Lysis of Adult Ventricular Myocytes by Cells Infiltrating Rejecting Murine Cardiac Allografts
Background Immunologic mechanisms that mediate myocardial cell injury during rejection are not fully understood. We therefore investigated whether cells that infiltrate rejecting cardiac allografts are capable of directly injuring myocytes and whether this injury resembles that produced by cytotoxic T lymphocytes (CTLs) that are generated in a mixed lymphocyte reaction (MLR).
Methods and Results Heart-infiltrating cells (HICs) were isolated from murine heterotopic BALB/c cardiac allografts undergoing rejection 6 to 8 days after transplantation into C57BL/6 mice. An in vitro model system of cultured adult murine ventricular myocytes was developed to facilitate investigation of cell-mediated myocyte injury. Isolated adult myocytes were incubated with either HICs or MLR effector cells, and myocyte death was quantified by counting the number of rod-shaped myocytes excluding trypan blue. The frequency of donor-reactive CTLs was similar in the HIC and MLR populations, as assessed by limiting dilution analysis. However, HICs were less efficient at killing donor-strain myocytes than were MLR cells. CTL-mediated cell lysis occurred by 6 hours, whereas myocyte injury produced by HICs was more gradual, with considerable cytotoxicity occurring between 12 and 24 hours. Furthermore, whereas MLR cells lysed only donor-strain myocytes, HIC lysed donor, third-party, and syngeneic myocytes. Treatment of MLR cells and HICs with anti-CD8 antibody plus complement produced a much greater inhibition of MLR cytotoxicity than of HIC cytotoxicity.
Conclusions These data demonstrate that only a small component of myocyte injury mediated by allograft-infiltrating cells can be ascribed to CTLs within the infiltrating cell population. These findings suggest that cell types associated with a delayed-type hypersensitivity response, as well as CTLs, cause myocyte injury during cardiac rejection.
Cardiac transplantation has been established as an effective therapy for end-stage cardiac disease.1 Improved immunosuppressive regimens have resulted in significant prolongation of survival, from 50% at 1 year in 1981 to >90% in some series in more recent years.2 However, despite improvements in immunosuppression, acute and chronic rejection continue to be among the major causes of morbidity and mortality after transplantation.
Many investigators consider parenchymal cell injury by CTLs to be a major contributor to acute allograft rejection.3 4 5 6 7 For example, Frisman et al7 found that the cytotoxic activity of the IL-2–responsive lymphocytes in endomyocardial human biopsy samples was closely related to clinical rejection. Similarly, Sell et al5 documented the presence of CTLs in cardiac biopsy specimens of patients after transplantation. However, work by other investigators8 9 10 has suggested that CD4+ cells may be more important in allograft rejection than CD8+ cells. The mechanism by which CD4+ cells assist in allograft rejection is not clearly understood but may be via cytokine production with activation of macrophages that results in a DTH response.9 11
The target cell used in many investigations of cytotoxicity has been the lymphoblast.8 9 Since the myocyte is an important target cell in allograft rejection in vivo, a more appropriate target cell for in vitro investigations would be isolated myocytes. We have reported previously12 that CTLs generated in an allogeneic MLR lyse cultured fetal murine myocytes in an alloantigen-specific fashion. In this system, fetal myocytes are killed within 3 to 5 hours (determined by 51Cr release), and myocyte injury is prevented by depleting CD8+ but not CD4+ T cells. These studies, along with those of other investigators, suggest that “classic” CD8+ CTLs could be the principal mediators of myocyte lysis during cardiac transplant rejection.13 However, MLR cells are stimulated in vitro and may not accurately reflect the mechanism of transplant rejection in vivo. In addition, injury of fetal myocytes produced by various effector cells may differ from injury of adult myocytes.
The present study addresses these issues by comparing adult and fetal myocyte injury induced by CTLs produced in an MLR and by HICs freshly isolated from rejecting cardiac allografts. Our results indicate that HICs induce adult ventricular myocyte damage via mechanisms distinct from those attributable to CTLs.
BALB/c (H-2d), C3H/HeN (H-2k), and C57BL/6 (H-2b) mice were obtained from the National Cancer Institute-Frederick (Md) Cancer Research and Development Center and used when they were between 6 and 12 weeks of age. Animal care was in accordance with NIH guidelines, and all experiments were approved by the University of Utah Animal Care and Use Committee.
Heterotopic cardiac transplantation in mice was performed as described by Shelby and Corry.14 Mice were anesthetized with 0.1 mL 3.6% solution of chloral hydrate per 10 g body weight. The abdomen of the recipient animal (C57BL/6) was incised, and the infrarenal abdominal aorta and vena cava were dissected free for a length of ≈2 mm. After anesthesia was administered, a midline incision was made in the donor animal (BALB/c) and 1 mL heparin (200 U/mL) was injected into the inferior vena cava. The donor heart was removed, and the inferior and superior venae cavae were ligated and divided. The aorta and pulmonary artery were divided, and the pulmonary veins were ligated. The donor heart was placed briefly in a cooled, lactated Ringer’s solution, then sutured in the abdomen of the recipient by joining the donor ascending aorta to the recipient abdominal aorta and the donor pulmonary artery to the recipient inferior vena cava in an end-to-side fashion with 10-0 nylon suture material. The success rate with this technique was >90%. In this model, the transplanted heart is perfused with the recipient mouse’s blood and resumes contractions. Rejection begins in this strain combination within ≈4 to 6 days,9 with evidence of severe histological rejection and complete loss of contractile function and myocyte necrosis by 12 days.15
Generation of MLR
As previously detailed,12 spleens were obtained from C57BL/6 mice 6 to 9 days after transplantation with BALB/c heterotopic cardiac allografts and processed into single-cell suspensions. Responder splenocytes (1×106/mL) were cocultured at 37°C and 5% CO2 with irradiated (3300 R) BALB/c splenocytes (1×106/mL) for 6 to 8 days in 25-mL volumes of RPMI-1640 medium supplemented with 10% fetal bovine serum (HyClone), 1 mmol/L sodium pyruvate, 2 mmol/L glutamine, 0.05 mmol/L 2-mercaptoethanol, 0.1 mmol/L nonessential amino acids, and penicillin and streptomycin antibiotics, as described by Lynch et al.13 This process, which includes in vivo and in vitro allostimulation of lymphocytes, produces an MLR in which the effector cells causing myocyte lysis are primary CTLs.12
Nonsensitized (Control) Lymphocytes
Spleens were obtained from 2 or 3 nontransplanted C57BL/6 (recipient strain) mice and processed into a single-cell suspension. The cell suspension was resuspended in RPMI-1640 medium and cultured as described above but without exposure to irradiated donor-strain lymphocytes.
Isolation of HICs
Heterotopic BALB/c cardiac allografts from 5 to 8 C57BL/6 mice transplanted 6 to 8 days previously were removed, pooled, and minced. The tissue suspension was serially digested four times with 1 mg/mL collagenase A (Boehringer Mannheim Biochemicals) for 20 minutes for each digestion at 37°C. After each digestion, the supernatant suspension containing HICs was removed with a pipette and placed in medium containing 10% fetal bovine serum to halt digestion. After the completion of four digestions, the red blood cells were lysed with sterile H2O. HICs were washed and resuspended in myocyte culture media. Viable leukocytes were identified by trypan blue exclusion and counted on a hemocytometer.
Fetal Ventricular Myocyte Culture
The fetal myocyte culture technique has been reported in detail previously.13 Fetal hearts were removed, minced, and serially digested with collagenase (75 U/mL) (Worthington Biochemical Corp) at 37°C. Myocytes were washed, suspended in culture medium, and plated in 96-well microtiter tissue culture dishes (1.5×105 cells/well) (Corning Glass Inc). Spontaneous contractions developed by day 2 of culture, and cells spread into a confluent layer that contracted synchronously by 4 to 5 days.
For detection of cell injury, cultures of fetal ventricular myocytes were labeled with 51Cr at a concentration of 2 μCi per well for 1 hour. Cells were washed and covered with fresh medium before lymphocytes were added to the myocytes at various E/T ratios. At the end of 6 hours, supernatants were collected, and total released 51Cr counts were measured on a gamma counter (Micro-Medic Systems, Inc). Cell lysis was calculated as percent 51Cr release=(51Cr release in each treatment well−spontaneously released counts)/(51Cr release in wells treated with NP-40 detergent−spontaneously released counts).16 17
Dissociation and Culture of Adult Mouse Ventricular Myocytes
Adult mouse myocyte isolation was performed with a modification of the method of Benndorf et al.18 Individual hearts were removed from anesthetized mice and immediately attached to an aortic cannula that provided continuous retrograde coronary artery perfusion at 37°C by a pump (Minipuls 2, Gilson Instrument Co) at a coronary perfusion pressure of 70 to 90 mm Hg and a flow rate of 1.8 mL/min. Sterile conditions were maintained, and the heart was perfused with Ca2+-free modified Tyrode’s bicarbonate buffer solution for 5 minutes, immediately followed by 12 minutes of perfusion with the same solution containing 0.5 mg/mL collagenase A (Boehringer Mannheim Biochemicals). Both cell isolation solutions contained (in mmol/L): NaCl 126, KCl 4.4, MgCl2 1.0, NaHCO3 18, glucose 11, HEPES 4, butanedione monoxime 30, and 0.13 U/mL insulin and were gassed with 5% CO2/95% O2 (pH 7.40). The heart was detached from the cannula, and ventricles were cut into small pieces in the same solution. Butanedione monoxime inhibits injury due to cutting and elevation of [Ca2+]i19 and enhances survival of these myocytes, which are prone to Ca2+ overload.
The tissue was separated by bubbling with 5% CO2/95% O2 for approximately 2 to 3 minutes, followed by gentle pipetting. The resulting cell suspension was pipetted into the same solution containing 50 μmol/L CaCl2 and 2% albumin. After 15 minutes of incubation at 37°C, the cell suspension was centrifuged at 300 rpm, and the supernatant was discarded. Cells were resuspended in a similar solution with 200 μmol/L CaCl2 and 2% albumin and incubated at 37°C for 30 minutes. Cells were centrifuged at 300 rpm, and the supernatant was discarded. Cells were then resuspended in myocyte culture medium.
Two different media were used in these studies to culture adult ventricular myocytes. Most experiments were performed in media composed of 5% heat-inactivated fetal bovine serum (HyClone), 47.5% MEM (Gibco Laboratories), 47.5% modified Tyrode’s bicarbonate-buffered balanced salt solution,12 0.1% penicillin-streptomycin, 10.0 mmol/L pyruvic acid, 4.0 mmol/L HEPES, and 6.1 mmol/L glucose. Antibody deletion experiments were performed with myocytes cultured in the same medium but without serum. No differences in responses of myocytes to effector cells in these different media were noted.
Measurement of Lysis of Adult Ventricular Myocytes
Suspensions of MLR cells, HICs, or NSLs (all 1.2×106/300 μL) or RPMI-1640 medium without added cells were added to 300 μL of cultured adult myocytes in 2-mL tubes. The concentration of rod-shaped adult myocytes was adjusted to give the desired E/T ratio. An additional 150 μL of RPMI-1640 medium per tube was added. Myocytes and effector cells settled to the bottom of the tube by gravity. After incubation in a 5% CO2/95% air atmosphere at 37°C for a predetermined amount of time, 625 μL supernatant was removed from each tube and 25 μL trypan blue was added. The number of viable myocytes was determined by counting non–trypan blue–stained rod-shaped cells with clear cross-striations in 25 μL cell suspension in a modified hemocytometer. The relative survival of myocytes incubated with MLR cells, HICs, and NSLs was determined by dividing the average number of viable myocytes per 25-μL sample in each tube incubated with each effector by the average number of viable myocytes per 25-μL sample in the tubes containing media alone. Percent survival was calculated as (number of viable myocytes incubated with lymphocyte preparation)/(number of viable myocytes incubated with medium)×100. Measurements in each tube of effector cells were performed in quadruplicate, and the average of the four values was determined.
LDA of Alloantigen-Reactive CTL
As described by Orosz et al,20 modified LDA techniques were used to quantify in vivo–stimulated donor alloantigen-reactive CTLs in MLR and HIC preparations. These in vivo–stimulated CTLs are referred to as “antigen-conditioned,” or cCTL. In addition, conventional LDA techniques were used to quantify tCTLs with the potential to respond to donor alloantigens.
To quantify tCTLs, appropriate dilutions of responder cells were added to round-bottomed microtiter plates along with 5×105 irradiated (2000 R) donor-strain splenocytes per well plus 10% EL4 supernatant in DMEM complete medium (adjusted to contain 10% fetal bovine serum). EL4 supernatant was used as a source of exogenous growth factor and was prepared as described by Farrar et al.21 Cultures were incubated for 7 days at 37°C in 10% CO2. To detect cytolytic activity, 50 μL (5×103 cells) target cells was added to each microwell. Target cells were prepared by incubating splenocytes with 1 μg/mL concanavalin A for 2 to 3 days, followed by incubation for 90 to 120 minutes in 51Cr (500 μCi/107 cells). After a 4-hour incubation, 150 μL supernatant was removed from each well. Supernatants were assayed for released 51Cr in a gamma counter. Microcultures were considered cytolytic if observed chromium release was greater (mean±SD) than the chromium release observed in wells that contained target cells, LDA stimulator cells, and EL4 but no responder cells.
To selectively quantify in vivo–stimulated cCTLs, the same culture conditions were used as described above, except that 5×105 irradiated (2000 R) syngeneic splenocytes were used in place of the allogeneic splenic stimulator cells. Since no stimulating alloantigens were present in the modified LDA microcultures, only those CTLs that had received an allogeneic stimulus before the analysis could proliferate and develop LDA–detectable cytolytic activity.20 Alloantigen specificity of these CTLs is defined by their ability to lyse only those 51Cr-labeled allogeneic target cells bearing graft alloantigens. Thus, the modified LDA technique detects only cCTLs that have been activated by alloantigen and not unstimulated precursor CTLs.
Minimal estimates of CTL frequency were obtained according to the Poisson distribution equation as the slope of a line relating the number of responder cells per microwell (plotted on a linear x axis) and the percentage of microwells that failed to develop cytolytic activity (plotted on a logarithmic y axis).22 23 The slope of this regression line was determined by computer by use of χ2 minimization analysis, as described by Taswell.24 This analysis yields minimal frequency estimate, 95% confidence interval of the frequency estimate, and a χ2 estimate of probability. Frequency estimates with overlapping confidence intervals are not statistically different.
Depletion of CD8+ Cells
MLR and HIC preparations were obtained as previously described, and 300 μL MLR cell or HIC suspensions were added to myocytes. Ascites fluid containing anti-CD8 monoclonal antibody (hybridoma cell line 2.43) was diluted to 1:103, and 240 μL was added to the reaction mixture. Rabbit serum complement (Low-Tox-M Rabbit Complement, Accurate Chemical & Scientific Corp) was reconstituted in 0.67 mL ice-cold medium and 60 μL added to each of the reaction mixtures to make a final complement concentration of 1:10. Control suspensions of HICs or MLR cells plus myocytes were treated with 300 μL of serum-free medium or with medium containing anti-CD8 antibody plus complement. The reaction mixtures were cultured in a 95% air/5% CO2 atmosphere at 37°C for 24 hours before determination of myocyte lysis.
Antibody-mediated and complement-mediated lysis of CTLs was carried out directly in the reaction suspension, rather than in the standard two-step process in multiwell plates previously described by our laboratory,12 because the washing and incubation steps involved appeared to diminish somewhat the cytotoxicity caused by HICs and decreased cell recovery.
Survival of fetal ventricular myocytes cultured with MLR cells, HICs, and NSLs from experiments performed on different days was compared (mean±SEM) by use of Student’s paired t tests. Survival of adult ventricular myocytes cultured with MLR cells, HICs, and NSLs from experiments performed on different days was compared (mean±SEM) by use of Student’s paired t tests and ANOVA.
Phenotypic and Functional Analysis of HICs and MLR Cells
We initially examined the cell types present in HIC and MLR populations. As expected, an analysis of centrifuge preparations stained with Wright’s stain revealed that lymphocytes were the primary cell type in the MLR population (81%). Although the HIC population also contained a large percentage of lymphocytes (44%), numerous macrophages (30%) and a significant number of polymorphonuclear leukocytes (20%) also were present.
The frequencies of donor-reactive CTLs present in HIC and MLR cell populations were determined by LDA. Specifically, modified LDA was used to quantify cCTL, which had been stimulated previously by alloantigen, as well as the total population of CTLs that had the potential to respond to donor alloantigens.20 The Table⇓ demonstrates that the frequencies of tCTLs capable of responding to donor alloantigens were not different between HIC (1/93) and MLR (1/72) populations. Furthermore, a significant number of CTLs in both populations were stimulated cCTLs as assessed by modified LDA. Thus, we anticipated that HICs would have a cytotoxic effect similar to that of MLR cells if the offending cell type in HICs is the CTL and if these CTLs were equally cytotoxic compared with CTLs in MLR.
Cytotoxicity of MLR Cells and HICs Against Fetal Myocytes
We initially compared the effects of HICs and MLR cells with fetal mouse myocytes; results are shown in Fig 1⇓. 51Cr release from cultured fetal mouse myocytes coincubated for 6 hours with either HICs or MLR cells at various E/T ratios differed. In a series of six separate experiments, MLR cells were quite cytotoxic and caused 79% 51Cr release at an E/T ratio of 12:1. This result is consistent with our previous work.13 25 However, HICs produced <30% 51Cr release even at much higher E/T ratios of up to 30:1.
The results of these experiments suggested that HICs conditioned in vivo that have infiltrated the target end-organ allograft are significantly less cytotoxic against myocytes than MLR cells prepared in vitro. However, the 51Cr assay is an indirect measure of cytotoxicity and, because of spontaneous release of 51Cr and reuptake of released 51Cr by effector cells, cannot be used reliably to quantify target-cell injury over long periods of time or at high E/T ratios. In addition, myocytes constitute only ≈50% of the cell population present in the culture,25 and the remaining cells include fibroblasts and vascular cells. Neonatal or fetal myocytes also may be a less appropriate target for this cytotoxicity assay than adult murine myocytes with respect to human in vivo transplant rejection. We therefore attempted to develop a model that used isolated adult murine ventricular myocytes as targets for a cytotoxicity assay.
Differences in Ability of HICs and MLR Cells to Lyse Adult Mouse Ventricular Myocytes
Isolated adult murine ventricular myocytes could be maintained in culture for up to 72 hours at 37°C. Viable ventricular myocytes are rod-shaped with clear cross-striations and exclude the vital dye trypan blue (Fig 2⇓). In contrast, lysed ventricular myocytes lose this rod-shaped morphology and are stained with trypan blue. Hence, by counting the number of unstained, rectangular myocytes, we quantified the survival of myocytes exposed to MLR cells (Figs 2⇓ and 3⇓). Whereas a significant rate of “spontaneous” cell death occurred over 36 hours when myocytes were cultured with medium or NSLs, almost all myocytes were killed when exposed to MLR cells. MLR-mediated myocyte death was rapid and was virtually complete by 6 hours. This time course of lysis of adult myocytes is comparable to 51Cr release by fetal murine myocytes exposed to MLR cells.12 Thus, the use of adult myocytes as target cells to evaluate the cytotoxicity of MLR cells yielded results very similar to those obtained previously by use of fetal cultured myocytes. We next used this approach to compare the time course and magnitude of cytotoxicity caused by MLR and HIC cells.
Cytotoxicity of HICs Against Donor-Strain Adult Myocytes
We first analyzed the cytotoxic effects of HICs and MLR cells against donor-strain myocytes at varying E/T ratios (Fig 4A⇓). Results were similar to those obtained when 51Cr-loaded fetal ventricular myocytes were used as targets (Fig 1⇑). However, we were able to examine much higher E/T ratios in the assay with adult ventricular myocytes because the adult myocyte cytotoxicity assay did not rely on 51Cr release. The cytotoxicity of MLR cells against adult myocytes was significantly greater than that of HICs at E/T ratios up to 125:1.
In a series of 10 separate experiments at an E/T ratio of 140:1, we found that HICs lysed donor-strain adult ventricular myocytes (41.6% survival at 12 hours) but not with the same intensity as MLR cells (7.1% survival at 12 hours) (see Fig 4B⇑). Thus, results obtained both with fetal and with adult donor-strain myocytes as targets cells indicated that the intensity of cell injury produced by HICs was less than that produced by MLR cells.
Effects of Collagenase Treatment of MLR Cells
We postulated that perhaps the process of harvesting HICs from allografts could interfere with the cytotoxic effect of CTLs in the HIC preparation by alteration of adhesion molecules or by a direct effect on cytotoxic activity, thus explaining the differences mentioned above. To evaluate this possibility, we exposed MLR cells prepared in the usual fashion to the same procedure used to harvest HICs, adding collagenase A and minced myocardium to the MLR preparation and incubating the cells. Comparisons were made of cytotoxicity caused by MLR cells exposed to collagenase and that caused by MLR cells prepared in the usual fashion. In a series of four separate experiments that used an E/T ratio of 150:1, no difference in cytotoxicity was observed; the survival of myocytes incubated with MLR cells at 12 hours was 4.4±1.3% compared with 12.1±3.7% for myocytes incubated with MLR cells exposed to collagenase. Thus, it does not appear that collagenase treatment used in the preparation of HICs alters adhesion molecules or other cell components that are necessary for target-cell interaction.
Time Course of Cytotoxicity Induced by HICs and MLR Cells
The time course of lysis of target cells by different effector cells may vary. Therefore, we examined the relative time courses of cytotoxicity induced by HICs and by MLR cells. Using data from a representative experiment, we plotted the cumulative cytotoxicity of HICs and MLR cells over time (Fig 5⇓). Most of the cytotoxic effect of MLR cells occurred in the first 6 hours, whereas the cytotoxic effect of HICs was more gradual and continuous, with a considerable cytotoxic effect occurring between 12 and 24 hours. These findings suggest that the mechanism as well as the intensity of cytotoxicity produced by HICs is different from that of MLR cells.
Alloantigen Specificity of Cytotoxicity Induced by HICs and MLR Cells
Because injury produced by CTLs is known to be alloantigen specific, we next compared the specificity of injury produced by HICs and MLR cells. Experiments were performed simultaneously on myocytes isolated from donor strain (BALB/c), recipient strain (syngeneic; C57BL/6), and unrelated third-party strain (C3Hf/HeN). Results of six separate experiments are shown in Fig 6⇓. Average E/T ratios for each strain were as follows: donor strain (BALB/c), 137:1; recipient strain (syngeneic; C57BL/6), 325:1; and third-party strain (C3Hf/HeN), 160:1. Whereas MLR cells were cytotoxic against only the donor strain and thus produced alloantigen-specific injury, HICs were cytotoxic against donor-strain, syngeneic, and third-party myocytes. Thus, the alloantigen specificity, as well as the intensity and time course of cytotoxicity of HICs, differs from that of MLR cells. Therefore, it seemed unlikely that injury of myocytes produced by HICs was due only to the CTL component.
Effects of Exposure of MLR and HIC Populations to Anti-CD8 Antibody Plus Complement
In these experiments, depletion of the CD8+ cells (CTLs) in the MLR and HIC populations was produced by exposure of cell suspensions to anti-CD8 antibody plus complement, as described in “Methods.” As shown in Fig 7A⇓, treatment of MLR cultures with anti-CD8 antibody plus complement caused almost complete inhibition of cytotoxicity. This finding is consistent with our previous results,12 in which the effects of CD8+ cell depletion on MLR killing of fetal cultured myocytes were investigated. Treatment of HICs with anti-CD8 antibody plus complement also caused a consistent and statistically significant decrease in cytotoxicity. However, the magnitude of the effect of treatment with anti-CD8 antibody plus complement on HIC cytotoxicity (9.23±1.89% increase in survival) was significantly less than on MLR cytotoxicity (59.4±3.71% increase, P<.0001). Treatment of MLR cells with anti-CD8 antibody alone had a partial inhibitory effect on cytotoxicity, whereas complement alone had no effect (data not shown).
Use of Adult Myocytes as Target Cells for Cytotoxicity Studies
Effector mechanisms in allograft rejection remain poorly understood. Traditionally, most investigators have believed that cardiac allograft rejection is mediated by CD8+ CTLs and directed toward donor vascular (endothelial) cells and parenchymal cells (myocytes).26 However, the target cell used in most studies of CTL–mediated cytotoxicity has been the lymphoblast, and thus, results of these studies may not be applicable to myocyte cytotoxicity. For example, Willebrand et al27 demonstrated that parenchymal cells may be more resistant to CTL-mediated injury than lymphoblasts. In addition, expression of major histocompatibility complex class I antigens may be relatively low in cardiac myocytes.28 To understand the mechanisms of contractile dysfunction and myocyte necrosis during allograft rejection, it is therefore important to consider the impact of putative effector cells on myocyte target cells.
Our previous studies12 25 that used cultured fetal ventricular myocytes as target cells demonstrated that CTLs generated in an MLR can cause myocyte damage as measured by contractile abnormalities and 51Cr release. This model, however, has several limitations, as previously discussed. Mann et al29 recently reported successful use of cultured feline adult myocytes in an assay to detect catecholamine-induced cytotoxicity. Their results prompted us to examine immune effector cell cytotoxicity against adult ventricular myocyte target cells. This approach has several advantages. First, adult murine ventricular myocytes are the cell type being injured in vivo during transplant rejection in this model. Neonatal and fetal myocytes are immature and may respond differently to effector cell cytotoxic mechanisms. Furthermore, the cytotoxic effects on adult myocytes can be measured at prolonged time points, and the direct method of quantifying non–trypan blue–stained rod-shaped cells allows a relatively unambiguous assay of cell death. Although dissociation and culture of murine adult ventricular myocytes proved technically difficult, the present study suggests that the cultured adult ventricular myocyte model for assessing cytotoxicity in allograft rejection has significant advantages over the fetal myocyte model.
Characteristics of HIC- and MLR Cell–Mediated Cytotoxicity
Both by LDA and by cytospin analysis, the HIC preparation contains a sizable fraction of lymphocytes, a significant portion of which are alloreactive CTLs. Indeed, by LDA, the frequency of cCTLs in the HIC preparation did not differ significantly from that in the MLR preparation (Table⇑). However, experiments that used either fetal or adult myocytes as target cells indicated that the cytotoxicity produced by HICs is not as marked as that produced by MLR cells, develops more slowly, and is not allospecific. These findings strongly suggest that CTLs in the HIC population are not as cytotoxic as CTLs in the MLR and that cell types other than allosensitized CTLs in the HIC population cause direct myocyte injury during cardiac allograft rejection.
These results are consistent with the recent report by Bishop et al,9 who found that whereas in vivo depletion of CD8+ cells in the recipient animal virtually eliminated CTLs from the infiltrating cell population in this murine model, rejection of the allograft occurred in a normal fashion. Depletion of CD4+ cells inhibited HTL as well as CTL infiltration, and no histological evidence of tissue damage occurred. More recent work by this group30 has shown that unmodified cardiac rejection is associated with increased myocardial levels of mRNA for Th1-type cytokines, IL-2, and interferon gamma. In hearts rejected by CD8+-depleted recipients, a more prominent eosinophil infiltrate was noted, and tissue levels of mRNA for Th2 cytokines IL-4, IL-5, and IL-10 were elevated as well. Previous studies from our laboratory12 have shown that CD4+ cells have no direct cytotoxic effect on myocytes. Therefore, these results indicate that CD4+ HTLs mediate cardiac allograft injury indirectly and probably via production of Th1-type cytokines, which are important in promoting both CTL-mediated and DTH-mediated parenchymal cell injury.30 Because allograft rejection clearly can occur after CD8+ cell depletion, additional pathways for myocyte injury, supported by Th1-type and perhaps Th2-type cytokine production by HICs, presumably exist. Thus, considerable evidence exists to support the hypothesis that cells other than CD8+ CTLs are important effector and/or regulatory cells responsible for parenchymal cell injury of the rejecting transplanted cardiac allograft.
Our results documenting the effects of CD8-antibody treatment of the HIC population support this idea. A small portion of myocyte cytotoxicity caused by HICs could be prevented by CTL depletion (Fig 7⇑), which is consistent with previously demonstrated cytolytic potential of CD8+ cells in the HIC population against lymphoblast target cells.9 However, a much greater degree of cytotoxicity was caused by non–CD8+ cells in HIC compared with MLR populations.
Cells in HIC Population With Cytolytic Potential
The types and functions of cells infiltrating rejecting allografts have been investigated by a number of laboratories.31 32 33 Ascher34 found that studies of the in vitro functions of infiltrating cells from a rejecting allograft may mimic the in vivo activity of such cells. Infiltrating cells from rejecting rat cardiac allografts include macrophages, T lymphocytes, B lymphocytes, neutrophils, basophils, and eosinophils.30 33 35 Likewise, the cells observed in our HIC population consist primarily of lymphocytes, macrophages, and polymorphonuclear leukocytes.36
Cell types other than CD8+ CTLs present in the HIC population that may cause myocyte dysfunction include CD4+ lymphocytes, macrophages, neutrophils, and eosinophils. As mentioned, we have shown previously that CD4+ HTLs present in an MLR do not induce lysis of cultured fetal myocytes,12 but CD4+ lymphocytes produce cytokines such as interferon gamma that can cause macrophage activation.36 Activated macrophages stimulate acute inflammation through mediators such as platelet-activating factor, prostaglandins, and leukotrienes, all of which can injure even normal cells in their vicinity.37 38 39 Cytokines secreted by macrophages, such as tumor necrosis factor, IL-1, and IL-6, augment the actions of T cells and endothelial cells and may have direct effects on myocytes.36 Macrophages may also injure cells by production of free radicals.40 Work by Christmas and MacPherson41 42 demonstrated that macrophages infiltrating a rejecting rat allograft did not cause direct neonatal cardiac myocyte lysis as detected by 51Cr release but were able to inhibit spontaneous contractions of myocytes. Similarly, Strom et al31 found that macrophages obtained from rejecting hearts have a relatively small cytolytic effect. These results suggest that macrophages are not the cause of direct myocyte lysis. However, more recent preliminary work by Pinsky et al43 indicated that a macrophage cell line may cause lysis of isolated adult rat ventricular myocytes in vitro.
Other cells in the HIC population that may injure isolated myocytes are polymorphonuclear leukocytes. Neutrophils are widely recognized as important mediators of tissue injury in inflammation.44 Neutrophil activation develops as a result of recognition of “foreign” antigen, such as that from microorganisms, but if parenchymal cells are identified as “foreign,” neutrophil activation could lead to cell injury in an allograft. Work by Entman et al45 showed that neutrophils may cause direct myocyte injury and that cytokines may be involved by inducing expression of the adhesion molecule ICAM-1 on the myocyte surface, which interacts with an integrin expressed on the surface of the activated neutrophil.46 As mentioned previously, in CD8+-depleted animals, eosinophils may be involved in parenchymal cell injury during allograft rejection.30 Further studies are needed to investigate whether neutrophils, activated macrophages, or eosinophils play a direct role as cytotoxic effector cells in the HIC population participating in rejection.
Several conclusions can be drawn from our work. First, cultured adult murine ventricular myocytes can be used as target cells to detect immune-mediated cytotoxicity. CTLs from an MLR cause allospecific lysis of adult ventricular myocytes within 6 hours. Cells directly isolated from rejecting hearts (HICs) can also injure adult myocytes, but only a small part of the cytotoxicity can be attributed to CTLs within the HIC population. The major component of the cytotoxicity observed is less intense, more delayed, and not alloantigen specific. Thus, we postulate that the cytotoxicity of adult ventricular myocytes by HICs in this murine model may be induced by both an alloantigen-specific CTL-mediated pathway and by cell types such as macrophages and neutrophils participating in a DTH response.
Selected Abbreviations and Acronyms
|CTL||=||cytotoxic T lymphocyte|
|HTL||=||helper T lymphocytes|
|LDA||=||limiting dilution analysis|
|MLR||=||mixed lymphocyte reaction|
|tCTLs||=||total CTLs (precursor plus in vivo–stimulated CTLs)|
|Th1, Th2||=||type 1 or 2 HTL|
This study was supported by NIH grants RO1-HL-42535 (Dr Barry) and R29-AI-30104-03 (Dr Bishop) and National Research Service Award 1S32-HL-08934-01 (Dr Wagoner).
Presented in part at the 2nd International Symposium of Heart Failure, Geneva, Switzerland, May 29, 1993, and the 66th Scientific Sessions of the American Heart Association, Atlanta, Ga, November 8, 1993, and published in abstract form in Circulation (1993;88:I-41).
- Received April 13, 1995.
- Revision received July 13, 1995.
- Accepted August 8, 1995.
- Copyright © 1996 by American Heart Association
Renlund DG, O’Connell JB, Gilbert EM, Hammond ME, Burton NA, Jones KW, Karwande SV, Doty DB, Menlove RL, Bristow MR. A prospective comparison of murine monoclonal CD3 (OKT3) antibody-based and equine antithymocyte globulin-based rejection prophylaxis in cardiac transplantation. Transplantation. 1989;47:599-605.
Sell KW, Kanter K, Rodey GE, Wang YC, Ansari AA. Characterization of human heart-infiltrating cells after transplantation, V: suppression of donor-specific allogeneic responses by cloned T-cell lines isolated from heart biopsy specimens of patients after transplantation. J Heart Lung Transplant. 1992;11:500-510.
Bishop DK, Shelby J, Eichwald EJ. Mobilization of T lymphocytes following cardiac transplantation: evidence that CD4-positive cells are required for cytotoxic T lymphocyte activation, inflammatory endothelial development, graft infiltration, and acute allograft rejection. Transplantation. 1992;53:849-857.
Woodley SL, McMillan M, Shelby J, Lynch DH, Roberts LK, Ensley RD, Barry WH. Myocyte injury and contraction abnormalities produced by cytotoxic T lymphocytes. Circulation. 1991;83:1410-1418.
Lynch DL, Weiland DJ, Rosenberg SA, Hodes RJ. Different specificities of cloned T cells assessed by in vitro proliferation assays and by the ability to mediate skin graft rejection in vivo. Transplantation. 1987;43:403-411.
Shelby J, Corry RJ. The primarily vascularized mouse heart transplant as a model for the study of immune response. J Heart Transplant. 1982;2:32-36.
Ikenouchi H, Zhao L, Barry WH. Effect of 2,3-butanedione monoxime on myocyte resting force during prolonged metabolic inhibition. Am J Physiol. 1994;267:H419-H430.
Farrar J, Fuller-Farrar J, Simon P, Hilfiker M, Stadler B, Farrar W. Thymoma production of T cell growth factor (interleukin 2). J Immunol. 1980;125:2555-2558.
Ryser JE, MacDonald HR. Limiting dilution analysis of alloantigen-reactive T lymphocytes, I: comparison of precursor frequencies for proliferative and cytolytic responses. J Immunol. 1979;122:1691-1696.
Taswell C. Limiting dilution assays for the determination of immunocompetent cell frequencies, I: data analysis. J Immunol. 1981;126:1619-1627.
Pfau S, Bender JR. Initiation of cardiac allograft rejection: new developments in cellular and molecular mechanisms. Trends Cardiovasc Med. 1993;3:196-203.
Mann DL, Kent RL, Parsons B, Cooper G. Adrenergic effects on the biology of the adult mammalian cardiomyocyte. Circulation. 1992;85:790-804.
Strom TB, Tilney NL, Paradysz JM, Bancewicz J, Carpenter CB. Cellular components of allograft rejection: identity, specificity, and cytotoxic function of cells infiltrating acutely rejecting allografts. J Immunol. 1977;118:2020-2026.
Hall BM. Cellular infiltrates in allografts. Transplant Proc. 1987;109:50-56.
Ascher NL. Effector mechanisms in allograft rejection. Transplant Proc. 1987;19:57-60.
Barry WH. Mechanisms of immune-mediated myocyte injury. Circulation. 1994;89:2421-2432.
Abbas AK, Lichtman AH, Pober JS. Cellular and Molecular Immunology. Philadelphia, Pa: WB Saunders; 1991:251-253.
Cramer DV. Cardiac transplantation: immune mechanisms and alloantigens involved in graft rejection. CRC Crit Rev Clin Lab Sci. 1987;7:1-30.
Babiar BM. The respiratory burst of macrophages. J Clin Invest. 1984;73:599-601.
Pinsky DJ, Cai B, Young X, Rodriquez C, Sciacca RR, Cannor PJ. Nitric oxide dependent killing of myocytes by adjacent macrophages. Circulation. 1994;90(suppl I):I-192. Abstract.
Entman ML, Youker K, Shoji T, Kukielka G, Shappel SB, Taylor AA, Smith CW. Neutrophil induced oxidative injury of cardiac myocytes. J Clin Invest. 1992;90:1335-1345.
Entman ML, Youker K, Shappel SB, Siegel C, Rothlein R, Dreyer WJ, Schmalstieg FC, Smith CW. Neutrophil adherence to isolated adult canine myocytes. J Clin Invest. 1990;85:1497-1506.