Human T cells expressing two additional receptors (TETARs) specific for HIV-1 recognize both epitopes

For an intended cellular therapy approach, T cells can be reprogrammed with a new specificity by T-cell receptor (TCR) transfer. This was performed successfully by retroviral transduction of virus- and tumor-specific TCRs^1,2  or alternatively by transient transfection with TCR-encoding mRNA.^3-8 

Mutation-prone targets such as HIV-1 or tumors can evade immune surveillance by CD8⁺ T cells through introduction of escape mutations within CTL epitopes^9-13  or complete antigen loss,¹⁴  respectively. Therefore, it would be of great advantage to target two epitopes simultaneously. The most direct approach would be to generate T cells expressing two additional receptors (TETARs) specific for different epitopes. Transferred TETARs with two different HIV-1 specificities could still recognize target cells even if one of the CTL epitopes was lost by mutation. To date, dual-specific human T cells have been generated by retroviral transfer of one TCR into T cells with a defined endogenous TCR specificity (eg, CMV, influenza), not to target the pathogen with two TCRs but to provide a constant low-dose triggering to promote T-cell survival in vivo.^15,16 

Here, we generate HIV-1–specific human TETAR by simultaneous introduction of two TCRs by TCR mRNA electroporation. We used TCRs against the HLA-A2–restricted gag epitope SLYNTVATL (SL9)¹⁷  and the HLA-B13–restricted nef epitope RQDILDLWI (RI9).¹⁸  We demonstrate that TETARs are generated by transfer of a human TCR together with a murinized TCR, that both TCRs are functional in the same CD8⁺ T cell, and that TETARs could sequentially kill targets that presented the gag epitope and then targets that presented the nef epitope, or vice versa. To the best of our knowledge, this is the first time that human T cells have been artificially equipped with two new specificities.

CD8⁺ T cells were isolated and cultured as described previously.⁶  The EBV-transformed B-cell line e_0361MS (HLA-A2⁺/HLA-B13⁺) was cultured in R20 medium.¹⁹  The HLA-A2–restricted HIV-1-p17 (gag) SLYNTVATL peptide (SL9) and the HLA-B13–restricted HIV-1-nef RQDILDLWI peptide (RI9) were used.

TCR-encoding sequences were cloned, transcribed in vitro, and transfected as described previously.^4,5  See supplemental Methods for details (available on the Blood Web site; see the Supplemental Materials link at the top of the online article).

The gagTCR was detected with PE-labeled gag/HLA-A2 Dextramer according to the manufacturer's instructions (Immundex). Cell sorting was performed with a FACSAria II.

EBV-transformed B cells were loaded with 2 μg/mL of peptide for 1 hour at 37°C. Stimulations and cytokine detection were performed as described previously.²⁰  Cytolytic activity was tested in standard 4- to 6-hour ⁵¹Cr-release assays as described previously,^5,6  with peptide-loaded B cells as targets. Alternatively, TCR-RNA–transfected T cells were prestimulated for 2 hours with peptide-loaded B cells. The mixture of T and B cells was used as effector cells in a cytotoxicity assay.

To generate TETARs, we electroporated RNAs that encoded a gag/HLA-A2–specific TCR (gagTCR) and an nef/HLA-B13–specific TCR (nefTCR) into CD8⁺ T cells either alone or together at different ratios (Figure 1A). These cells were stimulated with peptide-loaded EBV-immortalized B cells, and cytokine secretion was determined (Figure 1A). T cells transfected with each individual TCR recognized their cognate peptide specifically (Figure 1). In contrast, transfection of both TCRs simultaneously resulted in considerably lower specific cytokine production (Figure 1A). The use of a 5-fold excess of 1 TCR restored recognition of the cognate peptide but further reduced or abolished recognition via the other TCR (Figure 1A). The same was the case for the reverse experiment (Figure 1A). Hence, the mere coexpression of two additional TCRs in one cell does not result in reactivity to both epitopes, which was not expected given the assumption that either TCR can signal independently, and rather indicates that some mechanism of competition or mutual interference exists.

Competitive effects strongly depend on the chosen TCRs (unpublished observation), and here, the nefTCR appeared dominant over the gagTCR. To overcome this problem, we used an excess of the gagTCR but improved the proper pairing and the CD3 binding of the nefTCR by replacing the human constant domains with murine ones,²¹  which resulted in a murinized nefTCR. Consequently, T cells transfected with adapted amounts of mRNA of the human gagTCR and the murinized nefTCR were able to recognize both epitopes with similar efficiency as T cells transfected with only one receptor, as indicated by TNF production (Figure 1B). In contrast, the completely human nefTCR lost its function when combined with an excess of the gagTCR (Figure 1B). These results indicate that it was possible to generate TETARs by cotransfection of a human TCR and a murinized TCR.

To prove that one cell expresses both TCRs, we stained gag- and nef-specific TETARs with a gag/HLA-A2 Dextramer and sorted them by FACS. Those T cells that bound the gag/HLA-A2 Dextramer were subsequently able to recognize target cells loaded with gag peptide, a combination of gag and nef peptide, but also target cells loaded only with the nef peptide, as indicated by IL-2 secretion (Figure 2A). T cells transfected with only the gagTCR recognized only target cells that were loaded with gag peptide or the combination of gag and nef peptide (Figure 2A). These results clearly show that both TCRs were functional in the same cell. This could counteract the immune escape by antigen mutation, because both antigens would have to mutate simultaneously. HIV-1 reverse transcriptase mutates 1 per 1700 nucleotides.²²  In theory, the chance that a 9-mer CTL epitope is altered is roughly 1/100, disregarding amino acids conserved for functional properties; however, the chance that a viral variant arises that bears mutations in 2 recognized epitopes is much lower (ie, 1/100 × 1/100 = 1/10 000). If, in contrast, two single-specific T-cell populations were present, they would not necessarily encounter the target simultaneously, so the mutations could occur slightly after each other, making escape mutant selection presumably more effective. The same would apply to tumor-specific TETARs.

Because T cells transiently lose their ability to bind to their specific epitope after stimulation,²³  we examined whether TETARs could successively kill targets via their two specificities. We stimulated T cells that expressed the human gagTCR, the murinized nefTCR, or both for 2 hours with peptide-loaded B cells. Then, Cr⁵¹-labeled B cells, loaded with the same epitope or the other epitope, were added. T cells transfected with only one TCR displayed an abrogated or reduced lytic activity against their epitope when preincubated with cold targets that presented the same epitope. The same effect was observed with TETARs. In contrast, when TETARs were prestimulated with nef-peptide–loaded targets, they still lysed gag-peptide–loaded B cells. In turn, when TETARs were prestimulated with gag-peptide–loadedtargets, their ability to lyse a second wave of nef-loaded cells was better than that of TETARs prestimulated with nef peptide (Figure 2B-C). These data provide evidence that TETARs obtained the ability to kill target cells via both of the newly introduced receptors and that the down-modulation of a TCR after antigen encounter is a process that acts in cis on a TCR complex but not in trans from one TCR complex to another on the same cell. The fact that TETARs can kill via the second TCR after down-modulation of the first one should also clearly improve their ability to obliterate the intended targets. These findings confirm in the human system the results of Gladow and coworkers,²⁴  who showed that TCR down-modulation on murine transgenic dual-specific T cells depends primarily on binding of the specific ligand, and the results of Hah and coworkers,²⁵  who showed indirectly that tolerization is receptor specific in a related murine model.

To the best of our knowledge, this is the first work describing the reprogramming of human T cells with two new specificities, but whether this approach will result in better in vivo functionality needs to be addressed in further experiments. If not efficient against HIV-1, the treatment of cancer with TETARs appears promising, because tumors can also escape immune surveillance by generation of mutation variants.¹⁴ 

The authors thank Wolfgang Uckert for the murine TCR constant domains; Kris Thielemans for the pGEM4Z-5′UTR-sig-huSurvivin-DC.LAMP-3′UTR vector; Stefanie Baumann, Ina Müller, Tanja Moritz, Verena Wellner, and Kathrin Zitzelsberger for technical assistance; Barbara Schmidt for performing accompanying experiments not included in the manuscript; Kathrin Pritschet and Philipp Schuster for providing reagents; Stefan Schliep, Michael Erdmann, Stina Rosenheinrich, Sandra Schiemann, Margit Lamm, and Doris Schuster for collection of blood; and Katrin Birkholz, Jennifer Etschel, Stefanie Böhm, Christian Krug, Sandra Müller-Schmucker, Isabell Pfeiffer, and Sabrina Prommersberger for fruitful discussions.

Financial support was provided by the ELAN (Erlanger Leistungsbezogene Anschubfinanzierung und Nachwuchsförderung) fund of the Friedrich-Alexander-University of Erlangen-Nuremberg (DE-06.03.29.1), the DFG (German Research Foundation) Graduiertenkolleg 1071 (Viruses of the Immune System, project B1), DFG grant HA 2331/2-1, DFG grant SCHA1247/1-1, the German Competence Network for HIV/AIDS (HIVNET), the Interdisciplinary Center for Clinical Research (IZKF) Erlangen (T.H., project A27), and the Hector Foundation (T.H.).

Contribution: C.H., S.H., A.H., S.B., and E.H. performed research; C.H., G.S., J.D, N.S., and T.H. designed research; and C.H., J.D., and N.S. wrote the manuscript.

Conflict-of-interest statement: The authors declare no competing financial interests.

Correspondence: Niels Schaft, Department of Dermatology, University Hospital Erlangen, Hartmannstraße 14, 91052 Erlangen, Germany; e-mail: niels.schaft@uk-erlangen.de.