Plasma high density lipoproteins: Therapeutic targeting and links to atherogenic inflammation

a b s t r a c t
Plasma HDL levels have an inverse relationship to coronary artery disease (CAD) risk, which led to the idea that increasing HDL levels therapeutically would ameliorate atherosclerosis. Human genetic defi- ciency of CETP caused markedly elevated HDL and moderately reduced non-HDL cholesterol levels, suggesting that CETP inhibitors might produce cardiovascular benefit. The CETP inhibitor anacetrapib reproduced the phenotype of homozygous CETP deficiency and showed a highly significant benefit for CAD in the REVEAL trial. However, the magnitude of this effect was moderate, and the mechanism of benefit remains unclear. Insights into the mechanisms underlying macrophage cholesterol efflux and reverse cholesterol transport have come from monogenic human disorders and transgenic mouse studies. In particular, the importance of the ATP binding cassette transporters ABCA1 and ABCG1 in promoting cholesterol efflux from myeloid and other hematopoietic cells has been shown and linked to aberrant myelopoiesis and macrophage inflammation. Recent studies have shown that myeloid defi- ciency of ABCA1 and ABCG1 leads to macrophage and neutrophil inflammasome activation, which in turn promotes atherosclerotic plaque development and notably the formation of neutrophil extracellular traps (NETs) in plaques. In addition, clonal hematopoiesis has emerged as an important CAD risk factor, likely involving macrophage inflammation and inflammasome activation. Further elucidation of the mechanisms linking plaque accumulation of cholesterol and oxidized lipids to myeloid cell inflammation may lead to the development of new therapeutics specifically targeting atherogenic inflammation, with likely benefit for CAD.

1.HDL and atherosclerosis
In epidemiological studies, plasma HDL-cholesterol levels show a robust inverse relationship with coronary heart disease (CHD) independent of other risk factors [1]. Infusions of HDL or increased expression of the main HDL protein, apoA-1, consistently reduce atherosclerosis in animal models [2e5]. In most studies, the ability of HDL to promote cholesterol efflux from macrophage foam cells is inversely correlated with human coronary atheroma burden and with incident CHD [6e11]. In contrast, Mendelian Randomization studies have failed to find a relationship between SNPs that in- crease HDL cholesterol levels and coronary heart disease [12]. Nonetheless, preclinical studies and studies of human HDL func- tionality suggest that some approaches to increasing HDL levels or beneficial functions might be a way to reduce the substantial burden of CHD that remains in subjects optimally treated by LDLlowering [13].The idea that HDL might mediate protection from atheroscle- rosis by stimulating an overall process of reverse cholesterol transport was originally proposed by Glomset [14]. Over the sub- sequent four decades, the factors controlling HDL metabolism and the molecular underpinnings of the reverse cholesterol transport pathway have been elucidated with contributions by many labo- ratories [2,3,15e18]. The most important insights have been gained by the elucidation of human genetic deficiency states affecting HDL and by transgenic mouse studies. Fig. 1 illustrates the overall pro- cess of reverse cholesterol transport. This is initiated in the arterial wall by the ATP binding cassette transporters ABCA1 and ABCG1, which are induced in arterial wall macrophage foam cells by LXR activation and promote the efflux of cholesterol onto lipid poor ApoA-1 and HDL particles. Cholesterol in HDL may be esterified by lecithin:cholesterol acyltransferase (LCAT) and directly taken up in the liver by a process of selective free and esterified cholesterol removal mediated by scavenger receptor B1 (SR-BI).

In humans,The role of HDL in macrophage cholesterol efflux and reverse cholesterol transport. Macrophage cholesterol efflux and reverse cholesterol transport are initiated in the arterial wall by the ATP binding cassette transporters ABCA1 and ABCG1, which are induced in arterial wall macrophage foam cells by LXR activation and promote the efflux of cholesterol onto lipid poor apoA-1 or HDL particles.Cholesterol in HDL may be esterified by lecithin:cholesterol acyltransferase (LCAT) and directly taken up in the liver by a process of selective cholesteryl ester removal mediated by scavenger receptor B1 (SR-BI). This may be followed by the excretion of cholesterol into bile involving ABCG5/8. In humans, cholesteryl ester transfer protein (CETP) mediates the exchange of cholesteryl esters (CE) in HDL with triglycerides in VLDL or chylomicrons leading to a net transfer of CE form HDL to the triglyceride-rich lipoproteins and LDL and subsequent removal in the liver via the LDL receptor and other pathways. ABCA1 also initiates the formation of HDL particles in the liver and small intestine (not shown) by binding ApoA-1 and promoting its lipidation. HL, hepatic lipase; EL, endothelial lipase.cholesteryl ester transfer protein (CETP) mediates the exchange of cholesteryl esters (CE) in HDL with triglycerides in VLDL or chylo- microns, leading to a net transfer of CE from HDL to the triglyceride- rich lipoproteins and LDL and subsequent removal in the liver via the LDL receptor and other pathways. Research in my laboratory initially focused on elucidating the role of CETP in lipoprotein metabolism. More recent studies have used mouse models to investigate the role of ABCA1 and ABCG1 in cholesterol efflux and atherosclerosis.

In collaboration with colleagues in Japan, we first defined hu- man genetic deficiency of CETP, which was characterized by markedly elevated levels of HDL cholesterol (HDL-C), as well as reduced levels of LDL cholesterol (LDL-C) and ApoB, a profile that is typically associated with reduced atherosclerosis [19]. This led to the development of CETP inhibitors. These were shown to raise HDL-C and ApoA-1 levels, and for the more potent CETP inhibitors, there was also a lowering of HDL cholesterol and ApoB levels. Based on epidemiological observations, it was expected that this dramatic increase in HDL would deliver a marked anti-atherogenic effect. However, this was not the case for CETP inhibitors that were initially developed. In fact, the first CETP inhibitor to enter human clinical trials, torcetrapib, caused an excess of deaths and cardio- vascular disease [20]. This led many experts to conclude that the HDL itself was dysfunctional or harmful. However, significant off-target side effects, involving hyperaldosteronism and substantial hypertension, were attributed to torcetrapib [20]. The demonstra- tion of an overall anti-atherogenic effect of CETP inhibition was suggested by the majority of animal studies showing a pro- atherogenic effect of CETP expression [21]. Moreover, multiple large human genetic studies showed that SNPs in the CETP gene, that are associated with increased HDL and reduced LDL choles- terol, are associated with reduced CHD [22e24]. This includes SNPs that likely reduce the function of the promoter region upstream of the CETP gene [24], and most importantly, CETP protein truncating mutations that abrogate the function of CETP [23]. This has permitted further human clinical studies to be performed to eval- uate other members of this class of drugs. Subsequent studies trials with the relatively weak CETP inhibitor dalcetrapib [25] and with the potent inhibitor evacetrapib [26] were halted prematurely because of projected lack of efficacy.Finally, in the largest study of a CETP inhibitor, and the first to goto completion, the potent CETP inhibitor anacetrapib was shown to significantly reduce major coronary events [27].

This study involved 30,449 patients with atherosclerotic cardiovascular disease who were randomized to receive anacetrapib 100 mg daily or placebo on top of effective statin therapy and followed for a median of 4.1 years. The study showed a highly significant reduction (rate ra- tio ¼ 0.91, p < 0.004) in the composite primary endpoint of coro- nary death, myocardial infarction or coronary revascularization [27]. The modest degree of reduction in the primary endpoint likely reflected the fact that control patients, who were highly effectively treated with statins, had an LDL cholesterol of 61 mg/dl, making it ahigh hurdle to show a major incremental effect of CETP inhibition. Merck decided not to seek marketing approval for anacetrapib, seemingly based on a combination of its clinical profile, which included a modest effect on cardiovascular endpoints, prolonged accumulation in adipose tissue, as well as marketing consider- ations. As monotherapy, potent CETP inhibitors lower LDL choles- terol by about 40% and ApoB by about 30%, similar to homozygous CETP deficiency. Thus, other CETP inhibitors lacking adipose accu- mulation could potentially be used to treat CHD, either in combi- nation with statins or as monotherapy. However, further clinical trials of efficacy would likely be required.We and others showed that the ATP binding cassette trans- porters ABCA1 and ABCG1 promote cholesterol efflux from mac- rophages to lipid-poor apoA-1 and HDL particles, respectively [28,29]. Abca1 and Abcg1 are induced in cholesterol loaded mac- rophages as a result of direct promoter activation by LXRs [30,31]. Cholesterol efflux pathways mediated by these transporters exert anti-atherogenic effects by suppressing inflammatory responses in myeloid cells [32]. Abca1/g1 deficient macrophages show height- ened inflammatory responses to lipopolysaccharide, oxidized phospholipids and apoptotic cells in part related to increased cell surface expression of TLR4, increased signaling via MyD88 and TRIF dependent pathways and NADPH oxidase activation; these effects are dependent on increased cellular cholesterol content [33,34]. ABCA1/G1 suppress excessive proliferation of hematopoietic stem cells, monocytosis, neutrophilia and macrophage accumulation in atherosclerotic plaques of hypercholesterolemic mice [35e39]. We developed Abca1fl/flAbcg1fl/fl mice and showed increased athero- sclerosis with myeloid (Myl) or endothelial knockout of these transporters [40,41].

2.Cholesterol efflux pathways suppress inflammasomes and NETosis
Our recent studies in MylABCDKO mice have revealed a major role of ABCA1/G1-mediated cholesterol efflux pathways in suppressing the inflammasome [42]. MylABCDKO mice showed inflammasome activation in macrophages and neutrophils, involving both Nlrp3/ caspase-1 and noncanonical/caspase-11. Unexpectedly, MylABCDKO mice showed prominent neutrophil extracellular traps (NETs) in atherosclerotic lesions, while deficiency of Nlrp3 or Caspase- 1/11 abolished NETosis and reduced lesion area, suggesting a novel role for ABCA1/G1 in suppressing inflammation associated with atherosclerosis. Recent studies in the CANTOS trial have highlighted the importance of inflammasome activation and IL-1b production in human coronary heart disease (CHD) [43], while other studies have suggested a role of NETosis in atherogenesis [44] and plaque instability [45,46]. Thus, our studies showing that HDL and cholesterol efflux pathways can suppress these processes have significant potential translational relevance, especially in condi- tions where ABCA1/G1 are suppressed and HDL levels are low, which may include Type 2 diabetes, chronic kidney disease and ageing [47e53]. In these conditions, treatment of CHD by infusions of cholesterol-poor reconstituted HDL particles, which mediate cholesterol efflux by non-transporter dependent mechanisms, may be particularly beneficial.
The role of the NLRP3 inflammasome in atherosclerosis was first explored by Latz in Ldlr—/- mice [54] and by Tschopp in Apoe—/— mice [55]. Latz et al. found a significant impact on lesion develop- ment and related this to cholesterol crystal formation even in early foam cell lesions, while Tschopp et al. reported no effects on atherogenesis. While some subsequent studies appeared to confirm the initial reports [56,57], our own [42] and other studies [58] found no impact of deletion of key inflammasome components, Nlrp3 or Caspase1/11, on lesion area or morphology in WD-fed Ldlr_/ _ mice. However, when additional mutations that led to macro- phage inflammasome activation were introduced into the Ldlr_/_ model, joint deficiencies with Nlrp3 or Caspase 1/11 showed that the NLRP3 inflammasome does contribute to atherogenesis [42,58].

3.Inflammasome activation in human atherosclerosis
Our mouse studies showed that whole body Abca1 deficiency on the Ldlr—/— background induced inflammasome activation, and this effect was echoed by elevated IL-1 and IL-18 plasma levels (markers of inflammasome activation) in Tangier disease patients, indicating human relevance. This suggests that low HDL (decreasing ABCG1- mediated cholesterol efflux), defective apoA-1 and reduced expression of ABCA1/G1 in monocyte/macrophages may be suffi- cient to induce inflammasome activation in humans. A variety of studies suggest that these conditions may be commonly found in patients with Type 2 diabetes, chronic kidney disease and with ageing [47e53]. Clonal hematopoiesis involving common variants in hematopoietic genes (TET2, JAK2, ASXL1 and DNMT3a) that pre- dispose to hematological malignancy has recently emerged as an important novel risk factor for CHD especially in the elderly [59,60]. Recent studies in mice with myeloid Tet2 deficiency, modeling clonal hematopoiesis, have shown macrophage inflammasome activation, leading to increased IL-1b production and accelerated atherosclerosis [60,61]. Along with the positive outcome of the CANTOS trial [43], involving IL 1b neutralization, this suggests that inflammasome activation is an important contributor to human CHD. However, the conditions promoting inflammasome activation in atherosclerosis and the links between the inflammasome and NETosis need to be more clearly delineated.

4.Perspective
Athero-thrombotic disease remains the major cause of disability and death in the industrialized world. Despite the success of LDL lowering therapies in treatment, there remains a large burden of residual risk. Attempts to treat this residual risk with CETP inhibitors that dramatically raise HDL have only met with modest success. Other approaches that more effectively stimulate cholesterol efflux pathways may be more successful in the future. There is a great deal of excitement about new targets for lowering levels of triglyceride and cholesterol rich remnants that have emerged from human ge- netic studies. The CANTOS trial, in which IL-b inhibition was found to reduce CVD, strongly supports the concept of anti-inflammatory therapies in the treatment of athero-thrombotic disease. However, an excess of deaths from infections may limit the clinical impact of this or other broadly immunosuppressive therapies. This highlights the need to understand more clearly the inflammatory mechanisms that are specific to atherogenesis. Perhaps the most obvious candi- dates for future studies are inflammatory mechanisms that are linked to oxidized phospholipids and to cholesterol uptake and removal from macrophages. Inflammasome activation and neutro- phil NETosis, promoted by macrophage and neutrophil cholesterol accumulation, are emerging mechanisms underlying atherogenic inflammation. Clonal hematopoiesis has recently been discovered as a major non-traditional risk factor for CVD, and likely interacts with hyperlipidemia to promote atherogenic MK-0859 inflammation. In summary, many recent studies have linked hyperlipidemia, defective choles- terol efflux pathways and aberrant hematopoiesis to atherogenic inflammation. However, this area remains poorly understood, ther- apy is challenging, and there is a tremendous need for further research.