Construction and molecular characterization of a T-cell receptor-like antibody and CAR-T cells specific for minor histocompatibility antigen HA-1H


The genetic transfer of T-cell receptors (TCRs) directed toward target antigens into T lymphocytes has been used to generate antitumor T cells efficiently without the need for the in vitro induction and expansion of T cells with cognate specificity. Alternatively, T cells have been gene-modified with a TCR-like antibody or chimeric antigen receptor (CAR). We show that immunization of HLA-A2 transgenic mice with tetramerized recombinant HLA-A2 incorporating HA-1 H minor histocompatibility antigen (mHag) peptides and β2-microglobulin (HA-1 H/HLA-A2) generate highly specific antibodies. One single-chain variable region moiety (scFv) antibody, #131, demonstrated high affinity (KD=14.9 nM) for the HA-1 H/HLA-A2 complex. Primary human T cells transduced with #131 scFV coupled to CD28 transmembrane and CD3ζ domains were stained with HA-1 H/HLA-A2 tetramers slightly more intensely than a cytotoxic T lymphocyte (CTL) clone specific for endogenously HLA-A2- and HA-1 H-positive cells. Although #131 scFv CAR-T cells required >100-fold higher antigen density to exert cytotoxicity compared with the cognate CTL clone, they could produce inflammatory cytokines against cells expressing HLA-A2 and HA-1 H transgenes. These data implicate that T cells with high-affinity antigen receptors reduce the ability to lyse targets with low-density peptide/MHC complexes (~100 per cell), while they could respond at cytokine production level.


A subset of patients with relapsed hematologic malignancies following allogeneic stem cell transplantation (SCT) can be treated with donor lymphocyte infusions (DLI) where graft-versus-tumor (GVT) effect is mediated by donor T cells recognizing minor histocompatibility (H) antigens on tumor cells.1,2 Selective GVT reactivity with minimal risk of graft-versus-host disease is thought to be induced by targeting minor H antigens expressed only on patients’ hematopoietic cells. Among HLA-A*02:01-positive patients, minor H antigens such as HA-1 (ref. 3) and HA-2 have been shown to be associated with antitumor responses with minimal graft-versus-host disease.2 As preparation of T-cell clones or lines for adoptive immunotherapy (ACT) is labor-intensive and time consuming, T-cell receptor (TCR) gene transfer to polyclonal T cells has become an attractive approach for the rapid preparation of large numbers of T cells specific for target antigen of interest.4, 5, 6 However, problems such as induction of unexpected specificities caused by chimeric TCR formation with endogenous TCR or low expression of transduced TCR due to competition with endogenous TCR have been reported.7, 8, 9 These may be overcome by the use of γδT cells instead of αβT cells as effector cells,10 RNA interference specific for endogenous TCR5 or finally TCR gene deletion with designer zinc finger nucleases.11

Another mode of adoptive T-cell therapy is to engineer T cells to express chimeric antigen receptors (CARs).12, 13, 14, 15 The prototypical CAR uses a single-chain variable region moiety (scFv) made of the light chain and heavy-chain variable regions of a monoclonal antibody (mAb), an extracellular hinge, a transmembrane and intracellular signaling domains such as CD3ζ chain.13, 14, 15 This artificial structure results in precluding endogenous TCR from forming unwanted chimeric receptors. CARs enable T cells with MHC-independent, antibody-derived specificity and thus may target any surface molecule on target cells. However, conventional CAR strategy has the limitation of only targeting cell surface antigens on target cells. One possible way to attain intracellular antigen targeting with a CAR is to use a TCR-like mAb as a source of binding moiety for conferring antigen specificity to T cells.16, 17, 18 While such TCR-like mAbs can also be administered alone,18 if armed as CAR-T cells, the necessity of frequent administration can be avoided and long-lasting in vivo effect can be expected. Generation of TCR-like mAbs in small animals has been technically challenging due to the high xenogenic immunogenicity of HLA molecules. Screening of non-immunized phage libraries,1920 peptide-loaded MHC (pMHC) immunization,21,22 and their combination18,23,24 also have been explored to produce TCR-like mAbs targeting potentially therapeutic peptides presented on various HLA molecules. Among these reports, a promising scFv mAb against HLA-A2 complex incorporating PR1, a proteinase 3-derived peptide that can induce efficient complement-dependent cytolysis of myeloid leukemia progenitor cells but not normal hematopoietic cells, has been reported.18

In this study, we describe the efficient isolation of #131, a novel scFv mAb that binds with a conformational epitope of HA-1 H/HLA-A*02:01 by combining phage display screening of splenocytes from HLA-A2.1 (HHD) transgenic mice (Tgm) immunized with tetramerized pMHC consisting of HA-1 H peptide, HLA-A*0201 and β2-microglobulin. We hypothesized that the use of HLA-A2.1 (HHD) Tgm would be effective in offsetting xenogenic immunogenicity of HLA-A*02:01 complexes to which the murine T cells have been rendered tolerant during intrauterine development. We chose an HA-1 H minor H antigen epitope (VLHDDLLEA; the other allelic variant encodes VLRDDLLEA) derived from the HMHA1 gene product. The monomeric #131 scFv mAb with high affinity (dissociation constant [KD]=14.9 nM) successfully recognized HA-1 H/HLA-A2 but not HLA-A2 reconstituted with other known HLA-A2-binding peptides including HA-1 R. Next, we characterized primary human CD4 and CD8 T cells modified with #131 and other scFVs with lower KD coupled to the CD28 transmembrane and CD3ζ domains. Most of them stained with HA-1 H/HLA-A2 tetramers as strongly as a cytotoxic T lymphocyte (CTL) clone, EH6,25 specific for endogenously HLA-A2- and HA-1 H-positive cells. Unexpectedly, however, both CD4 and CD8 #131 scFv CAR-T cells required ~100-fold greater antigen density to be pulsed exogenously in order to exert cytotoxicity comparable to the cognate CTL clone, EH6. In addition, antibody-blocking experiments demonstrated that #131 CD8 CAR-T cells were less sensitive to CD8 blockade while they were completely blocked with HA-1 H/HLA-A2 tetramer. CAR-T cells with lower KD (446 nM for clone #4 and 64.3 nM for clone #9) did not exhibit cytotoxic activity superior to #131 CAR-T cells. These data suggest that T cells with higher affinity antigen receptors than TCRs (average KD ranging between 1100 μM) are less able to recognize low-density peptide/MHC antigens as reported in the case of affinity-matured TCR26 or CAR,12 and that particularly CD8+ CAR-T cells may not be necessarily CD8-dependent possibly due to failure to form complexes with CD3.


Isolation of scFv specific for HA-1 H/HLA-A2 complex

A total of 5 × 108 phage clones were initially generated from splenic B cells of three mice. After three rounds of panning, the recovered phages were used to infect Escherichia coli, which were spread on plates containing ampicillin without infection with helper phages (Figure 1a). A total of 144 clones randomly selected from 8.1 × 108 clones that had been recovered after the third panning were tested by parallel ELISA screening for binding to HA-1 H/HLA-A2 but not to MAGEA4/HLA-A2, as an irrelevant control pMHC complex. Clones showing at least 10-fold higher ELISA signal with HA-1H/HLA-A2 compared with MAGEA4/HLA-A2 were further tested and categorized according to their VH CDR3 amino-acid sequences. Among 144 clones screened, 18 (12.5%) showed preferential binding to HA-1/HLA-A2, 137 showed similar binding to both pMHC complexes, and 7 showed reactivity to neither of them (Figure 1b). As shown in Figure 1b, among 18 specific binders to HA-1H/HLA-A2 we identified seven different groups containing 1–8 identical clones per group based on their CDR3 amino-acid sequences. The binding activities of the representative cp3 clones from each group were assessed by ELISA. As shown in Figure 1c, binding of #131 to bottom-fixed HA-1H/HLA-A2 antigens was detected until 64-fold dilution when the cutoff OD value was set to 0.5, while that of #4 and #9 was detected until 16-fold dilution. We also used surface plasmon resonance to measure binding affinity of candidate scFv antibodies. Among three representative clones (#4, #9 and #131) tested, clone #131 bound with best affinity to soluble HA-1H/HLA-A2 monomer, mainly because of its balanced association and dissociation rate, with a dissociation constant (KD) of 14.9 nM (Figure 1d), compared with KD of 446 nM for clone #4 and 64.3 nM for clone #9, respectively. We also tested binding affinity to MAGE3A/HLA-A2, but we were unable to detect any resonance(data not shown). Collectively, we chose clone #131 for further study.

Figure 1

Schematic representation of the generation and selection of HA-1 H/HLA-A2-specific scFv mAb. (a) HLA-A2.1 (HHD)-transgenic mice were subcutaneously immunized with suspension composed of 50 μg HA-1 H/HLA-A2-Db tetramers emulsified with Montanide ISA51 adjuvant at 7-day intervals. One week after the third immunization, splenocytes were harvested and B-lymphocyte-rich fractions were used to construct a library by a phage-display system. The phage library was then concentrated by three-round panning against coated HA-1 H/HLA-A2 in the presence of irrelevant MAGEA4/HLA-A2 monomers. (b) The table shows the total number of screened scFv cp3 clones after the panning and their specificities. (c) The binding activities of the representative cp3 clones assessed by ELISA. Binding of the seven clones to bottom-fixed HA-1 H/HLA-A2 antigens was assessed by twofold serial dilutions down to 128-fold. (d) BIAcore surface plasmon resonance of selected clones. BIAcore measurements were performed to measure binding affinity of candidate-tetramerized scFv mAbs as described in the Materials and Methods. HA-1 H, HMHA1 histocompatibility (minor) HA-1-derivied peptide (amino-acid position 153–161), VLHDDLLEA. Note that strong spikes were recorded at buffer exchange due to the high sensitivity of BIAcore 3000 instrument.

Characterization of #131 scFv specificity and affinity

To further confirm #131 specificity, we examined whether monomeric #131 scFv could distinguish peptides presented by cell surface HLA-A2 molecules on T2 cells at a peptide concentration of 10 μM by using flow cytometry. As shown in Figure 2a, #131 scFv monomer detected T2 cells pulsed with HA-1H (bold solid line) but not other peptides that bound HLA-A2, such as Flu-A M1, MAGEA4, HBV Env and HER2/neu. On the other hand, T2 cells pulsed with HA-1R (human HA-1H counterpart) were weakly stained (dashed line) while those pulsed with 1 μM HA-1R were not stained (data not shown). The stabilized expression level of HLA-A2 on T2 cells loaded with the above-mentioned peptides was verified as increased MFI with anti-HLA-A2 mAb (Figure 2b). Interestingly, HA-1R peptide was almost incapable of stabilizing HLA-A2 molecule on T2 cells, confirming the previous report that HA-1R is a ~2 log inferior binder to HLA-A2 compared with HA-1H.3 These data suggested that under exogenously pulsed conditions, monomeric #131 binds to efficiently HA-1H peptide and weakly to HA-1R peptide presented by HLA-A2 molecules. Furthermore, monomeric #131 binding was concentration-dependent against T2 cells labeled at fixed concentration (10 μM) of HA-1H peptide (Figure 2c) and also on the concentration of HA-1H peptide added to T2 cells (Figure 2d). Use of a fluorescence amplification increased the detection limit of HA-1H presented on T2 from 1 μM to 0.1 μM, suggesting that the monomeric #131 scFv did bind to HA-1H peptide/HLA-A2 complexes at concentrations as low as 0.1 μM (see Supplementary Figure 1).

Figure 2

Characterization of #131 scFv specificity and affinity. (a) Specificity was evaluated by incubating T2 cells with 10 μM of the indicated peptides for 1 h, and cells were stained with PE-conjugated #131 scFv monomer and analyzed by flow cytometry. HA-1R, H151R polymorphic version of HA-1 H; Flu-A M1, influenza A matrix protein58-66; HBV Env, Hepatitis B virus envelope183-191. Non-pulsed T2 cells is shown as a shaded cytogram. Representative data from two independent experiments are shown. (b) MHC stabilization assay. T2 cells were pulsed with each of the indicated peptides (10 μM) at 27 °C overnight, followed by incubation at 37 °C for 3 h. Surface HLA-A2 molecules were then stained with Alexa 647-conjugated HLA-A2-specific mAb and analyzed by flow cytometry. Non-pulsed T2 cells is shown as a shaded cytogram. Representative data from two independent experiments are shown. (c) Binding characterization of #131 scFv monomer. Titrated concentrations of the PE-labeled #131 scFv monomer were tested with T2 cells pulsed with 10 μM HA-1 H peptide at 37 °C for 2 h. Representative data from two independent experiments are shown. (d) Sensitivity of #131 scFv monomer. T2 cells were incubated with the indicated concentrations of the HA-1 H or HA-1R peptides for 2 h and used for the binding assay with #131 scFv monomer. Representative data from one of at least two independent experiments are shown.

To further evaluate the affinity of the monomeric #131 scFv, we quantified HA-1H/HLA-A2 molecules on T2 cells as antibody-binding capacity (ABC) as previously reported.27 By interpolation on the calibration curve (Figure 3a) and the obtained mean fluorescence intensity (MFI) for individual T2 cells pulsed with HA-1H peptide diluted serially by 10-fold from 10 μM to 0.01 μM (Figure 3b), we could determine specific antibody-binding capacity (SABC) as number of cell surface antigens. For example, T2 cells pulsed with 10 μM HA-1H peptide possessed 29 815 HA-1H peptide/HLA-A2 complexes per cell with MFI shift from 24.67 (control T2 cells) to 176.81 (Figure 3b). On the other hand, using an MHC stabilization assay, the MFI shift assessed by anti-HLA-A2 mAb binding was from 75.5 to 157 (Figure 2b). These data indicate that the monomeric #131 scFv did not fail to bind to HA-1H-bound, stabilized HLA-A2 complexes. The peptide concentration detection limit of 0.1 μM corresponding to 1963 antigens per cell is not necessarily low when considering a report from Wang et al.28 demonstrating that the mean ABC value for CD20 on normal B cells is ~140 000 and that on B-cells from chronic lymphocytic leukemia is 22 000.

Figure 3

Quantification of copy number of HA-1 H/HLA-A2 complexes per cell. (a) Calibration curve using QIFI kit (DAKO). Beads coated with well-determined quantities of mouse mAb were detected with F(ab)2 fragment of FITC-conjugated goat anti-mouse IgG. The MFI and indicated mouse IgG molecules per bead (as antigen-binding capacity, ABC) were plotted. (b) Detection of HA-1 H/HLA-A2 complexes on T2 cells pulsed with 10-fold serially diluted HA-1 H peptide. After 2-h incubation with indicated concentration of HA-1H peptide, T2 cells were first incubated with saturating concentration of #131 scFv mAb and then stained with FITC-conjugated anti-mouse IgG mAb. These procedures were performed in parallel with bead preparation as in (a). (c) Estimation of specific antigen-binding capacity of T2 cells pulsed with HA-1H. The obtained MFI was interpreted to ABC using the calibration curve, and then SABC was calculated by subtraction of background antigen equivalent (MFI of stained with irrelevant scFv) from observed ABC.

Generation of CAR-T cells from primary human T cells

As retroviral vectors can transfer genes into rapidly dividing cells with high efficiency, we used anti-CD3/anti-CD28 coated magnetic beads29 instead of the rapid expansion protocol30 with anti-CD3 (clone OKT3) for activation and expansion of human primary T cells. Activated T cells (2 × 105) purified from a normal volunteer were infected with a retroviral vector containing #4, #9 or #131 construct (Figure 4a) on days 3 and 4 in parallel. We compared whether the transduced T cells with different constructs would exert any growth advantage or disadvantage. As shown in Figure 4b, not only non-transduced cells but also scFv-28z CAR-transduced T cells (designated to as #4–28z, #9–28z and #131–28z, respectively) demonstrated an average of ~1000-fold expansion after a single cycle of bead stimulation over the course of transduction and growth in 21 days. Addition of IL-7 or IL-15 to the IL-2-supplemented cultures did not change the growth capacity when CD3/CD28 beads29 were used for initial stimulation (data not shown). The percentages and MFI of T cells expressing the individual CAR with anti-CD8 mAb and HA-1H/HLA-A2 tetramers after magnetic bead-based separation for CD4 or CD8 are shown in Figure 4c, while they were not stained with irrelevant HIV/HLA-A2 tetramers (data not shown). Interestingly, the specific MFI value of CD4- and CD8-sorted T cells expressing individual CARs (designated to as #4–28z (CD4), #4–28z (CD8), #9–28z (CD4), #9–28z (CD8), #131–28z (CD4) and #131–28z (CD8), respectively) was slightly stronger than that of the cognate HA-1H/HLA-A2-specific EH6-CTL25 (Figure 4c) when a minimally saturating concentration of HA-1H/HLA-A2 tetramers was used. In addition, of note is that a subset of CD4 T cells tend to be resistant to retroviral transduction (Figure 4c, lower panel), but we used these lines throughout the experiments for this revision.

Figure 4

Retroviral transduction of primary human T cells with various scFv CAR. (a) A diagram of the recombinant retroviral LZRSpBMN-based vector encoding scFv-28z is shown (LTR, long terminal repeat; CD28, part of the extracellular region and all of the transmembrane and intracellular regions of CD28; CD3-ζ, the entire cytoplasmic region of the CD3-ζ molecule). (b) In vitro expansion of T cells following activation with anti-CD3/anti-CD28-coated magnetic beads and transduction with or without scFv CAR vector on day 2 and 3. Data are representative of two independent experiments. (c) Expression levels of individual scFv on sorted CD8 and CD4 T cells were compared with that of TCR on EH6-CTL, which is specific for cognate HA-1H epitope in the context of HLA-A*02:01 by staining with PE-conjugated HA-1H/HLA-A2 tetramer. Mock means that T cells were infected with supernatant from packaging cell line. Data are representative of two independent experiments.

Cytotoxicity of CAR-T cells

We evaluated the cytolytic abilities of six CAR-T cell lines, two mock-transduced T cell lines and EH6-CTL using a 4-h CRA. We conducted peptide reconstitution assays using 51chromium-labeled T2 cells (Figure 5) or K562/A2 cells (Supplementary Figure 2) by pulsing the peptides (HA-1H and HA-1R) for 2 h at room temperature, and then using as targets. The #131–28z (CD8) cells could lyse T2 cells pulsed with as low as 10 nM cognate HA-1H peptide (corresponding to the concentration at the half maximal lysis) while T2 cells pulsed with HA-1H allelic counterpart HA-1R were not recognized. Similar cytotoxic activities were observed for #4–28z (CD4), #4–28z (CD8) and #131–28z (CD4) cells, but the same trend was observed only in #4–28z (CD8) and #131–28z (CD8) cells when target cells were changed to K562/A2 (Supplementary Figure 2). Both #9–28z (CD4) and #9–28z (CD8) cells demonstrated cytotoxicity to not only HA-1H but also HA-1R peptide-pulsed T2 cells, suggesting its lower specificity. The cognate EH6-CTL25 lysed T2 cells pulsed with HA-1H, at concentrations as low as 10pM (its half-maximal lysis was obtained at the concentration of 100 pM), ~2 log lower than that required for the #131–28z (CD8) cells (Figure 4b).

Figure 5

Epitope reconstitution assays. T2 cells were labeled with 51Cr and distributed in 96-well round-bottomed plates, pulsed with serial dilutions of the indicated peptides for 2 h at room temperature, and then used as targets for CD4- or CD8-sorted human primary T cells that had been gene-modified with indicated scFvs and EH6-CTL as cognate CTL in a standard 51Cr release assay (E:T ratio 10:1). Data are representative of two independent experiments. Half-maximal lysis for EH6-CTL and #131-28z (CD8) were shown as dashed line with arrow head.

Although the expression levels of #131–28z (CD8) detected by HA-1H/HLA-A2 tetramer was stronger than that of EH6-CTL (Figure 4c), the minimal HA-1H peptide concentration necessary for T-cell recognition was nearly 2 logs different.

Reactivity to endogenously HA-1H-expressing target cells by selected CAR-T cells

We next studied whether CAR-T cells selected for positive lytic activity specific for HA-1H peptide-pulsed target cells were able to kill target cells that express HA-1H epitope endogenously. To prepare versatile artificial target cells, K562 cells that possessed only HA-1R alleles homozygously were stably transduced with HLA-A*02:01 followed by minigenes encoding either HA-1H or HA-1R 9-mer peptide (designated as to K562/A2/HA-1H, K562/A2/HA-1R, respectively). For this experiment, T2 cells were excluded because they cannot transport cytosolic peptides to endoplasmic reticulum. As shown in Figure 6a, EH6-CTL lysed not only K562/A2 cells pulsed with 10 nM HA-1H but also K562/A2/HA-1H. None of the CAR-T cells (that is, #4–28z (CD8), #4–28z (CD4), #131–28z (CD8) and #131–28z (CD4)) however were able to lyse K562/A2/HA-1H efficiently. In contrast, both #131–28z (CD8) and #131–28z (CD4) cells were able to produce interferon-γ (IFN-γ), tissue necrosis-α (TNF-α) and IL-2 against K562/A2/HA-1H (Figures 6b–d). Of note is that #4–28z (CD4) cells were able to produce large amounts of TNF-α and IL-2, but #4–28z (CD8) cells showed no significant trend. Collectively, although the selected CAR-T cells failed to exert evaluable cytotoxic activity by conventional 4-h CRA, it was found that at least #4–28z (CD4), #131–28z (CD8) and #131–28z (CD4) CAR-T cells were potent to produce inflammatory cytokines in response to endogenously presented HA-1H/A2 complexes.

Figure 6

Recognition of endogenously HA-1H expressing target cells assessed by cytotoxicity and cytokine production. (a) The human primary CD4 or CD8 T cells gene-modified with either #4–28z or #131–28z vectors and cognate EH6-CTL were co-cultured with 51Cr-labeled, HLA-A*02:01-transduced K562 cells (both HA-1 alleles are HA-1 R) stably expressing HA-1 H or HA-1R peptides from corresponding minigene vectors in a 4-h standard chromium release assays at an E:T ratio of 10:1. Positive controls are HLA-A*02:01-transduced K562 cells pulsed with 10 nM HA-1 H peptide. (bd) Cytokine production assessed by ELISA for (b) interferon-γ, (c) tissue necrosis factor-α and (d) interleukin-2. Number of effector T cells and stimulators was 10 000 and 5000 per well, respectively. Positive controls are HLA-A*02:01-transduced K562 cells pulsed with 1 nM HA-1 H peptide. *P<0.01; **P<0.05; NS, not significant.

Blocking studies for CD8 #131–28z CAR-T cells

To this end, we examined the mode of recognition of these two effector cells by blocking experiments (Figure 7). Lytic activity of CD8/#131–28z CAR-T cells (independently generated from positively sorted CD8 T cells at the beginning) against HA-1H-pulsed T2 cells was slightly blocked by CD8 mAb (clone Hit8a) while that of EH6-CTL was profoundly blocked. Anti-HLA-A2 mAb (clone BB7.2) showed partial inhibition to CD8/#131–28z CAR-T cells but not E6-CTL, suggesting that #131 scFv contacts HLA-A2 residues that are close to BB7.2 recognition. HA-1H/HLA-A2 tetramer profoundly inhibited lytic activity of CD8/#131–28z CAR-T cells and EH6-CTL, but its effect was significantly stronger in inhibition of #131 scFv binding (Figure 7). Considering that #131–28z (CD4) CAR-T cells also lysed T2 cells pulsed with 10 nM HA-1H peptide (Figure 5), CD8 coreceptor is unlikely involved in the facilitated recognition.

Figure 7

Inhibition of cytotoxicity by mAbs and tetramers. Chromium-labeled T2 cells were incubated with indicated mAbs or tetramers, and were tested by human primary CD8+ cytotoxic T lymphocytes gene-modified with #131-28z vector at an E:T ratio of 1:1. Pre-determined concentrations of mAbs (typically 10 or 20 μg ml−1) or tetramers (10 μg ml−1) were used. The mAbs used for this purpose were against the following antigens: HLA-A2 (BB7.2), CD4 (RPA-T4), CD8 (HIT8a). *P<0.05.


In this study we describe the first scFv that binds specifically to a complex formed by a minor H antigen peptide, HA-1H, and the HLA-A*02:01 molecule. It has been suggested that the majority of clones isolated from conventional phage mAb libraries may not be of high affinity because they should be naive to the antigens.31 Thus, we used a phage mAb library that was constructed from HLA-A2.1-Tgm B cells that had been immunized with HA-1H/HLA-A2 tetramer. We hypothesized that use of HLA-A2.1-Tgm would facilitate efficient induction of high-affinity mAbs because B cells were tolerant to xenogeneic HLA-A2 complexes presenting murine endogenous peptides. We also incorporated a panning procedure in the presence of irrelevant MAGEA4/HLA-A2 monomers to block binders to immunogenic streptavidin and peptide/HLA-A2 molecules other than HA-1H/HLA-A2 in addition to depletion of streptavidin binders originally used by Cohen et al.32 In the end we found that 12.5% (18 out of 144) of the resultant phage mAbs showed preferential binding to HA-1H/HLA-A2 and one of these, clone #131, showed a high binding affinity (KD =14.9 nM) to HA-1H/HLA-A2. Sergeeva et al.18 reported that 78% of hybridoma clones generated were HLA-A2-specific and only 1 out of 2850 clones recognized PR1/HLA-A2, indicating our screening approach was capable of concentrating targeted clones. As expected, monomeric #131 scFv did not bind to HA-1R (due to its low affinity to HLA-A2) or irrelevant HLA-A2-restricted peptides on T2 cells, but unexpectedly, did bind to T2 cells pulsed exogenously with HA-1Q, a nonameric peptide corresponding to murine Hmha1 (Supplementary Figure 3a). The binding to HA-1Q/HLA-A2 was reflected as cytotoxicity against T2 cells pulsed with HA-1Q by CD8/#131–28z CAR-T cells (Supplementary Figure 3b); however, EH6-CTL did not lyse T2 cells even pulsed with 1 μM HA-1Q (Supplementary Figure 3c). In addition, murine thymoma cell line, EL4 (H-2Db background), expressing HLA-A2 and human β2-microglobulin was not lysed at all by CD8/#131–28z CAR-T cells (Supplementary Figure 3d). These findings suggest that HA-1Q peptide can be presented in association with HLA-A*02:01 at a level comparable to HA-1H, but because of its very low cell surface copy number (discussed later) even though endogenously presented, the murine B cells were likely not tolerant and able to mount immune responses when they were challenged by HA-1H/HLA-A2 tetramers. As a result, the generation of high-affinity #131 scFv was eventually possible. Elucidation of the mechanism of cross-reactivity between HA-1H and HA-1Q could require three-dimensional analysis and is beyond the scope of current studies. Nevertheless, it is of note that both use of HLA-A2.1 Tgm for immunization and depletion of potential HLA-A2 binders other than HA-1H/HLA-A2 using irrelevant HLA-A2 monomers resulted in the efficient generation of scFv specific for targeted peptide/HLA complexes. As the degree of polymorphism among HLA-A alleles is low compared with amino acid differences between HLA and H-2 molecules, we surmise that the use of a phage library prepared from HLA-A2.1 Tgm can facilitate the efficient generation of TCR-like mAb specific for peptides presented at least on other HLA-A alleles rather than using non Tgm. This should be also the case when phage libraries from human B cells are to be constructed.

Next we questioned whether endogenously HLA-A*02:01 and HA-1H-expressing cells were susceptible to CAR-T cells when #131 scFv monomer failed to detect them by flow cytometry even using fluorescence amplification. We transduced T cells with #131 scFv with transmembrane and intracellular signaling domain consisting of CD28 and CD3-ζ. Although expression of #131 scFv on #131–28z (CD8) cells assessed by HA-1H/HLA-A2 tetramer staining was slightly stronger than that of HA-1H-specific CTL clone, EH6, the #131–28z (CD8) cells were unable to lyse K562/A2/HA-1H efficiently while they could produce pro-inflammatory cytokines. The failure in cytotoxicity with #131–28z (CD8) cells may be explained by several findings. First, the #131–28z (CD8) cells required 100-fold higher peptide concentration compared with EH6-CTL irrespective of slightly superior HA-1H/ HLA-A2 tetramer binding. Second, CD8 co-receptor was not employed effectively by CD8/#131–28z cells, unlike EH6-CTL as shown by blocking studies. The coreceptor CD8 is thought to contribute to the antigen-recognition process by binding to α3 domain of the MHC class I molecule and by promoting intracellular signaling, serving to enhance TCR stimuli triggered by cognate ligands.33 It is possible that the #131–28z molecule failed to recruit CD8 resulting in insufficient activation of #131–28z (CD8) cells. However, this may not be critical because #131–28z (CD4) cells showed similar cytotoxic activities against HA-1H-pulsed T2 cells. Third, cell surface HA-1H/HLA-A2 complexes have been reported to be ~10 per cell by comparing the signal intensity of the naturally occurring peptide with a known amount of synthetic HA-1H peptide.3 In the case of PR1/HLA-A2, it has been shown that T2 cells pulsed with 1μM PR1 present 10 000 copies per cell, which corresponds to 100 copies per T2 cell pulsed with 10 nM peptides.34 Fourth, in our quantification analysis based on antibody-binding capacity, the detection limit of 10 nM HA-1H on T2 cells corresponded to ~300 copies per cell. It appears that ~100 copies per cell may be the threshold for detection by scFv or probably CAR-T cells when evaluated by cytotoxicity. Thus, it is possible that #131–28z (CD8) cells could not lyse target cells that were expressing HA-1H peptides endogenously at ~10 copies per cell. In terms of cytokine-producing capacity, however, this range may be sufficient even with #131–28z (CD8) cells. If this is the case, it is intriguing to know why EH6-CTL lysed such target cells efficiently. A serial triggering model of T-cell activation has been postulated where one MHC–peptide complex can sequentially engage several TCR molecules to achieve a critical threshold of ~200 triggered TCRs that is required for sufficient T-cell activation (reviewed by Valitutti et al.35) As the affinity of TCR for peptide/MHC complexes is relatively low, with dissociation constants KD ranging from 1–100 μM,36 while that of #131 scFv is 14.9 nM; such a high-affinity receptor expressed on #131–28z (CD8) might preclude the T cells from attaining sequential engagement with the limited number of HA-1H/HLA-A2 complexes. The findings that #4–28z (CD4) CAR-T cells possessing scFv whose KD was 446 nM, 30-fold higher than that of #131, produced large amount of TNF-α and IL-2 may support this hypothesis. In contrast, when target cells expressing a sufficient number of target antigens where a single engagement of individual CAR is enough for T-cell activation, high affinity at 'antibody level' would not limit CAR-T cells to activate and lyse targets. Our data suggest that problems in efficient T-cell activation with high-affinity receptors could occur not only in TCR but also CAR as previously reported by Chmielewski et al.12 More extensive binding studies using other scFv CARs obtained throughout the experiments other than #4 and #9 will address the appropriate range of affinity to low-density cell surface HA-1H/HLA-A2 complexes.

In summary, we have demonstrated the generation of the first CAR-T cells specific for minor H antigen, HA-1H, presented on HLA-A*02:01. Our data confirm that T cells with too-high-affinity CAR paradoxically may fail to get appropriately activated and kill target cells expressing a very low number of target antigens as previously reported12 and that CD8 coreceptor may be dispensable. Further studies with T cells expressing scFv with lower affinity similar to TCR level will be warranted.

Materials and methods


HLA-A2.1 (HHD) Tgm; H-2Db−/ β2m−/− double knockout mice introduced with a human β2mHLA-A2.1 (α1, α2)-H-2Db (α3 transmembrane cytoplasmic; HHD) monochain construct gene were generated in the Département SIDA-Retrovirus, Unite d’ Immunite Cellulaire Antivirale, Institut Pasteur37,38 and kindly provided by Dr FA Lemonnier. The experimental design was approved by the Animal Care Committee of the Aichi Cancer Center Research Institute, and the animals were cared for in accordance with institutional guidelines as well as the Guidelines for Proper Conduct of Animal Experiments.

Blood samples and cell lines

The study design and purpose, after prior approval by the Institutional Review Board of the Aichi Cancer Center, were fully explained, and written consent was obtained from healthy blood donors. Peripheral blood mononuclear cells (PBMCs) were isolated and subjected to B-lymphoid cell line (B-LCL) induction and isolation of CD8+ T cells with a CD8+ T Cell Isolation Kit (Miltenyi Biotec, Bergisch Gladbach, Germany). B-LCLs, T2 (T-cell/B-cell hybridoma deficient for transporter associated with antigen processing), K562 (homozygous for the Arg allele of the HA-1 genotype), EL6 (murine lymphoma, H-2b), EL4S3-Rob-A2-HHD (EL4 transduced with human β2-microglobulin-HLA-A2-Db), 293T, Phoenix-Galv cells and their transfectants were cultured as previously described.39 The HA-1H/HLA-A2-specific cytotoxic T cell clone, EH6-CTL,25 was used as a positive control.

Synthetic peptides

Peptides VLHDDLLEA (HA-1H),3 VLRDDLLEA (HA-1R), VLQDDLLEA (murine HA-1 counterpart, HA-1Q), FLLTRILTI (HVB Env183–191), GILGFVFTL (Influenza A MP58-66, Flu-A M1), GVYDGREHTV (MAGEA4230-239), ALCRWGLLL (HER2/neu5-13) and SLYNTVATL (HIV Gag77-85) were purchased from Bio-Synthesis, Inc. (Lewisville, TX, USA). They were reconstituted in an appropriate solvent and frozen until use.

Generation of pMHC monomers and tetramers

The human β2m-HLA-A2.1 (α1, α2)-H-2Db (α3) single-chain gene construct was PCR-amplified from genomic DNA isolated from HLA-A2.1 (HHD) Tgm spleen cells. For monomer production, the leader sequence of the original construct was removed and substituted with a prokaryote ribosome binding site for expression in E. coli and BamHI site was introduced at the 3′ end of Db exon 4 corresponding to the C terminus of the α3 domain with the primers as follows (EcoRI and BamHI sites for cloning are underlined, respectively): sense 5′-IndexTermATGAATTCTAAGGAGGATATTAAAATGATCCAGCGTACTCCAAAG-3′ and antisense 5′-IndexTermATGGATCCCCATCTCAGGGTGAGGGGCTTGGGCAAACC-3′. The PCR product was cloned into a pHN1+ vector (kindly provided by the late Dr Don C Wiley, Howard Hughes Medical Institute, Harvard University, Cambridge, MA, USA) that had been modified to join a BamHI site followed by the sequence for the biotinylating enzyme BirA, expressed as inclusion bodies in XA-90 E. coli, purified and refolded with HA-1H peptide essentially as described previously.40,41 In addition, for screening purposes, HA-1H and MAGEA3 peptides were refolded with HLA-A2 heavy chain and β2-microglobulin in a conventional manner.41 Part of the resultant pMHC complexes and pMHC incorporating HIV peptide (as negative control) were biotinylated and then converted into tetramers with PE-labeled streptavidin.

Flow cytometry and MHC stabilization assay

The scFv gene-transduced CAR-T cells were incubated with optimized concentration (1–2 μg ml−1) of tetramer at room temperature for 15 min followed by FITC-conjugated anti-CD3 (BD Biosciences, San Diego, CA, USA) and Tricolor anti-CD8 mAb (Caltag, Burlingame, CA, USA) on ice for 15 min. Peptide-pulsed T2, K562/A2 cells or K562 transfectants were stained with PE-labeled #131 monomer (see below) on ice for 15 min. If necessary, the signal of #131 monomer-PE was amplified by FASER Kit PE Amplification (Miltenyi Biotec) according to the manufacturer's instructions. Copy number of HA-1H/HLA-A2 complexes per cell was measured as ABC using the QIFI kit (DAKO, Glostrup, Denmark) as described previously.27 Finally, cells were analyzed with a FACSCalibur flow cytometer and CellQuest software (BD Biosciences).

MHC stabilization assays were performed as described earlier.42 Briefly, T2 cells were pulsed with each of the peptides (10 μM) at 27 °C overnight, followed by incubation at 37 °C for 3 h. Surface HLA-A2 molecules were then stained with Alexa 647-conjugated HLA-A2-specific mAb (clone BB7.2, BD Biosciences). Expression was measured by FACSCalibur (BD Biosciences) and MFI was recorded.

Immunization of mice with HA-1H/HLA-A2-Db tetramers

HLA-A2.1 (HHD) Tgm mice at 6–8 weeks of age were subcutaneously immunized with a 300-μl suspension composed of 50 μg HA-1 H/HLA-A2-Db tetramers emulsified with Montanide ISA51 adjuvant (SEPPIC, Paris, France) at 7-day intervals. One week after the third immunization, splenocytes were harvested and frozen until use.

Phage library preparation and isolation of mAbs against HA-1 H/HLA-A2

A phage library was constructed as previously reported with modifications.43 In brief, frozen spleen cells (~107 cells) from the immunized HLA-A2.1 (HHD) Tgm were thawed and used as gene sources of mAbs. Following isolation of total RNA by the guanidine thiocyanate method, cDNA synthesis was performed using random hexamer and SuperScript III RT First-Strand Synthesis System (Life Technologies, Carlsbad, CA, USA). cDNA was amplified with murine JH primer mix for VH, and random hexamer (Life Technologies) for Vκ and Vλ, respectively, digested with appropriate restriction enzymes (SfiI and XhoI for VH, NcoI and AscI for Vκ and Vλ, respectively), and finally cloned into pSCCA5 vector. Using a phage-display system, the scFv form of a mAb fused to a truncated cp3 (Fab-cp3) was expressed on the phage surface.43

Selection of phage mAbs on biotinylated complexes

Selection of phages exhibiting HA-1 H/HLA-A2 binding activity was performed by a panning method that was essentially the same as that described previously with some modifications.44, 45, 46 Phages were first preincubated with 1% Triton X-100/PBS solution with 100 μg ml−1 streptavidin and 100 μg ml−1 non-biotinylated MAGEA4/HLA-A2 monomers for 1 h to block the streptavidin and HLA-A2 binders, followed by incubation in a Maxisorp Loose plate (Nunc, Roskilde, Denmark) that had been coated with biotinylated HA-1H/HLA-A2 via NeutrAvidin (Life Technologies) for 1 h. Then, the well was washed 10 times with 0.1% Tween-20/PBS, five times with PBS and finally once with high-purity Milli-Q water (Millipore, Billerica, MA, USA). Finally, bound phages were eluted with 50 mM glycine (pH 3.0), and the eluent was immediately neutralized with 1.5 M Tris-HCl, and then infected to E. coli, DH12S cells. The phage clones obtained through this process were used for the next round of panning. After the third round of panning, DH12S cells infected with the selected phages were spread on appropriate agar plate and incubated at 30 °C overnight.

Enzyme-linked immunosorbent assay (ELISA)

To determine the specificity of the mAbs presented on cp3 against HA-1H/HLA-A2, immunoplates (96-well; Thermo Fisher Scientific, Cambridge, MA, USA) were in parallel coated either with HA-1H/HLA-A2 or MAGEA3/HLA-A2 tetramers as previously reported.43 For the reactivity study, the plates were then incubated with 50 ng (or twofold serial dilution for titration ranging from 1 × to 128 × ) of purified Fab-cp3 in PBS(+) at 4 °C overnight. After the plates had been washed with PBS(+), a custom-made rabbit anti-cp3 polyclonal antibody (MBL) was added and incubated at room temperature for 1 h. After washing with PBS(+), peroxidase-conjugated goat anti-mouse IgG (H+L chain; MBL) was added. Finally, bound scFv-cp3 was detected by incubation with OPD substrate (Sigma-Aldrich, St Louis, MO, USA) reading absorbance at 492 nm.

Surface plasmon resonance

Kinetics and affinities of the monomeric HA-1H/HLA-A2 binding to three scFv antibodies (#4, #9 and #131) were measured by surface plasmon resonance using a BIAcore 3000 at Institute for Antibodies Co., Ltd. (Nagoya, Japan). Binding studies were performed at 25 °C using a HBS-EP running buffer (GE Healthcare, Uppsala Sweden) containing 10 mM HEPES, pH7.4, 150 mM NaCl, 3 mM EDTA, 0.005% P20 surfactant. The scFvpp antibodies were captured on CM5 surfaces at densities of 230–1140 RU (resonance units) by amine coupling procedure. The HA-1H/HLA-A2 analyte was diluted to 1,000 nM and tested at 25 nM, 50 nM, 100 nM, 150 nM and 200 nM.

Preparation of monomeric scFv mAbs

The scFv-cp3 molecules were purified with anti-cp3 mAb-conjugated Sepharose beads. After isolation of phage particles, the gene encoding an scFv cp3 molecule was converted into another expression vector(pYA208 ) encoding His-tag and a C-terminal Avi-tag fused form and biotin ligase, transformed into DH12S cells, cultured with 2x YTAIB (ampicillin 200 μg ml−1, IPTG 0.5 mM, biotin 2 μM) medium. As scFv-His-Avi-form antibodies were secreted in the culture supernatant, they were collected by centrifugation, followed by 50% ammonium sulfate precipitation. The pellet was suspended with PBS, applied on Ni-NTA-agarose column, washed with 0.1% Tween-20/PBS, PBS and 10mM imidazole/PBS, and finally eluted with 500mM imidazole/PBS. After several times of dialysis against PBS, they were concentrated by Amicon Ultra (Millipore). Protein concentration was estimated by SDS-PAGE and Coomassie blue staining and western blot analysis was performed with NeutrAvidin-HRP (Thermo Fisher Scientific). Finally, R-phycoerythrin-labeled monomeric scFv Abs were generated with direct conjugation (R-Phycoerythrin Labeling Kit—NH2; DOJINDO, Kumamoto, Japan).

Expression vectors

For expression of scFv on mammalian cells, the variable heavy-chain region (VH) and the variable and constant light-chain region (VL CL) of an Fab fragment were amplified by PCR. After digestion with SfiI and AscI for the VL CL region, the PCR products were subcloned into an original pYA128 vector. For transduction of T cells, an expression cassette was constructed (referred to as #clone-28z, Figure 4a) was generated by combining leader sequence of the murine immunoglobulin kappa light chain, scFv, the transmembrane and truncated intracellular signaling domains of CD28 and CD3-ζ as previously reported by Kochenderfer et al.30 with minor modifications. The cassette was subcloned into LZRSpBMN-Z retroviral vector (a kind gift from Dr G Nolan, Stanford University, Stanford, CA, USA) with modification of packaging signal sequence according to Lee et al.47

Preparation of CAR-T cell lines

The retrovirus producer was prepared by transfecting the construct into a retrovirus packaging cell line, Phoenix-GALV using X-tremeGENE 9 (Roche Applied Science, Mannheim, Germany). CD8 T cells, CD4 T cells or both were positively selected by means of anti-CD8, anti-CD4 magnetic beads (MACS system; Miltenyi Biotec, Auburn, CA, USA). Purified T cells were activated with CD3/CD28 T-cell expander Dynabeads (Life Technologies) at 1:1 ratio for 2 days in Advanced RPMI 1640 (Life Technologies) supplemented with 4% pooled human serum, 2 mM L-glutamine in the presence of 30 U ml−1 Interleukin-2 (IL-2) and 10 ng ml−1 IL-7 (all from R&D Systems, Minneapolis, MN, USA). For the transductions, retroviral supernatant was centrifuged onto 24- or 48-well coated with Retronectin (Takara Bio, Shiga, Japan) plates for 2 h at 2000 g at 32 °C as per manufacturer’s instruction. T cells that had been activated for 2 days as above were then plated, which was repeated on the following day. During transduction, T cells were cultured with 30 U ml−1 IL-2 and 10 ng ml−1 IL-7; after that T cells were expanded with 30 U ml−1 IL-2 with or without 10 ng ml−1 IL-7 or 5 ng ml−1 IL-15 for a total of 10–21 days before assays (that is, 8–19 days after completion of the transduction). For CD4/CD8 subset analyses, gene modified bulk T cells were positively sorted by means of anti-CD8 or anti-CD4 magnetic beads (Miltenyi Biotec).

Chromium release assay

Target cells were labeled with 0.1 mCi of 51Cr for 2 h, and 1 × 103 target cells per well were mixed with CTL at the E:T ratio indicated. After incubation for 4 h at 37 °C, the supernatant radioactivity was measured with scintillation cocktail using MicroBeta2 LumiJET (both from Perkin Elmer, Waltham, MA, USA). All assays were performed at least in triplicate. Percent specific lysis was calculated as follows: ((Experimental cpm−Spontaneous cpm)/(Maximum cpm−Spontaneous cpm))x100. For epitope reconstitution assay, 51Cr-labeled HLA-A*02:01-positive, TAP-deficient T2 cells and K562/A2 were incubated for 2 h in complete medium containing 10-fold serial dilutions of the synthesized peptides and then used as targets in standard 4-h cytotoxicity assays. The blocking experiments were performed using pre-determined concentrations of mAb (typically 10 or 20 μg ml−1) or tetramers (10 μg ml−1). The mAbs used for this purpose were against the following antigens: HLA-A2 (BB7.2), CD4 (RPA-T4) and CD8 (HIT8a) (all from BD Biosciences).

Statistical analysis

The unpaired two-tailed Student's t-test was used for data analysis (, and a P-value <0.05 was considered to be statistically significant.


  1. 1

    Kolb HJ, Mittermuller J, Clemm C, Holler E, Ledderose G, Brehm G et al. Donor leukocyte transfusions for treatment of recurrent chronic myelogenous leukemia in marrow transplant patients. Blood 1990; 76 (12): 2462–2465.

  2. 2

    Marijt WA, Heemskerk MH, Kloosterboer FM, Goulmy E, Kester MG, van der Hoorn MA et al. Hematopoiesis-restricted minor histocompatibility antigens HA-1- or HA-2-specific T cells can induce complete remissions of relapsed leukemia. Proc Natl Acad Sci USA 2003; 100: 2742–2747.

  3. 3

    den Haan JM, Meadows LM, Wang W, Pool J, Blokland E, Bishop TL et al. The minor histocompatibility antigen HA-1: a diallelic gene with a single amino acid polymorphism. Science 1998; 279: 1054–1057.

  4. 4

    Heemskerk MH, Hoogeboom M, de Paus RA, Kester MG, van der Hoorn MA, Goulmy E et al. Redirection of antileukemic reactivity of peripheral T lymphocytes using gene transfer of minor histocompatibility antigen HA-2-specific T-cell receptor complexes expressing a conserved alpha joining region. Blood 2003; 102: 3530–3540.

  5. 5

    Ochi T, Fujiwara H, Okamoto S, An J, Nagai K, Shirakata T et al. Novel adoptive T-cell immunotherapy using a WT1-specific TCR vector encoding silencers for endogenous TCRs shows marked antileukemia reactivity and safety. Blood 2011; 118: 1495–1503.

  6. 6

    Udyavar A, Geiger TL . Rebalancing immune specificity and function in cancer by T-cell receptor gene therapy. Arch Immunol Ther Exp 2010; 58: 335–346.

  7. 7

    Bendle GM, Linnemann C, Hooijkaas AI, Bies L, de Witte MA, Jorritsma A et al. Lethal graft-versus-host disease in mouse models of T cell receptor gene therapy. Nat Med 2010; 16: 565–570.

  8. 8

    Heemskerk MH, Hagedoorn RS, van der Hoorn MA, van der Veken LT, Hoogeboom M, Kester MG et al. Efficiency of T-cell receptor expression in dual-specific T cells is controlled by the intrinsic qualities of the TCR chains within the TCR-CD3 complex. Blood 2007; 109: 235–243.

  9. 9

    van Loenen MM, de Boer R, Amir AL, Hagedoorn RS, Volbeda GL, Willemze R et al. Mixed T cell receptor dimers harbor potentially harmful neoreactivity. Proc Natl Acad Sci USA 2010; 107: 10972–10977.

  10. 10

    van der Veken LT, Hagedoorn RS, van Loenen MM, Willemze R, Falkenburg JH, Heemskerk MH . Alphabeta T-cell receptor engineered gammadelta T cells mediate effective antileukemic reactivity. Cancer Res 2006; 66: 3331–3337.

  11. 11

    Torikai H, Reik A, Liu PQ, Zhou Y, Zhang L, Maiti S et al. A foundation for universal T-cell based immunotherapy: T cells engineered to express a CD19-specific chimeric-antigen-receptor and eliminate expression of endogenous TCR. Blood 2012; 119: 5697–5705.

  12. 12

    Chmielewski M, Hombach A, Heuser C, Adams GP, Abken H . T cell activation by antibody-like immunoreceptors: increase in affinity of the single-chain fragment domain above threshold does not increase T cell activation against antigen-positive target cells but decreases selectivity. J Immunol 2004; 173: 7647–7653.

  13. 13

    Dotti G, Savoldo B, Brenner M . Fifteen years of gene therapy based on chimeric antigen receptors: "are we nearly there yet?". Human Gene Ther 2009; 20: 1229–1239.

  14. 14

    Cartellieri M, Bachmann M, Feldmann A, Bippes C, Stamova S, Wehner R et al. Chimeric antigen receptor-engineered T cells for immunotherapy of cancer. J Biomed Biotechnol 2010; 2010: 956304.

  15. 15

    Park TS, Rosenberg SA, Morgan RA . Treating cancer with genetically engineered T cells. Trends Biotechnol 2011; 29: 550–557.

  16. 16

    Stewart-Jones G, Wadle A, Hombach A, Shenderov E, Held G, Fischer E et al. Rational development of high-affinity T-cell receptor-like antibodies. Proc Natl Acad Sci USA 2009; 106: 5784–5788.

  17. 17

    Weidanz JA, Hawkins O, Verma B, Hildebrand WH . TCR-like biomolecules target peptide/MHC Class I complexes on the surface of infected and cancerous cells. Int Rev Immunol 2011; 30: 328–340.

  18. 18

    Sergeeva A, Alatrash G, He H, Ruisaard K, Lu S, Wygant J et al. An anti-PR1/HLA-A2 T-cell receptor-like antibody mediates complement-dependent cytotoxicity against acute myeloid leukemia progenitor cells. Blood 2011; 117: 4262–4272.

  19. 19

    Chames P, Hufton SE, Coulie PG, Uchanska-Ziegler B, Hoogenboom HR . Direct selection of a human antibody fragment directed against the tumor T-cell epitope HLA-A1-MAGE-A1 from a nonimmunized phage-Fab library. Proc Natl Acad Sci USA 2000; 97: 7969–7974.

  20. 20

    Cohen CJ, Sarig O, Yamano Y, Tomaru U, Jacobson S, Reiter Y . Direct phenotypic analysis of human MHC class I antigen presentation: visualization, quantitation, and in situ detection of human viral epitopes using peptide-specific, MHC-restricted human recombinant antibodies. J Immunol 2003; 170: 4349–4361.

  21. 21

    Bernardeau K, Gouard S, David G, Ruellan AL, Devys A, Barbet J et al. Assessment of CD8 involvement in T cell clone avidity by direct measurement of HLA-A2/Mage3 complex density using a high-affinity TCR like monoclonal antibody. Eur J Immunol 2005; 35: 2864–2875.

  22. 22

    Weidanz JA, Nguyen T, Woodburn T, Neethling FA, Chiriva-Internati M, Hildebrand WH et al. Levels of specific peptide-HLA class I complex predicts tumor cell susceptibility to CTL killing. J Immunol 2006; 177: 5088–5097.

  23. 23

    Andersen PS, Stryhn A, Hansen BE, Fugger L, Engberg J, Buus S . A recombinant antibody with the antigen-specific, major histocompatibility complex-restricted specificity of T cells. Proc Natl Acad Sci USA 1996; 93: 1820–1824.

  24. 24

    Krogsgaard M, Wucherpfennig KW, Cannella B, Hansen BE, Svejgaard A, Pyrdol J et al. Visualization of myelin basic protein (MBP) T cell epitopes in multiple sclerosis lesions using a monoclonal antibody specific for the human histocompatibility leukocyte antigen (HLA)-DR2-MBP 85-99 complex. J Exp Med 2000; 191: 1395–1412.

  25. 25

    Miyazaki M, Akatsuka Y, Nishida T, Fujii N, Hiraki A, Ikeda K et al. Potential limitations in using minor histocompatibility antigen-specific cytotoxic T cells for targeting solid tumor cells. Clin Immunol 2003; 107: 198–201.

  26. 26

    Thomas S, Xue SA, Bangham CR, Jakobsen BK, Morris EC, Stauss HJ . Human T cells expressing affinity-matured TCR display accelerated responses but fail to recognize low density of MHC-peptide antigen. Blood 2011; 118: 319–329.

  27. 27

    Poncelet P, Carayon P . Cytofluorometric quantification of cell-surface antigens by indirect immunofluorescence using monoclonal antibodies. J Immunol Methods 1985; 85: 65–74.

  28. 28

    Wang L, Abbasi F, Gaigalas AK, Vogt RF, Marti GE . Comparison of fluorescein and phycoerythrin conjugates for quantifying CD20 expression on normal and leukemic B-cells. Cytometry B Clin Cytometry 2006; 70: 410–415.

  29. 29

    Milone MC, Fish JD, Carpenito C, Carroll RG, Binder GK, Teachey D et al. Chimeric receptors containing CD137 signal transduction domains mediate enhanced survival of T cells and increased antileukemic efficacy in vivo. Mol Ther 2009; 17: 1453–1464.

  30. 30

    Kochenderfer JN, Feldman SA, Zhao Y, Xu H, Black MA, Morgan RA et al. Construction and preclinical evaluation of an anti-CD19 chimeric antigen receptor. J Immunother 2009; 32: 689–702.

  31. 31

    Kesmir C, Nussbaum AK, Schild H, Detours V, Brunak S . Prediction of proteasome cleavage motifs by neural networks. Protein Eng 2002; 15: 287–296.

  32. 32

    Parker KC, Bednarek MA, Coligan JE . Scheme for ranking potential HLA-A2 binding peptides based on independent binding of individual peptide side-chains. J Immunol 1994; 152: 163–175.

  33. 33

    Laugel B, Cole DK, Clement M, Wooldridge L, Price DA, Sewell AK . The multiple roles of the CD8 coreceptor in T cell biology: opportunities for the selective modulation of self-reactive cytotoxic T cells. J Leuk Biol 2011; 90: 1089–1099.

  34. 34

    Molldrem J, Dermime S, Parker K, Jiang YZ, Mavroudis D, Hensel N et al. Targeted T-cell therapy for human leukemia: cytotoxic T lymphocytes specific for a peptide derived from proteinase 3 preferentially lyse human myeloid leukemia cells. Blood 1996; 88: 2450–2457.

  35. 35

    Valitutti S . The serial engagement model 17 years after: from TCR triggering to immunotherapy. Front Immunol 2012; 3: 272.

  36. 36

    van der Merwe PA, Davis SJ . Molecular interactions mediating T cell antigen recognition. Annu Rev Immunol 2003; 21: 659–684.

  37. 37

    Firat H, Garcia-Pons F, Tourdot S, Pascolo S, Scardino A, Garcia Z et al. H-2 class I knockout, HLA-A2.1-transgenic mice: a versatile animal model for preclinical evaluation of antitumor immunotherapeutic strategies. Eur J Immunol 1999; 29: 3112–3121.

  38. 38

    Pascolo S, Bervas N, Ure JM, Smith AG, Lemonnier FA, Perarnau B . HLA-A2.1-restricted education and cytolytic activity of CD8(+) T lymphocytes from beta2 microglobulin (beta2m) HLA-A2.1 monochain transgenic H-2Db beta2m double knockout mice. J Exp Med 1997; 185: 2043–2051.

  39. 39

    Kondo E, Topp MS, Kiem HP, Obata Y, Morishima Y, Kuzushima K et al. Efficient generation of antigen-specific cytotoxic T cells using retrovirally transduced CD40-activated B cells. J Immunol 2002; 169: 2164–2171.

  40. 40

    Torikai H, Akatsuka Y, Miyazaki M, Warren EH 3rd, Oba T, Tsujimura K et al. A novel HLA-A*3303-restricted minor histocompatibility antigen encoded by an unconventional open reading frame of human TMSB4Y gene. J Immunol 2004; 173: 7046–7054.

  41. 41

    Kuzushima K, Hayashi N, Kudoh A, Akatsuka Y, Tsujimura K, Morishima Y et al. Tetramer-assisted identification and characterization of epitopes recognized by HLA A*2402-restricted Epstein-Barr virus-specific CD8+ T cells. Blood 2003; 101: 1460–1468.

  42. 42

    Khanna R, Burrows SR, Nicholls J, Poulsen LM . Identification of cytotoxic T cell epitopes within Epstein-Barr virus (EBV) oncogene latent membrane protein 1 (LMP1): evidence for HLA A2 supertype-restricted immune recognition of EBV-infected cells by LMP1-specific cytotoxic T lymphocytes. Eur J Immunol 1998; 28: 451–458.

  43. 43

    Higo-Moriguchi K, Akahori Y, Iba Y, Kurosawa Y, Taniguchi K . Isolation of human monoclonal antibodies that neutralize human rotavirus. J Virol 2004; 78: 3325–3332.

  44. 44

    Cohen CJ, Denkberg G, Lev A, Epel M, Reiter Y . 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 2003; 16: 324–332.

  45. 45

    Marks JD, Hoogenboom HR, Bonnert TP, McCafferty J, Griffiths AD, Winter G . By-passing immunization. Human antibodies from V-gene libraries displayed on phage. J Mol Biol 1991; 222: 581–597.

  46. 46

    Iba Y, Ito W, Kurosawa Y . Expression vectors for the introduction of highly diverged sequences into the six complementarity-determining regions of an antibody. Gene 1997; 194: 35–46.

  47. 47

    Lee JT, Yu SS, Han E, Kim S, Kim S . Engineering the splice acceptor for improved gene expression and viral titer in an MLV-based retroviral vector. Gene Ther 2004; 11: 94–99.

Download references


We thank Dr W Ho for critically reading the manuscript; Ms Sayoko Ogata and Ms Hiromi Tamaki for their technical expertise. This study was supported in part by Grant-in-Aid for Scientific Research (C)(24591435), from the Ministry of Education, Culture, Science, Sports and Technology, Japan; Grants for Research on the Human Genome, Tissue Engineering Food Biotechnology and the Second and Third Team Comprehensive 10-year Strategy for Cancer Control, from the Ministry of Health, Labour and Welfare, Japan; and a grant from the Japan Leukemia Research Fund (2013). This study was supported in part by Grant-in-Aid for Scientific Research (C)(24591435), from the Ministry of Education, Culture, Science, Sports and Technology, Japan; Grants for Research on the Human Genome, Tissue Engineering Food Biotechnology and the Second and Third Team Comprehensive 10-year Strategy for Cancer Control, from the Ministry of Health, Labour and Welfare, Japan; and a grant from the Japan Leukemia Research Fund (2013).

Author information

Correspondence to Y Akatsuka.

Ethics declarations

Competing interests

The authors declare no conflict of interest.

Additional information

Supplementary Information accompanies this paper on Gene Therapy website

Supplementary information

Rights and permissions

Reprints and Permissions

About this article

Cite this article

Inaguma, Y., Akahori, Y., Murayama, Y. et al. Construction and molecular characterization of a T-cell receptor-like antibody and CAR-T cells specific for minor histocompatibility antigen HA-1H. Gene Ther 21, 575–584 (2014) doi:10.1038/gt.2014.30

Download citation

Further reading