Sustained apoptosis in human cardiac allografts despite histologic resolution of rejection1 : Transplantation

Journal Logo


Sustained apoptosis in human cardiac allografts despite histologic resolution of rejection1

Masri, Sofia C.2; Yamani, Mohamad H.3; Russell, Mary A.2; Ratliff, Norman B.4; Yang, Jiacheng2; Almasan, Alex5; Apperson-Hansen, Carolyn6; Li, Jianbo6; Starling, Randall C.3; McCarthy, Patrick7 8; Young, James B.3 8; Bond, Meredith2 9

Author Information
doi: 10.1097/01.TP.0000084824.70320.DA
  • Free


Apoptosis occurs during acute cardiac allograft rejection; however, the question of whether apoptosis is resolved during resolution of the rejection episode is unknown. Furthermore, the major cell type undergoing apoptosis is controversial. Depending on the target cells, apoptosis could play different roles in graft survival (1,2). Some studies show a direct correlation between increased cardiomyocyte apoptosis and grade of rejection (3,4), suggesting that apoptosis of the cardiac myocytes contributes to allograft damage and dysfunction, whereas other studies did not show evidence of apoptosis in cardiomyocytes, even in regions of cardiomyocyte damage but instead report most terminal deoxynucleotide transferase-mediated dUTP nick-end labeling (TUNEL)-positive cells among leukocyte infiltrates (5,6). Apoptosis of T lymphocytes could thus be a mechanism to decrease an ongoing immune response. In fact, some immunosuppressive drugs currently used in clinical transplantation such as steroids act, in part, through induction of apoptosis in infiltrating lymphocytes (7,8), suggesting another mechanism of increased programmed cell death in cardiac allografts.

The changes associated with apoptosis are attributable to actions of caspases that play a key role in the regulation and execution of apoptotic cell death (9,10). One of the apoptotic signals is mediated by interaction of Fas ligand with Fas receptor-expressing cells, leading to activation of caspase-8, which then activates downstream effectors such as caspase-3, and leads to the morphologic and biochemical features of apoptosis (10–12). In cardiac transplant rejection, activation of T cells induces expression of Fas ligand (13,14) and is paralleled by increased expression of inducible nitric oxide synthase (2,4).

We performed a comprehensive evaluation of apoptotic activity in human cardiac biopsy specimens. We studied biopsy specimens from allografts with no history of rejection, during an episode of moderate rejection, and after resolution of cellular infiltrates. Our results indicate that the major cells undergoing apoptosis are inflammatory cells and that apoptosis persists despite histologic resolution of the rejection episode. The prolonged elevation of caspase activity and TUNEL-positive staining after the rejection episode is resolved suggests that rejection may be controlled by apoptosis of inflammatory infiltrates.


Patient Characteristics

This study was approved by The Cleveland Clinic Foundation Institutional Review Board. Informed consent was obtained from all patients. Fifty-nine endomyocardial biopsy (EMB) specimens were obtained from 24 patients (19 men and 5 women; median age, 57±10 years; range, 33–70 years) after heart transplantation (range 1–6 months). Frozen and paraffin-embedded biopsy specimens were graded according to the criteria of the International Society for Heart Transplantation grading system (15). One extra sample of myocardium from each biopsy was snap-frozen in liquid nitrogen and stored at −80°C for research purposes.

Eight patients had a diagnosis of ischemic heart disease, 15 had a diagnosis of idiopathic dilated cardiomyopathy, and 1 had valvular heart disease, before heart transplant. Immunosuppressive therapy consisted of prednisone, cyclosporine A, and mycophenolate mofetil (21 patients); prednisone, tacrolimus, and mycophenolate mofetil (2 patients); and prednisone, cyclosporine A, and azathioprine (1 patient). Episodes of rejection were treated with a 3-day course of oral prednisone (100 mg/day).

The three experimental groups are presented in Table 1. On each frozen biopsy specimen (3–5 mg of tissue wet weight), we measured caspase-8 and caspase-3 activity, performed Western blot analysis for caspase-3 expression and, in paraffin-embedded samples, determined co-localization of procaspase-3 and TUNEL.

Table 1:
Experimental groupsa

Caspase-3 and Caspase-8 Activity

Caspase-3 and caspase-8 activity was measured in homogenates of frozen biopsy specimens. Samples were homogenized in HEPES buffered solution (10 mM, pH 7.4) containing EDTA (2 mM), CHAPS (0.1%), DTT (5 mM), pepstatin A (1 μmol/L), leupeptin (1 μmol/L), and PMSF (0.8 mmol/L). Supernatants were collected after centrifugation at 15,000 rpm for 10 min at 4°C. Fluorometric caspase-3 and caspase-8 activity assays were performed in 96-well plates by incubating 35 μg of total protein with the synthetic fluorescent substrate Ac-DEVD-AFC (for caspase-3) and Ac-LETD-AFC (for caspase-8) (BioRad Laboratories, Hercules, CA) at 37°C for 6 hr, according to the manufacturer’s instruction.

Caspase activity was measured by release of AFC from the caspase substrate Ac-DEVD-AFC (for caspase-3) or Ac-LETD-AFC (for caspase-8) (excitation, 400 nm; emission, 530 nm) (Cytofluor II TM; PerSeptive Biosystems, Inc., Framingham, MA). After correction for background activity, enzyme activity was expressed as change in activity (ΔS) minus baseline activity as follows: ΔS=[S(t1)−B(t1)]−[S(t0)−B(t0)], where S=sample signal at time t, B=blank signal at time t, t1=time of last measurement, and t0=time of initial measurement.

Immunoblot Analysis

Thirty-five micrograms of total protein from homogenates of biopsy specimens (as described above) were separated onto 12% sodium dodecyl sulfate-polyacrylamide gel electrophoresis and transferred onto polyvinylidine fluoride membrane. The membranes were blocked in 3% milk in Tris-buffered saline containing 0.1% Tween-20 (TBST). After two washes of 10 min in TBST, the membranes were incubated overnight at 4°C with a polyclonal rabbit anti–caspase-3 antibody (1:1,000) (PharMingen, San Diego, CA) or anti–glyceraldehyde-3-phosphate dehydrogenase (GAPDH) monoclonal antibody (1:3,000) (Research Diagnostics, Inc., Flanders, NJ). According to the manufacturer, the anti–caspase-3 antibody recognizes both the 32-kDa unprocessed procaspase-3 and the 17-kDa subunit of the active caspase-3. Protein expression was normalized to GAPDH expression on the same immunoblot. After caspase-3 immunoblot analysis, antibody was stripped from the membrane, which was then used for anti-GAPDH immunoblotting. The secondary antibody was an alkaline phosphatase-linked anti-mouse immunoglobulin (Ig) G (1:3,000) or anti-rabbit IgG (1:2,500). Membranes were developed with VISTRA ECF substrate and the chemiluminescence corresponding to the bands was quantified by StormImager (Molecular Dynamics, Mountain View, CA).

Co-localization of Caspase-3 with In Situ Detection of Apoptotic Cell Death

To identify the cell types undergoing apoptosis, immunostaining of tissue for caspase-3 and TUNEL analysis were performed on the same paraffin-embedded specimens. Biopsy specimens were fixed in 10% buffered formalin and embedded in paraffin. Four-micron-thick sections were then cut and mounted on sialine-coated slides. To determine extent and severity of rejection, sections were stained with hematoxylin-eosin for routine histologic examination. Grading of rejection according to the ISHLT classification was carried out independently by a blinded observer (N.B.R.). The remaining sections were used for combined immunohistochemistry of caspase-3 and in situ detection of apoptotic cells by TUNEL. The samples were first immunostained for caspase-3 using rabbit anti-human anti–caspase-3 polyclonal antibody (90 min at 37°C) (1:400; Pharmigen). After washing in phosphate-buffered saline, the slides were incubated with a horseradish peroxidase-linked secondary antibody: horseradish peroxidase anti-rabbit (1:200) (Vector Laboratories, Burlingame, CA). Immunoreactivity was visualized by a 10-min incubation with the horseradish peroxidase substrate diaminobenzidine. Negative controls included an irrelevant rabbit IgG antibody substituted for the primary antibody caspase-3 and preabsorption of the primary antibody with the corresponding synthetic antigens. After staining for caspase-3, the same slides were then processed for in situ detection and localization of apoptosis at the level of single cells. We used terminal deoxynucleotidyl transferase to incorporate fluorescein-labeled dUTP into DNA strand breaks (In Situ Cell Death Detection Kit; Boehringer Mannheim, Mannheim, Germany) according to the manufacturer’s recommendations. Sections were then stained with antifluorescein antibodies linked with alkaline phosphatase, developed with Fast Red substrate and counterstained with hematoxylin. The percentage of TUNEL-positive nuclei was calculated as follows: six randomly chosen fields per slide, corresponding to approximately 700 nuclei, were examined at high magnification (×200). As a positive control, normal rat heart tissue was treated with DNase I (1 μg/mL) for 10 min at room temperature to induce DNA strand breaks. For negative controls, sections were incubated in labeling solution without terminal deoxynucleotidyl transferase.

Statistical Methods

Caspase-8 and caspase-3 activity and Western blot analysis of caspase-3 protein were compared among groups I, II, and III using a mixed model and the procedure MIXED in SAS (SAS Institute, Inc., Cary, NC). Group differences were compared for each patient, treating the change of activity over baseline as a response variable, the patient as a random factor, and the group as a fixed factor. Data from caspase-8 and caspase-3 activity and Western blot analysis of caspase-3 were normalized with a square-root transformation before performing analysis of variance for each of the three measurements. When the model was significant (at the level of P <0.05), pairwise comparisons were performed on the basis of least-square means and variances. The reported means±SEM was on the original scale and performed with a back-transformation from the least-square estimates. Correlations among the three measurements were estimated on the basis of original data scale using parametric (Pearson) and nonparametric (Spearman or Kendall) methods.


Caspase-8 and Caspase-3 Activity

Compared with biopsy specimens demonstrating mild or no rejection and with no previous rejection episodes (group I; n=23), caspase-8–like activity and caspase-3–like activity were significantly increased in biopsy specimens with moderate rejection (group II; n=24) (caspase-8, P =0.02; caspase-3, P <0.05) (Fig. 1). These increases were sustained in the 12 EMB specimens showing evidence of histologic improvement within 1 to 2 weeks after a rejection episode (grades 0 and 1 of the ISHLT classification) (group III), that is, caspase-8 and caspase-3 activities were significantly elevated in group III (P <0.005 for caspase-8 and P <0.0025 for caspase-3) compared with group I.

Figure 1:
Caspase-8 and caspase-3 activity. (A) Hydrolysis of caspase-8, measured by release of AFC from Ac-DEVD-AFC caspase substrates (excitation, 400 nm; emission, 530 nm) is increased in groups II and III vs. group I (P <0.05 for group II vs. group I and P <0.0025 for group III vs. group I). Each bar represents mean±SEM. (B) Enzymatic hydrolysis of caspase-3 (excitation, 400 nm; emission, 530 nm) measured by release of AFC using Ac-DEVD-AFC as caspase substrate is also increased in groups II and III vs. group I (P =0.02 for group II vs. group I; and P <0.005 for group III vs. group I). Each bar represent mean±SEM.

Consistent with observations that caspase-8 is an upstream caspase activating caspase-3, a high correlation coefficient was observed between caspase-8 and caspase-3 activity during acute rejection (Pearson r =0.82; P <0.0001) (Fig. 2). Taken together, these results demonstrate that apoptotic activity is increased during an episode of moderate rejection and, notably, is still elevated after treatment.

Figure 2:
Correlation between caspase-8 and caspase-3 activities. Activity is measured by release of AFC from caspase substrates Ac-DEVD-AFC (caspase-3) and Ac-LETD-AFC (caspase-8) (excitation, 400 nm; emission, 530 nm).

Western Blot Analysis of Caspase-3

The proform (p32) of caspase-3 was detected by Western blot analysis in all specimens (Fig. 3A). Expression of procaspase-3 was increased 1.6-fold in biopsy specimens with moderate rejection (group II; n=24) compared with grades 0 and 1 with no prior episodes of rejection (group I; n=23). After immunosuppressive therapy and resolution of histologic evidence of rejection (group III; n=12), these differences achieved borderline significance (P <0.07) compared with group I (Fig. 3B). The activated form (p17) of caspase-3 was detected only in EMB specimens of group II (50% of samples) and group III (33% of samples) (P <0.01); no evidence of p17 was found in EMB specimens of patients from group I. Thus, consistent with the measurements of caspase-3 activity (see above), expression of the activated form of caspase-3, as detected by Western blot, was present only during episodes of moderate rejection and after immunosuppressive treatment.

Figure 3:
Expression of procaspase-3 (p32) and cleavage form of caspase-3 (p17). (A) Expression of caspase-3 protein in EMB specimens of patients 1, 2, and 3. The p32 form of caspase-3 was detected in all patients; p17 was detected only in groups II and III. Procaspase-3 levels were normalized to GAPDH. Overall, 50% of patients in group II and 33% of EMB specimens in group III showed the presence of the p17 cleavage form of caspase-3. No patients in group I showed p17 expression. (B) Analysis of p32 expression in groups I, II, and III. The data, normalized to GAPDH, are presented in arbitrary intensity units. The increased expression of procaspase-3 in groups II and III vs. group I achieved borderline significance (P <0.07). Each bar represent mean±SEM.

In Situ Detection of Apoptosis in the Rejecting Heart and Co-localization with Procaspase-3

TUNEL-positive cells were characterized by focal nuclear staining. During moderate rejection (group II), 16±2.6% (±SEM) of inflammatory cells and 1.2±0.2% of cardiomyocytes stained positive for TUNEL. After immunosuppression treatment (group III), 7.1±1.3% of inflammatory cells and 0.8±0.3% of cardiomyocytes were TUNEL positive. No TUNEL-positive cells were found in group I (P <0.02 for inflammatory cells vs. groups II and III and P <0.05 for cardiomyocytes vs. groups II and III).

Immunohistochemical analysis of procaspase-3, performed on the same tissue sections used to identify TUNEL-positive cells, revealed increased procaspase-3 expression in groups II and III, compared with group I (Fig. 4). Western blot analysis (Fig. 3A) showed the presence of p32 in all groups, although the activated form (p17) of caspase-3 was only present in groups II and III. Because the anti–caspase-3 antibody used detects the proform and the cleavage form of the enzyme, the higher expression of caspase-3, detected in groups II and III compared with group I, may result from immunodetection of both p32 and p17 (Fig. 4).

Figure 4:
Sequential EMB specimens of two patients showing immunohistochemical detection of caspase-3 (brown) and TUNEL (pink). Counterstaining was performed with hematoxylin (see Patients and Methods). (arrowheads) Positive TUNEL staining in inflammatory cells. (A) Patient 4: During rejection (grade 3A, group II), note the presence of coarse-grained caspase-3. (B) Patient 4: After treatment (group III), there is continued expression of caspase-3 and TUNEL-positive inflammatory cells. (C) Patient 5: During grade 1B rejection with no prior rejection (group I), weak diffuse fine granular caspase-3 immunoreactivity is found in myocytes, but TUNEL-positive cells are not detected. (D) This EMB specimen followed an episode of grade 3A rejection (group II). Sections show coarse-grained caspase-3 and TUNEL-positive cells in inflammatory cells and cardiomyocytes.

Consistent with our measurements of caspase activity and caspase-3 protein, increased numbers of TUNEL-positive inflammatory but few cardiomyocyte cells were observed during moderate rejection (group II) (Fig. 4). TUNEL-positive inflammatory cells, and to a much lesser extent, TUNEL-positive cardiomyocytes were also observed after treatment (group III). In contrast, there were no TUNEL-positive cells in EMB specimens of patients with mild rejection or no previous episodes of rejection (group I). (Figure 5B) also shows that during acute cellular rejection, the major cells undergoing programmed cell death were infiltrating inflammatory cells. Figure 6 shows negative controls for procaspase-3 and TUNEL (Fig. 6A) and positive control for TUNEL (Fig. 6B).

Figure 5:
(A) Patient 5: During grade 3A rejection (group II), note the presence of coarse-grained caspase-3 and a high proportion of TUNEL-positive cells (arrowheads) in the inflammatory cells. (B) Patient 6: With no rejection, grade 0 (group I), fine granular caspase-3 immunoreactivity is seen in most of the myocytes; TUNEL-positive cells are absent.
Figure 6:
(A) Negative control for procaspase-3 and TUNEL. (B) TUNEL-positive control.


We investigated whether apoptosis in human cardiac allografts is associated exclusively with the rejection episode itself. We found that the increased apoptosis in cardiac allografts during acute cellular rejection persists during histologic resolution of cellular infiltrates. The increase in TUNEL-positive inflammatory cells during acute rejection is consistent with previous observations (5,6,16). The finding that most TUNEL-positive cells are inflammatory cells may reflect the fact that apoptosis helps regulate an ongoing immune response in the allograft (12,16,17). In contrast, we found a low percentage of TUNEL-positive cardiomyocytes both during rejection and after resolution. These results differ from some previous studies (3,4). These differences highlight controversies regarding extent of programmed cell death in cardiomyocytes during rejection and its role in graft damage.

Human cardiac allograft recipients received immunosuppressive drugs that could act through induction of apoptosis in inflammatory cells. In our study, 80% (22 of 24) of patients received prednisone (100 mg/day for 3 days) as treatment for moderate rejection. Previous reports indicate that glucocorticoids may have a dual effect on T-cell activation and apoptosis, because they can simultaneously activate cell suicide or an anti-death program (8,18–20). CD4 and CD8 lymphocytes are exquisitely sensitive to glucocorticoid-mediated cell death, and even physiologic concentrations achieved during a stress response may be sufficient to trigger programmed cell death. Thus, another mechanism to explain apoptotic activity after treatment is activation of peripheral T cells, making them more sensitive to glucocorticoids through caspase activation (20). However, in the context of cardiac allograft rejection with increased cytokine production, it is also possible that other pathways contribute to regulation of apoptosis (16,21).

Activation of the cell surface death receptor is a well-known mechanism for activating caspase-8 (11,12,21,22). We found that caspase-8 and caspase-3 activation occurred during moderate rejection and was maintained after immunosuppression treatment, and there was a strong correlation between caspase-8 and caspase-3 activities.

The observation that increased apoptotic activity occurred mainly in inflammatory cells may have implications for graft survival, with the possibility that induction of apoptosis could be exploited in a beneficial way (23). Also, our finding of a sustained increase in apoptosis after histologic resolution may be useful in interpreting the results of Annexin V imaging, a noninvasive method proposed to diagnose rejection (24–26). On induction of apoptosis, binding of Annexin V to phosphatidylserine on the outer leaflet of cell membranes occurs in cells undergoing apoptosis. Our results, showing persistent apoptosis of inflammatory cells despite evidence of resolution of rejection (return to New York Heart Association class 0, 1A, or 1B), raises the interesting issue that in human cardiac allografts, Annexin V imaging may not discriminate between episodes of acute rejection and resolution of the rejection episode.


1. Sayegh M, Turka L. The role of T-cell costimulatory activation pathways in transplant rejection. N Engl J Med 1998; 338: 1813.
2. Miller LW, Granville DJ, Narula J, et al. Apoptosis in cardiac transplant rejection. Cardiol Clin 2001; 19: 141.
3. Laguens RP, Meckert PM, Martino JS, et al. Identification of programmed cell death (apoptosis) in situ by labeling of nuclear DNA fragments in heart biopsy samples during acute rejection episodes. J Heart Lung Transplant 1996; 15: 911.
4. Szabolcs M, Ravalli S, Minanov O, et al. Apoptosis and increased expression of nitric oxide synthase in human allograft rejection. Transplantation 1998; 65: 804.
5. Van Hoffen E, Van Wichen DF, Leemans JC, et al. T cell apoptosis in human heart allografts: Association with lack of co-stimulation? Am J Pathol 1998; 153: 813.
6. Bergese SD, Klenotic SM, Wakely ME, et al. Apoptosis in murine cardiac grafts. Transplantation 1997; 63: 320.
7. Li XC, Strom TB, Turka LA, et al. T cell death and transplantation tolerance. Immunity 2001; 14: 407.
8. Aswell JD, Lu FW, Vacchio M. Glucocorticoids in T cell development and function. Annu Rev Immunol 2000; 18: 309.
9. Budihardjo I, Oliver H, Lutter M, et al. Biochemical pathways of caspase activation during apoptosis. Annu Rev Cell Dev Biol 1999; 15: 269.
10. Reed JC. Mechanisms of apoptosis. Am J Pathol 2000; 157: 1415.
11. Nagata S. Apoptosis by death factor. Cell 1997; 88: 355.
12. Larsen CP, Alexander DZ, Hendrix R, et al. Fas mediated cytotoxicity: An immunoeffector or immunoregulatory pathway in T cell mediated immune response? Transplantation 1995; 60: 221.
13. Haunstetter A, Izumo S. Apoptosis: Basic mechanisms and implications for cardiovascular disease. Circ Res 1998; 82: 1111.
14. Thornberry NA, Lazebnik Y. Caspases: Enemies within. Science 1998; 281: 1312.
15. Billingham ME, Cary NR, Hammond ME, et al. A working formulation for the standardization of nomenclature in the diagnosis of heart and lung rejection. Heart Rejection Study Group: The International Society for Heart Transplantation. J Heart Transplant 1990; 9: 587.
16. Kabelitz D. Apoptosis, graft rejection, and transplantation tolerance. Transplantation 1998; 65: 869.
17. Wells AD, Li XC, Li Y, et al. Requirement for T-cell apoptosis in the induction of peripheral transplantation tolerance. Nat Med 1999; 5: 1303.
18. Wolf BB, Green D. Suicidal tendencies: Apoptotic cell death by caspase family proteinases. J Biol Chem 1999; 274: 20049.
19. Nagata S, Golstein P. The Fas death factor. Science 1995; 267: 1449.
20. Mann CL, Hughes FM Jr, Cidlowski JA. Delineation of the signaling pathways involved in glucocorticoid-induced and spontaneous apoptosis of rat thymocytes. Endocrinology 2000; 141: 528.
21. Ashkenazi A, Dixit VM. Apoptosis control by death and decoy receptors. Curr Opin Cell Biol 1999; 11: 255.
22. Takeuchi T, Ueki T, Nishimatsu H, et al. Accelerated rejection of FAS ligand-expressing heart grafts. J Immunol 1999; 162: 518.
23. Jollow KC, Sundstrom JB, Gravanis MB, et al. Apoptosis of mononuclear cell infiltrates in cardiac allograft: Biopsy specimens question studies of biopsy-cultured cells. Transplantation 1997; 63: 1482.
24. Vriens PW, Blankenberg FG, Stoot JH, et al. The use of technetium Tc 99m annexin V for in vivo imaging of apoptosis during cardiac allograft rejection. J Thorac Cardiovasc Surg 1998; 116: 844.
25. Blankenberg FG, Katsikis PD, Tait JF, et al. In vivo detection and imaging of phosphatidylserine expression during programmed cell death. Proc Natl Acad Sci USA 1998; 95: 6349.
26. Narula J, Acio ER, Narula N, et al. Annexin-V imaging for noninvasive detection of cardiac allograft rejection. Nat Med 2001; 7: 1347.
© 2003 Lippincott Williams & Wilkins, Inc.