Abstract
Lung transplantation remains the only effective therapy for patients with end-stage lung disease, but survival is limited by the development of obliterative bronchiolitis (OB). The chemokine receptor CXCR3 and two of its ligands, CXCL9 and CXCL10, have been identified as important mediators of OB. However, the relative contribution of CXCL9 and CXCL10 to the development of OB and the mechanism of regulation of these chemokines has not been well defined. In this study, we demonstrate that CXCL9 and CXCL10 are up-regulated in unique patterns following tracheal transplantation in mice. In these experiments, CXCL9 expression peaked 7 days posttransplant, while CXCL10 expression peaked at 1 day and then again 7 days posttransplant. Expression of CXCL10 was also up-regulated in a novel murine model of lung ischemia, and in bronchoalveolar lavage fluid taken from human lungs 24 h after lung transplantation. In further analysis, we found that 3 h after transplantation CXCL10 is donor tissue derived and not dependent on IFN-γ or STAT1, while 24 h after transplantation CXCL10 is from recipient tissue and regulated by IFN-γ and STAT1. Expression of both CXCL9 and CXCL10 7 days posttransplant is regulated by IFN-γ and STAT1. Finally, we demonstrate that deletion of CXCR3 in recipients reduces airway obliteration. However, deletion of either CXCL9 or CXCL10 did not affect airway obliteration. These data show that in this murine model of obliterative bronchiolitis, these chemokines are differentially regulated following transplantation, and that deletion of either chemokine alone does not affect the development of airway obliteration.
Lung transplantation remains the only effective therapy for the large number of patients with end-stage lung disease (1, 2). However, despite advances in immunosuppressive therapies and surgical techniques, overall 5-year survival remains <50% (3). Mortality following lung transplantation is mainly due to the development of chronic graft dysfunction from obliterative bronchiolitis (OB),3 which develops in 40–50% of lung transplant recipients after 5 years (4, 5, 6, 7). The histologic hallmark of OB is an inflammatory and fibroproliferative response in the airways that can lead to significantly reduced lung function and death. Clinical studies have implicated lung injury from ischemia-reperfusion and acute rejection (AR) as two of the major causative factors for the development of OB (8, 9, 10). Central to the development of both of these processes is recruitment of inflammatory cells into the lung.
Leukocyte recruitment into tissue sites of inflammation is orchestrated by numerous chemoattractants, predominantly from the chemokine family of proteins (11, 12, 13). Studies have demonstrated that chemokines are up-regulated in injured or inflamed tissues and are specific to the type of injury and the subsequent inflammatory response. T cell infiltration of the transplanted lung is a key component of ischemia-reperfusion injury, graft rejection, and the development of OB (14, 15, 16, 17). Studies in humans and animals following transplantation have demonstrated expression of specific chemokines during these different processes (18, 19). Furthermore, in animal models, targeted inhibition of some of these chemokines can dramatically improve graft survival. In these studies, two distinct patterns of chemokine expression were seen. First, an early (within 24 h) cascade of chemokines was triggered in response to surgical trauma and ischemia-reperfusion injury. This is followed by a second cascade of T cell-active chemokines occurring during the development of allograft rejection (20, 21, 22, 23).
CXCL9, CXCL10, and CXCL11 are chemokines that attract activated lymphocytes through their receptor CXCR3. Both CXCL9 and CXCL10 are expressed in organ transplants during AR, and their expression has been shown to correlate with lymphocyte recruitment into allografts (14, 19, 24). The importance of CXCL10 was demonstrated in animal models of heart and small bowel transplantation (25, 26). In these studies, inhibition or deletion of CXCL10 delayed the onset of rejection and increased graft survival. Other studies in the murine heart transplant model have shown that inhibition of CXCL9 or CXCR3 also significantly improves graft survival (27, 28). Similar findings in a murine model of OB have also been seen using antiserum to inhibit the activity of CXCL9, CXCL10, or CXCR3 (14). In this study, inhibition of CXCL9, CXCL10, or CXCR3 resulted in profound reductions in lymphocyte recruitment into tracheal allografts and attenuation of obliteration. These data suggest that expression of both CXCL9 and CXCL10 is needed for the development of OB, and that these chemokines have unique functional roles following transplantation. The unique roles for these two chemokines could result from differences in temporal or tissue-specific expression, which ultimately may be due to differences in gene regulation. We therefore examined the regulation and function of the CXCR3 ligands using more defined reagents. In the present study, we use established murine models of OB (17) to define the expression patterns of CXCL9 and CXCL10 following transplantation and to determine the transcriptional mechanisms controlling their expression. In the heterotopic tracheal transplant model, two tracheas are harvested from donor animals and transplanted subcutaneously into the back of the recipient. In the orthotopic tracheal transplant model, a single donor trachea is anastomosed to the recipient’s trachea, as has been described (17). This orthotopic model has several advantages over the standard heterotopic model, including the natural anatomic location of the graft and improved integrity for histologic analysis. We then use these models to further explore the respective roles of CXCL9, CXCL10, and CXCR3 in the pathogenesis of airway obliteration in this model.
Materials and Methods
Mice
Wild-type BALB/c and C57BL/6 mice were purchased from Charles River Laboratories. The TLR4-deficient mice (TLR4−/−, strain 002930, BALB/c), IFN-γ-deficient mice (IFN-γ−/−, strain 002286, BALB/c and strain 003288, C57BL/6), p50-deficient mice (p50−/−, strain 002849, B6/129 hybrid), and appropriate control mice were purchased from The Jackson Laboratory. STAT1-deficient mice (STAT1−/−, strain 0020405-M, B6/129 hybrid) and appropriate control mice were purchased from Taconic Farms. MyD88-deficient mice (MyD88−/−) were provided by S. Akira (Osaka University, Osaka, Japan) and D. Golenbock (University of Massachusetts Medical Center, Worcester, MA). CXCL9-deficient mice (CXCL9−/−) were generated, as described, and backcrossed eight generations into C57BL/6 (29), and CXCR3-deficient mice (CXCR3−/−) were generated and provided by C. Gerard (Children’s Hospital, Boston, MA) (27). CXCL10-deficient mice (CXCL10−/−) were generated in our laboratory and backcrossed into C57BL/6 and BALB/c strains for nine generations (30). Mice were used at 8–12 wk of age and were age and sex matched for all experiments. All protocols were approved by the Massachusetts General Hospital Subcommittee on Research and Animal Care.
BAL analysis from patients after lung transplantation
The protocol was reviewed and approved by the Massachusetts General Hospital Human Studies Committee. Excess fluid from bronchoalveolar lavage (BAL) samples taken from patients who were donating a lower lobe for a living related lung transplant was used. The BALs were taken for analysis as part of a screening process just before excision of the lung for transplant. Eighteen to 24 h after transplantation, the donated lobes were lavaged again in the transplant recipient as part of normal surveillance and excess fluid was again taken for analysis. All BALs were done using standard techniques, as previously outlined (17). Fluid was filtered through a 40-μm filter and then centrifuged for 5 min at 300 × g
Orthotopic and heterotopic tracheal transplants
Orthotopic transplants were performed, as described (17, 31). Briefly, donor mice received a lethal ketamine injection (100 mg/kg), followed by sterile removal of the trachea. Recipient mice were anesthetized using ketamine (80 mg/kg)-xylazine (12 mg/kg). The trachea was exposed, and openings in the distal trachea and the proximal trachea were made. The donor trachea was anastomosed to the openings using 10-0 polypropylene suture. For the heterotopic transplants, two donor tracheas were implanted s.c. in the back of the recipient mouse, as described previously (17, 32). Mice were sacrificed at various time points, and the transplanted tracheas were removed for analysis.
Lung ischemia
Mice were given a lethal injection of ketamine and then exsanguinated. The chest was opened and the lungs were flushed free of blood with 10 ml of PBS via a right ventricular injection. The right lung was removed and frozen for RNA analysis. The left lung was placed in warmed PBS and placed at 37°C for 3 h. The left lung was then frozen from RNA analysis. Some mice received an i.p. injection of 200 mg/kg pyrrolidine dithiocarbamate (PDTC; Sigma-Aldrich) in 0.25 ml 1–2 h before removing the lungs. In addition, the left lungs from these mice were incubated in PBS with 6 mg/ml PDTC.
Histopathologic examination
Tissue was placed into 10% buffered formalin. Multiple paraffin-embedded 5-μm sections were prepared and stained with H&E. The slides were evaluated by light microscopy, and the amount of fibroproliferation was assessed by a 0–4 score given by an investigator blinded to the genotype of the animals.
Tissue digests and isolation of single-cell suspensions
Tracheas were excised and minced into small pieces with a scissors. The pieces were digested for 45 min in RPMI 1640 with 0.28 Wunsch units/ml Liberase Blendzyme (Roche) and 30 U/ml DNase (Sigma-Aldrich) for 45 min at 37°C. The digested tissues were then extruded through a mesh strainer.
Flow cytometry and cell sorting
Cells recovered from suspensions of tracheas were stained and analyzed, as previously described (17). All Abs were from BD Pharmingen.
Real-time quantitative PCR (QPCR)
Total RNA was isolated and analyzed by real-time PCR, as previously described (17). RNA was purified using a purification column (RNeasy; Qiagen). After a DNase step, 1 μg of RNA was converted to cDNA (Applied Biosystems). Samples underwent amplification in the presence of SYBR green (Applied Biosystems). The reaction was analyzed in real time during amplification by the PCR machine (MX-4000; Stratagene). Specific primers used for sequence detection of message for the chemokine genes are listed on the web site 〈www.immunologynet.org/〉.
Fibroproliferation
Hydroxyproline was assayed, as described previously (33).
Statistical analysis
Data are expressed as mean ± SEM, unless otherwise indicated. Results were interpreted using a two-tailed Student’s t test. Differences were considered to be statistically significant when p < 0.05.
Results
CXCL9 and CXCL10 have different expression patterns after tracheal transplantation
We performed heterotopic tracheal transplants into syngeneic or allogeneic hosts and removed the grafts at various times for analysis of CXCL9, CXCL10, and CXCL11 mRNA expression by QPCR. After syngeneic transplantation, CXCL10, but not CXCL9 and CXCL11, was up-regulated with an early peak of expression at 1 day (Fig. 1⇓A). After allogeneic transplantation, CXCL10 was up-regulated with two peaks of expression, the first peak at 1 day posttransplant and the second peak at 7 days posttransplant (Fig. 1⇓B). CXCL9 was up-regulated only after allogeneic transplantation starting 5 days after transplant and peaking 7 days after transplant (Fig. 1⇓B). CXCL11 expression was not significantly up-regulated at any time point after allogeneic or syngeneic transplant. CXCR3 expression was progressively up-regulated following allogeneic transplantation (Fig. 1⇓C), consistent with recruitment of CXCR3-expressing cells into the allografts.
Expression patterns of CXCR3 and its ligands after tracheal transplantation. A, CXCL9, CXCL10, and CXCL11 RNA copies normalized to copies of GAPDH RNA in syngeneic heterotopic tracheal transplants removed from recipients after 0, 1, 3, 5, 7, 14, or 21 days (n = 3 sets of 2 tracheas per group). B, CXCL9, CXCL10, and CXCL11 RNA copies normalized to copies of GAPDH RNA in allogeneic heterotopic tracheal transplants removed from recipients after 0, 1, 3, 5, 7, 10, 14, or 21 days (n = 3 sets of 2 tracheas per group). C, CXCR3 RNA copies normalized to copies of GAPDH RNA in allogeneic heterotopic tracheal transplants removed from recipients after 0, 1, 3, 5, 7, 10, 14, or 21 days (n = 3 sets of 2 tracheas per group).
Early CXCL10 expression is derived from both the donor and host tissues
We performed heterotopic syngeneic tracheal transplants and removed the transplanted tracheas at 1, 3, 6, 18, and 24 h after implantation for analysis of CXCL9, CXCL10, and CXCL11 mRNA expression. Increased levels of CXCL10 RNA were noted at 3 h, and its levels increased over the next 24 h (Fig. 2⇓A), while levels of CXCL9 and CXCL11 RNA remained relatively unchanged. We next used CXCL10−/− mice as donors or as recipients to determine the source of the CXCL10. These experiments demonstrated that expression of CXCL10 was completely disrupted at 3 h when CXCL10−/− mice were used as tracheal donors (Fig. 2⇓B). In contrast, at 24 h after implantation, the expression of CXCL10 was disrupted only when CXCL10−/− mice were used as recipients (Fig. 2⇓B). We then used mice with deletion of IFN-γ, the NF-κB subunit p50, and STAT1 to investigate the mechanisms regulating the increased expression of CXCL10 (Fig. 2⇓, C and D). At 3 h, we found that deletion of IFN-γ, p50, or STAT1 had no effect on the levels of CXCL10 RNA. At 24 h postimplantation, however, disruption in either IFN-γ or STAT1 attenuated the increased expression of CXCL10. Disruption in p50 expression did not affect CXCL10 up-regulation at 24 h. These data demonstrate that both donor and recipient tissue produce CXCL10 early after transplantation, and that this expression is regulated by different mechanisms. Interestingly, the very early up-regulation (3 h) of CXCL10 in donor tissue after ischemia seems to be regulated through an IFN-γ-independent mechanism.
Early expression pattern and regulation of CXCL10 after tracheal transplantation. A, CXCL9, CXCL10, and CXCL11 RNA copies normalized to copies of GAPDH RNA in syngeneic heterotopic tracheal transplants removed from recipients after 0, 1, 3, 6, 18, or 24 h (n = 3 sets of 2 tracheas per group). B, Comparison of CXCL10 RNA expression using CXCL10−/− donor tracheas or CXCL10−/− recipients. RNA expression is expressed as fold compared with a completely wild-type transplant (n = 2 sets of 2 tracheas per group). C, Comparison of CXCL10 RNA expression using CXCL10−/− donor tracheas, IFN-γ−/− donor tracheas and IFN-γ−/− recipients, STAT1−/− donors, or p50−/− donors. RNA expression is expressed as fold compared with a completely wild-type transplant (n = 3 sets of 2 tracheas per group). D, Comparison of CXCL10 RNA expression using CXCL10−/− recipients, IFN-γ−/− donor tracheas and IFN-γ−/− recipients, STAT1−/− recipients, or p50−/− recipients. RNA expression is expressed as fold compared with a completely wild-type transplant (n = 3 sets of 2 tracheas per group).
Whole lung ischemia up-regulates CXCL10 expression
To better define the mechanisms leading to increased expression of CXCL10 in the earliest time period (3 h), we developed a model of lung ischemia. In this model, both lungs are flushed free of blood and removed, the right lung is frozen for analysis, and the left lung is placed in warm saline (37°C) for 3 h and then frozen. RNA was then isolated from the lungs for analysis by QPCR. We saw a significant increase in CXCL10 levels in all mouse species tested (Fig. 3⇓A). We then used mice with disruption in various genes that have been shown to affect CXCL10 expression. Disruption of MyD88, TLR4, STAT1, and p50 did not significantly reduce levels of CXCL10 RNA in ischemic lungs. However, treatment of the lungs with the chemical agent PDTC, an antioxidant that has been shown to inhibit NF-κB (34, 35, 36), resulted in 95% inhibition in CXCL10 expression (Fig. 3⇓B). These data suggest that early CXCL10 expression by donor tissue is stimulated through an oxidant-sensitive pathway, possibly through NF-κB, but not through the p50 subunit.
CXCL10 levels in mouse lungs after 3 h of warm ischemia and in human lungs after ischemia-reperfusion. A, CXCL10 RNA expression levels normalized to copies of GAPDH RNA in left lungs from three different strains of mice subjected to 3 h of warm ischemia compared with the right lungs, which experienced no ischemia (n = 3 lungs per group). B, CXCL10 RNA expression levels in left lungs subjected to 3 h of warm ischemia from the indicated genetically modified mice normalized to the expression of CXCL10 in lungs from wild-type mice of the same background strain (n = 3–6 lungs per group). C, Concentration of the indicated chemokines as measured in BAL samples taken from human lungs before removal and ischemia-reperfusion (Pre) and then 18–24 h after implantation (Post) into a transplant recipient (n = 6 samples per group).
CXCL10 is induced following ischemia-reperfusion of human lungs
To correlate our findings to human lung transplantation, we obtained BAL fluid from human lungs before and after ischemia-reperfusion injury and analyzed the fluid for the protein levels of eight different chemokines. Patients donating a single lower lobe for transplantation as part of a living donor transplant program had their lungs lavaged just before resection of the lower lobe and then 18–24 h after implantation into the recipient. Living related lung transplantation involves two healthy people donating one of the lower lobes from their lungs to a person in need of a lung transplant. The donors are extensively screened and do not have any lung disease. In addition, unlike deceased lung donors, they are not brain dead and have not suffered injuries, either of which could influence the production of cytokines and chemokines, such as CXCL10. On average, lungs were ischemic for 90 min before reimplantation. Excess fluid from the BAL taken pre- and postischemia was analyzed by routine ELISA or cytokine bead array. CXCL8, CXCL10, and CCL20 were all significantly up-regulated following ischemia-reperfusion (Fig. 3⇑C). Of the chemokines assayed in the lavage fluid, CXCL10 protein levels were the highest. These data suggest that our findings in mice may have some relevance to the situation in human lung transplantation.
Late expression of CXCL9 and CXCL10 is regulated by IFN-γ and STAT1
We next looked at the tissue origin and regulation of CXCL10 and CXCL9 at the 7-day peak of expression during AR. To do this, we performed heterotopic allogeneic tracheal transplants and removed the allografts after 7 days for analysis of CXCL9 and CXCL10 mRNA levels by QPCR. Compared with wild-type transplants, deletion of CXCL10 in the recipient mouse reduced expression levels by 50%, suggesting equal production of CXCL10 from both the donor trachea and recipient-infiltrating cells (Fig. 4⇓A). Consistent with this, when CXCL10−/− mice were used as tracheal donors into wild-type mice, we also found a ∼50% reduction in CXCL10 levels (data not shown). We then looked at regulation of CXCL10 RNA expression by performing allogeneic tracheal transplants with wild-type tracheas into IFN-γ−/−, STAT1−/−, or p50−/− mice. Deletion of IFN-γ or STAT1 in the recipients led to >98% reduction in the level of CXCL10 RNA. However, transplants with wild-type tracheas into p50−/− recipients did not reduce CXCL10 RNA levels. In contrast to CXCL10, CXCL9 was found to be largely recipient derived, as tracheal transplants with wild-type tracheas into CXCL9−/− recipients reduced expression levels by 97% (Fig. 4⇓B). Similar to the CXCL10 RNA expression data, transplants into IFN-γ−/− or STAT1−/− mice led to >99 and 93% reduction in levels of CXCL9 RNA, respectively. In contrast, transplants with wild-type tracheas into p50−/− recipients led to only mild reductions in CXCL9 expression (20–30%). These data demonstrate that during acute rejection, CXCL10 is derived from both donor and recipient tissue, while CXCL9 is primarily derived from recipient tissue. Furthermore, at this late time point, both chemokines are regulated by recipient tissue-derived STAT1 and IFN-γ.
Late expression pattern and regulation of CXCL9 and CXCL10 after tracheal transplantation. A, CXCL10 RNA expression levels in allogeneic heterotopically transplanted tracheas removed 7 days after transplantation using CXCL10−/− recipient mice, IFN-γ−/− recipient mice, STAT1−/− recipient mice, or p50−/− recipient mice. Copy number is normalized to the number of copies in strain-matched wild-type transplants (n = 3 sets of 2 tracheas per group). B, CXCL9 RNA expression levels in allogeneic heterotopically transplanted tracheas removed 7 days after transplantation using CXCL9−/− recipient mice, IFN-γ−/− recipient mice, STAT1−/− recipient mice, or p50−/− recipient mice. Copy number is normalized to the number of copies in strain-matched wild-type transplants (n = 3 sets of 2 tracheas per group).
Deletion of CXCR3, but not CXCL9 or CXCL10, reduces airway obliteration
We wanted to explore the effects of targeted deletion of CXCL9, CXCL10, and their receptor CXCR3 on the development of OB. To do this, we used mice with deletion of CXCL9, CXCL10, or CXCR3 in tracheal transplants. Because CXCL10 is produced in both donor and recipient tissue, to study the role of CXCL10 we used tracheas from CXCL10−/− BALB/c mice and transplanted them into CXCL10−/− C57BL/6 recipients. CXCL9 is mainly produced from recipient tissue, so to study the role of CXCL9 we were able to use wild-type BALB/c tracheas into CXCL9−/− C57BL/6 recipients. First, we performed orthotopic tracheal transplants, as previously described (17). When wild-type BALB/c tracheas were transplanted into wild-type C57BL/6 recipients, there were significant fibroproliferation and obliteration of donor tracheas 2 wk after transplant (Fig. 5⇓Ai). Transplants of wild-type BALB/c tracheas into CXCL9−/− recipient mice also developed significant obliteration (Fig. 5⇓Aii). Similarly, when tracheas from BALB/c CXCL10−/− mice were transplanted into C57BL/6 CXCL10−/− mice, the tracheas developed fibroproliferation and obliteration that were not different from wild-type transplants (Fig. 5⇓Aiii). In contrast, tracheas of wild-type BALB/c transplanted into CXCR3−/− recipients demonstrated reduced fibroproliferation (Fig. 5⇓Aiv). Quantification of tracheal obliteration by a researcher blinded to the genotype of the animals demonstrated significantly reduced obliteration in the CXCR3−/− recipients, but not in the transplants into CXCL9−/− recipients or with CXCL10−/− donors and recipients (Fig. 5⇓B). However, transplants into CXCR3−/− recipients still had increased obliteration when compared with syngeneic transplants. These data suggest that in this model both CXCL9 and CXCL10 contribute to the development of airway obliteration.
Airway obliteration in orthotopic tracheal transplants with CXCL9−/−, CXCL10−/−, or CXCR3−/− mice. A, Representative histology (n = 6 mice per group) from allogeneic orthotopic tracheal transplants 2 wk after transplantation using wild-type mice (i), CXCL9−/− recipient mice (ii), CXCL10−/− donor and recipient mice (iii), or CXCR3−/− recipient mice (iv). B, Fibroproliferation score of histologic sections taken from orthotopic tracheal transplants 2 wk after transplantation using syngeneic wild-type mice, allogeneic wild-type mice, allogeneic transplants into CXCL9−/− recipient mice, allogeneic transplants with CXCL10−/− donor and recipient mice, or allogeneic transplants into CXCR3−/− recipient mice (n = 6 mice per group). Sections were scored by a single investigator blinded to the origin of the tissue and based on a scale of 0 (no proliferation) to 4 (complete occlusion of the trachea with fibroproliferation).
Deletion of CXCR3, but not CXCL9 or CXCL10, reduces fibroproliferation
For a more quantitative measure of fibroproliferation, we measured hydroxyproline levels of tracheas removed 2 wk after heterotopic transplantation. Hydroxyproline is an amino acid unique to collagen and can be used as a quantitative measure of airway fibrosis (17, 33). Allogeneic transplants into CXCR3−/− recipients had significantly less hydroxyproline content than allogeneic wild-type transplants, but slightly more than syngeneic wild-type transplants (Fig. 6⇓). Allogeneic transplants into CXCL9−/− recipients or with CXCL10−/− donors and recipients had similar hydroxyproline content to the wild-type allogeneic transplants.
Fibroproliferation in heterotopic tracheal transplants with CXCL9−/−, CXCL10−/−, or CXCR3−/− mice. Hydroxyproline content in heterotopically transplanted tracheas removed after 2 wk from wild-type syngeneic transplants, wild-type allogeneic transplants, allogeneic transplants into CXCL9−/− recipient mice, allogeneic transplants with CXCL10−/− donor and recipient mice, or allogeneic transplants into CXCR3−/− recipient mice (n = 4 sets of 2 tracheas per group).
Deletion of CXCR3, but not CXCL9 or CXCL10, reduces lymphocyte recruitment
To explore potential mechanisms for the reduced fibroproliferation and obliteration seen with the tracheal transplants into CXCR3−/− recipients, we examined lymphocyte recruitment into allografts at 7 days, which has been shown to be the peak of lymphocyte recruitment during rejection of the transplanted trachea (17, 33). Transplants into CXCR3−/− recipients had significantly reduced CD4+ and CD8+ T cell recruitment into the allografts compared with wild-type allogeneic transplants (Fig. 7⇓A). We also evaluated the trafficking of effector T lymphocytes using CD25 as a marker of activated T cells. We saw similar reductions in CD4+/CD25+ and CD8+/CD25+ T cell recruitment into allogeneic tracheas transplanted into CXCR3−/− recipients (Fig. 7⇓B). The CD25 marker may also detect some T regulatory cells; however, CD25 expression levels were in the intermediate range, not the high range typical of regulatory T cells (37). Furthermore, a recent study using a heart transplant model demonstrated that there were very few regulatory T cells recruited into the transplanted organ during acute rejection (38). In contrast to the data with CXCR3−/− recipients, allogeneic transplants with CXCL10−/− donors and recipients did not reduce T cell recruitment compared with wild-type transplants. In fact, the numbers were increased in transplants with CXCL10−/− donor and recipient mice relative to wild-type transplants (Fig. 7⇓). This increase in T cell recruitment was significant for CD8+/CD25+ T cells, but not significant for other subtypes. The reasons for an increase in the number of T cells recruited into allografts in the absence of CXCL10 are unclear. Allogeneic transplants into CXCL9−/− recipients had a moderate reduction in the number of lymphocytes recruited into the transplanted tracheas, but the difference was not significant. Because AR is most likely the major factor leading to airway obliteration and is mediated by T cells recruited into the allograft, these data are consistent with the histology and hydroxyproline data presented earlier. Overall, these data suggest that deletion of either CXCL9 or CXCL10 chemokine alone does not reduce the numbers of pathogenic T lymphocytes recruited into tracheal allografts.
T cell recruitment into heterotopic tracheal transplants with CXCL9−/−, CXCL10−/−, or CXCR3−/− mice. A, Numbers of CD4+ and CD8+ T cells recruited into heterotopically transplanted tracheas removed after 7 days from wild-type syngeneic transplants, wild-type allogeneic transplants, allogeneic transplants into CXCL9−/− recipient mice, allogeneic transplants with CXCL10−/− donor and recipient mice, or allogeneic transplants into CXCR3−/− recipient mice (n = 4–8 sets of 2 tracheas per group). B, Numbers of CD4+/CD25+ and CD8+/CD25+ T cells recruited into heterotopically transplanted tracheas removed after 7 days from wild-type allogeneic transplants, allogeneic transplants into CXCL9−/− recipient mice, allogeneic transplants with CXCL10−/− donor and recipient mice, or allogeneic transplants into CXCR3−/− recipient mice (n = 4–8 sets of 2 tracheas per group).
Discussion
Chronic lung dysfunction from OB remains the major factor limiting success in lung transplantation. The recruitment of T lymphocytes into the lung allograft has been shown to be a crucial component of lung ischemia-reperfusion injury and graft rejection, both of which are significant risk factors for the development of OB (16, 18). It follows that a knowledge of the mechanisms that control the recruitment of pathogenic T cells into the lung following transplantation will provide important insights into the pathogenesis of OB and may identify new therapeutic targets for these disorders. Previous studies in both humans and in animal models of transplantation have identified the T cell chemokine receptor CXCR3 and two of its ligands, CXCL9 and CXCL10, as important mediators of chronic organ dysfunction following solid organ transplantation (14, 19, 24, 25, 27, 28, 39). In experiments using a murine model of OB, Belperio et al. (14) demonstrated that inhibition of CXCR3, CXCL9, or CXCL10 delayed airway obliteration. In this study, the investigators showed up-regulation of CXCL9 and CXCL10 in tracheal allografts during the period of acute rejection. They then used polyclonal antiserum to inhibit the activity of CXCL9, CXCL10, or CXCR3. They demonstrated strikingly similar reductions in lymphocyte recruitment into the allografts with each of these treatments used individually and attenuation of OB in the model. These data suggested that despite expression of high levels of both chemokines, lymphocyte recruitment into tracheal allografts was dependent on the combined expression of CXCL9 and CXCL10. This would imply different functional roles for CXCL9 and CXCL10 in the development of OB and/or unique expression patterns for these two chemokines. For example, one could envision a serial interaction between the two chemokines whereby expression of one chemokine mediates early T cell recruitment and expression of the second chemokine at a later time point mediates late T cell recruitment. Our results would seem to support this hypothesis by demonstrating that CXCL10 is induced early following ischemia-reperfusion and CXCL9 is induced later during AR. However, when we examined indices of airway inflammation and fibroproliferation, we found that deletion of either CXCL9 or CXCL10 alone did not affect T cell recruitment into the allograft and airway obliteration in this model. We did, however, confirm that deletion of CXCR3 reduces T cell recruitment and airway obliteration consistent with findings by others in models of OB and in other transplant models (14, 27). The differences between our findings using CXCL9−/− and CXCL10−/− mice with those of Belperio et al., who used polyclonal goat and rabbit antiserum to inhibit CXCL9 and CXCL10, may result from the limitations of using antiserum in vivo.
Ischemia-reperfusion and AR are two of the major causative factors for the development of OB (8, 9, 10). Studies into the pathogenesis of ischemia-reperfusion injury have suggested that CD4+ T lymphocytes may be important cellular mediators of this process and that these cells may ultimately determine the extent of organ injury (15, 40, 41, 42). In a model of liver ischemia-reperfusion, CXCL10 was up-regulated following reperfusion and was felt to be the major factor leading to the recruitment of pathogenic CD4+ T cells into the liver (43). Our data demonstrate a similar up-regulation of CXCL10 in the lung following ischemia-reperfusion. These data suggest that CXCL10 could have a pathogenic role in the development of lung ischemia-reperfusion injury after transplantation, which could potentially influence the incidence of OB (44, 45).
The early increase in CXCL10 levels in transplanted tracheas was initially derived from donor tissue and then subsequently from recipient tissue (likely infiltrating leukocytes). In addition, we were able to demonstrate the up-regulation of CXCL10 levels in mouse lungs with 3 h of ischemia and in human lungs 24 h after transplantation. This is the first demonstration of CXCL10 up-regulation in human lungs following ischemia-reperfusion injury and correlates well with our findings in mice. Interestingly, CXCL9 was not induced early following ischemia in mouse or human lungs. Although CXCL9 and CXCL10 are both up-regulated by IFN-γ, they have different regulatory elements in their promoters that differentially control their expression. Specifically, the CXCL10 promoter has an IFN-stimulated regulatory element and two NF-κB sites, while the CXCL9 promoter has two IFN-γ activation sites (46, 47, 48, 49, 50, 51, 52). Studies into the gene regulation have demonstrated that CXCL9 induction is completely dependent on the presence of the IFNs (mainly IFN-γ) (52), whereas CXCL10 may be induced by both IFN-dependent and independent pathways (53). Consistent with this, our findings suggest that the very early expression of CXCL10 in lung ischemia-reperfusion is potentially regulated by NF-κB, but not through the p50 subunit. Furthermore, the increase in CXCL10 was not mediated by MyD88 or TLR4, as had been demonstrated in a model of liver ischemia-reperfusion (43). Based on the fact that PDTC inhibited the increase in CXCL10 following lung ischemia-reperfusion, we suspect that oxygen-free radicals released by ischemia-reperfusion are directly activating the p65 homodimer form of NF-κB, which then stimulates CXCL10 expression (54, 55, 56, 57).
Although the role of T cells in the pathogenesis of ischemia-reperfusion injury remains relatively undefined, it is clear that AR is primarily orchestrated by activated graft-specific T lymphocytes (58). In this process, effector T lymphocytes specific for donor Ags traffic into the airway and initiate a cascade of inflammation that leads to damage to the airway epithelium (59, 60). Thus, donor-specific T lymphocyte infiltration in AR is most likely crucial for the subsequent development of OB (61, 62, 63). Support for this hypothesis has come from experiments that have demonstrated donor-specific T lymphocytes in BAL fluid and lung tissue taken from transplant recipients with OB (61, 64). It follows that interventions that block lymphocyte recruitment into allografts during AR should reduce the intensity of rejection and delay the development of OB. Similar to the findings of Belperio et al., we saw dramatic increases in CXCL9 and CXCL10 RNA levels during acute rejection of tracheal allografts. CXCL9 was mainly derived from recipient tissue, while CXCL10 was derived equally from both the donor and recipient tissue. However, induction of both during AR was dependent on recipient tissue-derived IFN-γ and STAT1, but not by the p50 subunit of NF-κB. In these experiments, we suspect that recipient leukocytes that have been recruited into the tracheal allograft produce IFN-γ, which initiates CXCL10 production from both donor and recipient tissue. For the requirement for recipient-derived STAT1, the issue is more complicated because STAT1 is most likely involved in multiple mechanisms in the pathogenesis of rejection. Thus, one could envision a situation in which the initiation of the rejection process and the initial production of CXCL10 are largely dependent on recipient-derived STAT1. Once rejection of the transplanted trachea is started, pathogenic lymphocytes are recruited into the graft and initiate injury, and the recipient and donor tissue responds with further production of CXCL10 and amplification of the response. Thus, although CXCL10 is derived from both tissues, recipient-derived STAT1 would initiate the entire process.
Despite distinctive patterns of expression following tracheal transplantation, our data demonstrate that deletion of either CXCL9 or CXCL10 alone does not reduce the development of airway obliteration. Although studies in humans post-lung transplantation and in murine models of disease have demonstrated the up-regulation of several other T lymphocyte chemoattractants, including CCL5/RANTES and leukotriene B4 (17, 20), the fact that deletion of CXCR3 in transplant recipients can reduce airway obliteration suggests that CXCL9 and/or CXCL10 must contribute significantly to this process. We were thus surprised that we were unable to demonstrate any effect of CXCL9 or CXCL10 deletion on the development of OB in our model. This was especially true of CXCL10, because it was up-regulated early in response to ischemia-reperfusion to the graft, a process thought to contribute to the development of OB. This could be due to the fact that CXCL10 is not a major mediator of airway injury following ischemia-reperfusion or, more likely, the subsequent unsuppressed rejection of the allograft that occurs later, as a result of CXCL9 or expression of some other chemokine, is sufficient to lead to airway obliteration and compensates for any reduction in early injury. Similarly, with tracheal transplants into CXCL9−/− recipients, we suspect that expression of CXCL10 or some other chemokine during AR recruits in enough pathogenic T cells to compensate for the absence of CXCL9. Because we only had CXCL9−/− mice in one background strain, we were unable to fully delete its expression, although transplants into CXCL9−/− recipients reduced expression by 97%. However, it is still possible that CXCL9 is the major CXCR3 ligand mediator of tracheal AR and OB and that the low level expression of CXCL9 in donor tissue was enough to initiate injury and cause obliteration.
Based on our data, it seems that a strategy that targets either CXCL9 or CXCL10 would not be effective to prevent AR or OB as deletion of either chemokine alone does not affect pathogenic T cell recruitment into allografts. However, our data do provide support for therapeutic strategies that target CXCR3 expression in recipients of lung transplants.
Acknowledgments
We thank Andrew Carafone and Carol Leary for technical support with this research.
Disclosures
The authors have no financial conflict of interest.
Footnotes
-
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
-
↵1 This work was supported by grants from the National Institutes of Health (K08 HL072775 to B.D.M.; R01 AI050892 to A.D.L.), the Roche Organ Transplant Research Foundation (to A.D.L.), the Dana Foundation (to A.D.L.), and the Nirenberg Fellowship in Advanced Lung Disease (to B.D.M.).
-
↵2 Address correspondence and reprint requests to Dr. Andrew D. Luster, Massachusetts General Hospital, CNY 8301, 149 13th Street, Charlestown, MA 02129. E-mail address: luster.andrew{at}mgh.harvard.edu
-
↵3 Abbreviations used in this paper: OB, obliterative bronchiolitis; AR, acute rejection; BAL, bronchoalveolar lavage; PDTC, pyrrolidine dithiocarbamate; QPCR, real-time quantitative PCR.
- Received November 9, 2005.
- Accepted March 20, 2006.
- Copyright © 2006 by The American Association of Immunologists
References
In this issue
