Link Between Immune Cell Infiltration and Mitochondria-Induced Cardiomyocyte Death During Acute Cardiac Graft Rejection



Acute cardiac graft rejection (ACGR) is associated with cardiomyocyte apoptosis. We investigated the respective role of the Fas/FasL and mitochondrial permeability transition pore (mPTP) pathways in cardiomyocyte apoptosis accompanying ACGR. Heterotopic cardiac transplantations were performed in 7–9-week old C57BL6 or C3H mice. Wild type or Fas-deficient (lpr) mice underwent syngeneic (GS) or allogeneic (GA) transplantation, and received either saline or NIM811, a specific inhibitor of the mPTP. At day 5, we assessed ACGR by histology, cardiomyocyte apoptosis by caspase-3 activity and cytochrome c release, Ca2+-induced mPTP opening by a potentiometric approach, and expression of Fas, FasL, TNFα, perforin, granzyme using RT-PCR. Myocardial infiltration of CD8+ T lymphocytes was performed by immunohistochemistry. Allogenic transplantation increased infiltration of inflammatory cells, upregulated FasL, perforin, granzyme, and TNFα, favored Ca2+-induced mPTP opening and increased caspase-3 activity and cytochrome c release in WT grafts. NIM811, but not Fas-deficiency, significantly reduced all these effects. NIM811 also limited infiltration of CD8+ into WT and lpr transplants. These data suggest that the mPTP pathway plays a major role in cardiomyocyte apoptosis associated with ACGR. Inhibition of mPTP opening may attenuate cardiomyocyte apoptosis either directly or indirectly via a limitation of CD8+ T-cell activation.


Acute allograft rejection is the manifestation of a complex series of coordinated events initiated by activated alloreactive T cells. Antigen-primed T helper cells (CD4+ T cells) and the cytokine milieu help drive the effector mechanisms of rejection that include the generation of cytotoxic T cells (CD8+ T cells). T-cell-mediated cytotoxicity (CTL) is established as a fundamental effector mechanism of allogeneic-targeted cell death (1). Beside the cytolytic death induced by the perforin/granzyme system driven by CD8+ T cells, cardiomyocytes may also undergo apoptosis, that has been recognized as an important mechanism of cell death and graft failure during acute graft rejection. Apoptosis can be induced by extrinsic and intrinsic signals produced following cellular stress. The two classical pathways of apoptosis are (i) the cell death receptor pathway, including Fas/Fas Ligand (FasL), and (ii) the mitochondria pathway.

The role of mitochondria in the apoptotic process has been clearly demonstrated (2). It involves opening of a nonspecific mega-channel in the inner mitochondrial membrane, called the mitochondrial permeability transition pore (mPTP), that causes a loss of the mitochondrial membrane potential (ΔΨm) and the efflux of small molecules such as cytochrome c and other proapoptotic factors (2). We recently demonstrated that mPTP opening plays a key role in cardiomyocyte apoptosis associated with acute allograft rejection (3). But, cardiomyocytes are also rich in Fas (4) and cytotoxic T-lymphocytes express FasL during acute graft rejection (5,6). It is not clear which of the Fas/FasL or the mitochondria pathway might predominately induce apoptosis during ACGR.

Therefore, in the present study, we combined the genetic inactivation of Fas receptor on one hand, to the pharmacological inhibition of the mPTP on the other hand, to investigate the respective importance of either pathway on cardiomyocyte apoptosis associated with ACGR.

Material and Methods


Seven- to nine-week-old male C57BL/6JICO +/+ (B6) and C3H/HEOUJICO (C3H) mice were purchased from Charles RIVERS laboratories (L'Arbresle, FRANCE). Seven- to nine-week-old male lpr (lymphoproliferative, C57BL/6 background) mice were obtained from CDTA-CNRS (Orléans, FRANCE). All animals had unrestricted access to water and food, and were housed in accordance with the 'The Guide for the Care and Use of Laboratory Animals' (NIH Publication No. 85-23, revised 1996).

Experimental preparation: Heterotopic abdominal cardiac transplantation

Heterotopic heart transplantations were performed, as previously described (7). The mice were anesthetized with 0.3 mL/10 g of a 1:1 mixture of fentanyl citrate (0.011 mg/mL) and midazolam (0.4 mg/mL; Roche, France) given intra-peritonealy. Briefly, donor ascending aorta and the pulmonary trunk from the heart graft was anastomosed end-to-side to the recipient abdominal aorta and inferior vena cava, respectively, using 10–0 sutures. Cold ischemia period lasted less than 20 min. In this experimental model, acute rejection usually occurs between day 5 and day 7 after transplantation (8). Acute cardiac rejection was defined by cessation of a palpable heartbeat and was confirmed by histology.

Experimental protocol

We addressed the contribution of Fas/FasL pathway in apoptosis during acute allograft rejection by using mice with loss-of-function mutation in the Fas gene, called lpr (for lymphoproliferation) as donor. The lpr mutation results from the insertion of an early transposable element (ETn) into intron 2 of the gene-encoding Fas (9).

We investigated whether NIM811, a nonimmunosuppressive derivative of cyclosporine A (CsA), that specifically inhibits mPTP opening, might also affect cardiomyocyte apoptosis. In the treated groups, NIM811 (40 mg/kg per 8 h) was administered intra-peritonealy, starting immediately at the end of the cardiac transplantation.

Hearts were assigned to sham or grafted groups. In grafted groups, hearts could be either syngenic or allogenic. Syngenic hearts were used as controls for the corresponding (wild-type or deficient mice) allogenic hearts. In all groups, grafted hearts were harvested at day 5.

Seven experimental groups were defined.

  • 1Sham group (S): hearts were from C57BL6 (wild-type, WT) mice that only underwent a laparotomy.
  • 2Grafted syngenic group (GS): hearts were from C57BL6 (WT) mice, and transplanted into a C57BL6 recipient.
  • 3Grafted syngenic group (GS-C3H): hearts were from C3H (WT) mice, and transplanted into a C3H recipient.
  • 4Grafted allogenic group (GA): hearts were from C57BL6 (WT) mice, and transplanted into a C3H recipient.
  • 5GA-NIM group (GA-NIM): hearts were from C57BL6 (WT) mice, and transplanted into a C3H recipient. They received NIM811 treatment, as described above.
  • 6Grafted allogenic-lpr group (GA-lpr): hearts were from lpr mice, and transplanted into a C3H recipient.
  • 7Grafted allogenic-lpr NIM group (GA-lpr-NIM): hearts were from lpr mice, and transplanted into a C3H recipient. They received NIM811 treatment, as described above.


Cardiac histology:  Cardiac grafts were fixed in 10% paraformaldehyde, embedded in paraffin, cut into 5 μM sections, and stained with hematoxylin and eosin. For each slide, the mean number of inflammatory cells in five infiltrate focuses (high-powered field) was measured using a ×40 objective. For immunohistochemical staining, harvested grafts were immediately frozen in OCT compound. Cryostat sections were fixed with acetone and stained with anti-mouse CD8 (Beckman Coulter, France) mAb by using the peroxydase and DAKO AEC substrate system. The mean number of CD8+ lymphocytes was calculated in five fields using a ×40 magnification.

Myocardial apoptosis

Caspase-3 activity:  Heart lysates were centrifuged at 16 000 g for 45 min at 4°C, and supernatant fractions were collected. Caspase activity was determined, as previously described (3), using a fluorometric assay system with DEVD-AMC (Bachem Biochimie, France) as fluorochrome-associated caspase specific substrate. Fluorescence was measured on a spectrofluorometer (Perkin Elmer LS-5) at an excitation wavelength of 342 nm and an emission wavelength of 441 nm after incubation at 37°C for 60 min. The quantity of released 7-amino-4-methyl coumarin (AMC) was calculated by transposing the fluorescence measurement of each point onto a scale with purified AMC.

Western blot analysis of cytochrome c:  At day 5, hearts were harvested and homogenized in buffer A (70 mM sucrose, 210 mM mannitol, 1 mM EDTA in 50 mM Tris/HCL pH 7.4). Following a centrifugation at 100 000 g for 1 h at 4°C, the final supernatants (cytosolic fractions) were collected. Cytosolic proteins equivalent to 50 μg protein were separated by 15% SDS-PAGE and transferred to nitrocellulose membrane, blocked with 5% nonfat dried milk in PBS for at least 1 h, washed with PBS/Tween, and incubated with anti-cytochrome c (clone 7H8.2C12, Pharmingen®) in PBS/Tween for 12 h at 4°C. Blots were developed with secondary antibody (anti-mice, Jackson Immuno Research®) diluted 1/1000 in PBS/Tween. After washing with PBS/Tween, the blots were developed with an enhanced chemiluminescence (Lumi Light® kit, Roche), and exposed for autoradiography. For stripping, blots were incubated for 5 min into a buffer containing 0.2 M Tris/HCl, pH 6.8, 8% SDS, and 20 mM mercaptoethanol at 100°C. Blots were washed three times for 10 min in PBS/Tween. The amount of protein to be used for detection was normalized using the GAPDH protein as loading control. The semiquantitative expression of cytochrome c was determined using a computerized software package (ImageJ®).

Mitochondrial permeability transition pore opening

Isolation of mitochondria:  Mitochondria were isolated from cardiac grafts by homogenization and differential centrifugations in isolation buffer A: (70 mM sucrose, 210 mM mannitol, 1mM EDTA in 50 mM Tris/HCL pH 7.4). The homogenate was centrifuged twice at 1300 g for 3 min. The supernatant was centrifuged at 10 000 g for 10 min. The mitochondrial pellet (1 mg mitochondrial proteins) was washed and resuspended in buffer B: (70 mM sucrose, 210 mM mannitol in 50 mM Tris/HCl pH 7.4). Protein content was assayed according to Gornall's procedure (10). Purity and integrity of isolated mitochondria were assessed by measuring specific activity of cytochrome c oxydase (EC as inner membrane marker enzyme.

Ca2+-induced mPTP opening:  mPTP opening was assessed following in vitro Ca2+ overload (11,12). Isolated mitochondria were added in 450 μL of buffer C (150 mM sucrose, 50 mM KCl, 2 mM KH2PO4, 5 mM succinic acid in 20 mM Tris/HCl pH 7.4), within a Teflon chamber equipped with a Ca2+-selective microelectrode, in conjunction with a reference electrode (13,14). A Synchronie® software allowed a continuous recording of extramitochondrial calcium variations. Mitochondria were gently stirred for 1.5 min of pre-incubation. At the end of the pre-incubation period, 10 μM CaCl2 pulses were performed every minute. Each 10 μM CaCl2 injection causes a peak of extramitochondrial Ca2+ concentration. Then, Ca2+ is rapidly taken up by mitochondria, which results in a return of extramitochondrial Ca2+ concentration to near baseline level (Figure 4A). Following sufficient Ca2+ loading, extramitochondrial Ca2+ concentration abruptly increases, indicating a massive release of Ca2+ due to mPTP opening, as previously described (3) (Figure 4A). The amount of Ca2+ required to trigger this massive Ca2+ release is used as an indicator of the susceptibility of mPTP to Ca2+ overload.

Figure 4.

Panel (A) Typical example of Ca 2+ -induced MPT pore opening. In the GA mitochondria, a Ca2+ overload of 40 μM (4 pulses of 10 μM) was required to induce mPTP opening versus 110 μM Ca2+ (11 pulses of 10 μM Ca2+) in the GS heart. Panel (B) Ca2+ overload required for mPTP opening in untreated and treated grafts. In the GA group, Ca2+ overload required for mPTP opening was significantly reduced versus GS animals. GA-lpr displayed Ca2+ load close to GA values. Mitochondria isolated from GA or GA-lpr grafts treated with NIM811 exhibited an increased resistance to Ca2+ overload. *p < 0.05 versus GS.

Reverse transcriptase-polymerase chain reaction (RT-PCR)

Total cellular RNA was extracted from frozen heart tissue using TRIzol® Reagent (Invitrogen™, France). One milliliter of TRIzol was added to 100 μg of heart sample and homogenized. Residual protein was removed by the addition of 200 μL of chloroform, and centrifugation for 15 min at 12 000 g and 4°C. The aqueous phase was precipitated in 500 μL of isopropanol, incubation for 2 h at −80°C and centrifugation for 15 min at 12 000 g and 4°C. The resulting RNA pellet was washed with 700 μL ethanol (−20°C) and centrifuged for 10 min at 12 000 g and 4°C. The RNA pellet was air dried, resuspended in 20 μL of DEPC-treated water, and stored at −80°C. One microgram of total RNA was reverse transcribed using poly dT12–18 primers and Superscript II RT (50 min at 42°C, 15 min at 70°C). The amount of RNA to be used for detection was normalized using the housekeeping gene hypoxanthine phosphoribosyltransferase (HPRT) as reference. The cDNA obtained was amplified using different sets of primers Fas (5′ primer, 5′-ATGCTGTGGATCTGGGCTGTC; 3′ primer, 5′-TGTCTTCAGCAATTCTCGGAGTG), for FasL (5′ primer, 5′-ACCAACCAAAGCCTTAAA; 3′ primer, 5′-ATACTTCACTCCAGAGAT), for granzyme B (5′ primer, 5′-CTCCACGTGCTTTCACCAAA; 3′ primer, 5′-GGAAAATAGTACAGAGAGGCA), for perforin (5′ primer, 5′-TGCTACACTGCCACTCGGTCA; 3′ primer, 5′-TTGGCTACCTTGGAGTGGGAG), for TNFα (5′ primer, 5′-ATGAGCACAGAAAGCATGATC; 3′ primer, 5′-CCAAAGTAGACCTGCCCCGAC) and for IL-2 (5′ primer, 5′-TGGAGCAGCTGTTGATGGACCTAC; 3′ primer, 5′-AGATAGTGCTTTGACAGAAGGCTAATC). The PCR was done for 33 cycles of denaturation at 94°C for 1 min, annealing at 60°C for 45 s, and extension for 90 s at 72°C. Polymerase chain reaction products (10 μL) were analyzed on 1% agarose gel, and the prominent bands of the correct size were visualized with ethidium bromide staining.

Proliferation assays

In order to compare the antiproliferative properties of NIM811 to that of CsA, isolated T cells were plated in triplicate at a density of 200 000 cells per well in a 96-well tissue-culture plates. Cells were activated with anti-CD3 mAb coated to microtiter plates in the absence or presence of CsA or NIM811. After 48 h of stimulation, cultures were pulsed with 0.5 μCi/well [3H]thymidine (Amersham France SA, Les Ulis, France). Eight hours later, [3H]thymidine uptake was measured using a Packard direct counter (Packard, Meriden, CT).

Treatment by cyclosporine NIM811

Both CsA and NIM811 used in the present study were a generous gift of Novartis® (Basel, Switzerland). CsA and N-methyl-4-isoleucine-cyclosporin (called NIM811) were dissolved for in vivo use in a mixture of Cremophor EL (polyethoxylated castor oil) with ethanol-94%. NIM811 binds to the mitochondrial cyclophilin D, but does not interact with the cytosolic cyclophilin A, and thus has no known effect on the cytosolic calcineurin-NFAT signaling pathways involved in the immunosuppressive action of CsA (15).

Statistical analysis

Results are expressed as mean ± SEM. Groups were compared using one-way analysis of variance (ANOVA) with Tukey's post-hoc test. A value of p < 0.05 was considered as indicative of a statistically significant difference.


A total of 106 mice were included in the present study. The 114 mice allowed 52 transplantations and 10 mice were used as sham. Fifteen hearts were excluded: nine because of surgical problems during reperfusion, and six because of technical failure during the evaluation of the mPTP opening. Results are, thus, presented based on 37 successful transplantations and 10 sham experiments.

Acute graft rejection

Conventional histology allowed identification of acute graft rejection in our model. As depicted in Figure 1A, at postoperative day 5 (n = 4 per group), GS and GS-C3H grafts displayed moderate infiltration with a number of inflammatory cells count that averaged 82 ± 6 and 60 ± 14 cells. In contrast, allogenic hearts (GA) exhibited histological evidence of grade IIIA acute rejection (International Society for Heart and Lung Transplantation classification), with an increased number of inflammatory cells that averaged 199 ± 6 (p < 0.05 vs. GS-C3H), including lymphocytic infiltration, as well as edema, fibrosis and myocyte damage. Fas deficient allografts (GA-lpr) had comparable inflammatory cells infiltration, averaging 157 ± 16 (p = NS vs. GA). As expected, GA and GA-lpr hearts treated with NIM811 were also quoted as IIIA grade acute rejection and did not exhibit a limited inflammatory cells infiltration (Figure 1B).

Figure 1.

Acute cardiac graft rejection. (A) Mean number of inflammatory cells per high power field (HPF, ×40) in grafted hearts. *p < 0.05 versus GS-C3H. (B) Conventional histology in GS-C3H (grafted syngenic), GA (grafted allogenic), GA NIM, GA-lpr and GA-lpr NIM. None of the sham, GS or GS-C3H hearts exhibited signs of rejection at day 5. At day 5, GA and GA-lpr hearts exhibited an increased of inflammatory cells with obvious myocyte damage and destruction of cardiac muscle fibers which were not attenuated under NIM811 treatment.

Cardiac apoptosis

Caspase-3 activity:  At day 5, caspase-3 activity (n = 4 per group) was significantly increased in GA hearts, averaging 84 ± 14 pmol-AMC/min/mg versus 6 ± 2, 18 ± 8 and 16 ± 10 pmol-AMC/min/mg in sham, GS and GS-C3H hearts, respectively (Figure 2). In GA-lpr hearts, caspase-3 activity was significantly increased when compared to GS, but remained significantly lower than in GA hearts, averaging 54 ± 6 pmol-AMC/min/mg (p < 0.05 vs. GA and GS). NIM811 treatment significantly reduced caspase-3 activity in GA group, down to 20 ± 10 pmol-AMC/min/mg (p < 0.05 vs. GA). NIM811 further reduced caspase-3 activity in GA-lpr hearts that averaged 25 ± 12 pmol-AMC/min/mg (p < 0.05 vs. GA-lpr, p = NS vs. GS and GA-NIM811).

Figure 2.

Caspase-3 activity in GS, GS-C3H, GA, GA NIM, GA-lpr and GA-lpr NIM grafts. Allogeneic grafting resulted in an increased caspase-3 activity in GA and GA-lpr (p < 0.05 vs. GS). NIM811 treatment attenuated this increased activity. *p < 0.05 versus GS; p < 0.05 versus GA.

Cytochrome c release:  Western blotting indicated that cytochrome c release was increased in GA and GA-lpr hearts (Figure 3). Relative densitometry (in arbitrary units, AU) averaged 4.7 ± 0.4 in GA and 3.1 ± 0.2 in GA-lpr versus 1.8 ± 0.1 in GS heart. NIM811 treatment reduced this increase in cytochrome c release that averaged 2.3 ± 0.2 and 1.8 ± 0.3 in GA-NIM and GAlpr-NIM, respectively (p < 0.05 vs. respective untreated groups).

Figure 3.

Cytochrome c release. Bar graph indicates the mean value of the density of western blotting for each group (n = 3 per group). Cytochrome c release was increased in GA and GA-lpr heart. NIM811 reduced this increased release of cytochrome c. *p < 0.05 versus GS and GS-C3H; p < 0.05 versus respective untreated group.

Ca2+-induced mPTP opening

We assessed the susceptibility of mPTP opening to Ca2+ overload (n = 7–8 per group), (Figure 4B). At day 5, the calcium load required to open the mPTP averaged 87 ± 11 and 80 ± 15 μM Ca2+/mg mitochondrial proteins in GS and GS-C3H groups versus 97 ± 12 μM Ca2+/mg mitochondrial proteins in the sham group (p = NS). In the GA group (Figure 4A), the Ca2+ load required to trigger mPTP opening was significantly reduced to 35 ± 9 μM Ca2+/mg mitochondrial proteins (p < 0.05 vs. GS). GA-lpr hearts (n = 7) displayed similar alteration of the sensibility to Ca2+ overload than GA, with a Ca2+ load averaging 41 ± 9 μM Ca2+/mg mitochondrial proteins (p = NS vs. GA). As expected, NIM811 increased the resistance of mPTP to Ca2+ overload in GA (133 ± 17 μM Ca2+/mg mitochondrial proteins) as well GA-lpr (137 ± 10 μM Ca2+/mg mitochondrial proteins) hearts (p < 0.05 vs. GA, p = NS vs. GS).

Expression of Fas, FasL, granzyme, perforin and TNFα by RT-PCR

A common mechanism of cytotoxicity in the process of allograft rejection and induction of myocyte damage is via the secretion of FasL, granzyme, perforin and TNFα. To evaluate whether this mechanism might play a role in the induction of the observed cardiomyocyte damage, expression of these CTL effector molecules were analyzed by semiquantitative RT-PCR (Figure 5). Fas mRNA was similar in all groups, and its level of expression was not significantly altered in allografts. The expression of FasL mRNA, essentially undetectable in sham, GS and GS-C3H groups, was significantly upregulated in allografts (GA). Similarly, expression of the transcripts of granzyme, perforin and TNFα was significantly increased in cardiac allografts at day 5 (p < 0.05 vs. GS). Fas deficient (GA-lpr) hearts displayed a similar modulation of the expression of these transcripts than GA (p = NS vs. GA).

Figure 5.

Semiquantitative RT-PCR of Fas, FasL, granzyme, perforin and TNFα in grafted hearts. Fas mRNA was similar in all groups. Expression of FasL, granzyme, perforin and TNFα was upregulated in both wild-type and lpr allografts, and was significantly reduced by NIM811. *p < 0.05 versus GS and GS-C3H; p < 0.05 versus respective untreated group.

Surprisingly, NIM811 dramatically decreased the expression of FasL, granzyme, perforin and TNFα, and to a similar extent in GA and GA-lpr grafts (Figure 5).

NIM811 and cytotoxic T lymphocytes

The fact that NIM811, a specific inhibitor of mPTP opening, dramatically reduced expression of FasL, granzyme, perforin and TNFα in allografts prompted us to further analyze its effects on CTL infiltration and proliferation. NIM811-treated GA and GA-lpr hearts displayed unchanged global cellular infiltration and myocyte damage when compared to untreated corresponding groups (Figure 1B). However, in NIM811-treated hearts, we noticed a mononuclear cells substitution by histiocytes in the population of inflammatory cells (data not shown). We thus measured the number of CD8+ T lymphocytes in cardiac grafts at day 5 (Figure 6). The number of infiltrated CD8+ T lymphocytes was markedly increased in GA (171 ± 18 cells per HPF) and GA-lpr (131 ± 23 cells per HPF) when compared to GS-C3H (5 ± 2 cells per HPF). The infiltration of CD8+ into the cardiac grafts was significantly reduced by NIM811, averaging 30 ± 7 in NIM811-GA and 21 ± 8 cells per HPF in NIM811-GA-lpr hearts (Figure 6).

Figure 6.

Infiltrated CD8+ lymphocytes in grafted hearts. The number of infiltrated CD8+ lymphocytes, per high power field (HPF) was markedly increased when compared to GS-C3H. Fas deficient (GA-lpr) hearts displayed comparable infiltration with CD8+ T lymphocytes than GA hearts. Infiltration of CD8+ into the cardiac grafts was significantly reduced by NIM811. *p < 0.05 versus GS-C3H. p < 0.05 versus respective untreated groups.

Since IL-2 can be considered as a marker of CD4+ T lymphocytes activity, we measured IL-2 mRNA levels by RT-PCR (Figure 7). Expression of IL-2 transcript was significantly upregulated in GA hearts (p < 0.05 vs. GS), and NIM811 treatment did not alter this upregulation of IL-2 (p = NS vs. GA). We eventually addressed whether NIM811 might modify T-cell proliferation when compared to cyclosporin A. T-cell proliferation was estimated in vitro as the [3H]thymidine incorporation after antigen stimulation. CsA treatment strongly inhibited T-cell proliferation at the low dose of 0.1 μg/mL. In contrast, NIM811 showed no significant effect at 0.1 μg/mL and even at 1 μg/mL. Only at the very high dose of 10 μg/mL, did NIM811 significantly reduce T-cell proliferation, yet to a lower extent than CsA.

Figure 7.

Semiquantitative RT-PCR of IL-2. IL-2 was significantly upregulated in both GA untreated and GA-NIM. *p < 0.05 versus GS.


We report here that mPTP inhibition limited apoptosis associated with ACGR more efficiently than Fas inactivation. This protective effect may be related to a direct effect on cardiomyocyte apoptosis and/or to the rerouting of the classical acute graft rejection pathway involving CD8+ T lymphocytes.

Classical activation of graft damage during acute rejection

Cardiomyocyte injury following acute rejection involves T-cell-mediated cytotoxicity (1). At day 5, we observed a rapid infiltration of lymphocytes, plasmocytes and macrophages into allogenic transplants, with surrounding edema, fibrosis and myocyte damage. Studies have determined that cytotoxicity can be mediated through several major molecular pathways, mainly the granule exocytosis or perforin-granzyme pathway, and the Fas (Apo-1, CD95)/Fas ligand (FasL, CD95L)-mediated apoptosis pathway (6,16,17). In our model, expression of perforin, granzyme B and TNFα was clearly upregulated in allogenic grafts, and concomitant with an infiltration of CD8+ T cells.

Activation of the Fas/FasL system can trigger apoptosis in various cell types, including cardiomyocytes (4). Expression of FasL, essentially undetectable in syngenic groups, was strongly upregulated in allografts and correlated with a significantly increased cytochrome c release and caspase-3 activity, a pivotal effector caspase in the apoptotic program of cell death (18). In order to investigate the role of Fas in cardiomyocyte injury during acute graft rejection, we used lpr mice that lack a functional Fas receptor (19). Lpr allografts did not develop different myocardial damage when compared to wild-type allografts, in agreement with Djamali et al. who demonstrated that lpr allografts are often rejected in a near-normal time course (20,21). The increase of caspase-3 activation in these Fas-deficient transplants remained lesser than in wild-type grafts, confirming a nonnegligible role of the Fas/FasL pathway in cardiomyocyte apoptosis activation in this experimental model (22,23). Yet, the fact that NIM811 equally reduced caspase-3 activity and cytochrome c release in allogenic and Fas-deficient grafts indirectly suggests that Fas-induced apoptosis may possibly act via a cross-talk with mitochondria through caspase-8 activation.

Alternative pathway for apoptosis activation in myocardial transplant

Mitochondrial permeability transition is recognized as a key event in cardiomyocyte death after ischemia-reperfusion (24–26). We previously demonstrated that mPTP is also involved in apoptosis during acute graft rejection (3). Opening of the mPTP results in the collapse of the membrane potential and efflux of cytochrome c and downstream effector caspases activation that may lead to either apoptosis or necrosis (27).

Acute graft rejection causes accumulation of cytosolic Ca2+ through activation of IP3 pathway (28). Accumulated cytosolic Ca2+ enters the mitochondria through the Ca2+ uniporter, and high matrix Ca2+ concentrations provide an ideal condition for mPTP opening (29). We demonstrated here that mitochondria isolated from allogenic grafts display a significant reduction in their resistance to Ca2+ overload. Involvement of mPTP opening in cardiomyocyte death was strongly suggested by the fact that the specific inhibitor NIM811 clearly inhibited Ca2+-induced mPTP opening and significantly reduced cytochrome c release and caspase-3 activity. We then first interpreted that inhibition of mPTP opening in cardiomyocytes limited cardiac apoptosis during acute rejection.

We were, however, surprised to observe that NIM811 dramatically modified the expression of FasL, perforin, granzyme B and TNFα, that represent major CTL effector molecules. NIM811 is known as a nonimmunosuppressive derivative ([Me-Ile4]-cyclosporine) of CsA. Several reports indicate that this CsA analogue does not inhibit the calcium/calmodulin-dependent protein phosphatase, calcineurin, and does not inhibit NFAT migration into the nucleus (30). Our results are in agreement with those of Rosenwirth et al., who demonstrated that NIM811 does not alter IL-2 up-regulation into allogenic transplant and consequently, does not alter CD4+ proliferation (31). Indeed, allogenic grafts that received NIM811 did not exhibit an inhibition of acute rejection. Despite this, we observed that infiltration of CD8+ T cells into allogenic grafts was significantly reduced by NIM811. Expression of the CTL effector molecules (perforin, granzyme, FasL and TNFα) was consequently markedly reduced by NIM811, and in vitro dose-response of NIM811 on T-cells proliferation was in agreement with in vivo results. In other words, the specific mPTP inhibitor NIM811 inhibited CD8+ (but not CD4+) T-cells proliferation and activation, and the protective effect of NIM811 might not be solely due to inhibition of mPTP in cardiomyocytes.

What is the link between mitochondrial permeability transition and CD8+ T-cell activation? One may first hypothesize that cardiomyocyte apoptosis may stimulate CD8+ T lymphocyte infiltration in the rejecting graft. Faouzi et al. reported that caspase-3 activation promotes inflammatory cells attraction through chemokines induction in mice liver (32). Using a murine model of renal ischemia-reperfusion injury, Daemen et al. demonstrated that caspase inactivators not only limited apoptosis but also prevented inflammation and organ injury (33). NIM811 might therefore attenuate T-cell infiltration during acute graft rejection via a reduction of cardiomyocyte apoptosis secondary to the inhibition of mPTP opening and subsequent cytochrome c release and caspase3 activation.

Second, Ca2+, that is essential for regulating a host of distinct processes involving enzyme control, exocytosis, gene regulation, cell growth and proliferation and apoptosis, may play an important role. Lichtman et al. demonstrated that calcium is required for the stimulation of lymphocytes (34). Calcium enters the mitochondria as a result of a respiration-dependent mitochondrial membrane potential. Mitochondria are Ca2+-excitable organelles involved in the transduction of cell Ca2+ signals, and seem to play a major part in the regulation of regulation of Ca2+-release-activated Ca2+ current (ICRAC) (11,35). Calcium release from the mitochondria may also occur via opening of the mitochondrial transition pore, under either a low-conductance or a high-conductance state (11,36,37). In light of our present results, and in agreement with Hoth et al. (38), we can hypothesize that mPTP opening may contribute to increase, or maintain at a high level, cytosolic Ca2+ concentration, and thereby favor CD8+ T-cell activation.

The present data confirm that cardiomyocyte apoptosis is associated with acute graft rejection. Although cardiomyocyte apoptosis may be induced by different pathways, mitochondria appear to be a predominant player. Inhibition of mitochondrial permeability transition may, in the case of ACGR, act on two different targets: first, CD8+ T lymphocyte infiltration via either reduced local apoptosis or limited calcium release in the lymphocyte cytosol; second, cardiac apoptosis via mPTP inhibition within jeopardized cardiomyocytes.