3  Results

In this thesis, the amino acid (aa) sequence 410 - 419 of the human c-myc protein (myc-tag) sought to be introduced into the structure of a TCR in a position where it can be recognized by a myc-specific antibody without interfering with TCR function. For this, crystal structures of human and murine TCRs [3 ] were inspected visually. Four different regions were identified, which (1) were protruding from the TCR structure and therefore seemed more likely to be accessible for an antibody and (2) were located outside of the CDR regions, which are primarily responsible for the binding to the peptide-MHC complex. Figure 5 depicts these four regions - namely the N-termini of the TCRα- and TCRβ-chain, the FG loop of Cβ and the c-strand of Cα - in the crystal structure of the 2C TCR [6 ].

	Figure 5 : Regions for insertion of a myc-tag in to a TCR.  Crystal structure of the 2C TCRα-chain (purple) and TCRβ-chain (blue). The arrows indicate the four regions selected for insertion of the myc-tag. (1: N-terminus of TCRα-chain, 2: c-strand of Cα, 3: N-terminus of TCRβ-chain, 4: FG loop of Cβ). Adapted from [6, ].

As a model, the murine P14 TCR (recognizing the glycoprotein-derived peptide 33 - 41 of lymphocytic choriomeningitis virus, gp33) was chosen. For modification of the four described TCR regions two different strategies were applied: either the myc-tag was directly inserted between two amino acids of the TCR sequence, or ten amino acids of the original TCR sequence were replaced by the tag. For some regions, it was tested whether insertion of two adjacent tags was superior to insertion of only one myc-tag regarding the binding of a myc-specific antibody. Finally, nine different myc-tagged TCRs were designed as described in Table 2 and visualized in Figure 6.

Table 2 : Positions of myc-tag insertion in the murine P14 TCR.

In case of the N-terminal modifications it had to be considered that the first amino acids of the translated protein comprise the signal peptide which is cleaved during TCR processing and export. Therefore, the myc-tag needed to be inserted between the signal peptide and the first amino acid of the mature protein. For identification of the signal peptide cleavage site the SignalP 3.0 software [135 ] was used. In this program, for each amino acid the probability to be part of a signal peptide (S score), and the probability to be part of a cleavage site (C score) is determined. From these two values the Y score is calculated which estimates where the probability of cleavage is highest. Figure 7A and B show the S, C and Y scores of the P14 TCR chains indicating the putative cleavage position.

	Figure 6 : Positions of myc-tag insertion in the murine P14 TCR. The 10 aa myc-tag sequence was incorporated into the sequence of the TCRα- (A) or TCRβ-chain (B). In five variants, sequences of the original TCR were exchanged by one or two myc-tag sequences (1). Numbers indicate the first and last aa position of the original sequence that was replaced. Four variants were generated, in which one or two myc-tags were inserted (2). Here, numbers indicate the amino acid position behind which the tag was introduced. The name of each variant according to Table 2 is given in brackets. (S: signal peptide, V: variable region, D: diversity region, J: joining region, C: constant region). The asterix marks the position [DAN] which was later used for modification of different human and murine TCRs.

Furthermore, it was analyzed whether addition of the myc-tag to the N-terminus of the mature proteins leads to changes in the cleavage probability. Therefore, the designed amino acid sequence of the N-terminal myc-tag-modified TCR chains was analyzed using the SignalP 3.0 software. As seen in Figure 7C and D the proposed cleavage site of the signal peptide is not affected by insertion of the myc-tag both in the TCRα- and in the TCRβ-chain.

For retroviral transduction the MP71-PRE vector was employed. This vector harbors the long terminal repeats (LTRs) of mouse myeloproliferative sarcoma virus (MPSV), a leader sequence of murine embryonic stem cell virus and the woodchuck hepatitis virus posttranscriptional regulatory element (PRE) and has been shown to lead to efficient expression of transgenes in murine and human T cell lines and primary lymphocytes [56 ,61 ,136 ,137 ,138 ].  Retroviral vectors were generated encoding either the non-modified wild-type (wt) TCR (P14/TCRwt) or the TCR with one of the myc-tag modifications (P14/TCRmyc[X], X being the described position of myc-tag insertion of Table 1). The myc-tag was inserted by site-directed mutagenesis PCR using pairs of overlapping primers. In TCRs with two myc-tags, the second tag was introduced by ligation of the TCR vector containing one myc-tag with a double-stranded oligonucleotide encoding the second tag. 

	Figure 7 :   Fusion of a myc-tag to the N-terminus of the mature P14 TCR α – and TCR β –chain s does not alter the signal peptide cleavage position. (A,B) The peptide sequence of the P14 TCRα– and TCRβ–chain was analyzed using SignalP 3.0 software. The position with the highest Y score indicates the putative cleavage site of the signal peptide (VNG-QQ for the α–chain and MEA-AV for the β–chain). (C,D) Similarly, the modified P14/TCRmyc[AN] and P14/TCRmyc[BN] amino acid sequences which carry a myc-tag between the proposed signal peptide and the mature protein were analyzed. The myc-tag sequence is boxed in red.

To analyze whether the myc-tag-modified P14 TCRs can be expressed on T cells, retroviral particles were generated by transfection of Plat-E cells with the TCR gene-containing MP71 vectors, and used to transduce the murine T cell line B3Z. Cells were enriched by FACS sorting using vα2- and vβ8-specific antibodies. Seven days after sorting, expression of the TCR was analyzed by flow cytometry with a vα2- and vβ8-specific antibody. 

	Figure 8 :   All myc-tagged P14 TCR s are expressed in the murine T cell line B3Z. B3Z cells were transduced with one of the P14/TCRmyc and subsequently sorted for vα2 and vβ8 expression. (A) Cells were stained with vα2- and vβ8-specific antibodies and analyzed by flow cytometry. The number of double-positive cells after sorting is given in percentage. Untransduced cells (neg) or cells transduced with P14/TCRwt (both boxed in grey) were used as a control. (B) Mean fluorescence intensity (MFI) of the TCRα-chain (black bars) and TCRβ-chain (grey bars) staining was determined. The depicted relative MFI values indicate the expression level of the myc-tagged TCR relative to expression of the wt TCR chain which was set to 100%.

Figure 8 shows the percentage of TCR-positive T cells as well as the mean fluorescence intensity (MFI) for each TCR chain, which indicates the level of expression. All myc-tagged P14 TCRs were expressed on B3Z cells. Some TCRs (e.g. P14/TCRmyc[AN, DAN and XL] retained a similar expression level of the TCR when compared to P14/TCRwt. However, the MFI of some modified TCR chains (e.g. the β-chain of P14/TCRmyc[DL] or the α-chain of P14/TCRmyc[CS]) was lower when compared to P14/TCRwt.

Next, it was analyzed whether the myc-tag modified TCRs can be detected by a myc-specific antibody. B3Z cells transduced with the different P14/TCRmyc or P14/TCRwt retroviral vectors were stained with a myc-specific antibody and analyzed by flow cytometry (Figure 9).

	Figure 9 : Only some myc-tag modified TCRs can be detected by a myc-specific antibody. (A) P14/TCRmyc-transduced and –enriched B3Z cells were analyzed by FACS using a myc-specific antibody. Untransduced cells (neg) or cells transduced with P14/TCRwt (both boxed in grey) were used as controls. Numbers indicate the percentage of sorted cells that stained positive with the antibody. (B) An overlay of the histograms shows distinct levels of myc-tag detection of different TCRs.

Only four of the nine TCRmyc (namely AN, DAN, XL, DL) could be detected with a myc-tag specific antibody.

To reproduce these findings, a second murine T cell line, TCR-deficient 58 cells, was transduced with P14/TCRwt or those myc-tagged P14 TCRs, which bound the myc-specific antibody in B3Z cells (AN, DAN, DL, XL). The cells were sorted by FACS for TCR expression and analyzed as before with vα2- and vβ8-specific antibodies (Figure 10).

	Figure 10 :   P14/TCRmyc [ AN, DAN, DL and XL ] are expressed in a TCR-deficient T cell line. 58 cells were transduced with P14/TCRmyc [AN, DAN, DL and XL] retroviral vectors and subsequently sorted for vα2 and vβ8 expression. Cells were stained with vα2- and vβ8-specific antibodies and analyzed by flow cytometry. The number of transduced cells after sorting is given in percentage. Untransduced cells (neg) or cells transduced with P14/TCRwt (both boxed in grey) were used as a control.

As in B3Z cells, P14/TCRmyc [AN, DAN, DL and XL] were expressed comparably to the wt TCR. In this experiment, the MFI of the TCR chains did not allow to determine their relative expression level as 58 cells are TCR-deficient. Thus, the introduced TCR does not have to compete with the endogenous TCR for CD3 and TCR export components. Therefore, protein instability due to the introduced modification does not have a similarly strong impact on TCR expression level as observed in B3Z cells, which possess an endogenous TCR [7 ].

Next, it was analyzed whether the myc-tagged P14 TCRs were also detected by a myc-specific antibody when expressed in 58 cells. Therefore, the transduced cells were stained with a myc-specific antibody and analyzed by flow cytometry (Figure 11).

	Figure 11 :   P14/TCRmyc [ AN, DAN, DL and XL ]   can be detected by a myc-specific antibody   when ex pressed in a TCR-deficient murine T cell line. (A) P14/TCR-transduced 58 cells that were enriched for TCR expression were analyzed by FACS using a myc-specific antibody. Untransduced cells (neg) or cells transduced with P14/TCRwt (both boxed in grey) were used as controls. Numbers indicate the percentage of sorted cells that stained positive with the antibody. (B) An overlay of the histograms shows distinct levels of myc-tag detection of different TCRs.

Consistent with the previous results, the four myc-tagged P14 TCRs [AN, DAN, DL and XL] could be detected by a myc-specific antibody in the T cell line 58. In both analyzed murine T cell lines the variant [DAN] showed the highest detection level whereas detection of the variant [DL] was lowest. Anti-myc antibody staining of P14/TCRmyc[XL] and P14/TCRmyc[AN] displayed some variation between 58 and B3Z cells. However, in both cell lines the fluorescence intensity of [XL] and [AN] was higher than that of [DL] and lower than that of [DAN].

To analyze whether murine T cell lines expressing P14/TCRmyc could be depleted in vitro, 58 cells enriched for P14 TCR expression were subjected to complement-mediated lysis. For this, the T cells were incubated first with a myc-specific antibody and subsequently with rabbit complement factors. If antibody has bound to its antigen on the cell surface, complement factors bind to the Fc part of the antibody and induce a signaling cascade leading to specific lysis of the cell. As a control for unspecific lysis, cells were also incubated with complement alone. After incubation, 7-AAD staining was performed to discriminate between viable and dead cells. Flow cytometry results are shown in Figure 12A. Specific lysis was calculated under consideration of unspecific lysis mediated by incubation with complement alone (Figure 12B).

	Figure 12 :   P14/TCRmyc[ DAN ] and [ DL ]   support complement d epletion of T cells. (A) 58 cells transduced with P14/TCRmyc [AN, XL, DL or DAN] and enriched for TCR expression were incubated with a myc-specific antibody and rabbit complement factors. Viable and dead cells were discriminated by staining with 7-AAD and FACS analysis. P14/TCRwt-transduced cells were used as a negative control (boxed in grey). Numbers indicate the percentage of dead cells for specific (blue) or unspecific lysis (grey). (B) From these data, percentage of specific depletion was calculated. The experiment was performed twice with reproducible results.

When incubated with complement alone, unspecific cell lysis of 9 to 16% was observed. Specific depletion could only be demonstrated for P14/TCRmyc[DAN] (83%) and [DL] (40%). P14/TCRwt-transduced cells and cells transduced with P14/TCRmyc[AN] and [XL] were not depleted by a myc-specific antibody. Because depletion in the case of P14/TCRmyc[DAN] was most efficient, solely this variant was further studied (now designated as “TCRmyc” only).

For the potential clinical application of myc-tagged TCRs it is essential that the receptor function is not impaired by the insertion of the tag. For functional characterization of P14/TCRmyc, antigen binding as well as cytokine secretion were analyzed. The first was accomplished by staining with specific peptide-MHC multimers, the latter by detection of IL-2 secretion upon antigen stimulation. For this, the α-chain and β-chain genes of P14/TCRwt and P14/TCRmyc were combined in one single MP71 retroviral vector and linked by the 2A element of porcine teschovirus (P2A) yielding MP71-P14α-P2A-P14β and MP71-P14αmyc-P2A-P14β, respectively. These vectors were used to transduce splenocytes of B6 mice which were subsequently stained with a P14-specific and an irrelevant tetramer. Both TCRs similarly bound the P14 tetramer as shown by comparable MFI in flow cytometry (Figure 13A).

	Figure 13 :   P14/TCRmyc function s comparable to P14/ TCR wt. (A) B6 splenocytes were transduced with P14/TCRwt or P14/TCRmyc. After 72 hours the cells were stained with a CD8-specific antibody, a P14-specific MHC-tetramer and an irrelevant tetramer (irr). Untransduced cells (neg, boxed in grey) served as a negative control. Cells shown are gated on CD8 expression. Numbers indicate the MFI of the specific tetramer staining. The experiment was repeated twice with similar results. (B) For peptide titration, 58 (CD8α+) cells were transduced with P14/TCRwt or P14/TCRmyc. Cells were stained with vα2- and vβ8-specific antibodies and analyzed by FACS. Transduction efficiency is given in percentage. Untransduced 58 (CD8α+) cells were used as a negative control (neg, boxed in grey). (C) 1 x 105 P14/TCRwt- or P14/TCRmyc-transduced or untransduced 58 (CD8α+) cells (neg) were stimulated for 24 hours with 1 x 105 B6 splenocytes pulsed with 100 µM to 100 pM gp33 peptide. IL-2 concentration of the culture supernatant was analyzed by ELISA. Unloaded splenocytes cells (w/o) or splenocytes loaded with irrelevant peptide (irr) served as negative controls. Data represent mean values of duplicates and error bars indicate the standard deviation (SD).

For detection of cytokine secretion, CD8α-positive 58 cells were transduced with the P2A-linked P14/TCRmyc or P14/TCRwt retroviral vectors with a similar transduction efficiency of 57-59% (Figure 13B) and stimulated with gp33 peptide-loaded B6 splenocytes. Secretion of murine IL-2 was detected in a peptide concentration-dependent manner and was similar for cells transduced with either TCR (Figure 13C).

To study whether a myc-tag introduced into the same position of a different TCR was also capable of serving as a target site for anti-myc antibody-specific depletion, the murine OT-I TCR (recognizing the ovalbumin-derived peptide 257 – 264, ova) was modified with a myc-tag in position [DAN] as described for the P14 TCR. The OT-I TCR α-chain (wt or myc) and β-chain were cloned into separate MP71 vectors which were used to transduce 58 cells. Cells were enriched for TCR expression with vβ5-specific antibodies by FACS. Flow cytometry analysis using vα2- and vβ5-specific antibodies showed similar expression of both TCRs (Figure 14A). Incubation with a myc-specific antibody revealed only binding to OT-I/TCRmyc-modified cells, but not to OT-I/TCRwt-transduced cells (Figure 14B).

	Figure 14 : Expression of OT-I/TCRmyc and OT-I/TCRwt is comparable. TCR-deficient 58 cells were transduced with OT-I/TCRmyc or OT-I/TCRwt retroviruses and enriched by FACS with a vβ5-specific antibody. TCR expression was detected by staining with vα2- and vβ5-specific antibodies (A) or a myc-specific antibody (B) and anlyzed by flow cytometry. Untransduced 58 cells (neg, boxed in grey) served as a negative control for TCR expression. Numbers indicate the percentage of sorted double-positive T cells.

Furthermore, it was analyzed whether 58 cells transduced with OT-I/TCRmyc can be depleted by a myc-specific antibody in vitro. Therefore, the T cells were enriched for OT-I/TCRmyc expression and subjected to complement mediated lysis as described. Depletion was analyzed by 7-AAD staining and flow cytometry (Figure 15A). Specific lysis was calculated under consideration of unspecific lysis mediated by incubation with complement alone (Figure 15B). As seen for P14/TCRmyc, OT-I/TCRmyc-transduced T cells were depleted with a high efficiency by incubation with a myc-specific antibody (78%).

	Figure 15 : OT-I/TCRmyc- transduced 58 cells can be depleted by complement. (A) TCR-deficient 58 cells were transduced with OT-I/TCRmyc and enriched with vβ5-specific antibodies. For depletion, cells were incubated with rabbit complement factors and a myc-specific antibody (blue) or rabbit complement alone as a control (grey). 7-AAD was used to discriminate between living and dead cells. Percentages represent 7-AAD-positive, dead cells. OT-I/TCRwt-transduced cells served as a negative control (boxed in grey). (B) From these data, percentage of specific depletion was calculated. The results represent data from one of at least two independent experiments with comparable results.

To determine whether OT-I/TCRmyc functions comparable to its wt counterpart, specific antigen binding and cytokine secretion upon antigen stimulus were analyzed. For this, B6 splenocytes were transduced with OT-I/TCRmyc or OT-I/TCRwt. 72 hours after transduction cells were stained with an OT-I-specific tetramer and an irrelevant tetramer as a control (Figure 16A).

Cytokine release was analyzed by transducing CD8α-positive 58 cells with the wt or the myc-tagged TCR (Figure 16B), and incubating the cells with ova peptide-loaded T2-K^b cells. After 24 hours IL-2 concentration of the supernatant was determined by ELISA (Figure 16C). OT-I/TCRmyc-transduced cells bound the specific peptide-MHC tetramer comparable to OT-I/TCRwt-transduced cells as indicated by a similar MFI in the FACS staining. Furthermore, IL-2 secretion was similar for both TCRs and was peptide concentration-dependent. Untransduced 58 cells or cells incubated with unloaded T2-K^b cells or an irrelevant peptide did not show significant cytokine release.

	Figure 16 : Myc-tagged OT-I TCR functions comparable to wt OT-I TCR. (A) B6 splenocytes transduced with OT-I/TCRwt or OT-I/TCRmyc were stained with a CD8-specific antibody, an OT-I-specific tetramer and an irrelevant tetramer. Cells shown are gated on CD8 expression. Untransduced splenocytes (neg) were used as a control (boxed in grey). Numbers indicate the MFI of the tetramer staining. (B) CD8α-positive 58 cells were transduced with the myc-tagged or wt OT-I/TCR, stained with vα2- and vβ5-specific antibodies and analyzed by FACS. Numbers show the percentage of transduction efficiency. (C) 1 x 105 OT-I/TCRwt- or OT-I/TCRmyc-transduced 58 (CD8α+) cells were stimulated for 24 hours with 1 x 105 T2-Kb cells pulsed with 100 µM to 100 pM ova peptide. As a control untransduced 58 cells were used (neg). IL-2 concentration of the culture supernatant was analyzed by ELISA. Unloaded T2-Kb cells (w/o) or T2-Kb cells loaded with irrelevant peptide (irr) served as negative target controls. Data represent mean values of duplicates and error bars indicate the SD.

Complement-mediated lysis experiments showed that TCRmyc-modified T cells could be depleted by a myc-specific antibody in vitro. To analyze whether depletion was also efficient in vivo, splenocytes of B6 mice were transduced with either OT-I/TCRwt or OT-I/TCRmyc retroviruses. One day after transduction, cells were injected i.v. into T and B cell-deficient Rag-1^-/- mice. This mouse strain was chosen because (i) adoptively transferred T cells can easily be tracked in the blood without the use of congenic markers and (ii) the lymphopenic situation of the mice resembles that of a pre-conditioned patient before adoptive transfer in the clinic. Blood samples were taken 13 days after injection and stained with a CD8- and a myc-specific antibody (for TCRmyc-transduced T cells) or CD8-, vα2- and vβ5-specific antibodies (for TCRwt-transduced T cells), respectively. Flow cytometry analysis demonstrated the presence of the adoptively transferred cells in the blood of all mice (Figure 17A). For depletion, 500 µg of a myc-specific antibody were injected i.p. and blood samples were analyzed one day later by flow cytometry as described before. In mice that received OT-I/TCRmyc-transduced T cells and antibody, no myc-positive cells could be detected, indicating that TCRmyc-transduced T cells were completely depleted. In contrast, in mice that received OT-I/TCRwt T cells and antibody or in mice, which did not receive antibody treatment, the population of adoptively transferred cells remained unchanged (Figure 17B).

	Figure 17 :   T c ells transduced with myc-tagged TCRs can be depleted in vivo . Splenocytes of B6 mice were transduced with OT-I/TCRmyc. 5 x 106 TCR-transduced cells were adoptively transferred i.v. into Rag-1-/- recipients. (A) After 13 days blood was stained for presence of CD8- and myc-positive cells. (B) One group of mice received 500 µg of a myc-specific antibody i.p. for depletion. One day after antibody injection blood samples were collected and analyzed again with CD8- and myc-specific antibodies. As a control, one group of mice received OT-I/TCRwt-transduced cells, which were stained with CD8-, vα2- and vβ5-specific antibodies, and anti-myc antibody treatment (boxed in grey). The stainings show cells gated on CD8 expression and are representative for one of two treated animals.

In this work, it has been shown that OT-I/TCRmyc-transduced T cells can be depleted by a myc-specific antibody in vitro and in vivo, and that their function was comparable to that of OT-I/TCRwt-transduced T cells. On that account, the RIP-mOVA mouse model was chosen to analyze whether myc-tagged TCRs can be used to prevent an auto-immune disease.

This mouse model has been extensively studied with regard to antigen-specific auto-immunity [60 ,130 ,139 ]. RIP-mOVA mice express ovalbumin under the control of the rat insulin promoter (RIP) in the β-islet cells of the pancreas [130 ]. If transgenic OT-I T cells are transferred into these mice, they develop auto-immune diabetes due to destruction of the insulin-producing cells by the T cells. In this model, disease onset is extremely rapid: as early as day two after adoptive transfer insulitis – defined by infiltration of the islets with lymphocytes – can be detected. Blood glucose values increase from normal to highly glycemic (>14 mM) within 24 hours at day four or five after adoptive transfer and mice have to be sacrificed at day six to ten due to severity of symptoms.

In publications, which show diabetes disease in RIP-mOVA mice due to adoptive transfer of T cells, usually OT-I T cells from OT-I transgenic mice were employed. Only de Witte et al. described the transfer of OT-I/TCR-transduced T cells, but here additional infection with ova-expressing viruses was needed to stimulate an ova-specific immune response [140 ].

Hence, it was first analyzed whether polyclonal B6 T cells which were transduced with the OT-I TCR are capable of inducing diabetes in RIP-mOVA mice. In a first attempt, different numbers of OT-I/TCRwt-transduced splenocytes were injected. However, no increase in blood glucose levels could be observed although the same number of transgenic T cells was sufficient for disease induction. Therefore, it was tested whether pre-treatment of the RIP-mOVA mice with cyclophosphamide or total body irradiation gave the transferred cells an advantage due to homeostatic proliferation in the lymphopenic recipient. Splenocytes of B6 mice were transduced with OT-I/TCRwt retroviruses. One day later, TCR expression was analyzed by flow cytometry using vα2- and vβ5-specific antibodies to determine the percentage of transduced T cells. Either 1.5 x 10⁶ or 1.5 x 10⁷ OT-I TCR-positive cells were injected i.v. into RIP-mOVA mice of which one group had been sub-lethally irradiated with 5 Gy one day before adoptive transfer and a second had received 50 mg cyclophosphamide per kg body weight two days before adoptive transfer. As a control, one mouse was treated either with irradiation or cyclophosphamide, but did not receive T cells. Expansion of the OT-I-transduced T cells was followed by staining of blood samples seven days after transfer (Figure 18A). Diabetes onset was controlled by measurement of blood glucose concentration (Figure 18B).

	Figure 18 : OT-I TCR-transduced T cells induce diabetes only in previously irradiated RIP-mOVA mice. B6 splenocytes were transduced with OT-I/TCRwt and 1.5 x 107 or 1.5 x 106 OT-I TCR-positive T cells were injected i.v. into RIP-mOVA mice (n=1) which had either been irradiated or treated with cyclophosphamide before adoptive transfer. (A) Seven days after transfer blood samples were stained with CD8-, vα2- and vβ5-specific antibodies and analyzed by FACS. As a control blood from mice, which had either been irradiated or received cyclophosphamide but were not injected with T cells, was tested. Blood from an untreated B6 mouse served as a control for the endogenous vα2- and vβ5-positive T cell population. Cells shown are gated on CD8 expression. (B) Blood glucose levels of all mice were determined and followed up to 30 days. The data of irradiated mice are depicted using filled symbols; that of mice treated with cyclophosphamide using open symbols.

Only in mice which had been irradiated, a significant number of vα2/vβ5-positive T cells could be detected in peripheral blood. However, of those mice only the one which had received the higher number of cells showed an increase in blood glucose concentration.

In a second experiment it was analyzed (i) whether it was possible to prevent or cure the T cell-induced diabetes by administration of a myc-specific antibody and (ii) which was the appropriate time-point for antibody treatment. As destruction of islet cells and increase in blood glucose level is very rapid in this model, the time-span between transfer of T cells and administration of the antibody might be crucial. B6 splenocytes were transduced with OT-I/TCRmyc retroviruses. One day later the percentage of vα2/vβ5-positive cells was determined by FACS (data not shown) and 1.5 x 10⁷ OT-I TCR-positive T cells were injected into RIP-mOVA mice which had been irradiated using 5 Gy one day earlier. As a control, one mouse did not receive T cells. Blood glucose levels were followed (Figure 19). On day four after T cell transfer, first mice showed increased blood glucose values and were injected i.v. with 500 µg of myc-specific antibody (group “late myc-ab”). One group of mice, which at this time-point still exhibited normal blood glucose concentration, was treated in the same way (group “early myc-ab”). As a control, a third group did not receive antibody treatment. To determine whether multiple administration of antibody was necessary to cure the diabetes, half of the animals in group “late myc-ab” were treated repeatedly (again on days 6, 10 and 13) with 500 µg of myc-specific antibody (group “late myc-ab (rep)”).

As seen in Figure 19, only the group of mice which had received the antibody before blood glucose values increased could be effectively treated. After administration, glucose level first increased, but dropped soon and reached normal values at day 30 to 40 after transfer. Mice of this group were further analyzed up to day 100 and no increase in glucose concentration or disease symptoms were observed. Animals which received the antibody at a time-point when they already exhibited high glycemia could not be treated successfully. All mice in the groups “late myc-ab”, “late myc-ab (rep)” and “no myc-ab” had to be sacrificed due to severe diabetes symptoms (weakness, loss of weight). These data show, that depletion of T cells via a myc-tagged TCR is able to treat mice suffering from auto-immune T cell-induced diabetes. However, because the chosen model system is so rapid, depletion of the T cells has to be carried out at an early time-point after T cell transfer.

	Figure 19 :   Auto-immune diabetes induced by OT-I/TCR-transduced T cells can only be cured by an early administration of antibody. B6 splenocytes were transduced with OT-I/TCRmyc and 1.5 x 107 transduced cells were transferred into irradiated RIP-mOVA mice. One mouse did not receive T cells as a negative control. Animals of both groups “late myc-ab” and “early myc-ab” (each n=2) were injected on day 4 after transfer with 500 µg myc-specific antibody. Note that mice in the first one are glycemic at this time-point, whereas in the second one mice show normal blood glucose levels. Mice in group “late myc-ab (rep)” (n=2) received additionally the same dose of antibody on days 6, 10, and 13. As a control animals in group “no myc-ab” (n=2) were not treated with antibody. Blood glucose levels were determined and are depicted as mean values from animals in one group, SD is indicated by the error bars.

Next, the depletion of OT-I/TCRmyc-transduced T cells in RIP-mOVA mice was analyzed in an experiment using larger groups of animals (n=5/group) and compared to OT-I/TCRwt-transduced cells. Furthermore, efficiency of depletion in the pancreas and in lymphoid organs (lymph nodes, spleen) was determined by FACS and IHC staining. For this, splenocytes of B6 mice were transduced with either OT-I/TCRmyc or OT-I/TCRwt retroviruses. One day later, cells were injected i.v. into sub-lethally irradiated RIP-mOVA mice as described. For treatment, 500 µg of a myc-specific antibody were injected i.p. two days after adoptive transfer. None of the animals which received OT-I/TCRmyc T cells and antibody treatment developed diabetes as measured by blood glucose concentration until the end of the observation period on day 100. In contrast, all animals in the control groups receiving either OT-I/TCRwt T cells plus antibody or OT-I/TCRmyc T cells but no antibody succumbed to the disease within four to five days after adoptive T cell transfer and had to be sacrificed two to six days after onset of disease due to severe symptoms (Figure 20).

	Figure 20 :   Treatment of auto-immune insulitis mediated by myc-specific antibody depletion of OT-I/TCRmyc transduced auto-reactive T cells. B6 splenocytes were transduced with either OT-I/TCRwt or OT-I/TCRmyc retroviruses and 2 x 107 TCR-positive cells were injected i.v. into sub-lethally irradiated RIP-mOVA mice. Mice which were irradiated but received no cells served as a negative control. 500 µg of a myc-specific antibody was administered i.p. into all mice that had received T cells harboring the TCRwt and half of the mice which had received T cells carrying the TCRmyc. Blood glucose concentration was determined. Depicted are mean values of all animals (n=5) in one group; error bars indicate SD.

	Figure 21 : Immunohisto logy shows T cell infiltration and destruction of islets in the pancreas. (A) Two days after adoptive transfer, pancrei of mice from each group were analyzed by IHC with ovalbumin- (red) and CD8- (green) specific antibodies. Nuclei were stained with DAPI (blue). (B) Pancreatic sections from diabetic and antibody-treated mice were stained on day six in the same way. Insets show larger parts of the tissue at a lower magnification.

IHC staining of pancreatic sections of injected mice (but not control animals) two days after transfer showed infiltration of pancreatic islets with CD8-positive T cells demonstrating the very early onset of insulitis (Figure 21A). In stainings from pancrei isolated at day six after transfer from severely sick, diabetic mice (that had either received OT-I/TCRwt-transduced cells plus antibody or OT-I/TCRmyc-transduced cells but no antibody) revealed complete lacking of ova-expressing islet cells due to destruction by OT-I T cells. Also, infiltrating CD8-positive T cells could still be found in the tissue. In contrast, animals that had received TCRmyc-transduced T cells and antibody treatment exhibited intact islet structure and lack of T cells in the pancreas (Figure 21B).

Furthermore, OT-I tetramer-positive T cells were detected in mesenterial lymph nodes (Figure 22A) and spleens (Figure 22B) of diabetic mice, but not in treated animals as shown by flow cytometry.

	Figure 22 :   Depletion of OT-I/TCRmyc T cells can be detected in mesenterial lymph nodes and spleen.  Lymphocytes were isolated from (A) mesenterial lymph nodes and (B) spleen of animals from all groups and analyzed with a CD8-specific antibody and OT-I specific tetramer using flow cytometry. Lymphocytes from mice that did not receive T cells served as a negative control (boxed in grey). Depicted cells are gated on positive CD8 expression.

So far, murine model TCRs modified with a myc-tag have been analyzed in this study. For the application of the myc-tag safeguard in the clinic, however, it is necessary to determine whether human myc-tagged TCRs retain their function and can be employed for specific cell depletion, as well. For this, a myc-tag was introduced into a human TCR which is reactive against the peptide 209 - 217 of the melanoma antigen gp100. In particular, the gp100 TCR was modified with two myc-tags at the N-terminus of the TCRα-chain – corresponding to the position [DAN] which mediated the depletion of murine T cells transduced with the P14 or OT-I TCR. The gp100 TCRα- and TCRβ–chain genes were cloned into separate MP71 vectors which were used to generate retroviral particles. For expression analysis, the TCR-deficient human T cell line Jurkat76 was transduced. The cells were enriched with a β-chain-specific antibody and sub-cloned by limiting dilution. Several clones were analyzed by flow cytometry using CD3-, myc- and vβ8-specific antibodies. Results of one representative clone are shown in Figure 23. Both, the modified and the wild-type TCR were expressed on Jurkat76 cells as detected with a vβ8-specific antibody (Figure 23A). Because no antibodies are available for the detection of the gp100 TCRα-chain, this staining could not be performed. Only Jurkat76 cells transduced with gp100/TCRmyc, but not cells transduced with gp100/TCRwt could be stained with a myc-specific antibody by flow cytometry (Figure 23B).

	Figure 23 : The human gp100/TCRmyc can be expressed comparable to gp100/TCRwt and is d e tected by a myc-specific antibody. (A) The TCR-deficient human T cell line Jurkat76 was transduced either with gp100/TCRwt or gp100/TCRmyc, enriched with vβ8-chain-specific antibodies and subcloned by limiting dilution. TCR expression was analyzed by flow cytometry staining with a vβ8-specific antibody. Untransduced cells (neg, boxed in grey) served as a negative control. (B) Jurkat76 cells transduced with gp100/TCRmyc were stained with a myc-specific antibody and analyzed by flow cytometry. Cells transduced with the unmodified wild-type receptor served as a control. The data show results of one representative clone of several that were tested.

For functional comparison of gp100/TCRwt and gp100/TCRmyc, human PBLs were transduced with the two TCR retroviral vectors. Antigen binding and cytokine response were analyzed by staining with peptide-MHC tetramers and measuring IFN-γ secretion upon cultivation with peptide-pulsed target cells. Both TCRs similarly bound the gp100 tetramer as demonstrated by comparable MFI in flow cytometry (Figure 24A). Upon stimulation with gp100 peptide-loaded T2 cells, PBLs transduced with either TCR secreted comparable amounts of IFN-γ (Figure 24B).

	Figure 24: The human gp100/TCRmyc functions comparably to its TCRwt counterpart. (A) PBLs were transduced with gp100/TCRwt or gp100/TCRmyc retroviruses and stained with a vβ8-specific antibody and a gp100-specific tetramer. Untransduced PBLs (neg, boxed in grey) show the background of endogenous vβ8- positive T cells. Numbers indicate the MFI of the tetramer staining. (B) gp100/TCRwt- or gp100/TCRmyctransduced PBLs were co-cultured with T2 cells pulsed with 10 μM to 10 pM gp100 peptide for 24 hours. Untransduced PBLs were used as a negative control (neg). Culture supernatant was analyzed for IFN-γ content by ELISA. Unloaded T2 cells (w/o) or T2 cells loaded with an irrelevant NYeso1-derived peptide (irr) served as negative target controls. Data represent mean values of duplicates and error bars indicate SD. The results were reproduced in two independent experiments and with two different donors.

To show that gp100/TCRmyc-modified T cells can be depleted by a myc-specific antibody in vitro, PBLs were transduced with gp100/TCRmyc retroviruses and enriched using myc-specific MACS beads. Subsequently, the sorted cells were specifically restimulated with gp100 peptide-pulsed T2 cells and IL-2. Seven days after restimulation, PBLs that were 85% to 99% positive for myc-expression were analyzed for depletion by two different effector mechanisms: CDC and ADCC.

To investigate CDC, a myc-specific antibody and complement factors were added subsequently to the myc-tag-enriched PBLs. 7-AAD staining was performed to discriminate between viable and dead cells and specific lysis was calculated (Figure 25A).

For analysis of ADCC, the T cells were incubated first with a myc-specific antibody and subsequently with NK cells which served as effectors to mediate T cell lysis. To avoid allo-reactions, autologous NK cell cultures from the same donor were employed in the assays. To determine specific lysis, the myc-tag-enriched PBLs were radioactively labeled with ⁵¹Cr and the release of radioactivity into the supernatant was measured. (Figure 25B). Depending on the assay, 31% to 65% of the gp100/TCRmyc-transduced cells were depleted in the presence of a myc-specific antibody, whereas cells incubated without antibody showed only low unspecific lysis.

	Figure 25 :   T cells transduced with gp100/TCRmyc can be depleted in vitro by complement- and cell-mediated lysis. PBLs were transduced with gp100/TCRmyc retroviruses, sorted for myc-positive cells and restimulated with gp100 peptide-pulsed T2 cells. (A) For complement-mediated depletion, cells were incubated with a myc-specific antibody and rabbit complement factors. 7-AAD staining was used to discriminate between living and dead cells. Cells incubated without antibody served as a control. (B) For cell-mediated lysis, autologous PBMCs enriched for NK cells were used as effector cells. 51Chromium-labeled TCRmyc-positive PBLs were incubated with effector cells in E:T ratios from 50:1 to 2:1. A myc-specific antibody and a secondary rabbit anti-mouse IgG1 antibody were added and lysis was measured in a standard four-hour chromium release assay. Samples without antibody served as a control. Data represent mean values of duplicates and error bars indicate the SD. Similar results were obtained in an independent experiment with a different donor.