Animals. Experimental animals were used at 8–12 weeks of age and had a body weight of 20–25 g. Male C57BL/6J mice (Janvier), T cell–specific CD73-KO mice (CD4-Cre+/– CD73fl/fl), and control mice (CD4-Cre–/– CD73fl/fl), also on a C57BL/6J background, as well asA2bR–/– mice and their littermate controls on a C57BL/6N background (generated by Deltagene, RRID: MGI:3812454 and provided by Alexander Pfeifer, University of Bonn, Bonn, Germany), were housed and/or bred at the “Zentrale Einrichtung für Tierforschung und Tierschutzaufgaben” (ZETT) at Heinrich-Heine-University (Düsseldorf, Germany).

Infarct model. Reperfused MI was induced as previously described (61). In brief, 8- to 12-week-old male mice were anesthetized (isoflurane 1.5% via respiration), and the left anterior descending coronary artery was ligated for 50 minutes followed by reperfusion. Left anterior descending coronary artery occlusion was ensured by ST-segment elevation in ECG recordings. Mice were kept under analgesic treatment for 3 days (buprenorphin 0.1 mg/kg, 3 times/day). All animals used in this study survived the ischemia and reperfusion period.

Quantification of gene expression by qPCR. mRNA from cardiac cells was prepared using the RNeasy Micro Plus kit (QIAGEN) according to the manufacturer’s protocol. The quality and quantity of extracted mRNA were analyzed with the Agilent Bioanalyzer 2100 (Agilent Technologies). cDNA was generated using the High Capacity cDNA kit (Applied Biosystems). Because of the low amounts of cDNA, a preamplification reaction was performed using Pre-Amp Master Mix (Applied Biosystems), together with the same TaqMan assays (QIAGEN) later used for gene expression analysis in a reaction with 14 cycles. Gene expression was analyzed using Gene Expression Master Mix (Applied Biosystems) on a StepOnePlus Real-Time PCR system (Applied Biosystems). The following TaqMan primers were used: mouse Rplp0 (housekeeping gene) (Mm00725448_s1), Il6 (Mm00446190_m1), Il11 (Mm00434162_m1), Adora2b (Mm00839292_m1), Entpd1 (CD39) (Mm00515447_m1), ENPP1 (Mm01193761_m1), ENPP3 (Mm01193723_m1), CD38 (Mm01220906_m1), and Nt5e (CD73) (Mm00501910_m1) (all from Applied Biosystems).

In situ hybridization. Formalin-fixed, paraffin-embedded hearts collected from mice 3 days after MI were stained with Alexa Fluor 488–conjugated WGA (Thermo Fisher Scientific) to visualize the infarct area. Additionally, sections from the same section plane were processed using the RNAscope 2.5 HD Duplex kit (Advanced Cell Diagnostics ADC) to visualize Il6 (probe Mm-IL6 no. 315891 or probe Mm-IL6-C2 no. 315891-C2), CD45 (Ptprc) (probe Mm-Ptprc-C2 no. 318651), and Postn (probe Mm-Postn no. 418581) mRNA according to the manufacturer’s instructions. Sections were counterstained with Gill’s Hematoxylin I (Santa Cruz Biotechnology). Fluorescent images were taken from fixed (4% PFA in PBS) frozen section of hearts 3 days after MI and processed with the RNAscope Multiplex Fluorescent Reagent Kit (version 2) according to the manufacturer’s instructions. Images were recorded on an Olympus Bx 61 Microscope.

Cardiac cell isolation and flow cytometry. Cardiac cells were isolated by a previously described modified protocol (21). In brief, hearts were digested by intracoronary delivery of a collagenase solution (12,00 U/mL; collagenase type II from Clostridium histolyticum, Worthington) until the perfusion pressure was below 10 mmHg. To simultaneously isolate EpiSCs from the surface of infarcted hearts, hearts were additionally bathed in collagenase solution as described previously (21). Isolated single-cell suspensions were used for cardiac cell isolation by flow cytometric cell sorting or magnetic bead–mediated enrichment. For flow cytometric cell sorting, FcR binding sites were blocked (Miltenyi Biotec), and cells were stained for flow cytometric sorting on a MoFlo XDP flow cytometer (Beckman Coulter) with the following fluorochrome conjugated antibodies: CD45 Alexa Fluor 647 (30-F11, BD Biosciences); CD3 PE-CF594 (145-2C11, BD Biosciences); CD11b APC-Cy7 (M1/70, Thermo Fisher Scientific); Ly6G FITC (1A8, BD Biosciences); and CD19 PE (1D3, BD Biosciences). The cell populations analyzed were defined as follows: T cells, CD45+CD3+CD19–CD11b–; B cells, CD45+CD3–CD19+CD11b–; granulocytes, CD45+CD3–CD19–CD11b+Ly6G+; and monocytes, macrophages, and DCs, CD45+CD3–CD19–CD11b+Ly6g–. ECs, CFs, and EpiSCs were isolated using the following antibodies: CD31 APC (Mec 13.3, BD Biosciences) and CD45 PE-Cy7 (30-F11, BD Biosciences), whereas CFs and EpiSCs were defined as being CD45–CD31– and ECs as CD45–CD31+. For the isolation of CFs by magnetic bead–mediated enrichment, a cardiac single-cell suspension was depleted of CD45+ and CD31+ cells by magnetic bead negative selection (Mojosort Nanobeads, BioLegend) according to the manufacturer’s instructions, with the addition of biotin-labeled anti-CD31 (clone MEC13.3, BioLegend) antibody.

IL-6 ELISPOT assay. ECs, granulocytes, macrophages, or CFs were isolated from the hearts of C57BL/6J mice 3 days after MI (see above) and were seeded (10,000 cells per well each) onto an ELISPOT membrane provided in a commercial ELISPOT Mouse/Rat IL-6 Kit (EL406, R&D Systems) according to the manufacturer’s instruction. Cells were cultivated overnight in IMDM containing 10% FCS, 1% penicillin and streptomycin, 1% GlutaMAX, and 25 μM β-mercaptoethanol. IL-6–positive spots were quantified using ImageJ software (NIH) (62). First, RGB images were split into hue, saturation, and brightness stacks; the image representing the brightness was used to set a threshold (Yen), and spots were counted using the “analyze particle” tool.

IL-6 expression and secretion assay. Murine CFs were isolated from healthy A2bR–/– mice or their littermate controls by collagenase digestion using the Langendorff methods and magnetic bead–mediated enrichment as described above. Isolated CFs were cultivated in DMEM complete (4,500 mg/L glucose, 20% FCS, 1% penicillin and streptomycin, 1% GlutaMAX). All experiments were conducted with cells in passage 2: 60,000 CFs were seeded per well of a 12-well plate and permitted to settle overnight (passage 2). Cells were washed with DMEM complete and incubated with 400 μL DMEM complete containing 33 μM erythro-9-(2-hydroxy-3-nonyl) adenine (EHNA) (an adenosine deaminase [ADA] inhibitor, Tocris) and 33 μM nitrobenzylthioinosine (NBMPR) (an inhibitor of equilibrative nucleoside transporter 1 ENT1, Tocris), with or without 1 μM FR900359 (a Gq inhibitor, provided by Evi Kostenic, University of Bonn, Bonn, Germany) for 10 minutes at 37°C. Afterwards, 100 μL DMEM complete was added and contained (besides 33 μM EHNA and 33 μM NBMPR with or without 1 μM FR900359) 100 μM adenosine (Tocris) or only the corresponding amount of DMSO (vehicle control), giving a final concentration of 20 μM adenosine. After incubation for 24 hours, supernatant was collected and used for cytokine analysis with Bioplex technology (Bio-Rad) or IL-11 ELISA (Thermo Fisher Scientific). Cells were washed with PBS and used for mRNA expression analysis.

ATP degradation assay. CFs isolated as described above (magnetic bead isolation) from hearts of healthy C57BL/6J mice were seeded at a density of 10,000 cells per well of a 96-well, flat-bottomed plate (1 well per time point) and allowed to settle overnight. Cells were incubated with 20 μM ATP (MilliporeSigma) dissolved in HBSS buffer (Life Technologies, Thermo Fisher Scientific). Supernatants were collected at the indicated time points, and ATP metabolite concentrations were determined by HPLC (Waters) at 254 nm.

Flow cytometric determination of ATP- und ADP-degrading enzymes. For FACS analysis of ATP- and AMP-degrading enzymes on CFs (healthy mice), aCFs, and EpiSCs (day 5 after MI), hearts were digested as described above and stained with CD45 APC (30-F11, BD Biosciences), CD45 PE-Cy7 (30-F11, BD Biosciences), CD31 BV510 (MEC13.3, BD Biosciences), CD39 PeCy7 (24DM51, BD Biosciences), CD38 APC (90, BioLegend), CD73 FITC (496406, R&D Systems), ENPP1 APC (YE1/19.1, BioLegend), MEFSK-4 PE (mEF-SK4, Miltenyi Biotec), CD90.2 PE (30-H12, Thermo Fisher Scientific), or PDGFRA PE (ADA5, eBioscience). Dead cells were excluded by 7AAD staining. Cardiac stromal cells were identified by gating on 7AAD–CD45–CD31– cells that were positive for the PE stain, which was a mixture of MEFSK-4, CD90.2, and PDGFRA.

Single-cell transcriptomics. For single-cell transcriptomics, hearts were digested as described above, followed by FACS for the following cell populations: CD31–CD45+ (ICs) and CD31–CD45– (stromal cells), whose identity (CFs or EpiSCs) was defined by cell localization (see above) (21). The sorted single-cell suspensions were directly used for the scRNA-Seq experiments (as was already reported for aCF and EpiSC populations; ref. 22). Here, the analysis reported in Hesse et al. (22) was combined with our IC analysis of the same hearts studied by Hesse et al., revealing a new clustering for these analyses. scRNA-Seq analysis was performed by 10x Genomics Chromium System (10x Genomics). Cell viability and cell number analyses were performed via trypan blue staining in a Neubauer counting chamber. A total of 2,000–20,000 cells, depending on cell availability, were used as input for the single-cell droplet library generation on the 10x Chromium Controller system utilizing the Chromium Single Cell 3′ Reagent Kit, version 2, according to the manufacturer’s instructions. Sequencing was carried out on a HiSeq 3000 system (Illumina) according to manufacturer’s instructions, with a mean sequencing depth of approximately 90,000 reads/cell for EpiSCs and approximately 70,000 reads/cell for aCFs as well as approximately 55,000 reads/cell for ICs. Differences in sequencing depth were necessary in order to achieve a similar sequencing saturation of approximately 70% between all samples.

Processing of scRNA-Seq data. Raw sequencing data were processed using 10x Genomics CellRanger software (version 3.0.2). Raw BCL files were demultiplexed and processed to Fastq-files using the Cell-Ranger mkfastq pipeline. Alignment of reads to the mm10 genome and unique molecular identifier (UMI) counting were performed via the CellRanger count pipeline to generate a gene-barcode matrix. The median number of detected genes per cell was 3,155 for EpiSCs, 3,265 for aCFs, and 2,028 for ICs. The median of UMI counts per cell was 10,689 for EpiSCs, 11,110 for aCFs, and 6,096 for ICs. Mapping rates (reads mapped to the genome) were approximately 89% for EpiSCs, 90.9% for aCFs, and 88% for ICs.

Filtering and clustering of scRNA-Seq data. Further analyses were carried out with the Seurat version 3.0 R package (63). Initial quality control consisted of the removal of cells with fewer than 200 detected genes as well as the removal of genes expressed in fewer than 3 cells. Furthermore, cells with a disproportionately high mapping rate to the mitochondrial genome (mitochondrial read percentages >5.0 for EpiSCs and aCFs and >7.5 for ICs) were removed, as they represented dead or damaged cells. Normalization we performed utilizing SCTransform. Biological replicates (n = 3) were integrated into 1 data set by identifying pairwise anchors between data sets and using the anchors to harmonize the data sets. Dimensional reduction of the data set was achieved by principal component analysis (PCA) based on identified variable genes and subsequent uniform manifold approximation and projection (UMAP) embedding. The number of meaningful principal components (PCs) was selected by ranking them according to the percentage of variance explained by each PC, plotting them in an “Elbow Plot,” and manually determining the number of PCs that represented the majority of variance in the data set. Cells were clustered using the graph-based clustering approach implemented in Seurat, version 3.0. Doublet identification was achieved with the DoubletFinder (version 2.0.2) (64) tool to generate artificial doublets using the PC distance to find each cell’s proportion of artificial k nearest neighbors (pANNs) and ranking them according to the expected number of doublets. Heatmaps were generated using Morpheus (https://software.broad-institute.org/morpheus).

A list of the marker genes for all clusters is available in Supplemental Data File 1. The average expression values of the combined clusters are provided in Supplemental Data File 2, and the average expression values of each underlaying analysis are available in Supplemental Data File 3 (aCFs), Supplemental Data File 4 (EpiSCs), and Supplemental Data File 5 (ICs). An interactive Shiny-based visualization of the single-cell data set generated via ShinyCell (26) is available at: https://visualisierung.gtl.hhu.de/data/52Publication/

Isolated perfused Langendorff heart. Hearts were excised from heparinized T cell–specific CD73-KO (CD4-Cre+/– CD73fl/fl) or control (CD4-Cre–/– CD73fl/fl) mice immediately after cervical dislocation, prepared in ice-cold standard Krebs-Henseleit-Buffer, mounted to the Langendorff perfusion set-up, and perfused with oxygenated Krebs-Henseleit buffer (37°C) at a constant coronary perfusion pressure of 100 cm H2O. After an equilibration duration of 15 minutes, cardiac effluent was collected for 15 minutes on ice in a beaker containing a protease inhibitor cocktail (cOmplete Tablets, Mini, EDTA-free, EASYpack, Roche). Cardiac effluent perfusate was concentrated using an Amicon Ultra 3K Centrifugal Filter Device to a final volume of approximately 200 μL. Cytokine concentrations were determined using a Bio-Plex Pro Mouse Cytokine 23-plex Assay (Bio-Rad) and normalized to the heart weight.

IL-6 and various purines released into the coronary effluent perfusate of isolated hearts. IL-6, ATP, ADP, AMP, adenosine, and inosine were measured in the coronary effluent perfusate from isolated saline-perfused hearts (Langendorff), which were subjected to 30 minutes of ischemia at 37oC. For IL-6 measurement, perfusates were collected prior to a 20-minute ischemia period (normoxic controls) and 1–20 minutes and 21–40 minutes after release of ischemia in WT and A2bR–/– mice. Perfusates were treated with proteinase inhibitor (Complete Ultra Tablets EDTA-free, Roche) and concentrated using 10 kDa centrifugation filters (Amicon). IL-6 was measured by BioPlex (Bio-Rad). For the measurement of purinergic metabolites, aliquots from coronary perfusate (100 μL) were taken (without proteinase inhibitor treatment) at baseline and 1, 10, and 30 minutes after ischemia. Purinergic metabolites were measured by HPLC directly from unconcentrated aliquots using an ACQUITY UPLC Bio H-Class System equipped with a Cortecs C18+ UPLC column (3.0 × 150 mm, particle size 1.6 μm) (Waters). Separation of purine compounds was performed using a liner gradient of buffer A (150 mM KH2PO4/150 mM KCl, pH 6) and buffer B (150 mM KH2PO4 150 mM KCl/7.5% acetonitrile, pH 6). Purine absorbance was detected at 254 nm with baseline correction (Absorbance-MBF) (WT, n = 5; A2bR–/–, n = 4).

Statistics. Data are presented as the mean ± SD with the exception of Figure 1, in which the values are shown as the median with the IQR. The number of biological replicates is indicated in the figure legends. In cases where technical replicate experiments were performed, the mean value were are indicated as 1 “n.” Data were analyzed using GraphPad Prism (GraphPad Software) with either a 2-tailed Student’s t test or 1- or 2-way ANOVA with the corresponding post test. A P value of less than 0.05 was considered statistically significant.

Study approval. Animal experiments were performed in accordance with the national guidelines for animal care and were approved by the Landesamt für Natur-, Umwelt- und Verbraucherschutz of the state of North Rhine Westphalia (LANUV NRW, Recklinghausen, Germany).