Differential T-cell activation by B7-1 expression

Abstract

T-cell receptor-mediated T-cell activation requires cosimulation signal, which can be provided by B7-1 molecule. Our previous study demonstrated that the coexpression of a covalent peptide/major histocompatibility complex class II molecule complex and costimulatory molecule B7-1 by recombinant adenovirus leads to synergy in peptide-specific T-cell activation. However, the viral antigen-specific T-cell activation is not enhanced by B7-1 expressed by the adenovirus. To verify the differential T cell activation by B7-1 and investigate its underlying mechanisms, we constructed an adenovirus coexpressing a covalent complex of hen egg lysozyme peptide/I-A^k (HEL_46–61/I-A^k) and B7-1 in the present study. In vivo studies revealed that HEL_46–61-specific T-cell response, but not viral antigen-specific T-cell response, was enhanced by B7-1 expression mediated by the adenovirus, suggesting that exogenous B7-1 expression may regulate T-cell response to these two different antigens through distinct mechanisms. Furthermore, our results revealed that antigen-presenting cells were unsusceptible to adenovirus infection in vivo. Based on these findings, the possible mechanism of differential B7-1 costimulation on peptide-specific and viral antigen-specific T-cell activation is discussed.

Introduction

Major histocompatibility complex (MHC) class II molecules are capable of binding to and presenting various peptide epitopes to T-cell receptors (TCRs).¹ Previous studies, however, demonstrated that a peptide fragment may be genetically fused to the N-terminus of the class II β chain through a flexible peptide linker.^2,3 This covalent peptide/class II complex can be expressed, processed and displayed on cell surface. The covalent complex is recognizable by specific T cells in a class II-dependent manner. This approach raises the possibility of selectively modulating T-cell response in an epitope-specific manner. Our previous study demonstrated that coexpression of a myelin basic protein (MBP) peptide/class II complex and costimulatory molecule B7-1 by recombinant adenovirus led to peptide-specific T-cell activation in a TCR-guided fashion.⁴ This approach of immunomodulation may be potentially applicable not only to the activation of peptide-specific T-cell response, but also to the down-regulation of peptide-specific T-cell response in conjunction with other inhibitory molecules, such as Fas ligand (FasL).^5,6 On the other hand, B7-1 expression by adenovirus did not lead to synergy in adenoviral antigen-specific T-cell activation in our previous study.⁴

To verify the phenomenon of differential T-cell activation by B7-1 and investigate its underlying mechanisms, we constructed several replication-deficient recombinant adenoviruses, which express a covalent hen egg lysozyme (HEL) peptide/class II complex. HEL peptide 46–61(HEL_46–61) was genetically fused to the N-terminus of the MHC class II β chain of the k haplotype ($\text{I}-{\text{A}}_{\beta }^k$) through a flexible linker. In this study, we present biochemical and functional characterization of these adenoviruses, and in vivo modulation of peptide-specific and viral antigen-specific T-cell response by B7-1 coexpression. Furthermore, we also present evidence to demonstrate that adenovirus does not infect antigen-presenting cells (APCs), and speculate the distinct antigen presenting pathways for HEL peptide and viral antigens.

Materials and methods

Cells and reagents

The 3A9 T-cell hybridoma (THy) was from Dr E. Unanue (Washington University School of Medicine, St. Louis, MO).⁷ Line 1 cell⁸ a murine lung carcinoma cell line, was a gift from Dr J. P.-Y. Ting (University of North Carolina at Chapel Hill, NC). The 1934.4 THy, the 293 cell line, HEL_46–61 peptide and MBP_1–11[4Y] peptide were described previously.⁴ All of the cells were maintained in Dulbecco's modified Eagle's minimal essential medium (DMEM; Gibco BRL, Gaithersburg, MD) supplemented with 10% heat-inactivated fetal calf serum (FCS), penicillin/streptomycin (100 U/ml), 2 mm glutamine and 5 × 10⁻⁵ M β-mercaptoethanol. Anti-$\text{I}-{\text{A}}_{\alpha }^k$ monoclonal antibody (mAb) 11-5.2, Anti-$\text{I}-{\text{A}}_{\beta }^k$ mAb 10-3.6, biotin-labelled anti-mouse B7-1 mAb 16-10A1, and phycoerythrin (PE)-labelled streptavidin were from BD PharMingen (San Diego, CA). F(ab′)₂ goat anti-mouse immunoglobulin (IgG)–PE was from Caltag Laboratories (Burlingame, CA).

Recombinant adenovirus

DNA sequence encoding for HEL_46–61 peptide was inserted into N-terminus of I-A^k by polymerase chain reaction (PCR), as illustrated in Fig. 1(a). Adenovirus shuttle plasmid (pAdTrack) expressing I-A^k or HEL_46–61/I-A^k was generated, as previously described.⁴ Recombinant adenoviruses listed in Fig. 1(b) were generated and purified, as described previously.^4,9 The absence of wild-type adenovirus contamination was confirmed by the negative results of PCR (up to 35 cycles) using the E1 primers.⁴ Purified viruses were aliquoted and stored at −80°. The viral titre (plaque-forming units (p.f.u.)/ml) for each adenovirus preparation was determined in 293 cells using the agarose overlay method described previously.¹⁰ Each plaque was also verified for the expression of green fluorescent protein (GFP) under an inverted fluorescent microscope.

Figure 1.

(a) Construct of HEL_46–61 and $\text{I}-{\text{A}}_{\beta }^k$ fusion protein. Only the amino acid sequences near the fusion site are shown here. (b) Constructs of recombinant adenoviruses. GFP, green fluorescent protein; P_CMV, human CMV immediate early promoter; P_RSV, respiratory syncytial virus leukotriene promoter; PAS, polyadenylation sequences from SV40 small T antigen (SV40), bovine growth hormone (BGH) or thymidine kinase (TK).

Immunofluorescence and Northern blot

Line 1 cells infected with, or lacking, adenoviruses at 100 MOI (multiplicity of infection or p.f.u./cell) for 24 hr were detached from culture plates and stained by indirect immunofluorescence for cell surface expression of I-A^k using either Anti-$\text{I}-{\text{A}}_{\alpha }^k$ or Anti-$\text{I}-{\text{A}}_{\beta }^k$ mAbs followed by an F(ab′)₂ goat anti-mouse IgG–PE, or biotin-labelled anti-B7-1 mAb followed by streptavidin–PE. Cells were analysed by Becton Dickinson FACSCalibur flow cytometer. For the Northern blot analysis, Line 1 cells were infected with adenoviruses at 100 MOI for 24 hr. Total RNA extraction and Northern blot analyses were performed as previously described.¹¹

Adenovirus infection and T-cell activation

To express I-A^k or HEL_46–61/I-A^k, Line 1 cells (2 × 10⁶ cells in a 100-mm dish) or other cells, as indicated, were infected with various purified recombinant adenoviruses (100 MOI) at 37° for 24 hr. Infected cells were harvested by pipetting or trypsin digestion, and washed twice with the culture medium. Antigen-presentation assays were performed in 96-well round-bottomed plates in complete DMEM in quadruplicate. Virus-infected cells (5 × 10⁴ cells/well or as otherwise indicated) were cocultured with the 3A9 THy (1 × 10⁵ cells/well) at 37° for 24 hr. Samples of the supernatants were assayed for interleukin-2, as described previously.⁴ Data were analysed by the paired t-test using GraphPad Instat program (San Diego, CA).

T-cell proliferation assay

Recombinant adenoviruses or phosphate-buffered saline (PBS) were injected into C3H/HeJ syngeneic mice (H-2^k) (male, 5–7 weeks old, from the Jackson Laboratories, Bar Harbor, ME) through the tail vein at 2 × 10⁹ p.f.u./mouse in 0·1 ml of PBS. Twelve days after the virus infection, spleen cells were isolated, washed, plated in 96-well plates at 1 × 10⁵ cells/well, re-stimulated in quadruplicate with PBS, HEL_46–61 (10 µg/ml), MBP_1–11 (10 µg/ml), Ad.GFP (50 MOI) or concanavalin A (Con A) (2 µg/ml) for 48 hr [³H]thymidine (1 µCi/well) was added to the medium. Cells were cultured for additional 16 hr and collected with a microplate cell harvester. T-cell proliferation was determined by [³H]thymidine incorporation using a Packard microplate scintillation counter.

Peritoneal murine macrophages

Macrophages were isolated from C57Bl/6 mice, as previously described.¹² Briefly, mice were injected i.p. with 3 ml of 2·98% thioglycollate broth (Difco, Detroit, MI). After 3 days, peritoneal exudate cells were harvested by washing the peritoneal cavity with cold Ca²⁺- and Mg²⁺-free Hanks' balanced salt solution (HBSS). The cells were washed three times in RPMI-1640 with 2% FCS. Peritoneal macrophages prepared in this manner were at least 90% macrophage, as assessed by expression of Mac-1 (not shown).¹³ Adherent peritoneal macrophages were infected with Ad.GFP at 1000 MOI and cultured for 2 days (35 mm dish, 4 × 10⁶ cells/dish) at 37°/5% CO₂ in RPMI-1640 medium supplemented with 10% heat-inactivated FCS, l-glutamine (2 mm) and penicillin/streptomycin (100 U/ml).

Results

Characterization of adenoviruses

Some recombinant adenoviruses generated in this study, such as Ad.I-A^k.B7-1 and Ad.HEL_46–61/I-A^k.B7-1 in Fig. 1(b), express up to four different proteins in the adenoviral E1 region, including GFP. Therefore, it is important to characterize the expression of the recombinant proteins. Northern blot analysis using RNA extracts from Line 1 cells infected with different viruses revealed high levels of $\text{I}-{\text{A}}_{\alpha }^k$ and $\text{I}-{\text{A}}_{\beta }^k$ transcripts for Ad.I-A^k, Ad.I-A^k.B7-1, Ad.HEL_46–61/I-A^k and Ad.HEL_46–61/I-A^k.B7-1, and B7-1 transcripts for Ad.I-A^k.B7-1 and Ad.HEL_46–61/I-A^k.B7-1 (Fig. 2a). The expression of $\text{I}-{\text{A}}_{\alpha }^k$, $\text{I}-{\text{A}}_{\beta }^k$ and B7-1 in Line 1 cells infected with adenoviruses was further confirmed by flow cytometry (Fig. 2b). The surface expressions of both $\text{I}-{\text{A}}_{\alpha }^k$ and $\text{I}-{\text{A}}_{\beta }^k$ were detected in the Line 1 cells infected with Ad.I-A^k, Ad.I-A^k.B7-1, Ad.HEL_46–61/I-A^k and Ad.HEL_46–61/I-A^k.B7-1. The surface expression of B7-1 was detected in the Line 1 cells infected with Ad.I-A^k.B7-1 and Ad.HEL_46–61/I-A^k.B7-1. All of these viruses contained a GFP expression cassette. The expression of GFP in the virus-infected Line 1 cells was verified by flow cytometry (Fig. 2b). Dose-dependent expression of GFP was observed in Line 1 cells infected with all adenoviruses (not shown).

Figure 2.

(a) Northern blot analysis. Line 1 cells were infected with indicated adenoviruses at 100 MOI for 24 hr. Total RNA extracts (5 µg/lane) from the infected Line 1 cells were analysed by Northern blot using different cDNA probes. (b) Flow cytometry analysis. Line 1 cells were infected with indicated adenoviruses, as described in (a). Cells were stained with Anti-$\text{I}-{\text{A}}_{\alpha }^k$ mAb (11-5.2) Anti-$\text{I}-{\text{A}}_{\beta }^k$ mAb (10-3.6) or biotin-labelled anti-B7-1 mAb (16-10A). Antibody binding was revealed by a subsequent incubation with either PE-labelled goat anti-mouse IgG antibody or PE–streptavidin. Cells were analysed by flow cytometry. ‘No Ab control’ represents secondary fluorescent conjugate staining without primary antibody. Cells in the upper right quadrant not only were infected with GFP-expressing adenovirus, but also had surface expression of the indicated molecules at detectable levels.

Activation of HEL-specific THy by the adenoviruses

The 3A9 THy, specifically recognizing HEL_46–61 in the context of I-A^k, were employed to evaluate the functional gene expression on the surface of virus-infected Line 1 cells. If $\text{I}-{\text{A}}_{\alpha }^k$ and HEL_46–61/$\text{I}-{\text{A}}_{\beta }^k$ were expressed, assembled and displayed appropriately on the cell surface, these cells should be able to present the antigenic peptide to T cells and activate them. Among all of the recombinant viruses tested, Ad.HEL_46–61/I-A^k demonstrated some capability of stimulating the THy (Fig. 3a), while Ad.HEL_46–61/I-A^k.B7-1-infected Line 1 cells led to strong T cell activation, suggesting that the simultaneous expression of covalent HEL_46–61/I-A^k complex and B7-1 leads to the synergy in T cell activation. In contrast, cells infected with Ad.I-A^k, Ad.I-A^k.B7-1 or Ad.GFP did not have any stimulation on the 3A9 THy. Control THy, 1934.4 cells recognizing MBP_1–11 peptide in the context of I-A^u, was not stimulated by any adenovirus-infected cells.

Figure 3.

(a) T-cell stimulation assays with adenovirus-infected Line 1 cells. Line 1 cells were infected with indicated adenoviruses at 100 MOI for 24 hr. Infected Line 1 cells were detached from plate, washed twice with the medium, re-plated in 96-well plates at 5 × 10⁴ cells/well, and incubated with either 3A9 or 1934.4 THy (1 × 10⁵ cells/well) for 24 hr. To analyse THy activation, interleukin-2 in culture medium was analysed by cytokine-specific enzyme-linked immunosorbent assay. The results are representative of four independent experiments, carried out in quadruplicate. (b) Functional expression of I-A^k and HEL_46–61/I-A^k on cell surface. Adenovirus-infected Line 1 cells were cocultured in 96-well plates with the 3A9 THy and incubated with HEL_46–61 or MBP_1–11 peptide at a final concentration of 10 µg/ml, as indicated, for 24 hr. THy activation was analysed, as described above.

Although the expression of HEL_46–61/I-A^k on the surface of adenovirus-infected Line 1 cells was demonstrated functionally by the activation of the 3A9 THy (Fig. 3a), the expression of I-A^k on the surface of cells infected with Ad.I-A^k was uncertain. To verify the surface expression of I-A^k, HEL_46–61 peptide was added to the adenovirus-infected Line 1 cells during the antigen presentation. The addition of HEL_46–61, but not control peptide MBP_1–11, to Line 1 cells infected with Ad.I-A^k fully restored the stimulatory capacity of these infected Line 1 cells (Fig. 3b). These results indicate that I-A^k expressed by Ad.I-A^k was processed and displayed on the cell surface with appropriate conformation. However, the addition of HEL_46–61 did not further enhance the activation of the 3A9 THy by Line 1 cells infected with Ad.HEL_46–61/I-A^k or Ad.HEL_46–61/I-A^k.B7-1, suggesting that the antigen-binding groove was fully occupied by the covalently fused HEL_46–61 peptide. Furthermore, the expression of B7-1 on the surface of Line 1 cells infected with Ad.HEL_46–61/I-A^k.B7-1 was revealed by the enhancement of the 3A9 THy activation (Fig. 3b). The addition of CTLA4/Fc was able to block B7-1 synergy (not shown), as previously reported.⁴

In vivo T-cell activation by adenoviruses

The main reason for us to choose recombinant adenovirus in this study is its high efficiency for in vivo gene transfer, allowing the investigation of the differential T-cell activation by B7-1 in the context of the immune system. For this purpose, Ad.HEL_46–61/I-A^k.B7-1, along with the control viruses, was injected into C3H/HeJ mice. Ten days after virus infection, spleen cells were isolated from these mice, and the T-cell recall response was analysed. T cells isolated from all of the mice responded to the stimulation with Con A strongly and at a similar level (data not shown), suggesting that T cells were functionally intact after the isolation procedure. Our results indicated that HEL_46–61-specific T cells were significantly primed in mice infected with Ad.HEL_46–61/I-A^k.B7-1 (P < 0·001) (Fig. 4). However, Ad.HEL_46–61/I-A^k did not induce HEL_46–61-specific T-cell activation in vivo, suggesting that naïve T-cell activation through TCR ligation requires costimulatory signals, as previously described.^5,14,15 These results suggested that coexpression of HEL_46–61/I-A^k and B7-1 induced synergistic HEL_46–61-specific T-cell activation in a TCR-guided fashion.

Figure 4.

In vivo T-cell activation by recombinant adenoviruses. C3H/HeJ mice (H-2^k) were infected with indicated adenoviruses through the tail vein injection at 2 × 10⁹ p.f.u./mouse. Spleen T cells were isolated 12 days later, plated in 96-well plates at 1 × 10⁵ cells/well and re-stimulated with different reagents, as indicated, for 48 hr. To determine T-cell proliferation, [³H]thymidine (1 µCi/well) was added to the cells. [³H]thymidine incorporation was determined 16 hr later. The results are representative of three independent in vivo studies, carried out in quadruplicate.

Infection with adenovirus can trigger a host immune response against adenoviral antigens.^16,17 This response was monitored as an antigen specificity control. All the mice infected with adenoviruses were primed for a T-cell immune response against adenoviral antigens, but at different levels (Fig. 4). Ad.GFP elicited a vigorous T-cell response against adenoviral antigens. Ad.I-A^k and Ad.HEL_46–61/I-A^k also induced significant T-cell activation against adenoviral antigens. However, neither Ad.I-A^k.B7-1 nor Ad.HEL_46–61/I-A^k.B7-1 further boost the T-cell response against viral antigens.

Adenovirus infects non-APCs, but not APCs

Previous study demonstrated that lipopolysaccharide (LPS)-activated macrophages are susceptible to adenovirus infection in culture. However, our study demonstrated that thioglycollate-solicited macrophage from mouse peritoneal cavity was not susceptible to adenovirus infection with the adenovirus as high as 1000 MOI (Fig. 5a). Furthermore, insusceptibility of APCs to adenovirus infection was verified by the negative expression of GFP in MHC class II⁺ cells isolated from mice infected with Ad.GFP (Fig. 5b). On the other hand, non-APCs, such as hepatocytes, were easily transduced by Ad.GFP. Based on the fact that B7-1 expression by adenovirus can only enhance T-cell response against HEL_46–61 peptide, but not adenoviral antigens, we propose that these antigens may be processed and presented by non-APCs and APCs, respectively. A detailed discussion about the deficient presentation of the antigens is presented in the next section.

Figure 5.

Susceptibility of APCs and non-APCs to adenovirus infection in vivo. (a) Macrophages were isolated from peritoneal cavity with thioglycollate, infected with Ad.GFP at 1000 MOI for 2 days in culture and analysed for GFP expression by flow cytometry. (b) C3H/HeJ mice were infected i.v. with either Ad.GFP or PBS control for 48 hr, as described in Fig. 4. Spleen cells were isolated, stained with PE-labeled Anti-$\text{I}-{\text{A}}_{\alpha }^k$ mAb, and analysed by flow cytometry. (c) GFP expression in mouse liver. Mouse liver was collected after mice were infected with Ad.GFP for 48 hr, as described in (b). GFP expression on frozen tissue sections was examined under a fluorescent microscope.

Discussion

Our previous study revealed that coexpression of the covalent peptide/class II complex and accessory molecules, such as B7-1 on the same cell surface by adenovirus leads to peptide-specific T-cell activation in a TCR-guided fashion.⁴ The covalent peptide/MHC class II complex, expressed by the adenovirus on the surface of infected cells, serves as bait or molecular capturing device, as it selectively bind to TCRs recognizing this peptide epitope. In addition, the same adenovirus will express B7-1 that serves as costimulatory molecule for the bound T cells. Therefore, only the T cells that specifically bind to the peptide/MHC class II complex will be selectively activated, while the T-cell population that does not recognize the antigen will be left untouched, remaining naïve. Intriguingly, adenoviral antigen-specific T-cell activation is not enhanced by B7-1 coexpression.⁴

The current study is designed to verify the differential T-cell activation by B7-1 coexpression and investigate its underlying mechanisms. We chose HEL as the antigen to test this hypothesis, because the mouse immune response to the protein is well characterized and its antigenic peptide epitopes are well defined.^7,18 HEL_46–61 is the dominant epitope presented by I-A^k. Line 1 cells were selected for this study because of their susceptibility to adenovirus infection at as low as 3 MOI (not shown). More importantly, the cells of the H-2^d background should not be able to present HEL_46–61 to the 3A9 THy, although the induction of endogenous MHC class II molecules by adenovirus infection has not been reported previously.

The coexpression of I-A^k and B7-1 in adenovirus-infected cells were confirmed not only by Northern blot, but also by flow cytometry. Although our results indicated that not every GFP⁺ cell is I-A^k positive or B7-1 positive, this is mainly due to antibody affinities and assay conditions, such as the wash condition to reduce the background. Among the five recombinant adenoviruses studied, only two, Ad.HEL_46–61/I-A^k and Ad.HEL_46–61/I-A^k.B7-1, were capable of activating the 3A9 THy without the addition of HEL_46–61 peptide. However, the former only possessed a weak capacity of activating T cells. When the costimulatory B7-1 molecule was coexpressed by Ad.HEL_46–61/I-A^k.B7-1, T-cell activation by the virus-infected cells was significantly boosted, indicating that the two-signal pathway is required in our system for optimal T-cell activation. It is noteworthy that the level of B7-1 expression in Line 1 cells infected with Ad.I-A^k.B7-1 was somewhat lower than the one with Ad.HEL_46–61/I-A^k.B7-1 (Fig. 2a, b), even though DNA sequencing revealed no difference between these two viruses. Consequently, peptide-specific THy activation by Ad.I-A^k.B7-1 in the presence of HEL_46–61 peptide was also less vigorous than that induced by Ad.HEL_46–61/I-A^k.B7-1 (Fig. 3b). In summary, our approach of expressing B7-1 and HEL_46–61/I-A^k by the same adenovirus virtually guarantees that the cells infected by the adenovirus express both molecules on the cell surface. Thus, this strategy may lead to significantly higher efficiency to activate epitope-specific T cells in animal studies than a mixture of two adenoviruses expressing antigen and B7-1 separately.

Adenovirus infects a wide range of host cells, including hepatocytes (Fig. 5c), endothelial cells and muscle cells¹⁹ the chance for adenovirus to infect APCs in vivo is far less than that for non-APCs. One of the intriguing questions is which cells, APCs or non-APCs, play a more important role for the peptide-specific T-cell activation in our in vivo studies. Professional APCs express endogenous costimulatory molecules, such as B7-1 or B7-2.^5,14,15 If the peptide expressed by either Ad.HEL_46–61/I-A^k or Ad.HEL_46–61/I-A^k.B7-1 is presented to T cells primarily by APCs, HEL_46–61-specific T-cell priming may be expected in the absence of exogenous expression of B7-1. However, the activation of HEL_46–61-specific T cells by Ad.HEL_46–61/I-A^k.B7-1, but not by Ad.HEL_46–61/I-A^k (Fig. 4), suggests that the HEL_46–61 peptide may be mainly presented by virus-infected non-APCs, but not by APCs. This was further confirmed by our in vivo study that MHC class II⁺ cells from mice infected with Ad.GFP did not express GFP (Fig. 5b), even though LPS-activated APCs in in vitro culture may be susceptible to adenovirus infection.²⁰ Furthermore, adenovirus fails to transduce the macrophages solicited by thioglycollate from mouse peritoneal cavity. Insusceptibility of naïve lymphocytes to adenovirus infection was also previously reported.²¹ Therefore, it can be assumed that most of the cells infected by adenovirus in vivo were non-APCs. An explanation for the differential T-cell activation by B7-1 coexpression can be offered by the fact that the covalent HEL_46–61/I-A^k complex expressed by virus-infected non-APCs possesses an intrinsic antigen presenting capacity, while the viral antigens have to be processed by the conventional antigen presentation pathways through endocytosis in APCs and presented by endogenous I-A^k, as depicted in Fig. 6.

Figure 6.

Proposed mechanisms for differential T-cell activation by exogenous B7-1. Co-expression of the covalent HEL_46–61/I-A^k complex and B7-1 in adenovirus-infected non-APCs activates HEL_46–61-specific T cells. Viral antigens have to be presented to and activate T cells by endogenous class II and B7-1 in APCs.

The CD4⁺ T cells express a growing list of accessory molecules, including CD28, cytotoxic T-lymphocyte antigen-4, Fas (CD95), and programmed death-1, etc.^{5,14,15,22,23} The role for many of these molecules in TCR-mediated T-cell activation is not fully defined. Our approach of coexpressing the covalent peptide/class II complex and a ligand for an accessory molecule on the surface of the same cell by the same adenovirus provides a unique approach to delineate the functional role of these molecules in TCR signalling in the context of the immune system. More importantly, recombinant adenoviruses coexpressing peptide/class II molecule and FasL or programmed death-l ligand may induce epitope-specific CD4⁺ T-cell deletion in a TCR-guided fashion and, subsequently, peptide-specific T-cell tolerance. With the available 3A9 TCR transgenic mice²⁴ our strategy provides a valuable approach not only to study the modulation of the T-cell response in a peptide-specific fashion in vivo, but also to investigate the role of other accessory molecules in TCR-class II ligation.

Abbreviations

Ad

adenovirus

APC

antigen-presenting cell

GFP

green fluorescent protein

HEL

hen egg lysozyme

MOI

multiplicity of infection

TCR

T-cell receptor

THy

T-cell hybridoma