Genetically Engineered Cell Factories Produce Glycoengineered Vaccines That Target Antigen-Presenting Cells and Reduce Antigen-Specific T-cell Reactivity.

Allergen-specific immunotherapy is a promising approach to reduce or remove allergic symptoms by inducing tolerance.1Jutel M. Kosowska A. Smolinska S. Allergen immunotherapy: past, present, and future.Allergy Asthma Immunol Res. 2016; 8: 191-197Crossref PubMed Scopus (56) Google Scholar As an alternative to current allergy vaccines derived from extracts of natural allergens, recombinant vaccines offer the opportunity for de novo antigen design,2Kitzmüller C. Kalser J. Mutschlechner S. Hauser M. Zlabinger G.J. Ferreira F. et al.Fusion proteins of flagellin and the major birch pollen allergen Bet v 1 show enhanced immunogenicity, reduced allergenicity, and intrinsic adjuvanticity.J Allergy Clin Immunol. 2018; 141: 293-299.e6Abstract Full Text Full Text PDF PubMed Scopus (20) Google Scholar with enhanced targeting to antigen-presenting cells (APCs) and potentially increased ability to induce tolerance.3Macri C. Dumont C. Johnston A.P. Mintern J.D. Targeting dendritic cells: a promising strategy to improve vaccine effectiveness.Clin Transl Immunol. 2016; 5: e66Crossref PubMed Scopus (115) Google Scholar, 4van Die I. Cummings R.D. Glycan gimmickry by parasitic helminths: a strategy for modulating the host immune response?.Glycobiology. 2010; 20: 2-12Crossref PubMed Scopus (159) Google Scholar Using the type 1 allergen Bet v 1 as our model antigen, we evaluated the effects of combining a nonglycosylated antigen to a glycomodule5Halim A. Carlsson M.C. Madsen C.B. Brand S. Møller S.R. Olsen C.E. et al.Glycoproteomic analysis of seven major allergenic proteins reveals novel post-translational modifications.Mol Cell Proteomics. 2015; 14: 191-204Crossref PubMed Scopus (28) Google Scholar to target the APC-expressed family of C-type lectin receptors (CLRs). CLRs are known to recognize glycan structures and promote rapid internalization of antigens on binding (see Fig E1, A, in this article's Online Repository at www.jacionline.org).6Unger W.W. van Kooyk Y. “Dressed for success” C-type lectin receptors for the delivery of glyco-vaccines to dendritic cells.Curr Opin Immunol. 2011; 23: 131-137Crossref PubMed Scopus (60) Google Scholar, 7Sirvent S. Soria I. Cirauqui C. Cases B. Manzano A.I. Diez-Rivero C.M. et al.Novel vaccines targeting dendritic cells by coupling allergoids to nonoxidized mannan enhance allergen uptake and induce functional regulatory T cells through programmed death ligand 1.J Allergy Clin Immunol. 2016; 138: 558-567.e11Abstract Full Text Full Text PDF PubMed Scopus (76) Google Scholar We used our recently developed glycoengineered cellular platform8Yang Z. Wang S. Halim A. Schulz M.A. Frodin M. Rahman S.H. et al.Engineered CHO cells for production of diverse, homogeneous glycoproteins.Nat Biotechnol. 2015; 33: 842-844Crossref PubMed Scopus (166) Google Scholar to produce Bet v 1 with defined carbohydrates and evaluated APC uptake, efficacy using patient-derived T cells, and induction of tolerance in a murine model. Currently, allergen-specific immunotherapy is primarily administered subcutaneously or sublingually. Focusing on human skin, we examined the CLR expression pattern in skin-localized APCs. We observed expression of the major CLRs, including mannose receptor (MR), dendritic cell–specific intercellular adhesion molecule 3–grabbing nonintegrin (DC-SIGN), and macrophage galactose-type lectin (MGL), in the dermal compartment of skin biopsy specimens (see Fig E1, B). Next, we examined expression of MR, DC-SIGN, and MGL in in vitro–differentiated human monocyte-derived dendritic cells (moDCs) and macrophages (M1 and M2) by means of transcriptomics and flow cytometry (see Fig E1, C-E). Expression of CLRs peaked after 3 days of differentiation, with simultaneous and significant expression of DC-SIGN, MR, and MGL in moDCs (see Fig E1, D and E). Cytokine measurements confirmed the nature of the APCs (see Fig E1, F) and their relevance as a cellular model. To identify the optimal carbohydrate structure or structures for specific APC targeting, we evaluated uptake of chemically synthesized carbohydrate derivatives (N-acetylgalactosamine polyacrylamide [GalNAc-PAA], N-acetylglucosamine–PAA, Lewis X–PAA, and mannose-PAA) and a GalNAc glycosylated molecule with varying numbers of GalNAc residues (see Fig E2, A and B, in this article's Online Repository at www.jacionline.org). Both mannosylated and high-density GalNAc glycosylated structures were readily taken up by the APCs, and therefore GalNAc and mannose structures were selected for further studies. We next designed Bet v 1 allergen constructs for the expression of full-length Bet v 1 fused to glycomodules, with consensus glycosylation motifs ensuring the incorporation of either O- or N-linked glycans. The fusion constructs were expressed in glycoengineered Chinese hamster ovary (CHO) cell lines and Pichia pastoris (Fig 1, A, and see Fig E2, C-G, and Table E1 in this article's Online Repository at www.jacionline.org). This strategy allowed us to generate fusion proteins decorated with high-mannose N-linked glycans from MGAT1 knockout CHO cells, linear α1-2 O-linked mannose residues from P pastoris (Pichia), or O-GalNAc residues from COSMC knockout CHO cells for efficient targeting of MR, DC-SIGN, and/or MGL on APCs (Fig 1, A, and see Fig E2, C-G). Additionally, we expressed the O-glycan fusion construct in Escherichia coli and inserted GalNAc residues by means of in vitro glycosylation. Glycosylation profiles were verified by using matrix-assisted laser desorption ionization time-of-flight mass spectrometry (MALDI-TOF) of released glycans supplemented with lectin profiling (see Fig E2, C-G). As a control for glycan-mediated effects, we treated fractions of the glycoallergens with periodate, a treatment known to effectively destroy lectin recognition (see Fig E2, E). Because moDCs were the only cell type to express high levels of DC-SIGN, as well as MR and MGL, we used moDCs to evaluate uptake of glycan-modified Bet v 1 fusion proteins. Internalization of all glycan-modified Bet v 1 fusion proteins was enhanced 1.5- to 6-fold compared with that of their nonglycosylated counterparts (Fig 1, B). This uptake was mediated through clathrin-coated pits, EEA-1- and Rab-5–positive endosomes. Additionally, live imaging revealed a rapid uptake of glycosylated allergen (see Video E1 and Fig E3 in this article's Online Repository at www.jacionline.org). Importantly, periodate treatment of the glycomodified allergens eliminated the increased uptake and supported the glycan-specific effect (Fig 1, B). This was confirmed by ex vivo uptake of in vitro–glycosylated GalNAc–Bet v 1 (E coli) in primary plasmacytoid dendritic cells (pDCs) isolated directly from human blood (see Fig E4 in this article's Online Repository at www.jacionline.org), exemplifying in vivo relevance. Additionally, the increased uptake was confirmed in human tissue by means of injection of glycosylated Bet v 1 intradermally ex vivo (Fig 1, C). Glycosylated products accumulated and localized to APCs, expressing MR and MGL, whereas the nonglycosylated allergens were virtually undetectable (Fig 1, C). In contrast, DC-SIGN+ cells only seemed to internalize GalNAc–Bet v 1 (CHO). The APC nature of the cells internalizing the glycosylated allergen was verified by means of extraction and quantification of Bet v 1 uptake in cells colabeled with CD11c (see Fig E3, C). The high-density linear mannose structures produced by P pastoris mediated the most significant increase in uptake, closely followed by GalNAc glycosylated Bet v 1. The single high-mannose N-glycan also promoted uptake, although less efficiently (Fig 1). These findings are consistent with previous reports suggesting that clustered glycans are more potent compared with their linear and nonclustered counterparts.9Lo-Man R. Vichier-Guerre S. Bay S. Dériaud E. Cantacuzène D. Leclerc C. Anti-tumor immunity provided by a synthetic multiple antigenic glycopeptide displaying a Tri-Tn glycotope.J Immunol. 2001; 166: 2849-2854Crossref PubMed Scopus (110) Google Scholar To facilitate comparison of the physiologic effects of the glycan-mediated increased uptake between human and murine models, we conducted CLR expression profiling of murine APCs (see Fig E5, A, in this article's Online Repository at www.jacionline.org). In murine bone marrow–derived dendritic cells (BM-DCs), we again found increased uptake of Lewis X–PAA and GalNAc-PAA but not mannose-PAA. This uptake pattern correlated with the very low MR expression seen in murine BM-DCs (see Fig E5). We next investigated the biodistribution of fluorescence-labeled glycosylated Bet v 1 in vivo and found a significant increase in allergen uptake with targeting to the popliteal lymph nodes for GalNAc and linear mannose glycoforms compared with the nonglycosylated Bet v 1. Interestingly, there was a tendency for greater uptake of GalNAc-modified antigen in murine cells in contrast to human cells, where linear clustered mannose seems to dominate (Fig 1, D-G, and see Fig E5). This uptake pattern correlates with the expression level of MGL and MR in APC populations (see Figs E1 and E5, A). The potential differences in uptake between species might be relevant when testing glycosylated human vaccine candidates in murine models and could make it difficult to conclude which glycan structure would be most relevant in the human setting. We next wanted to clarify the global effect of glycans on dendritic cell stimulation. We performed RNA sequencing analysis of moDCs stimulated with the Bet v 1 allergens by using LPS as a reference (see Fig E6 in this article's Online Repository at www.jacionline.org). Although LPS stimulation resulted in a well-defined transcriptional change, a comparison between moDCs stimulated with glycosylated Bet v 1 and periodate-treated Bet v 1 yielded no significant transcriptional change and thereby no indications of any major immune pathways differentially activated (see Fig E6). This suggests that the major effect of glycans on allergens is related to increased uptake. This is supported by cytokine measurements on protein level of ex vivo–stimulated human pDCs, showing that cytokines are not produced at detectable levels in pDC cultures stimulated with Bet v 1 for 24 hours (see Fig E7 in this article's Online Repository at www.jacionline.org). However, future studies could include several time points to fully clarify the immunomodulatory effect and compare early and late cytokine responses. To investigate whether this increased uptake could alter immune reactivity, we performed in vitro experiments using T cells from allergic patients and in vivo experiments in an animal model. First, we evaluated the proliferation and cytokine release of primary T-cell lines from donors with birch pollen allergy. Cocultures of T cells with autologous moDCs pulsed with glycosylated Bet v 1 induced significantly increased proliferation compared with nonglycosylated Bet v 1, regardless of the glycophenotype (Fig 2, A, and see Fig E8 in this article's Online Repository at www.jacionline.org). Regarding cytokines, we observed a mixed picture with increased levels of both proinflammatory cytokines and anti-inflammatory cytokines, such as IL-10, after glyco–Bet v 1 priming (Fig 2, B, and see Figs E9 and E10 in this article's Online Repository at www.jacionline.org). To assess function, we evaluated the in vivo effect of glycosylating Bet v 1 in a prophylactic mouse model of type 1 allergy and compared the efficacy to nonglycosylated E coli–produced Bet v 1 (Bet v 1 noGM [E coli]), as well as nonglycosylated CHO-produced Bet v 1. Because GalNAc products had the overall highest uptake in murine cells (Fig 1 and see Fig E5), we chose to prophylactically treat the mice with GalNAc–Bet v 1 (CHO). We included the HighMan-Bet v 1 (CHO) variant also to test a mannose-containing product and ensure that all proteins for the in vivo analysis were produced in the same production cell. We showed significantly reduced ex vivo spleen cell proliferation, with a uniform low proliferative response across all experiments when compared to the nonglycosylated Bet v 1 (Fig 2, C, and see Fig E11, A, in this article's Online Repository at www.jacionline.org). Reduced T-cell proliferation was also seen for GalNAc–Bet v 1 (CHO) after restimulation with the natural Bet v extract in different concentrations (Fig 2, C, and see Fig E11, A). In addition, prophylactic treatment with GalNAc–Bet v 1 (CHO) significantly reduced the IgE response compared with control treatment (PBS; see Fig E11, B). The reduction in IgE levels is at the same level as Bet v 1 (E coli), suggesting that the glycan module does not disrupt the positive effect of prophylactic Bet v 1 treatment on IgE level while significantly reducing the T-cell proliferative response. We cannot fully exclude that introduction of the glycan module could have an effect on the overall 3-dimensional structure of Bet v 1 and thereby conformational epitopes. However, because pretreatment with Bet v 1 containing the glycan module can protect mice sensitized with Bet v 1 without the module, we believe the effect is minimal. In summary, the use of a genetically engineered production platform combined with a glycomodule design provides a flexible solution with freedom to choose different posttranslational moieties and antigens. For allergen-specific immunotherapy, our results suggest that fusion proteins containing high-density GalNAc modification is the prime candidate for a clinically relevant Bet v 1 vaccine. However, the specific design will most likely be dependent on the biophysical and physiologic properties of the individual antigen. The described glycoengineered production platform will be useful for the design of next-generation therapeutic vaccines. We acknowledge the Core Facility for Integrated Microscopy, Faculty of Health and Medical Sciences, University of Copenhagen, for assistance obtaining the confocal microscopy data. Download .docx (.05 MB) Help with docx files Online Repository Text Download .docx (.02 MB) Help with docx files Table E1eyJraWQiOiI4ZjUxYWNhY2IzYjhiNjNlNzFlYmIzYWFmYTU5NmZmYyIsImFsZyI6IlJTMjU2In0.eyJzdWIiOiI3NmMzNzVhYTI3MTA5NzVmZGNmMWIxZjdmNGU1YzU3NiIsImtpZCI6IjhmNTFhY2FjYjNiOGI2M2U3MWViYjNhYWZhNTk2ZmZjIiwiZXhwIjoxNjc4ODE5Nzk3fQ.NLfDpG_sR5kWQeQ9wQwVy7QObx7-9N_HlJC2pKVzdCLtuoelE0ffG9jNMdBMLME6J0K2G11xW_D7syORojFWNINqcGKrBoJaiiWCcYI-dda3xqHqIMUeOJdW1dE3XQy5wBmmJfhcGciWo8eGjeBxZnT6EjXVngU2JEGInpjwCSlj8TpSU4uzlixdCMMx-divzofAJbhO3px5EwFodIaFd36Xhjg_j0USvk8txUtSDbVAqatoDPWM_oAJdd5N8Mf2TYu3Cy9dVXCOyV5GIMKspKsFLqUVUShNtqEa1GAEOv9cT595HdCkmQZHyVtIVt6dhxozbSp45vhBZCK4D2Eh6A Download .mp4 (15.1 MB) Help with .mp4 files Video E1Fig E2View Large Image Figure ViewerDownload Hi-res image Download (PPT)Fig E3View Large Image Figure ViewerDownload Hi-res image Download (PPT)Fig E4View Large Image Figure ViewerDownload Hi-res image Download (PPT)Fig E5View Large Image Figure ViewerDownload Hi-res image Download (PPT)Fig E6View Large Image Figure ViewerDownload Hi-res image Download (PPT)Fig E7View Large Image Figure ViewerDownload Hi-res image Download (PPT)Fig E8View Large Image Figure ViewerDownload Hi-res image Download (PPT)Fig E9View Large Image Figure ViewerDownload Hi-res image Download (PPT)Fig E10View Large Image Figure ViewerDownload Hi-res image Download (PPT)Fig E11View Large Image Figure ViewerDownload Hi-res image Download (PPT)