Herein we describe a novel method to generate antigen-specific T cell receptors (TCRs) by pairing the TCRα or TCRβ of an existing TCR, possessing the antigen-specificity of interest, with complementary hemichain of the peripheral T cell receptor repertoire. The de novo generated TCRs retain antigen-specificity with varying affinity.
T cell receptors (TCRs) are used clinically to direct the specificity of T cells to target tumors as a promising modality of immunotherapy. Therefore, cloning TCRs specific for various tumor-associated antigens has been the goal of many studies. To elicit an effective T cell response, the TCR must recognize the target antigen with optimal affinity. However, cloning such TCRs has been a challenge and many available TCRs possess sub-optimal affinity for the cognate antigen. In this protocol, we describe a method of cloning de novo high affinity antigen-specific TCRs using existing TCRs by exploiting hemichain centricity. It is known that for some TCRs, each TCRα or TCRβ hemichain do not contribute equally to antigen recognition, and the dominant hemichain is referred to as the centric hemichain. We have shown that by pairing the centric hemichain with counter-chains differing from the original counter-chain, we are able to maintain the antigen specificity, while modulating its interaction strength for the cognate antigen. Thus, the therapeutic potential of a given TCR can be improved by optimizing the pairing between the centric and counter hemichains.
T cell receptors (TCRs) are heterodimeric adaptive immune receptors expressed by T lymphocytes, composed of a TCRα and TCRβ chain. They are generated via somatic rearrangement of V(D)J gene segments, which produces a highly diverse repertoire capable of recognizing virtually unlimited configurations of HLA/peptide complexes. Clinically, T cells engineered to express clonotypic TCRs specific for tumor-associated antigens have demonstrated efficacy in a variety of cancers1. However, many TCRs cloned for this purpose lack sufficient affinity for the antigen of interest, which limit their therapeutic application.
Here, we describe a method to overcome this limitation for existing TCRs by exploiting chain-centricity. It has been reported that one TCR hemichain could play a more dominant role in recognition of the target antigen2, here termed centricity. Crystal structural analyses have shown that one centric hemichain of a TCR could account for the majority of the footprint on the MHC/peptide complex3,4. Using this concept, we have previously demonstrated that the SIG35α TCRα can pair with a diverse repertoire of TCRβ chains and maintain reactivity against the MART127-35 peptide presented by HLA-A25. Similar results were obtained with the TAK1 TCR, where the centric TCRβ hemichain paired with various TCRα chains and maintained reactivity for the WT1235-243 peptide presented by HLA-A246. Both MART1 and WT1 are tumor-associated antigens. Chain-centricity was also applied to study antigen recognition of CD1d-restricted invariant natural killer (iNKT) TCRs, by pairing the invariant Vα24-Jα18 (Vα24i) TCRα chain of human iNKT TCRs with different Vβ11 TCRβ chains7.
In all cases, we were able to generate a de novo repertoire of TCRs by transducing the centric TCR hemichain to peripheral blood T cells, where the introduced hemichain paired with the endogenous TCRα or TCRβ counter-chains. In essence, the centric hemichain serves as a bait that can be used to identify the appropriate counter-chains, which when paired together form TCRs that maintain the antigen specificity of interest, yet varying in affinity. From these novel repertoires, we were able to isolate clonotypic TCRs with improved interaction strength against the target antigen compared to pre-existing TCRs. Therefore, we believe this method will accelerate the pipeline of identifying optimal TCRs for clinical application.
1. Preparing Retroviral Construct Encoding TCR Hemichain of Interest
2. Generating Stable Packaging Cell Line
NOTE: Both 293GPG and PG13 cells are adherent. Culture cells in Dulbecco's Modified Eagle Medium (DMEM) supplemented with 10% fetal calf serum (FCS) and 50 μg/ml gentamicin. Culture 293GPG cells with 1 μg/ml tetracycline before transfection. Incubate cells at 37 °C with 5% CO2 between all steps.
3. Activation and Transduction of Human T cells
NOTE: Human samples are obtained and used in accordance with the institutional ethics committee approved protocols. Culture primary human cells in Roswell Park Memorial Institute (RPMI) medium supplemented with 10% human serum instead of FCS and 50 μg/ml gentamicin. Incubate cells at 37 °C with 5% CO2 between all steps.
4. Cloning De Novo TCR Counter-hemichains
5. Reconstituting Novel Antigen-specific TCR Clones
NOTE: Culture Jurkat 76 cells and subsequent cell lines in RPMI medium supplemented with 10% FCS and 50 μg/ml gentamicin. Incubate cells at 37 °C with 5% CO2 between all steps.
Without prior knowledge of which hemichain is chain-centric, the TCRα and TCRβ chain should be separately cloned and transduced to peripheral blood T cells, which was done in the case of HLA-A24/WT1 reactive TAK1 TCR (Figure 1). Transduction of TAK1β yielded a noticeably higher frequency of antigen-specific T cells. Conversely, transduction of a non-centric hemichain would not yield de novo multimer positive T cells, as seen with TAK1α chain (Figure 1). During analysis, gate on NGFR+ cells if the TCR gene was tagged to specifically analyze transduced cells (Figure 1–2). On the other hand, the single centric hemichain can be transduced if it is known to be dominant, for example the SIG35α and invariant Vα24 TCRα chains (Figure 2 and 3). T cells specific for the respective cognate antigen increased in frequency upon transduction of a centric TCRα hemichain. (Figure 2–3). Together, these data demonstrate that introduction of a centric TCRα or TCRβ hemichain to peripheral blood T cells can generate de novo TCRs with the antigen-specificity of interest.
Antigen-specific population can be sorted by flow-assisted cell sorting and TCR hemichains can be cloned as described in the Protocol section 4. Clonotypic TCR counter-chains can be individually reconstituted on a T cell line stably expressing the centric hemichain. Jurkat 76 transfectants, expressing newly cloned unique TCRs from TAK1β or SIG35α centric hemichain transduced T cells, possessed varying avidity for the respective cognate antigen, as determined by HLA/peptide multimer staining5,6. Specifically, we were able to isolate de novo TCRα counter-chains that when paired with TAK1β were stained with either lower or higher intensity by the WT1 antigen complex than the parental TAK1αβ pairing (Figure 4). Similarly, SIG35α paired with unique counter-chains recognized HLA-A2/MART1 with moderate to high avidity (Figure 5). SIG35α was cloned independent of a TCRαβ heterodimer5, therefore a parental pairing for this hemichain was unavailable for comparison. Importantly, these data indicate that it is possible to modulate the affinity for the target antigen by pairing the centric hemichain with various counter-chains.
Figure 1: TAK1β as an example of HLA-restricted TCRβ centric hemichain. TCRα or TCRβ hemichains of the HLA-A24/WT1 reactive TAK1 TCR was tagged with truncated nerve growth factor receptor (ΔNGFR), and individually transduced to peripheral blood T cells. Transfectants were stained with PC5-conjugated anti-CD8 (133x diluted) and FITC-conjugated anti-NGFR (20x diluted) monoclonal antibodies (mAbs), and indicated HLA-A24 multimer (5 μg/ml), following two times antigen-specific stimulation. Total of 6 donors were tested. All data were gated on NGFR+ cells. Figure re-printed with permission from reference6. Please click here to view a larger version of this figure.
Figure 2: SIG35α as an example of HLA-restricted TCRα centric hemichain. SIG35α TCRα chain tagged with ΔNGFR, or the tag alone, was transduced to peripheral blood T cells. Transfectants were stained with PC5-conjugated anti-CD8 (133x diluted) and FITC-conjugated anti-NGFR (20x diluted) mAbs, and indicated HLA-A2 multimer (8 μg/ml). Total of 4 donors were tested. All data were gated on NGFR+ cells. Figure re-printed with permission from reference5 (Copyright 2015. The American Association of Immunologists, Inc.). Please click here to view a larger version of this figure.
Figure 3: Invariant natural killer T (iNKT) cell receptor Vα24 (Vα24i) as an example of CD1d-restricted TCRα centric hemichain. Vα24i TCRα chain was transduced to peripheral blood T cells. Transfectants were stained with FITC-conjugated anti-CD3 (20x diluted) mAb and indicated CD1d multimer (5 μg/ml). Unloaded CD1d molecules were produced from HEK293 cells and present endogenous unknown self-lipids. iNKT TCR is defined by recognition of CD1d presenting α-GalCer or its analog PBS-57. Total of 4 donors were tested. Figure re-printed with permission from reference7. Please click here to view a larger version of this figure.
Figure 4: Representative TCRα counter-chains cloned from TAK1β transduced T cells. TCRα clones T262, A262, A186, A133, and T4 were cloned from HLA-A24/WT1 multimer positive TAK1β transduced peripheral T cells, and individually reconstituted on Jurkat 76 cells expressing CD8, along with the TAK1β chain. Transfectants were stained with PC5-conjugated anti-CD8 mAb (133x diluted) and indicated HLA-A24 multimer (5 μg/ml). Data are representative two independent experiments. Error bars indicate range. Figure re-printed with permission from reference6. Please click here to view a larger version of this figure.
Figure 5: Representative TCRβ counter-chains cloned from SIG35α transduced T cells. TCRβ clones 413, 523, 788, 1086, 794, and 830 were cloned from HLA-A2/MART1 multimer positive SIG35α transduced peripheral T cells, and individually reconstituted on Jurkat 76 cells expressing CD8, along with the SIG35α chain. DMF5 was a previously cloned high affinity TCR recognizing HLA-A2/MART1. Transfectants were stained with PC5-conjugated anti-CD8 mAb (133x diluted) and indicated HLA-A2 multimer (2 μg/ml). Data are representative of two independent experiments. Error bars indicate range. Figure re-printed with permission from reference5 (Copyright 2015. The American Association of Immunologists, Inc.). Please click here to view a larger version of this figure.
Table 1: Setup for pMX vector digestion. Reagents and respective volumes are shown. Please click here to view a larger version of this table.
Table 2: Setup for 1st round 5' RACE PCR. Reagents and respective volumes, PCR settings, and primer sequence are shown. Please click here to view a larger version of this table.
Table 3: Setup for 2nd round 5' RACE PCR. Reagents and respective volumes, PCR settings, and primer sequences are shown. Please click here to view a larger version of this table.
The first requirement for successful application of this method is achieving sufficient transduction efficiency of primary T cells with the hemichain of interest. In our experience, the combination of using PG13 as packaging cell line and pMX as retroviral vector results in stable and efficient expression of the introduced gene in human primary T cells. PG13 packaging cells can be single-cell cloned to select for high-titer packaging cells to improve transduction efficiency. Furthermore, proliferation of T cells is also required for high transduction efficiency by retrovirus. In the described protocol, stimulation with soluble anti-CD3 mAb OKT3, along with high concentration of rhIL-2, provides the necessary activation signals to induce cell division. Monocytes are required for the stimulatory capacity of soluble OKT3, likely through binding with Fc receptors. Therefore, bulk PBMC, and not purified T cells, is required for this particular stimulation protocol. T cells should be cultured in RPMI supplemented with human serum for optimal experimental results. All other cells described here can be cultured in the presence of fetal calf serum.
Second, de novo generated antigen-specific TCRs expressing cells must be detectable. For certain TCR hemichains, such as TAK1β, de novo TCR expressing cells are only detectable after stimulation with antigen-pulsed antigen-presenting cells (APCs). Thus, if antigen-specific TCRs are not detected immediately after transduction of centric hemichain, they could be detectable after expansion by APCs. Note that transduction of a non-centric hemichain would not yield de novo antigen-specific T cells even after antigen-specific expansion with APCs, as seen with the TAK1α hemichain (Figure 1). In our studies, HLA multimers are used for detection, however, this might not be available for some TCRs, especially HLA class II restricted TCRs. An alternative would be to detect T cell functional responses. Hemichain transduced T cells can be stimulated with APCs pulsed with the antigen of interest. Vehicle-pulsed APCs and untransduced T cells can be used as controls. Responding T cells can be detected by upregulation of surface activation markers such as CD107a, CD154, or CD137.
Lastly, it is important to validate that the cloned counter-chains do in fact recognize the antigen of interest when paired with the centric hemichain. This is particularly important when the centric hemichain is TCRβ and counter-chains are TCRα, since allelic exclusion for TCRα genes is leaky and it is estimated that a significant portion of peripheral αβ T cells express more than one TCRα chain28. Reactivity can be confirmed with multimer staining or measuring functional response after stimulation with APCs presenting the cognate antigen.
One limitation of this technique is that not every TCR will compose a centric hemichain, since both TCRα and TCRβ contribute comparably to antigen recognition in some cases3. Therefore, peripheral T cells transduced with such non-centric hemichains may not yield de novo antigen-specific TCRs.
Nevertheless, this method represents a conceptual advancement upon the conventional approach to cloning TCRs, where both TCRα and TCRβ genes must be cloned and correct pairing must be ensured29,30. In our technique, pairing between hemichains is no longer a main concern and only one of the TCR chains needs to be cloned, given that the other is fixed. In addition to HLA class I or CD1d-restricted TCRs, the described method can also be applied to the isolation of HLA class II-restricted TCRs for TCR gene therapy.
The authors have nothing to disclose.
This work was supported by NIH grant R01 CA148673 (NH); the Ontario Institute for Cancer Research Clinical Investigator Award IA-039 (NH); BioCanRX Catalyst Grant (NH); The Princess Margaret Cancer Foundation (MOB, NH); Canadian Institutes of Health Research Canada Graduate Scholarship (TG); Natural Sciences and Engineering Research Council of Canada Postgraduate Scholarship (TG); Province of Ontario (TG, MA); and Guglietti Fellowship Award (TO). HLA and CD1d monomers were kindly provided by the NIH tetramer core facility.
0.05% Trypsin-EDTA | Wisent Bioproducts | 325-043-CL | |
293GPG cells | Generated by Ory et al. (ref 8) | ||
Agar | Wisent Bioproducts | 800-010-CG | |
Agarose | Wisent Bioproducts | 800-015-CG | |
Ampicillin sodium salt | Wisent Bioproducts | 400-110-IG | |
Chloroform | BioShop | CCL402 | |
Deoxynucleotide (dNTP) Solution Mix | New England Biolabs | N0447L | |
DMEM, high glucose, pyruvate | Life Technologies | 11995065 | |
EcoRI | New England Biolabs | R0101S | |
EZ-10 Spin Column DNA Gel Extraction Kit | BS353 | ||
Fetal Bovine Serum | Life Technologies | 12483020 | |
Ficoll-Paque Plus | GE Healthcare | 17-1440-02 | |
Filter | Corning | 431220 | 0.45 mm pore SFCA membrane |
FITC-conjugated anti-human CD271 (NGFR) mAb | Biolegend | 345104 | clone ME20.4 |
FITC-conjugated anti-human CD3 mAb | Biolegend | 300440 | clone UCHT1 |
Gentamicin | Life Technologies | 15750078 | |
Gibson Assembly Master Mix | New England Biolabs | E2611L | used for multi-piece DNA assembly |
HLA-A2 pentamer | Proimmune | depends on antigenic peptide | HLA-A2/MART1 multimer used here was purchased from Proimmune |
HLA/CD1d monomers | NIH Tetramer Core Facility | multimerize monomers according to protocol provided by NIH tetramer core facility | |
Human AB serum | Valley Biomedical | HP1022 | |
human CD3 microbeads | Miltenyi Biotec | 130-050-101 | |
IOTest Beta Mark TCR V beta Repertoire Kit | Beckman Coulter | IM3497 | |
Jurkat 76 cells | Generated by Heemskerk et al. (ref 10) | ||
LB Broth | Wisent Bioproducts | 800-060-LG | |
LS MACS column | Miltenyi Biotec | 130-042-401 | |
NEB 5-alpha Competent E. coli | New England Biolabs | C2987I | |
NEBuffer 3.1 | New England Biolabs | B7203S | used for EcoRI and NotI digestion |
NotI | New England Biolabs | R0189S | |
NucleoBond Xtra Midi | Macherey-Nagel | 740410 | used for plasmid purification |
PC5-conjugated anti-human CD8 mAb | Beckman Coulter | B21205 | clone B9.11 |
PG13 cells | ATCC | CRL-10686 | |
Phusion HF Buffer Pack | New England Biolabs | B0518S | |
Phusion High-Fidelity DNA Polymerase | New England Biolabs | M0530L | |
pMX retroviral vector | Cell Biolabs | RTV-010 | |
polybrene | Sigma-Aldrich | H-9268 | |
Proleukin (recombinant human interleukin-2) | Novartis | by Rx only | equivalent product can be purchased from Sigma-Aldrich |
Purified anti-human CD3 antibody | Biolegend | 317301 | clone OKT3, used for T cell stimulation |
RPMI 1640 | Life Technologies | 11875119 | |
SA-PE | Life Technologies | S866 | used for multimerizing monomers from NIH tetramer core facility |
SMARTer RACE 5'/3' Kit | Clontech | 634858 | |
Sterile water | Wisent Bioproducts | 809-115-LL | |
SuperScript III First-Strand Synthesis System | Invitrogen | 18080051 | for cDNA synthesis |
Syringe | BD | 301604 | 10 mL, slip tip |
Tetracycline hydrochloride | Sigma-Aldrich | T7660 | |
TransIT-293 | Mirus Bio | MIR 2700 | used to transfect 293GPG cells |
TRIzol Reagent | Life Technologies | 15596026 |