Fcγ Receptors and Ligands and Cardiovascular Disease

Abstract

Fcγ receptors (FcγR) classically modulate intracellular signaling upon binding of the Fc region of IgG in immune response cells. How FcγR and their ligands impact cardiovascular health and disease has recently been interrogated in both preclinical and clinical studies. The stimulation of activating FcγR in endothelial cells, vascular smooth muscle cells and monocytes/macrophages causes a variety of cellular responses that may contribute to vascular disease pathogenesis. Stimulation of the lone inhibitory FγcR, FcγRIIB, also has adverse consequences in endothelial cells, antagonizing NO production and reparative mechanisms. In preclinical disease models, activating FcγR promote atherosclerosis whereas FcγRIIB is protective, and activating FcγR also enhance thrombotic and non-thrombotic vascular occlusion. The FcγR ligand C-reactive protein (CRP) has undergone intense study. Although in rodents CRP does not impact atherosclerosis, it causes hypertension and insulin resistance and worsens myocardial infarction. Massive data has accumulated indicating an association between increases in circulating CRP and coronary heart disease in humans. However, Mendelian randomization studies reveal that CRP is not likely a disease mediator. CRP genetics and hypertension warrants further investigation. Studies to date of genetic variants of activating FcγR are insufficient to implicate the receptors in coronary heart disease pathogenesis in humans. However, a link between FcγRIIB and human hypertension may be emerging. Further knowledge of the vascular biology of FcγR and their ligands will potentially enhance our understanding of cardiovascular disorders, particularly in patients whose greater predisposition for disease is not explained by traditional risk factors, such as individuals with autoimmune disorders.

Introduction

Fcγ receptors (FcγR) are plasma membrane-associated receptors for IgG and the pentraxins C-reactive protein (CRP) and serum amyloid P component (SAP). As a result of recent interest in the role for inflammation in cardiovascular disorders, FcγR and their ligands have been investigated in the context of cardiovascular disease. This review will highlight major observations made in both preclinical and clinical studies in that realm. Following a brief description of FcγR subtypes and their functions, their IgG and pentraxin ligands will be discussed. Investigations performed in cell culture will then be highlighted, followed by queries of the impact of FcγR and ligands in animal models of cardiovascular disorders. Attempts to understand the influence of IgG and pentraxins on cardiovascular disease risk and outcome in humans will then be summarized, including the voluminous studies of CRP as a risk factor and possible disease mediator. How genetic variants in FcγR impact human cardiovascular health will be presented, including a recent interrogation of a variant in the lone inhibitory FcγR, FcγRIIB, that alters receptor function in endothelium. Finally, the current knowledge gaps in this area and the rationale to fill them will be presented. By recognizing what is known and not yet known about the influence of FcγR on cardiovascular health, it is hoped that this review will stimulate further work in this area. Such future efforts are warranted because FcγR biology may underlie a considerable component of vascular disease predisposition that is not currently explained by traditional risk factors.

Fcγ Receptor Subtypes and Functions

Fcγ Receptor Subtypes

The classical function of FcγR is to invoke intracellular signaling upon IgG binding in immune response cells and thereby modulate numerous inflammatory processes. FcγR are categorized into activating receptors and inhibitory receptors (Table 1)¹. In humans, the activating Fc receptors are FcγRI (CD64), FcγRIIA (CD32a), FcγRIIC (CD32c), FcγRIIIA (CD16a) and FcγRIIIB (CD16b), and the sole inhibitory receptor is FcγRIIB (CD32b). In mice the activating FcγR are FcγRI, FcγRIII and FcγRIV, and in mice differential splicing of a single gene for inhibitory FcγRIIB yields FcγRIIB1, B2 and B3. FcγR belong to the large immunoglobulin superfamily and they are type I transmembrane glycoproteins with the exception of the human subtype FcγRIIIB, which is GPI-anchored. All Fcγ receptors display a high degree of sequence identity in their extracellular portion (50–96%), but differ significantly in their cytoplasmic domains^1–4.

Table 1.

Activating and inhibitory Fcγ receptors in humans and mice.

	Human		Mouse
Activating	FcγRI (ITAM, γ-chain)	CD64	FcγRI (ITAM, γ-chain)	CD64
	FcγRIIA (ITAM)	CD32a
	FcγRIIC (ITAM)	CD 32c
	FcγRIIIA (ITAM, γ-chain)	CD16a	FcγRIII (ITAM, γ-chain)	CD16
	FcγRIIIB (GPI-anchored)	CD16b
			FcγRIV (ITAM, γ-chain)	CD16-2
Inhibitory	FcγRIIB (ITIM)	CD32b	FcγRIIB (ITIM)	CD32b

Innate immune effector cells, such as monocytes, macrophages, dendritic cells (DCs), basophils and mast cells, express activating and inhibitory FcγR. In human leukocytes, FcγRI and FcγRIIA are abundant in monocytes, macrophages and granulocytes, FcγRIIIA are expressed primarily on macrophages but also in monocytes, and FcγRIIIB are found in granulocytes^{5, 6}. In mice, monocytes and macrophages express all activating and inhibitory FcγR, neutrophils mainly express the activating FcγRIII and FcγRIV and the inhibitory FcγRIIB, whereas the expression of FcγRI, FcγRIIB and FcγRIII dominates on DCs. There are two cell types that do not co-express activating and inhibitory receptors; NK cells solely express the activating receptor FcγRIII, and B cells only express the inhibitory receptor FcγRIIB. In B cells, FcγRIIB functions as a regulator of activating signals transmitted by the B cell receptor (BCR)³. In DC, the balance between activating and inhibitory FcγR activity influences their relative activation status⁷. In addition to hematopoietic cells, FcγR are expressed by follicular DCs, endothelial cells, microglial cells, osteoclasts and mesangial cells³. Interestingly, ligand binding impacts the abundance of cell surface FcγR. Immune complex binding to FcγRIIA leads to their endocytosis and ubiquitination and degradation, and engaged FcγRIIIA also undergo internalization and degradation⁸; in contrast, in human endothelial cells C-reactive protein (CRP) upregulates the surface expression of CD32 and CD64⁹.

Activating Fcγ Receptor Function

The activating FcγR, with the exception of the human GPI-anchored FcγRIIIB, activate signaling pathways through immunoreceptor tyrosine-based activation motifs (ITAMs) contained in their cytoplasmic regions. Human FcγRIIA and FcγRIIC are comprised of a single subunit with a cytoplasmic ITAM (Table 1), and they transduce activating signaling pathways autonomously. All other activating FcγR (human FcγRI, FcγRIIIA and mouse FcγRI, FcγRIII, FcγRIV) consist of a ligand-binding α-chain and a signal-transducing γ chain that contains an ITAM in its cytoplasmic domain. In some instances the signal transducing subunits for an FcγR can differ between cell types. Whereas human FcγRIIIA is associated with an ITAM-containing γ-chain in monocytes and macrophages, it associates with the ITAM-containing CD2 ζ chain in NK cells. Along with its signaling function, the γ-chain is important for the assembly and cell-surface transport of the respective α chains. Because of their expression in innate immune effector cells, activating FcγR have been proposed to link the specificity of antibodies generated by the adaptive immune system to the potent effector functions of the innate immune system^1–3.

After crosslinking by immune complexes, the signaling pathways initiated by the different activating FcγR are quite similar, beginning with tyrosine phosphorylation of the ITAM by kinases of the SRC family. This leads to the recruitment of SYK-family kinases to the ITAM, followed by the activation of various downstream targets, such as the linker of activation of T cells, multimolecular adaptor complexes and PI3 kinase. By generating PIP3, PI3 kinase creates membrane-docking sites for Bruton's tyrosine kinase (BTK) and phospholipase Cγ (PLCγ). Activation of PLCγ causes an increase in intracellular calcium and triggering of further downstream signaling events. In addition to calcium-dependent pathways, the RAS-RAF-MAPK-pathway is of central importance for cell activation following activating FcγR crosslinking. An additional major function of activating FcγR is to promote the endocytosis or phagocytosis of immune complexes, which include antibody-coated microorganisms and soluble proteins. In the case of granulocytes, monocytes and macrophages, this will mainly result in the rapid degradation of the engulfed material in lysosomal compartments^1–3.

Inhibitory Fγc Receptor Function

The only known inhibitory FcγR, FcγRIIB, is the most broadly expressed FcγR, and it is present on virtually all leukocytes with the exception of NK cells and T cells. Because of the broad expression pattern, global genetic deletion of FcγRIIB can result in complex phenotypic changes affecting either innate or adaptive immune responses. FcγRIIB transmits inhibitory signals through an immunoreceptor tyrosine-based inhibitory motif (ITIM) contained in its cytoplasmic region. FcγRIIB function is best understood in B cells, in which it serves as a checkpoint for humoral immunity, playing a critical role in preventing the generation of autoreactive antibodies. Partnering with BCR, the simultaneous triggering of the ITIM-containing FcγRIIB and the BCR results in the recruitment of phosphatases including SHIP (SH2 domain-containing inositol polyphosphate 5' phosphatase) and SHP1 (SH2-domain -containing protein tyrosine phosphatase 1) that interfere with activating signaling pathways by hydrolyzing phosphoinositide intermediates. This prevents the recruitment of pleckstrin homology-domain-containing kinases, such as BTK or PLCγ, to the cell membrane, thereby diminishing downstream events such as increases in intracellular calcium levels. Processes that disrupt FcγRIIB function in B cells result in a lower threshold for B cell activation and stronger activating signals after BCR crosslinking^1–3.

In addition to its participation in adaptive immunity by tempering antibody production by B cells, FcγRIIB is a regulator of innate immunity via its actions in mast cells, granulocytes and macrophages. As these cells have the capacity to trigger strong proinflammatory responses, their activation needs to be tightly controlled. In the case of antibody-mediated responses, such as phagocytosis, antibody-dependent cell-mediated cytotoxicity, allergic reactions and the release of pro-inflammatory mediators, this is the function of the inhibitory FcγRIIB. FcγRIIB contributes varying levels of negative regulation depending on the specific IgG subclass that is bound to the receptor (see below)^1–3.

Fcγ Receptor Ligands

Immunoglobulins

The complexity of the FcγR family is mirrored by the presence of four different IgG subclasses in humans (IgG1-IgG4) and in mice (IgG1, IgG2a/c, IgG2b and IgG3), which bind with varying affinity and specificity to different FcγR via the Fc portion of the IgG. Overall, in humans IgG1 and IgG3 are the most pro-inflammatory IgG subclasses, and in mice IgG2a and IgG2b are the most pro-inflammatory IgG molecules. FcγR vary in their affinity for IgG. FcγRI is the single high-affinity receptor in humans and in mice, particularly regarding binding of IgG1 and IgG3 in humans or IgG2a in mice. All other FcγR have a 100–1000 fold lower affinity and show a broader IgG subclass specificity. Relative binding affinity of mouse and human IgG subclasses to mouse and human FcγR is shown in Table 2¹⁰. The low affinity nature of IgG binding to most of the FcγR proteins serves an important function in that it prevents the binding by monomeric antibody molecules that are always present at high levels in the circulation, thereby avoiding the potential non-specific activation of pro-inflammatory responses. In contrast, the high-affinity FcγRI is constantly saturated with ligand. However, as has been described for the binding of IgE to the high-affinity FcεRI, cell activation only ensues after the FcγRI receptors have been crosslinked by antigen. Considering the existence of both activating and inhibiting FcγR, and the varying affinities for IgG subclasses, the summary in vivo actions of IgG-FcγR tandems can be difficult to predict. To deal with this complexity, the ratio of affinities of a given IgG subclass for the activating versus the inhibitory receptors has been termed the A/I-ratio and it has emerged as a helpful predictive value for the activity of a specific IgG subclass in vivo^1–3.

Table 2.

Binding affinity between IgG subclasses and Fcγ receptors¹⁰.

	Human					Mouse
	IgG	1	2	3	4	IgG	1	2a	2b	3
Activating	FcγRI	high	none	high	high	FcγRI	none	high	low	very low
	FcγRIIA	low	low	low	low
	FcγRIIC	low	very low	low	low
	FcγRIIIA	low	very low	low	low	FcγRIII	low	low	low	none
	FcγRIIIB	low	none	low	low
						FcγRIV	none	high	high	none
Inhibitory	FcγRIIB	low	very low	low	low	FcγRIIB	low	low	low	none

Pentraxins

In addition to IgG, the acute phase reactant C-reactive protein (CRP) is a ligand for FcγR. CRP is produced principally by the liver in response to a variety of pathologic conditions including inflammation, infection and trauma. CRP is a member of the pentraxin family of proteins, and it was originally described as a protein which binds to the C-polysaccharide of the cell wall of pneumococci. The protein consists of five identical 23 kDa subunits noncovalently associated in a flat pentameric disk. The primary stimulus for hepatocyte synthesis and secretion of CRP is the proinflammatory cytokine interleukin (IL)-6. Circulating CRP levels can rise 500-fold within 24 to 48h of the initiation of an inflammatory process. CRP was initially found to serve as an opsonin, binding to pathogenetic microorganisms and mediating the phagocytosis of sensitized erythrocytes. CRP also activates complement by binding to C1q, and studies in mice indicate that it is protective against infection and the development of autoimmunity. CRP levels are often used clinically as an indicator of infection and/or inflammation^11–14.

In addition to the liver being a source of CRP, transcript for the pentraxin has been detected in human atherosclerotic lesions, with mRNA levels 10-fold greater in plaques than in normal arteries¹⁵. CRP transcript has also been detected in areas of myointimal hyperplasia in a porcine arteriovenous graft model¹⁶. Expression of CRP has been demonstrated in endothelial cells, particularly upon treatment with macrophage conditioned medium¹⁷ or electronegative low density lipoprotein cholesterol¹⁸. In vascular smooth muscle, angiotensin II, endothelin-1 and homocysteine promote CRP expression^19–21. Adipose tissue may be an additional important source of CRP since the mRNA is detectable in human adipose tissue, where it is expressed in both adipocytes and stromal cells^22,23. CRP mRNA is increased in the adipose tissue of obese individuals compared with controls and it is upregulated in vitro in adipose tissue explants by LPS or IL-6²⁴.

Regarding the forms of CRP that impact the cardiovascular system, it is controversial whether the potential actions of the pentraxin of relevance to cardiovascular health are mediated by native, pentameric CRP, or monomeric (m)CRP. In some studies it has been reported that mCRP is detected in normal human blood vessels and in inflamed tissues^25–27, yet others indicate that mCRP is deposited in human atherosclerotic plaques but not in healthy vessels, and that pentameric CRP is not detectable in either healthy or diseased blood vessels²⁸. It has been reported that mCRP promotes a proinflammatory phenotype in cultured endothelial cells. However, in earlier investigations mCRP actions on endothelial cells were modified by anti-CD16 (FcγRIII) blocking antibody and not by anti-FcγRII antibody²⁹, yet genetic studies in mice indicate that FcγRIIB is critically involved in the actions of endogenous CRP on endothelium³⁰. In addition, there is a recent report that mCRP actually mediates responses in human endothelial cells via plasma membrane insertion rather than by binding to surface FcγRs³¹. Furthermore, it has been demonstrated that native CRP impairs endothelial function in vivo, whereas mCRP does not³². Activated platelets disassociate pentameric CRP to mCRP via lysophosphatidylcholine which is only present on activated platelets²⁸, suggesting that it may remain challenging to definitively determine which form of CRP impacts the cardiovascular system in vivo.

It was initially determined that CRP binds to the high affinity receptor for IgG, FcγRI, in human monocytes and transfected COS-7 cells^33,34. FcγRIIA, which is a low-affinity receptor for IgG, is also a high-affinity CRP receptor^35,36. FcγR were additionally identified as CRP receptors in mouse leukocytes using knockout lines lacking FcγR; blood cells from γ chain^−/− mice displayed less CRP binding than controls, FcγRIIB^−/− mice also had diminished CRP binding, and no CRP binding was observed in leukocytes from double-deficient mice³⁷. A role for FcγRIII in CRP binding is less likely since NK cells from both humans and mice, which express FcγRIII as their sole FcγR, display negligible CRP binding³⁶. Thus, in both humans and mice FcγRI and FcγRII most likely serve as the major receptors for CRP. The structurally similar pentraxin, serum amyloid P-component (SAP), which is the CRP-equivalent acute phase reactant in mice, also binds to FcγR including FcγRII³⁸. Structural studies of pentraxin-FcγR binding have demonstrated that the molecular basis for FcγR recognition is conserved between CRP, SAP and IgG³⁹.

Fcγ Receptor-Ligand Actions in Vascular Cells

Endothelial Cells

Based upon observed associations between chronic modest elevations in CRP and cardiovascular disease in humans (see below), a variety of studies have been performed in cultured endothelial cells evaluating the effects of CRP. One of the first demonstrated that CRP decreases endothelial NO synthase (eNOS) mRNA and protein abundance and enzymatic activity in human aortic endothelial cells, and that there is an associated promotion of monocyte adhesion to the endothelial cells (Figure 1A)⁴⁰. CRP upregulation of endothelium-derived plasminogen activator inhibitor-1 (PAI-1) was then demonstrated⁴¹, and this was followed by studies showing that CRP upregulates interleukin-8 (IL-8) and endothelial cell-monocyte adhesion by activating NF-kB, with these responses as well as CRP binding to endothelial cells being attenuated by anti-CD32 (FcγRII) and anti-CD64 (FγcRI) antibodies^9,42,43. Nuclear actions of CRP in endothelial cells were further delineated in studies of cells transfected with human eNOS 5’ flanking sequence fused to luciferase that indicated that CRP decreases eNOS gene transcription⁴⁴.

Figure 1.

Effects of Fcγ receptors and their ligands on vascular cells (A) and on preclinical models of cardiometabolic disorders (B). ((Ilustration credit: Ben Smith).

Acute effects of CRP on endothelial cell function were also demonstrated in studies that showed that the rapid activation of eNOS by diverse agonists is blocked by the pentraxin. SAP and aggregated IgG used to mimic immune complexes also inhibited eNOS activation. FcγRIIB mRNA expression was demonstrated in endothelial cells, and heterologous expression studies revealed that CRP antagonism of eNOS requires FcγRIIB. Demonstrating in vivo actions of CRP on the endothelium for the first time, it was found that acute CRP administration blunts acetylcholine-induced increases in carotid artery conductance in wild-type mice; in contrast, in FcγRIIB^−/− mice CRP actually enhanced the vasodilatory response to acetylcholine⁴⁵. Long-term actions of CRP impacting the endothelium have also been shown in vivo, with CRP transgenic mice (TG-CRP) displaying attenuated carotid artery reendothelialization following perivascular electric injury. In parallel, CRP blunts cultured endothelial cell migration⁴⁴. CRP administration to rats for 72 hours resulted in decreased eNOS activity and blunted vasodilatory responses to acetylcholine in isolated, pressurized mesenteric arterioles⁴⁶. Endothelial actions of CRP have additionally been interrogated in mice harboring an inflammation-sensitive CRP transgene. Turpentine activation of the CRP transgene yielded serum levels of human CRP of 276 µg/ml, with CRP concentrations being <1.0 ug/ml in healthy humans and rarely exceeding 2 ug/ml in mice even in response to inflammatory stimuli^14,47. Aortic rings isolated from the mice with elevated CRP displayed impaired endothelium-dependent responses to acetylcholine, and NO release and eNOS-Ser1179 phosphorylation were decreased. In addition, CRP-overexpressing mice had increased perivascular fibrosis, greater endothelial VCAM-1 and MCP-1 staining, and enhanced macrophage infiltration⁴⁸.

The features of FcγRIIB and related signaling events required for CRP to antagonize eNOS were revealed in studies of eNOS activation by insulin⁴⁹, which stimulates the enzyme to promote vasodilation and capillary recruitment in the skeletal muscle microvasculature, thereby enhancing skeletal muscle glucose disposal⁵⁰. In mice the activating phosphorylation of eNOS-Ser1179 and also Akt phosphorylation in response to insulin were decreased by CRP treatment, and reconstitution experiments with wild-type and mutant FcγRIIB in NIH3T3^IR cells revealed that these processes require the ITIM of the receptor. It was further shown that endothelial cells express SHIP-1, that CRP induces SHIP-1 stimulatory phosphorylation in endothelium in culture and in vivo, and that SHIP-1 knockdown by small interfering RNA prevents CRP antagonism of insulin-induced eNOS activation⁴⁹. How CRP causes the critical FcγRIIB ITIM phosphorylation in endothelial cells has additionally been elucidated. It was determined in cultured cells that FγcRI blocking antibodies prevent CRP antagonism of eNOS (Figure 1A), that CRP activates SRC via FcγRI, and that both ITIM phosphorylation and eNOS antagonism caused by CRP are SRC-dependent. Parallel processes were found to mediate the inhibition of eNOS by aggregated IgG used to mimic immune complex. As such, FcγRI-FcγRIIB coupling by SRC kinase underlies both CRP and IgG attenuation of endothelial NO production. The partnership between FcγRI and FcγRIIB in the actions of CRP on endothelium was demonstrated in vivo in studies of carotid artery reendothelialization in TG-CRP mice crossed with either mice lacking the γ subunit of FcγRI (FcRγ^−/−) or FcγRIIB^−/− mice. Whereas reendothelialization was impaired in TG-CRP vs wild-type mice, it was normal in FcRγ^−/−;TG-CRP mice and in FcγRIIB^−/−;TG-CRP mice³⁰. Thus, both FcγRI and FcγRIIB have been implicated in the actions of CRP on endothelium in vivo.

Whereas various forms of FcγRIIB loss-of-function have been employed to implicate the receptor in the actions of immune complex and pentraxin on endothelial cells, and FcγRIIB mRNA has been demonstrated in endothelium⁴⁵, evidence that FcγRIIB potentially impacts cardiovascular health in humans would be strengthened if FcγRIIB protein expression could be demonstrated in human endothelial cells. This has been perplexing due to the lack of specificity of antibodies and the challenges involved in reliable detection of an antibody receptor using antibodies as probes. Fluorescence-activated cell sorting (FACS) was therefore performed with human endothelial cells using an Alexa Fluor 488-conjugated monoclonal Ab (clone 8B5, MacroGenics) to human FcγRIIB generated by immunization of human FcγRIIA transgenic mice to prevent cross-reactivity with FcγRIIA. In 8B5, Asn297 in the Fc domain has been mutated to Gln to prevent Fc-FcγR binding⁵¹, thereby minimizing nonspecific binding. Using 8B5 Ab, FcγRIIB protein was readily detected in Raji positive control cells (Fig. 2A), and in human aortic, coronary artery and umbilical vein endothelial cells (Fig. 2B–D, HAEC, HCAEC, HUVEC). Further confirmation of Ab specificity was obtained by siRNA knockdown of FcγRIIB (Fig. 2E, green and red for scrambled and FcγRIIB-targeted siRNA, respectively). Immunoblotting detected FcγRIIB as a 40 kDa protein in plasma membranes from Raji cells (positive control) and HAEC (Fig. 2F). These findings indicate that FcγRIIB protein is expressed in human endothelial cells derived from a variety of vascular beds.

Figure 2.

FcγRIIB protein is expressed in human endothelial cells. A–D. FACS was performed in Raji cells (positive control) (A) and in human aortic endothelial cells (HAEC) (B), human coronary artery endothelial cells (C), and human umbilical vein endothelial cells (HUVEC) (D) using 8B5, which is an Alexa Fluor 488-conjugated monoclonal antibody to human FcγRIIB. Findings in the absence of antibody (Ab) and with 8B5 are shown. E. FACS with 8B5 was performed in HAEC transfected with control siRNA (D-001810–02–05, Dharmacon) or siRNA targeting FcγRIIB (GCUACAGGUUCAAGGCCAA, Dharmacon). F. Immunoblotting was performed with 8B5 on plasma membranes isolated from Raji positive control cells and HAEC.

Vascular Smooth Muscle Cells

As has been the case for studies of FcγR ligand actions on endothelial cells, CRP has been the principal ligand employed in investigations in vascular smooth muscle (VSM) cells. Using human coronary artery VSM cells it was demonstrated that CRP induces caspase-mediated apoptosis (Fig. 1A). Upregulation of the growth arrest- and DNA damage-inducible gene 153 (GADD153) was also observed, the impact of CRP was primarily on GADD153 mRNA stability, and GADD153 silencing decreased the pro-apoptotic effect of CRP. Potentially supportive of a role for GADD153 in processes occurring in atherosclerotic plaques, GADD153 was specifically localized to apoptotic VSM cells in human coronary artery lesions⁵². It has also been shown that CRP induces intracellular reactive oxygen species (ROS) generation in VSM cells by NADPH oxidase 4 isoform (Nox4), and AP-1/NF-kB activation, and MCP-1, IL-6 and ET-1 production^53,54. CRP additionally upregulates VSM cell matrix metalloproteinase-2 (MMP-2) synthesis and activity⁵⁵, and VSM cell tissue factor expression⁵⁶. Furthermore, CRP upregulates VSM cell angiotensin II type 1 receptor (AT1-R) expression, and it promotes VSM cell migration and proliferation and enhances the effects of angiotensin II on those processes^16,57.

Regarding the potential FcγR operative in the actions of CRP on VSM cells, FcγRIIA expression has been demonstrated in cultured VSM cells isolated form human coronary arteries, and the receptor has been colocalized with α-actin positive VSM cells in atheromatous regions in human coronary artery plaques⁵². FcγRIIA blocking antibody attenuates the effects of CRP on VSM cell ROS and inflammatory cytokine production, and apoptosis (Fig. 1A)⁵², FcγRIIA and FcγRIIIA transcripts have been detected in VSM cells, and the silencing of FcγRIIIA attenuates the upregulation of tissue factor expression by CRP⁵⁶. Thus, in studies limited to cell culture, CRP has direct adverse effects on VSM typical of responses observed with injury or inflammation, and the operative FcγR are FcγRIIA and FcγRIIIA.

Leukocytes and Platelets

Evidence for potential actions of FcγR ligands and receptors in leukocytes has primarily been obtained in studies of LDL cholesterol-containing immune complexes (LDL-IC). Human monocytes/macrophages exposed to LDL-IC display marked cholesterol ester accumulation such that they assume a foam cell phenotype (Fig. 1A), and the LDL-IC cause increased LDL receptor expression. The uptake of LDL-IC is mediated by FcγRI, and it results in the release of cytokines, particularly TNF-alpha, and matrix metalloproteinase-1^58,59. In addition, the LDL-IC-FcγRI tandem promotes the survival and proliferation of monocytes/macrophages^60,61. Since LDL-IC are abundant in atherosclerotic lesions^59,62–64, these processes mediated by FcγR in monocytes/macrophages may play important roles in atherosclerosis and its consequences (see below).

In human platelets the ITAM-containing FcγR FcγRIIA is expressed, and it participates in platelet activation in a number of immune-mediated thrombocytopenias and thrombosis syndromes. These include disseminated intravascular coagulation and bacterial sepsis-associated thrombocytopenia, and heparin-induced thrombocytopenia and thrombosis^65–70. FcγRIIA also enables platelets to participate in innate immunity, allowing them to bind to and be activated by antibody-opsonized pathogens, thereby promoting their clearance^66,67,71,72. Since there is not a murine homologue for human FcγRIIA, the receptor is studied in vivo in transgenic mice expressing the human receptor^65,73. Platelets express two additional ITAM receptors. These are C-type lectin 2, which is the receptor for the snake venom toxin rhodocytin and for podoplanin, and FcγR γ chain, which associates with the glycoprotein IV collagen receptor. Platelet ITAM receptors, their signaling mechanisms and functions have recently been extensively reviewed⁶⁵, and therefore they will be mentioned in limited fashion in the remainder of this review.

Fcγ Receptors, Ligands and CVD in Animal Models

Atherosclerosis

Atherosclerosis is considered to be a chronic inflammatory disease, and both innate and adaptive immunity play critical roles in its initiation and progression⁷⁴. FcγR are expressed in virtually all individual cell types participating in atherogenesis (see above), and as such it is logical for FcγR to potentially influence atherosclerosis severity. This has been tested in vivo primarily in four studies in mice. The first entailed experiments in apoE^−/− versus apoE^−/−;FcγR γ chain^−/− mice, with the latter having genetic loss-of-function of activating FcγR. Without affecting plasma lipids, FcγR γ chain deletion resulted in decreased atherosclerotic lesion size (Fig. 1B), and lesion macrophage and T-cell content were cut approximately in half⁷⁵. In a separate study in mice fed a high fat diet for 10 weeks, the global deletion of the FcγR γ chain caused improved acetylcholine-induced relaxation of isolated aortic rings, and it lowered superoxide abundance in the vascular wall⁷⁶. These findings suggest that FcγR γ chain-mediated processes may promote atherosclerosis by attenuating endothelial function or by increasing oxygen-derived free radical abundance. Alternatively, since the quantification of FcγR in the aorta in these studies revealed that the activating/inhibitory ratio changed from 2.5 to 0.4 with the deletion of the FcγR γ chain⁷⁵, it is possible that the decrease in atherosclerosis observed in the absence of the FcγR γ chain reflects an increase in the relative atheroprotective properties of the inhibitory FcγR FcγRIIB. In bone marrow reconstitution experiments, hematopoietic deficiency of FcγR γ chain in apoE^−/− mice resulted in decreased atherosclerotic lesion size, with fewer macrophages and T lymphocytes within lesions and also increased plaque stability. The expression of pro-inflammatory genes was attenuated and anti-inflammatory gene expression was increased. In vitro experiments using murine macrophages revealed decreases in foam cell formation, pro-inflammatory gene expression and oxidative stress in cells lacking FcγR γ chain, suggesting that the cell type in which the γ chain influences atherosclerosis severity is the macrophage⁷⁷ . Other work in a similar model suggested that with hypercholesterolemia, activating FcγR promote atherosclerosis by increasing antigen-presenting cell IL-6 secretion, resulting in an enhanced Th17 response ⁷⁸. The impact of loss-of-function of an activating FcγR was also evaluated in studies of LDLR^−/− versus LDLR^−/−;FcγRIII^−/− mice. At later stages of atherosclerosis development, FcγRIII silencing resulted in smaller lesions, and there was an associated increase in interferon-γ and IL-10 production by an expansion of CD4+ T cells⁷⁹. Potential modulation of atherosclerosis severity by FcγRIIB has also been evaluated, with either bone marrow reconstitution with FcγRIIB^+/+ versus FcγRIIB^−/− marrow in LDLR^−/− mice, or global deletion of the receptor in apoE^−/− mice. Independent of changes in plasma lipids, FcγRIIB omission by either strategy resulted in exaggerated atherosclerosis. In the study with global FcγRIIB silencing, there were both increased proinflammatory cytokines in the aorta and increased antibody titers to modified LDL. The latter finding may have been related to a loss of suppression of B cell production of antibodies to LDL^80,81. Thus, consistent with the inflammatory nature of atherosclerosis, and with impacts that parallel their activating versus inhibitory input into immune responses, FcγR modulate atherosclerosis in mice.

Additional potential evidence of FcγR modulation of atherosclerosis severity comes from studies of the effect of immunoglobulin treatment in apoE^−/− mice. The injection of polyclonal intravenous immunoglobulin preparations (IVIg) reduces fatty streak formation, and this is associated with reduced IgM antibodies to oxidized LDL and the inactivation of spleen and lymph node T cells⁸². Whereas the administration of intact immunoglobulin causes a decrease in lesion size and fewer macrophages within lesions, treatment with F(ab’)2 fragments of human immunoglobulin has no effect, thereby implicating involvement of the Fc region of IgG and FcγR.

The potential contribution of CRP to atherogenesis has also been investigated, using primarily gain-of-function strategies. Using generally similar strategies in male apoE^−/− mice, the transgenic expression of human CRP has been found to either accelerate atherosclerosis progression⁸³, or not affect the development or severity of atherosclerosis, with the increased atherosclerosis in the former study being attributed to excessively high CRP levels^84,85. In apoE*3-Leiden transgenic mice expressing a human CRP transgene, the CRP did not affect atherosclerosis⁸⁶. In atherosclerosis-prone ApoB^100/100;LDLR^−/− mice with human-like hypercholesterolemia, human CRP blunted atherosclerosis development⁸⁷. In apoE^−/− mice administered CRP via osmotic minipump for 4 weeks, atherosclerosis was unchanged⁸⁸. Compared with wild-type rabbits, transgenic rabbits expressing human CRP developed similar atherosclerotic lesions on a cholesterol-rich diet⁸⁹. Regarding loss-of-function studies, CRP-deficient mice displayed either equivalent or actually increased atherosclerotic lesions compared with controls⁹⁰, and whereas CRP antisense oligonucleotides effectively lowered plasma CRP levels in Watanabe heritable hyperlipidemic rabbits, they did not impact atherosclerosis in the rabbits⁹¹. Recognizing that SAP is the acute phase reactant and CRP equivalent in mice, SAP concentrations have been measured during the development of atherosclerosis in apoE^−/− mice, and they were found to be unchanged^84,85. Collectively these observations in mice and rabbits do not support a role for CRP in atherosclerosis pathogenesis. Gain- or loss-of-function studies of potential participation of SAP in modulating atherosclerosis severity in mice have not yet been reported.

Vascular Injury, Occlusion and Thrombosis

The potential participation of FcγR and their ligands in altering the response to vascular injury and degree of thrombosis has been interrogated in mice. Using human CRP transgenic mice, it was demonstrated that the pentraxin causes exaggerated neointima formation following carotid artery ligation (Fig. 1B)⁹². Followup studies were done crossing the human CRP transgenic mouse with FcγR null mice, and in them neointima development was decreased with either FcγRI or FcγR γ chain deletion. In contrast, the deletion of FcγRIIB or FcγRIII resulted in neointima formation that was equal to or greater than that observed without FcγR manipulation; the findings with FcγRIII excision may have resulted from decreased competition for the γ chain that is common between FcγRI and FcγRIII⁹³. Studies have also been performed in rats, with CRP administration causing increased neointima formation in a carotid artery angioplasty model⁵⁷. In addition to influencing neointima formation, it has been demonstrated that CRP promotes thrombosis, with CRP transgenic mice displaying exaggerated clot formation in both a transluminal wire injury model and a photochemical arterial injury model of thrombosis (Fig. 1B).⁹⁴ Furthermore, there is evidence that the impact of FcγR and their ligands on neointima development may relate to their modulation of thrombosis, since in a trans-femoral artery wire injury model in which thrombosis was controlled with aspirin and heparin, CRP transgenic mice displayed less neointima formation.⁹⁵ The interpretation of findings regarding neointima development with FcγR manipulation may be complex because the FcγR γ chain plays a key role in platelet activation by collagen, and platelet activation may be instrumental to neointima initiation and progression.⁹⁶ Overall, the animal model data available to date indicate that FcγR impact thrombotic and non-thrombotic vascular occlusion.

Hypertension

Since chronic elevations in CRP are associated with the development of hypertension in humans⁹⁷, to evaluate a possible causal relationship, blood pressure has been studied in transgenic mice expressing rabbit CRP under the regulation of the phosphoenolpyruvate carboxykinase promoter. Compared with controls, CRP transgenic mice had hypertension that was predominantly systolic, and the severity of hypertension varied in parallel with changes in CRP levels modulated by dietary carbohydrate manipulation (Fig. 1B). The regulated transgene made it possible to study CRP levels as low as 9ug/ml, and such mice were hypertensive, indicating that modest elevations in CRP are sufficient to alter BP in the mouse. The CRP transgenic mice displayed exaggerated BP elevation in response to angiotensin II but not in response to norepinephrine, and there was a reduction in vascular angiotensin II receptor subtype 2 (AT2) expression. In contrast, the decline in BP with angiotensin II receptor subtype 1 (AT1) antagonism and vascular AT1 abundance were unaltered, indicating a selective effect of CRP on AT2. Ex vivo experiments further showed that the CRP-induced decrease in vascular AT2 is a direct effect on the vascular wall not requiring systemic responses, and that it is reversed by an NO donor, suggesting a role for NO deficiency in the in vivo process. In parallel, the chronic inhibition of NO synthase in wild-type mice attenuated vascular AT2 expression without affecting AT1. These findings provided direct evidence for CRP-induced hypertension in mice⁹⁸. In rats, human CRP expression via an adeno-associated virus similarly resulted in hypertension, and there was also evidence of decreased NO production indicated by a fall in serum NO and urine cGMP and impaired endothelium-dependent relaxation⁹⁹. In a complementary study, human CRP was expressed in the livers of spontaneously-hypertensive rats using a transgene under the control of the apoE promoter. The rats with elevated CRP displayed both greater systolic and diastolic blood pressure¹⁰⁰. These collective findings indicate that CRP causes hypertension in rodents.

Myocardial Infarction

Since the relative rise in plasma CRP provoked by a myocardial infarction in humans is strongly associated with postinfarct morbidity and mortality^101–103, a potential contribution of CRP to disease severity has been evaluated. In studies in rats, the administration of human CRP following coronary artery ligation caused a 40% increase in infarct size (Fig. 1B)¹⁰⁴. A small molecule inhibitor of CRP was then identified, 1,6-bis (phosphocholine)-hexane, which binds to CRP and occludes its ligand-binding B-face and thereby blocks its function. When administered to rats given human CRP following coronary artery ligation, the agent attenuated the increase in infarct size and the cardiac dysfunction caused by CRP.¹⁰⁵ Using a rat CRP-specific antisense oligonucleotide, it was recently observed that lowering blood CRP resulted in reduced infarct size and improved cardiac function following myocardial infarction caused by ligation of the left anterior descending coronary artery in rats. In addition, in human CRP transgenic mice, declines in circulating human CRP induced by a human CRP-specific antisense oligonucleotide were associated with decreased neointima formation following carotid artery ligation¹⁰⁶. The specific actions of CRP as well as possible isoforms of FcγR that influence myocardial infarction severity or outcome are unknown.

Insulin Resistance

Individuals with insulin resistance have markedly greater risk of developing cardiovascular disease, and there is mechanistic linkage between insulin resistance and endothelial dysfunction and vascular disease⁵⁰. Since chronic elevations in CRP are associated with increased risk of both cardiovascular disorders and type 2 diabetes in humans^107–111, the impact of CRP on glucose homeostasis has been queried in mice. Using the mouse transgenic for rabbit CRP or the administration of human recombinant CRP to wild-type mice, it was discovered that elevations in CRP cause insulin resistance (Fig. 1B)¹¹². Paralleling these findings, spontaneously-hypertensive rats expressing CRP driven by an apoE promoter-regulated transgene display glucose intolerance and hyperinsulinemia¹⁰⁰. In the CRP transgenic mice, animals lacking FcγRIIB were protected from CRP-induced insulin resistance, and immunohistochemistry revealed that FcγRIIB is expressed in skeletal muscle microvascular endothelium. Consistent with the receptor distribution, the primary mechanism in glucose homeostasis disrupted by CRP was skeletal muscle glucose delivery, and CRP attenuated insulin-induced skeletal muscle blood flow. In contrast, CRP did not impair skeletal muscle glucose delivery in FcγRIIB^−/− mice or in eNOS knock-in mice with phosphomimetic modification of eNOS Ser1176 (Ser1179 in human eNOS); eNOS Ser1176/9 is normally phosphorylated by insulin signaling to stimulate NO-mediated skeletal muscle blood flow and glucose delivery, and is it dephosphorylated by CRP/FcγRIIB¹¹² Thus, CRP causes insulin resistance in mice through FcγRIIB-mediated inhibition of skeletal muscle glucose delivery, and this may represent an additional mechanism whereby FcγR and one of its ligands contribute to cardiovascular disease.

FcγR Ligands and CVD in Humans

IgG and Immune Complexes

Specific antigens and their respective IgG antibodies have been detected in the serum of patients with cardiovascular disease. In addition, IgG deposits have been detected in atherosclerotic lesions both in humans and in mice^114–115. Two identified antigens in humans are oxidized LDL and heat shock protein^116,117, and along with their corresponding antibodies they have the potential to form immune complexes which may contribute to cardiovascular disease due to their proinflammatory properties involving either FcγR or complement activation. In an investigation in 52 patients undergoing carotid artery endarterectomy, there was a correlation between the levels of IgG against oxidized LDL at the time of surgery and arterial wall thickness six months later¹¹⁸. In a study of over 500 patients undergoing coronary angiography, IgG oxidized LDL autoantibodies and IgG apolipoprotein B-100 immune complexes were positively associated with angiographically-determined coronary artery disease in logistic regression analyses. However, in this population the relationships were not apparent in multivariate analyses¹¹⁶, and inconsistencies in correlations between IgGs and manifestations of cardiovascular disease have been observed between studies¹¹⁹. A potential explanation for such inconsistencies may be the lack of reproducible antigens used to quantify the IgGs¹²⁰. Investigations in larger numbers of subjects, for example in the EPIC-Norfolk cohort, have failed to demonstrate a relationship between levels of IgG autoantibodies or immune complexes and coronary artery disease events¹²¹. Interestingly, for decades it has been recognized that serum levels of antibodies are elevated in patients with both essential and pregnancy-associated hypertension^122–125. The antigens targeted by these autoantibodies include the angiotensin II type-1 receptor, L-type voltage gated calcium channels, the alpha-1 adrenergic receptor, and the beta-1 adrenergic receptor¹²⁶. Recognizing that these proteins regulate blood pressure by governing sodium and water reabsorption by the kidney, vascular tone, and cardiac output, the potential contributions of the autoantibodies to disease pathogenesis have thus far logically been attributed to antibody recognition of their antigen targets, and not to FcγR activation. Systemic autoimmune diseases, including systemic lupus erythematosus, rheumatoid arthritis and the antiphospholipid syndrome, are characterized by accelerated atherosclerosis and greater risk of coronary artery disease that is not explained by traditional risk factors. The increase in vascular disease risk associated with these autoimmune conditions hypothetically may be related to the presence of autoantibodies and autoantigens and the subsequent formation of immune complexes¹²⁷. However, evidence for a causal link between immune complex formation and cardiovascular disease risk or severity is currently lacking in patients with autoimmune disorders, as well as in individuals without known complications of autoimmunity.

Pentraxins

Pentraxins and Coronary Heart Disease

Over the past few decades massive data has accumulated indicating an association between modest elevations in circulating CRP and incident coronary heart disease¹⁰⁷. This is perhaps best exemplified by two meta-analyses that have been performed, the first entailing 22 studies including over 7000 coronary heart disease cases and a mean follow-up period of 12 years. Comparing individuals in the top one-third versus the bottom one-third for baseline CRP level, the odds ratio for disease was 1.58 (95% CI 1.48–1.68)¹²⁸. The second analysis of 23 studies comparing subjects with CRP levels of >3.0 mg/l (>3.0 ug/ml) versus <1.0 mg/l (<1.0 ug/ml) yielded a similar odds ratio of 1.58 (1.37–1.83)¹²⁹.

Another approach that has been used to evaluate potential linkage between CRP and cardiovascular disease incidence or severity is to study an intervention that alters CRP and then determine how outcome is modified. This has been done in studies of statins which lower not only LDL cholesterol but also CRP¹³⁰. In a study of 502 individuals with angiographically documented coronary disease, the impact of moderate versus intensive statin therapy (40 mg pravastatin versus 80 mg atorvastatin daily, respectively) on atherosclerosis progression was evaluated using intravascular ultrasonography. Decreases in both LDL cholesterol and CRP predictably occurred with statin treatment, and after adjustment for changes in lipids, the declines in CRP were independently correlated with less atherosclerosis progression¹³¹. In the JUPITER trial involving almost 18,000 subjects without cardiovascular disease at baseline, LDL cholesterol levels below 130 mg/dl, and CRP levels greater than or equal to 2 mg/l, outcomes with placebo versus 20 mg rosuvastatin treatment daily were compared. The trial was stopped early at a median of 1.9 years of follow-up when a 44% decline in vascular events was observed with rosuvastatin. This was associated with a 50% fall in LDL cholesterol and a 37% decrease in CRP¹³². The event rate in the placebo group and the clinical impact observed with treatment and the associated fall in CRP suggested that elevated levels of CRP increased vascular disease risk in these subjects, even when the LDL cholesterol level was within acceptable range based on guidelines at the time of the trial. However, recognizing that statins have numerous actions besides lowering CRP, the statin-based intervention trials shed only modest light on whether there is causal linkage between CRP and cardiovascular disease.

In addition to the effects of inflammatory insults and behavioral and environmental factors on CRP levels, 20–50% of the differences in CRP between individuals has been attributed to genetic variability¹³³. Several single nucleotide polymorphisms (SNPs) in the CRP gene influence its abundance, and loci in other genes involved in inflammatory pathways have also been identified to impact CRP^133,134. The identification of genetic modifiers of CRP allowed numerous studies to be performed involving Mendelian randomization, in which the genetic variants serve as likely unconfounded proxies for CRP levels to evaluate if CRP has a causal role in coronary heart disease¹³⁵. The outcomes of the Mendelian randomization queries are exemplified by the collaborative study reported in 2011 involving close to 150,000 control subjects and more than 46,000 individuals with prevalent or incident coronary heart disease. Although four SNPs were demonstrated to influence CRP levels, and increasing CRP concentration was associated with increased risk of disease, no association was found between the SNPs and coronary heart disease¹³⁶. When combined with the overall neutral outcomes of animal studies of CRP and atherosclerosis, such findings indicate that CRP does not likely contribute to coronary heart disease pathogenesis in humans.

Pentraxins and Hypertension

In a number of reports CRP levels have been shown to predict the development of hypertension in previously normotensive individuals. This was the case in the Women’s Health Study, in which 5365 women out of 20,525 with normal blood pressure at study onset developed hypertension over a median follow-up duration of 8 years¹³⁷. Even after adjustment for a variety of risk factors, baseline CRP was an independent predictor of the development of incident hypertension. Comparable observations were made in the Framingham Offspring Study, in which 232 subjects out of 1456 initially normotensive individuals developed hypertension over a mean follow-up period of 3 years¹³⁸. Thus, there is epidemiologic evidence linking CRP levels with incident hypertension.

As has been done in queries of possible linkage of CRP with coronary artery disease but in far fewer studies, the genetics governing CRP level and hypertension prevalence or development have been investigated. In a British Women’s Heart and Health Study investigation of a single CRP SNP, although the SNP was associated with a marked difference in CRP level and CRP was associated with systolic BP, there was no relationship between the SNP and the prevalence of hypertension¹³⁹. In a study of almost 2000 Turkish subjects, CRP haplotypes were associated with hypertension in both men and women¹⁴⁰, and two CRP SNPs were associated with hypertension in a study of 1400 control and 1331 Han Chinese with elevated BP¹⁴¹. However, in a larger study of 2000 Han Chinese in which eight CRP SNPs were investigated, although CRP levels were associated with increasing systolic as well as diastolic BP, none of the SNPs were associated with prevalent or incident hypertension¹⁴². In a recent study a weighted genetic risk score was created to evaluate the combined effects of genetic variants associated with changes in CRP levels on blood pressure in almost 750 Korean subjects. Individuals with the highest genetic risk score had CRP levels that were approximately 2.5-fold higher than subjects with the lowest genetic risk score, and an elevated genetic risk score increased the likelihood of hypertension (OR 2.18)¹⁴³. Since it has been demonstrated that elevations in CRP cause hypertension in rodents (see above), additional human studies of the genetics of CRP regulation and hypertension may be warranted, including the application of the weighted genetic risk score for CRP to other populations.

Fcγ Receptors and CVD in Humans

Activating Fcγ Receptors

Activating FcγR, in particular FcγRI, FcγRIIA and FcγRIIIA, have been detected in human atherosclerotic lesions and in macrophages and other cell types in the medial and adventitial regions of the vascular wall¹⁴⁴. The observed functions of activating FcγR in cell types of relevance to vascular health and the findings with receptor deletion in animal models of atherosclerosis (Fig. 1B) provide sound rationale for studies determining if genetic variants in FcγR influence the incidence or severity of vascular disease in humans. In a study of almost 900 subjects undergoing coronary angiography, the polymorphism FcγRIIIA-F158V was related to coronary artery disease, with individuals with FcγRIIIA-158V/V having decreased risk of disease (OR 0.53, CI 0.32–0.90). From a functional perspective, FcRγIIIA-158V/V displays greater IgG1 and IgG3 binding than FcγRIIIA-F/F¹⁴⁵.

FcγRIIA is another activating FcγR for which genetic variation has been evaluated in a number of studies of cardiovascular disease. In a query of the polymorphism FcγRIIA-R131H involving over 700 patients with a first acute coronary syndrome (ACS) event compared to almost 500 individuals with stable angina pectoris, the FcγRIIA-131R/R genotype was associated with ACS as the first manifestation of coronary disease (OR 2.86, CI 2.06–3.99)¹⁴⁶. In 553 individuals with either stable angina pectoris or unstable angina pectoris, those with FcγRIIA-131R/R were more likely to have unstable angina (OR 4.02, CI 2.52–6.41)¹⁴⁷. In an evaluation of 430 individuals with peripheral atherosclerosis and 411 controls in the Rotterdam Study, FcγRIIA-131H heterozygous and homozygous subjects were protected against advanced peripheral atherosclerosis. The age- and gender-adjusted odds ratios were 0.77 (CI 0.54–1.12) and 0.65 (CI 0.44–0.98), respectively¹⁴⁸. In 78 hypercholesterolemic subjects, homozygous carriers of the H allele compared with the R allele displayed better endothelial dependent vasodilation as assessed by changes in forearm blood flow in response to intra-arterial acetylcholine infusion¹⁴⁹. In contrast to these four studies suggesting a cardiovascular health benefit of the H allele, in 1041 Finnish subjects, FcγRIIA-131H/H homozygotes had more premature atherosclerosis¹⁵⁰, and in a number of other investigations, the FcγRIIA-H131R polymorphism was found to have no impact on cardiovascular disease risk^{145,151–154}. Regarding impact on receptor function, FcγRIIA-131R/R has increased signal transduction upon CRP binding compared with FcγRIIA-131H/H¹⁵⁵. However, FcγRIIA-131H/H has higher binding efficiency for IgG2 and IgG3 than FcγRIIA-131R/R, resulting in decreased internalization of IgG2-opsonized particles by phagocytes in FcγRIIA-R/R individuals¹⁵⁶. In addition to these studies of FcγRIIA SNPs, differences in receptor expression have been investigated. An evaluation of FcγRIIA abundance on peripheral monocytes by flow cytometry found that the receptor expression is decreased in subjects with clinical atherosclerosis compared with controls¹⁵⁷. Recognizing that FcγRIIA participates with collagen in platelet activation, its abundance on platelets has also been assessed, and it was found to be increased in patients with acute myocardial infarction, unstable angina or ischemic stroke¹⁵⁸. Not surprising considering the complex nature of FcγR biology, the cumulative available information about genetic variation in activating FcγR or differences in their expression has not yet added clarity to our understanding of how the receptors may influence cardiovascular health and disease in humans.

Inhibitory Fcγ Receptor FcγRIIB

In an attempt to understand the potential participation of FcγRIIB in cardiovascular disease in humans, advantage has been taken of previous interrogations of genetic variation in the receptor in the context of lupus¹⁵⁹. One particular site of known single amino acid variation is amino acid 232, which is threonine (T) or isoleucine (I) depending on basepair 695 in exon 5 being C or T, respectively¹⁶⁰. The T allele is less common than the I allele, with the T allele frequency 0.13 in Caucasians and 0.29 in African Americans¹⁶¹. Recognizing that this SNP resides in the transmembrane domain and may thereby affect receptor function, the abilities of human FcγRIIB-T232 versus FcγRIIB-I232 to antagonize eNOS were compared in cultured endothelial cells. Doing so tests the impact of the variant in the cell type in which FcγRIIB actions potently influence vascular health^30,44,45. Cell context is critical because whereas prior studies of autoimmunity-related mechanisms showed less receptor function for FcγRIIB-T232 than FcγRIIB-I232 in monocytes, the opposite was found in B cells^161,162. Endogenous FcγRIIB was knocked down in bovine aortic endothelial cells (BAEC) by siRNA, followed by sham transfection or introduction of either human FcγRIIB-T232 or FcγRIIB-I232 to equal abundance by transient transfection (Fig. 3A). CRP (25 ug/ml) antagonism of eNOS activation by VEGF (100 ng/ml) was then evaluated. Whereas sham-transfected cells deficient in FcγRIIB displayed no antagonism by CRP, CRP inhibited eNOS activation in cells expressing FcγRIIB-T232 (Fig. 3B). In contrast, in cells expressing FcγRIIB-I232 there was no antagonism of eNOS by CRP. Thus, the identity of amino acid 232 in human FcγRIIB markedly affects the capacity of the receptor to alter endothelial cell function, providing a mechanism-based genetic entry point for the study of FcγRIIB in vascular disease pathogenesis in humans.

Figure 3.

The FcγRIIB-I232 variant has attenuated capacity to mediate CRP action in endothelium. Following siRNA-based knockdown of endogenous FcγRIIB (GAAACCAGCCUCUGAAU, Dharmacon), bovine aortic endothelial cells were transfected with sham plasmid or cDNA encoding human FcγRIIB-T232 or FcγRIIB-I232. A. FcγRIIB abundance was evaluated by immunoblotting, with eNOS detection providing assessment of protein loading. B. CRP (25 ug/ml) antagonism of eNOS activation by VEGF (100 ng/ml) was evaluated by measuring ¹⁴C-L-arginine conversion to ¹⁴C-L-citrulline conversion by intact cells over 15 min. Values are mean±SEM, n=4, *p<0.05 vs no CRP.

In light of the preclinical observations made regarding FcγRIIB, its ligands, eNOS, endothelial function and hypertension^30,44,45,98, FcγRIIB-I232/T232 and its potential influence on blood pressure were then interrogated in the Dallas Heart Study (DHS). DHS is a population-based epidemiologic study of Dallas County residents that has provided a cohort for successful genetic inquiry into processes underlying cardiovascular disorders and their risk factors^163–168. In the DHS, subjects between the ages of 30 to 65 years old have undergone extensive cardiovascular phenotyping¹⁶⁹, including detailed studies of blood pressure^163,170. 3493 subjects in DHS were genotyped for FcγRIIB-I232 versus FcγRIIB-T232, and due to high homology between FcγRIIB and FcγRIIC, the strategy entailed PCR followed by SNP genotyping by sequence-specific fluorescent hybridization probing¹⁶⁰. The prevalence of genotypes is shown in Table 3, and as previously observed¹⁶¹, the T allele was less common than the I allele, particularly in non-African Americans. The study population then consisted of 2925 subjects without active malignancy, inflammatory illness, or a diagnosis of lupus for which genotype, CRP and BP data were available. The relationship between FcγRIIB-I232/T232 genotype and systolic BP was then evaluated for all subjects with CRP>2.0 mg/L (n=2069), noting that FcγRIIB genotype was not associated with CRP level. The genotype distribution in the study population was similar to that of the total population, and both were within the parameters of Hardy-Weinberg. In DHS subjects with CRP>2.0 mg/L of any race/ethnicity, the IT or TT genotype was associated with higher systolic BP compared with the II genotype (Table 4, upper panel), which encodes an FcγRIIB that is less able to antagonize eNOS (Fig. 3). Recognizing that the IT or TT genotype is more prevalent in African Americans (Table 3), who in general have higher BP than non-African-Americans for a variety of reasons^171,172, separate analysis was performed in African Americans, and the association between genotype and systolic BP was again observed. Due to the low prevalence of the IT or TT genotype in Whites and Hispanics (Table 3), statistical power was insufficient to assess the genotype-phenotype link in those groups. Importantly, the 3–4 mmHg difference in systolic BP between genotypes observed for the SNP is similar in degree to the effect sizes for individual variants in known HTN susceptibility genes^173,174. Since a number of subjects in the DHS query were on antihypertensive treatment, the comparisons were repeated using the recommended strategy of adding a constant to the observed BP in treated subjects (15 mmHg added to systolic BP)¹⁷⁵. It was again observed that systolic BP is greater with IT or TT versus II genotype in the combined racial/ethnic groups and in African Americans (Table 4, lower panel). Importantly, subject age and sex distribution did not differ by genotype. These findings in a single population indicate that the FcγRIIB-I232/T232 variant impacts BP in the setting of elevated CRP in humans. Queries of the FcγRIIB-I232/T232 variant are now warranted in additional populations.

Table 3.

FcγRIIB-I232/T232 genotypes in the Dallas Heart Study population.

	II	IT	TT
Total Population (n=3493)	2297	1058	138
	66%	30%	4%
Blacks (n=1792)	948	721	123
	53%	40%	7%
Whites (n=1032)	806	218	8
	78%	21%	0.8%
Hispanics (n-595)	492	101	2
	83%	17%	0.3%
Other (n=74)	51	18	5
	69%	24%	7%

Table 4.

Impact of FcγRIIB-I232/T232 genotype on systolic blood pressure in Dallas Heart Study subjects with CRP>2.0 mg/L.

	Overall	African-American
IT or TT	130.6 ± 20.2	134.0 ± 21
	(725)	(536)
II	126.0 ± 18.9	131.0 ± 19.7
	(1344)	(608)
	P<0.05	P<0.05
Corrected for Antihypertension Therapy
IT or TT	134.7 ± 22.9	138.2 ± 23.7
	(725)	(536)
II	129.3 ± 21.4	135.5 ± 22.2
	(1344)	(608)
	P<0.05	P<0.05

(N values)

Conclusions and Current Unknowns

The recent interest in the contribution of inflammation to cardiovascular disorders has led to both preclinical and clinical studies of FcγR and their ligands in the context of cardiovascular disease. The activating FcγR, particularly FcγRI, FcγRIIA and FcγRIIIA, cause a variety of cellular responses in endothelial cells, vascular smooth muscle cells and monocytes/macrophages that may contribute to vascular disease pathogenesis (Fig. 1A). The lone inhibitory FγcR, FcγRIIB, which blunts proinflammatory processes in immune response cells, invokes detrimental processes in endothelial cells, inhibiting NO production, promoting adhesion and attenuating the capacity for endothelial monolayer repair. In animal models, activating FγcR promote atherosclerosis whereas FcγRIIB is atheroprotective, and activating FcγR additionally enhance thrombotic and non-thrombotic vascular occlusion (Fig. 1B).

Due to its upregulation under inflammatory conditions and perhaps also its stability in stored samples, the FcγR ligand CRP has been intensely studied. In mice and rats CRP causes hypertension and insulin resistance, and it exaggerates myocardial infarct size and related cardiac dysfunction (Fig. 1B). In contrast, CRP has been found to not impact atherosclerosis in animal models. A large volume of work has demonstrated an association between modest, chronic elevations in circulating CRP and coronary heart disease in humans. However, Mendelian randomization studies, which employ genetic variants influencing CRP abundance as unconfounded proxies for CRP levels, have shown that CRP is likely a marker of inflammation and not a mediator of coronary vascular disease. Studies evaluating CRP genetics and hypertension risk are currently insufficient to provide meaningful insight into a possible causal relationship.

Paralleling the genetic strategies assessing how CRP impacts cardiovascular health, inherited variants of activating FcγR have been studied. The information available to date does not yet add clarity to our understanding of how the receptors may influence vascular health and disease in humans. However, the first investigation of a loss-of-function variant of FcγRIIB in endothelium suggests that the receptor may contribute to the hypertension that often develops in individuals with chronically elevated circulating CRP (Fig. 3, Table 4).

At the same time that these new insights are recognized, multiple remaining knowledge gaps become apparent. Since the vast majority of mouse studies evaluating FcγR participation in vascular disease pathogenesis entailed global FcγR gene deletion, how individual FcγR in specific cell types impact cardiovascular health is unknown. This will require cell-specific gain- or loss-of-function studies in vivo. It is also poorly understood how individual FcγR ligands impact cardiovascular health in vivo, with different IgG subclasses having varying affinity for particular FcγR, and with FcγR additionally varying in their affinity for a given subclass of IgG. The identity of FcγR ligands that may be relevant to cardiovascular disease remains even more open-ended considering the potential diversity in circulating immune complexes, whose actions via FcγR may explain the greater predisposition for disease in individuals with autoimmune disorders. Finally, having observed that there is a common SNP in FcγRIIB that dramatically alters the endothelial actions of CRP (Fig. 3), it is critical to recognize that Mendelian randomization studies of FcγR ligands such as CRP or FcγR themselves may have limited utility because there is functionally-relevant genetic variation in multiple components in the FcγR ligand-receptor signaling pathway. Recognizing the various complexities that characterize FcγR biology, further attempts to understand how the receptors and their ligands impact vascular health will likely require the simplification that can only be afforded by discrete genetic manipulation or highly specific targeted pharmacologic intervention. Such efforts are warranted, however, because FcγR biology may underlie a considerable component of vascular disease predisposition that is not explained by traditional risk factors.

Non-standard Abbreviations and Acronyms

AT1

angiotensin II receptor subtype 1

AT2

angiotensin II receptor subtype 2

BCR

B cell receptor

BAEC

bovine aortic endothelial cells

BTK

Bruton’s tyrosine kinase

CRP

C-reactive protein

DC

dendritic cell

DHS

Dallas Heart Study

eNOS

endothelial NO synthase

FcγR

Fcγ receptors

GADD153

growth arrest- and DNA damage-inducible gene 153

ITAM

immunoreceptor tyrosine-based activation motif

ITIM

immunoreceptor tyrosine-based inhibitory motif

IVIG

intravenous immunoglobulin

LDL-IC

LDL cholesterol-containing immune complex

MMP-2

matrix metalloproteinase-2

NOX4

NADPH oxidase 4 isoform

PLCγ

phospholipase Cγ

SAP

serum amyloid P component

SHIP

SH2 domain-containing inositol polyphosphate 5′ phosphatase

SHP1

SH2-domain -containing protein tyrosine phosphatase 1

TF

tissue factor

TG-CRP

CRP-transgenic mice