Enhanced receptor expression and in vitro effector function of a murine-human hybrid MART-1-reactive T cell receptor following a rapid expansion

Abstract

Peripheral blood lymphocytes (PBL) genetically modified to express T cell receptors (TCR) specific to known melanoma antigens, such as melanoma antigen recognized by T cells-1 (MART-1), and gp100 can elicit objective tumor regression when administered to patients with metastatic melanoma. It has also been demonstrated that modifications within the constant regions of a fully human TCR can enhance surface expression and stability without altering antigen specificity. In this study, we evaluated the substitution of murine constant regions for their human counterpart within the DMF5 MART-1-specific TCR. Unlike previous studies, all modified TCRs were inserted into retroviral vectors and analyzed for expression and function following a clinical transduction protocol. PBL were transduced with retroviral supernatant generated from stable packaging lines encoding melanoma-specific TCRs. This protocol resulted in high levels of antigen-specific T cells without the need for additional peptide stimulation and selection. Both the human and murinized TCR efficiently transduced PBL; however, the murinized TCR exhibited significantly higher tetramer binding, mean fluorescence intensity, as well as, increased in vitro effector function following our clinical transduction and expansion protocol. Additional TCR modifications including insertion of a second disulfide bond or the linker modifications evaluated herein did not significantly enhance TCR expression or subsequent in vitro effector function. We conclude that the substitution of a human constant region with a murine constant region was sufficient to increase receptor expression and tetramer binding as well as antitumor activity of the DMF5 TCR and could be a tool to augment other antigen-specific TCR.

Fig. 1

Diagram of gammaretroviral vectors encoding DMF5 TCR modifications. The DMF5 (WT-F5) TCR was cloned into the pMSGV1 gammaretroviral backbone. The modified TCR containing murine constant regions (denoted by the gray shaded box) was designated Mur-F5. The 2A sequences are derived from foot and mouth disease virus (F2A) and Turkey rhinotracheitis virus (T2A)

Fig. 2

Comparison of TCR expression and tetramer binding following transduction with gammaretroviral vectors encoding the WT-F5 or Mur-F5 TCR. Bulk human PBL were transduced with WT-F5 or Mur-F5 retroviral supernatant and placed into culture with media containing low doses of IL-2. No additional specific stimulation was performed. A representative FACS analysis of TCR-transduced PBL under a S1 and b R2 conditions for the respective TCR (gated on CD3⁺/CD8⁺/MART-1 tetramer⁺ cells). c Percent tetramer expression in Mur-F5 TCR-transduced PBL relative to wild-type, expressed as mean ± SEM. All S1 p < 0.001, R2 CD3 p = 0.006, CD4 p = 0.0052, CD8 p = 0.0272. d Mean fluorescent intensity (MFI) of the MART-1 tetramer signal as a measure of surface expression in Mur-F5 TCR-transduced PBL relative to wild type, expressed as mean ± SEM. All p < 0.001. e CD3 MFI of the S1 and R2 CD3⁺/MART-1 tetramer⁺ cells (*p ≤ 0.05). The results of at least six independent experiments

Fig. 3

In vitro functional analysis of TCR-transduced PBL following an initial stimulation (S1) or a rapid expansion (R2). Bulk human PBL were transduced with WT-F5 or Mur-F5 retroviral supernatant and placed into culture with media containing low doses of IL-2. S1 cells were sampled 7–9 days after initial OKT3 stimulation. R2 cells were sampled 7–10 days after rapid expansion with high-dose IL-2, OKT3 and irradiated allogeneic feeders. a A representative assay showing IFNγ release following overnight co-culture with HLA-A2⁺/MART-1⁺ tumor targets (*p < 0.01). Data shown are from the PBL of three separate patients. Cells released no IFNγ against HLA-mismatched controls (data not shown). b The IFNγ ratio of Mur-F5:WT-F5 for S1 and R2 cell populations following overnight co-culture with HLA-A2⁺/MART-1⁺ tumor cell lines. Mur-F5 transduced cells released a median 2.2-fold more IFNγ (range 0.5–5.8). HLA-mismatched lines showed no specific release of IFNγ (data not shown), p values are indicated for both S1 and R2 for the given target cell line with the results from at least six independent experiments. S1 and R2 cells were evaluated for their ability to lyse c HLA-A2⁺/MART-1⁺ (mel624) or d HLA-A2⁻ (mel938) targets at decreasing ratios of effector to target (E:T) ratios. Data shown is from the PBL of three separate patients. No significant differences were observed between both S1 and R2 or F5 and mF5 TCR-transduced cells within or between groups

Fig. 4

Assessment of phenotypic markers of differentiation on WT-F5 and Mur-F5-transduced PBL following and initial stimulation (S1) or a rapid expansion (R2). There were no phenotypic differences between cells transduced with either the WT- or Mur-F5 TCR retroviral vector. However, there were distinct differences between S1 and R2 cells within a given transduced population. All plots are gated on CD3⁺/CD8⁺/MART-1 tetramer⁺ cells, representative of six patients. a S1 cells in both WT- and Mur-F5 cells had a sizeable CD45RA⁺/CD62L⁺ (or T naïve-like population). b After rapid expansion the T naïve-like population was no longer discernible with the majority of cells expressing a central memory (CD45RA⁻/CD62L⁺) or effector-like (CD45RA⁻/CD62L⁻) phenotype. The T cell subsets are described as follows: T _N naïve, T _CM central memory, T _EM effector memory, T _T terminally differentiated

Fig. 5

Evaluation of modifications to the high avidity wild-type DMF5 TCR following a REP. PBL were transduced with a gammaretroviral vector encoding the indicated TCR construct. a TCR modifications indicating the location of cysteine residues and substitution of the T2A linker for the optimized furinSGSGP2A linker are indicated. The P2A sequence was derived from the porcine teschovirus. b Tetramer expression relative to the mF5 TCR of transduced PBL 7 days post-REP as described. c IFNγ release of R2 cells relative to mF5, not normalized for tetramer expression. Cells released no IFNγ against antigen negative, HLA mis-matched controls (data not shown). Data representative of at least three independent experiments (one-way ANOVA, p > 0.05)