The Hypoxia-Adenosine Link during Myocardial Ischemia—Reperfusion Injury

Despite increasing availability and more successful interventional approaches to restore coronary reperfusion, myocardial ischemia-reperfusion injury is a substantial cause of morbidity and mortality worldwide. During myocardial ischemia, the myocardium becomes profoundly hypoxic, thus causing stabilization of hypoxia-inducible transcription factors (HIF). Stabilization of HIF leads to a transcriptional program that promotes adaptation to hypoxia and cellular survival. Transcriptional consequences of HIF stabilization include increases in extracellular production and signaling effects of adenosine. Extracellular adenosine functions as a signaling molecule via the activation of adenosine receptors. Several studies implicated adenosine signaling in cardioprotection, particularly through the activation of the Adora2a and Adora2b receptors. Adenosine receptor activation can lead to metabolic adaptation to enhance ischemia tolerance or dampen myocardial reperfusion injury via signaling events on immune cells. Many studies highlight that clinical strategies to target the hypoxia-adenosine link could be considered for clinical trials. This could be achieved by using pharmacologic HIF activators or by directly enhancing extracellular adenosine production or signaling as a therapy for patients with acute myocardial infarction, or undergoing cardiac surgery.


Introduction
Myocardial ischemia-reperfusion injury is most commonly caused by a mechanical obstruction of a coronary artery, for example by a plaque, thromboembolism or vasospasm [1][2][3]. The subsequent restoration of coronary blood flow will cause inflammatory cells to move into the ischemic myocardial tissues, which provides the immunologic cause of cardiac reperfusion injury [4]. Despite the advancement of clinical strategies to achieve earlier and more persistent reperfusion, myocardial ischemia-reperfusion injury continues to be a leading cause of morbidity and mortality in the USA and worldwide [2,3]. In addition, cardioprotection from ischemia-reperfusion is also critical for patients who are undergoing cardiac surgery, since those patients are at risk for myocardial ischemiareperfusion injury [5,6]. Novel pharmacologic approaches to render the myocardium more resistant to ischemic tissue injury or dampen myocardial inflammation during reperfusion would be highly desirable and are currently areas of intense research.
During myocardial ischemia, the occlusion of a coronary vessel causes profound changes in metabolic supply and demand within the area that is perfused by the specific artery (so-called "area at risk"). Due to the limited supply of metabolites and oxygen from the bloodstream, the area of risk becomes profoundly hypoxic, thus leading to the

Role of HIF in Regulating Adenosine Signaling during Myocardial Ischemia-Reperfusion Injury
Several previous studies have proposed linkages between hypoxia, HIF, and extracellular adenosine signaling as a means to providing tissue-adaptation, or to dampen hypoxia-driven inflammation [61][62][63]. In the extracellular compartment, adenosine is generated from precursor nucleotides, such as ATP or ADP [64][65][66][67]. Once adenosine is generated, it can signal through four distinct adenosine receptors, including the adenosine A 1 receptor (ADORA1), the adenosine A 2 A receptor (ADORA2A), the adenosine A 2 B receptor (ADORA2B), and the adenosine A 3 receptor (ADORA3) [68,69]. These G-protein coupled receptors have all been implicated in cardio-adaptive responses [53]. For example, the Adora1 is known to mediate the heart-rate slowing effects of intravenous adenosine, used for the treatment of supraventricular tachycardia [70]. In particular, the Adora2a and the Adora2b have been shown to dampen inflammatory responses [71][72][73][74]. For example, Adora2a signaling has been discovered on polymorphonuclear neutrophils (PMNs) [75] and contributes to attenuated inflammatory responses [76,77]. The subsequent uptake of adenosine from the extracellular compartment [49,78], and metabolism to inosine [13,[79][80][81] or AMP is implicated in terminating extracellular adenosine signaling. In the next section, we will discuss studies on how HIF-dependent alterations of gene transcription can alter extracellular adenosine signaling during myocardial injury, and how these responses have functional implications on cardioprotection during myocardial ischemia-reperfusion injury.

Impact of Hypoxia-Signaling on the Production of Extracellular Adenosine
During conditions of hypoxia, inflammation, or cellular stress, different cells release nucleotides, particularly in the form of ATP or ADP. For example, ATP can be released from inflammatory cells through specific molecular pathways [16,[82][83][84][85][86][87]. The extracellular release of ADP has been described extensively from platelets [68]. ATP or ADP can function as precursor molecules for the extracellular production of adenosine. This process is a two-step, enzymatically controlled pathway. As the first step, the ectonucleotidase CD39 converts extracellular ATP or ADP to AMP [88]. Studies in gene-targeted mice for cd39 (cd39 −/− mice) [89] show that these mice experience larger myocardial infarct sizes in the context of diminished levels of AMP and adenosine [90]. Moreover, cd39 −/− mice are not protected by ischemic preconditioning, where one or more preceding cycles of myocardial ischemia are associated with the attenuated size of injury [90][91][92]. Importantly, several studies demonstrate that the transcript and protein levels, and also the enzymatic function of CD39 are increased during ischemia, inflammation, or hypoxia [15,90,[93][94][95][96]. Studies on the transcriptional mechanism controlling CD39 expression during limited oxygen availability link the increased CD39 levels to transcriptional control of Sp1 [97,98]. The second step for the extracellular production of adenosine is under the control of the ecto-5 -nucleotidase CD73 [99]. This enzyme promotes the extracellular conversion of AMP to adenosine and can be considered as a "pace-maker" for extracellular adenosine generation. Similar to cd39 −/− mice, gene-targeted mice for cd73 [100] experience increased myocardial injury and are not protected by ischemic preconditioning [101]. As would be expected based on its enzymatic function, cd73 −/− mice experience attenuated concentrations of cardiac adenosine in conjunction with elevated cardiac AMP concentrations during myocardial injury [101]. Moreover, transcript, protein, and functional levels of CD73 are increased under hypoxia [101]. Several studies link these increases to a transcriptional program under the control of HIF1A. Studies with transcription factor binding assays and promoter constructs had demonstrated that CD73 is a classic HIF target gene and implicate HIF-dependent induction of CD73 in hypoxia-adaptive responses [93,94,100]. When exposed to myocardial ischemia-reperfusion, cd73 −/− mice experience larger infarct sizes and higher elevations of cardiac injury markers (troponin I) compared to control animals [101]. Subsequent studies during myocardial injury demonstrate that the protective effects of pharmacologic HIF activator treatment is attenuated in cd73 −/− mice [23], thereby directly implicating HIF-dependent CD73 regulation in cardioprotection. Together, these studies indicate that during myocardial ischemia, hypoxia signaling through Sp1 and HIF1A coordinate the transcriptional induction of CD39 and CD73, which leads to the increased production of extracellular adenosine and thereby contributes to attenuated myocardial infarct sizes ( Figure 1). [15,90,[93][94][95][96]. Studies on the transcriptional mechanism controlling CD39 expression during limited oxygen availability link the increased CD39 levels to transcriptional control of Sp1 [97,98].
The second step for the extracellular production of adenosine is under the control of the ecto-5′-nucleotidase CD73 [99]. This enzyme promotes the extracellular conversion of AMP to adenosine and can be considered as a "pace-maker" for extracellular adenosine generation. Similar to cd39 -/mice, gene-targeted mice for cd73 [100] experience increased myocardial injury and are not protected by ischemic preconditioning [101]. As would be expected based on its enzymatic function, cd73 -/mice experience attenuated concentrations of cardiac adenosine in conjunction with elevated cardiac AMP concentrations during myocardial injury [101]. Moreover, transcript, protein, and functional levels of CD73 are increased under hypoxia [101]. Several studies link these increases to a transcriptional program under the control of HIF1A. Studies with transcription factor binding assays and promoter constructs had demonstrated that CD73 is a classic HIF target gene and implicate HIF-dependent induction of CD73 in hypoxia-adaptive responses [93,94,100]. When exposed to myocardial ischemia-reperfusion, cd73 -/mice experience larger infarct sizes and higher elevations of cardiac injury markers (troponin I) compared to control animals [101]. Subsequent studies during myocardial injury demonstrate that the protective effects of pharmacologic HIF activator treatment is attenuated in cd73 -/mice [23], thereby directly implicating HIF-dependent CD73 regulation in cardioprotection. Together, these studies indicate that during myocardial ischemia, hypoxia signaling through Sp1 and HIF1A coordinate the transcriptional induction of CD39 and CD73, which leads to the increased production of extracellular adenosine and thereby contributes to attenuated myocardial infarct sizes ( Figure 1).

Figure 1.
Hypoxia increases extracellular adenosine during myocardial ischemia. In the context of hypoxia, different cell types such as inflammatory cells and platelets release large amounts of adenine nucleotides (particularly ATP or ADP). The ectonucleotidases CD39 and CD73 convert ADP/ATP to AMP and AMP to adenosine, respectively. Therefore, the level of extracellular adenosine during hypoxia or inflammation critically depends on the expression level and enzymatic activity of CD39 and CD73. Hypoxia promotes the induction of CD39 expression through SP1 signaling, and of CD73 expression through binding of the transcription factor hypoxia-inducible factor HIF1A to a hypoxia-response element (HRE) within the CD73 promoter. ATP: adenosine triphosphate; ADP: adenosine diphosphate; AMP: adenosine monophosphate. Hypoxia increases extracellular adenosine during myocardial ischemia. In the context of hypoxia, different cell types such as inflammatory cells and platelets release large amounts of adenine nucleotides (particularly ATP or ADP). The ectonucleotidases CD39 and CD73 convert ADP/ATP to AMP and AMP to adenosine, respectively. Therefore, the level of extracellular adenosine during hypoxia or inflammation critically depends on the expression level and enzymatic activity of CD39 and CD73. Hypoxia promotes the induction of CD39 expression through SP1 signaling, and of CD73 expression through binding of the transcription factor hypoxia-inducible factor HIF1A to a hypoxia-response element (HRE) within the CD73 promoter. ATP: adenosine triphosphate; ADP: adenosine diphosphate; AMP: adenosine monophosphate.

Role of HIF in Coordinating Extracellular Adenosine Signaling during Myocardial Ischemia-Reperfusion Injury
As described above, myocardial ischemia-reperfusion injury is associated with increased production of extracellular adenosine. Adenosine acts on four different receptor subtypes, including Adora1, Adora2a, Adora2b, and Adora3. All these receptors have been implicated in providing cardioprotection [101][102][103][104]. However, only the Adora2a and the Adora2b are transcriptionally regulated by HIF. They are highly expressed on a variety of different cellular sources, for example on cells of the innate immune system [96,105,106], erythrocytes [18,20], cardiac myocytes [56], stromal or epithelial cells [107][108][109][110], regulatory T-cells [111][112][113], and other immune cells [114]. Several previous studies have shown that the Adora2b promoter contains an HRE and can be directly induced by HIF1A during conditions of hypoxia [115], inflammation [108,116], or during myocardial ischemiareperfusion injury [23,56]. Similarly, the Adora2a has been previously identified as a target for hypoxia-signaling through the HIF2A isoform [117]. Studies in murine models of myocardial ischemia-reperfusion injury implicate both Adora2a and Adora2b in cardioprotection from ischemia-reperfusion. For example, murine studies demonstrate that infarct size-reducing effects of treatment with an Adora2a agonist are linked to Adora2a signaling on bone-marrow-derived T or B lymphocytes [118], which were subsequently identified to be most likely CD4+ T cells [119]. Similar to the Adora2a, several studies implicate the Adora2b in cardioprotection from ischemia-reperfusion injury. For example, Adora2b −/− mice are not protected by ischemic preconditioning and exhibit larger myocardial infarct sizes [101]. Moreover, treatment with a specific agonist for the Adora2b is associated with a significant reduction in infarct sizes in murine [101] or rat [120] models of myocardial ischemia-reperfusion injury. Studies using treatment approaches with the pharmacologic HIF activator dimethyloxalylglycine (DMOG) demonstrate abolished cardioprotection by this treatment in Adora2b −/− mice, thereby directly linking HIF1A and Adora2b signaling during cardioprotection [23]. Studies on the cellular source of the Adora2b receptor implicate myeloid-dependent Adora2b signaling in cardioprotection from ischemia-reperfusion injury [121,122]. Other studies suggest that Adora2b signaling on cardiac myocytes or inflammatory cells can interface with the stabilization of circadian rhythm signaling molecules, thereby contributing to the circadian oscillation of myocardial injury [53,56,[123][124][125]. In addition, a recent study demonstrated a regulatory function for Adora2b signaling in promoting epicardial stromal cells HIF stabilization after myocardial infarction as an additional crosstalk between Adora2b and HIF implicated in cardioprotection after myocardial infarction [109]. Taken together, these findings demonstrate HIF-dependent control of adenosine receptor expression and signaling in attenuating myocardial injury during ischemia-reperfusion ( Figure 2).

HIF-Dependent Promotion of Alternative Adenosine Receptor Activation
Several studies implicate the neuronal guidance molecule netrin-1 [126] in alternative mechanisms of adenosine receptor activation, particularly for the Adora2b [127,128]. Netrin-1 was discovered as a neuronal guidance molecule. Its function was originally described as netrin-1 secreted from cells of the floor plate of the mammalian embryonic neural tube [129][130][131]. Its secretion sets up a circumferential gradient of netrin-1, which in some instances attracts or in other instances repels other axons to the ventral midline [126]. Receptors for secreted netrin-1 include, for example, the receptor DCC (deleted in colorectal cancer) [132] and the UNC5 homologs (UNC5A, B, C, and D) [133] and neogenin-1 [134]. Importantly, the profound ability of the netrin-1 in guiding the repulse or outgrowth of neuronal cells makes it an ideal candidate molecule for the coordination of inflammatory cell migration [135].

HIF-Dependent Promotion of Alternative Adenosine Receptor Activation
Several studies implicate the neuronal guidance molecule netrin-1 [126] in alternative mechanisms of adenosine receptor activation, particularly for the Adora2b [127,128]. Netrin-1 was discovered as a neuronal guidance molecule. Its function was originally described as netrin-1 secreted from cells of the floor plate of the mammalian embryonic neural tube [129][130][131]. Its secretion sets up a circumferential gradient of netrin-1, which in some instances attracts or in other instances repels other axons to the ventral midline [126]. Receptors for secreted netrin-1 include, for example, the receptor DCC (deleted in colorectal cancer) [132] and the UNC5 homologs (UNC5A, B, C, and D) [133] and neogenin-1 [134]. Importantly, the profound ability of the netrin-1 in guiding the repulse or outgrowth of neuronal cells makes it an ideal candidate molecule for the coordination of inflammatory cell migration [135].
Studies utilizing a two-hybrid screen of a human brain cDNA library discovered a previously unreported interaction of netrin-1 with the Adora2b adenosine receptor [136]. Several studies using genetic and pharmacologic approaches demonstrate that netrin-1 can function to promote Adora2b signaling during inflammatory conditions outside of Studies utilizing a two-hybrid screen of a human brain cDNA library discovered a previously unreported interaction of netrin-1 with the Adora2b adenosine receptor [136]. Several studies using genetic and pharmacologic approaches demonstrate that netrin-1 can function to promote Adora2b signaling during inflammatory conditions outside of the brain, including acute lung injury [47,128,137,138], inflammatory peritonitis [127], intestinal inflammation [139,140], inflammatory kidney disease [141], corneal wound healing [142], and also myocardial ischemia-reperfusion injury [143]. However, one study found inconsistent results by showing that the Adora2b is actually not expressed in neurons, and is functionally not required for commissural axon guidance in the context of netrin-1 signaling [144]. At present, it is not well understood how netrin-1 and the Adora2b interact, including the possibility that netrin-1 could directly bind to Adora2b as its ligand, a role of netrin-1-dependent enhancement of extracellular adenosine levels, or indirect effects of netrin-1 by binding to a classic netrin-1 receptor and enhancing intracellular signaling cascades under the control of the Adora2b. A recent study found that netrin-1 levels were up-regulated in samples of patients who experienced myocardial ischemia-reperfusion injury [143]. Subsequent studies in mice with deletion of netrin-1 in the myeloid lineage (Ntn1 loxp/loxp LyzM Cre+ mice) revealed selectively larger infarct sizes and higher troponin levels, while other mouse lines with conditional deletion of netrin-1 in other tissues didn't experience a similar phenotype. Importantly, treatment studies with recombinant netrin-1 demonstrated that the interaction of netrin-1 with myeloid-dependent Adora2b signaling is critical in this pathway, suggesting an autocrine signaling pathway [143]. Previous studies have found that the promoter of netrin-1 contains an HRE and that HIF1A binding to the netrin-1 promoter dramatically increases netrin-1 expression of transcript and protein levels [145]. Subsequent studies in myeloid cells confirmed that finding [146], including recent studies showing that Hif1a-deficient myeloid cells fail to induce netrin during injury [47]. In conjunction with these studies, it is conceivable that hypoxia-signaling coordinates netrin-1 and Adora2b signaling in an autocrine loop where neutrophil-derived netrin-1 attenuates myocardial injury through signaling events on Adora2b receptors expressed on myeloid cells of the innate immune system (Figure 3).

Impact of HIF Signaling on Extracellular Adenosine Uptake and Metabolism
Previous studies have implicated HIF in modulating extracellular adenosine uptake and metabolism. In this context, the consequences of HIF transcriptional activity function towards attenuating extracellular adenosine uptake and intracellular metabolism, thereby enhancing extracellular adenosine signaling events. Adenosine signaling is terminated through equilibrative nucleoside transporters (ENTs), particularly ENT1 and ENT2 [21,78,147]. Those are channels that allow the bidirectional flow of adenosine across the cell membrane following its gradient. Extracellular production of adenosine is dramatically increased and the gradient for adenosine is directed from the extracellular compartment towards the intracellular side during ischemia-reperfusion injury. Therefore, deletion or inhibition of adenosine transporters with an ENT inhibitor such as dipyridamole will result in increased extracellular adenosine levels. Due to its function as an ENT inhibitor, dipyridamole treatment has been used clinically for many decades during pharmacologic stress echocardiography, where it increases coronary adenosine levels, and can unmask coronary artery stenosis [148,149]. Importantly, for the hypoxia-adenosine link during myocardial injury, previous studies have shown that HIF functions to repress both ENT1 and ENT2 during conditions of hypoxia or inflammation, and thereby functions to increase extracellular adenosine levels [17,[19][20][21]108,150]. Interestingly, mice with global deletion of Ent1 experience elevated plasma levels of adenosine, which can contribute to cardioprotection [151]. However, the individual contributions of ENT1 versus ENT2 during myocardial ischemia-reperfusion injury have not been addressed, for example by using genetic murine models. Nevertheless, global inhibition of ENTs with dipyridamole has been implicated in cardioprotection from ischemia-reperfusion injury [152,153]. Taken together, these studies highlight the likelihood that HIF-dependent repression of ENTs contributes to cardioprotection from ischemia-reperfusion injury. However, it would be important to define the individual contributions of ENT1 versus ENT2, as well as their tissue-specific functions and adenosine receptor signaling events in experimental studies of myocardial ischemia-reperfusion injury (Figure 4). sion of transcript and protein levels [145]. Subsequent studies in myeloid cells confirmed that finding [146], including recent studies showing that Hif1a-deficient myeloid cells fail to induce netrin during injury [47]. In conjunction with these studies, it is conceivable that hypoxia-signaling coordinates netrin-1 and Adora2b signaling in an autocrine loop where neutrophil-derived netrin-1 attenuates myocardial injury through signaling events on Adora2b receptors expressed on myeloid cells of the innate immune system (Figure 3).  . HIF coordinates alternative adenosine receptor signaling via increasing netrin-1 expression and signaling through Adora2b. During myocardial reperfusion injury, different types of inflammatory cells, such as neutrophils, monocytes, etc. infiltrate into the myocardial tissue. This further exacerbates tissue hypoxia and tissue damage. During reperfusion, the transcript and protein levels of Netrin-1 are robustly increased in patients with myocardial ischemia and in mice with myocardial IR injury. The increased expression of netrin-1 is mediated by HIF1A activity, which can bind to an HRE within the Netrin-1 promoter. The increased release of netrin-1 enhances Adora2b signaling by interacting with myeloid Adora2b in an autocrine manner, dampens the accumulation of inflammatory cells, and ultimately mediates cardioprotection against IR injury. The red arrowhead denotes increase, and the dark blue arrowhead denotes decrease. A2 B : Adenosine A2b Receptor. NTN1: Netrin-1.
In addition to the repression of ENT1 and ENT2, HIF1A has also been shown to repress a key metabolic process in intracellular adenosine metabolism. Adenosine can be metabolized intracellularly to AMP by the adenosine kinase (AK). Studies on hypoxia responses of AK demonstrate that hypoxia is associated with attenuated transcript and protein levels of AK. Moreover, studies in genetic models directly implicate HIF1A in its repression and demonstrate increased adenosine responses with AK repression [22]. Several studies implicate this pathway in cardioprotection. For example, experimental studies in rats treated with the AK inhibitor iodotubercidin demonstrate attenuated myocardial infarct sizes [154]. Together, these studies indicate the likelihood that HIF1A-dependent repression of AK contributes to adenosine-dependent cardio-protection from ischemia-reperfusion injury (Figure 4).  . HIF contributes to attenuated adenosine uptake, reduced adenosine metabolism and concomitant cardioprotection during myocardial ischemia-reperfusion injury. Equilibrative nucleoside transporters (ENTs) regulate the uptake of adenosine from the extracellular towards the intracellular compartment where the major routes of adenosine removal is based on phosphorylation to AMP via adenosine kinase, thereby modulating adenosine levels. During myocardial ischemia-reperfusion injury, HIF transcriptionally represses ENT1, ENT2 and adenosine kinase, leading to elevated extracellular adenosine levels. The inhibition of ENTs in mice with dipyridamole or global deletion of Ent1 showed decreased intracellular adenosine uptake and increased extracellular adenosine levels, ultimately exerting cardioprotective effects. These indicate the contribution of HIF-dependent repression of ENTs to adenosine-mediated cardioprotection. ENT: equilibrative nucleoside transporter; AK: adenosine kinase.

HIF-Dependent Cardioprotection beyond Purinergic Signaling Events
In addition to the cardioprotective functions of HIF1A, several studies revealed that HIF2A also contributes to cardioprotection from ischemia-reperfusion injury. A head-tohead comparison of mice with genetic deletion of Hif1a or Hif2a in cardiac myocytes revealed larger infarct sizes in Hif2a loxp/loxp Myosin Cre+ mice compared to Myosin Cre+ controls, whereas there was essentially no difference in infarct sizes between Hif1a loxp/loxp Myosin Cre+ mice and controls [52]. A subsequent microarray screen for HIF2A targets revealed the epidermal growth hormone amphiregulin (AREG) as the most differentially regulated gene [52]. The epidermal growth factor receptor (EGFR or ErbB1) ligand AREG has been identified to induce activation of the survival kinases Akt in the myocardium to protect against ischemia-reperfusion injury [52]. Previous studies had shown that AREG can be induced by hypoxia, independent of HIF1A [155][156][157]. Indeed, mice with global deletion of Areg or mice with myocyte-specific deletion of the Areg-receptor ErbB1 (ErbB1 loxp/loxp Myosin Cre+) demonstrated increased susceptibility to myocardial ischemiareperfusion injury [52,54]. Genetic studies in mice with Hif2a deletion confirmed the regulatory function for HIF2A for the transcriptional induction of Areg. Interestingly, HIF2A was also found to be critical for the induction of the AREG receptor ERBB1; however, this was independent of a transcriptional role of HIF2A [54]. Together, those findings demonstrate the cardioprotective functions of HIF2A expressed in cardiac myocytes by coordinating the induction of AREG and signaling through the ERBB1 receptor ( Figure 5), independent of purinergic signaling events. . HIF contributes to attenuated adenosine uptake, reduced adenosine metabolism and concomitant cardioprotection during myocardial ischemia-reperfusion injury. Equilibrative nucleoside transporters (ENTs) regulate the uptake of adenosine from the extracellular towards the intracellular compartment where the major routes of adenosine removal is based on phosphorylation to AMP via adenosine kinase, thereby modulating adenosine levels. During myocardial ischemia-reperfusion injury, HIF transcriptionally represses ENT1, ENT2 and adenosine kinase, leading to elevated extracellular adenosine levels. The inhibition of ENTs in mice with dipyridamole or global deletion of Ent1 showed decreased intracellular adenosine uptake and increased extracellular adenosine levels, ultimately exerting cardioprotective effects. These indicate the contribution of HIF-dependent repression of ENTs to adenosine-mediated cardioprotection. ENT: equilibrative nucleoside transporter; AK: adenosine kinase.

HIF-Dependent Cardioprotection beyond Purinergic Signaling Events
In addition to the cardioprotective functions of HIF1A, several studies revealed that HIF2A also contributes to cardioprotection from ischemia-reperfusion injury. A headto-head comparison of mice with genetic deletion of Hif1a or Hif2a in cardiac myocytes revealed larger infarct sizes in Hif2a loxp/loxp Myosin Cre+ mice compared to Myosin Cre+ controls, whereas there was essentially no difference in infarct sizes between Hif1a loxp/loxp Myosin Cre+ mice and controls [52]. A subsequent microarray screen for HIF2A targets revealed the epidermal growth hormone amphiregulin (AREG) as the most differentially regulated gene [52]. The epidermal growth factor receptor (EGFR or ErbB1) ligand AREG has been identified to induce activation of the survival kinases Akt in the myocardium to protect against ischemia-reperfusion injury [52]. Previous studies had shown that AREG can be induced by hypoxia, independent of HIF1A [155][156][157]. Indeed, mice with global deletion of Areg or mice with myocyte-specific deletion of the Areg-receptor ErbB1 (ErbB1 loxp/loxp Myosin Cre+) demonstrated increased susceptibility to myocardial ischemiareperfusion injury [52,54]. Genetic studies in mice with Hif2a deletion confirmed the regulatory function for HIF2A for the transcriptional induction of Areg. Interestingly, HIF2A was also found to be critical for the induction of the AREG receptor ERBB1; however, this was independent of a transcriptional role of HIF2A [54]. Together, those findings demonstrate the cardioprotective functions of HIF2A expressed in cardiac myocytes by coordinating the induction of AREG and signaling through the ERBB1 receptor ( Figure 5), independent of purinergic signaling events.
In contrast, the cardioprotective effects of HIF1A signaling have been suggested in other tissue compartments than myocytes, for example in vascular endothelial cells. Mice with Hif1a deletion specifically in vascular endothelial cells were not protected from ischemic preconditioning [158,159]. Moreover, previous studies have implicated HIF1A in mediating the effects of remote ischemic preconditioning [160]. This is an experimental strategy where short repetitive episodes of ischemia to an arm or a leg provide organ protection to the heart or the kidneys from a subsequent ischemic injury [161,162] and have been applied successfully in randomized trials of patients undergoing major surgery [163,164]. Experimental studies demonstrate that remote ischemic preconditioning leads to the stabilization of HIF1A, and subsequent induction of IL-10 as HIF target genes. Cardiac IL-10 signaling is subsequently responsible for the observed cardioprotection [160]. Together, these studies demonstrate that there are multiple functions of HIF1A and HIF2A to orchestrate cardioprotection. While HIF-dependent enhancement of extracellular adenosine signaling is central to its role in cardioprotection, there have also been pathways described that highlight HIF-dependent cardioprotection outside of purinergic signaling events.
OR PEER REVIEW 10 of 22 Figure 5. HIF2A induces AREG signaling in cardiac myocytes to provide cardioprotection. HIF2A contributes to cardioprotection during myocardial IR injury. The epithelial growth factor amphiregulin (AREG) has been identified as one of the target genes of HIF2A, which is significantly induced at both mRNA and protein levels in cardiomyocytes during hypoxia. HIF2A was also found to increase the expression of AREG receptor ERBB1 at the post-transcriptional level. These findings indicate HIF2A protects against myocardial IR injury through AREG signaling. ERBB: Epidermal growth factor receptor.
In contrast, the cardioprotective effects of HIF1A signaling have been suggested in other tissue compartments than myocytes, for example in vascular endothelial cells. Mice with Hif1a deletion specifically in vascular endothelial cells were not protected from ischemic preconditioning [158,159]. Moreover, previous studies have implicated HIF1A in mediating the effects of remote ischemic preconditioning [160]. This is an experimental strategy where short repetitive episodes of ischemia to an arm or a leg provide organ protection to the heart or the kidneys from a subsequent ischemic injury [161,162] and have Figure 5. HIF2A induces AREG signaling in cardiac myocytes to provide cardioprotection. HIF2A contributes to cardioprotection during myocardial IR injury. The epithelial growth factor amphiregulin (AREG) has been identified as one of the target genes of HIF2A, which is significantly induced at both mRNA and protein levels in cardiomyocytes during hypoxia. HIF2A was also found to increase the expression of AREG receptor ERBB1 at the post-transcriptional level. These findings indicate HIF2A protects against myocardial IR injury through AREG signaling. ERBB: Epidermal growth factor receptor.

Potential Therapeutic Approaches
Various drugs targeting the different steps of the HIF-adenosine link have been developed for myocardial protection. A summary of some of the published clinical trials and ongoing clinical trials on these medications are presented in Tables 1 and 2.

HIF Activators
Since the advent of a new generation of HIF activators, such as PHD inhibitors, many of them have finished phase 3 clinical trials and are currently applying for FDA approval for the treatment of anemia in patients with chronic kidney disease. For example, in phase III studies, roxadustat has been shown to increase hemoglobin levels and reduce cholesterol levels in chronic kidney disease patients with or without dialysis [59,60]. Vadadustat also underwent phase III clinical trials, showing an improved iron metabolism and anemia in patients undergoing dialysis [57,165]. Two other orally available PHD inhibitors, daprodustat and molidustat, have also completed phase III clinical trials for the treatment of anemia associated with chronic kidney disease [166][167][168]. Although most previous clinical research on these drugs has mainly focused on treating kidney disease, animal studies have also demonstrated myocardial protection after PHD inhibitor treatment in rodent models of myocardial ischemia-reperfusion injury or heart failure [169][170][171][172][173]. In addition, PHD inhibitors are orally available and have shown favorable short-term safety profiles that would make them ideal for the treatment of myocardial ischemia-reperfusion injury for prophylactic treatments of patients undergoing cardiac surgery. For example, a phase II clinical trial (ROXAMI, NCT04803864) on the efficacy and safety of roxadustat in the treatment of patients with acute myocardial infarction is currently in the stage of recruitment. Our anticipation is that more of these compounds will soon be trialed for cardioprotection in patients.

Adenosine
Adenosine plays an important role in ischemia-reperfusion injury by improving postischemic ventricular function, reducing neutrophil activity, and limiting myocardial necrosis and apoptosis [174]. Moreover, adenosine is crucial for ischemic preconditioningmediated cardiac protection in animal models by decreasing myocytes apoptosis after the reperfusion [175]. Large-scale clinical trials have proven the efficacy of adenosine infusion in reducing myocardial infarction size in patients experiencing acute MI [176][177][178][179][180]. The Acute Myocardial Infarction Study with Adenosine (AMISTAD) trial found that adenosine treatment resulted in a 33% relative reduction in infarct size, with a more profound beneficial effect in patients with anterior wall infarction, although no reduction in the composite endpoint (e.g., death, reinfarction, shock, congestive heart failure or stroke) was observed [176]. The AMISTAD-II trial, which was designed as a follow-up trial of the AMISTAD-I trial to focus on anterior wall ST-elevated patients, did not find adenosine infusion to improve clinical outcomes. However, high-dose adenosine infusion significantly reduced infarct size (infarct size: 11% in the high-dose group and 27% in the placebo group; p = 0.023) [177]. In post-doc analysis of the AMISTAD-II trial, early adenosine infusion with reperfusion therapy improved survival and the six-month composite outcome [178] ( Table 1). The Attenuation by Adenosine of Cardiac Complications (ATTACC) trial did not demonstrate that low-dose adenosine improved left ventricular function at discharge in patients with acute myocardial infarction receiving thrombolysis, but suggested a potential benefit on long-term survival [179] (Table 1).
In addition to patients with acute myocardial infarction, adenosine has also been evaluated in patients undergoing cardiac surgery. It was used either as an adjunct to intermittent blood cardioplegia or as an intra-aortic infusion before the release of aortic cross-clamp [181][182][183][184], and has been shown to be cardioprotective in some clinical studies (Table 1). Despite favorable trial results, the clinical utility of systemic adenosine is limited by its ultra-short intravascular half-life (<1 s) and its undesirable peripheral hemodynamic side effects such as bradycardia and hypotension. Therefore, a more selective adenosine activator will be highly valuable in providing myocardial protection and avoiding side effects of adenosine at the same time.

Adenosine Receptor Agonists
Neladenoson bialanate is a partial adenosine A1 receptor agonist. Preclinical studies have found that this medication could provide potential cardioprotection by improving mitochondrial function, preventing ventricular remodeling, and reducing fibrosis, thereby preventing ischemic injury [185]. There are two randomized, double-blind, placebo-controlled, dose-finding Phase 2b trials investigating the effects of short-term neladenoson treatment on cardiac structure, function or exercise capacity in patients with heart failure. However, no significant beneficial effects were found [186,187] (Table 1).
Selective A2 A receptor agonists are also investigated as a therapeutic approach for cardiovascular diseases. The adenosine A2 A receptor agonist regadenoson (Lexiscan; Astellas Pharma Inc., Deerfield, Illinois, U.S.) is a commonly used agent for myocardial perfusion imaging studies. The clinical utilization of this medication for myocardial protection has not been validated.
Methotrexate, originally used as an anti-inflammatory drug for the treatment of rheumatoid arthritis, has received attention in recent years for its anti-atherosclerotic effects by increasing adenosine release and activating A2 A receptors [188]. The cardiovascular beneficial effect of methotrexate was initially found through several large retrospective studies. Micha et al. [189], in a meta-analysis of 66,334 patients, found that methotrexate at a median dose of 13-15 mg/week improved cardiovascular outcomes in patients with systemic inflammation. Another meta-analysis by Roubille et al. [190] found that patients with rheumatoid arthritis, psoriasis, or psoriatic arthritis on anti-rheumatic drugs treated with methotrexate have a reduced risk of cardiovascular events. The additional study indicated that methotrexate reduces cardiovascular related death mortality in patients with rheumatoid arthritis [191]. Despite these promising findings, in a phase III clinical trial in patients with stable coronary artery disease and Type 2 diabetes or metabolic syndrome (CIRT), low-dose methotrexate failed to reduce the incidence of cardiovascular events to meet the primary endpoint [192]. However, this trial was stopped early because lowdose methotrexate failed to reduce levels of inflammatory mediators and the incidence of cardiovascular events was similar to the placebo group (Table 1). Whether a higher dose of methotrexate could provide myocardial protection in prospective RCT still needs to be determined.

Adenosine Reuptake (ENT) Inhibitor
Equilibrative nucleoside transporter (ENT) inhibitors could serve as potential therapeutics for heart protection by potentiating the protective effects of adenosine. However, currently, there are no clinical trials designed to investigate ENT inhibitors in myocardial infarction. Only one study has attempted to assess the diagnostic and prognostic value of serum Netrin-1 levels in patients undergoing coronary angiography for acute coronary syndromes ( Table 2). Future trials focusing on myocardial protection of ENT inhibitors will hopefully shed some light on this promising therapy. Adenosine infusion (150 µg/kg/min) for 10 min into the aortic root, starting 10 min before aortic cross-clamp removal.
Adenosine postconditioning group showed better cardiac function indices, lower cardiac enzyme levels, lower incidence of arrhythmia, less inotropic drug consumption, and shorter ventilation time and ICU stay. Low-dose methotrexate did not reduce inflammatory markers levels and cardiovascular events compared with placebo.

2009
Moreira et al. [193] Methotrexate Therapy on the Physical Capacity of Patients with Ischemic Heart Failure (METIS Trial) Patients with ischemic chronic heart failure 50 Methotrexate (7.5 mg/week) or placebo, plus folic acid (5 mg/week), for 12 weeks.
For patients receiving methotrexate, their NYHA score showed an improving trend, but no significant change in 6-min walk test results.
CABG: coronary artery bypass grafting; AMI: acute myocardial infarction; MI: myocardial infarction; STEMI: ST-elevation myocardial infarction; CHF: congestive heart failure; PCI: percutaneous coronary intervention; LVEF: left ventricular ejection fraction; N/A: not available, NYHA score: New York Heart Association score. To assess the effect of 6-h adenosine infusion started before revascularization on the recovery of myocardial akinesia and cardiac function at 48 h in patients with STEMI.

NCT04027127
Patients diagnosed with acute coronary syndrome (ACS) and received coronary angiography To determine the effect of serum Netrin-1 levels on diagnosis and prognosis in patients presenting to emergency department with ACS.

Summary and Future Perspectives
Purinergic signaling events through the activation of extracellular adenosine receptors have long been implicated in cardioprotection from ischemia-reperfusion injury [152,[194][195][196][197]. More recent studies using mice with the genetic deletion of adenosine receptors globally, or in individual tissue compartments, have provided additional insight into mechanisms of adenosine-dependent cardioprotection. Moreover, many of these studies were able to link purine metabolism and signaling with the activity of hypoxia-signaling and highlight regulatory functions of HIF in coordinating adenosine-mediated cardioprotection. We are now at a stage where multiple pharmacologic tools are available to modulate the hypoxiaadenosine link for the treatment or prevention of myocardial ischemia-reperfusion injury. These strategies include the use of orally available HIF activators, adenosine receptor agonists, or adenosine transport inhibitors. We anticipated that clinical trials in patients with myocardial infarction or in patients undergoing cardiac surgery will help to bring those pharmacologic interventions from the research laboratory to the patient's bedside.