Discovery of a series of 1H-pyrrolo[2,3-b]pyridine compounds as potent TNIK inhibitors
A B S T R A C T
In an in-house screening, 1H-pyrrolo[2,3-b]pyridine scaffold was found to have high inhibition on TNIK. Several series of compounds were designed and synthesized, among which some compounds had potent TNIK inhibition with IC50 values lower than 1 nM. Some compounds showed concentration-dependent characteristics of IL-2 inhibition. These results provided new applications of TNIK inhibitors and new prospects of TNIK as a drug target.
Protein kinases are enzymes that catalyze the phosphate transfer from ATP molecule to the specific residues of its substrate proteins. According to the phosphate-acceptor residues, kinases can be roughly classified as tyrosine kinases and serine or threonine kinases. With high throughput sequencing, at least 518 kinases are identified in human genome, and phylogenetically, these kinases can be distributed into five broad groups and 44 families. In the past 30 years, kinases have been intensively investigated as drug targets, with more than 50 kinase in- hibitors approved to date. Although this manifests a significant advance in kinase drug development field, these drugs are predominantly re- ceptor tyrosine kinase inhibitors for cancer treatment. There are large uncultivated kinases to study for understanding their diverse biological functions and possible utilities in modulation of human diseases.Traf2 and Nck-interacting kinase (TNIK) was identified as the novel germinal center kinase (GCK) family, a subgroup of the STE20 kinase family. TNIK mRNA is expressed predominantly in heart and skeletal muscle and less abundantly in brain, kidney, and pancreas. Taira et al. adopted an exciting approach to identify TNIK as a protein interacting with a small GTP-binding protein Rap2.1. Later, Shkoda et al. showed that TNIK was involved in the NF-κB and c-Jun N-terminal kinase pathway, and it could specifically activate the c-Jun N-terminal kinasepathway similar to many other GCKs.2. Mahmoudi et al. utilized the proteomics approach to uncover the essential activator role of TNIK in Wnt pathway.3.
They found that TNIK acted as a novel protein inter- acting with Tcf4 in the mouse intestinal crypt. TNIK is localized in the nuclei and is recruited to promoters of Wnt target genes in mouse crypts and in colorectal cancer cells in a β-catenin-dependent manner. Their study showed that TNIK interacted directly with both β-catenin and TCF4, phosphorylating TCF4 and leading to TCF/LEF-driven transcrip- tional activation of Wnt target genes. Recently Jaeger-Ruckstuhl et al.found that priming of CD8+ T cells induced a TNIK-dependent nucleartranslocation of β-catenin with consecutive Wnt pathway activation.4, showing that TNIK was an important regulator of effector and memory T cell differentiation and induced a population of stem cell-like memory T cells.Due to its diverse biological functions in the different cellular con- texts, several studies focused on developing TNIK inhibitors for probing its potential as a drug target. Masuda et al. reported their efforts in discovering the TNIK inhibitor.5. They screened an in-house kinase- focused library and identified a series of quinazoline compounds with high TNIK inhibitory activity. Further optimization led to the potent TNIK inhibitor NCB-0846 (1, Fig. 1), with a half-maximal inhibitory*concentration (IC50) value of 21 nM. Read et al. in 2019 reported another series of TNIK inhibitors harboring a naphthyridine scaffold, giving IC50 of the most active compound about 10 nM (2, Fig. 1).6. KY- 05009 (3, Fig. 1) was first disclosed in a patent US2010216795.
The inhibitory activity of KY-05009 against TNIK was confirmed by an ATPcompetition assay (Ki = 100 nM), and it significantly inhibited the TGF-β-activated EMT through the attenuation of Smad and non-Smad signaling pathways, including the Wnt, NF-kB, FAK-Src-paxillin- related focal adhesion, and MAP kinases (ERK and JNK) signaling pathways in A549 cells.7. Ho et al. reported a series of compounds based on a 4-phenyl-2-phenylaminopyridine scaffold that were potent and selective inhibitors of TNIK (4).8. These compounds were used as tools to evaluate the roles of TNIK kinase activity in signaling and proliferation in Wnt-activated colorectal cancer cells. To further investigate the bio- logical function of TNIK inhibitors, we launched a campaign to identify the TNIK inhibitors with different scaffolds, which might provide more options for both biological study and drug development.To identify the starting molecules for development, an in-house protein kinase inhibitor library containing 167 diversified molecules was subjected to TNIK inhibition assay. Compound 5 with 1H-pyrrolo [2,3-b]pyridine scaffold exhibited high inhibition on TNIK with an IC50 value of 9.2 nM. This compound was composed of a 1H-pyrrolo[2,3-b] pyridine scaffold, which was supposed to interact with the hinge region and form hydrogen bonds with Glu106 and Phe107. The docking study suggested that two aromatic rings, Ar1 and Ar2, directed to the solvent region and back-pocket, respectively (Fig. 2). Although compound 5 had good activity on TNIK inhibition, its simple structure manifests a high probability of improving bioactivity.
Besides, high hydrophobicity isexpected according to the structure, which may affect membrane permeability and pharmacokinetic properties. Since Ar1 directed to the solvent region, we introduced methylpiperazine moiety as a solubility- enhancing group into the structure (6, Table 1).According to the docking study, Ar2 is pointed to the back-pocket, which is essential to the improvement of binding activity. Then, the initial modifications were focused on the Ar2 part (Fig. 2B). Commonly used substitutions were introduced in the phenyl ring to explore the cavity size and potential hydrogen bonding interaction. Also, different bicyclic rings were adopted to replace the phenyl one (Table 1). Halogen substitutions at p-site had little effect on activities (7, 8). The intro- duction of a cyano group at m-site resulted in a significant increase in activity (9), then we performed the docking study on compound 9. As shown in Fig. 3A, it superimposed well with the original ligand in the template protein structure (PDB entry code: 5AX9), with the cyano group pointing to the structured water molecule that further formed hydrogen bonding interactions with residues Glu69 and Phe172. When the cyano group was replaced by amino, amide, or sulfonamide struc- tures (10–14), which might interact with water, the activities reduced to the same level of compound 9. The introduction of a meth- anesulfonamide group at the other m-site further increased TNIK inhi- bition (16–18).
We also docked compound 16 into the ATP site of TNIK, and the modeling indicated that compound 16 bound to TNIK essentially as compound 9 (Fig. 3B). The different part, methanesulfonamide group, directed to the A loop region, and made a hydrogen bond be- tween one of the oxygen atoms in sulfamide and the backbone nitrogen atom of residue Asn33. The bicyclic ring structure was well-tolerated(19). However, when the bicyclic ring connected to the core structureat o-site of the phenyl ring (20) the activities showed 6 times lower than connecting at m-site (19), which indicated the cavity of back-pocket might not be wide enough. When extra N atoms were introduced into indole such as indazole (21) and benzo[d]imidazole (22), different re- sults were obtained that the inhibition activity of indazole ring was about 10 times higher than benzo[d]imidazole. When the indazole ring was replaced by pyrazolo[3,4-b]pyridine (23), the activity reduced about 20 times. The activity enhancement of cyano group could also be observed when introduced in the indazole (24).
In this round of opti- mization compound 16 exhibited excellent TNIK inhibition in these compounds and was chosen for further investigation. In Table 1, we used the methylpiperazine group as the solubility- enhancing motif. Since m-cyano and m-methanesulfonamide have excellent inhibition effects on TNIK, intensive studies on different types and substitution sites of hydrophilic groups were carried on at Ar1 part (Fig. 2B). In compound 16 the hydrophilic group methylpiperazine connected to the core structure by a methylene group, thus we designed some different connection modes. First, we adopted amides as the linker (Table 2). Connection at m or p-site had little effect on activities (25 and 26). Different sizes of the hydrophilic groups (27–31, 35), together with carboxylic acids (32, 33), also showed no apparent differences. More- over, as the pyrazol ring was commonly used in medicinal chemistry, it was also adopted in our study (34, 36). However, the designed com- pounds showed slightly decreased activities with IC50 values over 0.5 nM. Although these compounds had different hydrophilic groups, most of them showed high inhibition activities with IC50 values below 0.5 nM. These results indicated that the Ar1 part had little effect on molecular activities. Jurkat cells were treated with tested compounds at a final concen- tration of 1 and 10 μM for 24 h. The levels of IL-2 in cell supernatants were detected by ELISA kit and analyzed from three independent ex- periments. Data were expressed as mean + SD. The annotation “+” meant the treatments with the indicated compounds and stimulating factors. NCB-0846 is a reported inhibitor of TNIK, while NFAT Transcription Factor (TF) Regulator-1 is an IL-2 synthesis inhibitor.
TNIK is a vital regulator of effector and memory T cell differentiation and T cells produce IL-2 after in vitro stimulation [4], we chose several compounds with good enzymatic activities and different types of Ar1 or Ar2 structures to examine their effects on inhibiting secretion of IL-2. NCB-0846 is a known TNIK inhibitor[5], while NFAT Transcription Factor Regulator-1 is an IL-2 synthesis inhibitor.9 (Fig. 4). Both two positive inhibitors were used as controls to determine the inhibitory activity of tested compounds against IL-2. The result indicated that PMA and anti-CD3 antibody could induce the expression of IL-2, increasing the level by 11.7 times. Under the condition of stimulant inducement, all tested compounds showed concentration-dependent characteristics on IL-2 inhibition with only one exception of compound 24 at 1 μM. Among them, four compounds, including the positive TNIK inhibitor, showed more potent inhibition (36, 16, NCB-0846, and 17, in a low-to-high order) compared with an IL-2 synthesis inhibitor. IL-2 level was inhibited to below 25 pg/ml by four out of seven tested compounds at 10 μM (16, 17, 24 and 36). It was worth noting that compounds 16 and 17 with relatively better binding activities (lower IC50) were more ad- vantageous in IL-2 inhibition, which suggested the potential value of TNIK inhibitors in inhibiting IL-2 synthesis.
In conclusion, from an in-house screening and following optimiza- tion, a series of 1H-pyrrolo[2,3-b]pyridine compounds were identified as novel TNIK inhibitors with potent enzymatic activities. Some of them showed INS018-055 IL-2 inhibition equivalent to the positive compound NCB-0846. These results provided new applications of TNIK inhibitors and new prospects of TNIK as a drug target.