be assessed in these cells. We then made use of monocyte-derived DC. Cells were incubated with Leb glycopeptide dendrimers for 1 h at either 4 or 37 ◦ C and DC-SIGN internalization was assessed by imaging flow cytometry of cells fixed, permeabilized, and stained

Fig. 2. Glycopeptide dendrimers are efficiently internalized. A, internalization score histogram of atto488-labeled generation 3 OT-II Leb glycopeptide dendrimers incubated with K562/DC-SIGN or K562/DC-SIGNLL/EEE cells. (B) Mean internalization score of DC-SIGN after incubation of monocyte-derived DCs with OT-II Leb glycopeptide dendrimers at either 4 ◦ C (filled square, dotted line) or 37 ◦ C (open square, continuous line). (C) Fluorescence intensity of monocyte-derived DC incubated with atto488-labeled generation 3 OT-II Leb glycopeptide dendrimers at either 4 ◦ C or 37 ◦ C; untreated monocyte-derived DC (black line) and atto488-labeled generation 3 OT-II Leb glycopeptide dendrimers pre-incubated with the blocking anti-DC-SIGN antibody AZN-D1 at 4 ◦ C (dotted line) were used as control. Data represent the mean + S.D. of triplicate measurements. This experiment is representative of at least three independent experiments.

with a polyclonal anti-DC-SIGN antibody. At 4 ◦ C, DC-SIGN was mainly on the membrane. Incubation with the Leb glycopeptide dendrimers at 37 ◦ C for 1 h resulted in the internalization of DCSIGN (Fig. 2B). The internalization appeared to be maximal for generations 3–7 Leb glycopeptide dendrimers, and suboptimal for generations 0, 1 and 2 Leb glycopeptide dendrimers (Fig. 2B),

J.J. García-Vallejo et al. / Molecular Immunology 53 (2013) 387–397

393

Fig. 3. Multivalent glycopeptide dendrimers are routed preferentially to the lysosomes. (A and C) Bivariate plots showing the internalization of atto488-labeled generation 3 OT-II Leb glycopeptide dendrimers vs their colocalization with the early endosomal marker EEA1 (A) or the lysosomal marker LAMP1 (C), the blue dots correspond to cells incubated with the dendrimers at 4 ◦ C and the red dots correspond to cells incubated with the glycopeptide dendrimers at 37 ◦ C. The percentage of cells in the upper-right quadrant (high internalization/high colocalization) of the sample incubated with the glycopeptide dendrimers at 37 ◦ C is indicated inside each plot. (B and D) Kinetic of dendrimer colocalization with EEA1 (B) or LAMP1 (D). (E and F) Some representative examples of cells showing high internalization and colocalization with EEA1 (E) or LAMP (F). Data represent the mean + S.D. of triplicate measurements. This experiment is representative of at least three independent experiments. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of the article.)

394

J.J. García-Vallejo et al. / Molecular Immunology 53 (2013) 387–397

indicating that multivalency is not only determinant for the avidity of the interaction, but also for the internalization process. Incubation of DCs with MH or NG dendrimers did not affect the localization of DC-SIGN (data not shown). We then investigated whether Atto488-labeled Leb glycopeptide dendrimers were also internalized by monocyte-derived DC. Standard fluorescence intensity histograms show that incubation of DC with generation 3 Leb glycopeptide dendrimers at 4 ◦ C resulted in an increase in the fluorescence signal (Fig. 2C), consistent with the binding of the Leb glycopeptide dendrimers to DC-SIGN on DC. DC incubated with Leb glycopeptide dendrimers at 37 ◦ C showed a 10-fold higher level of fluorescence, indicating that the cells had been actively internalizing the probe, which was not washed from the incubation media (Fig. 2C). As noted before, the internalization was DC-SIGN-dependent, as incubation with the blocking anti-DC-SIGN antibody AZN-D1 resulted in a drastic decrease in the binding of the Leb glycopeptide dendrimers (Fig. 2C). Internalization was calculated as described for K562/DC-SIGN cells and it, indeed, showed that incubation at 37 ◦ C resulted in an efficient internalization of Leb glycopeptide dendrimers, whereas cells incubated with Leb glycopeptide dendrimers at 4 ◦ C only showed a membrane-associated signal (Y-axis on Fig. 2A and B). Imaging flow cytometry also allows for the calculation of the degree of colocalization of two probes within the same cell. DCs were exposed to Leb glycopeptide dendrimers at 4 ◦ C for 60 min, washed and then incubated for a further 120 min at 37 ◦ C. DC were then washed, fixated, permeabilized and stained using antibodies against the early endosomal marker EEA1 (Mu et al., 1995) and the lysosomal marker LAMP1 (Viitala et al., 1988). The imaging flow cytometry analysis of these cells shows that at 4 ◦ C, the Leb glycopeptide dendrimers are localized to the membrane and show very poor colocalization with either EEA1 or LAMP1. Incubation at 37 ◦ C resulted in the shift of the signal toward a higher internalization and an increased colocalization with EEA1 and LAMP1 (Fig. 3). The internalization and colocalization of generation 3 Leb glycopeptide dendrimers with EEA1 (Fig. 3A, B, and E) and LAMP1 (Fig. 3C, D, and F) is shown as a representative example. As expected, colocalization with EEA1 showed a faster kinetics that peaked at approximately 15–30 min (Fig. 3B), where as colocalization of the dendrimers to the LAMP1+ compartment occurred at a slower rate (Fig. 3D). All other Leb glycopeptide dendrimers gave comparable internalization and colocalization scores. In summary, binding of Leb glycopeptide dendrimers to DC-SIGN triggers an efficient internalization process that results in the routing of the Leb glycopeptide dendrimers to lysosomal compartments. 3.3. Multivalent systems for DC-SIGN targeting enhance Ag-specific CD4+ and CD8+ T cell proliferation Lysosomal compartments are well known to be involved in the processing of antigen for presentation in MHC class II, so we hypothesized that the OT-II peptides included in the Leb OTII glycopeptide dendrimers should be processed and presented to OVA-specific CD4+ T cells. To investigate this, bone marrowderived DC obtained from DC-SIGN transgenic mice were pulsed with 1 nmol of each of the Leb OT-II glycopeptide dendrimer generations for 6 h. DC were then washed and set up in triplicate cultures with either 1:4 or 1:20 dilutions of CD4+ OT-II-specific T cells purified by negative selection from the spleens of OT-II mice (Barnden et al., 1998). DC-T cell cocultures were allowed for 48 h and [3 H]Thymidine was added for a further 18 h incubation. Cells were then harvested and [3 H]-Thymidine incorporation measured as an indicator of T cell proliferation. OVA or Leb -conjugated OVA (65 nmol) were used as controls. Interestingly, dendrimers with a low multivalency (generations 0 and 1) induced a poor T cell proliferation, not higher than the negative control (Fig. 4A). However, generations 2–7 induced a very strong T cell proliferation, well beyond the

Fig. 4. Multivalent glycopeptide dendrimers enhance Ag-specific T cell proliferation. (A and B) OT-II (A) and OT-I (B) T cell proliferation after co-culture with DC-SIGN-transgenic bone-marrow derived DC pulsed with the different generations of OT-II (A) or OT-I (B) Leb glycopeptide dendrimers. Untreated DC (−) were used as negative controls and OVA and OVA-Leb -treated cells as positive control. Data represent the mean + S.D. of triplicate wells. This experiment is representative of at least three independent experiments.

T cell proliferation induced by BM-DC pulsed with OVA or OVA-Leb , especially considering that control BM-DC were pulsed with 65 times more antigen epitope (Fig. 4A). The fact that there was no significant increase in T cell proliferation from generations 2 to 7 Leb OT-II glycopeptide dendrimers indicates that the amount of antigen might already be saturating at this concentration, but a certain degree of multivalency is needed for efficient targeting to antigen-processing compartments, since generations 0 and 1 gave a poor stimulation of T cell proliferation. MH and NG glycopeptide dendrimers did not induce any CD4+ T cell proliferation (data not shown), which could be explained by the lack of binding and intracellular uptake shown earlier. Recently, we described that antigen uptake via DC-SIGN also leads to cross-presentation (Singh et al., 2009). We therefore used the Leb OT-I glycopeptide dendrimers to pulse bone marrowderived DC from DC-SIGN-transgenic mice, as explained in the previous experiment, and co-cultured them with negatively isolated CD8+ splenocytes from OT-I mice (Hogquist et al., 1994). After 48 h of incubation, [3 H]-Thymidine was added to the cultures for a further 18 h, cells were harvested and [3 H] incorporation measured. All generations of Leb OT-I glycopeptide dendrimers tested induced massive levels of T cell proliferation, ranging 5–10 times higher than the controls, which were administered at 65-times higher concentration of antigen epitope (Fig. 4B). Again, MH and NG glycopeptide dendrimers did not induce any CD8+ T cell proliferation (data not shown). Interestingly, generations 0 and 1 were able to induce a powerful cross-presentation, and there was a consistent decrease in T cell proliferation inversely proportional to the amount of functional groups per dendrimer. In other words, it seems as if multivalency was detrimental for cross-presentation. Nevertheless, the amount of T cell proliferation observed for generation 7 was clearly higher than that of the controls. In summary, DC-SIGN

J.J. García-Vallejo et al. / Molecular Immunology 53 (2013) 387–397

395

secretion in the absence of LPS. In summary, DC-SIGN ligand multivalency increases the signaling capacity of DC-SIGN to a certain multivalency degree, suggesting that there is an optimal level of receptor aggregation required to induce the association of second messengers to the intracellular domain of DC-SIGN.

4. Discussion

Fig. 5. Multivalent glycopeptide dendrimers do not affect DC maturation and enhance LPS-mediated IL-10 secretion. (A) Mean fluorescence intensity for CD86 on human DCs incubated o/n with Leb glycopeptide dendrimers in the presence or absence of LPS (10 ng/ml). (B) Secretion of IL-10 (LPS-treated/untreated ratio) in human DCs incubated o/n with Leb glycopeptide dendrimers in the presence or absence of LPS (10 ng/ml). Data represent the mean + S.D. of triplicate wells. This experiment is representative of at least three independent experiments.

targeting glycopeptide dendrimers induce robust CD4+ and CD8+ antigen-specific T cell responses. 3.4. Multivalent systems for DC-SIGN do not induce DC maturation and enhance the LPS-mediated upregulation of IL-10 A previous report has demonstrated that mannosylated OVAcontaining dendrimers induced a moderated up-regulation of DC maturation markers (Sheng et al., 2008). We investigated if our Leb glycopeptide dendrimers induced DC maturation, but could not observe any changes in the expression of CD80, CD83 (data not shown) or CD86 (Fig. 5A), while LPS greatly up-regulated expression of these markers (data not shown and Fig. 5A). Also, incubation of DCs in the presence of the Leb glycopeptide dendrimers and LPS did not influence the LPS-mediated upregulation of the maturation markers (data not shown and Fig. 5A). DC-SIGN has been shown to carry intracellular motifs that mediate the association to signaling molecules that modulate TLR4 signaling (Gringhuis et al., 2007). The net result of this interaction is a synergistic increase in the TLR4-mediated IL-10 production when DC-SIGN and TLR4 are simultaneously triggered (Geijtenbeek et al., 2003). We were curious to explore whether increasing the multivalency of the DCSIGN ligand would have effects on IL-10 secretion. To investigate this, we tested the secretion of IL-10 after the exposure of human DC to different generations of Leb glycopeptide dendrimers and LPS. Results show that all Leb glycopeptide dendrimers increased the LPS-mediated IL-10 production, that there was a clear increase in this effect from generations 0 to 3 and that generations 3 to 7 appear to reach a plateau (Fig. 5B). All dendrimers were devoid of LPS, as they were not able to induce maturation or induce IL-10

We here demonstrate that multivalent glycopeptide dendrimers are powerful antigen probes that facilitate DC-SIGN targeting and internalization, and lead to a robust induction of CD4 and CD8T cells. Manipulation of DCs has been proposed for the immunotherapy of cancer, autoimmunity and several infectious diseases. The reasons that support this statement include, amongst others that tumors contain antigens that do not trigger the immune system within the tumor but become strongly immunogenic when presented by DCs; that DCs act on different arms of the adaptive immune system and are, therefore, able to induce a multifaceted long-lasting immune response and, finally, that DCs can be cultured in vitro, maintaining their therapeutic properties (Steinman and Banchereau, 2007). Irrespective from the antigen of choice or the desired outcome (immunity or tolerance) a necessary step in the achievement of an efficient DC-based inmmunotherapy is the delivery of the antigen to the DC. In this regard, many aspects need to be taken into consideration, both from the biological point of view (the desired effect on DCs) but also from the chemical engineering side (the synthesis and design of the antigen and its carrier). DCs are unique antigen presenting cells in that they are equipped with a set of receptors for the recognition of pathogenic motifs. Amongst these receptors, the family of the C-type lectin receptors is gaining importance as recent research demonstrates their important role in the maintenance of regulatory circuits within the immune system (van Kooyk and Rabinovich, 2008; Rabinovich and Croci, 2012). DC-SIGN, the best-studied C-type lectin receptor to date, constitutes an interesting candidate for the targeting of antigens to DCs (Unger and van Kooyk, 2011) since it is able to mediate both antigen uptake (Engering et al., 2002) and signaling (Gringhuis et al., 2007). In order to achieve an efficient targeting it is desirable to use highaffinity ligands that would be able to compete with endogenous DC-SIGN ligands. Since the affinity of the natural DC-SIGN ligands is within the same range (van Liempt et al., 2006) and synthetic ligands have not shown obvious advantages, the most logical way to improve the binding is by enhancing avidity through the presentation of the ligand in highly multivalent systems. Using this approach, it has been demonstrated that the use of a higher density and number of glycans/molecule results in a significant decrease in the binding threshold and large increases in affinity and crosslinking interactions (Dam and Brewer, 2010; Dam et al., 2009). A previous study has shown that mannosylated PAMAM dendrimers were able to target DCs and induce strong CD4 and DC8 immune responses (Sheng et al., 2008). However, the compounds used in this study could target simultaneously several CLRs and the dendrimers were conjugated directly to OVA, a molecule known to target other CLRs such as the mannose receptor (Burgdorf et al., 2007). Furthermore OVA has a higher molecular weight and size than the dendrimers and therefore avidity changes of glycans and antigen cannot be as clearly measured as our glycodendrimers that include OVA peptides and defined glycans. Thus, further insights in the requirements for multivalency in DC-SIGN-specific targeting for immunotherapy using peptides as antigen source were still necessary. We here show that is possible to use relatively simple multivalent targeting systems for the efficient delivery of antigen peptides to DCs. Our multivalent system is based on commercially available components that can be easily assembled without the need of specialized lab equipment. The component used to achieve

396

J.J. García-Vallejo et al. / Molecular Immunology 53 (2013) 387–397

multivalency is the PAMAM dendrimer. PAMAM dendrimers are symmetric highly branched monodisperse polymers with a compact spherical structure that ranges from 1.1 nm in diameter for the 1st generation (G0) to 9 nm for G7) (Biricova and Laznickova, 2009). The small change in internal diameter compared to the large increase in functional groups (4 in G0 and 512 in G7) implies that the differences amongst these 8 generations of dendrimers are not only related to the number of functional groups but also the mobility and flexibility of the ligands attached, specially as in the case of this work, when linear structures such as glycopeptides are used. Therefore, even though the call would be for the most multivalently available compound, it is necessary to consider other factors that will affect the biological outcome. Our data shows that there is an optimal level of multivalency for an efficient DC-SIGN targeting. Low multivalency (generations 0 and 1) showed a modest increase in binding, induced a poor DCSIGN internalization, and resulted in hardly any MHC-II-dependent T cell proliferation. Dendrimers with an intermediate multivalency (generations 2–5) showed a linear increase in the binding to soluble or cellular DC-SIGN with the increase in multivalency and a more efficient internalization and Ag-presentation, both in MHC-I and MHC-II. Higher generation dendrimers (generations 6 and 7) seem to have achieved a plateau in the binding, internalization, and Agprocessing and presentation. Therefore, generations 3 and 4 appear to have the optimal size and multivalency to achieve the most efficient DC-SIGN targeting. Assuming that, at the end of the synthesis, the site occupancy of functional groups on the dendrimers by the glycopeptides was homogeneous, we could estimate that the diameter of the generation 3 glycopeptide dendrimers would be of approximately 20 nm, similar to the distance between the carbohydrate recognition domains of DC-SIGN (Menon et al., 2009). Higher generation dendrimers have bigger molecular diameters but due to the spheric shape of the dendrimers, there will still be glycans fitting the required distance to trigger internalization and the conformational changes required to induce signaling. On the contrary, smaller glycopeptide dendrimers are more likely to cluster neighboring DC-SIGN tetramers rather than be able to activate single DC-SIGN tetramers. This interpretation would, therefore, favor a model in which the biological internalization and association with the internalization and routing machinery is triggered by occupation of the carbohydrate recognition domains of individual tetramers, rather than clustering of neighboring tetramers. Once the glycopeptide dendrimers are internalized, they are transported via early endosomes into lysosomes. This is a prerequisite for the processing of the antigen into peptides that can be loaded on to MHC class II molecules. We flanked the OT-II epitope by lysines in order to facilitate the action of lysosomal endoproteases. This strategy appeared to work efficiently, as judged by the Ag-specific CD4+ T cell proliferation assay. Interestingly, there seems to be a threshold in the multivalency needed to induce MHC-II-dependent T cell proliferation. One could argue that the concentration used was not sufficient for the delivery of enough peptide to induce T cell proliferation in the case of the generations 0 and 1 OT-II glycopeptide dendrimers. However, OT-I glycopeptide dendrimers were administered in the same concentration and, yet, generations 0 and 1 OT-I glycopeptide dendrimers induced the best Ag-specific T cell proliferation. An alternative explanation could be that since the association with the receptor is weaker in the case of the generations 0 and 1 dendrimers, the compounds are dissociated from the receptor earlier than needed for an efficient processing. In the case of the OT-I T cell proliferation, the response is remarkable; the T cell proliferation achieved for any of the glycopeptide dendrimers is 5–10 times higher than the T cell proliferation achieved by OVA, even though 65-times more OVA was administered. Perhaps the OT-I glycopeptide dendrimers are also degraded at the lysosome resulting in enhanced availability of OT-I peptide that is

exported into the cytoplasm by a yet unknown mechanism. Thus, glycopeptide dendrimers would be enhancing cross presentation by a double mechanism, on one hand the enhancement in uptake and concentration in the lysosome, and on the other hand an efficient endoprotease-mediated release of the MHC-I epitope that short-circuits the proteasome. Finally, Leb glycopeptide dendrimers reproduce the DC-SIGNmediated signaling effect described for several pathogens. The effect reached a maximum, once again, with generation 3 Leb glycopeptide dendrimers, supporting the hypothesis of the DC-SIGN tetramer triggering vs DC-SIGN tetramer clustering. In contrast to a previous report using mannosylated OVA dendrimers (Sheng et al., 2008), we could not observe any effects on the maturation of DCs. The difference in effect could be related to several factors. On one hand, mannosylated dendrimers may trigger not only DC-SIGN, but also other mannose-specific CLRs. Also, the authors detected minute amounts of LPS in their preparation, but did not test the bioactivity of the contaminating LPS. In summary, multimeric presentation of DC-SIGN ligands has a linear effect on the binding to the receptor. However, internalization and signaling are enhanced only until a certain level. Glycopeptide dendrimers are very efficient systems of antigen delivery for the induction of both CD4 and CD8T cell proliferation and should be considered as an interesting candidate for the development of DC-SIGN targeting immunotherapies. Funding The present work was funded by an ALW-NWO Veni grant to JJGV (863.08.020). Authors’ contribution JJGV, HK, and YvK conceived and planned the experiments. HK and MA designed and synthesized the compounds. JJGV, AO, HvR, KB, WU, and WWJU performed the experiments. JGB and KN prepared the peptides. JJGV analyzed and interpreted the data and wrote the manuscript. Conflict of interest The authors declare no conflict of interest for the present manuscript. Acknowledgments We would like to thank Tom O’Toole and Nico Blijleven for excellent technical help with the imaging flow cytometer. Also Sherree Friend and Thadeus Georg (Amnis Inc.) for suggestions and support with the imaging flow cytometry analysis. We thank the staff of our animal facility for the care of the animals used in this study. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.molimm. 2012.09.012. References Appelmelk, B.J., van Die, I., van Vliet, S.J., et al., 2003. Cutting edge: carbohydrate profiling identifies new pathogens that interact with dendritic cell-specific ICAM-3-grabbing nonintegrin on dendritic cells. Journal of Immunology 170 (4), 1635–1639. Barnden, M.J., Allison, J., Heath, W.R., et al., 1998. Defective TCR expression in transgenic mice constructed using cDNA-based alpha- and beta-chain genes under

J.J. García-Vallejo et al. / Molecular Immunology 53 (2013) 387–397 the control of heterologous regulatory elements. Immunology and Cell Biology 76 (1), 34–40. Basiji, D.A., Ortyn, W.E., Liang, L., et al., 2007. Cellular image analysis and imaging by flow cytometry. Clinics in Laboratory Medicine 27 (3), 653–670, viii. Berzi, A., Reina, J.J., Ottria, R., et al., 2012. A glycomimetic compound inhibits DCSIGN-mediated HIV infection in cellular and cervical explant models. AIDS 26 (2), 127–137. Biricova, V., Laznickova, A., 2009. Dendrimers: analytical characterization and applications. Bioorganic Chemistry 37 (6), 185–192. Bonifaz, L., Bonnyay, D., Mahnke, K., et al., 2002. Efficient targeting of protein antigen to the dendritic cell receptor DEC-205 in the steady state leads to antigen presentation on major histocompatibility complex class I products and peripheral CD8+ T cell tolerance. The Journal of Experimental Medicine 196 (12), 1627–1638. Boscardin, S.B., Hafalla, J.C.R., Masilamani, R.F., et al., 2006. Antigen targeting to dendritic cells elicits long-lived T cell help for antibody responses. The Journal of Experimental Medicine 203 (3), 599–606. Burgdorf, S., Kautz, A., Böhnert, V., et al., 2007. Distinct pathways of antigen uptake and intracellular routing in CD4 and CD8T cell activation. Science 316 (5824), 612–616. Caminschi, I., Proietto, A.I., Ahmet, F., et al., 2008. The dendritic cell subtyperestricted C-type lectin Clec9A is a target for vaccine enhancement. Blood 112 (8), 3264–3273. Cao, W., Zhang, L., Rosen, D.B., et al., 2007. BDCA2/Fc epsilon RI gamma complex signals through a novel BCR-like pathway in human plasmacytoid dendritic cells. PLoS Biology 5 (10), e248. Carbone, F.R.F., Belz, G.T.G., Heath, W.R.W., 2004. Transfer of antigen between migrating and lymph node-resident DCs in peripheral T-cell tolerance and immunity. Trends in Immunology 25 (12), 655–658. Dam, T.K., Brewer, C.F., 2010. Lectins as pattern recognition molecules: the effects of epitope density in innate immunity. Glycobiology 20 (3), 270–279. Dam, T.K., Gerken, T.A., Brewer, C.F., 2009. Thermodynamics of multivalent carbohydrate-lectin cross-linking interactions: importance of entropy in the bind and jump mechanism. Biochemistry 48 (18), 3822–3827. Dudziak, D., Kamphorst, A.O., Heidkamp, G.F., et al., 2007. Differential antigen processing by dendritic cell subsets in vivo. Science 315 (5808), 107–111. Engering, A., Geijtenbeek, T., van Vliet, S., et al., 2002. The dendritic cell-specific adhesion receptor DC-SIGN internalizes antigen for presentation to T cells. Journal of Immunology 168 (5), 2118–2126. Engering, A., van Vliet, S.J., Hebeda, K., et al., 2004. Dynamic populations of dendritic cell-specific ICAM-3 grabbing nonintegrin-positive immature dendritic cells and liver/lymph node-specific ICAM-3 grabbing nonintegrin-positive endothelial cells in the outer zones of the paracortex of human lymph nodes. The American Journal of Pathology 164 (5), 1587–1595. Fawcett, J., Holness, C.L., Needham, L.A., et al., 1992. Molecular cloning of ICAM-3, a third ligand for LFA-1, constitutively expressed on resting leukocytes. Nature 360 (6403), 481–484. Geijtenbeek, T.B., Kwon, D.S., Torensma, R., et al., 2000a. DC-SIGN, a dendritic cellspecific HIV-1-binding protein that enhances trans-infection of T cells. Cell 100 (5), 587–597. Geijtenbeek, T.B., Torensma, R., van Vliet, S.J., et al., 2000b. Identification of DC-SIGN, a novel dendritic cell-specific ICAM-3 receptor that supports primary immune responses. Cell 100 (5), 575–585. Geijtenbeek, T.B.H., van Vliet, S.J., Koppel, E.A., et al., 2003. Mycobacteria target DCSIGN to suppress dendritic cell function. The Journal of Experimental Medicine 197 (1), 7–17. Gringhuis, S.I., den Dunnen, J., Litjens, M., et al., 2007. C-type lectin DC-SIGN modulates toll-like receptor signaling via Raf-1 kinase-dependent acetylation of transcription factor NF-kappa B. Immunity 26 (5), 605–616. Hawiger, D., Inaba, K., Dorsett, Y., et al., 2001. Dendritic cells induce peripheral T cell unresponsiveness under steady state conditions in vivo. The Journal of Experimental Medicine 194 (6), 769–779. Heath, W.R., Carbone, F.R., 2001. Cross-presentation in viral immunity and selftolerance. Nature Reviews Immunology 1 (2), 126–134. Hogquist, K.A., Jameson, S.C., Heath, W.R., et al., 1994. T cell receptor antagonist peptides induce positive selection. Cell 76 (1), 17–27. Idoyaga, J., Lubkin, A., Fiorese, et al., 2011. Comparable T helper 1 (Th1) and CD8 T-cell immunity by targeting HIV gag p24 to CD8 dendritic cells within antibodies to Langerin, DEC205, and Clec9A. Proceedings of the National Academy of Sciences of the United States of America 108 (6), 2384–2389. Joffre, O., Nolte, M.A., Spörri, R., et al., 2009. Inflammatory signals in dendritic cell activation and the induction of adaptive immunity. Immunological Reviews 227 (1), 234–247. Kanazawa, N., 2007. Dendritic cell immunoreceptors: C-type lectin receptors for pattern-recognition and signaling on antigen-presenting cells. Journal of Dermatological Science 45 (2), 77–86. Kraal, G., Breel, M., Janse, M., et al., 1986. Langerhans’ cells, veiled cells, and interdigitating cells in the mouse recognized by a monoclonal antibody. The Journal of Experimental Medicine 163 (4), 981–997.

397

Kretschmer, K., Apostolou, I., Hawiger, D., et al., 2005. Inducing and expanding regulatory T cell populations by foreign antigen. Nature Immunology 6 (12), 1219–1227. Kurts, C., Robinson, B.W.S., Knolle, P.A., 2010. Cross-priming in health and disease. Nature Reviews Immunology 10 (6), 403–414. Li, S., Zeng, D., 2007. Chemoenzymatic enrichment of phosphotyrosine-containing peptides. Angewandte Chemie 46 (25), 4751–4753. Luczkowiak, J., Sattin, S., Sutkeviciute, I., et al., 2011. Pseudosaccharide functionalized dendrimers as potent inhibitors of DC-SIGN dependent Ebola pseudotyped viral infection. Bioconjugate Chemistry 22 (7), 1354–1365. Mahnke, K., Qian, Y., Knop, J., et al., 2003. Induction of CD4+/CD25+ regulatory T cells by targeting of antigens to immature dendritic cells. Blood 101 (12), 4862–4869. Menon, S., Rosenberg, K., Graham, S., et al., 2009. Binding-site geometry and flexibility in DC-SIGN demonstrated with surface force measurements. Proceedings of the National Academy of Sciences of the United States of America 106 (28), 11524–11529. Mitchell, D.A., Fadden, A.J., Drickamer, K., 2001. A novel mechanism of carbohydrate recognition by the C-type lectins DC-SIGN and DC-SIGNR. Subunit organization and binding to multivalent ligands. The Journal of Biological Chemistry 276 (31), 28939–28945. Mu, F.T., Callaghan, J.M., Steele-Mortimer, O., et al., 1995. EEA1, an early endosomeassociated protein. EEA1 is a conserved alpha-helical peripheral membrane protein flanked by cysteine fingers and contains a calmodulin-binding IQ motif. The Journal of Biological Chemistry 270 (22), 13503–13511. Neefjes, J., Jongsma, M.L.M., Paul, P., et al., 2011. Towards a systems understanding of MHC class I and MHC class II antigen presentation. Nature Reviews Immunology 11 (12), 823–836. Ortyn, W.E., Hall, B.E., George, T.C., et al., 2006. Sensitivity measurement and compensation in spectral imaging. Cytometry A 69 (8), 852–862. Ortyn, W.E., Perry, D.J., Venkatachalam, V., et al., 2007. Extended depth of field imaging for high-speed cell analysis. Cytometry A 71 (4), 215–231. Paul, P., van den Hoorn, T., Jongsma, M.L.M., et al., 2011. A genome-wide multidimensional RNAi screen reveals pathways controlling MHC class II antigen presentation. Cell 145 (2), 268–283. Rabinovich, G.A., Croci, D.O., 2012. Regulatory circuits mediated by lectin-glycan interactions in autoimmunity and cancer. Immunity 36 (3), 322–335. Robinson, M.J., Sancho, D., Slack, E.C., et al., 2006. Myeloid C-type lectins in innate immunity. Nature Immunology 7 (12), 1258–1265. Sallusto, F., Lanzavecchia, A., 1994. Efficient presentation of soluble antigen by cultured human dendritic cells is maintained by GM-CSF plus IL-4 and downregulated by TNF-␣. The Journal of Experimental Medicine 179, 1109–1118. Sancho, D., Mourão-Sá, D., Joffre, O.P., et al., 2008. Tumor therapy in mice via antigen targeting to a novel, DC-restricted C-type lectin. Journal of Clinical Investigation 118 (6), 2098–2110. Sattin, S., Daghetti, A., Thépaut, M., et al., 2010. Inhibition of DC-SIGN-mediated HIV infection by a linear trimannoside mimic in a tetravalent presentation. ACS Chemical Biology 5 (3), 301–312. Schaefer, M., Reiling, N., Fessler, C., et al., 2008. Decreased pathology and prolonged survival of human DC-SIGN transgenic mice during mycobacterial infection. Journal of Immunology 180 (10), 6836–6845. Schreibelt, G., Klinkenberg, L.J.J., Cruz, L.J., et al., 2012. The C-type lectin receptor CLEC9A mediates antigen uptake and (cross-)presentation by human blood BDCA3+ myeloid dendritic cells. Blood 119 (10), 2284–2292. Sheng, K.-C.K., Kalkanidis, M.M., Pouniotis, D.S.D., et al., 2008. Delivery of antigen using a novel mannosylated dendrimer potentiates immunogenicity in vitro and in vivo. European Journal of Immunology 38 (2), 424–436. Singh, S.K., Stephani, J., Schaefer, M., et al., 2009. Targeting glycan modified OVA to murine DC-SIGN transgenic dendritic cells enhances MHC class I and II presentation. Molecular Immunology 47 (2–3), 164–174. Steinman, R.M., Banchereau, J., 2007. Taking dendritic cells into medicine. Nature 449 (7161), 419–426. Unger, W.W.J., van Kooyk, Y., 2011. Dressed for success C-type lectin receptors for the delivery of glyco-vaccines to dendritic cells. Current Opinion in Immunology 23 (1), 131–137. Unger, W.W.J., van Beelen, A.J., Bruijns, S.C., et al., 2012. Glycan-modified liposomes boost CD4+ and CD8+ T-cell responses by targeting DC-SIGN on dendritic cells. Journal of Controlled Release 160 (1), 88–95. van Kooyk, Y., Rabinovich, G.A., 2008. Protein-glycan interactions in the control of innate and adaptive immune responses. Nature Immunology 9 (6), 593–601. van Liempt, E., Bank, C.M.C., Mehta, P., et al., 2006. Specificity of DC-SIGN for mannose- and fucose-containing glycans. FEBS Letters 580 (26), 6123–6131. Viitala, J., Carlsson, S.R., Siebert, P.D., et al., 1988. Molecular cloning of cDNAs encoding lamp A, a human lysosomal membrane glycoprotein with apparent Mr approximately equal to 120,000. Proceedings of the National Academy of Sciences of the United States of America 85 (11), 3743–3747. Wang, J., Zou, Z.-H., Xia, H.-L., et al., 2012. Strengths and weaknesses of immunotherapy for advanced non-small-cell lung cancer: a meta-analysis of 12 randomized controlled trials. PLoS ONE 7 (3), e32695.