Structural insights and clinical advances in small-molecule inhibitors targeting TGF-β receptor I

Abstract

The dysregulation of the transforming growth factor β (TGF-β) signaling pathway plays a critical role in the onset and progression of several diseases, including cancer. Notably, TGF-β has emerged as a significant barrier to effective outcomes in cancer immunotherapies, particularly those using immune checkpoint inhibitors. In response to this challenge, small-molecule inhibitors targeting the TGF-β receptor I (TGF-βRI) have garnered attention as promising candidates for modulating the TGF-β signaling pathway. This comprehensive review focuses on the development of small-molecule inhibitors targeting TGF-βRI. We provide a detailed analysis of the structural biology of TGF-βRI, highlighting key binding interactions and structural insights derived from high-resolution X-ray crystal structures. Additionally, we review the current landscape of TGF-βRI inhibitors in clinical trials, including eight promising inhibitors, and discuss their mechanisms of action, selectivity, and therapeutic potential. Our investigation extends to the patent literature, summarizing over 2 decades of innovation from leading pharmaceutical companies, spanning January 2000–May 2024. This consolidated structural and biochemical knowledge aims to facilitate the design of next-generation TGF-βRI inhibitors, addressing unmet clinical needs in oncology and fibrosis treatment. The synergistic potential of combining TGF-βRI and immune checkpoint inhibitors is also explored, offering promising avenues for enhancing cancer immunotherapy efficacy.

Graphical abstract

This comprehensive overview of TGF-βRI includes an analysis of 24 high-resolution three-dimensional experimental structures from the PDB database, all derived from Homo sapiens, with resolutions below 2.5 Å, obtained through X-ray diffraction. Additionally, it provides an extensive review of TGF-βRI inhibitors patented from 2000 to 2023.

Introduction

The transforming growth factor β (TGF-β) superfamily is a large group of structurally conserved growth factors that plays a pivotal role in the regulation of fundamental cellular properties and vital cellular processes, including morphogenesis, cell proliferation, apoptosis, metabolism, homeostasis, fibrosis, adhesion, inflammation, invasion, and differentiation, particularly in the context of diseases such as cancer.^1,2,3 The TGF-β superfamily is composed of several subfamilies, TGF-βs, growth and differentiation factors, bone morphogenetic proteins, activins, inhibins, nodals, and anti-Müllerian hormone.^2,3,4 Among these subfamilies, the role of TGF-βs in tumorigenesis has been intensely studied in the last several decades.^1,2,3,4,5

In early tumorigenesis, TGF-β signaling exhibits antitumor functions by inhibiting the growth of epithelial and lymphoid tissues, maintaining genomic stability, and suppressing paracrine mitogenic factors from stromal cells. It induces G1 cell-cycle arrest and apoptosis in tumor cells via upregulation of cyclin-dependent kinase inhibitors and apoptotic regulators. However, in later stages, TGF-β signaling exhibits protumorigenic roles due to mutations that disrupt its inhibitory effects. Therefore, it facilitates autocrine mitogenic growth factor production, enhancing tumor cell proliferation, invasion, migration, and metastasis. In the tumor microenvironment, nearly all cells potentially respond to TGF-β. However, these responses are heterogeneous, depending on the cellular context and type. Consequently, understanding the role of TGF-β in cancer progression requires a comprehensive analysis of the complex interplay between the responses of cancer and stromal cells to TGF-β (Figure 1).

Figure 1.

TGF-β signaling pathways and effects on immune cell populations

(A) TGF-β ligand homodimer isoforms bind to the extracellular domain of TGF-βRII, which then recruits and activates TGF-βRI. This activation triggers two signaling pathways: (1) Smad-dependent canonical pathway: Smad2 and Smad3 are phosphorylated at their C-terminal, enabling their association with Smad4 to form trimeric complexes (e.g., Smad2/3/4, Smad2/2/4, or Smad3/3/4). These complexes regulate target gene expression. (2) Non-canonical pathways: these pathways activate a phosphorylation cascade targeting serine and threonine residues in the linker region of Smad2 and Smad3, modulating their activity and functions in gene regulation. (B) TGF-β exerts distinct effects on immune cell populations: (1) inhibitory effects (in red): TGF-β inhibits the activation, maturation, antigen presentation, and proliferation of antigen-presenting cells, such as dendritic cells (DCs) and B cells. It also suppresses the functions of natural killer (NK) cells, polymorphonuclear leukocytes (PMNs), and M1-like macrophages. (2) Stimulatory effects (in green): TGF-β stimulates M2-like macrophages, promoting the release of anti-inflammatory cytokines (e.g., IL-4, IL-10) and supports the differentiation, expansion, and immune suppression of myeloid-derived suppressor cells (MDSCs). Its effects on T effector or helper (H) cells can be either stimulatory or suppressive. This figure was created using BioRender software (www.biorender.com).

The immunosuppressive activities of TGF-β on key cell types that orchestrate innate and adaptive immunity are increasingly recognized as major barriers to effective immune activation in the treatment of cancer and other diseases characterized by chronic inflammation and fibrosis.

Regarding innate immunity, TGF-β signaling inhibits the functions of antigen-presenting and phagocytic cells and disrupts the maturation of natural killer (NK) cells. In adaptive immunity, TGF-β suppresses B and T cell activity, induces B cell apoptosis, and impairs CD8⁺ cytotoxic T cells and CD4⁺ helper T cells. Dendritic cells (DCs) further impair T cell-mediated antitumor responses as TGF-β suppresses antigen presentation and DC migration. TGF-β also inhibits the expression of interferon-γ (IFN-γ) and cytolytic gene products essential for cytotoxic activity and regulates T regulatory cell populations in primary tumors and metastatic lymph nodes.^1,2,3,4,5

Myeloid cells, including macrophages, myeloid-derived suppressor cells (MDSCs), and neutrophils, are similarly affected by TGF-β. In response to TGF-β released by cancer cells or other cells in the tumor microenvironment, macrophages and neutrophils adopt tumor-promoting phenotypes (M2-like and N2-like, respectively). In addition, TGF-β suppresses the production of reactive oxygen species and nitric oxide, critical players in macrophage-mediated cytotoxicity.^1,2,3,4,5,6 It also influences processes such as epithelial-to-mesenchymal transition, cancer-associated fibroblasts (CAFs) activation, and angiogenesis.

TGF-β is a key regulator in the activation and differentiation of fibroblasts, which are central to fibrosis and the tumor microenvironment. It promotes transdifferentiation of fibroblasts into myofibroblasts, a cell type characterized by the high production of extracellular matrix (ECM) components, including collagen. In fibrotic diseases, TGF-β-driven fibroblast activation is crucial for wound healing; however, persistent TGF-β signaling favors an excessive and uncontrolled ECM deposition, leading to chronic fibrotic conditions affecting organs such as the lungs, heart, liver, and kidneys. In cancer, TGF-β activates CAFs, contributing to a dense, fibrotic stroma around tumors, known as desmoplasia. This rigid stroma provides structural support to the tumor but also forms a physical barrier, hindering immune cell infiltration and reducing the efficacy of drug delivery. Through paracrine signaling, CAFs further enhance cancer progression by stimulating angiogenesis, evading immune surveillance, and recruiting additional myofibroblasts.^7,8,9

Therefore, the clinical applications of TGF-β signaling blockade offer a promising research avenue to enhance the efficacy of current and forthcoming immunotherapies, especially in treating fibrosis and cancer, where TGF-β plays a role in pathophysiology.^10,11,12,13 Different strategies have been adopted to inhibit TGF-β signaling, ranging from antisense oligonucleotides to decrease TGF-β ligand production to blocking integrin-mediated activation of latent TGF-β ligands extracellularly, as well as sequestering TGF-β ligands in circulation using TGF-β receptors or monoclonal antibodies.^1,2,3,4,5 One notable approach, which is the focus of this comprehensive review, comprises the use of small molecules to inhibit TGF-β receptors (TGF-βRs).¹⁴

Small-molecule inhibitors targeting TGF-βRI have gained significant attention, with eight such inhibitors currently undergoing evaluation in clinical trials: galunisertib (LY2157299), vactosertib (TEW-7197), LY3200882, MDV6058 (PF-06952229), GFH018, YL-13027, AGMB-129 (ORG-129), and SH3051 (Figure 2).^{15,16,17,18,19,20,21,22,23,24,25,26,27,28,29} The development of this diverse array of TGF-βRI inhibitors has been facilitated by extensive structural insights into TGF-βRI, including the availability of several experimental X-ray crystal structures.^2,5,14,30

Figure 2.

Molecular structures of small-molecule TGF-βRI inhibitors in clinical trials with publicly released structures

In this comprehensive overview of small-molecule TGF-βRI inhibitors, we aim to compile and critically analyze all available structural data and known inhibitors, particularly those already patented by various pharmaceutical companies. We have consolidated critical structural information, currently dispersed across multiple sources, into a single document to aid the development of new inhibitors. Our analysis encompasses 24 high-resolution three-dimensional (3D) experimental structures obtained from the PDB database,³¹ all derived from Homo sapiens, with resolutions under 2.5 Å, acquired through X-ray diffraction. Additionally, we provide an extensive review of the TGF-βRI inhibitors patented between 2000 and 2023, building upon prior patent reviews covering patents from 2005 to 2008³² and 2015 to 2020.³³

TGF-βRI biology and structure

The TGF-βs form a subfamily comprising three mammalian ligand isoforms, TGF-β1, TGF-β2, and TGF-β3, which are synthesized by various cell types as large pre-pro-polypeptides.^2,34 The discovery of the first member, TGF-β1, in the 1980s initially highlighted its role in stimulating the growth of cultured fibroblasts.³⁵ These three active TGF-β ligand homodimer isoforms bind to the cysteine-rich N-terminal extracellular domain and signal via the TGF-βRII, which then recruits and activates TGF-βRI (Figure 1). The resulting heterohexameric complex initiates the intracellular signaling of TGF-β through the Smad-dependent canonical pathway and Smad-independent non-canonical pathways.

Canonical TGF-β signaling

The Smad-dependent canonical pathway begins with the phosphorylation of receptor-Smads (R-Smads), Smad2 and Smad3, by the activated TGF-βRI. Phosphorylated Smad2/3 dissociates from TGF-βRI and binds to the Co-Smad, Smad4, forming a heterotrimeric transcriptional complex. This complex translocates to the nucleus, binding to DNA to regulate TGF-β-responsive gene expression. TGF-β also induces the expression of inhibitory Smads Smad6 and Smad7, which serve as key negative regulators of the TGF-β/Smad signaling pathway. Smad7, in particular, exhibits a stronger suppressive effect than Smad6. By competitively binding to TGF-βRI and Smad4, Smad7 inhibits R-Smad phosphorylation and activation, blocking downstream signaling. Furthermore, Smad7 interferes with Smad-DNA complex formation and recruits E3 ubiquitin ligases, promoting the degradation of TGF-βRI, R-Smads, and Co-Smad through proteasomal or lysosomal pathways. This regulatory feedback mechanism helps maintain cellular homeostasis, prevents excessive TGF-β signaling, and mitigates pathological outcomes.^{8,36,37,38,39}

Non-canonical TGF-β signaling

In addition to the canonical pathway, TGF-β signaling can activate various non-canonical pathways, including extracellular signal-regulated kinase, Rho guanosine triphosphatase, p38 mitogen-activated protein kinase (MAPK), c-Jun N-terminal kinase (JNK) and nuclear factor-κB, phosphatidylinositol 3-kinase/alpha serine/threonine-protein kinase, Abelson murine leukemia viral oncogene homolog 1, and Janus kinase/signal transducer and activator of transcription (JAK/STAT), along with their respective downstream effectors. These pathways have been extensively reviewed in the literature.^8,39

TGF-β activation initiates a phosphorylation cascade that directly or indirectly targets serine and threonine residues in the linker region of Smad2 and Smad3, modulating their activity and cellular functions to regulate gene expression.^36,37,38 These non-canonical pathways regulate numerous processes, including cell proliferation, differentiation, migration, metabolism, and survival, and are implicated in diseases such as chronic inflammation, neurodegeneration, obesity, and cancers.⁸ Interestingly, TGF-β plays an immunosuppressive role by inhibiting the interleukin (IL)-12-mediated JAK2 and STAT activation in T lymphocytes, thereby reducing T cell proliferation and IFN-γ production. Consequently, the JAK/STAT pathway is critical to fibroblast activity and the regulation of the adaptive immune response.^8,40,41,42

At the structural level, TGF-βRI and TGF-βRII receptors consist of a cysteine-rich N-terminal extracellular domain for ligand binding, a transmembrane helix, and a C-terminal intracellular domain with serine/threonine kinase regions. TGF-βRI, encoded by the TGF-βRI gene on chromosome 9q22.33, has a molecular mass of about 53 kDa and consists of 503 amino acids, with the GS domain forming a helix-loop-helix structure (Figure 3). TGF-βRII, encoded by the TGFBR2 gene on chromosome 3p24.1, has a molecular mass of approximately 75 kDa and consists of 567 amino acids, lacking a GS domain.^30,43,44 TGF-βRIII, comprising betaglycan and endoglin, plays a supportive role, with betaglycan binding all TGF-β ligand isoforms and endoglin binding specifically to TGF-β1 and TGF-β3. TGF-β ligands interact with activin A receptor-like type 1 (ALK1), primarily in endothelial cells. This leads to Smad1, Smad5, and Smad8 phosphorylation and subsequent transcription regulation of specific TGF-β target genes.^2,45

Figure 3.

TGF-βRI structure

Crystallographic structure of TGF-βRI (PDB: 5QIM, in gray) highlighting key domains: GS region (light teal), αC helix (dark teak), phosphate binding loop (red), catalytic segment (yellow), and activation segment (blue). The close-up view displays the residues engaged in small-molecule interactions using a stick representation. Hydrogen atoms are not shown, nitrogen atoms are colored in blue, and oxygen atoms are red.

Exploring TGF-βRI: Structural insights from small-molecule inhibitor complexes

The first crystal structure of the unphosphorylated C-terminal cytoplasmic domain of human TGF-βRI was resolved by Huse et al.³⁰ in 1999 (PDB: 1B6C) at a resolution of 2.60 Å. This structure, showcasing the receptor in complex with the human protein inhibitor FKBP12, unveiled a canonical protein kinase fold. The structure comprises an N-terminal lobe consisting of a five-strand β sheet and a regulatory α helix (αC helix) alongside a predominantly helical C-terminal lobe (as illustrated in Figure 3), with the ATP-binding site located deep between these two lobes. The segment located between the β4 and β5 strands dictates Smad substrate specificity. It was revealed that TGF-βRI adopts an inactive conformation mediated by interactions between its unphosphorylated GS domain, the activation segment, and the N-terminal lobe in the presence of the inhibitor, preventing phosphorylation by TGF-βRII and contributing to the stabilization of the αC helix.

To further understand the phosphorylation mechanisms of the GS domain of TGF-βRI, Huse’s research group obtained the crystal structure of the unphosphorylated TGF-βRI in complex with the quinazoline derivative NPC-30345 in 2001 (PDB: 1IAS; resolution 2.9 Å).⁴⁶ However, the chemical structure of NPC-30345 was not disclosed in the refined crystal structure. NPC-30345, reported by Scios (WO/2000/12497), targeting both p38-α and TGF-β signaling pathways, occupies the ATP binding site, similar to the TGF-βRI/FK506-binding protein (FKBP)12 complex. Unlike FKBP12, which induces an inactive conformation by wedging between the αC helix and the N-lobe β sheet of the receptor, NPC-30345 extends toward the rear of the ATP binding pocket, fitting between Leu278 and Ser280. This hydrophobic cavity is not well conserved across other kinases, as most have a bulkier residue at the position corresponding to Ser280.

In 2003, Ling from Biogen and Sawyer et al. identified HTS466284 as a potent TGF-βRI inhibitor.⁴⁷ Ling’s research group performed a TGF-βRI shape-based virtual screening using five pharmacophore features of the 2,4,5-triarylimidazole SB203580, a p38 MAPK inhibitor also known to inhibit TGF-βRI inhibitor, with a half-maximal inhibitory concentration (IC₅₀) of 20 μM.⁴⁸ HTS466284 was identified as a potent TGF-βRI inhibitor, with an IC₅₀ of 27 nM.⁴⁹ Similarly, Sawyer et al. identified HTS466284 as a TGF-βRI inhibitor, with an IC₅₀ of 51 nM in an in vitro screening of an extensive library of compounds.⁴⁷

The X-ray structure of the TGF-βRI-HTS466284 complex (PDB: 1PY5; resolution 2.3 Å) reveals the binding of the inhibitor to the ATP site, establishing multiple hydrogen bonds and van der Waals interactions with TGF-βRI, thereby explaining its potency (Figure 4A). These interactions included hydrogen bonds between the pyrazole N2 and the side-chain nitrogen of Lys232, the quinoline and the backbone amide of His283, and the pyrazole N1 and the carboxylate oxygen of Asp351. Additionally, a tetrahedral complex involving the 2-pyridyl nitrogen, a receptor-bound water molecule, the phenol of Tyr249, and the backbone amide of Asp351 was observed. The quinoline moiety of the inhibitor mimics the adenine group of ATP, establishing van der Waals contacts with Ile211, Ala230, and Leu340. Furthermore, the pyridine ring interacts with a hydrophobic pocket involving Ala230, Lys232, Leu260, Leu278, Val279, and Ser280. Like NPC-30345, HTS466284 stabilizes the inactive conformation of the receptor, hindering ATP binding and substrate phosphorylation.

Figure 4.

Pyrazole-based TGF-βRI inhibitors

Molecular structures of SB203580, HTS466284, and compounds derived from the latter with improved activity. The half-maximal inhibitory concentration (IC₅₀) is displayed below each molecular structure for TGF-βRI and p38 MAPK. Crystallographic structures of TGF-βRI with (A) HTS466284 in blue (PDB: 1PY5) and (B) compound 2 in brown (PDB:1RW8) are displayed. The key domains are depicted in transparency mode: GS region (light teal), αC helix (teal), phosphate binding loop (red), catalytic segment (yellow), and activation segment (blue). Hydrogen bonds are represented in black, and water molecules are in red. Molecular structure of SB431542, an inhibitor of TGF-βRI, which was co-crystallized with TGF-βRI by Ogunjimi et al.⁵⁰ On the right, (C) displays the alignment and superimposition between the crystal structures of TGF-βRI bound to HTS466284 (blue) (PDB: 1PY5) and SB431542 (pink) (PDB) 3TZM, which were made using MOE version 2022.01.⁵¹ The orientation of the side chain of Asp351 is highlighted. Nitrogen atoms are colored in blue, oxygen atoms in red, and a fluorine atom is represented in violet.

Building on the structural insights of Sawyer et al., the team pursued a structure-activity relationship (SAR) study to refine HTS466284 further.^47,52 They confirmed that critical features for effective TGF-βRI inhibition included a 2-pyridyl ring at the 3-position of pyrazole and an aryl or heteroaryl substituent at the 4-position, incorporating a hydrogen bond acceptor. This confirmation resulted in TGF-βRI inhibitors 1, 2, and 3 with improved activity. The crystal structures of TGF-βRI in complex with HTS466284, 5,6-dihydro-4H-pyrrol[1,2-b]pyrazole compounds 1 and 2 (PDB: 1RW8; resolution 2.4 Å), and a 4-aryl-substituted pyrazole 3 elucidated their binding mode.^47,52 Compounds 1, 2, and 3 bind to the ATP binding site of TGF-βRI, such as HTS466284, establishing a hydrogen bond with the pyridyl nitrogen and a water molecule held in place by hydrogen bonds involving Asp351, Glu245, and Tyr249. The additional hydrogen bond acceptor interacting with the backbone amide of His283 (fluorine atom in 1, quinoline nitrogen in 2 [Figure 4], and the hydroxyl oxygen in 3) is the key determinant in achieving selectivity of these inhibitors for TGF-βRI over p38 MAPK (Figure 4B). Moreover, the TGF-βRI hinge region, connecting the N- and C-terminal kinase lobes, accommodates weaker acceptors like aromatic fluorine. In contrast, the more flexible hinge of p38 MAPK requires stronger acceptors. The pyridine nitrogen in TGF-βRI and p38 MAPK inhibitors interacts with active site water molecules. However, due to distinct hydrophobic pocket gatekeeper residues (Ser280 in TGF-βRI and Thr106 in p38 MAPK), the 4-phenyl-substituted series prefers the TGF-βRI domain over p38 MAPK.

The pivotal role of the interaction between small-molecule inhibitors and the gatekeeper residue Ser280 in driving TGF-βRI selectivity was further demonstrated by Ogunjimi et al.⁵⁰ Their findings revealed that the Ser280Thr mutant conferred resistance to SB431542 inhibition. In contrast, the ALK2 Thr283Ser mutant exhibited sensitivity to SB431542 inhibition (Ser280 is the only differing gatekeeper residue between TGF-βRI and ALK2).

Further investigation into the selectivity of SB31542 involved its co-crystallization with the TGF-βRI kinase domain (PDB: 3TZM; resolution 1.7 Å), revealing its binding site to the ATP binding site and adjacent hydrophobic areas near the hinge region (Ala230, Leu260, and Tyr282). Hydrogen bonding occurs between the benzodioxol oxygen of SB431542 and the amide nitrogen of His28 3and between Lys232 and the imidazole core of the inhibitor. Notably, the orientation of the side chain of Asp351 is the major structural distinction between the 1PY5 (HTS466284) and 3TZM (SB431542) crystal structures (Figure 4C), suggesting its pivotal role in accommodating a diverse range of inhibitors with varying steric profiles in the hinge region.

In 2004, Gellibert et al. from GlaxoSmithKline initiated a program focused on the discovery of novel small-molecule TGF-βRI inhibitors.⁵³ They identified the aminothiazole hit compound 4, with an IC₅₀ of 274 nM, against TGF-βRI autophosphorylation and found it to be inactive against p38 MAPK. The hit-to-lead optimization was facilitated by docking studies using GOLD version 1.1 and the TGF-βRI crystal structure (PDB: 1B6C). In these docking studies, a hydrogen bond constraint was applied to ensure interaction between the quinolone N1 or naphthyridine N5 and the backbone amide of His283. The active site center was determined using a flood fill radius of 12 Å from a dummy atom at the centroid created from the heavy atoms of Lys232, Leu260, Ser280, His283, and Asp351 residues. These studies proposed the naphthyridine system as a potential alternative to the quinoline moiety. To explore other heterocyclic alternatives for the central core, the team investigated 1,5-naphthyridin-4-yl aminothiazole and pyrazole derivates. The two pyrazole compounds 5 and 6 showed comparable activity against TGF-βRI and maintained selectivity over p38 MAPK, similar to the aminothiazole analogs 7 and 8 (Figure 5). Notably, compound 6 exhibited selectivity toward TGF-βRI over a range of other kinases, including p38 MAPK, JNK1, JNK3, ITK, MSK1, GSK3, MLK3, B-RAF, and LCK, at a concentration of 16 μM.

Figure 5.

Aminothiazole- and pyrazole-based TGF-βRI inhibitors

Molecular structures of the aminothiazole derivative compounds 4, 7, and 8, and of the two pyrazole derivatives compounds 5 and 6 (analogs of compounds 7 and 8, respectively), identified by Gellibert et al. The IC₅₀ against TGF-βRI autophosphorylation is indicated below each molecular structure. All these molecules exhibited selectivity for TGF-βRI over p38 MAPK at 16 μM.⁵³ In the lower-left corner, an amplified stereo view of compound 6 (colored in salmon) within the TGF-βRI binding site (PDB: 1VJY) highlights the residues in contact with compound 6. Hydrogen bonds are shown as black dashed lines, and a water molecule is depicted by a red sphere. The color scheme is consistent with that described in Figures 3 and 4.

A crystal structure of the TGF-βRI-compound 6 complex was obtained (PDB: 1VJY, resolution 2 Å) and showed that the inhibitor occupies the ATP-binding site of TGF-βRI (Figure 5). The N1 of the pyrazole establishes a hydrogen bond with Asp351’s side chain, and the pyrazole’s N2 hydrogen bonds with Lys232’s side chain. The N1 of the pyridine interacts with a water molecule that forms additional hydrogen bonds with the side chains of Tyr249, Glu245, and the backbone of Asp351, while the 1,5-naphthyridine N5 interacts with the backbone amide of His283.The 6-methyl substituent occupies a hydrophobic pocket, with the pyridine ring positioned close to Ser280’s side chain. The presence of a pyridine or phenyl ring substituted with small electron-withdrawing groups was deemed essential due to two interactions in this region: a hydrogen bond with the pyridyl N1 and an electronic interaction with Ser280’s oxygen.

In a subsequent study, Gellibert et al.⁵⁴ expanded their arsenal of TGF-βRI inhibitors by modifying the quinazoline compound 9, which exhibited moderate activity against TGF-βRI and p38 MAPK, with IC₅₀ values of 194 and 1,270 nM, respectively (Figure 6). The 2-phenyl group was substituted with a 6-methyl-pyridin-2-yl moiety, resulting in GW855857 (Figure 6). This modification significantly increased binding affinity (with an IC₅₀ of 25 nM) and selectivity to TGF-βRI. Additionally, the team explored replacing the 4-aminopyridine moiety in GW855857, yielding the compound N-1H-indazol-5-yl GW857175 (Figure 6), which exhibited potency comparable to GW855857 (with an IC₅₀ of 25 nM). Notably, GW857175 displayed superior selectivity toward TGF-βRI, showing no inhibitory activity against p38 MAPK, even at concentrations >10 μM. Crystallographic analysis by the Gellibert group of TGF-βRI/GW857175 (PDB: 3GXL; resolution 1.8 Å) and TGF-βRI/GW855857 (PDB: 3HMM; resolution 1.7 Å) complexes revealed interaction patterns similar to previous inhibitors. In the GW855857 complex, a hydrogen bond formed between the 4-pyridine nitrogen and the backbone amide of His283. Conversely, in the GW857175 complex, hydrogen bonds between nitrogen atoms N5 and N6 with the backbone amide and carboxyl groups of His283 were observed, respectively. Additionally, GW855857 demonstrated interaction with a water molecule, influencing the positioning of the amide spacer at the C4 position of the quinazoline ring.

Figure 6.

Quinazoline- and pyridine-based TGF-βRI inhibitors

Molecular structure of the quinazoline derivatives 9, GW855857, GW857175, and 4-pyridinoxy-2-anilinopyridine derivatives 10, 11, and 12 identified by Gellibert et al. and Goldberg et al., respectively.^54,55,56 The IC₅₀ against TGF-βRI autophosphorylation is indicated below each molecular structure. (A) Amplified stereo view of compound 10 (depicted in light brown) and compound 11 (depicted in green) within the binding site of TGF-βRI (PDB: 2WOU and 2WOT, aligned and superimposed using MOE version 2022.01⁵¹), highlighting the residues in contact with these compounds. (B) Stereo view of compound 12 (depicted in light blue) within the binding site of TGF-βRI (PDB: 5FRI). A chlorine atom is shown in brown, and a sulfur atom is shown in light green. Hydrogen bonds are represented by black dashed lines, and a water molecule is depicted as a red sphere.

In 2009, Goldberg et al. from AstraZeneca⁵⁵ unveiled the 4-pyridinoxy-2-anilinopyridine derivatives as a novel class of TGF-βRI inhibitors. These compounds were discovered through molecular modeling and docking studies using Glide software and the 1RW8 TGF-βRI crystal structure. Docking calculations were centered on 5,6-dihydro-4H-pyrrolo[1,2-b]pyrazole compound 2, a ligand of the 1RW8 structure. Among these compounds, compound 10, containing a bipyridyl moiety, exhibited an IC₅₀ of 44 nM.

Despite its potency, compound 10 displayed poor oral bioavailability in rat models (Figure 6). The crystal structure of compound 10 in complex with TGF-βRI (PDB: 2WOT; resolution 1.85 Å) validated its binding mode and guided SAR studies. The structure illustrated interactions between Lys232 and the first pyrimidine ring (ring 1), although the impact of this interaction on inhibitor potency remained unclear. In addition, the second ring (ring 2) was crucial for binding through interaction with a conserved water molecule, while the 2-amino-4-oxo pyridyl moiety interacted with the hinge region. Furthermore, the trimethoxyaniline moiety occupied a solvent-exposed channel within the binding pocket of the receptor.

Based on these findings and scaffold optimization, compound 11 was developed, with improved pharmacokinetic properties (Figure 6). Compound 11 demonstrated an IC₅₀ of 72 nM, high bioavailability, favorable oral exposure, low molecular weight of 370 Da, appropriate lipophilicity (logD7.4 = 2.5), and remarkable selectivity toward TGF-βRI over a panel of 80 representative kinases. Only 5 kinases showed more than 50% inhibition at 1 μM. The crystal structure of the TGF-βRI-compound 11 complex (PDB: 2WOU; resolution 2.3 Å) revealed notable differences from compound 10. Compound 11 did not interact with the Lys232 residue and instead formed an additional interaction between the amine of its sulfonamide group and the carboxyl group of Asp290 (Figure 6A).

In pursuit of new ALK1 inhibitors to block angiogenesis in cancer, Goldberg et al.⁵⁶ identified N-(4-anilino-2-pyridyl)acetamide 12, showing strong inhibitory activity (IC₅₀ of 1.1 nM) against ALK1, ALK2/3, and TGF-βRI, with less activity against ALK4 and ALK6. Despite sharing a similar ATP binding site, a key distinction lies in the gatekeeper residue, serine in ALK4/5/7 and threonine in ALK1/2/3/6. The crystal structure of the TGF-βRI-compound 12 complex (PDB: 5FRI; resolution 2 Å) revealed interaction patterns similar to previously described inhibitors. The chlorodioxopyridine moiety of compound 12 resided in the hydrophobic selectivity pocket of the receptor, establishing a hydrogen bond with a water molecule that interacted with Glu245 and Tyr249 side chains. The aminopyridine moiety engaged in a donor-acceptor interaction with the hinge residues Tyr282 and His283, while the cyclopropyl group localized itself in the solvent-exposed channel (Figure 6B).

In 2014, Czodrowski et al.⁵⁷ co-crystallized different fragments with TGF-βRI, aiming to identify fragments suitable for fragment-based screening. Compound 13, identified from the fragmentation of a tyrosine kinase receptor (TIE2) inhibitor, served as a core fragment, along with the four additional fragments (14–17) from their proprietary compound database (Figure 7). Their binding modes were validated through docking studies, molecular dynamics simulations, and SZMAP/WaterMap calculations, predicting water molecule positioning and assessing their role in the binding modes. Three binding modes were identified, with compounds 13 and 14 sharing one mode, compound 15 having another, and compounds 16 and 17 exhibiting a third. These modes correlated with the placement and features of substituents on the core fragment, further elucidated by considering water molecule replacements. Additionally, insights into the water patterns in the absence of bound ligands within the hinge region were provided by the apo structure of TGF-βRI (PDB: 4X2N; resolution 1.8 Å).

Figure 7.

Staurosporine selectivity toward TGF-βRI

Molecular structure of the fragments evaluated in complex with TGF-βRI by Czodrowski et al. The name of the fragment, the PDB code, and the resolution of the crystal structure are indicated below each molecular structure.⁵⁷ Molecular structures of 18 and staurosporine. The IC₅₀ against TGF-βRI and TGF-βRII are provided below each molecular structure. (A) Stereo view of compound 18 (depicted in light brown) and (B) staurosporine (depicted in orange) in complex with TGF-βRI (PDB: 5E8Z and 5E8W, respectively). The stereo view of the staurosporine-TGF-βRI complex illustrates only interactions contributing to TGF-βRI selectivity over compound 18. Hydrogen bonds are shown as black dashed lines, and a red sphere represents a water molecule.

Bristol-Myers Squibb significantly contributed to TGF-βRI structural elucidation by resolving several crystal structures, including those of the wild-type TGF-βRI (PDB: 5E8S; resolution 1.45 Å) and a constitutively active form resulting from the Thr204Asp mutation (PDB: 5E8T; resolution 1.7 Å).⁵⁸ Notably, the Thr204Asp mutation primarily affects the N terminus of TGF-βRI, with minimal impact on the active site. This mutation alters the positioning of Trp224 and Arg225 side chains, resulting in a root-mean-square deviation of 0.2 Å across 285 αC helix atom pairs.⁵⁹

To compare binding modes between TGF-βRI-Thr204Asp and TGF-βRII, the TGF-βRI-Thr204Asp complex was co-crystallized with two inhibitors, staurosporine (PDB: 5E8W; resolution 1.86 Å) and compound 18 (PDB: 5E8Z; resolution 1.51 Å) (Figures 7A and 7B).⁵⁹ Compound 18 inhibits both TGF-βRI and TGF-βRII, whereas staurosporine shows a 60-fold higher affinity for TGF-βRI. Analysis revealed hydrogen bonds with hinge residues His283 and Asp281, as well as hydrophobic interactions and hydrogen bonds mediated by a water molecule to Ser287. Compound 18 forms a hydrogen bond with Lys232, a feature also retained in the TGF-βRII structure. The selectivity of staurosporine for TGF-βRI over compound 18 is attributed to its methoxy group’s fitting into a pocket lined by TGF-βRI residues Asn338, Leu340, and Ala350 at the base of the ATP binding site, interactions that are not possible in TGF-βRII due to steric hindrances by Cys396. Staurosporine also forms a hydrogen bond with the backbone carbonyl of Lys337 (Figure 7B). Based on these findings, Tebben et al.⁵⁹ proposed that selective inhibitors for TGF-βRI should involve groups projecting into the Asn338/Leu340/Ala350 pocket and moieties interacting with gatekeeper Ser280. The crystal structures of TGF-βRII/compound 18 (PBD: 5E91) and TGF-βRI/SB431542 (PDB: 3TZM) revealed steric clashes with the TGF-βRII gatekeeper Thr325, explaining the selectivity of SB431542 for TGF-βRI.

Tebben et al. also reported novel heterobicyclic inhibitors targeting TGF-βRI to block the SMAD-dependent canonical pathway (Figure 8).⁶⁰ The crystal structure of compound 19 complexed with TGF-βRI-Thr204Asp (PDB: 6B8Y; resolution 1.65 Å), validated the binding mode as predicted by SAR studies. Key interactions include three hydrogen bonds: one between the aminopyridine nitrogen of the inhibitor and the backbone amide of His283, another between the pyrrole nitrogen and the side-chain carbonyl of Asp351, and the third between pyrimidine nitrogen N1 and Lys232. Additionally, the 2-pyridyl nitrogen interacts with a water molecule, which also engages with Asp351, Glu245, and Tyr249. However, this inhibitor demonstrated poor aqueous solubility.

Figure 8.

Selective TGF-βRI inhibitors

Molecular structure of compounds 19–24 and BMS-986260 evaluated by Bristol-Myers Squibb.^60,61 (A) Stereo views of compound 21 (depicted in brown) in complex with TGF-βRI (PDB: 5QIL) and (B) TGF-βRII (PDB: 5QIN). (C) Stereo view of compound 22 (depicted in light pink) in complex with TGF-βRI (PDB: 5QIM). Compound 22 exhibits a distinctive feature, with two water molecules crucially involved in the binding pocket. The second water molecule is positioned adjacent to Ala350, a position that would be occupied by the side chain of Cys396 in TGF-βRII. This amino acid variance likely contributes to the selectivity of compound 22 for TGF-βRI over TGF-βRII. Hydrogen bonds are represented by black dashed lines, and red spheres denote water molecules.

TGF-β signaling is increasingly recognized as a major obstacle to effective immune system activation, particularly in the context of treatments involving immune checkpoint inhibitors.⁶² This recognition has led to extensive research into targeting TGF-β signaling to enhance cancer immunotherapies, including those targeting programmed cell death protein 1/programmed cell death ligand 1 (PD-1/PD-L1).^1,3,4,63,64 TGF-β and PD-L1 mediate independent yet complementary immunosuppressive pathways. TGF-β suppresses immune cell activation and infiltration, while PD-1/PD-L1 inhibits T cell cytotoxicity against tumor cells. Dual inhibition enhances T cell infiltration into tumor cores, reduces regulatory T cells (Tregs), and restores sensitivity to anti-PD-L1 therapy. This dual blockade effectively modulates the tumor microenvironment, reducing immunosuppressive mediators and increasing immune cell accessibility, thereby creating conditions that are less favorable to tumor growth.^{63,64,65,66,67} Recognizing the synergistic potential of combining TGF-β signaling blockade with anti-PD-1 therapy in cancer treatment, researchers at Bristol-Myers Squibb undertook a study to develop a highly selective TGF-βRI inhibitor.⁶¹ This study involved the high-throughput screening of 12,000 small molecules, leading to the identification of a novel pyrrololactam core structure. However, the identified compound 20 showed high binding affinity against both TGF-βRI and TGF-βRII. Structural analysis of the TGF-βRI-Thr204Asp/compound 20 crystal structure (PDB: 5QIK; resolution 1.58 Å) revealed typical hydrogen bond interactions with the hinge residue His283 through the pyridinylacetamide of the inhibitor. The fluoropyridine nitrogen of the compound formed a hydrogen bond with a water molecule, which in turn interacted with Tyr249, Glu245, and Asp351. In addition, the lactam amide of the inhibitor established hydrogen bonds with Lys232 and Asp351.

Looking for TGF-βRI selectivity, the pyrrololactam core was replaced by a 4-azaindole moiety. This modification retained inhibitory activity while enhancing selectivity for TGF-βRI over a broad panel of 240 kinases. Finally, the 3-pyridyl azaindoles 21 and 22 exhibited 13-fold and 2,500-fold higher selectivity for TGF-βRI compared to TGF-βRII, respectively (Figure 8).

Crystal structures of these compounds with TGF-βRI/II were obtained (PDB: 5QIL, 5QIM, and 5QIN for TGF-βRI-Thr204Asp/compound 21, TGF-βRI-Thr204Asp/compound 22, and TGF-βRII/compound 21, respectively; resolutions 1.57–1.98 Å). Both compounds exhibited a consistent binding mode to TGF-βRI in these structures, forming a hydrogen bond with His283 and interacting with Lys232 (Figures 8A–8C). Compound 21 maintained the interaction with the conserved water molecule via the nitrogen of its 3-pyridyl group, while compound 22 interacted with a different water molecule near Ala350. The high selectivity of compound 22 is attributed to this unique water-mediated interaction, as the corresponding Ala350 residue in TGF-βRI is replaced by Cys396 in TGF-βRII, which could disrupt this interaction (Figure 8C). However, compound 22 also demonstrated potential inhibition of the cytochrome P450 isozyme 3A4. Compound 23 was selected for further preclinical toxicology studies. This inhibitor demonstrated effectiveness in overcoming Treg-mediated immune suppression, showed good oral bioavailability, a favorable pharmacokinetic profile, and enhanced antitumor immunity when combined with an anti-PD-1 antibody in the MC38 syngeneic mouse colon adenocarcinoma tumor model.

More recently, Bristol-Myers Squibb has focused on developing novel TGF-βRI inhibitors featuring an imidazole scaffold as potential therapeutic agents. They co-crystallized two such inhibitors (compound 24 and BMS-986260) with a mutated form of TGF-βRI, specifically Thr204Asp. The crystallographic structure with compound 24 (PDB: 5QTZ; resolution 1.83 Å) showed that the imidazo-pyridine nitrogen of the compound formed a standard hydrogen bond with the hinge residue His283. Additional hydrogen bonds were observed between the imidazole group and Lys232 and between the difluoromethyl group and Lys337. Furthermore, a water molecule mediated a hydrogen bond between the 2-pyridyl group and Tyr249. However, compound 24 demonstrated limited metabolic stability in rat liver microsomes.

BMS-986260 emerged as a more balanced candidate regarding potency, selectivity, solubility, pharmacokinetic profile, and efficacy. The crystal structure of the TGF-βRI-Thr204Asp complex with BMS-986260 (PDB: 5QU0; resolution 1.67 Å) revealed that it retained all the characteristic interactions observed with compound 24.⁶⁸ This interaction profile includes the hydrogen bond formed between the hydroxyethyl group of the imidazole and Lys337, indicating a consistent binding mode for these imidazole-containing imidazopyridine inhibitors within the TGF-βRI binding site. Overall, BMS-986260 exhibited potent inhibitory activity, with an IC₅₀ below 1 μM against only 10 of the 213 tested kinases, including ALK2, ALK4, and TGF-βRI. Moreover, it showed promising oral efficacy in combination with an anti-PD-1 antibody in the MC38 murine colon carcinoma syngeneic model.

Patented TGF-βRI inhibitors

Our TGF-βRI patent investigation was carried out using the PATENTSCOPE database.⁶⁹ We used a targeted search employing key words such as “ALK5,” “TGF-ΒRI,” “TGFBRI,” “TGF-BR1,” and “TGFBR1” within the “english_all” search field, and we limited the “Publication Date” range from January 1, 2000 to May 1, 2024. Inclusion criteria were patents in English that detail biochemical assays for TGF-βRI, specifically those reporting IC₅₀ values for TGF-βRI inhibitors. Our search yielded a dataset of 86 patents filed by 29 different entities.

GlaxoSmithKline Plc. (SmithKline Beecham Corporation and Glaxo Group Ltd.)

The SmithKline Beecham Corporation was a trailblazer in the pharmaceutical industry, dedicating significant resources toward developing small-molecule inhibitors targeting TGF-βRI. Their efforts led to a series of impactful patent filings from 2000 to 2004.

During the early phase, from 2000 to 2003, the company filed patents for a diverse range of compounds, including triarylimidazole, imidazolyl cyclic acetal, and triazole derivatives, all identified as potent TGF-βRI inhibitors. These compounds exhibited a wide range of IC₅₀ values, from 0.0001 to 10 μM.^{70,71,72,73,74,75}

Subsequent developments saw a more focused approach, with patents claiming compounds having a narrower IC₅₀ range between 0.4 and 300 nM. These compounds included a variety of chemical structures such as benzoxazine-, benzoxazinone-, and pyridinyl-substituted triazoles, aminothiazole, pyrazole, imidazopyridine, quinoline, quinazoline, and 2-phenylpyridin-4-yl heterocyclyl derivatives, all characterized as TGF-βRI inhibitors.^{76,77,78,79,80,81,82,83,84}

In 2002, Glaxo Group Ltd. made a significant contribution by introducing 84 TGF-βRI inhibitors, each demonstrating an IC₅₀ of 1 μM or lower. These inhibitors were based primarily on thiazolamine, thiazole, and pyrazole derivatives.^{85,86,87,88,89}

Further expanding the landscape of TGF-βRI inhibitors, in 2005, 30 derivatives of 1-amino-isoquinoline were patented, all showing IC₅₀ values at or below 10 μM.⁹⁰ Figure S1 provides a comprehensive overview of the general structures and specific examples of these patented TGF-βRI inhibitors, showcasing the wide-ranging efforts of GlaxoSmithKline Plc. in this area of pharmaceutical development.

Eli Lilly and Company

Eli Lilly and Company, a renowned player in the pharmaceutical industry, has been actively engaged in the development of TGF-βRI inhibitors. Their contributions to this field have been marked by a series of patent filings and the development of clinically significant compounds.

In 2002, the company patented 13 pyrrole derivatives as TGF-βRI inhibitors, all demonstrating an IC₅₀ of 20 μM or less. These compounds were distinctively characterized as not inhibiting other targets such as TGF-βRII, mitogen-activated protein kinase p38α, and vascular endothelial growth factor receptor-2.⁹¹ Progressing from 2004 to 2005, Eli Lilly and Company expanded their repertoire by patenting an array of novel inhibitors, including quinolinyl-pyrrolopyrazoles, pyrazoloazepine, condensed pyrazole, and fused pyrazole derivatives. These compounds maintained the same efficacy range in terms of IC₅₀ values.^92,93,94,95

A notable escalation in potency was observed in patents filed in 2006 and 2016, where they introduced compounds with an IC₅₀ of less than 1 μM.^96,97 Among these, galunisertib (identified in the literature as formula II/example 2 of WO/2004/048382) and LY3200882 (example 1 of WO/2016/057278) stand out for their clinical significance. Both compounds are being developed in clinical trials.^17,18,20 Particularly noteworthy is LY3200882, which demonstrated a remarkably low IC₅₀ value of 27 nM, indicating its high potency as a TGF-βRI inhibitor. Figure S2 offers a comprehensive overview of their significant contributions to the advancement of TGF-βRI inhibitor developments. It provides a detailed representation of the general structures and specific examples of these patented inhibitors from Eli Lilly and Company.

Pfizer (and Medivation)

Pfizer, in collaboration with Medivation, has been instrumental in advancing the field of TGF-βRI inhibitors, as evidenced by their diverse range of patented compounds and their application in clinical settings.

In 2004, Pfizer filed patents for compounds featuring core structures such as pyrazole, triazole, imidazole, isothiazole, isoxazole, oxazole, or thiazole rings, along with fused heteroaromatic derivatives. These compounds were characterized by their selectivity for TGF-βRI over TGF-βRII and TGF-βRIII, showcasing in vitro IC₅₀ values lower than 10 μM.^{98,99,100,101,102,103} This range indicates a broad spectrum of activity against TGF-βRI.

A significant advancement came in 2008, when Pfizer patented nine pyrazolyl thienopyridine compounds, with IC₅₀ values spanning 4.33–108 nM.¹⁰⁴ The efficacy of these compounds was further validated when Thirona Bio acquired a patent in 2022, using these nine compounds to develop several formulations for inhaled therapeutic delivery. These formulations targeted a range of pulmonary TGF-βRI diseases, including idiopathic pulmonary fibrosis, idiopathic interstitial pneumonia, scleroderma-associated interstitial lung disease, sarcoidosis, cystic fibrosis, lung cancer, and COVID infection, demonstrating the versatility and potential clinical applications of these TGF-βRI inhibitors.¹⁰⁵

In 2015, Medivation obtained a patent for aminopyridine derivatives designed to treat conditions associated with excessive TGF activity. These compounds underwent in vitro testing against several TGF family Ser/Thr kinases, including ALK1, ALK2, ALK3, ALK4, TGF-βRI, and ALK6, with some showing selectivity for TGF-βRI.¹⁰⁶ Notably, one of the patented compounds, PF-06952229/MDV6058, entered a phase 1 clinical trial. While the trial was ultimately terminated due to strategic decisions rather than safety concerns,²⁴ this compound has demonstrated an IC₅₀ of 24.7 nM against TGF-βRI; however, it also inhibits ALK1 and ALK4 (IC₅₀ of 7.29 and 56.7 nM, respectively).¹⁰⁶ Figure S3 provides a comprehensive overview of the general structures and examples covered by the patent claims of Pfizer and Medivation.

Novartis AG

Novartis AG has been a key player in the field of TGF-βRI and/or ALK4 receptors, as evidenced by their patents filed in 2008 and 2009. Their focus has been predominantly on compounds that inhibit the kinase activity of TGF-βRI, with less emphasis on ALK4 assays. The TGF-βRI compounds patented by Novartis AG include TGF-βRI 2,3,7-substituted imidazo-[1.2b]-pyridazines, pyrimidine, and pyridine derivatives. These compounds demonstrated impressive inhibitory activity, with most exhibiting IC₅₀ values below 2 μM and a significant number achieving values below 1 μM.^107,108,109 In addition to these, Novartis AG has patented pyrrolopyrimidines, pyrrolopyridines, imidazo[1,2-a]pyridine, and imidazo-pyridine derivatives. These compounds further demonstrate the company’s commitment to developing effective TGF-βRI inhibitors, with most showing IC₅₀ values lower than 10 μM, often less than 1 μM.^{110,111,112,113} Figure S4 provides a detailed overview of the general structures and examples of compounds covered by the patent claims of Novartis AG.

Merck Patent GmbH

Merck Patent GmbH also substantially contributed to this field between 2009 and 2010, developing compounds based on 5-cyano-thienopyridines, thienopyrimidines, and imidazothiadiazoles.^114,115,116 The kinase activity assessments for TGF-βRI of these compounds have shown promising results, with TGF-βRI IC₅₀ values lower than 820 nM. Continuing their efforts, between 2010 and 2011, Merck Patent GmbH expanded its range to include 68 alkoxy-thienopyrimidines, along with 93 heteroarylaminoquinolines and 1,8-naphthyridines, all of which exhibited IC₅₀ values below 1 and 10 μM, respectively.^117,118,119 Figure S5 provides a detailed overview of the general structures and examples of compounds covered by the patent claims of Merck Patent GmbH.

Takeda Pharmaceutical (and Millennium Pharmaceuticals)

In 2004, Millennium Pharmaceuticals made a significant contribution to the field of TGF-βRI research by reporting eight compounds that exhibited IC₅₀ values below 10 μM in TGF-βRI phosphorylation assays. These compounds were noteworthy for their selectivity, displaying a 50- to 100-fold preference for TGF-βRI over ALK6 and p38 kinases.¹²⁰ This discovery marked an important step in identifying potent and selective inhibitors for TGF-βRI, contributing to the advancement of targeted therapies in this area.

Continuing this trend, Takeda Pharmaceutical in 2011 patented a series of pyrazolo[4,3-b]pyridine-7-amine inhibitors. These inhibitors were distinguished by their robust inhibitory activity against TGF-βRI, with IC₅₀ values mostly below 100 nM and all under 100 μM.¹²¹ This achievement underscored Takeda’s commitment to developing highly potent inhibitors for TGF-βRI, potentially opening new avenues for therapeutic interventions.

Figure S6A presents a comprehensive overview of the general structures of the compounds covered by these claims, along with specific examples.

In2Gen, SK Chemicals, and Ewha University-Industry Collaboration Foundation

In 2005, a collaborative effort by In2Gen, SK Chemicals, and Ewha University-Industry Collaboration Foundation culminated in a patent describing 2-pyridyl substituted imidazoles as TGF-βRI inhibitors of TGF-βRI and/or ALK4. These compounds demonstrated good inhibitory activity, with IC₅₀ values less than 10 μM. Notably, several of these molecules exhibited IC₅₀ values lower than 1 μM, and a few even exhibited exceptionally potent IC₅₀ values of less than 50 nM.¹²²

Building on this process, SK Chemicals further expanded the scope of 2-pyridyl substituted imidazole inhibitors in patents filed in 2009 and 2013. These inhibitors maintained similar IC₅₀ ranges for TGF-βRI and/or ALK4, reinforcing the potential of this compound class in targeting these receptors.^123,124 In a notable development in 2012, Ewha University-Industry Collaboration Foundation introduced a new generation of more potent 2-pyridyl-substituted imidazoles as inhibitors of TGF-βRI and/or ALK4. These novel inhibitors displayed even lower IC₅₀ values against TGF-βRI, some under 1 μM, with certain compounds achieving values below 0.1 μM and a few demonstrating an IC₅₀ less than 10 nM. One particularly significant compound from this research is vactosertib, highlighted in their patent (example 2 of WO/2012/002680). Vactosertib stands out with an IC₅₀ of 6.68 nM against TGF-βRI and 17.3 nM against ALK4, while showing limited activity against p38α (IC₅₀ of 1.72 μM). Its kinase selectivity profile indicates inhibition of ALK1, ALK2, and ALK3 at 66%, 71%, and 27%, respectively, at a concentration of 10 μM.¹²⁵ Currently, vactosertib is being evaluated in phase 1/2 clinical trials.²⁵

Figure S6B provides a detailed overview of the general structures of these 2-pyridyl substituted imidazoles, along with specific examples.

Bristol-Myers Squibb

In 2015, Bristol-Myers Squibb embarked on a collaborative venture with Rigel Pharmaceuticals, focusing on the development of potent and selective TGF-βR inhibitors for potential use in immuno-oncology. This collaboration was marked by Bristol-Myers Squibb’s acquiring exclusive global rights to advance and market a range of orally bioavailable, potent, and selective small-molecule TGF-βRI inhibitors. These inhibitors, initially developed by Rigel, have demonstrated promising in vivo efficacy in preclinical animal cancer models, particularly targeting TGF-βRI and TGF-βRII receptors.¹²⁶

During the period 2016–2018, Bristol-Myers Squibb patented a variety of small molecules (Figure S7). These compounds were characterized by not only their potent inhibitory activity against TGF-βRI but also their selectivity, exhibiting significantly lesser activity against TGF-βRII (with an in vitro kinase activity assay-derived IC₅₀ greater than 15 μM for TGF-βRII). This selectivity aspect is crucial for potentially combining these inhibitors with immuno-oncology agents, notably anti-PD-L1 drugs, to enhance therapeutic efficacy in cancer treatment.^{127,128,129,130,131,132,133}

Figure S7 presents an overview of the general structures of these small molecules, along with specific examples that underscore their selectivity against TGF-βRI compared to TGF-βRII.

GenFleet Therapeutics and Medshine Discovery

In 2017, GenFleet Therapeutics (Shanghai), in partnership with Medshine Discovery, secured a patent for a series of benzotriazole-derived α and β unsaturated amide compounds, identified as TGF-βRI inhibitors. Notably, many of these compounds displayed IC₅₀ values below 500 nM, indicating their potent inhibitory activity against TGF-βRI.¹³⁴ Among the patented compounds, GFH018 (specified as embodiment 1) was particularly noteworthy. GFH018 entered phase 1/2 clinical trials, highlighting its potential as a therapeutic agent.^21,22,23 In 2019, GenFleet Therapeutics expanded on its initial findings regarding GFH018. This advancement included its crystal and salt forms. In vitro kinase assays revealed that GFH018 has an IC₅₀ of 40 nM. This level of potency positions GFH018 as significantly more effective than galunisertib, being approximately 5.2 times more potent based on the TGF-βRI binding activity protocol employed in this patent.¹³⁵ This marked increase in potency underscores the potential of GFH018 as a highly effective TGF-βRI inhibitor.

Figure S8A showcases the general structure of the compounds covered by the 2017 patent claim, focusing on GFH018.

Shanghai Yingli Pharmaceutical

In 2018 and 2019, Shanghai Yingli Pharmaceutical successfully secured patents for a novel series of heterocyclic aromatic derivative compounds specifically designed to inhibit TGF-βRI. Most of these compounds demonstrated potent inhibitory activities, as evidenced by their in vitro kinase IC₅₀ values being predominantly lower than 1 μM.^136,137,138 These values reflect the high efficacy of the compounds in inhibiting the TGF-βRI kinase activity, underscoring their potential therapeutic utility.

Figure S8B displays the primary structures of these compounds as covered by the patent claims of Shanghai Yingli Pharmaceutical.

Among the patented compounds, YL-13027, which is highlighted as example 9 in the patent WO/2018/019106, stands out due to its remarkably low IC₅₀ value of 16 nM. This value indicates an exceptionally high level of potency against TGF-βRI. Currently, YL-13027 is advancing through phase 1/2 clinical trials.^15,16,19 This progression into clinical testing marks a significant milestone in the development of YL-13027 and demonstrates Shanghai Yingli Pharmaceutical’s commitment to bringing forward new therapeutic options in the field of TGF-βRI inhibition. The ongoing clinical trials aim to further evaluate the safety, efficacy, and potential application of YL-13027 as a novel treatment modality.

Clavius Pharmaceuticals

In 2019, Clavius Pharmaceuticals introduced novel substituted imidazoles designed to inhibit the TGF-β signaling pathway. These compounds demonstrated good in vitro TGF-βRI inhibition efficacy, with IC₅₀ values below 1.37 μM.¹³⁹ Building on this success, in 2020, they further refined their substituted imidazoles, achieving even more potency, with IC₅₀ values below 244 nM.¹⁴⁰ The main structures from these patent claims (WO/2019/005241 and WO/2020/041562) and exemplary compounds are depicted in Figure S8C.

ARS Pharmaceuticals (and Silverback Therapeutics)

From 2019 to 2022, Silverback Therapeutics secured patents for several small molecules as TGF-βRI inhibitors.^141,142,143 These compounds demonstrated notable in vitro kinase inhibitory activity, typically with IC₅₀ values below 10 μM, and many as low as 100 nM (WO/2019/195278 and WO/2021/11834). In 2022, ARS Pharmaceuticals extended this work by patenting a distinct set of small molecules characterized by even greater potency, generally with in vitro kinase IC₅₀ values below 25 nM (WO/2022/006340 and WO/2022/076905).¹⁴⁴ The structures of these patented compounds are presented in Figure S9A.

Agomab Spain (and Origo Biopharma)

Origo Biopharma obtained a patent in 2021 for benzylamide derivative compounds with IC₅₀ values below 800 nM (WO/2021/105317).¹⁴⁵ In the following year, Agomab Spain patented 2-(3-pyridin-2-yl-4-quinolin-4-yl-pyrazol-1-yl)-acetamide derivatives, demonstrating even greater potency as TGF-βRI inhibitors, with IC₅₀ values below 250 nM (WO/2022/069509).¹⁴⁶ Figure S9B outlines the principal structures encompassed by these patent claims.

Chiesi Farmaceutici

In 2022, Chiesi Farmaceutici introduced a series of pyridazinyl amino and pyrido oxazine derivatives as exceptionally potent TGF-βRI inhibitors. These compounds consistently displayed in vitro kinase IC₅₀ values primarily below 10 nM and many impressively below 1 nM. Comparative testing affirmed that the unique structural scaffolding of these patented compounds significantly contributes to their enhanced inhibitory activity effect on TGF-βRI (WO/2022/013307, WO/2022/013311, WO/2022/013312, and WO/2022/136221).^{147,148,149,150} The key structures from these patent claims are shown in Figure S10.

Other pharmaceutical companies that contributed to the development of TGF-βRI inhibitors

Several other pharmaceutical companies have contributed to the development of small-molecule TGFβRI inhibitors (Figure S11).

In 2006, Scios claimed carboxamide small molecules with a range of inhibitory activities against TGF-βRI, with many IC₅₀ values below 1 μM (WO/2006/105222).¹⁵¹ AstraZeneca, in 2009, reported novel pyridine compounds with IC₅₀ values below 10 μM (WO/2009/022171).¹⁵² Hanmi Pharmaceutical, in 2018, patented pyrazole derivatives, mostly with IC₅₀ values below 100 nM (WO/2018/004290).¹⁵³ In 2020, Integral BioSciences patented heterocyclic core compounds mostly with IC₅₀ values below 5 μM (WO/2020/012357).¹⁵⁴ Additionally, Nanjing Sanhome Pharmaceutical patented naphthalene derivatives as TGF-βRI inhibitors with IC₅₀ values below 250 nM, exhibiting high selectivity against p38α (WO/2020/078402).¹⁵⁵

More recently, in 2022, Sumitomo Pharma Oncology described eight compounds with IC₅₀ values below 10 nM against TGF-βRI (WO/2022/204721), most with selectivity below 100-fold to ALK2, but with high selectivity to JAK2 (greater than 1,000-fold).¹⁵⁶ Also in 2022, the company described compounds with IC₅₀ values below 100 nM against TGF-βRI (WO/2022/204721 and WO/2022/126133), most with selectivity below 100-fold to ALK2143.¹⁵⁷

In 2023, Sichuan Kelun-Biotech Biopharmaceutical described pyrazole compounds with IC₅₀ values below 50 nM against TGF-βRI (WO/2022/063050-EP4219543), including six compounds with lesser affinity to TGF-βRII (IC₅₀ > 1000 nM) and eight compounds with low affinity to human ether-a-go-go-related gene (IC₅₀ > 10 μM), thus indicating selectivity toward TGF-βRI and a reduced risk of cardiotoxicity.¹⁵⁸

Final considerations

The TGF-β signaling pathway is pivotal in regulating cellular processes such as cell growth, tissue repair, and immune regulation. Dysregulation of this pathway has been implicated in various pathologies, making it an appealing target for therapeutic intervention. Inhibiting this pathway with TGF-βRI inhibitors is emerging as a promising approach for treating multiple diseases, particularly fibrosis and cancer.

Although no TGF-βRI inhibitors have yet received regulatory approval, eight are undergoing clinical trials for a spectrum of oncological and fibrotic conditions, representing significant progress in addressing clinical needs.

The portfolio of TGF-βRI inhibitors in clinical development is expected to expand further, driven by new patents and the availability of novel X-ray or cryoelectron microscopy structures of protein-inhibitor complexes. These structures provide valuable insights into key binding interaction patterns and action mechanisms, paving the way for the development of more potent therapeutics. Our analysis underscores the importance of “classical” interaction patterns of TGF-βRI inhibitors,²⁹ including hydrogen bonds with His283, Lys232, and Asp351 and critical water molecule-mediated interactions (hydrogen bonds) with Tyr249, Glu245, and Asp351. These inhibitors typically occupy a hydrophobic pocket between the αC helix and β1–β4 sheets of the ATP binding site, engaging with critical residues such as Ala230, Leu260, and Leu278. Targeting the Asn338/Leu340/Ala350 pocket with the Ser280 gatekeeper residue is deemed crucial for the development of selective TGF-βRI inhibitors.

The increasing availability of experimental 3D structures and patented compounds targeting TGF-βRI, coupled with structural insights, facilitates advanced computer-aided research techniques, including molecular docking, pharmacophore modeling, molecular dynamics simulations, de novo drug design, and SAR studies.

An interesting aspect of TGF-β signaling is its influence on the immune system, particularly its role in suppressing immune activation. This bears significant implications for immunotherapies and immune checkpoint inhibitor treatments. Current research efforts are directed toward developing multi-target inhibitors that block TGF-βRI and interact with other key immune regulatory molecules, potentially enhancing the efficacy of combination therapies and improving patient outcomes.

The ongoing advancements in TGF-βRI inhibitor research hold promising prospects for novel therapeutic developments in this field.

Acknowledgments

The authors gratefully acknowledge the following funding agencies for their generous support: the Fundação para a Ciência e a Tecnologia - Ministério da Educação, Ciência e Inovação (FCT) under the framework of the projects UID04138 - Instituto de Investigação do Medicamento, and 2023.08262.CEECIND/CP2843/CT0002 (to R.C.A.), PD/BD/145161/2019 (to C.L.-S.), and 2023.01201.BD (to R.B.). H.F.F. is thankful for the support from PTDC/BTM-SAL/4350/2021 (FCT), 'La Caixa' Foundation under the framework of the Healthcare Research call 2019 (LCF/PR/HR19/52160021; NanoPanther), and 2022 (LCF/PR/HR22/52420016; MultiNano@BBM). R.C.G. acknowledges funding from LISBOA-01-0246-FEDER-000017, funded by FEDER through COMPETE2020.

Author contributions

C.L.-S. and R.C.A. wrote and edited the manuscript, supervised by H.F.F. and R.C.G. C.L.-S., R.C.A., and R.B. designed the figures. All authors discussed and contributed to the final manuscript.

Declaration of interests

The authors declare no competing interests.