Growth factors as immunotherapeutic targets in cardiovascular disease

Abstract

Growth factors, such as colony-stimulating factors (CSFs), epidermal growth factors (EGFs), and fibroblast growth factors (FGFs) are signaling proteins that control a wide range of cellular functions. Although growth factor networks are critical for intercellular communication and tissue homeostasis, their abnormal production or regulation occurs in various pathologies. Clinical strategies that target growth factors or their receptors are used to treat a variety of conditions, but have yet to be adopted for cardiovascular disease. In this review, we focus on M-CSF, GM-CSF, IL-3, EGFR, and FGF21. We first discuss the efficacy of targeting these growth factors in other disease contexts (i.e. inflammatory/autoimmune diseases, cancer, or metabolic disorders) and then consider arguments for or against targeting them to treat cardiovascular disease.

Introduction

Ischemic heart disease, the leading cause of non-communicable death in the U.S. and abroad, typically stems from atherosclerosis.^{1, 2} Aging, lifestyle factors, and genetic susceptibility converge to promote the accumulation of lipid-laden atherosclerotic plaques in the intimal walls of arterial vessels.^3–5 Over decades, these plaques contribute to many life-threatening complications, including myocardial infarction (MI) or stroke.⁶ Atherosclerotic plaque buildup was once thought to be a slow, passive process driven by cholesterol accrual (i.e. a lipid storage disorder). However, the identification of inflammatory and immune biomarkers (e.g. C-reactive protein (CRP), interleukin (IL)-6, anti-oxidized low-density lipoprotein (oxLDL) antibodies) and evidence that innate and adaptive immune cells enter plaques and contribute to their formation has led to atherosclerosis’ re-classification as a lipid-driven chronic inflammatory disease.^7–9 It is now clear that leukocytes not only play important roles in driving inflammation in atherosclerosis and other forms of cardiovascular disease (CVD) but also contribute to cardiac development and heart function.^{10, 11}

Despite the considerable advances in our understanding of inflammation and leukocyte involvement in CVD, treatment options for targeting inflammatory mediators in atherothrombotic disorders remain limited. The currently preferred drugs for primary and secondary prevention of MI, stroke, and death from other CVD complications are statins, a class of cholesterol biosynthesis inhibitors.^{12, 13} Statins are well-known to reduce hyperlipidemia, a major CVD risk factor, by lowering low-density lipoprotein (LDL) levels.^{14, 15} In addition, statins perform many pleiotropic, immunomodulatory, functions, such as the ability to suppress T cell activation, inhibit MHCII and adhesion molecule expression on vascular cells, and lower CRP, an inflammatory biomarker and independent predictor of major adverse cardiovascular events (MACE)^{12–14, 16–18}. Because of the pleiotropic effects of statins, however, it has remained unclear whether targeting inflammatory mediators could prove beneficial for treating CVD independently of lipid-lowering therapies.

Recently, the Canakinumab Anti-inflammatory Thrombosis Outcome Study (CANTOS) trial was designed to test the inflammatory hypothesis of atherothrombosis by targeting IL-1β, an active inflammatory mediator in atherosclerosis. The trial was a success: IL-1β neutralization lowered recurrent cardiovascular events, providing support for the inflammatory hypothesis and rationale for developing focused therapies that target active inflammatory cascades.^19–21 Importantly, blocking IL-1β, which is active in the acute inflammatory response, reduced CRP and IL-6 without affecting lipid levels.^{22, 23} There is no question that lowering lipids with statins is an effective approach for mitigating cardiovascular risk²⁴, but CANTOS has opened up another dimension concerning treatment strategy that should be considered in more depth, namely, the development of focused immunotherapeutic approaches to treat CVD.

Growth Factors and Their Receptors as Potential Immunotherapy Targets to Treat Cardiovascular Inflammation

Growth factors are signaling proteins that mediate diverse functions related to cellular differentiation and proliferation. Though important to homeostasis, their abnormal production or regulation can promote disease. Immunotherapeutic approaches, including those that target growth factors and their receptors, have demonstrated clinical success against cancers and acute or chronic inflammatory/autoimmune disorders, but have not been widely adopted for CVD.^{24, 25}

Though many different kinds of growth factors abound, we will limit our discussion to a handful for which there is both clinical evidence for treating cancer, inflammatory/autoimmune diseases, or metabolic disorders and pre-clinical evidence for treating cardiovascular disease (Table 1). First, the colony-stimulating factors (CSFs) are glycoproteins that regulate leukocyte survival, differentiation, and function.^{26, 27} In particular, the four CSFs – macrophage colony-stimulating factor (M-CSF; also called CSF-1), granulocyte-macrophage colony-stimulating factor (GM-CSF; also called CSF-2), granulocyte colony-stimulating factor (G-CSF; also called CSF-3), and interleukin-3 (IL-3; previously called multi-CSF) – differentially promote the proliferation, differentiation, mobilization, and survival of myeloid cells, which are derived from hematopoietic stem and progenitor cells (HSPCs) found in bone marrow niches and, in cases of extramedullary hematopoiesis, the spleen. These growth factors also regulate the mature myeloid cell functions in the steady state and under pathologic, inflammatory conditions.^{26, 27} Second, the epidermal growth factor receptor (EGFR), which binds multiple EGF ligands, and fibroblast growth factor 21 (FGF21) have multiple functions that modulate vascular smooth muscle cells, cardiomyocytes, cardiac fibroblasts, endothelial cells, adipocytes, and immune cells, with new functions arising.^{28, 29}

Table 1.

Overview of Growth Factor Pathways in Cardiovascular Disease and Their Available Therapies

Growth Factor Pathway	Clinically Available Therapies	Potential Role in CVD	Rationale for Targeting in CVD	Caveats for Targeting in CVD
M-CSF/CSF-1R	Monoclonal antibodies Small molecule inhibitors (TKIs)	Pathogenica	• Clinically predicts atherosclerosis progression and adverse outcome [42, 43] • Induces monocyte recruitment, promotes macrophage differentiation and survival, drives aortic inflammation in atherosclerosis [48, 53, 56] • Drives lesional accumulation of monocyte-derived DCs in atherosclerosis [58] • Elevates monocytosis in the bone marrow due to poor sleep in atherosclerosis [56] • Alters expression of genes in macrophages to favor inflammation, cholesterol retention, and uptake [45, 54, 55] • Induces monocyte production and monocyte/macrophage accumulation in viral myocarditis [61] • Promotes vascular inflammation and macrophage accumulation in aortic aneurysm [62]	• Role in myocardial ischemia is protective – mediates proper cardiac function, repair, and angiogenesis in models of MI [63–67] • Important roles in host defense and homeostastic functions that may be compromised if inhibited [68–72] • Reducing inflammatory macrophages with CSF-1R therapies in models remains in question [27] • Unclear how inhibiting M-CSF may affect the functions of IL-34, an alternate CSF-1R ligand [32, 73]
GM-CSF/GMRα:βc	Monoclonal antibodies Recombinant protein	Pathogenic	• Promotes survival and/or proliferation of myeloid progenitors and progeny in atherosclerosis and MI [85–89] • Induces TH1 immunity via generation of DCs in the spleen in atherosclerosis [90] • Regulates the accumulation and proliferation of myeloid cells, DCs, and T cells in atherosclerotic lesions [91,92] • Promotes macrophage apoptosis and oxidative stress in advanced lesions via IL-23 in atherosclerosis [93] • Promotes cardiac dysfunction after MI [89] • Drives production of inflammatory cytokines, myeloid cell recruitment, and/or proliferation in KD, EAM, aortic aneurysm, and MI [89, 100, 102, 104] • Promotes DC infiltration into the myocardium in EAM and after MI [98, 99] • Induces pathogenic TH17 cells and by signaling on DCs to sustain inflammation in EAM [105,107]	• Autoantibodies against GM-CSF interfere with the priming of neutrophil antimicrobial functions in patients with pulmonary alveolar proteinosis [110] • Autoantibodies against GM-CSF are associated with inflammation in IBD [111] • Promotes intestinal barrier integrity [112]
IL-3/IL-3Rα:βc	Monoclonal antibodies	Pathogenic	• Promotes survival and proliferation of myeloid progenitors and progeny in atherosclerosis [85–87] • Drives monocyte recruitment and local T cell proliferation in EAM [109] • Stimulates monocyte adhesion and activation in vitro [108] • Stimulates SMC migration and proliferation in vitro [108] • Stimulates EC proliferation, adhesion molecule expression, and cytokine production in vitro [108]	• Role in atherosclerosis, MI, and other types of CVD has not yet been tested in vivo or requires further study
G-CSF/CSF-3R	Monoclonal antibodies Recombinant protein	Protectiveb	• Reduces fibrotic areas after MI [121, 122] • Accelerates vessel formation after MI [119, 121, 122] • Improves cardiac function after MI [119–121] • Protects against cardiac remodeling after MI [119] • Protects against apoptosis including in cardiomyocytes after MI [119, 120]	• Clinical trials utilizing G-CSF for post-MI repair yielded inconclusive effects [118] • Conflicting reports on the role of G-CSF infusion in models of atherosclerosis and modestly protective overall [125] • Mechanisms surrounding the effects of cells mobilized by G-CSF on CVD are still unclear/inconclusive [118, 123, 124]
EGFs/EGFR	Monoclonal antibodies Small molecule inhibitors (TKIs)	Pathogenic	• Promotes macrophage accumulation, oxLDL uptake, and inflammatory cytokine production in atherosclerosis [136, 138] • Induces T cell activation, proliferation, and cytokine production in peripheral lymphoid organs and in atherosclerotic plaques [137] • Stimulates monocyte migration in vitro [134] • Amphiregulin, an EGFR ligand, produced by macrophages promotes cardiac dysfunction and fibrosis after MI [140] • Promotes cardiac and vascular hypertrophy and dysfunction via effects on both SMCs and ECs (e.g. proliferation, migration, activation, etc.) in vascular diseases including hypertension and AAA [131, 133]	• Role in atherosclerosis, MI, and other types of CVD has not yet been tested in vivo or requires further study
FGF21/FGFR:KLB	Receptor agonists Engineered analogs	Protective	• Attenuates plaque formation and macrophage accumulation in part by inhibiting liver cholesterol biosynthesis in atherosclerosis [153] • Inhibits smooth muscle cell proliferation in atherosclerosis [153] • Prevents foam cell formation by increasesing cholesterol efflux and reducing expression of inflammatory mediators in vitro [159–161] • Counteracts AngII-induced hypertension and vascular inflammation by inducing ACE2 in adipocytes and renal cells [158] • Attenuates cardiac remodeling and promotes repair by promoting cardiomyocyte survival after Mi [154, 157] • Ameliorates cardiac fibrosis, hypertrophy, oxidative stress, and inflammation in part via signaling in cardiomyocytes in diabetic cardiomyopathy [156]	• Role in atherosclerosis, MI, and other types of CVD requires further study, especially with respect to FGF21 effects on leukocytes and the activities of specific FGFRs

^aProtective in settings of myocardial ischemia.

^bOverall appears to be protective, but conflicting data between animal models and clinical trials for G-CSF infusion post-MI. Conflicting data and only modest effects surrounding studies in models of atherosclerosis.

Abbreviations: AAA, abdominal aortic aneursym; ACE2, Angiotensin converting enzyme 2; AngII, Angiotensin II; CVD, cardiovascular disease; EAM, Experimental autoimmune myocarditis; EC, Endothelial cell; FGFR, Fibroblast growth factor receptor; GMRα, GM-CSFRα; IBD, Inflammatory bowel disease; IL-34, Interleukin-34; KD, Kawasaki disease; KLB, β-klotho co-receptor; MI, myocardial infarction; oxLDL, oxidized low-density lipoprotein; SMC, Smooth muscle cell; TH, T helper; TKIs, tyrosine kinase inhibitors.

Macrophage Colony-Stimulating Factor

Among CSFs, M-CSF is a key growth factor relevant to the production, proliferation, survival, and function of many mononuclear phagocytes. As such, M-CSF is readily detectable in the circulation and is produced locally in tissues by different cell types, including endothelial cells (ECs), epithelial cells, neurons, and osteoblasts. M-CSF’s receptor, CSF-1R (CD115), is expressed on most mononuclear phagocytes and their progenitors.^{27, 30–32} In Csf1r^−/− and Csf1^op/op mice (Csf1^−/−) (the latter have a null mutation in the Csf1 locus), many tissue macrophages and osteoclasts are severely compromised.^32–37 Moreover, dendritic cells (DCs), including the skin-resident Langerhans cells, depend on the M-CSF:CSF-1R axis for their development since all DCs augment CSF-1R during their differentiation and are highly reduced in Csf1^−/− mice.^{38, 39} Although blood monocyte development from bone marrow precursors does not entirely rely on M-CSF, it can boost monocyte production and modulate their functional capacity to express cytokines and cell surface receptors.^{37, 39}

To date, M-CSF and CSF-1R have been targeted for treating rheumatoid arthritis (RA) and cancer with monoclonal antibodies (mAbs) and oral kinase antagonists that inhibit the CSF-1R’s receptor tyrosine kinase (RTK) activity. Clinical trials for RA have reported minimal efficacy, failed to disclose information, or remain ongoing.²⁷ In cancer, these strategies are still being tested for solid tumors as monotherapies and in combination with immunotherapies, and have yielded some benefit in phase I trials.⁴⁰ The M-CSF:CSF-1R axis has not been targeted for cardiovascular disease despite some intriguing data. First, M-CSF is a causal inflammatory biomarker in coronary artery disease and strongly predicts atherosclerotic plaque progression and adverse outcome.^41–43 Second, M-CSF is augmented in experimental atherosclerotic lesions, can also be detected in human plaques, and its expression increases in endothelial cells and macrophages exposed to oxLDL.^44–47 Third, M-CSF is atherogenic in mouse models as both Csf1^−/− and Csf1^+/− mice on the Apoe^−/− and Ldlr^−/− backgrounds are protected from atherosclerosis in spite of elevated circulating cholesterol levels, which are curiously increased in fully-deficient animals.^48–51 Likewise, Apoe^−/− mice treated with anti-CSF-1R mAb or a CSF-1R kinase inhibitor develop smaller atherosclerotic lesions.^{52, 53} Fourth, work performed in vitro suggests that M-CSF signaling in macrophages promotes atherogenesis by altering metabolic pathways to favor cholesterol retention, inducing monocyte chemoattraction, and enhancing scavenger receptor activity, which aids LDL uptake.^{45, 53–55} Fifth, observations in vivo have revealed that signaling downstream of CSF-1R favors plaque progression by promoting aortic inflammatory gene expression and macrophage survival and by inducing monocyte production, differentiation, and recruitment from bone marrow precursors (Figure 1).^{48–50, 52, 53, 56} For example, we have recently discovered that poor sleep, a lifestyle risk factor for CVD, aggravates atherosclerosis through heightened monocyte production incited by M-CSF derived from pre-neutrophils in the bone marrow.^{4, 57} Sixth, M-CSF may act on monocytes in the aorta to drive the differentiation and accumulation of CD11c⁺ CD11b⁺ F4/80⁺ monocyte-derived DCs, which are highly reduced in Csf1^−/− aortas and, in contrast to classical DC subsets, operate independently of Flt3:Flt3L signaling.^{58, 59} Although the role of M-CSF-dependent myeloid-derived DCs in atherosclerosis is still unclear, this DC subset becomes dramatically elevated in the aortas of atherosclerotic mice along with CD11c⁻ macrophages and may perpetuate chronic inflammation by sustaining T cell activation, proliferation, and inflammatory cytokine production (Figure 1).^58–60 In acute models of inflammation in CVD, M-CSF also exerts pathologic activities. During myocarditis, M-CSF rises in the heart, where it influences the accumulation of inflammatory monocytes and macrophages, and it also drives bone marrow monocyte production, possibly by reaching the circulation.⁶¹ Following aortic aneurysm, M-CSF increases in the vasculature and promotes inflammation and macrophage accumulation, though whether M-CSF recruits monocytes to the vasculature during aortic aneurysm is unknown.⁶²

Figure 1.

Pathologic effects of colony-stimulating factors in cardiovascular disease. M-CSF, GM-CSF, and IL-3 promote hematopoiesis in the bone marrow and spleen by triggering hematopoietic stem and progenitor cell proliferation, survival, and differentiation into myeloid cell subsets (i.e. monocytes and neutrophils), especially in the setting of hypercholesterolemia. Myeloid cells are recruited from the blood into cardiovascular organs, including the arterial vessels and myocardial tissues, via chemokine gradients established in part by CSFs acting on cells in the tissue sites (e.g. cardiac macrophages). In the intima of arterial vessels, GM-CSF and M-CSF promote the differentiation of monocytes into macrophages and monocyte-derived DCs, macrophage proliferation, activation, and production of inflammatory mediators and chemokines, foam cell formation via scavenger receptor upregulation, and cell death in advanced lesions, all of which contribute to inflammation and the accrual of leukocytes. Although GM-CSF also promotes the accumulation of CD11c⁺ dendritic cells and T cells in plaques, it is unclear whether this occurs directly or indirectly. Within cardiac tissues in settings of myocardial inflammation (e.g. MI, KD, myocarditis), GM-CSF produced by fibroblasts can stimulate myeloid cell recruitment by signaling on cardiac macrophages’ production of chemoattractants as well as stimulate bone marrow hematopoiesis distally via the circulation. M-CSF secreted by cardiac macrophages can also promote distal bone marrow myelopoiesis in myocarditis. CD4⁺ T cells promote cardiac fibroblasts’ production of GM-CSF in part via the production of IL-17A. In myocarditis, IL-3 produced by autoreactive T cells stimulates cardiac macrophages to produce monocyte chemoattractants, which guides monocyte recruitment and differentiation into monocyte-derived APCs. In turn, monocyte-derived APCs stimulate local T cell proliferation which thereby fuels leukocyte accumulation and cardiac inflammation.

While the above observations suggest that inhibiting the M-CSF:CSF-1R axis may be beneficial in chronic and acute cardiovascular inflammation by dampening monocyte infiltration and the development of inflammatory macrophages and myeloid-derived DCs, challenges remain. Specifically targeting M-CSF during bouts of myocardial ischemia, such as in MI, may instead prove detrimental to host physiology based on multiple animal studies. For example, introducing recombinant M-CSF into animals after experimental MI improves their heart function, promotes cardiomyocyte survival, and stimulates infarct repair and angiogenesis.^63–65 Such positive effects of M-CSF appear to rely, in part, on the mobilization of bone marrow CXCR4⁺ cells and monocytes, which differentiate into macrophages and proliferate in healing infarcts, as well as the anti-inflammatory functions of cardiac-resident CSF-1R⁺ macrophages, which participate in reparative processes after an MI.^{64, 66, 67} Thus, hampering monocyte infiltration and macrophage differentiation and function by blocking M-CSF may interfere with proper myocardial tissue repair.

Apart from the Janus-like roles of M-CSF in cardiovascular inflammation and ischemia, other potential hurdles that may arise downstream of M-CSF inhibition also require consideration. First, suppressing M-CSF or its receptor may compromise important monocyte and tissue macrophage host defense and homeostatic functions, such as thermogenesis, neuron pruning, iron recycling, and gut motility.^68–72 Secondly, preclinical disease models show the actual efficiency of reducing inflammatory macrophages with CSF-1R-directed therapies requires further clarification.²⁷ Complicating things further, IL-34, a recently identified CSF-1R ligand that competes with M-CSF for binding, regulates mononuclear phagocytes with somewhat overlapping yet distinct functions and may independently contribute to inflammatory disorders.^{32, 73} Overall, it remains unclear how targeting the M-CSF:CSF-1R axis would fare as a treatment strategy for CVD. Rather, it may be more effective to block or deplete M-CSF’s cellular sources, such as ECs and myeloid cells, with carefully directed therapies for atherothrombotic disorders in which M-CSF is believed to stimulate pathologic, inflammatory myeloid cell activity.^{44, 57, 61}

Granulocyte-Macrophage Colony-Stimulating Factor and Interleukin-3

Although GM-CSF and IL-3 are classified as CSFs, they are also, like IL-5, β common (β_c) cytokines that rely on the β_c receptor (CD131; also CSF-2RB or IL-3Rβ) for signaling. To elicit signaling, the β_c subunit forms a heterodimeric surface receptor with cytokine-specific receptor α subunits, such as GM-CSFRα, IL-3Rα, and IL-5Rα.⁷⁴ Though produced in the steady state, particularly in the lung, GM-CSF production can be augmented during injury or infection by many cell types, including T cells, B cells, macrophages, and fibroblasts, among others, while GM-CSF’s cellular targets are primarily myeloid cells, hematopoietic progenitors, and some non-myeloid populations (e.g. neurons, epithelium, endothelium). Consequently, GM-CSF has been linked to several inflammatory and autoimmune disorders.^75–77 In clinical trials, directly targeting the GM-CSF:GM-CSFRα axis with mAbs in RA has been successful up to phase II, whereas phase I/II trials for multiple sclerosis, psoriasis, and asthma are either ongoing or deem the therapy tolerable for patients.^{27, 76} In the clinic, GM-CSF’s immunostimulatory effects can overcome immune suppression in sepsis, boost myeloid cell recovery after chemotherapy, and induce hematopoietic regeneration after bone marrow transplantation.^{26, 78} Inflammatory stimuli such as infection and injury also augment IL-3, which appears to be more T cell-restricted. However, human mast cells and innate response activator (IRA) B cells, which are B1-derived cells that induce myelopoiesis by secreting growth factor cytokines, can also produce IL-3.^{74, 79, 80} IL-3 has numerous targets since myeloid cells, including the granulocytic subtypes, plasmacytoid dendritic cells, hematopoietic progenitors, and ECs, can be regulated by IL-3.^{27, 74, 81–83} In contrast to GM-CSF, the IL-3:IL-3Rα axis has not been clinically targeted to inhibit inflammation, though IL-3Rα-specific mAbs have been used to eliminate leukemic cells, which highly augment IL-3Rα in hematologic malignancies.⁸⁴

Preclinical reports suggest that GM-CSF and IL-3 drive cardiovascular inflammation by promoting leukocyte accumulation in cardiovascular organs (Figure 1). In experimental atherosclerosis, plaque myeloid cell accumulation is influenced by the distal expansion of β_c receptor⁺ HSPCs, which are found in the bone marrow and spleen. Work using Apoe^−/− mice and Abca1^−/− Abcg1^−/− mice with defective cholesterol efflux showed that hypercholesterolemia elevates the β_c receptor on HSPCs. In turn, HSPC β_c receptor signaling via STAT5 induces their proliferation and survival, leading to monocytosis, neutrophilia, and, subsequently, leukocyte accumulation in plaques.^85–88 Similarly, β_c receptor signaling on bone marrow HSPCs gives rise to inflammatory monocytes and neutrophils, which mediate left ventricular rupture in the infarcted myocardium in experimental MI.⁸⁹ Though the potential sources and individual roles of GM-CSF and IL-3 have not been thoroughly addressed in atherosclerosis, experimental evidence suggests that both growth factors can promote HSPC proliferation and that IRA B cells can peripherally produce GM-CSF and IL-3.^{85–88, 90} GM-CSF derived from IRA B cells can also boost adaptive T_H1 immunity in atherosclerosis via inflammatory DC generation in peripheral organs.⁹⁰ In MI, however, bone marrow HSPC proliferation appears to rely distinctly on circulating GM-CSF derived from cardiac fibroblasts rather than from a peripheral cellular source.⁸⁹

Apart from its roles in driving myelopoiesis in the bone marrow and spleen in atherosclerosis and MI, GM-CSF also seems to control leukocyte accumulation and cardiovascular inflammation by acting on locally situated cells in plaques and cardiac tissue (Figure 1). For example, in nascent lesions, GM-CSF regulates the proliferation and accrual of intimal leukocytes, including CD11c⁺ dendritic cells (DCs), CD11b⁺ myeloid cells, and T cells, whereas in established lesions, it triggers inflammatory cytokine and reactive oxygen species (ROS) production and promotes apoptosis and plaque necrosis.^91–93 It remains unclear whether GM-CSF’s local actions are specific to the Ldlr^−/− background because of the divergent effects of GM-CSF deficiency and administration in Apoe^−/− models.^{94, 95} Nevertheless, in models of MI, Kawasaki disease (KD), myocarditis, and aortic aneurysm, GM-CSF promotes myeloid cell and DC infiltration, proliferation, and pro-inflammatory cytokine secretion, a result that aligns with GM-CSF’s pathogenic activities in atherosclerosis.^{89, 96–103} Although more work is needed to define how GM-CSF locally induces myeloid cell recruitment and instigates myocardial inflammation, cardiac fibroblasts produce GM-CSF, which stimulates cardiac macrophages to express monocyte and neutrophil chemoattractants (e.g. CCL2, CCL7, CCL12, CXCL1, CXCL3), leading to myeloid cell accumulation in KD and MI.^{89, 102} Moreover, models of MI and myocarditis show that GM-CSF secretion by cardiac fibroblasts depends upon the hallmark T helper 17 (T_H17) cytokine IL-17A.^{103, 104} Because GM-CSF also promotes pathogenic T_H17 cell differentiation by stimulating DCs to produce IL-23 and IL-6, it is possible that cardiac fibroblasts, DCs, and T_H17 cells form a collaborative loop via GM-CSF to both initiate and sustain myocardial inflammation and possibly trigger the development of heart failure.^105–107

IL-3 may locally aggravate atherogenesis by stimulating monocyte activation and adhesion, EC activation, and smooth muscle cell (SMC) proliferation in plaques,¹⁰⁸ but these observations have yet to be confirmed in vivo. Yet IL-3 does critically amplify myocardial inflammation.¹⁰⁹ Specifically, self-reactive, cardiac-infiltrating CD4⁺ T cells produce IL-3 that provokes cardiac inflammation, leukocyte accumulation, and myocardial dysfunction in autoimmune myocarditis. By locally stimulating IL-3R⁺ cardiac macrophages to express monocyte chemoattractants, IL-3 mediates Ly-6C^high monocyte recruitment and differentiation into monocyte-derived antigen-presenting cells (APCs), which subsequently amplify local T cell proliferation in cardiac tissue (Figure 1).¹⁰⁹ Ultimately, whether IL-3 directly affects atherosclerosis, MI, and other CVD complications warrants further study.

Taken together, observations on GM-CSF and IL-3 in CVD suggest that inhibiting either growth factor might dampen inflammatory myeloid cell production in peripheral lymphoid organs and prevent inflammatory leukocyte recruitment and accumulation in myocardial tissues. At this point in time, however, more evidence is needed to establish IL-3 as a decisive target in CVD treatment. In addition, caution should be exercised regarding GM-CSF blockade as a therapeutic strategy. Because GM-CSF is important for priming myeloid cell antimicrobial functions and maintaining mucosal barrier homeostasis,^110–113 blocking GM-CSF might predispose individuals to infection or inflammation at barrier sites. Therefore, depleting the relevant cellular sources of GM-CSF or defining a safe therapeutic dosage to inhibit GM-CSF may be more optimal for CVD treatment.

Granulocyte Colony-Stimulating Factor

G-CSF is produced by stromal-lineage cells such as ECs, epithelial cells, and fibroblasts as well as macrophages and other hematopoietic cells. Like GM-CSF and IL-3, G-CSF is typically induced by inflammation in response to host infection and injury. Because neutrophils express the most CSF-3R, the G-CSF receptor, its myeloid lineage regulation capacity is somewhat limited to controlling neutrophil production and function.^{27, 114} Accordingly, G-CSF primarily induces granulopoiesis by modulating the proliferation and mobilization of HSPCs, which can then give rise to mature neutrophils, and G-CSF also regulates neutrophil release from the bone marrow.^{114, 115} Thus, Csf3^−/− mice harbor fewer bone marrow granulocyte precursors and exhibit neutropenia.^{116, 117}

In the clinic, G-CSF is administered following chemotherapy for several types of cancer. G-CSF reduces the severity and duration of chemotherapy-induced neutropenia and protects against febrile neutropenia, infection-related mortality, and early mortality from all causes, with minor side-effects.^{26, 114} Although it is pathogenic in preclinical models of inflammatory/autoimmune diseases, such as arthritis, experimental autoimmune uveitis, and granulomatous disease, in which neutrophils and other myeloid cells drive harmful inflammatory responses, G-CSF is not yet a clinical target in such conditions. Additionally, G-CSF may perform potentially protective functions in various nervous system disease models.²⁷ Since G-CSF’s role in inflammation appears disease dependent, the effect(s) of G-CSF depletion or administration could be either beneficial or detrimental depending on the inflammatory condition.

G-CSF’s functional role in CVD remains in question because of inconsistencies among clinical trials and preclinical disease models. For example, due to its known ability to mobilize stem and progenitor cells, G-CSF was used to treat acute MI alongside stem cell-based regimens intended to improve cardiac regeneration and function in damaged myocardial tissue.¹¹⁸ G-CSF was considered an attractive therapeutic candidate because it had been shown to protect against cardiac remodeling, fibrosis, and cardiomyocyte apoptosis while promoting angiogenesis and re-endothelialization in experimental MI animal models.^119–122 Unfortunately, however, clinical trials utilizing G-CSF for post-MI repair were inconclusive, possibly due to differences in timing, administration route, and G-CSF dosage among the studies.¹¹⁸ At present, the mechanisms governing G-CSF’s effects after MI, including how G-CSF influences hematopoietic cell function during post-MI recovery, require further exploration.^{123, 124}

In preclinical models of atherosclerosis, conflicting reports demonstrate that G-CSF infusion either promotes or prevents plaque progression. While a pooled study analysis suggests that G-CSF likely inhibits atherogenesis by mobilizing progenitor cells,¹²⁵ whether those cells replace dysfunctional ECs and stimulate angiogenesis or give rise to neutrophils and monocytes that alter plaque evolution needs further mechanistic study.^125–128 Ultimately, differing animal models, dosages, administration methods, and lesion measurement time points may have caused inconsistent results among the atherosclerosis G-CSF studies.¹²⁵ Given its modest and inconsistent clinical effects, G-CSF infusion is unlikely to become a treatment strategy for cardiovascular inflammation. Additionally, since G-CSF’s function is potentially protective in CVD, blocking or altering its production or availability may not yield any substantial benefit.

The Epidermal Growth Factor Receptor Axis

EGFR is a cell surface receptor with downstream RTK activity that guides tissue development and maintains tissue homeostasis, principally in epithelial lineages. In response to seven EGF family ligands, which have distinct tissue-specific expression patterns, EGFR expands and sustains pools of undifferentiated, stem cell-derived progenitors by controlling their proliferation, migration, and survival.¹²⁹ Due to its pathological role in the outgrowth of several carcinomas and brain tumors, EGFR has become a critical therapeutic target in cancer. Typically, oncogenic mutations, in-frame deletions, and genetic rearrangements in the EGFR locus trigger EGFR over-expression or over-activation.¹³⁰ Consequently, uncontrolled EGFR signaling can promote tumor angiogenesis, proliferation, and metastasis, as well as inhibit apoptosis in malignant cells. Multiple blocking antibodies and small molecule inhibitors have been developed and safely utilized to counteract EGFR’s pro-tumorigenic activity.^{28, 129, 130}

While EGFR signaling appears to be important for cardiac development,¹³¹ EGFR and EGF family members have gained more attention as detrimental players in vascular disorders. Within the vascular system, EGFR and its ligands are expressed by SMCs and ECs and can regulate their proliferation, migration, survival, and production of angiogenic factors and ROS (Figure 2).^131–133 Most work on the EGF:EGFR axis in CVD has therefore focused on the role of EGFR stimulation or transactivation in non-immune cells. Although EGFR and EGFs are expressed in atherosclerotic plaques, more progress has been made in understanding how EGFR and EGF signaling promotes abdominal aortic aneurysm formation, cardiac remodeling, endothelial dysfunction, fibrosis, hypertension, and neointimal hyperplasia.^131–133 In vascular disease models, mAb therapies and small molecule inhibitors have successfully antagonized the pathologic vascular activities of SMCs and ECs.¹³³

Figure 2.

Pathologic effects of epidermal growth factor receptor in cardiovascular disease. Most work surrounding the pathologic effects of EGFR on vascular inflammation has focused on smooth muscle cell proliferation, migration, and survival, and endothelial cell activation and dysfunction, though more recent data has suggested EGFR also affects cardiomyocytes and fibroblasts. Regarding the immune effects of EGFR in CVD: (a) EGFR promotes macrophage accumulation, activation, foam cell formation, and macrophage production of inflammatory mediators in atherosclerotic vessels. Though EGF has been reported to mediate monocyte adhesion and chemotaxis as well as macrophage proliferation in vitro, it remains unknown whether EGF:EGFR interactions (including which EGFs) drive monocyte recruitment, differentiation, and macrophage proliferation in vivo. (b, c) EGFR promotes CD4⁺ T cell proliferation and cytokine production as well as T cell accumulation in peripheral lymphoid organs and atherosclerotic plaques. It is unknown whether EGFR signaling affects CD4⁺ T helper cell differentiation in atherosclerosis, drives T cell recruitment into atherosclerotic plaques, or affects local T cell proliferation and cytokine production in plaques.

Recent work on EGFR expression in certain leukocyte subsets, including cardiac macrophages and activated T cells, has indicated that blocking the receptor may be a worthwhile strategy since EGFR may drive detrimental leukocyte activity in CVD.^134–137 Mechanistic approaches suggest that EFGR signaling in macrophages provokes plaque progression in animal models of atherosclerosis (Figure 2). Ldlr^−/− mice reconstituted in vitro with LysM^Cre/+ Egfr^fl/fl bone marrow showed that myeloid-specific EGFR deficiency decreases macrophage accumulation in plaques, reduces macrophage oxLDL uptake by lowering CD36, and limits macrophages’ production of TNF-α and IL-6.¹³⁶ Similarly, inhibiting EGFR signaling in mice abates macrophage accumulation, reduces expression of pro-inflammatory mediators and adhesion molecules, and attenuates development of oxidative stress in aortic plaques, supporting the idea of targeting EGFR in atherosclerosis.¹³⁸ Moreover, oxLDL may promote foam cell formation in part by inducing EGFR signaling downstream of TLR4 activation.^{138, 139} EGFR may also control CD4⁺ T cell function and, by extension, promote atherosclerotic plaque development (Figure 2).¹³⁷ More specifically, EGFR co-localizes with CD4⁺ T cells in atherosclerotic plaques and modulates these cells’ capacity to undergo activation, proliferate, and produce cytokines, including IFN-γ. Importantly, blocking EGFR signaling with an EGFR-specific mAb or conditionally deleting EGFR expression in CD4⁺ T cells limited atherosclerosis by suppressing effector CD4⁺ T cell activation, infiltration, proliferation, and cytokine production.¹³⁷

Despite these recent discoveries, many questions regarding how EGFR and EGF ligands affect immunity and inflammation in CVD remain unanswered. Although monocytes express EGFR and EGF purportedly acts as a monocyte chemoattractant, it is unknown whether EGFR signaling guides monocytes into atherosclerotic plaques or affects their phenotype and function in plaques in vivo.¹³⁴ While there is evidence that amphiregulin, an EGFR ligand, secreted by cardiac macrophages enhances cardiac fibrosis and myocardial dysfunction in experimental MI, it is unclear if EGFR signaling affects macrophage functions during MI.¹⁴⁰ Additionally, whether EGFR and its ligands are produced by or signal in other cells in CVD remains unexplored. Overall, emerging work suggests that blocking EGFR might curtail adaptive and innate immune-driven cardiovascular inflammation and foam cell formation in atherosclerosis. How EGFR signaling affects other forms of CVD awaits exploration.

Fibroblast Growth Factor 21

Counteracting the metabolic syndrome is an urgent clinical need because its components (i.e. central obesity, insulin resistance, elevated fasting glucose, dyslipidemia, and systemic hypertension) are major risk factors for CVD, type 2 diabetes (T2D), and all-cause mortality.^{141, 142} In recent years, studies have shown that the FGF endocrine hormone FGF21 has beneficial effects on metabolism, thus making it an attractive therapeutic target to offset cardiometabolic risk. FGF21 is broadly expressed in many tissues, including the liver, adipose tissue, and pancreas, and acts on multiple cell types in a paracrine or endocrine manner by signaling via several FGF receptors (FGFRs) and its co-receptor β-klotho (KLB).¹⁴³ In diabetes and obesity models, FGF21 improves insulin sensitivity, lowers fasting glucose and triglycerides, induces weight loss, and promotes both energy expenditure and white adipose tissue browning by uncoupling protein (UCP)1-dependent and -independent mechanisms (Figure 3).^144–149 Moreover, human studies and early clinical trials in obese patients with T2D indicate that FGF21 and its analogs protect against obesity-related disorders through similar mechanisms, potentially by regulating brown adipose tissue activity and energy expenditure.^{29, 150, 151} Based on FGF21’s protective metabolic effects in humans and animal models, multiple FGF21 analogs and FGFR agonists are currently under development and used for treating obesity-related diseases.^{29, 143}

Figure 3.

Protective effects of fibroblast growth factor 21 in cardiovascular disease. Most work on the protective effects of FGF21 in CVD has focused on FGF21 signaling primarily in tissues, such as the liver, kidney, adipose tissue, and heart (i.e. effects on cardiomyocytes). The induction of angiotensin converting enzyme 2 (ACE2) in adipocytes and renal cells by FGF21 attenuates pathologic vascular remodeling, fibrosis, and inflammation in hypertension. The induction of adiponectin in adipocytes and the liver by FGF21 limits vascular inflammation in atherosclerosis and cardiac dysfunction in MI and suppresses smooth muscle cell proliferation and macrophage oxLDL uptake directly. In atherosclerosis, FGF21 directly inhibits cholesterol biosynthesis by signaling on hepatocytes. The immune effects of FGF21 have primarily been studied in vitro (see box). FGF21 suppresses macrophage pro-inflammatory cytokine production and foam cell formation by inducing the nuclear factor erythroid 2-related factor 2 (Nrf2) pathway and upregulation of cholesterol efflux machinery (e.g. ABCA1, ABCG1) in vitro. It is unknown whether FGF21 impacts myeloid cell and other leukocyte functions in cardiovascular disease in vivo.

Since FGF21’s identification as a protective hormone against cardiometabolic risk factors and a potential biomarker for atherosclerosis, multiple studies have focused on evaluating the FGF21’s effects in CVD models.¹⁵² FGF21 appears to suppress atherosclerotic plaque accumulation by reducing hypercholesterolemia, oxidative stress, and SMC proliferation via adiponectin-dependent and -independent mechanisms.^{152, 153} In addition, FGF21 directly inhibits cardiomyocyte apoptosis, attenuates pathological cardiac remodeling, and protects against cardiac hypertrophy, thus preventing myocardial dysfunction and injury in ischemic heart tissue and diabetic cardiomyopathy (Figure 3).^154–157 More recently, FGF21 was shown to inhibit hypertension and vascular inflammation by inducing angiotensin-converting enzyme 2 in adipocytes and renal cells.¹⁵⁸ However, FGF21’s impact on leukocytes in cardiovascular organs and metabolically relevant sites, such as the liver and adipose tissue, in the setting of CVD is largely unexplored. Since it prevents macrophage accumulation, inflammation, and fibrosis in atherosclerosis and hypertension models, FGF21 may prove to regulate monocyte and macrophage recruitment, proliferation, and inflammatory functions in vessels and myocardial tissues.^{153, 158} Indeed, preliminary in vitro evidence suggests that FGF21 regulates cholesterol efflux, oxLDL uptake, and foam cell formation in THP1 macrophages and inhibits macrophages’ inflammatory capacity through the Nrf2 pathway (Figure 3).^159–161 Moreover, recent data from diet-induced obesity and adaptive thermogenesis models indicate that FGF21 promotes anti-inflammatory macrophage polarization in adipose depots and white adipose tissue browning, which could effectively prevent adipose tissue from adopting pro-inflammatory profiles.^162–164 In context of cardiovascular health and disease, agonizing FGF21 pathways might be advantageous because the adipose tissue is an active endocrine organ that can favor cardiovascular disease progression by inducing chronic, low-grade inflammation via hormones and pro-inflammatory mediators (i.e. “adipokines”).^{142, 165} Likewise, leukocytes that infiltrate adipose tissue in obesity can further release an inflammatory milieu that impacts CVD and metabolic disease outcomes.¹⁶⁵ In sum, FGF21 may act directly on macrophages or other cell types in vivo and thereby suppress inflammation by altering macrophage polarization states, potentially at sites local and distal to cardiovascular organs. Although FGF21’s functional role on leukocytes in obesity and CVD remains unclear, preliminary evidence suggests administering either FGF21 or agonists that stimulate the FGF21 signaling pathway may help combat obesity-associated inflammation and thus offset cardiometabolic risk factors that promote cardiovascular inflammation and its downstream complications, including ischemic heart disease and stroke.

Conclusions

As the availability of immunotherapies and other “precision medicine” biologics continues to grow, physicians and scientists alike should evaluate novel targeted therapies in experimental models of CVD to provide a proof-of-concept for eventual adoption in cardiovascular medicine. Mounting evidence from preclinical models of atherosclerosis, myocardial infarction, and other cardiovascular disorders calls for a closer look at immunomodulatory growth factors, including the CSFs, EGF ligands, and FGF21, as potential immunotherapy targets for the treatment of CVD. Inhibiting the signaling of hematopoietic growth factors, such as M-CSF, GM-CSF, and IL-3, and the EGFR blunts the expansion and inflammatory functions of innate and adaptive immune cells in cardiovascular organs and at sites distal to the heart and coronary vasculature, such as the bone marrow and spleen. On the other hand, agonizing the FGF21 pathway induces a favorable cardiometabolic profile, which may protect against the development of CVD. Utilizing clinically available growth factor-targeted therapies to block or alter the bioavailability of candidate growth factors may ultimately help prevent adverse cardiac events in people with CVD risk factors and aid individuals who have already suffered from the complications of cardiovascular inflammation.

Highlights section.

The CANTOS trial provides rationale for designing focused immunotherapeutic strategies that target immunomodulatory mediators in cardiovascular disease.

Growth factors are immunomodulatory agents that control cellular functions, including cell proliferation, differentiation, survival, and mobilization, among others.

Growth factors have been successfully targeted in the clinic for a variety of inflammatory/autoimmune diseases, cancer, or metabolic disorders, yet they have not been adopted for cardiovascular disease treatment.

This review highlights recent and newly emerging immmunomodulatory effects of growth factors in cardiovascular inflammation, particularly for the colony-stimulating factors (M-CSF, GM-CSF, G-CSF, IL-3), the epidermal growth factor axis, and fibroblast growth factor 21.

Nonstandard Abbreviations

CSF-1R

colony stimulating factor-1 receptor

CVD

cardiovascular disease

DC

dendritic cell

EC

endothelial cell

EGF

epidermal growth factor

EGFR

epidermal growth factor receptor

FGF

fibroblast growth factor

HSPC

hematopoietic stem and progenitor cell

IRA

innate response activator

MI

myocardial infarction

oxLDL

oxidized low-density lipoprotein

SMC

smooth muscle cell