Development of a novel immunoproteasome digestion assay for synthetic long peptide vaccine design

Abstract

Recently, many autologous tumor antigens have been examined for their potential use in cancer immunotherapy. However, the success of cancer vaccines in clinical trials has been limited, partly because of the limitations of using single, short peptides in most attempts. With this in mind, we aimed to develop multivalent synthetic long peptide (SLP) vaccines containing multiple cytotoxic T-lymphocyte (CTL) epitopes. However, to confirm whether a multivalent vaccine can induce an individual epitope-specific CTL, the only viable screening strategies currently available are interferon-gamma (IFN-γ enzyme-linked immunospot (ELISPOT) assays using human peripheral blood mononuclear cells, or expensive human leukocyte antigen (HLA)-expressing mice. In this report, we evaluated the use of our developed murine-20S immunoproteasome (i20S) digestion assay, and found that it could predict the results of IFN-γ ELISPOT assays. Importantly, the murine-i20S digestion assay not only predicted CTL induction, but also antitumor activity in an HLA-expressing mouse model. We conclude that the murine-i20S digestion assay is an extremely useful tool for the development of "all functional" multivalent SLP vaccines.

Fig 1. Time-dependent cleavage of trivalent SLP in the murine-i20S digestion assay.

The i20S digested fragment was analyzed by MS/MS with a UPLC system. (a) Total ion chromatogram (TIC) traces of the bulk peptide products derived after 0, 1, 2, or 4 h digestion of the SLP (highlighted in black, purple, green, or red, respectively). (b) TIC shown in Fig 1A with change in the vertical axis scale (4 h time point) to optimize visualization of each peak. Numbers above each peak correspond to fragment numbers in Fig 1E. (c) Transformed MS spectrum of peak 36. The calculated mass of protonated DYLRSVLEDF is 1256.6157. (d) MS^E spectrum of peak 36. (e) “digestion map”. Each fragment is highlighted in purple, green, or red according to the first detected time point. The detected intact epitope is outlined in bold.

Fig 2. Comparison between the murine-i20S digestion assay and specific CTL induction using HLA-expressing mice.

(a)-(f) The digestion maps of trivalent SLP fragments after 1, 2, or 4 h (highlighted in purple, green, or red, respectively). Epitope-related fragments are outlined in bold. (g)-(l) Results of IFN-γ ELISPOT in trivalent SLP-treated mice. Data represent mean ± s.d. (n = 3–4); open bars, target epitope peptide stimulation; closed bars, negative control peptide stimulation; * Student’s t-test, p < 0.05. (m) IFN-γ ELISPOT results in each short epitope peptide-treated mice. (n) Comparison of the results between CTL induction and the murine-i20S digestion assay. CTL-induced epitopes are colored in dark grey, and i20S digested epitopes are colored in purple, green or red, based on their first detected time-point i.e. 1, 2, or 4 h, respectively.

Fig 3. Application of the murine-i20S digestion assay to predict CTL induction for the HLA-A3 supertype-restricted epitope.

(a) IFN-γ ELISPOT assay results from SART3_734-742 epitope-treated mice. Data represent mean ± s.d. (n = 3–4); open bars, target epitope peptide stimulation; closed bars, negative control peptide stimulation; * Student’s t-test, p < 0.05. (b)-(d) The digestion map of S3-W-S3b, S2-S3b-W, S3b-S2-W constructs from the murine-i20S digestion assay. (e)-(g) IFN-γ ELISPOT assay results from S3-W-S3b, S2-S3b-W, and S3b-S2-W-treated mice. Peptides used for in vitro re-stimulation are shown at the bottom. (h) Comparison of the results between CTL induction and the murine-i20S digestion assay.

Fig 4. Antitumor efficacy of multivalent SLP vaccines in the syngeneic tumor mouse model.

HLA-A24 KI mice (n = 10) were subcutaneously treated with distilled water (control), SART2_93-101 or multivalent SLPs that were emulsified with Montanide ISA-51VG. Seven days after the final vaccination, B16F10.A24/SART2_93-101 cells (5 × 10⁶ cells) were subcutaneously engrafted into vaccinated mice. Tumor sizes were monitored with calipers, twice per week. Data are presented as mean tumor volume ± s.d.; * Student’s t-test, p < 0.05 vs. control group.

Fig 5. Comparison of cleavage preferences between murine-i20S and human-i20S.

(a)-(c) Sequence alignments of the catalytic β-subunits from mouse-i20S and human-i20S. Important residues on the active site are displayed in red. The substrate-recognizing pocket for primed substrate was called S’, and those for unprimed substrate were called S1, S2, and S3 respectively, from the C-terminal anchor residue of the epitope. Residues contributing to substrate-specificity pockets are highlighted by colored boxes: S1 pocket, green; S2 pocket, blue; S3 pocket, brown; S’ pocket, yellow. Different residues between murine and human β-subunits are marked with asterisks. Secondary structures (S: β-sheet & Helix) are indicated under the sequence. (d) The digestion map of S2-S3b-W in the human-i20S digestion assay. The numbers of epitope-related fragments detected in the murine-i20S digestion assay (Fig 3F) are enclosed in circles.