Multiplex T Cell Stimulation Assay Utilizing a T Cell Activation Reporter-Based Detection System

Abstract

Recent advancements in single cell sequencing technologies allow for identification of numerous immune-receptors expressed by T cells such as tumor-specific and autoimmune T cells. Determining antigen specificity of those cells holds immense therapeutic promise. Therefore, the purpose of this study was to develop a method that can efficiently test antigen reactivity of multiple T cell receptors (TCRs) with limited cost, time, and labor. Nuclear factor of activated T cells (NFAT) is a transcription factor involved in producing cytokines and is often utilized as a reporter system for T cell activation. Using a NFAT-based fluorescent reporter system, we generated T-hybridoma cell lines that express intensely fluorescent proteins in response to antigen stimulation and constitutively express additional fluorescent proteins, which serve as identifiers of each T-hybridoma expressing a unique TCR. This allows for the combination of multiple T-hybridoma lines within a single reaction. Sensitivity to stimulation is not decreased by adding fluorescent proteins or multiplexing T cells. In multiplexed reactions, response by one cell line does not induce response in others, thus preserving specificity. This multiplex assay system will be a useful tool for antigen discovery research in a variety of contexts, including using combinatorial peptide libraries to determine T cell epitopes.

Keywords: T cell receptors; antigens; epitopes; multiplex assay; reporter.

FIGURE 1

Comparison of assistant promoters. (A) Schematic showing the relative positions of NFAT binding sites, assisting promoters, and ZsGreen-1 gene used in the NFAT-reporter constructs. (B–E) GSE.6H9 TCR 5KC T-hybridoma cells transduced with an NFAT-reporter construct containing the either miniCMV promoter or TATA-box were cultured with anti-CD3 antibody or different concentrations of the cognate peptide, insulin B:9-23, in the presence of K562 cells expressing HLA-DQ8. IL-2 production in culture supernatant was measured by ELISA (B), and frequency of ZsGreen-1-positive 5KC T-hybridoma cells was measured by flow cytometry using the gating strategy shown in Supplementary Figure S1 (C). Mean values ± standard error of the mean (SEM) from experiments performed in duplicate are shown. (D) Representative flow cytometry contour plots showing percentage of ZsGreen-1-positive cells containing either the miniCMV promoter or the TATA-box. Cells without an NFAT reporter construct are included to show the baseline fluorescent intensity that was used to establish gate placement for ZsGreen-1 positive cells. (E) Correlation between ZsGreen-1 positivity and IL-2 production for both cell lines. All results are representative of three independent experiments.

FIGURE 2

Comparison of number of NFAT binding site repeats. (A) Schematic showing the three NFAT-reporter constructs used in the experiments, containing either four, six, or eight copies of the NFAT binding site. (B–D) GSE.6H9 TCR 5KC T-hybridoma cells transduced with an NFAT-reporter construct containing four, six, or eight copies of the NFAT binding site were cultured with anti-CD3 antibody or different concentrations of the cognate peptide, insulin B:9-23, in the presence of K562 cells expressing HLA-DQ8. IL-2 production in culture supernatant was measured by ELISA (B), and frequency of ZsGreen-1-positive 5KC T-hybridoma cells was measured by flow cytometry using the gating strategy shown in Supplementary Figure S1 (C). Mean values ± SEM from experiments performed in duplicate are shown. *p-values calculated by unpaired t-test were < 0.01. (D) Representative flow cytometry contour plots showing percentage of ZsGreen-1-positive cells containing either the 4x NFAT, 6x NFAT, or 8x NFAT construct. Unstimulated cells containing each construct are included to show the baseline fluorescent intensity that was used to establish gate placement for ZsGreen-1 positive cells. All results are representative of three independent experiments.

FIGURE 3

Comparison of reporter fluorochromes. (A) Schematic showing the constructs that contained one of four potential fluorescent reporter genes. All constructs contain eight repeats of the NFAT binding site followed by a TATA-box. (B–D) 1E6 TCR 5KC T-hybridoma cells transduced with an NFAT-reporter construct containing each fluorescent reporter protein gene were cultured with or without anti-CD3 antibody overnight in the presence of K562 cells expressing HLA-A2. Values in panels (B) and (C) represent mean ± SEM from experiments performed in duplicate. (B) IL-2 production in culture supernatant was measured by ELISA. (C) Cells harvested from each culture condition were assessed on flow cytometer with 405, 488, and 633 nm laser beams. Fluorescent intensity of 5KC T-hybridomas, selected by the gating strategy shown in Supplementary Figure S1, was determined. Fold change in mean fluorescent intensity of 5KC T-hybridomas transduced with an NFAT-reporter construct containing each fluorochrome over 5KC T-hybridomas without an NFAT-reporter construct is shown. (D) Representative flow cytometry contour plots showing proportions of stimulated and unstimulated fluorescent cells, each expressing a different fluorescent reporter protein. Cells without reporter construct are included to show baseline fluorescent intensity that was used to establish gate placement of fluorescence-positive cells. All results are representative of three independent experiments.

FIGURE 4

Establishment of assay conditions. (A) and (B) 5KC T-hybridoma cells expressing an insulin B:9-23-responsive CD4 TCR GSE.6H9 (A) or a preproinsulin 15-24-responsive CD8 TCR 1E6 (B) were unstimulated or were stimulated with anti-CD3 antibody or the cognate peptide in the presence of K562 APCs expressing HLA-DQ8 or HLA-A2, respectively. Percentage of ZsGreen-1-positive cells, determined using the gating strategy shown in Supplementary Figure S1, was assessed via flow cytometry at 4, 8, 18, 32, and 42 h post-stimulation. Mean values ± SEM from experiments performed in duplicate are shown. (C–H) Different numbers (2 × 10⁴, 1 × 10⁵, or 2 × 10⁵ cells/well) of 5KC T-hybridoma cells expressing the GSE.6H9 TCR (C,E,G) or 1E6 TCR (D,F,H) were unstimulated or were stimulated with anti-CD3 antibody or the cognate peptide in the presence of K562 APCs (5 × 10⁴ cells/well) expressing HLA-DQ8 or HLA-A2, respectively. IL-2 production was measured by ELISA (C,D) and frequency of ZsGreen-1 positive 5KC T-hybridoma cells was assessed by flow cytometry using the gating strategy shown in Supplementary Figure S1 (E,F). Values in panels (C–F) represent mean ± SEM from experiments performed in duplicate. Panels G and H show correlation between ZsGreen-1 positivity and IL-2 production for each concentration of T-hybridoma cells expressing GSE.6H9 TCR (G) and 1E6 TCR (H). (I) 5KC T-hybridoma cells expressing the GSE.6H9 TCR or the 1E6 TCR with or without additional CD3 genes were stained with anti-mouse CD3, human CD4, human CD8, or mouse CD28 antibodies and assessed by flow cytometry. (J–M) 5KC T-hybridomas expressing the GSE.6H9 TCR (J,L) or the 1E6 TCR (K,M) were further transduced with additional CD3 complex genes and stimulated with a stimulation cocktail containing PMA and ionomycin or anti-CD3 antibody for 6–8 h (J,K), or cognate peptides or anti-CD3 antibody in the presence or absence of anti-CD28 antibody overnight (L,M). Frequency of ZsGreen-1 positive T-hybridoma cells was assessed by flow cytometry. All results are representative of three independent experiments.

FIGURE 5

Addition of fluorochrome identifiers and multiplexing. (A) ZsGreen-1 positivity of unstimulated cells expressing TCRs with different fluorochrome combinations. Each 5KC T-hybridoma cell line expresses a single unique TCRs along with one of eight different fluorochrome combinations. 5KC T-hybridoma cells were cultured overnight in the presence of autologous B cells transformed with Epstein-Barr virus. Frequency of ZsGreen-1-positive 5KC T-hybridoma cells expressing eight different fluorochrome combinations was determined by flow cytometry using the gating strategy shown in Supplementary Figure S2 and were compared using one-way ANOVA and Tukey’s Honestly Significant Difference (Tukey’s HSD). Numbers of TCRs that are included for each fluorochrome combination are shown in parenthesis. T-hybridoma cells expressing ametrine and mCherry (AM+mChe) have significantly higher ZsGreen-1 positivity compared to T-hybridomas expressing all other fluorochrome combinations except LO+CR (Turkey’s p = 0.001 in AM+mChe vs. AM+BFP, AM+CR, AM+TM, and Turkey’s p < 0.001 in AM+mChe vs. LO+BFP, LO+CR, LO+TM). (B) 1E6 TCR T-hybridoma cells expressing eight different fluorochrome combinations were unstimulated or were stimulated with the cognate preproinsulin 15-24 peptide at 0.1, 1, 10, and 100 μg/ml in the presence of K562 APCs expressing HLA-A2. Frequency of ZsGreen-1-positive 5KC T-hybridoma cells were determined by flow cytometry using the gating strategy shown in Supplementary Figure S2. Stimulation indexes were calculated by dividing ZsGreen-1 positivity of stimulated cells by mean values of ZsGreen-1 positivity of unstimulated cells. (C) 1E6 TCR T-hybridoma cells expressing each fluorochrome combination were co-cultured with the cognate peptide and HLA-A2 K562 cells in the presence (multiplexed, black inverse triangles) or absence (single, white inverse triangles) of T-hybridomas expressing irrelevant TCRs expressing the other seven fluorochrome combinations. Stimulation indexes of 1E6 TCR T-hybridomas in response to the peptide stimulation were calculated as described above based on frequency of ZsGreen-1-positive 1E6 TCR T-hybridoma cells determined by flow cytometry using a gating strategy shown in Supplementary Figure S2. (D) 1E6 TCR T-hybridoma cells expressing each fluorochrome combination were co-cultured with the cognate peptide and HLA-A2 K562 cells in the presence of T-hybridomas expressing irrelevant TCRs and the other seven fluorochrome combinations. Frequencies of ZsGreen-1-positive cells for T-hybridomas expressing each fluorochrome combination in each culture condition were determined by flow cytometry, and stimulation indexes were calculated as described above. (B–D) Values represent mean ± SEM from three independent experiments performed in duplicate.

FIGURE 6

Multiplex screening for a large panel of peptide libraries. (A) Schematic showing the breakdown of a protein of interest into pools of 8- to 11-mers such that each potential epitope of the peptide is represented in the peptide panel. Pools contain equimolar amounts of each truncated peptide. (B–E) 1E6 TCR T-hybridoma cells expressing LSSmOrange and tdTomato (B–D) or ametrine and E2-Crimson (E) were co-cultured with 103 truncated peptide pools generated from preproinsulin and HLA-A2 K562 cells in the absence [(B,C), single test] or presence [(D,E), multiplex test] of T-hybridomas expressing irrelevant TCRs with the other seven fluorochrome combinations. Stimulation indexes of 1E6 TCR T-hybridomas in response to the peptide stimulation were calculated as described in Figure 5 based on frequency of ZsGreen-1-positive 1E6 TCR T-hybridoma cells determined by flow cytometry using a gating strategy shown in Supplementary Figure S2. Results from two independent experiments testing a single T-hybridoma cell line individually are shown in (B,C). In all cell culture conditions testing 1E6 TCR T-hybridoma cell lines expressing different fluorescence identifiers, T-hybridomas responded to peptide pools 15–18 with a peak at 17 and 18. These pools contain the cognate peptide PPI:15-23 and PPI:15-24.

FIGURE 7

Identification of preferable amino acid residues in epitopes using a combinatorial peptide library (CPL). The CPL consists of 200 decamer peptide pools, each of which contains peptides with a specific amino acid residue at one position and randomized amino acids at the other nine positions (Supplementary Table S4). 1E6 TCR T-hybridoma cells expressing LSSmOrange and tdTomato (A–C) or ametrine and E2-Crimson (D) were co-cultured with each CPL peptide pool along with K562 APCs lentivirally expressing HLA-A2 in the absence [(A,B), single test] or presence [(C,D), multiplex test] of T-hybridomas expressing irrelevant TCRs and the other seven fluorochrome combinations. Stimulation indexes of 1E6 TCR T-hybridomas in response to the peptide stimulation were calculated as described in Figure 5 based on frequency of ZsGreen-1-positive 1E6 TCR T-hybridoma cells as determined by flow cytometry using the gating strategy shown in Supplementary Figure S2. Log2 fold changes of stimulation indexes in each culture condition were depicted as heat maps. Results from two independent experiments testing a single T-hybridoma cell line individually are shown in (A,B). Blue boxes represent the amino acids of the nominal epitope ALWGPDPAAA at each position.