The Biological Basis for Cardiac Repair After Myocardial
Infarction: From Inflammation to Fibrosis
Sumanth D. Prabhu and Nikolaos G. Frangogiannis
1
Division of Cardiovascular Disease, University of Alabama at Birmingham, and Medical Service,
Birmingham VAMC, Birmingham, AL
2
The Wilf Family Cardiovascular Research Institute, Department of Medicine, Albert Einstein
College of Medicine, Bronx NY
Abstract
In adult mammals, massive sudden loss of cardiomyocytes following infarction overwhelms the
limited regenerative capacity of the myocardium, resulting in formation of a collagen-based scar.
Necrotic cells release danger signals, activating innate immune pathways and triggering an intense
inflammatory response. Stimulation of toll-like receptor signaling and complement activation
induces expression of pro-inflammatory cytokines (such as interleukin-1 and tumor necrosis
factor-α) and chemokines (such as monocyte chemoattractant protein-1/CCL2). Inflammatory
signals promote adhesive interactions between leukocytes and endothelial cells, leading to
extravasation of neutrophils and monocytes. As infiltrating leukocytes clear the infarct from dead
cells, mediators repressing inflammation are released, and anti-inflammatory mononuclear cell
subsets predominate. Suppression of the inflammatory response is associated with activation of
reparative cells. Fibroblasts proliferate, undergo myofibroblast transdifferentiation, and deposit
large amounts of extracellular matrix proteins maintaining the structural integrity of the infarcted
ventricle. The renin-angiotensin-aldosterone system and members of the transforming growth
factor-β family play an important role in activation of infarct myofibroblasts. Maturation of the
scar follows, as a network of cross-linked collagenous matrix is formed and granulation tissue
cells become apoptotic. This review discusses the cellular effectors and molecular signals
regulating the inflammatory and reparative response following myocardial infarction.
Dysregulation of immune pathways, impaired suppression of post-infarction inflammation,
perturbed spatial containment of the inflammatory response, and overactive fibrosis may cause
adverse remodeling in patients with infarction contributing to the pathogenesis of heart failure.
Therapeutic modulation of the inflammatory and reparative response may hold promise for
prevention of post-infarction heart failure.
Keywords
Myocardial Infarction; Inflammation; Fibrosis; Immune Cells; Remodeling
Correspondence: Sumanth D. Prabhu, MD, Professor and Director, Division of Cardiovascular Disease,, University of Alabama at
Birmingham,, 311 Tinsley Harrison Tower,, 1900 University Blvd,, Birmingham, AL-35294-0006, sprabhu@uab.edu; Nikolaos G.
Frangogiannis, MD, The Wilf Family Cardiovascular Research Institute, Department of Medicine, Albert Einstein College of
Medicine, 1300 Morris Park Avenue, Forchheimer G46B, Bronx, NY 10461, nikolaos.frangogiannis@einstein.yu.edu.
DISCLOSURES: None.
HHS Public Access
Author manuscript
Circ Res
. Author manuscript; available in PMC 2017 June 24.
Published in final edited form as:
Circ Res
. 2016 June 24; 119(1): 91–112. doi:10.1161/CIRCRESAHA.116.303577.
Author Manuscript Author Manuscript Author Manuscript Author Manuscript
Subject Terms
Heart Failure
1. Introduction
Adverse left ventricular (LV) remodeling following myocardial infarction (MI) constitutes
the structural basis for ischemic heart failure (HF), and is comprised of complex short- and
long-term changes in LV size, shape, function, and cellular and molecular composition.
1, 2
While multiple pathophysiological factors converge to remodel the heart after MI, the
fundamental determinants of this process (and its progression to clinical HF) are the extent
of the initial infarction and the sufficiency of the post-MI reparative process. In clinical
practice, limiting infarction extent is routinely addressed by timely coronary reperfusion. In
contrast, therapeutic manipulation of the ensuing repair process, which is driven principally
by robust tissue inflammation and subsequently by its active suppression and resolution, has
proved much more challenging and elusive. Nonetheless, recent studies have suggested a
large number of potential therapeutic targets that may favorably influence cardiac wound
healing and repair. In this review, we will broadly consider the multiplicity of cellular and
molecular factors that influence post-MI repair, highlighting the translational implications
for these events in the amelioration of adverse remodeling and the development of ischemic
HF.
2. The phases of cardiac repair after myocardial infarction
Cardiac repair after MI results from a finely orchestrated and complex series of events,
initiated by intense sterile inflammation and immune cell infiltration (inflammatory phase)
that serve to digest and clear damaged cells and extracellular matrix tissue (~3–4 d in mice),
followed by a reparative phase with resolution of inflammation, (myo)fibroblast
proliferation, scar formation, and neovascularization over the next several days (Figure
1).
3, 4
Early inflammatory activation is a necessary event for the transition to later reparative
and proliferative programs. Appropriate and timely containment and resolution of
inflammation are further determinants of the quality of wound healing; a proper physiologic
balance needs to be achieved between these two phases for optimal repair.
5, 6
An
inflammatory phase that is disproportionately prolonged, of excessive magnitude, or
insufficiently suppressed can lead to sustained tissue damage and improper healing,
defective scar formation, and heightened cell loss and contractile dysfunction, thereby
promoting infarct expansion, adverse remodeling and chamber dilatation. To date, there has
been no large-scale immunomodulatory or anti-inflammatory therapeutic strategy post-MI
that has been successfully translated into clinical practice, no doubt a reflection of both the
exquisite complexity and our incomplete understanding of the healing process.
Prabhu and Frangogiannis Page 2
Circ Res
. Author manuscript; available in PMC 2017 June 24.
Author Manuscript Author Manuscript Author Manuscript Author Manuscript
3. The inflammatory phase
3.1. Molecular cascades implicated in the post-infarction inflammatory response
Hypoxia during ischemia impairs vascular endothelial cell integrity and its barrier function
thereby augmenting vessel permeability, facilitating leukocyte infiltration.
7
If the ischemic
period is sufficiently prolonged, parenchymal and cardiomyocyte cell death programs are
activated, primarily due to cell necrosis, but also secondary to apoptosis and autophagic
mechanisms.
3, 7
Restoration of blood flow may further augment tissue damage via
reperfusion injury, due to abrupt re-oxygenation, reactive oxygen species (ROS) generation,
and activation of the complement pathway.
3, 5, 7–9
Necrotic and stressed/injured cells, and
the damaged extracellular matrix, release substances that act as danger signals, termed
danger-associated molecular patterns (DAMPs). DAMPs bind to cognate pattern recognition
receptors (PRRs) of the innate immune system on surviving parenchymal cells and
infiltrating leukocytes (and also activate the complement pathway) to robustly activate a
cascade of inflammatory mediators, including inflammatory cytokines, chemokines, and cell
adhesion molecules.
8, 10–14
In addition to being passively released upon cell necrosis or
matrix damage, select DAMPs may also be upregulated and secreted by stressed
cardiomyocytes and fibroblasts, and by activated leukocytes.
8, 11, 12, 15–17
Several DAMPs can trigger the inflammatory response during MI (Table 1). These including
high mobility group box-1 (HMGB1), S100 proteins, fibronectin extra domain A,
interleukin(IL)-1α, heat shock proteins (HSPs), low molecular weight hyaluronic acid, ATP,
uric acid, mitochondrial DNA, dsRNA, ssRNA, and complement, among others.
7, 8, 11, 12, 14
The PRRs are primarily the membrane bound toll-like receptor/IL-1 receptors (TLR/
IL-1Rs), as well as cytosolic nucleotide-binding oligomerization domain (NOD)-like
receptors (NLRs) and the cell-surface receptor for advanced glycation end-products
(RAGE). The signaling pathways downstream of these PRRs have been comprehensively
detailed in recent reviews
13, 14, 18, 19
and briefly considered below. In the context of MI,
downstream signaling converges on the activation of mitogen-activated protein kinases
(MAPKs) and nuclear factor (NF)-κB. These pathways (NF-κB in particular) drive the
expression of a large panel of pro-inflammatory genes including inflammatory cytokines
(e.g, tumor necrosis factor-α [TNF], IL-1β, IL-6, IL-18); CXC chemokines containing the
glutamic acid-leucine-arginine (ELR) motif that act predominantly as neutrophil
chemoattractants; CC chemokines that attract monocytes and T-lymphocytes; cell adhesion
molecules (e.g., VCAM, ICAM, selectins), and complement factor B.
14, 20, 21
Subsequent
leukocyte recruitment further amplifies the inflammatory response, augments the production
of DAMPs, and promotes both efferocytosis of dying cells and tissue digestion via the
release of proteases and oxidases.
3, 22
Efficient efferocytosis of apoptotic cardiomyocytes is
particularly important for transitioning to the phase of inflammation resolution and wound
healing, and is mediated in part by macrophages expressing the myeloid-epithelial-
reproductive tyrosine kinase (Mertk).
23
3.1.1. DAMPs
HMGB1: HMGB1, a loosely-associated chromatin protein involved in DNA stabilization
and gene control,
24
is a potent mediator of inflammation following tissue injury.
25
HMGB1
Prabhu and Frangogiannis Page 3
Circ Res
. Author manuscript; available in PMC 2017 June 24.
Author Manuscript Author Manuscript Author Manuscript Author Manuscript
is passively released from necrotic cells (but not apoptotic cells),
25
actively secreted by
stimulated monocytes and macrophages,
26
and induced by peroxynitrite and oxidative stress
in ischemic cardiomyocytes.
27
HGMB1 engages and activates several TLRs (including
TLR2, TLR4, TLR9) and RAGE,
8, 15, 28, 29
to induce NF-κB nuclear translocation and pro-
inflammatory signaling. HMGB1 also promotes monocyte recruitment in a TLR- and
RAGE-independent manner via direct binding to CXCL12 (stromal cell-derived factor-1
[SDF-1]) and the formation of HMGB1-CXCL12 heterocomplexes that synergistically
enhance CXCR4 signaling in inflammatory cells.
30
In humans with acute MI, serum HMGB1 levels are elevated and predictive of subsequent
mortality, LV dysfunction, and effort intolerance.
31–33
In rodents with reperfused
15, 34
and
non-reperfused
33
MI, serum levels and myocardial HMGB1 expression increase very early
after injury. In reperfused MI, HMGB1 plays a pivotal role in the activation of MAPK and
NF-κB pathways, increasing leukocyte infiltration, and augmenting tissue injury, apoptosis,
and infarct size, in part via RAGE-dependent signaling.
15, 34
Interestingly and in contrast,
however, in acute and chronic non-reperfused MI models, augmenting HMGB1 via either
exogenous administration
35–37
or cardiomyocyte-specific overexpression
38, 39
improved LV
remodeling and function with augmented c-kit+ progenitor cell infiltration, reduced
dendritic cell accumulation, less collagen deposition, and better tissue angiogenesis.
Exogenous administration of HMGB1 prior to ischemia/reperfusion (I/R) also induces
preconditioning and cardioprotection.
40
Interestingly, antibody-mediated HMGB1
neutralization in non-reperfused MI reduced tissue inflammatory cytokine expression and
macrophage infiltration at day 3, but induced scar thinning and more pronounced LV
remodeling, underscoring the concept that inflammation, properly regulated and controlled,
is essential for optimal wound healing in the heart.
33
Hence, these results indicate that while
HMGB1 is an inflammatory mediator, the ultimate effects of HMGB1 modulation depends
on the underlying disease, its temporal context, and the degree of inflammatory response
referable to the pathophysiology. Compared to non-reperfused MI hearts, I/R hearts exhibit
greater magnitude but shorter duration of inflammatory cell infiltration.
9, 41
This profile of
augmented inflammation may explain the beneficial results with HMGB1 inhibition in
reperfused MI, and the divergence in response from non-reperfused MI in which angiogenic
effects may predominate.
Other DAMPs: S100A8 (calgranulin A) and S100A9 (calgranulin B) are members of the
calcium-binding S100 family expressed in phagocytic cells.
42
S100A8 and A9 rapidly
associate to form the S100A8/A9 heterodimer. S100A8/A9 complexes function as DAMPs
secreted by neutrophils and monocytes/macrophages during inflammatory conditions and
signal via RAGE and TLR4 receptors.
43, 44
Humans with acute MI exhibit elevated serum
S100A8/A9 levels that correlate with circulating neutrophil counts and the risk of
cardiovascular death and subsequent MI.
45, 46
In mice with reperfused MI, S100A8/A9 is
rapidly expressed and released after ischemia, primarily by inflammatory cells and
fibroblasts, and induces pro-inflammatory signaling, leukocyte infiltration, and cardiac
dysfunction in a RAGE-dependent manner,
47
suggesting that these DAMPs are central to
post-MI inflammation. Interestingly, S100A1, the S100 protein most abundant in
cardiomyocytes, is also released from damaged cardiomyocytes in both humans and mice
Prabhu and Frangogiannis Page 4
Circ Res
. Author manuscript; available in PMC 2017 June 24.
Author Manuscript Author Manuscript Author Manuscript Author Manuscript
with acute MI.
48
However, rather than promoting generalized inflammation, S100A1 is
taken up by endocytosis in adjacent cardiac fibroblasts to transiently activate TLR4-
endolysosomal signaling, resulting in an immunomodulatory and anti-fibrotic phenotype
with beneficial effects on post-MI LV remodeling
in vivo
.
48
This suggests that specific
DAMPs have unique cell targets and functional roles in the cardiac repair process that can be
either pro- or anti-inflammatory. In this regard, the β-galactoside binding lectin galectin-1 is
expressed by hypoxic cardiomyocytes and infiltrating leukocytes after MI, and also imparts
anti-inflammatory and cardioprotective effects in the remodeling heart.
49
Fibronectin is an extracellular matrix protein secreted by fibroblasts in response to tissue
injury and pro-inflammatory cytokines,
50
and includes an alternatively spliced exon coding
type III repeat extra domain A (EDA) that binds to TLR-4 to activate mast cells and
leukocytes.
51, 52
Mice with parenchymal myocardium-localized fibronectin-EDA deficiency
exhibited improved LV remodeling and function, less monocyte recruitment, and reduced
remote zone fibrosis after non-reperfused MI as compared with wild-type mice, indicating a
critical role for fibronectin-EDA in tissue inflammation and remodeling.
53
Conversely,
recent studies have demonstrated that necrotic cardiomyocytes (but not fibroblasts) release
IL-1α as danger signal that activates pro-inflammatory MAPK and NF-κB signaling in
cardiac fibroblasts in a MyD88-dependent but NLRP3- and TLR-independent manner, via
activation of the IL-1R pathway.
54
Hence, multiple danger signals act in a concerted fashion
on parenchymal and inflammatory cells in the infarcted heart to drive and/or modulate
inflammation. For a further discussion of DAMPs, the reader is referred to several
comprehensive reviews.
8, 10–12
3.1.2. TLRs, NLRs, and RAGE
TLRs: The TLRs comprise the major PRRs on mammalian cells.
14
Expressed most
prominently on leukocytes, TLRs are also expressed by parenchymal cells, including
cardiomyocytes, fibroblasts, and endothelial cells. Thus far, 13 functional mammalian TLRs
have been identified (10 in humans, TLRs 1–10)
13, 14
that recognize a variety of pathogen-
associated molecular patterns
14
and DAMPs
11, 12
to trigger innate immune responses. Of
these, TLRs 1, 2, 4, 5, 6, and 11 are cell-surface receptors, whereas TLR3 and TLRs 7–10
are expressed in endolysosomes.
13, 14
Signal transduction by TLRs and IL-1Rs occurs
through a conserved cytoplasmic Toll/IL-1R (TIR) domain that serves as the docking site for
TIR-containing cytoplasmic adaptor proteins (Figure 2). Except for TLR3, all TLRs (and
IL-1Rs) engage with the adaptor MyD88 (myeloid differentiation factor 88) either directly,
or, for TLR2 and TLR4, in combination with the adaptor TIRAP (TIR domain-containing
adaptor protein) to trigger receptor complex interactions with IRAKs (IL-1R associated
kinases) 4, 1 and 2, TRAF6 (TNF receptor associated factor 6), and the MAPKKK
transforming growth factor activated kinase (TAK)1. As fully reviewed elsewhere,
13, 14
these
signaling cascades ultimately activate NF-κB (p65 and p50) and MAPK pathways to
upregulate a broad array of pro-inflammatory mediators (Figure 2). TLR4 is also
endocytosed after ligand binding; endolysosomal TLR4 signals in a MyD88-independent
manner via the cytoplasmic adaptor TRIF (TIR domain-containing adaptor inducing
interferon [IFN]-β) and the bridging adaptor TRAM (TRIF-related adaptor molecule). This
pathway results in NF-κB nuclear translocation, and the induction of type I IFN via
Prabhu and Frangogiannis Page 5
Circ Res
. Author manuscript; available in PMC 2017 June 24.
Author Manuscript Author Manuscript Author Manuscript Author Manuscript