Central mechanisms of emesis: A role for GDF15

Abstract

Background

Nausea and emesis are ubiquitously reported medical conditions and often present as treatment side effects along with polymorbidities contributing to detrimental life‐threatening outcomes, such as poor nutrition, lower quality of life, and unfavorable patient prognosis. Growth differentiation factor 15 (GDF15) is a stress response cytokine secreted by a wide variety of cell types in response to a broad range of stressors. Circulating GDF15 levels are elevated in a range of medical conditions characterized by cachexia and malaise. In recent years, GDF15 has gained scientific and translational prominence with the discovery that its receptor, GDNF family receptor α‐like (GFRAL), is expressed exclusively in the hindbrain. GFRAL activation may results in profound anorexia and body weight loss, effects which have attracted interest for the pharmacological treatment of obesity.

Purpose

This review highlights compelling emerging evidence indicating that GDF15 causes anorexia through the induction of nausea, emesis, and food aversions, which encourage a perspective on GDF15 system function in physiology and behavior beyond homeostatic energy regulation contexts. This highlights the potential role of GDF15 in the central mediation of nausea and emesis following a variety of physiological, and pathophysiological conditions such as chemotherapy‐induced emesis, hyperemesis gravidarum, and cyclic vomiting syndrome.

Key points.

• Nausea and emesis are common medical conditions and treatment side effects, leading to poor nutrition, lower quality of life, and unfavorable patient prognosis.

• GDF15 is a stress response cytokine that is elevated in conditions associated with malaise and cachexia.

• GDF15 plays a central role in chemotherapy‐induced emesis and nausea, cancer anorexia cachexia syndrome, hyperemesis gravidarum, and potentially, cyclic vomiting syndrome.

1. INTRODUCTION

Nausea and vomiting (i.e., emetic behaviors) are primitive aspects of physiology and behavior that, although unpleasant, can promote mammalian survival. ¹ Paradoxically, emetic “side effects” are universally reported for FDA‐approved pharmacotherapeutics for obesity, diabetes, and cancer, during which they may reduce pharmacotherapy tolerance. Emetic behaviors are additionally present alongside polymorbidities that contribute to detrimental life‐threatening outcomes, such as poor nutrition, low quality of life, and unfavorable patient prognosis. ² , ³ Evidence‐based reports indicating that nausea and emesis are hurdles for optimizing the effective impact of therapeutics are critical to our understanding of emetic behaviors given that the pharmaceutical management of these conditions (e.g., obesity, diabetes, and cancer) requires prolonged interventions. ⁴ , ⁵ Without controlling emetic side effects, drug treatments are, in essence, “sub‐optimal.” ⁶ , ⁷ In fact, reports pertaining to profile and utility of now increasingly popular Glucagon‐like peptide 1 (GLP‐1) receptor agonists, concluded that many patients reported gastrointestinal‐related issues that “Made me feel sick” (64.4%) and “Made me throw up” (45.4%) as their major reasons for treatment discontinuation. ⁸

Nausea is often treated as a symptom fused with vomiting, when nausea is more debilitating to quality of life, far less controlled, and has a higher incidence than vomiting, all occurring despite a wide range of available antiemetic prophylactic drugs. ⁹ , ¹⁰ , ¹¹ , ¹² In fact, for patients suffering from cancers, chemotherapy‐induced nausea and anorexia remain important clinical problems, even when chemotherapy‐induced vomiting is relatively better controlled. ¹³ , ¹⁴ , ¹⁵ , ¹⁶ , ¹⁷ , ¹⁸ Such complexities may also explain why there are clear “disparities between patient experiences and physician perceptions” with a clear need to improve “gaps in physician–patient communication” regarding the incidence of illness following treatment. ⁸ These effects are not transient nor insignificant as they may lead to discontinuation of treatment. In fact, in a significant percentage of oncology patients (up to 30%), nausea and emesis can be so distressing that treatment discontinuation is contemplated, which could lead to an obvious negative impact on prognosis. ¹⁹ Perhaps even more compelling are data in women suffering from severe forms of nausea and vomiting behaviors during pregnancy (i.e., hyperemesis gravidarum [HG]), which in and of itself is a devastating experience while pregnant, that show roughly half of study participants considered termination of pregnancy, and 25% reported suicidal ideations as a consequence of nausea and emesis. ²⁰

Growth differentiation factor 15 (GDF15) is a relative newcomer to the field of emesis and its importance has been highlighted by our group and others in rodents, non‐human primates, and humans. ²¹ , ²² , ²³ , ²⁴ , ²⁵ , ²⁶ Exogenous GDF15 administration profoundly reduces food intake and body weight in multiple species, ²⁷ , ²⁸ , ²⁹ , ³⁰ , ³¹ , ³² , ³³ and similarly, overexpression of GDF15 in mice results in chronic hypophagia and a lean phenotype, ³⁴ whereas knockout (KO) of GDF15 results in increased adiposity. ³⁵ Relevant to the current review, exogenous GDF15 was recently found to be an emetic stimulus. ²² Conversely, GDNF family receptor α‐like (GFRAL) antagonism ³⁶ or neutralization of circulating GDF15 ²⁵ was reported to have anti‐emetic properties against chemotherapy. Additionally, in recent years, there has been profound development and characterization of GDF15 action as a potential primary driver of morning sickness and complications of pregnancy involving severe and uncontrolled emetic behaviors, such as in HG. ²¹ , ²⁶ In the following review, we will highlight GDF15's role in chemotherapy‐induced nausea and vomiting (CINV), cancer anorexia cachexia syndrome (CACS), morning sickness in pregnancy, and HG. Additionally, we will speculate on other conditions where GDF15 might be involved, such as mitochondrial dysfunction and cyclic vomiting syndrome.

2. NAUSEA AND VOMITING MEDIATED BY CENTRAL CIRCUITS: FOCUS ON THE HINDBRAIN

Strategies to mitigate the symptoms of nausea and vomiting must appreciate that counteracting emetic behavior necessarily requires focus on control within the “vomiting center” located in the hindbrain. ³ , ³⁷ , ³⁸ , ³⁹ , ⁴⁰ This emetic “hub” termed the dorsal vagal complex (DVC) is comprised of three distinct but interconnected nuclei—the area postrema (AP), the nucleus tractus solitarius (NTS), and the adjacent dorsal motor nucleus of the vagus (DMV). Physiological and pathological modulators of energy balance and emesis share many common neural substrates and anatomical nodes centrally, including the AP/NTS. ³ , ³⁷ , ³⁸ The AP/NTS is the primary axonal target of vagal afferents (10th cranial nerve) projections originating from the gut, while the presence of fenestrated capillaries in the AP allows neurons in the DVC to be reached by emetic/anorectic circulating agents that cannot cross the blood–brain barrier (BBB) in other brain regions. ³⁷ , ⁴¹ Increased neuronal activity in these nuclei is therefore not surprisingly associated with emesis and nausea behaviors. ³⁹ , ⁴² For example, the chemotherapy agent cisplatin, anti‐diabetic/obesity drugs like GLP‐1 analogs, beta‐blockers for cardiac arrhythmias, as well as other emetic stimuli activate neurons in the AP/NTS, subsequently activate other upstream central nervous system (CNS) targets from the DVC implicated in regulation of feeding and the development of aversions (e.g., the parabrachial nucleus [PBN] and central nucleus of the amygdala [CeA]). ⁴³ , ⁴⁴ , ⁴⁵ , ⁴⁶ , ⁴⁷ Surprisingly, the peripheral mediators acting in the AP/NTS in response to noxious agents, as well as the upstream neuronal circuitry contributing to nausea, emesis, and anorexia, however, remain poorly understood. Thus, a reassessment of the neural substrates that mediate nausea and emesis, like GDF15, using modern behavioral and neurogenetic approaches is required to achieve more effective, sustained control of nausea and vomiting in a variety of important clinical contexts.

3. GDF15‐GFRAL: HINDBRAIN PATHS OF NAUSEA AND VOMITING

3.1. GDF15: Discovery, expression, circulation, and transcriptional regulation

GDF15, formerly macrophage inhibitory cytokine‐1 (MIC‐1), was discovered by the laboratory of Samuel Breit in 1997, while identifying novel macrophage‐secreted biologically active molecules. ⁴⁸ Originally named MIC‐1 to reflect an ability to inhibit macrophage secretion of tumor necrosis factor‐α (TNF‐α) following lipopolysaccharide (LPS) stimulation, the gene encoded a protein bearing structural characteristics (i.e., three‐dimensional structure, including a highly conserved pattern of seven cysteine residues in its mature domain) of a transforming growth factor beta (TGF‐beta) superfamily cytokine. ⁴⁸ It was subsequently given the official designation of GDF15, ⁴⁹ although recent data indicate it is not a GDF, but most likely a member of the a glial‐derived neurotrophic factor (GDNF) family instead. ⁵⁰ In humans, GDF15 is the product of a 2‐exon gene, located on chromosome 19p13.11 encoding a transcript of ~1200 bp, which translates a mature peptide of 112 amino acids. ⁵¹ Orthologues have been annotated in mammals, reptiles, amphibians, bony fish, and birds with a high level of conservation observed in the C‐terminal region of the protein representing the mature peptide. ⁵²

Mature GDF15 circulates as a 25 kDa dimer linked by a single inter‐chain disulphide bond. ⁴⁸ In healthy individuals, GDF15 is expressed most abundantly in the placenta, followed by the prostate, kidney, colon, liver, and lung; it is also expressed at lower levels in the brain, heart, pancreas, gastrointestinal tract, and bone marrow. ⁵³ Unusual for a “classic” cytokine, GDF15 circulates in substantial amounts in healthy humans, with a normal range of about 200–1000 pg/mL, while in rodents, they are lower, at about 50–200 pg/mL. ⁵⁴ GDF15 concentrations also increase with age, following physical exercise, smoking, and drug use. ⁵² Blood GDF15 levels also rise in a wide range of diseases, including most types of cancer, chronic inflammatory diseases, cardiovascular, pulmonary and renal diseases, obesity, diabetes, infections, tissue injury (e.g., anoxia, ischemia), and other medical conditions such as pregnancy and as a consequence of many medical treatments (e.g., metformin and anticancer therapies such as chemotherapy and ionizing irradiation) (see, ⁵² , ⁵⁴ for review). Pathophysiological relevance in some of these medical conditions will be discussed below.

Considering the diverse conditions associated with increased GDF15, it is not surprising that its transcription can be influenced by numerous cellular regulators. Upstream of the GDF15 promotor site there are several binding sites for basal transcription factors (for comprehensive review, see ⁵² ). One of the first transcription factors identified as a regulator of GDF15 was p53. ⁵⁵ Another potential tumor suppressor protein that has been implicated in regulation of GDF15 is early growth response factor 1 (EGR1), a zinc‐finger protein induced in response to growth factor signaling and DNA damage. ⁵⁶ The integrated stress response (ISR), via the alpha subunit of eukaryotic translation initiation factor 2 (eIF2α), also seems to be of critical importance to increase GDF15 expression. ⁵⁷ Given its cytokine nature, another direct effector is the transcription factor complex NF‐κB. ⁵⁸

3.2. GFRAL: Discovery of the GDF15 receptor

The receptor for GDF15 had eluded identification until late 2017, when four research groups from independent pharmaceutical companies simultaneously identified it as GFRAL, a previously characterized distant orphan member of the GDNF receptor alpha (GFRα) family. ²⁷ , ³⁰ , ³² , ³³ The GFRAL gene consists of 9 exons and 8 introns and it is located on human chromosome 6p12.1. Its major transcript codes for a 393 amino acid long protein with a molecular weight of ~44 kDa and ~70% homology with its murine counterpart. ⁵⁹ In contrast to the other GFRα family members, GFRAL has a single transmembrane and very short cytoplasmic domains unlikely to have signaling functions, suggesting that it does not serve other functions than anchoring the receptor to the cell membrane. Each of the research groups who published in 2017 utilized distinct methodologies, and conducted independent, unbiased screens of thousands of human cell membrane‐associated proteins as well as a biased screenings of orphan members of the GDNF family. ²⁷ , ³⁰ , ³² , ³³ Importantly, these studies demonstrated that GFRAL did not bind with any structurally related ligands of the TGFβ and GDNF families. Conversely, GDF15 binds with high affinity to GFRAL, but to no other members of the GDNF family of receptors, clearly identifying the GFRAL receptor as the sole receptor for GDF15. Curiously, a shorter splicing variant creates a truncated protein (GFRAL‐B) that lacks the transmembrane region. This could form a soluble secreted protein that could function as a soluble receptor that blocks GDF15 signaling (competitive inhibitor) or even induces trans‐signaling on the rearranged during transfection (RET) tyrosine kinase co‐receptor but non‐GFRAL expressing cells, ⁵⁴ however this notion has not yet been examined.

3.3. Intracellular signaling cascades activated by GFRAL‐RET

The foundational studies on GFRAL identification, in line with the finding of a lack of proper cytoplasmatic domain, also importantly highlighted the necessary recruitment of the RET tyrosine kinase co‐receptor to activate the GDF15‐GFRAL intracellular signaling. The stoichiometry of GDF15:GFRAL:RET binding interaction is postulated to be 1 GDF15 homodimer to 2 GFRAL receptors to 2 RET receptors. ⁶⁰ Upon GDF15 binding to GFRAL, it induces conformational changes in the receptor that facilitates the dimerization of RET receptors, leading to its auto‐phosphorylation. ²⁷ , ³⁰ , ³² Phosphorylated RET acts as a docking site for various intracellular signaling molecules. ⁶¹ The recruitment and subsequent activation of these downstream effectors lead to the activation of protein kinase B (Akt), the proto‐oncogene Fos, extracellular signal‐related kinase ½ (Erk1/2) and phosphoinositide‐specific phospholipase C (PLC) that ultimately alter neuronal activity and cause gene expression changes (see Figure 1). ²⁷ , ³⁰ , ³² , ⁶²

FIGURE 1.

Overview of the GDF15/GFRAL system in the central mediation of nausea and emesis. Circulating GDF15 levels are elevated in various medical conditions, including cancer, treatment with chemotherapeutic agents, exposure to toxins, pregnancy, and mitochondrial stress/dysfunctions. Bloodborne GDF15 can directly reach the area postrema (AP), the nucleus tractus solitarius (NTS), which are key hindbrain structures for the mediation of nausea and emesis. Within the AP/NTS there is a unique neuronal population which co‐expresses both the GDF15 receptor, GDNF family receptor α‐like (GFRAL) and its co‐receptor rearranged during transfection (RET) tyrosine kinase. Upon GDF15 binding to GFRAL, GFRAL conformational changes occurs that facilitates the dimerization of RET, leading to the subsequent phosphorylation (i.e., activation) of several downstream effectors (protein kinase B [Akt], extracellular signal‐related kinase ½ [Erk1/2] and phosphoinositide‐specific phospholipase C [PLC]) that ultimately alter neuronal activity and gene expression profile. Consequently, this activation extends to other upstream central targets implicated in regulation of anorexia, nausea and emesis (e.g., the parabrachial nucleus [PBN] and central nucleus of the amygdala [CeA]).

3.4. GFRAL: Neuronal expression, phenotype, and projections

Humans, cynomolgus monkeys, rats, and mice all remarkably show GFRAL expression highly localized in the CNS, exclusively expressed in the AP and in the adjacent medial part of the NTS. In large‐scale screenings of tissues and cell lines, GFRAL expression has been found to be virtually absent outside the CNS, although limited evidence does suggest low expression in the gonads and adipose tissue. ²⁷ , ³⁰ , ³² , ³³ During development, a more widespread CNS Gfral expression was identified, suggesting potential roles of GFRAL signaling in neurogenesis. ⁵⁹ , ⁶³ The apparent conundrum of a cytokine capable of being expressed and released by virtually all cells/tissues but only “sensed” by a small subset of highly localized neurons within the hindbrain represents the potential uniqueness of this system and raises the possibility of the existence of a second, yet unidentified, receptor. While this seems unlikely given evidence to date, it represents a venue for further research and speculation.

Parallel studies leveraging single nuclei sequencing, RNA scope techniques, and murine transgenic lines, provided further and deeper phenotypical characterization of GFRAL neurons by identifying a transcriptomically distinct population of excitatory neurons within in the AP/NTS that co‐express both GFRAL and its coreceptor RET. ⁶⁴ , ⁶⁵ , ⁶⁶ , ⁶⁷ , ⁶⁸ Despite some minor incongruency likely due to species differences (i.e., rat vs. mouse) and the methods of analysis employed, collectively these reports indicate that GFRAL‐positive neurons are exclusively excitatory neurons and co‐express the neuropeptide cholecystokinin (CCK) in a relatively high percent, and to a lesser extent, the catecholaminergic marker tyrosine hydroxylase (TH). ⁶⁴ , ⁶⁵ , ⁶⁶ , ⁶⁷ , ⁶⁸ Tracing studies indicate that while some GFRAL neurons send axon projections locally within the DVC, GFRAL neurons heavily project to two brain nuclei known to play key roles in the regulation of feeding and homeostasis: the parabrachial nucleus (PBN) where they directly innervate CGRP‐positive neurons, and to the paraventricular nucleus of the hypothalamus (PVN) where they are most likely synaptic contact with CRH expressing neurons (see Figure 1). ⁶³ , ⁶⁸ , ⁶⁹

3.5. GDF15‐GFRAL activation: Effects on nausea, vomiting, and energy dysregulation

GDF15 administration reduces food intake and body weight in nonhuman primates and rodents. ²⁷ , ²⁸ , ²⁹ , ³⁰ , ³¹ , ³² , ³³ Similarly, overexpression of GDF15 in mice results in chronic hypophagia and a lean phenotype, ³⁴ whereas KO of GDF15 leads to a moderate increase in adiposity. ³⁵ A recent clinical trial conducted by Eli Lilly, Inc. in participants with overweight and obesity showed that chronic treatment with a long‐acting GDF15 analog significantly decreased hunger, but only modestly affected body weight, thus signifying challenges in clinical weight‐loss applications targeting GDF15‐GFRAL. ⁷⁰

We have reported that in addition to the well‐documented anorectic and body weight lowering effects, exogenous GDF15 administration induced anorexia through nausea and emesis, suggesting that the anorectic response to GDF15 is driven by sensations of nausea and malaise (i.e., feelings or visceral discomfort or “sickness”). ²² All anorectic doses of GDF15 induced pica behavior (i.e., the ingestion of non‐nutritive kaolin clay, a validated, powerful, and accurate proxy for nausea/emesis in non‐emetic laboratory rodents ⁷¹ , ⁷² , ⁷³ ), that preceded the onset of the anorectic response. Anorexia and malaise also occurred upon repeated exposure, denoting the lack of an improved tolerance over time. ²² , ²³ Similar studies that employed selective activation of GFRAL neurons via chemogenetic tools in mice yielded comparable outcomes, demonstrating the ability of GFRAL neuronal activation to drive anorexia and to cause conditioned taste avoidances in mice. ⁶³ , ⁶⁴ Using the shrew model, we also provided clear evidence that GDF15 is emetogenic and anorexigenic, supporting the notion that the hindbrain emetic centers were convincingly engaged by GDF15. ²² Interestingly, GDF15‐induced anorexia does not appear to impact hedonic and rewarding aspects of feeding, hypothalamic agouti‐related peptide (AgRP) signaling, or gastrointestinal satiation signals, thereby reinforcing the concept that GDF15 likely causes anorexia mainly via the induction of malaise. ²⁴ It could be speculated that, beyond its behavioral effects on feeding and energy homeostasis, activation of the GDF15/GFRAL system also enhances survival during inflammatory states through central induction of metabolic adaptation. In fact, data show that this GFRAL‐mediated response activates hepatic triglyceride export via sympathetic outflow to support cardiac function and robustly activates the hypothalamic–pituitary–adrenal (HPA) axis leading an increase in glucocorticoid hormones, which are essential for survival in the face of severe somatic stresses. ⁶⁹ , ⁷⁴

The exploration of GDF15 agonism as an obesity treatment has shown limited results to date, primarily due to its modest body weight reducing actions and its strong emetogenic potential, ²² , ⁷⁰ although the pursuit of its antagonistic properties does emerge as a very promising avenue and holds considerable potential as an effective and much‐needed intervention for medical conditions associated with severe and chronic nausea and emesis. Below, we have identified and highlighted four medical areas that may greatly benefit from GDF15‐GFRAL based therapeutic approaches.

4. THE ROLE OF GDF15 IN THE MEDIATION OF CINV

CINV is among the most recognized clinical examples highlighting the severity and prevalence of these side effects surrounding diseases and treatments. ³ Despite significant advancements in pharmacology within the field, particularly via serotonin type 3 receptor (5‐HT₃R) and neurokinin 1 receptor (NK1R) antagonists, ⁷⁵ satisfactory control of CINV remains unmet. Today, CINV continues to be a distressing set of side effects among patients undergoing classical/conventional chemotherapeutic treatments. ⁷⁶ Between 30% and 80% of cancer patients experience CINV even with the use of modern anti‐emetic medications ⁷⁷ , ⁷⁸ and although common, CINV is often underestimated in outpatient settings, which can exert a profound negative influence on prognosis, quality of life, and the ability to adhere to treatment. ⁷⁵ , ⁷⁹ Historically, CINV severity has even dissuaded some patients from proceeding with potentially curative chemotherapeutic regimes. ⁸⁰

In the early 2000s GDF15 was purported to be a potential circulating prognostic indicator of chemotherapy response and chemotherapy‐induced endothelial damage (e.g., damage to viscera or poor outcomes), and notably these reports appeared well before GFRAL discovery. ⁸¹ , ⁸² , ⁸³ , ⁸⁴ These studies did not specifically delve into GDF15 effects on energy homeostasis, however they importantly established evidence suggesting GDF15 in circulation may be stimulated by chemotherapy. Upon the discovery of GFRAL, it was confirmed in controlled experiments that cisplatin increases GDF15 secretion in vivo, an effect now well established in multiple preclinical models. ²² , ²⁵ , ²⁷ , ⁶⁸ Hsu et al. also remarkably observed that GFAL KO mice were insensitive to the anorexia and weight loss‐producing effects of cisplatin.

Exogenous GDF15 administration is now well known to activate AP/NTS neurons ²⁷ , ²⁹ , ⁶⁸ and we have previously shown that a significant portion of GFRAL neurons are activated upon cisplatin administration. ²² To provide deeper mechanistic and molecular insights underlying the behavioral and neuronal effects of GDF15 signaling in CINV, our group performed an unbiased systematic characterization of the AP/NTS cellular transcriptome using snRNA‐seq in rats following GDF15 or saline treatment. While more than 500 differentially expressed genes were identified in the AP/NTS, changes in RET/GFRAL co‐expressing neurons (i.e., direct effects of GDF15) included significant upregulation of Ret and Doc5, a component of the mitogen‐activated protein kinase (MAPK) signaling cascade known to be activated by GDF15‐GFRAL signaling (see ⁶² for details of the transcriptomic analyses). Lastly, in a recent study, using a dose of GDF15 that mimics endogenous levels of GDF15 following chemotherapy treatment, we tested the ability of several FDA‐approved antiemetics to counteract GDF15‐induced anorexia and kaolin consumption in rats. ⁸⁵ We found that all failed to show meaningful beneficial effects in preventing GDF15 actions, suggesting that mechanism(s) of sickness behaviors engaged by GDF15 may be largely independent of the signaling pathways targeted by the tested antiemetics. These results may also explain why emesis, nausea, and anorexia remain broad problems, which further stresses the importance of evaluating alternative strategies to counteract GDF15 action.

Work by our group and others have begun to report in preclinical models the anti‐emetic effects of a peptide‐based GFRAL antagonist ³⁶ and GDF15 antibody ²⁵ strategies, respectively. ²⁴ Using a potent long‐lasting antibody against GDF15, Breen et al. demonstrated a strong suppression of cisplatin‐induced emesis in cynomolgus monkeys following GDF15‐neutralization. ²⁵ Similarly, by leveraging an alternative unique strategy, we developed and characterized a novel small peptide‐based antagonist of the GFRAL‐RET complex, that by preventing GFRAL‐RET hetero‐dimerization, mitigates downstream signaling generated by RET recruitment. In our studies, we showed an attenuation of GDF15‐induced malaise as well as malaise resulting from cisplatin. In addition, co‐administration of our GFRAL antagonist with ondansetron led to a greater attenuation of the anorectic effect of cisplatin. ³⁶

5. THE ROLE OF GDF15 IN CANCER ANOREXIA CACHEXIA SYNDROME (CACS)

Cancer anorexia‐cachexia syndrome (CACS) is a complex multifactorial syndrome that is characterized by reduced eating (anorexia), hypermetabolism/catabolism, and body weight loss, which is the consequence of reduced adipose, muscle mass, and abnormal inflammation. ⁸⁶ , ⁸⁷ , ⁸⁸ CACS is associated with reduced survival and poor prognosis, ⁸⁶ , ⁸⁹ presumably due to increased susceptibility to infections, decreased tolerance to chemotherapies, and lack of essential micronutrients (e.g., iron). ⁹⁰ An increased prevalence of chemotherapy side effects may be likely observed in patients with CACS, leading to necessary dose reductions, treatment delays, or even treatment discontinuation. Thus, patients suffering from CACS likely do not receive the full potential benefit of current therapies. ⁹⁰ , ⁹¹ , ⁹² It has been assessed that CACS directly accounts for 20% of cancer deaths ⁹³ and it is estimated that CACS contributes to >7 million deaths worldwide each year. ⁹⁴ Several treatments and approaches have been used in the past with minor beneficial effects ⁹⁵ , ⁹⁶ and the most recently tested compounds such as cannabinoids and ghrelin analogs, despite high expectations, have failed to show tangible benefits. ⁹⁷ , ⁹⁸ , ⁹⁹ Therefore, the identification of the pathophysiological mechanisms mediating CACS and the development of effective pharmacological approaches, has a high clinical relevance.

CACS itself, similar to chemotherapy and radiation therapy, has been well known to cause profound nausea and emesis likely via muscle wasting, gastroparesis, and inflammation, which notably are stand‐alone factors that commonly lead to chronic moderate nausea. ¹⁰⁰ The prevalence of nausea and/or vomiting in terminal cancer patients who are no longer receiving anti‐tumor treatment ranges from 30% to 60%. ¹⁰¹ , ¹⁰² , ¹⁰³ , ¹⁰⁴ , ¹⁰⁵ Fundamental pre‐clinical work under the direction of Ilene Bernstein in the late 20th century provided the first evidence in a laboratory and controlled setting that rats bearing tumors avoid the diet eaten during tumor growth, thus contributing to the anorectic response, as anorexia is attenuated when a novel diet is presented. ¹⁰⁶ , ¹⁰⁷ , ¹⁰⁸ , ¹⁰⁹ , ¹¹⁰

The first link between GDF15 levels and cancer stem from the pioneering work of the Sam Breit laboratory ²⁸ upon discovery that chronic exposure to GDF15 caused anorexia and cachexia in experimental models, and that GDF15 levels and cancer‐associated weight loss in patients with advanced prostate cancer could be directly related. Importantly, the investigators found that tumor‐derived GDF15 effects on feeding, weight, lean and fat mass could be reversed by GDF15 monoclonal antibody treatment. ²⁸ These initial data have since been validated in 4 different xenografted murine models substantiating the role of GDF15 in the etiology of CACS. ¹¹¹ In subsequent screening, GDF15 was found to be the most prominently overexpressed circulating factor across various types of human cancers. ¹¹² Multiple groups have now linked substantially elevated serum levels of this cytokine to CACS and proposed GDF15 as a diagnostic biomarker for several types of malignancies. Rises in circulating GDF15 levels can reach up to 100‐fold and, importantly, often correlate with the severity of the anorexia/cachexia in both oncology patients and in rodent models, further stressing GDF15's clinical relevance as a mediator of CACS. ²⁸ , ⁵⁰ , ¹¹² , ¹¹³ , ¹¹⁴ , ¹¹⁵ The clinical literature in totality (see ¹¹⁶ for review) indicates that GDF15 is a promising prognostic marker for various type of cancers and importantly it could be to use to predict manifestation, progression, and clinical outcomes.

In 2017, prior to the discovery of the GDF15 receptor, we discovered that targeted specific lesion of the AP reduced anorexia, weight loss, and muscle wasting in a rat model for CACS. ¹¹³ Conversely, subdiaphragmatic vagal deafferentation (SDA) had no effect on these observations, providing one of the first neurobehavioral explanations/mechanism of GDF15 actions, which turned out to be consistent with a key role for GFRAL expression in the AP as the putative mechanism mediating GDF15 effects. ¹¹³ Thanks to the advancements in the GDF15 field, new studies have since employed transgenic lines and GFRAL or GDF15 antibodies to demonstrate the ability of GDF15/GFRAL inhibition to reverse body weight loss, excessive lipid oxidation, and importantly, to restore muscle function and physical performance in mouse models of cachexia. ¹¹⁷ , ¹¹⁸ Mechanistically, while increased food intake is the major factor responsible of the beneficial effects of GFRAL signaling inhibition, intake‐independent effects in muscle and adipose tissue also occurred, by modulating corticosterone levels (for muscles) and by sympathetic‐mediated effects (for adipose tissue). ¹¹⁸ , ¹¹⁹

In summary, these findings highlight the key role of GDF15 in the intricate etiology and pathophysiology of CACS. Several clinical trials (clinicaltrials.gov IDs: NCT05546476, NCT04815551, NCT05865535, NCT04068896, and NCT03392116) that leverage the GDF15/GFRAL system are currently ongoing and hold considerable potential as an effective intervention for CACS. Since both underlying disease and treatment therapeutics both produce nausea, emesis, and anorexia through the GDF15/GFRAL system, targeting this system offers the profound benefit of potentially mitigating both causes of these side effects.

6. HYPEREMESIS GRAVIDARUM AND GDF15: A SIGNIFICANT BREAKTHROUGH IN MECHANISTIC INSIGHT

Nausea and vomiting during pregnancy (NVP) affects ~70% of women with often debilitating consequences. ¹²⁰ It usually begins at approximately 6–8 weeks of gestation and generally diminishes in intensity by the end of the first trimester. Hyperemesis gravidarum (HG), a severe and highly dangerous emetic condition associated with pregnancy, has been reported to occur in up to 11% of pregnant women and it is characterized by nausea and emesis that are so severe that normal eating/drinking is negatively impacted and often leads to weight loss and electrolyte disturbances which carry significant risks to the longer‐term health of both mother and offspring. ¹²⁰ In fact, HG is the primary cause of hospitalization in the first trimester, and the second most common cause of pregnancy hospitalization overall in the United States. ¹²¹ Despite being a severe problem, only in the last few years have significant advances in the understanding of its underlying pathophysiological mechanism(s) been made. Traditional forms of interventions or management for HG encompass dietary supplementation (e.g., vitamin supplements), lifestyle adjustments (like consuming smaller and more frequent meals), and medical treatments (e.g., metoclopramide, ondansetron). ¹²⁰ The latter is used by ~20% of pregnant women in the United States. ¹²² While demonstrating some efficacy in alleviating HG symptoms, these medications were not originally designed for HG and are considered off‐label for treating other types of nausea and emesis, making them suboptimal choices at best and likely unable to qualitatively improve the condition.

The classically dominant theory that human chorionic gonadotropin (hCG) is the central mediator of morning sickness and HG ¹²³ contrasts with emerging evidence suggesting GDF15 may be the primary driver. High levels of GDF15 in maternal blood in human pregnancy was first reported decades before discovering GDF15 involvement in CACS and CINV by Breit et al. ¹²⁴ In more recent years, GDF15 mRNA was found highly expressed in the human placenta. ¹²⁵ Furthermore, GDF15 proved to be one of the most abundant peptides secreted from human trophoblast organoids, a three‐dimensional ex vivo culture model that can be used to study various aspects of placental development and physiology. ¹²⁶ Foundational literature clearly demonstrates that concentrations of GDF15 in maternal circulation are higher in women experiencing vomiting in pregnancy and were positively correlated with antiemetic use during the second trimester. ¹²⁷ Work led by Marlena Fejzo, a key leader in the field, showed that serum concentrations of GDF15 at the end of the first trimester were significantly increased in women hospitalized for HG compared to women experiencing morning sickness. ¹²⁸ Interestingly, genetic variants associated with altered expression of GDF15 segregated with disease in families affected by HG and were associated with recurrence of HG in subsequent pregnancies. ¹²⁹ Importantly, a recent genome wide association study (GWAS) of >53,000 women, reported several independent variants in the region (Loci chr19p13.11) containing the GDF15 gene as the most highly associated single nucleotide polymorphism (SNP) in the maternal genome with morning sickness and HG. ²⁶ Powerful data recently published in late 2023 elegantly confirmed that fetally‐derived GDF15 accounts for the elevations in GDF15 concentrations in the maternal blood. ²¹ Additionally, the study suggests that exposure to high GDF15 levels prior to pregnancy may be protective against the subsequent development of morning sickness and HG. ²¹

The concept that GDF15 may have a primary role in the etiology of HG, rather than occur as a consequence of the condition, is supported by preclinical findings emerging in other fields. As stated above, GDF15 causes, through malaise, profound anorexia and weight loss in various animal models, and increased chronic GDF15 circulating levels are considered a key driver of cachexia, all symptoms shared with HG. Several questions are still unanswered. GDF15 levels remain elevated for the whole gestational period, however morning sickness usually decreases in the second and third trimester. It will be important to understand whether the “desensitization effect” of GDF15 proposed in Fejzo et al. (i.e., reduced food intake suppression upon chronic administration in mice) ²¹ can be replicated and expanded in other animal models (e.g., rats, shrews, ferrets, and non‐human primates) using other behavioral responses indicative of malaise (e.g., vomiting, conditioned aversive responses, pica behavior). It will also be necessary to examine potential changes in GFRAL expression levels during pregnancy and what role this shift may play, as this also remains to be studied. Strikingly, while several studies reported up to 200‐fold increases in circulating GDF15 in humans and non‐human primates, ¹²⁴ , ¹²⁷ , ¹²⁹ common laboratory rodents (i.e., rats and mice) only show a very marginal increase in GDF15 during gestation. ¹³⁰ The underlying evolutionary mechanism behind the potential different role of GDF15 in rodents and primates remains unsolved and further studies in other species should be conducted. It also remains to be investigated whether GDF15 signaling during pregnancy serves other roles important to the gestation process. Overall, the recent work highlighted by Fejzo and colleagues represents a profoundly impactful step forward in GDF15 and HG research and treatment potential.

7. MITOCHONDRIAL DYSFUNCTIONS AND CYCLIC VOMITING SYNDROME: SPECULATION ON GDF15

Idiopathic or multicausal conditions of emesis inherently require further research to understand their origins and mechanisms. Mitochondrial dysfunctions (MDs) encompass a heterogenous group of clinical entities characterized by greatly impaired mitochondrial function due to mutations in either the mitochondrial or nuclear genome that impair the function of mitochondrially expressed proteins leading to reduced ATP production, and increased oxidative stress. ¹³⁶ , ¹³⁷ Such mitochondrial dysfunction can compromise the normal functioning of various organs causing myopathy, neurological disorders, developmental delays and many other symptoms including impaired gut motility, nausea and vomiting. ¹³⁶ , ¹³⁸ , ¹³⁹ GDF15 has recently been considered as a biomarker for MDs, as researchers found GDF15 mRNA levels to be increased up to 150‐fold in muscles from patients suffering from a myopathic mitochondrial DNA depletion syndrome, and GDF15 was constitutively secreted by skeletal muscle cells. ¹⁴¹ Additionally, significantly elevated circulating GDF15 was found in patients with a genetic diagnosis of mitochondrial disease. ¹⁴¹ Montero et al. examined children with mitochondrial disease and healthy controls ¹⁴¹ and found that children with mitochondrial diseases have an increase of GDF15 plasma levels by 11‐fold when compared to healthy controls. ¹⁴¹ Similarly, GDF15 concentration was 6‐fold higher in adult MD patients compared to healthy controls, and GDF15 was the most useful biomarker and predictor for MDs severity for MDs examined. ¹⁴² In summary, in patients with different forms of mitochondrial DNA depletion syndrome, GDF15 levels are elevated and seem to correlate with clinical and histopathological markers of disease severity. ¹⁴¹ , ¹⁴² Perhaps importantly, mitochondrial stress activates the integrated stress response pathway (ISR) and it is plausible that eIF2α mechanisms contributs to the elevated GDF15 levels observed in mitochondrial disorders. ⁵⁷ Since GI manifestations of MDs often include anorexia, vomiting and nausea, further studies of the impact of GDF15 on specific models of genetically mediated mitochondrial diseases are needed to determine whether treatments directed at the GDF15/ GFRAL system hold value in these largely currently untreatable group of diseases.

Cyclic vomiting syndrome (CVS) is a disorder comprised of severe and recurrent episodes of unexplained vomiting interspersed with periods without symptoms. ¹³¹ Currently defined as a disorder of brain‐gut interaction, CVS is a diagnosis of exclusion and, despite its name, is not always associated with a specific cyclic/recurring pattern. ¹³² CVS often presents itself first in childhood, and while many children recover once they reach adulthood, the duration can last from months to years. ¹³³ Given the lack of specific diagnostic tools, assessing its true prevalence is challenging; however, it is estimated to be between 1% and 2% in children and there has been a recent rise in diagnosed adults. ¹³⁴ The exact causes of this perplexing condition remain elusive and current and future research into CVS requires uncovering the neuroendocrine causes and substrates poorly understood. Given the emergence of GDF15‐GFRAL as a novel substrate in nausea and emetic circuits, exploring the potential role of GDF15 in CVS has great merit. GDF15 is well‐documented to be induced by mitochondrial dysfunction, and similarly, mitochondrial stress may be present in CVS, ¹³¹ , ¹³⁵ although it remains speculative whether links between CVS and GDF15 exist. Children with CVS may display likely maternally inherited (e.g., mtDNA) gastrointestinal and neurological disorders with neuroendocrine pathways. ¹⁴⁰ Adults with CVS also have a higher probability of maternal inheritance of functional disorders (12% vs. 1% controls). ¹⁴⁰ In a comprehensive study examining the entire exome/genome for its relevance to CVS, the authors identified 22 genes as highly/likely related to CVS. Notably, all 22 of these genes are either directly or indirectly involved in cation transport or energy metabolism, and several are specifically linked to mitochondrial function, ¹³¹ thus highlighting the intricate relationship between cellular energy metabolism and gastrointestinal function.

Mitochondrial dysfunctions (MDs) encompass a heterogenous group of clinical entities characterized by greatly impaired mitochondrial function due to mutations in either the mitochondrial or nuclear genome that impair the function of mitochondrially expressed proteins leading to reduced ATP production, and increased oxidative stress. ¹³⁶ , ¹³⁷ Such mitochondrial dysfunction can compromise the normal functioning of various organs causing myopathy, neurological disorders, developmental delays and many other symptoms including impaired gut motility, nausea and vomiting. ¹³⁶ , ¹³⁸ , ¹³⁹ GDF15 has recently been considered as a biomarker for MDs, as researchers found GDF15 mRNA levels to be increased up to 150‐fold in muscles from patients suffering from a myopathic mitochondrial DNA depletion syndrome, and GDF15 was constitutively secreted by skeletal muscle cells. ¹⁴¹ Additionally, significantly elevated circulating GDF15 was found in patients with a genetic diagnosis of mitochondrial disease. ¹⁴¹ Montero et al. examined children with mitochondrial disease and healthy controls ¹⁴¹ and found that children with mitochondrial diseases have an increase of GDF15 plasma levels by 11‐fold when compared to healthy controls. ¹⁴¹ Similarly, GDF15 concentration was 6‐fold higher in adult MD patients compared to healthy controls, and GDF15 was the most useful biomarker and predictor for MDs severity for MDs examined. ¹⁴² In summary, in patients with different forms of mitochondrial DNA depletion syndrome, GDF15 levels are elevated and seem to correlate with clinical and histopathological markers of disease severity. ¹⁴¹ , ¹⁴² Perhaps importantly, mitochondrial stress activates the integrated stress response pathway (ISR). The ISR, via alpha subunit of eukaryotic translation initiation factor 2 (eIF2α), may be of critical importance to increase GDF15 expression. ⁵⁷ It is plausible that this mechanism contributes to the elevated GDF15 levels observed in mitochondrial disorders. Since GI manifestations of MDs often include anorexia, vomiting and nausea, further studies of the impact of GDF15 on specific models of genetically mediated mitochondrial diseases are needed to determine whether treatments directed at the GDF15/GFRAL system hold value in these largely currently untreatable group of diseases.

In summary, while the association between GDF15 levels and MDs has been validated, the link between GDF15 and CVS remain speculative. However, given the presumed involvement of mitochondrial stress/dysfunction in the etiology of CVS, the potential role of GDF15 in the pathophysiology of CVS may further underscore the involvement of mitochondrial‐derived stress response mechanisms in the development and progression of this enigmatic disorder. Further research is warranted to elucidate the precise mechanisms underlying the link between MDs, GDF15, and CVS.

8. CONCLUDING REMARKS

Nausea and emesis are ubiquitously present alongside polymorbidities contributing to detrimental life‐threatening outcomes, such as poor nutrition, lower quality of life, and unfavorable patient prognosis. Our collective understanding of the GDF15/GFRAL system has enormously expanded in the last decade. While the mechanisms and roles of the GDF15/GFRAL system in evolution remain partially elusive, it is evident that GFRAL activation causes anorexia and body weight loss through the induction of malaise, with minimal impact on homeostatic control of feeding and energy metabolism. These mechanistic and behavioral insights are crucial for deepening our understanding of a variety of medical conditions characterized by cachexia and malaise concomitant with elevated circulating GDF15 levels. It is our hope that these advancements in the field will pave the way for effective GDF15/GFRAL‐based therapies in the coming decade. Such treatments could potentially alleviate conditions like CINV, morning sickness, HG, and other disorders associated with elevated GDF15 levels.

9. AUTHOR CONTRIBUTIONS

T.B and A.M.P. prepared the manuscript with inputs from B.C.D.J. All authors edited and approved the revised version of the manuscript.

FUNDING INFORMATION

This work was supported by DK130239 (B.C.D.J.).

CONFLICT OF INTEREST STATEMENT

TB receives research funding from Pfizer, Inc. BCD receives research funding from Gila Therapeutics, Boehringer Ingelheim, and Eli Lilly, Inc.

Borner T, Pataro AM, De Jonghe BC. Central mechanisms of emesis: A role for GDF15. Neurogastroenterology & Motility. 2025;37:e14886. doi: 10.1111/nmo.14886

DATA AVAILABILITY STATEMENT

Research data are not shared.