Identification of Noncytotoxic and IL-10–Producing CD8⁺AT2R⁺ T Cell Population in Response to Ischemic Heart Injury

After acute myocardial infarction (MI), the heart suffers, besides ischemia-induced direct myocardial injury, from a subsequent indirect damage through improper inflammatory reaction. For instance, experimental studies demonstrated that CD8⁺ T cells derived from rats with MI are cytotoxic to healthy cardiomyocytes (CMs), and adoptive transfer of postinfarct splenic lymphocytes results in T cell-mediated myocardial injury in recipient rats (1, 2). It is proposed that ischemia-induced myocardial necrosis releases or exposes normally sequestered constituents that may trigger an autoimmune response in the myocardium, as evidenced by clinical observations that autoantibodies against cardiac contractile proteins are indeed detectable in the sera of patients after MI (3).

A wealth of information indicates that the renin-angiotensin system can interfere with acute cardiac remodeling and inflammation processes during cardiovascular injury (4, 5). Most of the cardiovascular actions of angiotensin II (Ang II), such as blood pressure and osmotic control, are mediated by AT1R (6). In contrast, debatable observations on AT2R expression in cardiac fibroblasts of patients with heart failure argue whether AT2R is involved in the regulation of cardiac fibrotic process (7). Our recent data show that cardiac c-Kit⁺AT2R⁺ cell population exists and concerns adaptive cardioprotection (8), raising the importance of other potential cellular mechanisms, which may also contribute to AT2R-mediated repair processes in the heart.

Several previous studies have suggested that Ang II via its AT1R plays an important role in triggering inflammatory cardiovascular injury (4, 5). Indeed, the cardioprotective effects of AT1R blockade are related to reduced myocardial inflammation in patients with MI (9). Accumulating experimental evidence from AT2R knockout mice or from rats under AT2R agonists further indicates that AT2R, which may be exposed to enhanced Ang II level after AT1R blockade, afford protection against MI-induced cardiac injury (5, 10). Given the recent observations showing that angiotensin AT2R is abundantly expressed in immunocompetent cells and involved in cell-mediated inflammatory injury (11), it appears likely that AT2R exerts its actions through interfering with cellular inflammatory processes in the heart.

In this study, we have identified the noncytotoxic and IL-10–producing CD8⁺AT2R⁺ T cell population, increasing after ischemic heart injury. We further show that AT2R activation enhances cardioprotective CD8⁺AT2R⁺ T cells and IL-10 production in the infarcted myocardium, revealing an AT2R-mediated cellular mechanism in reducing inflammatory injury in the heart.

MI and treatment in male Wistar rats are performed as described previously (5). In brief, after left lateral thoracotomy, a suture was tightened around the proximal left anterior descending coronary artery. Sham-operated rats received the same procedure with the exception of coronary ligature. To determine the in vivo effect of AT2R, we treated rats with AT2R agonist compound 21 (C21; provided by A. Hallberg, University of Uppsala, Uppsala, Sweden). At day 7 after operation, hearts and spleens were rapidly harvested. Animal protocols followed the German law on animal protection.

Free-floating heart cryosections (10 μm) were incubated with 10% donkey serum, followed by mouse monoclonal anti-CD8 (1:150; AbD Serotec, Düsseldorf, Germany) and rabbit polyclonal anti-AT2R (1:150; Santa Cruz Biotechnology, Santa Cruz, CA) Abs. Sections were then incubated with donkey indocarbocyanin (Cy3) anti-mouse IgG and donkey FITC anti-rabbit IgG (each 1:150; Dianova, Hamburg, Germany), and counterstained with DAPI (1:1000; Invitrogen, Darmstadt, Germany). Stained sections were examined under Leica DMIRE2 microscope (Leica Microsystems, Wetzlar, Germany).

Cardiac cells were isolated as described previously (12), with some modifications. After digestion and density gradient sedimentation, CMs were collected and CD8⁺ cells were selected using MACS (Miltenyi Biotec, Auburn, CA). Splenic CD8⁺ cells were also isolated by MACS. The CD8⁺ cell suspension was then incubated with goat anti-AT2R (Santa Cruz Biotechnology), FITC-conjugated mouse anti-CD3 (BD Biosciences, Heidelberg, Germany) Abs, indirectly labeled to allophycocyanin (Jackson ImmunoResearch Laboratories, West Grove, CA) anti-goat secondary Ab and subjected to flow cytometry. Analysis and cell acquisition was performed on a FACSCalibur cytometer or cell sorting (CD8⁺AT2R⁺ and CD8⁺AT2R⁻ cells) on FACSAria. Data were analyzed using FlowJo software (Tree Star, Ashland, OR).

Adult CMs, isolated from sham-operated hearts, were cultured alone or cocultured with postinfarct cardiac or splenic CD8⁺AT2R⁺ or CD8⁺AT2R⁻ T cells at a ratio of 1:10 (CM/T), in a 96-well plate, with DMEM/RPMI (1:1, 1% FBS). Apoptotic CM was examined by an ethidium bromide/acridine orange method (8). A minimum of n = 100 CMs/well was counted. Percentage of apoptotic CM was calculated by the formula (apoptotic + necrotic CM) × 100/(total CM) and normalized to the apoptotic rate of CM cultured alone. Apoptotic CM cocultured with postinfarct splenic CD8⁺AT2R⁺ or CD8⁺AT2R⁻ T cells was also evaluated in the presence or absence of neutralizing anti–IL-10 Ab (50 μg/ml; PeproTech, Hamburg, Germany).

Rat fetal CMs (H9C2, from ATCC), prestained with CFSE (1 μM; Invitrogen), were cocultured with postinfarct splenic CD8⁺AT2R⁺ or CD8⁺AT2R⁻ T cells at a ratio of 1:10 (H9C2:T). Cocultured cells were kept for 4 h in DMEM/RPMI 1640(1:1, 1% FBS; Life Technologies, Rockville, MD). Propidium iodide (PI) was added before flow cytometry acquisition (FACSCalibur), and percentage of lysed CMs was calculated by the formula (CFSE + PI⁺)/(CFSE + PI⁺) + (CFSE + PI⁻) × 100.

RNA was isolated with TRIzol Reagent (Invitrogen) according to manufacturer’s indications. One microgram total RNA was reverse transcribed to cDNA using a Reverse Transcription kit (Promega, Madison, WI). Real-Time PCR (SYBR green; Applied Biosystems, Foster City, CA) was carried out for 40 cycles (95°C 15 s, 60°C 1 min). The primers as listed in Supplemental Table I were selected to be intron spanning. Expression levels were normalized to the expression of 18S gene.

CD8⁺ T cells, isolated from rat spleens 7 d after MI, were cultured at a density of 10⁶ cells/ml with RPMI 1640 (Life Technologies). Cells were stimulated with AT2R agonist C21 (10⁻⁷M) or Ang II (10⁻⁷ M; Sigma-Aldrich, St. Louis, MO), or Ang II (10⁻⁷ M) plus AT2R antagonist PD123319 (PD, 10⁻⁶M; Tocris Biosciences, Bristol, U.K.). Two days after treatment, cells were harvested for the analysis of IL-10 mRNA.

Cells were stimulated for 6 h with PMA (5 ng/ml; Calbiochem, Darmstadt, Germany) and ionomycin (500 ng/ml; Calbiochem) in the presence of monensin (2 μM; BD Bioscences) and blocked with 2% donkey serum (Jackson ImmunoResearch Laboratories). Cells were incubated with FITC-conjugated mouse anti-CD8 (AbD Serotec) and goat anti-AT2R (Santa Cruz) Abs. Cells were then labeled with anti-goat allophycocyanin (Jackson ImmunoResearch Laboratories). After fixation and permeabilization, cells were stained with PE-conjugated mouse anti–IL-10 Ab (1:50; BD Biosciences). IL-10–producing CD8⁺ T cells were acquired using FACSCalibur.

Splenic CD8⁺ T cells were isolated from donor Wistar rats 7 d after MI, and further sorted into CD8⁺AT2R⁺ and CD8⁺AT2R⁻ T cell populations, as described earlier. Freshly isolated CD8⁺AT2R⁺ or CD8⁺AT2R⁻ T cells were transplanted intramyocardially in recipient Wistar rats immediately after the induction of MI, while the recipient Wistar rats were still under anesthesia. Around 1.25 × 10⁵ cells suspended in 20 μl PBS were injected into the peri-infarct myocardium.

Infarct size was evaluated as previously described (13). In brief, recipient rats were sacrificed 2 wk after MI and syngeneic intramyocardial cell transplantation. After perfusion with PBS, the hearts were harvested and weighted, and then sliced into four transverse 3-mm sections from apex to base. Slices were stained freshly with 2% triphenyltetrazolium chloride (Sigma-Aldrich) and fixed with 4% paraformaldehyde. The caudal faces of each section were scanned using a flatbed color scanner, and the infarct volume was calculated using ImageJ software (National Institutes of Health, Bethesda, MD).

Buffy coats from healthy adult donors were obtained in accordance with local ethical committee approval. Rat peripheral blood was collected immediately before euthanasia 7 d after MI. Mononuclear cells were isolated by density gradient sedimentation using Lymphocyte Separation Medium (PAA Laboratories, Cölbe, Germany). CD8⁺ cells were enriched by MACS and further sorted (CD8⁺AT2R⁺ and CD8⁺AT2R⁻, respectively) on FACSAria.

Results were expressed as mean ± SEM. Two-group comparisons were analyzed by two-tailed Student t test. Multiple comparisons were analyzed with one-way ANOVA followed by Bonferroni post hoc test. Differences were considered significant at p < 0.05.

We first examined the cellular distribution of cardiac AT2R after acute ischemic injury. Seven days after MI, myocardial infiltrating CD8⁺ T cells were abundantly detected in the peri-infarct area by immunohistochemical staining (red, Fig. 1B). Notably, positive signals for AT2R (green, Fig. 1A) were localized in a fraction of these infiltrating CD8⁺ T cells (Fig. 1A, 1B, 1D, arrows), suggesting that AT2R is potentially involved in the cellular inflammatory response to ischemic heart injury.

To characterize the CD8⁺ T cell-mediated inflammatory response, we decided to isolate the CD8⁺ T cells from the heart and spleen 7 d after MI. Only CD8⁺ cells, which coexpressed CD3, were isolated (Fig. 2A); ∼30.57 ± 5.60 × 10³ CD8⁺ T cells per infarcted heart were obtained (n = 10), and the purity of these isolated cells was 95–99%, as determined by FACSCalibur. In contrast, only 6.31 ± 2.72 × 10³ CD8⁺ T cells were isolated from each sham-operated heart (n = 5). The CD8⁺ T cells in each infarcted and sham-operated heart represented ∼0.025% and ∼0.0058% of total cardiac cells, respectively. From the spleen, 1.96 ± 0.32 × 10⁶ CD8⁺ T cells were isolated after MI (n = 9), whereas only 0.55 ± 0.48 × 10⁶ CD8⁺ T cells were obtained after sham operation (n = 3) (Fig. 2B). The frequency of splenic CD8⁺ T cells was ∼1.0% and ∼0.23% of total splenic cells after MI and sham operation, respectively (Fig. 2C).

To clarify the role of AT2R in CD8⁺ T cell-mediated inflammatory response, we further sorted the whole cardiac and splenic CD8⁺ T cells into the CD8⁺AT2R⁺ (upper right quadrant) and CD8⁺ AT2R⁻ (lower right quadrant) populations (Fig. 3A, 3B). MI led to an increase of CD8⁺AT2R⁺ T cells in both heart and spleen. The CD8⁺AT2R⁺ cells occupied 33.53% and 4.01% of cardiac CD8⁺ T cells after MI and sham operation, respectively (Fig. 3C). In the spleen, the CD8⁺AT2R⁺ cells were 46.88% and 15.68% of CD8⁺ T cells after MI and sham operation, respectively (Fig. 3D).

To confirm the specificity of the cell isolation method, we further examined mRNA abundance of AT2R, CD8, and CD3 n postinfarct splenic CD8⁺AT2R⁺ T cells (versus CD8⁺AT2R⁻). AT2R mRNA was 13.7-fold higher in the CD8⁺AT2R⁺ T cells (n = 4; Supplemental Fig. 1). The predominant mRNA expression of CD8 and CD3 was also confirmed in CD8⁺ T cells, isolated from hearts and spleens after MI (data not shown).

To extend the animal findings to humans, we further analyzed the CD8⁺AT2R⁺ T cells from peripheral blood of healthy donors. As shown by the flow cytometric analysis in Supplemental Fig. 2A, AT2R distinguished human circulating CD8⁺ T cells into the CD8⁺AT2R⁺ and CD8⁺AT2R⁻ populations. The frequency of the CD8⁺AT2R⁺ T cell population represented ∼4.67% of CD8⁺ T cells. In addition, there was a tendency toward increased expression of IL-10 mRNA in the CD8⁺AT2R⁺ T cells in comparison with CD8⁺AT2R⁻ T cells (Supplemental Fig. 2B). These data confirm that the CD8⁺AT2R⁺ T cell population exists in humans as well.

To understand whether the CD8⁺AT2R⁺ and CD8⁺AT2R⁻ T cells may act on CMs differently in the context of cell-mediated cytotoxicity, we compared the cytotoxicity of these T cells to CMs. We observed that the apoptotic rate of adult CMs cocultured with either cardiac or splenic CD8⁺AT2R⁻ T cells was much higher (1.36-fold, p < 0.05, n = 4; or 1.39-fold, p < 0.05, n = 7; Fig. 4A, 4B) when compared with adult CMs cultured alone. In contrast, the apoptotic CMs cocultured with the CD8⁺AT2R⁺ T cells were not increased. Interestingly, increased apoptotic rate of CMs was again observed when neutralizing anti–IL-10 Ab was present in the coculture of CMs and splenic CD8⁺AT2R⁺ T cells (Supplemental Fig. 3), indicating that IL-10 is mainly responsible for the noncytotoxicity of CD8⁺AT2R⁺ T cells to adult CMs. To investigate whether this difference in cell-mediated cytotoxicity may generally apply to target cells with CM-related Ags, the cytotoxicity of CD8⁺AT2R⁺ and CD8⁺AT2R⁻ T cells was further evaluated on H9C2 fetal CMs. Indeed, the cytotoxicity to cocultured fetal CMs was detected only in the CD8⁺AT2R⁻ T cells but not in the CD8⁺AT2R⁺ T cells isolated from the spleens after MI (Fig. 4C, 4D). Thus, during adaptive immune reaction to ischemic heart injury, the CD8⁺AT2R⁺ T cells were induced and exerted noncytotoxicity to cells expressing CM Ags.

A noncytotoxic CD8⁺ T cell population may participate in local immune responses by the release of cytokines (14, 15). We next dissected the expression of anti-inflammatory cytokine IL-10 and proinflammatory cytokines including IL-1β, IL-2, and INF-γ in postinfarct cardiac and splenic CD8⁺AT2R⁺ and CD8⁺AT2R⁻ T cells. FACS analysis revealed that the percentage of IL-10–producing cells within CD8⁺AT2R⁺ T cells was significantly increased (Fig. 5A, 5C), whereas the percentage of IL-1β–producing cells within cardiac CD8⁺AT2R⁺ T cells was significantly reduced (Fig. 5B, 5D) when compared with that within cardiac CD8⁺AT2R⁻ T cells. Real-time PCR further showed that IL-10 mRNA was upregulated by 2.8-fold, whereas IL-2 and INF-γ were downregulated by 5.6-fold and 5.1-fold, respectively, in splenic CD8⁺AT2R⁺ T cells (versus CD8⁺AT2R⁻; p < 0.05; n = 6; Fig. 6A). There was no difference for IL-1β mRNA between splenic CD8⁺AT2R⁺ and CD8⁺AT2R⁻ T cells. We further explored whether AT2R may influence IL-10 production in splenic CD8⁺ T cells. AT2R stimulation with Ang II or C21 led to an increment of IL-10 mRNA by 1.3-fold (p < 0.05; n = 6) and 1.4-fold (p < 0.05; n = 6), respectively. However, PD123319 abolished Ang II-induced IL-10 expression (Fig. 6B), suggesting that noncytotoxic CD8⁺AT2R⁺ T cells are also involved in regulating local production of cytokines, especially anti-inflammatory IL-10.

To explore how AT2R may modulate cellular inflammatory reaction to cardiac injury in vivo, we examined the effects of AT2R activation on cardiac and splenic CD8⁺AT2R⁺ T cells in rats with MI. In vivo AT2R stimulation with C21 led to an enhanced accumulation of CD8⁺AT2R⁺ T cells into the infarcted myocardium. Flow cytometry analysis of isolated cardiac cells showed that the frequency of cardiac CD8⁺AT2R⁺ T cells was significantly increased after C21 treatment (Fig. 7A). A redistribution of CD8⁺AT2R⁺ T cells from blood (Supplemental Fig. 4) to postinfarct heart appears to contribute to increased cardiac accumulation of CD8⁺AT2R⁺ T cells after in vivo AT2R stimulation. Notably, the frequencies of IL-10–producing cells within cardiac and splenic CD8⁺AT2R⁺ T cells were also increased by C21 (Fig. 7B). These findings indicate that AT2R may reduce inflammatory injury in the infarcted heart through cardioprotective IL-10 via the CD8⁺AT2R T cells.

To demonstrate whether the CD8⁺AT2R⁺ T cell population is cardioprotective in vivo, we investigated the effect of postinfarct CD8⁺AT2R⁺ T cells on cardiac injury by evaluating infarct size in rats 14 d after MI. All six rats transplanted with postinfarct splenic CD8⁺AT2R⁺ T cells survived until 14 d after MI. In contrast, two of six rats transplanted with postinfarct splenic CD8⁺AT2R⁻ T cells could not survive until 14 d after MI. Importantly, infarct volume was significantly reduced by 54.66% in rats transplanted with CD8⁺AT2R⁺ T cells (n = 6) when compared with those survived rats transferred by CD8⁺AT2R⁻ T cells (n = 4; p < 0.05; Fig. 8). These results support a cardioprotective role of postinfarct CD8⁺AT2R⁺ T cells in vivo.

Emerging evidence suggests a cardioprotective role of the angiotensin AT2R, albeit the underlying cellular mechanisms are poorly understood. In this article, we defined the CD8⁺AT2R⁺ T cell population, which increased after acute MI. Furthermore, we developed a method to selectively isolate these cardiac CD8⁺AT2R⁺ T cells, which exhibited downregulated expression of proinflammatory cytokines and noncytotoxicity to CMs. AT2R activation engendered a significant increment in cardiac CD8⁺AT2R⁺ T cells and IL-10 production, likely contributing to reduced infarct size and improved cardiac function, as shown recently by our group (5). In addition, intramyocardial transplantation of CD8⁺AT2R⁺ T cells (versus CD8⁺AT2R⁻) led to reduced ischemic heart injury. These findings advance our understanding of the functional role of AT2R in the heart and provide a cellular mechanism via cardioprotective CD8⁺AT2R⁺ T cells.

The renin-angiotensin system is an important player in the regulation of the cardiovascular homeostasis. Most of the actions of Ang II, including blood pressure and osmotic control, are mediated by AT1Rs (6). In contrast, the upregulation of AT2R, which occurs during acute tissue injury, including MI (8, 16) and brain ischemia (17), speaks in favor of its potential role in the early cellular response to tissue injury. At the cellular level, we have previously shown that the number of AT2R-expressing CMs was not changed after acute MI, suggesting that noncardiomyocytes may account for the upregulated AT2R in the infarcted heart (16). We recently further demonstrated that increased cardiac c-Kit⁺AT2R⁺ cell population exists during acute ischemic injury. This study provides the essential evidence that postinfarct AT2R is also induced in cardiac-infiltrating CD8⁺ T cells, raising a concept that AT2R is linked to CD8⁺ T cell-involved immune response in the heart.

The early cellular response to MI is characterized by inflammatory infiltration of immunocompetent cells involving T lymphocytes at the injury site and phagocytosis of necrotic myofibers, which play a critical role not only in facilitating the scar development but also in causing subsequent cell-mediated inflammatory injury. Indeed, there is clear evidence showing that an inflammatory reaction can severely augment postinfarct cardiac damages (1, 2). For example, postinfarct splenic CD8⁺ T lymphocytes exert cytotoxicity to healthy CMs (2), and adoptive transfer of postinfarct splenic lymphocytes results in adaptive autoimmune myocarditis in recipient rats (1). Because of the complexity and heterogeneity of cardiac cell populations imbedded within cross-linked myocardial matrix, the functional roles of cardiac infiltrating T lymphocytes have not been explored. In this study, we developed a method, allowing the isolation and characterization of cardiac infiltrating T lymphocytes. Besides the cytotoxic CD8⁺AT2R⁻ T cells, we have defined a noncytotoxic CD8⁺AT2R⁺ T cell population.

Although accumulating evidence shows that Ang II is involved in inflammatory reactions during cardiovascular injury (4, 9, 18), the respective cellular contributions of the angiotensin receptors to cardiac inflammatory processes are still not well understood. These data further demonstrate that cardiac AT2R is associated with downregulated expression of proinflammatory cytokines and sustained IL-10 production, at least in part, via postinfarct CD8⁺AT2R⁺ T cells, providing a cellular mechanism behind cardiac AT2R-mediated actions.

It is well accepted that cardiac AT2R exerts protective actions against ischemic injury, as evidenced by experimental data from AT2R^−/− mice or from rats under AT2R agonists (5, 10). Our recent observations indicate that anti-inflammatory mechanisms are attributable to AT2R-mediated cardioprotection (5). Taking into account that IL-10–mediated immunomodulation has been clearly shown to inhibit pathological processes in autoimmune diseases including MI (19), the present data support an unrecognized immune regulatory role of AT2R, reducing cardiac inflammatory injury by cardioprotective IL-10 via the CD8⁺AT2R⁺ T cells.

Initially, CD8⁺ T cell has been defined as CTL. Later, convergent evidence suggests that CD8⁺ T cells can be differentiated into several subsets. Based on the pattern of cytokine production, the regulatory T cell subsets have been not only described in CD4⁺ T cells (20) but identified in CD8⁺ T cells (14, 15). These CD8⁺ regulatory T cells, which are characterized by the production of immune suppressive cytokines involving mainly IL-10, have been generated from native CD8⁺ T cells in peripheral blood of healthy donors (14) or detected in the spleen and lymph node of rats with CD4-dependent graft-versus-host disease (15), and in the livers of patients with chronic hepatitis C virus infection (21). In this study, the CD8⁺AT2R⁺ T cell population was characterized in infarcted rat heart and spleen, and was further analyzed in human peripheral blood. It will be interesting to investigate whether this CD8⁺AT2R⁺ T cell population belongs to those previously described IL-10–producing CD8⁺ regulatory T cells or is a novel CD8⁺ T cell subset with unexpected properties, and to what extent this CD8⁺AT2R⁺ T cell population may have any influence on other cell populations involved in inflammatory responses including CD4⁺ T cells, NKT cells, and macrophages.

Collectively, we have identified the CD8⁺AT2R⁺ T cell population, increasing in response to ischemic cardiac injury. AT2R activation exerts an anti-inflammatory action in the injured heart via cardioprotective CD8⁺AT2R⁺ T cells, hence providing a cardioprotective cellular mechanism. These findings are clinically relevant because therapeutic strategies using selective AT2R agonists may help reduce inflammatory tissue injury.

We are grateful to Melanie Timm for her excellent technical assistance. We also thank Dr. Kai Kappert for his helpful suggestions and discussions.

Disclosures The authors have no financial conflicts of interest.