HeLa, Lewis lung carcinoma, B16-F1, and B16-F10 cells, obtained from the American Type Culture Collection, were cultured in Dulbecco’s modified Eagle’s medium containing 10 % fetal bovine serum at 37 °C with 5 % CO2. Mouse embryonic stem (ES) cells (E14) were cultured in StemMedium Serum Free Media for Mouse ES Cell (DS PHARMA BIOMEDICAL, Suita, Japan) containing 0.1 mM β-mercaptoethanol and ESGRO Leukemia Inhibitory Factor (Chemicon, Merck Millipore, Darmstadt, Germany) on tissue culture dishes gelatinized using ESGRO Complete Gelatin Solution (Merck Millipore).
Generation of a rabbit anti-mouse FEAT antibody
A rabbit polyclonal antibody was produced and affinity purified by Eurofins Genomics (Tokyo, Japan) using His-tagged mouse FEAT purified with TALON Metal Affinity Resin (Clontech Laboratories, Takara Bio, Otsu, Japan) under denaturing conditions.
Selection and synthesis of major histocompatibility complex (MHC) class I-restricted peptides
Prediction of 9-mer and 10-mer peptides with high affinity for H-2Kb and H-2Db was performed using the mouse FEAT (Mettl13) sequence and BIMAS  and SYFPEITHI  software. Peptides with a purity of >90 % were synthesized by Eurofins Genomics.
Female C57BL/6J mice (6 weeks old) were purchased from Charles River Japan (Yokohama, Japan). Experiments were approved by an animal experiment committee at Kyushu University and performed in accordance with national and institutional guidelines for animal use in research.
Immunization was conducted subcutaneously with 20 μg peptide A (EWYGTYLEL) and/or 20 μg peptide B (ALLRNPELL) in 100 μl phosphate-buffered saline (PBS) containing 10 μg AbISCO-100 adjuvant (Isconova, Novavax, Gaithersburg, MD, USA) in the left flank twice at a 1-week interval. Control mice only received PBS or the adjuvant. One week after the second immunization the mice were injected subcutaneously with 1 × 105 B16-F10 cells into their right flank. The tumor size was monitored daily. Mice were sacrificed when the largest tumor in the experiment reached 10 mm in diameter. Vaccination experiments were conducted four times using three mice per treatment group.
Mice were euthanized by neck dislocation, dissected, and fixed with 3.7 % formaldehyde in PBS. Tissues were cut into 2-mm-thick sections and placed into tissue cassettes (Tissue-Tek Uni-Cassette, Sakura Finetek Japan). Fixed tissues were embedded in paraffin, sectioned with a microtome, and stained with hematoxylin and eosin (H&E) by the Laboratory of Technology, Medical Institute of Bioregulation, Kyushu University.
Blood samples were collected from the peri-orbital sinus of mice, allowed to clot, and centrifuged at 800×g for 15 min. Serum aspartate aminotransferase (AST)/glutamate oxaloacetate transaminase (GOT), alanine aminotransferase (ALT)/glutamate pyruvate transaminase (GPT), and creatinine were measured with a FUJI DRI-CHEM 3500v and FUJI DRI-CHEM slides (Fujifilm, Tokyo, Japan).
Paraffin-embedded sections were deparaffinized with Clear-Advantage (Polysciences, Warrington, PA, USA), rehydrated, treated with Citrate-based Antigen Unmasking Solution (Vector laboratories, Burlingame, CA, USA), and stained with a mouse anti-CD3-ζ monoclonal antibody (6B10.2) (1:500 dilution; Santa Cruz Biotechnology, Dallas, TX, USA) and a rabbit polyclonal anti-CD8 antibody (1:200 dilution; Bioss Antibodies, Woburn, MA, USA) using a Multiview (mouse-HRP/rabbit-AP) IHC kit and IHC background blocker (Enzo Life Sciences, Farmingdale, NY, USA). The sections were counterstained with Mayer’s Hematoxylin (Merck), and coverslips were mounted with CC/Mount Aqueous Permanent Mounting Medium (Diagnostic BioSystems, Pleasanton, CA, USA). Images were acquired using a BZ-9000 Fluorescence Microscope (KEYENCE, Osaka, Japan) and a Zeiss Axioskop 2 plus microscope equipped with a Zeiss AxioCam camera controlled by AxioVision software (Carl Zeiss Microscopy, Jena, Germany).
Heparinized human blood plasma from healthy volunteers
After receiving written informed consent, blood was collected from three healthy volunteers by venipuncture into a tube containing sodium heparin and centrifuged at 1200 rpm for 10 min at 21 °C. Five samples of normal human plasma in sodium heparin were purchased from Cosmo Bio (Tokyo, Japan).
ImmunoCruz IP/WB reagents (Santa Cruz Biotechnology) were used to immunoprecipitate FEAT from 1 mg of plasma.
Proteins were subjected to sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and semidry transfer as described previously . The membrane was stained with primary and secondary antibodies using Can Get Signal Immunoreaction Enhancer Solution (TOYOBO, Osaka, Japan). Signals were visualized with Amersham ECL Select Western blotting detection reagent and an ImageQuant LAS 500 (GE Healthcare Life Sciences, Little Chalfont, UK). Antibodies used were: a rabbit polyclonal antibody against peroxiredoxin 1 (Atlas Antibodies, Stockholm, Sweden) and mouse monoclonal antibodies against β-actin (Santa Cruz Biotechnology) and FEATΔC (7F2) (MBL, Nagoya, Japan).
Immunogold electron microscopy
Adherent cells were washed once with PBS and exposed to a prefixation solution (4 % paraformaldehyde, 0.4 % glutaraldehyde, 3.4 % sucrose, and 3 mM CaCl2 in 0.1 M cacodylate buffer, pH 7.4) for 10 min at room temperature. The cells were detached by a cell scraper (Nunc, Thermo Scientific) and centrifuged at 2000 rpm for 10 min in a swing-bucket rotor. After overnight fixation at 4 °C, the cell pellet was washed with PBS for 1 h at room temperature, dehydrated in a graded series of ethanol solutions on ice, embedded in LR White resin (Medium grade, Electron Microscopy Sciences, Hatfield, PA, USA) at −20 °C, exposed to ultraviolet radiation for 3 days at −20 °C, and then incubated for 24 h at 45 °C. Resin-embedded cells were sectioned at 100-nm thicknesses using a Leica EM UC7 Ultramicrotome (Leica Microsystems, Wetzlar, Germany) and collected on Formvar carbon-coated nickel grids. Grids were blocked with 3 % bovine serum albumin (BSA)-PBS for 15 min at room temperature, stained with rabbit anti-human FEATΔN or anti-mouse FEAT antibodies in 0.3 % BSA-PBS overnight at 4 °C, and then incubated with colloidal gold (10-nm)-conjugated goat anti-rabbit IgG (1:50 dilution in PBS; EY Laboratories, San Mateo, CA, USA) for 2 h at room temperature. Postfixation was performed with 0.5 % osmium tetroxide in PBS for 5 min at room temperature, followed by staining with 2 % uranyl acetate for 5 min to confer a light contrast. Cell sections were examined using a Tecnai 20 transmission electron microscope (FEI, Hillsboro, OR, USA). Images were acquired with an Eagle 2k CCD camera with a high resolution scintillator (FEI).
A solid phase sandwich ELISA kit for human FEAT (FEAT Assay Kit; Lot. 1H-512 and 1L-517) was produced by the contract manufacturing service of Immuno-Biological Laboratories (Fujioka, Japan). To 96-well plates coated with the rabbit polyclonal anti-human FEATΔC antibody, 100 μl EIA buffer (1 % BSA and 0.05 % Tween-20 in PBS) and 100 μl sample were added, followed by overnight incubation at 4 °C. The plates were treated with 100 μl/well mouse anti-human FEATΔC monoclonal antibody (MBL) for 30 min at 37 °C, 100 μl/well anti-mouse IgG (H + L) goat IgG Fab′ conjugated with horseradish peroxidase for 30 min at 37 °C, and then 100 μl/well tetramethylbenzidine solution for 30 min at room temperature. The reaction was stopped with 100 μl/well 1 N H2SO4. Optical density was read at 450 nm using an EnSpire Multimode Plate Reader (PerkinElmer, Waltham, MA, USA). Purified His-tagged recombinant human FEAT was used to generate a standard curve.
Plasma C-reactive protein (CRP) was measured with a Quantikine ELISA Human CRP Immunoassay kit (R&D Systems, Minneapolis, MN, USA).
Exosomes were purified from plasma using a Total Exosome Isolation (from plasma) kit (Invitrogen, Thermo Fisher Scientific, Waltham, MA, USA). Exosomes in the pellet were resuspended in SDS-PAGE sample buffer and incubated for 3 min at 95 °C.
Statistical analyses were performed using the Statcel4 add-in package (OMS Publishing, Tokorozawa, Japan) for Microsoft Excel. The Kruskal–Wallis test, a non-parametric one-way analysis of variance, was used for AST/GOT and ALT/GPT because homogeneity of variances was unlikely according to Bartlett’s test. The Kruskal–Wallis test followed by the Steel test, a multiple comparison test for non-parametric data, were used for ELISA data, because a normal distribution was unlikely according to Pearson’s Chi-square goodness-of-fit test.