Indirect Minor Histocompatibility Antigen Presentation by Allograft Recipient Cells in the Draining Lymph Node Leads to the Activation and Clonal Expansion of CD4⁺ T Cells That Cause Obliterative Airways Disease¹

Recent advances in cadaveric organ preservation techniques and immunosuppressive combination chemotherapy have increased 1-year allograft survival rates to >75%. Consequently, organ transplantation has become a viable option for many end-stage diseases (1, 2). Nevertheless, chronic allograft rejection remains a significant barrier to long term allograft acceptance in most transplant recipients (3, 4). In the case of lung transplantation, late graft failure secondary to small airway fibrosis (obliterative bronchiolitis) is responsible for reductions in 5-year graft survivals to <50% (5, 6, 7).

It has been well documented that the adaptive immune system plays a role in the development of chronic rejection. Graft-recipient CD4⁺ T cells have the potential to respond against foreign graft tissue Ags through two distinct mechanisms. A high frequency of T cells (1–10% of the repertoire) appears to directly recognize foreign class II MHC-encoded molecules expressed on the surface of donor cells (8, 9, 10). Less frequently (1 in 10⁴–10⁵), recipient CD4⁺ T cells recognize a peptide derived from a donor polymorphic protein (minor histocompatibility Ag (mHAg)³) bound to a self-class II MHC molecule (10, 11). Historically, it has been thought that the direct recognition of foreign MHC molecules plays the dominant role in acute graft rejection. However, during the last decade the role of indirectly activated mHAg-reactive CD4⁺ T cells in the development of chronic rejection has been appreciated (12, 13). Accordingly, class I-deficient graft recipients have been found to reject both class I- and class II-deficient skin allografts (14). Consistent with this, ELISPOT and limiting dilution analyses have determined that CD4⁺ T cells recognizing donor-derived mHAg can increase in frequency to as much as 1–5% of a graft recipient’s T cells in the spleen and peripheral blood after allotransplantation (11, 15, 16). This increased frequency also appears to be functionally important, as a direct correlation has been observed between the development of chronic lung graft dysfunction and the presence of T cells reactive to donor peptides within the recipient’s spleen (17).

Donor-derived mHAg are thought to drain into the graft recipient’s lymph nodes soon after transplantation due to ischemic cell death and alloantigen-targeted cytolysis. Host APC may even engulf whole cells derived from transplanted tissue. Nevertheless, it has been difficult to identify the relevant APC for graft mHAg or the CD4⁺ T cells in the lymph nodes draining the transplant site that are reactive to the mHAg. Such a characterization of the mHAg-reactive T cells (among the large population of uninvolved bystander T cells) would be an important first step toward determining which T cells are responsible for the chronic destruction of the allograft as well as those that have regulatory function and are protecting the graft from further immunological damage. In addition, a careful analysis of the APC that are relevant to mHAg presentation may determine how sharing of class II MHC alleles between the donor and recipient can promote graft acceptance. Therefore, we took advantage of chicken OVA-specific TCR-transgenic (TCR-Tg) T cells as well as Tg trachea donor mice that uniformly express a membrane form of OVA (B6-Act-mOVA-II Tg mice, also called B6-OVA in this report) in a heterotopic model of airway transplantation to track mHAg-reactive CD4⁺ T cells during the course of chronic airway allograft rejection. Our results now directly demonstrate the response of the graft-reactive CD4⁺ T cells after allotransplantation and during the development of obliterative airways disease (OAD).

Wild-type C57BL/6NCr (B6) (H-2^b), BALB/cAnNCr (BALB/c) (H-2^d), and BALB/cAnNCr-nu (BALB/c nu/nu) mice, 6–8 wk old, were purchased from Charles River Breeding Laboratories (Wilmington, MA) through a contract with the National Cancer Institute animal program of the National Institutes of Health (Frederick, MD). C57BL/6TacAbb^tm1 (class II-deficient) were purchased from Taconic Farms (Germantown, NY). C57BL/6-Cd4^tm1Mak (CD4-deficient, 2663), C57BL/6J-recombinase-activating gene 1 (Rag1)^tm1Mom (Rag1-deficient, 2216), B6.PL-Thy1^a/Cy (Thy1.1, 0406), and B6.C-H2^bm1/ByJ (bm1, 1060) mice were purchased from The Jackson Laboratory (Bar Harbor, ME).

The OT-II mice were backcrossed to Rag1^−/− mutant and B6.PL (Thy1.1 congenic) mice. OT-II TCR-Tg CD4⁺ T cells are reactive to OVA peptide 323–339 (OVAp) bound to I-A^b (18, 19). DO11.10 TCR-Tg mice were bred to homozygosity and maintained in our animal facility as previously described (20). DO11.10 T cells are CD4⁺, are reactive to chicken OVAp in the context of I-A^d, and express a clonotypic TCR detectable with the KJ1-26 mAb (21). B6-Act-mOVA-II (B6-OVA) mice constitutively express OVA and have been previously described (22). The B6-OVA class II^−/− and (BALB/c × B6-OVA)F₁ mice were also bred in our colony. Experimental mice were age-matched (6–8 wk old) and sex-matched except where indicated. Animal care was provided according to National Institutes of Health and University of Minnesota Institutional animal care and use committee guidelines, and animals were housed in specific pathogen-free conditions with free access to food and water.

The following Abs and fluorochrome conjugates were purchased from BD PharMingen (San Diego, CA): FITC-, PE-, PerCP-cyanin 5.5-, and allophycocyanin-labeled or biotinylated anti-CD4 (RM4-5); FITC-, PE-, PerCP-cyanin 5.5-, allophycocyanin-labeled or biotinylated anti-CD8α (53-6.7); FITC- or PE-labeled anti-CD69 (H1.2F3); PE-labeled anti-CD44 (IM7); PE-labeled anti-CD45RB (16A); and purified anti-CD16/CD32 (2.4G2). The following Abs and fluorochrome conjugates were purchased from eBioscience (San Diego, CA): allophycocyanin-labeled anti-CD90.1 (Thy1.1; HIS51) and PerCP- or allophycocyanin-labeled anti-CD90.2 (Thy1.2; 53–2.1). Allophycocyanin-labeled anti-DO11.10 TCR (KJ1-26) was purchased from Caltag Laboratories (Burlingame, CA).

T cells (2.5–5 × 10⁶) were prepared for adoptive transfer as described previously (23) and were injected into the tail vein of recipient mice in a volume of 0.3 ml 1 day before transplantation. Briefly, lymph node and spleen were harvested from mice and placed in cold medium consisting of a 1/1 mixture of Eagle’s Hanks’ amino acids medium (Biofluids, Rockville, MD) and RPMI 1640 (Life Technologies, Grand Island, NY) containing 10% FCS (Sigma-Aldrich, St. Louis, MO), 2 mM l-glutamine, 100 U/ml penicillin, 100 U/ml streptomycin, and 5 × 10⁻⁵ M 2-ME. The tissues were mashed onto sterile nylon screens (Sefar America, Depew, NY) in medium and were analyzed by flow cytometry to determine the number of CD8⁺ T cells transferred. In most experiments cells were labeled with CFSE (Molecular Probes, Eugene, OR) before transfer, using a modification of a technique previously described (24). Briefly, lymph node and spleen cells at a concentration of 1 × 10⁷ cells/ml were incubated in 2.5 μM CFSE for 5 min at 37°C. The labeling reaction was stopped with the addition of ice-cold medium. The CFSE-labeled T cells were washed twice in PBS before adoptive transfer.

This heterotopic tracheal transplantation system was originally developed by Hertz et al. (25) as a mouse model of chronic lung transplant rejection (15, 25, 26, 27, 28). Animals were transplanted as previously described (25, 27, 28). Briefly, donor mice were euthanized and skin-prepped with 70% ethanol (Sigma-Aldrich). The neck and upper chest were incised and bluntly dissected. The trachea was resected from the inferior border of the larynx to the carina and immediately placed in ice-cold PBS with penicillin G sodium (100 U/ml), streptomycin sulfate (100 μg/ml), and amphotericin B (0.25 μg/ml; Life Technologies, Grand Island, NY). Recipient mice were anesthetized with ketamine/xylazine (100 and 2 mg/kg, respectively, i.p.; Phoenix Pharmaceuticals, St. Joseph, MO), and the skin of the back was shaved and prepped with iodine solution from the base of the neck to the tail. A 1-cm horizontal incision was made, and s.c. pockets were formed by blunt dissection. Two or three trachea grafts were placed heterotopically into the pockets without primary vascularization, and the wound was closed with Nextband surgical glue (Veterinary Products Laboratories, Phoenix, AZ). Recipient animals received sulfamethoxazole and trimethoprim oral suspension (Qualitest Pharmaceuticals, Huntsville, AL) and acetaminophen (Altaire Pharmaceuticals, Aquebogue, NY) in their drinking water for 3 days, beginning on the day of transplantation. No immunosuppressive agents were given to any graft recipient.

Lymph nodes draining the site of transplantation (axillary and brachial) were harvested and mashed as described above. The cell suspensions were washed with staining buffer (PBS containing 2% FCS and 0.2% sodium azide), then stained with the combination of Abs for 30 min in the cold. Cells were electronically gated based on forward and side scatter properties, and at least 1000 events were collected using the FACSCalibur flow cytometer (BD Biosciences, Mountain View, CA) and analyzed using CellQuest (BD Biosciences) and FlowJo (Tree Star, San Carlos, CA) software. Graft tissues were harvested at various time points, minced with sterile scissors, and mashed onto sterile nylon screens. Three trachea grafts were pooled for each flow cytometric determination. Recovered cells were aliquoted for various staining combinations, and then entire samples were analyzed by flow cytometry as described above.

Grafts were harvested and snap-frozen to −60°C in embedding medium (Tissue-Tek OCT compound; Miles, Elkhart, IN) in a bath of 2-methyl butane (Sigma-Aldrich) and dry ice. Frozen sections (6 μm) were cut from tissue blocks onto coated 3-in. glass slides (Fisher Scientific, Pittsburgh, PA) at −19°C on a Leica CM1800 cryostat (Nussloch, Germany), fixed in acetone (Fisher Scientific, Pittsburgh, PA), and immediately stained with H&E (Fisher Scientific) or stored at −80°C until used for immunohistochemical staining. Slides were warmed to room temperature, and sections were hydrated with PBS for 15 min. Endogenous peroxidase activity was extinguished by incubation with 1% H₂O₂ for 30 min. Nonspecific Ab binding was blocked by incubation in anti-Fc mAb 2.4G2 for 15 min, followed by sequential 15-min incubation periods with avidin and biotin blocking reagents (Avidin/Biotin Blocking Kit; Vector Laboratories, Burlingame, CA). All slides were then incubated with 2 μg/ml biotinylated Abs for 30 min at room temperature. For conventional immunohistochemistry, sections were incubated with HRP-conjugated streptavidin-biotin complex (ABC Elite; Vector Laboratories) for 30 min, followed by a 5-min incubation with 3,3-diaminobenzidine substrate (Vector Laboratories) and diaminobenzidine enhancer (Vector Laboratories) according to the manufacturer’s instructions, mounted, and coverslipped with Permount (Fisher Scientific). For fluorescence immunohistochemistry, Abs were diluted in blocking buffer and tyramide reagents were diluted in smplification diluent (PerkinElmer, Boston, MA). After primary Ab incubation, slides were washed and sequentially incubated with HRP-labeled streptavidin for 20 min and with Cy3-labeled tyramide for 5 min at room temperature (PerkinElmer). All slides were mounted with Vectashield containing 4,6-diamidino-2-phenylindole (Vector Laboratories) to preserve fluorescence and counterstain nuclei. Slides were examined on an Axioplan 2 microscope (Zeiss, Thornwood, NY), and images were captured with a SPOT camera (Diagnostic Instruments, Sterling Heights, MI). All images were imported into Adobe Photoshop 7.0 (Adobe Systems, San Jose, CA) and prepared for presentation.

H&E-stained slides were coded and examined by two blinded reviewers (M.I.H. and D.L.M.) for the presence or the absence of fibrosis consistent with the development of OAD as previously described (27). The p values were determined using the nonparametric χ² test of statistical significance. Confidence intervals cited in the text reflect the SEM.

Our previous study of OAD after airway allotransplantation showed that (bm1 × bm12)F₁ tracheas bearing a single pair of class I (K^bm1) and class II (I-A^bm12) MHC alloantigens developed fibrosis only about half the time in B6 mice, whereas fully mismatched BALB/c tracheal grafts were nearly always fibrosed (27). This difference in the intensity of chronic rejection may simply have reflected the limited number of alloantigenic MHC targets present in the (bm1 × bm12)F₁ grafts. Alternatively, the weak response may have resulted from an absence of mHAg. To begin to assess the role of mHAg in the induction of OAD after tracheal allotransplantation, we examined male B6 tracheas expressing Y chromosome-encoded mHAg after transplantation into female B6 mice. Unlike sex-matched B6 isografts, male grafts transplanted into female B6 recipients typically developed an extremely intense and organized mononuclear cell infiltrate (Fig. 1, A and B). This infiltrate was comprised of CD4⁺ and CD8⁺ T cells as well as B cells and CD11c⁺ dendritic cells (DC) based on immunohistochemistry (data not shown). Most of the DC as well as the nearby luminal airway epithelial cells expressed high amounts of class II MHC (data not shown). Similar to the destructive inflammatory infiltrates previously observed in fully mismatched tracheal grafts (27), infiltrates in mHAg-disparate grafts tended to focus in areas adjacent to the tracheal cartilage and in many cases within the subepithelial space. In contrast to these male allografts, T cell and DC infiltration as well as high level class II MHC expression were only rarely seen in isografts (data not shown).

Flow cytometry was used to further characterize the T cells that infiltrated these tracheal allografts. Although large numbers of T cells could be recovered from mHAg-only disparate male grafts after transplantation into B6 female hosts, similar to BALB/c fully mismatched tracheas, there was a difference in the relative contributions of the CD4⁺ and CD8⁺ cells to the response. As previously reported (27, 28), CD8⁺ T cells were the dominant T cell observed to infiltrate BALB/c allografts (Fig. 2 A). In contrast, male B6 grafts typically accumulated more CD4⁺ than CD8⁺ T cells after transplantation into B6 female mice. Taken together, our results indicated that class II MHC-bearing DC accumulate in mHAg-disparate (but MHC-matched) allografts adjacent to a large number of CD4⁺ responder T cells. This finding raised the possibility that alloreactive mHAg-specific CD4⁺ T cells play a major role in the development of OAD.

Despite the large number of CD4⁺ T cells observed in the majority of these mHAg-disparate tracheal grafts, intense CD4⁺ T cell infiltration did not in itself predict the development of fibrosis secondary to OAD. Luminal fibrosis occurred in only 26% of male grafts transplanted into female recipients, although this was significantly more frequent than isograft fibrosis (9% fibrosed; p = 0.05; Fig. 1D). Accordingly, many of the most heavily infiltrated grafts demonstrated relatively normal-appearing airway epithelium and remained free of luminal fibrosis throughout the 56-day experiment (Fig. 1 B). Thus, mHAg was shown to be an important stimulus for graft infiltration. However, in the absence of MHC disparities, mHAg appeared to be insufficient to promote uniform and rapid fibrosis of the airway grafts.

In previous experiments direct class I alloreactivity by CD8⁺ T cells was shown to be an important stimulus for the development of OAD (27, 28). Therefore, we hypothesized that male mHAg-induced OAD often failed to occur because of the lack of a direct class I MHC target for CD8⁺ T cells within the tracheal allografts. Transplantation of male bm1 tracheas was performed in female B6 mice, and this resulted in near uniform destruction by OAD within 56 days (90% of the allografts; Fig. 1 D). This was a statistically significant increase in luminal fibrosis over that observed with the male mHAg alone (p < 0.01). Thus, the combination of a single class I MHC mismatch together with the mHAg differences associated with the Y chromosome proved to be a potent barrier to allotransplant acceptance because of the development of OAD.

This capacity of the K^bm1 mutant class I molecule to promote allograft luminal fibrosis in the presence of other mHAg differences supported a role for directly alloreactive CD8⁺ T cells in the development of OAD. Perhaps consistent with this, CD8-deficient B6 female mice transplanted with male bm1 tracheas can be shown to demonstrate a reduced potential to develop allograft luminal fibrosis (28). Nevertheless, it was formally possible that for the case of the chronic airway rejection seen in these experiments, CD8⁺ T cells were also the dominant responders to the mHAg, given that peptides derived from male mHAg may act as antigenic ligands for either CD8⁺ or CD4⁺ T cells and despite the high frequency of CD4⁺ T cells infiltrating these allografts (29, 30). To further investigate the phenotype of the male mHAg-reactive T cells that participated in the development of the OAD observed in this study, we examined chronic tracheal allograft rejection in CD4-deficient B6 recipient mice. Male tracheal grafts did appear to be better tolerated in CD4-deficient female mice (8% fibrotic) compared with wild-type recipients (26% fibrotic); however, this difference did not reach statistical significance (p = 0.1; Fig. 1 D). Interestingly, male bm1 grafts also demonstrated enhanced fibrosis-free survival in CD4-deficient B6 female recipients (40% fibrotic) compared with wild-type recipients (90% fibrotic), and this difference proved to be statistically significant (p = 0.03). Thus, CD4⁺ T cells can be shown to respond to mHAg differences, accumulate in the airway allograft tissue, and contribute to the induction of OAD.

To further investigate the role of CD4⁺ T cells in the development of OAD, we examined by flow cytometry the phenotype of lymphocytes isolated from tracheal grafts as well as from the lymph nodes draining the site of transplantation (Fig. 2). The dominant population of CD4⁺ T cells recovered from either mHAg-disparate male B6 tracheas or from fully mismatched BALB/c allografts in female B6 recipient mice was enriched for CD45RB^low and CD44^high expression (Fig. 2, B and C). This phenotype is consistent with an identification of effector/memory cells in these allografts (31, 32). In contrast, CD4⁺ T cells residing in the draining lymph node (DLN) of allografted animals or within isografts were predominantly naive (CD45RB^high and CD44^low; Fig. 2, B and C, and data not shown). This suggested a previous Ag recognition event by many of the male or BALB/c allograft-infiltrating CD4⁺ T cells and preferential homing by these effector/memory cells to the transplanted tracheal tissue. Interestingly, about half the CD44^high CD4⁺ T cells found in the fully mismatched BALB/c allografts coexpressed CD69 (Fig. 2 C), a marker of very recent Ag stimulation (33). This, however, was not seen in the graft tissue of female B6 mice receiving mHAg-disparate male allografts. Thus, an analysis of activation marker expression suggested that fully mismatched grafts were a more potent stimulus for the activation of graft-infiltrating effector/memory CD4⁺ T cells than either mHAg-disparate or isogeneic tracheal grafts.

Our failure to find evidence for recent or ongoing Ag stimulation of the majority of endogenous CD4⁺ T cells recovered from mHAg-disparate tracheal allografts (or from the lymph nodes draining these grafts) hampered our ability to further study the phenotype of the mHAg-reactive CD4⁺ T cells because we were unable to determine which of the CD4⁺ T cells was specific for an mHAg. To address this problem, we took advantage of several Tg mouse lines to develop a more robust model of mHAg transplantation that would allow the identification of mHAg-specific responder CD4⁺ T cells. Specifically, we used B6-Act-mOVA-II (B6-OVA) Tg mice that express a transmembrane form of chicken OVA in the majority of their tissues as a source of tracheal tissue for transplantation (22). OVA-Tg mice produced on the B6 background are isogeneic to B6 with respect to MHC and all other mHAg. Tg expression of OVA by this mouse strain thus allowed us the opportunity to model the immune system’s response to a unique protein mHAg present only in the transplanted tissue.

Similar to male tracheas, after transplantation into female B6 mice the B6-OVA tracheas became highly infiltrated with mononuclear cells, yet failed to develop OAD by day 28 after transplantation into B6 recipients (Fig. 1,D and data not shown). This inefficient rejection of mHAg-disparate airway grafts despite evidence of intense inflammation suggested either that the majority of infiltrating CD4⁺ T cells were not OVA-specific or, alternatively, that counter-regulatory mechanisms existed that inhibited the function of the graft-reactive T cells. To distinguish these two possibilities, OVA-specific TCR-Tg CD4⁺ T cells on the B6 or BALB/c background (OT-II or DO11.10, respectively) were used as responder T cells in grafting experiments. OVA-reactive OT-II CD4⁺ T cells were first labeled with CFSE, then adoptively transferred into syngeneic B6 recipients 1 day before heterotopic transplantation of B6-OVA mouse tracheas. B6 recipients adoptively transferred with OT-II cells before transplantation of B6-OVA tracheas developed OAD in 75% of the grafts by day 28 (Fig. 1, C and D). This result was in contrast to B6 recipients of both OT-II T cells and B6 isografts, which developed no evidence of either OT-II infiltration of the isograft or OAD (Fig. 1 D). Thus, it appeared that OAD would more frequently develop in response to a single mHAg difference within the airway graft tissue when a higher frequency of mHAg-reactive CD4⁺ responder T cells was present.

Beginning around day 7 after B6-OVA tracheal transplantation, a sizable fraction of OVA-specific OT-II CD4⁺ T cells within the DLN appeared to have diluted its CFSE dye and increased in size, based on the forward scatter profile (Fig. 3,A). Similar dye dilution and size enlargement were never observed in the DLN after transplantation of B6 tracheal isografts. In fact, when animals transplanted with B6-OVA grafts for a period of 7 days were compared with the B6 isografted animals, it was apparent that DLN OT-II CD4⁺ T cells had increased in number in response to the presence of the OVA transgene (Fig. 3,B). Furthermore, the OT-II CD4⁺ T cells undergoing cell division in the DLN of B6-OVA transplanted recipients uniformly expressed high levels of CD44, and many were CD69⁺ (Fig. 4, A and C). In contrast, the majority of OT-II CD4⁺ T cells in isografted recipients had not undergone any cell division and did not express high levels of either CD44 or CD69. Even those few OT-II cells that had undergone several rounds of Ag-nonspecific division in isografted mice were not CD69⁺, thus correlating ongoing Ag recognition with CD69 expression. It should be noted that this OT-II response differed from the endogenous DLN CD4⁺ T cells, which showed no detectable change in response to B6-OVA tracheal graft transplantation, presumably because of a relatively low frequency of OVA-specific T cells (Fig. 4, B and D).

By day 21, most of the responding OT-II cells seemed to have disappeared from the DLN; however, there continued to be significant enrichment of OT-II cells relative to endogenous CD4⁺ T cells (∼1.5%) that infiltrated the OVA-expressing allograft tissue (Fig. 3,B). These graft-infiltrating OVAp-specific OT-II CD4⁺ T cells had all undergone at least seven rounds of cell division and were uniformly high for CD44 expression (Fig. 4,A). About half these cells also coexpressed CD69, consistent with ongoing OVAp/self-class II MHC recognition and TCR stimulation of at least a portion of the mHAg-specific graft-infiltrating T cells (Fig. 4,C). In contrast, virtually all the endogenous CD4⁺ T cells found within the B6-OVA allografts were negative for CD69, even though most expressed a high level of CD44 (Fig. 4, B and D). Taken together, the results suggested that when MHC is matched between an airway allograft and the recipient, the graft-infiltrating endogenous CD4⁺ T cells are only infrequently specific for the mHAg differences. Furthermore, there was little evidence for counter-regulation of bona fide mHAg-specific CD4⁺ T cells that do enter these allografts.

The B6-OVA/OT-II experiments described above were designed to be analogous to the situation of the male trachea graft transplanted into a female recipient, with the matching of MHC alleles between the graft donor and recipient. Given the strong OT-II CD4⁺ T cell proliferative response observed within the DLN after allotransplantation of B6-OVA tracheas into B6 mice, we postulated that both the donor-derived APC carried into the host with the graft tissue as well as the recipient-derived APC would have had access to OVA, and that either group of cells could have been responsible for the initiation of an mHAg-specific CD4⁺ T cell response. To test the role of host APC in the initiation of chronic rejection, OVA-specific TCR-Tg DO11.10 CD4⁺ T cells were examined after adoptive transfer into BALB/c mice that also received fully MHC-mismatched B6-OVA tracheal grafts. This strain combination mimics the more relevant human cadaveric transplant situation where self-class II-restricted mHAg-specific CD4⁺ T cells would participate in the induction of OAD in the midst of a fully disparate allogeneic response (2, 34).

Similar to the results obtained in B6 mice with the OT-II CD4⁺ T cells and the MHC-identical, OVA-expressing grafts, >50% of the DO11.10 CD4⁺ T cells found in the DLN of BALB/c mice underwent blastogenesis and multiple rounds of cell division in response to transplantation of fully allogeneic B6-OVA tracheas (Fig. 5). These OVAp-specific CD4⁺ T cells also underwent modest clonal expansion in response to B6-OVA tracheal transplantation that was most obvious between days 3 and 14 after transplantation. Cell cycle progression was accompanied by progressively reduced expression of CD45RB, consistent with T cell differentiation into an effector/memory cell phenotype in response to recognition of OVA mHAg in the DLN (Figs. 5,E and 6,C). Essentially similar results were observed when (BALB/c × B6-OVA)F₁ donor tissue was transplanted into BALB/c recipients, demonstrating no enhancement of mHAg presentation when class II MHC molecules were matched between the donor and the recipient (Fig. 6,D). Furthermore, MHC class II^−/− B6-OVA tracheas were fully capable of eliciting activation and differentiation of the OVAp-reactive DO11.10 CD4⁺ T cells after transplantation (Fig. 6,E). In contrast, activation of DO11.10 T cells was not observed after transplantation of either BALB/c isografts or B6 allografts lacking the OVA transgene (Fig. 6, A and B). Thus, fully allogeneic airway tissue also appeared to be an effective stimulus for the proliferation and differentiation of recipient MHC-restricted and donor mHAg-reactive CD4⁺ T cells in lymph nodes draining the site of transplantation.

In studies reported here and elsewhere (27, 28, 35), neither a single donor allogeneic class I MHC molecule (K^bm1), nor a single source of mHAg (male Ag or Tg OVA) was sufficient on its own to stimulate chronic allograft rejection by endogenous polyclonal responder T cells that led to the consistent development of OAD in this heterotopic transplantation model. Nevertheless, the combination of this allogeneic K^bm1 molecule and male mHAg was sufficient to elicit OAD at a high frequency. Furthermore, experiments using graft-recipient mice genetically deficient for CD4⁺ demonstrated a requirement for CD4⁺ T cells to achieve the highest rate of destruction of mHAg-containing (e.g., male bm1 or BALB/c) tracheal allografts. These findings provide support for the commonly held belief that cooperativity exists in the immune system between directly alloreactive CD8⁺ T cells and mHAg-specific CD4⁺ T cells.

It remains to be determined whether the presence of allogeneic class I target molecules within mHAg-disparate allografts promotes OAD by enhancing the number or the responsiveness of indirectly reactive mHAg-specific CD4⁺ T cells, or whether the normal mHAg-specific CD4⁺ T cell repertoire is sufficient for OAD development only when directly alloreactive CD8⁺ T cells can also be recruited into the response. An intense CD8⁺ T cell-mediated cytotoxic response targeted against direct class I alloantigens may lead to greater graft injury and increased tissue mHAg release, with consequent enhancement of mHAg-reactive CD4⁺ T cell clonal expansion and differentiation within both the DLN and the graft. Such a model would require that expanded mHAg-reactive CD4⁺ T cells within an allograft have the capacity to directly elicit OAD. The increased development of OAD in MHC-matched B6-OVA tracheal grafts transplanted into B6 recipients that had received adoptive transfer of OT-II CD4⁺ T cells that recognize an OVA peptide Ag could support this possibility of direct pathogenicity by mHAg-specific CD4⁺ T cells. The elaboration of lymphokines such as TNF-α and IFN-γ is one potential mechanism by which CD4⁺ T cells may directly promote OAD development. Production of TNF-α and IFN-γ has, in fact, been observed in bronchoalveolar lavage fluid from patients undergoing lung rejection after allotransplantation and within the heterotopic tracheal allografts themselves (36, 37, 38, 39, 40). Such CD4⁺ T cell-derived lymphokines might activate and/or recruit those cells that are ultimately responsible for the production of the fibrosis-inducing cytokine platelet-derived growth factor, a fibroblast growth factor implicated in the pathogenesis of OAD (6, 41). Immunohistochemistry studies have shown that both macrophages and alveolar epithelial cells can be important sources of platelet-derived growth factor within pulmonary tissue (6, 42).

Alternatively, mHAg-specific CD4⁺ T cells in DLN may be responsible for producing molecules that assist in the activation, clonal expansion, and differentiation of the directly alloreactive CD8⁺ T cells, in particular the CTL. CD4⁺ T cells may activate or condition APC in the DLN through ligation of CD40 (43). APC activation can increase the expression of costimulatory molecules such as ICAM-1 and B7 on the surface of APC, and this process may be essential for the full activation of all Ag-specific T cells in the DLN (43, 44, 45). In addition, activated APC can also produce IL-12, an essential third signal necessary for the development of cytotoxic CD8⁺ T cells (46, 47, 48). CD4⁺ T cell production of TNF-α and IFN-γ within allografts could also be expected to induce the expression of allogeneic class I molecules within the graft and promote the direct recognition of allo-MHC on the parenchyma by CTL (49). Finally, mHAg-reactive CD4⁺ T cells may help overcome the development of activation-induced nonresponsiveness within the alloreactive CD8⁺ T cell population. CD8⁺ T cells naturally lose the capacity to make IL-2 after a few days of stimulation and become dependant upon IL-2 production by nearby CD4⁺ T cells for continued responsiveness (50, 51, 52). Activated mHAg-specific CD4⁺ T cells within the DLN and/or graft tissue may provide the IL-2 that allows for continued CTL development and responsiveness. Such a role for CD4⁺ T cell help in the avoidance of activation-induced nonresponsiveness has been demonstrated in other model systems, including CTL responses directed against viral pathogens as well as antitumor responses (53, 54, 55).

One of the other major differences observed between mHAg-disparate and fully mismatched tracheal grafts was the great preponderance of CD4⁺ T cells that was observed to accumulate within the mHAg-disparate, but MHC-matched allografts. Unlike fully mismatched grafts where CD8⁺ T cells are the dominant infiltrating cells, CD4⁺ T cells are the major responder lymphocyte population in grafts bearing only mHAg differences. Most of these polyclonal CD4⁺ T cells express high levels of CD44 and low levels of CD45RB within the mHAg-expressing allografts, consistent with previous Ag stimulation and the acquisition of homing receptors that allow trafficking to the peripheral tissues. However, few were found to express the CD69 activation molecule. This could indicate that the vast majority of these infiltrating CD4⁺ T cells have no inherent reactivity to graft mHAg and are simply recruited to a site of inflammation in an Ag-nonspecific fashion. Alternatively, the CD4⁺ T cells may indeed be graft-reactive, but fail to respond to the mHAg because of either defective Ag-presentation or some counter-regulatory influence (e.g., the presence of immunoregulatory cells). Regardless, an inability to establish the Ag specificity of the majority of mHAg-disparate tracheal graft-infiltrating CD4⁺ T cells prevented a more definitive analysis of this endogenous lymphocyte population.

To further investigate the nature of CD4⁺ reactivity to mHAg-disparate tracheal allografts, we took advantage of CD4⁺ TCR-Tg T cells of known Ag specificity together with Tg mice engineered to express a model mHAg (OVA) in donor tissues. In contrast to the majority of CD4⁺ T cells infiltrating mHAg-bearing allografts, OT-II CD4⁺ TCR-Tg T cells adoptively transferred into B6-OVA recipient mice were frequently CD69⁺ when recovered from the OVAp-expressing, MHC-matched tracheal allografts, and all had undergone multiple rounds of cell division. Their accumulation in allografts at only relatively modest frequency (1–2% of the infiltrating CD4⁺ T cells) nonetheless proved sufficient to provoke the development of OAD in almost all OVA⁺ tracheas. Thus, mHAg-reactive CD4⁺ T cells can be shown to participate in the chronic graft rejection of airway transplants when present initially at relatively high frequency. We conclude that the majority of endogenous polyclonal CD4⁺ T cells that infiltrate mHAg-mismatched allografts are not alloantigen-specific, based on their low expression of CD69 within the grafts and their frequent inability to induce OAD. Apparently those endogenous mHAg-reactive CD4⁺ T cells that do infiltrate mHAg-bearing, MHC-matched allografts (even in the absence of TCR-Tg T cells) are too infrequent or inefficient to induce OAD at a high rate on their own.

It is important to note that this conclusion does not rule out the possibility of additional counter-regulatory mechanisms that act to delay the induction of fibrosis by these low frequency, mHAg-specific, endogenous CD4⁺ responder T cells. Regulatory CD4⁺ T cells have been shown to inhibit the ability of APC to fully activate graft-responsive T cells and to secrete various cytokines, such as IL-10 and TGF-β, that act to suppress the immune response against allograft tissue (56). It has been observed in the trachea transplantation model that localized IL-10 production can prevent the development of airway fibrosis (57, 58). TGF-β also appears to play a role in regulatory T cell-induced graft survival based on in vivo blocking studies (59, 60). Of particular note, we have not identified by flow cytometry significant CD25 expression within the graft-infiltrating endogenous CD4⁺ T cell population (D. M. Richards, unpublished observation). Regardless, the failure to develop a high frequency effector CD4⁺ T cell clone in normal recipient mice appears to be one important factor limiting OAD development after airway transplantation using donor grafts disparate only for mHAg. The finding that an adoptive transfer of OVAp-specific CD4⁺ OT-II T cells into recipient mice can greatly enhance the development of OAD in OVA-Tg trachea grafts is consistent with this model.

Similar to the Ag-specific CD4⁺ TCR-Tg T cells observed within the transplanted allograft tissue, OT-II cells recovered from the lymph nodes draining the site of transplantation demonstrated clear evidence of alloantigen recognition. A subpopulation of these T cells responded with an increase in size, an up-regulation of CD69 and CD44 expression, and a dilution of their CFSE dye as a consequence of cell division. This same flow cytometric analysis of the polyclonal CD4⁺ T cell population within these same lymph nodes, however, demonstrated no such signs of alloantigen recognition. This again suggests that in the DLN, endogenous CD4⁺ T cells are responding to transplanted mHAg at a relatively low frequency that is apparently below the level of detection in our flow cytometric analysis. As mentioned above, other investigators have, in fact, identified endogenous mHAg-responding T cells within DLN using more sensitive, but less specific, techniques (11, 15, 17, 61, 62).

In this OT-II/B6-OVA Tg transplantation system, the recognition of graft-derived mHAg in the DLN could have occurred through the direct recognition of OVAp/class II MHC complexes on the surface of donor-derived APC, because the donor and recipient are syngeneic for MHC. Ag recognition may also have occurred in an indirect manner when donor-derived cells die, releasing OVA that can be picked up and presented by host-derived APC. After the transplantation of fully mismatched B6-OVA tracheas into BALB/c recipients, however, only host APC should have been capable of processing and presenting OVA in the context of I-A^d to the OVA-specific DO11.10 TCR-Tg CD4⁺ T cells. Using this transplantation combination (B6-OVA grafts and BALB/c recipient mice), DO11.10 CD4⁺ T cell reactivity to graft-derived OVA was easily observed within the DLN. Furthermore, the intentional inclusion of an I-A^d molecule within the graft tissue using F₁ allografts did not enhance this response. It was formally possible that DO11.10 CD4⁺ T cells may have had some low reactivity to OVAp presented in the context of I-A^b (21). Nevertheless, the transplantation of class II-deficient B6-OVA grafts onto BALB/c recipients also proved to be an effective stimulus for DO11.10 CD4⁺ T cell activation. Thus, the migration of intact allogeneic MHC-expressing donor-derived APC from the graft to the lymph node and direct presentation of mHAg is not necessary for, nor does it accelerate, the initial activation of mHAg-responsive T cells.

Taken together, these results support the hypothesis that the indirect presentation of donor-derived mHAg to recipient CD4⁺ T cells can be a major stimulus for the development of OAD after airway transplantation. Nevertheless, the data reported in this study may rely on a relatively high level of Tg protein expression or unique tissue distribution, and it is not known whether Tg OVA is representative of any natural mHAg. It should be noted that a dominant role for recipient (self) class II MHC in the priming and clonal expansion of mHAg-specific alloreactive CD4⁺ T cells after cardiac and skin transplantation in mice has also recently been reported (63, 64). In one study another Tg protein (in this case hemagglutinin) served as the model mHAg (63). In the other, TCR-Tg CD4⁺ T cell responses against the natural male mHAg H-Y were examined, with results similar to ours (64). Therefore, the use of a combination of Tg model mHAg and TCR-Tg T cells appears to provide reliable and consistent information regarding mHAg processing and presentation after solid organ transplantation.

There exists a strong association between the rate of development of OAD and the intensity of the clonal expansion and graft infiltration by activated effector/memory phenotype mHAg-specific CD4⁺ T cells. These data have important implications for efforts to reduce the activation of potentially dangerous mHAg-specific host CD4⁺ T cells by depleting grafts of passenger APC. The efficacy of such donor APC depletions is predicted to be limited to only those mHAg where the APC themselves are the major source (e.g., peptides derived from allogeneic class II MHC molecules). Accordingly, there appears to be little increased risk of priming mHAg-reactive recipient CD4⁺ T cells by intentionally matching class II MHC molecules between the donor and the recipient.

We thank E. Ingulli for the careful reading of this manuscript and her insight into the problem of chronic allograft rejection.