As a member of the Neuregulin family of growth factors, NRG-1, encoded by a gene spanning 2.4 million base pairs in mice and 2.6 million base pairs in humans and rats, is located on chromosome 8 in humans and mice, and on chromosome 16 in rats (9,10). The 6 known types of NRG-1 proteins (types I-VI) are classified by 6 different transcriptional initiation sites, and alternative splicing of the NRG-1 gene produces 33 different isoforms in humans (11,12). The 6 types of NRG-1 are defined by differences at the N-terminus; more specifically, these differences lay in the linker region between the transmembrane domain and the EGF-like domain encoded by exons (13,14). All 6 isoforms have an EGF-like domain, which is necessary and sufficient for the activation of ErbB receptors. However, only type I, II, IV and V isoforms have an additional immunoglobulin (Ig)-like domain, which is located between the N-terminal sequence and the EGF-like domain (Fig. 1A), while NRG-1 types III and VI are characterized by an N-terminal region connected directly to the EGF-like domain (Fig. 1B) (10,14). Of note, the N-terminal region of NRG-1 type III consists of a cysteine-rich domain (CRD) and an additional N-terminal transmembrane domain (TM), and this unique structure limits the functional fragment release (Fig. 1C) (14). NRG-1 protein is functionally shed by proteolytic cleavage of its membrane-bound precursor (pro-NRG-1) (15). A bioactive extracellular NRG1 fragment is released by type I transmembrane domain proteases [e.g., beta-secretase (BACE), a disintegrin and metallopeptidase domain (ADAM)17, ADAM19 and ADAM10], termed mature NRG-1, acting in a juxtacrine or paracrine manner, while mature NRG-1 type III is anchored by CRD rather than shedding from the membrane (16).

To exert vital biological functions, NRG-1 must bind to a family of ErbB to activate tyrosine kinase receptor proteins. The ErbB family encompasses four transmembrane tyrosine kinase receptors: ErbB1 [also termed epidermal growth factor receptor (EGFR)], ErbB2, ErbB3 and ErbB4 (10,13). Upon ligand binding, an ErbB receptor encounters structural modifications in the juxta-membrane domain (JMD) (17), which leads to the homodimerization or heterodimerization of ErbB receptors due to differences in receptor affinity. NRG-1 ligand binding triggers the homodimerization of ErbB4 and the heterodimerization of ErbB-2/3, ErbB-2/4 and ErbB-3/4 (18). Compared with NRG-1, EGF has more potent affinity to the binding site of ErbB1 (19). Among the ErbB receptors, only ErbB4 can form functional homodimers upon NRG-1 binding, as it consists of both ligand binding and active kinase domains (15). On the other hand, ErbB2 lacks the ligand binding pocket, while ErbB3 only has a pseudokinase domain. The absence of sufficient signal transmission requires the heterodimerization of both receptors to exert their function (13).

The dimerization of ErbB receptors (either homo- or heterodimerization) activates the tyrosine kinase domain and induces the phosphorylation of the cytoplasmic region of the ErbB partner (20). Subsequently, the phosphorylated tyrosine residues recruit various adaptors/effectors and trigger a switch in signal transmission. The main activated downstream signaling molecules are the phosphoinositide 3-kinase (PI3K)/Akt and Raf/MEK/extracellular regulated kinase (ERK) pathways (21), which are both important reperfusion injury salvage kinases (RISK). Other downstream kinases known to be activated by NRG-1 include Pyk2, c-Abl, JNK, Fyn and CDK5 (10). In the cardiovascular system, these pathways affect cell survival, migration, adhesion and differentiation properties and proliferation (22), and they play an indispensable role in maintaining cardiovascular homeostasis when cardiomyocytes encounter stimuli or insults, such as hypoxia, acidosis and oxidative stress, that contribute to myocardial IR injury.

With the initial activation of inflammation accompanied by the release of pro-inflammatory cytokines, as well as inflammatory leukocyte recruitment and infiltration, the heart attempts to defend itself against deleterious stimuli during IR. However, excessive inflammation in the endangered myocardial region can lead to exacerbated myocyte death via pro-apoptotic signaling pathways and further cardiac remodeling (28,29). Toll-like receptors (TLRs) and nuclear factor-κB (NF-κB) are the primary pathways associated with IR-induced inflammation (3). TLRs are expressed by inflammatory cells, endothelial cells and cardiomyocytes. In addition to defending against microorganisms, TLRs, particularly TLR-4, play a critical role in inflammation-induced apoptosis during IR (30). NF-κB pathways can be divided into canonical and alternative/non-canonical pathways. The canonical NF-κB pathway involves the phosphorylation and degradation of inhibitory κB (IκB), resulting in the nuclear translocation of p65/p50 NF-κB heterodimers and ultimately triggering the production of pro-inflammatory molecules (31). By contrast, the alternative NF-κB pathway promotes the generation of anti-apoptotic and anti-inflammatory molecules through the enhanced translocation of RelB/p52 NF-κB heterodimers by the activation of IKB kinase-α (IKK-α) (32,33).

Studies have demonstrated the anti-inflammatory effects of NRG-1 in the setting of sepsis-induced cardiac injury and ischemic stroke (34,35). NRG-1 alleviates the inflammatory response by both the inhibition of canonical NF-κB and the activation of alternative NF-κB. Consistent with the alleviation of the inflammatory response, the pro-inflammatory factors [e.g., tumor necrosis factor (TNF)-α, interleukin (IL)-6 and IL-1β] are downregulated and anti-inflammatory factors [e.g., granulocyte-colony stimulating factor (G-CSF) and IL-9] are upregulated (35). Furthermore, NRG-1/ErbB4 can inhibit the action of macrophages through the PI3K/AKT pathway, attenuating myocardium inflammation and fibrosis (36). Although there is limited evidence available to indicate that NRG-1 exerts its anti-inflammatory effects during IR to protect the heart, inflammation is a ubiquitous stress-response involving the entire body, and ischemic stroke shares similar features as myocardial infarction. Consequently, it can be inferred that NRG-1 modulates inflammation by regulating the NF-κB pathway and suppressing macrophage activation during IR injury, thus salvaging at-risk myocytes. On the other hand, repressing the maladaptive organelle response attributed to inflammation (e.g., oxidative stress, endoplasmic reticulum stress and calcium overload) may also be an approach to exert anti-inflammatory effects.

ER stress, as a transient adaptive response aimed at reducing the level of unfolded proteins and returning the ER to homeostasis, following the upregulation of molecular chaperones, ultimately increases the ability of the ER to regulate calcium content and protect myocardial cells (37,38). However, prolonged or severe ER stress, attributed to persistent hypoxia, oxidative stress and calcium overload, can abandon its pro-survival efforts and instead trigger apoptotic cell death by activating caspase-12 (an ER stress-specific proapoptotic molecule) and c-JUN N-terminal kinase and upregulating CCAAT/enhancer binding protein homologous protein (CHOP) (39). Increasing evidence has indicated that ER stress plays a crucial role in the pathogenesis of myocardial IR injury (40,41). It has been demonstrated that NRG-1/ErbB protects against cardiac IR injury by suppressing cardiac ER stress through the PI3K/Akt pathway and directly inhibits the upregulation of ER stress-related markers [e.g., glucose-regulated protein (GRP)78, CHOP and cleaved caspase-12] in both neonatal and adult ventricular myocytes to delay the onset of ER stress, reducing caspase-12-related apoptosis (39). On the other hand, NRG-1 also hinders the onset of ER stress by alleviating oxidative stress and calcium overload, which will be discussed in detail below.

The disequilibrium of intracellular Ca²⁺ homeostasis plays a significant role during the pathogenesis of IR injury and contributes to the impaired cardiac contractile function (3). ER stress-induced calcium overload is the upstream signal for IR injury (42). The increased level of calcium activates Ca²⁺-dependent xanthine oxidase (XO), which results in ROS generation and cellular apoptosis (43).

NO, recognized as an inorganic free radical gas, is synthesized from the amino acid L-arginine by nitric oxide synthases (NOSs), such as neuronal NOS (nNOS or NOS I), inducible NOS (iNOS or NOS II), and endothelial NOS (eNOS or NOS III) (44). NO functions as an important biological molecule attenuating myocardial IR injury via the regulation of cardiac contractility and vasodilation (45). Under calcium overload, NRG-1 rapidly enhances the level of NO in adult ventricular myocytes through the activation of the PI3K/Akt/eNOS pathway. The increase in NO generation leads to PKG activation with the phosphorylation of phospholamban, which promotes sarcoplasmic reticulum calcium ATPase (SERCA2a) activity as well as calcium uptake by the sarcoplasmic reticulum (46). It has been shown that NRG-1 regulates calcium through the production of NO. The generation of NO also enhances the open probability of mitochondrial adenosine triphosphate-sensitive K⁺ (mitoKATP) channels, but reduces the mPTP open probability. Notably, accumulated NO may exert a detrimental effect through the formation of peroxynitrite (a byproduct of NO degradation due to decreased NO bioavailability) (47), which aggravates oxidative stress and activates the apoptotic signaling pathway. It has been demonstrated that the hyperactivation of eNOS induced by NRG-1 can lead to an increase in ROS production (48), resulting in an increased MI size. The theory of eNOS uncoupling may explain this phenomenon. Briefly, under substrate (L-arginine) or cofactor deficiency, eNOS would transfer electrons to molecular oxygen rather than to substrate, resulting in the burst of ROS (49). It seems paradoxical due to the dual role of NO; however, its combination with L-arginine can reverse this deleterious effect.

As previously mentioned, a surge of ROS has been shown to stimulate NRG-1 secretion. Moreover, NRG-1/ErbB can defend against oxidative stress during IR. Morano et al revealed that the expression of ErbB3 receptor in H9c2 cells increased the cell survival rate and enhanced mitochondrial resistance to oxidative stress by maintaining mitochondrial membrane potential (25), which inhibits the release of ROS and cytochrome c through the opening of mitoKATP channels. In addition to inhibiting the network of maladaptive organelle responses that aggravate oxidative stress, NRG-1 also induces adaptations of transcription factors involved in redox regulation, including 11 factors [catalase (CAT), Cu-Zn-dismutase (SOD1), thioredoxin (TXN), TXNRD1, protein disulfide-isomerase (PDI)A1, PDIA4, PDIA3, glutathione S-transferase Pi 1 (GSTP1), glutathione peroxidase 1 (GPX1), glutamate-cysteine ligase catalytic subunit (GCLC) and thiosulfate sulfurtransferase (TST)], to exert its radical scavenging effects (50). Furthermore, NRG-1 increases glutathione reductase mRNA, whose translation products are known as important antioxidants (51).

Hypoxia-inducible factor-1 (HIF-1), a heterodimeric transcription factor consisting of a constitutively expressed β subunit and an oxygen-regulated α subunit, regulates angiogenesis, proliferation and cellular metabolism, assisting cells in the adaption to hypoxic environments (52,53). Under anoxic conditions, HIF-1α becomes stable and is accompanied by an upregulated transcriptional response to ROS, inflammatory mediators and somatotropic hormone (54,55). Increased levels of HIF-1α and the enhanced transcriptional activity of key HIF-1 target genes, such as vascular endothelial growth factor (VEGF), play a critical role in myocardial IR injury. The activated HIF-1α/VEGF signaling pathway promotes the proliferation of cardiac microvasculature endothelial cells, which leads to NRG-1 generation and confers cardio-protective effects against anoxia injury during IR (56,57). It has previously been demonstrated that hypoxia can induce NRG-1 secretion and the phosphorylation of ErbB (27), and it has been shown that ErbB3 is upregulated by HIF-1α (58). Thus, it can be inferred that HIF-1α may upregulate NRG-1 via unidentified pathways. Of note, NRG-1 can in turn mediate HIF-1α expression. It has been demonstrated that the PI3K/Akt/mammalian target of rapamycin (mTOR) pathway activated by NRG-1 can lead to HIF-1α activation and can regulate angiogenesis (59). Collectively, the interactions between HIF-1α and NRG-1 mediate angiogenesis by triggering an increase in the proliferation, migration and invasion of endothelial cells, as well as in the formation of a capillary-like tubular structure network. This helps cardiomyocytes to survive hypoxic conditions and under subsequent reperfusion injury.

The eventual and principal target of all therapeutic interventions is aimed at reducing the MI size and salvaging the myocardium. Thus, the prevention of cardiomyocyte death is crucial for cardiac function recovery and the estimation of intervention effects. Studies have demonstrated that the caspase-8-dependent Fas/FasL extrinsic death receptor pathway, caspase-9-related mitochondrial apoptosis and caspase-12-involved ER stress are associated with cardio-myocyte death in response to IR injury (60-62). Notably, NRG-1 has the ability to inhibit cell death by improving mitochondrial membrane potential (63), suppressing calcium overload (64), suppressing endoplasmic reticulum stress (38), alleviating the inflammatory response (35) and ultimately maintaining cellular viability during myocardial IR injury.

Apoptosis is an ATP-consuming form of programmed cell death characterized by chromatin accumulation, DNA fragmentation and apoptotic body formation, typically without an inflammatory response or membrane stability changes (2,65). Apoptosis can be initiated extrinsically by the activation of sarcolemma receptors (e.g., Fas and TNF-α) or intrinsically by the mitochondrial release of cytochrome c, which initiates a cascade of caspase activation and ultimately, intracellular proteolysis (66,67). NRG-1 has been demonstrated to directly suppress cardiomyocyte apoptosis by inhibiting mPTP opening, cytochrome c release and caspase-3 activation through PI3K/Akt signaling. Furthermore, the inhibition of ErbB2 and ErbB4 receptors leads to the induction of Bcl-xL splicing toward its pro-apoptotic protein Bcl-xS, thus activating mitochondrial dysfunction and apoptosis (68-70). This indirectly confirms the role of NRG-1/ErbB in myocyte apoptosis. Moreover, alleviating ER stress can inhibit the release of caspase-12, further reducing apoptotic death.

Autophagy denotes a regulated process of lysosomal degradation and the recycling of cytoplasm or mitochondrial proteins (mitophagy), characterized by the formation of double-membrane vesicles (autophagosomes) and elevated levels of light chain 3, beclin-1, autophagy-related gene (ATG) 5-12 complex, p62 and parkin (71). It has been demonstrated that autophagy acts as a 'double-edged sword' in the pathology of IR injury. To the best of our knowledge, autophagy may exert beneficial effects during the early period of ischemia, but detrimental effects during the late period of ischemia and reperfusion (72). The present review focuses on the prevention of the detrimental effects induced by autophagy. It has been illustrated that the PI3K/PKB/mTOR and mitogen-activated protein kinase (MAPK)/ERK1/2/mTOR pathways activated by NRG-1/ErbB via phosphorylation of phosphatidylinositol are involved in the negative regulation of autophagy (73,74). Moreover, the Akt-mediated reduction of ROS (Akt/ROS signaling) results in the upregulation of Bcl-2, which plays a role in the anti-autophagy effects of NRG-1 (73).

Preconditioning is the protective stimulus applied prior to the onset of a sustained episode of isch-emia (78). Generally, preconditioning comprises two temporal windows, as well as protection peaks. The first window (acute form) occurs several minutes following the stimulus and lasts for 1-3 h, while the second window (delay form) begins a few hours (peak at 24 h) after the stimulus and lasts several days but no longer than 72 h (79). The major interventions of preconditioning include ischemic conditioning (IPC), pharmacological preconditioning, remote ischemic preconditioning (RIPC) and physical preconditioning, which have been shown to reduce the MI size (78,80). Hereinafter, the role of NRG-1 in IPC and pharmacological preconditioning is described.

IPC refers to a process of repetitive non-lethal ischemia prior to sustained lethal myocardial ischemia, which increases cardiac resistance against IR injury (81). It has been shown that the reperfusion injury salvage kinase [e.g., the PI3K/AKT/glycogen synthase kinase (GSK)-3β and ERK1/2 pathways] and the survivor activator factor enhancement pathways (e.g., TNF-α, JAK2/STAT3 pathway) are the main pathways involved in IPC-induced cardioprotection (75,82). Notably, NRG-1/ErbB shares the same signaling pathways with IPC as above, and NRG-1 is rapidly upregulated during IPC and myocardial IR injury. This indicates that NRG-1 may at least partially play an important role in IPC-induced cardioprotection.

Pharmacological preconditioning is characterized by pre-treatment with drugs (78). In vivo, NRG-1 preconditioning protects the heart against IR injury by reducing myocardial necrosis and apoptosis mediated by a PI3K/Akt-dependent mechanism. Consistent with MI size limitations, NRG-1 preconditioning also decreases the level of plasma CK and LDH after 45 min of regional myocardial ischemia and 180 min of reperfusion (24). Thus, NRG-1 preconditioning is effective at reducing myocardium damage and represents a novel cardiac protective strategy that can be used in the setting of elective myocardial IR, as encountered during cardiac surgery and acute myocardial infarction.

Postconditioning, defined as cardioprotective intervention applied at the onset of reperfusion following sustained ischemia (83), can be achieved by short repeated occlusions of the vessel prior to permanent reperfusion [ischemic postconditioning (IP)] or by pharmacological interventions [pharmacological postconditioning (pPC)], which both have recently been shown to have potential as novel cardioprotective interventions against IR injury (77).

For conferring protection against IR injury, IP plays a key role in the alleviation of oxidative stress, inflammation and apoptosis through NO production and mitoKATP channels opening by salvage kinase pathways, including AKT, ERK1/2, 5'AMP-activated protein kinase (AMPK), protein kinase (PK)C and PKG (84-86). Of note, in vivo, IP promotes NRG-1 protein expression, as well as the upregulation/activation of ErbB3 and ErbB4, indicating that the cardioprotection may be mediated by the NRG-1/ErbB3 and ErbB4 signaling pathways, which has also recently been confirmed in ischemic local postconditioning (87).

pPC with NRG-1 concurrently with reperfusion has been shown to inhibit apoptosis and reduce MI size in a IR rat model or isolated murine heart (25), exerting a cardioprotective effect via the PI3K/Akt pathway (87,88). Therefore, pPC with NRG-1 may be an effective treatment following timely reperfusion, such as thrombolytic therapy or primary percutaneous coronary intervention (PPCI).

Recently, stem cell-based therapies have provided great promise for interventions in ischemic heart diseases (89,90). Among the forms of ischemic heart diseases, stem cells have been explored, particularly in the setting of myocardial infarction (91). In this connection, it has been reported that NRG-1 is a pivotal target of stem cell regulation, both in cultured cardiomyocytes and in whole embryos (5). In addition to mediating ventricular myocyte proliferation, NRG-1/ErbB has been found to be involved in embryonic stem cell (ESC) differentiation into cardiac myocyte lineages (92) and further, in the induction of cardiac conduction system cell differentiation. Moreover, the differentiation of stem cells into working-type cardiomyocytes can be modulated by NRG-1 (93). These effects can be interpreted as the upregulation of connexin with NRG-1 administration in ESC-derived cardiomyocytes, such as connexin 40 (Cx40) and connextin-45 (Cx45) (94). As intracellular channel proteins, connexins bridge gaps with cardiomyocytes to achieve synchronized contraction of the heart (95). The upregulated expression of connexins induced by NRG-1 also helps ESC differentiate into cardiac myocytes via MEK/ERK, and different cell lineages may be attributed to the different expression of connexin (94). It has been reported that treatment with cardiosphere-derived cells (CDCs) improves ventricular function in children with single ventricle physiology (96), which indicates the potential of cardiac self-repair. Since myocardial IR injury is inevitable in MI, further attention should be paid to restituting and restoring functional and structural components of the heart; stem cell transplantation administered with NRG-1 may be a strategic option.

Gene-based therapy utilizes gene delivery systems (e.g., viral and non-viral vectors) to modulate gene expression at the cellular level to treat pathological conditions (97). As regards IR injury, it has been reported that lentivirus-mediated hNRG-1 gene transduction establishes a stable expression system in infarcted hearts of rats and further activates the PI3K/Akt/eNOS pathway to promote neovascularization and angiogenesis, as manifested by enhanced expression of VEGF. In addition, the overexpression of hNRG-1 alleviates myocyte apoptosis through the Bcl-2/Bax signaling pathway (98). Collectively, the gene-based therapy of NRG-1 helps attenuate IR injury and eventually improves cardiac function. Although the application is still limited to animal experiments, and gene-based therapy has not yet been popularized, the significant protective effects suggest that gene delivery can be an alternative approach for NRG-1-dependent therapeutic strategies against myocardial IR injury.

The widespread clinical use of cardiovascular protein treatment may be hampered due to the limited stability and rapid degradation of protein, and novel formulation strategies that take into account sustained drug bioavailability in the infarcted border zone are urgently required (99,100). The application of microparticles (MPs) through catheter-based intramyocardial injection, with minimally invasive methods, may be a desirable approach for the clinical translation of cardiac regenerative medicine of MI (101). Briefly, cardiovascular protein molecules, such as NRG-1, are encapsulated into delicate bioresorbable scaffolds (PLGA) to form MPs and injected into target cardiac tissue with the guidance of visual cardiac mapping to achieve precise treatment (102). It has previously been demonstrated that NRG-1 plays critical roles in cardiac remodeling and MI size limitation through RISK and survivor activating factor enhancement (SAFE) pathways (103), and these benefits can be maximally utilized in the target infarcted zone. NRG-1-loaded MPs have been previously applied in a porcine model of IR over a period of months without severe side-effects; additionally, a prolonged and effective angiogenic stimulus was provided to the ischemic myocardium due to the sustained release, which failed to achieve success in clinical trials by applying pro-angiogenic factors (104). Notably, the transplantation of adipose-derived stem cells combined with NRG-1-loaded MPs stimulated cardiomyocyte proliferation and provided more complete healing in a rat myocardial infarction (105), suggesting that 'the whole is greater than the sum of its parts.' With the growing morbidity of AMI and the progress being made in precision medicine, NRG-1-loaded MPs may be a promising treatment for patients with MI by enhancing patient compliance and curative effects.

Cardiac transplantation, considered as the only effective therapy for end-stage heart failure, requires appropriate storage conditions for donor hearts to attenuate IR injury and preserve heart function during reperfusion (106). It has been demonstrated that rhNRG-1 mitigates left ventricular remodeling and sarcomere disorganization by upregulation of the RISK pathway (107,108). Notably, by combining organ-storage solution (Celsior) with rhNRG-1 in an isolated working rat heart model, additional cardiac preservation was observed after hypothermic storage, as evidenced by the reduction of myocyte apoptosis and necrosis during transplantation. Moreover, this recovery function could be enhanced by combination with other cardioprotective agents (e.g., glyceryl trinitrate and cariporide) (109). Accompanied by increased steady-state level of phosphorylated kinases [e.g., p-Akt, p-ERK1/2, p-signal transducer and activator of transcription 3 (STAT3) and p-GSK-3β] and reduced cleaved caspase-3, NRG-1 may exert these benefits by activating downstream pathways, including Akt, Ekr1/2 and JAK/STAT3, and involving caspase-3 related apoptosis (109,110). With rhNRG-1 supplementation, the goals of a longer storage time and higher cardiac vitality can be achieved during transplantation, and the success rate of surgery can be enhanced due to attenuation of IR injury. This suggests that NRG-1 may partially mitigate the contradiction between wanting donor hearts and growing clinical needs via the potential benefit against myocardial IR injury.