Chronic vascular inflammation, oxidative stress, and endothelial dysfunction are hallmarks of the development and progression of atherosclerotic lesions in coronary arteries resulting in CAD [43, 44, 45]. The imbalance of reactive oxygen species (ROS) and antioxidant defenses is one of the main causes of endothelial dysfunction [43]. Increased NADPH oxidase (Nox) activity uncouples endothelial NO-Synthase (eNOS), increases ROS, and decreases nitric oxide (NO) bioavailability [46]. NO is a strong vasodilator that also inhibits the expression of transcription factors such as NF-$\kappa$B and adhesion molecules, e.g., intercellular adhesion molecule-1 (ICAM-1) and vascular cell adhesion molecule-1 (VCAM-1) [47]. The relationship between inflammation and oxidative stress in early-stage of human atherosclerosis leads to a cyclic worsening of the condition, as inflammatory processes that attempt to repair oxidative damage increase oxidative stress, which in turn leads to endothelial dysfunction. The cytokine GDF-15 has been shown to enlarge atherosclerotic plaques, increase plaque vulnerability, impair ECs in plaques, and induce endothelial-to-mesenchymal transition (EndMT) [48, 49, 50, 51] (Fig. 1; Table 2, Ref. [52, 53, 54, 55, 56, 57, 58, 59, 60]).

Studies of vascular remodeling in pulmonary arterial hypertension (PAH), which is characterized by endothelial dysfunction with release of vasoactive mediators, growth factors, and cytokines [61], show that GDF-15 is increased in PAH lungs, predominantly located in vascular ECs [52]. PAH is characterized by pulmonary vascular remodeling, progressive arterial stiffening, increased vascular resistance, and right ventricular failure. Animal and human studies suggest with growing evidence that ROS and oxidative stress play a key role in the pathogenesis of PAH [61, 62]. In vitro analyses of human pulmonary microvascular endothelial cell (HPMEC) proliferation and apoptosis suggest a role for GDF-15 in endothelial cell homeostasis in PAH patients [52]. HPMEC showed marked upregulation of GDF-15 in hypoxia and laminar shear stress [52]. Recombinant (r) GDF-15 protein decreased apoptotic cell death of HPMEC. In contrast, proliferation was either increased or decreased depending on the concentration of rGDF-15 protein [52] (Table 2). Further studies showed that GDF-15 stimulated the proliferation of human umbilical veins endothelial cells (HUVECs) by upregulating cyclins D1 and E via the phosphoinositide 3-OH kinase (PI3K)/ protein kinase B (AKT) signaling pathway, extracellular signal-regulated kinases (ERK), and c-Jun N-terminal kinase (JNK)-dependent AP-1 and E2F activation signaling pathways [53, 54] (Fig. 2). The effect of GDF-15 against apoptotic cell death might be related to influence on PI3K/AKT/eNOS pathway and NF-$\kappa$B/JNK pathway. This was demonstrated by in vitro studies in HUVECs, where GDF-15 protected against apoptosis, which was induced by high glucose concentration via maintenance of the PI3K/AKT/eNOS pathway and attenuation of the NF-$\kappa$B/JNK pathway [55]. Clinical studies have shown that GDF-15 plasma levels correlate with the levels of other CV risk biomarkers such as cTnT, NT-proBNP, CRP, possibly indicating a relationship between GDF-15, inflammatory processes, and oxidative stress [30]. Kim et al. [56] also demonstrated a molecular relationship between CRP and GDF-15, reporting that GDF-15 expression was increased by CRP via the binding p53 to its promoter region in human aortic endothelial cells (HAECs). Thus, GDF-15 is a direct target gene of p53 through the mediation of CRP [56] (Fig. 2). These data support the predictive role of GDF-15 concerning CVD.

Fig. 2.

Downstream targeting and signaling of GDF-15 in ECs in experimental stress-induced models.

Cell senescence is a mechanism of aging and plays a vital role in the onset and prognosis of CVD [63]. Increasing evidence shows that cell senescence is indispensable in the formation and development of atherosclerosis [63]. To investigate GDF-15 expression, function and role during cellular senescence, Ha et al. [57] studied endothelial colony-forming cells (AB-ECFCs) as a model for ECs, because cell senescence is mainly involved in vascular stress and loss of endothelial function. They found that AB-ECFCs expressed higher levels of GDF-15 compared with cord blood colony-forming cells (CB-ECFCs) and that GDF-15 expression progressively increased as AB-ECFCs senescent [57]. Previous studies showed that GDF-15 was overexpressed in radiation-induced senescent HAECs [64]. The paracrine action of GDF-15 promotes AB-ECFC proliferation, migration, and NO production through activation of AKT, ERK, and Mothers against decapentaplegic homolog 2 (SMAD2) signaling pathways. It induces ROS production independently of nuclear factor-like 2 (NERF2), the major transcription factor regulating antioxidant response [57] (Fig. 2). Ha et al. [57] interpreted the paracrine effect of GDF-15 by senescent AB-ECFCs on non-senescent AB-ECFCs as a benefit and claimed that GDF-15 might play a beneficial role in a dysfunctional vasculature by limiting endothelial dysfunction associated with vascular stress.

An increase in endothelial permeability and micro vascularization in the plaque are critical factors in the atherogenesis. Regarding the angiogenic process, Whitson et al. [58] described that GDF-15 interacts with connective tissue growth factor 2 (CCN2), inhibits CCN2-mediated angiogenesis, and blocks CCN2-mediated tube formation in HUVECs. However, in hypoxic HUVECs Song et al. [59] described that GDF-15 promotes angiogenesis via the hypoxia-inducible factor 1-alpha (HIF-1$\alpha$)/VEGF-dependent signaling pathway. Furthermore, GDF-15 has been reported to increase the expression of VEGF in a time- and dose-dependent manner, stimulating proliferation and thereby promoting the vascular development of HUVECs [53, 54, 59]. Again, confirming GDF-15 enhances proliferation, migration, and NO production in various endothelial cell types, it also plays an essential role in angiogenesis [53, 54, 59, 60] (Fig. 2).

An induction of GDF-15 has been reported and described in numerous diseases, such as CVD, cancers, metabolic disorders, rheumatic diseases and viral infection [49, 65, 66, 67, 68]. The majority of these diseases are associated with inflammation and cellular stress.

TGF-ß family members, including GDF-15, have effects on cell proliferation, differentiation, apoptosis and inflammation as well as cellular motility and adhesion [69, 70]. The expression of GDF-15 was examined in the human monocytic cell lines U937, KG-1 and THP-1 [1, 71], whereby under (oxidative) stress conditions, such as incubation with trans-retinoic acid (RA) and phorbol 12 myristate 13-acetate (PMA), oxidized low-density lipoprotein (oxLDL), C6-ceramide, or H${}_2$O${}_2$ the GDF-15 transcript expression was upregulated in human myelomonocytic cell lines and cultured human macrophages (PBMCs) [1, 13] (Table 3, Ref. [1, 13, 48, 49, 71, 72, 73, 74, 75, 76, 77]). GDF-15, also named NAG-1 and MIC-1, is induced by several anti-inflammatory drugs [78]. Expression of GDF-15 is induced by cytokines involved in macrophages activation like interleukin (IL)-1ß, tumor necrotic factor (TNF)-$\alpha$, macrophage colony-stimulating factor (M-CSF) and IL-2 (Fig. 3), by nonsteroidal anti-inflammatory drugs (NSAIDs) and by the antidiabetic and anti-inflammatory drug troglitazone [6, 79, 80]. Whereas, interferon (IFN) $\gamma$ and lipopolysaccharide (LPS) have no effects on GDF-15 expression in U937 and KG-1 [1]. Another study about bacterial and viral infections, as well as sepsis, has described an increased transcription of GDF-15 in an early (1 h–3 h) response to LPS stimulation in CD11b${}^{+}$CD45${}^{+}$ myeloid cells in the liver and bone marrow-derived macrophages from mice [72] (Table 3). Interestingly, the autocrine/paracrine effect of GDF-15 suppresses the LPS-induced TNF-$\alpha$ release in U937 and KG-1 [1]. Therefore, GDF-15 limits LPS-stimulated macrophage activation and inflammation [1] (Fig. 3). Additionally, in serum of LPS-stimulated hNAG-1/GDF-15 transgenic mice, Kim et al. [73] have described a decreased level of pro-inflammatory cytokines. Moreover, GDF-15 reduces the increase of IL‑6, TNF‑$\alpha$, and IL‑1$\beta$ expression in serum and liver tissue, and inhibits the activation of the I$\kappa$B$\alpha$/NF‑$\kappa$B pathway by disrupting TGF-$\beta$-activated kinase 1 (TAK1) phosphorylation in Kupffer cells [74] (Table 3). Furthermore, GDF-15 prevents LPS/D-galactosamine (D-GalN)-induced cell death, increases inflammatory cell infiltration and serum alanine aminotransferase (ALAT) and aspartate aminotransferase (ASAT) activities in liver tissue [74]. Additionally, GDF-15 has been shown to coordinate tolerance to inflammatory damage through regulating triglyceride metabolism [73]. Kim et al. [73] have also described that GDF-15 does not directly inhibit the toll-like receptor (TLR)4/NF-kB pathway in RAW 264.7 and in NAG-1${}^{\mathrm{\text{Tg/Lox}}}$ peritoneal macrophages, as well as does not affect inflammatory cytokine production from mice Kupffer cells ex vivo. But they found that NAG-1${}^{\mathrm{\text{Tg/Lox}}}$ mice have less white adipose tissue and lower basal leptin level. Leptin is a hormone produced predominantly by adipose cells as a pro-inflammatory cytokine, that appears to be a pivotal mediator of inflammation in mice [81]. Apart from leptin, the adipokine adiponectin was shown to affect atherosclerosis, inflammation and oxidative stress pathways [82]. So, Kim et al. [73] have found a link between white adipose tissue mass and inflammation and suggest that GDF-15 may play an anti-inflammatory role in response to LPS stimulation in interaction with leptin (Table 3).

Fig. 3.

Downstream targeting and signaling of GDF-15 in monocytes/macrophages in experimental stress-induced models.

In relation to atherosclerotic research studies, immunohistochemical analyses of human atherosclerotic carotid arteries have demonstrated colocalization of GDF-15 with oxLDL, CD68 immunoreactive cells and apoptosis-relevant proteins [13, 75]. In this context, GDF-15 was up-regulated in murine atherosclerotic lesions during disease progression in a pattern similar to CD68${}^{+}$ macrophages [75]. Also, research studies on animals and human clinical trials have shown that oxLDL and induction of apoptosis correlate with an increased GDF-15 protein level and mRNA expression in human macrophages [13, 75] (Fig. 3; Table 3). Therefore, GDF-15 may contribute to oxidative stress-dependent modulation of pro-inflammatory processes in atherosclerotic lesions. Studies with RAW 264.7 macrophages have shown that rGDF-15 promotes the S/G2-phase arrest in a TGF-ßRII-dependent manner (Fig. 3; Table 3) and does not induce apoptosis [75]. Moreover, GDF-15–deficient macrophages appear less prone to oxLDL–induced apoptosis and necrosis [75]. GDF-15 deficiency results in a long-term reduction of atherosclerotic lesions by decreased occurrence of inflammatory CD11b${}^{+}$ or IL6${}^{+}$ leukocytes as well as an elevated percentage of macrophages, enhanced cell density and decreased p62-accumulation in atherosclerotic lesions of the brachiocephalic trunk in mice [48, 49]. Moreover, it has most recently been shown that, GDF-15 deficiency affects the morphology of atherosclerotic plaques in vessels with deoxygenated blood and low blood pressure, such as the pulmonary trunk (PT), to show a trend decrease of 6.7% in lumen stenoses in the PT of hypercholesterolemic GDF-15${}^{\mathrm{-}\mathit{}\mathrm{/}\mathrm{-}}$/ApoE${}^{\mathrm{-}\mathit{}\mathrm{/}\mathrm{-}}$ compared with ApoE${}^{\mathrm{-}\mathit{}\mathrm{/}\mathrm{-}}$ mice [50]. Additionally, a significant reduction of the necrotic area in the plaque of GDF-15-deficient mice, with concomitant increases in CD68${}^{+}$, $\alpha$-actin${}^{+}$, and TUNEL${}^{+}$ cells in the plaque of the PT, was demonstrated [50]. Therefore, GDF-15 is thought to be involved in development and progression of atherosclerotic lesions in the brachiocephalic trunk, but also in the PT, likely targeting different mechanisms (e.g., in apoptosis).

In this context, in vitro data have proved that GDF-15 deficiency leads to a decreased mRNA expression of apoptosis- or inflammation-relevant cytokines in cultured peritoneal macrophage of mice [49]. Additionally, data from human PMA-differentiated THP-1 macrophages have suggested that GDF-15 is involved in the regulation of lipid homeostasis by regulating autophagic processes [48, 71](Fig. 3; Table 3). Studies using small interfering RNA against GDF-15 (siGDF-15) and recombinant GDF-15 have demonstrated that GDF-15 directly affects autophagic activity in macrophages without affecting lysosomal activity [48, 71]. Also, in combination with oxLDL, GDF-15 affects autophagic processes with consequences for lipid homeostasis in human macrophages [71] (Fig. 3; Table 3), indicating its emerging important pathophysiological role in the development and progression of atherosclerotic plaques. In the context of foam cell formation, another study using THP-1 macrophages has demonstrated that GDF-15 might be a potential target to prevent foam cell formation via the PI3K/PKC/SP1 pathway and promote cholesterol efflux [76] (Fig. 3; Table 3). Therefore, GDF-15 has been shown to regulate apoptosis, autophagy and inflammatory processes of macrophages and is involved in configuring atherosclerotic lesion development.

In terms of the selectin-mediated leukocyte capturing and rolling, followed by the actual transmigration through the endothelium, resulting in chemokine-induced leukocyte arrest [83, 84], GDF-15 is essential to prevent the excessive chemokine-activated leukocyte arrest and transmigration through the endothelium [77]. Additionally, GDF-15 is an inhibitor of leukocyte ß${}_2$-integrin activation via Cdc42 and Rap1 [77] (Fig. 3; Table 3). The interaction of activated $\beta$${}_2$-integrins with ICAM-1 leads to leukocyte arrest on the endothelium and initiates trans-endothelial migration [85]. de Jager et al. [75] have concluded that a reduction of macrophage accumulation in plaques of GDF-15${}^{\mathrm{-}\mathit{}\mathrm{/}\mathrm{-}}$ chimeras mice results in an impaired cellular migration and mobility, possibly via C-C chemokine receptor type 2 (CCR2) [75] (Fig. 3; Table 3). CCR2 is a key chemokine receptor for monocyte recruitment at early stage of atherosclerosis and is decreased by GDF-15–deficient macrophages [75]. The direct interaction of GDF-15 with the chemokine receptor CCR2 function suggests that the GDF-15-induced macrophage mobility modulates the CCR2 response [75].

In the context of CAD, smooth muscle cells (SMCs) play a key role in the stability and progression of atherosclerotic plaques. In addition to macrophages and ECs, as cellular sources of GDF-15 production, VSMCs also secrete GDF-15 in response to metabolic and/or oxidative stress or stimulation by pro-inflammatory cytokines [28]. In studies concerning atherosclerotic plaques of the pulmonary trunk of GDF-15 deficient ApoE${}^{\mathrm{-}\mathit{}\mathrm{/}\mathrm{-}}$ mice after 20 weeks cholesterol-enriched diet, Bonaterra et al. [50] found an increase in the percentage of $\alpha$-actin${}^{+}$ SMCs with a higher percentage of CD68${}^{+}$ macrophages and a decreased necrotic core area compared to ApoE${}^{\mathrm{-}\mathit{}\mathrm{/}\mathrm{-}}$ mice.

After a high-fat meal, with elevated postprandial lipemia, a strong upregulation of GDF-15 expression in coronary artery SMCs (CASMCs) by triglyceride-rich lipoproteins (TRL) was observed [86]. The group of TRLs is composed of chylomicrons and very low-density lipoproteins (VLDLs). TRLs and their metabolites are involved in the pathogenesis of atherosclerosis by modulating inflammation, oxidative stress, and foam cell formation [87], as well as inducing cell proliferation [88] and monocyte chemoattractant protein-1 (MCP-1) expression in SMCs [89]. Likewise, TRLs and their metabolites have been detected in atherosclerotic plaques [90].

Hence, more research projects are necessary to understand the direct effects of GDF-15 on VSMCs in the context of atherosclerotic plaque development, progression, and stability.

Myocardial infarction, a condition associated with CAD, is associated with many deaths [91] and shows upregulation of GDF-15 after acute myocardial infarction [92]. GDF-15 is not constitutively expressed in adult myocardium. Cardiomyocytes produce and secrete GDF-15 only in response to oxidative stress, angiotensin II or inflammatory cytokines, ischemia, and mechanical stretch [28]. Increased plasma levels of GDF-15 can be detected in patients suffering from myocardial infarction, or as a result of injury and heart failure [93, 94]. Among other findings, data from the Women’s Health Study show that serum GDF-15 levels are an independent risk indicator for adverse CV events [93]. In this regard, GDF-15 has cardioprotective effects on cardiomyocytes in ischemic tissue and controls the conversion of cardiac fibroblasts to myofibroblasts during the development of fibrosis [92, 95]. Thus, understanding cellular GDF-15 signaling and crosstalk in cardiac metabolism is a research concern.

In vitro experiments with immune cells, ECs, and cardiomyocytes from the ventricle suggest that GDF-15 may act as a survival factor on one side and as an inducer of cell death factors on the other [49, 55, 95, 96, 97, 98, 99, 100]. Using GDF-15 gene-targeted mice, endogenous GDF-15 was shown to protect the heart from ischemic/reperfused (I/R) injury [92]. Similarly, cell culture experiments with recombinant GDF-15 showed that cardiomyocytes are protected from hypoxia-induced ischemic injury via PI3K and AKT-dependent signaling pathways [92]. GDF-15 promotes rapid activation by transient Ser473 phosphorylation of AKT in cardiomyocytes, which is accompanied by an increase Ser136 phosphorylation (inactivation) of the AKT downstream target Bcl-2 antagonist of cell death (Bad) [92] (Fig. 4; Table 4, Ref. [92, 95, 99, 101]), a pro-apoptotic protein of the Bcl-2 family [102]. In addition, other PI3K/AKT-independent pathways may be involved in the autocrine/paracrine effects of GDF-15 [95, 99], with GDF-15 transiently activating ERK1/2 in cardiomyocytes [92, 99], but not p38 or JNK [92] (Table 4).

Fig. 4.

Downstream targeting and signaling of GDF-15 in experimental stress-induced models that stimulate GDF-15 expression in cardiomyocytes and reveal GDF-15 as a cardioprotective via PI3K-AKT, ERK, and SMAD proteins.

Pathological myocardial hypertrophy leads to increased oxygen demand and decreased contractility of the affected ventricle [103]. This usually results in heart failure, as well as an increased risk of myocardial infarction [104, 105]. The hypertrophic signaling effect mediated by GDF-15 via the epidermal growth factor receptor (EGFR), PI3K, AKT, ERK, as well as SMAD proteins is controversial in this regard [95, 99, 101] (Fig. 4; Table 4). Analyses have shown that GDF-15 attenuates norepinephrine (NE)-induced myocardial hypertrophy as well as hypertrophy in cultured rat neonatal ventricular cardiomyocytes through induction of small body size (SMA) and SMAD2/3 phosphorylation and detectable induction of SMAD1/5/8 phosphorylation [95, 101] (Fig. 4; Table 4). NE is known to induce oxidative stress resulting in hypertrophy, apoptosis, and intracellular Ca${}^{2+}$ overload in the myocardium [106]. Moreover, in vivo and in vitro studies show that NE can stimulate the synthesis and release of GDF-15 [101]. Thus, GDF-15 negatively regulates NE-induced myocardial hypertrophy, activation of EGFR, and its signaling pathway [101]. Contrary to the findings of Xu et al. [95, 101], GDF-15 triggers hypertrophic growth in rat ventricular cardiomyocytes [99]. Heger et al. [99] investigated the different R-SMAD isoforms and found that GDF-15 does not stimulate R-SMAD2 but enhances the phosphorylation of R-SMAD1 (Fig. 4; Table 4). SMAD1 mediates bone morphogenetic protein (BMP) signaling, which is involved in various biological activities including cell growth, apoptosis, development, and immune responses [107]. Furthermore, activation and cardiac-specific overexpression of R-SMAD1 results in smaller myocardial infarct area and reduced apoptotic cell death in cardiomyocytes [108]. Heger et al. [99] suggests that GDF-15 gains its anti-apoptotic and pro-hypertrophic character through stimulation of R-SMAD1.