|Home | About | Journals | Submit | Contact Us | Français|
Early graft inflammation enhances both acute and chronic rejection of heart transplants, but it is unclear how this inflammation is initiated.
To identify specific inflammatory modulators and determine their underlying molecular mechanisms after cardiac transplantation.
We used a murine heterotopic cardiac transplant model to identify inflammatory modulators of early graft inflammation. Unbiased mass spectrometric analysis of cardiac tissue before and up to 72 hours after transplantation revealed that 22 proteins including haptoglobin, a known anti-oxidant, are significantly upregulated in our grafts. Through the use of haptoglobin deficient mice, we show that 80% of haptoglobin deficient recipients treated with peri-operative administration of the costimulatory blocking agent CTLA4 immunoglobulin exhibited > 100 days survival of full major histocompatibility complex mismatched allografts, whereas all similarly treated wild type recipients rejected their transplants by 21 days post transplantation. We found that haptoglobin modifies the intra-allograft inflammatory milieu by enhancing levels of the inflammatory cytokine IL-6 and the chemokine MIP-2 but impair levels of the immunosuppressive cytokine IL-10. Haptoglobin also enhances dendritic cell graft recruitment and augments anti-donor T cell responses. Moreover, we confirmed that the protein is present in human cardiac allograft specimens undergoing acute graft rejection.
Our findings provide new insights into the mechanisms of inflammation after cardiac transplantation and suggest that, in contrast to its prior reported anti-oxidant function in vascular inflammation, haptoglobin is an enhancer of inflammation after cardiac transplantation. Haptoglobin may also be a key component in other sterile inflammatory conditions.
Sterile inflammation occurs in several medical conditions, including after organ transplantation 1–3. In the latter, the harvest and subsequent preservation of an organ induces ischemic injury to the transplant, which is exacerbated at organ implantation by restoration of blood flow, a condition known as ischemia reperfusion injury (IRI) 4. IRI, in turn, induces intra-graft inflammation, which can have detrimental effects on long-term allograft function 5.
In organ transplantation, several putative inflammatory triggers including cellular chaperones, nuclear proteins, and components of the extracellular matrix have been associated with acute allograft rejection 6–8. These putative triggers may be released by the cellular necrosis that accompanies organ implantation and reperfusion. However, it is unclear what role these triggers play in inflammation induction after organ transplantation. Importantly, inhibition of some of these triggers only modestly extends cardiac allograft survival 8, indicating that other yet undiscovered factors may contribute to inflammation after cardiac transplantation.
As graft inflammation has detrimental effects on graft outcomes 9, the identification of novel inflammatory mediators is therefore key for identifying lead candidates for development of novel therapies to reduce inflammation after organ transplantation. Such lead candidates are particularly relevant to organ transplantation as the harvest and preservation of an organ provides a therapeutic window of opportunity to treat the graft prior to implantation while avoiding the inhibition of host defense pathways in recipients, an important strategy given that organ transplant recipients are intentionally immune suppressed.
Here, we used an unbiased proteomic screen to identify novel inflammatory modulators in transplanted tissue after cardiac transplantation in a murine model. Our study revealed that by 72 hours after transplantation 22 proteins are significantly upregulated within the graft including haptoglobin, a heme-binding protein with anti-oxidative properties. In our murine cardiac transplant model, haptoglobin remarkably prevents the effects of costimulatory blockade therapy to induce indefinite allograft survival. We also found the protein in human cardiac allograft specimens undergoing acute graft rejection. Haptoglobin therefore may be a potential candidate for the development of novel therapeutics to reduce inflammation within cardiac allografts and possibly other organ transplants.
Note that detailed experimental procedures and methods are reported in the Methods section of the online data supplement.
C57BL/6J (stock# 000664), BALB/cJ (stock# 000651) mice were purchased from Jackson Laboratories (Bar Habor ME). C57BL/6 haptoglobin deficient (Hp−/−), C57BL/6 MyD88−/− (each backcrossed ten times) and relevant wild type littermate controls 2–4 months of age, both males and females, were maintained in our colony. The Hp−/− mice were initially provided by Dr. Maffei (Dulbecco Telethon Institute, University Hospital of Pisa, Pisa Italy). Use of vertebrate animals was approved by the Yale School of Medicine IACUC.
A murine heterotopic heart transplant model was employed to assess intra-graft inflammatory responses 10 (detailed in the supplemental material). Anesthesia was induced by isofluorane via a pressurized vaporizer and maintained with ketamine (0.1mg/g body weight) and xyalazine (0.1mg/g body weight).
To determine if cardiac transplantation led to an increase in inflammatory modulators, we procured cardiac lysates pre- and post-transplantation and cultured them with bone marrow-derived dendritic cells (DCs) as previously reported 11. Cardiac lysates were free of LPS (Limulus assay) and did not contain any microbial peptides as assessed by mass spectrometry. Lysates were treated with pronase (0.5U/ml, Calbiochem-Behring Corp, La Jolla, CA), DNase I (60 unit/ml, Roche) or RNase (10ug/ml, Thermo Scientific) at 37°C for 1h and then added to the DC cultures overnight as indicated. Target cytokines and chemokines analytes in the cardiac lysates were measured by ELISA for IL-6, IL-10, MCP-1, (eBioscience), MIP-2, (R&D Systems), HO-1 (Enzo Life Sciences, Farmingdale, NY), and haptoglobin (Immunology Consultants Laboratory, Tampa, FL). We enriched subpopulations of cells within cellular suspension from cardiac tissue with a serial magnetic enrichment protocol (magnetic beads and columns from Milenyi Biotec, Inc., CA) as detailed in the online data supplement. Purified human haptoglobin was obtained from Sigma and found to contain <40pg/ml of LPS. Mixed lymphocyte cultures consisted of magnetically enriched T cells (Stem Cell Technologies, Vancouver, British Columbia) cultured in a 1:5 ratio with irradiated donor spleen cells. Anti-donor T cell responses were assessed either by measuring cytokine production by ELISPOT, ELISA or proliferation by thymidine incorporation. Cellular suspensions (i.e., spleen cells and cells harvested from cardiac tissue) were stained with relevant fluorescent-tagged antibodies and data acquired on an LSR II (BD Bioscience) flow cytometer and analyzed with FlowJo software (Treestar, Ashland, OR).
To determine the presence of intra-graft haptoglobin after human heart transplantation, we stained for haptoglobin in human, archived endomyocardial specimens from the Pathology Department of the Yale New Haven Hospital. The samples had been histologically scored as either no evidence of cellular graft rejection or moderate cellular graft rejection according to the diagnostic criteria of the International Society for Heart and Lung Transplantation 12. These specimens were paraffin-embedded at the time of routine protocol biopsy after heart transplantation and were de-paraffinized and counterstained with an anti-human haptoglobin antibody (AbCam, Cambridge, MA) or isotype control via immune histochemistry. Slides were read blindly. Use of clinical specimens and accessing patient data from de-identified medical records was approved by the Human Regulatory Board at Yale School of Medicine.
To assess histological inflammation after cardiac transplantation in mice, native and transplanted hearts were harvested and fixed by immersion in 10% Neutral Buffered Formalin, bisected lengthwise, processed, sectioned, and stained for hematoxylin and eosin (HE) by routine methods. An anti-murine haptoglobin antibody (Proteintech, Chicago, IL) was used to stain for haptoglobin in murine hearts. Murine histological data was read in a blinded fashion.
Statistical analysis is detailed in the on line data supplement.
After organ implantation, graft inflammation occurs within hours in mice and humans 13. To establish a model to assess early intra-graft inflammation after cardiac transplantation, we subjected murine BALB/c hearts to increasing cold storage and then implanted these organs into C57BL/6 murine recipients. This model differs by a full major histocompatibility complex (MHC) mismatch between the donor and the recipient.
When we measured inflammatory cytokines and chemokines within the transplant 24h after implantation, we found that regardless of the exposure of the organ to cold storage, the inflammatory mediators MCP-1, MIP-2, and IL-6 were increased two to ten-fold over those in non-transplanted hearts (Online Figure IA), indicating an early phase of intra-graft inflammation after organ implantation.
Characterization of the intra-graft levels of cytokines and chemokines demonstrated similar levels between syngeneic and allogeneic transplants at 24h post transplantation (Figure 1B). At this time point the T cell response to transplantation was not yet evident within the transplant (Online Figure II). Our observations also held true when the hearts were exposed to 4h of cold storage prior to implantation (Online Figure II), indicating that the early inflammatory phase within the graft after cardiac transplantation is antigen-independent.
To determine the kinetics of the early, antigen-independent (i.e., syngeneic transplants), inflammatory phase for cardiac transplants, we analyzed the levels of inflammatory mediators including IL-6, MCP-1 and MIP-2. These three mediators peaked within the graft between 6h and 48h after transplantation and then subsided by 72h post transplantation (Online Figure IC). In parallel, the levels of inflammatory mediators within the recipient’s native heart (denoted as native heart), which was exposed to peri-operative stressors, but not to the effects of organ procurement and implantation, were comparable to those in the donor heart prior to implantation (Online Figure IB + C), indicating that the early antigen-independent phase of graft inflammation is mainly driven by the effects of organ procurement and implantation.
To characterize intra-graft inflammation further after cardiac transplantation, we adapted an in vi assay in which dendritic cells (DCs) are cultured with tissue lysates 11. We then assessed the production of the proinflammatory cytokine IL-6 by DCs during the culture to determine the degree of DC activation induced by the lysates. We measured IL-6 in the assay, as there are prior reports that IL-6 contributes to the tempo of cardiac allograft rejection in mice 14–16. The IL-6 response by DCs was increased after the culture with the lysates obtained from cardiac transplants at 24h and 48 post transplantation but not with the lysates obtained from non-transplanted hearts or recipient native hearts (Online Figure IIIA). These findings indicate that cardiac transplantation increases the concentration of inflammatory mediators within the graft.
To evaluate the nature of the inflammatory mediators within cardiac grafts, we added either RNase, DNase or pronase separately to lysates from syngeneic cardiac grafts at 24h post transplantation in order to degrade RNA, DNA, and proteins, respectively. Our findings suggest that proteins are the major contributors to the early antigen-independent phase of inflammation after cardiac transplantation as digestion of proteins abrogated the induced IL-6 response of DCs (Online Figure IIIB). In contrast, digestion of RNA or DNA did not reduce the IL-6 response. (Concentrations of RNase and DNase employed in our experiments effectively digested known RNA and DNA-based DC activators, and the concentration of pronase employed in the assay did not impair DCs to respond to LPS, Online Figures III C–D).
Based on the results from our in vitro DC assay, we identified proteins that are differentially regulated after cardiac transplantation by comparing non-transplanted cardiac tissue to tissue from syngeneic cardiac transplants at various time points after transplantation via mass spectrometry. By 6h post transplantation, only one protein, NADP transhydrogenase was differentially (down) regulated in cardiac grafts, whereas by 24 to 72h after transplantation, several proteins were differentially down and up regulated within the transplant (Online Figure IV A–B). The entire proteomic screen identified a total of 1318 up and down regulated proteins (Online Table I). Among these proteins, bioinformatic analyses indicated enrichment for biological and molecular processes including immune response, phagocytosis, and cytoskeletal re-organization (Online Tables II–III) and cellular components including mitochondrial components and myofilaments (Online Table IV), reflective of the complex alterations that occur within the graft after cardiac transplantation.
Our proteomic screen identified 22 proteins that were significantly upregulated in the graft after cardiac transplantation (Online Table V). The only significantly upregulated proteins at both 24h and 72h after transplantation were calgranulin A, haptoglobin, chitinase-like 3 protein, and vimentin (Online Table V). We had previously employed an unbiased proteomic screen of skin transplants and noted that haptoglobin, a protein with anti-oxidant and immune modulatory properties 17–19, was upregulated in the graft after skin transplantation 11. Our earlier findings indicated that haptoglobin increases the tempo of minor mismatched skin graft rejection 11. Given this result, we explored the role of haptoglobin in cardiac transplantation.
We confirmed the upregulation of intra-graft haptoglobin in the syngeneic heart transplant model via an ELISA test (Figure 1A), and found that haptoglobin was highly induced in the serum 3h after cardiac transplantation in mice (Figure 1A). At 24h after transplantation, haptoglobin levels were six-fold higher in the cardiac transplant than in the native heart (Figure 1B), indicating that haptoglobin preferentially enters sites of inflammation.
Although intra-graft haptoglobin levels were similar between allografts and syngeneic grafts 1 day after transplantation, by 3 days post transplantation, haptoglobin levels were two-fold higher in allografts than in syngeneic grafts (Figure 1C), although by day 7, post transplantation levels in allografts declined to the levels of syngeneic grafts (Figure 1C). We also observed that haptoglobin levels increased in the liver of murine heart transplant recipients and that recipients of allografts exhibited higher levels of haptoglobin in the liver than recipients of syngeneic grafts (Figure 1D). These results indicate that intra-graft haptoglobin is highly induced in both allografts and syngeneic grafts but peaks at a higher level in allografts during the first week after transplantation. Furthermore, concentrations of haptoglobin in the liver, a known site of haptoglobin production, increase after cardiac transplantation.
Given these findings, we examined the effect of haptoglobin on the tempo of acute allograft rejection after cardiac transplantation. For this purpose, we transplanted wild type (WT) or Hp−/− recipient mice with a WT BALB/c cardiac allograft with or without peri-operative CTLA4 Ig. CTLA4 Ig is related to belatacept, a drug that is used currently in clinical organ transplantation and which inhibits costimulation between T cells and antigen presenting cells, thus increasing transplant survival 20. Without CTLA4 Ig treatment, Hp−/− recipients exhibited a 3 day significant extension of allograft survival compared to WT recipients (Figure 2A). With CTLA4 Ig treatment, this difference was dramatically increased as Hp−/− recipients displayed significantly longer cardiac allograft survival (median survival time > 100 days) than WT recipients (median survival <21 days) (Figure 2B), indicating that recipient haptoglobin impairs the ability of CTLA4 Ig to induce prolonged allograft survival. Hp expression in the donor allograft did not impact the tempo of acute allograft rejection in our model (Online Figure V). At 21 days after transplantation, the majority (8/9) of Hp−/− recipients treated with CTLA4 Ig exhibited beating heart grafts and thus allograft survival, whereas all similarly treated WT recipients had rejected their allografts (Online Figure VI and online Videos I and II). At this time point, Hp−/− recipients treated with CTLA4 Ig also exhibited reduced histological evidence of graft necrosis (Figure 2C). Immune histochemical staining for haptoglobin within the allografts of WT recipients showed that haptoglobin co stained with some cardiomyocytes, macrophages and exhibited scant perivascular staining (Figure 2D).
We previously demonstrated that haptoglobin activates bone marrow derived DCs in vitro to induce the production of proinflammatory cytokines, specifically the production of IL-6 11. Therefore, we evaluated how recipient haptoglobin impacted the intra-graft inflammatory milieu after cardiac transplantation and peri-operative treatment with CTLA4 Ig. During the first 3 weeks after transplantation, we noted lower intra-graft levels of IL-6 and MIP-2 in Hp−/− than in WT recipients treated with CTLA4 Ig (Figure 3A–B). In addition, graft levels of the immunosuppressive cytokine IL-10 were higher in the allografts of Hp−/− recipients treated with CTLA4 Ig than in similarly treated WT recipients (Figure 3C).
This elevation of IL-10 in the grafts of Hp−/− recipients was surprising given that prior work found that haptoglobin-hemoglobin complexes are internalized in macrophages to induce heme-oxygenase-1 (HO-1) and IL-10 production 19, 21. We determined that intra-graft levels of HO-1 were not altered by haptoglobin after cardiac transplantation and treatment with CTLA4 Ig (Online Figure VII). Furthermore, Hp−/− recipients of WT and thus haptoglobin sufficient allografts treated with CTLA4 Ig exhibited abrogated intra-graft levels of haptoglobin (Figure 3D), further supporting the idea that haptoglobin, which impairs the ability of CTLA4 Ig to enhance graft survival (Figure 2B), originates from the recipient. Finally, we found that haptoglobin levels were significantly increased in the livers of WT cardiac transplant recipients treated with CTLA4 Ig but not in non-transplanted recipients (Online Figure VIII), indicating that the liver is a contributing source to haptoglobin after cardiac transplantation.
We further characterized how haptoglobin altered the inflammatory response of specific cells within heart allografts by enriching immune cells, fibroblasts and endothelial cells (ECs) from allografts at day 21 post transplantation from WT and Hp−/− recipients treated with CTLA4 Ig. The total numbers of these different cell types were obtained from the allografts of WT or Hp−/− recipients and the cells were then cultured ex vivo. As IL-6 and MIP-2 were reduced in the allografts of Hp−/− recipients (Figure 3A–B), we measured IL-6 and MIP2 in the supernatants after 12h of cell culture. Compared to cells obtained from non-transplanted WT BALB/c hearts, immune cells enriched from BALB/c allografts from WT recipients produced significantly elevated (i.e., 5–10 fold) levels of both IL-6 and MIP-2 compared to immune cells enriched from Hp−/− recipients (Figure 3E–F). We also noted that there were significantly more immune cells (i.e., four fold higher) in the allografts of WT recipients as compared to Hp−/− recipients at this time point after transplantation (Online Figure IX). These data suggest that infiltrating immune cells are a likely source of proinflammatory cytokines within the allograft and that haptoglobin enhances this response by increasing the recruitment of immune cells into the allografts.
Fibroblasts and ECs from allografts did not produce IL-6 and MIP-2 above levels from cells obtained from non-transplanted hearts (Online Figure X). Immune cells enriched from allografts also produced more IL-10 after culture than cells obtained from non-transplanted hearts (Online Figure XI). Furthermore, immune cells enriched from the allografts of Hp−/− recipients exhibited somewhat lower levels (i.e., 50% reduced) of IL-10 than immune cells enriched from the allografts of WT recipients (Online Figure XI). ECs, but not fibroblasts, enriched from allografts had higher levels of IL-10 production than cells obtained from non-transplanted hearts, although we did not observe significant differences between ECs enriched from the allografts of Hp−/− recipients and WT recipients. These results suggest that in addition to immune cells, ECs contribute to the graft levels of IL-10 after cardiac transplantation and treatment with CTLA4 Ig.
As DCs have been shown to be critical for activating CD4+ T cells to induce acute cardiac allograft rejection 22, 23, we next assessed if there were alterations in the activation phenotype and accumulation of intra-graft DCs (defined as CD11c+ MHC class II+ cells) in the cardiac allografts of Hp−/− compared to WT recipients treated with CTLA4 Ig at day +7 post transplantation. We found that the upregulation of the costimulatory molecule CD80 was reduced three-fold on the surface of intra-graft DCs in the Hp−/− recipients as compared to WT recipients treated with CTLA4 Ig (Figure 4A–B). There was also a lower number of DCs within the allografts of Hp−/− recipients treated with CTLA4 Ig than in WT recipients treated with CTLA4 Ig (Figure 4C). Similar to what we observed at day +21 post transplantation, we found a two fold lower total number of immune cells within the allografts of Hp−/− recipients than in WT counterparts at day +7 post transplantation (Figure 4D). Analysis at day +7 post transplantation also found reduced numbers of neutrophils and T cells in the allografts of Hp−/− recipients as compared to their WT counterparts (Online Figure XII).
We previously demonstrated that haptoglobin activates bone marrow derived DCs to produce IL-6 in vitro via MyD88 11. Furthermore, we and others have shown that MyD88 is critical for IL-6 production by DCs and macrophages in response to microbial stimulation 24. We therefore examined the impact of MyD88 expression within the allograft on the development of graft inflammation. For this purpose, WT or MyD88−/− C57BL/6 hearts were transplanted into WT BALB/c recipients that were treated with CTLA4 Ig. At day 14 after transplantation, the allografts were obtained and graft levels of IL-6, MIP-2 and IL-10 were measured. We found that the graft levels IL-6 and MIP-2 were significantly reduced three to five fold in MyD88−/− allografts as compared to WT allografts (Figure 5A–B). IL-10 levels were significantly elevated three fold in the MyD88−/− allografts as compared to WT allografts (Figure 5C). These results phenocopy the graft inflammatory alterations observed in Hp−/− recipients of WT allografts (Figure 3A–C).
To determine the cellular targets of haptoglobin within resident heart cells, we isolated immune cells, fibroblasts and ECs from WT and MyD88−/− non-transplanted hearts and stimulated the cells in vitro with haptoglobin. We found that only immune cells enriched from hearts produced IL-6 in response to haptoglobin and this response was mostly abrogated in immune cells obtained from MyD88−/− hearts (Figure 5D). (Immune cells, fibroblasts and ECs produced IL-6 in response to in vitro stimulation with LPS, indicating that these populations of cells enriched from hearts were capable of producing IL-6, Online Figure XIII). Overall, these data show that donor expression of MyD88 enhances allograft inflammation after cardiac transplantation and treatment with CTLA4 Ig and that immune cells are the likely targets within the heart that respond to haptoglobin in a MyD88-dependent fashion.
As prior in vitro studies have indicated that haptoglobin may impact T cell function to nominal antigens or non-specific stimulation 18, 25, we assessed whether haptoglobin alters intrinsic T cell responses to allostimulation in a mixed lymphocyte culture. We found that purified WT and splenic Hp−/− polyclonal T cells stimulated in vitro with irradiated allogeneic spleen cells exhibited similar production of the Th1 cytokine, IFN-γ, and similar IL-2 levels as WT T cells (Figure 6A–B). The levels of proliferation measured in the mixed lymphocyte culture were also similar between WT and Hp−/− T cells (Figure 6C). However, when we assessed anti-donor T cell responses in T cells obtained from the spleens of WT and Hp−/− recipients treated with CTLA4 Ig and transplanted with cardiac allografts, Hp−/− recipients exhibited a 2–3 fold reduction in splenic anti-donor T cell IFN-γ and IL-2 responses (Figure 6D–E). Recipient haptoglobin had no impact on the numbers of splenic CD4+ FoxP3+ regulatory T cells either before, or after transplantation and treatment with CTLA4 Ig (Figure 6F). Thus, haptoglobin appears to amplify intra-graft inflammation (Figure 3A–B) and enhance anti-donor Th1 T cell alloimmunity to impair the graft prolonging effects of costimulatory blockade.
To provide an initial determination of the presence of haptoglobin in human heart transplants, we performed a case-controlled study in which 10 endomyocardial biopsies that had no histological evidence of rejection on routine endomyocardial biopsy obtained at various time points post transplantation were compared to 9 biopsies that exhibited moderate histological evidence of rejection. We found that haptoglobin staining was present in seven out of nine human cardiac transplant specimens that exhibited evidence of moderate allograft rejection but only in two of ten cardiac transplant specimens that were free of acute rejection (p = 0.02, Fisher’s exact test) (Figure 7A–B). Examination of the seven biopsies that exhibited moderate rejection and haptoglobin staining suggested that haptoglobin was present within some cardiomyocytes (Online Figure XIV) and within some scattered endothelial cells (Figure 7B). In this small sample size we did not observe haptoglobin staining with immune cells. (The clinical characteristics of patients who had moderate rejection are compared with those of patients with no rejection in Online Table VI). These findings suggest that intra-graft haptoglobin is also associated with acute allograft rejection in humans.
Our study identified haptoglobin as a novel enhancer of cardiac graft inflammation that prevents the allograft prolonging effects of CTLA4 Ig. Although haptoglobin has important physiological functions, such as an anti-oxidant in the setting of hemolysis 17, our study has found that it also enhances the levels of inflammatory cytokines, such as IL-6, which have been shown to impair cardiac allograft survival 14.
In contrast to a skin transplant model 11, our current study found that only recipient, but not donor, expression of haptoglobin accelerates the tempo of acute cardiac allograft rejection (Figure 2 and Online Figure V). In previous studies, the skin has been shown to be a source of haptoglobin 26. Unlike the skin, however, which undergoes neovascularization after transplantation, cardiac transplants are immediately vascularized and subjected to a rapid ischemia reperfusion injury. Our study suggests that the liver is a source of haptoglobin in our cardiac transplantation model. Haptoglobin production from the liver may be induced by the initial inflammation and hypoxia that occurs after graft implantation. In turn, haptoglobin enters the graft within 24h after transplantation via the vasculature and amplifies inflammation.
Prior work has indicated that haptoglobin enhances the clearance of Salmonella and increases inflammation in delayed-type hypersensitivity reactions of the skin 18. There is also evidence that haptoglobin impacts intrinsic T cell function in particular Th1 responses, at least to nominal antigens or non-specific stimulation in vitro 18,25. Furthermore, haptoglobin increases the cellularity of secondary lymphoid organs in mice 18, which could impact immune responses including the tempo of cardiac allograft rejection. Other studies have found that haptoglobin dampens inflammation induced by innate immune activation via lipopolysaccharide 27. In addition, haptoglobin reduces oxidative stress within the myocardium and reduces the rate of late ventricular rupture after acute myocardial infarction induced by coronary ligation in mice 28. Moreover, when haptoglobin binds to hemoglobin, this complex induces the production of the immune suppressive cytokine IL-10 by macrophages 19. Given that innate immune activation is implicated in IRI 29, we expected haptoglobin to reduce inflammation after cardiac transplantation. However, our study found that haptoglobin enhanced graft inflammation and impaired the graft prolonging properties of CTLA4 Ig, although it did not impact intrinsic T cell function during in vitro stimulation with allogeneic spleen cells (Figure 6A–C). Together, our current study and prior reports imply that haptoglobin exhibits either anti or proinflammatory effects, and the dominant phenotype may depend on the disease context or the model employed.
Our current study indicates that in the context of organ transplantation haptoglobin enhances the recruitment of immune cells into the allograft including DCs to amplify graft inflammation. We believe that the initial IRI after cardiac transplantation induces graft inflammation. This leads to the production of haptoglobin in the recipient, which is released into the circulation and subsequently enters the graft to amplify inflammation, possibly through activating intra-graft immune cells, such as DCs. In turn, activated immune cells within the transplant release cytokines and chemokines to enhance immune cell recruitment and promote inflammation. Future studies will be needed to determine how haptoglobin impacts other possible cellular sources of pro and anti-inflammatory molecules with the heart, such as cardiomyoctyes or vascular smooth muscle cells. Regardless of the source of inflammatory mediators, recipient haptoglobin enhances the intra-graft inflammatory milieu and increases anti-donor T cell immunity after cardiac transplantation. Our study suggests that the inflammatory alterations mediated by haptoglobin renders T cells less susceptible to the effects of costimulatory blockade (Online Figure XV).
We previously demonstrated that haptoglobin activates bone marrow derived DCs in vitro via MyD88, a signal adaptor downstream of TLR and IL-1 and 18 receptors, to induce inflammatory responses 11. In the current study, we found that the expression of MyD88 within the allograft enhances graft inflammation (i.e., elevating levels of IL-6 and MIP-2 but impairing IL-10 levels) similarly to the impact recipient haptoglobin has on graft inflammation. We also provide evidence that immune cells enriched from cardiac tissue respond to haptoglobin in vitro via MyD88. Together, these data imply that recipient derived haptoglobin is sensed by immune cells in the graft to enhance alloimmunuty. Future studies will be required to identify the MyD88 sensing cell within the allograft that is activated by haptoglobin in vivo to enhance intra-graft inflammation after cardiac transplantation.
The cellular recognition mechanisms and pathways by which haptoglobin alters the intra-graft inflammatory milieu after vascularized organ transplantation will require further investigations to discern detailed underlying molecular pathways. Moreover, haptoglobin was not the only upregulated protein in the graft after cardiac transplantation. Aside from haptoglobin, calgranulin A, chitinase-like 3 protein and vimentin were also significantly upregulated at both 24h and 72h after transplantation (Online Table V). Prior work has shown that calgranulin A suppresses DC priming and slows the tempo of MHC Class II mismatched cardiac allograft rejection in mice 30. A recent study identified that chitinase like 1 protein, which is similar to the chitinase-like 3 protein, correlates with the degree of graft injury after kidney transplantation 31, although the role of either protein in cardiac transplantation is not known. In addition, there is evidence that vimentin is an important antigen in organ transplantation and antibodies to it are associated with cardiac allograft rejection 32. These other proteins therefore warrant further exploration as they may potentially also contribute to inflammation after heart transplantation.
Aside from the murine data that implicates haptoglobin as an amplifier of inflammation after cardiac transplantation, our study also showed that there is an association between intra-graft haptoglobin expression and acute allograft rejection in archived human biopsies that were obtained at various time points after heart transplantation. Our murine data indicate that without immune suppression, haptoglobin is upregulated within the graft during the first week after transplantation (Figure 1C) and is present in cardiac transplants three weeks post transplantation in recipients treated with CTLA4 Ig (Figure 3D). Human heart transplant biopsies are not routinely available before one week after transplantation due to patient safety issues. We therefore could not determine if haptoglobin is upregulated within the immediate peri-operative period after human heart transplantation. As our small case-controlled study did not determine where haptoglobin localizes within human cardiac transplants, future clinical studies will be required to identify the time course and localization of haptoglobin within human hearts transplants, although it will be challenging to assess if haptoglobin is induced in the perioperative period.
In summary, we show that haptoglobin is an important amplifier of inflammation after cardiac transplantation. Our results from the murine experimental data combined with our human archived samples suggest that haptoglobin enhances allograft inflammation after cardiac transplantation. The protein and its associated pathways may be key to identifying potential candidates for future anti-inflammatory treatments and therapies.
After organ transplantation, inflammation is induced by the innate immune system. However, the mechanisms that induce and maintain graft inflammation remain unclear. We performed a proteomic analysis of murine cardiac transplants and showed that the protein haptoglobin, which is known for its anti-oxidant properties, is upregulated in these grafts. Recipient, but not donor, expression of haptoglobin impairs the ability of immune modulators to induce indefinite cardiac allograft survival in mice. Furthermore, haptoglobin in cardiac allografts enhances the production of the pro-inflammatory cytokine IL-6, the chemokine MIP-2 but impairs the immune suppressive cytokine, IL-10. We also showed that haptoglobin activates resident immune cells in heart tissue via the innate immune adaptor, MyD88. Importantly, the presence of haptoglobin in human cardiac allograft biopsies correlates with acute allograft rejection. Together, these findings indicate a novel and surprising function for haptoglobin in cardiac allografts: activation of innate immunity to impair indefinite allograft survival. Our study reveals that the inflammatory pathway amplified by haptoglobin may be therapeutically targeted to reduced inflammation after organ transplantation and possibly in other sterile inflammatory conditions.
SOURCES OF FUNDING
This study was supported by NIH grants AI101918, AI098108, AG028082 and an Established Investigator Award from the American Heart Association to DRG.