T cell antigen discovery using signaling and antigen presenting bifunctional receptors (SABRs)

CD8+ T cells recognize and eliminate tumors in an antigen-specific manner. Despite progress in characterizing the antitumor T cell repertoire and function, identifying their target antigens remains a challenge. Here, we describe the use of chimeric receptors called Signaling and Antigen-presenting Bifunctional Receptors (SABRs) in a novel cell-based platform for T Cell Receptor (TCR) antigen discovery. SABRs present an extracellular peptide-MHC complex and induce intracellular signaling via a TCR-like signal upon binding with a cognate TCR. We devised a strategy for antigen discovery using SABR libraries to screen thousands of antigenic epitopes. We validated this platform by identifying the targets recognized by public TCRs of known specificities. Moreover, we extended this approach for personalized neoantigen discovery. The antigen discovery platform reported here will provide a scalable and versatile way to develop novel targets for immunotherapy. vectors and used to transduce NFAT-GFP-Jurkat cells. NFAT-GFP-Jurkat cells expressing the SABR library were co-cultured with Jurkat cells expressing an ‘orphan’ TCR. GFP+CD69+ NFAT-GFP-Jurkat cells were sorted using fluorescence activated cell sorting (FACS), followed by genomic DNA extraction. The epitope portion of the SABRs was amplified and subjected to high throughput sequencing. The sequencing reads were aligned with the SABR vector backbone using Burrows-Wheeler alignment. Aligned reads were translated to reveal the epitope. The number of reads corresponding to each epitope was counted and reported in a list. A minimum of three replicates of the co-incubation assay were performed. For each replicate, a numerical rank was given to each epitope based on descending order of the number of reads. The rank from three replicates for each assays was averaged and reported as ‘Average Rank’. The top ranked epitopes were putative antigens for that TCR and are subsequently validated by constructing individual SABRs presenting each of the epitopes and measuring GFP expression in co-culture assays.


Introduction
Here, we sought to develop novel antigen discovery techniques to address the unmet need. We present a cell-based platform for T cell antigen discovery. In this study, we describe chimeric receptors called Signaling and Antigen-presenting Bifunctional Receptors (SABRs). By virtue of genetically linking the peptide epitope with MHC, SABRs can be used to present a defined antigen and to report its successful recognition by a TCR. Therefore, we asked if SABR libraries presenting a large number of epitopes can be used to screen successful TCR-pMHC interactions. An orphan TCR may be obtained by single cell sequencing on tumor infiltrating lymphocytes isolated from a patient sample.
We designed a strategy to construct and use SABR libraries for T cell antigen discovery of 'orphan' antitumor TCRs with unknown antigens. First, a list of target epitopes was generated from an existing database or from tumor exome data followed by prediction of MHC binding. The protein sequences of the target epitopes were backtranslated to generate oligonucleotide sequences. Overhangs of sequences with 15bp overlap with the SABR vector were added to the oligonucleotide sequences corresponding to the epitopes. The entire list of oligonucleotides for the library was synthesized using pooled synthesis. The pooled library was then amplified and cloned using ligation-free cloning using the 15bp overhangs into the SABR vector plasmid. The SABR libraries were packaged into lentiviral

1.
Starting with a list of desired peptides to express in a SABR library, generate backtranslated nucleotide sequences with the IDT Codon Optimization Tool. Select "Amino Acids" as the input sequence type, "gBlocks Gene Fragments" as the Product Type, "Homo sapiens (human)" as the Organism, and enter a tab-delimited list of peptides on the "Bulk Entry" page.

3.
Append random nucleotides to the end of the epitope oligo for normalization + "GGCGGCCGCTCATCTCGTGCGACATCAAGCTATACTCTAATATAGCATTCCGTTCGAGTATAGCAG" 4. Submit the first 100 bp of each nucleotide sequence with appended random nucleotides for synthesis =Left(, 100).
Pause point: The synthesized pooled library can be stored at -20°C indefinitely

Cloning Oligonucleotides in SABR vectors
The plasmids for HLA-A0201-SABR backbone (  If using Ethidium bromide to visualize gels: Ethidium bromide is a carcinogen.
Dispose of it appropriately.

9.
Amplify library oligos: Dilute or resuspend single-stranded oligo pool to 10 ng/ul with Elution Buffer (supplied in the PCR purification kit). Prepare 600 ul library amplification mix (300 ul of 2X KOD master mix, 24 ul of 10 uM Oligo-Amplify-Fwd, 24 ul of 10 uM Oligo-Amplify-Rev, 2.4 ul of 10 ng/ul single-stranded oligo pool, and 250 ul of Nuclease-free water), aliquot mix to 24 wells of a PCR plate or strips, and amplify according to the following program: Initial Denaturation at 95°C for 2 minutes, 6 cycles of 95°C for 20 seconds (denature), 61°C for 10 seconds (anneal), and 70°C for 15 seconds (extension), then a final extension at 70°C for 1 minute, and store at 4°C for up to a month or at -20°C indefinitely. Troubleshooting: If the oligonucleotides do not amplify, perform the reaction using a gradient of annealing temperatures from 50-70°C. Cycle number can be increased to 10-15 at the optimal temperature. Verify on gel to make sure alternate products do not exist, which can occur for high cycle numbers after primers are exhausted.

10.
Pool all amplification reactions in a 1.5 mL Eppendorf tube. PCR purify with NucleoSpin Gel and PCR purification kit, using one column, and elute in 17 ul Elution 8 Buffer (supplied). Expect above 10 ng/ul of double-stranded oligos.
Maxi prep the 1.5 L of liquid media across three maxi columns. Pool maxi preps in a 2 mL eppendorf tube. Pause point: Store plasmids at 4°C for up to one month or at -20°C indefinitely.

Packaging Lentiviral Vectors
Caution: All the steps in this subsection are to be performed at BSL2 with lentiviral vector precautions.

15.
Prepare D10 media (55 mL of FBS and 5.5 mL Penicillin Streptomycin added to a 500 mL bottle of DMEM), warm D10 to room temperature.

16.
Coat a 10 cm Tissue-culture treated dish with poly-L-Lysine. Add 10 ml of poly-L-Lysine to the plate and swirl to ensure the entire surface is covered. Aspirate the remaining poly-L-Lysine. Note: Poly-L-Lysine can be reused multiple times.
18. Add 6 ml of Trypsin and ensure that it covers the entire surface.

19.
Incubate plate at 37°C for 2-3 min until the cells detach from the surface.

20.
Add 15 ml of D10 and wash all the cells off the surface using a 10-ml serological pipet.

21.
Form a single cell suspension by repeatedly pipetting up and down the cells.

22.
Count the cells using a hemocytometer and adjust the concentration to 0.5e6 cells/ml with D10.

23.
Add 10 ml of cell suspension to the poly-L-Lysine coated plate. Shake the plate forward-backward and side-by-side to ensure even distribution of the cells.

28.
After the 20 minute incubation, combine both mixtures and mix well by pipetting up and down gently. Critical Step: ensure that the incubation is atleast 15 min, but no more than 30 min. Ensure that the mixing is performed gently. Do not vortex.

29.
Incubate at room temperature for 20 minutes.

30.
During this incubation, aspirate the medium from the plated HEK-293T cells from Day 1. Replace the medium with 10 ml of fresh D10.

31.
After the incubation in step 29, add the mixture dropwise to the plated HEK-293T

cells. Critical
Step: ensure that the incubation is atleast 15 min, but no more than 30 min. Do not disturb the cell monolayer.

33.
Remove the plate from the incubator. Note: the cell-free medium from this plate will be the lentiviral vector to be used for transduction.

34.
Harvest the medium with a 10 ml serological pipet without disturbing the cell monolayer. Filter the medium through a 0.45 micron PVDF filter. Critical step: Use only 0.45 micron PVDF filters. Using 0.22 micron or using other polymers will result in reduction of viral titers. Do not mix the cell-free medium while harvesting.

35.
Aliquot the filtered vector in 1.5ml microcentrifuge tubes and store at -80°C until further use.
Pause point: Store filtered virus at -80°C indefinitely. Troubleshooting: Viral titers may be negatively affected by plating a higher or lower cell number per plate, by prolonging or shortening the incubation beyond the recommended time, by filtering through the wrong filter, by using high passage number HEK-293T cells, by using plasmids that have degraded, by using incorrect amounts and ratios of the plasmids.

Transducing NFAT-GFP-Jurkat or Jurkat Cells
Caution: All the steps in this subsection are to be performed at BSL2 with lentiviral vector precautions.

Count NFAT-GFP-Jurkat cells or Jurkat cells with FACS or hemocytometer. Dilute to
1e6 cells/ml in R10.

Thaw the viral vector from
Step 35 at room temperature or 37°C. Mix by vortexing briefly.

39.
In 12 well Tissue-culture treated plates, distribute 500 ul of cell suspension from Step 37.

40.
Add 500 ul of the thawed viral vector. Mix well by pipetting up and down.

42.
If using NFAT-GFP-Jurkat cells, add 1 ml of R10+ 2mg/ml G-418 at 48 hours posttransduction. Troubleshooting: Transduction efficiency may be increased by using fewer cells or by using higher amount of the viral vector.

Co-culture Assay
Caution: All the steps in this subsection are to be performed at BSL2 with lentiviral vector precautions.

43.
Prepare R10 media (55 mL of FBS and 5.5 mL Penicillin Streptomycin added to a 500 mL bottle of RPMI), warm R10 to room temperature.  Pour a 2% agarose gel (mix 2.4 g agarose and 120 mL TAE buffer, heat to above 95°C until fully dissolved, cool to 60°C, add 10 ul gel green, mix and cast in a gel tray with wide combs until solidified). Combine 1 ul purified amplicon and 1 ul 6X loading dye, then load mixture on gel. Run gel at 120 V for 20 minutes, or until dye front has run 30% of the gel. Image the gel in a gel imager at 302 nm UV wavelength. Expected size of amplicon is 200 bp. Note: Amplicon might be very faint depending on the number of cells sorted. When amplifying very few cells, off-target amplification may occur. Analysis will ignore those products, but be sure to account for them while loading the flowcell. Caution: If using Ethidium bromide to visualize gels: Ethidium bromide is a carcinogen. Dispose of it appropriately.

59.
Run on BioAnalyzer following manufacturer's instructions. In brief: Load 9 ul gel-dye on to chip, add 5 ul Nano Marker to each sample well, load sample and ladder into appropriate wells, vortex chip, place chip in BioAnalyzer, and start analysis.

60.
Perfom Illumina sequencing on HiSeq2500 at a sequencing facility according to their instructions. Obtain the final data as FASTQ files.

Anticipated Results
To demonstrate proof-of-concept of the approach detailed in Fig 1, we asked if SABR libraries can identify the cognate antigens of known public TCRs in an unbiased manner. We constructed a SABR library encoding 12,055 epitopes presented on HLA-A0201 (A2-SABR-library) consisting of all known HLA-A0201-restricted epitopes from the Immune Epitope Database24 (IEDB). We first interrogated if the A2-SABR-library allows identification of the cognate antigen for two TCRs with known specificities (F5, which recognizes EAAGIGILTV, and SL9, which recognizes SLYNTVATL). We transduced NFAT-GFP-Jurkat cells with the A2-SABR-library, and incubated them with Jurkat cells expressing F5 or SL9 TCRs. After 10 hours of co-culture, we sorted GFP+CD69+ cells by FACS (Fig 3a), extracted genomic DNA, sequenced the epitopes, and calculated average ranks for each epitope as described in Fig 41 First, we plotted the average ranks of each of the epitopes from the SL9 sort against those from F5 sort (Fig 3b). Six epitopes formed a distinct cluster by their rank in the F5 sort.
In the SL9 sort, the top ranks were outliers, but did not form a separate cluster. The top six epitopes in the F5 sort were analogs of EAAGIGILTV, indicating successful identification of its antigen (Fig 3c).
Six out of the top ten epitopes from the SL9 sort were analogs of SLYNTAVATL (Fig 3d). Epitopes enriched for their corresponding TCRs were not enriched in mismatched TCRs (Fig 3c and d). The average fold-enrichment of the top hits from the F5-and SL9-sorts over the Mock-sort was 296 and 70 respectively. The noise observed in the SL9 sort is possibly due to the higher number of analogs of the SLYNTVATL peptide. The A2-SABR library contains 22 analogs EAAGIGILTV and 60 analogs of SLYNTVATL. A higher number of recognized analogs would lower the Average Rank for each of the epitopes because of competition among the epitopes. We compared the ranks of all the analogs of EAAGIGILTV and SLYNTVATL in the sorts. Six out of twenty-two EAAGIGILTV analogs were identified in the F5 sort (Fig 3e), whereas nine out of sixty SLYNTVATL analogs were identified in the SL9 sort (Fig   3f). The lack of identification of all the analogs is presumably due to reduced cross-reactivity of the F5 or SL9 TCRs towards them. Indeed, analogs SLYNTIATL (V6I) and SLFNTVATL (Y3F) are documented escape mutations in the SLYNTVATL epitope. We validated the top six hits from the F5 sort by in vitro cytotoxicity assays. We observed that all six analogs of the Mart1 peptide were specifically recognized by the F5 TCR, leading to induction of cytotoxicity (Fig 3g). Nevertheless, these experiments showed that a SABR library approach could identify the cognate antigen of a TCR by 17 screening thousands of epitopes. Example of expected outcome for two TCRs a. Sorting A2-SABR library cells based on reporter gene expression. Co-culture assays using 9 million library cells with 9 million TCR-21 transduced Jurkat cells were set up. At 10 hours post co-culture, cells were stained for CD69 and sorted using FACS. Representative flow cytometry plots from one replicate are shown.
The rectangle in the top right corner of each flow plot shows the gate used for the sort.
Frequency of cells in the sort gate is indicated as percentage. b. Average ranks from F5 and SL9 sorts. Each dot represents the average rank for a unique epitope as calculated using the procedure described in