Adoptive transfer of Ag-specific T lymphocytes is an attractive form of immunotherapy for cancers. However, acquiring sufficient numbers of host-derived tumor-specific T lymphocytes by selection and expansion is challenging, as these cells may be rare or anergic. Using engineered T cells can overcome this difficulty. Such engineered cells can be generated using a chimeric Ag receptor based on common formats composed from Ag-recognition elements such as αβ-TCR genes with the desired specificity, or Ab variable domain fragments fused with T cell–signaling moieties. Combining these recognition elements are Abs that recognize peptide-MHC. Such TCR-like Abs mimic the fine specificity of TCRs and exhibit both the binding properties and kinetics of high-affinity Abs. In this study, we compared the functional properties of engineered T cells expressing a native low affinity αβ-TCR chains or high affinity TCR-like Ab–based CAR targeting the same specificity. We isolated high-affinity TCR-like Abs recognizing HLA-A2-WT1Db126 complexes and constructed CAR that was transduced into T cells. Comparative analysis revealed major differences in function and specificity of such CAR-T cells or native TCR toward the same antigenic complex. Whereas the native low-affinity αβ-TCR maintained potent cytotoxic activity and specificity, the high-affinity TCR-like Ab CAR exhibited reduced activity and loss of specificity. These results suggest an upper affinity threshold for TCR-based recognition to mediate effective functional outcomes of engineered T cells. The rational design of TCRs and TCR-based constructs may need to be optimized up to a given affinity threshold to achieve optimal T cell function.

Adoptive transfer of Ag-specific T lymphocytes is an attractive form of immunotherapy for hematologic malignancies and solid cancers. This approach, in which tumor reactive T cells undergo ex vivo expansion and are then infused back to patients, has proven to be effective in metastatic melanoma patients (14). However, the widespread use of this approach is limited by the need to isolate Ag-specific T lymphocytes for individual patients. To overcome this difficulty, the strategy of engineered T cells has been developed, which involves genetic modifications based on two different approaches reviewed in (5).

In the first approach, tumor targeting by T cells is achieved using a cloned αβ-TCR that is introduced into the cells and enables specific MHC-restricted targeting of tumor cells (68). This approach has been proved effective in clinical trials in melanoma (911), and several groups are currently working to improve the expression of the exogenous TCR on the surface of T cells (12). The second strategy involves redirection of T cells based on Ab variable fragments (Fv). The availability of antitumor Abs targeting a variety of tumors prompted the idea of incorporating the recognition domain of these Abs in the form of single-chain Fv (scFv) domain in a chimeric receptor construct (13). This chimeric Ab-based or Ag receptor (CAR) is based on linking the recognition elements of an Ab to signaling moieties for T cell activation, thereby redirecting T cells to a desired Ag in a non-MHC restricted manner. This “new” specificity redirects T cells independently of MHC, and it has proved to be effective against tumor cells that lost their HLA expression because of tumor escape mechanisms (14). After several studies in the last decade demonstrated the effectiveness of this approach in inducing tumor regression in mouse models, this approach is currently being evaluated in clinical trials (1517). Furthermore, this strategy has shown dramatic responses in a pilot clinical trial in patients with chronic lymphocytic leukemia treated with CD19-specific CAR T cells (18). Targeting tumor-associated Ags (TAAs) with engineered T cells that express a specific CAR is an ideal approach combining the unique expression pattern of TAAs with potent killing mediated by the engineered effector T cell (19).

Wilms tumor suppressor gene 1 (WT1) is one of the most important TAAs classified by the National Cancer Institute (20). WT1, a zinc finger transcription factor, is highly expressed in many solid cancers and leukemia cells (21, 22), but not in normal tissues (including hematopoietic progenitor and stem cells). Several studies have suggested that WT1 has an essential role in leukemogenesis and tumorigenesis, and it is required to maintain the transformed phenotype/function; therefore, tumor escape from immune surveillance as a result of downregulation of WT1 expression is unlikely to occur (22, 23), marking WT1 as an attractive and important target for immunotherapy.

The WT1Db126 (RMFPNAPYL) peptide was identified by screening the WT1 aa sequence for 9-mer peptides that include major anchor motifs for binding to HLA-A2 (24). In vitro immunization elicits WT1 peptide-specific CTLs that mediate lysis of WT1-expressing tumor cells, indicating that this peptide constitutes a highly immunogenic epitope. The αβ genes of a TCR that recognizes HLA-A2-WT1Db126 complexes were isolated using an allogeneic repertoire, and they were introduced with retroviral transduction into human T cells (7). This TCR exhibited efficient and specific reactivity with HLA-A2-WT1Db126 complexes, enabling specific cytolytic reactivity by CTLs expressing the TCR toward target cells (25).

Specific peptide-MHCs (pMHCs) can also be targeted by generating recombinant Abs mimicking TCR specificity, which are termed “TCR-like Abs” (26). We and others have isolated Abs that recognize HLA-A2 complexes bearing peptides derived from tumor and viral Ags by means of phage display and hybridoma strategies (2730). These TCR-like Abs, which exhibit both binding properties and kinetics of Abs (e.g., high affinity), while mimicking the specificity of TCRs, are being used as a novel research tool to study Ag presentation and immunotherapy targets.

The major difference between native αβ-TCRs and TCR-like Abs lies in the binding affinity and avidity. Whereas the αβ-TCRs possess relatively low affinity in the mM range, their avidity is high because multiple TCRs are present on the surface of T cells (31, 32). In contrast, TCR-like Abs possess high affinity in the range of medium-to-low nM, but their avidity is limited to monovalent or bivalent formats as a recombinant scFv or Fab Ab fragment or IgG, respectively. The present study investigates how these differences in affinity and avidity of an αβ-TCR versus a TCR-like Ab–based CAR play a functional role in engineered T cells that carry these recognition moieties. We have isolated TCR-like Fab Abs that recognize the HLA-A2 molecule bearing the WT1-derived peptide (WT1Db126; RMFPNAPYL) with different affinities (400 and 30 nM). CARs were generated based on the isolated Fabs and were used to redirect T cells toward HLA-A2-WT1Db126. In comparison, we used engineered T cells that carry αβ-TCR genes targeting the same WT1-specific epitope. These redirected T cells exhibited efficient and specific reactivity with HLA-A2-WT1Db126 complexes.

Using these TCR-like Ab–based CARs and the recombinant TCR, we directly compared, to our knowledge for the first time, two different approaches for redirecting T cells. This study enhances the understanding of the influence and relationships of affinity and avidity on the biological functions of T cells being redirected by either cloned αβ-TCRs or TCR-like Ab–based CARs. We demonstrate herein that the combination of high affinity and avidity of a TCR-like Ab displayed on the surface of the engineered T cells has dramatic effects on the specificity and function of these T cells compared with engineered T cells carrying low-affinity αβ-TCRs. We present evidence for a TCR-based affinity threshold that limits the maximal T cell effective function, and we suggest that rational design of improved TCRs or TCR-like Abs for T cell redirection may need to be optimized up to a given affinity threshold to achieve optimal T cell function without risking cross reactivity. Our comparison enables better understanding of limits and thresholds in T cell recognition using these two different approaches, and it clarifies the optimal recognition properties required for most effective and specific T cell retargeting toward tumor cells.

Unless otherwise stated, all culture medium was RPMI 1640 supplemented with 10% heat inactivated FCS, 1% penicillin and streptomycin, and 1% l-glutamine. Cell lines used were: Jurkat 76 cell line, TAP-deficient-HLA-A2+ T2 cell line that can be efficiently loaded with exogenous peptides, 501A and Skmel5 melanoma cell lines, Loucy and BV-173 ALL cell line, DG75 lymphoma cell line, MDA-MB-231 human breast carcinoma, SW620 and colo-205 human colon cancer, Panc-1 human pancreatic carcinoma, and A431 epidermoid carcinoma cell line. The cell lines UMUC3 human bladder transitional cell carcinoma and Fibroblast were cultured in DMEM supplemented with 10% heat-inactivated FCS, 1% penicillin and streptomycin, and 1% l-glutamine. Caco-2 human colon cancer cell line was cultured in DMEM supplemented with 20% heat-inactivated FCS, 1% penicillin and streptomycin, and 1% l-glutamine. PBMCs were obtained from volunteer donors from the National Blood Service (Colindalea, London, U.K.).

Flow cytometry Abs (Abs) were anti-human PE (Jackson ImmunoResearch), CD3 PerCp, CD8 APC, CD8 APC-Cy7, IFN-γ–APC, IL-2 PE (Bactlab Diagnostic), CD107a eFluor660, and PE-labeled HLA-A2/WTDb126 tetramers (Beckman Coulter). The peptides used for this study were pWT1Db126 (RMFPNAPYL), pWT1235 (CMTWNQMNL); gp100: G2-209-2M (IMDQVPFSV) and gp100-280 (YLEPGPVTA); and HIV: Gag (SLYNYVATL), MDM2 (LLGDLFGV).

Single-chain MHC (scMHC)-peptide complexes were produced by in vitro refolding of inclusion bodies produced in Escherichia coli upon IPTG induction, as described previously (33). An scMHC containing the β2-microglobulin and the extracellular domains of the HLA-A2 gene connected to each other by a flexible linker was engineered to include the BirA recognition sequence for site-specific biotinylation at the carboxyl-terminus. In vitro refolding was performed in the presence of peptides as described. Correctly folded pMHCs were isolated and purified by anion exchange Q-Sepharose chromatography (Pharmacia), followed by site-specific biotinylation using the BirA enzyme (Avidity).

Selection of phage Abs on biotinylated complexes was performed as described previously (27). A large human Fab library containing 3.7 × 1010 different Fab clones was used for the selection. Phages were first preincubated with streptavidin-coated paramagnetic beads (200 μl; Dynal) to deplete the streptavidin binders. The remaining phages were subsequently used for panning with decreasing amounts of biotinylated scMHC-peptide complexes. The streptavidin-depleted library was incubated in solution with soluble biotinylated scHLA-A2-WT1 complexes (500 nM for the first round, and 100 nM for the subsequent rounds) that were added to the mixture and incubated for 30 min at room temperature. Streptavidin-coated magnetic beads (200 μl for the first round, and 100 μl for the subsequent rounds) were added to the mixture and incubated for 10–15 min at room temperature. The beads were washed extensively 12 times with PBS/0.1% Tween 20 with an additional two washes with PBS. Bound phages were eluted with triethylamine (100 mM, 5 min at room temperature), followed by neutralization with Tris-HCl (1 M, pH 7.4), and used to infect E. coli TG1 cells (OD600 = 0.5) for 30 min at 37°C.

Fab Abs were expressed and purified as described previously (27). E. coli BL21 cells were grown to OD600 = 0.8–1.0 and were induced to express the recombinant Fab Ab by the addition of IPTG for 3–4 h at 30°C. Periplasmic content was released using the B-PER solution (Pierce), which was applied onto a prewashed TALON column (Clontech). Bound Fabs were eluted using 0.5 ml of 100-mM imidazole in PBS. The eluted Fabs were dialyzed twice against PBS (overnight, 4°C) to remove residual imidazole.

Kinetic studies for affinity measurements of the F2 and F3 Fabs to HLA-A2/WT1Db126 complexes were performed on a ProteOn XPR36 Protein Interaction Array System (Bio-Rad Laboratories, Hercules) as described previously (34).

The number of specific pMHCs on the surface of tumor cell lines was determined as previously described (35). Specific binding of F3 Fab to HLA-A2/WT1Db126 complexes was detected using PE-labeled anti-λ L chain mAb. To transform the florescent signal obtained by flow cytometer into the number of HLA-A2/WT1Db126 sites, we used the QuantiBRITE PE kit (BD Biosciences) according to the manufacturer’s instructions.

scFv DNA of the TCR-like Fabs F2 and F3 were generated by connecting the carboxyl-terminus of the VL region and the N terminus of the VH region by a peptide linker. For the chimeric receptor construct, the F2 and F3 scFvs were connected via the carboxyl-terminus of the VH region to a CD28- FcγRI γ-chain construct (36). The TCR-like chimeric receptor DNA constructs were cloned into retroviral pBullet vector followed by IRES and the GFP gene (37).

A total of 2 × 106 Phoenix amphotropic packaging cells were cultured in 10-cm culture plates for 24 h at 37°C with 5% CO2. The cells were transfected with the vector constructs and pCL-ampho using calcium phosphate precipitation (Invitrogen Life Technologies). After culturing for 24 h and replacing the medium, the viral supernatant was harvested. Twenty-four hours before retroviral transduction, Jurkat cells were split and T cells were enriched using Ficoll and RosetteSep Human T cell enrichment kit followed by an activation for 48 h using 24-well nontreated plates coated with anti-CD3 Ab OKT3 at 1 μg/ml and anti-CD28 at 5 μg/ml with the addition of and IL-2 (600 U/ml; Chiron). For retroviral transductions, RetroNectin-coated (Takara) 24-well plates were seeded with cells at 1 × 106 per well in 1 ml, cultured for 30 min, and transduced with 1 ml of the constructs viral supernatant. For PBMCs (T cells), the transductions were conducted in culture medium supplemented with IL-2 at 600 U/ml. After 24 h at 37°C with 5% CO2, the culture medium for Jurkat cells was replaced; for PBMCs (T cells) the replaced medium was supplemented with IL-2 at 100 U/ml. After an additional 48 h culture period, flow cytometry analysis was performed by a LSR II flow cytometer (BD Biosciences).

Transduced T cells and T2 cells loaded with WT1Db126 or WT1235 were added at 2 × 105 cells/well in 200 μl of culture medium containing brefeldin A (Sigma-Aldrich) at 1 μg/ml. After 16 h at 37°C with 5% CO2 staining of the cells for surface CD8 was performed, followed by fixation, permeabilization, and staining for intracellular IFN-γ and IL-2 (Fix & Perm Kit; Caltag). Cells were then washed and analyzed by a LSR II flow cytometer (BD Biosciences).

Transduced T cells were stimulated with irradiated T2 cells (at 1:1 ratio) loaded with the WT1126 or Gp100-280 peptide. The assay was conducted in triplicates in 200 μl medium. After 18 h of incubation at 37°C with 5% CO2, the supernatant was harvested and tested for secreted IFN-γ using a human ELISA kit (BD Biosciences).

T cell subpopulation depletion was obtained by incubating the T cell population with anti-CD4 or anti-CD8 for 20 min. After washing with PBS 0.1% BSA, cells were mixed with anti-mouse IgG-coated magnetic beads (Dynal, Invitrogen) for an additional 30 min followed by magnetic depletion for 5 min. The negative fraction was then washed three times with PBS 0.1% BSA and incubated for 24 h recovery in 37°C. Flow cytometry analysis of purified subpopulations revealed purify >90%.

The EBV-transformed B cell line T2 was labeled with [35S] methionine for 18 h at 37°C with 5% CO2. The cells were washed three times and loaded with the various concentrations of WT1126 or Gp100-280 peptide for 2 h at 37°C with 5% CO2. After incubation, the cells were washed of excess peptide. For different E:T ratios, peptide-loaded [35S]-labeled T2 cells were added to 2-fold dilutions of transduced T cells. For the peptide titration assays, transduced cells and peptide-loaded [35S]-labeled T2 cells were cocultured at a ratio of 5:1 (E:T). After an incubation period of 4 h at 37°C with 5% CO2, 25 μl of supernatant were harvested, diluted with 200 μl of scintillation fluid, and counted using a β counter (BD Biosciences). The percentage of specific lysis was calculated as ([Experimental [35S]-release − Spontaneous [35S]-release] / [Maximum [35S]-release − Spontaneous [35S]-release]) × 100, with spontaneous release being the [35S]methionine released from target cells in the absence of effector cells and maximum release being the [35S]methionine released from target cells lysed with 0.05 M NaOH. The percentage of maximal lysis was determined as the highest cytotoxic activity of target cells. In the cytotoxic assays of the CD4, CD8-enriched T cells, the amount of cells used was normalized respective to the highest frequency of specific transduced cells determined by tetramer or Vβ Ab staining.

Transduced T cells were cocultured with T2 cells (at 1:1 ratio) loaded with the WT1126 or Gp100-280 peptide in 200 μl medium. After 4 h of incubation at 37°C with 5% CO2, the cells were stained with anti-CD8 PerCP and anti-CD107a Abs for 30 min. The cells were washed, resuspended in PBS 0.1% BSA and the expression of CD107a on CD8+ transduced T cell was determined by flow cytometry (FACSCalibur, Becton Dickinson).

Recombinant TCR-like Fab Abs recognizing HLA-A2-WT1Db126 complexes were isolated by screening a large naive phage Fab library as described previously (27, 28). For panning, we used recombinant single-chain HLA-A2-WT1Db126 complexes expressed in E. coli as insoluble inclusion bodies. Subsequently, HLA-A2-WT1Db126 complexes were refolded and purified using established redox-shuffling strategies (28). The phage display screening was analyzed using differential binding to specific HLA-A2-WT1Db126 versus control complexes and identified specific clones. Of 96 phage clones tested, 13 exhibited specific binding to recombinant HLA-A2-WT1Db126 complexes, compared with control complexes displaying irrelevant peptides (Fig. 1A).

FIGURE 1.

Binding properties of F2 and F3 anti-HLA-A2/WT1Db126 TCR-like recombinant Abs. (A) ELISA of anti-HLA-A2- WT1Db126 soluble purified Fabs with immobilized HLA-A2-WT1Db126 complexes versus control complexes displaying control peptide. Anti-HLA mAb W6/32 was used to determine the correct folding and stability of the bound complexes during the binding assay. (B) Flow cytometric analysis of the binding of F2 and F3 Fabs to T2 cells loaded with WT1Db126 peptide or control peptides. (C) SPR analysis of F2 and F3 Fabs for affinity to HLA-A2- WT1Db126 complexes, determined as 400 and 30 nM, respectively.

FIGURE 1.

Binding properties of F2 and F3 anti-HLA-A2/WT1Db126 TCR-like recombinant Abs. (A) ELISA of anti-HLA-A2- WT1Db126 soluble purified Fabs with immobilized HLA-A2-WT1Db126 complexes versus control complexes displaying control peptide. Anti-HLA mAb W6/32 was used to determine the correct folding and stability of the bound complexes during the binding assay. (B) Flow cytometric analysis of the binding of F2 and F3 Fabs to T2 cells loaded with WT1Db126 peptide or control peptides. (C) SPR analysis of F2 and F3 Fabs for affinity to HLA-A2- WT1Db126 complexes, determined as 400 and 30 nM, respectively.

Close modal

Based on specific recognition to HLA-A2-WT1Db126 complexes, two clones, F2 and F3, were selected for further characterization and tested for their ability to bind the HLA-A2-WT1Db126 complexes in their native form. To this end, we used the TAP-deficient HLA-A2 positive T2 cells that were loaded with the WT1Db126RMFPNAPYL peptide or control peptides. As shown in Fig. 1B, flow cytometry assays using the anti-HLA-A2-WT1Db126-specific soluble purified Fabs, F2 or F3, exhibited specific binding to WT1Db126-loaded APCs, but not to APCs loaded with control peptides. The binding affinity of the HLA-A2/WT1Db126–specific Fab TCR-like Abs was determined by SPR using soluble HLA-A2/WT1Db126 complexes on chip-immobilized Fabs coupled through an anti-human Fab, and revealed an apparent affinity of 400 nM and 30 nM for F2 and F3, respectively (Fig. 1C).

To determine whether the WT1-specific TCR-like Abs recognize the authentic endogenously derived HLA-A2-WT1Db126 complexes on the surface of target cells, we tested binding using a large panel of tumor cell lines. As shown in Fig. 2, the F2 Fab recognized HLA-A2-WT1Db126–positive cells. No reactivity was observed with WT1-negative fibroblasts cells or HLA-A2-negative A431 cells. Similar data were observed with the F3 Fab (data not shown). The endogenous expression of WT1 transcript in these tumor cell lines was determined by mRNA analysis (Table I). We quantified the number of HLA-A2-WT1Db126 complexes expressed on the surface of tumor cell lines using the F3 Fab and the total number of HLA-A2 molecules expressed on the cell surface using the anti–HLA-A2 Ab BB7.2 (Table I). Overall, our data indicate that Fabs F2 and F3 exhibit properties of TCR-like Abs; they bind with HLA-A2 restriction and WT1Db126 peptide specificity to peptide-loaded APCs as well as to tumor target cells that present the Ag in an HLA-restricted manner and with high affinity (30 and 400 nM, respectively) compared with native TCRs.

FIGURE 2.

Reactivity of F2 anti–HLA-A2-WT1Db126 TCR-like Fabs with tumor cell lines. Detection of HLA-A2-WT1Db126 complexes on the surface of HLA-A2+ WT1+ cell lines: 501A and Skmel5 melanoma cell lines; Loucy ALL cell line; CCRF-SB and DG75 lymphoma cell line; MDA-MB-231 human breast carcinoma; SW620, Caco-2, and colo-205 human colon cancer; Panc-1 human pancreatic carcinoma; Hep-G2 liver hepatocellular carcinoma; UMUC3 human bladder transitional cell carcinoma. Cells were incubated with anti–HLA-A2-WT1Db126 TCR-like Fabs, followed by incubation with PE-labeled anti-human Ab. HLA-A2–negative (WT1-positive) A431 epidermoid carcinoma cell line and WT1-positive (HLA-A2–negative). Fibroblast cells were used as controls.

FIGURE 2.

Reactivity of F2 anti–HLA-A2-WT1Db126 TCR-like Fabs with tumor cell lines. Detection of HLA-A2-WT1Db126 complexes on the surface of HLA-A2+ WT1+ cell lines: 501A and Skmel5 melanoma cell lines; Loucy ALL cell line; CCRF-SB and DG75 lymphoma cell line; MDA-MB-231 human breast carcinoma; SW620, Caco-2, and colo-205 human colon cancer; Panc-1 human pancreatic carcinoma; Hep-G2 liver hepatocellular carcinoma; UMUC3 human bladder transitional cell carcinoma. Cells were incubated with anti–HLA-A2-WT1Db126 TCR-like Fabs, followed by incubation with PE-labeled anti-human Ab. HLA-A2–negative (WT1-positive) A431 epidermoid carcinoma cell line and WT1-positive (HLA-A2–negative). Fibroblast cells were used as controls.

Close modal
Table I.
mRNA expression and quantitation of the number of HLA-A2/WTDb126 complexes presented on the surface of tumor cell lines
CellsWT1 mRNAHLA-A2aHLA-A2/WT1b
501A (melanoma) 30,191 3,719 
Skmel5 (melanoma) 17,382 1,030 
Loucy (T-ALL) 3,449 45 
CCRF-SB (lymphoma) 14,251 2,542 
DG-75 (lymphoma) 15,137 314 
MDA-231 (breast) 16,500 2,051 
SW620 (colon) 3,227 81 
Caco-2 (colon) 4,538 
Colo-205 (colon) 2,962 76 
Panc-1 (pancreas) 9,115 680 
Hep-G2 (liver) 3,440 328 
UMUC3 (bladder) 9,941 649 
Fibroblasts − 21,800 
A431 (epidermal) 
CellsWT1 mRNAHLA-A2aHLA-A2/WT1b
501A (melanoma) 30,191 3,719 
Skmel5 (melanoma) 17,382 1,030 
Loucy (T-ALL) 3,449 45 
CCRF-SB (lymphoma) 14,251 2,542 
DG-75 (lymphoma) 15,137 314 
MDA-231 (breast) 16,500 2,051 
SW620 (colon) 3,227 81 
Caco-2 (colon) 4,538 
Colo-205 (colon) 2,962 76 
Panc-1 (pancreas) 9,115 680 
Hep-G2 (liver) 3,440 328 
UMUC3 (bladder) 9,941 649 
Fibroblasts − 21,800 
A431 (epidermal) 
a

The number of HLA-A2 molecules was determined with labeled MAb BB7.2 and calibration beads that extrapolate fluorescence intensity to a number of sites.

b

The number of HLA-A2/WT1 molecules was determined with labeled HLA-A2/WT1 TCRL F2 and calibration beads that extrapolate fluorescence intensity to a number of sites.

Previous CAR studies revealed that the use of single-chain CARs is an efficient strategy to redirect T cells using Ab fragments, as it requires only a single retroviral transduction step (38). Therefore, we generated scFv forms of the F2 and F3 TCR-like Fab Abs by fusing the carboxyl-terminus of the VL domain to the N terminus of the VH domain using a flexible Gly4Ser peptide linker. The complete CAR molecule was composed of the F2 or F3 scFv connected via the carboxyl-terminus of the VH domain to a CD28–γ-chain construct. This CAR construct was introduced into a retroviral vector (pBULLET), which also harbors a GFP reporter gene to visualize cells transduced with the vector (37). Fig. 3 represents an analysis of TCR-negative Jurkat-76 cells transduced with the CAR retroviral vectors encoding the F2 or the F3 chimeric receptors. We examined the staining of GFP-positive transduced cells with WT1 and control tetramers and found that GFP-positive cells representing transduced cells were costained with a high frequency of >20% with HLA-A2-WT1Db126–specific tetramers, but not with tetramers displaying irrelevant peptides. Data are shown for the F2 CAR and similar results were obtained for F3.

FIGURE 3.

Expression of TCR-like chimeric receptor on transduced Jurkat cells. Jurkat cells were transduced with retroviral vector encoding the 400-nM F2 anti–HLA-A2-WT1Db126 chimeric receptor construct; 96 h after transduction, cells were stained with PE-labeled HLA-A2 tetramers presenting the WT1Db126 specific peptide or control peptides and were analyzed by flow cytometry. GFP expression represents positive transduced Jurkat cells.

FIGURE 3.

Expression of TCR-like chimeric receptor on transduced Jurkat cells. Jurkat cells were transduced with retroviral vector encoding the 400-nM F2 anti–HLA-A2-WT1Db126 chimeric receptor construct; 96 h after transduction, cells were stained with PE-labeled HLA-A2 tetramers presenting the WT1Db126 specific peptide or control peptides and were analyzed by flow cytometry. GFP expression represents positive transduced Jurkat cells.

Close modal

Next, we tested the transduction efficiency in human primary T cells. F2 (400 nM) and F3 (30 nM) TCR-like CARs were retrovirally transduced into HLA-A2–positive human T cells. Fig. 4A shows that CD8+ human T cells were efficiently transduced to express significant levels of the F2 TCR-like CAR, as evident by staining of transduced cells with HLA-A2-WT1Db126 tetramers. Human CD8+ T cells transduced with the F3 TCR-like CAR expressed a high level of CAR, as revealed by tetramer staining; however, the viability of the transduced cells was very low. As shown in Fig. 4A, the remaining fraction of viable CD8+ transduced T cells expressed the receptor at high levels (>60%), as determined by tetramer staining. Retroviral transduction of the 30-nM TCR-like F3 CAR into HLA-A2 negative cells yielded cell surface expression comparable to that observed for F2 (Fig. 4B), but with much higher viability rate (most transduced cells were viable). These results suggest that the elevated affinity of the 30-nM TCR-like CAR, F3, in combination with the high avidity of the receptor present on the T cell surface, may lead to some loss of specificity and consequently to decreased cell survival. The F2 TCR-like CAR, having moderate affinity of 400 nM, exhibited expression properties and characteristics that are more suitable for the comparison of T cells redirected by either CAR based on TCR-like Ab fragments versus a cloned αβ-TCR.

FIGURE 4.

Expression of TCR-like chimeric receptors on transduced human T cells. HLA-A2+ (A) and HLA-A2 (B) activated primary T cells were transduced with either vector encoding the F2 or F3 TCR-like chimeric receptors; 48 h after transduction, the cells were stained with anti-CD3 and CD8 Abs and tetramer. Cells were gated on the CD3+ T cell populations.

FIGURE 4.

Expression of TCR-like chimeric receptors on transduced human T cells. HLA-A2+ (A) and HLA-A2 (B) activated primary T cells were transduced with either vector encoding the F2 or F3 TCR-like chimeric receptors; 48 h after transduction, the cells were stained with anti-CD3 and CD8 Abs and tetramer. Cells were gated on the CD3+ T cell populations.

Close modal

To compare the specificity and function of T cells retargeting by either αβ-TCRs or TCR-like Ab–based CARs, we used an αβ-TCR specific to the HLA-A2-WT1Db126 epitope, which has been previously isolated and characterized by Prof. Stauss’s group (University College of London) and was inserted into the retroviral vector MP71 (7). We studied the transduction efficiency of the αβ-TCR versus that of the F2 TCR-like CAR, using retroviral transductions into primary human T cells. Fig. 5 demonstrates efficient transductions of HLA-A2+ T cells transduced with either the αβ-TCR construct (as determined by Vβ2.1 or tetramer staining) or the F2 TCR-like CAR construct (as determined by tetramer staining). The αβ-TCR construct has the addition of an internal disulfide bond to minimize the mispairing of the exogenous TCR-α and -β chains with the endogenous TCR chains (39). However, tetramer staining of the αβ−TCR transduced T cells pointed to low levels of expression. This observation may have been due to either a low expression of functional αβ-TCR, which will not facilitate tetramer binding, or TCR mispairing.

FIGURE 5.

Expression of F2 TCR-like chimeric receptor and WT1Db126 αβ-TCR construct on transduced human T cells. HLA-A2+ activated primary T cells were transduced with either vector encoding the αβ-TCR or F2 chimeric receptor; 48 h after transduction, the cells were stained with anti-CD3, CD8, Vβ2.1 Abs, or HLA-A2-WT1Db126 tetramer. Cells were gated on the CD3+ T cell populations.

FIGURE 5.

Expression of F2 TCR-like chimeric receptor and WT1Db126 αβ-TCR construct on transduced human T cells. HLA-A2+ activated primary T cells were transduced with either vector encoding the αβ-TCR or F2 chimeric receptor; 48 h after transduction, the cells were stained with anti-CD3, CD8, Vβ2.1 Abs, or HLA-A2-WT1Db126 tetramer. Cells were gated on the CD3+ T cell populations.

Close modal

As a first step for functional evaluation and comparison of T cells transduced with αβ−TCR or TCR-like Ab CAR, we tested their ability to undergo proper activation. αβ-TCR or F2 TCR-like Ab CAR-transduced T cells were stimulated with the human TAP-deficient T2 cells loaded with either the specific peptide (WT1Db126) or an irrelevant peptide (WT1235) as a control. After overnight stimulation, the cells were stained for CD8 and for intracellular IFN-γ and IL-2. As shown in Fig. 6A, CAR-transduced T cells and αβ-TCR–transduced T cells were activated in an Ag-specific manner as indicated by the comparable number of transduced T cells that were intracellularly stained with anti–IL-2, IFN-γ, or both.

FIGURE 6.

Response of CARs transduced T cells to Ag-specific stimulation. (A) FACS analysis of freshly transduced T cells with either 400 nM F2 TCR-like chimeric receptor or WT1Db126 αβ-TCR. Transduced cells were stimulated for 18 h with T2 cells loaded with 100 μM either relevant (WTDb126) or control (WT235) peptide. Cells were stained for CD8 and then fixed, permeabilized, and stained with anti–IFN-γ and anti–IL-2 Abs followed by flow cytometric analysis. Shown is the percentage of CD8+ T expressing IFN-γ and IL-2. The percentage of positive transduced CD8+ cells was 30% and 42% for TCR and F2, respectively. (B) ELISA assay for IFN-γ release. Transduced T cells were stimulated for 18 h with T2 cells loaded with decreasing dilutions of relevant (WTDb126) peptide (100 to 0.1 μM) or control (Gp100-280) peptide. IFN-γ release was determined by ELISA. The percentage of positive transduced CD8+ cells was 57% and 41% for TCR and F2, respectively. The percentage of positive transduced CD4+ cells was 48% and 30% for TCR and F2, respectively. The data shown are representative of three independent experiments showing similar results. (C) Flow cytometry for CD107a expression on CD8+ T cell. Transduced T cells were stimulated for 4 h with T2 cells loaded with decreasing dilutions of relevant (WTDb126) peptide (100 to 0.1μM) or control (Gp100-280) peptide. Percent of CD107a was determined with flow cytometry. Data shown are representative of three independent experiments showing similar results.

FIGURE 6.

Response of CARs transduced T cells to Ag-specific stimulation. (A) FACS analysis of freshly transduced T cells with either 400 nM F2 TCR-like chimeric receptor or WT1Db126 αβ-TCR. Transduced cells were stimulated for 18 h with T2 cells loaded with 100 μM either relevant (WTDb126) or control (WT235) peptide. Cells were stained for CD8 and then fixed, permeabilized, and stained with anti–IFN-γ and anti–IL-2 Abs followed by flow cytometric analysis. Shown is the percentage of CD8+ T expressing IFN-γ and IL-2. The percentage of positive transduced CD8+ cells was 30% and 42% for TCR and F2, respectively. (B) ELISA assay for IFN-γ release. Transduced T cells were stimulated for 18 h with T2 cells loaded with decreasing dilutions of relevant (WTDb126) peptide (100 to 0.1 μM) or control (Gp100-280) peptide. IFN-γ release was determined by ELISA. The percentage of positive transduced CD8+ cells was 57% and 41% for TCR and F2, respectively. The percentage of positive transduced CD4+ cells was 48% and 30% for TCR and F2, respectively. The data shown are representative of three independent experiments showing similar results. (C) Flow cytometry for CD107a expression on CD8+ T cell. Transduced T cells were stimulated for 4 h with T2 cells loaded with decreasing dilutions of relevant (WTDb126) peptide (100 to 0.1μM) or control (Gp100-280) peptide. Percent of CD107a was determined with flow cytometry. Data shown are representative of three independent experiments showing similar results.

Close modal

Next, we examined the dose dependency and peptide specificity of IFN-γ secretion by transduced T cells in an ELISA assay. As shown in Fig. 6B, T cells transduced with the αβ-TCR construct were more sensitive to lower peptide concentrations than T cells transduced with the F2 TCR-like Ab CAR were. T cells expressing the αβ-TCR secreted higher levels of IFN-γ in response to T2 cells loaded with peptide concentration ranging from 0.3 to 100 μM, whereas T cells expressing the F2 TCR-like CAR secreted considerably reduced levels of IFN-γ at peptide concentrations starting as low as 0.30 μM. At 1 μM, T cells transduced with native TCR secreted significant levels of IFN-γ, whereas the TCRL-transduced T cells did not secret at all. Notably, at 10 μM, the αβ-TCR–transduced T cells exerted high and close to maximal IFN-γ secretion, whereas the TCR-like Ab CAR-transduced T cells showed ∼50% less secretion. In addition to measuring IFN-γ secretion, we accessed the expression of CD107a, a marker of CD8+ T cell degranulation following stimulation, on the surface of the transduced T cells (Fig. 6C). Confirming the cytokine secretion assays, we observed significant differences in the expression level of CD107a as a function of WT1Db126 peptide concentration. The F2 TCR-like Ab–transduced T cells expressed significant lower levels of CD107a compared with the TCR-transduced T cells, which indicates a much lower degranulation and stimulation levels. At low peptide doses (1 and 3 μM), a 3–10-fold difference in CD107a expression was observed (Fig. 6C). These results indicate major differences in Ag sensitivity of the αβ-TCR versus TCR-like Ab CARs and consequently in T cell stimulation and cytokine secretion.

Next, we compared the F2 TCR-like Ab CAR to the engineered αβ-TCR for their cytolytic activity toward [35S]-methionine–labeled T2 cells loaded with either a relevant (WT1Db126) or irrelevant control peptide. As shown in Fig. 7A, T cells transduced with the αβ-TCR showed greater specific cytolytic activity than T cells transduced with the F2 TCR-like Ab CAR did. Both transduced T cells maintained their specificity toward the WT1 peptide, as their background cytotoxic activity toward a control nonspecific peptide was similar and low (Fig. 7A). The transduced T cells exhibited also low accepted background cytotoxicity to T2 APCs without any loaded peptide (Fig. 7B). To examine the Ag sensitivity of both receptors, we performed a killing assay using [35S]-methionine–labeled T2 cells loaded with decreasing concentrations of the WT1Db126-specific peptide.

FIGURE 7.

Ag-specific cytotoxic activity of CARs transduced T cells. (A) Transduced T cells were cultured in a 4-h assay with [35S]-labeled T2 cells loaded with either 100 μM relevant (WT1Db126) or control (Gp100-280) peptide at the stated E:T ratios. The data shown are representative of three independent experiments showing similar results. (B) Transduced T cells were cultured in a 4-h assay with [35S]-labeled T2 cells, loaded with decreasing dilution of WT1Db126 peptide (100 to 0.1μM) or control peptide Gp100-280 (100 μm), at an E:T ratio of 5:1. The data shown are representative of three independent experiments showing similar results. In both assays, the amount of positive transduced cells added was normalized respectively to the highest percent of transduced cell. (C) Transduced T cells were cultured in a 4-h assay with [35S]-labeled 501A Melanoma and Breast MDA231 (HLA-A2+WT1Db126+) or control A431 (HLA-A2-WT1Db126+) at the stated E:T ratios. In all assays, the percentage of positive transduced CD8+ cells was 57% and 41% for TCR and F2, respectively. The percentage of positive transduced CD4+ cells was 48% and 30% for TCR and F2, respectively. (A) and (C) represent a typical result of three independent repetitions.

FIGURE 7.

Ag-specific cytotoxic activity of CARs transduced T cells. (A) Transduced T cells were cultured in a 4-h assay with [35S]-labeled T2 cells loaded with either 100 μM relevant (WT1Db126) or control (Gp100-280) peptide at the stated E:T ratios. The data shown are representative of three independent experiments showing similar results. (B) Transduced T cells were cultured in a 4-h assay with [35S]-labeled T2 cells, loaded with decreasing dilution of WT1Db126 peptide (100 to 0.1μM) or control peptide Gp100-280 (100 μm), at an E:T ratio of 5:1. The data shown are representative of three independent experiments showing similar results. In both assays, the amount of positive transduced cells added was normalized respectively to the highest percent of transduced cell. (C) Transduced T cells were cultured in a 4-h assay with [35S]-labeled 501A Melanoma and Breast MDA231 (HLA-A2+WT1Db126+) or control A431 (HLA-A2-WT1Db126+) at the stated E:T ratios. In all assays, the percentage of positive transduced CD8+ cells was 57% and 41% for TCR and F2, respectively. The percentage of positive transduced CD4+ cells was 48% and 30% for TCR and F2, respectively. (A) and (C) represent a typical result of three independent repetitions.

Close modal

As shown in Fig. 7B, the sensitivity of the αβ-TCR–transduced T cells was indeed greater compared with the TCR-like Ab–transduced cells, as the αβ-TCR cells were more efficient in mediating killing of WT1-loaded T2 cells at low peptide concentration. As observed previously, this result was most significant at the lower peptide doses of 1–3 μM, with a 4–20-fold difference.

We also tested the killing sensitivity of the two types of transduced T cells on tumor cells that express the native HLA-A2/ WT1Db126 complex. For this purpose we used HLA-A2+ WT1 [35S]-methionine-labeled Ag positive 501A and MDA231 cells and HLA-A2-/WT1+ A431 cells as control. As shown in Fig. 7C, we observed a marked and significant difference in the killing sensitivity of the αβ-TCR–transduced T cells compared with the TCR-like Ab CAR T cells at various E:T ratios. Although the αβTCR transduced T cells exhibited E:T-dependent killing of 501A and MDA231 tumor target cells, the TCR-like Ab–transduced cells showed only marginal minor killing activity on these cells (Fig. 7C). These results correspond to the aforementioned observation that T cells transduced with the αβ-TCR exhibited higher Ag sensitivity compared with the TCR-like Ab–transduced cells, and consequently results in more efficient cytolytic activity toward tumor cells.

Next, we attempted to investigate the differences between CAR and TCR-transduced T cells concerning the relative contribution of the CD8 and CD4 subpopulations to overall T cell–mediated cytotoxic activity. Thus, we transduced primary T cells with the CAR or TCR constructs and derived the CD4+ and CD8+ T cell subpopulations by depletion with anti-CD8 or anti-CD4, respectively. Flow cytometric analysis of purified subpopulations revealed purity >95% (data not shown). Using the isolated purified CD4+ and CD8+ subpopulations, we tested their sensitivity and function in killing assays in which peptide-pulsed [35S]-methionine–labeled T2 cells were used as targets (Fig. 8). We observed that the differences in sensitivity and function of TCR versus CAR –transduced T cells was maintained also for the purified subpopulation of CD8+ T cells when peptide sensitivity (Fig. 8A) and E:T ratio (Fig. 8B) were tested. TCR-transduced CD8+ T cells were more active compared with CAR-transuded cells similarly with what was observed with the intact whole T cell population. We also observed that most killing activity was mediated by CD8+ T cells (comparing data in Fig. 8A and 8B to Fig. 8C and 8D) and CD4+ cells exhibited minor killing activity of up to 30% killing compared with efficient killing (up to 100% of relative maximal killing) with CD8+ T cells. Interestingly, purified CD4+ cells of CAR-transduced T cells was somewhat more efficient compared with the TCR-transduced cells in most of the peptide doses tested (Fig. 8C) and at high E:T ratios (Fig. 8D). This result demonstrates that the CAR construct confers upon the T cell a TCR-like specificity that is not CD8 dependent.

FIGURE 8.

Ag-specific cytotoxic activity of CD8 and CD4 transduced T cells. Purified CD4 and CD8 subpopulations were obtained by incubating the T cells with anti-CD4, anti-CD8, and anti-mouse IgG-coated magnetic beads. (A) Transduced purified CD8 T cells were cultured in a 4-h assay with [35S]-labeled T2 cells and loaded with decreasing dilution of WT1Db126 peptide (100 to 0.1 μM) or control peptide Gp100-280 (100 μm), at an E:T ratio of 5:1. (B) Transduced purified CD8 T cells were cultured in a 4-h assay with [35S]-labeled T2 cells loaded with either 100 μM relevant (WT1Db126) or control (Gp100-280) peptide at the stated E:T ratios. (C) Transduced purified CD4 T cells were cultured in a 4-h assay with [35S]-labeled T2 cells, loaded with decreasing dilution of WT1Db126 peptide (100 to 0.1 μM) or control peptide Gp100-280 (100 μm), at an E:T ratio of 5:1. (D) Transduced purified CD4 T cells were cultured in a 4-h assay with [35S]-labeled T2 cells loaded with either 100 μM relevant (WT1Db126) or control (Gp100-280) peptide at the stated E:T ratios. The data shown are representative of three independent experiments showing similar results. In all assays, the amount of positively transduced F2 and TCR CD8 and CD4 T cells added was normalized respectively to 60%, which was the highest percentage of transduce CD8 TCR transduced cell.

FIGURE 8.

Ag-specific cytotoxic activity of CD8 and CD4 transduced T cells. Purified CD4 and CD8 subpopulations were obtained by incubating the T cells with anti-CD4, anti-CD8, and anti-mouse IgG-coated magnetic beads. (A) Transduced purified CD8 T cells were cultured in a 4-h assay with [35S]-labeled T2 cells and loaded with decreasing dilution of WT1Db126 peptide (100 to 0.1 μM) or control peptide Gp100-280 (100 μm), at an E:T ratio of 5:1. (B) Transduced purified CD8 T cells were cultured in a 4-h assay with [35S]-labeled T2 cells loaded with either 100 μM relevant (WT1Db126) or control (Gp100-280) peptide at the stated E:T ratios. (C) Transduced purified CD4 T cells were cultured in a 4-h assay with [35S]-labeled T2 cells, loaded with decreasing dilution of WT1Db126 peptide (100 to 0.1 μM) or control peptide Gp100-280 (100 μm), at an E:T ratio of 5:1. (D) Transduced purified CD4 T cells were cultured in a 4-h assay with [35S]-labeled T2 cells loaded with either 100 μM relevant (WT1Db126) or control (Gp100-280) peptide at the stated E:T ratios. The data shown are representative of three independent experiments showing similar results. In all assays, the amount of positively transduced F2 and TCR CD8 and CD4 T cells added was normalized respectively to 60%, which was the highest percentage of transduce CD8 TCR transduced cell.

Close modal

Overall, the data with total unseparated T cells transduced with TCR or CAR as well as the CD8+ and CD4+ subpopulations further support the findings that an upper affinity threshold for TCR-based recognition is required to mediate effective and optimal functional activity in killing of target cells. When these principals are applied to the functional outcomes of engineered T cells, the rational design of TCRs and TCR-based constructs may need to be optimized to a given affinity threshold to achieve optimal T cell function.

Engineered T cells constitute a powerful tool to redirect T cells to a desired target for immunotherapy, and they are being tested in clinical trials (1, 3, 5, 7, 9, 10, 15, 18, 40, 41). Recent advances in TCR/Ab engineering led to two promising approaches in adoptive cell transfer for cancer therapy. One approach is the ability to modify TCR sequences to increase their affinity for cognate tumor Ag epitopes (4244). For example, various strategies such as phage-display of TCR libraries have led to the generation of TCR variants with supraphysiologic binding affinity (4nM–26pM) toward epitopes derived from NY–ESO-1 or HTLV-1 (43). The second major advance was the ability to generate high-affinity TCR-like Abs that bind the HLA-peptide complex with high affinity and specificity in the low nanomolar affinity range (2730). However, in contrast to native TCRs that require further affinity engineering and sequence manipulation, these are native Abs made by variable domain recombination using Ab phage display libraries or native Ab germ-line sequences made by hybridoma approaches from immunized mice.

This study has aimed to compare directly, to our knowledge for the first time, engineered T cells that carry two forms of HLA-peptide complex-based recognition moieties: a naturally cloned αβ-TCR versus a corresponding recombinant TCR-like Ab–based CAR, both recognizing the same antigenic epitope derived from TAA WT1. The ability of engineered T cells based on the cloned αβ-TCR strategy to recognize and specifically kill tumor cells expressing the desired target has been demonstrated previously (611). In this study, we investigated the transduction efficiency of the two constructs into primary human T cells and compared the biological functions of the transduced cells.

To generate a chimeric Ag receptor based on TCR-like Ab fragments, two TCR-like Fab Abs were isolated that target the HLA-A2-WT1Db126 complex with affinities of 400 and 30 nM. These TCR-like Fabs exhibited high specificity by their ability to discriminate between HLA-A2 complexes displaying the specific WT1Db126 peptide and HLA-A2 displaying irrelevant control peptides. Furthermore, the WT1-specific TCR-like Abs recognized tumor cells that displayed the naturally processed and presented WT1 epitope in the context of HLA-A2. Recognition was peptide- and HLA-specific, which merits further evaluation of these TCR-like Abs for Ab-based tumor targeting using Ab arming strategies, such as Ab-drug conjugates (45) or T cell engagement approaches through the use of Ab-bispecific constructs with anti-CD3 or anti-CD16 (46).

Using these TCR-like Fabs in the context of CARs, as designed by Eshhar et al. (38), revealed that retroviral transduction of the high-affinity F3 TCR-like Ab (30 nM) into HLA-A2–positive cells induced massive cell death of transduced cells, with a small fraction of viable cells expressing the receptor. We hypothesize that this CAR lost its specificity and mediated nonspecific killing of neighbor lymphocytes by recognizing HLA-A2 complexes regardless of their presented peptides. It is likely that the avidity of the chimeric receptors is increased because they are expressed on the cell surface at relatively high levels. This high avidity, combined with the high affinity of this receptor, affects the ability of the receptor to maintain its specificity, leading to a leakage in its discrimination ability (i.e., cross-reactivity with HLA-A2 complexes presenting irrelevant peptides). This notion is supported by transduction experiments in HLA-A2–negative T cells in which most of the cells were viable and a fraction expressed the CAR. In this case, because of an absence of HLA-A2 complexes on the surface of the cells, no cytotoxic activity was observed by transduced T cells that expressed the 30-nM F3 TCR-like CAR. Additional supporting evidence includes our previous finding demonstrating that HLA-A2–specific TCR-like Abs can cross-react with HLA-A2 molecules presenting other HLA-A2-restricted peptides, but not with other HLA alleles (unpublished data). Studies with affinity-matured native TCRs demonstrated that although some of the identified affinity enhanced variants showed superior T cell function, the increase in affinity oftentimes led to a loss of target cell specificity (44, 47).

The data presented herein are supported by a recent study assessing the relationship of TCR affinity, TCR–pMHC binding parameters, and T cell function. This study tested a panel of sequence-optimized HLA-A2/NY–ESO-1 specific TCR variants with affinities lying within physiologic boundaries to preserve antigenic specificity and avoid cross-reactivity, as well as two variants with a very high and low affinities. Primary human CD8 T cells transduced with these TCRs demonstrated robust correlations between binding measurements of TCR affinity and avidity and the biological response of the T cells, such as TCR cell-surface clustering, intracellular signaling, proliferation, and target cell lysis (48). Strikingly, and as observed by us herein, above a threshold of TCR–pMHC affinity (KD < ∼5 μM), T cell function could not be further enhanced, revealing a plateau of maximal T cell function, compatible with the notion that multiple TCRs with slightly different affinities participate equally (codominantly) in immune responses. Similar to our present study, TCRs displaying affinity above the defined threshold exhibited nonspecific recognition by the TCR. Our observation with the TCR-like Ab–based CAR is consistent with this study and indicates a functional threshold for optimal affinity of engineered TCR or Ab-based CARs. The work performed on the NY–ESO-1 αβ-TCR CARs defined the upper affinity limit of TCR for specific Ag recognition as ∼280–450 nM (44, 48).

The affinity of the F2 TCR-like CAR as soluble monovalent Ab fragment was 400 nM, at the uppermost border of this limit. The data presented herein for the WT1 TCR-like Ab–based CAR demonstrate that specificity for the WT1 peptide epitope was maintained. However, in other studies that examined high-affinity TCR-based CARs alterations in peptide specificity were observed (48). Thus, the possibility that high-affinity TCR or TCR-like Ab–based CARs may still bind to other self-peptide complexes at lower affinity and densities compared with the specific peptide cannot be excluded. Each CAR specificity should be examined carefully as these properties are crucial for clinical applications.

Based on the transduction data of the F3 (30 nM) and F2 (400 nM) TCR-like CARs, we further compared the F2 TCR-like Ab CAR with the cloned αβTCR receptors. Each one of the receptors was introduced into human primary T cells by retroviral transduction. The expression of functional TCR or TCR-like CAR on the transduced cells was measured by specific WT1 tetramer binding and revealed that expression of functional TCR-like CAR was higher compared with the TCR (38% versus 12%). Despite the difference, these data were reproducible and consistent, enabling the comparison of their ability to activate and mediate the biological functions of the transduced T cells. We found that both constructs were capable of properly activating the engineered T cells as indicated by intracellular cytokine expression; however, cytokine secretion assays revealed that the αβ-TCR construct was more sensitive to peptide target concentrations. These results were surprising because the 400-nM F2 TCR-like CAR exhibited higher affinity than the native TCR did, which appears to have an affinity of 1 μM. The higher-affinity receptor was expected to bind to lower concentrations of Ag than the native TCR (i.e., higher levels of cytokine secretion at lower target peptide concentrations). Moreover, as indicated above, tetramer staining revealed that the TCR-like CAR was expressed well on the cell surface compared with the TCR, which further strengthen this observation. These results may be explained by impaired T cell activation caused by the high-affinity TCR-like Ab CAR with agreement with the serial triggering model. This model suggests that for efficient T cell activation, multiple TCRs on the cell surface should sample the pMHC, a process depending on adequately short dissociation rates of the receptor–ligand interaction (32). Therefore, the high affinity of the TCR-like CAR may have reduced the dissociation rates leading to Ag sensitivity reduction. Similarly, a recently published study using affinity-matured TCRs demonstrated that TCRs displaying affinity above the physiological range can exhibit reduced sensitivity to their corresponding ligand (49).

The cytotoxic activity of the transduced engineered cells, whether tested as intact unseparated population of cells or after separation to CD8+ and CD4+ subpopulations, supported the aforementioned data. The TCR-like Ab–based CAR transduced cells exhibited a reduced cytotoxic activity compared with the αβ-TCR transduced cells, which exhibited significant and specific cytotoxic activity. Both transduced T cells expressing the αβ-TCR or the CAR exhibited specific killing of target cells expressing the specific WT1 epitope, and they did not recognize cells presenting an irrelevant epitope of WT1 nor Ag-negative tumor cells. Finally, CD8 is not likely to play the same role in the CAR transduced T cells as it does with the conventional TCR/CD3 complex. This also can explain why the higher affinity CAR did not operate with the sensitivity of the TCR.

The major conclusion from this work is that high-affinity TCR-like Abs are not suitable for optimal construction of CARs designed to retarget T cells. Another important aspect of this work is the relationships between affinity and avidity when comparing T cell and Ab-based immunotherapeutic approaches. The affinity of an Ab or TCR is defined as the binding strength of a single molecule to a cognate Ag. Distinguishably, TCR binding avidity is defined as the binding strength of multiple cell surface TCRs to their respective Ag, whereas the avidity of soluble recombinant Abs may be monovalent or bivalent. Hence, when using a TCR-like Ab as a CAR, the natural avidity of the soluble Ab is out of context from its natural properties. Therefore, all the potential effects arising from the cellular context including cell surface receptor density, MHC coreceptors, and T cell activation state, should also be considered when constructing high-affinity TCR-like Ab–based CARs (32). The overall sensitivity of T cells to Ag density is termed “functional avidity,” and it includes the relative affinity of the TCR–pMHC interaction and the subsequent efficiency of the downstream signal transduction. Because “functional avidity” refers to the actual sensitivity of the cellular response to pMHC Ag density, all of which are aspects of a TCR binding to pMHC (i.e., kinetic constants, avidity, relative affinity), have direct implications for efficient T cell function. Studies have demonstrated that CTLs that exhibit high functional avidities, and thus higher sensitivity, are more effective and therefore essential for antitumor response (50). In this study, we showed that the high-affinity TCR-like Ab in a monovalent or bivalent context (data with IgG not shown) maintained high specificity. However, when the avidity of this TCR-like Ab was increased to a high level, as expressed on the surface of the engineered T cells as a CAR, it had significant consequences on T cell functions. We have isolated two Fabs with high and moderate affinity toward the WT1 epitope. Perhaps the isolated Fabs bind preferentially to WT1, but still cross-react with lower affinity for other self-peptides. Other Fabs against other targets may be better in this regard, and such studies should be expanded to have a more complete picture on the relationships among affinity, avidity, and specificity.

In summary, this study strongly supports the notion that optimal TCRs or TCR-based constructs possess a threshold affinity beyond which no improvement in T cell function is achieved. When TCR-based constructs or TCR-like Ab affinity is enhanced to high and supraphysiologic affinities, engineered T cells carrying such CARs react with many different pMHCs and may change T cell functional properties, leading to unresponsiveness or in some cases loss of Ag specificity, also leading to dangerous cross-reactivity. Therefore, rational design of improved TCRs to generate optimal CARs may need to be optimized up to a given affinity threshold to achieve optimal T cell function without risking T cells to undergo unresponsiveness or cross-reactivity. In this context, TCR-like Ab–based CARs are less suitable and attractive than native αβ-TCR for the design of CARs; however, they are most suitable as bivalent or monovalent Ab forms to be used for Ab-based tumor targeting when coupled with a potent effector function, such as drugs (i.e., Ab–drug conjugates) or T cell engagers in the form of bispecific Ab-based molecules.

This work was supported in part by grants from the European Union Framework Project 6 Attack Consortium and from the Israel Science Foundation.

Abbreviations used in this article:

Fv

variable fragment

pMHC

peptide-MHC

scFv

single-chain Fv

scMHC

single-chain MHC

TAA

tumor-associated Ag.

1
Rosenberg
S. A.
,
Restifo
N. P.
,
Yang
J. C.
,
Morgan
R. A.
,
Dudley
M. E.
.
2008
.
Adoptive cell transfer: a clinical path to effective cancer immunotherapy.
Nat. Rev. Cancer
8
:
299
308
.
2
Dudley
M. E.
,
Wunderlich
J. R.
,
Robbins
P. F.
,
Yang
J. C.
,
Hwu
P.
,
Schwartzentruber
D. J.
,
Topalian
S. L.
,
Sherry
R.
,
Restifo
N. P.
,
Hubicki
A. M.
, et al
.
2002
.
Cancer regression and autoimmunity in patients after clonal repopulation with antitumor lymphocytes.
Science
298
:
850
854
.
3
Besser
M. J.
,
Shapira-Frommer
R.
,
Treves
A. J.
,
Zippel
D.
,
Itzhaki
O.
,
Hershkovitz
L.
,
Levy
D.
,
Kubi
A.
,
Hovav
E.
,
Chermoshniuk
N.
, et al
.
2010
.
Clinical responses in a phase II study using adoptive transfer of short-term cultured tumor infiltration lymphocytes in metastatic melanoma patients.
Clin. Cancer Res.
16
:
2646
2655
.
4
Rosenberg
S. A.
,
Yang
J. C.
,
Sherry
R. M.
,
Kammula
U. S.
,
Hughes
M. S.
,
Phan
G. Q.
,
Citrin
D. E.
,
Restifo
N. P.
,
Robbins
P. F.
,
Wunderlich
J. R.
, et al
.
2011
.
Durable complete responses in heavily pretreated patients with metastatic melanoma using T-cell transfer immunotherapy.
Clin. Cancer Res.
17
:
4550
4557
.
5
Restifo
N. P.
,
Dudley
M. E.
,
Rosenberg
S. A.
.
2012
.
Adoptive immunotherapy for cancer: harnessing the T cell response.
Nat. Rev. Immunol.
12
:
269
281
.
6
Cohen
C. J.
,
Zheng
Z.
,
Bray
R.
,
Zhao
Y.
,
Sherman
L. A.
,
Rosenberg
S. A.
,
Morgan
R. A.
.
2005
.
Recognition of fresh human tumor by human peripheral blood lymphocytes transduced with a bicistronic retroviral vector encoding a murine anti-p53 TCR.
J. Immunol.
175
:
5799
5808
.
7
Xue
S. A.
,
Gao
L.
,
Hart
D.
,
Gillmore
R.
,
Qasim
W.
,
Thrasher
A.
,
Apperley
J.
,
Engels
B.
,
Uckert
W.
,
Morris
E.
,
Stauss
H.
.
2005
.
Elimination of human leukemia cells in NOD/SCID mice by WT1-TCR gene-transduced human T cells.
Blood
106
:
3062
3067
.
8
Stanislawski
T.
,
Voss
R. H.
,
Lotz
C.
,
Sadovnikova
E.
,
Willemsen
R. A.
,
Kuball
J.
,
Ruppert
T.
,
Bolhuis
R. L.
,
Melief
C. J.
,
Huber
C.
, et al
.
2001
.
Circumventing tolerance to a human MDM2-derived tumor antigen by TCR gene transfer.
Nat. Immunol.
2
:
962
970
.
9
Morgan
R. A.
,
Dudley
M. E.
,
Wunderlich
J. R.
,
Hughes
M. S.
,
Yang
J. C.
,
Sherry
R. M.
,
Royal
R. E.
,
Topalian
S. L.
,
Kammula
U. S.
,
Restifo
N. P.
, et al
.
2006
.
Cancer regression in patients after transfer of genetically engineered lymphocytes.
Science
314
:
126
129
.
10
Johnson
L. A.
,
Morgan
R. A.
,
Dudley
M. E.
,
Cassard
L.
,
Yang
J. C.
,
Hughes
M. S.
,
Kammula
U. S.
,
Royal
R. E.
,
Sherry
R. M.
,
Wunderlich
J. R.
, et al
.
2009
.
Gene therapy with human and mouse T-cell receptors mediates cancer regression and targets normal tissues expressing cognate antigen.
Blood
114
:
535
546
.
11
Robbins
P. F.
,
Morgan
R. A.
,
Feldman
S. A.
,
Yang
J. C.
,
Sherry
R. M.
,
Dudley
M. E.
,
Wunderlich
J. R.
,
Nahvi
A. V.
,
Helman
L. J.
,
Mackall
C. L.
, et al
.
2011
.
Tumor regression in patients with metastatic synovial cell sarcoma and melanoma using genetically engineered lymphocytes reactive with NY-ESO-1.
J. Clin. Oncol.
29
:
917
924
.
12
Uttenthal
B. J.
,
Chua
I.
,
Morris
E. C.
,
Stauss
H. J.
.
2012
.
Challenges in T cell receptor gene therapy.
J. Gene Med.
14
:
386
399
.
13
Gross
G.
,
Waks
T.
,
Eshhar
Z.
.
1989
.
Expression of immunoglobulin-T-cell receptor chimeric molecules as functional receptors with antibody-type specificity.
Proc. Natl. Acad. Sci. USA
86
:
10024
10028
.
14
Eshhar
Z.
2008
.
The T-body approach: redirecting T cells with antibody specificity.
Exp. Pharmacol.
329
342
.
15
Kershaw
M. H.
,
Westwood
J. A.
,
Parker
L. L.
,
Wang
G.
,
Eshhar
Z.
,
Mavroukakis
S. A.
,
White
D. E.
,
Wunderlich
J. R.
,
Canevari
S.
,
Rogers-Freezer
L.
, et al
.
2006
.
A phase I study on adoptive immunotherapy using gene-modified T cells for ovarian cancer.
Clin. Cancer Res.
12
:
6106
6115
.
16
Pule
M. A.
,
Savoldo
B.
,
Myers
G. D.
,
Rossig
C.
,
Russell
H. V.
,
Dotti
G.
,
Huls
M. H.
,
Liu
E.
,
Gee
A. P.
,
Mei
Z.
, et al
.
2008
.
Virus-specific T cells engineered to coexpress tumor-specific receptors: persistence and antitumor activity in individuals with neuroblastoma.
Nat. Med.
14
:
1264
1270
.
17
Kochenderfer
J. N.
,
Wilson
W. H.
,
Janik
J. E.
,
Dudley
M. E.
,
Stetler-Stevenson
M.
,
Feldman
S. A.
,
Maric
I.
,
Raffeld
M.
,
Nathan
D. A.
,
Lanier
B. J.
, et al
.
2010
.
Eradication of B-lineage cells and regression of lymphoma in a patient treated with autologous T cells genetically engineered to recognize CD19.
Blood
116
:
4099
4102
.
18
Porter
D. L.
,
Levine
B. L.
,
Kalos
M.
,
Bagg
A.
,
June
C. H.
.
2011
.
Chimeric antigen receptor-modified T cells in chronic lymphoid leukemia.
N. Engl. J. Med.
365
:
725
733
.
19
Jena
B.
,
Dotti
G.
,
Cooper
L. J.
.
2010
.
Redirecting T-cell specificity by introducing a tumor-specific chimeric antigen receptor.
Blood
116
:
1035
1044
.
20
Cheever
M. A.
,
Allison
J. P.
,
Ferris
A. S.
,
Finn
O. J.
,
Hastings
B. M.
,
Hecht
T. T.
,
Mellman
I.
,
Prindiville
S. A.
,
Viner
J. L.
,
Weiner
L. M.
,
Matrisian
L. M.
.
2009
.
The prioritization of cancer antigens: a national cancer institute pilot project for the acceleration of translational research.
Clin. Cancer Res.
15
:
5323
5337
.
21
Inoue
K.
,
Ogawa
H.
,
Sonoda
Y.
,
Kimura
T.
,
Sakabe
H.
,
Oka
Y.
,
Miyake
S.
,
Tamaki
H.
,
Oji
Y.
,
Yamagami
T.
, et al
.
1997
.
Aberrant overexpression of the Wilms tumor gene (WT1) in human leukemia.
Blood
89
:
1405
1412
.
22
Oji
Y.
,
Ogawa
H.
,
Tamaki
H.
,
Oka
Y.
,
Tsuboi
A.
,
Kim
E. H.
,
Soma
T.
,
Tatekawa
T.
,
Kawakami
M.
,
Asada
M.
, et al
.
1999
.
Expression of the Wilms’ tumor gene WT1 in solid tumors and its involvement in tumor cell growth.
Jpn. J. Cancer Res.
90
:
194
204
.
23
Yamagami
T.
,
Sugiyama
H.
,
Inoue
K.
,
Ogawa
H.
,
Tatekawa
T.
,
Hirata
M.
,
Kudoh
T.
,
Akiyama
T.
,
Murakami
A.
,
Maekawa
T.
.
1996
.
Growth inhibition of human leukemic cells by WT1 (Wilms tumor gene) antisense oligodeoxynucleotides: implications for the involvement of WT1 in leukemogenesis.
Blood
87
:
2878
2884
.
24
Oka
Y.
,
Elisseeva
O. A.
,
Tsuboi
A.
,
Ogawa
H.
,
Tamaki
H.
,
Li
H.
,
Oji
Y.
,
Kim
E. H.
,
Soma
T.
,
Asada
M.
, et al
.
2000
.
Human cytotoxic T-lymphocyte responses specific for peptides of the wild-type Wilms’ tumor gene (WT1 ) product.
Immunogenetics
51
:
99
107
.
25
Gao
L.
,
Bellantuono
I.
,
Elsässer
A.
,
Marley
S. B.
,
Gordon
M. Y.
,
Goldman
J. M.
,
Stauss
H. J.
.
2000
.
Selective elimination of leukemic CD34(+) progenitor cells by cytotoxic T lymphocytes specific for WT1.
Blood
95
:
2198
2203
.
26
Dahan
R.
,
Reiter
Y.
.
2012
.
T-cell-receptor-like antibodies - generation, function and applications.
Expert Rev. Mol. Med.
14
:
e6
.
27
Cohen
C. J.
,
Denkberg
G.
,
Lev
A.
,
Epel
M.
,
Reiter
Y.
.
2003
.
Recombinant antibodies with MHC-restricted, peptide-specific, T-cell receptor-like specificity: new tools to study antigen presentation and TCR-peptide-MHC interactions.
J. Mol. Recognit.
16
:
324
332
.
28
Denkberg
G.
,
Cohen
C. J.
,
Lev
A.
,
Chames
P.
,
Hoogenboom
H. R.
,
Reiter
Y.
.
2002
.
Direct visualization of distinct T cell epitopes derived from a melanoma tumor-associated antigen by using human recombinant antibodies with MHC- restricted T cell receptor-like specificity.
Proc. Natl. Acad. Sci. USA
99
:
9421
9426
.
29
Griffiths
A. D.
,
Williams
S. C.
,
Hartley
O.
,
Tomlinson
I. M.
,
Waterhouse
P.
,
Crosby
W. L.
,
Kontermann
R. E.
,
Jones
P. T.
,
Low
N. M.
,
Allison
T. J.
, et al
.
1994
.
Isolation of high affinity human antibodies directly from large synthetic repertoires.
EMBO J.
13
:
3245
3260
.
30
Sergeeva
A.
,
Alatrash
G.
,
He
H.
,
Ruisaard
K.
,
Lu
S.
,
Wygant
J.
,
McIntyre
B. W.
,
Ma
Q.
,
Li
D.
,
St John
L.
, et al
.
2011
.
An anti-PR1/HLA-A2 T-cell receptor-like antibody mediates complement-dependent cytotoxicity against acute myeloid leukemia progenitor cells.
Blood
117
:
4262
4272
.
31
Davis
M. M.
,
Boniface
J. J.
,
Reich
Z.
,
Lyons
D.
,
Hampl
J.
,
Arden
B.
,
Chien
Y.
.
1998
.
Ligand recognition by alpha beta T cell receptors.
Annu. Rev. Immunol.
16
:
523
544
.
32
Stone
J. D.
,
Chervin
A. S.
,
Kranz
D. M.
.
2009
.
T-cell receptor binding affinities and kinetics: impact on T-cell activity and specificity.
Immunology
126
:
165
176
.
33
Denkberg
G.
,
Cohen
C. J.
,
Segal
D.
,
Kirkin
A. F.
,
Reiter
Y.
.
2000
.
Recombinant human single-chain MHC-peptide complexes made from E. coli By in vitro refolding: functional single-chain MHC-peptide complexes and tetramers with tumor associated antigens.
Eur. J. Immunol.
30
:
3522
3532
.
34
Bronner
V.
,
Denkberg
G.
,
Peled
M.
,
Elbaz
Y.
,
Zahavi
E.
,
Kasoto
H.
,
Reiter
Y.
,
Notcovich
A.
,
Bravman
T.
.
2010
.
Therapeutic antibodies: Discovery and development using the ProteOn XPR36 biosensor interaction array system.
Anal. Biochem.
406
:
147
156
.
35
Oved
K.
,
Ziv
O.
,
Jacob-Hirsch
J.
,
Noy
R.
,
Novak
H.
,
Makler
O.
,
Galit
D.
,
Keren
S.
,
Segal
D.
,
Gefen-Dor
C.
, et al
.
2007
.
A novel postpriming regulatory check point of effector/memory T cells dictated through antigen density threshold-dependent anergy.
J. Immunol.
178
:
2307
2317
.
36
Friedmann-Morvinski
D.
,
Bendavid
A.
,
Waks
T.
,
Schindler
D.
,
Eshhar
Z.
.
2005
.
Redirected primary T cells harboring a chimeric receptor require costimulation for their antigen-specific activation.
Blood
105
:
3087
3093
.
37
Eshhar
Z.
,
Waks
T.
,
Bendavid
A.
,
Schindler
D. G.
.
2001
.
Functional expression of chimeric receptor genes in human T cells.
J. Immunol. Methods
248
:
67
76
.
38
Eshhar
Z.
,
Waks
T.
,
Gross
G.
,
Schindler
D. G.
.
1993
.
Specific activation and targeting of cytotoxic lymphocytes through chimeric single chains consisting of antibody-binding domains and the gamma or zeta subunits of the immunoglobulin and T-cell receptors.
Proc. Natl. Acad. Sci. USA
90
:
720
724
.
39
Thomas
S.
,
Xue
S. A.
,
Cesco-Gaspere
M.
,
San José
E.
,
Hart
D. P.
,
Wong
V.
,
Debets
R.
,
Alarcon
B.
,
Morris
E.
,
Stauss
H. J.
.
2007
.
Targeting the Wilms tumor antigen 1 by TCR gene transfer: TCR variants improve tetramer binding but not the function of gene modified human T cells.
J. Immunol.
179
:
5803
5810
.
40
Kessels
H. W.
,
Wolkers
M. C.
,
van den Boom
M. D.
,
van der Valk
M. A.
,
Schumacher
T. N.
.
2001
.
Immunotherapy through TCR gene transfer.
Nat. Immunol.
2
:
957
961
.
41
Kantoff
P. W.
,
Higano
C. S.
,
Shore
N. D.
,
Berger
E. R.
,
Small
E. J.
,
Penson
D. F.
,
Redfern
C. H.
,
Ferrari
A. C.
,
Dreicer
R.
,
Sims
R. B.
, et al
IMPACT Study Investigators
.
2010
.
Sipuleucel-T immunotherapy for castration-resistant prostate cancer.
N. Engl. J. Med.
363
:
411
422
.
42
Dunn
S. M.
,
Rizkallah
P. J.
,
Baston
E.
,
Mahon
T.
,
Cameron
B.
,
Moysey
R.
,
Gao
F.
,
Sami
M.
,
Boulter
J.
,
Li
Y.
,
Jakobsen
B. K.
.
2006
.
Directed evolution of human T cell receptor CDR2 residues by phage display dramatically enhances affinity for cognate peptide-MHC without increasing apparent cross-reactivity.
Protein Sci.
15
:
710
721
.
43
Li
Y.
,
Moysey
R.
,
Molloy
P. E.
,
Vuidepot
A. L.
,
Mahon
T.
,
Baston
E.
,
Dunn
S.
,
Liddy
N.
,
Jacob
J.
,
Jakobsen
B. K.
,
Boulter
J. M.
.
2005
.
Directed evolution of human T-cell receptors with picomolar affinities by phage display.
Nat. Biotechnol.
23
:
349
354
.
44
Robbins
P. F.
,
Li
Y. F.
,
El-Gamil
M.
,
Zhao
Y.
,
Wargo
J. A.
,
Zheng
Z.
,
Xu
H.
,
Morgan
R. A.
,
Feldman
S. A.
,
Johnson
L. A.
, et al
.
2008
.
Single and dual amino acid substitutions in TCR CDRs can enhance antigen-specific T cell functions.
J. Immunol.
180
:
6116
6131
.
45
Reichert
J. M.
,
Dhimolea
E.
.
2012
.
The future of antibodies as cancer drugs.
Drug Discov. Today
17
:
954
963
.
46
May
C.
,
Sapra
P.
,
Gerber
H. P.
.
2012
.
Advances in bispecific biotherapeutics for the treatment of cancer.
Biochem. Pharmacol.
84
:
1105
1112
.
47
Holler
P. D.
,
Chlewicki
L. K.
,
Kranz
D. M.
.
2003
.
TCRs with high affinity for foreign pMHC show self-reactivity.
Nat. Immunol.
4
:
55
62
.
48
Schmid
D. A.
,
Irving
M. B.
,
Posevitz
V.
,
Hebeisen
M.
,
Posevitz-Fejfar
A.
,
Sarria
J. C.
,
Gomez-Eerland
R.
,
Thome
M.
,
Schumacher
T. N.
,
Romero
P.
, et al
.
2010
.
Evidence for a TCR affinity threshold delimiting maximal CD8 T cell function.
J. Immunol.
184
:
4936
4946
.
49
Thomas
S.
,
Xue
S. A.
,
Bangham
C. R.
,
Jakobsen
B. K.
,
Morris
E. C.
,
Stauss
H. J.
.
2011
.
Human T cells expressing affinity-matured TCR display accelerated responses but fail to recognize low density of MHC-peptide antigen.
Blood
118
:
319
329
.
50
Appay
V.
,
Douek
D. C.
,
Price
D. A.
.
2008
.
CD8+ T cell efficacy in vaccination and disease.
Nat. Med.
14
:
623
628
.

Y.R. is a scientific advisor for Applied Immune Technologies. The other authors have no financial conflicts of interest.