The Journal of Clinical Investigation   RESEARCH ARTICLE CD69 expression on regulatory T cells protects from immune damage after myocardial infarction Rafael Blanco-Domínguez,1 Hortensia de la Fuente,2,3 Cristina Rodríguez,3,4,5 Laura Martín-Aguado,1 Raquel Sánchez-Díaz,1,3 Rosa Jiménez-Alejandre,1 Iker Rodríguez-Arabaolaza,1 Andrea Curtabbi,1 Marcos M. García-Guimaraes,6 Alberto Vera,6 Fernando Rivero,6 Javier Cuesta,6 Luis J. Jiménez-Borreguero,3,6 Alberto Cecconi,6 Albert Duran-Cambra,7 Manel Taurón,7 Judith Alonso,3,5,8 Héctor Bueno,1,3,9,10 María Villalba-Orero,1,11 Jose Antonio Enríquez,1,12 Simon C. Robson,13 Fernando Alfonso,3,6 Francisco Sánchez-Madrid,1,2,3 José Martínez-González,3,5,8 and Pilar Martín1,3 Vascular Pathophysiology Area, Centro Nacional de Investigaciones Cardiovasculares (CNIC), Madrid, Spain. 2Department of Immunology, IIS Princesa, Hospital Universitario de la Princesa, Universidad 1 Autónoma de Madrid, Madrid, Spain. 3CIBER de Enfermedades Cardiovasculares (CIBERCV), Madrid, Spain. 4Institut de Recerca del Hospital de la Santa Creu i Sant Pau, Barcelona, Spain. 5Instituto de Investigación Biomédica Sant Pau (IIB-Sant Pau), Barcelona, Spain. 6Department of Cardiology, IIS Princesa, Hospital Universitario de la Princesa, Universidad Autónoma de Madrid, Madrid, Spain. 7Hospital de la Santa Creu i Sant Pau, Barcelona, Spain. 8Instituto de Investigaciones Biomédicas de Barcelona–Consejo Superior de Investigaciones Científicas (IIBB-CSIC), Barcelona, Spain. 9Cardiology Department, Hospital Universitario 12 de Octubre and Instituto de Investigación Sanitaria Hospital 12 de Octubre (imas12), Madrid, Spain. 10Facultad de Medicina, Universidad Complutense de Madrid, Madrid, Spain. 11 Departamento de Medicina y Cirugía Animal, Facultad de Veterinaria, Universidad Complutense de Madrid, Madrid, Spain. 12CIBER de Fragilidad y Envejecimiento Saludable (CIBERFES), Madrid, Spain. 13 Department of Medicine, Harvard Medical School, Transplantation Research Center, Beth Israel Deaconess Medical Center, Boston, Massachusetts, USA. Increasing evidence has pointed to the important function of T cells in controlling immune homeostasis and pathogenesis after myocardial infarction (MI), although the underlying molecular mechanisms remain elusive. In this study, a broad analysis of immune markers in 283 patients revealed significant CD69 overexpression on Tregs after MI. Our results in mice showed that CD69 expression on Tregs increased survival after left anterior descending (LAD) coronary artery ligation. Cd69–/– mice developed strong IL-17+ γδT cell responses after ischemia that increased myocardial inflammation and, consequently, worsened cardiac function. CD69+ Tregs, by induction of AhR-dependent CD39 ectonucleotidase activity, induced apoptosis and decreased IL-17A production in γδT cells. Adoptive transfer of CD69+ Tregs into Cd69–/– mice after LAD ligation reduced IL-17+ γδT cell recruitment, thus increasing survival. Consistently, clinical data from 2 independent cohorts of patients indicated that increased CD69 expression in peripheral blood cells after acute MI was associated with a lower risk of rehospitalization for heart failure (HF) after 2.5 years of follow-up. This result remained significant after adjustment for age, sex, and traditional cardiac damage biomarkers. Our data highlight CD69 expression on Tregs as a potential prognostic factor and a therapeutic option to prevent HF after MI. Introduction CD4+CD25+Foxp3+ Tregs have emerged as important players in Myocardial infarction (MI) is an acute ischemic myocardial protection against cardiac damage after experimental MI (5). insult that clinically represents the global leading cause of mor- MI induces a cardioprotective T cell response (6) and a rapid tality. Inflammation plays an essential role in the pathophysiol- recruitment of Tregs to the injured myocardium, mediated by ogy of atherosclerosis, plaque formation, and plaque rupture, the CXCL12/CXCR4 axis (7) and galectin-1 (Gal-1) (8). In par- the main trigger of MI, but also in the process of myocardial allel, Tregs become activated 7 days after MI in the heart-drain- healing and repair after ischemia-induced damage (1). The role ing lymph nodes (9) with increased percentages and absolute of innate immunity and myeloid cells after MI has been wide- numbers, protecting against uncontrolled inflammation and ly studied (1, 2), but that of the adaptive immunity remains cardiac deterioration. Indeed, Treg depletion aggravates cardi- scarcely explored. Recent studies have shown that CD4+ T ac inflammation and worsens prognosis early after experimen- cells are important contributors to inflammatory mechanisms tal MI (9, 10). However, the specific molecular mechanisms and, more specifically, to myocardial damage repair mech- that operate in these processes remain elusive. anisms after ischemia-reperfusion processes (3, 4). Indeed, Different studies in patients indicate that the proportions of Tregs change in peripheral blood after MI (11–13). In addition, some evidence indicates that Tregs accumulate in coronary Conflict of interest: The authors have declared that no conflict of interest exists. thrombi (14) and in the heart (6) of patients with MI, reinforc- Copyright: © 2022, Blanco-Domínguez et al. This is an open access article published ing the notion that Tregs play a major role at the site of injury. under the terms of the Creative Commons Attribution 4.0 International License. Submitted: June 16, 2021; Accepted: September 1, 2022; Published: November 1, 2022. Studies in larger cohorts of patients and exploration of different Reference information: J Clin Invest. 2022;132(21):e152418. Treg regulatory molecules may further clarify the function of https://doi.org/10.1172/JCI152418. these cells after MI. 1 RESEARCH ARTICLE The Journal of Clinical Investigation   CD69 is a C-type lectin whose gene is encoded within the NK https://doi.org/10.1172/JCI152418DS1. More than 90% of patients cell gene complex on chromosome 6 in mice and chromosome 12 presented with at least 1 risk factor, most frequently dyslipidemia, in humans. CD69 promotes immune homeostasis by regulating arterial hypertension, and a history of smoking. Supplemental the differentiation and function of Tregs (15, 16). On the basis of Table 1 also summarizes the results of biomarkers, ECG, coronary CD69 expression, Tregs are classified into 2 subtypes: CD69+ and angiography (number of diseased vessels), and echocardiography CD69– Tregs. CD69+ Tregs express higher levels of suppression (left ventricular [LV] segmental wall motion abnormalities and markers such as CTLA-4, ICOS, CD38, and GITR than to CD69– ejection fraction). Most patients presented with ST-segment eleva- or Cd69 –/– Tregs, resulting in enhanced effector T cell suppression tion MI, and the culprit lesion was more often located in the left and tolerance (15). Cd69 –/– mice exhibit uncontrolled Th17 cell anterior descending (LAD) coronary artery or the right coronary (CD3+CD4+IL-17+) responses and aggravation of multiple autoim- artery, with 38.5% of patients having multivessel coronary artery mune and inflammatory disease models (17), including those of disease. LV systolic function was generally preserved, and only the cardiovascular system such as atherosclerosis (18), myocardi- one-third of patients had a reduced LV ejection fraction. tis, and inflammatory dilated cardiomyopathy (19). Other studies We performed an extensive analysis of lymphoid and myeloid highlight that IL-17 exacerbates fibrosis and cardiac remodeling cell populations by FACS in peripheral blood in the first 24 hours after MI (20, 21), although the involvement of Th17 cells after car- after the ischemic event (Supplemental Figure 1A). Elevation of diac injury remains controversial. CD3+TCRγδ+ (γδT) cells, which serum cardiac damage markers, such as troponin T (TnT) and cre- rapidly accumulate in the infarcted myocardium (22), are the main atine kinase (CK), after MI were correlated with CD69 expression source of IL-17 during the first week after MI and appear to be on circulating CD4+ T cells and CD4+CD25+Foxp3+ Tregs as well associated with the morbidity of MI in mice (23) and patients (24). as with the ratios of effector T cells versus CD69+ Tregs, suggest- There is evidence linking CD69 expression on T cells to ing that CD69 expression was induced upon myocardial damage MI. Interestingly, natural ligands of CD69, such as Gal-1 (25) (Supplemental Table 2 and Supplemental Figure 1A). Indeed, the or S100A8/A9 (26), are upregulated after MI and contribute to analysis of CD4+ T cells by flow cytometry revealed that a subset immune cell recruitment (8, 27). Furthermore, CD69 is overex- of CD69+CD25+Foxp3+ Tregs appeared after MI (Figure 1A). CD4+ pressed on infiltrating T cells in the atherosclerotic plaques (28) T cell and Treg populations were decreased and augmented, and circulating CD4+ T lymphocytes of patients with MI, in whom respectively, in patients with MI compared with healthy controls it has been associated with increased apoptosis driven by lympho- (Figure 1B and Supplemental Figure 1B). CD69 is virtually absent cyte activation (29). Therefore, we postulated that CD69 expres- in circulating blood lymphocytes due to the high concentration of sion on lymphocytes might be a player in the post-MI progression sphingosine-1-phosphate (S1P), which maintains S1P1 receptors at toward chronic scenarios. high levels (31), thereby suppressing CD69 expression (32). There- Many classical clinical and anatomic factors have been fore, we analyzed CD69 expression in circulating CD4+ T cells in involved in the prognosis of patients with MI. In these patients, the blood after overnight bystander activation with anti-CD3 anti- early reperfusion followed by guideline-based medical therapy body. The percentage of CD69+ Tregs increased in the circulation represents the cornerstone of clinical management. However, in patients with MI (Figure 1C and Supplemental Figure 1C), as predictive models, capable of identifying MI patients at higher determined by an overall increase in CD69 expression on Tregs risk for adverse long-term clinical outcomes, remain limited (30). after infarction in most of the patients (Figure 1D). However, we Consequently, the use of additional markers capable of improving divided these patients into 2 groups according to their expression risk stratification after MI to select patients most likely to bene- levels of CD69 on Tregs: those with high CD69 expression (65% fit from closer clinical follow-up, or more aggressive therapeutic of patients) and those with low CD69 expression (35% of patients) strategies, would be desirable. and conducted further analyses (Figure 1D). Patients with high This study reveals that increased expression of CD69+ Tregs is CD69 expression had lower percentages of CD4+ T cells and a common feature of mouse and patient peripheral blood shortly IL-22+CD4+ T cells but higher percentages of Tregs. In addition, after MI. In addition, we describe a mechanism by which CD69+ patients with high CD69 expression had more naive and few- Tregs controlled inflammation and cardiac damage after coronary er memory Tregs, according to the expression of CD45RA+ and artery ligation through inhibition of pathogenic IL-17+ γδT cells. CD45RO+, respectively (Figure 1E and Supplemental Figure 1D). Finally, in the clinical setting, we found that Tregs with low CD69 Unsupervised hierarchical clustering of patients based on myeloid expression after acute MI were associated with an increased risk and lymphoid cell types analyzed by FACS (Supplemental Figure of rehospitalization for heart failure (HF) at mid-term follow-up. 1A) revealed that patients with high CD69 expression had differ- Therefore, our data suggest that increasing the expression of ent immune phenotype than did those with low CD69 expression, CD69 on Tregs could be a promising strategy to improve post-MI as they clustered in distinct groups on the t-distributed stochastic management of patients. neighbor embedding (t-SNE) plot (Figure 1F). CD69 expression improves survival and recovery after LAD liga- Results tion in mice. To evaluate the role of CD69 expression in recovery Enhanced peripheral CD69+ Treg response in patients with acute MI. after MI, we analyzed the survival of mice after ligation of the LAD The blood immune phenotype was determined in a cohort of 283 coronary artery. Survival was significantly reduced in Cd69 –/– mice participants with acute MI and 80 healthy volunteers (Figure 1). compared with their wild-type littermates (Figure 2A). Cd69- Demographic and clinical data are summarized in Supplemental deficient mice had worse recoveries after infarction, as they were Table 1; supplemental material available online with this article; unable to reach the baseline weight 1 week after LAD ligation 2 J Clin Invest. 2022;132(21):e152418 https://doi.org/10.1172/JCI152418 The Journal of Clinical Investigation   RESEARCH ARTICLE J Clin Invest. 2022;132(21):e152418 https://doi.org/10.1172/JCI152418 3 RESEARCH ARTICLE The Journal of Clinical Investigation   Figure 1. Patients with MI have a strong peripheral CD69+ Treg response. IL-17+ γδT cells are the main source of peripheral IL-17A shortly (A) t-SNE plots of CD4+ T cells from PBLs from a representative healthy after MI and are increased in Cd69 –/– mice. It is well known that control and a patient with MI, considering the indicated markers measured the number of CD4+CD25+Foxp3+ Tregs increases in the heart by FACS. Color bars indicate the relative intensity of the markers. Dots represent individual cells. (B) Quantification of the percentages of CD4+ T during the first day after infarction (37), indicating an antigen-in- cells and CD4+CD25+Foxp3+ Tregs in peripheral blood from healthy donors (n dependent migration of Tregs into the injured myocardium. We = 51) and patients with MI (n = 215) at the time of hospital admission. (C) observed a specific mobilization of CD69+ Tregs in wild-type mice Percentages of CD69+ Tregs and CD69– Tregs among total PBLs. (D) CD69 after infarction, with a 2.5-fold increase in peripheral blood in the expression on Tregs, quantified as the percentage of CD69+ cells after gating first 24 hours (Figure 3A), mimicking the peripheral response in on Tregs. Two groups of patients with MI were differentiated according to CD69 expression: patients with high levels of CD69 (High CD69), shown patients with MI (Figure 1). This mobilization was not observed in as black circles, and patients with low levels (Low CD69), shown as red either wild-type CD69– or Cd69 –/– Tregs (Figure 3A). Since Cd69 –/– circles. Representative histograms and percentages of CD69 expression on mice have exacerbated Th17 responses in different inflammatory Tregs are shown. In B–D, data indicate the mean ± SEM, and significance diseases (38), we analyzed the kinetics of IL-17A+ cells in the blood was analyzed by Mann-Whitney U test. (E) Heatmap shows the levels of 1 week after infarction. We found that Cd69 –/– mice had a signifi- different cell populations analyzed by FACS and cardiac damage markers in patient with MI expressing high levels of CD69 and low levels of CD69. cantly increased IL-17 response during the first 2–7 days after LAD Each column represents 1 patient. Data were normalized by subtracting the ligation, which was not observed in Cd69+/+ mice (Figure 3B). Most mean and dividing by the SD. Color bar denotes the relative levels of each IL-17A–producing cells in the peripheral blood after infarction are parameter, with black indicating high expression and red indicating low CD3+ T cells (Figure 3C), with the majority of them being γδT cells, expression. Differences between patients with CD69hi and CD69lo expression which are the main source of IL-17 in the blood after infarction, were analyzed by Mann-Whitney U test; *P < 0.05, **P < 0.01, and ****P < 0.0001. (F) t-SNE plot was generated based on the percentages of cell instead of Th17 cells (Figure 3C). Cd69 –/– γδT cells expressed high- populations shown in E and analyzed by FACS. Circles represent individual er levels of IL-17A even in steady state, and Cd69–/– mice had an patients with MI. Black circles indicate patients with MI who had high CD69 amplified IL-17+ γδT cell response in peripheral blood shortly after expression, and red circles indicate patients with low CD69 expression. MI (Figure 3, D and E). As IL-17+ γδT cells are well known as ini- tiators of inflammation (39) and induce apoptosis of cardiac myo- cytes (40), we postulate that this population may contribute to the (Figure 2B). In addition, Cd69 –/– mice showed elevated myocardial increased damage observed in Cd69 –/– mice after LAD ligation. damage (Figure 2C) and increased heart weight/body weight and IL-17+ γδT cells rapidly accumulate in the infarcted myocardium heart weight/tibia length ratios 2 days after LAD ligation, indicat- in Cd69 –/– mice. Next, we analyzed the leukocyte populations ing that Cd69 –/– hearts may be swollen, perhaps as a consequence infiltrating the myocardium to characterize inflammation in of greater inflammation (Figure 2D). The assessment of infarct size infarcted tissue from Cd69 –/– mice. Quantification of the total and ischemic area at risk by double staining with triphenyltetra- number of leukocytes per milligram of infarcted tissue showed zolium chloride (TTC) and Evans blue is a widely used method for that the hearts of Cd69 –/– mice exhibited increased inflamma- the quantification of myocardial damage after infarction (33). In tion on day 2 after MI compared with their Cd69+/+ littermates t his study, the infarct size of Cd69+/+ mice ranged from 40%–80 (Figure 4A). Consistent with previous studies, we found that %, in agreement with previous reports in this model (34–36). The Tregs were recruited to the heart after ischemia. However, absence of CD69 led to a significantly increased infarct size 2 days we detected an increased number of CD69+ Tregs and higher after infarction, supporting the observation of greater tissue dam- CD69 expression on Tregs in the myocardium of MI mice ver- age in these mice (Figure 2, E and F). Two days after MI, wild-type sus that of sham-operated mice, indicating a selective migra- and CD69-deficient mice had several cardiac rhythm abnormali- tion of these cells to the myocardium after LAD ligation (Fig- ties, including ventricular premature depolarizations, first-degree ure 4B). The analysis of CD69 expression in different cardiac atrioventricular block, and ST-segment elevation. However, under cell populations revealed that CD45 –CD31+ endothelial cells, cardiac stress conditions induced by isoproterenol injection, the CD45+CD11b+ myeloid cells, and CD45+CD11b –CD4+Foxp3 – T Cd69 –/– mice had an enhancement of the above-mentioned cardiac effector cells did not have upregulated expression of CD69 after arrhythmias compared with their wild-type littermates, and, even- MI to the same extent as Tregs (Supplemental Figure 4, A and tually, bradyarrhythmias led to complete heart block, ventricular B). These data support a dominant role of CD69 in the Treg escape rhythms, and death (Supplemental Figure 2). Evaluation of compartment after MI. In the infarcted myocardium, γδT cells cardiac dysfunction by echocardiography indicated that the ven- peak 1 week after ischemia and are the main producers of IL-17 tricular wall motion score index increased in Cd69 –/– mice as the (23). Indeed, we confirmed that the majority of IL-17A–produc- disease progressed, reaching significance compared with wild-type ing cells were TCRγδ+ cells (~75%) in both Cd69+/+ and Cd69 –/– mice 1 month after infarction, supporting the idea of a worse prog- mice, although γδT cell and IL-17+ γδT cell infiltration into the nosis in Cd69 –/– mice after LAD ligation (Figure 2G). myocardium was significantly increased in Cd69 –/– mice as early Since most patients with MI are eventually reperfused during as 2 days after MI (Figure 4C). Neither Cd69+/+ nor Cd69 –/– mice admission, we analyzed the consequences of CD69 depletion in showed a significant recruitment of Th17 cells at this early time a mouse model of MI induced by ischemia and reperfusion (I/R). point (Supplemental Figure 4C). Furthermore, although we Similar to the data obtained with the permanent LAD ligation observed no differences in the number of CD11b+ myeloid cells model, Cd69–/– mice had increased cardiac dysfunction and worse or CD11b+Gr1hi granulocytes between genotypes, Cd69 –/– mice weight recovery after I/R, resulting also in slight decrease in sur- had a higher number of inflammatory Ly6Chi cells accumulated vival (Supplemental Figure 3). in the infarcted myocardium (Figure 4D and Supplemental 4 J Clin Invest. 2022;132(21):e152418 https://doi.org/10.1172/JCI152418 The Journal of Clinical Investigation   RESEARCH ARTICLE Figure 2. CD69 deficiency worsens heart damage and decreases survival after MI in mice. (A) Survival curve of mice after LAD ligation (n = 23–29 MI mice, n = 11–20 sham-operated mice). Data were pooled from 5 independent experiments and were analyzed by long-rank (Mantel-Cox) test. (B) Kinetics of the percentage of body weight loss after LAD ligation (n = 9–17 mice). Data represent the mean ± SEM and were analyzed by 2-way ANOVA with Šidák’s multiple-comparison test. (C) Representative images of infarcted hearts collected after intravenous injection of Evans blue dye 2 days after surgery. (D) Heart weight was normalized to body weight and tibia length 2 days after LAD ligation (n = 3–4 sham-operated mice; n = 5–6 MI mice). Data are repre- sentative of 3 independent experiments and indicate the mean ± SEM. Statistical significance was analyzed by 1-way ANOVA with Tukey’s post hoc test. (E) Representative images of heart slices showing the AAR (negative for Evans blue dye) in the upper panels and the extent of necrosis (negative for TTC staining) in the lower panels. (F) Histological quantification of the percentage of the LV AAR and the percentage of infarct size (IS) (n = 5–6 mice). Data are expressed as the mean ± SEM and were analyzed by unpaired Student’s t test. (G) Time course of LV dysfunction according to the WMSI measured by echocardiography (n = 6–16 MI mice, n = 4–8 sham-operated mice). Data were pooled from 3 independent experiments, represent the mean ± SEM, and were analyzed by 2-way ANOVA with Šidák’s multiple-comparison test. Asterisks denote differences between MI and sham-operated mice (light gray for Cd69+/+ mice, light red for Cd69–/– mice); ampersands denote differences between Cd69+/+ and Cd69–/– MI mice; plus signs denote differences between each day and day 0. &P < 0.05, **/&&P < 0.01, ***/&&&P < 0.001, and ****/++++P < 0.0001. Figure 4D), indicating a more proinflammatory phenotype (41). may be involved in limiting γδT cell activity. To test this hypoth- In parallel, the pathogenic IL-17+ γδT cell population was signifi- esis, we cocultured sorted Cd69+/+ or Cd69 –/– Tregs with Cd69+/+ cantly increased in the mediastinal lymph nodes draining the γδT cells from peripheral lymph nodes (Supplemental Figure 5, heart after infarction in Cd69 –/– mice (Supplemental Figure 4E). A and B). Our data indicate that CD69+ Tregs induced apoptosis CD69+ Tregs induce apoptosis and reduce IL-17A production in γδT cells more efficiently than did Cd69 –/– Tregs (Figure 5A). in γδT cells in a CD39-dependent manner in mice. Previous stud- Both Cd69+/+ and Cd69 –/– Tregs successfully decreased IL-17A ies provide evidence for antigen-independent inhibition of γδT production in a dose-dependent manner. However, at lower cells by Tregs (42, 43), although the mechanisms remain poorly numbers of Tregs, Cd69+/+ Tregs showed a more significant understood. Our data suggest that CD69 expression on Tregs reduction of IL-17A production (Figure 5B). J Clin Invest. 2022;132(21):e152418 https://doi.org/10.1172/JCI152418 5 RESEARCH ARTICLE The Journal of Clinical Investigation   CD39 is a membrane ectonucleotidase expressed in most and T effector cells, although CD69+ Tregs exhibited the highest Tregs that, together with CD73, transforms extracellular ATP into CD39 levels (Supplemental Figure 6B). Interestingly, infiltrating adenosine (44, 45). It has been shown that CD39 on Tregs can IL-17A+ γδT cells were reduced in the myocardium after transfer mediate the inhibition of innate (46) and nonimmune cells (47) of Cd69+/+ iTregs but not Cd69 –/– iTregs, supporting CD69-medi- by immunosuppression-independent mechanisms upon tissue ated inhibition of this population by Tregs (Figure 6D). In paral- injury. We evaluated whether the observed inhibition of γδT cells lel, myeloid cells, including proinflammatory Ly6Chi monocytes, by CD69+ Tregs after infarction was mediated by CD39. Our data were also reduced (Supplemental Figure 6D). All these data sug- show that CD69+ Tregs expressed higher levels of membrane gest that CD69-expressing Treg transfer reverted the immune- CD39 than did CD69 – or Cd69 –/– Tregs in peripheral blood, medi- mediated cardiac damage after MI in Cd69 –/– mice. Interesting- astinal lymph nodes, and cardiac infiltrate 2 days after MI (Figure ly, when ARL 67156 was administered in vivo to inhibit CD39, 5C). Sorted Cd69+/+ Treg also showed an increase in extracellu- Cd69+/+ iTregs lost their ability to restore survival, myocardial lar ATP consumption compared with Cd69 –/– Tregs. The addition inflammation, and IL-17A+ γδT cell accumulation in the heart (Fig- of ARL 67156, a chemical inhibitor of CD39, reduced ATP con- ure 6, E–G), suggesting that the therapeutic effects of Cd69+/+ iTreg sumption of Cd69+/+ Tregs and abolished the differences between transfer was CD39 dependent. Transfer of Cd69+/+ iTregs into genotypes (Figure 5D). These results support the idea that CD39 wild-type infarcted mice (Supplemental Figure 7A) also improved activity was impaired in CD69-deficient Tregs. Regarding the myocardial inflammation and survival to a greater extent than did observed reduction of IL-17A production in γδT cells by CD69+ the transfer of Cd69 –/– iTregs, although without reaching signifi- Tregs (Figure 5B), it was reverted in the presence of the inhibi- cance, probably due to the presence of endogenous CD69+ Tregs tor ARL 67156, whereas no effect was observed on Cd69 –/– Tregs (Supplemental Figure 7, B and C). Probably for the same reason, (Figure 5E). A direct effect of ARL 67156 on γδT cells was ruled we found no differences in myocardial γδT cells in the Cd69+/+ out, since no γδT cell inhibition was observed when ARL 67156 iTreg–transferred wild-type mice (Supplemental Figure 7D). was added in the absence of Tregs. Early expression of CD69 on Tregs is associated with a lower risk of The aryl hydrocarbon receptor (AhR) is one of the main factors HF development in patients with MI. We tested the prognostic value that mediates CD39 expression. Ligand-dependent activation of of CD69 expression after MI in those patients who completed fol- the AhR induces the expression of Entpd1 (CD39-encoding gene) low-up at the time this manuscript was written. The mean clini- through direct binding to XRE regions of the Entpd1 promoter cal follow-up time of our primary study population was 2.5 years (48). Our data also validated that Entpd1 mRNA levels, as well as (Supplemental Table 3). CD69 expression on Tregs, measured at CD39 activity, were reduced in Ahr–/– Tregs, supporting the notion the time of hospital admission during the index MI, was lower in that AhR is a dominant transcription factor for Entpd1 expression patients who were subsequently rehospitalized for HF (Figure 7A). in this cell type (Supplemental Figure 5, C and D). We have pre- Interestingly, the LV ejection fraction, but not CK or troponin T viously described that CD69 allows the uptake of AhR ligands levels, also decreased during MI in the patients who subsequently such as l-tryptophan in T cells by association with the aromatic– developed HF (Supplemental Figure 8A). After stratifying patients amino acid transporter complex on the membrane, allowing AhR according to high or low CD69 expression, as in Figure 1D, we activation (49). We observed that Tregs lacking CD69 exhibited observed that most patients who developed HF within the first 2.5 decreased expression of AhR transcriptional targets, includ- years of MI belonged to the group expressing low membrane CD69 ing Cyp1b1 and, importantly, Entpd1 (Figure 5F). Additionally, levels on Tregs (Figure 7B). Determination of CD69 mRNA expres- AhR-deficient CD69+ Tregs exhibited decreased CD39 expres- sion by quantitative PCR (qPCR) in total peripheral blood leuko- sion compared with wild-type Tregs (Supplemental Figure 5E), cytes (PBLs) also revealed an association with disease outcome, as supporting the idea that AhR signaling mediates CD39 expression most of those patients expressing low CD69 mRNA levels devel- in CD69+ Tregs. Together, these data suggest that CD69 expres- oped HF (Figure 7C). CD69 mRNA levels correlated with surface sion on Tregs allowed AhR-mediated expression of CD39, increas- CD69 levels, associating CD69 protein and transcript expression ing ATP conversion to adenosine, which triggered the inhibition of (Supplemental Figure 8B), and with FOXP3 mRNA levels, high- pathogenic IL-17+ γδT cells and controlled inflammation after MI. lighting a predominant CD69 expression by Tregs (Figure 7D). The Adoptive transfer of CD69-sufficient Tregs into Cd69 –/– mice equivalence in measuring CD69 by qPCR or by FACS supports the improves survival and reduces inflammation and cardiac dam- validity of using either method for the prediction of HF develop- age after MI. Next, we performed adoptive transfer exper- ment. Importantly, the percentage of CD69+ Tregs remained as an iments to assess the specific role of CD69 on Tregs in the independent predictor of the development of HF after adjusting for control of inflammation and recovery after infarction. In vitro– the levels of cardiac damage, age, and sex (Table 1). differentiated Tregs (iTregs) from naive Cd69+/+ and Cd69 –/– In parallel, we prospectively analyzed the association of early CD4+ T cells (Supplemental Figure 6A) were injected intrave- CD69 expression (during the first 24 hours of hospitalization for nously into Cd69 –/– mice 4–5 hours after LAD ligation, and mice MI) with clinical outcome after MI in an additional independent were monitored for 1 week (Figure 6A). Therapy with Cd69+/+ validation cohort of 84 patients with similar follow-up times (Sup- iTregs improved survival of Cd69 –/– mice after infarction (Figure plemental Tables 4 and 5). Quantification of CD69 mRNA levels in 6B). Furthermore, the heart/body weight ratio was preserved in frozen PBLs confirmed that patients who developed HF expressed the Cd69 –/– mice transferred with Cd69+/+ iTregs, as was the total lower CD69 levels at acute MI (Figure 7, E and F). Although data on number of leukocytes infiltrating the myocardium (Figure 6C). CD69 expression on the surface of Tregs were not available in this CD69 expression drives CD39 expression in myocardial Treg cohort, CD69 mRNA expression correlated significantly with Foxp3 6 J Clin Invest. 2022;132(21):e152418 https://doi.org/10.1172/JCI152418 The Journal of Clinical Investigation   RESEARCH ARTICLE J Clin Invest. 2022;132(21):e152418 https://doi.org/10.1172/JCI152418 7 RESEARCH ARTICLE The Journal of Clinical Investigation   Figure 3. Treg and IL-17A responses in the blood of mice after LAD Acute MI is the leading cause of mortality worldwide, with HF ligation. (A) Fold increase of the percentages of wild-type CD69+ Tregs, secondary to MI as one of the most important complications (51). wild-type CD69– Tregs, and Cd69–/– Tregs among CD4+ cells in peripher- In a follow-up study of 2 independent clinical cohorts, we found al blood 1 day after LAD ligation or sham surgery, compared with the percentages on day 0 (dotted line). Representative density plots of Tregs that patients who were rehospitalized because of HF had lower on day 0 and day 1 after MI are shown on the right (n = 10–20). Histograms CD69 expression levels within the first hours of acute MI. Even indicate the mean ± SEM, and data were analyzed by 1-way ANOVA with after adjusting for age, sex, and the level of cardiac damage, CD69 Tukey’s post hoc test. (B) Kinetics of IL-17A+ cells in peripheral blood after expression on Tregs remained as an independent predictor of HF surgery, expressed as a percentage of total cells. (C) Left: Representative development after MI. Thus, in agreement with the phenotype dot plots of IL-17A+ cells, with the percentages of total cells indicated in the outlined box. Right: Representative dot plots showing the main cell observed in mice, CD69 expression during MI was associated with populations positive for IL-17A. (D). Kinetics of the percentages of γδT a better clinical outcome in these patients. Although these results cells and IL-17+ γδT cells in peripheral blood after LAD ligation or sham have been validated in 2 independent cohorts with multivariate surgery. (E) Representative density plots of γδT cells in peripheral blood, adjustment, the number of patients who developed HF was small. with the percentages of cells indicated in the outlined box. Data in B and Therefore, our findings should be reproduced in larger cohorts, D are representative of 4 independent experiments and indicate the mean ± SEM (n = 6–10). Statistical significance was analyzed by 2-way ANOVA ideally including a longer post-MI time interval. with Šidák’s multiple-comparison test. Asterisks denote differences In order to shed light on the regulatory mechanism of CD69 between MI and sham-operated mice (light gray for Cd69+/+ mice, light red after MI, we explored the inflammatory signature in a mouse model for Cd69–/– mice); ampersands denote differences between Cd69+/+ and of LAD ligation. We found that the absence of CD69 resulted in a Cd69–/– MI mice; plus signs denote differences between each day and day rapid and detrimental excess of inflammatory infiltrating leukocytes 0. +/&P < 0.05, **/++/&&P < 0.01, ***/&&&P < 0.001, and ++++/&&&&P < 0.0001. in mice as early as 2 days after LAD ligation. Increased numbers of IL-17+ γδT cells characterize the cellular heart infiltrate of Cd69–/– expression, establishing an association between CD69 expression mice, being the main source of IL-17A early after MI. The increased and the Treg subset, as in the main patient cohort (Figure 7G). Thus, recruitment of IL-17+ γδT cells could explain the worse prognosis of these data support the idea that CD69 expression during acute MI is Cd69–/– mice after MI, as γδT cells induce cardiomyocyte apoptosis associated with a lower risk of developing HF after MI. (40), and their production of IL-17 promotes fibrosis, sustains neu- trophil/monocyte infiltration, and polarizes macrophages toward Discussion a proinflammatory phenotype (23). We observed that peripheral This study identifies CD69 as a key Treg receptor for controlling Cd69–/– γδT cells expressed markedly higher levels of IL-17A under immune-mediated myocardial damage early after MI. We basal conditions, suggesting that CD69 may also be playing a role in observed a specific deployment of CD69+ Tregs in the peripher- the generation of this T cell subset. We found that CD69+ Tregs were al blood of MI patients and LAD ligation mice. Moreover, CD69- substantially recruited to the infarcted myocardium, but Cd69–/– deficient mice and patients with low CD69 levels shortly after mice showed an even higher total number of Tregs in the myocardi- acute MI had a worse prognosis after the ischemic event. um at this time point, probably due to a parallel compensatory anti- The general idea is that Tregs are recruited to the myocardium inflammatory response to balance the tissue damage. during the first few days after MI to control excessive inflammation Interestingly, the accumulation of IL-17+ γδT cells in the myo- and prevent cardiac deterioration (7, 9, 10, 37, 50). We have recently cardium was impaired when Cd69+/+, but not Cd69–/–, Tregs were shown that CD69 expression on lymphocytes prevents atheroscle- transferred, suggesting that Tregs inhibited IL-17+ γδT cells by a rosis progression in mice and that low CD69 levels in PBLs pre- mechanism involving CD69. It is well established that Tregs sup- dict subclinical atherosclerosis in asymptomatic individuals after press effector CD4+ T cells and CD8+ T cells (52, 53). However, these adjustment for traditional cardiovascular risk factors (18). Here, we cell populations play a minor a role after MI. It has been described describe for the first time to our knowledge that CD69 expression that Tregs can reduce the proliferation of human phosphoantigen– is pivotal for the antiinflammatory properties of Tregs after isch- expanded γδT cells (43) and the proliferation and cytokine produc- emic myocardial damage. CD69 expression in T cells is an inflam- tion of murine intestinal γδT cells (42) in an antigen-independent matory brake that promotes Treg differentiation and suppressor manner, suggesting that Tregs are also capable of inhibiting this T function and prevents proinflammatory T cell responses in multiple cell subset. Our data show that mouse γδT cells underwent apop- disease scenarios (38). In this study, we describe a mechanism by tosis and decreased IL-17A production in vitro in the presence of which CD69 expression on Tregs promoted antigen-independent Tregs, indicating an inhibitory effect of Tregs on γδT cells. This inhibition of proinflammatory γδT cells after MI. CD69 induced effect was dose dependent and was mediated by CD69, as Cd69–/– AhR-dependent CD39 expression and extracellular ATP conversion Tregs showed a lower inhibitory capacity. These results may explain to adenosine, inhibiting effector cytokine production and inducing the decrease in the number of IL-17A+ γδT cells in the infarcted myo- apoptosis in γδT cells infiltrating the ischemic myocardium. cardium after the adoptive transfer of Cd69+/+ Tregs. Mice lacking CD69 showed increased myocardial inflamma- In order to further explore the molecular basis of this rapid tion and dysfunction, leading to a rapid decrease in survival during inhibition of γδT cells, we examined the CD39-mediated, anti- the first week after infarction. Therapy with CD69+ Tregs in the gen-independent mechanisms of Treg suppression. CD39 is an first hours after LAD ligation reduced myocardial inflammation ectonucleotidase that hydrolyzes extracellular ATP, released from and improved the survival of Cd69 –/– mice, indicating that CD69 damaged tissue, to AMP, which is further degraded to adenosine expression specifically on the Treg subset was sufficient to allevi- by CD73. Uptake of adenosine by innate and T effector cells pro- ate the recovery after MI. motes apoptosis and cell inhibition (45, 46), but uptake by Tregs 8 J Clin Invest. 2022;132(21):e152418 https://doi.org/10.1172/JCI152418 The Journal of Clinical Investigation   RESEARCH ARTICLE Figure 4. Myocardial accumulation of CD69+ Tregs and Il-17+ γδT cells after LAD ligation. (A) Leukocyte cell numbers per milligram of heart tissue in the myocardium 2 days after infarction. (B) Quantification of the number of Tregs (CD4+Foxp3+) and CD4+Foxp3+CD69+ cells per milligram of heart tissue and CD69 mean fluorescence expression on Tregs in the heart. Representative density plots showing gating on CD45+CD11b–CD4+ cells. (C) Representative density plots gated on CD45+CD11b–CD3+ cells and numbers of γδT cells and Il-17+ γδT cells per milligram of tissue. (D) Quantification of total cell numbers per milligram of CD11b+ myeloid cells, CD11b+Gr1hi cells, and CD11b+F4/80loLy6Chi cells in the heart. Heart cell–infiltrating populations were evaluated 2 days after infarction (n = 6–11 animals per group). Data are representative of 4 independent experiments and indicate the mean ± SEM. Statistical significance was analyzed by 1-way ANOVA with Tukey’s post hoc test. P values for significant differences are shown. J Clin Invest. 2022;132(21):e152418 https://doi.org/10.1172/JCI152418 9 RESEARCH ARTICLE The Journal of Clinical Investigation   induces antiinflammatory properties (54). Tregs from patients with had a confirmed MI after extensive diagnosis and met the inclusion MI overexpress CD39 early after revascularization (10). Our work criteria were included in the analyses. Data from patients with an shows that CD69+ Tregs infiltrating the peripheral blood, draining uncertain diagnosis were excluded. Blood samples from patients lymph nodes, and heart expressed higher levels of membrane CD39 processed more than 24 hours after extraction were also excluded than did CD69– or Cd69–/– Tregs 2 days after MI. Since CD69 facil- from the study, since degradation of blood components might have itates the entrance of AhR ligands (49), the increased expression altered the sample. The number of human samples was determined and activity of CD39 (a transcriptional target of AhR) in CD69-suf- according to an estimation of clinical relevance. The incidence of ficient Tregs could be explained by the observed increased activa- HF and death was assessed in the patients with MI for whom FACS tion of the AhR pathway. In agreement, different studies described was measured and a follow-up was available at the time of the writ- that CD39 confers protection after infarction, showing that CD39 ing of this manuscript. A similar number of healthy volunteers with overexpression ameliorates the progression of myocardial damage normal ECG and heart function measured by echocardiography after MI (55–57) or that Cd39–/– mice develop exacerbated cardiac were included in the study as comparators. damage (58). Moreover, adoptive transfer of Cd39–/– iTregs fails to Human blood sample collection included in the study. Blood samples prevent inflammation and achieve cardiac protection after MI (10). from patients with MI were prospectively collected from the arterial Thus, Tregs deficient in either CD69 or CD39 present similar dys- sheath used for coronary angiography and percutaneous coronary functional properties early after MI. Consistently, we report that intervention prior to the administration of heparin, between March chemical inhibition of CD39 in vitro and in vivo resulted in impaired 2017 and May 2019. All demographic, clinical, angiographic, and proce- suppression of γδT cells by CD69-sufficient Tregs, suggesting that dural characteristics of these patients with MI were also prospectively CD69+ Tregs exert their inhibition of γδT in a CD39-dependent collected. Revascularization and subsequent clinical management manner. All this evidence suggests a novel molecular mechanism in the coronary care unit and in the cardiology ward were performed of Treg protection against inflammation-mediated cardiac impair- following the guideline recommendations for MI (30) (Supplemental ment, although further exploration in a clinical setting is needed. Tables 1 and 3). All patients with MI were systematically followed at a In conclusion, we have identified CD69 as a crucial orches- dedicated outpatient clinic. trator of the cardioprotective function of Tregs after MI. Although In parallel, we analyzed blood samples from 80 healthy donors additional studies are needed to prospectively confirm the diag- without any cardiac disease, as assessed by normality of cardiac mark- nostic and therapeutic value of circulating CD69+ Tregs, we ers, electrocardiography, and echocardiography. believe this target opens new, specific cellular avenues for preci- The validation cohort consisted of a consecutive series of patients sion medicine to improve the management of patients with MI. admitted for MI to the Hospital de la Santa Creu i Sant Pau, Barcelo- na, Spain. Baseline clinical characteristics, clinical presentation, and Methods in-hospital evolution are summarized in Supplemental Table 4. All Study design. The overall objective of our study was to investigate the patients were followed, and the predefined adverse clinical outcomes relevance of CD69-expressing Tregs in the progression of myocardi- (HF and all-cause death) are summarized in Supplemental Table 5. al damage after MI. For this purpose, we analyzed the expression of Blood samples were collected from patients within the first 24 hours of CD69 on Tregs in peripheral blood from patients with MI. In addition, their hospital admission and prior to heparin administration. to further elucidate the mechanism of action of these cells, we per- Mice. CD69-deficient mice were generated as described previ- formed permanent LAD ligation to induce MI in Cd69 –/– mice and their ously (15). Wild-type or CD69-deficient mice used for experiments Cd69+/+ littermates. Peripheral immune response and heart inflam- were 8- to 12-week-old males on a C57Bl/6 background. Foxp3-RFP/ mation, necrosis, and function were evaluated. Cd69 –/– infarcted mice IL17-eGFP reporter mice were provided by R.A. Flavell (Yale Universi- were also transferred with CD69-sufficient iTregs to narrow down the ty, New Haven, Connecticut, USA) and crossed with Cd69+/+ or Cd69+/+ protective effect of CD69 to the Treg compartment. Further in vitro mice. Different experimental groups were sex and age matched with experiments adding Tregs to γδT cell cultures were performed to prove the offspring of littermate mice. Animals were housed and used under a direct CD69-dependent inhibitory effect on γδT cells. specific pathogen–free (SPF) conditions at the CNIC animal facility. The genotyping of mice was recorded throughout the entire period Mouse model of MI. For MI induction, a permanent ligation of of the study. Mice were randomly assigned to each experimental group, the LAD coronary artery was performed. Briefly, mice were anes- and the only criteria used to stratify them was the genotype. Infarct thetized with sevoflurane (4%) and intubated using a 24 gauge cath- size quantification was performed in a blinded manner. For survival eter with a blunt end. Mice were artificially ventilated with a mixture plots, mice that died during the surgical procedure were excluded. of O2 and air (1:1, vol/vol) using a rodent ventilator (Minivent 845, Animals were excluded that had undetectable/marginal infarct zones Harvard) with 160 strokes/minute in a total volume of 250 μL. Mice by echocardiography 2 days after surgery, a result that was probably were placed on a heating pad to maintain body temperature at 37°C. due to an unsuccessful LAD ligation. For all experiments, sample sizes A thoracotomy was performed through the fourth left intercostal were determined by our previous data from similar experiments and space, the pericardium was opened, and the heart was exposed. The prior literature to ensure statistical relevance. At least 3 biological rep- proximal LAD coronary artery was localized and permanently ligat- licates were used for each assay, and each experiment was performed ed by passing a 7-0 silk suture around the artery. Sham-operated at least 3 times when applicable. mice were analyzed in parallel as surgical controls. All blood samples from patients suspected of meeting all the MI Mouse infarct size quantification. Animals were anesthetized and 0.5 inclusion criteria were processed in a blinded manner upon hospital mL of 1% (w/v) Evans blue dye was infused intravenously through the admission of the patient, during the study period. Only patients who vena cava 36 hours after surgery. We rapidly opened the thoracic cavity 10 J Clin Invest. 2022;132(21):e152418 https://doi.org/10.1172/JCI152418 The Journal of Clinical Investigation   RESEARCH ARTICLE J Clin Invest. 2022;132(21):e152418 https://doi.org/10.1172/JCI152418 11 RESEARCH ARTICLE The Journal of Clinical Investigation   Figure 5. Expression of CD39 by CD69+ Tregs after MI mediates the inhibition frame rate of more than 230 frames/second. Mice were lightly anes- of γδΤ cells. (A) Sorted wild-type γδT cells were cocultured for 24 hours with thetized with 0.5%–2% isoflurane in oxygen, adjusting the isoflu- Cd69+/+ or Cd69–/– sorted Tregs at the indicated γδT/Treg ratios. Apoptosis of rane delivery to maintain the heart rate at 450 ± 50 bpm. Mice were γδT cells, represented as the fold increase of annexin V+ (Ann. V+) γδT cells versus γδT cells alone (ratio 1:0) (n = 4–10). A representative zebra plot of the placed in the supine position and maintained at normothermia using 1:0.5 ratio, gated on γδT cells is shown. (B) Inhibition of IL-17A production by a heating platform and warmed ultrasound gel. A base apex ECG was γδT cells, plotted as the fold change of IL-17A+ γδT cells for each ratio versus continuously monitored. Longitudinal and short-axis views of the the 1:0 ratio (n = 6–12). Representative zebra plots of the 1:0.5 ratio, gated on left ventricle were acquired at the papillary muscle level for M-mode γδT cells, are shown. Data in A and B were pooled from 4 independent experi- and also medium and apical levels for 2D. For the analysis of the wall ments. Data indicate the mean ± SEM. Statistical significance was deter- mined by 2-way ANOVA with Sidak’s multiple-comparison test. Significant P motion score using echocardiography, the regional LV function was values are shown (black for Cd69+/+ Tregs and red for Cd69–/– Tregs). (C) CD39 evaluated in the parasternal long-axis view. The LV wall was subdivid- expression on Tregs in PBLs, mediastinal lymph nodes (Med-LN), and heart ed into 6 segments (basal, mid, and apical in the anterior and posterior was measured by FACS, 2 days after MI (n = 4–5). Data indicate the mean ± walls). Each segment was scored in a blinded manner by an indepen- SEM. Statistical significance was determined by 1-way ANOVA with Tukey’s dent evaluator on the basis of its motion and systolic thickening, using post hoc test. (D) Extracellular ATP was measured in the supernatant of isolated Tregs, in the presence or absence of ARL 67156 (ARL) at the indicated the following guidelines of the American Society of Echocardiography time points after ATP supplementation (n = 3–5). Data are from 1 representa- (59): (a) normal or hyperkinetic; (b) hypokinetic (reduced thickening); tive independent experiment of 4 experiments and indicate the mean ± SEM. (c) akinetic (absent or negligible thickening, e.g., scar); and (d) dys- Statistical significance was determined by mixed-effects 2-way ANOVA with kinetic (systolic thinning or stretching, e.g., aneurysm). The number Šidák’s multiple-comparison test. (E) IL-17A production by sorted γδT cells in of dysfunctional segments was quantified, and the total wall motion the presence of Tregs and/or ARL 67156, fold change versus the percentage of IL-17A+ γδT cells alone (n = 3–7). Data are representative of 3 independent score index (WMSI) representing the sum of the score of the 6 individ- experiments and indicate the mean ± SEM. Statistical significance was deter- ual segments in each heart was calculated. mined by 2-way ANOVA with Šidák’s multiple-comparison test. (F) Quanti- Isolation of infiltrating cells from mouse heart. Hearts were perfused fication of Ahr, Cyp1b1, and Entpd1 mRNA levels in Tregs by qPCR (n = 6–7). with 10 mL PBS and removed from the chest cavity. The hearts were then Data were pooled from 2 independent experiments and represent the mean ± minced and digested with collagenase IV (100 U/mL; Gibco, Thermo SEM. Statistical significance was determined by unpaired Student’s t test. Fisher Scientific) for 45 minutes at 37°C under constant agitation. The resulting cell suspensions were filtered through 40 μm cell strainers (BD to perfuse hearts with PBS and washout the excess of Evans blue dye. Falcon) and washed twice with PBS, 0.5% BSA, and 1 μM EDTA. Eryth- The heart was then harvested and rinsed, and the atria were removed. rocytes were removed using hypotonic buffer, and the number of leuko- The hearts were photographed before being cut into transverse slices cytes was assessed. Single-cell suspensions were stained as described (n = 4–5 per ventricle). Then, slices were weighed. The palish negative below, and the different cell populations were analyzed by FACs. area for Evans blue delineated the area at risk (AAR), i.e., myocardium FACS analysis. PBLs were isolated from mouse or human blood lacking blood flow. In order to differentiate infarcted from viable tis- samples using Ficoll-Isopaque (density = 1.121 g/mL) gradient centrifu- sue, the same slices were incubated in TTC (1% [w/v] in PBS) at 37°C gation. Human/mouse PBLs or single-cell suspensions of mouse lymph for 15 minutes. The sections were then rephotographed and weighed. nodes or heart-infiltrating leukocytes were incubated in PBS plus 0.05% After the incubation with TTC, Evans blue staining cleared out, and BSA plus 0.01% EDTA buffer with fluorochrome-conjugated antibod- the sections showed 2 areas: 1 necrotic (palish negative for TTC stain- ies. PBLs were stained in fresh for myeloid cell markers and cultured ing: infarcted myocardium) and 1 alive (red positive for TTC staining). overnight for bystander activation in plates coated with 3 μg/mL puri- Regions negative for Evans blue staining (AAR) and for TTC (infarcted fied anti-CD3 antibody (for human cells: ΟΚΤ3 clone, BioLegend; for myocardium) were quantified in a blinded manner using ImageJ with mouse cells: 145-2C11 clone, BD Pharmingen) in complete RPMI medi- FIJI image processing package (NIH). The percentages of AAR and um (20% FBS, Gibco, Thermo Fisher Scientific) prior to T cell staining. infarcted myocardium were weighted and corrected independently for For human Foxp3 evaluation, nuclear staining was performed each slice, and the total milligrams of AAR and TTC-negative regions using the Foxp3 staining buffer set (Miltenyi Biotec) according to the was calculated for each heart. The AARs were determined according manufacturer’s instructions. to the mg/mg ratio of AAR/ventricle weight, and the infarct size was For cytokine production assessment, cells were stimulated with determined by the infarcted myocardium/AAR ratio. 50 ng/mL PMA (MilliporeSigma), 1 μg/mL ionomycin (MilliporeSig- Electrocardiography in mice. Two days after LAD ligation, mice ma), and 1 μg/mL GolgiPlug (BD Pharmingen) in complete culture were anesthetized with 0.5%–2% isoflurane in oxygen, administered medium for 4 additional hours. For cytokine-producing Th cell eval- via the nose cone, with adjustment of isoflurane delivery to maintain uation, cells were fixed with PBS 2% paraformaldehyde for 10 min- the heart rate around 500 ± 50 bpm. Surface ECGs were obtained utes at room temperature and intracellularly stained with conjugated by using unipolar and bipolar limb leads for 5 minutes under basal antibodies in PBS with 0.05% BSA, 0.01% EDTA, and 0.5% saponin. conditions (Biopac Systems). Then, 1.5 mg/kg isoproterenol (Milli- The following monoclonal antibodies were used for staining of poreSigma) was injected intraperitoneally, and ECGs were monitored different markers in human samples: anti-CD14 (clone: M5E2, Bio- for another 15–25 minutes. ECGs were blindly analyzed by an expert Legend); anti-CD16 (clone: DJ130, Dako); anti-CD25 (clone: 2A3, BD using Acknowledge, version 4.1.1, for MP36R (BIOPAC Systems). Biosciences); anti-CD4 (clone: SK3, BD Biosciences); anti-CD45RA Echocardiography in mice. Transthoracic echocardiography was (clone: H100, BD Biosciences); anti-CD45RO (clone: UCHL1, BD Bio- performed in a blinded manner by an expert operator using a high-fre- sciences); anti-CD66b (clone: G10F5, BioLegend); anti-CD69 (clone: quency ultrasound system (Vevo 2100, Visualsonics) with a 30 MHz FN50, BD Biosciences), anti-Foxp3 (clone: 3G3, Miltenyi Biotec); linear probe. 2D and M-mode echocardiography were performed at a anti–HLA-DR (clone: G46-6,BD Biosciences); anti–IFN-γ (clone: B27, 12 J Clin Invest. 2022;132(21):e152418 https://doi.org/10.1172/JCI152418 The Journal of Clinical Investigation   RESEARCH ARTICLE Figure 6. Adoptive transfer of CD69-sufficient Tregs into Cd69–/– mice reduces myocardial inflammation and improves survival after MI. (A) Sche- matic workflow of the iTreg adoptive transfer after LAD ligation. (B) Survival after LAD ligation (n = 7–24). Black arrow depicts the time of iTreg inocula- tion (4–5 hours after infarction). Cd69–/– mice without cell transfer were used as controls. The P value was calculated by long-rank (Mantel-Cox) test. (C) Heart/body weight ratio and total leukocyte numbers per milligram of heart tissue 7 days after LAD ligation (n = 5–6). (D) Representative density plots (gated on CD45+CD11b– cells) and numbers of γδT cells and IL-17+ γδT cells per milligram of myocardial tissue. Data in C and D correspond to 1 represen- tative independent experiment of 3 experiments. Data indicate the mean ± SEM and were analyzed by 1-way ANOVA with Tukey’s post hoc test. (E) Survival after LAD ligation (n = 12–13), Cd69+/+ iTreg transfer, and ARL 67156 or saline administration. Black arrow depicts the time of iTreg inoculation (4–5 hours after infarction). The P value was calculated by long-rank (Mantel-Cox) test. (F) Total leukocyte numbers per milligram of heart tissue 7 days after LAD ligation (n = 9). (G) Numbers of γδT cells and IL-17+ γδT cells per milligram of myocardial tissue (n = 9). Data in F and G were pooled from 4 independent experiments and indicate the mean ± SEM. Statistical significance was determined by unpaired Student’s t test. J Clin Invest. 2022;132(21):e152418 https://doi.org/10.1172/JCI152418 13 RESEARCH ARTICLE The Journal of Clinical Investigation   Figure 7. High CD69 expression in patients early after MI is associated with a decreased risk of developing HF. (A) After 2.5 years of clinical follow-up, patients from the main study cohort were stratified according to whether or not they developed HF. CD69 expression on Tregs at the time of hospital admission for acute MI in patients who developed HF (n = 7) or did not develop HF (n = 180). Data were analyzed by Mann-Whitney U test. (B) Percent- age of patients with low or high levels of surface CD69 expression on Tregs, measured by FACS in the main study cohort. The P value was calculated using a χ2 test. (C) Frequency of patients with low or high levels of CD69 mRNA expression, measured by qPCR in the main study cohort (the mean nor- malized CD69 2–ΔCt values were used to discriminate patients expressing low or high levels of CD69). (D) Correlation of FOXP3 and CD69 mRNA expres- sion in PBLs from individuals in the main study cohort. Spearman’s correlation coefficient (r) and P values are shown. (E) CD69 mRNA levels measured by qPCR in total PBLs from the independent validation cohort of patients (n = 75 with no HF and n = 9 with HF). Data were analyzed by Mann-Whitney U test. (F) Frequency of patients with low or high levels of CD69 mRNA, measured by qPCR in the validation cohort (the mean normalized CD69 2–ΔCt values were used to discriminate patients with low or high CD69 expression). The P value was calculated using a χ2 test. (G) Correlation of FOXP3 and CD69 mRNA expression in PBLs from individuals in the validation cohort. Spearman’s correlation coefficient (r) and P values are shown. BD Biosciences); anti–IL-17A (clone: SCPL1362, BD Biosciences); and bodies against CD44, CD8, MHC class II (I-Ab), CD19, B220, IgM, anti–IL-22 (clone: 142928, R&D Systems). For mouse samples, the fol- CD11b, CD11c, and DX5 and subsequently with Streptavidin Microbe- lowing antibodies were used: anti-CD11b (clone: M1/70, BD Bioscienc- ads (MACS, Miltenyi Biotec). Naive CD4+ T cells were negatively select- es); anti-CD3 (clone: 145-2C11, BioLegend); anti-CD4 (clone: RM4-5, ed in MACS LD columns (Miltenyi Biotec) according to the manufactur- BD Biosciences); anti-CD45.2 (clone: 104, BioLegend); anti-CD69 er’s instructions. The naive status was confirmed by expression of CD4 (clone: H1.2F3, BD Biosciences); anti-F4/80 (clone: BM8, BioLegend); and CD62L by flow cytometry (data not shown). Naive CD4+ T cells anti-Gr1 (clone: RB6-8C5, BD Biosciences); anti–IL-17A (clone: TC11- (106 cells/mL) were cocultured for 72 hours with irradiated (30 Gy) anti- 18H10, BD Pharmingen); anti-Ly6C (clone: AL-21, BD Biosciences); gen-presenting cells in the presence of plate-bound anti-CD3 (2 μg/mL) and anti-TCRγδ (clone: GL3, BioLegend). and soluble anti-CD28 (2 μg/mL) plus recombinant TGF-β1 (10 ng/mL) Cells were analyzed in a LSRFortessa or FACSymphony Flow and IL-2 (2 ng/mL) in RPMI medium (Gibco, Thermo Fisher Scientific). Cytometer, and the data were processed with FlowJo, version After determining the purity of CD4+CD25+Foxp3+ cells, 1.5 × 106 iTregs 10.0.4 (TreeStar). were intravenously injected into Cd69–/– mice 4–6 hours after LAD liga- Adoptive transfer of Tregs. Naive CD4+ T cells were purified from tion. In the indicated experimental groups, ARL 67158 (0.26 mg/kg single-cell suspensions of the spleen and mesenteric lymph nodes of mice, Tocris) was intraperitoneally administered in PBS 4–6 hours, 48 Foxp3-mRFP/IL17-eGFP/Cd69+/+ or Foxp3-mRFP/IL17-eGFP/Cd69–/– hours, and 96 hours after MI. Mice treated with PBS alone were used as reporter mice. Cell suspensions were incubated with biotinylated anti- controls. Mice were sacrificed 7 days after infarction. 14 J Clin Invest. 2022;132(21):e152418 https://doi.org/10.1172/JCI152418 The Journal of Clinical Investigation   RESEARCH ARTICLE tific) for Cd69 (forward: 5′-ATTGTCCAGGCCAATACACATT-3′, Table 1. Multivariable logistic regression model to discriminate reverse: 5′-CCTCTCTACCTGCGTATCGTTTT-3′); Foxp3 (for- patients with and without HF, adjusting for potential confounders ward: 5′-GAGAAGCTGAGTGCCATGCA-3′, reverse: 5′-GGAGC- OR (95% CI) P value CCTTGTCGGATGAT-3′); Gapdh (forward: 5′-GAAGGTGAAG- (Intercept) 0.000 (0.000–0.136) 0.0938 GTCGGAGTC-3′, reverse: 5′-GAAGATGGTGATGGGATTTC-3′); Troponins (normalized) 0.992 (0.974–1.001) 0.2049 and Actb (forward: 5′-CATCGAGCACGGCATCGTCA-3′, reverse: LV ejection fraction (%) 0.907 (0.759–1.044) 0.1888 5′-TAGCACAGCCTGGATAGCAAC-3′). Real-time qPCR analyses CK (U/I) 1.001 (0.999–1.005) 0.4089 were performed with an ABI Prism 7900HT SDS 384-well thermal CD69+ Tregs (%) 0.929 (0.838–0.980) 0.0409 cycler (Applied Biosystems). Relative gene expression was deter- Sex 11.1 (0.404–4765.4) 0.2471 mined using the 2–ΔCt method and normalized to both Actb and Gapdh Age 1.430 (1.128–2.482) 0.0444 housekeeping genes. A general linear model was used to assess the capability of variables to Statistics. For statistical analysis, the normality of the distributions predict patients who would develop HF (n = 7) versus those who would not was first evaluated using the Shapiro-Wilk test for mouse experiments (n = 180). The model considered normalized troponin values, LV ejection and the D’Agostino-Pearson test for patient samples with higher n fraction, percentage of peripheral CD69+ Tregs, sex, and age. values. If normal, an unpaired Student’s t test was used for 2-group comparisons or a 1-way ANOVA with Tukey’s post hoc test when more than 2 groups were compared. If distributions were non-normal, a Cocultures of Tregs and γδT cells. Single-cell suspensions of periph- Mann-Whitney’s U test was used for the analysis of 2 groups and a Kru- eral (axillary, inguinal, and submaxillary) lymph nodes from Foxp3- skal-Wallis test with Dunn’s post hoc test for multiple comparisons. In mRFP/IL17-eGFP/Cd69+/+ or Foxp3-mRFP/IL17-eGFP/Cd69 –/– report- the kinetic experiments, a 2-way ANOVA with Šidák’s multiple-com- er mice were stained with fluorochrome-conjugated antibodies. Tregs parison post hoc test was performed. Spearman′s correlation coeffi- (CD3+Foxp3-mRFP+) and γδT cells (CD3+ TCRγδ+ cells) were sorted cient was used to analyze the independence of continuous variables. using a BD FACSAria II cell sorter. Both cell populations were cocul- Associations between nominal variables were calculated by χ2 test. tured at different ratios for 24 hours in the presence of plate-bound Data analyses were performed with GraphPad Prism 8.0 (GraphPad anti-CD3 (2 μg/mL) and soluble anti-CD28 (2 μg/mL) plus recombi- Software). To assess the association of CD69+ Tregs with the develop- nant IL-2 (10 ng/mL) with or without ARL 67158 (250 μM, Tocris). ment of HF, multivariate logistic regression analysis was performed, Apoptosis of γδT cells was evaluated by flow cytometry using annexin V adjusting for sex, age, troponin levels (normalized values), CK, and (BD Biosciences). For IL-17A–eGFP production assessment, cells were ejection fraction with R statistical software, version 3.6.2. In general, incubated for 4 additional hours in the presence of PMA (MilliporeSig- P values below 0.05 were considered significant. ma), ionomycin (MilliporeSigma), and GolgiPlug (BD Pharmingen). Study approval. Verbal informed consent was obtained from Extracellular ATP consumption assay. Sorted Tregs (CD4+- patients who required emergency coronary angiography and primary Foxp3-mRFP+; 5 × 104 cells) from spleens were activated for 3 days angioplasty, and written informed consent was obtained immediately in the presence of plate-bound anti-CD3 (3 μg/mL) and soluble anti- after the procedure. All participants included in the study were identi- CD28 (2 μg/mL) plus recombinant IL-2 (10 ng/mL) in 96-well U-fond fied by number and provided authorization for the use of their medi- plates. After washing with PBS, cells were incubated for 40 minutes cal records for research. The study was approved by the research ethics at 37°C in complete RPMI medium (20% FBS, Gibco, Thermo Fisher committees of the Universitario de la Princesa and Santa Creu i Sant Pau Scientific) in the presence of ATP (50 μM) and, when indicated, ARL hospitals. All animal procedures were approved by the ethics committee 67158 (250 μM, Tocris). Supernatants (50 μL) were collected at 20 and of the Comunidad Autónoma de Madrid and conducted in accordance 40 minutes for extracellular ATP measurements. Cells were washed with the institutional guidelines that comply with European Institutes of and used for FACS or qPCR analyses. Health directives (European Institutes of Health, 2010). The ATP concentration was quantified from the supernatants by firefly luciferase assay in a 96-well luminescence reading plate Author contributions (Costar). Cell supernatant (50 μL) was diluted in 130 μL Buffer A (150 PM and RBD contributed to the study design, data analysis, and mM KCl, 25 mM Tris HCl, 1 mM EDTA, 0.1 % BSA fatty acid free, manuscript preparation. RBD, LMA, and A Curtabbi executed and 10 mM KH2PO4, pH 7.4). The reaction was started by adding 20 μL analyzed preclinical experiments. RBD, IRA, RSD, RJA, and LMA luciferin/luciferase cocktail (0.5 M tris-acetate, 0.8 mM luciferin, 10 contributed to patient blood sample processing and FACS analysis. μg/mL luciferase, pH 7.4). Bioluminescent signal was read by Orion RBD analyzed the human data and coordinated the clinical data- Microplate Luminometer with Simplicity 4.2 software. A regression base. MVO, JAE, SCR and FSM provided scientific guidance. HDLF, curve for the ATP concentration was calculated by a standard ATP CR, MMGG, AV, FR, JC, LJJB, A Cecconi, ADC, MT, JA, HB, FA, curve measured in parallel in the same assay buffer. and JMG provided clinical guidance and/or contributed to patients’ RNA extraction and qPCR. RNA was extracted from frozen sample collection and data analysis. PM contributed to study con- human PBLs with a MiRNeasy Mini Kit (QIAGEN) as recommend- ception. All authors read and approved the final manuscript. ed by the manufacturer. Reverse transcription was performed with 200 ng DNAse-treated RNA using the High Capacity cDNA Acknowledgments RT Kit (Applied Biosystems). Then, gene expression was mea- We thank Richard A. Flavell (Yale University) for providing the sured by real-time qPCR using SYBR Green PCR Mix (Applied Foxp3-mRFP/Il17 eGFP double-reporter mice. We thank members Biosystems) and mRNA-specific primers (Thermo Fisher Scien- of the Veterinary Service of CNIC’s Comparative Medicine Unit, J Clin Invest. 2022;132(21):e152418 https://doi.org/10.1172/JCI152418 15 RESEARCH ARTICLE The Journal of Clinical Investigation   Insourcing Solution Spain Charles River Laboratories for perform- by a Human Frontier Science Program grant (RGP0016/2018) ing the LAD ligation surgeries in mice. We thank Ana Vanesa Alon- and a Leducq Transatlantic Networks grant (17CVD04), to JAE. so, Lorena Flores, and Eva Sánchez de Rojas for their excellent AC is supported by a Marie Skłodowska-Curie grant (agreement technical work with the echocardiography and analysis in mice. no. 713673). RBD is supported by the Formación de Profesorado We also thank all the patients and volunteers for agreeing to partic- Universitario (FPU16/02780) program of the Spanish Ministry ipate in this study. This study was supported by competitive grants of Education, Culture, and Sports. The CNIC is supported by the from the Ministerio de Ciencia e Innovación (MCIN), through the Instituto de Salud Carlos III (ISCIII), the Ministerio de Ciencia e Carlos III Institute of Health (ISCIII) – Fondo de Investigación Innovación (MCIN), and the Pro CNIC Foundation) and is a Seve- Sanitaria (PI22/01759), to PM; grant PID2021-122509OB-I00 ro Ochoa Center of Excellence (grant CEX2020-001041-S fund- funded by MCIN/ AEI/10.13039/501100011033 and by “ERDF ed by MICIN/AEI/10.13039/501100011033). A way of making Europe” to JMG; Comunidad de Madrid grants S2017/BMD-3671-INFLAMUNE-CM, to PM and FSM; a grant Address correspondence to: Pilar Martín, Vascular Pathophysiology from the Fundació La Marató TV3 (20152330 31), to JMG and Area, Centro Nacional de Investigaciones Cardiovasculares (CNIC). FSM; Ministerio de Ciencia e Innovación (MCIN) grant RTI2018- Calle Melchor Fernández Almagro, 3. 28029 Madrid, Spain. Phone: 099357-B-I00 and a CIBERFES grant (CB16/10/00282); and 34.914531200; Email: pmartinf@cnic.es. 1. Swirski FK, Nahrendorf M. Leukocyte behavior in 15. Cortes JR, et al. Maintenance of immune toler- with unstable angina and acute myocardial atherosclerosis, myocardial infarction, and heart ance by Foxp3+ regulatory T cells requires CD69 infarction. Atherosclerosis. 2003;168(1):73–80. failure. Science. 2013;339(6116):161–166. expression. J Autoimmun. 2014;55:51–62. 29. Pasqui AL, et al. T cell activation and enhanced 2. Swirski FK, Nahrendorf M. Cardioimmunology: 16. Sanchez-Diaz R, et al. Thymus-derived regula- apoptosis in non-ST elevation myocardial infarc- the immune system in cardiac homeostasis and tory T cell development is regulated by C-type tion. Clin Exp Med. 2003;3(1):37–44. disease. Nat Rev Immunol. 2018;18(12):733–744. lectin-mediated BIC/microRNA 155 expression. 30. Ibanez B, et al. 2017 ESC Guidelines for the man- 3. Hofmann U, et al. Activation of CD4+ T lympho- Mol Cell Biol. 2017;37(9):e00341. agement of acute myocardial infarction in patients cytes improves wound healing and survival after 17. Martin P, Sanchez-Madrid F. CD69: an unexpect- presenting with ST-segment elevation: The Task experimental myocardial infarction in mice. Cir- ed regulator of TH17 cell-driven inflammatory Force for the management of acute myocardial culation. 2012;125(13):1652–1663. responses. Sci Signal. 2011;4(165):pe14. infarction in patients presenting with ST-segment 4. Hofmann U, Frantz S. Role of T-cells in myocar- 18. Tsilingiri K, et al. Oxidized low-density lipopro- elevation of the European Society of Cardiology dial infarction. Eur Heart J. 2016;37(11):873–879. tein receptor in lymphocytes prevents atheroscle- (ESC). Eur Heart J. 2018;39(2):119–177. 5. Weiß E, et al. Myocardial-Treg crosstalk: how to rosis and predicts subclinical disease. Circula- 31. Shiow LR, et al. CD69 acts downstream of tame a wolf. Front Immunol. 2022;13:914033. tion. 2019;139(2):243–255. interferon-alpha/beta to inhibit S1P1 and lym- 6. Rieckmann M, et al. Myocardial infarction triggers 19. Cruz-Adalia A, et al. CD69 limits the severity of phocyte egress from lymphoid organs. Nature. cardioprotective antigen-specific T helper cell cardiomyopathy after autoimmune myocarditis. 2006;440(7083):540–544. responses. J Clin Invest. 2019;130(11):4922–4936. Circulation. 2010;122(14):1396–1404. 32. Bankovich AJ, et al. CD69 suppresses sphin- 7. Wang Y, et al. C-X-C motif chemokine receptor 4 20. Barry SP, et al. Enhanced IL-17 signalling follow- gosine 1-phosophate receptor-1 (S1P1) function blockade promotes tissue repair after myocardial ing myocardial ischaemia/reperfusion injury. Int through interaction with membrane helix 4. J Biol infarction by enhancing regulatory t cell mobili- J Cardiol. 2013;163(3):326–334. Chem. 2010;285(29):22328–22337. zation and immune-regulatory function. Circula- 21. Mora-Ruiz MD, et al. Role of interleukin-17 33. Bohl S, et al. Refined approach for quantification tion. 2019;139(15):1798–1812. in acute myocardial infarction. Mol Immunol. of in vivo ischemia-reperfusion injury in the 8. Seropian IM, et al. Galectin-1 controls cardiac 2019;107:71–78. mouse heart. Am J Physiol Heart Circ Physiol. inflammation and ventricular remodeling 22. Yan X, et al. Temporal dynamics of cardiac immune 2009;297(6):H2054–H2058. during acute myocardial infarction. Am J Pathol. cell accumulation following acute myocardial 34. Stieger P, et al. Targeting of extracellular RNA 2013;182(1):29–40. infarction. J Mol Cell Cardiol. 2013;62:24–35. reduces edema formation and infarct size and 9. Weirather J, et al. Foxp3+ CD4+ T cells improve 23. Yan X, et al. Deleterious effect of the IL-23/ improves survival after myocardial infarction in healing after myocardial infarction by modulat- IL-17A axis and γδT cells on left ventricular mice. J Am Heart Assoc. 2017;6(6):e004541. ing monocyte/macrophage differentiation. Circ remodeling after myocardial infarction. J Am 35. Sano S, et al. Tet2-mediated clonal hematopoiesis Res. 2014;115(1):55–67. Heart Assoc. 2012;1(5):e004408. accelerates heart failure through a mechanism 10. Xia N, et al. Activated regulatory T-cells atten- 24. Chen XM, et al. Gene expression pattern of TCR involving the IL-1β/NLRP3 inflammasome. J Am uate myocardial ischaemia/reperfusion injury repertoire and alteration expression of IL-17A Coll Cardiol. 2018;71(8):875–886. through a CD39-dependent mechanism. Clin Sci gene of γδ T cells in patients with acute myocar- 36. Stevenson MD, et al. NADPH oxidase 4 regulates (Lond). 2015;128(10):679–693. dial infarction. J Transl Med. 2018;16(1):189. inflammation in ischemic heart failure: role of 11. George J, et al. Regulatory T cells and IL-10 levels 25. de la Fuente H, et al. The leukocyte activation soluble epoxide hydrolase. Antioxid Redox Signal. are reduced in patients with vulnerable coronary receptor CD69 controls T cell differentiation 2019;31(1):39–58. plaques. Atherosclerosis. 2012;222(2):519–523. through its interaction with galectin-1. Mol Cell 37. Bansal SS, et al. Dysfunctional and proinflamma- 12. Sardella G, et al. Frequency of naturally-occur- Biol. 2014;34(13):2479–2487. tory regulatory T-lymphocytes are essential for ring regulatory T cells is reduced in patients with 26. Lin CR, et al. Glycosylation-dependent inter- adverse cardiac remodeling in ischemic cardio- ST-segment elevation myocardial infarction. action between CD69 and S100A8/S100A9 myopathy. Circulation. 2019;139(2):206–221. Thromb Res. 2007;120(4):631–634. complex is required for regulatory T-cell differ- 38. Gonzalez-Amaro R, et al. Is CD69 an effective 13. Mao X, et al. IL-37 plays a beneficial role in entiation. FASEB J. 2015;29(12):5006–5017. brake to control inflammatory diseases? Trends patients with acute coronary syndrome. Media- 27. Sreejit G, et al. Neutrophil-derived S100A8/A9 Mol Med. 2013;19(10):625–632. tors Inflamm. 2019;2019:9515346. amplify granulopoiesis after myocardial infarc- 39. Papotto PH, et al. IL-17+ γδ T cells as kick-starters of 14. Klingenberg R, et al. Clonal restriction and pre- tion. Circulation. 2020;141(13):1080–1094. inflammation. Nat Immunol. 2017;18(6):604–611. dominance of regulatory T cells in coronary throm- 28. Hosono M, et al. Increased expression of T cell 40. Huber SA. T cells expressing the gamma delta T bi of patients with acute coronary syndromes. activation markers (CD25, CD26, CD40L and cell receptor induce apoptosis in cardiac myo- Eur Heart J. 2015;36(17):1041–1048. CD69) in atherectomy specimens of patients cytes. Cardiovasc Res. 2000;45(3):579–587. 16 J Clin Invest. 2022;132(21):e152418 https://doi.org/10.1172/JCI152418 The Journal of Clinical Investigation   RESEARCH ARTICLE 41. Swirski FK, et al. Identification of splenic reser- CD39-mediated protection from renal injury. Eur 54. Ohta A, Sitkovsky M. Extracellular adenos- voir monocytes and their deployment to inflam- J Immunol. 2012;42(9):2441–2451. ine-mediated modulation of regulatory T cells. matory sites. Science. 2009;325(5940):612–616. 48. Mascanfroni ID, et al. Metabolic control of type Front Immunol. 2014;5:304. 42. Park SG, et al. T regulatory cells maintain intesti- 1 regulatory T cell differentiation by AHR and 55. Cai M, et al. Transgenic over expression of ecto- nal homeostasis by suppressing γδ T cells. Immu- HIF1-α. Nat Med. 2015;21(6):638–646. nucleotide triphosphate diphosphohydrolase-1 nity. 2010;33(5):791–803. 49. Cibrian D, et al. CD69 controls the uptake of protects against murine myocardial ischemic 43. Kunzmann V, et al. Inhibition of phosphoanti- L-tryptophan through LAT1-CD98 and AhR-de- injury. J Mol Cell Cardiol. 2011;51(6):927–935. gen-mediated gammadelta T-cell proliferation pendent secretion of IL-22 in psoriasis. Nat 56. Wheeler DG, et al. Transgenic swine: expression by CD4+ CD25+ FoxP3+ regulatory T cells. Immu- Immunol. 2016;17(8):985–996. of human CD39 protects against myocardial inju- nology. 2009;126(2):256–267. 50. Sharir R, et al. Experimental myocardial infarc- ry. J Mol Cell Cardiol. 2012;52(5):958–961. 44. Borsellino G, et al. Expression of ectonucleoti- tion induces altered regulatory T cell hemostasis, 57. Smith SB, et al. Impact of cardiac-specific expres- dase CD39 by Foxp3+ Tregs: hydrolysis of extra- and adoptive transfer attenuates subsequent sion of CD39 on myocardial infarct size in mice. cellular ATP and immune suppression. Blood. remodeling. PLoS One. 2014;9(12):e113653. Life Sci. 2017;179:54–59. 2007;110(4):1225–1232. 51. Pfeffer MA, et al. Development and prevention 58. Kohler D, et al. CD39/ectonucleoside triphos- 45. Deaglio S, et al. Adenosine generation catalyzed of congestive heart failure following myocar- phate diphosphohydrolase 1 provides myocardial by CD39 and CD73 expressed on regulatory T dial infarction. Circulation. 1993;87(5 sup- protection during cardiac ischemia/reperfusion cells mediates immune suppression. J Exp Med. pl):Iv120–IV125. injury. Circulation. 2007;116(16):1784–1794. 2007;204(6):1257–1265. 52. Josefowicz SZ, et al. Regulatory T cells: mecha- 59. Lang RM, et al. Recommendations for cardiac 46. Fabbiano S, et al. Immunosuppression-indepen- nisms of differentiation and function. Annu Rev chamber quantification by echocardiography in dent role of regulatory T cells against hyperten- Immunol. 2012;30:531–564. adults: an update from the American Society of sion-driven renal dysfunctions. Mol Cell Biol. 53. Shevach EM. Mechanisms of foxp3+ T regu- Echocardiography and the European Association 2015;35(20):3528–3546. latory cell-mediated suppression. Immunity. of Cardiovascular Imaging. Eur Heart J Cardio- 47. Wang YM, et al. Regulatory T cells participate in 2009;30(5):636–645. vasc Imaging. 2015;16(3):233––270. J Clin Invest. 2022;132(21):e152418 https://doi.org/10.1172/JCI152418 17