Journal Logo

Clinical and Translational Research

Quantitative and Functional Diversity of Cross-Reactive EBV-Specific CD8+ T Cells in a Longitudinal Study Cohort of Lung Transplant Recipients

Mifsud, Nicole A.1,4; Nguyen, Thi Hoang Oanh1; Tait, Brian D.2; Kotsimbos, Tom C.1,3

Author Information
doi: 10.1097/TP.0b013e3181ff4ff3
  • Free

Abstract

Human lung transplantation is now a well-accepted therapeutic option for selected patients with end-stage chronic lung disease. However, despite often excellent early results, a major obstacle to long-term success is late allograft dysfunction with chronic rejection, clinically manifested as bronchiolitis obliterans syndrome, occurring in 40% of lung transplant recipients (LTRs) at 2 years and leading to high mortality rates of 50% at 5 years (1–3). In addition, advances in both surgical and clinical management protocols during the past decade have had little impact on the rates of bronchiolitis obliterans syndrome. Possible explanations for this include the lung allograft's predisposition and susceptibility to injury arising from ischemia-reperfusion, the ever-present risk of alloreactivity, relatively high rates of latent DNA virus reactivation, and constant exposure to environmental infections (community respiratory viruses, bacteria, and fungi), antigens, and pollutants (4, 5).

The antigen-specific memory T-cell pool is shaped by previous immunologic exposure to environmental pathogens that can influence future immune responses through T-cell receptor (TCR) cross-reactivity (heterologous immunity) (6–9). Recent evidence highlights that TCR cross-reactivity is a form of molecular mimicry underpinned by a similarity in major histocompatibility complex (MHC) or peptide landscapes despite fundamental differences of the MHC molecule, the displayed peptide, or a combination of both (10, 11). Moreover, it is increasingly recognized that molecular mimicry, denoted by TCR cross-reactivity, is associated with human T-cell alloreactivity and is contributed by both the naive and memory T-cell compartments (12–14).

Dissection of the alloresponse has revealed that reactivity to donor MHC alloantigens can be mediated by the presence of host T cells representing 1% to 10% of total T-cell repertoire (15–17) and by a high proportion, up to 60%, of MHC-restricted cytotoxic T-lymphocyte (CTL) clones (18). Given the relatively high frequency of circulating memory T cells toward latent DNA viruses (19–22) and the cross-reactive nature of these T cells toward unrelated human leukocyte antigen (HLA) alloantigens (23–29), the potential for graft rejection particularly in association with viral reactivation is heightened and adds another level of complexity to the already recognized links between allograft rejection and DNA virus reactivation, after solid organ transplantation (30–32).

A well-characterized model of human TCR cross-reactivity is the HLA-B8-restricted CD8+ T-cell response toward the Epstein-Barr virus (EBV) epitope FLRGRAYGL (FLR), derived from the EBNA3A latent protein. This cross-reactive T-cell response uses a public immunodominant TCR (LC13) that is virtually invariant between unrelated HLA-B8+ individuals (23, 24, 33), which contrasts to other viral responses that are oligoclonal and are expressed among HLA-divergent individuals (29, 34, 35). Cross-reactive FLR-specific CD8+ T cells have demonstrated allorecognition of HLA-B*4402, B*4405 but not B*4403 (23, 24, 36) and more recently HLA-B*5501 (29, 37). Recent structural analyses revealed that LC13 TCR recognition of HLA-B*4402 and B*4405 is contributed to by presentation of the allotope EEYLQAFTY, derived from the ATP-binding cassette protein, ABCD3 (38). Interestingly, in HLA-B8+, B*4402+ heterozygous healthy individuals, self-tolerance drives T-cell alloreactivity of the FLR-specific CTL response toward other alloantigens, namely HLA-B14 and -B35 (23).

Taken together, human lung transplantation and the EBV model of TCR cross-reactivity provided an opportunity to quantitatively measure the longitudinal dynamics of FLR-specific CD8+ T cells in EBV+ HLA-B8+ LTR who received a B*4402 allograft and to determine their contribution to posttransplant outcomes. This lung transplant study, for the first time, demonstrated a functional divergence between cross-reactive FLR-specific CD8+ T cells isolated from the periphery compared with the allograft. We showed that ex vivo allograft-isolated cross-reactive FLR-specific CD8+ T cells were incapable of interferon (IFN)-γ production but could express a cell surface cytotoxicity marker compared with peripheral FLR-specific CD8+ T cells that had a dual phenotype of cytokine production and cytotoxicity. In addition, we dissected the total alloresponse using HLA-B8/FLR tetramers and revealed the contribution of both cross-reactive FLR-specific (tetramer+) and newly generated anti-B*4402 (tetramer−) CD8+ T cells to the alloresponse and a modulation of the alloreactive hierarchy. Finally, although we demonstrate the presence of the cross-reactive FLR-specific CD8+ T cells both in the periphery and within the allograft, under potent immunosuppressive conditions, there was no observable clinical divergence in EBV seropositive LTR who received a B*4402 (n=5) compared with those who received a B*4403 (n=4) allograft.

RESULTS

Subtyping of HLA-B44 Alloantigens and Monitoring of EBV Reactivity

The study cohort consisted of 11 LTRs (HLA-B8+) who had received a HLA-B44 allograft. Further allelic subtyping of the B44 antigenic group (performed by Victorian Transplantation and Immunogenetics Service, Parkville, Victoria, Australia) revealed that six LTRs (Tx19, Tx26, Tx39, Tx41, Tx42, and Tx66) received a B*4402 allograft, four LTRs (Tx6, Tx7, Tx59, and Tx90) received a B*4403 allograft, whereas one LTR (Tx85) received a heterozygous B*4402, B*4403 allograft. Considering that Tx85 expressed HLA-B molecules, B8 and B*4402, this LTR would not be expected to generate the cross-reactive LC13 TCR-recognizing HLA-B*4402 (23).

All LTRs were EBV seropositive except for one who was EBV seronegative but received an EBV-seropositive organ (Tx39). EBV primary infection or reactivation events were prospectively monitored in plasma samples at all time points used for T-cell cross-reactivity investigations. Results demonstrated no detectable presence of EBV DNA in any of the plasma samples tested.

Measurement of Preexisting EBV-Specific Memory CD8+ T Cells

We initially investigated whether EBV-specific CD8+ T cells in LTR, under immunosuppressive conditions, were measurable ex vivo compared with healthy controls. By using MHC/peptide tetrameric complexes, we quantitated both the HLA-B8-resticted cross-reactive epitope FLR and a noncross-reactive epitope QAKWRLQTL (QAK), both from the latent EBNA3A protein. In EBV-seropositive healthy controls (n=7, H1-H4 and H6-H8), the %FLR-specific CD8+ T cells and %QAK-specific CD8+ T cells were 0.6%±0.5% (range 0.1%–1.7%) and 0.2%±1.4% (range 0.1%–3.9%), respectively (data not shown). In LTR (n=11), longitudinal measurements of EBV-specific CD8+ T cells were quantitated both pretransplant and up to 36 months posttransplant. The percent of FLR-specific CD8+ T cells and QAK-specific CD8+ T cells were 0.95%±0.8% (range 0%–3.5%) and 1.25%±4.9% (range 0%–20.1%), respectively (data not shown). These results indicated that frequencies of measurable CD8+ T cells for both the cross-reactive epitope FLR and noncross-reactive epitope QAK could be detected pretransplant and posttransplant and were comparable with healthy controls.

Expansion of HLA-B8-Restricted Memory FLR-Specific CD8+ T Cells

We next explored whether the in vitro expansion of FLR-specific CD8+ T cells could be achieved in LTR at various time intervals pretransplant and posttransplant (n=11), and whether these frequencies were comparable with those of healthy controls (n=7). Peripheral blood mononuclear cells (PBMCs) were stimulated using autologous PBMC pulsed (2:1 ratio) with the FLR peptide. We demonstrated that the peak response, as measured by IFN-γ production, occurs on day 13 after peptide stimulation (data not shown). Thus, day 13 in vitro cultures were generated to measure the expansion of FLR-specific CD8+ T cells using MHC/peptide (B8/FLR) tetramers. In six of seven healthy controls (H2-H4 and H6-H8), FLR-specific CD8+ T cells were generated (53.5%±18.8%, range 20.6%–68.9%), with the exception of H1 in which no expansion of FLR-specific T cells was observed (data not shown). Expansion profiles of FLR-specific CD8+ T cells on day 13 within the LTR group (n=11) varied considerably both across the LTR and also within individual LTR measured at various posttransplant time intervals (6.3%±19.9%, range 0.3%–65.8%, data not shown). These results demonstrated that in the LTR group, FLR-specific CD8+ T cells can be generated, with a range of frequencies comparable with those exhibited by healthy controls.

Cross-Reactivity of HLA-B8-Restricted FLR-Specific CD8+ T Cells Toward the Alloantigen HLA-B*4402

Previously published studies have demonstrated the in vitro potential of HLA-B8-restricted FLR-specific memory CD8+ T cells to recognize the alloantigen HLA-B*4402 but not the closely related HLA molecule B*4403 in healthy individuals (23, 24, 33, 36, 38). In all four of the healthy controls (H4 and H6-H8) studied, we observed cross-reactivity of FLR-specific CD8+ T cells toward B*4402 as measured by IFN-γ production, after a 6-hr stimulation with a panel of reporter cell lines and FLR peptide alone in (1) ex vivo PBMC (B*4402: 48.0%±25.7%, range 1.9%–59.2%, B*4403: 0.2%±1.1%, range 0%–2.3%) and (2) day 13 T-cell cultures (B*4402: 26.1%±25.7%, range 11.9%–70.1%, B*4403: 0.5%±0.5%, range 0.3%–1.4%; Fig. 1A).

FIGURE 1.
FIGURE 1.:
Cross-reactive potential of FLRGRAYGL (FLR)-specific CD8+ T cells. Quantitation of FLR-specific CD8+ T-cell cross-reactivity producing interferon (IFN)-γ was measured by flow cytometry after a 6 hr stimulation with a panel of transfected cell lines and FLR peptide in ex vivo peripheral blood mononuclear cell (PBMC), day 13 T cells, and ex vivo BAL mononuclear cell (MNC). Both the positive controls (C1R.B8/FLR, FLR peptide) and negative controls (C1R, C1R.B8, T cells alone) responded as expected. Cross-reactivity of FLR-specific CD8+ T cells to recognize human leukocyte antigen (HLA)-B*4402 was observed in healthy donors with varying frequencies (A). In lung transplant recipient, longitudinal pretransplant and posttransplant measures of FLR-specific CD8+ T cells revealed differential recognition profiles toward B*4402 within the ex vivo PBMC population (B), with cross-reactivity becoming more evident in day 13 T-cell cultures (C). Analysis of ex vivo BAL MNC indicated that these cells were incapable of IFN-γ production after stimulation with HLA-B*4402 molecules, although data obtained were limited by cell availability (D).
FIGURE 1.
FIGURE 1.:
(Continued).

We further explored the mechanism of cross-reactivity in LTR to determine whether antiviral memory T cells, in an immunosuppressed environment, were capable of HLA alloantigen cross-recognition. Ex vivo PBMCs (11 LTRs), day 13 T cells (11 LTRs), and bronchoalveolar lavage (BAL) mononuclear cells (MNCs; 5 LTRs) were stimulated against a panel of cell lines (C1R.parental, C1R.B8, C1R.B8/FLR, and C1R.B*4402) and FLR peptide alone for 6 hr, with immune reactivity determined by IFN-γ production. Cross-reactive CD8+FLR+IFN-γ+ T cells that responded to B*4402 were present in the ex vivo PBMC pool in 4 of 11 LTR (1%±9.8%, range 0%–25.5%; Fig. 1B), and in a further 2 of 11 LTRs on the day 13 T-cell cultures (1%±17.8%, range 0%–50.2%; Fig. 1C). Although, the negative controls in these assays were all appropriate, we note that there was a significant variation in the positive control response both between different LTR and within an individual LTR over time, with peptide alone stimulation and the day 13 assay generally being much more sensitive. Interestingly, FLR-specific CD8+ T cells isolated from BAL MNCs were unable to recognize the B*4402 alloantigen or efficiently respond to positive controls (C1R.B8/FLR and FLR peptide alone) as measured by IFN-γ production (Fig. 1D) in all five of the LTRs studied, suggesting that T cells within the allograft were functionally distinct from peripheral CD8+ T cells.

Distinct Functional Subsets of FLR-Specific CD8+ T Cells Within Allograft and the Periphery

To unravel functional differences between FLR-specific CD8+ T cells isolated from within the allograft (represented by BAL MNC) or the periphery (represented by PBMC), we sought to examine three functional FLR-specific CD8+ T-cell subsets, which were determined by IFN-γ cytokine production alone (subset 1), a combination of IFN-γ production and CD107a cell surface expression (degranulation marker as a measure of cytotoxicity; subset 2), or CD107a cell surface expression alone (subset 3). Note that stimulation with FLR peptide alone was not analyzed for subsets 2 and 3 because there are no target cells present for measuring cytotoxicity. Initial studies were conducted in seven healthy controls (H1-H4 and H6-H8) for ex vivo PBMC and day 13 T cells, the latter being generated after in vitro stimulation with FLR-pulsed autologous PBMC. Stimulation for each T-cell population was performed against a panel of cell lines (C1R.parental, C1R.B8, C1R.B8/FLR, C1R.B*4402) and FLR peptide alone for 6 hr and then analyzed by flow cytometry. The ex vivo PBMC data suggested that cross-reactivity of FLR-specific CD8+ T cells toward B*4402 was dynamic as four of seven healthy controls (H1-H3 and H7) almost exclusively comprised subset 3 CD8+ T cells, whereas the remaining three healthy controls (H4, H6, and H8) comprised all three subsets (Fig. 2A). Similar results were observed after further analysis was performed on day 13 T-cell cultures. Except for one donor (H7) whereby the FLR-specific CD8+ T-cell pool now consisted of all three subsets.

FIGURE 2.
FIGURE 2.:
Functional dissection of peripheral cross-reactive FLRGRAYGL (FLR)-specific CD8+ T cells in healthy donors. Flow cytometric analyses of expression of interferon (IFN)-γ or CD107a enabled a functional characterization of cross-reactive FLR-specific CD8+ T cells in healthy donors, after stimulation with cell lines or peptide for 6 hr before measurement of IFN-γ cytokine production and CD107a cell surface expression. Three functional subsets of FLR-specific CD8+ T cells were defined as IFN-γ+ single positive (subset 1), IFN-γ+CD107a+ double positive (subset 2), and CD107a+ single positive (subset 3). Ex vivo peripheral blood mononuclear cell cross-reactivity was represented by all subsets (1–3) in n=3 of 7 and only subset 3 in n=4 of 7 (A). Similar data were obtained in day 13 T-cell cultures with all subsets (1–3) in n=4 of 7 and only subset 3 in n=3 of 7 (B).

Longitudinal dynamics in LTR of the three T-cell subsets for cross-reactivity toward HLA-B*4402 were monitored and measured in (1) ex vivo PBMC (n=9), (2) day 13 FLR-specific CD8+ T cells (n=11), and (3) ex vivo BAL MNC (n=5). Analysis of ex vivo PBMC demonstrated that there was functional diversity displayed across the FLR-specific CD8+ T-cell subsets, with three of nine LTRs exhibiting subsets 1 to 3 (Tx41, Tx59, and Tx90), three of nine LTRs exhibiting subsets 2 and 3 (Tx26, Tx42, and Tx66), one of nine LTR exhibiting subsets 1 and 3 (Tx85), and one of nine LTR exhibiting subset 3 (Tx19; Fig. 3A). A similar analysis of day 13 FLR-specific CD8+ T cells also confirmed functional diversity across T-cell subsets, with two of eleven LTR exhibiting subsets 1 to 3 (Tx6 and Tx59), four of eleven LTR exhibiting subsets 1 and 3 (Tx19, Tx66, Tx85, and Tx90), two of eleven LTR exhibiting subsets 2 and 3 (Tx41 and Tx42), and two of eleven LTR exhibiting subset 3 (Tx7 and Tx26). No cross-reactivity was observed in ex vivo PBMC and day 13 T cells for Tx39 because this individual was EBV seronegative (Fig. 3B). In addition, we investigated the functionality of FLR-specific CD8+ T cells isolated from the allograft (BAL MNC) in five LTRs. Interestingly, the data suggested that cross-reactivity toward B*4402 was measurable in subset 3 in four of five LTRs (range 3.3%–14.1%, Tx41, Tx42, Tx59, and Tx66), with a low response measured in Tx19 (0.6%; Fig. 3C). These results indicated that peripheral FLR-specific CD8+ T cells display functional T-cell subset diversity, which are capable of producing cytokine (IFN-γ) or have a cytotoxic phenotype alone or exhibit dual function of cytokine production and cytotoxicity, whereas a skewed subset phenotype was observed for allograft-isolated FLR-specific CD8+ T cells, which were predominantly cytotoxic.

FIGURE 3.
FIGURE 3.:
Functional divergence between peripheral and allograft-isolated cross-reactive FLRGRAYGL (FLR)-specific CD8+ T cells in lung transplant recipient (LTR). Longitudinal functional profiling in LTR (n=11) of cross-reactive FLR-specific CD8+ T-cell subsets was achieved after a 6-hr stimulation with cell lines or FLR peptide as described in Figure 2. Both ex vivo peripheral blood mononuclear cell (PBMC; A) and day 13 T-cell cultures (B) showed subset variation across different LTR (n=9). However, generally, ex vivo PBMCs were more skewed toward a cytotoxic T-cell profile compared with day 13 T cells, which produced greater interferon (IFN)-γ. BAL mononuclear cells (n=5) were unable to produce IFN-γ (subsets 1 and 2) but demonstrated a cytotoxic phenotype (subset 3) in response to cross-recognition of human leukocyte antigen-B*4402 (C).
FIGURE 3.
FIGURE 3.:
(Continued).

The Alloresponse Is Driven by the Presence of Cross-Reactive FLR-Specific CD8+ T Cells

Considering that up to 60% of the T-cell pool is composed of memory CTL (18), we examined whether the cross-reactivity of FLR-specific CD8+ T cells toward B*4402 contributes to or dominates the alloreactive hierarchy. Initial investigations were conducted with PBMC from healthy controls (n=5, H1-H5) who were in vitro stimulated with allogeneic PBMC (n=4, H9-H12) and the cell line C1R.B*4402 for 13 days. Restimulation of the T-cell cultures with cell lines specific for mismatched HLA molecules enabled the dissection of the alloresponse (39). By using the fine specificity of MHC/peptide (HLA-B8/FLR) tetramers, we segregated tetramer+ and tetramer− T-cell populations, which were representative of the cross-reactive FLR-specific and allospecific CD8+ T-cell populations, respectively. The data strongly demonstrated in three of five healthy controls (H3-H5) that the immune reactivity (IFN-γ production) toward the B*4402 alloantigen was predominantly contributed by the cross-reactive FLR-specific CD8+ T cells (tetramer+) compared with the generation of anti-B*4402 allospecific T cells (tetramer−; Fig. 4A). Moreover, the additive effect of both cross-reactive FLR-specific CD8+ T cells and newly generated anti-B*4402 allospecific CD8+ T cells potentially skewed the alloreactive hierarchy toward B*4402 over other mismatched HLA alloantigens. Interestingly, although H1 and H2 are EBV seropositive, B*4402 reactivity was not observed in the tetramer+population.

FIGURE 4.
FIGURE 4.:
Both cross-reactive FLRGRAYGL (FLR)-specific and allospecific CD8+ T cells contribute to human leukocyte antigen (HLA)-B*4402 alloreactivity. Allogeneic stimulation of peripheral blood mononuclear cell (PBMC) from healthy donors or lung transplant recipient (LTR) were performed to assess the contribution of cross-reactive FLR-specific (tetramer+) and allospecific (tetramer−) CD8+ T-cell populations to the total alloresponse. Healthy donors (H1: HLA-A1, -A2, -B8, -B*4403; H2: HLA-A2, -B8, -B15; H3: HLA-A2, -B8, -B62; H4: HLA-A2, -B8, -B57; and H5: HLA-A2, -A3, -B8, -B35) were each stimulated with irradiated allogeneic PBMC or cell line (a=H9: HLA-A1, -A29, -B8, -B*4402; b=H10: HLA-A2, -A32, -B7, -B*4402; c=H11: HLA-A1, -A2, -B7, -B*4402; d=H12: HLA-A2, -B15, -B*4402; and e=C1R.B4402) for 13 days before a 6-hr restimulation with cell lines, representing the HLA-mismatched alloantigens. (a) C1R, C1R.A1, C1R.B8, C1R.B*4402, 721.221, 721.221.A29; (b) C1R, C1R.A2, C1R.A3, C1R.B4402, 9063 (A*3201, B*4402), 9082 (A*0301, B*0702); (c) C1R, C1R.A1, C1R.A2, C1R.A3, C1R.B*4402, 9082 (A*0301, B*0702); (d) C1R, C1R.A2, C1R.B*4402, 9031 (A*0201, B*1501); and (e) C1R.B*4402 (A). Individual LTR (Tx19: HLA-A2, -A3, -B8, -B39; Tx26: HLA-A1, -A3, -B7, -B8; Tx41: HLA-A1, -A24, -B8, -B60; and Tx42: HLA-A1, -A2, -B8, -B62) were stimulated with irradiated donor PBMC (donor HLA typings for Tx19=HLA-A1, -A2, -B8, -B*4402; Tx26= HLA-A1, -A2, -B8, -B*4402; Tx41= HLA-A3, -A24, -B7, -B*4402; and Tx42= HLA-A1, -A2, -B8, -B*4402) for 13 days before cultures before a 6 hr restimulation as described for A. Tx19, Tx26, and Tx42=C1R, C1R.A1, C1R.A2, C1R.B8, C1R.B*4402; Tx41=C1R, C1R.A3, C1R.B*4402, 9002 (A*2402, B*1402), 9082 (A3, B7) (B). Background interferon (IFN)-γ production was subtracted from the parental C1R and 721.221 cell lines.

We next evaluated the combined effect of both cross-reactive FLR-specific CD8+ T cells and newly generated anti-B*4402 CD8+ T cells to elicit immune reactivity in HLA-B8+ EBV+ LTR who had received a B*4402 allograft. PBMC (Tx19, Tx26, Tx41, Tx42, and Tx85) were in vitro stimulated with donor-derived allogeneic PBMC for 13 days. Dissection of the alloresponse was the same as that described for the healthy controls. The results demonstrated that both cross-reactive FLR-specific (tetramer+) and anti-B*4402 (tetramer−) CD8+ T cells were able to induce in vitro alloreactivity, with varying magnitudes, toward the B*4402 molecule expressed on donor PBMC. Moreover, functional analysis of T-cell subsets suggested that the cross-reactive FLR-specific CD8+ T cells were of subset 1 or subset 2 lineages, whereas the anti-B*4402 CD8+ T cells were heterogenous with all three subsets being represented (Fig. 4B).

Clinical Implications of TCR Cross-Reactivity

The investigation of our study cohort of 11 LTRs (HLA-B8+) who had received a HLA-B44 allograft occurred in the context of their clinical setting. All these LTRs were EBV seropositive except for one who was EBV seronegative but received an EBV-seropositive organ (Tx39). Further allelic subtyping of the B44 antigenic group revealed that six LTRs (Tx19, Tx26, Tx39, Tx41, Tx42, and Tx66) received a B*4402 allograft, four LTRs (Tx6, Tx7, Tx59, and Tx90) received a B*4403 allograft, whereas one LTR (Tx85) received a heterozygous B*4402, B*4403 allograft. In addition, Tx85 was also HLA-B8, B*4402, thus would not be expected to generate the cross-reactive FLR-specific CD8+ T cells (23).

There were no differences between the five EBV- seropositive HLA-B8+ LTRs (Tx19, Tx26, Tx41, Tx42, and Tx66) who received a B*4402 allograft compared with the four HLA-B8+ LTRs (Tx6, Tx7, Tx59, and Tx90) who received a B*4403 allograft in the rates of histopathologically defined acute rejection (Tx39, Tx66, and Tx90), lung function decline of more than 10% at 12 months posttransplant (Tx19, Tx42, Tx66 and Tx6, and Tx90), or long-term survival (four deaths: Tx41 because of infection at day 1119, Tx6 because of nonlymphoid malignancy at day 790, and Tx7 because of nonspecific pulmonary failure at day 189, and Tx59 because of fibrotic organizing pneumonitis at day 180).

DISCUSSION

Our phenotypically well-characterized study cohort of 11 LTRs (EBV status; HLA-B8+ recipients/HLA-B44+ allograft; clinical outcomes) enabled us to quantitate and functionally dissect cross-reactive FLR-specific CD8+ T cells within a clinical transplantation framework. We demonstrated the presence of peripheral FLR-specific CD8+ T cells in LTR under immunosuppressive conditions with frequencies who were comparable with healthy controls. Although the cross-reactive potential of these FLR-specific CD8+ T cells varied between individual LTR and over time within the first year posttransplantation, there was functional diversity between peripheral and allograft-isolated cross-reactive cells. In addition, our findings suggest that the presence of cross-reactive FLR-specific CD8+ T cells may influence the alloreactive hierarchy directed against the allograft although they were not associated with poorer short- or long-term clinical outcomes in the absence of any EBV reactivation in the HLA-B8+ LTR studied and in the setting of current immunosuppression and antiviral prophylaxis protocols.

Although the detection and expansion of peripheral FLR-specific CD8+ T cells in a subset of LTR is in agreement with previous reports in the literature (23, 24, 29, 36, 37), we have for the first time examined the cross-reactive potential of these cells within a clinical context in both the peripheral blood and allograft during the first year posttransplantation.

Having initially confirmed the cross-reactivity of FLR-specific CD8+ T cells from all the healthy controls studied toward B*4402 but not B*4403, as measured by IFN-γ production (23, 24, 36), we then demonstrated the cross-reactive potential of these cells in the peripheral blood from some but not all of our LTRs with differential dynamics in individual LTR, all of whom were managed according to routine immunosuppression and antiviral prophylaxis protocols. However, in all five LTRs studied from whom FLR-specific CD8+ T cells were isolated from allograft-derived BAL MNC, these cells were unable to recognize the B*4402 alloantigen or efficiently respond to positive controls (C1R.B8/FLR and FLR peptide alone) as measured by IFN-γ production, suggesting that T cells within the allograft were functionally distinct from peripheral CD8+ T cells.

Further exploration of functionally distinct cross-reactive FLR-specific CD8+ T-cell subsets in healthy controls yielded no differences between ex vivo- and in vitro-derived day 13 T-cell assays. However, functional profile differences were observed in different LTR and in different compartments using this experimental analysis. In particular, ex vivo analysis of cross-reactive CD8+ T cells in the periphery demonstrated a tendency to a more cytotoxic profile compared with a day 13 analysis shift toward a combined cytokine production and cytotoxicity profile. Although allograft-derived FLR-specific CD8+ T cells were strongly skewed to a cytotoxic profile with no IFN-γ production in all five LTRs studied, this was only examined in a ex vivo assay and not using a 13-day in vitro stimulation assay. Whether skewing toward a cytotoxicity marker profile represents a form of destructive alloreactivity or operational active tolerance in our LTR remains to be ascertained but is a question that will be greatly informed by an analysis of alloreactivity profiles and clinical outcomes as discussed later.

We also examined whether the cross-reactivity of FLR-specific CD8+ T cells toward B*4402 contributes to or dominates the alloreactive hierarchy. Profiles of cross-reactive (tetramer+) and noncross-reactive (tetramer−) CD8+ T cells indicated that the alloresponse in both populations was dominated by the anti-B*4402 T-cell response compared with other HLA-mismatched alloantigens. This was observed for both healthy controls and LTR.

Although our study results are unique given the clinical context in which they are framed, significant limitations remained. Despite the number of HLA-B8+ LTR receiving a HLA-B44 lung allograft in our study, these numbers remain relatively small for clinical outcome comparisons where the frequency and phenotype of cross-reactive FLR-specific CD8+ T cells varied significantly between the LTR (Figs. 1B and C and 3A-C) and where there were many significant confounders for lung function and survival outcomes in particular. Hence, at best, our results are limited to excluding an overwhelming negative effect of a HLA-B8+ LTR receiving a B*4402 allograft in the absence of any EBV reactivation in the LTR studied and in the setting of current immunosuppression and antiviral prophylaxis protocols.

In conclusion, the longitudinal measurement of FLR-specific CD8+ T cells in the peripheral blood and allograft compartments of HLA-B8+ LTR receiving a HLA-B44+ organ is the first step in dissecting the potential role of these cells in influencing clinical outcomes. Although it may seem paradoxical that these cross-reactive CD8+ T cells are not dramatically associated with worsening allograft outcomes, it is worthwhile remembering that all LTRs received mismatched donor organs because of logistic limitations, that routine immunosuppression protocols have been successful in minimizing allograft rejection syndromes, and that an active peripheral tolerance phenotype is operational in many stable LTRs. Nevertheless, we would postulate that the potential for allograft damage from EBV-specific cross-reactive T cells would be greatest in the setting of uncontrolled EBV reactivation, and as such, this model may potentially be more broadly extrapolated to any common infection that has a tendency for reactivation or reinfection in the lung allograft.

MATERIALS AND METHODS

Cohort Demographics and Ethics Approval

Eleven HLA-B8+ LTR receiving HLA-B44 single/bilateral lung allografts between January 2006 and December 2008 (six men and five women, age 21–65 years) were recruited to the study. Primary disease status consisted of cystic fibrosis (n=1), chronic obstructive pulmonary disease (n=3), emphysema (n=1), idiopathic pulmonary fibrosis (n=3), systemic lupus erythematosus (n=1), and sarcoidosis (n=2). All LTRs underwent standard triple-therapy immunosuppression and routine surveillance bronchoscopy at 0.5, 1, 2, 3, 6, 9, and 12 months posttransplant or if clinically indicated as outlined previously (40). All LTRs except one were EBV seropositive at the time of transplant, who latter received an EBV-seropositive organ and therefore being a primary mismatch for EBV. Both LTR and EBV-seropositive healthy donors (n=8) provided written consent, with ethics approval granted by The Alfred Hospital (Victoria, Australia) and the Australian Bone Marrow Donor Registry (New South Wales, Australia).

EBV DNA Testing

Plasma samples for individual LTR at all time points used for T-cell cross-reactivity investigations were sent to the Victorian Infectious Diseases Reference Laboratory (North Melbourne, Victoria, Australia) for EBV DNA testing. Briefly, the EBV polymerase chain reaction targets a conserved region of the EBNA gene. The assay was performed on ABI 7500 fast real-time instrument (Applied Biosystems, Foster City, CA), using a FAM-labeled Taqman MGB probe with ABI fast mastermix (Applied Biosystems). The assay gives semiquantitative results relating to cycle threshold and copy number in the sample. From experiments with plasmids to this region, approximate copy numbers can be determined for any positives.

Blood and BAL Samples

Peripheral blood samples from healthy donors, LTR (at the time of bronchoscopy) and deceased donors were collected in heparinized vacutainer tubes. PBMC were isolated by Ficoll-Paque (GE Healthcare, Uppsala, Sweden) density gradient centrifugation. BAL samples were obtained as described previously (41), filtered through a 70-μm cell strainer (Becton Dickinson, San Jose, CA), and centrifuged to recover all MNC present. PBMC and BAL MNC samples were cryopreserved until required. LTR samples were chosen for a hierarchy of experiments, which was purely dependent on limited sample amounts, and hence, there was no selection bias.

Cell Lines and Culture

B-lymphoblastoid cell lines (9002: A*2402, B*1402, C*0202, and 0802; 9031: A*0201, B*1501, and C*0304; 9063: A*3201, B*4402, and C*0501; 9082: A*0301, B*0702, and C*0702) and transfected cell lines (C1R.parental/A*0101/A*0201/A*0301/B*0801/B*4402/B*4403, and 721.221.parental/A*2902) were maintained as described previously (42).

HLA/Peptide Tetramers and Peptides

Tetramers specific for HLA-B8-restricted EBV EBNA3A epitopes, FLR, and QAK were generated essentially as described previously (43) and were synthesized by Genscript (Piscataway, NJ).

Autologous/Allogeneic Stimulations and Intracellular Cytokine Staining

T-cell cultures were generated by stimulating healthy donor or LTR PBMC with FLR-pulsed autologous PBMC, allogeneic PBMC, or C1R.B*4402 (stimulators were irradiated at 3000 Rad except for the latter at 10,000 Rad) for 13 days at a 2:1 ratio as described previously (39, 42).

Quantitation of EBV-specific CD8+ T cells in ex vivo PBMC (healthy donors and LTR) and BAL MNC (LTR) was performed on day 0 to measure baseline frequencies. Cells were stained with anti-CD8 PE-Cy5 (1:20, clone HIT8a, Becton Dickinson) and HLA-B8/FLR or HLA-B8/QAK PE (Phycoerythrin) tetramers for 30 min at 4°C, then fixed in 1% paraformaldehyde (ProSciTech, Queensland, Australia), and acquired by flow cytometry (FACSCalibur, Becton Dickinson).

Intracellular cytokine staining was performed as described (39) with minor modifications. Briefly, PBMC, BAL MNC, or day 13 T-cell cultures (2×105 cells) were stimulated with the appropriate cell line (105 cells) or peptide (1 μM) for 6 hr. CD107a-FITC (1:20, clone H4A3, Becton Dickinson), monensin (3.5 μg/mL, Sigma, St Louis, MO), and Brefeldin A (10 μg/mL, Sigma) were added at 0, 1, and 2 hr time points, respectively. Cells were then labeled with anti-CD8 PE-Cy5 (1:20, clone HIT8a, Becton Dickinson) and HLA-B8/FLR PE tetramer, fixed in 1% paraformaldehyde, and permeabilized with 0.3% saponin (Sigma) containing anti-IFN-γ APC (1:1000, clone B27, Becton Dickinson) and acquired on a FACSCalibur. All flow cytometry data were analyzed using FlowJo software (Tree Star Inc., Ashland, OR).

ACKNOWLEDGMENTS

The authors thank the generous support of Prof. Greg Snell and all the clinicians, nurses, and allied health professionals associated with the lung transplant team at The Alfred Hospital and all the patients involved with this study. Tetramers were kindly provided by Dr. L. Sullivan (The University of Melbourne, Victoria, Australia).

REFERENCES

1. Estenne M, Hertz MI. Bronchiolitis obliterans after human lung transplantation. Am J Respir Crit Care Med 2002; 166: 440.
2. Estenne M, Maurer JR, Boehler A, et al. Bronchiolitis obliterans syndrome 2001: An update of the diagnostic criteria. J Heart Lung Transplant 2002; 21: 297.
3. Boehler A, Estenne M. Post-transplant bronchiolitis obliterans. Eur Respir J 2003; 22: 1007.
4. Ison MG. Influenza, including the novel H1N1, in organ transplant patients. Curr Opin Infect Dis 2010; 23: 365.
5. Liu M, Mallory GB, Schecter MG, et al. Long-term impact of respiratory viral infection after pediatric lung transplantation. Pediatr Transplant 2010; 14: 431.
6. Burrows SR, Khanna R, Silins SL, et al. The influence of antiviral T-cell responses on the alloreactive repertoire. Immunol Today 1999; 20: 203.
7. Welsh RM, Selin LK. No one is naive: The significance of heterologous T-cell immunity. Nat Rev Immunol 2002; 2: 417.
8. Adams AB, Pearson TC, Larsen CP. Heterologous immunity: An overlooked barrier to tolerance. Immunol Rev 2003; 196: 147.
9. Adams AB, Williams MA, Jones TR, et al. Heterologous immunity provides a potent barrier to transplantation tolerance. J Clin Invest 2003; 111: 1887.
10. Mariuzza RA, Poljak RJ. The basics of binding: Mechanisms of antigen recognition and mimicry by antibodies. Curr Opin Immunol 1993; 5: 50.
11. Oldstone MB. Molecular mimicry and autoimmune disease. Cell 1987; 50: 819.
12. Merkenschlager M, Terry L, Edwards R, et al. Limiting dilution analysis of proliferative responses in human lymphocyte populations defined by the monoclonal antibody UCHL1: Implications for differential CD45 expression in T cell memory formation. Eur J Immunol 1988; 18: 1653.
13. Lombardi G, Sidhu S, Daly M, et al. Are primary alloresponses truly primary? Int Immunol 1990; 2: 9.
14. Macedo C, Orkis EA, Popescu I, et al. Contribution of naive and memory T-cell populations to the human alloimmune response. Am J Transplant 2009; 9: 2057.
15. Lindahl KF, Wilson DB. Histocompatibility antigen-activated cytotoxic T lymphocytes. II. Estimates of the frequency and specificity of precursors. J Exp Med 1977; 145: 508.
16. Suchin EJ, Langmuir PB, Palmer E, et al. Quantifying the frequency of alloreactive T cells in vivo: New answers to an old question. J Immunol 2001; 166: 973.
17. Baker RJ, Hernandez-Fuentes MP, Brookes PA, et al. The role of the allograft in the induction of donor-specific T cell hyporesponsiveness. Transplantation 2001; 72: 480.
18. Ashwell JD, Chen C, Schwartz RH. High frequency and nonrandom distribution of alloreactivity in T cell clones selected for recognition of foreign antigen in association with self class II molecules. J Immunol 1986; 136: 389.
19. Hislop AD, Annels NE, Gudgeon NH, et al. Epitope-specific evolution of human CD8(+) T cell responses from primary to persistent phases of Epstein-Barr virus infection. J Exp Med 2002; 195: 893.
20. Levitsky V, de Campos-Lima PO, Frisan T, et al. The clonal composition of a peptide-specific oligoclonal CTL repertoire selected in response to persistent EBV infection is stable over time. J Immunol 1998; 161: 594.
21. Maecker HT, Maino VC. Analyzing T-cell responses to cytomegalovirus by cytokine flow cytometry. Hum Immunol 2004; 65: 493.
22. Benninger-Doring G, Pepperl S, Deml L, et al. Frequency of CD8(+) T lymphocytes specific for lytic and latent antigens of Epstein-Barr virus in healthy virus carriers. Virology 1999; 264: 289.
23. Burrows SR, Khanna R, Burrows JM, et al. An alloresponse in humans is dominated by cytotoxic T lymphocytes (CTL) cross-reactive with a single Epstein-Barr virus CTL epitope: Implications for graft-versus-host disease. J Exp Med 1994; 179: 1155.
24. Burrows SR, Silins SL, Cross SM, et al. Human leukocyte antigen phenotype imposes complex constraints on the antigen-specific cytotoxic T lymphocyte repertoire. Eur J Immunol 1997; 27: 178.
25. Elkington R, Walker S, Crough T, et al. Ex vivo profiling of CD8+-T-cell responses to human cytomegalovirus reveals broad and multispecific reactivities in healthy virus carriers. J Virol 2003; 77: 5226.
26. Koelle DM, Chen HB, McClurkan CM, et al. Herpes simplex virus type 2-specific CD8 cytotoxic T lymphocyte cross-reactivity against prevalent HLA class I alleles. Blood 2002; 99: 3844.
27. Gamadia LE, Remmerswaal EB, Surachno S, et al. Cross-reactivity of cytomegalovirus-specific CD8+ T cells to allo-major histocompatibility complex class I molecules. Transplantation 2004; 77: 1879.
28. Rist M, Smith C, Bell MJ, et al. Cross-recognition of HLA DR4 alloantigen by virus-specific CD8+ T cells: A new paradigm for self-/nonself-recognition. Blood 2009; 114: 2244.
29. Amir AL, D'Orsogna LJ, Roelen DL, et al. Allo-HLA reactivity of virus-specific memory T cells is common. Blood 2010; 115: 3146.
30. Cainelli F, Vento S. Infections and solid organ transplant rejection: A cause-and-effect relationship? Lancet Infect Dis 2002; 2: 539.
31. Westall GP, Brooks AG, Kotsimbos T. CD8+ T-cell maturation following lung transplantation: The differential impact of CMV and acute rejection. Transpl Immunol 2007; 18: 186.
32. Westall GP, Michaelides A, Williams TJ, et al. Bronchiolitis obliterans syndrome and early human cytomegalovirus DNAaemia dynamics after lung transplantation. Transplantation 2003; 75: 2064.
33. Argaet VP, Schmidt CW, Burrows SR, et al. Dominant selection of an invariant T cell antigen receptor in response to persistent infection by Epstein-Barr virus. J Exp Med 1994; 180: 2335.
34. Turner SJ, Doherty PC, McCluskey J, et al. Structural determinants of T-cell receptor bias in immunity. Nat Rev Immunol 2006; 6: 883.
35. Wynn KK, Fulton Z, Cooper L, et al. Impact of clonal competition for peptide-MHC complexes on the CD8+ T-cell repertoire selection in a persistent viral infection. Blood 2008; 111: 4283.
36. Burrows SR, Silins SL, Moss DJ, et al. T cell receptor repertoire for a viral epitope in humans is diversified by tolerance to a background major histocompatibility complex antigen. J Exp Med 1995; 182: 1703.
37. D'Orsogna LJ, Amir AL, Zoet YM, et al. New tools to monitor the impact of viral infection on the alloreactive T-cell repertoire. Tissue Antigens 2009; 74: 290.
38. Macdonald WA, Chen Z, Gras S, et al. T cell allorecognition via molecular mimicry. Immunity 2009; 31: 897.
39. Mifsud NA, Purcell AW, Chen W, et al. Immunodominance hierarchies and gender bias in direct T(CD8)-cell alloreactivity. Am J Transplant 2008; 8: 121.
40. Westall G, Kotsimbos T, Brooks A. CMV-specific CD8 T-cell dynamics in the blood and the lung allograft reflect viral reactivation following lung transplantation. Am J Transplant 2006; 6: 577.
41. Kotsimbos AT, Sinickas V, Glare EM, et al. Quantitative detection of human cytomegalovirus DNA in lung transplant recipients. Am J Respir Crit Care Med 1997; 156: 1241.
42. Macdonald WA, Purcell AW, Mifsud NA, et al. A naturally selected dimorphism within the HLA-B44 supertype alters class I structure, peptide repertoire, and T cell recognition. J Exp Med 2003; 198: 679.
43. Nguyen TH, Sullivan LC, Kotsimbos TC, et al. Cross-presentation of HCMV chimeric protein enables generation and measurement of polyclonal T cells. Immunol Cell Biol 2010; 88: 676.
Keywords:

Cross-reactivity; T cells; Lung transplant recipients

© 2010 Lippincott Williams & Wilkins, Inc.