Lymph node-targeting adjuvant/neoantigen-codelivering vaccines for combination glioblastoma radioimmunotherapy

Abstract

Glioblastoma multiforme (GBM) is the most common and lethal type of adult brain cancer. Current GBM standard of care, including radiotherapy, often ends up with cancer recurrence, resulting in limited long-term survival benefits for GBM patients. Immunotherapy, such as immune checkpoint blockade (ICB), has thus far shown limited clinical benefit for GBM patients. Therapeutic vaccines hold great potential to elicit anti-cancer adaptive immunity, which can be synergistically combined with ICB and radiotherapy. Peptide vaccines are attractive for their ease of manufacturing and stability, but their therapeutic efficacy has been limited due to poor vaccine co-delivery and the limited ability of monovalent antigen vaccines to prevent tumor immune evasion. To address these challenges, here, we report GBM radioimmunotherapy that combines radiotherapy, ICB, and multivalent lymph-node-targeting adjuvant/antigen-codelivering albumin-binding vaccines (AAco-AlbiVax). Specifically, to codeliver peptide neoantigens and adjuvant CpG to lymph nodes (LNs), we developed AAco-AlbiVax based on a Y-shaped DNA scaffold that was site-specifically conjugated with CpG, peptide neoantigens, and albumin-binding maleimide-modified Evans blue derivative (MEB). As a result, these vaccines elicited antitumor immunity including neoantigen-specific CD8⁺ T cell responses in mice. In orthotopic GBM mice, the combination of AAco-AlbiVax, ICB, and fractionated radiation enhanced GBM therapeutic efficacy. However, radioimmunotherapy only trended more efficacious over radiotherapy alone. Taken together, these studies underscore the great potential of radioimmunotherapy for GBM, and future optimization of treatment dosing and scheduling would improve the therapeutic efficacy.

Keywords: DNA engineering; albumin; glioblastoma immunotherapy; neoantigen vaccine; vaccine codelivery.

Scheme 1

Lymph-node-targeting AAco-AlbiVax for combination cancer radioimmunotherapy. Upon s.c. administration, AAco-AlbiVax binds to endogenous albumin to form nanocomplexes, which are drained to LNs with prolonged retention. In LNs, AAco-AlbiVax are internalized by APCs, resulting in efficient presentation of antigenic epitopes, upregulated the expression of co-stimulation signals, and enhanced the production of proinflammatory cytokines, all of which are essential for T cell priming. As a result, AAco-AlbiVax, when combined with ICB and radiotherapy, improved the therapeutic efficacy of murine orthotopic GBM.

Figure 1

Construction of Y-shaped DNA-scaffolded vaccines. (A) Representative HPLC chromatograms of A3-NH₂, A3-SMCC, and A3-Rtn2. (B) An agarose gel electrophoresis image showing the gel retarding of A3-Rtn2, A3-Ntrk1, and A3-Imp3 conjugates relative to A3-NH₂, suggesting the successful conjugation of A3-NH₂ with these neoantigen peptides. (C) An agarose gel electrophoresis image showing the formation of Y-shaped structures via the hybridization of A1-MEB, A2, with A3-Rtn, A3-Ntrk, and A3-Imp3, respectively. StdMix: hybridization of unmodified A1, A2, and A3; A1 + A2: mixed A1 and A2; A1 + A3: mixed A1 and A3; A2 + A3: mixed A2 and A3; Y-Rtn2: hybridized A1-MEB, A2, with A3-Rtn. Y-Ntrk1: hybridized A1-MEB, A2, with A3-Ntrk1. Y-Imp3: hybridized A1-MEB, A2, with A3-Imp3. (D) Negligible IFN-β responses by DNA scaffold in mouse BMDCs and human THP-1 monocytes (100 nM, 24 h). ISG DNA served as a positive control. (E) IL-1β and IL-18 responses showed negligible AIM2 activation by dsDNA scaffold in mouse BMDCs (100 nM, 24 h). Poly(dA:dT) dsDNA served as a positive control. Data: means ± SD. ns: non-significant; and ****p < 0.0001, by Two-way ANOVA. Transfection vehicle: lipofectamine 3000.

Figure 2

In vitro immunostimulation and in vivo LN delivery of AAco-AlbiVax. (A) Confocal microscopy images showing the uptake of Y-shape scaffold into DC2.4 cells after a 3-h incubation. Scale bar: 20 µm. (B) ELISA results of IL-6 secretion from DC2.4 cells treated with free SIINFEKL + free CpG or Y-shape scaffold with CpG (100 nM) for 24 h. (C) Flow cytometric analysis of the SIINFEKL/H-2K^b complex levels on DC2.4 cells treated with SIIFENKL, free SIINFEKL + free CpG, or Y-shape scaffold with or without CpG, respectively, (CpG: 100 nM) for 24 h. (D) The activity of B3Z CD8⁺ T cells co-cultured with DCs pre-treated with the indicated vaccines. (E, F) Balb/c mice (n = 4) were s.c. injected (at the tail base) with IR800-labeled Y-shaped scaffold, with PBS, IR800 dye in IFA or Y-shape scaffold without MEB as controls, followed by IVIS imaging of mice. Shown are representative images illustrating the biodistribution at a series of time points post administration (E) and the quantified fluorescence signals in the same region of interest in draining inguinal LNs (F). Asterisks in (F) indicate statistical analysis between IFA IR800 or Y-shaped scaffold and Y-shaped scaffold with MEB. Data: means ± SEM. *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001 (Student's t-test).

Figure 3

GL261-specific neoantigen-based AAco-AlbiVax elicited T cell responses in mice. (A) Study design of T cell response studies elicited by AAco-AlbiVax in C57BL/6 mice (n = 5). (B) Frequencies of CD8⁺ T cells in peripheral blood by flow cytometry over 56 days post priming. (C) Flow cytometry plots of cytokine⁺ in CD8⁺ T cells from peripheral blood, measured by intracellular staining of IFN-γ and TNF-α on day 35. (D) Intracellular cytokine staining results showing the percentages of cytokine⁺ CD8⁺ T cells on total PBMC CD8⁺ T cells on day 35. (E) MFI (Median Fluorescence Intensity) quantification of PD-1 levels on peripheral CD8⁺ T cells on day 35. (F, G) Representative flow cytometry quantification of CD45⁺CD11c⁺ DCs and CD8⁺ DCs (CD11c⁺CD8⁺B220^-) in the above immunized mice on day 35. (H) Flow cytometry quantification of CD8⁺ memory T cells (central memory: CD62L⁺CD44⁺, effector memory: CD62L^-CD44⁺ and naive T cells: CD62L⁺CD44^-) in the above immunized mice on day 35. (I) ELISA results of serum anti-dsDNA IgG and IgM titers (day 35) in as-treated C57Bl/6 mice (n=5). AlbiCpG did not significantly induce anti-dsDNA IgG or IgM. (J) Serum anti-dsDNA IgG and IgM antibody titers for Group #5 over 42 days after priming. (K) Tumor growth curves in the above immunized mice challenged with GL261 on the right flank on day 89. Data: mean ± SEM; * p < 0.05; ** p < 0.01; *** p < 0.001 (one-way ANOVA).

Figure 4

Radioimmunotherapy of orthotopic GL261 GBM. (A) Study design of radioimmunotherapy that combined AAco-AlbiVax, ICB, and radiotherapy for orthotopic GL261 in C57BL/6 mice (n = 6). αPD-1, αCTLA-4: 200 μg each, i.p. injection; vaccine: 3 nmole CpG, 1 nmol antigen each, s.c. injection at tail base. Radiation: 5 Gy. (B, C) Representative bioluminescence images and quantified tumor bioluminescence signal intensities in orthotopic GL261 GBM-bearing C57BL/6 mice. (D, E) and mouse body weight (D) and Kaplan-Meier overall survival curves (E) of the above treated mice. Asterisks in (E) indicates statistical analysis for radioimmunotherapy vs PBS, radioimmunotherapy vs immunotherapy, and radioimmunotherapy vs radiotherapy. (F, G) Frequencies of CD8⁺ T cells and CD8⁺/CD4⁺ T cell ratio in peripheral blood from the above treated mice by flow cytometry on day 27 after tumor inoculation. (H) Flow cytometry quantified percentages of cytokine⁺ CD4⁺ T cells and cytokine⁺ CD8⁺ T cells as measured by intracellular staining of IFN-γ and TNF-α on day 27. (I) Representative flow cytometry quantification of CD45⁺CD11c⁺ DCs among PBMCs of the above treated mice on day 27. (J) Representative flow cytometry quantification of major CD11c⁺ DC subsets of CD8⁺ DCs (CD11c⁺CD8⁺B220^-), pDCs (CD11c⁺B220⁺) and migratory and residents CD8^- DCs (CD11c⁺CD8^-B220^-) among PBMCs of the above treated mice on day 27. (K) The percentage of CD45⁺CD11b⁺F4/80⁺ macrophages among PBMCs measured by flow cytometry on day 27. Data: mean ± SEM; * p < 0.05; ** p < 0.01; *** p < 0.001 (Student's t-test).