Emerging Roles of Growth Differentiation Factor 15 in Immunoregulation and Pathogenesis

Growth differentiation factor 15 [GDF-15; also known by other names such as macrophage inhibitory cytokine 1 (1), nonsteroidal anti-inflammatory drug (NSAID)–activated gene-1 (2), or PLAB (3)] is an emerging biomarker for the severity of diverse conditions, such as cardiovascular diseases (4, 5), mitochondrial disorders (6), cancers (7, 8), sepsis (9), and more recently, coronavirus disease 2019 (COVID-19) infection (10). Late-stage cancer patients experience anorexia and cachexia that is linked to high circulating levels of GDF-15 (7). Shimada and Mitchison (11), using machine learning of aggregated toxicology studies, identified GDF-15 as a putative central mediator of tissue injury induced by a large array of xenobiotics, especially in the kidney and liver. Wang et al. (12), using a murine model of pediatric cardiomyopathy, observed that distressed cardiomyocytes produce GDF-15, which inhibits insulin-like growth factor signaling in the liver to limit body growth. During a parasitic infection, high systemic and tissue GDF-15 levels correlated with decreased disease tolerance, independent of parasite burden (13). And more recently, GDF-15 has been proposed as an indicator of COVID-19 severity and fatality (14, 15). These studies, altogether, highlight the broad significance of GDF-15 in disease pathogenesis.

GDF-15 is a distant member of the TGF-β superfamily, with central and peripheral functions. The central actions of GDF-15 involve regulating energy metabolism in nonhomeostatic conditions; i.e., it regulates energy intake and expenditure amid highly stressful conditions, such as nutrient deprivation (16), infectious and noninfectious diseases (17, 18), and strenuous activities (19). Circulating GDF-15 accesses the hindbrain, where its only known receptor, the GDNF Family Receptor α Like (GFRAL) receptor, is located. GFRAL expression is restricted to the neurons of the area postrema (AP) and nucleus solitary tract (NTS) (20–22). GDF-15 engagement of the GFRAL receptor and its coreceptor, Rearranged during Transfection receptor, mediates its central actions (21–24). In contrast, GDF-15 is also able to function in peripheral tissues and cells, such as adipose tissues (25) and immune cells (26) (see (Fig. 1). However, because GFRAL expression is restricted to the AP/NTS neurons, the GDF-15 receptor in peripheral tissues is still unknown. Nevertheless, an emerging literature suggests that GDF-15 plays an immunoregulatory role during infection and inflammation. In this review, we will explore the immunoregulatory roles of GDF-15 in the host immune response. We will also discuss what is currently known about how GDF-15 is regulated and propose that the immune response may be a pivotal player in this process.

GDF-15 regulates whole-body energy balance in nonhomeostatic conditions (see (Fig. 1). Chrysovergis et al. (27) demonstrated that GDF-15 overexpression in mice given a high-fat diet improves glucose tolerance and insulin sensitivity, increases lipolysis, and promotes thermogenesis in brown adipose tissues. Chung et al. (28) observed that mice with a muscle-specific defect in the oxidative phosphorylation system have high levels of GDF-15, which confers resistance to high-fat diet–induced weight gain by increasing glucose tolerance, insulin sensitivity, lipolysis, and energy expenditure. Interestingly, Aguilar-Recarte et al. (29) demonstrated that activation of PPARβ/δ, a well-known regulator of glucose and lipid metabolism, induces GDF-15, which mediates the beneficial metabolic effects of PPARβ/δ. Altogether, these studies demonstrate that GDF-15 exerts a positive effect on the metabolism of hosts with disorders such as obesity by improving glucose metabolism and promoting a more negative energy balance.

Anorexia worsens the morbidity and mortality of patients suffering from chronic diseases (30). GDF-15 is an anorectic agent because it can suppress appetite. Coll et al. (23) demonstrated that metformin, an antidiabetic drug, causes weight loss through a GDF-15–dependent decrease in food intake. The same group also observed that nutritional deficiency in mice leads to endoplasmic reticulum stress-induced GDF-15 upregulation, which causes taste aversion (16). GDF-15–mediated taste aversion occurs as a result of GFRAL-expressing NTS neurons exciting the calcitonin gene–related peptide–producing parabrachial neurons, which are also known contributors to aversive behaviors (31). Furthermore, GDF-15 is also known to cause nausea and emesis, especially in the case of pregnant individuals (32, 33). Thus, two ways that GDF-15 causes weight loss is to (1) diminish food intake by triggering an aversive response and (2) induce nausea. In totality, these studies describe the role of GDF-15 in regulating energy balance during nonhomeostatic conditions. Next, we will discuss the known roles of GDF-15 in the host immune response.

Numerous studies have used mouse sepsis models to understand the role of GDF-15 during inflammation. Sepsis is a life-threatening condition involving the dysregulation of the host immune response, which can lead to multiorgan failure. Cytokines play an important role in the pathology of sepsis, where the balance between the levels of proinflammatory and anti-inflammatory cytokines can determine host survival (34). TNF-α (35) and IL-1 (36) are proinflammatory cytokines that activate immune cells and can also cause cell death in tissues, leading to multiorgan dysfunction. In contrast, IL-4 (37) and IL-10 (38) are anti-inflammatory cytokines that antagonize proinflammatory cytokines; e.g., IL-10 can inhibit TNF-α production by macrophages (39). Thus, immunosuppression caused by an overabundance of anti-inflammatory cytokines can be detrimental during sepsis by promoting bacterial outgrowth (34, 40). Among the major sources of cytokines during sepsis are the innate immune cells, such as macrophages and neutrophils (41). Indeed, the roles of macrophages (42) and neutrophils (43) in sepsis have been extensively studied.

What role does GDF-15 play during sepsis? Luan et al. (18) using an LPS mouse model of sepsis, concluded that GDF-15 is a mediator of disease tolerance. Disease tolerance mechanisms during infection are mechanisms to maintain proper tissue function and integrity despite direct insults to the host by the pathogen and the host’s own immune response (44). However, tolerance mechanisms do not necessarily have to affect pathogen burden or the immune response itself (45). During sepsis, Luan et al. (18) discovered that GDF-15 promotes the mobilization of lipids, which were a critical energy source for the heart.

In the Luan et al. study (46), Ab blockade of GDF-15 had no apparent effect on the proinflammatory cytokine response to LPS. However, there are numerous studies describing an immunoregulatory role of GDF-15 during sepsis. Using the cecal ligation puncture model of sepsis, Santos et al. (47) observed that GDF-15–deficient mice have increased systemic CXCL5, neutrophilic inflammation, and neutrophil-dependent mortality. Abulizi et al. (48) observed that GDF-15 deficiency aggravated cardiac and renal injury during sepsis because of increased expression of inflammatory cytokines, TNF-α and IL-6. In two separate studies focusing on lung (49) and liver (50) injury during sepsis, treatment with recombinant GDF-15 suppressed the LPS-induced production of proinflammatory cytokines by alveolar epithelial cells and Kupffer cells. In addition to its effects on NF-κB–dependent proinflammatory cytokines, GDF-15 also exerts a suppressive effect on IL-1β responses by decreasing expression of key inflammasome components, ASC and caspase-1 (51). Altogether, these studies suggest that GDF-15 indeed suppresses inflammation in different tissues by impairing production of proinflammatory cytokines and recruitment of immune cells to the organs during sepsis. In this regard, GDF-15 functions in parallel with IL-10, protecting the host from tissue immunopathology mediated by dysregulated levels of Th1 cytokines (52).

GDF-15 also inhibits immune cell migration. Studies focused on atherosclerosis, where immune cell recruitment and activation contribute to the pathology, revealed the mechanisms by which GDF-15 regulates immune cell recruitment. In humans, circulating levels of GDF-15 increase as atherosclerosis progresses, which suggests that GDF-15 plays a role in the progression of the disease (53). Indeed, de Jager et al. (53) observed that GDF-15–deficient macrophages express lower levels of CCR2, a chemokine receptor that is required for macrophage migration (54), which diminished the ability of the cells to infiltrate the atherosclerotic lesion (53). Intriguingly, Kempf et al. (55) demonstrated that GDF-15 can modulate leukocytes migration through regulation of integrin activation. Integrins are adhesion molecules that require activation through conformational changes and clustering to increase affinity to their ligands (56). GDF-15 inhibits β₂ integrin activation by inhibiting the activity of Rap1, a GTPase that mediates integrin activation, through the activation of Cdc42, another GTPase (55). Although the studies cited provide conflicting data on the effects of GDF-15 on immune cell migration, overall, they support the idea that, indeed, GDF-15 is a regulator of immune cell migration.

The mechanism by which GDF-15 suppresses LPS-induced cytokine production is still unknown. IL-10 and TGF-β are classically known suppressors of LPS-induced responses in immune cells, albeit through different mechanisms, as first described by Bogdan et al. (57). This study found that concurrent LPS and IL-10 stimulation of macrophages can immediately inhibit TNF-α production, whereas TGF-β has a delayed effect. Conaway et al. (58) found that IL-10 can suppress the immediate enhancer activation in early LPS-induced genes. IL-10 can also suppress NF-κB nuclear translocation either through expression of Bcl-3 (59), which interacts with the p50 subunit, or through suppression of IκB degradation (60). In contrast, a mechanism by which TGF-β can suppress inflammation is by promoting mRNA decay through the suppression of p38 activation, which is an LPS-induced factor that stabilizes mRNA (61). We speculate that GDF-15 may mimic the effects of both IL-10 and TGF-β to suppress inflammation. Indeed, Ratnam et al. (62) found that GDF-15 can suppress TNF-α production by macrophages through the inhibition of NF-κB signaling.

It is interesting to note there are studies that suggest that GDF-15 can regulate inflammation and/or pathogen clearance. GDF-15 promotes inflammation in human periodontal ligaments fibroblasts (HPdLFs) according to a study by Stemmler et al. (63) HPdLFs produce IL-6, IL-8, and TNF-α in response to either mechanical stress or bacterial infection. In the absence of GDF-15, HPdLFs downregulate gene and protein levels of IL-6. Moreover, GDF-15 can also promote production of antimicrobial peptides. Majhi et al. (64) observed that uroepithelial cells upregulate GDF-15 in response to metformin. The uroepithelial cell–derived GDF-15 can contribute to the resistance to bacterial infection by inducing the production of antimicrobial peptides by macrophages (64). However, overexpression of GDF-15 in the airway has been shown to promote viral replication through suppression of IFN-λ expression (65). Thus, GDF-15 may increase antibacterial defense but at the expense of antiviral innate immunity.

GDF-15 also plays a role in a type 2 inflammatory environment. Excess adipose tissues can trigger a chronic proinflammatory response, which can exacerbate metabolic disorders (66). Th2 cytokines, such as IL-13, are known to mediate improvement in insulin resistance and glucose homeostasis (67). Lee et al. (68) found that GDF-15 mediates the IL-13–dependent improvement in glucose homeostasis in obese mice fed a high-fat diet. Interestingly, the study also suggests that GDF-15 plays a role in the polarization of macrophage to alternatively activated or M2 macrophages (68). Jung et al. (69) demonstrated that GDF-15 promotes oxidative phosphorylation, causing the macrophages to adapt an M2 phenotype. Allergic inflammation is characterized by a type 2 immune response, wherein IL-13 induces airway hyperresponsiveness, among many other effects (70). GDF-15 can also promote allergic inflammation in the airways. Harb et al. (26) observed that T_reg-derived GDF-15 promotes IL-13 production by innate lymphoid cell 2 after exposure to ultra-fine particles. Interestingly, elevated GDF-15 levels in trauma patients correlated with decreased NK activation by IL-12, suggesting that GDF-15 may also be involved in cross-regulation of type 1 immunity by type 2 cells (71).

T_regs are a subset of CD4⁺ T cells that are highly implicated in autoimmune diseases because of their immunoregulatory role (72)_. Studies using mouse models of autoimmune disease suggest that GDF-15 can regulate T_reg formation and function. Lorenz et al. (73) observed using a lupus model that GDF-15 deficiency leads to a decrease in the number of T_regs in the spleen and an increase in the overall number of T cells and activated CD8 T cells. Interestingly, Wang and colleagues (74) found that, in the presence of IL-2, GDF-15 can induce T_reg differentiation of CD4 T cells in vitro to a comparable degree as TGF-β. This study also demonstrated that GDF-15 downregulates STUB-1 expression, a ubiquitin ligase, preventing the ubiquitination of Foxp3, the master regulator of the development and function of T_regs (75), thereby stabilizing the transcription factor and promoting CD4 T cell differentiation into T_regs (74). Moon et al. (76) observed that GDF-15 can augment IL-10 production by T_regs, which enhances the suppression of T cell effector functions. Carbon tetrachloride (CCl₄) causes hepatic injury through the formation of free radicals after being metabolized in the liver (77). Chronic CCl₄ exposure causes a dysregulated inflammation in the liver that causes fibrosis (78). GDF-15 is upregulated in CCl₄-treated mice, which ameliorates fibrosis by limiting the number of T cells and suppressing the activation of CD8 T cells (50). Given the known role of T_regs in preventing aggravated liver fibrosis (79), a potential mechanism through which GDF-15 suppresses hepatic inflammation and fibrosis is by promoting the formation and function of T_regs. Two outstanding questions in this field pertain to the critical sources of GDF-15 that enhance T_reg function and what cytokines/transcription factors drive GDF-15 expression by certain T_reg subsets.

As previously mentioned, GDF-15 mediates tissue tolerance by promoting lipid mobilization during sepsis. GDF-15 is also a regulator of iron homeostasis, which can be another potential mechanism through which GDF-15 mediates immune response and tissue tolerance. Animals can obtain iron either from the diet (absorbed by enterocytes) (80) or from senescent RBCs (mediated by macrophages) (81). The release of intracellular iron from both enterocytes and macrophages is mediated by ferroportin and negatively regulated by hepcidin (82, 83). Interestingly, GDF-15 has been shown to inhibit hepcidin expression, which would lead to increased systemic iron availability (84). During the blood stage of malaria infection, hepcidin is induced, thereby withholding iron from the parasite (85). In contrast, increased intracellular stores of iron in macrophages could favor Mycobacterium tuberculosis growth (86). Thus, upregulation of GDF-15 could exert effects that are favorable to the host or to the pathogen, through its regulation of the ferroportin-hepcidin. Overall, the ability of GDF-15 to regulate iron availability presents another potential area of investigation for the role of GDF-15 in infection and inflammation.

Infection-induced anorexia impacts the host response to a pathogen (87, 88). In a fasted state, the host uses alternative energy sources, from carbohydrates to lipids (89). Free fatty acids are liberated from stores and are used by the liver to produce ketogenic metabolites, including β-hydroxybutyrate (BHB). BHB is an alternative energy source used by immune cells to properly function during nutrient deficiency because of infection (90). Karagiannis et al. (91) found that impaired ketogenesis contributed to T cell dysfunction in COVID-19 patients with acute respiratory distress syndrome. BHB supplementation improved the function of T cells from COVID-19 acute respiratory distress syndrome patients. GDF-15 is known to induce ketogenesis in the liver, as shown by Zhang et al. (92). Thus, GDF-15 activation of ketogenesis represents another potential mechanism by which GDF-15 regulates inflammation and the immune response.

There are several known signaling pathways involved in the induction of Gdf15 (see (Fig. 2). The integrated stress response (ISR) is a signaling cascade activated in response to a variety of stress stimuli, including nutrient deprivation and infection (93). These stressors can result in overwhelming amounts of unfolded proteins, which can cause cell death. The ISR can protect the cells by temporarily shutting down translation to prevent further accumulation of improperly folded proteins (94). Key factors involved in this pathway include the α subunit of eukaryotic translation initiation factor 2, activating transcription factor 4, and C/EBP homologous protein (95). Patel et al. (16) showed that nutritional deprivation leads to Gdf15 upregulation via the activating transcription factor 4/C/EBP homologous protein pathway. p53, a tumor-suppressor gene implicated in a variety of different cancers (96), is also a regulator of Gdf15. p53 binds the Gdf15 promoter, as shown by Osada et al. (97). GDF-15 is also known as NSAID-activated gene-1, first described by Baek et al. (2). Recently, Eisenstein et al. (98) found that NSAID triggered myeloid cell Gdf15 upregulation through the induction of NF erythroid 2–related factor 2.

Inflammatory cytokines induce Gdf15 expression. Bootcov et al. (1) first identified GDF-15 after screening activated monocytes for unique factors that regulate macrophage function and named it macrophage inhibitory cytokine 1. Further, the study demonstrated that proinflammatory cytokines, including IL-1β, IL-2, and TNF-α, itself, could induce Gdf15 expression, suggesting that GDF-15 might be an autocrine signal that inhibits a proinflammatory response. Ratnam et al. (62) also demonstrated that GDF-15 is regulated by NF-κB signaling. Transformed fibroblasts produce GDF-15 via the p65–NF-κB axis. It is likely that the proinflammatory cytokines tested by Bootcov et al. (1) induced Gdf15 expression in an NF-κB–dependent manner. In contrast, Lee et al. (68) showed that either IL-4 or IL-13 can directly control Gdf15 expression in adipocytes. Therefore, both type 1 and 2 cytokines can induce Gdf15 expression.

GDF-15 levels can be regulated on the level of the mRNA. mRNA stability and turnover can influence the amount of protein produced from a gene. mRNA is stabilized through the addition of multiple adenosine [poly(A)] at the tail of the transcribed gene (99). Conversely, mRNA degradation is important to maintain homeostatic levels of a protein. Deadenylases aid in mRNA turnover by removing the poly(A) tail, making the mRNA unstable and more susceptible to degradation. Interestingly, Katsumura et al. (100) found the CCR4-NOT deadenylase complex at steady state can degrade Gdf15 mRNA and prevent its translation, thereby limiting the circulating levels of GDF-15. Inhibiting the deadenylase complex stabilizes the Gdf15 mRNA, which results in more production of the protein and recovery of the anorectic effect of GDF-15 (100).

Late-stage cancer patients often experience anorexia and cachexia (101, 102). Circulating GDF-15 is elevated in advanced-stage cancer patients and is strongly correlated with cancer-induced cachexia (103). In line with this, S. Breit’s group discovered a mechanism by which circulating GDF-15 levels are regulated. GDF-15 is produced as a precursor protein, containing a propeptide and the mature peptide, similar to other TGF-β family members (104). Pro–GDF-15 contains a cleavage site recognized by members of the proprotein convertase subtilisin/kexin (PCSK) protein family, including PCSK-3,5,6 (105). Bauskin et al. (106) showed that the GDF-15 propeptide not only facilitates the proper folding of cytokine but also allows binding to heparin (106), which is abundant in the extracellular matrix (102). Thus, immature GDF-15 bound to heparin and stored in the extracellular matrix represents an extracellular stromal reservoir, which can be released into the circulation when triggered. However, the exact mechanism for the liberation of GDF-15 is still unknown. Likely, the GDF-15 maturation, involving proteolytic cleavage of the propeptide, will be a required step to liberate GDF-15 from these stromal stores.

Because GDF-15 is a good candidate for therapeutic interventions, especially in obese (107) and cancer patients (108), ways to intervene with its central function have been investigated. Antagonism of the GFRAL receptor can indeed abolish the central effects of GDF-15. This has been achieved by surgical ablation of the AP/NTS region of the brainstem (22) and by Ab blockade (21). Furthermore, a compelling study by Chow et al. (109) demonstrated that posttranslational modification of the GFRAL can also be a mechanism to regulate GDF-15 function. The study found that matrix metalloproteinase 14 can proteolytically cleave GFRAL, thereby impeding GDF-15 interaction with the receptor, abolishing the hormone’s weight-regulation function. Thus, various mechanisms that regulate the GDF-15 receptor can also be viable options in regulating the effect of GDF-15 on energy homeostasis.

GDF-15, a distant member of the TGF-β superfamily, is primarily known as a biomarker of the severity of diverse disease states and as an induced regulator of energy metabolism under nonhomeostatic conditions. Nonetheless, there are many studies revealing potent immunoregulatory roles of GDF-15. Numerous published studies describe the inhibitory role of GDF-15 on the production of proinflammatory cytokines such as TNF-α. Moreover, GDF-15 induces a tissue-protective program during inflammation. Thus, in a type 1 inflammatory setting, GDF-15 behaves like IL-10, a suppressor of type 1 response and a mediator of tissue tolerance. Numerous studies also reveal that GDF-15 promotes a type 2 immune response and can mediate improvement in energy metabolism, similar to the effects of IL-4 or IL-13. Future studies should address the protective and pathogenic roles of GDF-15 in the immune response to diverse intracellular or extracellular pathogens, which induce a type 1 and 2 immune response, respectively.

The regulation of GDF-15 is complex and occurs at multiple levels. Stress signals, through pathways such as the ISR, induce GDF-15 as a host adaptive mechanism that alters metabolism to maintain host integrity. Inflammatory cytokines also induce GDF-15 expression, although the precise pathways involved are not well understood. However, it is likely that immune cytokines induce GDF-15 expression through the same canonical pathways used by stress signals (see (Fig. 2). Thus, TNF-α can potentially induce GDF-15 expression through diverse pathways, including the p38 (110) and p53 (111) signaling pathways. Raines et al. (112) demonstrated that IL-4 triggers the ISR in macrophages, specifically upregulating the PERK pathway. Thus, aside from the direct IL-4–STAT-6 pathway, IL-4 induction of ISR could potentially be a pathway that induces GDF-15. GDF-15 can also be regulated on the mRNA and protein levels, but the involvement of the immune response is not clear. Can inflammatory cytokines regulate deadenylases that destabilize Gdf15 mRNA or proteolytic enzymes involved in GDF-15 maturation? Interestingly, PCSK-3 is regulated by IL-12 in Th1 cells, as shown by Pesu et al. (113). Whether the IL-12–dependent induction of furin family convertases is involved in the maturation process of GDF-15 and its liberation from tissue in the stromal stores is yet to be determined.

Determining the identity of the functional GDF-15 receptor in the periphery, including in diverse immune cell types, is a clear priority. Multiple studies have already confirmed that GFRAL is restricted to AP and NTS neurons (18, 114). However, it is undeniable that cells elsewhere can respond to GDF-15 despite the apparent lack of GFRAL expression.

Interestingly, Artz et al. (115) demonstrated that GDF-15 modulation of integrin activation in neutrophils is mediated by GDF-15 binding to the TGF-β receptor I/II complex, suggesting that this receptor might also be the GDF-15 receptor in other immune cells. Indeed, exposure of leukocytes to GDF-15 induces canonical Smad activation that is dependent on the TGF-β receptor (71). Moreover, a more recent study by Wang et al. (75) demonstrated that CD48 (also known as signaling lymphocytic activation molecule 2) can bind GDF-15 in T cells. Critical evaluation and validation of these candidate GDF-15 receptors used by diverse lymphoid and myeloid cells will be important in advancing our understanding of the immunoregulatory actions of GDF-15.

After considering the systemic and immune effects of GDF-15, we conclude that GDF-15 is primarily a stress-induced cytokine that promotes host fitness by (1) downregulating inflammation and (2) preserving cellular integrity after tissue injury. It also represents a mechanism for how tissues can communicate distress to the brain and other organs and regulate metabolism. In most instances, these responses are likely to be adaptive and facilitate a return to homeostasis. However, under severe or chronic disease conditions, GDF-15–triggered responses might become maladaptive and contribute to further pathogenesis. It remains an open question whether targeting GDF-15 can be used as an adjunctive therapy for cancer or severe infectious diseases (116, 117).

The authors have no financial conflicts of interest.