The role of T and B cells in human atherosclerosis and atherothrombosis

Abstract

Far from being merely a passive cholesterol accumulation within the arterial wall, the development of atherosclerosis is currently known to imply both inflammation and immune effector mechanisms. Adaptive immunity has been implicated in the process of disease initiation and progression interwined with traditional cardiovascular risk factors. Although the body of knowledge regarding the correlation between atherosclerosis and immunity in humans is growing rapidly, a relevant proportion of it derives from studies carried out in animal models of cardiovascular disease (CVD). However, while the mouse is a well-suited model, the results obtained therein are not fully transferrable to the human setting due to intrinsic genomic and environmental differences. In the present review, we will discuss mainly human findings, obtained either by examination of post-mortem and surgical atherosclerotic material or through the analysis of the immunological profile of peripheral blood cells. In particular, we will discuss the findings supporting a pro-atherogenic role of T cell subsets, such as effector memory T cells or the potential protective function of regulatory T cells. Recent studies suggest that traditional T cell-driven B2 cell responses appear to be atherogenic, while innate B1 cells appear to exert a protective action through the secretion of naturally occurring antibodies. The insights into the immune pathogenesis of atherosclerosis can provide new targets in the quest for novel therapeutic targets to abate CVD morbidity and mortality.

Introduction

Cardiovascular disease (CVD) is currently the most common cause of death globally, accouting for nearly one-third of deceases worldwide 1. While many diseases can affect the cardiovascular (CV) system, atherosclerosis is by far the most common, and atherosclerosis-related conditions, including acute myocardial infarction (AMI) and stroke, dominate death and disability statistics throughout the world 2. Atherosclerosis is a chronic disease that affects medium- and large-sized arteries. By causing severe stenosis or thrombosis in arteries it leads to regional ischaemic damage, which may result in life-threatening conditions 3. According to the prevailing view, cholesterol accumulation in the vessel wall, in particular the fraction assembled in low-density lipoproteins (LDL), has a central role in atherogenesis. The association between cholesterol and human atherosclerosis has been proved by epidemiological studies 4,5, and further research into cholesterol metabolism has led to the implementation of preventive programmes that avert thousands of deaths worldwide 6. Notably, the introduction of hydroxy-methyl glutaryl CoA (HMG-CoA) reductase inhibitors, i.e. statins, a class of cholesterol-lowering drugs, decreased the incidence of major CV events, including stroke and AMI, by approximately one-third 6. Similar results have been obtained with campaigns aimed to reduce systemic hypertension and cigarette consumption 7,8. Despite the outstanding results obtained by such preventive programmes, a substantial residual risk needs to be addressed to further abate CVD incidence and related mortality. New insights in pathogenic elements of atherosclerosis beyond lipid profile have assumed major interest in the quest for novel and effective therapeutic interventions to treat atherosclerosis. Population-based observations are useful for this purpose: using genomewide association studies, for example, a locus on chromosome 9p21 has been identified repeatedly as one of the main determinants of complex CVD, and risk-associated polymorphisms seem to be responsible for up to 21% of attributable risk of AMI independently of well-established CV risk factors (CVRFs) 9,10. Most interestingly, compelling evidence building continuously over recent years links inflammation and adaptive immune response to atherogenesis 11. Immune mechanisms also appear to underlie the pathogenesis of traditional CVRFs, including hypertension, diabetes mellitus (DM) and metabolic syndrome 12,13, and to be dysregulated on a genetic basis [i.e. interleukin (IL)-6 receptor variant] 14, at least in some individuals. Although the body of knowledge regarding the correlation between atherosclerosis and immunity is growing at an untoward speed, a relevant proportion of it derives from studies carried out in animal models of CVD, the mouse being the most widely studied due to its low cost of maintenance, fast reproduction rate and easy genetic manipulation 15. While experimental research has provided key mechanistic insights, it remains unclear whether the identified pathways may contribute to human atherosclerosis, due mainly to the intrinsic genomic and environmental differences between humans and animal models 16. The present review addresses established knowledge and evidence gaps in the aetiology of atherosclerosis with particular focus on human findings, which can be attained through a number of techniques summarized in Table 1, most notably including pathological examination of post mortem or surgical atherosclerosis samples or through the analysis of the immunological profile in peripheral blood, and we will try to underline their clinical significance and potential therapeutic implications, alongside some key pathogenic evidence derived from experimental studies.

Table 1.

Summary of the main techniques and biological materials employed in the study of the relationship between adaptive immune response and atherosclerosis in humans

Technique	Material	Attainable results
Classical histology	Surgical or post mortem pathological sample of atherosclerotic lesions	Purely morphological analysis of the plaque and its cellular components
Immunohistochemistry and immunofluorescence	Surgical or post mortem pathological sample of atherosclerotic lesions	Identification-specific cellular subtypes or presence of proteins by exploiting antibody–antigen reaction
Real-time polymerase chain reaction (RT–PCR)	Peripheral blood mononuclear cells (PBMC)-derived cDNA	Quantification of mRNA of proteins relevant to adaptive immunity, generally cytokines, as an indirect measure of their expression by circulating immune cells
Enzyme-linked immunosorbent assay (ELISA)	Plasma	Direct quantification of relevant soluble proteins, generally cytokines
Flow cytometry	Whole blood or PBMC	Characterization and quantification of immune cells subpopulations through the identification of surface markers

Innate immunity recruits adaptive response in atherosclerosis

The recognition of disease-specific antigens by germline-encoded pattern recognition receptors initiates a local inflammatory response against pathogens or components of the host itself. Inflammatory responses are usually adaptive, either killing the eliciting pathogen or contributing to homeostasis by stimulating the removal of damaged tissue and dead cells by phagocytes. Conversely, inflammation may be pathogenic by fuelling local tissue damage itself: such a maladaptive inflammatory process appears to be a relevant driving process in atherosclerosis initiation and progression 17. The antigens involved in inflammation initiation in atherosclerosis are only recently beginning to be elucidated. Infectious agents may contribute to provoking a response, but a primary role is currently considered unlikely 18–20, although epidemiological studies provided evidence of an association between recent systemic respiratory and urinary tract infections and the first AMI or stroke based on 39 546 first CV events 21. By contrast, influenza, tetanus and pneumococcal vaccinations do not produce a detectable increase in the risk of vascular events 21. Other potential antigens, such as heat shock proteins (HSPs) 22 and cholesterol crystals 23, have been implied, but as both require pre-existing tissue damage they are not likely to initiate the primary inflammatory cascade. At present, however, much evidence suggests that potential major antigens involved in atherosclerosis consist of neoepitopes generated by either enzymatic or non-enzymatic oxidation reactions that occur when oxidized LDL (oxLDL) form in the vessel wall or when cells undergo apoptotic death 24. The resulting oxidation-specific epitopes (OSEs) constitute a signature of oxidative damage and are thus the target of a common set of pattern recognition receptors, which renders them proinflammatory and immunogenic 24. Furthermore, other potential antigens derived by apoptotic cells in the plaques can further promote the progression of the atherosclerotic plaque, and defects in apoptotic cell clearance can sustain atherogenesis 25. For instance, the CDKN2B gene, encoding the cyclin-dependent kinase inhibitor 2B that has reduced expression in human carriers of the 9p21 risk allele, has been demonstrated to regulate the clearance of apoptotic debris 26,27. Its loss of function can increase the size of plaque and its lipid core in mice, decreasing the successful engulfment process (efferocytosis) mediated by phagocytic cells 27, thus leading to an inflammatory danger response by apoptotic cells evading clearance 28. Importantly, the innate immune response is accompanied by secretion of chemokines and cytokines as well as antigen presentation, which together co-operate to initiate the definitive adaptive immune response. Dendritic cells (DCs) in particular are the specific cells in charge of initiating of antigen-specific immunity and tolerance 29. DCs drive the maturation and polarization of naive T cells through surface exposure of antigenic peptides in the context of a major histocompatibility complex (MHC) class II molecule, alongside the exposure of appropriate membrane-bound signalling molecules and the secretion of cytokines 29. Mature DCs, i.e. those that have undergone stimulation of their pattern recognition receptors, induce T cell activation and polarization by providing appropriate co-stimulation 30. Conversely, immature DCs as well as DCs receiving strong anti-inflammatory signals, i.e. through the action of cytokines such as transforming growth factor (TGF)-β or IL-10, do not provide co-stimulation and thus induce T cell anergy or death, or skew the response towards a regulatory phenotype 30. In the mouse, a cellular immune response to apolipoprotein (Apo) B100 of native LDL has been identified as an important player in atherosclerosis 31. Selective MHC-II deficiency on DC in LDL receptor null (LDLR^−/−) chimeric mice showed reduced plaque formation, further supporting a pro-atherogenic role for DCs and identifying a critical role for MHC-II-restricted antigen presentation by DCs in driving pro-atherogenic T cell immunity 32. In humans, patients with symptomatic coronary artery disease (CAD) showed increased lesional levels of so-called homeostatic chemokines (CCL19 and CCL21) that, together with their receptor CCR7, play a pivotal role in lymphocyte trafficking by promoting infiltration of T cells and encounter DCs in lymphoid tissue 33. Furthermore, DCs are found in the intima and adventitia of normal human arteries, and their number was shown to increase with the formation and the progression of atherosclerotic lesions 34,35, supporting the hypothesis that a stronger activation of adaptive immune response parallels the progression of atherosclerotic disease.

T cell response in atherosclerosis

The hypothesis that adaptive immunity operates in human atherosclerosis has spread from studies carried out in the 1980s, demonstrating widespread expression of the MHC-II molecule human leucocyte antigen D-related (HLA-DR) in human atheroma 36, alongside a large number of CD3⁺ T cells 37. Most T cells within human atherosclerotic plaques display an effector memory phenotype, a large proportion of which show signs of activation, and approximately two-thirds of them are CD4⁺ T helper cells (Th) bearing the αβ T cell receptor (TCR) 38. CD8⁺ cytotoxic T cells are also abundant in human atherosclerotic plaques. T cells are among the first cells to be recruited within the atheroma, and they are enriched in unstable plaques, i.e. atherosclerotic plaques prone to rupture and subsequently causing thrombosis, embolism and acute cardiovascular manifestations 39,40. Beyond macrophage infiltration, atherosclerotic lesions of patients suffering acute coronary syndromes (ACS) are infiltrated with oligoclonal T cells, demonstrating an ongoing, antigen-driven immune response 41, and specific activated T cell subsets have been recognized with monoclonal TCR, such as HLA-DR⁺ and CD28^null T cells 42,43. As underlined previously, a key unresolved question is the identification of relevant antigens recognized by these cells. While the inflammatory milieu of the plaque may recruit an unspecific heterogeneous population of polyclonal cells, most T cells isolated and cloned from human plaques respond to oxLDL or other antigens under current investigation, i.e. antigenic determinant of the bacterial wall 44, in a HLA-DR-restricted manner 45. Furthermore, oral viridans streptococci DNA was measured in 78% of thrombi of patients with AMI, underscoring the fact that activation of inflammatory pathways in ACS is not confined to coronary lesions but is also contained in the thrombus 46. Many such antigens can be found systemically. Thus, effector T cell responses are most probably initiated in secondary lymphoid organs such as lymph nodes and the spleen, and activated T cells are thought to recirculate to atherosclerotic lesions where putative antigens are present. The finding that patients with CAD have an increase of chemokine receptors involved in T cell plaque recruitment, i.e. CCR5 and CXCR3, on the CD4⁺ T cell surface supports this hypothesis 47. Of note, inhibition of chemokine receptors that are involved in the recruitment of circulating T cells into human plaque such as CCR5 and CXCR3 48,49 attenuates atherosclerotic lesion formation in animal models 50,51. Therefore, atherogenic T cell responses can be considered systemic, providing the rationale for the study of T cells both within the plaque and systemically, through the analysis of T cells profile in peripheral blood (Fig. 1). Table 2 summarizes the main findings concerned with T cell responses in human atherosclerosis.

Figure 1.

T cells and atherosclerotic plaque development. Within secondary and tertiary lymphoid organs, activated dendritic cells polarize naive CD4⁺ T helper (Th) cells in different types of effector T cell through antigen presentation alongside secretion of cytokines and co-stimulation. Th1, Th2 and Th17 subsets are categorized by the cytokine they release. Tolerogenic dendritic cells prime naive CD4⁺ cells toward a regulatory phenotype (regulatory T cells: T_regs). The ontogeny of CD28^nullCD4⁺ is currently a matter of active investigation. Upon acquisition of an effector memory phenotype, characterized by the expression of CD45RO, human leucocyte antigen D-related (HLA-DR) and chemokines receptor CCR5, activated T cells recirculate in peripheral blood. The secretion of cytokines and chemokines from the plaque drives their homing to the atherosclerotic lesion, where they exert diverse actions. Th1 cells promote endothelial disfunction, lipid accumulation in macrophages with subsequent foam cell formation, and cell death through the secretion of their master cytokine interferon (IFN)-γ, exerting a pro-atherogenic effect. The role of Th2 has not yet been elucidated, but appears to be atheroprotective, due possibly to inhibition of Th1 cells. The role of Th17 is currently under intense investigation. T_regs appear to be atheroprotective through the induction of tolerance, the inhibition of atherogenic T cell subsets and suppression of inflammation. CD28^nullCD4⁺ exert a pro-atherogenic action through the secretion of proinflammatory cytokines and cell-mediated cytotoxicity. They were shown to resist T_reg inhibition. Human atherosclerotic plaques are also infiltated with CD8⁺ cytotoxic T cells, which exert an atherogenic effect through induction of cell death. SMC = smooth muscle cells; IL-4: interleukin 4; IL17: interleukin 17; FoxP3: forkhead box protein 3; OX40: CD 131.

Table 2.

Summary of main findings concerned with T cell-mediated immune response in human atherosclerosis

Intraplaque T cells
Subtype	Finding		Number of patients	Sample	Reference
Th1	Most abundant subtype, oligoclonal expansion correlates with lesion progression and inflammation		10 (9 carotid arteries, 1 femoral artery)	Carotid artery and femoral artery endarterectomies	Frostegard, 1999 52
Th17	Controversial, local production of IL-17A, a master cytokine, correlates with inflammation and instabilization		79	Carotid endarterectomy	Erbel, 2011 53
CD4+CD28null	Oligoclonal expansion in plaque inflammation and instability		1	Coronary artery post mortem specimen	Liuzzo, 2000 43
Tregs	Reduced in early stage lesions		45	Aorta, carotid artery and femoral artery specimens	De Boer, 2007 54
	Increased in unstable, symptomatic lesions		57	Carotid endarterectomy	Patel, 2011 55
	Increased with TCR restriction in thrombus aspirates from ACS patients		16	Thrombus aspirates	Klingenberg, 2014 56
NK T cells	Likely to be pro-atherogenic		7	Carotid endarterectomy	Bobryshev, 2005 57

Circulating T cells
Subtype	Markers	Investigated condition	Finding	Number of patients	References
Th1	CD3+IFN-γ+	ACS and stable CAD	Higher activation in ACS compared with stable CAD. Higher activation in ACS and stable CAD compared to controls	28 ACS; 18 stable CAD; 16 controls	Methe, 2005 58
	CD3+CD56−CD4+IFN-γ+	ACS and stroke	Inverse correlation with ACS and stroke risk evaluated prospectively	700	Engelbertsen, 2013 59
	CD4+CCR7−	Stable and unstable angina	Increased proportion in stable and unstable angina when compared to controls	40 stable angina; 40 unstable angina; 20 controls	Damas, 2007 33
	HLA-DR+CD3+CD4+CD45RA−CD45RO+CCR7−	Carotid IMT and CAD	Correlation with IMT and association with CAD and cardiovascular risk factor	183 carotid group; 130 CAD group	Ammirati, 2012 60
	CD3+HLA-DR+	Unstable and stable angina	Increased in unstable angina compared to stable angina and controls	23 unstable angina; 13 chronic stable angina; 6 controls	Caligiuri, 2000 42
	CD3+HLA-DR+	Unstable and stable angina	Increased in unstable angina compared to stable angina and controls	29 stable angina; 36 stable angina; 30 controls	Neri-Serneri, 1997 61
CD4+CD28null	CD4+CD28null	Unstable angina and stable CAD	Increased in unstable angina compared to stable CAD and controls	25 unstable angina; 25 stable angina; 21 hospitalized controls; 20 healthy controls	Liuzzo, 1999 62
	CD4+CD28null	ACS, stable angina and CVRF without CAD	No differences among groups and with controls	20 ACS; 30 stable angina; 22 CVRF no CAD; 16 controls	Teo, 2013 63
TCR zeta-dim	TCR zeta-dim	ACS and chronic stable angina	Increased proportion in ACS patients compared to chronic stable angina. Increase was higher in patients with higher C-reactive protein. Subsets enriched with TCR zeta-dim where CD8+, NK and CD4+CD28null	66 ACS; 32 stable angina	Ammirati, 2008 64
Th17	CD4+IL-17A+	ACS and Stable CAD	Increased in AC with respect to stable CAD and controls	43 (ACS); 22 (stable CAD); 20 (controls)	Cheng, 2008 65
	CD4+IL-17 +	ACS and stable CAD	No association	44 (ACS); 20 (stable CAD); 24 (controls)	Zhao, 2011 66
Th17/Th1	CD4+IL-17+IFN-γ+	ACS and stable CAD	Increased in ACS	44 (ACS); 20 (stable CAD); 24 (controls)	Zhao, 2011 66
Th2	CD3+CD56−CD4+IL-4 +	Carotid mean IMT and area, stroke and ACS	Inverse correlation with carotid mean IMT and area, and with the risk of stroke and ACS evaluated prospectively	700	Engelbertsen, 2013 59
		ACS recurrence	Elevated in patients suffering from ACS recurrence during 2-year follow-up	120	Liuzzo, 2007 67
Tregs	CD4+CD25+FoxP3+	ACS and stable CAD	Reduced number in ACS, no correlation with stable CAD	40 ACS; 18 stable CAD; 20 controls	Han, 2005 68
	CD4+CD25high	ACS and stable CAD	Reduced number and functionally compromised in ACS, no correlation for stable CAD	40 ACS; 28 stable CAD; 28 controls	Mor, 2006 69
	CD4+CD25hiCD127lo	Carotid IMT, stable CAD, ACS (ST-elevation and non-ST-elevation AMI)	No correlation with IMT or IMT progression. No correlation in stable CAD. Reduced in non-ST elevation AMI, increased in ST-elevation AMI	113 carotid group; 200 coronary group (75 controls, 36 stable angina, 50 non-ST elevation AMI, 39 ST-elevation AMI)	Ammirati, 2010 70
	CD3+CD4+CD25hiCD127loCCR5+	Carotid IMT, stable CAD, ACS (ST-elevation and non-ST-elevation AMI)	No correlation with either CAD or carotid IMT	113 carotid group; 150 coronary group (25 controls, 36 stable angina, 50 Non-ST elevation AMI, 39 ST-elevation AMI)	Ammirati, 2010 70
T effector memory	CD3+CD4+CD45RA−CD45RO+CCR7−	Carotid IMT and CAD	Correlation with IMT and association with CAD and cardiovascular risk factor	183 carotid group; 130 CAD group	Ammirati, 2012 47
	CD4+CD45RA-CD45RO+	Carotid IMT and coronary arteries calcifications	Positive correlation with carotid IMT and coronary arteries calcifications	912	Olson, 2013 71
	CD4+CD45RO+	Carotid IMT	Positive correlation with carotid IMT	557	Tanigawa, 2003 72

IFN = interferon; IL = interleukin; Th = T helper; NK = natural killer; TCR = T cell receptor; ACS = acute coronary syndromes; CAD = coronary artery disease; HLA-DR = human leucocyte antigen D-related; IMT = intima medial thickness; CVRF = cardiovascular risk factor; FoxP3 = forkhead box P3; AMI = acute myocardial infarction.

Intraplaque T cell subsets

CD4⁺ Th cells are abundant in atherosclerotic plaques. In response to combined stimulation with antigens, co-stimulators and specific cytokines, naive T cells differentiate into distinct effector or Th subsets, distinguished by cytokines they produce. The three best-characterized Th subsets are Th1, which secretes interferon (IFN)-γ, Th2, which secretes IL-4, IL-5 and IL-13, and Th17, which secretes IL-17 and IL-22. Chronic or repeated antigen exposure, which occurs during the course of atherosclerosis, generally results in the emergence of a dominant Th subset. In particular, compelling evidence has pointed to a role for Th1 and IFN-γ in promoting atherosclerosis and inflammation. Th1 is the most abundant T cell subtype in human atherosclerotic lesions 52, where they display signs of activation. In particular, they have been shown to secrete IFN-γ, TNF-α and IL-2, and to proliferate in situ 38. Of importance, these cells secrete IFN-γ upon stimulation with oxLDL and LDL 45. IFN-γ was in fact shown to promote atherosclerotic lesion development and destabilization in a variety of ways, including altering endothelial function, recruiting inflammatory cells within the lesion and interfering with cholesterol export from cells within the lesions 73. IL-4, one of the master cytokines of Th2 subset, effectively inhibits Th1 differentiation and subsequent IFN-γ secretion 74, pointing to a possible protective role with regard to atherosclerosis, but definitive pathological evidence in humans is currently lacking 75. The association between variants near the IL-5 gene locus and CAD has been demonstrated 76, suggesting a potential role of the Th2 subset in the modulation of development and progression of CAD. A putative atheroprotective effect has been suggested for IL-5 due to its negative correlation with carotid intima media thickness (IMT), a marker of subclinical atherosclerosis 77. Th17 subsets were reported more recently to be associated with atherosclerosis 78,79, but their role remains debated. In particular, the production of IL-17A by lesional T cells was shown to associate with plaque inflammation and instability 53. In humans, a CD28^null subset of CD4⁺ T cells proliferate in the setting of inflammatory disease, cytomegalovirus infection and advanced age. These cells elaborate proinflammatory cytokines, including IFN-γ, and exhibit cytotoxicity 80. Clonally expanded CD28^null CD4⁺ cells were found in unstable coronary plaques, where they may sustain and augment the inflammatory process 43. Many of these cells specifically recognize HSP60, and display a high expression of members of the TNF family receptors, e.g. OX40 (CD 131), which may act as alternative co-stimulatory receptors 81. In addition, these cells were shown to be resistant regulatory T cells (T_regs) to suppression in vitro 81. They comprise a number of CD4⁺ T cell subsets with immunosuppressive properties that play an essential role in self-tolerance and protection against autoimmunity 82. T_regs comprise approximately 1–5% of all T cells within atherosclerotic lesions, which is less than ≤25% found generally in other chronically inflamed tissues 54. While some studies show a reduced number of T_regs in unstable plaques 83, suggesting an atheroprotective role for their anti-inflammatory immune regulator function, others challenge this view, reporting a higher number of T_regs in lesions at risk 55, which may be attributable to an altered T_reg functional state or to a compensatory increase of T_regs to counterbalance a T cell activation at the plaque level. A recent report by Klingenberg and colleagues showed an increased number of T_regs in coronary artery thrombus aspirate from 16 ACS patients compared to circulating T_regs from the same patients or healthy controls 56. Interestingly, T cells within the thrombus showed restricted TCR expression: together these data support differential, antigen-driven trapping of T_regs in the thrombus as a consequence of ACS 56.

CD8⁺ cytotoxic T cells are generally less abundant than CD4⁺ cells in human atherosclerotic plaques. None the less, they can constitute up to 50% of cells in advanced lesions, suggesting a potential role in plaque inflammation and instability 84.

Natural killer T (NK T) cells constitute a distinct subset of T cells expressing both natural killer and T cell markers. They are activated upon stimulation by lipid antigens presented through the MHC-I-like molecule CD1d, which renders them particularly interesting in the study of atherosclerosis 85. Human atheroma contains cells that display CD1 86 as well as NK T cells 57, suggesting a pro-atherogenic role for this T cell subtype.

Circulating T cell subpopulations

Considering the systemic nature of the immune response in atherosclerosis, analysis of the circulating T cell subpopulation by means of flow cytometry provides a viable option for its characterization in humans. While this approach has gained widespread popularity due to its reliability and low intrinsic invasiveness, methodological caveats need to be taken into consideration: in particular, trafficking of T cells between lesions, blood and secondary lymphoid organs may hamper the interpretation of the blood count at a certain time-point 87. Accordingly, thoracic lymph nodes of patients with CAD revealed a lymphocyte subpopulation profile differing substantially from that of blood, including a higher proportion of B cells, lower proportions of CD8⁺ T cells, a twofold higher CD4/CD8 ratio and an enrichment of CD4⁺CD69⁺ cells, as well as T_regs 88. Peripheral blood of CVD patients shows an increased proportion of effector memory T cells (TEM), defined as CD3⁺CD4⁺CD45RA⁻CD45RO⁺CCR7⁻, which correlated with the extent of atherosclerotic disease in the coronary and carotid districts 47 (see Table 1). Consistent with this result, other studies documented an increase of CCR7⁻ T cells in patients with CAD 33 and an increase of memory T cells in patients with subclinical carotid atherosclerosis 71,72. TEM emerged as the T cell subset with the strongest association with atherosclerosis in carotid and coronary vascular districts at different stages of disease. TEM correlated significantly with plasma total cholesterol and LDL cholesterol, although the association between TEM and carotid atherosclerosis was independent of the classical CVRFs, supporting the relevance of adaptive immune response in CV disorders 47,89. TEM cells, together with central memory T cells (TCM), persist in the memory pool once the antigen that elicited an immune response has been eliminated. They retain the memory of (i) antigen specificity, (ii) the array of cytokines they have produced and (iii) the site where their effector function is needed. Upon antigen re-exposure, TEM display immediate effector functions in inflamed peripheral tissues (in this case the atherosclerotic plaque), mainly by the expression of CCR5 and CXCR3 90. The expression of HLA-DR constitutes a marker of effector function acquisition, and several studies documented an increase of activated HLA-DR⁺ T cells in patients with CAD 42,47,61. Th1 cells were shown to be more abundant in the blood of patients suffering from ACS 58,66, but it is yet to be determined whether this reflects an acute response to myocardial damage or the underlying CAD. In addition, a subset of INF-γ-secreting Th17 cells, namely Th1/Th17 cells, was shown to be associated with the development of ACS 66, further supporting the role for IFN-γ in atherosclerosis. Th17 themselves have been associated with an increased risk of CVD, although inconsistently 65,66. A recent retrospective study demonstrated a negative correlation between circulating Th2 cells and common carotid intima media thickness (IMT) and the risk of cardiovascular events, along with a negative association between Th1 cell count and the development of atherosclerosis-related complications 59. During ACS, the classical immunological synapse mediated by antigen–TCR engagement (signal 1) and co-stimulatory receptors such as CD28, that mediate signal 2, are perturbed in circulating T cells 91. In fact CD3⁺CD4⁺TCR zeta-dim, i.e. a T cell subset with decreased levels of the zeta subunit of TCR, also called CD247, which couples the engaged TCR–CD3 complex to downstream intracellular signal transduction pathways, and CD4⁺CD28^null T cells were found to be higher in patients suffering from ACS 62,64. Interestingly, higher circulating levels of CD4⁺CD28^null cells were found to associate with poor prognosis upon recurrence of ACS 67. Down-regulation of the TCR zeta chain and CD28 generally occur after antigen engagement or in response to inflammatory stimuli as a feedback mechanism aimed at tuning the immune response. Together with CD28 down-regulation, TCR zeta chain reduction was observed in T cells isolated in several chronic diseases, including cancer, autoimmune, e.g. systemic lupus erythematosus (SLE), and infectious diseases 92. Given that intact TCR signalling is critical to maintain immune homeostasis through the generation and functioning of regulatory T cell subsets, alterations in signal 1 pathways could result in increased TCR zeta-dim T cells which, in turn, could dampen modulator feedback signals, thus potentially limiting CD4⁺CD28^null T cell responsiveness to inhibition. Both TCR zeta-dim T cells and CD4⁺CD28^null T cells can respond to stimuli independently of the antigen-mediated TCR pathway 93,94. Furthermore, human circulating or plaque CD4⁺CD28^null T cells from ACS patients express IL-12 receptors even in the absence of antigenic stimulation, and up-regulate the expression of the chemokine receptor CCR5 and the C-type lectin receptor CD161, both implicated in regulating tissue homing of effector T cells after IL-12 stimulation 93. This suggests that CD4⁺CD28^null T cells could functionally resemble NK cells, with proinflammatory activity even in the unprimed state and increased tissue trafficking and homing after IL-12-inducing host infection associated with accrual in inflammatory lesions. Both antigen-dependent and -independent mechanisms are therefore thought to be critical to elicit responses in CD28 and/or TCR zeta chain defective memory T cell subsets, thus promoting a proinflammatory and pro-atherosclerotic response 91.

Adaptive immunity can operate during atherogenesis as a ‘double-edged sword’, exerting both promoting and inhibitory effects on plaques 95. Regarding the anti-inflammatory or anti-atherosclerotic side of T cell function, analysis of circulating T_regs has yielded contrasting results. ACS patients were shown to have lower levels of CD4⁺CD25⁺forkhead box protein 3 (FoxP3⁺) circulating T cells 68, and T_regs isolated from the blood of ACS patients displayed a reduced ability to suppress oxLDL-induced CD⁺CD25⁻ proliferation 69. No significant association was found, however, with the extension of atherosclerotic disease in stable CAD patients 68,69. A subsequent report confirmed the lack of association between stable CAD extent and progression and the levels of circulating T_regs, defined as CD4⁺CD25^hiCD127^lo, and established an association between ST-elevated AMI and high levels of T_regs 70. A more prominent inflammatory activation in ST-elevated AMI patients, as documented by increased IL-6 levels 60, might account for a proportional compensatory T_reg counterbalance, similarly to the observed IL-10 increase 60. In contrast, patients suffering from non-ST-elevated ACS were shown to have reduced circulating levels of T_regs 70. Finally, as CCR5 not only drives effector T cells but also T_regs homing and trafficking in inflamed non-lymphoid tissue 96, suggesting that CCR5⁺ T_regs constituted a subgroup of ‘effector’ T_reg cells, levels of circulating CCR5⁺ T_regs were assessed in both patients with subclinical carotid and CAD. Analysis of 313 individuals did not show any association between CCR5⁺ T_regs and atherosclerosis 70. Despite challenging findings in humans, the complex role of T_regs in atherosclerosis is currently under investigation, supported by promising data on the atheroprotective role of this T cell subset obtained in murine models 97.

B cells and humoral response in atherosclerosis

While the role of T cells in atherosclerosis has been studied extensively over decades, the role of B cells has only recently begun to gain attention. The first cue for B cell involvement in the pathogenesis of atherosclerotic lesions has spread from animal studies 98, but currently evidence is also building in humans. A recent network-driven integrative analysis of data from genomewide association studies and whole blood gene expression profiles from Framingham Heart Study participants identified B cell immune responses as causative for CAD 99. In contrast to T cells, only few B cells can be detected locally within the atheroma 44, while a large number can be found in the adventitial layer of atherosclerotic vessels, where they display a structural organization closely resembling a tertiary lymphoid organ 100, which is consistent with the presence of chronic immune response. B cells in atherosclerotic lesions were shown to be oligoclonal and undergoing antigen-driven proliferation 101. Such an antigen-driven B cell response is Th-cell dependent, delayed, and results in the production of high-affinity antibodies undergoing class switch. The entire process occurs in specialized structures within lymphoid organs, namely germinal centres. A specific subset of Th cells, known as T follicular helper cells (Tfh), was shown to be responsible for germinal centre organization and for providing B cells with the help required for proliferation and affinity maturation 102. Tfh cells were shown to express lower amounts of cytokines, but also a higher variety of surface receptors, such as CD40L (CD154) or OX-40, when compared to other Th cell subsets 102. The role of Tfh cells in atherosclerosis has not yet been investigated. B cells responsible for this type of response originate from the bone marrow, and are known as B2 cells 103. Antibodies secreted by B2 cells comprise all human immunoglobulin (Ig) classes, namely IgM, IgG, IgE and IgA. IgG antibodies directed against OSEs, in particular aldehyde-modified peptide sequences of apolipoprotein B-100, can be detected readily in the sera of atherosclerotic patients 104. In addition, self-reacting IgG against transgelin (TAGLN), a cytoskeletal protein, were shown to be secreted by B2 cells located within carotid artery plaques 44. Interestingly, these antibodies were shown to cross-react with antigenic determinants of the bacterial wall of Gram-negative bacteria belonging to the Enterobacteriaceae family, again supporting a potential role for infection in the development of atherosclerosis 44. Further studies are needed to deepen our understanding of the role and association with the CV risk of IgG and IgM against OSEs and other antigens that can be detected in the atherosclerotic plaques 75. Apart from the production of atherogenic antibodies, experimental studies showed that B2 cells appear to aggravate atherogenesis through antibody-independent mechanisms that augment the action of proinflammatory cytokines 105.

IgA immunoglobulins can be found on mucosal surfaces, where they provide the first line of defence against pathogens, and at lower concentrations in the circulation. Although there is little information about the role of IgA in atherosclerosis, there appears to be an association between high serum IgA titres and advanced vascular disease and myocardial infarction 106. While, currently, no mechanism has been proposed to explain such association, recent data in the role of gut microbiome in CVD 107,108 may potentially provide new insights in the role of IgA in atherosclerosis.

Alongside B2 cells, humans have a minor B cell subset, called B1 cells, comprised of long-lived, non-circulating cells found preferentially in the spleen and the peritoneal or pleural cavity 103. These cells secrete poorly specific natural IgM antibodies, setting up a rapid and T cell-independent humoral response. B1 secreted antibodies are polyreactive and constitute a first line of defence against pathogens. Natural IgM antibodies make up a substantial proportion of IgM in the uninfected human, and up to 30% of them are directed specifically against OSEs 24. Several clinical studies have shown that titres of such naturally occurring OSE-specific IgM correlate inversely with atherosclerotic burden estimated by carotid artery IMT 104,109,110, as well as with the risk of stroke and AMI 111. The atheroprotective mechanism of natural IgM is yet to be elucidated, but experimental studies suggest that these antibodies prevent oxLDL internalization by macrophages and limit the accumulation of apoptotic cells by augmenting efferocytosis 112 (Fig. 2). Table 3 summarizes the main findings concerned with B cell and humoral response in atherosclerosis.

Figure 2.

Role of B cells and Immunoglobulins in atherosclerotic lesion development. Under chronic inflammatory conditions, B2 cells become activated by T follicular helper cells within lymphoid-like structures in the vessel wall. They undergo maturation into antibody secreting cells within the plaque where they directly secrete immunoglobulins. Further immunoglobulins, including B1 cell-secreted naturally occurring antibodies, are secreted in other districts and reach the plaque through the systemic circulation. The presence of B1 cells within the plaque is currently a matter of debate. Immunoglobulins show a wide range of specificities, including oxidized low-density lipoproteins (oxLDL) and heat shock protein 60 (HSP60). Naturally occurring immunoglobulin (Ig)M appears to exert a protective action by inhibiting the uptake of oxLDL by macrophages and by favouring the clearance of apoptotic bodies, which is aided by local deposition of complement. Conversely, IgG directed against oxLDL may enhance the uptake of cholesterol, thus favouring atherosclerosis. A protective effect, however, was shown for IgG specific for peptide 210 of ApoB100. IgG cross-reacting with bacterial antigens were shown to be produced within lesione, suggesting a role for infectious agents. IgG recognizing endothelial antigens, e.g. HSP60, may directly cause endothelial disfunction. SMC = smooth muscle cells.

Table 3.

Summary of main findings regarding humoral response in human atherosclerosis

B cell type	Immunoglobulin class	Specificity	Finding	Number of patients	Reference
B1	IgM	Naturally occurring, specific to various OSEs	Inverse correlation with carotid artery IMT and number of plaques	1022	Karvonen, 2003 109
			Association with lower CVD and stroke risk	765	Tsimikas, 2012 111
B2	IgM	ApoB100 peptide 210	Negative correlation with carotid artery IMT severity and progression during 30 months	3430	McLeod, 2014 113
	IgG	ApoB100 peptide 210	Negative correlation with carotid artery IMT severity and progression during 30 months	3430	McLeod, 2014 113
		Polyclonal anti-OSEs	Positive correlation with presence of angiographically determined CAD. Loss of significance upon correction for other risk factors	504	Tsimikas, 2007 114
		TAGLN (human cytoskeletal protein)	Locally produced within carotid artery plaques. Cross-react against bacterial atigens	4	Canducci, 2012 44
	IgE	–	Correlation with risk of AMI and sudden death in dyslipidaemic men	135	Kovanen, 1998 106
			Higher in CAD patients than in controls. Among CAD patient, higher in AMI than in unstable angina, and higher in unstable than in stable angina	709 CAD; 273 controls	Wang, 2011 115
	IgA	–	Correlation with risk of AMI and sudden death in dyslipidaemic men	135	Kovanen, 1998 106
			Increased in ACS patients compared to controls	145 ACS; 34 controls	Muscari 1988 116

AMI = acute myocardial infarction; ACS = acute coronary syndromes; CAD = coronary artery disease; Ig = immunoglobulin; IMT = intima medial thickness; CVD = cardiovascular disease; OSE = oxidation-specific epitope.

Platelet adhesion, thrombosis and adaptive immunity

Thrombosis is a critical event in the natural history of atherosclerosis. Rupture or erosion of advanced, vulnerable lesions exposes the highly thrombogenic subendothelial layer and initiates platelet adhesion and thrombosis, resulting in acute complications such as ACS or stroke 117. In addition, many indications suggest that platelets may contribute actively to neointimal formation and atherosclerotic lesion initiation and progression 118 119. Several lines of evidence functionally link lymphocytes and platelets in the development and clinical manifestations of atherosclerosis. In particular, lymphocyte master cytokines such as IFN-γ and IL-4 were shown to associate significantly with residual platelet reactivity in ACS patients on dual anti-platelet therapy 120, pointing to a role for T cell effector function in the development of thrombosis. Experimental studies have shown that IFN-γ can enhance platelet-dense granule secretion and conjugation with lymphocytes 121, and IL-2 was shown to reduce platelet adhesion and increase α-granule secretion 122. Furthermore, lymphocyte ecto-ATPase may convert ATP released extracellularly by platelets themselves or other cell types into ADP, which subsequently enhances platelet aggregation 123. Notably, however, the interaction of platelets and lymphocytes appears to be bidirectional, and platelets appear to modulate various aspect of lymphocyte function and to influence their engagement in atherosclerosis. Platelets can adhere to lymphocytes to form platelet–lymphocyte conjugates 124, which may facilitate the adhesion of lymphocytes under sheer stress conditions 125 and recruit them in sites of arterial thrombi 126. Platelets were shown to enhance T cell cytokine production, mainly through a cell-to-cell interaction mediated by CD40–CD40L (CD154) ligation 127, and to arrest CD4⁺ T cell clonal expansion by favouring their differentiation into effector cells by multiple cell-to-cell interactions and soluble mediators 128. Platelets were also shown to directly stimulate B cell proliferation and supprt isotype switching 127.

Atherosclerosis and disorders of the immune system

The profound involvement of adaptive immunity in atherosclerotic disease is confirmed further by the increased risk of CVD in patients affected by disorders of the immune system, including autoimmune diseases and HIV. Autoimmune diseases are a group of disorders characterized by loss of immunological tolerance with subsequent humoral and cell-mediated immune responses against self-constituents. Such disorders, which include SLE and rheumatoid arthritis (RA) among others, are associated with an accelerated progression of atherosclerosis and excess CV morbidity and mortality not accounted for fully by traditional CVRF 129,130. The specific underlying mechanism of increased risk is yet to be defined, but appears to involve both aspecific inflammation and immune activation. In SLE, titres of autoantibodies, correlate with atherosclerosis development and progression and the risk of ACS 129. A thorough analysis of circulating T cells showed a greater proportion of activated cells, i.e. HLA-DR⁺ and CCR5⁺ cells, in SLE patients, which paralleled a faster progression of atherosclerosis, but no significant association was established between T cell profile and CVD 131. Similarly, RA is associated with higher risk for CAD and with higher ACS mortality and recurrence rates 130. Interestingly, RA patients suffering from atherosclerosis were shown to have a disproportionate expansion of CD28^nullCD4⁺ T cells, alongside Th17/T_reg imbalance 132. In addition, patients positive for anti-cyclic citrullinated peptide IgG autoantibodies were shown to be at increased risk for atherosclerosis progression with respect to RA patients negative for the same autoantibody and healthy individuals 133. Citrullinated peptides, which form in the face of active inflammation, were shown to be abundant in atherosclerotic lesions 133.

Chronic HIV infection, which is well known to depress immune function of the host, is associated with a high risk of vascular atherosclerotic disease, progressing rapidly from endothelial dysfunction to subclinical atherosclerosis and, in some cases, to advanced disease and ACS 134. The drivers of the spectrum of vascular disease in HIV patients are complex and incompletely understood, and appear to involve both medication-related elements 135 and chronic inflammation with lymphocyte activation secondary to the loss of CD4⁺ T cells induced by the virus 136. In particular, generalized CD4⁺ T cell activation as well as CD8⁺ T cell activation and senescence were shown to correlate positively with subclinical atherosclerosis in HIV-infected individuals treated with an effective anti-retroviral therapy 137.

Immune system modulation for the treatment and prevention of atherosclerosis

Apart from new insights into the pathogensis of CVD, study of the immunological mechanisms of atherosclerosis has provided clues for the development of novel treatments aimed at modulating immunity to improve patient outcomes. The identification of autoantigens involved in atherogenesis led to the proposal that immunization against them may lead to the alleviation of atherosclerotic disease 138. Vaccination against a number of oxLDL-associated OSEs was shown to reduce lesions in animal models, although whether a similar approach would be effective in humans is yet to be proved 138. Experimental studies in mice identified antigen-specific T_regs as a critical component of immunization-induced atheroprotection 139. OxLDL is a complex particle, and may contain many relevant epitopes. Screening for immunogenic LDL-derived epitopes has identified a set of peptides which were used for immunization in mice 140. The few displaying an atheroprotective activity are currently undergoing extensive study to develop experimental vaccines.

Experimental studies provide evidence for the reduction of atherosclerotic vascular disease by direct inhibition of pathogenic cytokines or their receptors 75. Consistent with this strategy, the ongoing Canakinumab Anti-inflammatory Thrombosis Outcome Study is aimed at evaluating the impact of IL-1β blockage on AMI recurrence and atherosclerosis progression in AMI patients 141. Preliminary data have shown that Canakinumab, a human monoclonal antibody that neutralizes interleukin-1β, significantly reduces inflammation, measured as C-reactive protein and IL-6 levels, without major effects on LDL or HDL cholesterol 142. In addition, prospective observational data from patients affected by autoimmune diseases treated with anti-cytokine medications will probably provide insights into the impact of such interventions on CVD. Other potential stategies that have currently proved effective in animal studies include modulation of co-stimulatory and co-inhibitory signals and T cell deple tion, although the latter is unlikely to ever be translated into clinical practice due to excessive immunosuppression. The advent and clinical application of B cell-targeted compounds in autoimmune disease and onco-haematology offers a great opportunity for novel therapeutic strategies in atherosclerosis. Among these, rituximab is a B cell-depleting antibody currently approved for use in lymphomas and RA that acts through the cross-linking of B cell-specific surface antigen CD20 143. Belimumab is a monoclonal antibody that neutralizes the cytokine B cell activating factor (BAFF), which is essential for B cell survival and maturation 144. The effect of both of these therapies on CVD, in particular in patients with autoimmune disorders, is currently unknown, but experimental evidence demonstrated an atheroprotective effect for both 145,146.

Conclusions

The evidence summarized above supports a role for the operation of adaptive immune response in atherosclerotic lesion development and clinical manifestations. Apart from providing new fascinating insights in the pathogenesis of CVD, recognition of the immunological mechanisms in atherogenesis has provided the rationale for the development of novel therapeutical strategies for the treatment of atherosclerosis and its clinical manifestations. However, in the clinical setting this approach presents major challenges, as global interference with host defence mechanisms can yield an immunosuppressed individual, at high risk for cancer and opportunistic infections. Furthermore, to date no trial with any anti-inflammatory agent was able to reduce CV mortality. The recent negative results of darapladib 147, an oral selective inhibitor of the lipoprotein-associated phospholipase A2 (Lp-PLA2), an enzyme able to generate OSEs and inflammatory mediators directly from LDLs, challenges some hypotheses derived from animal models, where LDLs modified by Lp-PLA2 acquire potent pro-atherogenic activities 148. Finally, it should not be forgotten that hypercholesterolaemia per se remains the single most important driver for the primary inflammatory cascade and subsequent immune response, and that clinical studies have demonstrated that lowering plasma cholesterol effectively reduces inflammation 149. In addition, the measurement of C-reactive protein with a 2 mg/l cut-off would not have predicted 41% of unequivocally documented ST-elevation AMIs, thus indicating both its limitations as an individual prognostic marker and as an indicator of a generalized inflammatory pathogenetic component in AMI 150. Further research is therefore needed to unwind the complexity of atherosclerosis-associated immune response and the inter-relation between immunity and dyslipidaemia, in order to limit atherosclerotic disease most effectively.

Disclosure

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Author contributions

E. A. and F. M. wrote the report. M. M. and P. G. C. crically revised the manuscript.