Background
Cancer immunotherapy is becoming a cornerstone in the clinical care of cancer patients due to the breakthrough of immune checkpoint blockade antibodies [
1]. Adoptive transfer of ex vivo-engineered autologous T-cells has also been successful, especially when T-cells are engineered with a chimeric antigen receptor (CAR) against CD19 and used for treatment of B cell malignancies [
2‐
5]. A CAR molecule has an extracellular single chain antibody fragment for tumor antigen recognition and intracellular signaling domains from the T-cell receptor (TCR) CD3z chain and co-stimulatory molecules. CAR T-cells provide the advantage of major histocompatibility complex (MHC)-independent binding to tumor-associated antigens on the surface of tumor cells. However, the requirement of surface expression of the antigen limits the available targets making it difficult to identify new and specific CAR targets. Additionally, CAR T-cells do not require very high antigen expression to execute their effector function [
6‐
8], which may be a drawback when targeting antigens that are over-expressed but not restricted to tumor cells.
T-cells engineered with a novel TCR can target antigenic peptides presented by MHC molecules on the surface of tumor cells. This means that also intracellular mutated neoepitopes can be recognized and targeted. As the T-cell naturally uses the TCR for binding, the affinity of the interaction of the introduced TCR would be in the physiological range in which the TCR normally binds. That would at least partly avoid toxicity to normal cells with low expression level of the targeted antigen. As tumor-associated antigens are processed a variety of tumor-associated peptides is available for targeting with TCR-engineered T-cells. TCR-engineered T-cells are now starting to show progress for solid tumors [
2]. With novel techniques involving exon and RNA sequencing for identification of mutated neoepitopes, TCR T-cell therapy can be developed on an individual basis [
9].
The apparent strength of the interaction between the TCR and MHC/peptide complex is dependent on the affinity, avidity and functional avidity of the TCR for its target. Measuring the dissociation rate or the half-life of a single TCR-MHC/peptide interaction reflects the residence time of the TCR/MHC interaction. Such biochemical measurement is technically challenging and often does not reflect the kinetics of multiple receptor/ligand interactions (avidity), which are needed for the activation of the T-cell. Higher avidity T-cells are found to be more rapidly activated and exert a better cytotoxic function [
10], though that is still debated, reviewed in [
11]. However, high avidity T-cells may have a lower sensitivity to their ligand. Treating cancer patients with adoptive T-cell therapy that result in targeting normal tissue and severe side effects. In addition, over-stimulated T-cells are short-lived and prone to apoptosis upon antigen engagement. Because TCR-MHC/peptide binding is strengthened by co-receptor binding to MHC, co-receptor expression may affect the subsequent activation of the T-cells. To measure how well T-cells respond to antigen-specific stimulation, the term functional avidity is used. Functional avidity is directly related to the ability of T-cells to adhere to and kill cells expressing the target antigen.
Many reports indicate that virus-specific T-cells are of high affinity (often in the nM range), in contrast to T-cells directed towards the body’s own proteins that are of low affinity due to deletion of high affinity clones in the thymus to avoid targeting self-tissue and autoimmunity [
12]. We have cloned two TCRs, one TCR recognizes an HLA-A2-restricted peptides from the prostate and breast cancer-associated antigen TARP [
13] and in this paper one TCR recognizing an HLA-A2-restricted peptide from the cytomegalovirus (CMV) pp65 antigen. These TCRs are compared side-by-side. As TARP is an antigen highly expressed in prostate cancer cells it has a potential to be used in the clinics. Therefore, characterization of its affinity, avidity and function as compared to a virus-directed TCR would provide basic information for its effectiveness and safety.
Methods
Cells and cell lines
Target cells: Mel526 cells (HLA-A2+) was cultured in Dulbecco’s Modified Eagle Medium (DMEM), supplemented with 10 % fetal bovine serum (FBS), 1 % penicillin/streptomycin (PEST) and 1 mM sodium pyruvate. T2 cells (HLA-A2+) were grown in RPMI-1640 containing 10 % FBS and 1 % PEST. Producer cells: HEK-293 T-cells was cultured in DMEM, 10 % FBS, 1 % PEST and 500 μg/ml geneticin (used only during culturing but not during virus production). Peripheral blood mononuclear cells (PBMCs) from buffy coat blood of healthy volunteers (unidentified; bought from the Uppsala University Hospital blood center) were isolated using Ficoll Paque (GE Healthcare, Uppsala, Sweden). PBMCs were cultured in RPMI-1640 supplemented with 10 % human AB-serum (own production), 100 IU/ml interleukin (IL)-2 (Proleukin, Novartis, Basel, Switzerland), 2 mM L-glutamine, and 10 mM HEPES. The cell culture reagents were purchased from Life Technologies (Carlsbad, CA).
Peptides, peptide pulsing and multimers
The HLA-A2-restricted peptides: TARP(P5L)
4–13 (amino acid sequence FLPSPLFFFL) [
14], CMV pp65
495–503 (NLVPMVATV) [
15] and as negative control VMAT-1
31–39 (LLLDNMLFT) were synthesized to purity above 95 % (Genscript, Piscataway Township, NJ). Target cells were pulsed for 2 h with peptide at concentrations of 50 μM, 10 μM, 1 μM, 100 nM, 10 nM, 1 nM, 100 pM and 10 pM (only the highest concentration was used for the VMAT-1
31–39 control peptides). Phycoerythrin (PE)-conjugated TARP(P5L)
4–13/HLA-A*0201 dextramer was purchased from Immudex (Copenhagen, Denmark) and PE-conjugated pp65
495–503/HLA-A*0201 tetramer was purchased from MBL International (Woburn, MA).
TCR cloning and viral vector construction
We have previously reported cloning of the TCR recognizing the TARP
4–13/HLA-A2 complex. [
13] The TCR recognizing the CMV pp65
495–503/HLA-A2 complex was cloned using the same approach. In brief, PBMCs, isolated from a HLA-A2 positive, CMV-seropositive donor, were stained with the pp65
495–503 tetramer and tetramer-positive T-cells were isolated using magnetic beads. T-cells were cloned at 0.6 cells/well in 96-well plates and expanded with IL-2 and irradiated feeder cells before testing for peptide reactivity. The TCR chains of ten reactive clones were sequenced and found to contain one unique TCR-α chain and one unique TCR-β chain. A recombinant sequence with the identified TCR α chain linked to the β chain was cloned into a lentiviral vector under the
Spleen Focus-Forming Virus (SFFV) promoter. The α and β chains were separated by a 2A self-cleaving peptide sequence from
Thosea Asigna Virus (T2A). Mouse constant domains of TCR α and β were used to improve the pairing between the chains of the introduced TCR chains and avoid mispairing with endogenous TCR α and β chains. Vesicular stomatitis virus (VSV)-G pseudotyped lentiviral particles were produced in HEK 293-T-cells and concentrated by ultracentrifugation as described previously [
13].
T-cell activation, transduction and sorting of TCR-transduced T-cells
T-cells in a pool of freshly isolated PBMCs were activated for 48 h using 100 ng/ml OKT3 antibody (Nordic Biosite, Täby, Sweden) and 100 IU/ml IL-2. One million activated PBMCs were then transduced for 4 h with 50 μl concentrated lentivirus, encoding the pp65-TCR or TARP-TCR as described previously [
13]. After transduction the cells were plated in 24-well plates, rested overnight and re-transduced 24 h later. The transduced cells were tested for transduction efficiency using multimers and flow cytometry analysis 7 days after transduction.
To purify TCR-engineered T-cells, the transduced cells were stained with PE-conjugated pp65
495–503/HLA-A*0201 tetramer or PE-conjugated TARP(P5L)
4–13/HLA-A*0201 dextramer for 30 min at 4 °C. Anti-PE magnetic beads (Miltenyi Biotec, Bergisch Gladbach, Germany) were then used to separate the PE-labeled T-cells according to manufacturer’s instructions. The purity was estimated by flow cytometry (FACSCanto II BD Biosciences, Franklin Lakes, NJ) using PE-conjugated tetramer/dextramer and antibodies (Biolegend, San Diego, CA) against the following markers: CD3 conjugated with allophycocyanin (APC) or Pacific Blue, CD8 conjugated with fluorescein isothiocyanate (FITC), CD4 conjugated with APC. The results were analyzed using FACS Diva 8 and Flow Jo software (Ashland, OR). The sorted TCR-engineered T-cells were then expanded using a rapid expansion protocol as described earlier [
13]. The expanded T-cells then reassessed by flow cytometry and were in all cases found to be > 90 % multimer positive.
Ligand Tracer® measurement of T-cell binding to target cells
One million mel526 target cells in 2 ml of culture medium were let to adhere overnight to a tilted 10-cm Petri dish. The target cells were then pulsed with peptides as described above. The Petri dish was then inserted on the tilted rotating platform of the Ligand Tracer® instrument (Ridgeway Instruments AB, Uppsala, Sweden) and background measurement of fluorescence was done in real time during rotation (1 rpm) for 30 min.
Transduced and expanded TCR-engineered T-cells were labeled with Carboxyfluorescein succinimidyl ester (CFSE) according to manufacturer’s instructions (Thermo Fisher, Uppsala, Sweden) and then washed thoroughly with serum-containing medium. CFSE-labeled TCR-engineered T-cells (1.5 × 105 cells) were then added to the Petri dish with peptide-pulsed target cells. Rotation started again and T-cell binding (association) to the target cells was measured in real time through detection of fluorescent signal from the target cells (T-cell binding) with subtraction of the fluorescent signal from the opposite side of the Petri dish without target cells. After 90 min another 3 × 105 T-cells were added and the measurement continued.
ELISA and killing assays
For specific TCR activation experiments, the transduced T-cells were co-cultured with target cells pulsed with relevant peptide and control peptide as described above. To detect IFN-γ secretion, 1 × 105 peptide-pulsed T2 target cells were co-cultured overnight with TCR-engineered T-cells (effector to target ratio 1:1). Supernatants were collected and IFN-γ was measured using ELISA (Mabtech, Nacka Strand, Sweden). For killing assays, 1 × 105 peptide-pulsed luciferase-expressing mel526 target cells were co-cultured with TCR-engineered T-cells at 1:1 ratio (effector to target ratio 1:1) or target cells were pulsed with 10 μM peptide and mixed with effector T-cells at different effector to target ratios (10:1, 5:1, 1:1, 1:5 and 1:10). The results in terms of luciferase activity were analyzed after 72 h using Steady-Glo Luciferase Assay System (Promega, Madison, WI) according to manufacturer’s instructions. For the xCELLigence cell growth assay 2 × 104 mel526 target cells were plated in an E-Plate View (ACEA Biosciences, San Diego, CA). Cell growth was measured overnight using the xCELLigence RCPA DP Instrument (ACEA Biosciences). The next day, the target cells on the E-plate View were pulsed for 2 h with TARP(P5L)4–13 or pp65495–503 peptide at 10 μM, 100nM or 1nM. The effector T-cells were then added to the target cells (approximately 26 h after plating of the target cells) in 1:1, 5:1 and 10:1 effector to target ratios and the killing of the target cells was continuously measured for an additional 48 h.
Discussion
We have previously identified a novel TCR against an HLA-A2-restricted peptide from the prostate and breast cancer-associated antigen TARP [
13] and in this study, we have further characterized its interaction and binding to HLA-A2
+ targets cells. We also report the cloning of a TCR against an HLA-A2-restricted peptide from CMV pp65 and demonstrate its strong ability to recognize and kill pp65-pulsed HLA-A2
+ target cells. The purpose of this study was to investigate the difference in target recognition between T-cells with a virus-directed TCR (pp65-specific) and tumor self antigen-directed TCR (TARP-specific) using both established and novel analyses methods.
It is known that high and low affinity TCRs needs different peptide concentrations to exert cytotoxicity and we predicted that to be the case for the pp65-specific and TARP-specific TCRs. From our functional in vitro studies (IFN-γ release and cytotoxicity) it is clear that T-cells with the pp65-specific TCR respond to a lower peptide concentration than T-cells with the TARP-specific TCR. We verified that pp65-TCR T-cells were also able to kill target cells at low peptide concentration (100 pM) as previously described for T-cells specific to a different antigen [
25,
26]. The bioinformatics predictive value of the binding strength of the peptide to HLA-A2, as estimated by NetMHC was comparable for the TARP(P5L)
4–13 or pp65
495–503 peptides. Both were categorized as strong binders. Therefore, we believe that a direct comparison between the two TCRs is valid.
When using multimers to assess the TCRs we found that the pp65-TCR resulted in a higher transduction rate (%-wise) of T-cells and that the T-cells had higher MFI than the TARP-TCR. There were slightly more CD4+ T-cells than CD8+ T-cells that were multimer positive for both TCRs. The reason for this is unknown but it may indicate a slight preference of transducing CD4+ T-cells over CD8+ T-cells, that CD4+ T-cells have better viability after transduction with lentiviral vectors or simply this occurs because there are more CD4+ T-cells than CD8+ T-cells in peripheral blood. CD8+ T-cells had a higher MFI than CD4+ T-cells. This is expected as the transferred TCRs interact with HLA-A2 (MHC class I) for which CD8 is the co-receptor. The fact that CD4+ T-cells could also be detected by multimers indicates that both TCRs had a rather high affinity since they could be detected without the CD8 co-receptor. In the therapeutic setting of adoptive T-cell transfer it may be advantageous to have T-cells that have a good binding for the target antigen but do not get activated too rapidly, as that may lead to their exhaustion.
When using IFN-γ release and in vitro killing assays pp65-TCR T-cells were superior to TARP-TCR T-cells. This is probably due to higher affinity. TARP is a self-antigen and the TARP-TCR was cloned from a healthy HLA-A2 positive donor [
13]. Although TARP has restricted expression in differentiated cells it is possible that the TARP
4–13 epitope has been presented by thymic antigen-presenting cells during T-cell development and the high affinity T-cell clones against it were deleted during negative selection. Most T-cells recognizing self-antigens (including tumor-associated self antigens) are of lower affinity and avidity or anergic. A recent study, addressing the responses of several TCRs directed against tumor-associated antigens, concludes that T2 cells pulsed with nanomolar concentrations of peptide generally corresponds to the number of epitopes that are physiologically presented by tumor cells [
27]. For the TCRs tested in that study, partial response was detected at 10 nM peptide for all antigens tested in terms of ELISPOT IFN-γ production [
27]. Although it is difficult to directly compare our results with theirs, small amounts of IFN-γ were detected in our study at 10 nM TARP peptide by ELISA. Of note, the amounts detected exceeded the limit for reactive clones that is used in adoptive T-cell transfer of expanded tumor-infiltrating lymphocytes to cancer patients of some investigators [
28]. One also need to keep in mind that T-cells with a high affinity TCR have caused severe side effects in experimental treatment of cancer with adoptively transferred T-cells [
29]. One suggested mechanism is through TCR signaling enhancement through Tim-3 [
30].
In this study, we introduce LigandTracer technology as a method to monitor the binding of T-cells to target cells in real time. It is a convenient method to demonstrate the increase in the binding as the signal is proportional to the number of the cells and increases in a linear fashion. TCR binding to the cognate peptide was demonstrated both for the TARP-TCR and pp65-TCR but peptide concentration on target cells needs to be adjusted for individual TCRs for optimal result. It is therefore difficult to directly compare the pp65-specific T-cells and the TARP-specific T-cells. However, LigandTracer data suggest that the better in vitro performance of the pp65-TCR T-cells (IFN-γ release and cytotoxicity) cannot be explained solely with better binding to target cells. To complement the binding data from the LigandTracer experiments, we looked at target cell inactivation in real time using the xCELLigence cell growth assay technology to assess the killing potency of the TCR-engineered T-cells. The TARP-TCR T-cells were less efficient in killing their targets and required both higher peptide concentrations and higher effector to target ratios to achieve efficient response. The killing kinetics differed between the TARP-TCR T-cells and pp65-TCR T-cells, and it may indicate that there are different mechanisms that are employed if the T-cell-mediated killing is slower. The TARP-TCR T-cells were able to kill target cells, which growth plateaued at Cell Index of approximately 1, whereas pp65-TCR T-cells had fast and efficient killing shown by the Cell Index going back to 0. For TARP-TCR T-cells, at all peptide concentrations, and for pp65-TCR T-cells, at a low peptide concentration, targets cells started to grow back again, but were then killed in something that looks like a second wave of T-cell killing. In the case of pp65-TCR T-cells, after the re-growth of target cells, T-cells took control again and the Cell Index dropped to 0. Taken together the data indicate a more rapid and efficient response of pp65-TCR T-cells, and that dependent on the TCR the T-cells may need more time to kill their target cells. It is also possible that because of different binding avidity the target cells could undergo apoptosis at a different speed. Because of the real time fine resolution of the xCELLigence assay (Fig.
5c) and possibility to long term follow up (Fig.
5a, b) we suggest that this assay can efficiently replace the standard
51Cr-release assay for T-cell mediated killing.
Abbreviations
CAR, chimeric antigen receptor; CD, cluster of differentiation; CFSE, carboxyfluorescein succinimidyl ester; CMV, cytomegalovirus; DMEM, Dulbecco’s Modified Eagle Medium; ELISA, enzyme-linked immunosorbent assay; ELISPOT, enzyme-linked immunosorbent spot-forming cell assay; FACS, fluorescence-activated cell sorting; HLA, human leukocyte antigen; IFN, interferon; IL, interleukin; MHC, major histocompatibility complex; PBMC, peripheral blood mononuclear cells; PE, phycoerythrin; TARP, T-cell receptor gamma alternate reading frame protein; TCR, T-cell receptor; VSV-G, Vesicular stomatitis virus glycoprotein
Acknowledgements
The authors want to thank Berith Nilsson for help with virus preparation and Sina Bondza for help with LigandTracer data analysis. The research was funded by the Swedish Cancer Society, Erik, Karin o Gösta Selanders foundation and the Swedish Research Council.
Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (
http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (
http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.