Development of allograft rejection continues to be the major determinant of morbidity and mortality postlung transplantation. We have recently demonstrated that a population of donor-derived mesenchymal stem cells is present in human lung allografts and can be isolated and expanded ex vivo. In this study, we investigated the impact of lung resident mesenchymal stem cells (LR-MSCs), derived from allografts of human lung transplant recipients, on T cell activation in vitro. Similar to bone marrow-derived MSCs, LR-MSCs did not express MHC II or the costimulatory molecules CD80 or CD86. In vitro, LR-MSCs profoundly suppressed the proliferative capacity of T cells in response to a mitogenic or an allogeneic stimulus. The immunosuppressive function of LR-MSCs was also noted in the absence of direct cell contact, indicating that LR-MSCs mediated their effect predominantly via a soluble mediator. LR-MSCs isolated from lung transplant recipients demonstrated PGE2 secretion at baseline (385 ± 375 pg/ml), which increased in response to IL-1β (1149 ± 1081 pg/ml). The addition of PG synthesis inhibitors (indomethacin and NS-398) substantially abrogated LR-MSC-mediated immunosuppression, indicating that PGE2 may be one of the major soluble mediators impacting T cell activity. This is the first report to demonstrate that human tissue-derived MSCs isolated from an allogeneic environment have the potential to mediate immunological responses in vitro.
Allograft rejection continues to be the factor limiting successful outcomes following lung transplantation. Acute allograft rejection is the most common complication following lung transplantation, seen in up to 50–60% of lung transplant recipients by 1 year (1), and predisposes to the development of chronic rejection or bronchiolitis obliterans syndrome (BOS)4 (2, 3, 4, 5). BOS develops in 60% of transplant recipients by 5 years, causes fixed obstruction of the airways, and is the major cause of mortality after 1 year (6). T cell-based allorecognition is important in driving the development of acute lung rejection and BOS (7). Acute allograft rejection is a T cell-mediated response against alloantigens (1), and both CD4+ and CD8+ T cell populations are important in mediating the complex events leading to chronic allograft rejection (8, 9).
Mesenchymal stem cells (MSCs) are a unique subset of progenitor cells defined by their capacity to differentiate into multiple mesenchymal lineages (10, 11). MSCs isolated from bone marrow (BM-MSCs) have attracted interest as key modulators of local immune responses (12), having been reported to profoundly inhibit T cell proliferation in vitro (13, 14, 15, 16, 17). We recently described a population of MSCs that can be isolated and expanded from bronchoalveolar lavage (BAL) fluid of human lung allograft patients (18). Studies of sex-mismatched lung allograft donors and recipients demonstrated that these cells are lung resident and donor derived. Furthermore, gene chip analysis of these lung resident mesenchymal stem cells (LR-MSCs) revealed that they possess a unique repertoire of cytokines and growth factors that differ from those of BM-MSCs (18).
Because LR-MSCs represent an endogenous population of lung progenitor cells that are distinctly different from BM-MSCs and are components of the transplanted lung cellular milieu, we aimed to investigate whether these cells possess immunoregulatory potential. This study demonstrates that, in vitro, LR-MSCs inhibit the proliferation of diverse subsets of allogeneic T cells in response to both mitogenic and antigenic stimuli, and they seem to mediate their effects via both contact with and elaboration of soluble mediators. Our data indicate that one such mediator contributing to T cell unresponsiveness is the prostaglandin PGE2.
Materials and Methods
Isolation and identification of lung-derived MSCs
Mesenchymal stem cells were obtained from BAL of lung transplant recipients as previously described under a protocol approved by the Institutional Review Board of the University of Michigan Health Care System (18). BAL samples were used only if a bronchoscopy was performed for routine surveillance within 6 mo of transplantation with no evidence of acute rejection or infection on transbronchial biopsies and microbiological cultures, respectively. Cells were maintained in culture in DMEM with penicillin/streptomycin and 10% FCS at 37°C in 5% CO2 and used at passages 2–6. LR-MSCs obtained from individual BAL samples were treated as separate cell lines.
PBMCs from healthy volunteers were fractionated by density centrifugation using Lymphoprep gradient (Axis-Shield). The buffy coat was isolated and subpopulations of T cells (Pan T cell, CD4+, CD4+CD25−, and CD8+) were purified using the relevant magnetic MicroBead kits (Miltenyi Biotec) according to manufacturer’s instructions.
Proliferation assays and cytokine measurement
LR-MSCs (2 × 104; 30 Gy irradiated) were plated per well in a 96-well flat-bottom culture-treated plates 4 h before the addition of Pan T cells or subpopulations of T cells (2 × 105 T cells/well) in complete DMEM supplemented with 10% FCS. For mitogenic stimulation, the cultures were stimulated with PHA (12.5 μg/ml; Sigma-Aldrich) for 5 days. Eighteen hours before harvest, [3H]thymidine (1μCi) was added. [3H]Thymidine incorporation is expressed as the mean of triplicates in cpm. For dose-response experiments, 2 × 105 responder T cells were incubated with various numbers of MSCs. For allogeneic stimulation, 2 × 104 MSCs/well were plated in 96-well, round-bottom, culture-treated plates 4 h before the addition of the stimulator (1 × 105 PBMC/well; 15 Gy irradiated) and 1 × 105cells/well of third party responder PBMCs or responder Pan T cells. Cells were cocultured in complete RPMI 1640 supplemented with 10% FCS for 8 days with the addition of [3H]thymidine for the last 18 h of culture. To mimic APCs, CD28/CD3/CD2 Ab-coated beads (Miltenyi Biotec) were used to induce proliferation according to the manufacturer’s instructions at a ratio of 1:2 bead to T cell. The cells were cultured in the presence and absence of LR-MSCs and assessed for proliferation 3 days later. For cytokine measurement, third party 30 Gy-irradiated MSCs (8 × 104 MSC/well) were plated into 24-well, flat-bottom plates in DMEM 4 h before the addition of 4 × 105 responder T cells/well in the presence or absence CD28/CD3/CD2 Ab-coated beads. After 72 h, cell-free supernatants were collected and analyzed for IL-2 and IL-10 levels using a highly specific enzyme immunoassay technique (eBioscience and R&D Systems, respectively).
LR-MSCs (30 Gy irradiated) (2 × 105 cells/well) were plated in the upper chamber of a 96-well Transwell plate with a 0.5-μm pore size membrane (Corning) 4 h before the addition of Pan T cells (2 × 10 4 cells/well) to the lower chamber. Cultures were stimulated by the addition of 12.5 μg/ml PHA and [3H]thymidine was added for the last 18 h of culture.
Determination of PGE2 synthesis and immunoblot analysis
To measure PGE2 in BAL, lipids were extracted from BAL fluid samples via solid phase extraction using Sep-Pak C-18 cartridges (Waters). Eluted sample was reconstituted in assay buffer and PGE2 was quantified in using a highly sensitive and specific enzyme immunoassay kit from Cayman Chemicals according to the manufacturer’s suggestions. For measurement of PGE2 synthesis by LR-MSCs, LR-MSCs were cultured in the absence or presence of IL-1β (10 ng/ml) for 24 h. PGE2 was quantified in the supernatant by ELISA. In separate experiments, protein was isolated and cyclooxygenase (COX)-1 and COX-2 proteins were analyzed by Western blotting as previously described (19).
Data are presented as mean values ± SEM. Statistical significance was analyzed using the GraphPad Prism 3 software. Significance was assessed with a Student’s t test for comparisons of two groups or with ANOVA and a post hoc Bonferroni test for three or more groups. p < 0.05 was considered significant.
LR-MSCs derived from human lung allografts fail to stimulate allogeneic T cells
We have previously demonstrated that donor LR-MSCs can be isolated as late as 11 years postlung transplantation, clearly indicating that these cells can survive in an allogeneic environment (18). Because donor LR-MSCs are exposed to recipient circulating cells, we first investigated whether LR-MSCs cultured with third party T cells lead to T cell proliferation in vitro. No proliferation was noted when Pan T cells were cultured with irradiated allogeneic LR-MSCs (2,125 ± 178 cpm and 2,311 ± 211.9 cpm mean T cell [3H]thymidine incorporation in the absence and presence of MSCs, respectively; p = 0.53; Fig. 1,A). Similar results were noted when LR-MSCs were cocultured with peripheral blood PBMCs and a subpopulation of T cells (CD4 and CD8 T cells; data not shown). We have previously demonstrated that LR-MSCs express HLA-I but not HLA II (DR or DQ) (18). Examination of costimulatory molecules demonstrated that, in addition, LR-MSCs do not express CD80 or CD86, indicating that they do not function as classical APCs (Fig. 1 B).
LR-MSCs inhibit T cell proliferation in response to mitogen stimulation in vitro in a dose-dependent manner
Because LR-MSCs were not immunogenic, we next sought to determine whether, similarly as BM-MSCs, they had an immunosuppressive impact on lymphocytes in response to mitogenic stimulation in vitro. LR-MSCs were cocultured for 4 days with peripheral blood T cells of third party healthy volunteers in the presence of PHA at a ratio of 1:10 LR-MSCs to T cells. Upon examination of the coculture under light microscopy, we observed that wells containing T cells alone in the presence of PHA displayed typical T cell clustering, indicative of a robust proliferative response. In contrast, wells containing a monolayer of LR-MSCs in addition to T cells and PHA lacked T cell clustering, with the T cells appearing to be distributed singularly on top of the LR-MSC monolayer. Additionally, T cells in the coculture appeared alive and healthy, and their viability was confirmed by trypan blue exclusion (not shown). Upon quantitative analysis obtained by [3H]thymidine incorporation, there was a significant reduction in PHA-induced T cell proliferation in the presence of LR-MSCs ((84 ± 7%; p = 0.0001; Fig. 2,A and B). Lower inhibition was noted with a decreasing ratio of LR-MSCs to responder T cells (Fig. 2 C).
LR-MSCs inhibit both CD4 and CD8 T cell populations in vitro and do this independently of preexisting regulatory T cell (Treg) populations
Because both CD4+ and CD8+ T cells have been shown to mediate allograft rejection, we aimed to determine whether LR-MSCs can block T cell proliferation of these individual cell populations in vitro. Both CD4+ and CD8+ T cells stimulated by PHA exhibited a markedly decreased proliferative capacity when cocultured with LR-MSCs, with [3H]thymidine incorporation reduced to unstimulated baseline levels (p < 0.001 and 0.002, respectively; Fig. 3, A and B). To determine whether LR-MSCs require the presence of preexisting CD4+CD25+ regulatory T cells to mediate this effect, peripheral blood CD4+ T cells populations were depleted of CD4+CD25+ T cells, cocultured with LR-MSCs, and stimulated with PHA as previously described. LR-MSCs significantly inhibited the proliferation of the CD4−CD25− population, indicating that preexisting regulatory T cells are not necessary for LR-MSCs to inhibit T cell activation in vitro (p < 0.001; Fig. 3 C).
LR-MSCs suppress T cell proliferation in response to allogeneic stimuli
Although LR-MSCs were effective at regulating T cell activation in response to mitogenic stimulation, allogeneic stimulation may be more relevant to the transplanted lung. To examine this in vitro, we performed a MLR assay in which T cell-depleted and irradiated PBMCs were used to stimulate allogeneic Pan T cells in the presence or absence of third party MSCs. The presence of LR-MSCs led to significant suppression of T cell proliferation in response to allogeneic stimulation (Fig. 4,A). LR-MSC inhibitory effects in an MLR could be exerted directly on effector T cells or mediated indirectly via effects on the APCs. To differentiate between these possibilities, we examined the effect of LR-MSCs on T cell proliferation in response to CD28/CD3/CD2 Ab-coated beads, circumventing the need for APCs. The presence of LR-MSCs caused significant inhibition of T cell proliferation in response to CD28/CD3/CD2 Ab-coated beads (p = 0.008; Fig. 4 B), demonstrating that the LR-MSC effect on T cells in an MLR is not mediated via its action on the APCs.
Soluble mediators play an important role in mediating T cell suppression by LR-MSC
To determine whether MSCs require direct contact with effector T cells to abrogate their proliferative response, we cultured the two populations in a Transwell apparatus that separated the two populations by a porous membrane. In the absence of direct cell contact, LR-MSCs were still able to inhibit T cell proliferation in response to PHA, albeit to a lesser extent (T cell proliferation was reduced by 58 ± 10%; Fig. 5,A). This inhibition was lower than what was seen with direct cell-cell contact (inhibition of 84 ± 7%). Transwell experiments were repeated with the use of both irradiated and nonirradiated LR-MSCs (n = 3), and no significant differences were noted (p = 0.33; Fig. 5 B). These data suggest that soluble mediators are important for the T cell-inhibitory affects of LR-MSCs but that direct cell contact further facilitates the suppressive ability of LR-MSCs.
Prostanoid production by LR-MSCs contributes to their ability to inhibit T cell proliferative responses in vitro
PGE2 is an immunomodulatory mediator that has been implicated as one of the potential candidates responsible for T cell inhibition by MSCs (16, 20). Like all prostanoids, PGE2 is synthesized from arachidonic acid via the actions of either the constitutive COX-1 or the inducible COX-2 enzyme (21). PGE2 has an important role in pulmonary diseases and has been demonstrated to be present in BAL obtained from human lungs (22). PGE2 was quantitated in BAL obtained from lung transplant recipients (n = 14), and measurable PGE2 was detected in all BAL samples (mean = 420 pg/ml BAL fluid; range = 142 to 1212 pg/ml). Next, we examined the PGE2 synthetic capacity of lung allograft-derived MSCs isolated from the BAL of lung transplant recipients. LR-MSCs (n = 10) demonstrated PGE2 secretion at baseline (385 ± 375 pg/ml) with up-regulation in response to IL-1β (1149 ± 1081 pg/ml) (Fig. 6,A). Up-regulation of COX-2 in response to IL-1β was also demonstrated by Western blot analysis (Fig. 6 B). Because all coculture experiments were performed with irradiated LR-MSCs, the effect of irradiation on LR-MSC PGE2 secretion was studied in four separate cell lines. Irradiation of LR-MSCs did not alter their PGE2 secretory ability at baseline (273 ± 81 and 258 ± 89 pg/ml in the absence and presence of irradiation, respectively; p = 0.56) or in response to IL-1β stimulation (790 ± 102 and 783 ± 92 pg/ml in the absence and presence of irradiation, respectively; p = 0.78).
As has been demonstrated previously, the addition of PGE2 to T cells inhibited their proliferation in response to both mitogenic and allogeneic stimulation (Fig. 6, C and D). To determine whether prostanoid secretion by LR-MSCs is important for mediating T cell suppression, irradiated LR-MSCs, T cells, or T cell/LR-MSC cocultures stimulated with PHA were treated with either the nonselective COX inhibitor indomethacin (10 μM) or the selective COX-2 inhibitor NS-398 (5 μM). T cell suppression in the presence of LR-MSCs was substantially reversed by the inhibition of PG generation by both of those drugs, demonstrating that an endogenous prostanoid, with PGE2 being a likely candidate, represents an important soluble mediator by which LR-MSCs exert T cell inhibition (Fig. 6 E).
Prostanoid production by LR-MSCs modulates T cells cytokine secretion
IL-2 and IL-10 are important cytokines secreted by T cells (23, 24) that can alter T cell proliferation and function. To determine whether LR-MSCs modulate T cell secretory ability, Pan T cells were activated by CD28/CD3/CD2 Ab-coated beads in the presence or absence of LR-MSCs. Cytokine determinations by ELISA in the supernatants from the cocultures demonstrated a significant increased production of the regulatory cytokine IL-10 and a decreased production of the proinflammatory cytokine IL-2 in the presence of LR-MSCs. The addition of the COX inhibitor indomethacin completely reversed IL-2 inhibition and partially ameliorated IL-10 increase, demonstrating that PGE2 secreted by LR-MSCs contributes to the modulation of cytokine secretion by T cells (Fig. 7, A and B).
In this study we examined interactions between T cells and donor-derived LR-MSCs isolated from the BAL of human lung transplants. We demonstrate that, in vitro, LR-MSCs significantly inhibited proliferation of third party, HLA-mismatched T cells. This effect was demonstrated in the context of both mitogen and alloantigen stimulation and across CD4 and CD8 T cell populations and did not require the presence of preexisting CD4+CD25+ T cells. LR-MSC-mediated suppression of T cells was partially contact dependent and largely explained by the actions of a soluble mediator. COX inhibitors blocked the immunosuppressive potential of LR-MSCs on T cells, suggesting that PGE2 may be a major mediator of the immunomodulatory effects of LR-MSC in vitro. This is the first report demonstrating that MSCs isolated from a transplanted lung have immunosuppressive capacity in vitro, a finding that contributes to our understanding of the immunosuppressive potential of tissue resident MSCs.
Transplantation offers the only definitive therapy for a variety of end stage lung diseases. However, long-term outcomes for lung transplant recipients are poor, with a 10-year lung survival rate of only 26% (25). A major challenge in lung transplantation continues to be allograft rejection orchestrated by recipient derived-inflammatory cells, primarily T cells, which infiltrate the graft during acute rejection (1). Recipient-derived T lymphocytes also play a seminal role in the pathogenesis of BOS. Lymphocytic bronchitis, a condition characterized by lymphocytic infiltration of the small airways, is considered to be a harbinger of BOS in humans (26) and precedes the development of luminal obliteration in animal models of tracheal transplantation (27, 28, 29). Although host responses to allograft are very well characterized, the role of graft-derived cells in a transplant milieu remains to be elucidated. Our study suggests that resident donor-derived cells might have a role in modulating the local microenvironment in the lung allograft, a novel paradigm that begets further investigation.
We demonstrate that lung resident MSCs derived from allografts possess immunosuppressive potential. MSCs are a unique, well-characterized population of progenitor cells characterized by their ability to differentiate into multiple mesenchymal lineages (30). Although an emerging body of data indicates that MSCs possess immunomodulatory properties both in vitro and in vivo (12, 31, 32), our study is unique is several ways. As we have clearly demonstrated that MSCs isolated from allografts are donor derived and hence resident in a human adult lung (18), this is the first study to report that tissue-resident populations of MSCs possess immunoregulatory potential. It is important to note that we failed to isolate meaningful numbers of BM-MSCs from the BAL of lung allograft recipients, indicating that they were not major contributors to the lung MSC population (18). The facts that LR-MSCs possess immunoregulatory potential, are the major population isolated from a lung allograft, and can be easily isolated from BAL make them an attractive therapeutic option for prolonging lung allograft acceptance. Further, our study demonstrates for the first time that MSCs isolated from an allogeneic milieu retain their immunosuppressive potential. The majority of prior reports on the immunoregulatory role of BM-MSCs have used cells isolated from healthy volunteers. In a single report, Bacigalupo et al. demonstrated that BM-derived MSCs from patients with sickle cell anemia were deficient in their ability to down-regulate T cell proliferation (33). In this study we examined cells isolated from lungs posttransplantation and demonstrate their ability to suppress T cell proliferation. Variability was noted in the degree of suppression by LR-MSCs isolated from different donors, signifying the need to investigate further the mechanism of this variability and its clinical implications. Furthermore, it should be noted that the cells used in this study were isolated early posttransplantation in the absence of histological evidence of rejection. Whether the T cell suppressive ability of LR-MSCs changes over time and predicts the development of acute and chronic allograft rejection in human lung transplant recipients remains to be established. The ability to isolate LR-MSCs with a minimally invasive procedure used routinely in the management of lung transplant recipients (bronchoscopy and BAL) offers a unique opportunity to study these cells in vitro and correlate their phenotypic properties to clinical outcomes in future studies.
The mechanisms by which MSCs mediate T cell suppression remain to be entirely elucidated (12, 31). Various studies, including our own, support the idea that a significant contributor to the immunosuppressive effects of MSC is the production of soluble mediators (13, 16). PGE2, an immunomodulatory lipid mediator, is presently one of the leading candidates for MSC-induced immune suppression (16, 20). PGE2 is a well-established inhibitor of T cell proliferation (34, 35). It has been demonstrated that PGE2 inhibits IL-2 secretion and the subsequent proliferation of T cells via a cAMP (protein kinase A)-dependent mechanism (36, 37). Similarly, by increasing intracellular cAMP, PGE2 has been demonstrated to augment IL-10 expression by T cells (38). The G protein-coupled receptors EP2 and EP4 are thought to be important in modulation of cellular immune responses by PGE2 (39). Recently PGE2 has been recently demonstrated to both induce Tregs (40) and potentiate the action of Tregs (41). Our results demonstrate that PGE2 is an important secretory product of LR-MSCs and that inhibition of a PG synthetic pathway attenuates LR-MSC-induced inhibition of T cells, pointing to a role for PGE2 in the immunosuppressive actions of LR-MSCs. Furthermore, our data on IL-2 and IL-10 modulation by LR-MSCs suggests that LR-MSC-derived PGE2 likely acts through similar pathways as those which have been characterized previously. However it should be noted that this reversal of inhibition by indomethacin was partial (∼60%), and the remaining inhibition of T cells by LR-MSCs in presence of a COX inhibitor could likely be secondary either to other soluble factors or to a contact-mediated interaction. It is interesting to note that this degree of reversal of inhibition is similar to the percentage of inhibition of T cells in the Transwell experiment, indicating that among soluble mediators, PGE2 is the likely dominant factor. The mechanism of increased inhibition in cocultures with direct contact among LR-MSCs and T cells remains to be investigated and will be the focus of future studies.
The findings of this study also provide a possible explanation for our previous observation that donor MSCs can be isolated as far out as 11 years postlung transplantation (18). Their lack of MHC-II and costimulatory molecules, failure to stimulate allogeneic T cells, and ability to induce a suppressive local microenvironment by secretion of PGE2 likely explain the in vivo capacity of LR-MSCs to survive in an allogeneic environment. These results are similar to what has been described for BM-MSCs (42).
In summary, we demonstrate that resident MSCs derived from lung allografts inhibit T cell proliferation via the elaboration of soluble mediators such as PGE2. These findings point to a possibly important immunoinhibitory role of this novel resident endogenous progenitor cell population in a lung allograft milieu, demonstrating the need to study the contribution of local graft-derived cells in the pathogenesis of allograft rejection and providing hope for the potential use of LR-MSCs as a therapeutic means of inducing long-lived allograft tolerance.
The authors have no financial conflict of interest.
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.
This work was supported by National Institutes of Health Grants K23 HL077719 (to V.N.L.), R01 HL085149-01 (to D.J.P.), and R01 HL55397 (to D.J.P.), an American Society of Transplantation and Chest Foundation clinical research award in lung transplantation (to V.N.L.), P01 HL 89407 (to D.P.J.), Taubman Medical Research Institute (to D.P.J.), American Thoracic Society Research Award (to V.N.L.), and a Scleroderma Research Foundation award (to V.N.L. and D.J.P.).
Abbreviations used in this paper: BOS, bronchiolitis obliterans syndrome; BAL, bronchoalveolar lavage; BM-MSC, bone marrow-derived MSC; COX, cyclooxygenase; LR-MSC, lung resident MSC; MSC, mesenchymal stem cell; Treg, regulatory T cell.