Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2022 Sep 1.
Published in final edited form as: J Mol Cell Cardiol. 2021 May 21;158:38–48. doi: 10.1016/j.yjmcc.2021.05.008

Dimethyl fumarate preserves left ventricular infarct integrity following myocardial infarction via modulation of cardiac macrophage and fibroblast oxidative metabolism

Alan J Mouton 1,2,*, Elizabeth R Flynn 1, Sydney P Moak 1, Nikaela M Aitken 1, Ana C Omoto 1,2, Xuan Li 1,2, Alexandre A da Silva 1,2, Zhen Wang 1,2, Jussara M do Carmo 1,2, John E Hall 1,2
PMCID: PMC8522337  NIHMSID: NIHMS1708170  PMID: 34023353

Abstract

Myocardial infarction (MI) is one of the leading causes of mortality and cardiovascular disease worldwide. MI is characterized by a substantial inflammatory response in the infarcted left ventricle (LV), followed by transition of quiescent fibroblasts to active myofibroblasts, which deposit collagen to form the reparative scar. Metabolic shifting between glycolysis and mitochondrial oxidative phosphorylation (OXPHOS) is an important mechanism by which these cell types transition towards reparative phenotypes. Thus, we hypothesized that dimethyl fumarate (DMF), a clinically approved anti-inflammatory agent with metabolic actions, would improve post-MI remodeling via modulation of macrophage and fibroblast metabolism. Adult male C57BL/6J mice were treated with DMF (10mg/kg) for 3–7 days after MI. DMF attenuated LV infarct and non-infarct wall thinning at 3 and 7 days post-MI, and decreased LV dilation and pulmonary congestion at day 7. DMF improved LV infarct collagen deposition, myofibroblast activation, and angiogenesis at day 7. DMF also decreased pro-inflammatory cytokine expression (Tnf) 3 days after MI, and decreased inflammatory markers in macrophages isolated from the infarcted heart (Hif1a, Il1b). In fibroblasts extracted from the infarcted heart at day 3, RNA-Seq analysis demonstrated that DMF promoted an anti-inflammatory/pro-reparative phenotype. By Seahorse analysis, DMF did not affect glycolysis in either macrophages or fibroblasts at day 3, but enhanced macrophage OXPHOS while impairing fibroblast OXPHOS. Our results indicate that DMF differentially affects macrophage and fibroblast metabolism, and promotes anti-inflammatory/pro-reparative actions. In conclusion, targeting cellular metabolism in the infarcted heart may be a promising therapeutic strategy.

Keywords: myocardial infarction, immunometabolism, inflammation, cardiac fibrosis, macrophage, fibroblast

Graphical Abstract

graphic file with name nihms-1708170-f0001.jpg

1.0. INTRODUCTION

Approximately 1 million US citizens experience a myocardial infarction (MI) each year. MI is characterized by an acute inflammatory response, followed by remodeling of the left ventricle (LV) and formation of a collagenous scar.[1, 2] It has recently been demonstrated clinically that targeting inflammation is an effective strategy to attenuate deleterious post-MI outcomes.[3] Both macrophages and fibroblasts play crucial roles in remodeling of the left ventricle, and therapeutic strategies have aimed at shifting these cell types towards reparative rather than inflammatory phenotypes.[4] Modulating cellular metabolism from glycolysis towards oxidative phosphorylation (OXPHOS) promotes anti-inflammatory and pro-reparative macrophage and fibroblast phenotypes.[5–8]

Dimethyl fumarate (DMF) is an anti-inflammatory drug clinically approved for treatment of psoriasis and multiple sclerosis (MS), and exerts cardioprotective actions in rodent models of MI.[9, 10] Recently, a novel role of DMF was discovered in its ability to covalently succinate glyceraldehyde-3-phosphate dehydrogenase (GAPDH) and inhibit its activity in activated macrophages during progression of MS.[11] However, whether DMF exerts cardioprotective actions after MI by modulating macrophage and fibroblast metabolism has not been investigated. Thus, we hypothesized that DMF would protect against MI-induced cardiac injury by promoting reparative macrophage and fibroblast phenotypes via metabolic reprogramming in these cell types.

2.0. METHODS

Mouse model of myocardial infarction and DMF treatment.

All procedures were approved by the Institutional Animal Care and Use Committee at the University of Mississippi Medical Center. Surgical ligation of the left coronary artery was performed to produce MI in adult male C57BL/6J mice (4–5 months age) as described previously.[1, 4] Mice were anesthetized with 2% isoflurane, and intubated through the trachea and connected to a ventilator. A small incision was made on the left flank, and the heart was visualized between the 3rd and 4th ribs. The left coronary artery was permanently ligated with 7–0 non-absorbable suture, and MI was confirmed by blanching of the LV below the suture. Mice were given post-operative analgesics (0.05mg/kg buprenorphine) immediately and 24hr following surgery. DMF (Sigma Aldrich #242926) was diluted in saline + 10% DMSO and injected into the intraperitoneal cavity (10mg/kg body weight) 2 hr after MI, and daily for 3, 7, or 28 days after MI. This dose was chosen based on previous MI studies in rodents and a tolerable dose for humans.[9, 10]

Echocardiography.

LV function was assessed using echocardiography (VEVO 3100, VisualSonics). Mice were anesthetized (2% isoflurane) and laid in the supine position on a heating surface (37°C) for the duration of the procedure. LV long-axis and short-axis B mode and M-mode images were obtained. Images were analyzed using VEVOLAB software (version 3.2.6). Ejection fraction and LV volumes were calculated from long-axis B mode images using the LV trace function, and wall thickness and LV diameters were measured from short-axis M mode tracings.

Infarct Size and Interstitial Fibrosis.

Mid-LV sections were fixed overnight in zinc-formalin, paraffinized, and sectioned at 5μm and attached to microscope slides. The sections were then stained with picrosirius red to detect collagen. Infarct size was assessed in whole LV cross-sectional images (Lionheart FX; Biotek; Winooski, VT) and measured as the percent of total LV area occupied by collagen (ImageJ). Interstitial fibrosis was assessed in remote and infarct areas at 40X magnification in at least 10 random images per field using inForm software.

Immunofluorescence.

Numbers of macrophages and fibroblasts were quantified by immunofluorescence in fixed mid-LV sections. Sections were cleared and rehydrated in decreasing concentrations of ethanol. Immunofluorescence was performed using the Opal 7-color Automation Kit (Perkin Elmer #NEL80100KT; Boston, MA) according to the manufacturer’s instructions. Samples were incubated with primary antibodies against alpha-smooth muscle actin (α-SMA; 1:1000; Sigma Aldrich #2547), inducible nitric oxide synthase (iNOS; 1:100; Novus Biologicals #NB300), or CD31 (1:100; Cell Signaling Technologies #77699) followed by HRP-conjugated secondary antibody supplied by the kit, and finally conjugated to Opal 520 fluorophore and counterstained with DAPI. At least 10 random fields per sample were obtained at 40X using the Mantra System (Perkin Elmer; Boston MA) with the observer blinded to the sample, and cells counts were automated using inForm software (Perkin Elmer).

Immunoblotting.

LV protein expression was determined by immunoblotting. LV tissue (~25mg) was homogenized in T-PER buffer with protease and phosphatase inhibitors. Protein expression was assessed by BCA assay. Protein (20μg) was loaded onto polyacrylamide gels and electrophoresed, then transferred to nitrocellulose membranes. After blocking for 1 hr at room termperature, total protein was stained (Revert™ 700 Total Protein Stain Kit; Li-Cor #P/N 926–11010) and membranes were incubated with the following primary antibodies overnight at 4°C: phospho-AMP-activated kinase (phospho-AMPK threonine-172, 1:1000; Cell Signaling Technologies #2535), AMPK (1:1000; Cell Signaling Technologies #5831), phospho-acetyl-CoA carboxylase (phospho-ACC, serine 79; 1:1000; Cell Signaling Technologies #11818), ACC (1:1000; Cell Signaling Technologies #3662), phospho-pyruvate dehydrogenase (phospho-PDH, serine 293; 1:1000; Cell Signaling Technologies #31866), PDH (1:1000; Abcam #155096), and α-SMA (1:1000; Sigma #2547) followed by incubation with secondary antibody (IRDye® 8000CW Donkey Anti-Mouse or Anti-Rabbit IgG; 1:10,000; LiCor) for 1 hr at room temperature. Membranes were then visualized using an Odyssey® CLx Imaging System (LiCor). Protein expression was normalized to the total amount of protein in each lane.

Quantitative Polymerase Chain Reaction.

RNA was extracted by the UMMC Molecular Genomics Core (PureLink RNA Mini-Kit; Fisher). RNA (1μg) was transcribed into cDNA (iScript Reverse Transcription Supermix for RT-qPCR; Bio-Rad). cDNA (10–50ng) was loaded into 96-well PCR plates, and target genes were amplified using SYBR® Green Master Mix (Bio-Rad) and the following primers (Integrated DNA Technologies): Ccl2, Col1a1, Col3a1, Eln, Fh1, Fn1, Hif1a, Idh1, Il1b, Il6, Myh6, Myh7, Nrf2, Sdha, Sdhb, and Tnf. Gapdh was used as a housekeeping gene for LV tissue, while Actb was used for isolated macrophages and fibroblasts. The reaction was performed using the StepOnePlus™ Real-Time PCR System (Applied Biosystems) for 40 cycles. Changes in gene expression were assessed by the ΔΔCt method and displayed as fold change.

Cardiac Cell Isolation.

Macrophages and fibroblasts were isolated from post-MI hearts by immunomagnetic separation. Hearts were enzymatically digested at 37°C for 1 hr (600U/mL collagenase, Worthington; 60U/mL DNase I, Sigma Aldrich; diluted in HBSS). Digested tissue was filtered through 30μm columns to generate single cell suspensions, then incubated with anti-Ly6G microbeads and filtered through magnetic columns to remove neutrophils (Miltenyi). Flow through was then incubated with anti-CD11b microbeads to purify macrophages, which were immediately plated in RPMI media with 0.1% fetal bovine serum (FBS) for 2 hr. For Seahorse® assays, 5 × 105 cells were plated in Seahorse® 24-well cell culture plates, while 1.0–1.5 × 106 cells were plated for RNA extraction and collected in 1mL Trizol. For fibroblast isolation, single cell suspensions were plated on T25 flasks for 2 hr in DMEM media with 10% FBS. Non-adherent cells were then removed and replaced with fresh DMEM + 10% FBS. Cells were allowed to grow and collected at 80–90% confluence in Trizol for RNA extraction (passage 0) or passaged with 0.25% trypsin and plated in Seahorse® 24-well cell culture plates (2 × 105) overnight in DMEM + 1% FBS for Seahorse assay. For in vitro treatment, fibroblasts were exposed to 10, 25, or 50 μM DMF for 24 hr or equivalent volume of vehicle control (DMSO).

Bone Marrow-Derived Macrophage Culture.

Bone marrow-derived macrophages (BMDMs) were harvested from the bone marrow by flushing the tibias and femurs with cell culture media (RPMI + 10% FBS) using a 27-gauge needle. The effluent was centrifuged and re-suspended in media, and plated using 5 × 106 cells in 10cm cell culture dishes. CSF-1 (Biolegend #576404) was added to the medium at a concentration of 25 ng/μL, and non-adherent cells were washed off every 3 days. The remaining BMDMs were used for in vitro experiments. To stimulate macrophage activation, BMDMs were exposed to lipopolysaccharide (LPS, 1 μg/mL, Sigma-Aldrich) for 24 hours and either 10, 25, or 50μM DMF or equivalent volume of vehicle control (DMSO).

Seahorse Metabolic Flux Analysis.

Glycolysis and mitochondrial oxidative phosphorylation was assessed in live cells using a Seahorse XFe24 analyzer (Agilent). Glycolysis was measured by the extracellular acidification rate (ECAR), and the glycolysis stress test was performed using the following injections: glucose (10mM) to measure basal glycolysis, oligomycin (1μM) to measure maximal glycolysis, and 2-deoxyglucose (50mM) to measure glycolysis-specific ECAR. The mitochondrial stress test was performed to assess oxygen consumption rate (OCR) using the following injections: oligomycin (1μM) to measure ATP-linked respiration and proton leak, FCCP (1μM) to measure maximal respiration and spare capacity, and rotenone/antimycin A (0.5μM) to measure non-mitochondrial OCR. To account for inter-assay variability, results are displayed as fold change over control.

GAPDH Activity Assay.

GAPDH activity was measured in BMDM or cardiac fibroblast cell lysates using a colorimetric assay (Abcam #204732). Cells were plated in 6-well plates, allowed to adhere overnight, and treated with LPS (BMDMs) and increasing doses of DMF (10, 25, and 50 μM; BMDMs and cardiac fibroblasts) for 24 hours. Cells were then washed with ice-cold PBS, and collected. The assay was performed according to the manufacturer’s instructions, and the concentration of NADH produced over time was measured using a multi-well spectrophotometer at a wavelength of 450nm.

Cell Proliferation Assay.

Fibroblast proliferation was assessed by CyQuant Cell Proliferation Assay (ThermoFisher #C7026). Cells were collected, lysed, and the number of proliferating cells was measured with the CyQuant dye using a microplate reader by measuring the fluorescence at 485nm excitation and 520nm emission.

RNA-Sequencing and Bioinformatic Analysis of the Transcriptome.

cDNA libraries were developed using the TruSeq Total Stranded RNA with RiboZero Kit (Ambion), then sequenced with the NextSeq 500 High Output Kit (150 cycles, paired end 75 bp) on the Illumina NextSeq 500 platform (Illumina, San Diego, CA). Alignment and differential expression using DESeq2 was performed on the Cloud Computing Platform (Basespace) to generate transcripts per million (TPM). Analysis of major affected biological pathways in the transcriptome was performed using Ingenuity® Pathway Analysis (Qiagen).

Statistics.

Data was analyzed using GraphPad 8 (Prism) and is presented as mean ± standard error margin (SEM). Survival rate was assessed by Kaplan-Meier analysis followed by the log-rank test. A parametric (Gaussian distribution) and unpaired (similar standard deviation) Student’s t-test was performed to assess significant differences between vehicle and DMF groups. A p-value of less than 0.05 was considered statistically significant.

3.0. RESULTS

3.1. DMF improves LV functional outcomes and pulmonary congestion after MI.

Day 7 survival and infarct size were not significantly different between groups (Figure 1A). At day 7 in surviving mice, DMF significantly attenuated thinning of both the infarcted and non-infarcted walls, and attenuated LV dilation as assessed by the end-diastolic volume, while not affecting ejection fraction (Figure 1B). DMF did not affect LV or RV mass (normalized to tibia length), but significantly decreased lung wet and dry mass (Figure 1C). DMF also improved ejection fraction at day 3, and attenuated wall thinning at both day 3 and day 28 (Supplementary Figure 1). Mice treated with DMF also had a higher body weight at day 3, and a trend for decreased wet lung mass (Supplementary Table 1; p=0.054). No other differences in morphometric parameters were observed at day 3 or day 28 (Supplementary Table 1).

Figure 1. Survival, infarct size, LV function, and morphometric parameters at post-MI day 7.

Figure 1.

Open in a new tab

(A) No differences in survival or infarct size were detected between groups. (B) DMF attenuated infarct and non-infarct wall thinning and LV dilation, without affecting ejection fraction. (C) DMF did not affect LV or RV mass, but significantly decreased wet lung and dry lung mass. *p<0.05 versus Vehicle.

3.2. DMF increases collagen deposition and angiogenesis in the infarcted LV.

DMF significantly increased collagen density in the infarct region (Figure 2A). No significant differences in collagen content within the remote region were observed. This effect was limited to day 7, as no differences in remote or infarct collagen were observed at day 28 (Supplementary Figure 1). To assess whether increased collagen content at day 7 was associated with increased myofibroblast activation, α-SMA+ cells were quantified at day 3 and day 7 (Figure 2B). No differences were observed at day 3, but DMF significantly increased α-SMA+ cells in the infarct region at day 7. DMF also improved angiogenesis in the infarct region at day 7, as assessed by CD31 (endothelial marker) immunofluorescence (Figure 2C).

Figure 2. Effects of DMF on infarct fibrotic and vascular remodeling.

Figure 2.

Open in a new tab

(A) DMF significantly increased collagen deposition in the infarct region, without affect the remote region. (B) DMF significantly increased the number of α-SMA+ myofibroblasts in the infarct region at day 7. (C) DMF significantly increased angiogenesis as assessed by CD31 expression in the infarct region at day 7. *p<0.05 versus Vehicle.

3.3. DMF decreases LV inflammatory markers and enhances Nrf2 expression.

To assess whether DMF decreased LV inflammation after MI, we quantified numbers of M1 macrophages in the infarct region by inducible nitric oxide synthase (iNOS) positivity. DMF did not alter the number of iNOS+ cells in the infarct region at day 3 or 7 (Figure 3A). We assessed expression of inflammatory cytokines, including Il1b, Ccl2, Tnf, and Il6 in the remote (non-infarcted) area of the LV (Figure 3B). DMF significantly decreased Tnf in the remote LV at day 3 (inflammatory phase), and trended to decrease Ccl2 (p=0.08), Il6 (p=0.07) and Nrf2 (p=0.06). No differences in inflammatory markers were observed at day 7 (maturation phase), coinciding with resolution of inflammation.[1] Furthermore, expression of the myosin heavy chains Myh6 and Myh7 were not different at day 7, indicating that the increased LV wall thickness observed in DMF groups was not associated with pathological hypertrophy. At day 7, we observed an increase in LV Nrf2 expression in DMF-treated mice, consistent with previous reports.[9]

Figure 3. Effects of DMF on LV macrophages and pro-inflammatory cytokines.

Figure 3.

Open in a new tab

(A) DMF did not affect the number of iNOS+ macrophages in the infarct region at day 3 or day 7. (B) DMF significantly decreased Tnf mRNA levels at day 3, and increased Nrf2 mRNA at day 7. N=6 per group. *p<0.05 versus Vehicle.

3.4. DMF alters LV metabolic pathways after MI.

To assess whether DMF-induced cardioprotection was associated with changes in markers of glucose and fatty acid oxidation,[12] we assessed phosphorylation and total expression levels of ACC, PDH, and AMPK. Representative total protein stains for each blot are depicted in Supplementary Figure 2. At day 3, DMF significantly decreased total ACC expression, but did not alter phosphorylation; no changes in PDH or AMPK were observed (Figure 4A). At day 7, no differences in ACC or PDH were observed; however, phosphorylation of AMPK, an indicator of activated stress signaling in the post-MI heart,[12] was decreased with DMF.

Figure 4. Effects of DMF on LV metabolic and stress pathways (3 and 7 days after MI).

Figure 4.

Open in a new tab

(A) DMF significantly decreased ACC expression, but not phosphorylation, at day 3. (B) DMF significantly decreased AMPK phosphorylation at day 7. ACC—acetyl CoA carboxylase. PDH—pyruvate dehydrogenase. AMPK—AMP activated kinase. N=6 per group. *p<0.05 versus Vehicle.

3.5. DMF promotes LV macrophage metabolic reprogramming towards OXPHOS in association with decreased inflammatory gene expression.

To assess the impact of DMF on macrophage phenotypes, we extracted macrophages from LV infarcts at post-MI day 3 for Seahorse and gene expression analysis. DMF did not affect macrophage glycolytic metabolism as assessed by ECAR (Figure 5A), but significantly increased the spare capacity (maximal OCR – basal OCR), indicating a metabolic shift towards OXPHOS. In cultured BMDMs, DMF significantly decreased basal glycolysis at lower doses (10 and 25 μM), but not at a higher dose (50 μM) (Supplementary Figure 3A). However, 50 μM DMF significantly decreased GAPDH activity (Supplementary Figure 3B). We also assessed whether DMF attenuated glycolysis and GAPDH activity in activated macrophages stimulated with LPS (1 μg/mL for 24 hr). As expected, LPS stimulated an increase in glycolysis and GAPDH activity (Supplementary Figure 3A and 3B). Both 25 and 50 μM DMF attenuated GAPDH activity in LPS-treated BMDMs, while only 50μM DMF attenuated glycolysis. We also investigated whether DMF could alter oxidative metabolism in vitro (Supplementary Figure 3C). DMF did not directly affect OCR, but 25 μM attenuated LPS-mediated decreases in OCR (Supplementary Figure 3). In macrophages isolated from infarcted hearts from DMF-treated mice, there was a significant decrease in Il1b expression, as well as a trend for decreased Il6 expression (Figure 5B). We also assessed expression of several tricarboxylic acid (TCA) cycle enzymes, including Sdha and Sdhb, Fh1, and Idh1, which were not significantly affected by DMF (Figure 5B). However, Hif1a, the master regulator of macrophage metabolic reprogramming away from OXPHOS and toward glycolysis,[13] was significantly decreased by DMF treatment. HIF-1α also activates IL-1β expression in macrophages,[14] indicating that DMF attenuated the HIF-1α-IL-1β axis. In contrast to our results in LV tissue at day 7, Nrf2 was significantly decreased in macrophages at day 3 (Figure 5B). In activated BMDMs, Nrf2 protein expression was significantly decreased by DMF in vitro, whereas Keap1 was not affected by DMF (Supplementary Figure 4).

Figure 5. Effects of DMF on infarct macrophage metabolism and gene expression 3 days after MI.

Figure 5.

Open in a new tab

(A) DMF did not affect macrophage glycolysis but significantly increased spare capacity. (B) DMF significantly decreased Il1b, Hif1a, and Nrf2 mRNA levels in macrophages. ECAR—extracellular acidification rate; OCR—oxygen consumption rate. *p<0.05 versus Vehicle.

3.6. DMF alters the cardiac fibroblast transcriptome characterized by attenuated inflammatory responses and improved ECM remodeling.

To further assess mechanisms of DMF-mediated improvements in infarct remodeling, we isolated fibroblasts at post-MI day 3 and performed RNA-sequencing to decipher changes in the transcriptome. After performing quality control measures in which genes with TPM < 1 for both group averages, genes with TPM = 0 in over half of samples (high variability), or genes with less than 1.2 fold change difference between groups were removed, we found that 907 genes were differentially expressed (Figure 6A). Of these, 466 were upregulated and 441 were downregulated, as displayed in the volcano plot (Figure 6A). Several genes were confirmed by PCR analysis by a significant correlation between fold change (Supplementary Figure 5). We performed pathway analysis (Ingenuity® Pathway Analysis, QIAGEN) to assess the major pathways that were upregulated, downregulated, or dysregulated but the change was not predictable (Figure 6B). Within the top 5 upregulated pathways (ranked by −log (p value)), 4 were related to cholesterol biosynthesis, including geranylgeranyldiphosphate biosynthesis and mevalonate pathway. The top 5 downregulated pathways included inhibition of matrix metalloproteinases, LPS/IL-1 mediated inhibition of RXR function, gamma-glutamyl cycle, Nrf2-mediated oxidative stress response, and Wnt/Beta-catenin signaling. The top 5 pathways that were significantly altered, but in which the change was not predictable, were axonal guidance signaling, granulocyte adhesion/diapedesis, hepatic fibrosis signaling pathway, hepatic fibrosis/stellate cell activation, and agranulocyte adhesion/diapedesis. Within the granulocyte adhesion/diapedesis pathway, several chemokines were significantly decreased, including Ccl2, Ccl7, Ccl20, Cxcl1, Cxcl2, Cxcl3, and Cxcl16 (Supplementary Table 2). Within the hepatic fibrosis signaling pathway, several ECM genes were upregulated, including COL3A1, COL5A1, COL6A1, COL6A2, COL6A3, ELN, and FN1. CSF1, another pro-inflammatory mediated, was also significantly decreased. We also performed analysis of upstream pathways. By p-value, several pro-inflammatory mediators were predicted to be inhibited upstream of these pathways, including TNF, lipopolysaccharide, interferon-gamma (IFNG), NFκB, IL1B, and IL1A, all of which were in the top 10 of upstream inhibited pathways (Supplementary Table 3). Dexamethasone, a pharmacological corticosteroid with powerful anti-inflammatory properties, was the highest ranked predicted upstream activator. TGFB1 was the highest ranked upstream activator without a predicted change, although the activation z-score was positive (indicating a trend towards activation). These results indicate that DMF promotes an anti-inflammatory, pro-reparative phenotype in cardiac fibroblasts following MI. Furthermore, we investigated whether DMF was associated with changes in signaling pathways associated with cardiac fibroblast proliferation and migration into the infarct area. In addition to increased cholesterol biosynthesis, integrin, p70S6K, STAT3, and Notch signaling were all predicted to be activated in fibroblasts isolated from DMF-treated mice (Supplementary Table 4). In vitro, DMF did not affect fibroblast proliferation at 10 μM but decreased proliferation at 25 and 50 μM (Supplementary Figure 6), indicating that the proliferative phenotype observed in vivo may be due to indirect effects of DMF on other cell types and the ischemic environment.

Figure 6. Effects of DMF on the cardiac fibroblast transcriptome 3 days after MI.

Figure 6.

Open in a new tab

(A) Flow chart for assessing differentially expressed (DE) genes and volcano plot (right) showing DE genes. (B) Top 5 dysregulated canonical pathways affected by DMF in cardiac fibroblasts as assessed by p-value (Ingenuity Pathway Analysis).

3.7. DMF reduces OXPHOS metabolism in cardiac fibroblasts after MI.

We then assessed whether DMF could affect cardiac fibroblast metabolism in association with a heightened myofibroblast phenotype. By Seahorse analysis, we found that DMF did not significantly affect cardiac fibroblast glycolytic metabolism, but significantly decreased both the maximal OCR and spare capacity (Figure 7A). Furthermore, DMF decreased several genes involved in mitochondrial electron transport (Figure 7B). In cultured fibroblasts, in vitro treatment with DMF at doses of 25 and 50 μM decreased glycolysis (Supplementary Figure 7A), while all doses of DMF decreased maximal OCR and spare capacity (Supplementary Figure 7B). However, DMF administered in vitro had no effect on cardiac fibroblast expression of α-SMA (Supplementary Figure 7C), indicating that while DMF may have direct effects on cardiac fibroblast metabolism, the increased myofibroblast density that we observed in the infarcted area may be due to indirect effects of DMF.

Figure 7. Effects of DMF on cardiac fibroblast metabolism 3 days after MI.

Figure 7.

Open in a new tab

(A) DMF did not significantly affect cardiac fibroblast glycolysis (ECAR), but decreased maximal OCR and spare capacity. (B) DMF significantly decreased several genes associated with oxidative phosphorylation (RNA-Seq). ECAR—extracellular acidification rate; OCR—oxygen consumption rate. *p<0.05 versus Vehicle.

4.0. DISCUSSION

Our results indicate that DMF improves infarct remodeling through actions on macrophages and fibroblasts, two of the major cell types that directly and actively participate in remodeling of the infarcted myocardium. In both cell types, DMF inhibited pro-inflammatory phenotypes, and increased collagen deposition and attenuated infarct wall thinning through activation of a myofibroblast phenotype, without promoting fibrosis in the remote myocardium. In infarct macrophages, DMF promoted mitochondrial OXPHOS in association with an anti-inflammatory phenotype. Furthermore, DMF decreased pulmonary congestion, an important heart failure outcome which decreases long-term survival and quality of life.[15] While this may partly be due to improved cardiac function, DMF also directly alleviates pulmonary injury and inflammation during pulmonary hypertension.[16] While previous studies have demonstrated cardioprotective effects of DMF via actions on cardiomyocytes during reperfusion injury, [9, 10] ours is the first study, to our knowledge, to demonstrate that DMF improves infarct remodeling in a permanent occlusion model through metabolic actions on non-myocyte cellular phenotypes.

Immunometabolism has attracted significant attention in inflammatory diseases, but studies investigating immunometabolism during cardiovascular disease are lacking.[17, 18] In particular, metabolic reprogramming between glycolysis and OXPHOS is a critical mechanism by which macrophages polarize between pro- and anti-inflammatory phenotypes.[19] While the anti-inflammatory properties of DMF are well known, its mechanism of action has not been clear. Recently, a novel immunometabolic role for DMF was discovered in its ability to inhibit GAPDH and thus inhibit glycolytic reprogramming in activated macrophages.[11] We found that DMF did not inhibit glycolysis, but enhanced OXPHOS at day 3 post-MI. DMF significantly decreased expression of the Hif1a gene, which is considered to be the major upstream activator of glycolytic metabolism,[1, 17, 19] and also promotes inflammatory actions largely via activation of the Il1b gene,[20] which was decreased by DMF in infarct macrophages. This did not result in decreased glycolysis, possibly because glycolysis is essential during the acute inflammatory phase, although it is possible that glycolytic metabolism would have been decreased in DMF-treated mice at a later time point. However, HIF-1α also inhibits mitochondrial gene expression and metabolism,[21] thus DMF may promote OXPHOS in macrophages via downregulation of HIF-1α. Thus, DMF-induced anti-inflammatory actions in macrophages may be mediated by direct inhibition of HIF-1α (Figure 8).

Figure 8.

Figure 8.

Open in a new tab

Proposed mechanisms of how DMF may affect metabolism and transcriptional pathways in macrophages and fibroblasts during MI.

Although the cardioprotective actions of DMF have been implicated as a result of Nrf2 activation,[9] this mechanism may not account for its actions in immune cells.[22] Although Nrf2 was unchanged in LV tissue at day 3 and increased at day 7, it was decreased in macrophages at day 3. Since Nrf2 is activated in response to oxidative stress,[22] it is possible that DMF reduced oxidative stress and thus reduced Nrf2 activation. Nrf2 has also been shown to act as an upstream activator of HIF-1α and regulate metabolic reprogramming in pluripotent stem cells.[23] Our results indicate that activation of Nrf2 does not account for the immunomodulatory effects of DMF in the infarcted heart, and may even depend on Nrf2 downregulation.

We found that DMF increased collagen deposition and angiogenesis in the infarct region by day 7. Angiogenesis is critical for restoring blood flow to the infarcted region, and vascular cells are a key component of granulation tissue to support formation and maintenance of scar tissue.[4] Improving angiogenesis in the infarcted LV is associated with improved outcomes and is a therapeutic target for MI.[24] Cardiac fibroblasts have been implicated in regulating angiogenesis following MI through secretion of angiogenic factors.[4] Remodeling of the cardiac extracellular matrix (ECM) is also a key component of the response to MI.[25] This response is characterized by early degradation of ECM during the acute inflammatory phase, followed by formation of a provisional matrix and finally a stable collagen-based scar by day 7 in the mouse.[26, 27] Fibroblasts begin populating the infarct region and rapidly proliferating 3 days after MI, and by day 7 achieve a maximally activated myofibroblast phenotype, characterized by increased expression of alpha-smooth muscle actin (α-SMA).[4, 27] In line with this, we found that DMF increased the number of α-SMA+ myofibroblasts in the infarct region and increased collagen deposition at day 7, which is associated with increased scar stability, decreased wall thinning, and decreased LV dilation.[27–29] Importantly, this pro-fibrotic effect of DMF was observed only in the infarct region, and not the remote region.

To assess whether this increase in myofibroblast transformation was associated with earlier changes in gene expression, we assessed changes in the transcriptome at day 3 post-MI, at which point the inflammatory phase has subsided and fibroblasts begin to repopulate the infarcted LV.[4] Our results indicate that DMF decreased fibroblast inflammatory pathways and activated ECM remodeling pathways.

In addition to their roles in ECM remodeling, fibroblasts have also been implicated in directly participating in the inflammatory response via secretion of chemokines and cytokines.[17, 30] We found that several inflammatory molecules predicted to act upstream of cardiac fibroblasts were decreased by DMF, including TNF-alpha and IL-1 beta. Likewise, several chemokines known to be expressed by inflammatory cardiac fibroblasts, including CCL2, CCL7, and several CXCL chemokines,[31] were decreased by DMF. Whether this anti-inflammatory effect was directly caused by DMF actions on fibroblasts or indirect effects through macrophage paracrine actions remains to be investigated. We also found that DMF upregulated several genes related to cholesterol metabolism, indicating that DMF increased cholesterol synthesis. Cholesterol synthesis is critical for cardiac fibroblast and myofibroblast function and survival, as inhibitors of cholesterol synthesis (statins) promote cardiac fibroblast apoptosis,[32] while decreasing cholesterol-rich lipid raft content in cardiac fibroblasts impairs proliferation.[33] Thus, DMF may promote cardiac fibroblast survival and proliferation via upregulation of cholesterol synthesis.

DMF also upregulated several other pathways associated with fibroblast proliferation, including integrin, p70S6K, STAT3, and Notch signaling. Beta-3 integrin (ITGB3), which was upregulated, mediates fibroblast adhesion and proliferation during the progression of cardiac fibrosis.[34] The p70S6K signaling pathway mediates proliferation in response to Akt activation by angiotensin II.[35] The STAT3 pathway mediates survival and proliferation in response to TGF-β signaling.[36] Notch signaling has also been reported to mediate differentiation of epicardial cells into reparative fibroblasts after MI.[37] Our in vitro results indicated that DMF did not affect proliferation at lower doses and even inhibited proliferation at higher doses, indicating that DMF may promote fibroblast proliferation in the infarcted heart indirectly, or may promote migration/differentiation of fibroblast progenitors.

We also found that several reparative ECM genes were upregulated in cardiac fibroblasts from mice treated with DMF. In particular, Col3a1, the gene for collagen type III, was upregulated by DMF. Along with collagen I, collagen III is one of the major fibrillar collagens that make up the collagenous cardiac ECM, and confers elastic strength to the ECM.[38] DMF also upregulated Eln, which encodes for elastin, another ECM protein with elastic properties.[39] Interestingly, this pro-reparative effect of DMF was associated with decreased activation of the Wnt/β-catenin signaling pathway. Canonical activation of this pathway has been associated with myofibroblast differentiation in the infarcted heart,[40] and inhibition of Wnt/β-catenin in a mouse model of MI improves infarct remodeling and integrity.[41] Overall, our results indicate that DMF supports an anti-inflammatory, pro-reparative cardiac fibroblast phenotype in the infarcted heart.

In addition to pro-reparative changes in the cardiac fibroblast transcriptome, we also found that DMF treatment resulted in decreased cardiac fibroblast OXPHOS metabolism ex vivo, without affecting glycolytic metabolism. Furthermore, in vitro treatment with DMF decreased OXPHOS, without affecting glycolysis at the lowest dose used. While the precise role of cardiac fibroblast metabolism in health and disease is not fully understood, some studies have indicated that switching between glycolytic and oxidative metabolism underlies phenotypic switching between quiescent fibroblast and active myofibroblast phenotypes, suggesting that quiescent phenotypes are more reliant on OXPHOS while myofibroblast phenotypes switch predominantly to glycolysis.[7, 8, 19, 42] Disrupting mitochondrial function genetically in cardiac fibroblasts promotes myofibroblast differentiation in models of cardiac injury,[8] while pharmacologically enhancing mitochondrial function inhibits myofibroblast transition and fibrosis in a model of pulmonary hypertension.[7] Furthermore, we found that similar to macrophages, ―NRF2-mediated oxidative stress response‖ was a major downregulated process in cardiac fibroblasts from mice treated with DMF, which could be due to decreased reactive oxygen species generation from OXPHOS metabolism.[23] Oxidative stress has also been shown to support the pro-inflammatory function of cardiac fibroblasts, and pro-inflammatory stimuli also induce oxidative stress through NF-κB,[43] which our results suggested was a major downregulated signaling pathway, indicating that DMF may also act through upstream inflammatory mediators to reduce oxidative stress (Figure 8). However, further studies are required to investigate cross-talk between fibroblast metabolism and oxidative stress during cardiac remodeling. Overall, DMF may enhance myofibroblast differentiation in the infarct region via suppression of oxidative phosphorylation. Whether this is a direct effect of DMF or an indirect result of improvements in the cardiac microenvironment remains to be determined in future studies.

5.0. CONCLUSIONS

In conclusion, our results demonstrate that DMF, a clinically approved anti-inflammatory agent, improves post-MI LV remodeling and heart failure outcomes via novel metabolic mechanisms in cardiac macrophages and fibroblasts. Our study provides important therapeutic implications for the rapidly blossoming fields of non-myocyte metabolism in the injured heart.

Supplementary Material

1

Highlights.

  • Dimethyl fumarate (DMF) improves left ventricular (LV) function and pulmonary congestion following myocardial infarction (MI)

  • DMF improves collagen deposition and angiogenesis in the infarcted LV

  • DMF attenuates LV inflammation and increases macrophage oxidative phosphorylation (OXPHOS) metabolism in association with decreased IL-1β expression

  • DMF promotes a reparative cardiac myofibroblast phenotype in association with decreased OXPHOS metabolism

ACKNOWLEDGMENTS

We acknowledge the excellent contributions of the UMMC Histology Core and Molecular and Genomics Core.

FUNDING SOURCES

The authors’ research was supported by National Heart, Lung, and Blood Institute (P01 HL51971), National Institute of General Medical Sciences (P20 GM104357 and U54 GM115428), National Institute of Diabetes and Digestive and Kidney Diseases (R01 DK121411), and the American Heart Association (18POST34000039).

Footnotes

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

DECLARATIONS OF INTEREST

None.

Supplementary data

Supplementary material

REFERENCES

  • 1.Mouton AJ, et al. , Mapping macrophage polarization over the myocardial infarction time continuum. Basic Res Cardiol, 2018. 113(4): p. 26. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2.Mouton AJ, Rivera OJ, and Lindsey ML, Myocardial infarction remodeling that progresses to heart failure: a signaling misunderstanding. Am J Physiol Heart Circ Physiol, 2018. 315(1): p. H71–H79. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3.Everett BM, et al. , Anti-Inflammatory Therapy With Canakinumab for the Prevention of Hospitalization for Heart Failure. Circulation, 2019. 139(10): p. 1289–1299. [DOI] [PubMed] [Google Scholar]
  • 4.Mouton AJ, et al. , Fibroblast polarization over the myocardial infarction time continuum shifts roles from inflammation to angiogenesis. Basic Res Cardiol, 2019. 114(2): p. 6. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.Lewis AJM, et al. , Noninvasive Immunometabolic Cardiac Inflammation Imaging Using Hyperpolarized Magnetic Resonance. Circ Res, 2018. 122(8): p. 1084–1093. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6.Zhang S, et al. , Efferocytosis Fuels Requirements of Fatty Acid Oxidation and the Electron Transport Chain to Polarize Macrophages for Tissue Repair. Cell Metab, 2019. 29(2): p. 443–456 e5. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7.Tian L, et al. , Epigenetic Metabolic Reprogramming of Right Ventricular Fibroblasts in Pulmonary Arterial Hypertension: A Pyruvate Dehydrogenase Kinase-Dependent Shift in Mitochondrial Metabolism Promotes Right Ventricular Fibrosis. Circ Res, 2020. 126(12): p. 1723–1745. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8.Lombardi AA, et al. , Mitochondrial calcium exchange links metabolism with the epigenome to control cellular differentiation. Nat Commun, 2019. 10(1): p. 4509. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.Ashrafian H, et al. , Fumarate is cardioprotective via activation of the Nrf2 antioxidant pathway. Cell Metab, 2012. 15(3): p. 361–71. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Meili-Butz S, et al. , Dimethyl fumarate, a small molecule drug for psoriasis, inhibits Nuclear Factor-kappaB and reduces myocardial infarct size in rats. Eur J Pharmacol, 2008. 586(1–3): p. 251–8. [DOI] [PubMed] [Google Scholar]
  • 11.Kornberg MD, et al. , Dimethyl fumarate targets GAPDH and aerobic glycolysis to modulate immunity. Science, 2018. 360(6387): p. 449–453. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.Karwi QG, et al. , Targeting the glucagon receptor improves cardiac function and enhances insulin sensitivity following a myocardial infarction. Cardiovasc Diabetol, 2019. 18(1): p. 1. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Wang T, et al. , HIF1alpha-Induced Glycolysis Metabolism Is Essential to the Activation of Inflammatory Macrophages. Mediators Inflamm, 2017. 2017: p. 9029327. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Palsson-McDermott EM, et al. , Pyruvate Kinase M2 Regulates Hif-1alpha Activity and IL-1beta Induction and Is a Critical Determinant of the Warburg Effect in LPS-Activated Macrophages. Cell Metab, 2015. 21(2): p. 347. [DOI] [PubMed] [Google Scholar]
  • 15.Dayeh NR, et al. , Echocardiographic validation of pulmonary hypertension due to heart failure with reduced ejection fraction in mice. Sci Rep, 2018. 8(1): p. 1363. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.Grzegorzewska AP, et al. , Dimethyl Fumarate ameliorates pulmonary arterial hypertension and lung fibrosis by targeting multiple pathways. Sci Rep, 2017. 7: p. 41605. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17.Mouton AJ, et al. , Inflammatory cardiac fibroblast phenotype underlies chronic alcohol-induced cardiac atrophy and dysfunction. Life Sci, 2020. 245: p. 117330. [DOI] [PubMed] [Google Scholar]
  • 18.Zhang S, et al. , Immunometabolism of Phagocytes and Relationships to Cardiac Repair. Front Cardiovasc Med, 2019. 6: p. 42. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19.Mouton AJ and Hall JE, Novel roles of immunometabolism and nonmyocyte metabolism in cardiac remodeling and injury. Am J Physiol Regul Integr Comp Physiol, 2020. 319(4): p. R476–R484. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.Corcoran SE and O’Neill LA, HIF1alpha and metabolic reprogramming in inflammation. J Clin Invest, 2016. 126(10): p. 3699–3707. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.Semenza GL, Hypoxia-inducible factor 1: regulator of mitochondrial metabolism and mediator of ischemic preconditioning. Biochim Biophys Acta, 2011. 1813(7): p. 1263–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Schulze-Topphoff U, et al. , Dimethyl fumarate treatment induces adaptive and innate immune modulation independent of Nrf2. Proc Natl Acad Sci U S A, 2016. 113(17): p. 4777–82. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.Hawkins KE, et al. , NRF2 Orchestrates the Metabolic Shift during Induced Pluripotent Stem Cell Reprogramming. Cell Rep, 2016. 14(8): p. 1883–91. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Kobayashi K, et al. , Dynamics of angiogenesis in ischemic areas of the infarcted heart. Sci Rep, 2017. 7(1): p. 7156. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25.Nielsen SH, et al. , Understanding cardiac extracellular matrix remodeling to develop biomarkers of myocardial infarction outcomes. Matrix Biol, 2019. 75–76: p. 43–57. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26.Frangogiannis NG, The extracellular matrix in myocardial injury, repair, and remodeling. J Clin Invest, 2017. 127(5): p. 1600–1612. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27.Fu X, et al. , Specialized fibroblast differentiated states underlie scar formation in the infarcted mouse heart. J Clin Invest, 2018. 128(5): p. 2127–2143. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28.Jugdutt BI, Ventricular remodeling after infarction and the extracellular collagen matrix: when is enough enough? Circulation, 2003. 108(11): p. 1395–403. [DOI] [PubMed] [Google Scholar]
  • 29.Trueblood NA, et al. , Exaggerated left ventricular dilation and reduced collagen deposition after myocardial infarction in mice lacking osteopontin. Circ Res, 2001. 88(10): p. 1080–7. [DOI] [PubMed] [Google Scholar]
  • 30.Shinde AV and Frangogiannis NG, Fibroblasts in myocardial infarction: a role in inflammation and repair. J Mol Cell Cardiol, 2014. 70: p. 74–82. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31.Lindner D, et al. , Cardiac fibroblasts support cardiac inflammation in heart failure. Basic Res Cardiol, 2014. 109(5): p. 428. [DOI] [PubMed] [Google Scholar]
  • 32.Copaja M, et al. , Simvastatin induces apoptosis by a Rho-dependent mechanism in cultured cardiac fibroblasts and myofibroblasts. Toxicol Appl Pharmacol, 2011. 255(1): p. 57–64. [DOI] [PubMed] [Google Scholar]
  • 33.Nishiga M, et al. , MicroRNA-33 Controls Adaptive Fibrotic Response in the Remodeling Heart by Preserving Lipid Raft Cholesterol. Circ Res, 2017. 120(5): p. 835–847. [DOI] [PubMed] [Google Scholar]
  • 34.Balasubramanian S, et al. , beta3 integrin in cardiac fibroblast is critical for extracellular matrix accumulation during pressure overload hypertrophy in mouse. PLoS One, 2012. 7(9): p. e45076. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 35.Olson ER, et al. , Inhibition of cardiac fibroblast proliferation and myofibroblast differentiation by resveratrol. Am J Physiol Heart Circ Physiol, 2005. 288(3): p. H1131–8. [DOI] [PubMed] [Google Scholar]
  • 36.Cao W, Shi P, and Ge JJ, miR-21 enhances cardiac fibrotic remodeling and fibroblast proliferation via CADM1/STAT3 pathway. BMC Cardiovasc Disord, 2017. 17(1): p. 88. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 37.Russell JL, et al. , A dynamic notch injury response activates epicardium and contributes to fibrosis repair. Circ Res, 2011. 108(1): p. 51–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 38.Mouton AJ, et al. , Exposure to chronic alcohol accelerates development of wall stress and eccentric remodeling in rats with volume overload. J Mol Cell Cardiol, 2016. 97: p. 15–23. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 39.El Hajj EC, et al. , Cardioprotective effects of lysyl oxidase inhibition against volume overload-induced extracellular matrix remodeling. Exp Biol Med (Maywood), 2016. 241(5): p. 539–49. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 40.Hermans KC, Daskalopoulos EP, and Blankesteijn WM, Interventions in Wnt signaling as a novel therapeutic approach to improve myocardial infarct healing. Fibrogenesis Tissue Repair, 2012. 5(1): p. 16. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 41.Saraswati S, et al. , Pyrvinium, a potent small molecule Wnt inhibitor, promotes wound repair and post-MI cardiac remodeling. PLoS One, 2010. 5(11): p. e15521. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 42.Gorski DJ, et al. , Cardiac fibroblast activation and hyaluronan synthesis in response to hyperglycemia and diet-induced insulin resistance. Sci Rep, 2019. 9(1): p. 1827. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 43.Matilla L, et al. , Soluble ST2 promotes oxidative stress and inflammation in cardiac fibroblasts: an in vitro and in vivo study in aortic stenosis. Clin Sci (Lond), 2019. 133(14): p. 1537–1548. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

1
NIHMS1708170-supplement-1.docx (2.1MB, docx)

RESOURCES