Adoptive transfer of tracer alloreactive CD4⁺ TCR-transgenic T cells alters the endogenous immune response to an allograft

Abstract

T cell receptor transgenic (TCR-Tg) T cells are often used as tracer populations of antigen-specific responses to extrapolate findings to endogenous T cells. The extent to which TCR-Tg T cells behave purely as tracer cells or modify the endogenous immune response is not clear. To test the impact of TCR-Tg T cell transfer on endogenous alloimmunity, recipient mice were seeded with CD4⁺ or CD8⁺ TCR-Tg or polyclonal T cells at the time of cardiac allograft transplantation. Only CD4⁺ TCR-Tg T cells accelerated rejection, and unexpectedly led to a dose-dependent decrease in both transferred and endogenous T cells infiltrating the graft. In contrast, recipients of CD4⁺ TCR-Tg cell exhibited enhanced endogenous donor-specific CD8⁺ T-cell activation in the spleen and accelerated alloantibody production. Introduction of CD4⁺ TCR-Tg T cells also perturbed the intra-graft accumulation of innate cell populations. Thus, transferred CD4⁺ TCR-Tg T cells alter many aspects of endogenous alloimmunity, suggesting that caution should be used when interpreting experiments utilizing these adoptively-transferred cells, as the overall nature of allograft rejection may be altered. These results may also have implications for adoptive CD4⁺ T cell immunotherapy in tumor and infectious clinical settings as cell infusion may have additional effects on natural immune responses.

Introduction

In animal studies, T cell receptor transgenic (TCR-Tg) T cells, which are monoclonal T cells specific for a known antigen, are often used as a convenient tool for the study of antigen-specific T cell responses. It is easy to obtain sufficient numbers of antigen-specific T cells from TCR-Tg mice and these cells, once transferred, can be identified by their particular TCR Vα or Vβ subunit, a clonotypic antibody, or a congenic marker. This approach has been used to track the fate and function of antigen-specific T cells including their deletion, proliferation, differentiation, or cytokine production (1,2). Furthermore, this approach with its ability to transfer a set number of cells has allowed estimation of the number of antigen-specific cells required to clear an infection, reject an organ or tumor, or to overcome a therapeutic intervention (1–4). The fate and numbers of the TCR-Tg cells are then often extrapolated to endogenous T cell responses. Importantly, it has been demonstrated in models of infection that TCR-Tg T cells can compete with endogenous T cells specific for the same antigen (5). These observations suggest that the introduced TCR-Tg T cells are not simply recruited into the immune response but they may also in fact alter the immune responses on which they are supposed to report.

In this study, we transferred TCR-Tg and polyclonal T cells into mice receiving fully allogeneic cardiac allografts to investigate their fate and effects on the endogenous alloimmune response. The transfer of 10⁴–10⁵ CD4⁺ TCR-Tg T cells/mouse, a similar dose to what has been frequently used for mechanistic studies of allograft rejection and tolerance, significantly modified the endogenous immune response to the allograft, increasing both CD8 and alloantibody responses, while paradoxically limiting the accumulation of intra-graft T cells and altering the profile of graft-infiltrating innate immune cells. These pleiotropic alterations in the endogenous immune response as a result of CD4⁺ TCR-Tg T cell transfer urge caution for the use of these cells in animal models and patient therapies.

Materials and Methods

Mice

C57BL/6 and BALB/c mice were purchased from Envigo RMS, Inc. (Indianapolis, IN). OT-I Rag-KO, MD4, µMT^−/− and CD45.1 mice were purchased from the Jackson Laboratory (Bar Harbor, ME). Ovalbumin-transgenic (OVA-Tg) mice on a BALB/c background, that express membrane OVA under the control of the β-actin promoter, were a generous gift by Elizabeth Ingulli (when at the University of Minnesota). 2W1S-OVA mice were provided by James Moon (Harvard) and crossed once to BALB/c mice to generate F1 mice. TCR75 TCR-Tg mice were obtained from R. Pat Bucy (University of Alabama-Birmingham) and crossed to CD45.1 mice. TEa TCR-Tg mice were obtained from Alexander Rudensky (when at the University of Washington) and crossed to CD45.1 mice. Mice were housed under specific pathogen-free conditions and used in agreement with the University of Chicago’s Institutional Animal Care and Use Committee, according to the National Institutes of Health guidelines for animal use.

Heart transplantation

Transplantation of cardiac allografts was performed using a technique adapted from that originally described by Corry et al. (6). Cardiac allografts were transplanted in the abdominal cavity by anastomosing the aorta and pulmonary artery of the graft end-to-side to the recipient's aorta and vena cava, respectively. The day of rejection was defined as the first day of cessation of heartbeat in the graft. Graft rejection was verified in selected cases by necropsy.

Adoptive transfer of cells

Spleen and lymph node cells were isolated from donor mice. Cells were enumerated with an Accuri C6 flow cytometer (BD Biosciences) and a subset of cells were stained for CD4 or CD8, Vβ or Vα, the congenic marker CD45.1 or CD45.2 and CD44. The percentage of CD44^lo, congenically marked TCR-Tg T cells was used to calculate the total number of cells for the adoptive transfer. Cells were injected retro-orbitally in 200 µl of phosphate-buffered saline (PBS) on the day of transplantation unless otherwise indicated.

Isolation of graft-infiltrating cells

Cardiac grafts were rinsed in situ with Hanks balanced salt solution (HBSS) containing 1% heparin. Explanted hearts were cut into small pieces and digested for 40 min at 37°C with 400 U/ml collagenase IV (Sigma), 10 mM N-2-hydroxyethylpiperazine-N′-2- ethanesulfonic acid (HEPES, Cellgro) and 0.01% DNase I (MP Biomedicals) in HBSS (Cellgro). Digested suspensions were washed with an equal portion of complete Dulbecco's Modified Eagle medium (DMEM) containing 5% fetal bovine serum, passed through a nylon mesh and centrifuged. Cells were either used for flow cytometry or incubated for 4 h with phorbol myristate acetate (PMA) and ionomycin in the presence of brefeldin A.

Intracellular staining and flow cytometry

Lymphocytes were isolated from spleens and heart grafts and processed into single cell suspensions. Cells were stained first with a fixable live/dead stain (Aqua, Invitrogen) and then with anti-CD4 (L3T4), anti-CD8 (Ly2), anti-CD45.1 (A20), anti-Vβ8-3 (1B3.3) or anti-Vα2 (B20.1) and anti-CD45.2 (104), anti-CD11b (M1/70), anti-Gr-1 (RB6-8C5), anti-TCRβ (H57–597), anti-CD19 (eBio1D3), anti-CD45 30-F11) mAbs. For interferon gamma (IFNγ) intracellular staining, cells were stimulated with PMA (Sigma), ionomycin (Sigma) and brefeldin A (BD Pharmingen) for 4 h, prior to staining for viability and surface markers. Surface-stained cells were then fixed with the Foxp3 fixation permeablization buffer kit (eBioscience) for 15 min at room temperature, washed with 1 × permeabilization buffer, stained using anti-IFNγ (XMG1.2) for 30 min at room temperature, washed again, and analyzed by flow cytometry. For T follicular helper (Tfh) analysis, cells were stained with surface antibodies and CXCR5-bio for 2 hours at 4°C, followed by two washes with FACS buffer and 30 minute staining with streptavidin-Percp-e710 at 4°C. For staining B cells, anti-B220 (RA3-6B2), anti-Fas (15A7) and anti-GL-7 (GL7) were used as well as a dump gate including anti-CD3 (145-2C11), anti-Ter119 (TER119), anti-F4/80 (BM8), anti-CD49b (DX5), and anti-Gr-1 (RB6-8C5). All mAbs were from BD Biosciences or eBioscience.

Multimers and monomers

K^b:OVA pentamers were used to stain OVA-specific CD8⁺ T cells (ProImmune) following T-cell magnetic enrichment by negative selection to remove CD19⁺ and CD11b⁺ cells. One test was used to stain 5 × 10⁶ cells resuspended at 50 × 10⁶ cells/ml.

H2-K^d- biotin monomers, H2-I-E^d -biotin monomers, H2-K^d or H2-I-E^d tetramers were obtained from the NIH tetramer Core Facility (Atlanta, GA). The peptides bound to H2-K^d and H2-I-E^d were SYIPSAEKI (Plasmodium berghei circumsporozoite peptide 252–260) (7) and SSIEFARL (herpes simplex virus glycoprotein B peptide 498 to 505) (8), respectively. In some experiments, magnetic enrichment of tetramer positive B cells was performed before cell surface staining as previously described (9). For staining 5 × 10⁶ cells, saturating (0.1 µg) concentrations of each tetramer were used.

ELISpot

For detection of total IgG secreting B cells, plates (Millipore, Cat: MAIPSWU10) were coated with anti-mouse IgG F(ab)’ specific antibody (Jackson Immunoresearch, Cat: 115-005-072) overnight at 4 °C, then washed with PBS, and blocked with DMEM (10% FBS) 2 hours at 37°C. Cells were plated and cultured overnight, then washed and alkaline phosphatase conjugated anti-mouse IgG Fc specific antibody was added (Jackson Immunoresearch, Cat: 115-055-071). After 1 hour at room temperature, plates were washed and developed with substrate BCIP/NPT (Sigma, Cat: B5655-25tab). For detecting H2-K^d specific IgG secreting plasma cells, plates were coated with anti-mouse IgG Fc-gamma specific antibody overnight at 4°C, washed and blocked. Cells (10³, 10⁴, and 10⁵) were then plated and cultured overnight. Plates were washed and H2-K^d-biotin monomers were added. After one hour at room temperature, plates were washed and streptavidin-ALP (Mabtech, Cat:3310-10) was added. After one additional hour incubation at room temperature, plates were then washed and developed. The numbers of spots per well were enumerated using the ImmunoSpot Analyzer (CTL Analyzers LLC).

Donor-specific antibody (DSA) assay

For donor-specific antibody detection, serum was collected before and after transplantation and was diluted (1:50) and incubated with 1 × 10⁶ BALB/c splenocytes resuspended at 20 × 10⁶ cells/ml for 1 hour at 4°C. After washing twice, the cells were incubated with anti-IgG (Southern Biotech, Cat. # 1030-02) or anti-IgM (eBioscience, Cat. # 125790-82) and anti-B220 (BD Biosciences, Cat. # 561226) for 30 minutes. Mean fluorescence intensity (MFI) was measured by flow cytometry gating on B220^lo cells.

Results

CD4⁺ TCR-Tg T cell transfer leads to reduced intra-graft T cells

Adoptive transfer of antigen-specific TCR-Tg T cells is a tool most often used to follow monoclonal T cell responses and extrapolate conclusions to endogenous responses. In order to determine whether the adoptive transfer of TCR-Tg cells alters endogenous immune responses, a model of fully MHC-mismatched cardiac allograft transplantation with and without TCR-Tg cell transfer was examined. 1,000–250,000 (1K-250K) naïve CD4⁺ TCR75 T cells that recognize a BALB/c K^d peptide presented on I-A^b were transferred into C57BL/6 mice on the day of transplantation with a BALB/c heart (Figure 1A). The time of rejection was slightly accelerated in mice receiving 100K TCR75 cells (MST Day 6 vs Day 8 with no transfer, Figure 1B) and the majority of CD4⁺ cells present in the graft were TCR75 (79.4%±13.4%, Figure 1C). Unexpectedly, the TCR75 cell transfer led to a dose-dependent decrease in the total number of T cells within the rejected allografts, with the most striking reduction being in the number of endogenous CD8⁺ T cells (Figure 1D). Whereas graft-infiltrating CD8⁺ T cells vastly outnumbered CD4⁺ T cells in unmodified acute rejection (AR), the addition of 100K or more TCR75 cells not only reduced the numbers of T cells in the graft but also shifted the ratio of CD8:CD4 T cells closer to 1:1 (Figure 1D).

Figure 1. Adoptive transfer of CD4⁺ TCR-Tg T cells prevented accumulation of transferred and endogenous CD4⁺ and CD8⁺ T cells.

A. Experiment design. B. Graft survival following transplantation without T cell transfer (∅, n=14) or with 10⁵ TCR75 (100K, n=7), 2.5 × 10⁵ TEa (250K, n=6), 10⁵ OT-I (100K, n=6), or 2.5 × 10⁵ CD4⁺ plus 2 × 10⁵ CD8⁺ polyclonal T cells (poly, n=6). C. Representative flow cytometry plots of intra-graft T cells of mice receiving TCR75 (10K–250K cells transferred, n=16), TEa (250K cells transferred, n=6), OT-I (100K cells transferred, n=6) and polyclonal T cells (250K CD4⁺ + 200K CD8⁺ cells transferred, n=6). Numbers are the mean percentage of transferred cells out of CD4⁺ or CD8⁺ graft-infiltrating T cells ± SD. D. Numbers of endogenous CD4⁺, TCR-Tg, and CD8⁺ cells infiltrating the grafts 6–8 days post-transplantation, and ratio of CD8:CD4 T cells; ∅(no transfer, n=14), TCR75: 1K (n=6) 10K (n=3) 100K (n=6) 250K (n=7), TEa: 250K (n=6), OT-I: 100K (n=6), polyclonal (250K CD4⁺ + 200K CD8⁺ cells transferred, n=6). E. Numbers of total CD4⁺ and CD8⁺ T cells infiltrating the grafts on days 3, 5, or 7–8 post-transplantation; ∅ (no transfer, n=3,3,14), +100K TCR75: (n=3,3,6). Data from Day 7–8 is the same as in 1D. Results are pooled from 2–10 independent experiments. Survival data was analyzed by log-rank test, and mean values were compared using one-way analysis of variance (ANOVA) (D) or two-way ANOVA (E) with Bonferroni correction for pairwise comparisons, ns=not significant, *p<0.05, **<p<0.01, ***p<0.001

We also examined the effects of an adoptive transfer of a second CD4⁺ TCR-Tg T cell, TEa, which recognizes an I-E^d peptide presented on I-A^b (Figure 1D). In contrast to the TCR75 cells, TEa cells comprised only a small proportion of the total CD4⁺ T cells within the allografts (4.4%±5.1%, Figure 1C). Despite this difference, TEa cells also mediated accelerated graft rejection that was characterized by reduced total numbers of graft-infiltrating T cells compared to rejection in the absence of TCR-Tg CD4⁺ T cells.

Accelerated graft rejection and reduction in total intra-graft T cells was specific to adoptive transfer of antigen-specific CD4⁺ TCR-Tg T cells, as transfer of similar numbers of naïve CD8⁺ TCR-Tg OT-I T cells into recipients of OVA-expressing BALB/c allografts or transfer of wild-type polyclonal CD4⁺ CD45.1⁺ T cells, which contain lower percentages of alloreactive T cells and a majority of T cells of irrelevant specificities, did not have these effects (Figure 1B,D). In fact, adoptive transfer of OT-I cells resulted in an increase in overall numbers of intra-graft T cells (Figure 1D).

One possibility for observing reduced numbers of T cells in Day 7-grafts from mice with CD4⁺ TCR-Tg T cell transfer was that we may have missed an earlier peak of T cell infiltration and T cells may have died or exited the graft by the time the mice were analyzed on Day 7. However, a time course analysis revealed that mice transferred with 100K TCR75 cells had few intra-graft T cells on days 3 and 5 post-transplantation, similarly to non-transferred animals (Figure 1E). Thus the addition of modest numbers of graft-specific CD4⁺ but not CD8⁺ TCR-Tg T cells profoundly limited the total accumulation of intra-graft CD4⁺ and CD8⁺ T cells.

CD4⁺ TCR-Tg T cells enhance graft-specific CD8⁺ T cell priming in the spleen

Another possibility for the observed reduction in intra-graft T cell numbers following CD4⁺ TCR-Tg T cell transfer was that the TCR-Tg cells were interfering with the priming of endogenous T cells in the spleen. To investigate this possibility, activated CD44^hi TCR-Tg and endogenous T cells were enumerated in the spleen 7–8 days following adoptive transfer of TCR75, TEa, OT-I or polyclonal T cells (Figure S1 and Figure 2). Rather than inhibiting T cell priming, transfer of CD4⁺ TCR-Tg T cells, but not OT-I or polyclonal cells, provided additional help to endogenous CD8⁺ T cells, with overall increased percentages and numbers of CD44^hi (Figure 2A) and interferon gamma (IFNγ)-producing (Figure 2B) CD8⁺ T cells in the spleens of those mice receiving TCR75 or TEa cells. There was no significant change in the percentages or numbers of regulatory T cells in the spleens of mice receiving CD4⁺ TCR-Tg T cells (Figure S2), suggesting that the CD4⁺ TCR-Tg T cells are not promoting stronger CD8⁺ T cell responses by affecting total numbers of Tregs or ratios of Tregs to T effector cells.

Figure 2. Adoptive transfer of CD4⁺ TCR-Tg T cells augmented endogenous CD8⁺ T cell responses in the spleen.

Percentages and numbers of CD44^hi (A) or IFNγ⁺ (B) endogenous CD8⁺ T cells in the spleens of mice receiving no transfer of cells (n=22), adoptive transfer of TCR75 (10K–250K cells, n=12–13), TEa (250K cells, n=5–6), OT-I (100K cells, n=6) or polyclonal (poly, 250K CD4⁺ + 200K CD8⁺ cells, n=6), and did (acute rejection, AR, with or without transfer) or did not (naïve) receive a heart transplant analyzed on day 7–8 post-transfer or post-transplantation. Mean values were compared using one-way analysis of variance (ANOVA) with Bonferroni correction for pairwise comparisons, ns=not significant, *p<0.05, ***p<0.001 for TCR75 groups #p<0.05 for TEa groups.

CD4⁺ helper T cells are so named for their ability to help other immune cells, including CD8 responses. To determine if TCR75 cell transfer could impact antigen-specific endogenous CD8⁺ T cells that recognize a different alloantigen than the K^d antigen recognized by the TCR75 cells, B6 mice were transplanted with BALB/c hearts expressing OVA so that endogenous CD8⁺ T cell responses could be tracked using K^b:OVA multimers (Figure S3, 3A and 3B). Adoptive transfer of 100K TCR75 cells resulted in a 14-fold increase in the number of OVA-specific CD8⁺ T cells within the spleen (Figure 3B and 3C). These results indicate that TCR75 transfer can alter endogenous CD8⁺ T cell responses that do not require the recognition of the same alloantigen. However, despite the increased accumulation of splenic K^b:OVA-reactive cells in the spleen, the TCR75 transfer resulted in a log-decrease in the numbers of intra-graft K^b:OVA-reactive CD8⁺ T cells (Figure 3C). Thus TCR75 transfer enhanced non-cognate endogenous allospecific CD8⁺ T cell accumulation in the spleen but inhibited their accumulation in the graft in a similar manner to what was observed for polyclonal CD8⁺ T cells.

Figure 3. TCR75 cell transfer affected endogenous CD8⁺ T cell responses even when they recognized distinct antigens.

A. C57BL/6 mice were transplanted with OVA-expressing BALB/c × B6 F1 hearts and a subset of mice were adoptively transferred with 100K TCR75 cells. B. Percentage of K^b:OVA⁺ cells in the spleen ± SD (n=5–6 per group). C. Numbers of endogenous K^b:OVA⁺ cells identified in the spleen (n=5–6 mice per group) and graft (n=6 per group) 8–9 days post transplantation. All results pooled from 2 independent experiments. Mean values were compared using Student’s t test, *p<0.05, **p<0.01.

CD4⁺ TCR-Tg T cells enhance allograft-specific B cell responses

In addition to helping CD8⁺ T cell responses, CD4⁺ T cells are also known to provide B cell help. To determine whether the transferred CD4⁺ TCR75 cells were similarly providing help to alloreactive B cells, splenocytes were first stained with CXCR5 and PD-1, which are markers of T follicular helper cells (Tfh cells). A high proportion of transferred TCR75 cells differentiated into Tfh cells while the number of endogenous Tfh cells remained similar in transferred versus non-transferred mice (Figure S4); thus, in mice receiving TCR75 cells, the total numbers of Tfh cells were effectively doubled (Figure 4A). This finding correlated with an expansion of allospecific B cells that recognized K^d, which reflects their ability to engage in cognate interaction with the TCR75 cells (Figure 4B). The K^d–specific B cells were not enriched in cells expressing germinal center markers (GL-7⁺, Fas⁺, data not shown). Instead, they had already begun to differentiate into plasmablasts that downregulate B220 (Figure 4C) and CD19 (Figure S5) by the time of acute rejection. Indeed, as measured by ELISpot, TCR75 transfer increased antibody-producing splenocytes at day 7 post-transplantation (Figure 4D) and serum IgM and IgG alloantibodies as early as 6 days post-transplantation (Figure 4E).

Figure 4. CD4⁺ TCR-Tg T cell transfer led to increased cognate and non-cognate alloantibody responses.

A. Percentages and total numbers of Tfh cells (PD-1^hi, CXCR5^hi) B. Numbers of K^d-specific (MHC I) and I-E^d specific (MHC II) B cells in the spleens of naïve mice or mice undergoing acute rejection with or without TCR75 transfer. C. Percentages of B220^lo cells amongst MHC I- or MHC II-specific B cells. For A–C, n=4–6 per group pooled from two independent experiments. D. ELISpot of antibody-producing cells that made alloantibodies (total IgG) or to specifically to K^d. 10⁴ cells were plated per well. Data are representative of two experiments of triplicate wells for each of 2–3 mice per group. E. Donor-specific antibodies were measured in the serum one day prior to transplantation and on days 4 and 6 post-transplantation. n=3 mice per group, representative of 2 independent experiments. F and G. Survival of hearts post-transplantation in wild-type (WT) or µMT^−/− mice with or without 10⁵ TCR75 cells (+100K) (F) and numbers of endogenous CD4, CD8, and TCR-Tg cells infiltrating the grafts 6–8 days post transplantation (G); ∅(no transfer, n=14), 100K TCR75 (n=6) µMT^−/− (n=6), µMT^−/− + 100K TCR75 (n=6). For F & G, data from WT and WT+100K TCR75 are the same as those shown in Figure 1, and data are pooled from two experiments. Mean values were compared using Student’s t-test or one-way ANOVA with Bonferroni correction for pairwise comparisons where appropriate **p<0.01, ***p<0.001.

We also quantified the numbers of Class II (I-E^d)-specific B cells, which are not able to engage in cognate interactions with TCR75 cells. While the total numbers of I-E^d-specific B cells were not significantly increased, the proportion of these B cells that downregulated B220 was significantly augmented (Figure 4C). In addition, the ELISpot revealed many more total antibodies produced than those specific for K^d (Figure 4D), illustrating that TCR75 transfer enhances the differentiation of alloreactive B cells that can engage in both cognate and non-cognate interactions with TCR75 cells.

To further illustrate the potency of the help provided by transferred CD4⁺ TCR-Tg T cells, we transplanted the quasi monoclonal MD4 BCR-Tg mice with or without transferred TCR75 cells. As 90% of B cells in MD4 mice are specific for Hen Egg Lysozyme (HEL) (10), these animals provide a model in which most B cells are not capable of producing alloantibodies. Remarkably, transfer of 100K TCR75 cells was able to drive allospecific IgG production by day 6 in these mice (Figure S6A). These observations underscore the potency of transferred TCR75 cells at providing B cell help, even in the face of a dramatically reduced alloreactive B cell repertoire. Notably, TCR75 transfer into MD4 hosts still resulted in reduced intra-graft T cell infiltration (Figure S6B).

Biopsies of patients undergoing acute antibody-mediated rejection often contain minimal cellular infiltrates (11). Because of the reduced intra-graft cellular infiltrate upon CD4⁺ TCR-Tg T cell transfer, we tested the possibility that the allografts in the presence of TCR75 underwent antibody-mediated instead of T cell-mediated rejection. To test whether alloantibodies and their downstream effects were inhibiting intra-graft T cell accumulation in animals that had received CD4⁺ TCR-Tg cell transfer we used µMT^−/− mice, which lack membrane-bound IgM and mature B cells. µMT^−/− mice did not make allo-IgM or allo-IgG upon transplantation and transfer of 100K TCR75 cells (Figure S6C). However, accelerated graft rejection (Figure 4F) and a reduction of T cells within the allografts were still observed (Figure 4G). Collectively these observations support the conclusion that early production of alloantibodies was not required for the reduction of intra-graft T cell accumulation following CD4⁺ TCR-Tg T cell transfer.

The transfer of CD4⁺ TCR-Tg T cells alters intra-graft accumulation of innate immune cells

To determine if CD4⁺ TCR-Tg T cell transfer had additional consequences on the endogenous alloresponse and gain insight into the mechanism of graft rejection, we analyzed the phenotype of other intra-graft leukocytes. TCR75 cell transfer resulted in a dose-dependent decrease in B cells and CD11c⁺ dendritic cells in the grafts (Figure 5A), and a striking increase in neutrophils (Figure 5B). Intra-graft CD11c⁺ cells were also decreased and neutrophils increased when TEa cells were transferred (data not shown). This reduction in graft-infiltrating antigen-presenting cells was unique to the CD4⁺ TCR-Tg T cell transfer, as OT-I transfer had no effect on B cell accumulation, and increased CD11c⁺ cells within the allografts (Figure S7A), although similarly to the CD4⁺ TCR-Tg T cell transfer, OT-I transfer also resulted in increased neutrophils in the graft (Figure S7A). Antigen specificity was required for T cell transfer to have an effect on accumulation of endogenous immune cell populations, as transfer of 250K polyclonal CD4s plus 200K polyclonal CD8s had no effect on B cell, CD11c⁺ cell or neutrophil accumulation compared to mice without transfer (Figure S7B).

Figure 5. Transfer of TCR75 cells altered the numbers of non-T cell populations in the graft.

C57BL/6 mice were transplanted with BALB/c heart grafts. Six to eight days post-transplantation intra-graft antigen-presenting cells and neutrophils were enumerated by flow cytometry. A. Numbers of intra-graft B cells and CD11c⁺ cells in mice with no transfer (∅, n=14), 10K TCR75 cells (n=3) or 100K TCR75 cells (n=6). B. Representative flow cytometry plots of mean percentages of graft-infiltrating neutrophils (Gr-1^hi, CD11b^hi) ± SD after gating on CD45⁺TCRβ⁻CD19⁻ cells and their total numbers in mice with no transfer (n= 17), 10K TCR75 cells (n=3) or 100K TCR75 cells (n=6). For A and B, mean values were compared using one-way ANOVA with Bonferroni correction for pairwise comparisons *p<0.05, **p<0.01, ***p<0.001. C. Numbers of intra-graft CD11c⁺ cells versus total T cells are positively correlated, Pearson r=0.66, p<0.0001. Results are pooled from 2–10 experiments.

As the total numbers of graft-infiltrating T cells was lowest in mice with TCR75 transfer and highest in mice with OT-I transfer, and this trend was recapitulated with intra-graft CD11c⁺ cells, we tested whether there was a correlation between the two factors. To this end, data from 38 animals were combined from each of the groups of non-transfer, TCR75, TEa, OT-I and polyclonal T cell transfer. There was a robust positive correlation between the numbers of CD11c⁺ cells and the total number of intra-graft T cells (Figure 5C, Pearson r=0.66, p<0.0001), suggesting that transfer of CD4 or CD8 TCR-Tg cells may affect the numbers of intra-graft CD11c⁺ cells, and these intra-graft CD11c⁺ cells may in turn affect the total numbers of intra-graft T cells, as has been described in tumor settings (12).

Discussion

TCR-Tg T cells are useful tools for following antigen-specific immune responses in vivo but they must be used cautiously. Using infectious disease models, it has been previously shown that OVA-specific OT-I cells can effectively compete for antigen with endogenous OVA-reactive CD8⁺ T cells, such that only transfer of less than 500 TCR-Tg OT-I cells can preserve the endogenous T cell response to the same antigen (5). Our data is consistent with this conclusion as few endogenous OVA-specific T cells were detectable in the graft following OT-I transfer (K^b:OVA⁺CD45.1⁺, Figure 1C). Badovinac et al as well as others reported that CD8 memory T cell differentiation kinetics depended on initial precursor frequency (5,13,14). Ford and colleagues have demonstrated different levels of resistance of TCR-Tg T cells to costimulation blockade based on their initial seeded frequency in a skin transplant model (2). For CD4⁺ TCR-Tg cells, Marshall et al compared the magnitude of the response of CD4⁺ SMARTA T cells with that of endogenous T cells reactive to the same LCMV antigen, but in separate mice (15). This study concluded that SMARTA T cells were representative of the endogenous response but no analysis of endogenous responses in the mice that received SMARTA T cells, compared to those that did not receive SMARTA cells, was shown (15).

Our study shows, in contrast, that the addition of naïve CD4⁺ TCR-Tg T cells not only likely competes with T cells recognizing the same antigen but also alters the response of CD4⁺ and CD8⁺ T cells recognizing other donor-specific antigens following allograft transplantation. The numbers and types of non-T cell subsets able to accumulate within the allografts were also altered by the addition of modest numbers of CD4⁺ TCR-Tg T cells, further demonstrating a significant reshaping of the endogenous alloresponse by the transferred graft-reactive T cells. Of the hematopoietic cell populations that were altered in the graft, the reduction in intra-graft T cells correlated most strongly with a reduction in the number of intra-graft CD11c⁺ cells. This result is interesting in the context of recent data from Spranger et al who showed in a melanoma model that the presence of CD11c⁺ cells positively correlated with intra-tumoral T cells, and that the addition of dendritic cells intra-tumorally allowed for increased T cell accumulation (12). Antigen-specific CD4⁺ T cells have been shown to have the ability to kill antigen-presenting cells (APCs) (16,17), and APCs have been shown to be particularly sensitive to CD4-mediated killing in the presence of high numbers of CD4⁺ TCR-Tg T cells (18). Therefore, a possible explanation for the reduction in CD11c⁺ cells observed in allografts with TCR75 transfer is that the TCR-Tg cells may have killed some of the APCs. The reduced number of APCs could then lead to a reduction in the chemokines produced by APCs that attract T cells or retain them within the allograft.

The reshaped endogenous immune response observed with TCR75 cell transfer was also observed with transfer of T cells from a second CD4⁺ TCR-Tg mouse, TEa, although the TEa cells infiltrated the grafts at a lower frequency (Figure 1) and accumulated in the spleen at lower numbers (Figure S1). Differences in intra-graft accumulation of these two types of TCR-Tg T cells may be due to the respective antigens recognized by each TCR (K^d for the TCR75 cells which is expressed by all cells in the allograft versus I-E for TEa cells which is only expressed on professional antigen-presenting cells and activated endothelial cells within the graft), or to a difference in the TCR affinity of each monoclonal population. Despite these differences, our data suggest that transfer of TEa cells is also associated with a decrease in CD11c⁺ cells, correlating with a reduction in the accumulation of endogenous allograft-specific T cells in the allograft.

Many groups utilizing TCR-Tg T cells in transplantation studies have transferred much higher numbers of T cells than those used here, up to 2 × 10⁷ T cells (2,19–21). In this study, as few as 10⁴ TCR75 cells had profound effects on the endogenous immune response, with elevated alloantibody responses detected with as few as 10³ cells (data not shown), and with greater perturbations with increasing numbers of transferred TCR-Tg T cells. In light of these observations, some past studies should be revisited and perhaps re-interpreted. For example, if a new immunosuppressive agent were to be tested in the mice used in this study it would not only have to overcome the effects of the adoptively transferred CD4⁺ T cells specific for the graft, but also would have to counter an enhanced (in magnitude and kinetics) endogenous CD8 and alloantibody response. Thus, the efficacy of the immunosuppressive agent may be significantly underestimated. Future investigations using these cells should limit the numbers of TCR-Tg cells used for transfer so studies may more accurately reflect endogenous precursor frequencies.

Interestingly the naïve CD4⁺ TCR-Tg T cell transfer was able to promote both cognate and non-cognate allospecific B cell responses. This result supports published data showing that B cells are directly able to receive help from non-cognate CD4⁺ T cells provided these B cells express MHC-II to present the antigen and have acquired the non-cognate antigen from the same cell that expresses the cognate antigen (22). Our data also show that transfer of CD4⁺ TCR-Tg T cells promotes a non-cognate alloreactive CD8⁺ T cell response, which invokes an indirect and/or semi-direct model of help. This could occur when the CD4⁺ T cell activates APCs that present both alloantigens, such that the APCs, in turn, can stimulate the CD8⁺ T cells that recognize a distinct alloantigen from that recognized by the CD4⁺ TCR-Tg T cells.

Overall, the data give insight into the “cooperativity” of rejection first described by Bucy and colleagues, who when adoptively transferring TCR75 cells into C57BL/6 recipients of B6.K^d hearts noted a non-linear relationship between the number of cells transferred and the kinetics of rejection (3). This is likely due to reaching a threshold number of CD4⁺ T cells that not only facilitate rejection themselves but also coordinate the rejection event by promoting more CD8⁺ T cell responses in the spleen and a strong alloantibody response. Indeed, in B cell-competent mice, CD4⁺ but not CD8⁺ T cells have been shown to be essential for cardiac allograft rejection (23). However, in B cell-deficient mice depleted of CD8⁺ T cells, the remaining naïve CD4⁺ T cells were incapable of rejecting a cardiac allograft (24), suggesting that CD4 help to either CD8⁺ T cells or B cells is necessary to elicit graft rejection. In contrast, memory CD4⁺ T cells might lead to allograft rejection by other mechanisms, as they could still trigger rejection in B cell-deficient mice depleted of CD8⁺ T cells (24). Our data reveal increasing the numbers of naïve alloreactive CD4⁺ T cells provides enhanced help to CD8⁺ T cells and allospecific B cells but it also leads to an overall reduction in intragraft T cells as well as CD11c⁺ cells. The mechanism of rejection is most likely altered by the presence of these increased numbers of CD4⁺ T cells, as graft rejection ensues without ever achieving high numbers of T cells within the graft and can still be accelerated without alloantibodies as the results of the µMT^−/− experiments showed. It is possible that the transferred CD4⁺ TCR-Tg T cells provide help to allospecific CD8⁺ T cells, which in turn may destroy donor endothelial cells, causing graft rejection before many T cells infiltrate the graft.

Our data showing that transfer of naïve TCR75 cells which results in reduced accumulation of T cells, B cells and dendritic cells in the allograft contrasts with data obtained by Martin-Orozco and colleagues in a tumor model. In this B16-OVA melanoma model, transfer of activated, Th17-skewed CD4⁺ TCR-Tg OT-II T cells enhanced endogenous CD4, CD8, antigen-presenting cell and neutrophil accumulation into the tumor (25). Whether the differences are due to the tumor microenvironment or to the TCR-Tg T cells being pre-activated rather than naïve remains to be established.

CD8⁺ T cell transfer also perturbed endogenous alloimmunity, albeit in different ways than the CD4⁺ T cell transfer. CD8⁺ T cell transfer led to an overall increase in the numbers of graft-infiltrating T cells, CD11c⁺ and neutrophils, though it did not accelerate graft rejection. Taken with CD4⁺ T cell transfer data, the alterations in endogenous alloimmunity following transfer of CD8⁺ T cells provides evidence that there are unique effects of distinct adoptively transferred populations on endogenous alloimmune responses.

Assessing the effects of adoptive transfer of antigen-specific T cells in a clinical setting should be the subject of future investigations as recently completed and ongoing clinical trials are evaluating the efficacy of transferring very large numbers of expanded antigen-specific T cells in settings of tumor and of viral infection, (for example NCT00393029, NCT02210104, NCT00880789, NCT00110578) (26). To date, endogenous immune responses have not been monitored in these patients and it is not known if the adoptive cell therapy would have unintended long-term consequences on the endogenous immune response against the tumor or virus being targeted, or against a concurrent infection. Clinical trials are also ongoing in transplantation, transferring expanded antigen-specific Tregs to prolong allograft survival, anti-tumor T cells to combat post-transplant lymphoproliferative disorder (PTLD) or viral-specific T cells to control infection (NCT02371434, NCT00063648, NCT00880789). Supporting our results, one recent preclinical study using adoptive Treg therapy in a non-human primate model of heart transplantation revealed that transfer of Tregs into a lymphopenic host can result in enhanced alloantibody responses and T cell memory, as the transferred cells provided help to endogenous alloreactive cells (27). Whereas our study assessed the consequences of transferring naïve T cells, many clinical trials use transfer of activated T cells, which may alter the endogenous immune response differently. In addition, humans have a higher proportion of endogenous memory T cells than laboratory mice, which may react differently that naïve T cells to the adoptive cell therapy. The consequences of T cell transfers on endogenous immune responses should be investigated in human clinical trials.

Supplementary Material

Supp Fig S1. Figure S1: Transferred T cells are activated during acute rejection but do not perturb endogenous CD4⁺ T cells.

Percentages and numbers of CD44^hi (A) or IFNγ⁺ (B) transferred and endogenous CD4⁺ T cells in the spleens of mice receiving no transfer of cells (n=22), adoptive transfer of TCR75 (10K–250K cells, n=12–13), TEa (250K cells, n=5–6), OT-I (100K cells, n=6) or polyclonal (250K CD4⁺ + 200K CD8⁺, “poly CD4” and “poly CD8”, n=6) T cells, and did (acute rejection, AR, with or without transfer) or did not (naïve) receive a heart transplant analyzed on Day 7–8 post-transfer. Mean values were compared using one-way and two-way ANOVA with Bonferroni correction for pairwise comparisons, ns=not significant, *p<0.05, ***p<0.001.

Supp Fig S2. Figure S2: Regulatory T cell percentages and numbers in the spleen are not changed by adoptive transfer of CD4⁺ TCR-Tg T cells.

The percentages (A) and total numbers (B) of endogenous and TCR-Tg Foxp3⁺ Tregs amongst CD4⁺ T cells in the spleen 7–8 days post-transplantation in mice receiving no transfer (∅, n=14), 10K–250K TCR75 cells (n=13) or 250K TEa cells (n=6).

Supp Fig S3. Figure S3: Gating strategy of K^b:OVA⁺ cells.

Spleen cells from mice undergoing acute rejection of OVA-expressing BALB/c × B6 F1 hearts without transfer or with 100K TCR75 cells were magnetically enriched for T cells and stained with K^b:OVA multimers. As a negative control, CD8⁺ T cells from an irrelevant TCR-Tg mouse were used (TEa on a RAG-sufficient background). Representative of n=5–6 per group from two independent experiments.

Supp Fig S4. Figure S4: Gating of T follicular helper cells.

Representative flow cytometry plots of T follicular helper cells (PD-1^hi, CXCR5^hi) from endogenous CD4⁺ T cells and TCR75 cells in naïve untransplanted mice, or in mice undergoing acute rejection without or with 100K TCR75 cells. Representative of n=4–6 per group from two independent experiments.

Supp Fig S5. Figure S5: Transfer of TCR75 cells increased the numbers of CD19^int cells in the spleen.

A. The mean percentages ± SD (A) and total numbers (B) of CD19^int cells in the spleen day 7–8 post-transplantation. No transfer: ∅ n=6, TCR75: 10K n=3, 100K n=6 pooled from 3–4 independent experiments. Mean values were compared using one-way ANOVA with Bonferroni correction for pairwise comparisons, *p<0.05, ***p<0.001.

Supp Fig S6. Figure S6: Alloantibody production in MD4 and µMT^−/− mice.

A. Allospecific IgM and IgG production in the serum of indicated mice on day 6 post-transplantation. B. Intra-graft T cells from day 6–8 post-transplantation in indicated mice. Data from WT and WT+100K TCR75 are the same as those shown in Figure 1D. C. Allospecific IgM and IgG production in the serum of indicated mice on day 6 post-transplantation. Mean values were compared using one-way ANOVA with Bonferroni correction for pairwise comparisons, **p<0.01, ***p<0.001.

Supp Fig S7. Figure S7: Transfer of CD8⁺ OT-I TCR-Tg cells but not polyclonal T cells changed non-T cell accumulation in the allograft.

The total numbers of endogenous CD19⁺ B cells, CD11c⁺ APCs and Gr-1^hi, CD11b^hi neutrophils in mice with no transfer, (∅, n=14–17) and mice with 100K OT-I transfer (n=6, A) or 250K polyclonal CD45.1⁺ CD4⁺ T cell transfer (n=6, B). Means were compared with Student’s t test and results are pooled from at least two independent experiments. No transfer group is the same in panel A and B. OT-I cells were transferred either 1 day prior to or on the day of transplantation. CD45.1⁺ T cells were transferred on the day of transplantation.

Abbreviations

ANOVA

analysis of variance

APC

antigen-presenting cell

AR

acute rejection

DMEM

Dulbecco modified Eagle medium

DSA

Donor specific antibody

HEL

Hen Egg Lysozyme

HBSS

Hanks balanced salt solution

HEPES

N-2-hydroxyethylpiperazine-N′-2- ethanesulfonic acid

IFNγ

Interferon gamma

MFI

Mean fluorescence intensity

n

number in group

NA

not applicable

ns

not significant

P

probability

PBS

phosphate-buffered saline

PMA

phorbol myristate acetate

PTLD

post-transplant lymphoproliferative disorder

Th

T helper [cell]

OVA

ovalbumin

TCR-Tg

T cell receptor transgenic

Tfh

T follicular helper