The Role of Erythropoietin in Metabolic Regulation

Erythropoietin (EPO) is a crucial regulator of erythroid progenitor cells, promoting their growth, survival, and differentiation, and acts by binding to the cell surface erythropoietin receptor (EPOR) to activate downstream signaling pathways. While the primary role of EPO is to regulate red blood cell production, increasing attention has been given to the broader functions of the EPO-EPOR axis, particularly in non-hematopoietic tissues [ 1 , 2 , 3 ]. Studies have shown that transgenic mice overexpressing human EPO (Tg6) exhibit significantly reduced body weight [ 4 ]. In contrast, mice lacking EPOR in adipose tissue display increased body weight and fat mass compared to wild-type controls [ 5 , 6 ]. This suggests that EPO-EPOR signaling influences glucose and fat metabolism. Enhanced EPO signaling in Tg6 mice correlates with improved metabolic regulation, while loss of EPO signaling in non-hematopoietic tissues, as seen in ΔEPORE mice with erythroid restricted EPOR, impairs these processes [ 6 , 7 ]. Recent findings revealed that EPO impacts adipose tissue, skeletal muscle, and the liver by regulating the expression of the transcription factor RUNX1 through EPOR signaling [ 6 ]. RUNX1 is a key regulatory protein in vertebrates, known for its role in controlling hematopoiesis and angiogenesis [ 8 ]. Beyond hematopoietic functions, RUNX1 participates in regulating the proliferation of mouse epithelial hair follicle stem cells, human epithelial cancer cells, and lipid metabolism. RUNX1 can drive lipid composition changes in hair follicle stem cells by inhibiting lipase activity, thereby reducing the proliferation of mouse epithelial cells and human squamous cell carcinomas [ 9 ]. Despite these advances, the precise mechanisms by which EPO-EPOR signaling modulates glucose and lipid metabolism via RUNX1 remain unclear. Understanding this pathway is critical for identifying therapeutic opportunities. This review explores the tissue-protective roles of EPO and its functions in lipid metabolism regulation. We also highlight the emerging EPO-EPOR-RUNX1 axis in metabolic disorders, such as obesity, diabetes, and related conditions. A non-erythropoietic EPOR agonist ARA290 activates EPOR signaling and associated energy-sensing networks providing further evidence of metabolic response to EPO stimulation in non-erythroid tissue. Therefore, in addition to its requisite role in determining erythroid response, endogenous and exogenous EPO-EPOR signaling contributes importantly to energy homeostasis.

Animal models have been instrumental in uncovering the roles of EPO-EPOR signaling in both erythroid and non-erythroid tissues. Studies using genetically modified mice with targeted deletions of EPO or EPOR demonstrated that while these factors are not essential for primitive erythropoiesis, they are critical for definitive fetal liver and adult erythropoiesis. Mature red blood cell formation requires EPO and EPOR for the survival, proliferation, and differentiation of maturing erythroid progenitor cells [ 20 , 21 ]. In adults, EPO, produced by interstitial cells in the kidney, is inducible under hypoxic conditions and increases in response to anemia, high altitude, or ischemic stress [ 22 , 23 , 24 ]. Additionally, animal models have revealed that EPOR expression extends beyond hematopoietic and erythroid cells to select non-hematopoietic tissues. This broader distribution of EPOR expression enables EPO to exert diverse physiological effects, including responses to ischemic and traumatic brain injury, neurocognitive recovery, wound healing, bone remodeling, and the regulation of metabolic homeostasis ( Figure 1 ) [ 1 , 3 , 25 ]. Expression of EPOR and EPO production in mouse brain indicates EPO activity on the other side of the blood–brain barrier that contributes to neural cell proliferation, viability, and protection against ischemia and glutamate damage [ 26 , 27 , 28 , 29 , 30 ]. In rodent models of diabetes, EPO treatment contributed to improved cognitive dysfunction associated with hippocampal neurodegeneration [ 31 , 32 ]. Potential benefit in the nervous system of EPO treatment is suggested by clinical trials in patients with ischemic stroke and other neurologic disorders [ 33 , 34 , 35 ]. Long-term darbepoetin-α treatment in the ApoE-/- mouse model for atherosclerotic disease reduced oxidative stress that is characterized by lipid profile, inflammation, endothelial injury, lipid peroxidation, and protein oxidation [ 36 ]. During development, EPOR expression in non-hematopoietic tissue includes heart and brain and EPOR knockout mice exhibit brain and heart hypoplasia prior to death at embryonic day 13.5, adding further support for direct non-hematopoietic tissue response to EPO [ 1 , 37 , 38 , 39 ]. Fetal liver erythropoiesis is supported directly by EPO production in the fetal liver. At birth, EPO production shifts from the fetal liver to the adult kidney while hepatocytes retain the ability to produce EPO in a hypoxia-dependent manner, albeit at a much lower level than the kidney. Hepatocyte EPO production cannot compensate for the reduction in EPO production by the kidney in chronic kidney disease [ 40 ]. EPO directly impacts liver health in animal models. EPO treatment or increasing EPO signaling exhibit protective activity in animal models of liver ischemia-reperfusion injury associated with transplantation including bone marrow-derived mesenchymal stem cell transplantation in a mouse liver fibrosis model [ 41 , 42 , 43 , 44 ].

EPO treatment in mice increases hematocrit and improves glucose tolerance on both normal chow and high-fat diets, and in both male and female mice [ 3 ]. In contrast, EPO regulation of body weight and fat mass readily observed in male mice exhibits a sex-differential response related to the anti-obesity activity of estrogen in female mice [ 57 ]. The glucose intolerance observed in male and female mice lacking estrogen receptor α was improved by EPO treatment, independent of changes in body weight and fat mass [ 58 ]. In male mice, particularly those on a high-fat diet, EPO treatment increased WAT oxidative metabolism, fatty acid oxidation, brown fat-associated gene expression, mitochondrial content, and uncoupled respiration, mediated by peroxisome proliferator-activated receptor (PPAR) α and SIRT1 activation [ 5 ]. In contrast, mice with targeted deletion of EPOR in adipose tissue on a high-fat diet showed decreased glucose tolerance, increased body and fat mass, elevated serum leptin and insulin levels, and reduced expression of brown fat-associated genes, mitochondrial activity, and AKT activity that did not respond to EPO stimulation [ 5 ]. EPO administration in high-fat diet-fed mice improved glucose metabolism and insulin resistance, mediated in part by PPARγ and SIRT1 activity and the PI3K/AKT pathway in the liver [ 59 ]. Other tissues that contribute to EPO activity to improve obesity and glucose homeostasis in obese mice include brown adipose tissue (BAT), in which EPO increased interscapular fat mass, temperature, and secretion of FGF21, a metabolic regulator of glucose homeostasis and insulin sensitivity [ 60 ]. In young male mice, elevated EPO via transgenic expression or EPO treatment improved glucose tolerance and insulin sensitivity, along with changes in lipid metabolism-associated gene expression [ 6 ]. In a type 1-like diabetic mouse model, elevated EPO treatment improved blood glucose in a dose-dependent manner, increased insulin sensitivity, and reversed the decreased GLUT4 levels in skeletal muscle [ 61 ]. EPO treatment in male rats on a high-fat diet improved glucose tolerance without changing body weight or fat mass. Furthermore, EPO administration in a streptozotocin diabetic rat model improved glucose tolerance, insulin tolerance, pancreatic β-cell dysfunction, and increased antioxidant activity [ 62 , 63 ]. In a mouse model of non-alcoholic fatty liver disease, EPO administration reduced body weight, reversed glucose intolerance and insulin resistance, reduced lipid accumulation in the liver and WAT, suppressed lipid synthesis genes in the liver, and increased lipolysis proteins in adipose tissue [ 64 ]. In a female rat model of polycystic ovary syndrome with impaired glucose tolerance and elevated plasma insulin, EPO treatment improved glucose tolerance, decreased serum insulin levels, and reduced disease-associated reproductive manifestations, although to a lesser extent than metformin [ 65 ]. In domestic pigs, transgenic EPO expression lowered fasting glucose levels and improved glucose tolerance on both normal chow and high-fat diets [ 66 ]. While elevated brain EPO did not increase hematocrit due to protection by the blood–brain barrier, improvement in glucose metabolism was observed. Increased brain EPO in mice, via transgenic expression or intracerebral ventricular EPO pump, improved glucose tolerance in both male and female mice on normal chow and high-fat diets, independent of body weight or fat mass [ 3 ]. Hypothalamic ventricular administration of EPO in aged obese mice decreased food intake and obesity, improved metabolic function, and reversed impairments in glucose tolerance and insulin sensitivity without increasing hematocrit [ 67 ]. These animal models demonstrate that EPO treatment improves glucose metabolism independent of gender, changes in body weight or fat mass, and increased hematocrit from EPO-stimulated erythropoiesis.

Increased EPO signaling in WAT and adipocytes, mediated by elevated EPO binding to its receptor, promotes the expression of mitochondrial biogenesis genes, the key metabolic regulator PGC-1α, and brown fat-associated genes such as UCP1, and enhances oxygen consumption [ 5 , 6 ]. EPO stimulation also modulates lipid metabolism gene expression, promoting a lean phenotype by increasing the expression of lipolytic genes and suppressing the expression of lipogenic genes [ 6 ]. In the subcutaneous WAT (scWAT) of transgenic mice with high EPO expression and EPO-treated wild-type mice, increased EPO activity resulted in the upregulation of lipolytic genes, PPARγ and lipoprotein lipase (Lpl), along with downregulation of lipogenic genes such as acetyl-CoA carboxylase 1 (Acc1), acetyl-CoA carboxylase 2 (Acc2), Fas receptor (Fas), Lipin1, sterol regulatory element binding transcription factor 1 (Srebf1), and stearoyl-CoA desaturase 1 (Scd1). In epididymal WAT, transgenic mice with high EPO did not show increased expression of brown fat-associated genes; however, both transgenic mice with high EPO and EPO-treated mice showed increased expression of lipolytic genes and decreased expression of lipogenic genes. Similar changes in lipid metabolism gene expression were observed in corresponding BAT, skeletal muscle, and liver in mice with elevated EPO [ 6 ]. In contrast, in mice with EPOR restricted to erythroid tissue and in mice with targeted deletion of EPOR in adipose tissue, EPO treatment resulted in the expected increase in hematocrit but no change in lipid metabolism gene expression in subcutaneous or epididymal WAT, BAT, skeletal muscle, or liver. The lack of EPO regulation of lipid metabolism gene expression in these tissues suggests that the ability of EPO to regulate lipid metabolism and promote a lean phenotype is largely mediated by non-erythroid EPO activity, particularly by EPOR expression in WAT, and is independent of EPO-stimulated erythropoiesis [ 6 ].

As a transcription factor, RUNX1 is a key regulatory protein in vertebrates. It plays a crucial role in blood formation and vascular development. During embryogenesis, RUNX1 is involved in the emergence of hematopoietic sites, and RUNX1 is essential for the differentiation of bone marrow hematopoietic stem cells in adults [ 96 ]. Beyond its role in the blood system, RUNX1 also influences lipid metabolism. It can affect the proliferation of mouse skin epithelial cells and human skin and oral squamous cell carcinoma by modifying the activity of lipid enzymes such as Scd1 and Soat1 [ 9 ]. In lipid metabolism, RUNX1 regulates gene expression by binding to promoters of lipogenic and lipolytic genes that contain a RUNX1 binding motif ( Figure 2 ) [ 6 ]. In the liver, CBFA2/RUNX1 partner transcriptional co-repressor 3 (CBFA2T3) can be activated by PPARα, influencing glucose tolerance and insulin resistance in mice [ 97 ]. In mice, elevated EPO increases RUNX1 protein stability in scWAT through activation of EPO-EPOR signaling [ 6 ]. The 5’ flanking regions of lipolytic and lipogenic genes in scWAT that respond to EPO stimulation to promote a lean phenotype contain RUNX1 binding motifs. In reporter gene assays, these RUNX1 binding motifs from lipolytic genes exhibit RUNX1-dependent enhancer activity, while those from lipogenic genes exhibit RUNX1-dependent silencer activity [ 6 ]. Mice expressing high transgenic EPO treated with the RUNX1 inhibitor Ro5-3335 show a decrease in glucose tolerance, and the improvement in lipid metabolism gene expression associated with elevated EPO is no longer evident in subcutaneous or epididymal WAT, BAT, skeletal muscle, or liver [ 6 ]. Core binding factor β (CBFβ) partners with RUNX1 to enhance RUNX1 binding to DNA and transcriptional regulation [ 98 , 99 ]. RUNX1 stability is regulated by post-translational ubiquitylation, leading to rapid degradation through the ubiquitin-proteasome pathway [ 100 ]. E3 ubiquitin ligases are required for ubiquitination, and in scWAT, elevated EPO-EPOR signaling specifically decreases FBXW7 E3 ubiquitin ligase mRNA and protein, coincident with increased RUNX1 protein stability ( Figure 2 ) [ 6 ]. EPO-EPOR signaling affects both RUNX1 and the expression of CBFβ, although increasing CBFβ did not appear to protect RUNX1 from degradation by increased FBW7 [ 6 ]. FBW7 protein may regulate adipocyte metabolism via RUNX1 ubiquitylation to increase RUNX1 proteasome degradation, and EPO-EPOR signaling decreases FBW7 expression, resulting in increased RUNX1 stability and RUNX1 transcription activity. This ultimately regulates lipid metabolism by increasing lipolytic gene expression and decreasing lipogenic gene expression to promote a lean phenotype. EPO administration in mice increased oxidative metabolism, fatty acid oxidation, and lipolytic gene expression, and promoted mitochondrial content and uncoupled respiration in adipose tissue [ 5 , 6 ]. Increased expression of lipolytic genes increases triglyceride catabolism, breaking down stored fat into free fatty acids in adipose tissue and numerous nonadipose tissues, and promoting tissue lipid mobilization and utilization such as energy production for skeletal muscle, and affecting gene transcription, the cell cycle, and cell growth [ 101 ]. These activities associated with increased lipid metabolism likely contribute to EPO activity in non-erythroid tissue including enhanced wound healing in skeletal muscle [ 55 , 56 ]. Metabolism coregulators, peroxisome proliferator-activated receptor (PPAR) α and Sirt1, along with RUNX1 combine to mediate, in part, EPO regulation of metabolism, including gene expression associated with lipid metabolism [ 5 , 6 ]. Decreases in fat mass in mice treated with EPO are consistent with EPO-stimulated down-regulation of lipogenic genes that decrease expression of enzymes crucial for fatty acid production, reduce fatty acid synthesis, lipid production, and fat storage [ 6 , 7 ]. EPO-regulated reduction of lipogenic genes could potentially help reduce fat synthesis that is associated with high triglyceride levels in liver and adipose tissues in metabolic disease or obesity [ 102 ]. Together, EPO as a novel regulator of energy homeostasis including lipolytic and lipogenic gene expression provides a potential strategy to enhance metabolic health and protect against obesity.