Specialized pro-resolving mediators in cardiovascular diseases

Abstract

The resolution of inflammation is a highly regulated process enacted by endogenous mediators including specialized pro-resolving lipid mediators (SPMs): the lipoxins, resolvins, protectins and maresins. SPMs activate specific cellular receptors to temper the production of pro-inflammatory mediators, diminish the recruitment of neutrophils, and promote the clearance of dead cells by macrophages. These mediators also enhance host-defense and couple resolution of inflammation to subsequent phases of tissue repair. Given that unresolved inflammation plays a causal role in the development of cardiovascular diseases, an understanding of these endogenous pro-resolving processes is critical for determining why cardiovascular inflammation does not resolve. Here, we discuss the receptor-dependent actions of resolvins and related pro-resolving mediators and highlight their emerging roles in the cardiovascular system. We propose that stimulating resolution could be a novel approach for treating chronic cardiovascular inflammation without promoting immunosuppression.

1.1 Introduction

Inflammation in its simplest definition is an immune response to a given stimulus. Acute inflammation is a protective host response that has evolved to thwart pathogens and to repair tissue injury¹. This protective program is self-limited which means that the inflammation is tightly coupled to a termination/resolution phase, which is now known as inflammation-resolution². When a stimulus is either excessive or persistent or when an immune response is mounted inappropriately to non-pathogenic endogenous antigens, this once protective response can become maladaptive, which can lead to chronicity and eventual tissue damage. It was previously believed that inflammation ceased due to passive dissipation of inflammatory stimuli. We now appreciate that the resolution of inflammation is an active and highly coordinated process that involves several chemical mediators and cell types². Some of the mediators involved in the resolution of inflammation include polyunsaturated fatty acid-derived lipoxins, resolvins, protectins and maresins (collectively referred to as specialized pro-resolving mediators or SPMs), as well as protein mediators like Annexin A1. Gaseous and nucleotide mediators are also emerging as important players in the resolution response^{3, 4}. Importantly, the discovery that resolution of inflammation is an active process is crucial because this knowledge has the potential to lead to new approaches to selectively target a safer termination of inflammation without compromising host defense².

Uncontrolled inflammation, and thus a failure of the inflammation-resolution response, is the underpinning of several prevalent human diseases. Atherosclerotic vascular disease, the leading cause of death worldwide, is a prominent example of failed inflammation-resolution in humans. Atherosclerosis is a disease of the vascular wall that is driven by the persistent subendothelial retention of ApoB lipoproteins⁵. This persistent stimulus deranges host cells in a manner that leads to arterial tissue injury^6–11. Because of the chronicity of inflammation that is characteristic of atherosclerosis, clinical trials are currently testing the hypothesis that blocking inflammation could provide additional clinical benefit beyond lipid lowering strategies that are the current standard of care. Specifically, the Cardiovascular Inflammation Reduction Trial (CIRT) is utilizing low dose methotrexate to determine whether blocking inflammation will lower vascular events rates^{12, 13}. Similarly, the Canakinumab Anti-inflammatory Thrombosis Outcomes Study (CANTOS) is testing whether an interleukin 1β antibody will decrease chronic vascular inflammation and cardiovascular events without affecting cholesterol levels^{14, 15}. While these clinical trials could provide evidence for a causal role of inflammation in atherosclerosis in humans, targeting pro-resolution pathways may offer additional advantages over traditional anti-inflammatory approaches that can be immunosupressive². Here, we discuss the role of pro-resolving lipid mediators in the cardiovascular system and highlight recent proof-of-concept studies in rodents demonstrating that stimulating resolution has beneficial effects on atheroprogression.

1.2 Inflammation-resolution and cardiovascular diseases (CVD)

One of the first insights into the role of SPMs in atherosclerosis came from a study by Serhan and colleagues in which they reported that angioplasty increases intracoronary levels of lipoxins in humans¹⁶. Building on these observations in animal models of atherosclerosis, Chan and colleagues identified that atherogenesis is significantly thwarted in rabbits overexpressing 15-lipoxygenase (15-LOX), a key biosynthetic enzyme for SPMs^{17, 18}. Atherogenesis is also delayed in transgenic mice overexpressing 15-LOX and crossed with ApoE^−/− mice fed a chow diet, compared with wild type controls. Isolated peritoneal macrophages from the 15-LOX -ApoE^−/− transgenic mice exhibited an increased production of key SPMs, compared with wild type control macrophages. Importantly, 15-LOX, can promote the biosynthesis of SPMs or pro-atherogenic oxidation of LDL^18–22 and one determinant of whether these pathways promote SPM production or pro-atherogenic lipids is diet composition. Along these lines, 15-LOX -ApoE^−/− transgenic mice fed a high-fat, high-cholesterol diet had worse atherosclerosis compared with wild type controls²¹. These studies suggest that diet is a critical determinant to whether these enzymatic pathways can promote or suppress plaque progression.

More recent studies have been focused on the therapeutic actions of SPMs in cardiovascular diseases. Several SPMs and pro-resolving proteins/peptide have been explored as therapeutics in experimental atherosclerosis. The sections below will highlight each SPM or pro-resolving protein/peptide and their roles in cardiovascular diseases, specifically atherosclerosis and myocardial infarction (Table 1).

Table 1.

Treatment with pro-resolving mediators in CVD

CVD	Actions
Atherosclerosis
RvE1	Decreased atherogenesis (rabbit)23 Attenuated atheroprogression and further decreased atheroprogression in the presence of atorvastatin (mice)89
RvD1	Halted atheroprogression, enhanced lesional efferocytosis and lesion stability (mice); decreased RvD1 in human vulnerable plaque regions34
RvD2 and Mar1	Decreased atheroprogression and enhanced lesion stability (mice)35
Aspirin triggered lipoxin A4 (ATL)	Decreased atheroprogression and enhanced lesion stability (mice)36
Ac2-26 plaque-targeted nanoparticles	Decreased atheroprogression and promoted plaque stability (mice) 37
Annexin A1	Decreased leukocyte migration/atheroprogression (mice)27 Decreased atheroprogession90
IL-10 plaque-targeted nanoparticles	Decreased atheroprogression and promoted plaque stability (mice)91
Myocardial infraction
RvE1	Reduced infarct size (rats) 48
RvD1	Reduced PMN infiltration and fibrosis (mice)43
Neointimal hyperplasia
Aspirin triggered lipoxin A4 (ATL)	Reduce neointimal hyperplasia (mice)92
RvD2	Reduced neointimal hyperplasia (mice) 93, 94
RvD1	Reduced neointimal hyperplasia (rats) rats87
Mar1	Reduce neointimal hyperplasia (mice)94

1.2.1 SPMs and pro-resolving proteins/peptides in atherogenesis

In rabbits fed a high-fat, high-cholesterol diet, the SPM resolvin E1 (RvE1), reduces atherogenesis, which is associated with a reduction in CRP²³. Importantly, RvE1 was administered at the onset of high-fat, high-cholesterol diet feeding and served as a proof-of-concept that higher levels of RvE1 early on may be important for delaying atherogenesis. In this particular study, RvE1 was given topically to the periodontium, which implies that locally administered RvE1 has systemic effects (Table 1). Importantly, there is a known link between atherosclerosis and periodontal disease²⁴, and this study also indicated that RvE1 could thwart atherogenesis in the context of periodontal disease²³. More information regarding periodontal disease and SPMs can be seen in an accompanying review in this series.

Annexin A1 is a protein-derived endogenous mediator of inflammation-resolution that has also been studied in atherosclerosis. Annexin A1 levels are high in asymptomatic plaques compared with symptomatic carotid artery plaques²⁵. This decrease in expression at both the mRNA and protein levels in symptomatic plaques suggests that advanced, symptomatic plaques may have a general defect in inflammation-resolution programs. Interestingly, Annexin A1 expression was found to be higher in smooth muscle cells (SMCs) isolated from human asymptomatic compared with symptomatic carotid plaques²⁶, which suggests a potential cell specific role for the generation of Annexin A1. Annexin A1 can be synthesized by several cell types, and so a deeper investigation of which cell type makes Annexin A1 in plaques would be invaluable. In other words, there could be certain cell types that are more dysregulated in their ability to synthesize Annexin A1, and this information may prove important for therapeutic intervention. In this regard, Annexin A1 and its bioactive peptide, called Ac2-26, were also recently tested in murine models of atherosclerosis. Annexin A1 plasma levels negatively correlated with lesion area²⁷. Upon further probing, Annexin A1 was found to decrease atherogenesis and regulate key chemokines involved in leukocyte migration into plaques when given at the onset of high-fat, high-cholesterol feeding in ApoE^−/− mice, largely through its ability to signal through its G-protein receptor called ALX/FPR2²⁷. These studies collectively demonstrate that increasing the levels of Annexin A1 early on is beneficial in thwarting the onset of the disease. It is important to note that some of the ligands for ALX/FPR2 also exert pro-inflammatory actions. As an example, the ALX/FPR2 ligand, CRAMP (in mice) or LL-37 (in humans) was shown to promote leukocyte extravasation into the arterial wall in an ALX/FPR2-dependent manner²⁸. GPCRs can form dimers that affect their activation and downstream signaling²⁹ and emerging evidence suggests that dimerization may play a key role in fine-tuning signaling pathways²⁹. Along these lines, Cooray et al. observed that pro-resolving ligands, like Annexin A1 initiated homodimerization of ALX/FPR2, which led to a pro-resolution feed-forward circuit, and an increase in IL-10³⁰. Ac2-26, which a has anti-inflammatory and pro-resolving properties³¹, initiates a heterodimer complex with ALX/FPR2 and another formyl peptide receptor called FPR1. This heterodimer complex neutralizes the actions of the ALX/FPR2 pro-inflammatory ligand called serum amyloid A (SAA)³⁰. Therefore, the way in which ligands interact with their receptors on the plasma membrane may provide new mechanistic clues as to why several ligands can bind a common receptor and regulate distinct signaling programs.

1.2.2 SPMs and pro-resolving proteins/peptides in advanced atherosclerosis

In humans, the subset of atherosclerotic plaques that are at risk for precipitating acute atherothrombotic clinical events, including myocardial infarction and stroke³², are typically referred to as vulnerable plaques. These plaques have distinct features, including heightened inflammation and oxidative stress; large areas of necrosis, which are composed of apoptotic cells not efficiently cleared by efferocytosis; and thinning of a protective layer of collagen that overlies the areas of necrosis in more stable lesions^{10, 32, 33}. All of these features are indicative of defective inflammation-resolution and recently, we found that vulnerable plaque regions from human carotid arteries have a marked imbalance in SPMs to pro-inflammatory lipid mediators (e.g. leukotrienes), compared with stable plaque regions. We and others recapitulated these findings in mice and found that advanced plaques of Ldlr^−/− or ApoE^−/− mice fed a high-fat, high-cholesterol diet also exhibited a marked imbalance in the SPM:pro-inflammatory mediator ratio compared with early plaques^{34, 35}. To prove causation, we and others administered key defective SPMs (RvD1, RvD2 and MaR1, or aspirin-triggered lipoxin A₄) during the critical time period where atherosclerosis progressed to the advanced stage in mice (Table 1). In general, SPMs delayed atherosclerosis progression and promoted a more stable-like plaque phenotype^34–36. One of the most striking observations of the plaques in athero-prone mice treated with SPMs was an increase in fibrous cap thickness and collagen synthesis in plaques^34–37 (Fig. 1). Even though mechanisms underlying the formation of the protective cap still remain unknown, it is possible that efferocytosis can participate in the repair process. Efferocytosis stimulates the biosynthesis of SPMs and the release of other tissue reparative molecules like TGFβ^{10, 38–40}. Furthermore, RvE1 increases collagen in the periodontal ligament post injury⁴¹, and 17R-RvD1 stimulates matrix synthesis in a model of arthritis⁴². Conversely, RvD1 decreases cardiac collagen in the setting of myocardial infarction⁴³, and there other reports suggesting that SPMs exert anti-fibrotic actions^{44, 45}. Thus, the role of SPMs in collagen synthesis and fibrosis is context-specific, and an understanding of the mechanisms whereby SPMs increase collagen synthesis in advanced atherosclerotic plaques is likely to be a fruitful area of future investigation.

Figure 1. SPMs promote features of plaque stability and temper platelet activation.

Several features of advanced plaques include large necrotic cores, thin fibrous caps, increased oxidative stress, and a deregulated activation of MMPs, defective efferocytosis, and an imbalance in the ratio of SPM:LTB₄. In the event of a plaque rupture, endothelial cell erosion, or arterial injury, platelets become activated which can lead to devastating thrombotic events. Administration of SPMs potently reduces necrosis, oxidative stress, and MMPs activation and increases fibrous cap thickness, efferocytosis and the SPM:LT ratio. SPMs also prevent platelet aggregation and thrombosis.

Recent results also suggest that pro-resolving drugs can be targeted to plaques. In this regard, Ac2-26 was encapsulated into plaque-targeted nanoparticles (NPs) and administered to mice with established atherosclerosis. The Ac2-26 plaque targeted NPs delayed plaque progression and promoted several features that are characteristic of stable plaques³⁷. Specifically, these targeted NPs enhanced the thickness of the fibrous cap and decreased lesional MMPs, which is similar to the actions of the RvD1 and aspirin-triggered lipoxin A₄ (ATL) ^{34, 36.} Importantly, RvD1, ATL and Ac2-26 bind and signal through the same receptor, ALX/FPR2. These studies used Fpr2/Fpr3^−/− mice whereas the above study²⁷ used Fpr2^−/− mice. Therefore, understanding more mechanisms associated with ALX/FPR2 may be advantageous for new drug design. With that said, other SPMs like RvD2, MaR1 and RvE1 that all bind and signal through distinct receptors also exert a similar protective phenotype on advanced plaques, which is suggestive of common downstream pathways for resolution-mediators and plaque stability.

This new explosion of work is only the tip of the iceberg for the inflammation-resolution and atherosclerosis fields. Currently, little is known about the mechanisms, cell types, and signaling pathways associated with SPM-mediated protection in advanced atherosclerosis. For example, T regulatory cells (T_reg) are emerging as crucial cellular players against atherosclerosis progression. Key D-series SPMs, such as those identified in human and murine plaques, have been shown to stimulate T_reg in other contexts⁴⁶ and therefore specific SPM in plaques may potentially play a role in regulating plaque T_reg.

1.3 SPMs in ischemia and myocardial infarction

Ischemia can lead to tissue injury prompting a sterile inflammatory response. During this process, PMN rapidly infiltrate and expel reactive oxygen species (ROS) and other mediators, damaging healthy tissue⁴⁷. Therefore, controlling unwarranted PMN accumulation, or promoting their removal, is important for protecting healthy tissues such as the myocardium during ischemia. Because SPMs promote both of these processes, recent studies have hypothesized that administration of SPMs could potentially improve tissue recovery during ischemic injury. Indeed, exogenous delivery of RvD1 after myocardial infarction (MI) in mice reduces accumulation of PMN and fibrosis, leading to improved cardiac function⁴³. It was similarly shown that RvE1 decreases infarct size in a rat model of MI/reperfusion injury⁴⁸. In addition to the heart, protective roles of SPM have also been described in other tissues during ischemia. For example, ischemia/reperfusion in the kidneys leads to biosynthesis of D-series resolvins and protectins. Treatment with resolvins prior to ischemia decreases PMN infiltration, and tissue fibrosis, and administration shortly after reperfusion is also organ protective⁴⁹. In a model of second-organ reperfusion injury, resolvins protect against PMN recruitment into the lungs^{50, 51}. The endogenous protective role of SPMs was demonstrated in mice lacking the Alx/Fpr2 or the Gpr18 (also known as DRV2) receptors, in which ischemia/reperfusion resulted in excessive leukocyte accumulation^{51, 52}. In addition to well-described detrimental roles of PMN during ischemia, there is also emerging evidence that PMN participate in tissue repair⁵³. Indeed, PMN depletion impairs the development of a tissue reparative macrophage phenotype, resulting in increased apoptotic cell accumulation in the heart after MI⁵⁴. Along these lines, recent studies by Gronert et al. identified a distinct tissue resident and lymphoid homing PMN subset that produces SPM to blunt activation of Th1 and Th17 T cells in a murine model of dry eye disease⁵⁵. Clearly, the roles of PMN in inflammation, resolution and tissue repair warrant further study.

Another complication from vascular inflammation and injury is neointimal hyperplasia and restinosis. This topic will be covered extensively in an accompanying review within this series but the roles of SPMs in this process are highlighted in Table 1.

1.4 Pro-resolving mediators in arteriogenesis

Arterial occlusion impairs tissue perfusion and can even lead to critical limb ischemia and limb loss in the case of peripheral atherosclerosis^{56, 57}. Arteriogenesis (i.e. collateral vessel growth) is a process that partially compensates for defects in tissue perfusion⁵⁸. Several therapeutic strategies to increase arteriogenesis have met limited success in large clinical trials for several reasons, notwithstanding the fact that most factors that promote revascularization are pro-inflammatory and can exacerbate atherosclerosis. Recent results indicate that RvD2 is generated in ischemic tissue from both humans and mice⁵⁹. Importantly, RvD2 enhances tissue perfusion and arteriogenesis in mice that had undergone hind limb ischemia in a receptor (GPR18/DRV2)-dependent manner⁵⁹. The improved revascularization by RvD2 was associated with decreased pro-inflammatory mediators and enhanced skeletal muscle regeneration. Chronic inflammation and metabolic disease impair the normal process of tissue revascularization and diabetic patients are particularly susceptible to defective tissue perfusion and wound healing^{56, 57}. Importantly, RvD2 treatment to diabetic mice significantly rescued the defect in revascularization⁵⁹. These studies highlight a new role for SPMs in the vasculature as well as a protective role for SPMs against skeletal muscle damage induced by ischemia.

1.5 SPMs in thrombosis and platelet activation

Myocardial infarction and stroke are the clinical manifestations of atherothrombosis⁶⁰. Lipid mediators have long been recognized to play important roles in platelet activation and thrombosis, with the thromboxane/prostacyclin balance being a critical determinant of this process. A role of pro-resolution mediators in regulating platelet activation has also recently emerged. In healthy humans receiving aspirin, an epimeric aspirin-triggered form of lipoxin A₄ generated from acetylated COX-2 (denoted ATL) was inversely correlated with pro-thrombotic thromboxane^{61, 62}. Plasma levels of 15-epi-LXA₄ were increased most significantly with low dose aspirin (i.e., 81mg), which is the recommended dose for patients with CVD. Interestingly, omega-3 fatty acid derived RvE1, another product of acetylated COX-2, blocks ADP and thromboxane-stimulated platelet aggregation^{63, 64}. Further mechanistic studies revealed that RvE1 blocked ADP-initiated signals downstream of its receptor, P2Y₁₂, and that the effects of RvE1 are mediated through its receptor called ERV1/ChemR23⁶⁴. More recently, RvD2 was found to prevent thrombosis of the deep dermal vascular network and subsequent dermal necrosis in a mouse burn model⁶⁵. SPM have also been shown to increase macrophage uptake of blood clots in vitro, indicating that SPM may play a role in clot remodeling during resolution⁶⁶. Interestingly, platelet:PMN aggregates formed during self-limited, acute inflammation lead to SPM biosynthesis, which have been shown to reduce human platelet:PMN aggregates^{67, 68}. Given that platelet:PMN aggregates contribute to plaque inflammation⁶⁹, these results suggest that these aggregates in CVD may be deficient in their ability to generate SPMs. In this regard, new results indicate that SPMs, like Mar1, promote a pro-resolving phenotype of platelets and prevent thrombin-activated platelets from releasing several known CVD targets like soluble CD40L, thromboxanes, CD62P and platelet microparticles⁷⁰. Therefore, these results collectively demonstrate that SPM may play key roles in regulating the resolution of thrombosis during inflammation, which could have important therapeutic implications for CVD.

1.6 SPMs enhance efferocytosis, an essential program for the resolution of inflammation

Efferocytosis, or the clearance of dead cells, is essential for a successful tissue resolution response. During efferocytosis, SPM (e.g., RvD1, RvD2, RvE2) production increases, which enhances further clearance of debris and apoptotic cells in an autocrine and paracrine manner^{38, 40}. However, defective efferocytosis is observed in several chronic inflammatory diseases like asthma, obesity/diabetes, atherosclerosis, and interestingly, these conditions are also associated with defective production of several SPMs^{10, 71, 72}. While the specific mechanisms underlying impaired SPM production and efferocytosis in these diseases are incompletely understood, the subcellular localization of 5-lipoxygenase (5-LOX) was recently shown to be a critical determinant of whether this enzyme produces pro-inflammatory leukotrienes or SPM in macrophages. Nuclear localization of 5-LOX favors the production of pro-inflammatory LTB_4, while non-nuclear localization favors pro-resolving LXA₄⁷³. Interestingly, RvD1 enhances this non-nuclear localization, promoting a positive feed-forward regulation of the SPM:leukotriene balance. This regulation of 5-LOX nuclear localization was found to be governed by signaling through the MER Proto-Oncogene Tyrosine Kinase (MerTK) receptor, coupling efferocytosis to SPM biosynthesis⁷⁴. In mice genetically engineered to resist cleavage (and therefore inactivation) of this receptor, SPM biosynthesis is increased and resolution of acute inflammation and efferocytosis are improved⁷⁴. In the context of atherosclerosis, these mice also had improved efferocytosis, decreased necrosis and increased lesional SPMs compared with controls⁷⁵. In acute MI, mice deficient in efferocytosis receptors, Mertk and/or Mfge8, have decreased cardiac function and remodeling that was specifically due to abrogation of efferocytosis^{76, 77}. Thus, defective efferocytosis in CVD could be related to defective production of SPM.

Defective efferocytosis was also reported in a mouse model of diabetic atherosclerosis. This defect, however, was reversed by a fish-oil diet rich in SPM precursors, EPA and DHA⁷⁸. As mentioned above, RvD1 administration to advanced lesions showed an improvement in lesional efferocytosis and an enhancement in plaque SPMs³⁴. Interestingly, it was also recently shown that RvD1, acting through ALX/FPR2, protects macrophages from oxidative stress-induced apoptosis during efferocytosis, in part by regulating NADPH oxidase activation and expression of apoptotic proteins, Bcl-xL and Bcl-2⁷⁹. In the context of atherosclerotic lesions, the ability of macrophages to survive and continue to phagocytose debris despite the highly oxidative environment is critically important.

Macrophages are highly plastic cells that adapt to their local tissue microenvironments⁸⁰. It is becoming increasingly clear that the phenotype of a macrophage within an atherosclerotic plaque dictates its function^{81, 82}. In the case of plaque regression, macrophages polarize to an M2-like phenotype⁸³, which characteristically quells inflammation and promotes tissue repair. In another context, M2 macrophages biosynthesize SPMs more readily than M1-like macrophages, which is consistent with a tissue reparative phenotype³⁸. Along these lines, RvD2 and MaR1 co-treatment promote a plaque tissue reparative macrophage phenotype³⁵. How exactly SPMs regulate macrophage phenotype in plaques in is an important area of research and warrants further investigation.

1.7 Summary and future directions

It is now widely accepted that chronic inflammation plays a causal role in the development of CVD and finding new ways to target the inflammatory response is gaining traction as a new therapeutic approach to treating CVD in conjunction with the current standard of care (e.g., aspirin, statins, anti-platelet therapies)¹⁵. Unlike traditional anti-inflammatory strategies that blunt the production of inflammation “initiators” and could potentially lead to immunosuppression, pro-resolving mediators resolve inflammation without compromising host-defense². In fact, pro-resolving mediators enhance host defense to both viral and bacterial infections and lower the threshold for antibiotic therapy^{84, 85} and stimulate distinct processes necessary for tissue repair and regeneration (see accompanying reviews in this series). Thus, new strategies to promote resolution of inflammation may offer unique opportunities to combat chronic inflammation associated with CVD.

In the future, it will be important to determine if SPM can be effectively targeted to sites of local inflammation (e.g., atherosclerotic plaques). Along these lines, it is interesting to point out that microparticles released from activated immune cells carry SPMs and prior work showed that humanized nanoparticles could potentially be vehicles for targeted delivery⁸⁶. In the setting of vascular surgery, unidirectional delivery of SPM through biodegradable vascular wraps have been successful in animal models, preventing restenosis and local inflammation^87,88. Further mechanistic studies on how resolution is perturbed in CVD, including an understanding of how SPM biosynthesis and receptor-mediated actions might be deregulated, will help to inform the development of novel pro-resolution therapeutics for CVD.