Albumin/vaccine nanocomplexes that assemble in vivo for combination cancer immunotherapy

Abstract

Subunit vaccines have been investigated in over 1000 clinical trials of cancer immunotherapy, but have shown limited efficacy. Nanovaccines may improve efficacy but have rarely been clinically translated. By conjugating molecular vaccines with Evans blue (EB) into albumin-binding vaccines (AlbiVax), here we develop clinically promising albumin/AlbiVax nanocomplexes that self-assemble in vivo from AlbiVax and endogenous albumin for efficient vaccine delivery and potent cancer immunotherapy. PET pharmacoimaging, super-resolution microscopies, and flow cytometry reveal almost 100-fold more efficient co-delivery of CpG and antigens (Ags) to lymph nodes (LNs) by albumin/AlbiVax than benchmark incomplete Freund's adjuvant (IFA). Albumin/AlbiVax elicits ~10 times more frequent peripheral antigen-specific CD8⁺ cytotoxic T lymphocytes with immune memory than IFA-emulsifying vaccines. Albumin/AlbiVax specifically inhibits progression of established primary or metastatic EG7.OVA, B16F10, and MC38 tumors; combination with anti-PD-1 and/or Abraxane further potentiates immunotherapy and eradicates most MC38 tumors. Albumin/AlbiVax nanocomplexes are thus a robust platform for combination cancer immunotherapy.

Fig. 1

Schematic of albumin/AlbiVax nanocomplexes for efficient vaccine delivery and combination cancer immunotherapy. a Upper: structure of HSA (PDB ID: 2BXH) and chemical structure of MEB; lower: schematic structure of albumin/MEB nanocomplexes (left) and 3D molecular structure predicted by molecular docking (right). Sticks represent MEB and the amino acid residues in the binding site I of HSA. Green dashed lines represent hydrogen bonds between MEB and amino acids. b Working mechanism of albumin/AlbiVax nanocomplexes as potent T cell vaccines. Left box: modular structures of AlbiCpG and albumin-binding Ag (AlbiAg). AlbiCpG were engineered by site specifically conjugating MEB and thiol-modified CpG, with hexaethyloxy-glycol (HEG) as tunable linkers; AlbiAg was synthesized by conjugating MEB and cysteine-modified Ags, including TAA and tumor-specific neoantigen discovered via exome sequencing. Left lower: locally administered AlbiVax binds to endogenous albumin and assembles into albumin/AlbiVax nanocomplexes, which were efficiently delivered to LNs due to lymphatic drainage and prolonged retention in LNs. Right: harnessing the endocytosis pathway of albumin, albumin/AlbiCpG and albumin/AlbiAg nanocomplexes were co-delivered into APCs and activated APCs for antigen cross presentation and clonal expansion of antigen-specific CD8⁺ CTLs, thereby eliciting robust and durable antitumor immunity. While albumin/AlbiVax nanocomplexes upregulated the expression of PD-1 on these CD8⁺ CTLs, combination of albumin/AlbiVax nanocomplexes with anti-PD-1 dramatically enhanced immunotherapeutic efficacy in established primary and metastatic tumors. The pharmacological behaviors of albumin/AlbiVax nanocomplexes were studied by quantitative PET imaging, light sheet fluorescence microscopy in whole cleared tissue, and super-resolution imaging in single APCs

Fig. 2

Quantitative screening of albumin/AlbiCpG nanocomplexes for LN delivery. a Structures, radiolabeling, and formulations of CpG derivatives for PET-based screening. Shown in the middle is the molecular structure of NMEB used for radiolabeling of AlbiCpG. Bioconjugations were conducted using thiol on Cp, maleimide on NMEB (for AlbiCpG), or NOTA (for CpG and IFA(CpG)), as well as maleimide and alkyne on bifunctional PEG20K and azide on NOTA for PEG–CpG. b Upper: representative coronal (coro) PET images showing FVB mice at 6 h post s.c. injection (dose: 4.4–5.5 Mbq) of CpG derivatives at tail base. Nanocomplexes of albumin with four MEB–CpG derivatives, respectively, efficiently delivered CpG to LNs, relative to free CpG, PEG–CpG, and IFA(CpG). Lower: representative transverse (trans) PET images showing LN delivery of MH₂C at 6 h post injection. White arrow heads mark IN and AX LNs. c Amounts of CpG derivatives in IN and AX LNs quantified from three-dimensional (3D)-reconstructed, decay-corrected PET images (n = 4–8). The IFA(CpG) signal in LNs was too low to visualize for quantification. d Biodistribution of tested compounds in resected organs measured by γ counting 3 days post injection (n = 4). (Liver/gal: liver with gallbladder; intest.: intestine). e An AFM image of HSA/AlbiCpG nanocomplexes (premixed AlbiCpG: HSA = 1:1). Scale bar: 200 nm. f, g In C57BL/6 mice, AlbiCpG s.c. injected at the tail base ameliorated the systemic toxicity of CpG (n = 4) as shown by lower IL-6 and IL-12p40 titers in blood (f) (dose: 5 nmol CpG equivalents) and ameliorated splenomegaly (g) on day 6 (dose: 5 nmol CpG equivalents on day 0 and day 3). Scale bar for spleen: 1 cm. Wt: weight. Data show mean ± s.e.m. of two–three independent experiments. ***p < 0.001, **p < 0.01, *p < 0.05, ns: not significant (p > 0.05) by one-way ANOVA with Bonferroni post test. Asterisks in c indicate statistically significant differences between the corresponding compounds with MH₂C

Fig. 3

Super-resolution observation of intranodal and intracellular co-delivery of albumin/AlbiVax nanocomplexes. a Photographs of non-cleared and PACT-cleared mouse LNs. Scale bar: 2 cm. b, c Light sheet fluorescence microscopy images showing the 3D intranodal distribution of albumin/AlbiCpG–Alexa488 nanocomplexes in a whole LN (b) and a close-up of the subcapsular sinus areas (c left: 3D distribution; right: cross sections) 1 day after injection. Substantial albumin/AlbiCpG nanocomplexes were located within or near the subcapsular sinus and around B cell follicles (Supplementary Videos 2 and 3). Scale bar in b: 400 µm; Scale bar in c: 200 µm. d Fractions of IN LN B220⁺ B cells, CD11c⁺ DCs, and F4/80⁺ macrophages that took up AlbiCpG-Alexa555 and/or MSA-FITC, on day 1 and day 3 post s.c. injection of premixed AlbiCpG + MSA into C57BL/6 mice (n = 4). e Deconvoluted confocal microscopy images showing AlbiCpG (200 nM) in the endolysosomes of one single BMDC after 2-h incubation. Inset: 2 endolysosomes showing AlbiCpG primarily on the endolysosome membrane. (Red: AlbiCpG; green: LysoTracker Green.) Scale bar: 2 µm; inset: 500 nm. f Instant SIM super-resolution images showing co-localization of MSA-Alexa555 (3 mg mL⁻¹) and AlbiCpG-Alexa488 (200 nM) in one BMDC after 2-h incubation. Scale bars: 4 µm. g Fractions of IN LN B220⁺ B cells, CD11c⁺ DCs, and F4/80⁺ macrophages that took up AlbiCpG-Alexa555 and AlbiCSIINFEK_(FITC)L on day 1 and day 3 post s.c. injection of AlbiCpG + AlbiSIINFEKL into C57BL/6 mice (n = 4). h Instant SIM images showing intracellular co-delivery of AlbiCpG-Alexa555 (200 nM) + AlbiCSIINFEK_(FITC)L (200 nM) in one BMDC after 2-h incubation. AlbiCSIINFEK_(FITC)L was located together with AlbiCpG as well as separately in the cytosol. Scale bars: 2 µm. i Confocal microscopy images showing efficient antigen presentation of AlbiCSIINFEK_(FITC)L (500 nM) + AlbiCpG (500 nM), compared with CSIINFEK_(FITC)L (500 nM) + CpG (500 nM) in BMDCs. For incubation time longer than 6 h, fresh medium was substituted after 8-h incubation. (Blue: Hoechst 33342; green: FITC; red: LysoTracker Red.) Scale bars: 10 µm

Fig. 4

Albumin/AlbiVax nanocomplexes induced potent and durable antitumor T cell responses. a–g C57BL/6 mice (n = 5–7) were s.c. vaccinated with AlbiVax (2 nmol AlbiCpG equivalents + 10 µg OVA) at the tail base on day 0, day 14, and day 28, followed by immune analysis on day 21, day 35, and day 70, and 1° tumor challenge on day 71. a Representative flow cytometry plots (left) and frequency (right) of SIINFEKL⁺CD8⁺ T cells in peripheral blood on day 21 stained using phycoerythrin (PE)-labeled H-2K^b-SIINFEKL tetramer. b Percentage of cytokine-producing CD8⁺ T cells in peripheral blood, measured by intracellular staining of IFN-γ and TNF-α on day 21. c Higher level (MFI) and frequency of PD-1 expression on SIINFEKL⁺CD8 T cells than that on total CD8⁺ T cells in peripheral blood on day 21. (Two-tailed paired t test.) d Frequencies of SIINFEKL⁺CD8⁺ CTLs in peripheral blood over 70 days post priming. e Representative flow cytometry results (upper) and percentage (lower) of effector memory T cells (Tem, CD62L⁻CD44⁺), central memory T cells (Tcm, CD62L^highCD44⁺), and naive T cells (CD62L⁺CD44⁻) in peripheral blood on day 70, showing AlbiVax-induced T cell memory. f, g Tumor growth curve (f) and mouse survival (g) after s.c. challenging vaccinated mice with EG7.OVA cells. 1° challenge: 3 × 10⁵ cells on the right shoulder on day 71 post priming vaccination; 2° challenge: 1 × 10⁶ cells on the right flank on day 211. h AlbiCpG + OVA regressed established EG7.OVA tumor. C57BL/6 mice (n = 6–8) were s.c. inoculated with 3 × 10⁵ EG7.OVA cells on day 0, and treated with AlbiCpG + OVA (2 nmol CpG equivalents, 20 µg OVA) on day 6, day 12, and day 18. Lymphocyte depletion by anti-CD8, but not anti-CD4 or anti-NK1.1 (200 μg, on day 6, day 9, day 12, day 15, and day 18) abrogated the therapeutic efficacy of AlbiVax. ***p < 0.001, **p < 0.01, *p < 0.05, ns: not significant (p > 0.05), by one-way ANOVA with Bonferroni post test, unless denoted otherwise. Data show mean ± s.e.m. of two–three independent experiments. Asterisks in a, b, f–h indicate statistically significant differences between AlbiCpG and other groups

Fig. 5

Albumin/AlbiVax nanocomplexes for melanoma combination immunotherapy. a Transformation of AlbiTrp2 nanoparticles into albumin/AlbiTrp2 nanocomplexes in the presence of HSA (molar ratio of HSA: albumin = 1:1). Insets: an AFM image (left) and a TEM image showing the corresponding nanoparticles. Scale bars: 200 nm. b Representative coronal, transverse, and 3D projection of PET images at 3 h post injection (left), and representative coronal PET images at 3, 6, 24, and 48 h post s.c. injection of AlbiTrp2, free Trp2, and IFA(Trp2) at the tail base of FVB mice. (Dose: 4.4–5.5 Mbq.) c Quantification of albumin/AlbiTrp2 nanocomplexes and Trp2 in draining LNs (IN + AX). White arrows mark LNs. d–f B16F10 tumor growth after treatment with AlbiVax (d), double combination of albumin/AlbiVax nanocomplexes and anti-PD−1 (e), and triple combination of albumin/AlbiVax nanocomplexes, anti-PD-1, and Abraxane (f). C57BL/6 mice were s.c. inoculated with 3 × 10⁵ B16F10 cells, treated with AlbiVax (2 nmol CpG equivalents + 20 µg AlbiTrp2) (day 6, day 12, and day 18), anti-PD-1 every 3 days from day 6 for five times (200 µg), and Abraxane on day 6, day 12, and day 18 (20 mg kg⁻¹). g–j C57BL/6 mice were i.v. injected with 1 × 10⁵ B16F10-fLuc cells, treated with AlbiVax (2 nmol CpG equivalents + 20 µg AlbiTrp2) on day 6, day 12, and day 18 and anti-PD-1 (200 µg) every 3 days from day 6 for six times. g Representative bioluminescence images on day 14, day 17, and day 20, and photographs of lungs on day 20. i Quantified bioluminescence intensities of lungs. i, j Numbers of tumor nodules (i) and lung weights (j) on day 20. Data show mean ± s.e.m. of two–three independent experiments. ***p < 0.001, **p < 0.01, *p < 0.05, ns not significant (p > 0.05) by one-way ANOVA with Bonferroni post-test

Fig. 6

Neoantigen-based albumin/AlbiVax nanocomplexes for personalized cancer immunotherapy. a Upper: representative coronal and transverse PET images at 6 h post injection, and lower: quantification of albumin/AlbiAdpgk nanocomplexes and Adpgk in IN and AX LNs of FVB mice (n = 4). White arrows mark LNs. IFA(Adpgk) was undetectable in LNs. (Dose: 4.4–5.5 Mbq) b–e C57BL/6 mice (n = 5) were vaccinated with AlbiVax (2 nmol AlbiCpG + 20 µg AlbiAdpgk) on day 0 and day 14, followed by flow cytometric analysis of H-2D^b-ASMTNMELM tetramer⁺CD8⁺ T cells (b, c) and PD-1 expression on peripheral CD8⁺ T cells (d) on day 21, and CD8⁺ T cell central memory (CD62L^highCD44⁺) on day 50 (e). f Exome-sequencing results verifying Adpgk variant in MC38 cells. g, h MC38 tumor growth after treatment with AlbiVax alone or in combination with anti-PD-1. C57BL/6 mice were s.c. inoculated with 3 × 10⁵ MC38 cells, treated with AlbiVax (2 nmol AlbiCpG + 20 µg AlbiAdpgk) on day 6, day 12, and day 18), and with anti-PD-1 (200 µg) every 3 days from day 6 for six times. h Depletion of CD8⁺ T cells, but not CD4⁺ T cells or NK cells abrogated the therapeutic efficacy of AlbiVax (n = 5). i–k Immunotherapy of lung metastatic MC38 tumor with AlbiVax alone or in combination with anti-PD-1. C57BL/6 mice were i.v. inoculated with 1 × 10⁵ MC38 cells, treated with AlbiVax (2 nmol AlbiCpG + 20 µg AlbiAdpgk) on day 10, day 16, and day 22), and with anti-PD-1 (200 µg) every 3 days from day 10 for six times. On day 40, mice were injected with FDG tracer (100 µCi). Mice were killed and lungs and tumors were collected (i), weighted (j), and radioactivity measured by γ-counting (k). Data show mean ± s.e.m. of two–three independent experiments. ***p < 0.001, **p < 0.01, *p < 0.05, ns: not significant (p > 0.05) by one-way ANOVA with Bonferroni post-test