Discussion

↓71

Adoptive T cell therapy with TCR-modified T cells is superior to previous cellular immuno-therapy approaches for several reasons. First, generation of a large number of gene-modified lymphocytes is rapid and efficient, taking less than two weeks with the current protocols. Second, tumor-reactive T cells can be produced for every patient regardless of pre-existence of TILs. Third, transfer of TCR genes allows endowing T cells with specificities that can be more efficient than those naturally occurring, e.g. high-affine receptors for auto-antigens.

Most likely, results of several clinical trials will be published in the upcoming years. They will not only allow to assess the efficacy of TCR-redirected T cells in humans, but also to define which possible side effects of the therapy are actually relevant in a patient. From one initial clinical study no severe adverse effects were reported [55 ]. However, the transferred T cells were of low avidity and did not persist long-term in most of the patients [55 ,141 ]. Accordingly, a tumor response was detectable, but limited. Progress in the recent years, however, has made it possible to generate TCR-modified T cells with high-affinity TCRs expressed at a high level [52 ]. Also, culture conditions have been defined to induce long-persistent central memory T cells [138 ,142 ]. Increase in efficiency, however, also bears the risk of enhanced side effects, especially auto-immunity. Certainly, with more clinical data available, the need for an inducible termination of therapy will become more obvious. Currently described safety strategies are for many reasons not suitable for TCR-redirected T cells. In this thesis, a novel safety approach was analyzed, based on a TCR-intrinsic mechanism. This safeguard does not impair the function of the T cells and allows their highly specific and efficient elimination in vivo.

4.1 Generation and expression of myc-tagged TCRs

Initially, nine different myc-tagged P14 TCR were generated, in which the myc-tag was inserted at several positions in the TCR sequence, either as a single or – to increase anti- myc antibody avidity – as a double tag. The chosen sites were selected (i) to protrude from the compact structure of the variable and constant regions to facilitate antibody accessibility and (ii) not likely to be involved in antigen-recognition which might hamper the function of the modified TCR. When the expression of the myc-tagged TCRs was analyzed in a murine T cell line with antibodies specific for both TCR chains, all could be stained on the cell surface. However, for some TCRs (e.g. P14/TCRmyc[L1] or P14/TCRmyc[BN]) the detection level was reduced indicating that myc-tag insertion interfered with the stability of the TCR structure leading to a lower expression rate or changed the epitope of the TCR-specific antibody which thus bound with a lower affinity. Staining with a myc-specific antibody was successful in only four of the nine myc-tagged TCRs. In the others, the incorporation into the TCR structure might have forced the myc-tag into a conformation in which it might not be recognized by the antibody. For the myc-specific antibody clone 9E10 it is known that it requires an extended-loop conformation of the epitope to mediate proper binding (personal communication W. Höhne, Charité, Berlin, Germany). Perhaps, the myc-tag adopts this conformation only in some of the TCRs, but not in others. When the tag was fused to the N-terminus of the TCRα-chain, detection with a myc-specific antibody was augmented when two tags were inserted as compared to one. This indicates that employing two adjacent tags might indeed increase antibody avidity to the T cell.

↓72

The myc-tagged TCR that was recognized best by the myc-specific antibody and later found to mediate sufficient depletion (P14/TCRmyc[DAN]) showed no decrease in expression level. Similar results were obtained when the murine TCR OT-I and the human TCR gp100 were modified with two myc-tags in the same position.

4.2 Depletion of TCRmyc-transduced T cells in vitro

The four myc-tagged P14 TCRs that were detected by a myc-specific antibody were analyzed for their ability to mediate depletion of transduced T cells. Performing a complement depletion assay, significant cell lysis was only achieved by the TCR vectors that carried two myc-tags (P14/TCRmyc[DL] and P14/TCRmyc[DAN]). It is known that the amount of epitopes available for the antibody is crucial for complement-mediated depletion because the cross-linking of several antibody constant regions augments activation of C1q, the initial factor of the classical complement pathway [143 ]. Most likely, the presence of two tags provides more epitopes in close proximity and may facilitate the parallel binding of two antibody molecules which is beneficial for depletion.

Complement depletion of T cells transduced with P14/TCRmyc[DAN] was of higher efficacy compared to T cells expressing P14/TCRmyc[DL]. Therefore, this position of myc-tag insertion was studied further and used to modify additional TCRs. Transduction of human and murine T cells with gp100/TCRmyc[DAN] and OT-I/TCRmyc[DAN], respectively, also allowed efficient elimination of the lymphocytes by treatment with myc-specific antibody and complement. Compared to the murine T cell lines, the percentage of depletion was lower for human PBLs (about 80% versus 65%). One reason for this is that the purity of myc-positive cells after enrichment was generally higher for the T cell lines and that the remaining myc-negative cells in the culture could not be eliminated. Probably using a second sorting step would overcome this problem. Another point may be the influence of T cell-intrinsic factors, such as activation state, cell cycle phase and expression of anti-apoptotic molecules. In this regard, a T cell line comprises a much more homogenous population of cells than PBLs. Furthermore, the capacity of rabbit complement factors to lyse murine T cells might be higher as compared to human. Still, given the time after which depletion was analyzed (two hours), the overall efficacy of T cell lysis by the TCRmyc safety approach is very high. Assays in which the potential of the HSV-TK gene is analyzed are usually performed for several days to achieve similar results [105 ,106 ,144 ].

↓73

To demonstrate that depletion of TCRmyc-transduced T cells can also employ other effector mechanisms, cell-mediated lysis assays were performed with TCRmyc-transduced PBLs as targets and autologous NK cells as effectors. Specific lysis of the PBLs was observed though this was lower as compared to the complement assays. This, however, might be due to the intrinsic property of the antibody to elicit specific effector functions. Ideally, a myc-specific antibody used for depletion of adoptively transferred T cells in a patient should be able to induce several effector mechanisms with a high efficiency (see also chapter 4.6.1).

4.3 Function of TCRmyc-transduced T cells

A prerequisite for any safeguard for adoptive T cell therapy is that is does not interfere with the function of the T cell. Although numerous crystal structures of TCRs or parts thereof have been published, it still remains unclear how exactly signal transduction within the TCR/CD3 complex is managed [3 ]. In this study, the structure of the TCR was genetically modified. This might on one hand change the mode of antigen binding or, on the other hand, have an impact on the conformation of the whole molecule rendering it incapable of signal transduction. Therefore, the functionality of each modified receptor was analyzed. Antigen recognition was assessed by staining transduced primary T cells with fluorescently labeled peptide-MHC tetramers. Both, the number of T cells that bound the tetramer and the intensity of binding were not reduced for the myc-tagged TCRs compared to their wt counterparts suggesting that the myc-tag insertion did not interfere with the recognition of the cognate peptide-MHC complex. Signal transduction was analyzed by measuring the amount of secreted cytokines upon antigenic stimulation of TCRmyc-transduced T cells. For the murine TCRs, the T cell line 58 was employed which secretes detectable amounts of IL-2 upon stimulation. Similarly, in the human system, IFN-γ production by gp100/TCRmyc- transduced primary human T cells was assessed. For every analyzed myc-tagged TCR, the quantity of cytokines upon stimulation with different amounts of antigen was comparable to that of the wt receptor indicating that antigen specificity and signal transduction are not affected by the safeguard.

The finding that function of the T cell is maintained in three different investigated receptors suggests that the position chosen for insertion of the myc-tag might be suitable for most TCRs.

↓74

 

4.4 Depletion of TCRmyc-transduced T cells in vivo

The efficiency of antibody-mediated depletion may differ in vitro and in vivo. First, the local concentration of the antibody, complement components or effector cells can vary. Second, permeability of tissue may be an important issue in vivo. Third, while in vitro only one effector mechanism at a time was analyzed, several may act together in a patient thereby enhancing depletion. Therefore, the possibility to eliminate auto-reactive T cells by the myc-tag safeguard was determined in the RIP-mOVA mouse model. In contrast to RIP-OVAlow mice, these animals have a high expression level of ovalbumin in the pancreas. Hence auto-reactivity is not transient and very rapid. This model was chosen to demonstrate the efficiency of the TCRmyc safety approach under drastic conditions.

Here, auto-immune diabetes by transfer of OT-I TCR-transduced T cells was only inducible by prior total body irradiation of the animals. This is in accordance with data from clinical trials or other mouse models showing that rendering the host lymphopenic provides a proliferative advantage for subsequently transferred cells by eliminating regulatory T cells and other cells competing for a limited pool of cytokines [145 ,146 ,147 ]. For treatment, administration of the myc-specific antibody was required before increase of blood glucose levels were detectable in order to rescue mice from lethal diabetes. If the antibody was injected at a later time-point, prevention of lethality was not possible. In a clinical setting, this pre-emptive treatment, however, is not feasible. Here, T cell therapy would only be terminated in case severe auto-immune side effects become obvious. Still, several facts underline that this necessity is model-specific. Disease onset in RIP-mOVA mice is extremely rapid: as early as day two after adoptive transfer infiltrating lymphocytes were detected in the pancreatic islets (insulitis). As therapeutic anti-myc antibody administration at this time-point was still possible, it can be assumed that the antibody was able to penetrate the pancreatic tissue and eliminate the TCRmyc-modified T cells in situ. Without treatment, blood glucose values increased from normal to highly glycemic (>14 mM) within 24 hours at day four or five after adoptive transfer and mice had to be sacrificed at day six to ten due to severity of symptoms. In a patient, careful analysis of several indicators of auto-immunity should be feasible allowing an earlier time-point of treatment. Also, additional administration of immuno-suppressive drugs – an option which was not studied in the RIP-mOVA model – can support antibody-mediated depletion in case of rapidly progressing auto-reactivity.

↓75

Nevertheless, it would be desirable to analyze the myc-tag safeguard system in a mouse model in which disease onset is slower and early symptoms of auto-reactivity can easily be followed. Then, therapeutic instead of prophylactic treatment should be feasible.

4.5 Advantages of the TCRmyc safeguard over others

In chapter 1.4 the properties of existing safety approaches have been discussed. All of them comprise several drawbacks rendering them more or less inappropriate for the use in adoptive transfer of TCR-modified T cells emphasizing the need to develop a more suitable strategy.

The most compelling advantage of using myc-tagged TCRs as a safeguard is that this strategy is TCR-intrinsic. The expression of the transgene is directly linked to the suicide mechanism and no additional genes need to be transferred. This avoids purification steps after transduction as adverse effects are only expected by TCR-modified T cells which automatically carry the safety switch. Furthermore, downmodulation of suicide genes, which has been observed for example for HSV-TK [77 ,103 ,125 ,148 ]. In the case of TCRmyc, downregulation of the safeguard is coupled to the loss of transgenic TCR expression. As auto-immunity is caused by activation of the introduced TCR chains, their down-regulation most likely also terminates the side effects.

↓76

Immunogenicity, which has been shown to lead to unwanted elimination of cells expressing suicide genes, is unlikely to occur for TCRmyc as the molecule consists entirely of human proteins. Immune reactions against the myc-specific antibody can be avoided by producing partially or fully human antibodies.

A further requirement for a safety system is high specificity and low cytotoxicity meaning the absence of side effects on other cells. Although not yet tested in patients, administration of a myc-specific antibody is not expected to cause tissue damage. First, c-myc is not expressed on the cell surface of a normal cell. Second, as a cell cycle regulator and proto-oncogene its expression is tightly regulated and at a relatively low level in normal cells. Still, it has been shown that the shortest peptide sequence needed to give a strong antibody binding signal of the 9E10 myc-specific antibody clone is KLISEEDL [149 ] and it cannot be excluded that this sequence is part of an MHC-presented peptide.

In sum, many of the obstacles known for other safety systems can be overcome by the use of myc-tagged TCRs. Important issues related to the implementation of the safeguard into a clinical setting and several limitations of the developed approach are discussed in chapters 4.6 and 4.7.

4.6 Implementation of the safeguard into a clinical setting

↓77

Protocols for efficient generation of TCR-modified T cells for clinical use have been successfully established [55 ] and will not be discussed here. This chapter focuses instead on the specific issues related to the applicability of the TCRmyc safety strategy in patients.

4.6.1 Availability of a myc-specific depleting antibody

Three general effector mechanisms of antibodies have been described: (i) blocking of signal transduction of receptor molecules, (ii) depletion by activation of the complement system and (iii) ADCC. Ideally, an antibody used for elimination of auto-reactive T cells in vivo should elicit all of the three responses to achieve a maximum effect. Which of the two last mechanisms is induced, largely depends on the antibody isotype. Human IgG1 and IgG3 antibodies are both capable of recruiting human NK cells by binding to Fc receptors and inducing ADCC as well as initiating a complement cascade. As IgG3 exhibits a shorter half-life, most antibodies in clinical studies are of the IgG1 isoform [150 ]. When murine antibodies are used in humans, ADCC is best triggered by the isotype IgG3, and complement lysis by IgG2a [151 ,152 ]). IgM antibodies, which are most efficient in complement activation, are not of clinical relevance due to their short serum half life. A disadvantage of the administration of murine antibodies is their immunogenicity. Frequently, the induction of human anti-mouse antibody (HAMA) responses has been observed which may impede subsequent treatment with the same antibody [153 ]. Hence, attempts have been made to either humanize the antibodies by replacing murine with human sequences, or to obtain fully human antibodies from phage display or human Ig locus-transgenic mice [154 ].

An alternative option is to make use of an antibody which has already been tested for safety in a clinical trial. In 2005, 18 different monoclonal antibodies had been approved for clinical use in the US and Europe, and more than 150 were in clinical trials [155 ]. The majority of them targets overexpressed oncoproteins (tumor therapy), cell surface molecules of immune cells (immuno-modulation) or viral proteins (treatment of infections). However, most of these antibodies will not be suitable for the safety strategy described in this work. First, the epitopes of many antibodies have not yet been defined. This, however, is necessary, as only a short target sequence can be introduced into a TCR. Second, high specificity of the antibody for the adoptively transferred T cells is desired. Most likely, antibodies that recognize oncoproteins (e.g. Her2, EGFR) or molecules of the immune system (e.g. CD4, CTLA-4) will provoke unwanted side effects on other cell types. Third, the epitope tag itself should not be immunogenic which can not be excluded if parts of viral proteins are employed. Nevertheless, if one finds an appropriate alternative antibody-tag-combination, their applicability to the TCRmyc safeguard system needs to be analyzed.

4.6.2 Universality of the safeguard for different TCRs

↓78

Ideally, the TCRmyc safety approach should be applicable to every TCR employed in clinical studies. In this study, the myc-tag has been fused to one TCR chain N-terminus which is part of the variable region. Every TCR used for therapy, however, has its unique variable domain; and although all are expected to adopt an Ig-like fold, minor differences in the tertiary structure exist and not all TCR variable segments have been crystallized. It is not clear whether the insertion of a myc-tag will maintain expression and function of every therapeutic TCR and will allow efficient depletion of transduced T cells. In this thesis, two mouse TCRs with two different TCRα-chains belonging to the same variable region familiy (vα2) and one human TCR have been analyzed; and all tested TCRs were suitable for application of the safeguard. Still, it would be interesting to test TCRs with various TCRα variants to see if differences exist. Probably, for some TCRs the insertion of only one tag might be sufficient for elimination or even required for maintenance of function. If the approach is found not to be universal for all TCRs, it might be necessary to re-evaluate the TCRmyc strategy for every therapeutic receptor going into a clinical trial anew. Instead of introducing the myc-tag into the variable region of a TCR, one might also consider the modification of the constant part to achieve universality. Unfortunately, in this work none of the constant region mutations led to sufficient recognition of the tag by a myc-specific antibody that allowed depletion.

4.7 Eventual limitations of myc-tagged TCRs as a safeguard

4.7.1 Immunogenicity of TCRmyc

A disadvantage with many safety approaches is their immunogenicity as they are usually of viral or bacterial origin or comprise artificial fusion proteins [99 ,101 ,102 ]. An immune reaction against the adoptively transferred cells might lead to their unwanted premature elimination and also prevent the survival of a second graft. Myc-tagged TCRs consist entirely of human protein sequences: the rearranged αβ TCR chains and two 10 aa stretches of the c-myc protein. Therefore, the risk of immunogenicity is rather low. Still, it can not be excluded that at the fusion site between the tandem tags or between the tag and the N-terminus of the TCRα chain an immunogenic peptide in context with some MHC alleles is generated which is recognized by the immune system of the patient. Furthermore, c-myc is a nuclear protein, and central humoral tolerance to nuclear proteins is not that strict [156 ,157 ]. Berger et al. injected T cells modified with a Fas-FKBP suicide construct into macaques and observed an immune response against the transferred cells. Detailed analysis, however, revealed that this was directed against epitopes that differed between the human and macaque sequence, but not against the fusion sites [109 ]. Such an approach could also be used to test the immunogenicity of myc-tagged TCRs. Still, this would not give conclusive data about the reaction of a large number of patients with a multitude of different MHCs. In this work, murine T cells that were modified with murine TCRs fused to the human myc-tag were injected into mice. The sequence of the human myc-tag differs from the corresponding mouse sequence in 3 of 10 aa. Hence, there is the possibility that the transgene is recognized by the mouse immune system as foreign. In the performed experiments, however, the recipient mice were always immuno-deficient at the time-point of T cell transfer, either because of the genetic background of the employed mouse strain (Rag1-/- mice) or because of previous treatment with cyclophosphamide or irradiation. Thus, these experiments do not allow assessing the immunogenicity of myc-tagged TCRs.

Nevertheless, it was experienced in clinical trials that adoptively transferred T cells engraft better in lymphopenic patients [41 ,158 ] and pre-treatment with non-myeloablative, lymphodepleting drugs before transfer seems to become a standard procedure. In lymphocyte-depleted patients the immunogenicity of any transgene might not be of importance, though. And even repopulating endogenous lymphocytes are likely to be ignorant of foreign peptides due to peripheral tolerance mechanisms.

4.7.2 Elimination of activated T cells

↓79

Upon activation of a T cell, the TCR/CD3 complex becomes downmodulated which has been shown to be caused by increased intracellular degradation of the constantly recycling molecules [159 ]. Thus, efficient elimination of TCRmyc-transduced T cells by a myc-specific antibody might be hampered if the cells are activated due to auto-reactivity and the number of TCRmyc molecules on the surface is reduced. It has been demonstrated that the amount of epitopes is critical for mediating depletion [143 ]; and also the results presented in this work show that efficient complement lysis was only achieved with the TCRmyc variant that showed highest anti-myc antibody staining suggesting that the quantity of myc-tag epitopes is of importance. In the in vitro data presented in this study, however, TCR downmodulation could not be analyzed. The T cells subjected to complement lysis assays were either a T cell line (58 cells) or primary human PBLs. The 58 T cell line barely shows receptor downmodulation upon stimulation (data not shown) and the PBLs were in resting phase, which employs culture conditions with low IL-2 concentration, at the time-point of the depletion assay. Accordingly, the in vitro data do not allow drawing conclusions about the efficiency of elimination of activated TCRmyc T cells. In contrast, it can be supposed that the T cells depleted in the in vivo experiments were – at least to some extend – activated. First, the T cells were injected after three days of mitogenic stimulation with anti-CD3/CD28 antibodies and high-dose IL-2. Second, when injected into Rag-1-/- mice, the T cells underwent homeostatic lymphopenia-induced expansion which correlates with an activated phenotype [160 ]. Third, in RIP-mOVA mice the depletion was carried out at a time-point when the transferred T cells had already migrated to the pancreas and most likely encountered their antigen. Probably, several effector mechanisms are involved in antibody-induced depletion in vivo which allows the elimination of activated T cells. Still, it might be interesting to compare the in vitro depletion efficacy of activated and resting T cells side by side and to characterize the phenotype of in vivo depleted T cells by staining for activation markers.

4.7.3 Elimination of transformed T cells

Retroviral insertion into the genome of a host cell bears the risk of malignant transformation through activation of oncogenes. In this case, it is desirable to have the possibility to eliminate the leukemic T cells in the patient by a safeguard mechanism. In this work, it has not been analyzed whether the TCRmyc safety approach is applicable to treat integration-induced leukemia. The only clinical trial for adoptive cell therapy, in which transformation was observed, is the genetic modification of stem cells from X-SCID patients with the gene for the cytokine receptor common gamma (γc) chain [78 ,79 ]. However, several aspects argue against insertional mutagenesis by TCR gene transfer. First, in contrast to the X-SCID studies, the transduced T cells are not hematopoietic progenitors but differentiated lymphocytes. Recent data indicate that transfer of various single oncogenes is detrimental in stem cells, but does not lead to transformation of T cells (personal communication D. v. Laer, Georg-Speyer-Haus, Frankfurt a.M., Germany). Second, it is not yet known whether the TCR transgene does provide a direct proliferative advantage for the T cells whereas expression of the γc chain in stem cells is clearly essential for stimulation by many growth-promoting cytokines [161 ]. 

Still, it is not clear whether transformed T lymphocytes might escape myc-specific antibody depletion by either downregulation or loss of TCRmyc expression or acquisition of complement resistance. Antigen-loss variants have been described in antibody-treated B cell lymphomas [162 ]; and resistance to antibody therapy has been observed even when the antigen is still expressed [114 ]. In a transformed T cell, expression of the transgenic TCR might not be essential for T cell survival. Hence, loss of TCRmyc expression might not lead to reduced proliferation but abrogates the possibility of elimination. Some tumor cell lines have been demonstrated to overexpress complement inhibitory molecules rendering them insensitive to complement-mediated lysis [163 ,164 ,165 ]. Further mechanisms of resistance include an elevated apoptotic threshold or altered susceptibility to cellular cytotoxicity.

↓80

Thus, it remains desirable to determine the influence of retroviral transfer of TCR genes into human T cells; e.g. by analysis of integration loci, loss or maintenance of polyclonality and growth behavior over an extended period of time. Another possibility is to transfer an oncogene additionally to the TCR genes – as to mimic activation of endogenous oncogenes – and to study whether the T cells become transformed. Then, anti-myc antibody-mediated depletion of these T cells could be analyzed.

4.7.4 Elimination of T cells expressing TCR heterodimers

If a T cell is genetically modified with a therapeutic TCR four different receptor combinations can be expressed: the transgenic TCR, the endogenous TCR and mispaired heterodimers of the α- and β-chain of the transgenic and endogenous TCR. Depending on the “strength” of the TCRs – meaning its intra- and inter-chain stability and interaction with CD3 subunits – only one, several or all of the combinations are found on the T cell surface [61 ,62 ]. The safeguard proposed in this thesis work relies on expression of the myc-tagged transgenic TCRα-chain. In an unfortunate scenario, a heterodimer of the endogenous α-chain and the introduced β-chain recognizes an auto-antigen and causes auto-immunity. If this heterodimeric TCR is dominant over the other variants, the myc-tagged TCRα-chain might not be expressed at all, hence providing no possibility of T cell elimination. To exclude this, one might also introduce a tag into the TCRβ-chain. In this study, however, modification of the P14 TCRβ-chain with one myc-tag did not lead to recognition by a myc-specific antibody. Therefore, it needs to be analyzed whether a double myc-tag (as in the P14 TCRα-chain), a flexible linker between the tag and the N-terminus of the TCR chain or the choice of a different tag support the depletion of auto-reactive T cells via the TCRβ-chain.

4.7.5 Activation of auto-reactive T cells by the myc-specific antibody

Several monoclonal antibodies specific for the TCR/CD3 complex have been identified that have both depleting and activating capacity. One example is the anti-human CD3 antibody OKT3 which stimulates human PBLs in vitro, but can be used for depletion of CD3-positive T cells in the presence of complement [166 ,167 ]. Similarly, the anti-murine CD3ε antibody 145-2C11 acts as a mitogen in T cells in vitro and in vivo, but is also described to induce immuno-suppression by lymphocyte depletion [168 ,169 ]. Some factors that determine the effects of an anti-TCR/CD3 antibody are (i) the antibody isotype, (ii) the presence of serum complement factors or cells capable of mediating ADCC and (iii) the dose of the administered antibody [170 ].

↓81

Preliminary experiments revealed that an immobilized myc-specific antibody can – in the absence of complement – induce activation of a TCRmyc-transduced indicator cell line in vitro (data not shown). Hence, there is the theoretical risk that administration of a myc-specific antibody might further stimulate auto-reactive TCRmyc T cells in vivo. In the performed animal experiments, however, no activation of the transferred T cells was observed. Instead, application of the antibody led to a rapid depletion of the T cells (< 1 day) and prevented onset of auto-immune disease. It might be interesting though to evaluate the activating capacity of a myc-specific antibody on TCRmyc T cells, e.g. by injection of a low dose of antibody, a non-depleting antibody or the use of antibody fragments which lack the parts necessary for depletion. In the absence of auto-reactivity, this would provide a tool to specifically stimulate adoptively transferred tumor-reactive T cells in a patient thereby enhancing an anti-tumor immune response.

4.8 Future prospect

Certainly, several clinical trials using TCR-modified T cells will be carried out in the upcoming years. When efficiency of this therapy improves, an increase in the risk of auto-immune side effects is expected emphasizing the need for a reliable safety strategy. This study demonstrated that the introduction of a short peptide sequence into a TCR molecule allows the specific elimination of T cells that express the TCR while maintaining full function. However, some limitations of this approach have been discussed.

For translation into a clinical setting it will be most important to obtain a clinically approved antibody. Currently, only murine myc-specific antibodies of IgG1 or IgG2a isotype are commercially available. Of those, only the hybridoma of clone 9E10 (an IgG1 antibody) can be obtained from ATCC and the variable fragments of this antibody have been cloned into bacterial expression vectors [171 ]. However, application of a myc-specific antibody in a clinical trial bears several economic hurdles. First, clinical scale production of the antibody under good manufacturing practice (GMP) conditions will be very cost-intensive. Second, before administration in a clinical trial with TCR-modified T cells it might be necessary to test the general safety of the antibody in humans in a separate study. Third, humanization of the antibody or production of a fully human antibody requires further expenses. With our available means, generation of a murine, humanized or fully human GMP-grade myc-specific antibody is difficult. Therefore, emphasis will be laid on finding an alternative peptide with properties comparable to those of the myc-tag that is recognized by a clinically characterized antibody. If a suitable sequence is found, a model TCR will be modified with the alternative tag and analyzed for maintenance of function and efficiency of elimination. Another important point will be to show the universality of the approach. It would be very interesting to compare myc-tag modified TCRs with different Vα segments. Plenty of TCRs have been isolated from various labs and it should be possible to obtain model TCRs of the most common segments. Also, candidate TCRs that are suggested for clinical trials will be analyzed.

↓82

The application of the myc-tag strategy is not limited to TCRs. In fact, several gene therapy approaches using cell surface-expressed transgenes could benefit from a specific safety modality. In the laboratory of H. Abken (Division Tumorgenetics and Immunology, Uniklinik Köln, Cologne, Germany) CAR molecules are currently being modified with a myc-tag and analyzed for their capability to mediate depletion of T cells. Similar to TCRs, CAR-transduced T cells can be employed for adoptive cell therapy, but bear the risk of auto-immune side effects. Additionally, in our laboratory Nicole Scheumann started to introduce a myc-tag into the γc chain and could demonstrate that B cells and T cells modified with the myc-tagged transgene can be depleted in vivo and in vitro. Gene therapy with this molecule has recently led to the development of T cell leukemia in some patients due to retroviral integration; and trials have been discontinued until safety is ensured. A third adoptive therapy with an urgent need for a safeguard is the transfer of allogeneic T cells into patients that had received a hematopoietic stem cell transplant from the same donor. Though this treatment is highly effective against relapse and virus-induced lymph-proliferative diseases, a high incidence of severe, often lethal GvHD requires a possibility to eliminate allo-reactive T cells in the patient. The modification of the lymphocytes with a membrane-anchored tag prior to transfer could allow treating GvHD while probably maintaining the anti-tumor effect.

Apart from acting as a safeguard, the introduction of a tag into the TCRα-chain offers the possibility of staining the molecule with a specific antibody, e.g. for FACS analysis without loss of functionality. As currently only very few monoclonal antibodies for different Vα variants are available this for the first time provides the means to differentially detect the transgenic TCRα-chain among the endogenous ones. Initial data from Simone Reuss in our laboratory show that a myc-tag and an HA-tag can be employed to discriminate between two different TCRα-chains of the same subfamily which has not been possible so far.


© Die inhaltliche Zusammenstellung und Aufmachung dieser Publikation sowie die elektronische Verarbeitung sind urheberrechtlich geschützt. Jede Verwertung, die nicht ausdrücklich vom Urheberrechtsgesetz zugelassen ist, bedarf der vorherigen Zustimmung. Das gilt insbesondere für die Vervielfältigung, die Bearbeitung und Einspeicherung und Verarbeitung in elektronische Systeme.
DiML DTD Version 4.0Zertifizierter Dokumentenserver
der Humboldt-Universität zu Berlin
HTML-Version erstellt am:
23.08.2010