The role and mechanisms of microvascular damage in the ischemic myocardium

Abstract

Following myocardial ischemic injury, the most effective clinical intervention is timely restoration of blood perfusion to ischemic but viable myocardium to reduce irreversible myocardial necrosis, limit infarct size, and prevent cardiac insufficiency. However, reperfusion itself may exacerbate cell death and myocardial injury, a process commonly referred to as ischemia/reperfusion (I/R) injury, which primarily involves cardiomyocytes and cardiac microvascular endothelial cells (CMECs) and is characterized by myocardial stunning, microvascular damage (MVD), reperfusion arrhythmia, and lethal reperfusion injury. MVD caused by I/R has been a neglected problem compared to myocardial injury. Clinically, the incidence of microvascular angina and/or no-reflow due to ineffective coronary perfusion accounts for 5–50% in patients after acute revascularization. MVD limiting drug diffusion into injured myocardium, is strongly associated with the development of heart failure. CMECs account for > 60% of the cardiac cellular components, and their role in myocardial I/R injury cannot be ignored. There are many studies on microvascular obstruction, but few studies on microvascular leakage, which may be mainly due to the lack of corresponding detection methods. In this review, we summarize the clinical manifestations, related mechanisms of MVD during myocardial I/R, laboratory and clinical examination means, as well as the research progress on potential therapies for MVD in recent years. Better understanding the characteristics and risk factors of MVD in patients after hemodynamic reconstruction is of great significance for managing MVD, preventing heart failure and improving patient prognosis.

Introduction

In recent decades, global environmental changes, modification of modern lifestyles, such as the Western diet and physical inactivity, and increased population aging are all closely related to people's health [1, 2]. As a result, the prevalence of metabolic risk factors, including hypertension, dyslipidemia, hyperglycemia/insulin resistance, and overweight/obesity, has increased significantly worldwide [3]. These risk factors lead to the development of a single metabolic disorder into a metabolic syndrome, and eventually to the development of many comorbidities, one of which is cardiovascular disease (CVD) [3]. Each year, more people die from CVD than from any other causes [4]. In 2019, 17.9 million people died from CVD, representing 32% of global deaths [5]. Moreover, according to the World Health Organization report, it is estimated that about 23.6 million people will die from CVD by 2030 [4]. Globally, ischemic heart disease (IHD) has the highest burden and the mortality among all forms of CVD [3, 6], including acute myocardial infarction (AMI), unstable angina, chronic IHD and its associated heart failure [7, 8]. Such stark facts raise an alarm and call all governments and cardiologic professionals around the world working together to seek and establishing effective strategies for the prevention and treatment of IHD.

The coronary arteries consist of epicardial coronary arteries (diameter > 400 µm) and coronary microcirculation [9]. The latter are usually present in vessels < 400 μm in diameter and are not visible on coronary angiography. Different from the epicardial coronary arteries, coronary microcirculation is composed of small arteries (diameter < 400 µm), arterioles (diameter < 100 µm), and capillaries (diameter < 10 µm) [10]. Epicardial coronary arteries carry out functions that include delivery of blood, oxygen, and nutrients and removal of waste products from the blood [11]. Coronary microcirculation is responsible for regulating blood flow distribution, modulating vascular resistance, and exchanging oxygen and nutrients in the heart [12]. Figure 1 elucidates the mechanisms of myocardial ischemia induced by a variety of pathophysiological conditions. Atherosclerotic plaque formation and rupture are the primary cause of coronary artery disease (CAD)/myocardial ischemia, according to an intravascular ultrasound substudy of the WISE study, 79% of patients had coronary atherosclerosis [13]. Non-obstructive CAD including epicardial coronary spasm and microvascular dysfunction account for 20–50% of myocardial ischemia [14, 15]. Epicardial CAD is commonly classified into atherosclerotic disease and vasospastic disease. Atherosclerotic lesion includes stable plaque and vulnerable plaque. The former reduces coronary flow reserve (CFR) and eventually leads to demand ischemia/angina, while vulnerable plaques are prone to rupture which results in thrombosis/embolism, the primary cause of acute coronary syndromes (ACS)/permanent MI. Coronary vasospasm can trigger prinzmetal angina by focal/transient vasospasm or permanent MI by persistent vasospasm. Coronary microcirculation dysfunction, mainly caused by pathological stimulations, such as hypertension, diabetes, dyslipidemia, psychological and mental factors as well as increased sympathetic nervous system reactivity etc., can impair coronary physiology, decrease CFR and damage endothelial function. Coronary microcirculation dysfunction not only contributes to myocardial ischemia in CAD and cardiomyopathy, but also induces severe acute ischemia, e.g., ‘Takotsubo’. Moreover, these three mechanisms can overlap, leading to permanent myocardial necrosis [9]. The coronary perfusion territory distal to the site of the occlusion is referred to as the area at risk of infarction. Within a given area at risk, both the duration and the severity of coronary blood flow reduction determine the nature and amount of injury [16]. The most effective clinical intervention currently available is timely restoration of blood perfusion to ischemic but viable myocardium to reduce irreversible myocardial necrosis, limit infarct size and prevent cardiac insufficiency [17–19]. However, reperfusion itself may exacerbate cell death and myocardial injury, a process commonly referred to as myocardial ischemia/reperfusion (I/R) injury [19], which primarily involves cardiomyocytes and cardiac microvascular endothelial cells (CMECs) [20] and is characterized by myocardial stunning, microvascular damage (MVD), reperfusion arrhythmia, and lethal reperfusion injury [21]. MI size is the major determinants of mortality, heart failure and arrhythmias after AMI [22–24]. Sustained damage to myocardial tissue and coronary microvasculature by reperfusion is a complex multifactorial process that is estimated to cause up to 50% of the final infarct size [25]. Therefore, myocardial I/R injury substantially limits the clinical benefit of reperfusion strategies [20, 26].

Fig. 1.

Mechanisms of myocardial ischemia. (modified from [9] with a permission license 5,623,560,707,414) CFR coronary flow reserve, ACS acute coronary syndromes, AMI acute myocardial infarction, HT hypertension, SNS sympathetic nervous system, CMP cardiomyopathy

I/R-induced coronary microcirculatory disturbances have been a neglected problem compared to myocardial injury [26–28]. Clinically, the incidence of microvascular angina and/or no-reflow due to ineffective coronary perfusion accounts for 5–50% in patients after acute revascularization [26]. MVD limiting drug diffusion into the damaged myocardium, is strongly associated with the development of heart failure, and few drugs are currently available for its treatment.

Over the past few decades, scientists have investigated underlying mechanisms and various therapeutic strategies to protect the heart against I/R injury. To date, the translation of cardioprotection from basic research findings to the clinic is still much less than expected. One reason for this is due to the focus of research has been largely on the cardiomyocyte itself, with a considerable degree of neglect in coronary microcirculation which actually plays a critical role in cardiomyocyte survival and maintenance of cardiac function. Therefore, to understand the characteristics and risk factors of coronary MVD owing to I/R injury as well as the pathological changes in CMECs is of great significance for preventing coronary microvascular dysfunction and improving patient prognosis [18, 29].

Classification of microvascular damage after myocardial I/R injury

All human vascular lumens consist of a single layer of ECs, vascular smooth muscle cells (VSMCs) and pericytes [10]. The surface area of ECs on the human surface is > 1000 m² [30], which plays an important role in vascular tone, fluid balance between the lumen and tissues, macromolecule and cellular permeability, as well as in coagulation and angiogenesis [31]. CMECs account for > 60% of the cardiac cellular components, and their role in myocardial ischemic injury cannot be ignored [32]. A systematic review provides comprehensive analysis on coronary microvascular regulatory mechanisms involving mechanical determinants, direct endothelium effectors, humoral mediators and neuronal mechanisms [33]. The molecular mechanisms of coronary MVD following I/R injury are not fully understood, and its manifestations may include microvascular obstruction (MVO), also known as "no-reflow" phenomenon, and microvascular leakage (MVL).

Microvascular obstruction

According to the latest European Society of Cardiology international guidelines, direct percutaneous coronary intervention (PCI) is the preferred reperfusion strategy in patients with acute ST-segment elevation MI (STEMI) [34], and up to 95% of clinically occluded coronary arteries can be reopened in this manner [35]. However, despite the restoration of infarct-related arterial blood flow, epicardial blood flow is slow or inadequate due to the onset of MVO and consequent inadequate tissue perfusion, a phenomenon known as coronary artery “non-reflow” [18, 36]. I/R induced swelling of CMECs, reduced blood perfusion, and a change in local fluid dynamics from laminar to turbulent flow, which promotes erythrocyte aggregation, platelet plug and thrombosis as well as inflammatory cell embolism to block blood and nutrients diffuse to myocytes (Fig. 2) [37]. In addition, Heusch and coworkers reported that the accumulation of atherosclerotic debris and soluble substances (such as serotonin, thromboxane B2 and TNFα) released from ruptured plaques also plays a key role in microvascular perfusion [38]. Clinical data further support that an increase in the size of the no-reflow zone limits blood flow in the area at risk to about 44% of the baseline flow before the onset of myocardial ischemia [39], and similar correlations have been reported in animal models [40, 41]. Statistically, MVO can affect 11–41% of STEMI patients treated with PCI, significantly reduces the beneficial effects of reperfusion therapy, leads to poorer clinical and functional outcomes, and is an independent predictor of increased infarct extent and 5-year mortality [42–44], while the MI size is not independently associated with adverse events [45, 46]. In addition, clinical studies have shown that advanced age, delayed reperfusion, lower systolic blood pressure (SBP), left ventricular ejection fraction (LVEF) and myocardial blush grade (MBG) at admission, and higher Killip's grade also increase the risk of no-reflow [47].

Fig. 2.

Pathological mechanisms of microvascular damage following myocardium ischemia/reperfusion

Microvascular leakage

The inner surface of the coronary microvasculature consists of a single layer of CMECs [10]. The barrier function and permeability of microvasculature to water or other molecules depend on the gap junction structure between CMECs, the degree of encapsulation of the basement membrane and pericytes outside CMECs [48, 49]. Under physiological conditions, water and solute transport across capillaries is mainly carried out through gap junctions between CMECs. The gaps between CMECs account for approximately 0.2% of the cell surface area, and the permeability of the gaps is maintained by an intercellular junction device and adherents, composed of multiple cell adhesion molecules and various connexins of CMECs, which form a tight junction structure [48, 49]. These molecules bind to each other and are maintained with cytoskeletal proteins, which in turn form stable intercellular junctions. Experimental studies have found that myocardial I/R injury damages CMECs and barrier function which led to hyperpermeability of microvasculature. Extravasation of blood components and immune cells through the dysfunctional barrier results in intramyocardial hemorrhage (IMH), interstitial edema and local inflammatory responses, a phenomenon known as MVL (Fig. 2) [50]. In addition, the MVO, namely, the no-reflow phenomenon, will be further aggravated by IMH and the force of external compression owing to interstitial edema caused by MVL. It seems from the histological analysis [51] that no-reflow is always in the area of the infarcted tissue [52], but the size of MVL is significantly larger than the infarcted or no-reflow area [50].

MVD worsens in a time-dependent manner during I/R injury [10], which expands the infarct size and increases perioperative mortality in patients treated with revascularization [53]. Thus, despite the great success of treatments to reduce post-ischemic cardiomyocyte injury, exploration of more mechanistic studies and therapeutic agents are needed to solve the cardiac microvascular dysfunction during myocardial I/R injury. So far, most studies on MVD after myocardial I/R have been limited to single-center studies. Table 1 shows the incidence of MVD after I/R and its corresponding diagnostic method in clinical studies searched on PubMed over the past 10 years. The incidence of MVO varied among diagnostic methods, with 39.5% detected by ECG and 2.1–42.0% by CAG, while cardiac magnetic resonance (CMR) was the most sensitive, detecting 35.2–72.2% of MVO. Currently, MVL is only indirectly reflected in the detection of IMH and edema by CMR. The incidence of IMH was 33.6–50.0% and edema was 30.8–72.1% (refer to Table 1).

Table 1.

Incidence of MVD after myocardial I/R and diagnostic methods in clinical studies over the past 10 years

Events	No. of patients	Event present	Diagnostic method	Incidence %	References
MVO	569	225	ECG	39.5	Erkol et al. (2014) [54]
	16	6	CAG	37.5	Grygier et al. (2014) [55]
	569	179		31.5	Erkol et al. (2014) [54]
	120	14		11.7	Li et al. (2015) [56]
	5997	128		2.1	Cenko et al. (2016) [57]
	105	15		14.3	Chen et al. (2016) [58]
	86	31		36.0	Durante et al. (2017) [59]
	733	54		16.1	Liang et al. (2017) [60]
	203	38		18.7	Li et al. (2018) [61]
	100	27		2.7	Mahmoud et al. (2019) [62]
	1658	491		42.0	Yang et al. (2020) [63]
	173	24		13.9	Rossington et al. (2020) [64]
	118	41		34.7	Sadeghian et al. (2022) [65]
	909	101		11.1	d'Entremont et al. (2023) [66]
	774	381	CMR	49.2	Eitel et al. (2013) [67]
	281	99		35.2	Hadamitzky et al. (2014) [68]
	151	100		66.0	Kandler et al. (2014) [69]
	108	78		72.2	Ding et al. (2015) [70]
	55	39		70.9	Kim et al. (2015) [71]
	55	35		63.6	Desch et al. (2016) [72]
	245	133		54.3	Carrick et al. (2016) [73]
	63	38		60.3	Nazir et al. (2016) [74]
	86	58		67.4	Durante et al. (2017) [59]
	1688	960		56.9	Kloner et al. (2017) [75]
	98	60		61.2	Broch et al. (2021) [76]
	234	125		53.4	Bulluck et al. (2022) [77]
	170	94		55.3	Tiller et al. (2022) [78]
	122	51		41.8	Li et al. (2022) [79]
IMH	699	243	CMR	34.8	Eitel et al. (2013) [67]
	304	102		33.6	Husser et al. (2013) [80]
	151	76		50.0	Kandler et al. (2014) [69]
	108	53		49.0	Ding et al. (2015) [70]
	38	16		42.1	Nazir et al. (2016) [74]
	30	13		43.0	Carrick et al. (2016) [81]
	245	101		41.2	Carrick et al. (2016) [73]
	170	59		34.7	Tiller et al. (2022) [78]
	234	96		41.0	Bulluck et al. (2022) [77]
Edema	197	109	CMR	55.3	Nazir et al. (2016) [74]
	86	62		72.1	Durante et al. (2017) [59]
	800	246		30.8	Garg et al. (2017) [82]

MVD microvascular damage, MVO microvascular obstruction, I/R ischemia/reperfusion, ECG electrocardiogram, CAG coronary artery angiography, CMR cardiac magnetic resonance, IMH intramyocardial hemorrhage

Molecular mechanisms of microvascular damage after myocardial I/R injury

The overall mechanism of coronary MVD including MVO and MVL is not fully understood, and microcirculatory blockage and microvascular hyperpermeability are the possible driven forces involved. At present, MVO following myocardial I/R injury have been extensively studied, but few studies focus on MVL [83]. In the following sections, we will provide an overview on the possible mechanisms based on current knowledge.

Impaired microvascular function

To date, many studies have recognized that dysfunction or death of CMECs is an essential cause of cardiac MVD following I/R injury [37, 84, 85]. When coronary artery obstruction occurs, blood supply and oxygen are drastically reduced and aerobic ATP is decreased, causing dysfunction of the cell membrane Na⁺–K⁺ pump, which in turn leads to intracellular water and sodium retention. Thus, during reperfusion the cells in the ischemic area swell and compress the micro-vessels, while the swollen CMECs protrude into the lumen causing luminal narrowing and obstructing blood perfusion. The development of edema during reperfusion is characterized by bi-modal changes, with an early peak after 120 min which coincided with the beginning of the inflammatory response and followed by a near disappearance of edema after 24 h, and a second peak occurred after 7 days, which coincided with the deposition of collagen [86].

The tension of coronary microvasculature is regulated by the balance between vasodilators, e.g., bradykinin and nitric oxide (NO) and vasoconstrictors, e.g., endothelin-1 (ET-1), both are released by CMECs [10]. CMECs promote vascular relaxation by synthesizing and releasing NO via endothelial nitric oxide synthase (eNOS). However, after myocardial I/R injury, mitochondria-derived superoxide induces the oxidation of the eNOS co-factor tetrahydrobiopterin (BH4) to dihydrobiopterin, which prevents BH4 from binding to eNOS, thereby uncoupling eNOS and reducing the production of NO [87]. In addition, nicotinamide adenine dinucleotide phosphate oxidases (NOXs), the main enzymes that generate reactive oxygen species (ROS), are expressed in ECs as four different NOX isoforms, superoxide-generating NOX1, NOX2, and NOX5 as well as hydrogen peroxide-generating NOX4 [88]. NOX-derived ROS can reduce NO bioavailability by oxidizing NO to peroxynitrite, a highly reactive and vasoconstrictive ROS species [10]. Consequently, downregulation of phosphorylated eNOS expression, rapid upregulation of endothelial-type vasoconstrictor ET-1 expression, and decreased NO production and utilization lead to microvascular diastolic dysfunction. Moreover, increased levels of ROS in CMECSs promote angiotensin II type-1 (AT-1) receptor-mediated vasoconstriction [89]. Although, in general, coronary artery constriction during hypoxia is an adaptive response to accelerate blood flow, MVO limits sufficient nutrients to cardiomyocytes. Clinically, Neil and coworkers have observed an elevated levels of vasoconstrictor (thromboxane A2, ET-1, and neuropeptide Y) in coronary sinus blood immediately after PCI, and the level of neuropeptide Y was closely associated with norepinephrine-mediated vasoconstriction and the index of microcirculatory resistance (IMR) elevation, which in turn led to more severe MVO [90].

Besides the dysfunction of CMECs, it was found that pericytes, a type of cell that surrounds endothelial cells in microvasculature and veins [91] are also involved in MVD and repair after I/R through the following approaches (1) monitoring and stabilizing the maturation process of CMECs through cellular communication via physical contact and paracrine signaling [92]; (2) directly contract to reduce microvascular blood flow during I/R [93], which can be reversed by adenosine [93]; (3) participating in the inflammatory response after I/R by releasing inflammatory inhibitory factors, chemokines, and others [94, 95]; (4) attenuating leukocytes recruitment and transendothelial trafficking [10]; (5) promoting angiogenesis by increasing the production of angiogenic factors, such as vascular endothelial growth factor (VEGF) and angiopoietin-1 (Ang-1) [96].

Endothelial barrier function damage and microvascular leakage

In addition to the impaired vasodilatory function described above, loss of barrier function is another key reason for the occurrence of MVL. The endothelial barrier function is made up by the glycocalyx, endothelial cells, pericytes and junction proteins [16]. Under physiological conditions, junction structures and myofilament–myosin skeleton fibers between CMECs determine the permeability of endothelial barrier [97]. Junction structures are maintained by an intercellular tight junction, adherens junction, and gap junction, including occludin, claudin-1/5, zonula occludens protein-1 (ZO-1), VE–cadherin, and connexin 43 (Fig. 3) [48, 97–100]. After myocardial I/R injury, the production of large amounts of ROS promotes leukocyte adhesion and increased vascular wall permeability. In the presence of C5a-activated neutrophils [101], the cytoskeleton proteins and gap junctions of CMECs are reorganized and resulting in an increased permeability of CMECs, which are essential for the development of MVL [102]. VEGF, a highly specific pro-vascular endothelial cell growth factor, promotes increased vascular permeability, extracellular matrix degeneration, vascular endothelial cell migration, proliferation, and angiogenesis. It has been shown that ischemic/hypoxia could lead to upregulation of VEGF in diabetic mice through activation of hypoxia-inducible factor 1 (HIF-1) [103] and also can increase vascular permeability by inducing phosphorylation of tight junction proteins, such as occludin and ZO-1 [104, 105].

Fig. 3.

Simplified and incomplete scheme showing the molecular composition of endothelial junction complexes. Occludin, Claudin1/5, and ZO-1 constitute tight junction that are essential for endothelial barrier permeability. The central structural and functional component of adherens junction is VE–cadherin, which is uniquely expressed in ECs. Gap junction consist of connexin and allow the passage of molecules. Decrease in junction proteins and loosening of interendothelial cell junction structures due to ischemia/reperfusion are the main causes of leakage of blood components from microvasculature. ZO-1, zonula occludens protein-1

Angiopoietin (Ang)/Tie-2 is another pathway distinct from VEGF that has been clearly demonstrated to maintain vascular integrity and is important for the regulation of endothelial barrier function [106], especially Ang-1 and Ang-2. The former is a natural ligand of Tie-2 receptor and maintains the integrity of the microvascular endothelial barrier by regulating the phosphorylation of VE–cadherin [107]. In contrast, Ang-2 is an important biomarker of MI [108]. It has been shown to act as an antagonist to Ang-1 at the Tie-2 receptor [107], which can promote MVL by decreasing VE–cadherin [105, 109]. In addition, sphingosine 1 phosphate (S1P), a blood-derived lysophospholipid, stabilizes endothelial cell–cell junctions by interacting with the type 1 receptor (S1P1) in ECs [110]. The main effects are increase of VE–cadherin binding, enhancement of junction proteins, e.g., claudin 5 transport to the cell periphery [111], and promotion of phosphorylation of the gap junction protein connexin43 [112].

Numerous studies have confirmed the importance of interendothelial cell junction structures. Glycocalyx degradation contributes to the reduction of endothelial barrier function and edema formation, and tumor necrosis factor α (TNF-α) is an essential mediator of glycocalyx degradation. Glycocalyx degradation can promote leukocyte and platelet adhesion, which increases microthrombosis and thus aggravates MVO [16]. It is evident that downregulation of various endothelial cell junction proteins and degradation of glycocalyx can lead to loss of barrier function of CMECs, which in turn to promotes extravasation of leukocytes and blood cells and leakage of plasma proteins [48, 49, 52, 98, 105, 109, 113, 114]. In addition, hemorrhage-induced interstitial iron deposition can also induce an inflammatory response to exacerbate ischemia/reperfusion injury [52].

Coronary microembolization and microthrombosis

Microembolization and microthrombosis are also important in the development of MVD after I/R. Spontaneous or iatrogenic (PCI-induced) rupture of epicardial atherosclerotic plaques can release plaque debris into the microvascular along with many soluble substances [52]. These substances, which cannot be captured by filter devices or meshes, directly contribute to the physical obstruction of microvessels (microembolization) and subsequent signaling or inflammatory cascades [115]. In addition, damaged ECs are dislodged from the microvessel wall and enter the circulation during MI. Apoptotic or necrotic circulating ECs can promote microvascular inflammatory responses by interacting with the endothelium [10], play an active role in pro-coagulant and immune processes, and have been suggested to be a biomarker of thrombotic risk [116]. The levels of circulating ECs were 4.5-fold higher in patients with MI than in healthy subjects [117]. Detached ECs after I/R also induce phosphatidylserine exposure and cytoskeletal protein remodeling, and then release endothelial-derived microparticles (EMPs, in size of 100 nm-1 μm) [118, 119]. It was demonstrated that EMPs downregulated the expression of tight junction proteins, such as ZO-1 and claudin 5, and upregulated the expression of transcellular transport-related proteins, such as caveolin-1. Moreover, EMPs promoted the production of platelet-derived microparticles (PMPs), and both synergistically mediated disruption of endothelial integrity and subsequent MVL after I/R [120]. Therefore, the number of circulating ECs and EMPs may serve as potential biomarkers of acute and chronic endothelial dysfunction [10].

After I/R injury, it is worth mentioning that the linear morphology of erythrocytes is concomitantly altered and promotes their aggregation in the microvasculature [37]. The platelet surface contains various biomolecules, such as GPIbIX-V, αIIbβ3, ICAM-1, p-selectin and α2β1 integrin [121]. On one hand, these molecules facilitate platelets to bind to the corresponding receptors or collagen on the surface of CMECs, including von Willebrand Factor (vWF), αVβ3 integrin and p-selectin glycoprotein ligand (PSGL-1). On the other hand, neutrophils can be activated and rapidly influx between CMECs and vascular elastic lamina [122], releasing neutrophil extracellular traps (NETs) [123] after reperfusion. This process leads to increased ROS production and a series of inflammatory cascades by enhancing leukocyte–platelet–CMECs adhesion/aggregation, promoting CMECs to highly express a series of inflammatory molecules, such as intercellular adhesion molecule (ICAM), vascular adhesion molecule (VCAM), p-selectin, thromboxane A2, and ET-1 [124, 125]. Rodrigues et al. specifically described in their study that about 25% of platelets bind directly to CMECs, while the remaining 75% of adherent platelets attach to leukocytes (mainly neutrophils) bound to the vascular wall, resulting in blood stasis [114]. Moreover, neutrophils have been shown to increase vascular permeability and promote MVL by releasing matrix metalloprotein-9 [126]. These findings illustrate that I/R leads to MVO and MVL via microembolization and microthrombosis [126].

Calcium overload in microvascular endothelial cells

Research evidence have showed that calcium overload is an early sign of cell apoptosis and necrosis after myocardial I/R injury [127]. Persistent calcium oscillation triggering mitochondrial dysfunction which leads to oxidative stress and the opening of the mitochondrial permeability transition pore (mPTP), and then activates mitochondrion-dependent apoptosis [128]. In a mouse cardiac I/R model, restriction of CMECs mitochondrial calcium ([Ca²⁺]_m) influx increased microvascular perfusion, reduced MVO and inflammatory cell infiltration and limited infarct size, which associated with mitochondrial morphological and functional protection [85, 129]. The [Ca²⁺]_m influx is primarily mediated by the mitochondrial calcium uniporter (MCU) complex, which consists mainly of four MCU subunits and essential MCU regulators (EMREs) [130, 131] The regulatory proteins that interact with MCU subunits include mitochondrial calcium uptake 1/2 (MICU1/2), which is responsible for mitochondrial calcium uptake [84], while the MCUb is a paralog of MCU, characterized as dominant negative subunit of uniporter complex, and acts as a negative regulator of MCU activity [132]. The four MCU subunits are upregulated in response to I/R injury, causing [Ca²⁺]_m overload and mitochondrial dysfunction [133]. Another critical intracellular Ca²⁺ regulator, sarco/endoplasmic reticulum Ca²⁺-ATPase (SERCA), catalyzes Ca²⁺ transport across membranes in an ATP-dependent manner [134]. Among more than 10 known isoforms of SERCA, the 110 kDa monomeric SERCA2a is the major isoform expressed in adult cardiac muscle, accounting for approximately 40% of the sarcoplasmic reticulum (SR) protein [135]. SERCA2a performs two major functions: it decreases cytoplasmic Ca²⁺ levels to initiate muscle relaxation and reloads Ca²⁺ required for muscle contraction into SR [135, 136]. In a recent study, cardiac-specific overexpression of SERCA in mice significantly reduced [Ca²⁺]_m overload and myocyte apoptosis and improved coronary microvascular function following cardiac I/R injury by preventing MCU upregulation [129]. However, the detailed mechanisms by which the MCU complex regulates mitochondrial function and microvascular protection have not been fully explored. It was recently reported that histidine triad nucleotide-binding 2 (HINT2), located in mitochondria and belonging to the HIT superfamily, could improve CMECs mitochondrial dynamics and MCU complex-associated [Ca²⁺]_m homeostasis by regulating the MCU complex [137], and thereby attenuating I/R-induced myocardial MVD [84].

Endoplasmic reticulum (ER) stress

ER is an important organelle for Ca²⁺ storage, lipid synthesis, protein translation, modification and folding in eukaryotic cells [138]. I/R injury leads to oxidative stress and increased intracellular ROS in ECs, causing unfolded or misfolded proteins to accumulate and aggregate within the ER, known as endoplasmic reticulum stress (ERS) [139]. ERS is generally considered to be the initial response or co-initial pathway that precedes other cellular stresses, such as mitochondrial oxidative stress. ERS is a cellular protective response that reduces the accumulation of unfolded proteins and restores the normal function of the ER. However, persistent ERS can shift its pro-survival mechanism to a pro-apoptotic pathway, leading to apoptosis [139]. Battson et al. provides a systematic review of ERS and the development of endothelial cell dysfunction, including the activation of protein kinase RNA-like endoplasmic reticulum kinase (PERK), inositol requiring enzyme 1 (IRE1), and activating transcription factor 6 (ATF6), as well as up-regulation of ER chaperon protein 78 kDa glucose-regulated protein (GRP78), which plays a role in regulating multiple aspects of protein synthesis, degradation, oxidative stress, inflammation and apoptosis [140]. It is noteworthy that mitochondria are an important source of ROS generation during ERS. ER Ca²⁺ released by cells under ERS is taken up by the mitochondria and ultimately results in cyt-c release and ROS generation [141, 142]. This functional relationship is mediated by an anatomical connection between two organelles called the mitochondria-associated ER membrane (MAM). The MAM is essential for the inter-organelle exchange of Ca²⁺, ATP, and ROS, as well as for coordinating cellular functions [140, 143]. He et al. demonstrated that Ca²⁺ accumulated in the cytoplasm after I/R transfers along the MAM ultimately leading to the opening of mPTP, which induces endothelial cell death [144].

Mitochondrial dysfunction

Mitochondria are involved in endothelial growth, proliferation and senescence [145]. In addition to CMECs dysfunction owing to [Ca²⁺]_m influx addressed above, a number of studies have identified mitochondrial damage as a central causative factor of MVD caused by cardiac I/R injury [12, 146, 147]. Mitochondria are dynamic organelles with self-regulations via fission and fusion, two diametrical processes, one separating the damaged part from the whole mitochondrion and another merging the two divided bodies into a healthy mitochondrion. The balance of fission and fusion is essential for maintaining mitochondrial and endothelial function [37, 148]. It is now believed that structural disorders of mitochondria precede their dysfunction [26]. When myocardial blood flow was restricted, the energy metabolism of CMECs was reduced, so mitochondrial fission was activated to increase the number of mitochondria to meet cellular metabolic demands, and mitochondrial fusion was inhibited instead [149]. At the molecular level, mitochondrial fission is a process of phosphorylation of dynamin-related protein 1 (Drp1) and subsequent recruitment by mitochondrial receptors, including mitochondrial fission factor (Mff) and mitochondrial fission protein 1 (Fis1) [150]. Studies have shown that I/R injury led to phosphorylation of Drp1 and Fis1, but had no effect on phosphorylation of Mff [26]. Phosphorylation of Drp1 accelerated its transfer from the cytoplasm to the mitochondrial surface [151], and phosphorylation of Fis1 increased its affinity with cytoplasmic Drp1 [152].

This pathological mitochondrial fission also damages mitochondrial DNA (mtDNA) during mitosis, decreases mitochondrial membrane potential (MMP), further exacerbates the accumulation of mitochondrial ROS (mROS) and the subsequently increases in cytoplasmic ROS [84, 153]. It is generally recognized that mitochondrial complexes I and III are the main sources of ROS generation under pathological conditions, and NOXs are the main enzymes for mROS production in ECs. Besides the vasoconstrictor function described above, NOXs-derived ROS promote apoptosis, inflammation, and senescence, which in turn lead to coronary MVD [154]. Its main downstream targets in ECs include NF-κB, activated protein-1 (AP1), HIF-1, p53, p38, c-Jun N-terminal kinase, and Src kinases [155].

Notably, antioxidant molecules such as glutathione (GSH), superoxide dismutase (SOD), and glutathione peroxidase (GPX) are significantly depleted, whereas ROS are unbalancedly increased in CMECs during I/R injury [10]. CMECs suffer from a complete loss of MMP and an increased opening of mPTP [37], which ultimately leads to oxidative phosphorylation uncoupling, ATP depletion or cell death [18, 156]. Moreover, when mitochondrial pro-apoptotic proteins such as cytochrome c (cyt-c) and second mitochondria-derived activator of caspases (Smac) entered the cytosol, they also activate members of the caspase family to promote apoptosis [26]. In this case, mitochondria switches from generators to endothelial apoptosis inducers [157], and CMECs become dysfunctional or died, eventually leading to microvascular swelling and vessel wall collapse, further blocking the blood flow to the ischemic myocardium [158, 159]. Such a pattern of cell death leads to many types of CVD, including I/R injury and heart failure [84, 160].

Impaired mitophagy in microvascular endothelial cells

The benefits of mitophagy in CMECs have been recognized [157, 161, 162]. Compared to aforementioned imbalance of mitochondrial fission/fusion provoked CMECs apoptosis [157, 163–165], mitophagy is an important protective mechanism that restores mitochondrial fission/fusion balance and thus maintains mitochondrial function and structure in CMECs after I/R injury [84, 166]. In general, the enhancement of mitochondrial fission mediated by phosphorylation of Drp1 is accompanied by increased mitophagy and mitochondrial biogenesis. Under normal physiological conditions, PTEN-induced putative kinase 1 (PINK1) is imported into mitochondria, where it is cleaved by proteases [167]. Low membrane potential mitochondrial fragments are stabilized by PINK1 on the outer mitochondrial membrane (OMM) due to inhibitory input, which in turn recruits E3 ubiquitin ligase (Parkin) to mitochondria. Meanwhile, the mitochondrial outer membrane localized mitophagy receptor FUN14 domain-containing protein 1 (FUNDC1) is activated by post-transcriptional phosphorylation. Mitochondria interacts with the autophagy marker microtubule-associated protein 1 light chain 3 (LC3) on lysosomes via Parkin and FUNDC1, forming autophagosomes. Therefore, mitophagy promotes the degradation of mitochondrial debris and maintained the dynamic homeostasis of mitochondria [10]. It has been shown that mitophagy not only attenuated high glucose-induced CMECs injury by inhibiting oxidative responses [168], but also inhibited oxidative LDL-induced [169] and/or oxidative stress-induced CMECs apoptosis [29, 159]. During I/R injury, the expression of pro-inflammatory factors such as IL-8, matrix metalloprotein-9 and monocyte chemotactic protein 1 (MCP-1) was increased in CMECs, which could be reversed by mitophagy [159]. This suggests that under hypoxic conditions, the MMP is reduced and Parkin in CMECs translocates to mitochondria, which can induce mitophagy [170] and may have an anti-inflammatory effect. It is a proposed hypoxic preconditioning mechanism.

Several studies have found that reoxidation after hypoxia inhibits mitophagy, leading to the accumulation of damaged mitochondria and mediating mitochondrial apoptosis in CMECs [171] and cardiomyocytes [172]. NR4A1 is a member of the nuclear receptor subfamily 4 group A (NR4A) family, which exerts transcriptional regulation to participate in life activities, such as energy metabolism and protein interactions [159]; Casein kinase 2α (CK2α), a constitutive Ser/Thr kinase [173], was originally described as an inhibitor of FUNDC1 and is associated with FUNDC1 phosphorylation at Ser13 [174]. In recent studies, significant upregulation of NR4A1 [159] and CK2α [175] has been observed in murine myocardial I/R model. Their expression levels were positively correlated with microvascular collapse, endothelial apoptosis and mitochondrial damage [159], as well as further cardiomyocyte death, infarct area enlargement and cardiac dysfunction [175], whereas deletion of these two genes mitigated I/R induced cardiac injury through maintaining FUNDC1-mediated mitophagy [159, 175]. NR4A1 has the ability to simultaneously promote Mff-induced mitochondrial fission and FUNDC1-required mitophagy, which can disrupt the dynamic homeostasis of mitochondria, promote endothelial cell apoptosis and trigger microvascular dysfunction [159], while the latter enhances the phosphorylation (inactivation) of FUNDC1 through the post-transcriptionally modified Ser13 site, thus effectively inhibiting mitophagy [175]. As a result, impaired mitophagy is unable to clear damaged mitochondrial debris, which leads to a series of pathophysiological changes, such as mitochondrial genome collapse, blocked mitochondrial biogenesis, oxidative stress, mPTP opening and mitochondrial debris accumulation, ultimately to CMECs injury and microvascular dysfunction [159, 175]. Thus, it is generally accepted that ischemia/hypoxia induces mitophagy to attenuate mitochondrial damage, whereas ischemia–reperfusion/hypoxia–reoxygenation inhibits mitophagy, which in turn leads to mitochondrial dysfunction and cell death [10].

In summary, these studies have demonstrated the latest research findings on the mechanisms of MVD development after myocardial I/R, especially the regulation of the balance between vasodilators and vasoconstrictors; CMECs swelling and loss of vascular barrier function; mitochondrial metabolism and dysfunction within CMECs. Among these mechanisms, microembolization, microthrombosis, inflammatory cell infiltration, energy metabolism and dysfunction of CMECs are relatively well-elucidated; while aspects such as loosening of junctions between CMECs, [Ca²⁺]_m overload, ERS and impaired mitophagy need to be further explored, particularly as the upstream and/or downstream signals of the known signaling pathways are not fully defined. MVD after myocardial I/R is an intricate process involving multifaceted and multifactorial interactions at molecular, cellular and tissue levels (Fig. 4). A better understanding of the pathophysiological mechanisms of MVD may pave a path for more precise management and improvement of the prognosis of myocardial I/R injury and reduction of CVD mortality.

Fig. 4.

A Complex process of multifaceted and multifactorial interactions of microvascular damage in molecules, cellular and tissue levels after myocardial ischemia/reperfusion (I/R). After I/R injury, endothelial cells swelling, vasodilation dysfunction, and microthrombosis cause luminal narrowing, known as microvascular obstruction (MVO), whereas loosening of the interendothelial cell junction structures, massive death of endothelial cells and increased vascular permeability cause microvascular leakage (MVL). B Mitochondrial dysfunction leads to endothelial cells apoptosis. Due to I/R injury, mitochondrial fission/fusion was imbalanced and pathological mitochondrial fission damages mtDNA, promoting cytochrome c (cyt-c) release from mitochondria to the cytoplasm and activating the caspase family to promote endothelial cells apoptosis. Moreover, impaired mitophagy is unable to remove damaged mitochondrial debris, leading to a series of pathophysiological changes that eventually lead to endothelial cells injury and apoptosis. p-eNOS phosphorylated endothelial nitric oxide synthase, ET-1 endothelin-1, vWF von Willebrand Factor, PSGL-1 p-selectin glycoprotein ligand, ICAM intercellular adhesion molecules, CECs circulating endothelial cells, EMPs endothelial-derived microparticles, Ang-2 angiopoietin-2, Tie-2 TEK receptor tyrosine kinase a receptor of Ang-2, VEGF vascular endothelial growth factor, VEGFR vascular endothelial growth factor receptor, PINK1 PTEN-induced putative kinase, Parkin E3 ubiquitin ligase, FUNDC1 FUN14 domain-containing protein 1, LC3 light chain 3, Drp1 dynamin-related protein 1, Mff mitochondrial fission factor, Fis1 mitochondrial fission protein 1, mROS mitochondrial ROS, MCU the mitochondrial calcium uniporter complex

Detections of microvascular damage after myocardial I/R injury

Experimental assays of microvascular damage after myocardial I/R injury

Detection of microvascular obstruction

The thioflavin-S and Evans blue dual staining is a common method used in experimental animal studies to determine the extent of MVO after I/R surgery [41, 50, 176]. The brief experimental procedure is as followings: first, after establishment of mouse myocardial I/R model, the thorax is reopened at the end of reperfusion and the left coronary artery is re-ligated at the same position of the occlusion, and then 4% thioflavin-S fluorescent dye via the inferior vena cava injected immediately, and subsequently perfused with 3% Evans blue retrogradely via the ascending aorta to stain the non-ischemic myocardium. Second, the heart is harvested, frozen on dry ice, the LV is serially sectioned at a thickness of 1 mm and then compressed between two slides. Finally, LV sections are exposed to UV light and digital image acquired. The area of ischemic region (area at risk, AAR, Evans blue non-stained region) and MVO region (thioflavin-S fluorescence diminished or absent region) were determined and the area of MVO was expressed as a percentage of AAR [50, 176–179] (Fig. 5A).

Fig. 5.

Experimental assays of microvascular damage after myocardial I/R injury. A Extent of MVO by the absence of thioflavin-S fluorescence (indicated by the thin yellow dotted line) within the AAR (indicated by the thick yellow dotted line) and NIZ by EB. B Left image detects MVL by EB within the AAR (indicated by the yellow dotted line) and NIZ by picrosirus red (PSR). The right image determines moderate MVL (yellow filled area) and severe MVL (red filled area). C Representative serial CMR images from the apex to the base of a mouse heart showing LGE (arrows) and serial cardiac transverse sections showing regions with MVL indicated by EB staining after I/R (1 h/24 h) (modified from [50])

High-precision imaging techniques include non-invasive echocardiograph and CMR imaging are promising ways for detecting MVO in live animals. Ultra-high-frequency (UHF) cardiac ultrasound systems can be used to study myocardial local perfusion by injecting ultrasound microbubbles in combination with a laser device. This technique has been developed for the popular Vevo 3100 high-frequency cardiac ultrasound system. However, the extremely transient presence of microbubbles in the myocardium, the high technical requirements, individual and operational errors, and the immaturity of fast three-dimensional imaging technologies have limited the application of this technique for the detection of MVO. With the development of small animal cardiac MRI technology, it is expected that a method to dynamically detect the extent of MVO in live animals by detecting the size of the "no-reflow" region will be achieved in the future.

Histological measurement of microvascular leakage

In experimental animal studies, MVL is mainly measured by tail vein injection of Evans blue as an MVL tracer to measure the size of myocardial blue-stained areas or by extracting Evans blue from the myocardium for chemical quantification [50, 179]. Briefly, following the establishment of cardiac I/R model in mice, Evans blue is injected via the tail vein to infiltrate through the damaged micro-vessel into the myocardial interstitium. After a period of time (several hours) the thorax is re-opened and the aorta canulated and retrogradely perfused with saline to flush away blood containing Evans blue in the cardiac vasculature. Next the LV is isolated and serially sliced for digital imaging. Total blue-stained areas and dark blue (severe MVL) or light blue-stained areas (light–moderate MVL) can be measured using image analysis software and expressed as a percentage of the blue-stained area to total LV section area [50] (Fig. 5B).

Quantitative detection of microvascular leakage

Quantitative chemical assay can be conducted by extracting Evans blue dye from blue-stained myocardium. Briefly, following the digital imaging, blue-stained portion of the LV sections is separated, minced and digested in 50% trichloroacetic acid (TCA) to extract the blue dye. The absorbance of Evans blue in the supernatant of digested tissue after centrifugation is measured at 620 nm using a spectrophotometer and calculated values are normalized by the wet weight of the myocardial tissue to quantify the Evans blue content in certain amount of myocardium (μg/mg) [50, 177, 178].

CMR detection of microvascular leakage

CMR imaging can be used for detecting MVL. We have previously documented the usefulness of this technique in small size animals [50]. Following the establishment of cardiac I/R model in mice, Evans blue as a tracer and gadolinium-based contrast agents were injected intravenously and CMR imaging for the isolated LV was performed using a 9.4 T Agilent animal scanner to assess myocardial MVL within couple of hours after the euthanasia. Microvascular hyperpermeability increases contrast agent retention within the ischemic tissue, that is, late gadolinium enhancement (LGE) in the acute phase of MI. Results surprisingly demonstrated that the area of MVL, i.e., LGE area was larger than both infarct size and no-reflow areas. Contour analysis of CMR–LGE images and histological MVL images showed a high degree of spatial co-localization (r = 0.959, P < 0.0001), indicating the accuracy of CMR imaging in the detection of MVL [50]. Notably, the size of the MVL detected by CMR imaging was highly correlated with the degree of LV enlargement and reduced ejection fraction, suggesting that MVL size could be used to predict LV function and remodeling [50] (Fig. 5C). Currently, the CMR technique for MVL is only performed in the static heart, the application in a live animal needs to be further developed.

Clinical diagnoses of MVD after myocardial I/R injury

Electrocardiogram (ECG)

ECG is a widely used, convenient, inexpensive tool for indirect assessment of MVD in clinic. Previous studies reported that by comparison the difference of ST-segment elevation from abnormal leads of ECG before and after PCI in patients with STEMI to assess microcirculatory status [180]. ST-segment regression (STR) is one of the highly accurate parameter for assessment of the effect of PCI, and at least 50% of patients with STEMI have a peri-procedural STR < 50%, indicating that microcirculatory perfusion may be impaired [181]. The great advantage of perioperative ECG analysis is that it can be performed in real time during the procedure, thus allowing for immediate and new interventions and so having an important clinical value [181]. Therefore, STR is a useful tool to monitor the response to treatments, such as fibrinolysis or PCI in STEMI patients, both in the acute phase and during follow-up.

Coronary artery angiography (CAG)

Coronary artery angiography (CAG) can be applied to evaluate the opening status of a diseased coronary artery received intervention. The Thrombolysis in Myocardial Infarction (TIMI) flow grade is a direct indicator of coronary flow after intracoronary thrombolysis or intervention in AMI with four grades of 0–3 [182, 183]. Grade 0 indicates no distal reperfusion, i.e., no passage of the distal contrast agents or complete occlusion of the culprit coronary artery. Grade 1 designates that a small amount of contrast agents passes through the obstruction but the distal artery is not visualized. Grade 2 denotes that the culprit coronary artery is completely visualized but the blood flow is slow compared to normal vessels. Grade 3 means that the culprit coronary artery is completely visualized with a normal distal perfusion [184, 185]. There are clinical studies reported a close association between TIMI flow grade and the risk of major adverse cardiovascular events (MACE) [186]. Mortality was significantly higher in patients with TIMI flow grade 0/1 than in patients with TIMI flow grade 2/3 following interventions [187]. Furthermore, the initial TIMI flow grade before intervention was also an independent predictor of infarct size, MVO, myocardial salvage index (MSI) and the risk of clinical events [188]. Although TIMI flow grade can evaluate degree of coronary artery flow, its ability to accurately reflect status of microvascular perfusion has been challenged, as nearly half of patients with TIMI flow grade 3 present MVD detected by CMR [186]. In practice, TIMI flow grade is somewhat subjective, so an alternative and more objective angiographic method has been established. Since the left anterior descending artery is longer than the circumflex and right coronary arteries, a correction factor has been generated for compensation. Usually, the number of frames from beginning of contrast agent coloring to their passage through the left anterior descending branch is divided by 1.7 for a correction of the vessel length, therefore, obtaining a corrected TIMI frame count (CTFC). CTFC expresses the time required for epicardial blood flow to the distal end of the damaged micro-vessel as a number of frames from 0 to 100 (when the angiographic rate is 30 frames/second), allowing for a more accurate depiction of flow rate [189].

Real-time myocardial contrast echocardiography (RTMCE)

Real-time myocardial contrast echocardiography (RTMCE) is a relatively new technique that utilizes microbubbles smaller than the diameter of erythrocyte to simulate erythrocyte rheology within a vessel [190], thereby displaying the entire myocardial microvascular perfusion, RTMCE is expected to quantify in real time the total volume of blood and estimate the velocity of blood flow through a localized region of the myocardium, including small myocardial arteries, capillaries, and small veins [191]. Specifically, by standard echocardiography the ventricular wall motion analysis can be conducted using the American Society of Echocardiography 17-segment LV model. RTMCE is performed in the setting of tissue harmonic imaging in three standard apical views (four-chamber, two-chamber, and three-chamber) with a continuous intravenous infusion of lipid-coated microbubbles as the contrast agent, and real-time images acquired at a mechanical index of 0.15–0.19 [192]. In clinical practice, normal microvascular perfusion (MVP) can be defined if all segments of myocardium supplied by the infarct-related artery are completely replenished within 4 s after a high mechanical index pulse. Delayed microvascular perfusion (dMVP) is defined if a perfusion defect was still observed in the infarct area within 4 s but was completely replenished within 10 s. If the plateau strength is reached and two or more parts of the infarct zone still have persistent defects, it is defined as MVO [192]. A recent RTMCE study reported that MVP, dMVP and persistent MVO were detected in 36%, 29% and 35% STEMI patients after successful PCI, respectively [192]. In fact, more than 60% of STEMI patients received emergency PCI still have different degrees of microvascular perfusion abnormality after 24–48 h, which was associated with poor recovery of LV systolic function at the 6-month follow-up [192]. Analyses from RTMCE studies indicate a superiority of this technique over clinical biomarkers, ECG and coronary angiographic parameters in assessment of microvascular perfusion status and prediction of LV functional recovery after PCI [190, 192, 193].

Cardiac magnetic resonance (CMR) imaging

CMR imaging technology is now commonly used in the diagnosis of CVD. With good soft tissue contrast resolution, a large scanning field of view, and cross-sectional images in all directions and different angles, it allows for unique in vivo assessment of cardiovascular abnormalities in STEMI patients [194, 195]. Currently, CMR is considered the best diagnostic imaging modality to detect MVD, including MVO, myocardial edema, and IMH [196–198].

From a technical point of view, cardiac MRI characterizes different tissues by the evaluation of proton relaxation times T1 and T2 [198]. The T1 value expresses the rapidity of the longitudinal relaxation of the tissue, while the T2 value describes the rapidity of the transverse relaxation of the tissue. Since the main magnetic field cannot reach absolute homogeneity, the transverse magnetization decays more rapidly, yielding a smaller value than T2, the so-called T2* value [198, 199]. Increased myocardial water content raises the signal on T2-weighted images, such as inflammation. Myocardial edema at the acute phase of MI shows up as a bright signal on T2-weighted images [200–202]. The main advantage of this technique is to quantify the proportion of myocardial salvage assessed retrospectively by comparing T2-weighted edema size with late-enhanced images [203]. Gadolinium (Gd) is a metal element with magnetic resonance paramagnetism and is often used as an extracellular contrast agent to improve the quality of images produced by CMR, which exhibits significantly high signal under specific conditions, such as myocardial necrosis or fibrosis, and conversely low signal in surviving myocardium [204, 205]. LGE images are T1-weighted inversion recovery sequences acquired 10 min after intravenous Gd injection, and inversion time is selected to null myocardial signal using “inversion time scout” or “look locker” sequences [206]. In I/R injury, increasing micro-vessel permeability causes myocardial edema and expansion of myocardial interstitial volume, resulting in spread distribution of Gd contrast agent in these areas, which manifests as abnormal enhancement in the infarcted regions. As the disease progresses, the damaged myocardial cells are replaced by extracellular collagen fibers and fibrosis, which leads to an expansion of the interstitial volume and prolonged Gd contrast agent influx/clearance time, so that the contrast agent appears concentrated in the fibrous scar areas, the so-called LGE [207], which has also been validated in animal models [208]. The LGE–CMR technique is now widely used to assess disease severity, risk stratification, prognosis, and the effectiveness of therapeutic interventions [209–212].

MVO can be evaluated using CMR first-pass perfusion and delayed post-contrast sequences [213]. On contrast-enhanced CMR images, MVO appears as a dark hypointense core within the hyper-enhanced areas on either early gadolinium enhancement (referred to as early MVO) or conventional late gadolinium enhancement (referred to as late MVO) [45]. Clinically, early and late MVO are generally assessed at approximately 1 and 15 min after gadolinium injection, respectively. Since the presence and extent of MVO decreases over time, early MVO is more sensitive to less pronounced MVD, which bears a low prognostic value in predicting post-infarction adverse events. Conversely, late MVO reflects severe myocardial arterial circulation disturbances and correlates closely with clinical outcome [45, 214]. In one of the earliest experimental studies, Judd et al. investigated the correlation of MRI findings with pathological changes by applying contrast-enhanced fast MRI to a dog cardiac I/R (90 min/2 days) model, and subsequently removed the heart and delineated the no-reflow zone by thioflavin-S staining. They found that a hypo-enhancement zone within the bright regions of LGE on delayed post-contrast sequences, so-called “dark zones”, were “no-reflow” areas [208]. In humans, Lima et al. performed rapid enhancement CMR on patients with recent MI and observed a well-defined time-intensity profile of the enhancement in different regions of MI, which was described based on the wash-in/wash-out kinetics of the contrast agent [215]. In addition, persistent hypo-enhancement (10–20 min after injection of a contrast agent) on delayed enhancement sequences was also found to appear to be characteristic of persistent MVO [216]. Recently, Li et al. and Alessandro et al. performed LGE–CMR in STEMI patients after 2–7 days of PCI and confirmed that the MVO was defined as a hypo-enhanced region within LGE area, which occurred in 41.8–67% STEMI patients [59, 79].

Except MVO, CMR can only indirectly assess MVL by detecting/measuring degree and extent of myocardial edema and IMH due to lack of proper indicator/tracer for MVL. As stated above, T2-weighted images can be used to assess myocardial edema and increased active water content within the ischemic myocardium, which appears as a bright signal on the images [198, 217, 218]. IMH can be detected by CMR as the hemoglobin catabolites are paramagnetic and IMH may attenuate the T2-weighted signal intensity of the edematous area [16, 52, 219]. Pankaj et al. performed CMR and obtained extracellular volume (ECV)-maps at the acute (24–72 h) and recovery (3 months) phases in STEMI patients who underwent PCI within 12 h. Adapted from the 17 segments of the American Heart Association (AHA) model [199], after excluding the apex, 16 segments of each patient's LV could be classified as normal segments (no edema on T2-weighted imaging and no infarction on LGE imaging); edema segments (predominantly edema on T2-weighted imaging and no infarction on LGE imaging) and infarct segments (presence of any infarction on LGE imaging with or without edema on T2-weighted imaging) [82]. The changes in myocardial extracellular volume and ventricular wall thickening were analyzed in a total of 800 segments from 50 patients during the acute phase. Among them, 325 segments (40.6%) were found to be normal, 246 segments (30.8%) were edematous, and 229 segments (28.6%) were infarcted [82]. At the recovery phase, it was found that normal myocardial extracellular volume mildly expanded and that edematous segments were associated with deterioration of cardiac systolic function through comparison of ECV-maps at baseline and follow-up scans [82].

IMH, owing to endothelial barrier disruption and blood cells leakage, is another indirect parameter can be applied to assess MVL after ischemic insult [69]. Clinical studies have shown that T2* imaging is the preferred CMR method for comprehensive assessment of IMH [69]. The specific imaging for IMH is based on dephasing of T2* relaxation times [81]. The area of IMH after perfusion is presented as a hypointense infarct core with a T2* value < 20 ms [81, 220]. Kandler et al. performed CMR in 151 STEMI patients within 8 days after PCI. LGE sequences were used to visualize MVO and T2*-weighted sequences were used to detect IMH. They found that 50% patients presented with IMH and those with a TIMI flow grade ≤ 1 before PCI were more likely to have IMH, which was associated with impaired LV function and increased myocardial infarct size [69].

With the development of new indicators/tracers, it is expected that the degree of MVL will be accurately reflected by CMR techniques, as shown in our experimental studies [50], which providing novel methods and tools for a direct diagnosis and assessment of MVL in clinic.

In summary, MVD after myocardial I/R injury can be detected clinically by ECG, CAG, RTMCE [126]. CMR allows visualization of MVO, myocardial edema, and IMH with LGE imaging and T2* mapping. Although MVL can be indirectly reflected by edema and IMH, a direct method of measuring MVL has not yet been developed. As mentioned in Sectinon Experimental assays of microvascular damage after myocardial I/R injury, our research team, for the first time, characterized MVL in a mouse model of cardiac I/R through a series of histological, biochemical/semi-chemical quantitative assays and CMR imaging. This is currently the only way to detect MVL straightforwardly in an animal model. However, in clinical practice, factors such as the high technique demand in ultrasound technology and the cost of expensive CMRs have limited the clinical performance of many of these examinations. These reasons have led to the underestimation of MVD and especially MVL in a significant proportion of patients in the clinical setting.

Potential therapeutic approaches for MVD after myocardial I/R injury

There have been numerous clinical studies evaluating the effects of several interventions on infarct size and cardiomyocyte protection, but few preclinical studies have explored potential therapeutic approaches for MVD [45]. Currently, some therapeutic modalities have been shown to ameliorate MVD due to endothelial cell dysfunction after I/R in animal models or at the endothelial cell level, but the mechanisms still need to be further explored and the therapeutic efficacy also requires a clinical validation.

Mechanical means

Transient/repeated myocardial ischemic stimuli before sustained ischemia (ischemic preconditioning, IPC) or at the beginning of reperfusion following ischemia (ischemic postconditioning, IPostC), or even carried on such intervention in the distal region or distal organs, e.g., kidneys, skeletal muscle (remote ischemic conditioning, RIC), can reduce myocardial infarct size [221]. These mechanical interventions augmenting resistance of the heart or other targeted organs to severe ischemic insult is named as ischemic conditioning. Heusch systematically reviewed the role of ischemic conditioning in I/R and related mechanisms, including physical stimuli, chemical stimuli, and intracellular signaling, especially in mitochondria [222].

IPC has been shown to have clinical benefit on perioperative outcomes in 22 clinical trials comprising 933 patients [221]. It has also been confirmed in animal models can activate hypoxic signaling pathways, resist inflammation, prevent oxidative stress, which in turn promotes endogenous neuroprotection [223], attenuates renal I/R injury [224], facilitates flap survival after skin surgery [225], and protects cardiac function [226]. In studies on ECs, IPC has been mainly shown to counteract MVD by regulating NO and ET levels, protecting hepatic microvascular endothelial barrier function and blocking leukocyte adhesion [227]. Lin et al. also found that IPC reduces neurovascular damage after cerebral I/R by inhibiting neuronal and vascular endothelial cell apoptosis [228].

Due to the clinical unpredictability of AMI, the application of IPostC is more suitable for clinical application [229]. IPostC exerts neuroprotective effects by modulating central and peripheral glutamate in a cerebral I/R model [230], and acts by altering autophagy activity and expression of Beclin-1 in a time-dependent manner in a rat cardiac I/R model [231]. Albeit cardioprotection of IPostC has been confirmed in animal studies [229], but its translation to the clinic has not been promising, with contradictory results of " improved [232] and not improved [71]" on the effect of IPostC on endpoint events in different studies.

Compared to the limitation of IPC and IPostC in clinical practice, RIC has the advantage that distal regions such as the extremities are more ischemia-tolerant, easier to operate with lower risk to cause new I/R injury [233]. RIC has been found to protect brain [234], kidney [235] and cardiac [236] following sustained I/R. Its cardioprotective effects have been suggested to be possibly mediated by microparticles [237, 238]. However, microparticles have been shown to be associated with MVD. Boulanger et al. reported that circulating microparticles in rats with AMI caused endothelial dysfunction and vasodilatory dysfunction by selectively inhibiting eNOS expression [239]. A clinical study also demonstrated that the level of microparticles in the coronary arteries was the marker of persistent microthrombosis after I/R and correlation between microparticles and indices of MVD (CAG and ST resolution) implicated a critical role of microparticles in mediating MVD [240]. However, there are still some evidences that RIC can reduce MVO or edema/MVL [241], and such benefits have also been confirmed in clinical investigations [242, 243]. Overall, IPC exerts strongest cardioprotection not only for cardiomyocytes but also for CMECs, while both IPostC and RIC show ambiguous results, further investigations of precise mechanisms of these mechanical means on MVD protection are needed.

Potential pharmacological therapies

Many pharmacological agents currently have been tested in experimental models/or clinical trials, but most focus on myocardial protection rather than microvascular or CMECs. In this part, we focus on medicines or agents possessing actions of microvascular protection including conventional clinical drugs and some potential agents that still need to be explored and developed.

Sodium glucose co-transporter 2 inhibition

Sodium glucose co-transporter 2 (SGLT2) inhibitors, including canagliflozin, dapagliflozin, and empagliflozin, are a class of medicines for type 2 diabetes treatment that reducing glucose reabsorption and increases urinary glucose excretion via inhibition of SGLT2 in the proximal tubules of the kidney, hereby decreasing plasma glucose levels [244, 245]. Recently, numerous clinical studies have demonstrated multiple cardioprotective effects of this type of drugs with reduced risk of nonfatal MI, heart failure and mortality from all causes [246–250], and these effects appear to be independent of glycemic control [251, 252].

SGLT2 inhibition yields conspicuous effects in microvascular protection in various pathological conditions. In a mouse cardiac I/R injury model, empagliflozin [26] and dapagliflozin [253] were administered within 7 days before surgery. Both empagliflozin and dapagliflozin treatment showed the beneficial effects in maintaining microvascular architecture, reducing luminal stenosis, preventing microthrombosis and attenuating the inflammatory response resulted from the endothelial hyperpermeability [26, 253]. In a rat model of diabetic cardiomyopathy, administration of dapagliflozin through drinking water for 8 consecutive weeks improved myocardial fibrosis by blocking endothelial-to-mesenchymal transition [254]. Moreover, empagliflozin treatment in heart failure patient with preserved ejection fraction improved endothelial and cardiomyocyte function thereby protect the heart [255].

Endothelial protection by SGLT2 inhibition has also been documented in ex vivo cell experiments [26, 253]. In human coronary artery endothelial cell (HCAECs) hypoxia and reoxygenation (H/R) culture model, it was found that addition of empagliflozin before H/R could inhibit Drp1 phosphorylation and DNA–PKCs-induced Fis1 phosphorylation, thereby attenuating mitochondrial fission and normalizing mitochondrial fission/fusion balance [26]. Moreover, addition of dapagliflozin before hypoxia improved endothelial barrier function, eNOS activity and angiogenic capacity and reduced H/R-induced apoptosis by preventing cofilin-dependent F-actin depolymerization and cytoskeletal degradation [253]. Furthermore, dapagliflozin pre-treatment also down-regulated xanthine oxidase activity induced by H/R, inhibited the oxidation and inactivation of SERCA2, and suppressed intracellular calcium overload, thereby effectively protecting cytoskeletal integrity and endothelial cell activity [253]. These findings suggest that SGLT2 inhibitors could be a potentially effective drug to improve MVD and deserve further experimental and clinical investigations.

Relaxin

Relaxin is a natural peptide hormone and is considered to be one of the most pleiotropic hormones in the human body, capable of inducing a wider range of biological effects in addition to its role in reproductive function [256–258]. Relaxin-like family peptide receptor 1 (RXFP1) is the primary receptor for relaxin [257, 259, 260]. Since major cardiac cells, including cardiomyocytes, fibroblasts and CMECs all express RXFP1, so the heart is recognized as an effector organ of relaxin [257, 260]. Relaxin has been proposed as a therapeutic strategy for a number of CVD, such as heart failure, atrial fibrillation (AF), acute MI, ischemic heart disease, and hypertension [261].

In a guinea pig model of smoking-induced cardiovascular damage, administration of recombinant human relaxin-2 (serelaxin) through the entire induction period reversed smoking-induced abnormalities, such as aortic vessel wall thickening and massive loss of endothelial cells [262]. In a model in which guinea pig aortic endothelial cells were exposed to cigarette smoke extract solution, serelaxin pretreatment could maintain ECs viability and eNOS expression, thereby reducing smoking-mediated vascular injury and dysfunction [262]. Moreover, in a mouse cardiac I/R model, serelaxin treatment immediately after coronary artery occlusion alleviated MVD including both MVO and MVL measured at the end of reperfusion with maintenance of capillary opening and inhibition of local inflammatory responses, such protection led to a smaller infarct size and improved cardiac function and remodeling [176]. Further cell experiment documented that serelaxin treatment significantly reduced permeability of microvascular endothelial cell monolayers induced by H/R intervention with the underlying mechanism of preservation of VE–cadherin expression. These findings suggest that relaxin may be as a microvessel protective agent for MVD treatment.

Traditional Chinese medicine

Salvia miltiorrhiza (named Danshen in Chinese) is an important natural medicine (Chinese herb) which improves circulation and blood stasis [263]. Due to the effect of coronary artery dilatation [264], it is widely used in the treatment of CVDs in Asian countries to prevent myocardial ischemia [265], myocardial infarction [266], improve microcirculation [267] and reduce myocardial oxygen consumption [268, 269]. Tanshinone IIA (Tan IIA) is mainly isolated from salvia miltiorrhiza and is a representative of the lipolytic components. In a mouse cardiac I/R model, Tan IIA possesses following beneficial effects: (1) reduced endothelial cell apoptosis; (2) preserved the phosphorylation of eNOS, reduced ET-1 levels and mitigated vasoconstriction; (3) inhibited ICAM1 expression; (4) downregulated transcriptional expression of inflammatory mediators, including matrix metalloprotein-9, TNFα, IL-6 and MCP-1, thereby maintaining microvascular structure, decreasing susceptibility to thrombosis; attenuating regional inflammatory response and protecting microvascular barrier function [12]. Such protection is essential to prevent leakage of blood components into myocardial interstitium. Similar results were also observed by Chen et al. in a rat model of chronic intermittent hypoxia by continuous 8-week intraperitoneal injection of Tan IIA (20 mg/kg/d) [270]. Moreover, in CMECs H/R challenge model, Tan IIA intervention exerted anti-oxidative action to prevent mitochondrial damage evidenced by preserving expression of Sirtunin 1 and PGC1α, scavenging mitochondrial ROS, maintaining MMP and reducing mPTP opening [12]. Tan IIA also inhibited pro-apoptotic enzyme, caspase-9 activity and limited mitochondrial cytochrome C release [12]. Thus, maintenance of mitochondrial homeostasis is an important protective mechanism of Tan IIA.

Scutellarin is a source of an herb called scutellaria altissima, which is commonly used clinically for the treatment of paralysis after cerebrovascular disease. In 2018, Mo et al. reported that scutellarin could antioxidize and prevent atherosclerosis through up-regulation of SOD and down-regulation of NOX4 [271]. In addition, scutellarin protects vascular ECs from hyperglycemia-induced injury by upregulating mitophagy through the PINK1/Parkin signaling pathway [272]. Another herbal medicine called tetrahydrocurcumin is the main bioactive component of curcuma longa. It has been shown to ameliorate mitochondrial dysfunction in brain cells and thereby reduce brain damage and MVL in vivo [273] and ex vivo cell experiments [274] in mice. Quercetin and procyanidin B2 are both flavonoids widely distributed in plants. The former showed endothelial protective effects in both HUVECs [275] and cerebrovascular ECs [276], and the main mechanisms included promoting endothelial cell proliferation, migration and angiogenesis, maintaining MMP, inhibiting ERS, and inhibiting apoptosis. Nie et al. found that procyanidin B2 attenuated endothelial cell ERS through PPARδ-dependent pathway [139].

FX-06

FX-06 (fibrin-derived peptide Bβ15-42) is a fibrin Bβ chain-derived peptide and is constitutively liberated by the plasmin degradation of cross-linked fibrin [277]. It has been shown to reduce infarct size, maintain the endothelial barrier, and improve hemodynamic changes and survival in rodent and porcine I/R models [278]. FX-06 was found to bind to VE–cadherin and inhibit leukocyte migration, thereby initiating VE–cadherin-mediated signaling. The mechanisms of stabilizing endothelial cell connectivity and barrier function include the following: (1) counteracting stress fiber formation by preventing myosin light chain phosphorylation, and thereby rebalancing the activation state of RhoA/Rac 1 to inhibit dissociation between endothelial cells and VE–cadherin; (2) competing with the binding of the fibrin E1 fragment to VE–cadherin and preventing the dissociation of intercellular VE–cadherin complexes and preventing leukocyte migration [278]; and (3) binding to the VLDL receptor separates Fyn from VE–cadherin and prevents endothelial cell contraction [31].

CU06-1004

CU06-1004 (formerly Sac-1004), a pseudosaccharide derivative of cholesterol, is also known as an endothelial dysfunction blocker. CU06-1004 has been shown to improve prognosis by inhibiting the reduction of MVL in a variety of diseases [279]. In a rat I/R model, CU06-1004 enhances the endothelial barrier through the cAMP/Rac/cortical actin pathway and increases the formation of cortical actin rings, thereby preventing VEGF-induced destabilization of actin filaments and ultimately decreasing I/R-induced disruption of the blood–brain barrier [280]. A recent study found that CU06-1004 also significantly inhibited astrocyte end-feet swelling after I/R by decreasing the levels of connexin 43 and aquaporin 4. In addition, CU06-1004 inhibited the degradation of β1-integrin and β-dystroglycan anchored to the cortical actin ring of ECs, as well as suppressed the loss of laminin and type IV collagen [279]. In a mouse model of diabetic retinopathy, it similarly prevented retinal vascular leakage [281]. These findings illustrate the multiple MVL-targeted protective functions of CU06-1004, including maintaining vascular integrity, inducing vascular normalization [282], improving cardiac remodeling, and inhibiting edema and inflammation.

Potential agents targeting angiopoietin/Tie-2 pathway

Section “Impaired microvascular function” has described the important role of the angiopoietin/Tie-2 pathway in maintaining vascular integrity, and recent studies have identified a number of negative regulators of this pathway as potential therapeutic targets for MVL. The overall activity of the Ang-1/Ang-2/Tie-2 system can be variably controlled by the dephosphorylating activity of vascular endothelial–protein tyrosine phosphatase (VE–PTP) [283]. Razuprotafib (AKB-9778) is a potent and selective inhibitor of VE–PTP that activates Tie-2, enhances Ang-1-induced Tie-2 activation, and stimulates phosphorylation of signaling molecules in the Tie-2 pathway, including Akt, eNOS, and ERK. AKB-9778-induced Tie-2 phosphorylation strongly inhibits neovascularization in a mouse model [283]. Another way to target the angiopoietin/Tie-2 system is through the application of vasculotide, a short synthetic peptide (HHHRHSF), which binds with high affinity to Tie-2 and activates Ang-1-like receptors. Vasculotide prevents I/R injury-induced renal vascular leakage, improves renal function and reduces mortality in mice [31]. In addition, there are drugs, such as Ang-1 variant (COMP-Ang-1), which have been shown to have protective effects on peritoneal vascular permeability stabilization and peritoneal transport function in a rat sepsis model [284]. The mechanism is mainly increased phosphorylation of Tie-2 receptors, enhanced pericyte coverage and expression of endothelial cell junction proteins.

S1P1 agonist

SAR247799 is a g-protein-selective S1P1 agonist. SAR247799 activates S1P1 on ECs without inducing receptor desensitization and is endothelium-protective in rats and pigs. In a rat model of renal I/R injury, SAR247799 was endothelium-protective and preserved renal structure and function at doses that did not induce S1P1 desensitization. Conversely, clinically used functional antagonists of S1P1 provide only minimal renal protection and desensitize S1P1 resulting in side effects (endothelial injury) [285]. Thus, pharmacologically S1P1 can be sustainably activated with no effect on the immune response, which provides a novel approach for the treatment of diseases associated with endothelial dysfunction and vascular hyperpermeability. The product is currently in Phase 2 clinical trials [31].

As summarized in Table 2, different drugs have multifaceted protective effects on the microvasculature in different models. Considering the crucial role of endothelial barrier and subcellular organelles, e.g., mitochondrion and ER in the prevention of MVD, especially MVL, targeting interventions by increase mitophagy flux, coping with ERS in CMECs and maintaining endothelial cell connectivity may have beneficial effect on MVD protection after I/R. Table 3 lists some of the targeted therapeutic strategies to improve mitochondrial/ER/endothelial barrier function and the corresponding underlying mechanisms. Nevertheless, these interventions are still poorly studied in ECs and MVD. In the future, further clarification of the synergistic roles of mitochondrial fission/fusion, mitophagy, ERS and endothelial dysfunction in MVD pathogenesis is required. In addition, how to rapidly and effectively detect subcellular organelle damage after reperfusion and select the appropriate time node to use targeted drugs are also key challenges to be solved in the clinical setting. There are a number of clinical trials tested currently used drugs with improvement of ischemic myocardial salvage or acute and chronic outcomes accompanied by mitigation of MVO, including β-blockers; statins; adenosine, a potent vasodilator of coronary microvasculature [45]; atrial natriuretic peptide which suppressing ET-1 production, exenatide, a glucagon-like peptide-1 agonist and antiplatelet therapy [286]. Physical interventions such as therapeutic hyperoxaemia/super-saturated oxygen (SSO2) and therapeutic hypothermia have also been shown to be cardioprotective and attenuate MVD after I/R [18]. As inflammation is the primary driving force for further tissue damage following ischemia–reperfusion, many clinical trials also evaluate the protective effects of agents targeting on a number of established inflammatory mediators, including IL-1β, IL-, IL-6, CRP at ligand or receptor level [286, 287]. Most these clinical trials end up with partial protection or improvement in infarct size, cardiac function or acute and chronic prognosis and in MVO reduction, but none of evaluated agents and therapeutic targets achieved complete protection. Thus, identification of novel targets, proper model selection, comprehensive large scale and multicenter clinical trial implementation are the future directions of research.

Table 2.

Multifaceted protective effects of the drugs mentioned in the section of Potential pharmacological therapies on the microvasculature of different models

Model	Species/ cell type	Drug (dose)	Vascular wall thickening/ lumen narrowing	Inflammation	Thrombosis	Vasodilation/ eNOS activity	Pathological mitochondrial fission	Mitophagy	Endothelial cells death	Ref No.
I/R (45 min/2 h)	Mice HCAECs	Empagliflozin (10 mg/kg/d) (10 µM)	↓	↓	↓	↑	↓	N	↓	[26]
I/R (45 min/2 h)	Mice	Empagliflozin (10 mg/kg/d)	↓	↓	↓	↑	↓	↑	↓	[37]
HFpEF	ZDF rats	Empagliflozin (0.25 mg/kg)	N	↓	N	↑	↓	N	N	[255]
I/R (45 min/2 h)	Mice HCAECs	Dapagliflozin (40 mg/kg/d) (10 µM)	↓	↓	N	↑	↓	N	↓	[253]
DCM	Rats HUVECs	Dapagliflozin (1 mg/kg/d) (1 μM)	N	N	N	N	N	N	↓	[254]
Smoking- induced CVD	guinea pigs GPAECs	Serelaxin (10 μg/day) (100 ng/ml)	↓	↓	N	↑	N	N	↓	[262]
I/R (1 h/24 h) (6 h/6 h)	Mice H5V	Serelaxin (50 μg/kg) (100 ng/ml)	↓	↓	N	↑	N	N	N	[176]
I/R (45 min/4 h)	Mice	Tan IIA (25 mg/kg)	N	↓	↓	↑	N	N	↓	[12]
CIH	Rats	Tan IIA (20 mg/kg/d)	↓	N	N	↑	N	N	↓	[270]

I/R ischemia/reperfusion, HCAECs human coronary artery endothelial cells, HUVECs human umbilical vein endothelial cells, GPAECs guinea pig aortic endothelial cells H5V, mouse heart immortalized endothelial cells line, HFpEF heart failure with preserved ejection fraction, ZDF rats Zucker Diabetic Fatty rats, DCM Diabetic cardiomyopathy, CIH chronic intermittent hypoxia, Tan IIA Tanshinone IIA, ↑ increased or upregulated, ↓ decreased or downregulated, N information not available/mentioned

Table 3.

Targeted therapeutic strategies to improve mitochondrial/endoplasmic reticulum/barrier function in endothelial cells

Model	Species/ cell type	Drug (dose)	Target point	Mechanism	Ref No.
Cardiac I/R (45 min/2 h)	Mice	Empagliflozin (10 mg/kg/d)	Mitochondrial function	Empagliflozin attenuates cardiac microvascular ischemia/reperfusion through activating the AMPKα1/ULK1/FUNDC1/mitophagy pathway in cardiac microvascular endothelial cells	[37]
Cardiac I/R (45 min/2 h)	Mice HCAECs	Dapagliflozin (40 mg/kg/d) (10 µM)		Dapagliflozin reduces pathological mitochondrial division during cardiac I/R injury by normalizing the XO–SERCA2–CaMKII–coffilin pathway, thereby protecting endothelial and microvascular function	[253]
Hyperglycemia	HUVECs	Liraglutide (10–80 nM)		Liraglutide prevents high glucose-induced dysfunction of endothelial cells by inhibiting PINK1/Parkin-dependent mitophagy	[288]
Oxidative stress	Rats HUVECs	Melatonin (40 mg/kg) (100 µM)		Melatonin increases Parkin-dependent mitophagy through activating the AMPK–TFEB signaling pathway, attenuating oxidative stress and endothelial cell apoptosis, which in turn promotes random pattern skin flaps survival	[289]
Hyperglycaemia Hyperlipidaemia	RAECs	Hydrogen sulphide		Hydrogen sulphide reduces endothelial cell apoptosis in a high-glucose and high-palmitate environment by inhibiting oxidative stress, reducing mitochondrial fragmentation and promoting mitophagy	[290]
Hyperhomocysteinemia	HUVECs	Tan IIA (100 µM)		Tan IIA protects endothelial cells from hyperhomocysteine-induced vascular endothelial damage by inhibiting intracellular oxidative stress and mitochondrial dysfunction through activation of the NNMT/SIRT 1-mediated NRF2/HO-1 and AKT/MAPKs signaling pathways	[291]
Hyperglycemia	HUVECs	Scutellarin (30 μM)		Scutellarin protects vascular endothelial cells from hyperglycemia-induced injury by upregulating mitophagy through the PINK1/Parkin signaling pathway	[272]
Type 2 diabetic and lung I/R (1.5 h/4 h)	Rats	Adiponectin (100 μg/kg)		Adiponectin attenuates I/R injury-induced oxidative stress, inflammation, endothelial cell apoptosis, and mitochondrial dysfunction by activating mitophagy mediated by the SIRT1–PINK1 signaling pathway	[292]
Hyperhomocysteinemia	bEnd3	Tetrahydrocurcumin (15 μM)		Pretreatment of endothelial cells with tetrahydrocurcumin ameliorated Homocysteine-induced oxidative damage, mitochondrial fission/fusion and mitophagy	[274]
High glucose and palmitate	HAECs	Ginseng-Sanqi-Chuanxiong extracts (GSC)		GSC extracts reverse mitophagy damage caused by high glucose and palmitate via AMPK pathway, regulates mitophagy, and thus ameliorates diabetes-induced endothelial cell senescence	[293]
Cerebral I/R	Mice	Tetrahydrocurcumin (25 mg/kg)		Tetrahydrocurcumin eliminates cerebral edema and MVL in the brain parenchyma by improving mitochondrial function in cerebrovascular endothelial cells during ischemic stroke	[273]
H/R (8 h/2 h)	HUVECs	Acetylcholine (10 μM)	MAM function	Acetylcholine inhibits H/R-induced intracellular Ca2+ overload in endothelial cells; inhibits ER Ca2 + depletion and mitochondrial Ca2+ overload during H/R; suppresses mitochondria–ER-dependent apoptosis by inhibiting M3AChR; and protects against I/R injury by inducing disruption of MAM in a NO-dependent manner	[144]
Sepsis	HUVECs	Adiponectin (10–20 μg/mL)	ER function	Adiponectin treatment reduces sepsis-induced endothelial cell apoptosis by attenuating oxidative stress and activating ERS pathways	[294]
H/R (12 h/8 h)	HBMECs	Quercetin (1 μM)		Quercetin has protective effects on HBMECs, including inhibiting apoptosis by promoting endothelial cell proliferation, migration and angiogenesis, maintaining MMP, and inhibiting ERS	[276]
Hyperglycemia	HUVECs	Procyanidin B2 (10 μM)		Procyanidin B2 attenuates endothelial cell ERS through a PPARδ-dependent mechanism	[139]
Hyperglycemia	HCAECs	Liraglutide (1 μM)		Liraglutide protects endothelial cells by attenuating ERS by inhibiting the phosphorylation of IRE1α and PERK and the expression of ATF6 and GRP78	[295]
Tunicamycin induced ERS	HCAECs	Dapagliflozin (1 μM)		Dapagliflozin protects endothelial cells by attenuating ERS by inhibiting the phosphorylation of IRE1α and PERK and the expression of ATF6 and GRP78	[295]
Hyperglycemia	HCAECs	Metformin (2.5 mM)		Metformin protects endothelial cells by attenuating ERS by inhibiting the phosphorylation of IRE1α and PERK and the expression of ATF6 and GRP78	[295]
ox-LDL-induced ERS	HUVECs	Melatonin (5 μM)		Melatonin attenuates ox-LDL-induced endothelial dysfunction by inhibiting the upregulation of CHOP and GRP78 to reduce ERS, which in turn inhibits JNK/Mff signaling	[296]
ZDF rats	Rats	Tongxinluo 50 mg/kg	Barrier function	Tongxinluo protects the integrity of the endothelial barrier by activating Angptl, significantly reducing endothelial cell apoptosis, and enhancing tight junction protein and VE–cadherin expression, thereby preventing MVL and IMH	[297]
Cardiac I/R (20 min/2 h)	Rats	FX-06 (2.4 mg/kg)		FX-06 inhibits dissociation between endothelial cells and VE–cadherin by rebalancing the activation state of RhoA/Rac 1; prevents migration of leukocytes by blocking dissociation of the intercellular VE–cadherin complex and prevents leukocyte migration; and prevents contraction of endothelial cells by binding to the VLDL receptor	[277]
Cerebral I/R (1.5 h) OGD/R (6 h/16 h)	Mice HBMECs	CU06-1004 (10 mg/kg)		CU06-1004 not only enhances the endothelial barrier via the cAMP/Rac/cortical actin pathway, but also reduces I/R-induced disruption of the blood–brain barrier by decreasing the levels of connexin43 and aquaporin 4	[279]
H/R (12 h/8 h)	HBMECs	Quercetin (0.1–10 μM)		Quercetin can increase the level of endothelial intercellular junction proteins such as Claudin 5 to maintain the integrity of the blood–brain barrier	[275]
Ischemic retinopathy Ang2 induced injury	Mice HUVECs	Razuprotafib (20 mg/kg) (0.17–50 μM)		Razuprotafib (AKB-9778) selectively inhibits VE–PTP, activates Tie-2, and stimulates phosphorylation of signaling molecules in the Tie-2 pathway, including Akt, eNOS, and ERK, which strongly inhibits neovascularization and protects the endothelial barrier	[283]
Renal I/R (35 min/24 h)	Mice	HHHRHSF (200 ng)		HHHRHSF activates Tie-2 signaling and blocks the vascular endothelial barrier in I/R ischemic kidneys without damaging renal tubules, thereby improving renal function and reducing mortality in mice	[298]
Subtotal nephrectomy induced uremia	Rats	COMP-Ang-1 (1 × 1010 pfu/ml per)		COMP-Ang-1 protects peritoneal vascular permeability stability and peritoneal transport function by increasing Tie-2 receptor phosphorylation, enhancing pericyte coverage and endothelial cell junction protein expression	[284]
Renal I/R (25 min/24 h)	Rats	SAR247799 (5 mg)		In a rat model of renal I/R injury, SAR247799 is endothelium-protective and preserves renal structure and function at doses that do not induce S1P1 desensitization	[285]

I/R ischemia/reperfusion, HCAECs human coronary artery endothelial cells, RAECs rat aortic endothelial cells, HUVECs human umbilical vein endothelial cells, bEnd3 mouse brain endothelial cells, HAECs human aortic endothelial cells HBMECs human brain microvascular endothelial cells, ox-LDL- oxidized low-density lipoprotein, H/R hypoxia/reoxygenation, Tan IIA Tanshinone IIA, MAM mitochondria-associated ER membrane, OGD/R oxygen–glucose deprivation/reoxygenation, rats Zucker Diabetic Fatty rats

The role and impact of microvascular damage in other diseases

Cerebrovascular disease

Currently, cerebrovascular disease is also the major leading cause of death globally, behind IHD [299]. There were 6.55 million people died of stroke in 2019, accounting for 11.6% of all deaths [300]. Cerebrovascular disease mainly includes ischemic stroke, hemorrhagic stroke, and subarachnoid hemorrhage (SAH). Acute ischemic stroke (AIS) is an infarction of brain tissue due to occlusion of cerebral arteries, accompanied by neuronal damage with high morbidity and mortality [301, 302]. Similar to myocardial I/R injury, oxidative stress, inflammatory response and microvascular endothelial cell dysfunction occur during cerebral ischemia and hypoxia [303–305], while oxygen supply is rapidly restored during reperfusion with a surge of ROS and secondary mitochondrial dysfunction [306–308]. These series of pathological changes eventually lead to massive microvascular endothelial cells death, cerebrovascular dysfunction, and more severe brain tissue infarction.

Currently, tissue-type plasminogen activator (tPA) is the only approved thrombolytic drug for the treatment of AIS [309], with significant clinical benefits when administered within 4.5 h after stroke onset [310–312]. However, tPA administration, especially delayed administration, may also increase cerebral edema, intracranial hemorrhage (ICH), and mortality [311, 313]. In addition to its own neurotoxicity, tPA, more importantly, can activate matrix metalloproteinase that degrade extracellular matrix integrity and increase risks of neurovascular cell death, blood–brain barrier leakage, edema, and hemorrhage [314–316], ultimately limiting the therapeutic effect in approximately 50% of patients in clinical practice [317]. In a model of lacunar infarction induced by injecting ET-1 into the rat internal capsule, transplantation of healthy endothelial cells into the internal capsule of the brain produced beneficial effect on the outcome of ischemic brain injury [318]. This result indicates the significance of maintenance of normal endothelial cell numbers and function in AIS treatment.

Diabetes

Diabetes mellitus (DM) is a metabolic disorder characterized by hyperglycemia, with an estimated global prevalence of 10.5% (536.6 million people) aged 20–79 years in 2021, rising to 12.2% (783.2 million people) in 2045 [319, 320]. Currently, approximately 90% of adult diabetics are type 2 diabetes [321]. Asia has become the main region, where DM is rapidly increasing [321]. The most common damage of DM is the diabetic microvascular complication (DMC), which significantly affects the quality of life, prognosis and life expectancy of diabetic patients [322, 323]. DMC represents a “syndrome”, occurring in multiple organs and tissues, such as retina, kidney, heart, nerve tissue, skin and extremities [324, 325]. It manifests as blindness due to diabetic retinopathy, end-stage renal failure due to diabetic nephropathy, numbness and pain in the extremities due to diabetic peripheral neuropathy, and finger/toe-end necrosis and long-term incurable skin ulcers due to MVD [322–324, 326]. DMC is characterized by vascular endothelial damage, increased blood viscosity, erythrocyte aggregation, platelet adhesion, micro-thrombosis and MVO [324, 327]. The mechanism is similar to that of ischemic cardiac MVD. Microvascular ECs mainly utilize glycolysis to produce energy [328]; however, excess glucose can damage microvascular ECs by generating large amounts of ROS via the polyol pathway and large amounts of advanced glycosylation end products (AGEs) via the sorbitol pathway [328, 329]. Moreover, chronic hyperglycemia also inhibits NO production by endothelial cells, stimulates the release of vasoconstrictive factors and accelerates apoptosis of microvascular ECs, which leads to abnormal constriction of micro-vessels [330]. Besides, degradation of the vascular basement membrane is another mechanism of DMC. Matrix metalloproteinases play an important role in the regulation of cellular homeostasis by degrading the extracellular matrix [331]. Excessive ROS stimulates the synthesis and secretion of matrix metalloproteinases, accelerating the degradation of vascular basement membranes and the development of DMC. In a cell model of mouse microvascular endothelial cell damage caused by a high glucose (30 mM) stimulation for 72 h, inhibition of ROS levels and matrix metalloproteinase activity by ropivacaine could attenuate high glucose-induced endothelial cell injury [332].

Kidney disease

Acute kidney injury (AKI) is a major risk factor for progressive renal insufficiency. Currently, more than 20% of hospitalized adults worldwide experience certain degree of AKI, most commonly due to kidney I/R injury and sepsis, AKI can lead to irreversible renal function loss as well as end-stage renal fibrosis [333, 334]. Damage to the renal microcirculation, particularly to the peritubular capillary (PTC) network, has been reported to be a key factor in the development of interstitial fibrosis in patients with kidney disease [335]. There are substantial evidences that PTC injury, sparing and interstitial fibrosis are typical pathological changes whatever the etiology of the kidney disease [335, 336]. PTC endothelial damage after renal I/R is similar to MVD of myocardial I/R injury, which manifests as early and progressive endothelial swelling, subendothelial edema, endothelial thickening and increased capillary permeability [333, 337]. After kidney I/R injury, damaged renal parenchymal cells and various inflammatory mediators upregulate the expression of adhesion molecules and mediate the migration of various types of inflammatory cells to the site of kidney injury, and the progressively increasing inflammatory response and excessive ROS production can lead to renal MECs dysfunction [338]. In addition to causing MVD, hypoxia can also cause an overexpression of HIF-1α in tubular epithelial cells, which stimulates proliferation of inflammatory cells and their recruitment to the kidney, induces maladaptive expression of matrix-modifying factors in hypoxic tubular epithelial cells and promotes epithelial-to-mesenchymal transition, eventually resulting in renal fibrosis [335, 339].

Conclusion

In industrialized countries, due to the spread of preventive measures in society and advances in medical technology, the majority of patients survive AMI and the mortality rate continues to decline [340]. However, the prevalence of heart failure after MI is on the rise, and the main reason for this lies in the weakened myocardium [341] due to complications after revascularization, such as MVD, including MVO and MVL. MVD has a significant impact on the severity, development, and prognosis of IHD [342, 343]. Clinically, most patients are of advanced age and have a long history of comorbidities, such as hypertension, dyslipidemia, diabetes mellitus and chronic inflammatory diseases. It is difficult for clinical studies to avoid many influential and confounding factors. Moreover, MVD itself involves a variety of clinical manifestations and molecular mechanisms, and the interaction between them is a complex signal transduction cascade. Experimental studies usually focus on “single” mechanism/intervention in young, male and healthy animals, ignoring the reciprocal or synergistic roles of other pathways. Therefore, the current array of clinical tests and intervention strategies for MVD after myocardial I/R injury has not yielded satisfactory clinical results. In addition, majority of previous studies have focused on the cardiomyocyte injury, but few studies on CMECs, the most abundant cardiac cellular components. The protection of the microvasculature has continued to receive less attention despite the overwhelming pathophysiologic evidence of the significance of MVD, especially MVL. Limited experimental benefits in MVL intervention still require clinical confirmation. In the future, targeting CMECs protection may represent a new direction in the battle against cardiac I/R injury, rather than focusing on cardiomyocyte only, which call for more comprehensive studies to explore molecular mechanisms and novel therapeutic interventions to prevent MVD.

Author contributions

BHZ, AR, LZ and QLL drafted the manuscript, JT, NX and AXD generated the images, MTG and XFS edited the manuscript and XMG revised the manuscript and responsible for funding support. All of the authors have read and approved the content of the manuscript.

Funding

This work was supported by grants of the Natural Science Foundation of China (No. U1903212, 81870272), opening project of the Xinjiang Key Laboratory (2021D04020), the State Key Laboratory of Pathogenesis, Prevention and Treatment of Central Asia High Incidence Diseases fund (SKL-HIDCA-2021-XXG1, SKL-HIDCA-2022-XXG1) and the key project of the Natural Science Foundation of Xinjiang Uygur Autonomous Region (2023).

Availability of data and materials

Not applicable.

Declarations

Conflict of interest

There are no conflicts of interest to declare.

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.