Immunity, atherosclerosis and cardiovascular disease

Low density lipoprotein modified by oxidation or enzymatic modification (OxLDL) is present at an early stage. Low density lipoprotein (LDL) penetrates into the intima at the earliest stages of atherosclerosis and binds to the proteoglycan matrix, enabling further modification through oxidation and/or enzymatic modification (OxLDL). Also at later stages, oxLDL and related compounds are ubiquitous in lesions [5, 6]. Therefore, oxLDL could play a role both in atherogenesis and in plaque complications. OxLDL is immunogenic and activates endothelial cells, monocytes/macrophages and T cells [7‐9]. Further, oxLDL is toxic at higher concentrations and could thus be a cause of cell death in lesions [7‐9]. Enzymatically modified LDL could play a major role, and phospholipase 2 (PLA2), which causes such modification, is expressed in both normal arteries and atherosclerotic lesions [10] and can induce activation of dendritic cells [11]. The proinflammatory and immune stimulatory effects of oxLDL are mimicked by inflammatory phospholipids, such as lysophosphatidylcholine (LPC), which is a major phospholipid in atherosclerotic lesions [12, 13]. Other proinflammatory phospholipids implicated in oxLDL, such as LPC, have phosphorylcholine (PC) as an important epitope, which cause these different phospholipids, to different degrees, to interact with the platelet activating factor (PAF)-receptor, which is one mechanism by which oxLDL exerts its effects [14, 15]. Other mechanisms include interaction with Toll-like receptors and scavenger receptors [16, 17].

Not only oxidized and/or enzymatically modified phospholipids are implicated as causes of oxLDL’s pro-atherogenic and pro-inflammatory effects; there are several other possibilities. It has even been suggested that epitopes, such as those exposed during LDL-modification and/or oxidation, represent an evolutionary conserved system of danger-associated molecular patterns (DAMP) in parallell with pathogen-associated molecular patterns (PAMP) [17]. One important example of DAMP in addition to phospholipid (PL)-related epitopes, such as PC, is malondialdehyde (MDA) which is generated during oxidation of LDL. MDA forms adducts on proteins and peptides, carbohydrates and DNA [17].

Modified and oxidized apolipoprotein B (apoB) and cholesterol could also be implicated [17] although putative mechanisms are not as well described as for PC-exposing epitopes.

While much evidence from epidemiological studies indicates that smoking is associated with atherosclerosis and CVD [18, 19], the exact mechanisms by which smoking could cause inflammation in the arteries are not fully elucidated, although increased oxidation of lipids is one interesting possibility [20]. Interestingly, smoking is associated with increased lipid oxidation [20]. Different animal models of smoking and atherosclerosis indicate that smoking promotes atherogenesis [21‐23], and one underlying mechanism is reported to be oxidative stress [24].

In support of inflammatory phospholipids as causes of atherosclerosis are data from clinical studies, where we reported that in conditions with increased atheroslerosis, such as hypertension and systemic lupus erythmatosus (SLE), PC-exposing LDL is increased [25].

Oxidation of LDL, and also increased oxidative stress in diabetes, could promote atherogenesis [26]. Formation of advanced glycation end products (AGEs) could contribute to atherosclerosis (and CVD) in diabetes since AGEs have proinflammatory and potentially atherogenic properties [27‐29].

The role of cell death as a cause of inflammation in atherosclerosis and plaque rupture is complicated and depends most likely on stage of disease, and naturally, on whether cell death is organized as in apoptosis, or not, as in necrosis. It is likely that a defective apoptotic clearance and ensuing necrosis could contribute to inflammation.

Dying cells can activate the innate immune system and induce an inflammatory response with release of the proinflammatory cytokine IL-1beta, activating the inflammasome [30].

According to the danger hypothesis, endogenous factors (DAMPs) released during cell death induce inflammation. DAMPs include high-mobility group protein B1 (HMGB-1), double-stranded DNA, amyloid-β-peptide and heat-shock proteins (HSP) [31]. Even though cell death is not an early event in atherogenesis (fatty streak formation and infiltration of monocytes/macrphages and T cells apparently come first), it could play a role at a later stage to promote inflammation. Whether it plays a role in plaque rupture is possible but not proved.

Antibodies against phospholipids (aPL), especially against cardiolipin (aCL) are well known to be associated with CVD, especially in SLE patients but it has been difficult to find unequivocal evidence of aPL as atherogenic, and there are both negative and positive reports [25]. In a recent study, we did not find an association between aPL and prevalence of atheroslerotic plaques in patients with SLE [32]. Typical of pathogenic aPL is that they are dependent on plasma co-factors, such as beta2-glycoprotein I (beta2GPI) , to promote CVD. Mechanisms could include a direct effect on endothelium and also interference with anti-coagulant proteins such as Annexin A5 [32, 33].

CL has a unique double structure, with four fatty acid chains and is present in bacteria and the inner mitochondrial membrane of eukaryotic cells [34], which is interesting since mitochondria apparently have a bacterial origin [35]. In contrast to aCL, we recently reported that antibodies against oxidized forms of CL (aOxCL) are negatively associated with CVD, low levels giving high risk and high levels low risk [36]. In contrast to aCL, aOxCL are not beta2GPI-dependent [36]. One mechanism could be inhibition of binding and uptake of oxLDL in macrophages [36]. Further, aOxCL and also antibodies against oxidized phosphatidylserine (aOxPS) are negatively associated with atherosclerosis development (unpublished data).

Another example of natural antibodies is those against PC (anti-PC). Several lines of evidence imply that anti-PC could play a role in atherogenesis, both from animal studies, other experimental studies and from clinical cohort studies [37]. Immunization in a mouse model with pneumococcae caused a decrease in atherosclerosis development in parallell with an increase in anti-PC levels, among other antibodies [38]. Both passive and active immunization with PC ameliorates atherosclerosis in mouse models [39, 40]. We have reported in several papers that immunoglobulin M (IgM) anti-PC is negatively associated with atherosclerosis development and risk of CVD. Typically, low levels give rise to increased risk and, in some cases, high levels were also associated with decreased risk [37]. We reported for the first time that anti-PC is a protective marker for atherosclerosis development in humans which was also the case with antibodies against malone dialdehyde LDL (anti-MDALDL) and anti-OxLDL [41].

Mechanisms by which human anti-PC could ameliorate atherosclerosis and CVD include: anti-inflammatory effects, inhibition of pro-inflammatory effects by inflammatory phospholipids [42], inhibition of uptake of oxLDL through scavenger receptors [43] and inhibition of cell death induced by LPC, a major inflammatory phospholipid [44]. In another paper, the inflammatory effect of anti-PC is also confirmed in a mouse model, in which facilitating phagocytosis is described as one mechanism. It is possible that natural IgM, such as anti-PC, could counter atherosclerosis development by binding to dead and dying cells in the lesions, increasing phagocytosis and clearance of obnoxious pro-inflammatory compounds [45].

Low levels of anti-PC could thus be a cause of the inflammation in atherosclerosis, although it is less clear by what mechanisms. We have suggested that a Western life style could contribute, and one underlying factor could be some types of infections which are not prevalent in the West which could raise anti-PC [46]; another could be factors which are relatively new from an evolutionary point of view such as gluten [47]. Genetic factors could also play a role in addition to environmental influences since heritability of anti-PC is 37% [48].

HSPs, especially HSP60 but potentially also other ones, such as HSP70 and HSP90, represent another interesting potential cause of inflammation in atherosclerosis. This could be of great interest since HSPs are immunogenic and T-cell clones recognizing HSP60 are present in both early and late atherosclerotic plaques [49, 50]. HSPs may also activate immune reactions through cross-reactivity with HSP from microorganisms, such as bacteria. This notion is supported by both clinical data with associations between antibodies against HSP60/65 and atherosclerosis, and experimental data where immunization with HSP60/65 aggravates atherosclerosis [51, 52].

HSPs could promote inflammation by other mechanisms as well. Besides being specific T-cell antigens per se, present on antigen presenting cells, HSP and/or peptides thereof could promote immune activation by other mechanisms. HSPs are chaperones and can form immune complexes with other antigens including tumor-derived ones, and these can be presented through class I or class II antigen presenting pathways [53]. HSP can be passively released which may occur in cell necrosis but also actively through exosomes. HSPs could thus play a role in the extracellular space where they could be endogenous ligands, activating the innate immune system, through Toll-like receptors or by association with endotoxin [54]. The mechanisms by which hypertension could cause inflammation in the artery wall are not clear. One possibility is a direct effect on the endothelium, which could become dysfunctional as a response to injury leading to proinflammatory changes [55]. We suggested in previous studies that hypertension could cause inflammation by induction of immunogenic HSP60/65, which is also induced by oxLDL [56, 57]. We reported that HSP70 is a protective factor for development of atherosclerosis among hypertensives, but a putative underlying mechanism is not clear [58].

Infections have been much discussed as potential causes of immune activation and inflammation in atherosclerosis. In early studies, before the rise of the lipid hypothesis, pathologists and others who observed the lesions thought they could be of infectious origin based on the microscopic and macroscopic features of atherosclerosis.

Among the most promising candidates that are present in plaques, promote atherosclerosis in animal studies and have associations with disease in humans are Chlamydia pneumoniae (CP), periodontal organisms including Porphyromonas gingivalis (PG) and Aggregatibacter actinomycetemcomitans (AA), Helicobacter pylori (HP) and cytomegalovirus (CMV) [59].

One important starting point underlying the hypotehsis that infections play a role in atherosclerosis was early studies of CP, demonstrating the presence of this pathogen in atherosclerotic plaques [60] and an association between cardiovascular and antibody titer [61‐63] by P Saikku and coworkers and others. As is the case with other pathogens, also discussed, such as CMV, there are also studies in which no such associations were demonstrated [59].

A more formal test of the hypothesis has been done with treatment with antibiotics which have an effect on CP. However, four large studies were not positive and did not support the notion of a causative effect of CP on CVD [64‐66]. However, there could be other reasons for the negative results. One is that chronic CP may be difficult to treat irrespective of CVD; another is that the treatment trials have been performed on patients with late stage disease, and it is possible that earlier stages would be more responsive [59].

Periodontal microorganisms, such as PG and AA, are also interesting candidates. Even though clinical/epidemiological studies show associations, there are many confounding factors which are difficult to control for, maybe more for these agents than is the case with other pathogens in this context. A recent scientific statement from the American Heart Association supports an independent association between periodontal disease but available data do not support causation, although intervention does decrease systemic inflammation and improves endothelial function [67].

CMV belongs to the Herpes virus group which is very common in the general population; this makes studies of associations difficult to interpret. There are interesting reports of an association between active CMV infection and transplant complications including vasculopathy [68]. A causal relationship between CMV infection and transplant vascular complications seems to be more plausible than associations with atherosclerosis per se. CMV is reported to be present in atherosclerotic lesions in many but not all studies [59]. Of note, CMV has also been documented in healthy arteries [69] which could also be taken as an indication that CMV may be an innocent by-stander. On the other hand, there are interesting properties in CMV which could make it a plausible candidate to be a contributing factor in atherosclerosis. For example, CMV infection induces migration of arterial smooth muscle cells [70].

HP, a well known cause of gastritis and gastric ulcer, may also be implicated in CVD and atherosclerosis, although animal experiments are less clear than is the case with CP. Viable bacteria have not been unequivocally demonstrated from lesions. However, there are positive reports of a reduction of CVD after eradication of HP. HP does not promote a local inflammatory reaction to the same extent as do CP, CMV and several other implicated pathogens [59]. Of note, viable HP have not been isolated from atherosclerotic plaques, and mouse experiments have not given any clear indication that HP is a cause of atherosclerosis [71].

Interestingly, treatment regimens for both CP and HP had a positive effect on clinical CVD events [72]. However, as discussed, larger studies for CP were not successful and are needed also for HP.

Other infectious agents that have been discussed and reported as potential causes of CVD and atherosclerosis include HIV, Epstein-Barr virus (EBV), influenza, Mycoplasma pneumoniae and Streptococcus pneumoniae. Another interesting case, in which there is evidence from human studies, is Borrelia [73], although experimental data or plaque data are not available to the best of my knowledge. However these appear to be less supported by evidence; especially, there is no convincing data from human studies [59].

Taken together, even though the infectious hypothesis in CVD and atherosclerosis has been studied for some decades, there is still little direct, though rather much circumstantial, evidence of a causative role of infectious agents.

Another interesting development, in which infections and atherosclerosis/CVD could be related in a more indirect way, is recent findings implicating the intestinal bacterial flora [74, 75]. Intestinal microbiota metabolism of choline and phosphatidylcholine produces trimethylamine (TMA), which is metabolized to proatherogenic trimethylamine-N-oxide (TMAO). Recently, it was demonstrated that metabolism by intestinal microbiota of dietary l-carnitine in red meat produces TMAO and accelerates atherosclerosis in mice. This finding suggests another interesting link between gut flora and atherosclerosis [76].

Another aspect of immunity and inflammation that could be relevant for atherosclerosis is modulation of the host by microorganisms in order to promote their own survival. PC has a central role, being exposed on different types of pathogens including Gram-positive and –negative bacteria, nematodes and protozoa. In this context, the pathogens use PC to create a favorable environment for themselves in the host. One interesting possibility is therefore that PC exposed in the atherosclerotic plaque in fact modulates the local immune reaction into an unresolved chronic inflammation, in a similar manner as PC-exposing pathogens do in chronic infections [77].

Further research is needed to clarify if the described associations are causative in humans.

Even though it remains to be demonstrated that infections play a direct role in atherogenesis, they could still be of great importance indirectly. By being present in lesions, they could stimulate an ongoing inflammatory process, which in the long run may lead to increased atherosclerosis and ensuing CVD. Further, it is possible that the total infectious burden could be a risk factor for increased atherosclerosis and CVD, possibly through promotion of systemic inflammation, platelet aggregation and endothelial dysfunction, which in turn could influence atherogenesis [59, 78].

Raised levels of C-reactive protein (CRP) have been implicated as a risk marker for atherosclerosis and CVD in many studies, although it is not clear if CRP could play a causative role. Also, cytokines, such as IL-6, raised systemic levels of oxLDL and have been implicated in many studies [4]. Another example of an interesting group of inflammatory compounds is lipid mediators, such as leukotrienes, which are present in advanced atherosclerotic plaques [79, 80].

Another interesting example of indirect effects by other types of inflammation is the relationship between chronic inflammatory and autoimmune diseases and atherosclerosis (and CVD). Here the evidence is strongest for the prototypic autoimmune disease SLE where the risk of CVD is very high. According to one report the risk increased 50 times [81] and several other studies have also reported a high risk (although from low levels since young and middle aged women without SLE (or familial hyperlipidemia) very seldom develop CVD, and, even less, advanced atherosclerosis [25]. Also in rheumatoid arthritis (RA), an increased prevalence of atherosclerosis has been reported, although the evidence appears less clear than in SLE. In other rheumatic diseases, an increasing number of papers indicate an increased atherogenesis [25]; which is confirmed in a recent meta-analysis which demonstrates that rheumatic diseases increase the risk of atherosclerosis [82].

It is interesting to note that this discussion is not new. Already in the first half of the 19th century, the legendary pathologists Rokitansky and Virchow (the former as the first) reported that atherosclerosis is an inflammatory process although their opinions differed somewhat, since Rokitansky thought inflammation in atherosclerosis was a secondary phenomenon, but Virchow proposed it was a primary factor [83]. Non-traditional risk markers, such as low anti-PC and inflammation per se appear to play a role, in addition to traditional ones, such as hypertension, dyslipidemia and to a varying degree, smoking [25].

Although opinions may differ somewhat as to what comes first in the development of atherosclerosis, it appears likely that different factors could act in concert. It is also possible that there are differences between different individuals, where some may have atherosclerosis with more inflammation than others. It is also possible that infection could play an important role among subgroups of individuals and patients.

Recently, new aspects of immune mechanisms in atherosclerosis have been discussed in addition to intimal immune reactions which have been in focus. Even though the adventitial inflammation in atherosclerosis was noted already by Virchow and Rokitansky [83], and cellular immune infiltrates were described in the early 1960s [84], this phenomenon has only recently been more thoroughly studied [2, 85]. The adventitia appears to be more complex than the intima and media also in normal arteries. Many cell types, including fibroblasts, dendritic cells, monocytes/macrophages, mast cells and T-cells are present. There are also nerve endings, small vessels (vasa vasorum), endothelial protenitor cells and the like in the intima, a matrix [86].

Data from mouse models of atherosclerosis are interesting, with B cells and T cells present in the adventitia forming inflammatory follicle-like structures [87]. An important role played by adventitial lymphocytes is suggested by several studies [2]. For example, proinflammatory IL-17A–producing T cells are present in the adventitia. Blockade of IL-17A led to a reduction in aortic macrophage accumulation and atherosclerosis [88].

Less is known about the role of the adventitia in human atherosclerosis,although immune competent cells including T cells and B cells are present in adventitial lymphoid follicles both in the aorta and in coronary arteries [2]. One important question is how the adventitia interacts with the intima in atherosclerotic lesions. It is interesting to note that B-cells are not common in the intima, as compared to findings from adventitia in atherosclerotic lesions. The role of B-cells in atherosclerosis most likely depends on subsets, at least according to studies in mouse models, with B2 lymphocytes being atherogenic and B1 lymphocytes protective against atherosclerosis [2].

Another factor in CVD, plaque rupture and late stage atherosclerosis may be intimal hemorrhage. Erythrocyte membranes are abundant in late stage lesions and could be a proinflammatory factor, increasing the risk of plaque rupture [89, 90].

Interestingly, statins, one of the most successful medicines in history from a commercial point of view, and clearly representing also a therapeutic improvement, may work to an unknown extent due to their pleiotropic, especially anti-inflammatory, properties. This effect is a ‘side-effect’ of these inhibitors of HMG CoA reductase, influencing prenylation as one mechanism. Further, the immunomodulatory effects of statins, directly interfering with major histocompatibility complex MHC class II presentation, have been described [102]. Interestingly, statins have been reported to be beneficial in RA; other anti-inflammatory mechanisms could be decreasing LDL-oxidation, and immune modulatory effects, decreasing MHC class II interaction with antigen [103]. In line with this, the Jupiter study demonstrated that statin treatment may be beneficial for individuals with raised high sensitivity CRP but normal LDL [104]. In a recent review by Ridker, it is stated that: ‘it is impossible in any statin trial to establish whether the clinical benefits of treatment are due to LDL-reduction alone, to inflammation inhibition, or to a combination of both processes’ [105]. It is striking that the possibility is raised that a novel and successful therapy developed to decrease cholesterol and specifically LDL, may work partly or even only, by another mechanism, namely inhibition of inflammation.

Other interesting potential therapies include both unspecific and specific approaches. In RA, treatment with methotrexate weekly is routinely used, with beneficial effects on patients. A recent meta-analysis indicates that the risk of CVD is decreased in patients with RA treated with methotrexate [106]. Animal studies in which methotrexate decreases atherogenesis add support to methotrexate as a possible anti-inflammatory therapy in CVD [107]. In the Cardiovascular Inflammation Reduction Trial (CIRT) low dose methotrexate (target dose 20 mg/week) is tested for reduction of major CVD events among post-myocardial infarction patients with diabetes or metabolic syndrome [104].

Initially in RA, treatment with biologics, such as tumor necrosis factor (TNF)-inhibitors, has proven to be successful, and novel biologics also include inhibitors of other cytokines. However, there are different opinions as to whether biologics, such as anti-TNF, are beneficial from a cardiovascular point of view, although a recent study in which there was a decrease of CVD in RA adds support to this possibility [108, 109].

The Canakinumab Anti-Inflammatory Thrombosis Outcomes Study (CANTOS) investigates if IL-1β inhibition could reduce the risk of MI, stroke, and cardiovascular death among stable coronary artery disease patients with high risk of CVD due to persistent elevations of CRP (≥2 mg/L) [105, 110].

Other potentially interesting treatment moieties include inhibition of inflammatory lipid mediators, such as PAF [111]. We recently reported that Annexin A5, an anti-thrombotic plasma protein, is anti-inflammatory and inhibits atherosclerosis development and also improves endothelial function in a mouse model. Annexin A5 could thus represent another possible therapy [112]. Likewise, inhibition of PLs, such as lipoprotein associated PLA2, is of interest. Studies of the inhibition of Lp-PLA(2) activity with darapladib in patients after an acute coronary syndrome are under way [113].

One example of possible immune modulatory, and thus potential anti-inflammatory, treatment against atherosclerosis and/or CVD is immune therapy against epitopes from oxLDL. In the mid-1990s, it was demonstrated that immunization with modified forms of LDL ameliorated atherosclerosis [114], which showed for the first time the possibility of inhibition of atherosclerosis development by immunization.

As discussed, during LDL-oxidation, various chemical moieties are formed, including both fragmented apoB and oxidized phospholipids. One approach is to target the apoB component. Peptides from apoB with promising immunomodulatory properties have been demonstrated to decrease atherosclerosis development in animal models [115‐117]. However, a recent clinical study did not show any positive effect in humans [118]. The primary endpoint was the relative change in inflammatory activity in an index arterial vessel after twelve weeks, as measured by fluorodeoxyglucose-positron emission tomography/computed tomography (FDG-PET/CT) imaging. Further studies with other end points and using other, perhaps more established, techniques could clarify if this approach could still be promising in humans.

Another line of research and potential treatment is based on the phospholipid moiety as a target for treatment with monoclonal antibodies, specifically PC. As discussed above, this notion is supported by cohort studies, animal and in vitro experiments [37, 39, 40, 42‐44]. Mechanisms include anti-inflammatory [42, 119], inhibition of cell death [44] and decreased uptake of oxLDL in macrophages [43].

A more unspecific method of immunomodulation is administration of Igs. This approach has shown promising results in animal studies with human Ig from pools of many donors, such as used in intravenous Ig treatment (IVIG) [120].

Yet another interesting possibility is to ameliorate atherosclerosis by immunomodulation with HSP. To the best of my knowledge, HSP-immunization demonstrated for the first time that atherosclerosis can be influenced by immunization. In a study using a rabbit model, such therapy increased atherosclerosis development [51].

Interestingly, nasal immunization with HSP65 led to induction of immune tolerance in a rabbit model, suggesting that immunization mucosally with Hsp65 protein could be a promising therapeutic method for atherosclerosis [121].

In another experiment, a vaccine was designed to target both HSP65 and cholesteryl ester transfer protein (CETP) in order to obtain both a positive effect on the immune reactions relevant in atherosclerosis and on blood lipids. Here, also, nasal immunization ameliorated atherosclerosis in a rabbit model [122].

There are also conflicting data. In a recent study, subcutaneous immunization with HSP-65 in apoE—mice actually reduced atherosclerosis [123]. Further, in another interesting study immunization with human HSP60 (with or without combination with apoB peptides) led to decreased atherosclerosis [124]. It is thus presently not clear how HSP-immunization could work as therapy against atherosclerosis development even in animal models. It is possible that differences in modes of presentation of the antigen could play a role. The potential anti-inflammatory treatments for atherosclerosis are summarized in Table 1.