We used RAW264.7 murine monocyte/Mφ cell line as a model for evaluating the polarization of Mφ phenotypes. To induce M1 polarization, RAW264.7 cells were treated with 20 ng/mL recombinant murine interferon-gamma (mIFNγ; 315-05, PeproTech, Cranbury, NJ, United States) for 24 h. To induce M2 polarization, RAW264.7 cells were treated with 20 ng/mL recombinant murine interleukin-4 (mIL-4; 214-14, PeproTech) for 48 h. Cellular morphology was observed and imaged daily post-induction with the EVOS M7000 microscope (ThermoFisher Scientific, Waltham, MA, United States). Flow cytometry was used to confirm the outcome of the induction. For M1 polarization, we labeled cells with a primary rat anti-mouse CD86 antibody conjugated with Alexa Fluor 488 (105,017, BioLegend, San Diego, CA, United States). For M2 polarization, we labeled cells with a primary rat anti-mouse CD206 (MMR) antibody conjugated with Alexa Fluor 488 (141,709, BioLegend, San Diego, CA, United States).

To examine the phenotypic effects of differential dosage of IL-10 on the classic M1/M2 polarization of Mφ in vitro, we used rhIL-10 (200-10, PeproTech). We chose rhIL-10 because it is functionally active on both murine and human cells, according to the manufacturer’s instructions and previous literature (Nishio et al., 1999; Stumpf et al., 2008). We implemented two distinct experimental protocols to investigate the impact of IL-10 on 1) phenotypes of established Mφ polarization or 2) prospective programming of Mφ polarization. In protocol #1, naïve RAW 264.7 cells were first induced with mIFNγ and mIL-4 for 24 and 48 h to establish the classic M1 and M2 phenotypes, respectively, followed by individual treatment with different doses of IL-10 (0, 25, 100, 250, 500, and 1,000 ng/mL) for 24 h. In protocol #2, naïve RAW 264.7 cells were co-stimulated with a phenotype-polarizing agent (mIFNγ for M1 or mIL-4 for M2) in conjunction with one of the six IL-10 doses (0, 25, 100, 250, 500, or 1,000 ng/mL) for 36 h. Cellular morphology was observed and imaged daily with the EVOS M7000 microscope. Flow cytometry and cytokine secretion were used to quantify the phenotypic and functional outcomes of these experiments, respectively.

Cells labeled with fluorescence-conjugated primary antibodies were analyzed with an Accuri C6 flow cytometer, equipped with 405, 488, and 561 nm lasers (BD Biosciences, San Jose, CA, United States). Ten thousand events per sample were collected for analysis. Sphero rainbow calibration particles with eight peaks (RCP-30-5A, Spherotech, Lake Forest, IL, United States) were used for instrument normalization and MEFL calculation. Briefly, RAW264.7 cells were detached by scraping, washed and resuspended in phosphate-buffered saline (PBS), and quantified using a Countess 3 FL automatic hemocytometer (AMQAF 2000, Thermo Fisher Scientific Cellular Analysis, Eugene, OR, United States). The cell concentration was then adjusted to 1 × 10⁶ cells/mL before incubating them for 20 min on ice with fluorescence-conjugated primary antibodies against specific cell lineage markers that were diluted in FACS buffer (PBS with 1% bovine serum albumin) and protected from light. Cells were then washed twice with PBS to remove excessive antibodies and resuspended in 500 μL of FACS buffer for analysis. For each sample, 10,000 events were collected to ensure statistical significance of the data. Flow cytometry data were analyzed and quantified with FlowJo (FlowJo LLC, Ashland, OR, United States) and represented in histograms (single parameter).

Enzyme-Linked Immunosorbent Assay (ELISA) was performed to quantify the amounts of TNF-α and murine IL-10 (mIL-10) in the cell culture supernatants that were secreted by the induced RAW264.7 cells treated with different doses of IL-10. The sandwich ELISA development kits specific for capturing murine TNF-α (900-T54) and mIL-10 (900-T53) were used with ELISA TMB Buffer Kit (900-T00) (all from PeproTech) for the colorimetric detection in prepare 96-well standard ELISA microplates (Greiner Bio-One North America, Monroe, NC, United States). ELISA was performed following the manufacturer’s instructions. Absorbance readings at 450 nm were recorded with an EnSpire Multilabel Reader (Perkin Elmer, Shelton, CT, United States).

The Institutional Animal Care and Use Committee (IACUC) at the University of Pittsburgh (UPitt) and University of South Dakota (USD) approved the animal usage and surgical procedures performed in this study (UPitt IACUC Protocol 12010140 and USD IACUC Protocol 03-01-22-25D). A total of 73 male 10–12 weeks old BALB/cJ mice (Jackson Laboratory, Bar Harbor, ME, United States) were used for this study.

The same rhIL-10 employed in vitro was also used for in vivo studies. The induction of AMI and intramyocardial injections were performed as we previously described (Figure 3A) (Xiang et al., 2011; Chen et al., 2013; Chu et al., 2013) Briefly, after the induction of anesthesia with 4% isoflurane gas, mice were intubated and inhalationally anesthetized with 2% isoflurane gas throughout the surgery. MI was microscopically induced by permanent ligation of the left anterior descending (LAD) coronary artery. Mice were then randomly assigned to one of the six IL-10 treatment groups: 25, 100, 250, 500, 1,000, and 2000 ng of rhIL-10 (all diluted in 30 µL of sterile 0.9% saline). Ten minutes after the induction of MI, injections were microscopically performed at three sites of the ischemic myocardium (at the center and two borders of the infarct area, each receiving 10 µL solution). Control mice received 30 µL saline (Baxter Healthcare, Deerfield, IL, United States) at the same three sites, following the same procedures.

Echocardiographic studies were performed by a blinded investigator repeatedly before surgery and at 5 days, 2 weeks, and 6 weeks post-surgery to assess the cardiac function as described previously (Chen et al., 2013; Chu et al., 2013; Okada et al., 2012). Concisely, mice were initially anesthetized with 2% isoflurane gas and subsequently maintained at 1%–1.5% isoflurane gas throughout the echocardiographic study. Mice were then immobilized on a heated stage equipped with electrocardiography. Heart rate, respiratory rate, and body temperature were continuously monitored and maintained. Echocardiographic parameters were measured using high-frequency ultrasound scanners (Vevo 770 and 2,100 at the Center for Ultrasound Molecular Imaging and Therapeutics, University of Pittsburgh Medical Center, and Vevo 3,100 at the University of South Dakota (USD) Animal Resource Center; both from FUJIFILM VisualSonics Inc., Canada). End-systolic dimension (ESD) and end-diastolic dimensions (EDD) were determined from the short-axis images of the LV using M-mode. At least eight consecutive beats were measured. End-systolic area (ESA) and end-diastolic area (EDA) were measured from short-axis images of the LV using B-mode. Contractile parameters, including LV fractional shortening (LVFS) and LV ejection fraction (LVEF), were determined as previously described (Manning et al., 1994; Pollick et al., 1995; Wandt et al., 1999). Mice that were sacrificed for histological analyses or died prior to the end point of 6 weeks post-injection were excluded from the echocardiographic studies.

Animals were sacrificed at 6 weeks post-surgery. Hearts were harvested and arrested in diastole by immediately immersing in 1 M potassium chloride (KCl) and processed as previously reported (Xiang et al., 2011; Chen et al., 2013; Chu et al., 2013). Hearts were flash-frozen in 2-methylbutane (Sigma-Aldrich, St. Louis, MO, United States) that was pre-cooled in liquid nitrogen or dry ice. Frozen hearts were stored at −80°C and subsequently serially cryosectioned at 8-µm thickness from the apex to the ligation level (approximately 0.5 mm in length). Heart slices at 1-mm thickness were collected immediately below the ligation level for molecular analyses from randomly selected heart samples. Each series contains 18–24 heart sections, which are roughly 200 µm apart natively and collected on one glass slide. Sections were fixed in a pre-cooled (−20°C) mixture of methanol and acetone (1:1) for 5 min or in 4% paraformaldehyde for 8 min at room temperature (RT) immediately prior to staining (all from Sigma-Aldrich). For immunohistochemistry, non-specific antibody binding was blocked with 10% donkey or goat serum for 1–2 h at RT and, if necessary, with the Mouse-on-Mouse (M.O.M.) antibody staining kit (Vector Laboratories, Burlingame, CA, United States). For evaluation of chronic inflammation, sections were incubated overnight at 4°C with rat anti-mouse CD68 primary antibody (1:200, Abcam, Cambridge, MA, United States), followed by goat anti-rat-Alexa488 IgG (1:400, Life Technologies, Grand Island, NY, United States). To examine vascular endothelial cells (ECs) and smooth muscle cells (SMCs), sections were first incubated overnight at 4°C with rat anti-mouse CD31 (platelet endothelial cell adhesion molecule, PECAM-1) antibody (1:100, Becton-Dickinson Biosciences, Franklin Lakes, NJ, United States), followed by donkey anti-rat-Alexa594 IgG (1:400, Life Technologies) at RT for 1 h, and subsequently incubated with mouse anti-alpha smooth muscle actin-FITC (1:100, Sigma-Aldrich). Nuclei were stained with Hoechst 33,342 diluted in PBS (1:1,000, Life Technologies) at RT for 5 min.

To evaluate chronic inflammation within the infarct region, immunofluorescent staining of anti-mouse CD68 was performed on the serial cryosections of flash-frozen mouse hearts. The macrophage infiltration index, indicated by the number of CD68⁺ phagocytic cells per mm², was subsequently estimated by averaging CD68⁺ cell counts in 6-8 randomly selected images of the entire infarct region of each heart at the mid-infarct level. To quantify revascularization in the heart, double immunofluorescent staining with antibodies against a mouse EC marker, CD31, and a mammalian SMC marker, αSMA, was sequentially performed on the serial cryosections of mouse hearts. The capillary density, indicated by the number of CD31⁺ capillary ECs per mm², was subsequently estimated by averaging CD31⁺ cell counts in 6 randomly selected images of the infarct region or peri-infarct area of each heart at the mid-infarct level; the SMC density, indicated by the number of perivascular (i.e., adjacent to CD31⁺ ECs) αSMA + cells per mm², was subsequently computed by averaging αSMA + cell counts in 6 randomly selected images of the infarct region or peri-infarct area of each heart at the mid-infarct level, as described previously (Chen et al., 2013; Okada et al., 2012). All images were analyzed using ImageJ.

Masson’s trichrome staining kit (IMEB, San Marcos, CA, United States) was used to reveal collagen deposition on heart serial cryosections, following the manufacturer’s instructions. The area of the collagen deposition (indicating cardiac fibrosis) and the area of the entire left ventricular cardiac tissue (including the septal area and excluding void space in the chamber cavity) were measured using a digital image analyzer (ImageJ). Fibrotic area fraction was estimated as the ratio of fibrotic tissue to the entire cross-sectional area of left ventricular tissue and averaged from 6 randomly selected sections at comparable infarct levels per heart.

Quantitative polymerase chain reaction (qPCR) was performed to quantify relative expression levels of key genes in the infarcted hearts 5 days after the induction of MI. The SYBR Green Master Mix for qPCR (A25741; Applied Biosystems, Waltham, MA, United States) was used with forward and reverse primer pairs (Integrated DNA Technologies, Newark, NJ, United States) for the qPCR gene expression analysis (Table 1). Glyceraldehyde-3-phosphate dehydrogenase (GAPDH) was used as the endogenous control gene for delta Ct calculation. Genes investigated include: TNF-α, IL-6, IFNγ, inducible nitric oxide synthase (iNOS), IL-1β, mIL-10, mannose receptor (CD206), arginase 1 (Arg1), AXL receptor tyrosine kinase (Axl), T-cell membrane protein 4 (Tim4), MER proto-oncogene tyrosine kinase (MerTK), chemokine (C-X3-C motif) ligand 1 (CX3CL1), vascular endothelial growth factor-A (VEGF-A), hyaluronidase 3 (Hyal3), alpha smooth muscle actin (α-SMA), v-set immunoglobulin domain-containing 4 (Vsig4), Fc-gamma receptor 1 (CD64), scavenger receptor cysteine-rich type 1 protein M130 (CD163), platelet-derived growth factor (PDGF), periostin (Postn), and type I collagen (Col1a1) (Jung et al., 2017). The fold change in gene expression was calculated using the 2^(−ΔΔCt) method (n = 5 per group).

All measured data are presented as mean ± standard error (SE). Statistical differences between groups were analyzed by one-way analysis of variance (ANOVA) or two-way repeated ANOVA (for repeated echocardiographic measurements) with a 95% confidence interval. Statistical significance was set at p≦0.05. For multiple comparisons post hoc analysis, Bonferroni or Dunnett test was performed for one-way ANOVA, and Tukey test was performed for two-way ANOVA. Statistical analyses were performed with GraphPad Prism 10 statistics software (Dotmatics, Boston, MA, United States).

The authors declare that all relevant data supporting the findings of this study are available within the paper and its supplementary Information. Source data files for the quantitative data generated in this study are available upon request.

To assess the efficacy of IL-10 for modulating the phenotypes of polarized Mφ, we first differentiated naïve RAW264.7 monocytes to classic M1 and M2 Mφ with mIFNγ and mIL-4, respectively (Supplemental Figure 1). We then treated polarized Mφ individually with differential doses of IL-10 (0, 25, 100, 250, 500, and 1,000 ng/mL) for 24 h. Flow cytometry analysis showed that the 250 group has significantly lower CD86 (classic M1 marker) expression when compared with no IL-10 treatment (0 ng/mL); the 500 and 1,000 groups have notably higher CD86 expression when compared with the 250 group (Figures 1A, B). The ELISA assay revealed that M1 Mφ in the 250 group secreted notably less pro-inflammatory cytokine murine TNF-α (mTNF-α) when compared with 0, 100, and 1,000 groups (Figure 1C). On the other hand, the 25, 100, and 250 groups had the highest CD206 (classic M2 marker) expression when compared with no IL-10 treatment while the 500 and 1,000 groups had ∼45–50% and ∼40–45% lower CD206 expression than the 100 and 250 groups, respectively (Figures 1D, E). The ELISA assay showed that all IL-10 treated groups have substantially increased secretion of native murine IL-10 (mIL-10), with the 250 group showing the highest mIL-10 secretion (Figure 1F), consistent with our flow cytometry data. No antibody cross-reaction between mIL-10 and exogenous hIL-10 was observed (Supplemental Figure 2). These results suggest that one can induce drifting of established Mφ phenotypes from the pro-inflammatory/destructive toward the anti-inflammatory/resolutive by administering appropriate doses of IL-10 (Yunna et al., 2020).

To investigate how different exogenous IL-10 doses would impact the Mφ polarization process, we co-stimulated naïve RAW264.7 monocyte differentiation with one of the six doses of IL-10: 0, 25, 100, 250, 500, and 1,000 ng/mL and a standard polarizing agent (mIFNγ for M1 and mIL-4 for M2). For M1 polarization, flow cytometry analysis showed that all groups treated with ≥100 ng/mL IL-10 have significantly decreased M1 polarization, but not the 25 ng/mL, when compared with the 0 ng/mL (Figures 2A, B). The ELISA assay indicated that all IL-10 treated groups have notably less mTNF-α secretion when compared with the 0 group (Figure 2C). It appeared that higher doses of IL-10 may have stronger effects in reducing M1 polarization. For M2 polarization, the 100 group exhibited the highest CD206 expression when compared with the 0, 250, and 1,000 groups (Figures 2D, E). The ELISA assay revealed that all IL-10 treated groups secrete substantially more native mIL-10 when compared with the 0 group, with the 500 group having the highest secretion (Figure 2F). These data suggested that when co-administered with a polarizing agent, IL-10 has notable dose-dependent effects over Mφ polarization, with the potential to achieve a distinct balance of M1 or M2 phenotype with the addition of an appropriate dose of IL-10.

To evaluate the potential of IL-10 as an immunomodulatory treatment for enhancing cardiac repair post-MI, we examined the cardiac function after single intramyocardial administration of different IL-10 doses in a mouse MI model (Figure 3A). A dose range of 25–2000 ng IL-10 was chosen for testing based on the outcomes of the in vitro studies. The surgical mortality rate (death during the operation or within 60 min after the surgery) was ∼17% (1 in 7) in all groups. Among all animals that fully recovered from the surgery, one died before the terminal time point in each of these four groups: saline control, 25, 100, and 2000, and was excluded from this study. In total, five animals (n = 5) per group with no notable behavioral abnormality were included in the 6-week repeated echocardiographic assessment.

Cardiac contractile function was estimated by LV fraction shortening (LVFS, Figure 3B) and LV ejection fraction (LVEF, Figure 3C), whereas LV chamber size was determined by LV end-diastolic area (LVEDA, Figure 3D) and end-systolic area (LVESA, Figure 3E). At 5 days post-MI, the 2000 group had significantly lower LVFS than all groups, except the 500 and 1,000 groups (Supplemental Figure 3). Particularly, the 1,000 and 2000 groups exhibited notably lower LVEF than the 250 group (Figure 3C). At 6 weeks post-MI, the 25, 250, and 500 groups showed considerably higher LVFS than the saline control (Figure 3B); however, only the 250 group exhibited notably higher LVEF than the saline control (Figure 3C, p = 0.0067) while showing significantly higher LVFS than the 2000 group (Figure 3B; p = 0.0202). These results suggest that single intramyocardial treatment with ≥1,000 ng IL-10 has an early negative impact in LV contractile function while injecting 250 ng IL-10 appears to offer more consistent LV contractile protection than other doses in the long term.

Moreover, at 6 weeks post-MI, the 25, 250, and 2000 groups exhibited significantly smaller LVEDA (Figure 3D) and LVESA (Figure 3E) when compared with the saline control. The 100 group notably reduced LVEDA (p = 0.007) but not LVESA (Figures 3D, E). These data suggest that single intramyocardial injection with select IL-10 doses can help ameliorate LV dilatation. Overall, our echocardiographic analyses suggest that 1) 250 ng IL-10 treatment resulted in the most significant long-term functional benefits among all doses, and 2) IL-10 has noteworthy dose-dependent effects when administered intramyocardially, including early adverse effects in LV contractile function with excessive exogenous IL-10.

To assess the early impact of intramyocardial IL-10 administration in modulating myocardial inflammation, Mφ function, and cardiac remodeling, we investigated the expression of 16 essential genes involved in relevant functional pathways in the infracted mouse hearts treated with 0 (saline control), 250, and 1,000 ng IL-10 at 5 days post-MI. Our qPCR analysis revealed that most of these genes exhibit distinct expression patterns after treating with 250 and 1,000 ng IL-10 (Figure 4).

Comparing with the saline control, despite inducing notable upregulation in multiple beneficial genes, including the anti-inflammatory CD206 (Figures 4B–H, p = 0.0166) and Arg1 (Figures 4B–I, p < 0.0001), the Mφ efferocytotic Tim4 (Figures 4C–L, p = 0.0005) and MerTK (Figures 4C–M, p < 0.0001), the chemotactic CX3CL1 (Figures 4D–N, p < 0.0001), the angiogenic VEGF-A (Figures 4D–O, p < 0.0001), and the anti-fibrotic Vsig4 (Figures 4E–Q, p < 0.0001) genes, the 1,000 group also considerably upregulated pro-inflammatory and pro-fibrotic genes, including TNF-α (Figure 4A, p = 0.0009), IL-6 (Figures 4A, B, p = 0.0159), iNOS (Figures 4A–D, p = 0.0015), Axl (Figures 4C–K, p < 0.0001), α-SMA (Figures 4E–P, p < 0.001), the hyaluronan-degrading Hyal3 (Figures 4E–R, p = 0.0009), PDGF (Figures 4E–S, p < 0.0001), Postn (Figures 4E–T, p = 0.0004), and Col1a1 (Figures 4E–U, p = 0.9989).

In contrast, comparing with the saline control, the 250 group showed notable upregulation of the immune-activating IFNγ (Figures 4A–C, p < 0.0001), CX3CL1 (Figures 4D–N, p < 0.0001), and Hyal3 (Figures 4E–R, p = 0.0348) genes significant downregulation of macrophage M1 phenotype marker CD64 (Figures 4A–F, p = 0.021) as well as modest downregulation of iNOS and Col1a1 genes. These results suggest that distinct IL-10 doses differentially regulate essential genes involved in the pro- and anti-repair pathways, potentially contributing to the sustained long-term LV function in the 250 group as well as the short-term contractile dysfunction observed in high-dose IL-10 treatment at 5 days post-MI.

Based on the results of the cardiac functional studies above, we selected a dose range of 25–1,000 ng IL-10 per heart to investigate whether single intramyocardial IL-10 treatment has any dose-dependent immunomodulatory effect on chronic inflammation. At 6 weeks post-MI, phagocytic cells at the infarct site were detected with anti-CD68 immunofluorescent staining (Figure 5A). Unaffected myocardium contained few CD68⁺ cells (Figures 5A, A). All IL-10-treated groups exhibited a trend of diminished numbers of CD68⁺ cells within the infarct region (Figures 5A, C–F) compared with the saline-injected control (Figures 5A, B). Quantitatively, three IL-10-treated groups showed significantly less CD68⁺ cell infiltration than the saline control group (969.81 ± 120.02/mm²): 100 (522.20 ± 71.62/mm²), 250 (379.41 ± 20.50/mm²), and 1,000 (356.83 ± 60.17/mm²) (Figure 5B, n = 3 per group). These results suggest that single intramyocardial administration with ≥100 ng IL-10 soon after MI has a long-lasting modulatory effect on MI-induced chronic inflammation.

To investigate the effect of early intramyocardial IL-10 treatment on cardiac fibrosis, we evaluated LV scar tissue formation (stained in blue/purple) with Masson’s trichrome histological staining. Healthy hearts had minimal deposition of collagen (Figures 6A, A). At 6 weeks post-infarction, hearts treated with different doses of IL-10 (Figures 6A, C–F) exhibited a trend of smaller infarct regions and less collagen deposition than the saline-injected control at the mid-infarct level (Figures 6A, B). Quantitative estimation of the LV fibrotic area fraction revealed a common trend of reduction in fibrosis in all IL-10 treated groups compared with the saline control group (Figure 6B, n = 3 per group). Only the 250 group exhibited a statistically significant 53% reduction in LV fibrosis (Figure 6B), suggesting that single intramyocardial IL-10 administration has a dose-dependent, long-term anti-fibrotic effect.

We examined whether single administration of IL-10 enhances long-term revascularization within the ischemic myocardium. At 6 weeks post-infarction, the presence of CD31⁺ endothelial cells (ECs) and αSMA + perivascular cells (i.e., pericytes and vascular smooth muscle cells adjacent to CD31⁺ ECs) at the infarct region were simultaneously detected with dual immunofluorescent staining (Figure 7A). Quantitative analysis (n = 3 per group) showed that only the 250 group had a significantly higher density of CD31⁺ ECs at the infarct region than the saline control (Figure 7B); all other IL-10 treatment groups had marginal benefits (Figure 7B). No IL-10 treatment groups exhibited a notable increase in CD31⁺ EC density in the peri-infarct areas compared with the saline control group (Figure 7C).

On the other hand, only the 1,000 group had a significantly higher density of αSMA + perivascular cells at the infarct region than the saline control group (Figure 7D); all other IL-10 treatment groups contributed to the repopulation of perivascular cells post-MI marginally (Figure 7D). None of the IL-10 treatment groups showed significant increases in αSMA + cell density in the peri-infarct areas compared with the saline control group (Figure 7E). These data suggest that when intramyocardial administered, select doses of IL-10 have notable effects in enhancing long-term repopulation of myocardial microvasculature at the infarct region, but not in the peri-infarct areas.