Cell line and reagents
The EBV-transformed human lymphoblastoid B-cell line JY (ECACC HLA-type collection, Sigma Aldrich) was cultured in RPMI 1640 (GIBCO, Invitrogen, Carlsbad, CA, USA) supplemented with 1% GlutaMAX (GIBCO, Invitrogen, Carlsbad, CA, USA) and 10% heat inactivated foetal bovine serum (HI-FBS, GIBCO, Invitrogen, Carlsbad, CA, USA).
Streptavidin (Streptomyces avidinii, affinity purified, lyophilized from 10 mm potassium phosphate, ≥13 U/mg protein) was purchased from Sigma-Aldrich (Saint Louis, Missouri, USA).
Biotin-conjugated anti-human HLA-A, B, C clone w6/32 was purchased from Biolegend (San Diego, CA, USA) for analysis.
The following peptides were purchased from Ontores Biotechnologies Co., Ltd., were used throughout the study: KVLEYVIKV (gene name MAGE A1), ILDKKVEKV (gene name HSP90), and QLVDIIEKV (gene name PSME3).
Additionally, the following peptides were purchased from Chempeptide (Shangai, China):
VIMDALKSSY (gene name NNMT), FLAEGGGVR (gene name FGA) and EVAQPGPSNR (gene name HSPG2).
Ovarian tumour biopsy and ethical consideration
The ovarian tumour biopsy was collected from a patient with ovarian metastatic tumour (high grade serous), who signed an informed consent, under the studies approved by the Research Ethics Committee of the Northern Savo Hospital District with the approval number 350/2020. The samples were chopped in small pieces and treated with a digestion buffer containing collagenase type D (Roche) 1mg/ml, Hyaluronidase (Sigma Aldrich) 100 mg/ml and DNase I (Roche) 1mg/ml for 1h at 37°C. The cell suspension was sequentially passed through a 500 mm and 300 mm cell strainer (pluriSelect) to obtain single cells.
Renal cell carcinoma and bladder tumour samples and ethical considerations
Patient tissue samples for organoid cultures were obtained from the DEDUCER study (Development of Diagnostics and Treatment of Urological Cancers) at Helsinki University Central Hospital with approval number HUS/71/2017, 26.04.2017, ethical committee approval number 15.03.2017 Dnro 154/13/03/02/2016, and patient consent. The kidney sample was obtained from a nephrectomy of an adult male with a clear cell renal cell carcinoma (ccRCC, pTNM stage pT3a G2). The benign kidney tissue sample was used for the experiments. The carcinoma urothelial (bladder cancer, high grade, gradus III, 1x1 cm) was obtained from adult female, and the cancer tissue sample was used for the organoid culture.
Clear cell renal carcinoma and bladder tumour organoid culture
Cells were isolated from the original tissue instantly after surgery by dissociating the tissue into small pieces and treating it with collagenase (40 units/ml) for 2–4 h. Benign and cancer cells of the kidney of a clear cell renal cell carcinoma patient cells were grown as organoids in F-medium [3:1 (v/v) of F‐12 nutrient mixture (Ham) - DMEM (Invitrogen), 5% FBS, 8.4 ng/mL cholera toxin (Sigma), 0.4 μg/mL hydrocortisone (Sigma), 10 ng/mL epidermal growth factor (Corning), 24 μg/mL adenine (Sigma), 5 μg/mL insulin (Sigma), 10 μM ROCK inhibitor (Y‐27632, Enzo Life Sciences, Lausen, Switzerland) and 1% penicillin-streptomycin with 10% Matrigel (Corning). The bladder tumour-derived organoids were grown in hepatocyte calcium medium (Corning)(15) supplemented with 5% CSFBS (Thermo Fisher Scientific), 10 μM Y-27632 RHO inhibitor (Sigma), 10 ng/mL epidermal growth factor (Corning), 1% GlutaMAX (Gibco), 1% penicillin-streptomycin and 10% Matrigel. 6x106 cells were collected by centrifugation, washed in PBS to remove Matrigel and snap frozen before analyses.
HLA typing
The clinical HLA typing of tumour samples (ccRCC and bladder) was performed by the European Federation for Immunogenetics (EFI) -accredited HLA laboratory of the Finnish Red Cross Blood Service. Allele determination of three classical HLA-I genes HLA-A, -B and -C was performed by targeted PCR based next generation sequence (NGS) technique according to the protocol provided by the manufacturer (NGSgo® Workflow, GenDx, Utrecht, The Netherlands).
The allele assignment at 4-field resolution level was implemented by NGSengine Version: 2.11.0.11444 (GenDx, Utrecht, The Netherlands) using IPD IMGT/HLA database Release 3.33.0, https://www.ebi.ac.uk/ipd/imgt/hla/.
Flow cytometry analysis
The following antibodies were used to analyse the cell surface expression of HLA-A2 and HLA-A, B, and C: PE-conjugated anti-human HLA-A2 (clone BB7.2, BioLegend 343306 San Diego, CA, USA), PE-conjugated anti-human HLA-A, B, and C (clone W6/32, BioLegend 311406, San Diego, CA, USA), and Human TruStain FcX block (BioLegend B247182, San Diego, CA, USA).
The data were acquired using a BDLSR Fortessa Flow Cytometer. Flow cytometric analysis of renal cell carcinoma and bladder tumour-derived organoids was performed using a BD Accuri 6 plus (BD Biosciences) and analysed with FlowJo software (Tree Star, Ashland, OR, USA).
Immobilized biotinylated pan-HLA antibody titer assay
The amount of immobilized biotinylated pan-HLA antibody has been tittered comparing the amount of the antibody in the feed solution versus the output solution. In detail, 12.5mg of anti-panHLA (Biolegend, cat. 311434, clone W6/32, biotin conjugated) in 25ml was added into microchip at each cycle and incubated for 15 minutes at room temperature. After the incubation time, the microchip was washed three times with 200mL of PBS and the elute collected. The antibody in the output solution was then quantified by ELISA. Briefly, maxisorb ELISA Nunc plates were coated with the output solution overnight at +4°C. After washing, 4%BSA (BioTop Oy) in PBS was added and incubated for 2h at 37°C, followed by washing steps in 0.05%Tween20 (Sigma Aldrich). Streptavidin/HRP (Pierce) was added for 30 minutes, followed by additional washing steps. Finally, TMB (Pierce) solution was applied for 20 minutes and Sulfuric Acid (Sigma Aldrich) 0.16M was used to stop the reaction and the plate read at 450 nm. The amount of biotinylated pan-HLA antibody was quantified by extrapolating the signal into a linear range (signal vs concentration) of a standard curve.
Purification and concentration of HLA class I peptides
HLA class I peptides were immunoaffinity purified from the JY human cell line using biotin-conjugated anti-human HLA-A, B, and C antibodies (clone W6/32, BioLegend 311434 San Diego, CA, USA). For sample preparation, the snap-frozen cell pellet was pipetted up and down 20 times in lysis buffer. The lysis buffer contained 150 mM NaCl, 50 mM TRIS-HCl, pH 7.4, protease inhibitors (A32955 Thermo Scientific Pierce, Waltham, Massachusetts, USA) and 1% Igepal (I8896 Sigma Aldrich, St. Louis, Missouri, USA). The lysates were first cleared by slow centrifugation for 10 min at 500xg, and then the supernatant was centrifuged for 30 min at 25,000xg. Next, HLA-I complexes were immunoaffinity purified from the cleared lysate with anti-human HLA-A, HLA-B, and HLA-C biotin-streptavidin bound to the micropillars of the biotinylated thiol-ene CHIP. The CHIPs were first washed three times with PBS, and then the HLA molecules were eluted at room temperature by adding acetic acid 7% (A113 Fisher Scientific, Leicestershire UK) in 50% MetOH (10402824 Fisher Scientific, Leicestershire UK).
Eluted HLA peptides and the subunits of the HLA complexes were desalted using SepPac-C18 cartridges (Waters) according to the protocol previously described by Bassani et al. (16). Briefly, the cartridge was prewashed with 80% acetonitrile in 0.1% trifluoroacetic acid (TFA) and then with 0.1% TFA. The peptides were purified from the HLA-I complex by elution with 30% acetonitrile in 0.1% TFA. Finally, the samples were dried using vacuum centrifugation (Eppendorf).
LC-MS/MS analysis of HLA class I peptides
Each dry sample was dissolved in 10 μL of LCMS solvent A (0.1% formic acid). The nanoElute LC system (Bruker, Bremen, Germany) injected and loaded the 10 μl of sample directly onto the analytical column (Aurora C18, 25 cm long, 75 µm ID, 1.6 µm bead size, Ionopticks, Melbourne, Australia) constantly kept at 50℃ by a heating oven (PRSO-V2 oven, Sonation, Biberach, Germany). After washing and loading sample at a constant pressure of 800 bar, the LC system started a 30 min gradient from 0 to 32% solvent B (acetonitrile, 0.1% formic acid), followed by increase to 95% B in 5 min, and finally a wash of 10 min at 95% B, all at a flow rate of 300 nL/min. Online LC-MS was performed using a Tims TOF Pro mass spectrometer (Bruker, Bremen, Germany) with the CaptiveSpray source, capillary voltage 1500V, dry gas flow of 3L/min, dry gas temperature at 180℃. MS data reduction was enabled. Mass Spectra peak detection maximum intensity was set to 10. Mobilogram peak detection intensity threshold was set to 5000. Mass range was 300-1100 m/z, and mobility range was 0.6-1.30 V.s/cm2. MS/MS was used with 3 PASEF (Parallel Accumulation – Serial Fragmentation) scans (300ms each) per cycle with a target intensity of 20000 and intensity threshold of 1000, considering charge states 0-5. Active exclusion was used with release after 0.4 min, reconsidering precursor if current intensity is >4 fold the previous intensity, and a mass width of 0.015 m/z and a 1/k0 width of 0.015 V.s/cm2. Isolation width was defined as 2.00 m/z for mass 700 m/z and 3.00 m/z for mass 800 m/z. Collision energy was set as 10.62 eV for 1/k0 0.60 V.s/cm2 and 51.46 eV for 1/k0 1.30 V.s/cm2. Precursor ions were selected using 1 MS repetition and a cycle overlap of 1 with the default intensities/repetitions schedule.
Proteomics database search
All MS/MS spectra were searched by PEAKS Studio X+ (v10.5 build 20191016) using a target-decoy strategy. The database used was the Swissprot Human protein database (including isoforms, 42373 entries, downloaded from uniprot.org on 20191126).
A precursor mass tolerance of 20 ppm and a product mass tolerance of 0.02 Da for CID-ITMS2 were used. Enzyme was none, digest mode unspecific, and oxidation of methionine was used as variable modification, with max 3 oxidations per peptide. A false discovery rate (FDR) cut-off of 1% was employed at the peptide level. The mass spectrometry proteomics data have been deposited to the ProteomeXchange Consortium via the PRIDE partner repository with the dataset identifier PXD022194. The dataset is currently hidden but will be made public upon eventual acceptance of the current manuscript.
Algorithms used for prediction of peptide ligands
Affinity to the corresponding HLA alleles was predicted for all eluted peptides identified in the JY cell line using NetMHC4.0. The threshold for binding was set to rank 0.5% to include only the strong binding partners.
GIBBS clustering analysis
Clustering of peptides into groups based on sequence similarities was performed using the GibbsCluster-2.0 tool with the default setting.
PBMC stimulation protocol
PBMCs from healthy donors were either purchased from Immunospot (Bonn, Germany) or isolated from whole blood of healthy donors using a Ficoll (Merck Millipore) density gradient and cultured in RPMI-1640 supplemented with 10% FBS (Gibco), 1% penicillin-streptomycin (Gibco), 1% GlutaMAX (Gibco), 15 mM HEPES, 50 µM b-mercaptoethanol (Gibco), and 1 mM sodium pyruvate (Gibco). The PBMCs were cultured and stimulated according to the following schedule:
- Day 0: thawing of the PBMCs and addition of 10 ng/mL IL-4, 800 IU/mL GM-CSF, 10 ng/mL IL-7 and 5 ng/mL IL-15.
- Day 2: addition of 10 ng/ml LPS, 50 IU/ml IFN-γ, 10 ng/ml IL-4, 800 IU/ml GM-CSF, 60 ng/ml IL-21, and 5 mg/ml peptides.
- Day 5 and 7: addition of 5 ng/ml IL-15, 5 ng/ml IL-7, 60 ng/ml IL-21, and 5 mg/ml peptides.
- Day 9: addition of 5 ng/ml IL-15, 5 ng/ml IL-7, and 5 mg/ml peptides.
PBMCs from patients were stimulated as previously described (17) with slightly modifications. 0.2x106 cells were allocated per well and cultured in RPMI-1640 (GIBCO, Invitrogen, Carlsbad, CA, USA) supplemented with 20%FBS (HI GIBCO, Invitrogen, Carlsbad, CA, USA), 1% penicillin-streptomycin (GIBCO, Invitrogen, Carlsbad, CA, USA), 1% GlutaMAX (GIBCO, Invitrogen, Carlsbad, CA, USA) and in presence of IL-4 (Biotechne) and GM-CSF (Biotechne) for 24h. 1 µM Peptide (Chempeptide), 0.5 ng/ml IL-7 (Biotechne) and 20 mg/ml Poly-I:C (Invitrogen) were added after 24 h. The stimulation lasted 9 days.
CD8+ T cell isolation
CD8+ T cells were isolated by MACS depletion (Miltenyi Biotec, Bergisch Gladbach, Germany) from PBMCs stimulated according to the aforementioned protocol.
Real-time impedance-based cytotoxicity assay
The cytotoxicity assay was performed using the xCELLigence real-time cell analysis system (ACEA Biosciences Inc.). Briefly, 80,000 JY cells or 75,000 ccRCC cells were seeded in a total volume of 50 ml per well (in an antiCD19 pre-coated 8-well plate in case of the JY) and cultured for 24 h at 37°C in 5% CO2. After 24 h, the effector cells (purified CD8 T cells) were added at a target (E/T) ratio of 1:1. The effector and target cells were cocultured for 36 h, and the CI of the target cells was measured every 1 h. The normalized cell index (NCI) was used for the analysis, and the following formula was applied:
% Specific Cytolysis at time t = 1- NCI(s)t/ NCI(M)t
where NCI is the normalized cell index, s is the Sample and M is Mock Effector Control.
Microchip design, fabrication and functionalization
The microchips used in this work incorporated a 30´4´0.2 mm3 (length´width´height) microchannel featuring an array of ca. 14 400 micropillars (diameter 50 μm, interpillar distance 100 mm from centre to centre) in a hexagonal lattice (Fig. 1). The internal volume of the microchannel was ca. 25 µL. The microchips were made of off-stoichiometric thiol-enes (OSTE) polymer composition as previously described by Tähkä et al. (17) and functionalized with biotin prior to use. Briefly, the OSTE prepolymer was prepared mixing a tetrafunctional thiol (pentaerythritol tetrakis(3-mercaptopropionate), Thiocure® PETMP, Bruno Bock, Marschacht, Germany) and a trifunctional ‘ene’ (triallyl-1,3,5-triazine-2,4,6(1H,3H,5H)-trione, 98%, Sigma Aldrich, St. Louis, MO) monomers at a ratio that yielded a 50% molar excess of thiol functional groups (i.e., 12.5% molar excess of the tetrathiol monomer) in the bulk solution. The monomer mixture was then poured onto a premade polydimethylsiloxane (Sylgard 184, Down Corning Corporation, Midland, MI) mould, incorporating a negative replica of the micropillar array, and kept under vacuum for ca. 5 min before curing the monomer mixture under UV for 5 min (Dymax 5000-EC Series UV flood exposure lamp, nominal power 225 mW/cm2, Dymax Corporation, Torrington, CT). After curing, the OSTE-based micropillar array was sealed by laminating a planar cover layer of the same composition, prepared in the same manner, on top of the micropillar array, and finalized by an additional UV exposure for 2 min (Dymax 5000-EC). The biotinylation of the micropillar array was achieved by filling the microchannel with 0.1 mg/mL biotin-PEG4-alkyne (Sigma Aldrich) in ethylene glycol, with 1% (m/v) Irgacure® TPO-L (BASF, Ludwigshafen, Germany) as the photoinitiator, after which the cross-linking reactions between biotin-PEG4-the alkyne and the surface thiols were initiated by UV (LED, λ = 365 nm, nominal intensity 14 mW/cm2). After UV exposure (1 min), the microchannel was rinsed sequentially with methanol (Sigma Aldrich) and Milli-Q water (3-5 mL each) and dried before use. The structural fidelity of the micropillar arrays was confirmed by scanning electron microscopy (Quanta™ 250 FEG, FEI, Hillsboro, OR) using a platinum coating (ca. 10 nm coating thickness).
Before loading the biotinylated pan-HLA antibody (1.6 mg/mL in PBS, BioLegend), the biotin-functionalized micropillars were precoated by filling the micropillar array with streptavidin (0.1 mg/mL in PBS, Sigma Aldrich), incubating for 15 min, and rinsing with 200uL of PBS three times.
Whenever fluorescent labelled streptavidin or antibodies were used, the quantitation of the fluorescence signal arising from on-chip immobilized biomolecules was performed through the top layer of the chip using a Zeiss Axioscope A1 upright epifluorescence microscope (Carl Zeiss Oy, Espoo, Finland) equipped with a HAL100W broadband lamp (Carl Zeiss) and a Hamamatsu R5929 photomultiplier tube coupled with a Cairn Integra signal amplifier module (Cairn Research, Faversham, UK). The on-chip fluorescence signals (excitation 488±5 nm, emission 500-700 nm) were quantitated and averaged from total of 3 locations along the micropillar array. Typically, 3-4 technical replicates (chips) were used.
Bioinformatic analysis
The functional annotation and visualisation was performed by using the clusterProfiler (19) Bioconductor package (v. 3.12.0) in the RStudio server environment (v. 3.6.0). ClusterProfiler implements a hypergeometric test to evaluate the statistical enrichment of the input gene list over the desired functional classes. Nominal p-values were adjusted by applying the Benjamini-Hochberg method (20) and the threshold was set to padj=0.01. The mapping between different human gene identifiers was performed through the use of the org.Hs.eg.db Bioconductor library (21) . The analysis of the Molecular Signatures Database (MSigDB) [URL: https://www.gsea-msigdb.org/gsea/msigdb] was performed by using the msigdbr CRAN package, while the visualisation of the results was obtained by employing the ComplexHeatmap Bioconductor package (22).
Statistical analyses
Statistical analysis was performed using GraphPad Prism 8.0 software (GraphPad Software Inc.). Details about the statistical tests for each experiment can be found in the corresponding figure legends.