• Graft arterial disease;
  • heart transplant;
  • interferon-γ;
  • tumor necrosis factor


  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References

Graft arterial disease (GAD) remains the leading cause of long-term solid organ allograft failure. Tumor necrosis factor (TNF) promotes multiple aspects of allograft rejection via binding to type 1 (p55) and type 2 (p75) receptors. We used TNF type 1 receptor deficient (TNFR1KO), type 2 receptor deficient (TNFR2KO) and receptor double-deficient (TNFRDKO) mice to assess the relative roles of TNFR in acute rejection and GAD. Heterotopic cardiac transplantation was performed between C57BL/6 (B/6) and Balb/c (B/c) mice (total allomismatches) to assess the effects on graft survival; B/6 and Bm12 mice (class II mismatches) were used to assess the effects on GAD 8 weeks after transplantation. We found that graft survival in the total allomismatch combinations was the same regardless of TNFR status. In class II mismatches, wild-type (WT) combinations showed severe GAD, and GAD was not diminished when WT hearts were transplanted into TNFRDKO hosts. TNFR1KO donors or TNFR2KO donors had GAD comparable to WT donors, however, GAD was significantly diminished in B/6 TNFRDKO donor hearts. We conclude that both p55 and p75 signals on donor vascular wall cells are involved in the development of GAD, and either TNFR is capable of mediating a response that will culminate in GAD.


  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References

Despite advances in immunosuppression, acute rejection and graft arterial disease (GAD) remain significant limitations to successful long-term allograft transplantation (1,2). Several cytokines and adhesion molecules enhance acute rejection. Graft arterial disease is characterized by intimal thickening comprised of smooth muscle cells (SMCs) and extracellular matrix, accumulating under the influence of numerous inflammatory cytokines (3,4); in particular, IFN-γ is necessary (5,6) and sufficient (7) to induce GAD in various transplant models.

The proinflammatory cytokine tumor necrosis factor (TNF)-α derives primarily from macrophages activated by inflammatory stimuli. However, lymphocytes, neutrophils, and dendritic cells can also produce this cytokine (8). TNF-α mediates a wide range of inflammatory and immunological responses including the elaboration of other proinflammatory cytokines (9,10) and chemokines (11). Consequently, this mediator may contribute to a wide array of immune disorders including graft-vs.-host disease (GVHD) and allograft rejection (8,10).

Two homologous receptors, TNF receptor type 1 (TNFR1; p55) and type 2 (TNFR2; p75), which share similar extracellular domains (12,13), mediate the biological activity of TNF-α. TNFR1 expression is ubiquitous, while TNFR2 is found more selectively on cells of hematopoietic lineage (9,10). TNFR1 is the major receptor for soluble TNF-α and mediates many of the proinflammatory activities of TNF-α (9,11). TNFR2 may enhance stable ligation of TNFR1 (14), but TNFR1 signaling does not require TNFR2 activation (12). Nevertheless, TNFR2 plays a critical role in some experimental models of disease (15,16).

With regard to solid organ transplantation, the two TNFRs play discrete roles in mediating rejection of murine corneal allografts (17). Using TNFR1-deficient mice, we previously demonstrated that TNFR1 single deficiency on either donor or host does not affect cardiac allograft survival or the development of GAD (18). In the present report, we further show that TNFR2 single deficiency does not affect acute rejection or GAD. However, while TNFR double deficiency does not diminish acute rejection, lack of both receptors on donor hearts leads to significantly reduced GAD. Notably, this attenuation of GAD occurs in the presence of normal levels of host inflammatory cell IFN-γ production, although diminished IFN-γ expression is seen in the vascular wall cells comprising the GAD lesion. The results suggest a pathway for GAD pathogenesis occurring through TNFR ligation on graft endothelial or medial smooth muscle cells that affects recruitment and activation of host smooth muscle cell precursors.

Materials and Methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References


Inbred male C57BL/6 (B/6, H-2b), BALB/c (B/c, H-2d) and B6.C-H-2bm12 KhEg (Bm12, H-2bm12) mice were obtained from Taconic Farms, Inc. (Germantown, NY). Male TNFR1-deficient (TNFR1KO) mice and TNFR2-deficient (TNFR2KO) mice were obtained from Jackson Laboratory (Bar Harbor, ME). TNFR double-deficient (TNFRDKO) mice, generated as described previously, were generously provided by Immunex Corporation (17). Animals were backcrossed to B/6 mice for at least eight generations. Male wild-type (WT) B/6 mice, WT Bm12 mice, WT B/c mice, TNFR1KO B/6, TNFR2KO B/6, and TNFRDKO B/6 mice were used as donors or recipients. All mice were 8–12 weeks old and weighed 20–25 g. The mice were maintained at the animal facilities, accredited by the American Association of Accreditation of Laboratory Animal Care, of Brigham and Women's Hospital and Harvard Medical School.

Heterotopic cardiac transplantation

Allografts were heterotopically transplanted in an intra-abdominal location using the microsurgical technique previously described (19,20). Briefly, this technique involves anastomosing the end of the donor aorta to the side of the recipient abdominal aorta, after which the donor pulmonary artery is connected to the inferior vena cava of the recipient mice to return myocardial blood flow. Ischemic time averaged 20 min, and the overall success rate was greater than 90%. Graft function was assessed by daily palpation, and graft failure was defined as the absence of detectable beating.

Graft harvesting

Allografts of total allomismatched combinations (B/6 and B/c) were harvested at the time of graft failure. Allografts of class II mismatched combinations (B/6 and Bm12) normally function for at least 12 weeks, and were harvested at 2, 4 or 8 weeks after transplantation. This class II mismatch combination is used because it reliably develops GAD lesions within 8 weeks of transplantation without immunosuppression (5); the histology and characteristics of GAD in this model are the same as for MHC minor mismatches, MHC class I mismatches, or total allomismatches, and recapitulate those seen in human GAD (6). Harvested allografts were transversely sectioned into three parts. The basal section was fixed with 10% phosphate buffered formalin and embedded in paraffin for morphologic analysis. The mid-portion was frozen in OCT compound (Tissue Tek; Miles, Elkhart, IN) and stored at − 80 °C for immunohistochemistry. The apical portion was used to extract RNA for RNase protection assay (RPA) (21).

Histological evaluation

Grafts were analyzed in a blinded fashion using multiple H&E and elastic stained sections, and the magnitudes of parenchymal rejection (PR) and GAD were scored. Parenchymal rejection was graded using a scale modified from the International Society for Heart and Lung Transplantation (0, no rejection; 1, focal mononuclear cell infiltration without necrosis; 2, focal mononuclear cell infiltration with necrosis; 3, multifocal mononuclear cell infiltration with necrosis; 4, widespread infiltrates with hemorrhage and/or vasculitis). Graft arterial disease scores (0, vascular stenosis < 10%; 1, 10–25%; 2, 25–50%; 3, 50–75%; 4, > 75%) of all epicardial and intramyocardial arteries and arterioles in each graft were averaged, as described in detail previously (5,22).


Five-µm-thick frozen sections were fixed in 4% paraformaldehyde dissolved in PBS for 8 min at 4 °C. To stain CD4, CD8, CD11b (monoclonal antibodies from PharMingen, San Diego, CA), VCAM-1, TNFR1, or TNFR2 (monoclonal antibodies from Santa Cruz Biotechnology, Inc., Santa Cruz, CA), sections were incubated with unlabeled primary antibodies (each at 1–10 µg/mL) for 90 min at RT, washed in PBS, followed by biotinylated secondary antibodies (PharMingen) at 5 µg/mL for 45 min at RT. To stain for IFN-γ, sections were incubated with unlabeled primary antibody [antimouse IFN-γ (XMG1.2), PharMingen) at 5 µg/mL] for 8 h at 4 °C, washed in PBS, followed by Histofine Simple Stain Mouse MAX-PO conjugate (Nichirei, Tokyo, Japan) for 30 min at RT. After washing in PBS, the sections were incubated and visualized by incubating with an aminoethylcarbazole (AEC) complex (Nichirei). To stain for α-smooth muscle actin (SMA), sections were incubated with fluorescein isothiocyanate (FITC)-labeled primary antibody (anti α-SMA (1A4), Sigma Chemical Co., St. Louis (MO) at 5 µg/mL) for 8 h at 4 °C. After washing in PBS, the sections were observed using a fluorescent microscope. Sections were counterstained with hematoxylin solution (Sigma) (23,24). CD4-, CD8-, or CD11b-stained graft infiltrating cells were enumerated as the mean numbers of positive cells per ten 10×-power microscopic fields. Immunohistochemical analyses of vascular wall cell expression of VCAM-1 or TNFRs were performed by blinded, independent observers (J. S., S. E. C, and S. B.) using qualitative scoring (0, absent; 1, weak, focal; 2, weak, diffuse; 3, strong, focal; 4, strong, diffuse). Scores uniformly fell within one grade of each other and were averaged.

RNase protection assay

The harvested allografts were homogenized in Trizol reagent (Life Technologies, Grand Island, NY) and frozen at −80 °C. Cytokine mRNA expression was measured by RNase protection assay (RPA) using mCK1, mCK2b, mCK3b, and mCK5 template (Riboquant kit, PharMingen, San Diego, CA). Levels of mRNA expression were quantified and normalized to GAPDH mRNA using densitometry (21,23,25).


Single-cell splenocyte suspensions were prepared by dissociating tissue with frosted glass slides. For primary MLR, a total of 5 × 105 naive responder cells were cultured with an equal number of mitomycin-C treated stimulator cells in 96 well plates in C/10 media (22). For primed MLR, splenocytes from recipients previously transplanted with class II mismatch allografts for 8 weeks were used as responder cells and were cultured with an equal number of mitomycin-C treated stimulator cells. Proliferation was assessed on days 2 through 4 by incubating cultures with 1 µCi/well of [3H] thymidine in C/10 medium for 6 h. Incorporated radioactivity was measured with a liquid scintillation counter (21,22).

IFN-γ elisa

Sandwich ELISA was performed using paired antibodies (purified and biotinylated anti-IFN-γ mAbs (PharMingen), and streptavidin-HRP) according to the manufacturer's instructions. Fifty µL of purified anti-IFN-γ antibody (1 µg/mL) was coated overnight on 96-well microtiter plates (Becton Dickinson, Mountain View, CA) at 4 °C. After blocking the plates with 2% BSA in PBS at RT for 2 h, 50 µL of the sample supernatant from MLR culture was added followed by incubation for 6 h at 4 °C and washing. Plates were then incubated with 50 µL of biotinylated anti-IFN-γ antibody (1 µg/mL) for 45 min at RT, followed by peroxidase-labeled streptavidin for 45 min at RT, and then 2,2′−azinobis (3-ethylbenzthiazoline-6-sulfonic acid) (Sigma) for colorimetric reaction. Absorbance data were collected using a microplate reader Emax (PharMingen) (21), and cytokine content was measured by comparing to standard curves run in parallel.

Flow cytometry of graft infiltrating cells

Graft infiltrating cells were collected from B/6 WT or B/6 TNFRDKO donor allografts transplanted into Bm12 WT recipients at 2 weeks after transplant and used for flow cytometry as previously described (22,23,25). Extracted cells were stimulated for 4 h at 37 °C with 25 µmol/L ionomycin (Sigma) and 10 ng/mL phorbol myristate acetate (Sigma) in the presence of 10 µg/mL brefeldin A (Sigma) to block cytokine secretion. The cells were permeabilized with a saponin buffer and incubated with CD16/CD32 mAb (PharMingen) to block Fc receptors. For intracellular cytokine staining, biotin-labeled anti-IFN-γ mAb or an isotype-matched control antibody were used. After 30 min of staining, the cells were incubated with allophycocyanin (APC)-conjugated streptavidin for a further 30 min. Surface staining was performed using anti-CD11b-FITC, anti-CD8-PE, and anti-CD4-PerCP. Flow cytometry was performed on a four-color FACScan flow cytometer (Becton Dickinson), using CellQuest software.

Statistical analysis

Comparative analysis of cardiac graft survival was accomplished using the Kaplan-Meier cumulative survival method; survival differences between two groups were determined using the log-rank (Mantel-Cox) test. Comparisons between two groups for PR, GAD, immunohistochemistry, ELISA, MLR, and RPA data were accomplished by one-way ANOVA using StatView 4.5 for Macintosh (Abacus Concepts, Berkeley, CA). Data are expressed as mean ± SE. p < 0.05 was considered statistically significant (21).


  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References

TNFR double deficiency does not prolong allograft survival

To examine the significance of TNFR type 1 and type 2 during cardiac allograft rejection in vivo, we first assessed allograft survival in a total allomismatch combination. B/c WT hearts in B/6 TNFRDKO mice or B/6 TNFRDKO hearts in B/c WT mice showed survival and histologic grades of rejection comparable to B/c WT hearts in B/6 WT hosts or B/6 WT hearts in B/c WT mice (Table 1). Similarly, both TNFR1 and TNFR2 single-deficient graft and host combinations showed graft survival comparable to WT animals (data not shown).

Table 1.  Allograft survival in total allomismatch combination
DonorRecipientNumberGraft survival (POD)
B/c WTB/6 WT96, 6, 7, 7, 7, 9, 9, 9, 13
B/c WTB/6 TNFRDKO77, 7, 10, 10, 10, 12, 16
B/6 WTB/c WT86, 9, 10, 11, 11, 12, 12, 13
B/6 TNFRDKOB/c WT68, 10, 12, 14, 17, 17

TNFRDKO responders show comparable proliferation to WT in MLR

To assess how TNF–TNFR interactions modulate T-cell alloresponses, we examined in vitro MLR using various combinations of WT- and TNFR-deficient stimulators and responders. In total allomismatch combinations, TNFRDKO B/6 and WT B/6 lymphocytes proliferated comparably in response to WT B/c spleen stimulator cells. Similarly, WT B/c lymphocytes proliferated comparably to TNFRDKO B/6 and WT B/6 spleen stimulator cells. In class II mismatch combinations, TNFRDKO and WT cells likewise showed no difference in MLR proliferation. IFN-γ levels from MLR cultures did not differ among these combination groups (data not shown).

TNFR double deficiency on donors diminishes GAD

Because TNF may activate graft arterial endothelial cells, we examined whether either TNFR type 1 and/or type 2 had a role on GAD development in vivo. Severe GAD developed by 8 weeks following transplantation using MHC class II mismatches with either Bm12 allografts in WT B/6 hosts or WT B/6 allografts in Bm12 hosts. Moreover, WT Bm12 allografts implanted in TNFRDKO recipients, and either TNFR1KO or TNFR2KO donor hearts in WT Bm12 mice developed severe GAD comparable to WT combinations. However, TNFRDKO B/6 allografts implanted to WT Bm12 recipients showed significantly less GAD compared with WT B/6 donor hearts. Parenchymal rejection did not differ statistically among these groups at 2, 4, or 8 weeks post transplantation (Figure 1, Table 2, and data not shown).


Figure 1. Representative GAD lesions in allografts 8 weeks after transplantation (elastic tissue stain). Severe GAD develops by 8 weeks in both WT B/6 donor hearts in Bm12 hosts (A), as well as in Bm12 hearts transplanted into WT B/6 hosts (B). MHC class II mismatched TNFRDKO B/6 allografts implanted into WT Bm12 recipients showed significantly less GAD (C) compared to WT B/6 donor hearts. However, the WT Bm12 allografts implanted into TNFRDKO recipients showed severe intimal thickening (D), comparable to WT B/6 hosts. Both TNFR1KO (E) and TNFR2KO (F) donor hearts transplanted into Bm12 hosts showed severe GAD comparable to WT combinations. (Arrows indicate internal elastic lamina; original magnification = 200X).

Download figure to PowerPoint

Table 2.  Graft arterial disease and parenchymal rejection scores in class II mismatch combination
  • Eight weeks after transplant.

  • *

    p < 0.05 vs. wild-type.

  • GAD = graft arterial disease, PR = parenchymal rejection.

B/6 WTBm12 WT92.1 ± 0.62.4 ± 0.2
B/6 TNFRDKOBm12 WT100.9 ± 0.2*2.3 ± 0.1
B/6 TNFR1KOBm12 WT101.8 ± 0.42.4 ± 0.2
B/6 TNFR2KOBm12 WT81.6 ± 0.52.5 ± 0.3
Bm12 WTB/6 WT92.3 ± 0.52.8 ± 0.2
Bm12 WTB/6 TNFRKO92.7 ± 0.33.4 ± 0.2

TNFR double deficiency does not alter graft infiltrating cell populations

Eight weeks after transplantation of WT B/6 donor hearts into WT Bm12 mice, the allograft myocardium exhibited diffusely scattered CD4-positive (277.8 ± 57.0 per 10 10X-powered fields) and CD8-positive (220.2 ± 38.8 per 10 10×-powered fields) cells, with CD11b-positive cells (249.0 ± 39.9 per 10 10×-powered fields) localized primarily in a periarteriolar distribution. TNFRDKO allografts transplanted into WT Bm12 recipients demonstrated CD4-positive (262.4 ± 43.7), CD8-positive (284.6 ± 80.6), or CD11b-positive (273.5 ± 59.2) infiltrates comparable to the WT combinations (n = 6 for each strain combination).

TNFRDKO allografts show less VCAM-1 arterial endothelial expression

To verify that TNFR expression is increased in WT allografts, and to assess the effect of TNFR signaling on VCAM-1 expression, we performed immunohistochemical staining. Native hearts and isografts did not express detectable TNFR1, TNFR2 or VCAM-1. Wild-type class II mismatch allografts 2 weeks after transplant demonstrated enhanced expression of TNFR1 and TNFR2, as well as VCAM-1, even in the absence of marked intimal thickening or significant cell infiltration (data not shown). As anticipated, TNFR1 and TNFR2 immunostaining was absent on TNFRDKO donor hearts; endothelium in these grafts also expressed VCAM-1 only weakly. At 8 weeks in WT allograft combinations, expression of VCAM-1 and both TNF receptors increased markedly on endothelial cells. In comparison, TNFRDKO donor hearts showed significantly reduced VCAM-1 expression on the vascular wall cells relative to WT combinations (Figure 2).


Figure 2. Immunohistochemical analyses of VCAM-1 and TNFR expression in allografts 8 weeks after transplantation. Panels A, B and C show representative immunohistochemical staining results for VCAM-1; the bar graph (C) indicates the semi-quantitative analysis of VCAM-1 expression. Increased expression of VCAM-1 was observed in WT combinations 8 weeks after transplantation (A). TNFRDKO donors (B) showed less VCAM-1 compared to WT donors. However, Bm12 hearts transplanted into TNFRDKO hosts showed VCAM-1 expression comparable to WT hosts (C). Panels D and E indicates semi-quantitative results of TNFR expression on allograft arteries. WT combinations demonstrated that both TNFR1 and TNFR2 are strongly expressed on endothelial cells; WT hearts in TNFRDKO hosts showed comparable TNF receptor expression on the arterial endothelial cells (n = 8 for each strain combination).

Download figure to PowerPoint

IFN-γ mRNA levels are reduced in TNFRDKO donors

Because TNFR-mediated signals may play a pivotal role in cytokine production, we used RPA to examine cytokine mRNA expression in the allografts of the TNFR1KO, TNFR2KO, and TNFRDKO groups 8 weeks after transplantation. IFN-γ mRNA expression decreased significantly in grafts from the TNFRDKO donor hearts compared with the WT, TNFR1KO, and TNFR2KO donors (Figure 3). All other cytokines (IL-1α, IL-1β, IL-2, IL-4, IL-6, IL-10, IL-12 (p40), IL-13, and IL-15) showed comparable expression among all allograft groups. Expression of the mRNA for the chemokines lymphotactin, RANTES, eotaxin, MIP-1, IFN-inducible protein-10, and monocyte chemoattractant protein-1, as well as their chemokine receptors, likewise did not differ between the TNFRDKO group and the WT groups. RNA from single receptor deficiency allografts showed a cytokine expression profile comparable to WT combinations (data not shown).


Figure 3. Allograft IFN-γ mRNA expression by RPA 8 weeks after transplantation. The bar graph (A) indicates quantitative analysis of IFN-γ expression normalized against glyceraldehyde phosphate dehydrogenase (GAPDH) housekeeping gene expression. IFN-γ mRNA expression was significantly reduced in TNFRDKO B/6 donor hearts transplanted in Bm12 hosts compared to WT B/6 donor hearts, TNFR1KO B/6 donor hearts, and TNFR2KO donor hearts transplanted in Bm12 hosts. Representative RPA results (B) also show suppressed IFN-γ mRNA expression in TNFRDKO donor hearts compared to WT donor hearts (n = 6 for each strain combination).

Download figure to PowerPoint

Graft-infiltrating inflammatory cells in WT and TNFRDKO donor hearts show comparable IFN-γ production

Using intracellular cytokine staining, we assessed the fraction of IFN-γ-producing graft-infiltrating lymphocytes in the presence or absence of TNFR on the donor heart grafts. As the total number of graft infiltrating cells in 8-week allografts were not sufficient to yield adequate flow cytometric analyses, intracellular cytokine staining for IFN-γ was performed on 2-week allografts. There was no significant difference in IFN-γ production by graft-infiltrating cells in WT or in TNFR1- and TNFR2-double-deficient allografts (Figure 4).


Figure 4. IFN-γ expression in graft-infiltrating cells by flow cytometry, 2 weeks after transplantation. Representative results for IFN-γ expression in graft-infiltrating cells (n = 2 donor hearts for each experiment; experiment repeated twice with comparable results). Deficiency of both TNFR1 and TNFR2 on donor hearts minimally decreased (not statistically significant) the IFN-γ producing capability of total graft-infiltrating cells recovered 2 weeks after transplantation. TNFRDKO donor hearts (A); WT donor hearts (B). Similar analysis attempted on 8 week allografts yielded insufficient cell numbers for adequate flow cytometric analysis.

Download figure to PowerPoint

IFN-γ expression in GAD intima cells

As shown earlier, absence of TNFR1 and TNFR2 did not affect MLR proliferative responses or IFN-γ production. In addition, TNFR-doubly deficient allografts showed comparable inflammatory cell infiltrates and inflammatory cell IFN-γ production. However, TNFR-doubly deficient allografts exhibited diminished IFN-γ mRNA expression. We therefore examined the pattern of IFN-γ expression in WT allografts 8 weeks after transplantation to identify other potential sources. Notably, IFN-γ expression was observed in spindle-shaped cells in the thickened intima of GAD in WT allografts that colocalize with smooth muscle α-actin positivity; these IFN-γ positive cells are negative for CD4, CD8, and CD11b, indicating that they are not inflammatory cells (Figure 5). As GAD is largely attenuated in TNFRDKO grafts, such IFNγ-positive intimal cells are not present to be stained.


Figure 5. IFN-γ expression in GAD by immunohistochemistry 8 weeks after transplantation. Representative immunohistochemical and immunofluorescence staining in 8 week allografts for IFN-γ smooth muscle α-actin, CD4, CD8, and CD11b expression. IFN-γ positive cells in the thickened intima of GAD lesions were spindle-shaped and co-localized with smooth-muscle α-actin-positive cells. The vast majority of these cells do not co-localize with CD4-, CD8-, or CD11b-positive inflammatory cells.

Download figure to PowerPoint


  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References

TNF-α critically regulates endothelial cell function and the emigration of circulating cells into target tissues. TNF-α stimulation augments many inflammatory functions of endothelial cells including the expression of adhesion molecules, cytokines, chemokines, and growth factors via activation of transcription factor NFκB (26,27). Two receptors, TNFR1 and TNFR2, regulate these biological activities of TNF-α (12,13). Studies in TNFR-deficient mice have shown that interruption of TNF signaling prevents selected pathological conditions. Thus, single deficiency of TNFR1 reduces some diseases such as toxoplasmosis (28) and hepatitis (29), while TNFR2 single deficiency decreases TNF-induced hypersensitivity (15) and cerebral malaria (16). Absence of both TNFR1 and TNFR2, but not either receptor alone, prevents ischemia-induced myocyte apoptosis in myocardial infarction (30), and some kinds of inflammation (e.g. listeria monocytogenes infection and endotoxic shock) (12). Conversely, TNFR deficiency promotes the progression of other diseases; for example, absence of TNFR1 accelerates atherosclerosis (31). Thus, TNFR signaling has variable effects depending on the underlying nature of the lesions.

Organ transplantation elevates serum levels of TNF-α and soluble TNFR (32,33). Moreover, administration of anti-TNF-α antibody delays graft failure in transplanted rat hearts (34). Thus, TNF-α and its receptors play a critical role in acute allograft rejection. However, there are few reports regarding the role of allograft TNFR-deficiency mediated signals in the development of GAD. TNFR1 single deficiency prolongs graft survival, but TNFR2 deficiency did not affect graft survival in either corneal (17) or bone marrow transplantation (35). In our previous work, TNFR1 deficiency did not alter graft survival or GAD (18).

This study documented increases in both TNFR1 and TNFR2 expression on the vascular wall cells of rejected cardiac allografts of WT combinations. However, TNFR status did not influence graft survival in total allomismatch combinations. Our results differ from these showing prolonged graft survival after anti-TNF-α antibody administration (34), and are attributable to TNF blockade on only donor or host individually when congenitally deficient mice were used.

Graft arterial disease declined significantly in allografted TNFRDKO donor hearts in contrast with WT or single-receptor-deficient donors. Several papers have shown involvement of both TNFRs in TNF-mediated cytotoxicity (36), cell activation (37), and signal transduction (38) of endothelial cells. We have shown previously that GAD development in this WT combination is modified not only by T-cell activation but also by selected secreted cytokines (5,6,18,22,23). We demonstrate here that endothelial VCAM-1 expression is diminished on transplanted TNFRDKO donor hearts, an observation consistent with prior work showing a relationship between TNFR, endothelial cell activation, and VCAM-1 expression (39,40). TNFR signaling contributes to endothelial cell adhesion by virtue of its regulatory effects on NF-κB, a transcription factor involved in VCAM-1 and other adhesion molecule gene regulation; we previously demonstrated that VCAM-1 is involved in the development of GAD (19). TNFR activation in vascular endothelial cells also promotes secondary mediator production. Paleolog and colleagues reported that TNFR engagement enhances expression of tissue factor, interleukin-8, and granulocyte-macrophage colony-stimulating factor, and that antibodies blocking receptor activation decrease production of these mediators (41).

Both p55 and p75 signals contribute to the development of GAD; we show here that either TNFR on donor vascular wall cells can mediate a response that will culminate in GAD. In addition, we determined that TNFRDKO donor hearts had diminished intragraft IFN-γ mRNA expression compared with WT combinations, correlating with diminished GAD. This finding agrees with our previous work demonstrating the necessity for IFN-γ in the development of GAD (5).

However, although TNFRDKO transplanted hearts have less IFN-γ by RPA, the number of graft infiltrating cells is the same as WT grafts and inflammatory cell IFN-γ production both in vivo and in vitro is likewise comparable to WT combinations. Consequently, we conclude that the absence of graft TNFR affects IFN-γ production by another cell type. In this study, we clearly demonstrated that IFN-γ is produced in GAD lesions in WT allografts; it is produced by spindle-shaped intimal cells of GAD that colocalize with smooth-muscle α-actin positivity, and is not attributable to infiltrating host inflammatory cells. In light of this finding, the recent demonstration that vascular SMC can produce IFN-γ is particularly intriguing (42). In that work, IL-18 and TNF stimulated SMC IFN-γ production; TNF and IL-12 increased the expression of IL-18R. Thus, the absence of TNFR1 and TNFR2 would conceivably lead to diminished recruitment and activation of the intimal SMC precursors of GAD (43). It is possible that lack of TNF-mediated signaling in graft vessel wall cells will directly cause diminished medial SMC IFN-γ, which in turn alters the local environment to reduce intimal SMC recruitment. Alternatively, the absence of TNF signaling in donor vascular wall cells may change the chemokine or adhesion molecule expression necessary to recruit the SMC precursors. Fewer recruited SMC leads to reduced GAD and secondarily to the diminished IFN-γ signal seen in the TNFRDKO grafts. This conclusion is supported by the observed change in VCAM and TNF receptor expression on endothelial cells, although other molecules may be involved.

Previous data indicated that IFN-γ is necessary (5,6) and sufficient (7) to induce GAD. The data presented here also suggests an important component in the overall development of GAD that is regulated via TNFR ligation. Because absence of TNF receptors on host cells has no effect on GAD, we conclude that TNF signaling on host inflammatory cells or smooth muscle cell precursors per se is not critical to drive the development of the lesions. Rather, we postulate that the absence of TNF signaling in donor vascular wall cells changes the ability of allograft vessels to recruit the SMC precursors that form the intimal lesion. However, it is not clear at this point whether the critical role of IFN-γ is to synergize in some manner with TNF to drive GAD development, or whether autocrine production of IFN-γ by intimal SMC recruited via TNF-dependent pathways is necessary to perpetuate the process.

Although the definitive mechanism for suppression of GAD in TNFRDKO grafts is not yet fully elucidated, the observation that TNFR expression on donor vascular wall cells mediates GAD suggests new potential therapeutic interventions involving TNFR blockade.


  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References

This study was supported by National Institutes of Health Grants RO1 HL 43364 (to P. L and R. N. M.). We thank Ms. Eugenia Shvartz, Ms. Elissa Simon-Morrissey, and Ms. Karen E. Williams for their skillful assistance. Jun-ichi Suzuki is a recipient of a Banyu Fellowship Award in Lipid Metabolism and Atherosclerosis.


  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References
  • 1
    Hosenpud JD, Bennett LE, Keck BM, Boucek MM, Novick RJ. The Registry of the International Society for Heart and Lung Transplantation: seventeenth official report. J Heart Lung Transplant 2000; 19: 909931.
  • 2
    Johnson DE, Alderman EL, Schroeder JS et al. Transplant coronary artery disease: histopathologic correlations with angiographic morphology. J Am Coll Cardiol, 1991; 17: 449457.
  • 3
    Suzuki J, Isobe M, Aikawa M et al. Nonmuscle and smooth muscle myosin heavy chain expression in rejected cardiac allografts. A study in rat and monkey models. Circulation 1996; 94: 11181124.
  • 4
    Hosenpud JD. Immune mechanisms of cardiac allograft vasculopathy: an update. Transpl Immunol 1993; 1: 237249.
  • 5
    Nagano H, Mitchell RN, Taylor MK, Hasegawa S, Tilney NL, Libby P. Interferon-gamma deficiency prevents coronary arteriosclerosis but not myocardial rejection in transplanted mouse hearts. J Clin Invest 1997; 100: 550557.
  • 6
    Nagano H, Libby P, Taylor MK et al. Coronary arteriosclerosis after T-cell-mediated injury in transplanted mouse hearts: role of interferon-gamma. Am J Pathol 1998; 152: 11871197.
  • 7
    Tellides G, Tereb DA, Kirkiles-Smith NC et al. Interferon-gamma elicits arteriosclerosis in the absence of leukocytes. Nature 2000; 403: 207211.
  • 8
    Eigler A, Sinha B, Hartmann G, Endres S. Taming TNF: strategies to restrain this proinflammatory cytokine. Immunol Today 1997; 18: 487492.
  • 9
    Kondo S, Sauder DN. Tumor necrosis factor (TNF) receptor type 1 (p55) is a main mediator for TNF-alpha-induced skin inflammation. Eur J Immunol 1997; 27: 17131718.
  • 10
    Tracey KJ, Cerami A. Tumor necrosis factor: a pleiotropic cytokine and therapeutic target. Annu Rev Med 1994; 45: 491503.
  • 11
    Luster AD. Chemokines – chemotactic cytokines that mediate inflammation. N Engl J Med 1998; 338: 436445.
  • 12
    Peschon JJ, Torrance DS, Stocking KL et al. TNF receptor-deficient mice reveal divergent roles for p55 and p75 in several models of inflammation. J Immunol 1998; 160: 943952.
  • 13
    Bazzoni F, Beutler B. The tumor necrosis factor ligand and receptor families. N Engl J Med 1996; 334: 17171725.
  • 14
    Tartaglia LA, Pennica D, Goeddel DV. Ligand passing: the 75-kDa tumor necrosis factor (TNF) receptor recruits TNF for signaling by the 55-kDa TNF receptor. J Biol Chem 1993; 268: 1854218548.
  • 15
    Erickson SL, De Sauvage FJ, Kikly K et al. Decreased sensitivity to tumour-necrosis factor but normal T-cell development in TNF receptor-2-deficient mice. Nature 1994; 372: 560563.
  • 16
    Lucas R, Juillard P, Decoster E et al. Crucial role of tumor necrosis factor (TNF) receptor 2 and membrane-bound TNF in experimental cerebral malaria. Eur J Immunol 1997; 27: 17191725.
  • 17
    Yamada J, Streilein JW, Dana MR. Role of tumor necrosis factor receptors TNFR-I (P55) and TNFR-II (P75) in corneal transplantation. Transplantation 1999; 68: 944949.
  • 18
    Nagano H, Tilney NL, Stinn JL et al. Deficiencies of IL-4 or TNF-alpha receptor-1 do not diminish graft arteriosclerosis in cardiac allografts. Transplant Proc 1999; 31: 152.
  • 19
    Suzuki J, Isobe M, Morishita R et al. Prevention of graft coronary arteriosclerosis by antisense cdk2 kinase oligonucleotide. Nat Med 1997; 3: 900903.
  • 20
    Ono K, Lindsey ES. Improved technique of heart transplantation in rats. J Thorac Cardiovasc Surg 1969; 57: 225229.
  • 21
    Shimizu K, Schonbeck U, Mach F, Libby P, Mitchell RN. Host CD40 ligand deficiency induces long-term allograft survival and donor-specific tolerance in mouse cardiac transplantation but does not prevent graft arteriosclerosis. J Immunol 2000; 165: 35063518.
  • 22
    Furukawa Y, Becker G, Stinn JL, Shimizu K, Libby P, Mitchell RN. Interleukin-10 (IL-10) augments allograft arterial disease: paradoxical effects of IL-10 in vivo. Am J Pathol 1999; 155: 19291939.
  • 23
    Furukawa Y, Mandelbrot DA, Libby P, Sharpe AH, Mitchell RN. Association of B7–1 co-stimulation with the development of graft arterial disease. Studies using mice lacking B7–1, B7–2, or B7–1/B7–2. Am J Pathol 2000; 157: 473484.
  • 24
    Hasegawa S, Becker G, Nagano H, Libby P, Mitchell RN. Pattern of graft- and host-specific MHC class II expression in long- term murine cardiac allografts: origin of inflammatory and vascular wall cells. Am J Pathol 1998; 153: 6979.
  • 25
    Stinn JL, Taylor MK, Becker G et al. Interferon-gamma-secreting T-cell populations in rejecting murine cardiac allografts: assessment by flow cytometry. Am J Pathol 1998; 153: 13831392.
  • 26
    Briones MA, Phillips DJ, Renshaw MA, Hooper WC. Expression of chemokine by human coronary-artery and umbilical-vein endothelial cells and its regulation by inflammatory cytokines. Coron Artery Dis 2001; 12: 179186.
  • 27
    Denk A, Goebeler M, Schmid S et al. Activation of NF-kappa B via the I-kappa B kinase complex is both essential and sufficient for proinflammatory gene expression in primary endothelial cells. J Biol Chem 2001; 276: 2845128458.
  • 28
    Deckert-Schluter M, Bluethmann H, Rang A, Hof H, Schluter D. Crucial role of TNF receptor type 1 (p55), but not of TNF receptor type 2 (p75), in murine toxoplasmosis. J Immunol 1998; 160: 34273436.
  • 29
    Wolf D, Hallmann R, Sass G et al. TNF-alpha-induced expression of adhesion molecules in the liver is under the control of TNFR1 – relevance for concanavalin A-induced hepatitis. J Immunol 2001; 166: 13001307.
  • 30
    Kurrelmeyer KM, Michael LH, Baumgarten G et al. Endogenous tumor necrosis factor protects the adult cardiac myocyte against ischemic-induced apoptosis in a murine model of acute myocardial infarction. Proc Natl Acad Sci U S A 2000; 97: 54565461.
  • 31
    Schreyer SA, Peschon JJ, LeBoeuf RC. Accelerated atherosclerosis in mice lacking tumor necrosis factor receptor p55. J Biol Chem 1996; 271: 2617426178.
  • 32
    Grenz A, Schenk M, Zipfel A, Viebahn R. TNF-alpha and its receptors mediate graft rejection and loss after liver transplantation. Clin Chem Lab Med 2000; 38: 11831185.
  • 33
    Dorge SE, Roux-Lombard P, Dayer JM, Koch KM, Frei U, Lonnemann G. Plasma levels of tumor necrosis factor (TNF) and soluble TNF receptors in kidney transplant recipients. Transplantation 1994; 58: 10001008.
  • 34
    Scheringa M, De Bruin RW, Jeekel H, Marquet RL. Anti-tumor necrosis factor alpha serum prolongs heart allograft survival in rats. Transplant Proc 1991; 23: 547548.
  • 35
    Hill GR, Teshima T, Rebel VI et al. The p55 TNF-alpha receptor plays a critical role in T cell alloreactivity. J Immunol 2000; 164: 656663.
  • 36
    Lucas R, Garcia I, Donati YR et al. Both TNF receptors are required for direct TNF-mediated cytotoxicity in microvascular endothelial cells. Eur J Immunol 1998; 28: 35773586.
  • 37
    Broide DH, Stachnick G, Castaneda D, Nayar J, Sriramarao P. Inhibition of eosinophilic inflammation in allergen-challenged TNF receptor p55/p75 – and TNF receptor p55-deficient mice. Am J Respir Cell Mol Biol 2001; 24: 304311.
  • 38
    Schmid EF, Binder K, Grell M, Scheurich P, Pfizenmaier K. Both tumor necrosis factor receptors, TNFR60 and TNFR80, are involved in signaling endothelial tissue factor expression by juxtacrine tumor necrosis factor alpha. Blood 1995; 86: 18361841.
  • 39
    Mackay F, Loetscher H, Stueber D, Gehr G, Lesslauer W. Tumor necrosis factor alpha (TNF-alpha)-induced cell adhesion to human endothelial cells is under dominant control of one TNF receptor type, TNF-R55. J Exp Med 1993; 177: 12771286.
  • 40
    Murakami T, Mataki C, Nagao C et al. The gene expression profile of human umbilical vein endothelial cells stimulated by tumor necrosis factor alpha using DNA microarray analysis. J Atheroscler Thromb 2000; 7: 3944.
  • 41
    Paleolog EM, Delasalle SA, Buurman WA, Feldmann M. Functional activities of receptors for tumor necrosis factor-alpha on human vascular endothelial cells. Blood 1994; 84: 25782590.
  • 42
    Gerdes N, Sukhova GK, Libby P, Reynolds RS, Young JL, Shoenbeck U. Expression of interleukin (IL) -18 and functional IL-18 receptor on human vascular endothelial cells, smooth muscle cells and macrophages: implications for atherogenesis. J Exp Med 2002; 195: 245257.
  • 43
    Shimizu K, Sugiyama S, Aikawa M et al. Host bone-marrow cells are a source of donor intimal smooth-muscle-like cells in murine aortic transplant arteriopathy. Nat Med 2001; 7: 738741.