The animal study was approved by Yangtze University of Medicine Animal Care and Use Committee and was performed in accordance with Yangtze University Guidelines for the Use of Animals. Male C57BL/6J and C57BL/6.SJL mice (8- to 10-week old) were ordered from the Animal Laboratory Center of Wuhan University. Mice were anesthetized with 2% halothane (Sigma-Aldrich) with 40% oxygen and followed by 0.5% halothane with 40% oxygen. Artificial ventilation (0.3 ml tidal volume, 120 breaths/min) was provided by tracheotomy. An 8.0 nylon surgical suture was used for ligation of the left coronary artery 1.0 mm distal from the tip of the left auricle. The left coronary artery was then occluded for 60 min before reperfusion for 3–7 days. The same surgical procedure was done on sham-operated mice, except that left coronary artery was not occluded.

Cells were isolated from infarcted myocardium according to our previous method (14). Briefly, mouse hearts were rapidly excised and placed in ice-cold Krebs–Henseleit (KH) buffer containing (in g/L) 2 glucose, 0.141 MgSO₄, 0.16 NaH₂PO₄, 0.35 KCl, 6.9 NaCl, 2.1 NaHCO₃, 0.373 CaCl₂, 1 NaN₃ at pH 7.4. Subsequently, the aorta was cannulated with a 22-gage tubing adaptor and flushed with ice-cold KH buffer to remove residual cells in the coronary vasculature. The infarcted tissue were harvested and minced with fine scissors and placed into a cocktail of 0.25 mg/ml Liberase Blendzyme 3 (Roche Applied Science), 20 U/ml DNase I (Sigma-Aldrich), 10 mmol/L HEPES (Invitrogen), 0.1% sodium azide in HBSS with Ca²⁺ and Mg²⁺ (Invitrogen) and shaken at 37°C for 40 min. Cells were then passed through a 40-µm nylon mesh (BD Falcon), centrifuged for 10 min at 500 g at 4°C, and resuspended in 0.1% sodium azide solution in HBSS without Ca²⁺ and Mg²⁺. Isolated cells were kept on ice before further processing. In some experiments, infarcted tissues from 5 to 6 mice were pooled for cell isolation.

Anti-mouse antibodies were used: APC anti-CD3 (17A2), APC anti-B220 (RA3-6B2), APC anti-Gr1 (RB6-8C5), APC anti-CD90 (OX-7), FITC anti-CD11b (M1/70), PE anti-F4/80 (T45-2342), and PE-Cy7 anti-Ly-6C (AL-21) were purchased from BD Biosciences. PE-Cy5 anti-CD80 (16-10A1), PE-Cy7 anti-CD86 (GL1), and APC-eFluor^® 780 anti-CD45.1 (A20) were purchased from eBioscience. Biotin anti-CD163 (6E10.1G6) was purchased from Novus Biologicals. For cell surface staining, cells were incubated with corresponding antibodies in PBS for 15 min on ice before analysis on a BD LSRII flow cytometer. Dead cells were excluded with propidium iodide (2 µg/ml) staining. For phospho-flow assay, PE anti-phospho-4EBP1 (Thr36/45, clone # V3NTY24, eBioscience) was used to stain cells following the manufacturer’s instructions. All flow cytometry data were analyzed with Flowjo 7.6.1 software. Cell sorting was performed on a BD FACSAria cell sorter based on cell surface marker staining.

Monocyte-derived macrophages were generated according to pervious reports (15–17). Briefly, normal mouse blood was taken into Falcon tubes with 2 mM EDTA. Spleens were collected from normal mice and pushed through 40-µm nylon meshes to prepare splenocyte suspension. The splenocyte suspension and blood were mixed before density gradient isolation with Ficoll-Paque™ PLUS (GE Healthcare), and mononuclear cells were collected and incubated in RPMI medium with 10% fetal calf serum, 4 mM l-glutamine, and penicillin/streptomycin. The cell density was 1 × 10⁶ cells/ml. 2 × 10⁶ cells were then seeded in each well of 6-well culture plates and were incubated overnight in a humid incubator with 5% CO₂. Floating cells were then aspirated. Adherent cells were treated for 6 days either with 50 ng/ml GM-CSF and 10 ng/ml IFN-γ (for M1 polarization), or with 50 ng/ml M-CSF and 10 ng/ml IL-4 and 10 ng/ml IL-10 (for M2 polarization). All cytokines were purchased from R&D Systems.

For lipopolysaccharides (LPS) stimulation, polarized macrophages were gently washed with PBS once before digestion with 0.25% trypsin-EDTA for 5 min at 37^oC. Cells were then plated at 1 × 10⁵ per well in 6-well plates in the presence or absence of 1 µg/ml LPS for 3 h. In some experiments, cells were pretreated with 2 µM MHY1485 (mTOR activator, Sigma-Aldrich) for 1 h before LPS treatment.

Lentiviral transduction of monocyte-derived macrophages was conducted according to previous successful reports (18–20). Briefly, polarized macrophages were gently washed with PBS once before digestion with 0.25% trypsin-EDTA for 5 min at 37^oC. Cells were resuspended in growth media at the density of 5 × 10⁶ cells/ml. Three milliliters of cell suspension were added to each well of 6-well plates. Lentiviral particles (SESN2 lentiviral activation particles or control lentiviral particles, Santa Cruz Biotechnology) were diluted in growth medium containing 8 µg/ml polybrene (Sigma-Aldrich) and were added into each well at the MOI of 8. Cells were then incubated overnight with the lentiviral particles before wash once with PBS. Cells were then cultured in fresh growth medium in the presence of above cytokines for additional 5 days before further experiments. To monitor the lentiviral transduction efficiency, copGFP lentiviral particles containing GFP sequence (Santa Cruz Biotechnology) were transduced under the same condition as for SESN2 lentiviral activation particles. copGFP lentiviral particles has the same envelope proteins as SESN2 lentiviral activation particles.

For selective depletion of monocytes/macrophages, each C57BL/6J mouse were injected i.v. with 0.1 ml of Clophosome^® (FormuMax Scientific) 24 h before MI. Exogenous macrophages were derived from C57BL/6.SJL mice and were labeled with carboxyfluorescein succinimidyl ester (CFSE, Thermofisher) or CellTrace Violet (Thermofisher) following the manufacturer’s instructions. Immediately before MI, 1 × 10⁶ or 4 × 10⁶ labeled macrophages in 200 µl of PBS were transferred into each C57BL/6J recipient mouse through retro-orbital injection.

For assessment of proteolytic activity, 5 nmol of MMPsense-680 or Prosense-680 (both from Perkin Elmer) in 150 µl of PBS were retro-orbitally injected into each mouse 24 h before sacrifice. Fluorescence of each probe was evaluated on a BD LSRII flow cytometer.

For quantification of infarct size, mice were anesthetized again. The heart was exposed and left coronary artery was ligated in the same spot. 0.3 ml of 1% Evans blue was perfused into the aorta to stain the heart for discrimination of the ischemic area from the non-ischemic area. The left ventricle of each heart was excised and was frozen and cut transversely into five slices of equal thickness followed by incubation in 1% triphenyltetrazolium chloride (TTC, sigma)/PBS at 37°C for 8 min and then were fixed in 10% formalin-PBS for 24 h. Fixed slices were then scanned, and ImageJ software (NIH) was used to measure the infarct area.

Collage deposition was analyzed by Masson staining. Briefly, hearts were excised and fixed with 10% formalin followed by dehydration and paraffin embedding. Six-micron sections were prepared and stained using Trichrome Stain (Masson) Kit (Sigma-Aldrich) following the vendor’s manual. The collagen fibers were quantified using collagen volume fraction (CVF), which is calculated as the ratio between the area of collagen staining and the area of the whole field. Five fields were randomly chosen and Image-pro plus 6.0 was used for the quantification.

Cellular or tissue RNAs were extracted using Trizol reagent (Invitrogen). cDNA synthesis was conducted with extracted RNAs and SuperScript^® III First-Strand Synthesis System (Invitrogen) according to the manufacturers’ instruction. q-PCR was conducted using Fast SYBR^® Green Master Mix (Invitrogen) on a 7300 q-PCR System (Invitrogen). Data were evaluated with 7300 system software. Primer sequences for each gene are listed in Table S1 in Supplementary Material. Relative abundance of gene expression was assessed using 2^−ΔΔCt method.

The following primary antibodies were used: rabbit monoclonal anti-mTOR and anti-phospho-mTOR (Ser2448) were purchased from Cell Signaling Technology. Mouse monoclonal anti-β-actin was purchased from Santa Cruz Biotechnology. Rabbit monoclonal anti-Sestrin2 was purchased from Abcam. Membranes were developed with SuperSignal West Pico Chemiluminescent Substrate (Thermo Scientific) and were visualized and the optical density of the identified protein bands on membranes was analyzed using a Biospetrum 500 imaging system (UVP, LLC).

At day 14 after MI, mice were anesthetized with isoflurane inhalation. Left ventricle function was assessed by echocardiography with tracheal intubation. Echocardiographic data were collected by a HP/Philips Sonos 5500 ultrasound system with 15-MHz linear array transducer. Five consecutive cardiac cycles were used for each measurement. Left ventricle end-systolic diameter (LVESD) and end-diastolic diameter (LVEDD) were measured at end-diastole and end-systole, respectively. Percentage fractional shortening (%FS) and ejection fraction (EF) were calculated as: %FS = [(LVEDD − LVESD)/LVEDD] × 100; EF = 100 × [(LVEDD³ − LVESD³)/LVEDD³].

Data were analyzed and results were presented as mean ± SD. Student’s t-test or one-way ANOVA was used for comparison of mean between the groups, and p < 0.05 was considered significant.

To characterize cardiac macrophages in normal and ischemic myocardium, cells were isolated from infarcted area after ligature of the left anterior descending artery, as well as from the corresponding area in sham hearts. As shown in Figure S1 in Supplementary Material, a significant increase of CD3⁻B220⁻Gr-1⁻CD90⁻CD11b⁺ population, which represents macrophages, was observed in infarcted myocardium in comparison with sham control. Further analysis of expression of Ly-6C and F4/80 indicated that more than 50% of cardiac macrophages were pro-inflammatory F4/80^loLy-6C⁺ cells in the acute phase of MI (post-MI day 1 to day 3). However, the phenotype of cardiac macrophages shifted toward F4/80^hiLy-6C⁻ cells from post-MI day 5 to day 7, suggesting the decrease of pro-inflammatory macrophages at the chronic stage of MI (Figure 1A). Our data were consistent with the previous phenotypic characterization of cardiac macrophages after MI (4). Interestingly, Sestrin2 expression was significantly enhanced in cardiac macrophages after MI and peaked at post-MI day 3 (Figures 1B,C). However, only a non-significant increase of Sestrin2 protein was seen in circulating monocytes after MI, suggesting that local factors rather than systemic factors were responsible for the increase of Sestrin2 in macrophages (Figures 1B,C). Moreover, careful analysis of Sestrin2 expression in F4/80^hi Ly-6C⁻, F4/80^loLy-6C⁻, F4/80^loLy-6C⁺, and F4/80^hiLy-6C⁺ subpopulations (subpopulation I–IV, respectively) revealed that Sestrin2 was predominantly expressed in Ly-6C⁺ macrophages at day 3 after MI (Figure 1D), suggesting that Sestrin2 was closely associated with the pro-inflammatory phenotype of cardiac macrophages. We also stained CD163 and CD80 to distinguish M1 and M2 macrophage subtypes. We found that about 50% of macrophages were CD163^-CD80⁺ M1 macrophages, 30% were CD163⁺CD80^- M2 macrophages, 15% were CD163⁻CD80⁻ macrophages, and only 7% were CD163⁺CD80⁺ macrophages (Figure S2 in Supplementary Material). Sestrin 2 expression was upregulated in CD163⁻CD80⁺ and CD163⁻CD80⁻ macrophages, as compared with CD163⁺CD80⁻ macrophages (Figure S2 in Supplementary Material), suggesting that Sestrin 2 was mainly expressed in M1 macrophages.

To explore the effects of Sestrin2 on macrophages, we cultured monocyte-derived macrophages under M1 (with GM-CSF and IFN-γ) or M2 (with M-CSF and IL-4) polarization condition. Flow cytometry analysis indicated that nearly 70% of cultured cells were CD80⁺CD163^lo/− in the presence of GM-CSF and IFN-γ, suggesting the majority was M1 macrophages (Figure 2A). However, in the presence of M-CSF and IL-4, 70% of macrophages were CD80⁻CD163⁺, suggesting the successful M2 polarization (Figure 2A). These cells were then transduced with SESN2 lentiviral activation particles to induce ectopic expression of Sestrin2. Flow cytometry analysis demonstrated that more than 65% of either M1 or M2 macrophages were successfully transduced with lentivirus (Figure 2B). And the ectopic Sestrin2 expression in either M1 or M2 macrophages was verified by Western blot (Figure 2C). We then determined the cytokine production in macrophages with or without LPS stimulation. As shown in Figures 2D,E, ectopic Sestrin2 expression alleviated TNF-α and IL-1β mRNA levels in M1 macrophages in the presence or absence of LPS. However, in M2 macrophages, ectopic Sestrin2 expression only moderately reduced IL-1β level, while TNF-α production was not significantly altered. In addition, Sestrin2 overexpression promoted IL-10 mRNA abundance only in M1 macrophages (Figure 2F).

To test the effect of Sestrin2 on M1 macrophages in vivo, monocytes were collected from C57BL/6.SJL mouse blood and spleens, induced to be M1 macrophages, and transduced with lentiviruses. Macrophages transduced with control lentivirus (L-C) and SESN2 lentiviral activation particles (L-S) were labeled with CellTrace Violet and CFSE, respectively. These two macrophage populations were then mixed at the ratio of 1:1 and adoptively injected into recipient C57BL/6 mice immediately before MI was conducted. Three days after MI, cardiac macrophages were analyzed. As shown in Figure 3A, exogenous CD45.1⁺ macrophages were present in infarcted myocardium, and the ratio of Violet-positive cells vs CFSE-positive cells was close to 1:1, suggesting that Sestrin2 overexpression did not influence recruitment of M1 macrophages into infarcted myocardium. Analysis of macrophage surface markers indicated that Sestrin2 overexpression down-regulated CD80 and CD86 expression while upregulating CD163 expression on these macrophages (Figures 3B,C), suggesting that Sestrin2 induced a phenotypic shift from M1 type toward M2 type. Additionally, Sestrin2 overexpression impaired the MMP activity of macrophages, whereas the cathepsin activity was not changed (Figures 3D,E). Moreover, Sestrin2 overexpression decreased production of pro-inflammatory cytokines such as TNF-α and IL-1β, and increased repair-promoting cytokines like IL-10 and VEGF (Figure 3F).

To test the effect of Sestrin2 on M2 macrophages in vivo, monocytes were collected from C57BL/6.SJL mouse blood and spleens, induced to be M2 macrophages, and transduced with lentiviruses as done for M1 macrophages. M2 macrophages were then labeled and adoptively transferred the same way as for M1 macrophages. Similarly, Sestrin2 overexpression did not profoundly impact the recruitment of M2 macrophages into infarcted myocardium (Figure 4A). In addition, M2 macrophages transduced with L-S had higher CD163 expression, suggesting that Sestrin2 maintained or enhanced M2 polarization (Figures 4B,C). However, Sestrin2 overexpression did not considerably alter the mRNA levels of TNF-α, IL-1β, and IL-10, whereas VEGF expression was moderately increased (Figure 4D). Taken together, our data imply that the effects of Sestrin2 on M2 macrophage phenotype and function are limited.

Because above data suggest the significant role of Sestrin2 in modulating M1 macrophage response, we then determined the effect of Sestrin2-overexpressing M1 macrophages on the postischemic myocardial inflammation. Endogenous macrophages were temporarily depleted using Clophosome^®, and exogenous Sestrin2-overexpressing M1 macrophages were adoptively transferred into mice immediately before MI. As shown in Figure S3 in Supplementary Material, over 90% of endogenous blood monocytes were depleted 24 h after administration of Clophosome^®. Similar depletion of endogenous cardiac macrophages was also observed in Clophosome^®-injected mice (Figure 5A). Adoptive transfer of exogenous macrophages partially restored cardiac macrophage abundance in Clophosome^®-injected mice (Figure 5A). Analysis of cytokine expression in infarcted myocardium revealed that MI itself profoundly induced expression of pro-inflammatory cytokines like TNF-α, IL-1β, and IL-6 (Figure 5B). Noteworthy, compared with mice receiving L-C-transduced macrophages, mice receiving L-S-transduced macrophages had lower expression of these cytokines in infarcted myocardium (Figure 5B). In addition, expression of anti-inflammatory IL-10 was higher in the myocardium of mice receiving L-S-transduced macrophages, in comparison with mice receiving L-C-transduced macrophages (Figure 5B). However, the expression of chemokines MCP-1 and IP-10, which have been shown to mediate leukocyte recruitment, was not significantly changed between mice receiving L-C-transduced macrophages and mice receiving L-S-transduced macrophages (Figure 5C). To further characterize the post-MI myocardial inflammation, the infiltrating neutrophils were quantified in the hearts. As shown in Figures 5D,E, very little neutrophils were recovered from sham-operated hearts, while substantial amount of neutrophils was found in infarcted hearts. Depletion of macrophages decreased infiltrating neutrophils. Furthermore, in comparison with mice receiving L-C-transduced macrophages, the mice receiving L-S-transduced macrophages had less infiltrating neutrophils in infarcted hearts, suggesting that Sestrin2-overexpressing macrophages induced myocardial inflammation to a less extent. In addition, Masson staining showed that L-S-transduced macrophages caused less collagen staining in infarcted myocardium, as compared with L-C-transduced macrophages (Figure 5F), suggesting that Sestrin2-expressing macrophages might result in a scar with low collagen content.

It has been previously shown that Sestrin2 inhibits mTORC1 signal pathway (21, 22), and mTORC1 is crucial for M1 macrophage polarization and function (23, 24). Therefore, we checked activation status of mTORC1 signal pathway in adoptively transferred L-S-transduced macrophages and L-C-transduced macrophages. Interestingly, activating phosphorylation of mTOR at serine 2448 was decreased in L-S-transduced macrophages (Figure 6A). This was also seen in in vitro cultured L-S-transduced macrophages (Figure 6B). Consistently, phosphorylation of 4EBP1 in L-S-transduced macrophages was lower than that in L-C-transduced macrophages (Figure 6C). To further confirm the role of mTORC1 signaling in Sestrin2-induced changes, we pretreated L-C or L-S-transduced macrophages with an mTOR activator, MHY1485, and then stimulated macrophages with LPS. As shown in Figure 6D, MHY1485 increased mTOR phosphorylation in both L-C-transduced and L-S-transduced macrophages, in comparison with non-treated L-C-transduced macrophages. In addition, MHY1485 pretreatment promoted TNF-α and IL-1β expression in in both L-C-transduced and L-S-transduced macrophages, suggesting that forced mTORC1 activation counteracted Sestrin2’s effect and enhanced inflammatory response of macrophages (Figures 6E,F). However, IL-10 expression was not dramatically impacted by MHY1485 pretreatment, suggesting that mTORC1 signaling might be irrelevant to IL-10 production (Figure 6G). Thus, Sestrin2-induced upregulation of IL-10 might not be related to mTORC1 signaling.

Although post-MI myocardial inflammation was moderately mitigated by L-S-transduced macrophages, TTC staining demonstrated comparable infarct size in mice receiving L-C-transduced macrophages and mice receiving L-S-transduced macrophages (Figures 7A–C). Macrophage depletion caused a trend of decrease in infarction size (Figure 7C). Meanwhile, the survival rate was also similar between mice receiving L-S-transduced macrophages and those receiving L-C-transduced macrophages (Figure 7D). Evaluation of left ventricle function revealed that depletion of macrophages increased LVESD and LVEDD after MI in comparison with the control group. Correspondingly, FS and EF were decreased when macrophages were depleted (Figures 7E–I). Transfer of exogenous macrophages caused partial but insignificant decrease in LVESD and LVEDD, as well as increase in FS. However, LVESD, LVEDD, FS, EF, and heart rates were comparable between mice receiving L-S-transduced macrophages and mice receiving L-C-transduced macrophages (Figures 7E–I).