Cardiac ATP production and contractility are favorably regulated by short-term S100A9 blockade after myocardial infarction

Graphical abstract

Highlights

• •S100A9 blockade during the first 48 h post-MI attenuates the early metabolic changes by efficient ATP production.
• •S100A9 blockade positively modulates the mitochondrial oxidative phosphorylation post-MI.
• •S100A9 blockade improves cardioprotection post-MI by higher availability of ATP for ventricular contraction.

Abstract

Introduction

The infarcted heart is energetically compromised exhibiting a deficient production of adenosine triphosphate (ATP) and the ensuing impaired contractile function. Short-term blockade of the protein S100A9 improves cardiac performance in mice after myocardial infarction (MI). The implications upon ATP production during this process are not known.

Objectives

This study evaluates whether S100A9 blockade effects ATP synthesis and cardiac contractility in C57BL/6 mice at seven days post-MI.

Methods

Three experimental groups were used: (i) mice with MI, induced by permanent left coronary ligation, (ii) mice with MI, short-term treated with the S100A9 blocker ABR-238901, and (iii) sham (control) mice. After removing the left ventricle, mass spectrometry, pathway enrichment analysis, Western blot, RT-PCR and pharmacological network analysis were performed.

Results

A number of 600 differentially abundant proteins (DAPs) was significantly altered by the S100A9 blocker in MI-treated mice compared with MI mice. Some of these proteins were associated with oxidative phosphorylation, citrate cycle (TCA), mitochondrial fatty acid beta-oxidation, glycolysis and cardiac muscle contraction pathways. In the ischemic ventricle, ABR-238901 treatment increased (1.8- to 38-fold) the abundance of proteins NDUFAB1, UQCRC1, HADHA, ACAA2, ALDOA, PKM1, DLD, DLAT, PDHX, ACO2, IDH3A, FH1, CKM, CKMT2, TNNC1, crucial for early cellular metabolic changes, ATP distribution and contractility. The cardiac level of ATP increased (1.8-fold, p < 0.05) in MI mice treated with ABR-238901 compared to MI mice. The network pharmacology analysis uncovered potential pharmacologic targets of ABR-238901 that may interact with DAPs related to ATP production and contractility.

Conclusion

Short-term S100A9 blockade effectively regulates the proteins implicated in ATP production and cardiac contractility post-MI, providing a framework for future cardiac energy metabolism studies.

Introduction

The heart produces large amounts of adenosine triphosphate (ATP) that sustains the contractile function [1]. The continuous production of ATP in the heart is achieved by metabolizing a variety of substrates, including fatty acids, glucose, lactate, ketones, pyruvate, and amino acids, primarily by mitochondrial oxidative phosphorylation [1]. After substrate oxidation, reducing equivalents generated by the tricarboxylic acid (TCA) cycle are used by the respiratory chain to generate a mitochondrial transmembrane potential that drives ATP synthesis [2]. The normal adult heart relies mostly on fatty acids to fuel oxidative phosphorylation, while 10 % to 30 % of total ATP derives from glucose [2]. The energy production in the heart can be compromised with catastrophic consequences on cardiac function when the energy metabolic pathways that produce ATP are disturbed [1].

Many cardiac diseases are linked with maladaptive changes in energy metabolism worsening the disease progression [3]. Pathological hypertrophy and ischemic heart disease are associated with reduced contractile function in parallel with a shift in energy substrate preference from fatty acids to glucose and the gradual decline of mitochondrial oxidative phosphorylation, mitochondrial TCA cycle activity, and overall oxidative metabolism [3], [4]. Previous clinical studies showed increased alarmin heterodimer S100A8/A9 levels in the infarcted myocardium and the blood of patients with acute myocardial infarction (MI) [5], [6]. S100A8 and S100A9 are Ca²⁺ binding proteins belonging to the S100 family that form homodimers, heterodimers, and tetramers [7]. Under physiological conditions, the S100A8/A9 heterodimer complexes are the most stable and predominant forms of these proteins, exhibiting both intracellular and extracellular functions [8]. The pathophysiological implications of S100A8/A9 proteins in the injured heart are complex. In MI murine models, the release of S100A8/A9 proteins from neutrophils and macrophages promotes cardiomyocyte death and mitochondrial dysfunction by Toll-like receptor 4 in ischemic myocardial tissues [9]. Previous preclinical studies targeting the binding between S100A9 and its receptors during the acute inflammatory phase of MI demonstrated improved cardiac function [10], [11]. We have recently shown that the short-term pharmacological S100A9 blockade preserves myocardial structural proteins, prevents compensatory hypertrophy, and reduces cardiac markers of post-ischemic stress and pressure overload in mice with MI [12].

Lately, metabolomic and proteomic analyses revealed that energy metabolism, amino acid metabolism, vascular smooth muscle contraction, gap junction, cytoskeleton reorganization and neuroactive ligand-receptor interaction are influenced by the myocardial injury after MI [13], [14], [15].

In this study, we questioned whether the short-term blockade of S100A9 could improve ATP synthesis and cardiac contractility at 7 days post-MI. The results identified the proteomic response of the infarcted left ventricle following the short-term early inhibition of the S100A9 protein. The S100A9 blockade favorably modulates the abundance of various proteins involved in the cellular metabolic pathways related to oxidative phosphorylation, TCA cycle, glycolysis and fatty acids beta-oxidation. Our data have further depicted significant differences in the proteomics of the left ventricular contraction under S100A9 inhibition post-MI, reflecting the positive effects of S100A9 protein inhibition. These results correlated well with the increased level of ATP detected in treated compared to untreated MI mice. In addition, potential pharmacological targets of ABR-238901 that may interact with proteins related to ATP production and cardiomyocytes contractility have been exposed.

Materials and methods

Reagents

All reagents for liquid chromatography (LC) and mass spectrometry (MS) analyses were of LC-MS grade and purchased from Merck (Darmstadt, Germany). Trypsin Gold was purchased from Promega (Madison, WI, USA). Millipore Protease Inhibitor Cocktail Set I was purchased from Merck (Darmstadt, Germany). Invitrogen antibodies cytochrome b-c1 complex subunit 1 mitochondrial (PA5-82037), alpha tubulin (#62204), Goat Anti-Rabbit IgG-peroxidase antibody (#31460), Pierce BCA Protein Assay Kit, and Pierce ECL chemiluminescence kit were purchased from Thermo Scientific (Rockford, IL, USA). Cytochrome c oxidase subunit 4 isoform 1 (#4850), NADH-ubiquinone oxidoreductase 75 kDa subunit, mitochondrial (#60153) were purchased from Cell Signaling Technology (Danvers, Massachusetts, USA). Goat Anti-Mouse IgG (Fab specific)-peroxidase antibody (A2304) were bought from Sigma (Saint Louis, Missouri, USA). ATP Assay Kit was purchased from Abbexa (Cambridge, UK). The S100A8/A9 inhibitor (ABR-238901) was a gift from Active Biotech AB (Lund, Sweden).

Ethics statement

All animal experiments were approved by the Ethics Committee of ICBP “N. Simionescu” (Approval no. 425/22.10.2018) and conducted in accordance with the National Institutes of Health guide for the care and use of Laboratory animals (NIH Publications No. 8023, revised 1978), EU Directive 2010/63/EU for animal experiments and Romanian Law no. 471/2002.

Animal model

Myocardial infarction (MI) was initially induced in twenty male and female C57BL/6 mice (8–12 weeks, 19–25 g) by permanent left coronary artery ligation procedure, as previously described [16]. The detailed experimental protocol is provided in the Supplementary material. Infarction was confirmed through echocardiography examination, which revealed that animals exhibited the left ventricular ejection fraction ≤ 40 % calculated on parasternal long-axis (pLAX) view. The animals with left ventricular ejection fraction > 40 % were excluded from the study.

Experimental design

Animals were divided into 3 groups: (i) sham operated mice (Sham, n = 4); (ii) mice subjected to MI intraperitoneally treated with phosphate-buffered saline (PBS) (n = 4); (iii) mice with MI and treated with the S100A9 inhibitor ABR-238901 (30 mg/kg) diluted in PBS (MI + ABR, n = 5). Both treatments were administered at the time of the intervention, at 24 and 48 h after MI. After seven days all animals were euthanized under general anesthesia with ketamine-xylazine (100/20 mg/kg body weight). The hearts were excised and washed in cold PBS containing a protease inhibitor cocktail. The left ventricle (LV) was detached, snap-frozen in liquid nitrogen and preserved at –80 °C for further analyses.

RNA and protein extraction

Total RNA and proteins were extracted from the LV using TRIzol Reagent (Sigma-Aldrich, MO, USA) according to the manufacturer’s protocol. Briefly, the LV was homogenized in TRIzol Reagent using the Polytron PT 1300 D homogenizer (Kinematica, Lucerne, Switzerland), incubated with chloroform and centrifuged. The upper aqueous phase of each sample was used for RNA precipitation and the lower organic one was utilized for protein precipitation. NanoDrop Lite spectrophotometer (Thermo Scientific) was used to assess the purity and concentration of total RNA. The protein pellet was solubilized in a buffer containing 8 M urea, 1 % sodium deoxycholate (DOC), 100 mM Tris-HCl (pH 7.5) and a protease inhibitor cocktail. Protein concentrations were determined by the BCA Protein Assay Kit.

LC-MS/MS analysis

An aliquot of 50 µg-protein from each sample was purified by acetone precipitation, reduced with DTT (20 mM) in a buffer containing 8 M urea, 0.1 M Tris-HCl (pH 8.8) and 0.1 mM EDTA, and alkylated using 80 mM iodoacetamide in 0.1 M Tris-HCl and 0.1 mM EDTA buffer, in the dark, under agitation. Samples were proteolyzed overnight at 37⁰C using Trypsin Gold (1:20 w/w). Formic acid was added to the resulting peptide mixtures up to a pH of 2.5, in order to inhibit the enzymatic activity and promote DOC precipitation. Next, the digested peptides were desalted on Sep-Pak C18 columns (Waters Corporation, Milford, MA, USA) and concentrated with the Concentrator plus system (Eppendorf, Hamburg, Germany). To minimize technical variability, three replicates of each sample were separated and analyzed using the Easy nLC II liquid chromatograph (Thermo Fisher Scientific, Waltham, MA, USA) coupled to the LTQ-Velos Orbitrap mass spectrometer (Thermo Fisher Scientific). Firstly, the peptides were loaded onto a trap column (Thermo Scientific Easy Column − 2 cm length, 100 μm inner diameter, 5 μm particle size, 100 Å pore size). Analytical separation was achieved using the 10 cm EASY column (Thermo Scientific − 75 μm inner diameter, 3 μm particle size, 100 Å pore size), employing a linear gradient from 3 % to 25 % solvent B (0.1 % formic acid in acetonitrile, while solvent A was 0.1 % formic acid in water) at 300 nL/min over 90  min. The mass spectrometer was operated using a top 12 data-dependent configuration at 60 k resolving power for a full scan across the 350–1700 m/z domain, while precursor fragmentation was performed by collision-induced dissociation (CID).

Protein inference and bioinformatic analysis

The resulted MS raw data were processed by Proteome Discoverer V2.4 (Thermo Fisher Scientific), using the UniProtKB/Swiss-Prot Mus musculus reference protein database for protein identification. Search parameters included: enzyme: trypsin; maximum miss cleavages: 2; fixed modification: cysteine carbamidomethylation; variable modifications: oxidation of methionine and deamidation of asparagine and glutamine. The criteria for protein identification also contained the detection of at least 2 unique peptide per protein and the protein false discovery rate (FDR) target to be below 0.05. Label-free relative protein quantitation on the precursor level was performed using the same software platform, which aligns chromatograms, extracts ion peaks and integrates them to compare peptide/protein spectral abundance along the different datasets. We enabled 30 % of replicate features and the ANOVA hypothesis test to determine the statistical significance associated with the protein ratio calculation. Spectral normalization was achieved using the total peptide amount on controls average.

The differentially abundant proteins (DAPs) were identified with an abundance ratio ≥ 1.5 or ≤ 0.667 and an adjusted P-value ≤ 0.05, based on the Benjamini-Hochberg algorithm, was set as statistical cut-off criterion. The volcano plot showing differential abundances of proteins was generated by GraphPad Prism software version 8 for Windows (San Diego, CA, USA). Cytoscape software version 3.9 [17], with STRING (Search Tool for Retrieval of Interacting Genes/Proteins) [18] as plugin, was used to construct the protein–protein interactions (PPI) network and classify DAPs from the left ventricle according to their enriched Gene Ontology Biological Process (GO BP), Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways, Reactome Pathways and WikiPathways. P < 0.05 was considered to indicate a statistically significant difference. The redundancy cut-off parameter of the STRING enrichment analysis was set at 0.5 and the redundant terms were removed. PPI networks were built using Mus musculus as species and a minimum required confidence score ≥ 0.4. Network analyzer, a plugin for Cytoscape, was used to identify influential nodes (hubs or bottlenecks) based on degree values and betweenness centrality [19]. Multivariate statistical Principal Component Analysis (PCA) was employed based on the abundance of proteins involved in the investigated pathways and performed using Proteome Discoverer 2.4.

In this paper, we use the nomenclature of the universal protein knowledgebase (UniProtKB) for the organism Mus musculus. Throughout the paper, the protein symbols used are the same as the gene symbol, are not italicized, and all are in the upper case.

RT-PCR

Eighty-four genes involved in mouse mitochondrial energy metabolism were analyzed with RT² Profiler™ PCR Array (PAMM-008Z), a pre-designed array from Qiagen Sciences (Maryland, USA). Reverse transcription and real-time PCR were performed using RT² First Strand Kit (Qiagen) and RT² SYBR Green qPCR Mastermixes (Qiagen), respectively. According to the manufacturer’s recommendations, the cycling conditions were adjusted for LightCycler 480 System (Roche, Mannheim, DE). For data analysis, Ct values were generated using the “Second Derivative Maximum” analysis method of LightCycler 480 Software 1.5 (Roche, Mannheim, DE). Data normalization was carried out against beta-actin (Actb), beta-2-microglobulin (B2m), and heat shock protein 90 alpha family class B member 1 (Hsp90ab1). Gene expression profiles and statistical analysis (two-tailed Student’s t-test with the significance threshold set to P < 0.05) were evaluated by RT2 Profiler PCR Array Data Analysis Spreadsheet 1904, version5.1 (Qiagen, geneglobe.qiagen.com/us/product-groups/rt2-profiler-pcr-arrays). The fold change in gene expression was calculated using the 2^(− Delta Delta Ct) method.

Western blot analysis

Equal protein amounts (20 µg/ lane) of left ventricle homogenates from MI and MI + ABR animal groups were separated by electrophoresis on 10 % sodium dodecyl sulfate polyacrylamide gels (SDS-PAGE) and blotted onto nitrocellulose membranes. The membranes were blocked with Tris-Buffered Saline (TBS) containing 2 % BSA for 1 h at room temperature and incubated with antibodies (1:1000) against NADH-ubiquinone oxidoreductase 75 kDa subunit, mitochondrial (NDUFS1), cytochrome b-c1 complex subunit 1, mitochondrial (UQCRC1), cytochrome c oxidase subunit 4 isoform 1 (COX4I1), troponin C (TNNIC), troponin I (TNNI3), and alpha tubulin (TUBA1A) at 4 °C overnight, followed by incubation with the appropriate horseradish peroxidase (HRP)-conjugated secondary antibodies (1:4000), at room temperature for 1 h. The resulting immune complexes were detected using the ECL chemiluminescence kit and visualized with the GE Healthcare ImageQuant™ LAS 4000 system. Densitometric analyses were performed using the Image J software (National Institute of Health).

Results

Pathway enrichment analysis based on the differentially abundant proteins in the left ventricle of MI mice and MI mice treated with S100A9 blocker

The LC-MS/MS analysis revealed 1808 proteins, while 958 proteins were confidently identified with a Sequest Score ≥ 10 and ≥ 2 unique peptide matches in the left ventricle samples of Sham, MI and MI + ABR mice. PCA performed on filtered proteins confirmed the homogeneity of biological replicates and the clear separation of the MI from MI + ABR and Sham samples (Fig. 1A). The quantitative analysis revealed 600 DAPs between MI and MI + ABR groups, out of which 411 were down-regulated and 189 were up-regulated (Fig. 1B).

Fig. 1.

Proteomic analysis. (A) Principal component analysis (PCA) score plot based on the LC-MS/MS data of proteins identified in the left ventricle samples of MI mice (gray), MI + ABR mice (dark yellow), and sham operated mice (green) with a Sequest Score ≥ 10 and unique peptide matches ≥ 2. An ellipse highlights the data points coming from one group samples, which were run in three technical replicates (dots) and is added for visual purposes only. (B) The Volcano plot revealed 600 differentially abundant proteins (DAPs) with a 1.5-fold change abundance and an adjusted P value ≤ 0.05 in the MI + ABR group compared with the MI group. Red dots represent the up-regulated proteins, blue dots represent the down-regulated proteins, and grey dots indicate the proteins that were not significantly changed. (C) Bar graph of top 20 statistically enriched KEGG pathways, Reactome Pathways and WikiPathways terms in the LV of MI + ABR vs MI groups across input regulated proteins’ lists, colored by FDR-values. A higher –log10 (FDR-value) indicates increased enrichment. (D) Enriched clusters of all 44 significantly enriched KEGG pathways, Reactome Pathways and WikiPathways terms detected in LV of MI + ABR vs MI presented in a network format. The most statistically significant term within a cluster was chosen to represent the cluster. Each node within a cluster represents one term and colored based on FDR-value. The size of the node correlates with the number of proteins within that term. Hierarchical clustering heatmaps of DAPs associated with oxidative phosphorylation and cardiac muscle contraction pathways between MI, MI + ABR, and Sham groups (n = 4 – 5). The clustering is constructed using the complete-linkage method together with the Euclidean distance. Each column in the heatmap represents a sample, and each row represents the abundance of a DAP. The color scale beside the heatmap represents the raw Z-score ranging from blue (low abundance) to red (high abundance). (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

The pathway enrichment analysis revealed that DAPs were associated with 44 significantly enriched pathways where oxidative phosphorylation (FDR = 1.55E-26) was the most over-represented pathway (Fig. 1C). Among the significantly enriched pathways, cardiac muscle contraction, the mitochondrial fatty acid beta-oxidation and TCA cycle were also associated with the analysed DAPs (Fig. 1C). AutoAnotation of the pathway enrichment analysis revealed 8 clusters, named after the most statistically significant term within a cluster, including oxidative phosphorylation and cardiac muscle contraction (Fig. 1D). The DAPs’ hierarchical clustering analysis revealed a clear separation of MI and MI + ABR groups, while the quantitative fingerprints of Sham and MI + ABR groups were interlaced (Fig. 1D).

Inhibition of S100A9 positively modulates the mitochondrial oxidative phosphorylation post-MI

The oxidative phosphorylation (OXPHOS) system comprises five complexes embedded in the inner mitochondrial membrane, whose main function is to provide energy in the form of ATP [20]. The short-term blockade of S100A9 after MI significantly down-regulated 31 DAPs and up-regulated 10 DAPs associated with the Oxidative phosphorylation pathway (Fig. 2A).

Fig. 2.

Post − myocardial infarction (7 days) components of the oxidative phosphorylation (OXPHOS) are modulated by S100A9 inhibition in the left ventricle. (A) Forty-one out of 600 DAPs achieved through the comparison between the MI + ABR mice and MI mice, were associated with the Oxidative phosphorylation KEGG pathway. Each column represents a single DAP. The up-regulated proteins are represented in red, while the down-regulated proteins are represented in blue. (B) Principal component analysis (PCA) score plot based on the LC-MS/MS data of DAPs involved in the Oxidative phosphorylation KEGG pathway for MI + ABR mice (dark yellow), MI mice (gray), and sham operated mice (green). An ellipse highlights the data points coming from one group samples, which were run in three technical replicates (dots) and is added for visual purposes only. (C) Representative Western blots and semi-quantitative analysis show the protein levels of NDUFS1 (D), UQCRC1 (E) and COX4I1 (F) in MI (n = 3) and MI + ABR (n = 3). Data represent the mean ± SD; *P < 0.05 using Student’s t test. (G) Relative expression of eighty-four genes encoding subunits of mitochondrial complexes. Each column represents −fold change-regulation value of a single gene. The p-values are calculated based on a Student’s t-test. *P < 0.05; MI + ABR (n = 5), MI (n = 4). (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

As shown in Fig. 2A, NADH dehydrogenase [ubiquinone] 1 alpha subcomplex subunit 6 (NDUFA6; 39.1 −fold change), NADH dehydrogenase [ubiquinone] 1 beta subcomplex subunit 4 (NDUFB4; 30.6 −fold change), Cytochrome b-c1 complex subunit 8 (UQCRQ; 102.8 −fold change), Cytochrome c oxidase subunit 6C (COX6C; 19.3 −fold change), ATP synthase subunit epsilon (ATP5E; 24.4 −fold change) and V-type proton ATPase subunit E1 (ATP6V1E1; 13.87 −fold change) were the most down-regulated subunits of the mitochondrial complexes. Notably, these DAPs were the most up-regulated when the MI group was compared to the Sham group (Table 1 and Supplementary File 1).

Table 1.

List of proteins linked to oxidative phosphorylation with the largest −fold change difference of abundance between mice with myocardial infarction (MI), mice with MI treated with the specific S100A9 blocker ABR-238901 (MI + ABR) and control (Sham).

Protein Symbol	Accession	Description	fold change abundance ratio
			MI+ABR/ Sham (Adj. P-Value)	MI+ABR/ MI (Adj. P-Value)	MI/ Sham (Adj. P-Value)
UQCRQ	Q9CQ69	Cytochrome b-c1 complex subunit 8	0.278 (6.59E-06)	0.01 (6.86E-12)	28.593 (5.00E-12)
NDUFA6	Q9CQZ5	NADH dehydrogenase [ubiquinone] 1 alpha subcomplex subunit 6	0.352 (7.83E-04)	0.026 (6.86E-12)	13.792 (4.15E-11)
NDUFB4	Q9CQC7	NADH dehydrogenase [ubiquinone] 1 beta subcomplex subunit 4	0.621 (0.004)	0.033 (6.86E-12)	19.021 (5.00E-12)
ATP5E	P56382	ATP synthase subunit epsilon, mitochondrial	0.578 (0.109)	0.041 (6.86E-12)	14.127 (5.00E-12)
COX6C	Q9CPQ1	Cytochrome c oxidase subunit 6C	0.353 (7.15E-05)	0.052 (4.15E-14)	6.834 (2.51E-10)
ATP6V1E1	P50518	V-type proton ATPase subunit E 1	3.5 (1.69E-08)	0.072 (6.86E-12)	48.557 (5.00E-12)
PPA2	Q91VM9	Inorganic pyrophosphatase 2, mitochondrial	0.911 (0.257)	16.478 (6.86E-12)	0.055 (5.00E-12)
NDUFAB1	Q9CR21	Acyl carrier protein, mitochondrial	0.659 (0.004)	38.601 (6.86E-12)	0.017 (5.00E-12)

Among the up-regulated proteins associated with oxidative phosphorylation, acyl carrier protein (NDUFAB1; 38.6 −fold change) and inorganic pyrophosphatase 2 (Ppa2; 16.4 −fold change) were the most up-regulated subunits (Fig. 2A). These proteins were the most down-regulated when the MI group was compared to Sham (Table 1 and Supplementary File 1).

In addition, when the MI + ABR group was compared to the Sham group (Table 1 and Supplementary File 1), UQCRQ (3.59 −fold change), NDUFA6 (2.84 −fold change), NDUFB4 (1.61 −fold change), COX6C (2.83 −fold change) and NDUFAB1 (1.51 −fold change) were significantly down-regulated. The analysis also revealed that ATP6V1E1 (3.5 −fold change) was significantly up-regulated, whereas ATP5E and PPA2 were not significantly regulated.

PCA analysis displayed the complete separation of the MI group from the MI + ABR group (Fig. 2B) based on DAPs involved in the oxidative phosphorylation pathway. However, the identified OXPHOS proteomes of MI + ABR and Sham groups overlap. Thus, the PCA analysis showed a specific DAPs pattern that could be used to distinguish between the experimental groups entirely and confirmed the positive response of proteins under the inhibition of S100A9 alarmin.

Western blot assays confirmed the up-regulation of NDUFS1, UQCRC1 and down-regulation of COX4 in ABR + MI vs MI (Fig. 2C-F).

Subsequently, the left ventricle tissue samples from day 7 post-MI were investigated to detect whether the S100A9 blockade may affect the expression of genes involved in OXPHOS. The analysis showed the significant up-regulation of cytochrome c oxidase subunit 4 isoform 2, mitochondrial (Cox4I2; 1.62 −fold change) gene expression in MI + ABR compared to MI groups (Fig. 2G and Supplementary File 2).

These data revealed a consistent and enhanced effect of the S100A9 blockade over the protein profile of OXPHOS complexes towards higher energy production in the infarcted ventricle.

Regulation of TCA cycle by inhibition of S100A9 under ischemic condition

The TCA cycle provides reducing equivalents for mitochondrial OXPHOS [21]. Based on the pathway enrichment analysis, which highlighted the association of DAPs with the TCA cycle, we further examined the specific components of the TCA cycle regulated by S100A9 inhibition.

Four proteins, aconitate hydratase, mitochondrial (ACO2, 8.49 −fold change), isocitrate dehydrogenase [NAD] subunit alpha, mitochondrial (IDH3A, 3.75 −fold change), dihydrolipoyl dehydrogenase (DLD, 3.3 −fold change) and fumarate hydratase, mitochondrial (FH1, 1.88 −fold change) were significantly up-regulated while isocitrate dehydrogenase [NAD] subunit gamma 1, mitochondrial (IDH3G, 8.13 −fold change) was significantly down-regulated in the LV tissues of the MI + ABR group compared to the MI group (Fig. 3 and Supplementary File 1).

Fig. 3.

Cardiac energy metabolism following inhibition of S100A9 post-MI. (A) The differentially abundant proteins (DAPs) involved in the TCA cycle, fatty acids oxidation, pyruvate oxidation and glycolysis regulated by S100A9 blockade in the infarcted left ventricle at seven days post-MI. The red color represents up-regulated proteins, the blue color, the down-regulated proteins, and the green color represents proteins whose abundance does not change significantly (an adjusted p-value > 0.05 or a −fold change value lower than ± 1.5) when the MI + ABR group was compared to the MI group. The white color indicates the unidentified proteins. (B) Hierarchical clustering heatmap of DAPs associated with oxidative phosphorylation, TCA cycle, fatty acids oxidation, pyruvate oxidation, and glycolysis between MI (n = 4), MI + ABR (n = 5), and Sham (n = 4) groups. The clustering is constructed using the complete-linkage method together with the Euclidean distance. Each column in the heatmap represents a sample, and each row represents the abundance of a DAP. The color scale beside the heatmap represents the raw Z-score ranging from blue (low abundance) to red (high abundance). (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

The abundance of these proteins was similar when the MI + ABR mice were compared to Sham mice, except IDH3A which was significantly (∼1.33 −fold change) down-regulated (Supplementary File 1).

These results show that the inhibition of S100A9 post-MI may prevent the impairment in the activity of the TCA cycle.

S100A9 blockade preserves unaltered the post-MI mitochondrial oxidation of fatty acids

In aerobic conditions, fatty acid oxidation is the biggest contributor to ATP production and decreases in heart failure associated with ischemia [1], [22]. In our experimental conditions, we found two enzymes involved in fatty acids oxidation, trifunctional enzyme subunit alpha, mitochondrial (HADHA, 3.39 −fold change) and 3-ketoacyl-CoA thiolase, mitochondrial (ACAA2, 2.54 −fold change) displaying a significantly higher abundance in the LV of the MI + ABR mice compared to MI mice (Fig. 3 and Table 2).

Table 2.

Abundance variations of proteins involved in the oxidation of fatty acids in the left ventricle of mice with myocardial infarction (MI), mice with MI treated with the specific S100A9 blocker ABR-238901 (MI + ABR) and control (Sham) on day 7 post-MI.

Protein Symbol	Accession	Description	fold change abundance ratio
			MI+ABR/ Sham (Adj. P-Value)	MI+ABR/ MI (Adj. P-Value)	MI/ Sham (Adj. P-Value)
ACAA2	Q8BWT1	3-ketoacyl-CoA thiolase, mitochondrial	0.761 (0.002)	2.548 (1.01E-6)	0.299 (2.38E-11)
HADHA	Q8BMS1	Trifunctional enzyme subunit alpha, mitochondrial	1.212 (0.999)	3.391 (0.02)	0.357 (0.006)
DECR1	Q9CQ62	2,4-dienoyl-CoA reductase, mitochondrial	0.775 (0.821)	0.513 (0.002)	1.512 (0.048)
ECI1	P42125	Enoyl-CoA delta isomerase 1, mitochondrial	0.619 (1E-04)	0.215 (6.85E-12)	2.88 (1.54E-11)
ECI2	Q9WUR2	Enoyl-CoA delta isomerase 2, mitochondrial	0.789 (0.065)	0.387 (1.66E-06)	2.04 (0.009)

When the MI group was compared to the Sham group, HADHA (2.8 −fold change) and ACAA2 (3.34 −fold change) exhibited significantly lower abundance. Notably, additional enzymes required for unsaturated fatty acids oxidation such as 2,4-dienoyl-CoA reductase, mitochondrial (DECR1, 1.94 −fold change), enoyl-CoA delta isomerase 1, mitochondrial (ECI1, 4.65 −fold change), and enoyl-CoA delta isomerase 2, mitochondrial (ECI2, 2.58 −fold change) were significantly down-regulated in the MI + ABR group compared to the MI group (Table 2). In contrast, these enzymes were significantly up-regulated in the MI group compared to Sham group (Table 2). The analyzed enzymes presented lower (ACAA2, ECI1) or unaltered (HADHA, DECR1, ECI2) abundances in the LV of the MI + ABR group compared to the Sham group (Table 2).

The abundance of glycolytic enzymes is improved after blocking the S100A9 protein

Given that reduced oxygen and substrate availability during myocardial ischemia diminishes mitochondrial oxidative processes and accelerates anaerobic glycolysis [22], we examined the impact of S100A9 inhibition on the level of glycolytic enzymes in the infarcted ventricle. Fructose-bisphosphate aldolase A (ALDOA, 1.88 −fold change) and isoform M1 of pyruvate kinase (PKM1, 8.42 −fold change), enzymes involved in glucose oxidation, were present at significantly higher abundances in the MI + ABR group compared to the MI group (Fig. 3 and Table 3). Instead, their abundances remained nearly unchanged when the MI + ABR and Sham groups were compared (Table 3). Other two enzymes that participate in glucose oxidation, phosphoglycerate mutase 1 (PGAM1, 6.25 −fold change) and phosphoglycerate mutase 2 (PGAM2, 6.84 −fold change) were significantly down-regulated in the MI + ABR group compared to the MI group (Fig. 3 and Table 3). When the MI group was compared with the Sham group, PGAM1 (11.74 −fold change) and PGAM2 (4.34 −fold change) were significantly up-regulated (Table 3).

Table 3.

Abundance ratios of proteins involved in the glucose oxidation in the left ventricle of mice with myocardial infarction (MI), mice with MI treated with the specific S100A9 blocker ABR-238901(MI + ABR) and control (Sham) on day 7 post-MI.

Protein Symbol	Accession	Description	fold change abundance ratio
			MI+ABR/ Sham (Adj. P-Value)	MI+ABR/ MI (Adj. P-Value)	MI/ Sham (Adj. P-Value)
PKM1	P52480-2	Isoform M1 of Pyruvate kinase PKM	0.89 (0.953)	8.429 (0.001)	0.106 (2E-04)
ALDOA	P05064	fructose-bisphosphate aldolase A	0.944 (0.491)	1.886 (0.008)	0.5 (1E-04)
PGAM1	Q9DBJ1	Phosphoglycerate mutase 1	1.878 (0.001)	0.16 (6.85E-12)	11.743 (5E-04)
PGAM2	O70250	Phosphoglycerate mutase 2	0.635 (0.006)	0.146 (6.85E-12)	4.345 (5E-04)
DLD	O08749	Dihydrolipoyl dehydrogenase, mitochondrial	1.01 (0.57)	3.305 (0.019)	0.306 (5E-04)
DLAT	Q8BMF4	Dihydrolipoyllysine-residue acetyltransferase component of pyruvate dehydrogenase complex, mitochondrial	0.862 (0.217)	15.204 (1.08E-09)	0.057 (1.31E-11)
PDHX	Q8BKZ9	Pyruvate dehydrogenase protein X component, mitochondrial	0.675 (0.055)	4.713 (2.41E-09)	0.143 (7.55E-04)

Pyruvate oxidation is catalyzed by pyruvate dehydrogenase complex (PDC), which occupies a key position in the oxidation of glucose by connecting glycolysis to the TCA cycle [23]. We detected 3 enzymes from the PDC, dihydrolipoyl dehydrogenase, mitochondrial (DLD, 3.3 −fold change), dihydrolipoyllysine-residue acetyltransferase component of pyruvate dehydrogenase complex (DLAT, 15.2 −fold change) mitochondrial, and pyruvate dehydrogenase protein X component, mitochondrial (PDHX, 4.7 −fold change) with significantly increased abundances in the ABR + MI group compared to the MI group (Fig. 3 and Table 3). In contrast, there were no significant changes in the abundance of DLD, DLAT and PDHX when the MI + ABR group was compared to the Sham group (Table 3). These data suggest that the temporary pharmacological inhibition of S100A9 enhances glycolysis in the infarcted ventricle. Additionally, the heatmap analysis depicts that DAPs associated with oxidative phosphorylation, TCA cycle, fatty acids oxidation, pyruvate oxidation, and glycolysis could be used to precisely distinguish MI from MI + ABR and Sham samples, the later ones sharing a similar expression pattern (Fig. 3B).

Effect of short-term inhibition of S100A9 on left ventricular contraction post-MI

Because metabolic changes leading to inadequate production of ATP can impair contractile function, we have further examined DAPs associated with the cardiac muscle contraction (CMC) pathway (FDR = 1.02E-11).

The comparison between the MI and MI + ABR groups revealed 21 CMC − related DAPs, out of which 7 were significantly up-regulated and 14 were significantly down-regulated (Fig. 4A and Supplementary File 1). According to the GO analysis for cellular component, CMC − related DAPs were mainly associated with cytochrome complex, sarcomere, and cation-transporting ATPase complex enriched categories (Fig. 4A). The abundance of sarcomere proteins TNNC1 and TNNI3 were validated by Western blot (Fig. 4B).

Fig. 4.

(A) Differentially abundant proteins (DAPs) associated with cardiac muscle contraction (CMC) pathway of MI + ABR relative to MI mice, as depicted in their network by STRING and visualized by Cytoscape software. In the network, the nodes correspond to DAPs up (red)- and down (blue) − regulated and the edges represent the interactions. The size of a node indicates the node degree; a smaller size denotes a lower degree, whereas a larger size denotes a higher degree. (B) Representative Western blots (top) and semi-quantitative analysis (graph, bottom) show the protein levels of TNNC1 and TNNI3 in MI and MI + ABR groups (n = 3 per group; mean ± SD; *P < 0.05 using Student’s t test). (C) PCA of DAPs associated with CMC pathway demonstrating the essential differences between MI, MI + ABR and Sham samples. (D) Cardiac concentration of ATP in the MI (n = 3), MI + ABR (n = 4), and Sham (n = 3) groups; mean ± SD; *P < 0.05 using Student’s t test. (E) The network between potential targets of ABR-238901 (pink diamond) and DAPs related to ATP production and contractility (green circle). (F) Biological process enrichment analysis of potential targets of ABR-238901 that connects DAPs related to ATP production and contractility based on FDR value. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

The PCA of DAPs associated with the CMC pathway confirmed the distinction between MI and MI + ABR/ Sham samples, while the MI + ABR and Sham samples were only partially separated (Fig. 4C).

The PPI network build from CMC − related DAPs revealed the connected components (21 nodes) and 87 interactions (edges). According to the topological analysis of the network (Supplementary File 3), the degree (14 interactions) and the betweenness centrality (BC = 0.5) values of COX6A2 were higher than those of other proteins. These results demonstrate COX6A2 as a hub-bottleneck protein within the constructed protein network, underscoring its pivotal role in the CMC pathway.

The relative gene expression levels of cytochrome complex proteins (Fig. 2G) and several sarcomeric proteins related to CMC pathway were not significantly changed between MI and MI + ABR groups (Fig. S2 in the Supplementary Material).

These data indicate that pharmacological inhibition of S100A9 influences a network of proteins from various cellular compartments, working in concert to enhance cardiac muscle contraction following myocardial infarction.

S100A9 blockade increases ATP level and the abundance of creatine kinase enzymes in the ischemic tissue

The reversible interaction of creatine and ATP under the control of creatine kinase (CK) enzymes is required to facilitate energy storage in the form of phosphocreatine and to sustain the energy transfer from producing sites towards the contractile apparatus [24], [25]. We detected significantly lower abundances for creatine kinase M−type (CKM, 2.9 −fold change) and mitochondrial creatine kinase S-type (CKMT2, 2.7 −fold change) in the MI group compared to the Sham group. Instead, the isoenzymes were found at significantly increased (∼2.2 −fold change) abundances when the MI + ABR group was compared to the MI group (Supplementary File1). In addition, ATP levels in the cardiac tissue of mice from MI + ABR and Sham groups were significantly higher (1.8 −fold change) compared to those of MI group (Fig. 4D).

Potential targets of the ABR-238901 interact with DAPs related to cardiac ATP production and contractility

Next, we performed a pharmacological network analysis to detect other possible molecular mechanisms exerted by ABR-238901 on the analyzed DAPs. According to Swiss Target Prediction, a total of 92 potential targets with a probability larger than 0 were detected. The overlap between all potential targets of ABR-238901 and all 958 identified proteins revealed one common protein, namely cathepsin B, which was upregulated (∼6.6 −fold change) in the ABR + MI group vs MI group (Supplementary File 1). Even though there was no intersection between the potential targets of ABR-238901 and DAPs associated with ATP production and contractility pathways, the STRING analysis disclosed a PPI network based on 60 proteins from the two categories as seen in Fig. 4E. In this network, 28 out 60 nodes were potential targets of ABR-238901, while 32 were DAPs related to ATP production and contractility. The bioinformatic analysis detected that IDH1 and PKM1, DAPs associated with ATP production, connect more than 10 potential targets of ABR-238901. The GO BP enrichment analysis using a redundancy cut-off of 0.2 and an edge cut-off of 0.04 showed that the 28 potential targets of ABR-238901 were associated with 14 pathways, including regulation of MAPK cascade, regulation of generation of precursor metabolites and energy, positive regulation of macrophage proliferation and positive regulation of muscle contraction (Fig. 4F and Fig. S4 in the Supplementary Material).

This pharmacological network with 60 nodes linked via 73 edges revealed the synergistic multicomponent and multitargeted effects of ABR-238901 in the infarcted tissue seven days post-MI.

Discussion

In this study, we conducted a global proteomic analysis of ATP production and contractility in the infarcted left ventricle of mice temporarily treated with the pharmacological inhibitor ABR-238901, commonly used to block the binding of S100A9 to its receptors, Toll-like receptor 4 and Receptor for advanced glycation end products. Additionally, we uncovered potential molecular mechanisms through which ABR-238901 regulates proteins involved in energy metabolism and contraction pathways.

Previous studies showed that ABR-238901 administration during the first 3 days post-MI reduces infiltration of myeloid cells to the myocardium and inflammatory processes, and improves cardiac function in acute MI [11], [26]. Furthermore, a significantly reduced infarcted area and increased mitochondrial complex I activity and its cardiac expression were observed one day post myocardial ischemia/reperfusion (I/R) in mice treated with a S100A9 neutralizing antibody before and after I/R [10]. In contrast, extended S100A9 blockade negatively impacts cardiac recovery [11].

Our results indicate that 41 proteins linked to OXPHOS were significantly modulated by S100A9 inhibition at 7 days post-MI. Half of these molecules were associated with mitochondrial complex I. Mitochondrial complex I protein NDUFAB1 was one of the most up-regulated proteins observed in the infarcted tissue of mice that received the S100A9 inhibitor. Previous data showed that the overexpression of NDUFAB1 has a cardioprotective effect when the heart is subjected to I/R injury [27]. In addition, NDUFAB1 has an essential role in the assembly and stability of complex I and maintaining normal cardiac functions [20]. The overexpression of UQCRC1, a core protein subunit of complex III, was revealed to have a key role in cardioprotection against I/R injury [28], while increased levels of UQCRQ were detected in mouse heart after ischemic cardiomyopathy [29]. Comparable results were detected in our study, further indicating the effects of the S100A9 blockade over mitochondrial complex III in the infarcted tissue. Taken together, the modulation of complex III protein profile by S100A9 blockade, seven days post-MI, may represent a key important event in mitigating early metabolic changes following myocardial infarction. We detected that S100A9 blockade down-regulated the level of all DAPs linked to mitochondrial complex IV in the left ventricle of MI mice, including the hub protein COX6A2. Our analysis revealed that COX6A2 controls information flow between the sarcomere and cytochrome complex proteins. We speculate that the elevated level of COX6A2 observed in the infarcted tissue may be linked to a cellular mechanism aimed at compensating for impaired mitochondrial complex IV activity.

The largely elevated levels of ATP5E detected in the infarcted ventricle, compared to tissue samples from Sham and MI mice treated with the S100A9 blocker, are noteworthy. ATP5E is one of the smallest and least characterized subunits of the complex V enzyme [30]. However, silencing of the ATP5E gene in HEK293 cells resulted in down-regulation of activity and content of the mitochondrial ATP synthase complex, the diminished mitochondrial oxidative phosphorylation function and the accumulation of F0 subunit c [30].

Herein, we provide the first evidence that the extremely high levels of the oxidative phosphorylation − associated proteins, PPA2 and ATP6V1E1, were restored in the infarcted tissues of animals that received the S100A9 inhibitor. Previous studies reported that glucose increases V-ATPase assembly and activity, while alterations in V-ATPase activity influence downstream glycolysis through the hypoxia inducible factor 1 [31].

In a healthy heart, mitochondrial oxidative phosphorylation primarily relies on fatty acid oxidation for ATP production, with glucose metabolism playing a lesser role [3]. In ischemia, reduced oxygen delivery to the heart diminishes cardiomyocyte capacity to break down fatty acids which, in turn, lowers cellular citrate levels and indirectly activates glucose uptake and glycolysis [32]. Increased glycolysis may be accompanied by uncoupling of glucose oxidation and elevation of lactate and proton levels, contributing to loss of contractile force during ischemia [33]. In experimental models, changes in cardiac energy metabolism may vary depending on experimental settings and the severity of ischemia. At 7 days post-MI, we found in the ischemic tissue low levels of several enzymes (HADHA, ACAA2, ALDOA, PKM1, DLD, DLAT, PDHX, ACO2, IDH3A and FH1) involved in fatty acid oxidation, glucose oxidation and TCA cycle. The up-regulations of PGAM1 and PGAM2, two important glycolytic enzymes, in the infarcted left ventricle, are in line with previous studies [34], [35]. In addition, mitochondrial IDH3G is the only TCA cycle enzyme found to be up-regulated following ischemia. Notably, the short-term blockade of S100A9 in the ischemic tissue restored the levels of all these enzymes, highlighting the significant role of S100A9 in the post-ischemic cardiac metabolism.

Published data demonstrated that cardiac contractile dysfunction was abolished in S100a9 KO mice [10]. Here, we have pinpointed the proteins responsible for contractile function improvement, by blocking the interaction of S100A9 with its receptors. In this context, remarkable changes were detected in the abundance of MYL4, ATPA2B, TNNC1 and ASPH proteins. Furthermore, our findings on ATP1B3 and ASPH levels across the experimental groups may hold greater significance for understanding their roles in ventricular contractility after ischemia, as no other experimental data on this topic have been reported to date.

The potential mechanisms of ABR-238901 in MI, beyond blocking the interaction between S100A9 and its receptors, have not yet been elucidated. In this study, we uncovered additional potential molecular mechanisms of action for ABR-23890 through pharmacological network analysis. Among the potential targets of ABR-238901, cathepsin B was the only one identified as a DAP in our study. Endogenous cathepsin B has been reported to affect myocardium injuries, pathological cardiac hypertrophy and remodeling [36], [37], while reparative neutrophils at day 5 and 7 post-MI present an increased protein expresion of cathepsin B [38]. The upregulation of cathepsin B in the infarcted ventricle of ABR-238901- treated mice compared with the MI group at seven days post-MI is intriguing, warranting further research to elucidate the role of cathepsin B in this context.

A potential limitation of the study could be the myocardial infarction (MI) experimental model employed, as it does not replicate the events leading up to MI, such as the progression of atheromatous plaque buildup in coronary arteries to stenosis and complete occlusion. However, the C57BL/6 mouse model with surgical ligation of the left anterior descending coronary artery remains the most efficient and widely used animal model for MI [39].

Conclusion

The inhibition of S100A9 during the first 48 h post-MI has a key role in attenuating the early metabolic changes following myocardial infarction, maintaining efficient ATP production and distribution, increasing contractility, and providing cardioprotection. Our data expand the knowledge of critical players in the recovery of the energy metabolism − contractility axis post-MI. The potential targets of ABR-238901 that may be linked to proteins involved in ATP production and contractility in MI should be carefully analyzed further, taking into account factors such as tissue expression, molecular function, and the biological processes in which they participate. We conducted this study using a more homogenous experimental animal model to generate detailed data and establish a foundation for future preclinical research.

Compliance with ethics requirements

All Institutional and National Guidelines for the care and use of animals (fisheries) were followed.

CRediT authorship contribution statement

Raluca M. Boteanu: Methodology, Validation, Formal analysis, Investigation, Writing – original draft. Viorel I. Suica: Methodology, Data curation. Elena Uyy: Methodology, Formal analysis. Luminita Ivan: Methodology. Diana V. Uta: Formal analysis. Razvan G. Mares: Methodology. Maya Simionescu: Funding acquisition, Conceptualization, Supervision, Writing – review & editing. Alexandru Schiopu: Funding acquisition, Conceptualization. Felicia Antohe: Funding acquisition, Conceptualization, Supervision, Writing – review & editing.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Appendix A. Supplementary material

The following are the Supplementary data to this article: