Small-Molecule Modulators of the Hypoxia-Inducible Factor Pathway

Subscriber access provided by AUT Library

Perspective

Small-Molecule Modulators of the Hypoxia-Inducible Factor Pathway: Development and Therapeutic Applications Zhihong Li, Qi-Dong You, and Xiaojin Zhang J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.8b01596 • Publication Date (Web): 25 Jan 2019 Downloaded from http://pubs.acs.org on January 26, 2019

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 68 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Journal of Medicinal Chemistry

Small-Molecule Modulators of the Hypoxia-Inducible Factor Pathway: Development and Therapeutic Applications Zhihong Li a,b, Qidong Youa,*, and Xiaojin Zhanga,b* a

State Key Laboratory of Natural Medicines and Jiangsu Key Laboratory of Drug Design and

Optimization, China Pharmaceutical University, Nanjing 210009, China. b

Department of Chemistry, School of Science, China Pharmaceutical University, Nanjing 211198,

China *Corresponding authors. E-mail addresses: [email protected] (X. Zhang), [email protected] (Q. You)

ABSTRACT Hypoxia-inducible factor (HIF) is a central regulator involved in detection and adaption to cellular oxygen stress through regulation of the hypoxic transcriptional program in angiogenesis, erythropoiesis, and metabolism. The HIF pathway is involved in many diseases. On one hand, overexpression of the HIF pathway is associated with solid tumors such as renal cell carcinoma (RCC). On the other hand, suppression of the HIF pathway is correlated with inflammatory, anemia, and other hypoxic-ischemic diseases. Therefore, modulation of the HIF pathway has been perceived as a promising strategy for treating HIF-related diseases. Recent advances in understanding of the biochemistry underlying the HIF pathway have stimulated small-molecule drugs development, and therapeutically, manipulation of the HIF-mediated response has been to have shown considerable medicinal potential. This review will summarize and provide insight into recent advances in research that have expanded our knowledge of the HIF pathway, including its structural basis and biology, 1

ACS Paragon Plus Environment

Journal of Medicinal Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 68

small-molecule modulators of the pathway, including inhibitors and activators, and the potential therapeutic applications of these modulators.

KEYWORDS HIF pathway; Small-molecule modulators; Inhibitors; Activators

1. INTRODUCTION The response to hypoxia is thought to be related to the α, β-heterodimeric hypoxia inducible factor (HIF).1 Under hypoxia, the oxygen-sensitive HIF-α subunit accumulates, enabling heterodimer complex formation between HIF-α and HIF-β. The HIF-α, β dimer further interacts with the transcriptional co-activator (p300) protein and binds to the hypoxia responsive element (HRE), which is part of the DNA nucleus, thereby regulating the expression of multiple genes, especially those related to hypoxic effects. HIF is a crucial oxygen-sensitive transcription factor that coordinates cellular adaptation to oxygen stress by regulating orchestrated transcriptional programs, such as mitochondrial metabolism, cell viability, proliferation, erythropoiesis, and angiogenesis.2 The coordination of all these programs is generally associated with human disease processes related to anemia, ischemia, most tumors, and other hypoxic-ischemic diseases, making HIF an attractive target for therapeutic application.3, 4 On one hand, according to a published report, a considerable proportion of solid tumors, which are induced by hypoxia, can exist in the low concentration of oxygen if the HIF pathway is stablized.5 Over-expression of the HIF pathway has been correlated with poor patient prognosis, aggressive tumor growth, and resistance to radiation.6 Therefore, the inhibition of HIF activity by small-molecule 2


Page 3 of 68 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


inhibitors has been identified as an effective and attractive therapeutic strategy for various solid tumors, such as breast cancer, lung cancer, multiple myeloma, and renal cell carcinoma (RCC). To date, an increasing number of HIF inhibitors with small-molecules have displayed obvious anticancer potency. For many of these, the mode of action has been developed on the basis of the HIF signaling pathway, mainly including the promotion of the HIF-1α protein degradation, disruption of the HIF-α/ HIF-β interaction, and the blockade of the HIF-α/p300 interaction. Recently, blockade of the HIF-α/p300 interaction has been demonstrated as a superior strategy for the inhibition of the HIF target.7 On the other hand, hypoxia occurs in ischemia reperfusion, inflammation, and neurodegenerative diseases; this has prompted researchers to pay significant attention to HIF activation. HIF activation is now considered as an innovative means to treat HIF-related diseases, such as anemia, ischemia, and Parkinson’s disease.8,9 Based on the HIF biological pathway, the current small-molecule activators of the HIF pathway can be classified into three main types: prolyl hydroxylase domain (PHD) inhibitors, von Hippel-Lindau (VHL) inhibitors, and factor inhibiting HIF (FIH) inhibitors. Among these activators of HIF, PHD inhibitors play the dominant role in up-regulating the transcription genes expression of the HIF signaling pathway, resulting in the recognition of the PHD inhibitors as a unique and rational approach for the treatment of anemia. Consequently, manipulation of the HIF pathway by small-molecule modulators has been widely perceived as a very promising strategy for treating numerous diseases closely associated with HIF dysfunction. Substantial research efforts to date have investigated HIF transcriptional inhibition for the treatment of solid tumors and HIF pathway activation as an effective anti-anemia therapy.10 This review will summarize and provide insights into the recent research advances with respect to the HIF pathway, including the structural basis and biology of the pathway, the crucial factors involved in its 3



Page 4 of 68

down- and up-regulation, small-molecule modulators of the pathway and their potential therapeutic applications.

2. HIF STRUCTURE AND BIOLOGY

The transcription factor HIF was first discovered by Semenza et al. in 1992; it is a multimeric factor that binds directly to a hypoxia-inducible enhancer with 50-base-pair sequence 3′ to the erythropoietin (EPO) gene. Further, it has been reported that HIF is a heterodimeric transcription factor, which consists of a highly oxygen-regulated subunit, HIF-α, and a constitutively expressed subunit, HIF-β. HIF-α contains three isoforms: HIF-1α, HIF-2α, and HIF-3α,11, 12 while HIF-β is the only subtype of the β subunit. HIF-1, consisting of HIF-1α and HIF-β subunits, is the most extensively investigated HIF isoform. HIF-β is a 91-94 kD subunit identical to the known aryl hydrocarbon receptor nuclear translocator (ARNT); it is capable of forming a heterodimer together with the aryl hydrocarbon receptor (AhR), thereby modulating genes associated with the transcriptional response to oxidant stress and various environmental xenobiotics.13 HIF-1α is a 120-130 kD protein; it is the most widely expressed HIF-α subunit in mammalian tissues, and is also highly conserved in many other species. The other HIF-α subunits, HIF-2α and HIF-3α, appear to have more specialized and tissuespecific functions. Among the three known HIF-α subunits, HIF-1α and HIF-2α share high sequence identity. HIF-1α and HIF-2α both contain a basic helix-loop-helix (b-HLH) domain, two PAS domains, a PAC domain, an oxygen-dependent degradation (ODD) domain, an N-terminal transactivation domain (N-TAD), and a C-terminal TAD (C-TAD) (Figure 1).14 It has been revealed that both the CTAD and N-TAD are essential for the transcriptional activity of HIF-1α. Interestingly, the C-TAD domain interacts with CREB-binding protein (CBP)/p300, a co-activator, to modulate transcriptional 4


Page 5 of 68 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


activity of downstream gene under hypoxia, and N-TAD contributes to HIF-1α stabilization against proteasomal degradation.15 However, in the case of the HIF-3α subunit, no intrinsic transactivation activity has been observed. This may be because HIF-3α is missing the C-TAD region, as compared to the structures of HIF-1α and HIF-2α (Figure 1). In addition, it has been found that HIF-3α can suppress HIF-1α, HIF-2α, and HIF-β-mediated transcription, by interfering with the interaction between HIF-1α/2α and DNA.16 HIF-1α is critical in vascular and metabolic responses to hypoxia, whereas HIF-2α is mainly related to erythropoiesis and vascular systems.17 Recent studies have revealed that HIF-2α rather than HIF-1α, might be of greater importance for hypoxia-induced expression of EPO in the kidneys.18 It has also been found that HIF-2α plays a critical role in the development of RCC and is a potentially effective novel target for the RCC treatment.19 Compared to HIF-1α and HIF-2α, HIF-3α has received much less research attention due to the existence of multiple HIF-3α variants.20 Although all of the HIF subunits are involved in generating diverse transcriptional responses under a hypoxic atmosphere, the key components are the HIF-1α and HIF-β subunits. Under the condition of normoxia, HIF-1α has a transient half-life because of content ubiquitination-dependent proteasomal degradation through the VHL E3 ligase protein. The VHL protein specifically recognizes the proline hydroxylated (Pro402-OH/Pro564-OH) HIF-1α on two independent sites: a carboxyl terminal oxygen-dependent degradation domain (CODD) and an amino terminal oxygen-dependent degradation domain (NODD).20,

21

The oxygen-sensitive PHD is an indispensable activator for

catalyzing these hydroxylation reactions, establishing hydroxyl action as a novel functional posttranslational modification in the HIF signaling pathways.22 Additionally, K532 residue acetylation is achieve by ARD1 acetyltransferase for the binding of VHL inside normoxic cells; this contributes to 5



Page 6 of 68

ubiquitination of the HIF protein for degradation.23 Moreover, HIF-1α can be also controlled by the oxygen-sensitive asparaginyl hydroxylase, factor inhibiting HIF (FIH). FIH is able to hydroxylate the Asn803 residue in the C-TAD domain of the HIF-1α subunit, affecting the interaction between the CTAD domain and the transcriptional co-activators p300/CBP, thereby inhibiting various HIF transcriptional activities.24

Figure 1. Functional domains of the HIF subunits. HIF-1α and HIF-2α share high amino acid sequence similarities, while C-TAD is missing in HIF-3α. For a better understanding of the progress in the HIF research field to date, we have compiled a historical timeline of core experimental results and key components of the HIF network (Figure 2). A tight interactive relationship has been found between hypoxia and chemoresistance, radioresistance, invasiveness, angiogenesis, and so forth. Over-expression of HIF pathway is associated with solid tumors such as RCC and triple-negative breast cancer (TNBC).25 On the other hand, the deficient expression of HIF is mainly correlated with inflammatory, anemia, and other hypoxic-ischemic diseases. Therefore, HIF modulation is recognized as a rather effective means to 6


Page 7 of 68 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


treat diseases related to HIF. HIF activators have shown therapeutic potential in treating ischemia, such as chronic kidney disease (CKD) related anemia.26 Meanwhile, HIF inhibitors have been considered as therapies for treating all kinds of diseases related to overactivation of the HIF pathway, including solid tumors. Thus it can be seen that the progress in HIF structure and biology over the past decades has contributed significantly to the development of HIF modulators for treating various diseases.

Figure 2. Profile of the main findings related to the critical components of the HIF pathway and the major modeled behaviors.

3. SMALL-MOLECULE INHIBITORS OF THE HIF PATHWAY

Almost all solid tumors have a common feature, that is, hypoxia, resulting in over-expression of the HIF pathway in the tumor microenvironment.27 In the case of tumor patients, over-expression of HIF usually results in poor prognosis and invasive tumor growth. Therefore, HIF inhibition has been suggested to be an attractive and promising strategy to treat cancer. In this section, we will summarize 7



Page 8 of 68

the recently described small-molecule inhibitors that down-regulate the HIF pathway. These smallmolecule inhibitors have been classified and elucidated as follows according to their different mechanisms of action related to the HIF signaling pathway (Figure 3): 1) inhibitors of the HIF-1α/p300 protein-protein interaction; 2) inhibitors of HIF-α/HIF-β dimerization; 3) inhibitors of HIF-1 DNA binding; 4) inhibitors that down-regulate the expression of HIF-1α mRNA; 5) small molecules that reduce the HIF-1α protein level; and 6) small molecules that inhibit HIF at multiple pathways.

Figure 3. Small-molecule inhibitors of the HIF pathway targeting different sites. 3.1. Inhibitors of the HIF-1α/p300 Protein-Protein Interaction When oxygen concentration is low enough, HIF-1α could escape the degradation and subsequently accumulates in the nucleus, thereby dimerizing with the HIF-β subunit. This allows the recruitment and binding of a transcriptional co-activator, p300, a multi-domain protein that plays a key part in the activation of HIF-1.28 It is worth noting that HIF-1α over-expression has been found in many cancers and is correlated with NF-kB pathway, aberrant p53 accumulation, and tumor progression.29 The NMR structure of the HIF-1α C-terminal activation domain (CAD) when complexed with the cysteine/histidine-rich region (CH1) domain of CBP/p300 was identified by Eck’s group (Figure 4).30 The elucidated interactions demonstrate that disrupting the HIF-1α/p300 proteinprotein interaction with α-helix mimetics is one strategy for blocking HIF-1 activation and could 8


Page 9 of 68 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


potentially be beneficial for tumor treatment.31, 32 Moreover, Wright et al. have also examined the complex structure of CBP/p300 and the CAD of HIF-1α; their study revealed the molecular interface between the CBP/p300 CH1 domain and HIF-1α in detail.33 Further, it has demonstrated that hydroxylation of the Asn803 residue by FIH in the C-TAD of HIF-1α hampers the interaction between HIF-1α and p300.33,

34

Inhibition of the HIF-1α pathway by disrupting its interaction with the

transcriptional co-activator p300 hinders tumor angiogenesis, tumor growth, and resistance to radiation therapy.35 As shown in Figure 5, the compound YC-1 (1), which is an activator of platelet guanylate cyclase, has been found to exert HIF-1 inhibitory activity by enhancing FIH binding to HIF-α C-TAD and subsequent hydroxylation of Asn803.36 It has been revealed that 1 and its derivatives (compounds 2 and 3)37,38 interrupt the HIF-α and p300 interaction, thus inhibit the transcriptional activity of HIF1. Cell-based HIF-responsive-luciferase (HRE-Luc) assay in HeLa cell lines revealed that compound 1, with an IC50 value of 2.0 μM, is a representative member of this type of HIF-1 inhibitors; it suppresses the HIF target genes such as VEGF and EPO. However, most of these derivatives showed similar activity to both HIF-1α and HIF-2α; no selectivity between HIF-1α and HIF-2α isoforms was observed.7

9



Page 10 of 68

Figure 4. NMR structure of CH1domain of p300 (red) in complex with the HIF-1α CAD peptide (yellow) (PDB ID:1L3E). Subsequently, researchers have devoted their efforts to searching for selective HIF-1α inhibitors. HIF-1α transcriptional activity mainly relies on the protein-protein interaction between HIF-1α C-TAD and the co-transcription factor p300/CBP, which can act on the HRE sequence of the promoter of the target gene, subsequently initiating transcription of the target gene. Thus, inhibition of the HIF-1α and p300/CBP protein-protein interaction can directly inhibit HIF-1α-dependent transcriptional activity; this mechanism is a potential and frequently investigated drug target in current medicinal research.39, 40

Initially, with a high-throughput screening (HTS) assay, De Munari et al. discovered that the natural

product chaetocin (4) hampers the HIF-1α/p300 complex in some cancer cells and the expression of HIF-1α target genes was selectively down-regulated.41, 42 Time resolved fluoroimmunoassay (TRFIA) revealed that compound 4 is able to inhibit the HIF-1α/p300 interaction with an IC50 value of 12.50 μM, while enzyme linked immunosorbent assay (ELISA) showed that it interferes with the secretion 10


Page 11 of 68 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


of downstream VEGF protein in Hep3B with an IC50 value of 0.10 μM; this resulted in remarkable and specific inhibition of tumor growth. Further, it was determined that the important factors for the HIF1α/p300 interaction include the C-TAD of HIF-1α and CH1 of p300.43 Thus, targeted blockade of HIF1a C-TAD binding to the CBP/p300 CH1 domain would constitute a novel and specific approach for cancer treatment via the suppression of tumor angiogenesis and metastasis. Kung et al. reported that the natural product chetomin (5)41 and its derivatives are capable of inhibiting the co-activator CBP/p300 and HIF-1α C-TAD interaction, and then effectively reduce HIF1 target proteins such as EPO, VEGF, and Glut1. However, further clinical application of these HIF1α inhibitors was limited due to their high toxicity in normal tissues. Furthermore, Figg’s group identified heterocyclic alkaloids eudistidine A (6) and eudistidine B (7), which were isolated from the marine ascidian Eudistoma sp, as novel HIF-1α/p300 inhibitors.44 Homogeneous time-resolved fluorescence (HTRF) assay revealed that eudistidine A (6) is a modest inhibitor of the HIF-1α/p300 interaction through effectively blocking the binding of p300-CH1 to immobilized HIF-1α C-TAD with an IC50 of approximately 75 µM; this effect occurred in a dose-response manner. In addition, this research group performed a high-throughput screen from 170298 crude natural product extracts and identified that pyrroloiminoquinone alkaloids 8-13 are HIF-1α/p300 inhibitors with comparable inhibitory activity to positive control 5. These natural alkaloids belong to a novel potential of HIF-1α inhibitors which interfere with the HIF-1α and p300 protein-protein interaction, thereby decreasing HIF-1α-related transcription.45

11



Page 12 of 68

Figure 5. Natural products as inhibitors of the HIF-1α/p300 protein-protein interaction. Consequently, inspired by the intriguing structures of the natural products chaetocin (4) and chetomin (5), Block et al. reported on a chetomin analogue, ETP 3 (14, Figure 6), which uses a benzene ring as the linker to connect two dithiopiperazine pharmacophoric moieties.46 Fluorescence polarization (FP) assay revealed that EP 3 can bind to the p300-CH1 protein and is able to selectively disrupt the interaction of the p300/CBP co-activator, with an IC50 value of 1.50 µM; this leads to down-regulation of the HIF-related genes critical for cancer progression. Encouraged by the success of 14, this research group further designed and synthesized a novel dimeric chetomin analogue, ETP2 (15), with an extended linker, which was found to exert a higher FP affinity with an IC50 value of 0.6 µM.47 It is worth noting that the designed natural product12


Page 13 of 68 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


like compound 15 was found to be less toxic to normal cells than natural lead chetomin (5). Verified by the in vivo intravital microscopy (IVM) imaging model of breast cancer, ETP2 was found to effectively inhibit tumor growth with no damage or acute toxicity in mouse models.

Figure 6. Small-molecule inhibitors involved in the HIF-1α/p300 protein-protein interaction. Additionally, indandione compound (16) and quinone compound (17) have been shown to possess the ability to break the tight interaction between HIF-1α CAD and p300.48 Further mechanism studies have revealed that these compounds inhibit the binding of p300 to HIF-1α by inducing Zn2+ ejection from the p300 protein. This mode of action results in a lack of specific-target selectivity for their potential interaction with other Zn2+ containing proteins. Ortho-quinone compound (18) and thiazolidinone compound (19) have shown a superior inhibitory activity for HIF-1α/p300 interaction.49 These two types of compounds has been found to restrain the expression of VEGF with IC50 values ranging from 0.10 to 100 μM in a hypoxic cell model. Moreover, Kwon et al. reported that KST012174 13



Page 14 of 68

(20) with a novel chemical type could interrupt the HIF-1α/p300 interaction with an IC50 value of 107.0 μM in an affinity-based FP assay.50 Wang et al. found that the interaction of HIF-1α with cofactors p300/CBP was efficiently interrupted by arylsulfonamide KCN1 (21) in vitro using a surface plasmon resonance (SPR) assay.51 Of note, it was observed that KCN1 inhibits the hypoxia-induced transcriptional activation of HIF-1 in LN229HRE-luc/LacZ cells with an IC50 value of 590 nM. Studies have shown that 21 inhibits cell growth and leads to cell cycle arrest in human pancreatic cancer cells in a dose-dependent manner in vitro; it also possesses promising anticancer efficacy in mice bearing Panc-1 or Mia Paca-2 tumor xenografts in vivo. Notably, 21 has good pharmacological properties and bioavailability, which lays a good foundation for the treatment of human tumors displaying overexpression of HIF.52, 53 With research advances in the clinical application of drugs, new drug indications are often revealed. Thus, there are many conventional drugs that find new clinical applications. Bortezomib (22, Figure 6), which is clinically used to treat solid tumors and multiple myeloma, has been reported to break tumor adaption in a hypoxia atmosphere by functionally stimulating the FIH-mediated inhibition of HIF-1α,54 by means of inhibiting the association between HIF-1α and the p300 co-activator, thereby down-regulating HIF-1α dependent gene transcription. It has been reported that bortezomib blocks the hypoxic induction of EPO and VEGF in both drug-sensitive U299 cell lines and a resistant Hep3B cell line with a sub-nanomolar concentration.54 Menadione (23) has a significant curative effect in hemostasis and detoxification, while ethacrynic acid (24) is regarded as a diuretic. Interestingly, recent research has revealed that these compounds could interfere with the HIF pathway by disrupting the interaction between HIF-1α and p300, thereby down-regulating the expression of VEGF.55 In addition, Wu et al. reported that novobiocin (25), a coumarin antibiotic, directly hinders the interaction between 14


Page 15 of 68 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


the CH1 region of p300/CBP and HIF-1α C-TAD.56 Meanwhile, another study has shown that, by regulating the m-TOR signaling pathway, 25 is capable of restraining the function of HIF-1α and its corresponding downstream target protein Further, several peptoid inhibitors have been designed, based on the structure of HIF-1α, to be HIF-1α/p300 inhibitors with high selectivity and less cytotoxic side effects. Arora et al. revealed that the HIF residues Leu818, Leu822, Asp823, and Gln824 play a significant role in the binding of HIF1α to p300.57 Based on this, a peptidomimetic HBS1 (26, Figure 7) was designed. Peptoid 26 showed a preferable affinity with p300-CH1 and blocked its binding to the HIF-1α C-TAD. The dissociation constant (Kd) between peptoid 26 and p300 was 690 nM.57 Peptoid 26 inhibits HIF-1 target genes, including VEGF and GLUT1, in HeLa cells under hypoxic conditions. Furthermore, 26 has shown remarkable efficacy in reducing tumor growth in an RCC xenograft mouse model. Importantly, it provides a reasonable approach for discovery of HIF-1α/p300 inhibitors through topographical mimics of residues on the secondary structures of proteins that are energetically important. Computational and structural analyses of the HIF-1α/p300 complex have revealed that the CH1 domain of p300/CBP interacts with the helical residues of HIF-1α helix814-824. Based on these results, peptoid OHM1 (27), simulating the key amino acids Leu818, Leu822, and Gln824, was found to exert a highly selective inhibition of the HIF-1α/p300 interaction both in vitro and in vivo.58 The intrinsic tryptophan fluorescence spectroscopy assay revealed that OHM1 targets the p300 CH1 domain with an affinity Kd = 530 nM.58

15



Page 16 of 68

Figure 7. Representative peptoid inhibitors of the HIF-1α/p300 protein-protein interaction. 3.2. Inhibitors of HIF-α/HIF-β Dimerization Interrupting HIF-α/HIF-β dimerization has recently been considered as an attractive strategy to down-regulate the HIF pathway. As shown in Figure 8, acriflavine (28), initially reported to possess antiviral, antibacterial, and trypanocidal activities, is considered as a HIF dimerization inhibitor that interacts with both HIF-1α and HIF-2α in their PAS-B subdomain.59 It was discovered that 28 can suppress HIF transcriptional activity, which further inhibit vascularization and the growth of tumors. However, compound 28, similar to most HIF inhibitors, targets both HIF-1 and HIF-2 without clear selectively between the two subunits. A high-throughput screening from a large library of over three million cyclic peptides using genetically encoded technology has led to the discovery of a cyclopeptide cyclo-CLLFVY (29), which is considered to be an effective inhibitor of the HIF-1α/HIF-β interaction.60 Notably, compound 29 has indicated a high selective to the PAS-B domain of HIF-1α, thus inhibiting HIF-1 dimerization and transcription activity without interfering with HIF-2α. HIF-2α has long been considered as non-targetable and undruggable.61 However, recent studies have revealed that the PAS-B domain of HIF-2α possesses a large cavity that carries a hydrophobic core and it is 16


Page 17 of 68 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


quite different from the corresponding PAS-B region of HIF-1α; this provides potential opportunities to selectively target-specific small-molecule regulation of HIF-2α (Figure 9).62-64 Achiral compound (30) antagonizes endogenous HIF-2 heterodimerization and DNA-binding activity in vitro in Hep3B cells, significantly down-regulating HIF-2 target gene expression. Further, Tambar et al. have developed a series of chiral inhibitors of the HIF-2α PAS-B domain, exemplified by compound 31, which can trigger greater changes in protein conformation than the achiral HIF-2 inhibitors reported previously.64 Moreover, selective small-molecule antagonists of HIF-2α have been reported, establishing a novel approach to treating RCC with HIF-2α over-expression. PT2385 (32) and PT2399 (33) are the most active and representative inhibitors of HIF-2α/HIF-β dimerization for the treatment of RCC.65-67 PT2399 can directly bind to the PAS-B domain of HIF-2α (Figure 9), crippling the ability of HIF-2α to bind to HIF-β. PT2385 (32) and PT2399 (33) represent a new class of therapeutic candidates for treating RCC and VHL disease; they have good preclinical efficacy and their tolerability and safety are improved relative to the existing agents.65 Of particular note is that PT2385 (32) exhibited no adverse effects on the cardiovascular system. A phase I clinical study for the treatment of RCC is currently underway.68 In addition, treatment with PT2385 (32) was able to markedly reduce the ceramide levels in the intestine and serum. Further experiments have shown that PT2385 has a positive effect on nonalcoholic fatty liver disease by increasing the expression of the neuraminidase 3 (Neu3) gene.69

17



Page 18 of 68

Figure 8. Small-molecule inhibitors interrupting HIF-α/HIF-β dimerization.

Figure 9. The binding of the HIF-2α PAS-B and HIF-β PAS-B domain (PDB ID:5TBM) and the binding mode of PT2385 (30) with HIF-2α. 3.3. Inhibitors of HIF-1 DNA Binding The binding of HIF to DNA is the final step in the mechanism of HIF’s function as a transcription 18


Page 19 of 68 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


factor. Inhibition of HIF binding to HRE-DNA sequence is a promising target for small-molecule inhibitors. Dervan’s group has revealed that sequence-specific DNA-binding molecules (34 and 35, Figure 10) can disrupt the connection between the HIF-1α/HIF-β heterodimer and the corresponding cognate DNA sequence, thus inhibiting the expression of VEGF and other genes related to hypoxia in the HIF pathway.2 In addition, it has been illustrated that echinomycin (36), a cyclic octapeptide, can affect the interaction between HIF-1 and DNA with a high affinity, thereby influencing expression of almost every gene that is induced by hypoxia.70 Additionally, Melillo et al. reported that NSC-50532 (37, Figure 10), a semi-synthetic pyrrolidone methyl tetracycline natural product, selectively breaks HIF-1 and DNA binding.71 Similarly, adriamycin (38) and daunorubicin (39) were able to inhibit HIF1 bioactivity by blocking its binding to the downstream HRE-DNA sequence.72 In human prostate cancer xenografted mouse models, both of 38 and 39 showed significant inhibitory activity on tumor growth and vascularization. In addition, DJ12 (40) was initially reported to be an inhibitor of HIF function via blockade of HIF-1α and HRE-DNA binding, resulting in down-regulation of the downstream target mRNA of HIF. However, recent research has shown that DJ12 does not directly interrupt the association between the HIF-DNA complex; rather, it likely inhibits the tight complex formation between HIF and p300/CBP transcription.73 Recently, the binding mode of the HIF-1/2 and HRE DNA sequences have been resolved with the use of X-Ray diffraction crystallography (Figure 11).74 It was shown that the binding between HIF and DNA is mediated by the b-HLH and PAS-A domain; these corresponding structural binding details will be beneficial for future design of novel small inhibitors of HIF-1 or HIF-2 DNA binding.

19



Page 20 of 68

Figure 10. Small-molecule inhibitors of HIF-1 DNA binding.

20


Page 21 of 68 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 11. Architecture of HIF bHLH-PAS heterodimers. A) X-ray crystal structure (PDB ID:4ZPR) of the HIF-1α/HIF-β bHLH-PAS heterodimer bound to DNA. B) Crystal structure (PDB ID:4ZPK) of the HIF-2α/HIF-β bHLH-PAS heterodimer bound to DNA. 3.4. Inhibitors that Down-Regulate the Expression of HIF-1α mRNA Aminoflavone (41, Figure 12), a ligand of AhR which is currently under investigation in a phase II clinical trial, was reported to be an inhibitor of HIF-1α mRNA expression.75 In addition, GL331 (42), a podophyllotoxin derivative, has been found to exhibit strong cell proliferation inhibition to many tumor cells in phase II clinical trials. GL331 was found to inhibit the expression of HIF-1α mRNA but did not cause substantial changes in HIF-1α protein.76 EZN-2968, a 16-Mer locked nucleic acid (LNA) antisense oligonucleotide, is another agent that was found to selectively reduce HIF-1α mRNA expression.77 In human prostate (15PC3, PC3, and DU145) and glioblastoma (U373) cells, inhibition of HIF-1α by EZN-2968 was found to successfully attenuate HIF-1α mRNA and protein level in an oxygen concentration independent fashion, which was associated with inhibition of tumor cell growth 21



Page 22 of 68

with an IC50 value of approximately 1-5 nM. Beyond this, the endogenous expression of HIF-1α mRNA in the livers of mice was found to be continuously inhibited by EZN-2968 in the nanomolar range. However, the comprehensive utilization of EZN-2968 in tumor models needs further study because of its long half-life in the liver and kidneys.

Figure 12. Small-molecule inhibitors that down-regulate HIF-1α mRNA expression. 3.5. Small-Molecules that Reduce HIF-1α Protein Levels The endogenous expression of the HIF-1α protein can be modulated by its upstream signaling pathways. For instance, the acceleration of the rate of HIF-1α protein biosynthesis induced by the PI3K-Akt-mTOR signaling pathway makes a significant difference in up-regulation of the expression of HIF-1 in human cancer cell lines. Thus, suppressing specific proteins in the PI3K-Akt-mTOR signaling pathway can indirectly inhibit the biosynthesis of HIF-1α protein. It has been reported that the HIF-1α protein synthesis process is inhibited by wortmannin (43, Figure 13), a PI3K-specific inhibitor, in a dose-dependent manner in the prostate cancer cells PC-3, P19, and DU145.78 Further, it has been revealed that inhibition of HIF-1α by suppression of the PI3K-Akt pathway can hinder the proliferation, migration and invasion of gastric cancer.79 Additionally, KC7F2 (44) and rapamycin (45), specific mTOR1 inhibitors, have been shown to reduce both the accumulation of HIF-1α and HIF-1 dependent transcription in PC-3 prostate cancer cells.80-82 Moreover, KC7F2 (44) has been shown to 22


Page 23 of 68 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


inhibit the proliferation of many cancer cells, an effect that is increased in hypoxic conditions, whereas non-tumor cells are less sensitive.81 Besides the PI3K-Akt-mTOR signaling pathway, other pathways have also been found to be capable of reducing HIF-1α protein levels. LY-294002 (46) has been shown to regulate the expression of the functionally BCR/ABL-induced VEGF gene by inhibiting BCR/ABLdependent HIF-1α expression.83 This may be of benefit for treating chronic myelogenous leukemia and for developing new treatment strategies. Gefitinib (47), an EGFR inhibitor, was revealed to increase the proteasomal degradation and reduce the protein biosynthesis of HIF-1α; thus, it may overcome the hypoxia-induced drug resistance.84 Genistein (48), a non-specific kinase inhibitor, was found to completely block the biosynthesis of the HIF-1α subunit and the binding of the HIF-1 dimer to DNA in A549 lung cancer cells; this implicated 48 as a promising candidate for the treatment of lung cancer.85 PD-98059 (49), an inhibitor of the MEK pathway, inhibits HIF-1-mediated target gene activation.86 Additionally, NS-398 (50), a COX-2 inhibitor, seems to facilitate to the inhibition of HIF1α expression in both COX-2-negative HCT116 cells and COX-2-positive PC-3 cells.87 Recent research has indicated that flavopiridol (51) down-regulates HIF-1α expression in combination with a proteasome inhibitor, a substance that causes the accumulation of HIF-1α under normal conditions. If flavopiridol treatment is combined with a proteasome inhibitor, the down-regulation of HIF-1α expression may be very promising for the treatment of cancers.88 2ME2 (52) and ENMD-1198 (53) have been confirmed to possess potent inhibitory effects on HIF-1α translation, nuclear translocation, and downstream elements responsible for microtubule disruption; this effect was related to antiangiogenic activity.89-91 NSC-609699 (54), a Topo I inhibitor, has been clinically used to treat small-cell lung cancer and ovarian cancer. This study indicated that 54 decreases the expression of HIF-1α, and suppresses angiogenesis and the growth of tumors in a human glioma xenograft model.9223



93

Page 24 of 68

Representative small-molecule inhibitors that decrease the cellular levels of HIF-1α protein are

summarized and listed in Table 1. Table 1. Small Molecules that Suppress HIF-1α Expression agent name

target

indications a

current status a

wortmannin (43)

PI3K/Akt inhibitor

inflammatory bowel disease

preclinical

solid tumors

preclinical

renal diseases

approved

leukemia

preclinical

non-small cell lung cancer

approved

prostate cancer

phase II

KC7F2 (44) mTOR complex 1 inhibitor rapamycin (45) LY-294002 (46) gefitinib (47) genistein (48)

EGFR inhibitor non-specific kinase inhibitor

PD-98059 (49)

MEK inhibitor

stroke

preclinical

NS-398 (50)

COX-2 inhibitor

thrombosis and stroke

preclinical

multiple myeloma

phase II

prostate cancer

discontinued

solid tumors

phase I

lung Cancer

approved

flavopiridol (51) 2ME2 (52) ENMD-1198 (53) NSC-609699 (54) a

BCR-ABL inhibitor

CDK inhibitor microtubules inhibitor Topo-I inhibitor

Indications and current status registered at integrity.thomson-pharma.com.

24


Page 25 of 68 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 13. Representative small-molecules that suppress HIF-1α expression. Decreasing HIF-1α protein stability by promoting the degradation of HIF-1α protein is also an important strategy for down-regulating the HIF-1α pathway. These kinds of small molecules increase the ubiquitination level of HIF-1α via specific pathways, thereby facilitating its subsequent degradation by proteasomes. In the upstream regulatory signaling pathway of HIF-1α, extensive 25



Page 26 of 68

research has revealed that the stability of HIF-1α is regulated by heat-shock protein 90 (Hsp90), p53MDM2 proteins, PHDs, and so forth. As shown in Table 2, inhibitors of the HIF-1α protein via decreasing its stability primarily included Hsp90 inhibitors, p53-MDM2 inhibitors, and PHD2 activators. Geldanamycin (GA, 55, Figure 14) and its corresponding analogues 17-AAG (56) and 17DMAG (57), representative Hsp90 inhibitors, can promote ubiquitination of HIF-1α independent from VHL; the subsequent degradation of HIF-1α is mediated by proteasomes in renal cell carcinoma.94, 95 Furthermore, STA-9090 (58), a second-generation Hsp90 inhibitor blocked HIF-1 activity and inhibits TNBC orthotopic tumor growth, invasion, and metastasis.95 In addition, the p53 tumor suppressor protein is a potent negative regulator of HIF-1α and VEGF in hypoxic conditions. RITA (59) was found to induce p53-dependent phosphorylation of eIF-2, and this was related to its ability to block HIF-1α induction in hypoxia.96 Further, KRH102053 (60) and KRH102140 (61), potent activators of the PHD2 enzyme, have been found to efficiently suppress HIF-1α in human osteosarcoma cells under hypoxia.97, 98

Moreover, R59949 (62), an activator of HIF prolyl hydroxylases which increases the affinity of the

PHD enzyme with oxygen and stimulated PHD activity,99 is capable of inhibiting the accumulation of the HIF protein, leading to suppression of the neovascular response and prevention of glial degeneration.100 Table 2. Inhibitors of the HIF-1α Protein by Decreasing Its Stability agent name

indications a

current status a

multiple myeloma

discontinued

multiple myeloma

phase III

breast cancer

discontinued

multiple myeloma

phase II

p53-MDM2 inhibitor

solid tumors

preclinical

PHD activator

solid tumors

preclinical

target

GA(55) 17-AAG(56) 17-DMAG(57)

Hsp90 inhibitor

STA-9090(58) RITA(59) KRH102053(60) KRH102140(61) R59949(62)

26


Page 27 of 68 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


a

Indications and current status registered at integrity.thomson-pharma.com.

Figure 14. Representative small molecules that decrease the stability of the HIF-1α protein. In addition, some inhibitors can probably bind to HIF-1α, thus facilitating its degradation. LW6 (63, Figure 15), a novel small-molecule inhibitor of HIF-1α with an IC50 value of 2.6 µM in an HRELuc assay in Hep3B cell lines which bears the unique adamantyl group, has been shown to decrease the stability of HIF-1α and then suppress the expression of HIF-1α target genes VEGF and EPO in a dose-dependent manner.101 Further, according to a detailed structure-activity relationship (SAR) investigation, the unique adamantyl group functions as the pharmacophoric motif responsible for the activity of 63. Based on lead compound 63, several derivatives such as isoniccotinic acid derivatives 64-66, benzimidazole derivative 67, and (E)-phenoxyacrylic amide derivative 68, were designed and synthesized to exert satisfactory bioactivity for directly decreasing the stability of the HIF-1α 27



protein.102,

103

Page 28 of 68

Further, our group reported a series of N-(benzofuran-5-yl)aromaticsulfonamide

derivatives as novel HIF-1α inhibitors.103 For instance, compound 69 was shown facilitate HIF-1α degradation and hinder the accumulation of HIF-1α protein in MCF-7 and HUVEC cells. Of particular note was that 69 could effectively suppress angiogenesis in chicken chorioallantoic membrane (CAM) model in vivo.104 Later, Moreno-Manzano et al. reported that FM19G11 (70) down-regulates the HIF1α protein in several tumor cells and in embryonic and adult stem cells. HRE-Luc assay in HeLa cells revealed that compound 70 exhibits significant inhibitory activity against HIF-1α with an IC50 value of 2.6 µM.105

Figure 15. Representative small molecules that facilitate HIF-1α degradation. 3.6. Small Molecules that Inhibit HIF at Multiple Pathways PX-478 (71, Figure 16), a phase I clinical agent against advanced solid tumors, was shown to decrease the mRNA level, protein stability, translation and transactivation activity of HIF-1α in various cancer cells under hypoxic conditions.106 Compound 71 was found to be highly selective for HIF-1α when compared with HIF-2α and is considered to be a promising drug to target the chemotherapy resistance of tumors by inducing cell cycle arrest in G2 stage, promoting apoptosis, and reducing COX28


Page 29 of 68 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


2 and PD-L1 expression of esophageal squamous cell cancer cells.107

Figure 16. Small molecule that inhibits HIF at multiple pathways.

4. SMALL-MOLECULE ACTIVATORS OF HIF PATHWAY

Under normoxic conditions, VHL Cullin RING E3 ubiquitin ligase complex poly-ubiquitinates the HIF-α subunit efﬁciently, resulting in proteasomal degradation of HIF-α.108, 109 The hydroxylation of HIF-α proline residues by PHD enzymes is critical for the recognition of HIF-α by VHL. There is another process of HIF regulation that involves hydroxylation of asparagine in HIF-α CAD by FIH. The hydroxylation of asparagine prevents HIF from recruiting the transcriptional coactivator p300/CBP to the HIF-α CAD. Based on the HIF biological pathway, the up-regulation of the HIF pathway can be categorized as follows (Figure 17): 1) PHD inhibitors; 2) inhibitors of the VHL: HIFα protein-protein interaction; 3) FIH inhibitors.

29



Page 30 of 68

Figure 17. The oxygen-dependent regulation process of HIF-α and the potential strategies for activating the HIF pathway. 4.1. PHD Inhibitors PHD enzymes primarily contain three isozymes in mammalian cells: 407-amino-acid PHD1, 426amino-acid PHD2, and 239-amino-acid PHD3.110 Since the genes coding PHD1, PHD2, and PHD3 are isogenous with the Egl-9 gene of the nematode, these three isoforms were respectively named EGLN2, EGLN1, EGLN3.111 A fourth isoform, PHD4, has been reported, however its function is much less elucidated.112 PHDs show a clear intracellular distribution pattern: PHD1 is primarily distributed in the nucleus, PHD2 is mainly distributed in the cytoplasm, and PHD3 is found both in the nucleus and cytoplasm.113 PHDs belong to the nonheme Fe(II)-containing dioxygenase family; 2-oxoglutarate (2OG) and oxygen are adopted as substrates and ascorbate and iron are used as cofactors. Under normoxia, PHDs hydroxylate certain proline residues of HIF-α (Pro402 and Pro564 of HIF-1α, Pro531 of HIF-2α, and Pro490 of HIF-3α) in the ODD region. Subsequently, p-VHL, which is critical in the 30


Page 31 of 68 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


process of recognizing the ubiquitin ligase complex, will bind to the hydroxylated HIF-α to initiate the proteasomal degradation of HIF-α. It is worth noting that all of three isoforms of the PHD1-3 subunits can hydroxylate residue Pro564 of HIF-1α, but residue Pro402 can only be hydroxylated by PHD1 and PHD2. In addition, the Km values of PHD1 and PHD2 for residue Pro402 are much higher than for Pro564, indicating that Pro564 plays a significant role in HIF regulation by PHDs, as compared to residue Pro402.114 More importantly, the Km values of HIF PHDs for oxygen ranged from 230 μM to 250 μM, as reported. This is only little higher than the concentration of oxygen which is physiologically dissolved at atmospheric pressure (nearly 200 μM),115 suggesting that PHDs are effective and sensitive oxygen sensors.116 Therefore, PHD inhibitors can efficiently up-regulate HIF levels. It has been revealed that PHD1/3 exhibits tissue-specific expression, while PHD2 is the most widely distributed subtype and of the most significance in preserving steady-state levels of HIF-α.117, 118

Moreover, researchers have revealed that hypoxia affects PHD1 expression probably via a non-

HRE-mediated manner, while two PHD1 isoforms have been identified in the nucleus, each showing a different function in cell proliferation.119 Since PHD2 plays a crucial role in controlling the HIF-α content in mammalian cells, it has been considered as a promising target for treating numerous conditions related to HIF, including ischemia, stroke, and inflammation. The complex structure of PHD2 with HIF-1α CODD and endogenous 2-OG has been determined (Figure 18), this provides a structural biology foundation for the exploration of target small-molecule inhibitors. PHD2 forms a stable complex with HIF-1α CODD, which is composed of a ligand 2-OG and a cofactor ferrous ion; the Pro564 residue that is close to the carboxyl of 2-OG of HIF-1α can recognize the catalytic pocket of PHD2 (Figure 18A).120 31



Page 32 of 68

Figure 18. A) Crystal structures of the PHD2 (red) and HIF-1α peptide substrate (green) complex (PDB ID:3HQR). B) Crystal structure of PHD2 and 2-OG complex (PDB ID:3OUJ). Mechanically, PHDs require 2-oxoglutarate,121 Fe (II), and O2 for their catalytic activities to hydroxylate HIF-α. Based on this, compounds that affect the concentration of O2, Fe (II), and 2-OG or their binding to PHD2,120 can obviously inhibit the hydroxylation function of PHD2, thereby upregulating HIF levels. PHD2 has been identified as a promising therapeutic target in erythropoiesis and iron metabolism. Up to now, many small-molecule PHD2 inhibitors have been explored, and these mainly fall into two categories: (1) non-competitive and non-selective iron chelators, and (2) competitive and selective 2-OG mimetics. Representative examples of iron chelators include deferoxamine (DFO, 72), 3,4-dihydroxybenzoic acid (DHB, 73), and 1,10-phenanthroline (74). These iron chelators can chelate with Fe (II), reduce the concentration of ferrous iron, hinder PHD catalytic activity, stabilize the HIF-α and up-regulate of the expression of EPO, VEGF, and transferrin receptor.122, 123 However, iron chelators lack selectivity as they can interact with all iron-containing proteins; this may cause toxic side effects such as heart failure, liver cirrhosis, and fibrosis. 32


Page 33 of 68 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Consequently, competitive and selective PHD inhibitors have been developed. Almost all of the reported PHD2 inhibitors are 2-OG derivatives or mimetics, sharing key pharmacophores similar to the binding mode of 2-OG with the PHD2 enzyme (Figure 18B).124 As shown in Figure 19 and Figure 20, with the example of FG-2216 (75),125 a ferrous iron (Fe2+) coordination fragment and a fragment that forms ionic bonds with Arg383 are both essential for FG-2216’s inhibitory activity against PHDs.126 Recently, IOX-2 (76) was developed; this potent PHD2 up-regulates endogenous EPO with high affinity (IC50 = 22.0 nM).125 Further studies revealed that IOX-4 (77), a promising PHD2 inhibitor, induces HIF-α in human cells (HeLa, Hep3B, and RCC4) and in the brain tissue of wildtype mice, thereby stabilizing HIF for treating cerebral diseases such as stroke.127 To date, at least six PHD2 inhibitors have been examined in clinical studies in humans (Table 3) and have shown clinically meaningful effects for the up-regulation of erythropoietin via stabilization of HIF-α , thereby providing a promising strategy for treating anemia induced by CKD.128 The reported structures are shown in Figure 19, including FG-2216 (75), FG-4592 (roxadustat) (78), AKB-6548 (vadadustat) (79), GSK1278863 (daprodustat) (80), BAY-85-3934 (molidustat) (81),129 and JTZ-951 (82).130 As clinical trial of FG-4592 has just been completed in China to explore the safety and efficacy of this compound in patients with anemia associated with CKD, including patients who were dialysis-dependent (DD) and those who were not dialysis-dependent (NDD). The clinical data indicated that FG-4592 is an orally active PHD inhibitor that can be used therapeutically for the treatment of anemia in CKD patients. Recent studies indicated that FG-4592 could potentially correct anemia with an acceptable safety profile, regardless of inflammation or initial iron repletion status. Besides, it can reduce hepcidin and cholesterol levels in patients with NDD-CKD.131 Clinical trials of representative small-molecule inhibitors of PHD are summarized and listed in Table 3. 33



Page 34 of 68

In addition, several new potent PHD inhibitors have been reported in recent years. Based on ASMS screening methodology, Hale et al. reported a series of triazaspiro[4.5]decanones as potent PHD inhibitors; these compounds have great clinical potential due to their short-acting PHD inhibition, leading to great downstream in vivo efficacy.132 Most recently, our group reported a series of triazolecontaining picolinoylglycines as new PHD2 inhibitors using a click-chemistry based strategy. Notably, the orally active inhibitor 83 has an IC50 of 62.3 nM, as determined by an FP assay, which is almost ten times more potent in vitro than the recently approved drug FG-4592.133 Compound 83 can stabilize HIF-α to stimulate EPO formation, thereby up-regulating hemoglobin to normal levels in cisplatininduced anemic mice; this was associated with favorable safety proﬁles for treating anemia. 133 Further, Merck recently used a high throughput screen method to report on a novel class of PHD pan inhibitors with promising pharmacokinetics and pharmacodynamics. The representative example was MK-8617 (84) with an encouraging IC50 value of 1.0 nM for PHD2 inhibition, as measured in time-resolved fluorescence (TRF) assay;134 the structure of this compound is totally different to the above-mentioned inhibitors as it is missing the carbonyl acid moiety. Molecular modelling studies revealed that the heterobiaryl moieties of 84 chelate Fe2+ and one of the methoxy substituted aryl rings stacking over Arg322 attributes to cation-π interaction with the Arg322 residue of PHD2 (Figure 19).134 Moreover, MK-8617 has shown improved potency and bioavailability in rats and its t1/2 determined in monkeys is 10 to19 h, suggesting that it has preferable pharmacological and pharmacokinetic properties for targeting PHD in vivo.

34


Page 35 of 68 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 19. Structures of representative small-molecule inhibitors of PHD.

35



Page 36 of 68

Figure 20. Crystal structure of the PHD2 and FG-2216 (75) complex (PDB ID:2HBT). Table 3. Clinical Trial Information of Small-Molecule PHD Inhibitors molecular FG-4592 roxadustat (ASP-1517)

phase

NCT reference

study design

treatment

renal disease

duration

category a

randomized open label Approved 12/2018

02278341

verus epoetin αand

4 years

darbepoetin α,

(11/2014-7/2018)

03303066

verus placebo 175 patients

3 years

MDS or risk

(11/2017-

with

Small-Molecule Modulators of the Hypoxia-Inducible Factor Pathway

Recommend Documents