The fully synthetic MAG-Tn3 therapeutic vaccine containing the tetanus toxoid-derived TT830-844 universal epitope provides anti-tumor immunity

Abstract

Malignant transformations are often associated with aberrant glycosylation processes that lead to the expression of new carbohydrate antigens at the surface of tumor cells. Of these carbohydrate antigens, the Tn antigen is particularly highly expressed in many carcinomas, especially in breast carcinoma. We designed MAG-Tn3, a fully synthetic vaccine based on three consecutive Tn moieties that are O-linked to a CD4+ T cell epitope, to induce anti-Tn antibody responses that could be helpful for therapeutic vaccination against cancer. To ensure broad coverage within the human population, the tetanus toxoid-derived peptide TT830-844 was selected as a T-helper epitope because it can bind to various HLA-DRB molecules. We showed that the MAG-Tn3 vaccine, which was formulated with the GSK proprietary immunostimulant AS15 and designed for human cancer therapy, is able to induce an anti-Tn antibody response in mice of various H-2 haplotypes, and this response correlates with the ability to induce a specific T cell response against the TT830-844 peptide. The universality of the TT830-844 peptide was extended to new H-2 and HLA-DRB molecules that were capable of binding this T cell epitope. Finally, the MAG-Tn3 vaccine was able to induce anti-Tn antibody responses in cynomolgus monkeys, which targeted Tn-expressing tumor cells and mediated tumor cell death both in vitro and in vivo. Thus, MAG-Tn3 is a highly promising anticancer vaccine that is currently under evaluation in a phase I clinical trial.

Keywords: Antibody response; Anticancer vaccine; MAG-Tn3; TT830-844; Universal epitope.

Fig. 1

Immunizing mice of different H-2 haplotypes with MAG-Tn3 induce a specific anti-Tn IgG response associated with a TT-specific T cell response. a HLA-DR1*A2 transgenic mice were im immunized on days 0, 21, 42, 63 and 84 with 15 (n = 12), 45 (n = 12) or 60 μg (n = 10) of MAG-Tn3 with 1/20, 1/7 and 1/5 of the AS15 immunostimulant, respectively, or with the AS01B immunostimulant alone. b C57BL/6J, C3H/HeN, BALB/c mice (n = 6) and NMRI and CD-1 mice (n = 8) were im immunized on days 0, 21 and 42 with 15 μg of MAG-Tn3 with 1/20 of the AS15 immunostimulant (middle and right panels), or with the AS01B immunostimulant alone (left panel). Sera were collected on days 21, 28 and 49 and tested for Tn-specific IgG by ELISA using Tn3-G6K (Biot) G. Antibody titers are expressed as the mean of Log10 individual antibody titers ± SEM. c. IFN-γ, IL-5, IL-13 and IL-17 production was analyzed by ELISA on supernatants of splenocytes stimulated with (black bars) or without (white bars) 50 μg/mL of TT for 72 h. The results are expressed as the means of individual mice ± SEM. The statistical significance of differences was determined by the Student t test (*P < 0.05, **P < 0.01). ND not detectable

Fig. 2

The anti-Tn IgG response induced by the MAG-Tn3 vaccine is CD4⁺ T cell dependent. a, b C3H/HeN mice (n = 6/group) that were left untreated or treated with anti-CD4 or isotype control antibodies (300 μg ip) on days −1, 0, 1, 20, 21 and 22 were im immunized on days 0 and 21 with 15 μg of MAG-Tn3 with 1/20 of the AS15 immunostimulant. CD4⁺ T cell depletion was confirmed by FACS analysis of the percentages of CD4⁺ and CD8⁺ cells among CD3⁺ cells on blood samples collected on day 21, prior to vaccination. Dot plots of one representative mouse per group are shown (a). The results are expressed as the mean of individual mice ± SEM (b). c Sera were collected on days 21 and 28 and tested for Tn-specific IgG by ELISA, using Tn3-G6K (Biot) G. Antibody titers are expressed as the mean of Log10 individual antibody titers ± SEM. The statistical significance of differences was determined by the Student t test (**P < 0.01)

Fig. 3

The TT epitope can be recognized and presented by different murine MHC class II and human HLA-DRB1*01:01 molecules. a–c. C3H/HeN mice (n = 5) were sc immunized with 50 μg of TT or HEL with CFA. On day 11, TNF-α production was analyzed by ICS on individual lymph nodes (a) and IFN-γ (b) and IL-2 (c) production was analyzed by ELISA on pooled lymph nodes stimulated with TT (black bars) or HEL (white bars). The results are expressed as the mean of individual mice ± SEM. d, e, f. HLA-DR1*A2 (n = 3), C57BL/6J (n = 3), BALB/c (n = 3), NMRI (n = 3) and CD-1 (n = 6) mice were sc immunized with 50 μg of TT or with immunostimulant alone on days 0 and 14. On day 28, TNF-α production was analyzed by ICS (d) and IFN-γ (e) and IL-2 (f) production was analyzed by ELISA on splenocytes stimulated with TT (black bars) or HEL (white bars). The results are expressed as the mean of triplicates ± SEM. The statistical significance of differences was determined by the Student t test (*P < 0.05, **P < 0.01)

Fig. 4

TT can be presented by different human HLA-DRB molecules. a The HLA binding assay was performed using TT against human HLA-DRB molecules of various HLA-DRB alleles. b The frequency of HLA-DRB molecules capable of binding TT is given for various populations, based on data published at www.allelefrequencies.net. NDA no data available

Fig. 5

Immunization of cynomolgus monkeys with the MAG-Tn3 vaccine induces a specific anti-Tn antibody response associated with a T cell response to TT. Cynomolgus monkeys (n = 10/group) were immunized on days 0, 21, 42, 63, 84 and 105 with NaCl, AS15 or MAG-Tn3 formulated with AS15. a Sera were collected on the day of each immunization and tested for Tn-specific IgM and IgG responses by ELISA. b, c T cell responses were measured on cynomolgus PBMCs harvested after five immunizations (day 105), stimulated with (black bars) or without TT (white bars). IFN-γ production was evaluated by ELISPOT and is expressed as the mean of individual SFC per 10⁶ cells ± SEM (b). Pictures are shown of one representative well per group, seeded with 10⁶ cells (c). The statistical significance of differences was determined by the Student t test (**P < 0.01)

Fig. 6

Anti-Tn antibodies are able to induce the death of Tn-expressing tumor cells in vitro and in vivo. a, b. Pre- and post-immunized sera, obtained from cynomolgus monkeys (n = 10/group) after the immunization protocol described in Fig. 5, were analyzed for their ability to recognize Tn-expressing tumor cells. The binding of sera to Jurkat cells was revealed using double labeling with anti-human-IgM-PE and anti-human-IgG-FITC (a). Sera of cynomolgus monkeys vaccinated with MAG-Tn3 were collected at day 105 and tested for Tn-specific binding to different Tn-expressing tumor cells, compared with naive sera (b). Antibody titers are expressed as the mean of Log10 individual antibody titers ± SEM. c, d CDC activity of pre-immunized (day 1) and post-immunized (day 126) sera of cynomolgus monkeys was analyzed on various Tn-expressing cells. Representative pictures of green and red fluorescence and associated transmission images for CDC on Jurkat-GFP cells are shown (c). The results are expressed as the mean of individual % of CDC ± SEM (d). e BALB/c mice grafted with TA3Ha cells were injected with CTX (50 mg/kg, day 1 after graft) and then treated (six injections) with the 8D4 murine anti-Tn mAb (n = 8), the Chi-Tn mAb (n = 8) or the Herceptin mAb (n = 6), and mouse survival was followed for 48 days. The statistical significance of differences was determined by the Student t test (*P < 0.05; **P < 0.01)