The present review aimed to provide a comprehensive analysis of prolyl hydroxylase domain (PHD) enzymes, detailing their subtypes, localization, regulatory mechanisms and functional roles, as well as their involvement in erythropoiesis and the development of related pharmacological agents. Under hypoxic conditions, the body initiates a cascade of adaptive biological responses, numerous of which are mediated by transcriptional complexes of the hypoxia-inducible factor (HIF) family. The precise equilibrium among HIF-1α, HIF-2α and HIF-3α is essential for orchestrating the transcriptional regulation of genes associated with erythropoiesis, angiogenesis, vascular homeostasis, metabolic pathways, and cellular proliferation and survival. Simultaneously, HIF exerts indirect inhibitory effects on the transcription of genes linked to other biological processes (1,2). Functionally, the HIF transcription factor exists as a heterodimer consisting of an oxygen-sensitive α and β subunit. The α subunit interacts with the nuclear transporter of the constitutively active aromatic hydrocarbon receptor and its activity is suppressed under normoxic conditions by the dioxygenase family, which requires oxygen, Fe, and 2-oxoglutarate (2-OG) for enzymatic function (3).

In response to hypoxia, HIF-1α levels rise rapidly, playing a pivotal role in acute hypoxic adaptation, particularly in regulating erythropoietin (EPO) synthesis, while HIF-2α predominantly governs long-term hypoxic responses (1,4). Hydroxylation of two conserved prolyl residues within the HIF-α subunit by PHD1-3 facilitates its recognition by the von Hippel-Lindau (VHL) tumor suppressor protein E3 ubiquitin ligase complex, leading to polyubiquitination and subsequent proteasomal degradation. In humans, three distinct PHD isoenzymes (PHD1-3) and an asparaginyl hydroxylase, known as factor-inhibiting HIF (FIH), have been identified. Hydroxylation of asparagine residues within the C-terminal transactivation domain (C-TAD) of HIF-α directly inhibits its interaction with the co-activator p300, thereby preventing the transcriptional activation of its target genes (5,6). Moreover, post-translational modifications such as acetylation and phosphorylation have been reported to further regulate HIF-α activity, adding additional layers of complexity to its control mechanisms (6).

PHDs exhibit substrate specificity and differential tissue distribution, particularly in their hydroxylation of HIF proline residues. By modulating intracellular metabolism, reactive oxygen species (ROS) generation, iron (Fe), homeostasis, nitric oxide (NO) signaling, and redox balance, PHDs serve as critical regulators of HIF activity, influencing a broad spectrum of physiological processes under hypoxic conditions. Although inhibitors targeting the HIF/PHD axis have been successfully integrated into clinical practice, the development of HIF/PHD activators or functional restorers remains limited due to significant technical hurdles. To date, no studies have reported the successful discovery of HIF/PHD activators, highlighting a major gap in current research and a potential avenue for future therapeutic advancements.

Therefore, the present review aimed to comprehensively summarize the subtypes, localization, regulatory mechanisms and biological functions of PHDs, along with the research and development of related pharmacological agents and their role in erythropoiesis. Activators or functional restorers targeting the HIF/PHD pathway represent a promising avenue for therapeutic intervention, particularly in addressing conditions such as polycythemia and chronic mountain sickness. Advancing research in this area could pave the way for groundbreaking developments, offering novel strategies to enhance clinical management and improve patient outcomes.

PHD enzymes function as critical oxygen sensors, belonging to a family of oxygen-dependent enzymes responsible for facilitating the proteasomal degradation of HIF-1α under normoxic conditions (7). The hydroxylation of HIF-1α is highly dependent on molecular oxygen, making these enzymes essential regulators of cellular oxygen homeostasis. Currently, three PHD isoforms, PHD1, PHD2 and PHD3, have been identified, each exhibiting distinct oxygen-sensing properties (8). In vitro studies suggest a hierarchical hydroxylation efficiency among these isoforms, with PHD2 demonstrating the highest activity, followed by PHD3 and then PHD1.

PHDs exhibit both substrate specificity and tissue-specific expression in their hydroxylation of HIFs. PHD2 preferentially hydroxylates HIF-1α over HIF-2α and HIF-3α, whereas PHD1 and PHD3 display higher efficiency in hydroxylating HIF-2α. Their expression patterns also vary across tissues: PHD2 is predominantly found in adipose tissue, PHD1 is highly expressed in testicular cells, and PHD3 is primarily located in cardiac cells (8). In terms of subcellular localization, PHD2 is mainly cytoplasmic, PHD3 is predominantly nuclear, and PHD1 is distributed in both the cytoplasm and nucleus. Encoded by the EGLN1 gene, PHD2 serves as the principal oxygen sensor responsible for modulating HIF-2α activity (9,10).

Among the three PHD isoforms, PHD2 was initially identified as the most functionally significant hydroxylase in most cell types (11). Differences in their intracellular distribution have been elucidated through studies involving the overexpression of labeled proteins (12,13). Specifically, fusion protein analyses have revealed that GFP-tagged EGLN1/HPH-2/PHD2 localizes predominantly to the cytoplasm, whereas EGLN2/HPH-3/PHD1 is primarily nuclear. Conversely, EGLN3/HPH-1/PHD3 is distributed across both the nucleus and cytoplasm in a relatively balanced manner (13).

Notably, PHD expression has been implicated in inflammatory conditions such as ulcerative colitis (UC), a subtype of inflammatory bowel disease (IBD). In patients with UC, PHD1 and PHD2 are highly expressed in the lamina propria and colonic epithelium, while PHD3 is predominantly localized to the vascular endothelium. Moreover, PHD3 expression is significantly elevated in inflamed biopsy samples and is positively correlated with inflammatory cytokines such as IL-8 and TNF-α, as well as the apoptotic marker caspase-3 (14).

HIF proteins are heterodimeric transcription factors consisting of a tightly regulated α subunit (HIF-1α, HIF-2α, or HIF-3α) and a constitutively expressed β subunit. Under normoxic conditions, HIF-α subunits undergo oxygen-dependent hydroxylation by PHDs, allowing their recognition by the VHL tumor suppressor protein, a key component of an E3 ubiquitin ligase complex. This interaction marks HIF-α for ubiquitination and subsequent degradation via the proteasome pathway (15,16). However, under hypoxic conditions, hydroxylation is suppressed, preventing VHL-mediated degradation and enabling HIF-α to accumulate and exert its transcriptional effects (12).

Within mammalian cells, the functions of the PHD protein family (PHD1, PHD2 and PHD3) are diverse, with PHD2 emerging as the predominant oxygen sensor regulating HIF-1α stability and degradation. In addition to PHDs, FIH-1, an oxygen-dependent asparaginyl hydroxylase, plays a critical role in regulating the C-TAD of HIF. FIH-1 provides a direct mechanistic link between oxygen sensing and HIF-mediated transcription, further modulating cellular responses to hypoxia (17).

HIF-1α exhibits dynamic nucleocytoplasmic shuttling, whereas HIF-1β remains permanently localized within the nucleus. The VHL tumor suppressor protein (pVHL) is distributed across both the nuclear and cytoplasmic compartments. The nuclear translocation of PHD1 is mediated by classical nuclear localization signals (NLS), utilizing the importin α/β receptor pathway. By contrast, PHD2 nuclear import relies on a non-classical NLS, following an alternative import pathway, which results in PHD2-mediated hydroxylation of HIF-1α predominantly occurring within the nucleus. The nuclear export of PHD2 is facilitated by an N-terminal nuclear export signal and requires the export receptor chromosome region maintenance 1. Meanwhile, the nuclear import of PHD3 is regulated via the importin α/β receptor pathway and also depends on a non-classical NLS (18).

In the central nervous system, NG2 cells represent a fourth class of glial cells alongside astrocytes, microglia, and oligodendrocytes. HIF-2 activation in NG2 cells plays a pivotal role in promoting neurovascular expansion and remodeling independently of EPO. The regulatory control of HIF-2 activity in NG2 cells involves both PHD2 and PHD3. Under conditions where PHD2 and PHD3 are inactive, PHD1 assumes a compensatory role in regulating the HIF-2 transcriptional response, thereby facilitating vascular expansion and remodeling. However, the simultaneous inactivation of PHD1, PHD2 and PHD3 leads to robust HIF-2 activation, further enhancing neurovascular remodeling independent of EPO signaling (19).

The genetic inactivation of PHD1, PHD2 and PHD3 results in HIF activation, triggering the reprogramming of myofibroblast-transformed renal EPO-producing cells (MF REPs) to resume EPO secretion. Specifically, the loss of PHD2 in REPs restores EPO gene expression in damaged renal tissue, leading to erythrocytosis. By contrast, the simultaneous deletion of PHD1 and PHD3 prevents the suppression of EPO expression without inducing erythrocytosis (20). These findings underscore the predominant role of PHD2 in EPO regulation and red blood cell (RBC) production. Given the tissue-specific regulatory mechanisms of HIF by different PHD isoforms, their physiological roles vary significantly across organ systems. As research progresses, pharmacological agents targeting angiogenesis and EPO synthesis have been integrated into clinical practice for the treatment of related disorders. Therefore, further investigation into PHD enzyme functions holds substantial clinical significance.

The study of PHD enzyme localization primarily involves the fusion of PHD to the N-terminus of a fluorescent protein, followed by transient transfection of the fusion construct into human osteosarcoma (U2OS) cells. Subsequent visualization is performed using three-dimensional two-photon confocal fluorescence microscopy (13).

PHDs belong to the family of α-ketoglutarate (α-KG)-dependent dioxygenases and serve as key regulators of cellular oxygen sensing and metabolic control. They play a crucial role in maintaining oxygen homeostasis by mediating the hydroxylation of target proteins, thereby influencing their stability and function. A summary of the major PHD isoforms, subcellular localization, tissue distribution, substrates, and biological functions is provided in Table I.

The oxygen-dependent hydroxylation of HIF-α subunits serves as a key signal for ubiquitination and subsequent proteasomal degradation, playing a crucial role in regulating HIF protein abundance and maintaining oxygen homeostasis (21–25). All three PHD isoforms (PHD1-3) are capable of hydroxylating HIF-α subunits, although their functional significance varies (12,26). Among them, genetic studies have highlighted PHD2 as the most critical regulator of HIF-1α stability. Notably, PHD2 expression is upregulated under hypoxic conditions, establishing an HIF-1-mediated autoregulatory feedback loop that fine-tunes oxygen-dependent responses (11).

Beyond their role in HIF-α hydroxylation, PHDs can also modulate HIF activity through hydroxylase-independent pathways. While these enzymes primarily control the stability of HIF-α proteins, PHD2 and PHD3 themselves are subject to upregulation via HIF-dependent feedback mechanisms. Functionally, different PHD isoforms contribute to distinct physiological and pathological processes, including angiogenesis, erythropoiesis, tumorigenesis, cell proliferation, differentiation and survival. As a result, disruptions in PHD expression or function can lead to diverse biological consequences, reflecting the unique regulatory roles of each isoform (27). The selectivity of PHD isoforms toward different HIF-α subunits exhibits a degree of specificity. PHD2 is preferentially induced by HIF-1α, whereas PHD3 responds to both HIF-1α and HIF-2α (28). Each PHD isoform regulates the HIF system in a functionally distinct, non-redundant manner. Under specific experimental conditions, the relative contribution of each PHD enzyme depends significantly on its expression levels, with the hydroxylation rate being directly correlated with the abundance of active PHD proteins (29). The oxygen-dependent hydroxylation of specific proline residues in HIF-1α allows recognition by the VHL tumor suppressor complex, a key component of the E3 ubiquitin ligase system. This interaction leads to the polyubiquitination of HIF-1α, triggering its rapid degradation via the ubiquitin-proteasome pathway (26,30–32).

All three PHD isoforms contribute to HIF regulation, and their cell-type specificity and inducibility provide HIF with the flexibility to fine-tune hypoxia responses across different tissues. The relative specificity of pharmacological PHD inhibition presents an opportunity for selective modulation of the HIF system, which holds considerable clinical significance. For instance, inhibiting PHD2 broadly enhances HIF activation across multiple cell types, even under normoxic conditions. By contrast, selective inhibition of PHD3 predominantly amplifies the hypoxic response in tissues where this enzyme is highly expressed (29). This approach could be leveraged to activate the HIF system for treating ischemic or hypoxic diseases. Future research must further elucidate the in vivo specificity of PHD effects on different HIF-α subunits, as well as the distinct roles of PHD isoforms in various biological processes. A deeper understanding of these mechanisms is essential for minimizing adverse effects associated with PHD inhibitors and optimizing their therapeutic efficacy.

When PHD function is inhibited, the oxygen-dependent degradation of VHL and HIFα is disrupted. Similar to PHDs, the HIF-associated regulatory factor FIH, an asparaginyl hydroxylase, suppresses the transactivation function of HIFα by hydroxylating its TAD, thereby inhibiting the p300/CBP signaling pathway, which plays a crucial role in cell cycle regulation and apoptosis. Although the pVHL/PHD/HIF axis is one of the most extensively studied oxygen-sensing mechanisms, numerous aspects remain to be elucidated (33). The mechanistic action of the pVHL/PHD/HIF pathway, the alterations in HIF2α-ARNT dimerization under hypoxia, and the downstream targets of HIF are illustrated in Fig. 1.

PHD1 and PHD2 are large enzymes, each consisting of >400 amino acids (407 and 426, respectively, in humans), with a highly conserved hydroxylase domain located in their C-terminal region, sharing ~55% sequence identity in this region. By contrast, their N-terminal regions exhibit greater sequence diversity and comparatively lower activity (26). PHD3, the shortest of the three isoforms, contains only 239 amino acids in humans. Although it retains a hydroxylase domain, its N-terminal region is markedly distinct, comprising only a short, unique segment.

Transition metal ions, particularly Fe, play a pivotal role in the hypoxia-regulated PHD-HIF-EPO signaling axis, which governs erythropoiesis, angiogenesis, anaerobic metabolism, adaptation, cell survival and proliferation, thereby maintaining cellular and systemic homeostasis (43). The regulation of HIF is primarily mediated by the highly conserved EGLN/PHD family, which hydroxylates HIF in an oxygen-dependent manner. This hydroxylation facilitates recognition by VHL family proteins, leading to ubiquitination and proteasomal degradation of HIF. Interestingly, PHD3 has been found to enhance HIF signaling through the hydroxylation of the glycolytic enzyme pyruvate kinase muscle subtype 2 (PKM2). Beyond its role in the HIF pathway, PHD also influences synaptic transmission by modulating α-amino-3-hydroxy-5-methyl-4-isoxazolepropanoic acid (AMPA) receptor trafficking and regulating transient receptor potential cation channel member A1 (TRPA1) activity in response to oxygen levels in sensory neurons. Moreover, PHD activation is modulated by poly(rC)-binding protein 1 (PCBP1), which functions as an Fe chaperone, as well as by the (R)-enantiomer of 2-hydroxyglutarate (2-HG), highlighting the intricate interplay between multiple regulatory pathways in PHD function (8).

The hydroxylation of HIF-α proline residues is catalyzed by Fe- and 2-OG-dependent dioxygenase enzymes known as PHDs, whose activity is strictly oxygen-dependent. As a result, inhibiting PHD function leads to the stabilization of HIF-α, thereby activating hypoxia-associated signaling pathways even under normoxic conditions. This mechanism has been leveraged for therapeutic purposes, with small-molecule PHD inhibitors developed as clinical treatments for renal anemia. Beyond their application in anemia management, recent research has highlighted the broader medical potential of PHD inhibitors in various disease contexts. Emerging evidence suggests that modulating PHD activity may have therapeutic implications for ischemic disorders, chronic inflammatory diseases, metabolic syndromes and even neurodegenerative conditions. These findings underscore the expanding role of PHD-targeted therapies and warrant further investigation into their diverse clinical applications (70).

PHD inhibitors have also been reported in the treatment of various other diseases (70). There is increasing evidence supporting the potential of targeting the HIF pathway in acute myeloid leukemia (AML) therapy. Studies suggest that the selective PHD2 inhibitor IOX5 exerts its effects by stabilizing HIF-1α, thereby triggering a cascade of events detrimental to AML cells. By stabilizing HIF-1α, IOX5 disrupts pro-leukemogenic signaling pathways, shifts metabolic processes to a state less favorable for cancer cells and induces apoptosis through BNIP3 upregulation, highlighting its potential as a non-toxic therapeutic strategy for AML. Furthermore, when combined with BCL-2 inhibitor venetoclax, which is already used in clinical settings, IOX5 enhances anti-AML efficacy. Notably, IOX5 selectively targets PHD without interfering with other enzymes such as FIH-1, a key feature for minimizing off-target effects (71).

Roxadustat, another PHD inhibitor, has shown promise in alleviating anemia in patients with lower-risk myelodysplastic syndrome, reducing their dependence on RBC transfusions. Additionally, roxadustat has been identified as a novel therapeutic candidate for Fe-refractory Fe-deficiency anemia (IRIDA). In mouse models of IRIDA, it activates the HIF-2α-ferroportin (FPN) axis, demonstrating significant therapeutic efficacy and clinical translation potential. Mechanistically, roxadustat stabilizes HIF-2α in the duodenum, leading to FPN transcriptional activation and increased intestinal Fe absorption, independent of hepcidin levels, ultimately ameliorating hepcidin-activated anemia (72).

Moreover, in ischemia/reperfusion injury models, roxadustat preconditioning has been shown to induce ischemic tolerance by shifting aerobic respiration to anaerobic metabolism, thereby maintaining ATP production under hypoxic conditions and reducing myocardial ischemic damage (73). Additionally, roxadustat activates the HIF-1α/VEGF/VEGFR2 pathway, promoting angiogenesis, and has demonstrated therapeutic potential in diabetic wound healing in rat models (74). Furthermore, roxadustat induces the accumulation of HIF-1α and WNT7a in the tibialis anterior muscle, facilitating muscle regeneration following cyclophosphamide-induced muscle injury and promoting the formation of significantly larger muscle fibers (75). These findings suggest that roxadustat could be a promising drug for diabetic wound healing and the prevention of age-related skeletal muscle atrophy, though clinical validation is still required.

Given the broad biological functions of HIF, roxadustat's ability to stabilize HIF has opened new avenues for treating hypoxia-related diseases, including neuroprotection in nerve injury (76), oxygen-induced retinopathy (77), pulmonary fibrosis (78), acute lung injury (79), fracture healing (80) and radiation-induced apoptosis (81).

The therapeutic development of HIF-PHD modulators has primarily focused on inhibitors designed to elevate hemoglobin (Hb) levels, particularly for the treatment of renal anemia. To date, no studies have reported on optimal recovery doses or the development of PHD activators.

A comprehensive meta-analysis encompassing 30 studies and a total of 13,146 patients has assessed the long-term efficacy and safety of HIF-PHD inhibitors, including roxadustat, daprodustat, vadadustat, molidustat, desidustat and enarodustat, in the management of anemia associated with CKD. The findings demonstrate that HIF-PHD inhibitors significantly enhance Hb levels, total Fe-binding capacity, and transferrin concentrations compared with placebo or erythropoiesis-stimulating agent (ESA) treatments. Additionally, these inhibitors contribute to a reduction in cholesterol levels, indicating potential metabolic benefits.

Regarding safety, patients receiving HIF-PHD inhibitors exhibit a higher incidence of serious adverse events compared with those in the placebo group. However, the overall risk profile is comparable to that observed with ESA therapy. Adverse effects commonly associated with HIF-PHD inhibitors include diarrhea, nausea, peripheral edema, hyperkalemia and hypertension, with a higher likelihood of vomiting, headaches and thrombotic events when compared with ESA treatment. Despite these risks, HIF-PHD inhibitors remain effective in elevating Hb levels, optimizing Fe metabolism, and demonstrating favorable long-term tolerability in CKD-associated anemia. To mitigate adverse effects related to excessive Fe utilization, it is recommended that HIF-PHD inhibitors be administered in conjunction with Fe supplementation for extended treatment regimens (82).

The research studies on HIF-PHD inhibitors for renal anemia are as follows [Table II;(83–94)]: Beyond clinical trials, investigations have explored the pharmacokinetic interactions of HIF-PHD inhibitors both in vitro and in healthy volunteers. In vitro studies have demonstrated that the formation of chelates between vadadustat and Fe-containing compounds varies depending on the specific type of Fe reagent in aqueous and simulated intestinal fluid environments. Notably, when vadadustat is co-administered with oral Fe supplements, its area under the plasma concentration-time curve (AUC_0-∞) and maximum plasma concentration (C_max) are reduced due to gastrointestinal chelation. This interaction suggests that Fe supplements may impair vadadustat absorption, necessitating a dosing interval to optimize its therapeutic efficacy (95).

Additionally, pharmacokinetic evaluations in healthy volunteers have assessed the interaction between roxadustat and warfarin. The combination has been found to be well-tolerated, with only mild treatment-related adverse events. Importantly, co-administration does not necessitate warfarin dose adjustments, indicating that roxadustat does not significantly alter warfarin metabolism or anticoagulant activity (96).

The potential impact of HIF-PHIs on tumor development and progression has been a subject of significant concern, particularly given their close association with VHL disease. VHL disease is a syndrome characterized by detectable genetic abnormalities in the VHL gene, frequently associated with tumor formation (97). HIF regulates the transcription of numerous target genes, including VEGF, a key factor that promotes tumor growth. Additionally, two Phase 3 clinical trials conducted in China have reported a higher incidence of hyperkalemia in the roxadustat group. However, further analysis of potassium levels measured by central laboratories does not confirm an increased risk of hyperkalemia in patients treated with Roxadustat (88,98).

In the TREAT trial, patients in the darbepoetin alfa group (target Hb 130 g/l) exhibit a higher risk of fatal or non-fatal stroke, venous thromboembolism, and arterial thromboembolism compared with the placebo group (where darbepoetin alfa is only administered when Hb levels fell below 90 g/l) (99). Furthermore, in 2019, the first reported case of roxadustat-induced pulmonary arterial hypertension is documented (100).

Beyond its role as an HIF prolyl hydroxylase inhibitor, roxadustat has been suggested to exert HIF-independent effects, indicating the presence of potential off-target activities. Notably, it may increase the risk of pulmonary arterial hypertension and vascular calcification, as well as exacerbate inflammation and infection (101). Although no significant off-target effects have been observed in clinical trials or real-world applications thus far, HIF regulates a vast array of genes and has pleiotropic effects. Therefore, strict control of HIF activation levels and duration is essential, and careful consideration of roxadustat's dosage and treatment regimen is warranted to minimize unintended side effects resulting from excessive HIF activation.

Human survival depends on a continuous and adequate oxygen supply to meet cellular metabolic demands, primarily through oxidative phosphorylation, which generates ATP. HIFs play a central role in regulating gene transcription to maintain oxygen homeostasis by balancing oxygen supply and demand. The activity of HIFs is tightly regulated through oxygen-dependent hydroxylation, primarily mediated by PHD proteins and VHL proteins. Mutations in VHL, HIF-2α and PHD2 genes can lead to hereditary polycythemia, a condition characterized by excessive RBC production and PHA due to aberrantly elevated HIF activity. Furthermore, genetic adaptations involving variations in PHD2 and HIF-2 contribute to high-altitude acclimatization by reducing erythropoiesis and pulmonary vascular reactivity to hypoxia, enabling improved survival in low-oxygen environments (102).

Population genomic studies have identified EPAS1 (HIF-2α) and EGLN1 (PHD2) as the primary genes responsible for high-altitude adaptation. These genes exhibit strong associations with lower Hb levels in the Tibetan population, suggesting an evolutionary advantage in hypoxic environments (103). PHD2, encoded by EGLN1, functions as a critical oxygen sensor, facilitating HIF-α hydroxylation and subsequent degradation under normoxic conditions. In addition to its catalytic domain, PHD2 contains a highly conserved zinc finger domain, which independently interacts with HIF-α in vitro. A C36S/C42S EGLN1 knockout mutation, which disrupts the zinc finger function, leads to increased EPO gene expression, excessive erythropoiesis, and an enhanced hypoxic ventilatory response in vivo. These physiological effects are attributed to loss-of-function mutations in EGLN1, resulting in prolonged HIF stabilization and activation (104).

A gain-of-function mutation in EGLN1 (PHD2 D4E: C127S), along with polymorphisms in EPAS1 (HIF-2α), contributes to reduced Hb levels among Tibetan highlanders. While the influence of EGLN1 haplotypes on Hb concentration varies with age in low-altitude populations, Tibetan highlanders consistently exhibit lower Hb levels due to adaptation to the EPAS1 rs142764723 C/C allele. The high-altitude-adapted EGLN1 haplotypes, c.12C>G and c.380G>C, are unique genetic variations found in Tibetans, conferring resistance to polycythemia while enabling survival in chronic hypoxia. Moreover, EPAS1 haplotypes have been incorporated into the Denisovan genome, suggesting that these genetic modifications play a critical role in shaping the lower Hb concentration observed in Tibetans living on the plateau (105).

The EPAS1 variant prevalent in the Tibetan population has been shown to reduce gene expression in human umbilical endothelial cells and the placenta. Additionally, mice with heterozygous EPAS1 knockout exhibit a delayed physiological response to prolonged hypoxia. This Tibetan EPAS1 variant is associated not only with lower Hb levels but also with a diminished pulmonary vasoconstriction response. These findings suggest that EPAS1 downregulation serves as the molecular basis for Tibetan adaptation to high-altitude hypoxia, minimizing the risk of chronic mountain sickness while optimizing oxygen utilization in a low-oxygen environment (106).

PHD2 serves as a key regulator of HIFs, mediating their degradation and thereby controlling physiological responses to hypoxia, including RBC production. The PHD2 p.[Asp4Glu; Cys127Ser] variant exhibits lower oxygen Km values, which enhances the degradation of HIFs even under low-oxygen conditions. In wild-type erythroid progenitor cells, hypoxia typically stimulates proliferation; however, in patients carrying the EGLN1 c.[12C>G; 380G>C] mutation, progenitor cell proliferation is significantly suppressed under hypoxic culture conditions. This mutation, originating from a common haplotype ~8,000 years ago, has been linked to high-altitude adaptation. Specifically, the EGLN1 c.[12C>G; 380G>C] variant prevents the hypoxia-induced, HIF-mediated increase in erythropoiesis, thereby acting as a molecular defense mechanism against high-altitude erythrocytosis in Tibetans. This genetic adaptation plays a crucial role in maintaining normal erythrocyte levels in high-altitude populations, allowing them to thrive in oxygen-deprived environments (107).

Whole-genome sequencing of 10 Tibetan patients with high-altitude polycythemia (HAPC) has identified two significant genetic loci associated with erythrocytosis: the egln3/phd3 homolog (14q13.1, rs1346902) and the PPP1R2P1 (protein phosphatase 1 regulatory inhibitor subunit 2) gene (6p21.32, rs521539). The egln3/phd3b gene regulates HIF-α stability, while PPP1R2P1 is involved in ROS homeostasis and oxidative stress regulation (108).

Furthermore, the zinc finger domain of PHD2 binds specifically to a Pro-Xaa-Leu-Glu (PXLE) motif found in the ribosomal chaperone nascent polypeptide complex-α (NACA). This interaction facilitates the recruitment of PHD2 to the translation machinery, enabling co-translational hydroxylation of HIF-α. Notably, this co-translational modification is enhanced by the presence of a translational pause sequence within HIF-α, which increases PHD2's efficiency in targeting the protein. Mice carrying a mutation in Naca that eliminates the PXLE motif exhibit erythrocytosis, highlighting the functional importance of this co-translational regulatory mechanism in RBC production (109).

Based on the aforementioned theoretical framework, HIF stabilization may be detrimental in various disease conditions, including certain cancers and PHA. Consequently, pharmacological strategies to inhibit HIFs have been proposed. Early pharmacological interventions have focused on suppressing HIF expression or promoting HIF degradation. More recently, small-molecule inhibitors such as PT2977 have been developed to block the formation of transcriptionally active HIF-2α/HIF-1β heterodimers. Depending on the specific HIFα isoform being targeted, pharmacological interventions can be classified into non-specific inhibitors targeting both HIF-1α and HIF-2α or isoform-selective inhibitors specifically targeting either HIF-1α or HIF-2α.

Non-specific HIF inhibitors include 2-methoxyestradiol, digoxin, and berberine acetate. However, these compounds have faced significant limitations in clinical development. 2-Methoxyestradiol, an endogenous estrogen metabolite, has a limited clinical window; digoxin requires HIF inhibition at toxic concentrations due to its narrow therapeutic index; and berberine acetate has potential side effects and a short half-life, preventing further clinical progress (110).

Selective HIF-1α inhibitors include topotecan (NSC-609699), PX-478 (S-2-amino-3-[4′-N,N-bis(chloroethyl)amino] phenylpropionic acid N-oxide dihydrochloride), cyclic peptide inhibitors (cyc-CLLFVY), and locked nucleic acid antisense oligonucleotides such as EZN-2968/RO7070179. While these compounds demonstrate potential therapeutic benefits, further research is required to establish their safety and efficacy in humans (110).

Selective HIF-2α inhibitors include C76 (methyl 3-{2-[cyano(methylsulfonyl)methylene]hydrazinyl}thiophene-2-carboxylate), and PT2385, the first clinical HIF-2α inhibitor. Later, the FDA approved PT2977 (belzutifan, MK-6482) for the treatment of ccRCC in patients with VHL disease. Additionally, several HIF-2α inhibitors from Nikang (NKT2152), Novartis (DFF332), and Arcus (AB521) are in various stages of clinical development (110).

Given that the EPAS1A529V mutation is a key driver of Pacak-Zhuang syndrome, belzutifan has been investigated as a potential therapeutic option for a single patient with severe polycythemia, hypertension and multiple PGL (110). Targeting HIF-2α may offer therapeutic benefits for PHA, retinal neovascularization and polycythemia.

While significant advancements have been made in HIF inhibition, research on enhancing PHD enzyme activity to counteract HIF accumulation remains scarce. This area presents challenges similar to small-molecule p53 reactivators, which are further complicated by the diverse isoforms and subcellular localization of PHDs. One promising future direction involves RNA-based therapies, including siRNA and gene editing technologies, alongside the development of targeted delivery mechanisms. These approaches could enable precise modulation of HIF activity while minimizing off-target effects on unrelated tissues or cell types, providing a novel avenue for HIF activation or inhibition in disease therapy.

The link between polycythemia and high-altitude adaptation was first reported over a century ago, with subsequent discoveries highlighting the central role of HIF-1 in oxygen homeostasis. As a heterodimeric transcription factor composed of an α and β subunit, HIF-1 is primarily regulated by oxygen-dependent post-translational hydroxylation of its α subunit. This regulatory mechanism is mediated by non-heme Fe hydroxylases, which specifically hydroxylate the proline and asparagine residues of HIF-1 through dioxygenation, utilizing 2-OG as a co-substrate. Three PHD enzymes and FIH-1 function as essential cellular oxygen sensors, fine-tuning HIF activity in response to fluctuating oxygen levels (111).

Genetic studies of high-altitude Tibetans have identified EPAS1 (HIF-2α), EGLN1 (PHD2), and PPARA as key genes shaped by natural selection for high-altitude adaptation. Among these, HIF signaling serves as the primary regulator of erythropoiesis and other critical physiological functions (112). Hypoxia-induced responses, spanning cardiovascular, pulmonary and metabolic pathways, provide crucial insights into disease mechanisms. Understanding these adaptive pathways may pave the way for targeted therapies for hypoxia-related conditions, including cardiopulmonary diseases, neurodegenerative disorders, metabolic syndromes such as diabetes and obesity, and high-altitude illnesses. To date, no PHD activators or functional restorers have been developed to enhance HIF-α degradation, thereby attenuating downstream gene expression and mitigating associated pathological effects. In Tibetans, genetic variants of EPAS1 downregulate EPAS1 transcription, serving as a protective mechanism against erythrocytosis. This natural adaptation helps maintain relatively low Hb levels and is associated with reduced pulmonary vasoconstriction responses (106). Building upon these genetic insights, strategies aimed at downregulating HIF-α transcription or enhancing PHD enzymatic activity could provide therapeutic avenues for HAPC and hypoxia-induced PHA. Such an approach may help maintain optimal RBC counts, improve oxygen-carrying capacity, balance erythropoiesis and apoptosis, and reduce complications arising from increased blood viscosity, ultimately alleviating patient symptoms and improving prognosis.

Despite the advancements in developing targeted inhibitors, the development of PHD activators or functional restorers remains a significant challenge (113). Although small-molecule p53 reactivators have been explored in clinical trials for restoring the function of certain p53 mutations, similar approaches have yet to be translated into clinical applications for hypoxia-related diseases (114).

Theoretical models suggest that HIF-α inhibitors or PHD functional restorers could be valuable in treating HAPC by counteracting excessive erythropoiesis. While extensive research has been conducted on HIF inhibitors in oncology, no studies have specifically investigated HIF-α suppression as a therapeutic strategy for HAPC. Furthermore, efforts to enhance or restore PHD enzyme function face formidable technological challenges, as identifying compounds with strong biological efficacy suitable for clinical development remains difficult.

Additionally, both HIFs and PHDs exhibit subtype-specific effects in hypoxic and inflammatory environments, adding another layer of complexity to drug development. However, promising advancements in targeting hypoxia-driven inflammatory pathways suggest that novel therapeutic strategies are on the cusp of entering clinical trials (69).