Discovery of Rogocekib (CTX-712): A Potent and Selective CLK Inhibitor for Cancer Treatment

Abstract

Cdc2-like kinase (CLK) inhibitors represent an innovative class of small molecules designed to modulate RNA splicing patterns, offering a novel avenue for therapeutic intervention in diseases where dysregulated splicing contributes to pathogenesis, particularly in oncology. Here, we describe the discovery of Rogocekib (CTX-712), a promising therapeutic candidate as a CLK inhibitor, which is currently in clinical development. Our medicinal chemistry research involved structure-based drug design-guided scaffold hopping from an initial chemical scaffold and subsequent chemical optimization to generate a novel 1H-imidazo­[4,5-b]­pyridine series. Treatment with CTX-712 reduced the phosphorylation of serine- and arginine-rich proteins in a dose-dependent manner, leading to potent in vitro cell growth suppression and in vivo antitumor activity in lung cancer NCI-H1048 xenograft model. These findings highlight the promise of CTX-712 as a novel CLK inhibitor and its potential as a therapeutic for cancers, particularly those characterized by RNA splicing alterations.

Alternative RNA splicing is a critical and fundamental process in gene expression, where precursor mRNA is transformed into mature mRNA for translation into proteins. Dysregulation of alternative RNA splicing is frequently observed across a wide spectrum of human diseases, including neurodegenerative disorders, autoimmune diseases, viral replication, and cancer. − Notably, cancer often exhibits significant dysregulation of alternative RNA splicing, which contributes to disease progression. Over the past few decades, advancements in genomic DNA and RNA sequencing have unveiled a multitude of mechanisms, through which RNA splicing is pathologically disrupted in cancer. These mechanisms include mutations in the regulatory sequences of tumor suppressor genes, recurrent somatic mutations in genes encoding core spliceosomal proteins and regulatory components, and tumor-specific changes in the expression of certain RNA splicing factors. Among the potential therapeutic targets, Cdc2-like kinases (CLKs) have recently emerged as promising targets due to their pivotal role in regulating alternative RNA splicing. − The CLK family belongs to the CMGC group of 62 kinases, which also includes cyclin-dependent kinases, mitogen-activated protein kinases, glycogen synthase kinases, serine-arginine protein kinases (SRPKs), and dual-specificity tyrosine-regulated kinases (DYRKs). Among these kinases, CLKs share close biological function, protein sequence identity, and similarity with the DYRK family.

The CLK family comprises four homologous members, CLK1, CLK2, CLK3, and CLK4, dual-specificity protein kinases that are evolutionarily conserved in eukaryotes. CLKs phosphorylate serine- and arginine-rich (SR) proteins, also known as SR splicing factors (SRSF1–12), which trigger the relocation of SR proteins from nuclear speckles to spliceosomes, facilitating exon recognition in the spliceosome molecular machinery. The past decade has seen a significant increase in publications describing CLKs as therapeutic targets, along with the discovery of new CLK inhibitors, such as TG-003, CC-671, and SM08502. − Our group has contributed to this growing body of work by introducing a series of small molecules, designated as Cpd-1 to -3, T3, and T-025. − These compounds exhibit diverse chemical structures, offering different kinase selectivities over the SRPK and DYRK families. Importantly, a set of in vitro and in vivo experiments evaluating these compounds highlighted the therapeutic impact of CLK inhibition on targeted cancer treatment. Based on these insights, we focused on delivering a clinical drug candidate using T-025 as the chemical starting point. In this paper, we describe the discovery of a highly potent and selective CLK inhibitor, Rogocekib (CTX-712), which modulates alternative RNA splicing, for cancer treatment.

In a previous study, T-025 (1) was identified as a potent CLK inhibitor through a chemical exploration guided by its potency against CLK2, showing an IC₅₀ value of 1.0 nM (Figure A). This compound has a U-shaped chemical structure comprising a pyrazolopyrimidine-based hinge binder and a unique quinoline moiety. X-ray cocrystal structure analysis revealed several key interactions of T-025 with the CLK2 protein: three intermolecular hydrogen bonds with amino acid residues Glu244, Leu245, and Leu246 in the CLK2 hinge region; an intermolecular hydrogen bond between the quinoline moiety and Lys193; and appropriate spatial occupancy of the pyrimidine moiety within the hydrophobic pocket (PDB code: 5UNP).

1.

Structure of T-025 and SBDD direction in our study. (A) Chemical structure of T-025 and its in vitro potency. (B) X-ray crystal structure of T-025 bound to the ATP-binding pocket of CLK2. Top-view illustration of SBDD-guided new scaffold generation, highlighting a newly designed vector that can access the hydrophobic region originally engaged by the pyrimidine moiety. (C) Side-view illustration corresponding to (B), showing the spatial orientation of the new vector.

In addition to the chemotype series of T-025, which exhibited unfavorable ADME-Tox profiles, such as low metabolic stability, CYP inhibition, and suboptimal pharmacokinetics (PK), we used an alternative chemotype generation strategy by scaffold hopping from T-025, leveraging structure-based drug discovery (SBDD) methodology. As depicted in Figures B and C, our conceptual design aimed to replace the quinoline ring with a versatile [6, 5]-membered heterocyclic ring as a new scaffold. This new scaffold needed to meet two requirements: (1) it should bear an N-atom capable of polar interactions with Lys193, and (2) it should provide a novel trajectory for accessing the hydrophobic pocket. Based on these criteria, we designed two new scaffolds, 3H-imidazo­[4,5-b]­pyridine and 1H-imidazo­[4,5-b]­pyridine, which were selected based on their chemical tractability and lipophilicity. Subsequently, we connected these scaffolds to a simplified pyrrolopyrimidine hinge binder and a methylene-tethered m-fluorophenyl unit (instead of the pyrimidine in T-025). Both modifications were deemed acceptable based on the structure–activity relationship (SAR) of the T-025 series. Compounds 2 and 3 were evaluated for in vitro enzymatic CLK2 inhibition, growth inhibition of HCT-116 human colon cancer cells, experimental logD values, and solubility at pH 6.8 (Table ). Both compounds exhibited CLK2 inhibition with IC₅₀ values of 1.6 nM for compound 2 and 0.4 nM for compound 3. In particular, compound 3, with its 1H-imidazo­[4,5-b]­pyridine scaffold, was considered a promising starting point owing to its comparable or better profile than that of the original compound T-025. Thus, compound 3 was selected as a lead for further chemical optimization.

1. Scaffold Hopping Results.

^aCLK2 enzyme potency was measured in the LANCE Ultra kinase assay system (PerkinElmer) at ATP concentration of 1 mM. The value of 95% confidence interval is listed in the Supporting Information. All experiments to determine IC₅₀ values were performed in at least duplicate at each compound concentration dilution unless otherwise noted.

^bCell growth IC₅₀ was measured by Cell titer-Glo after 3 days of compound treatment.

^cKinetic aqueous solubility run in Japanese Pharmacopeia second (JP2nd) fluid for disintegration and dissolution (PBS at pH6.8).

Next, various hinge binders were investigated to expand the chemical space (Table ). The primary focus was on modifying the NH group in pyrrolopyrimidine 3, which acts as a hydrogen bond donor. This H-bonding can enhance CLK2 potency through polar interactions with the hinge region. However, preliminary data suggested that it might also pose a risk of an unfavorable ADME-Tox profile, as generally reported in pyrrolopyrimidine-based kinase drug discovery. , Thus, we initially designed N atom-switched pyrrolopyrimidines 4–6 with OMe, NH, and Cl groups installed at the 7-position of the fused rings. These substitutions were guided by the SAR of T-025, and the order of potency matched the SAR [OMe > NH > Cl (>H)]. Therefore, OMe was selected as the optimal substituent for this position. Further exploration of hinge binders highlighted the versatility of non-NH-type hinge binders, as exemplified by compounds 7–9. These compounds exhibited very potent enzymatic CLK2 and cell growth inhibition, with compound 9 showing remarkable potencies: enzymatic CLK2 IC₅₀ of 0.24 nM and HCT-116 growth IC₅₀ of 4.2 nM.

2. SAR Exploration of the Hinge Binder .

^aSee notes in Table . NT = not tested.

As the final step in lead optimization, we modified the m-F-phenyl ring. At this point, compound 9 was a slightly hydrophobic molecule with a logD value of 2.78, demonstrating mouse liver microsomal (MLM) clearance rate: 15.3%/min/mg protein, as shown in Table . Therefore, we designed various heteroaromatic rings to reduce the lipophilicity of these compounds. We synthesized compounds with pyrimidines (10), pyrazoles (11 and 12), oxazoles (13 and 14), and oxadiazole (15), all of which exhibited moderate potency and a lowered logD value, as expected. Notably, the solubility of these compounds was greatly improved, compared to that of the parent compound 9. Furthermore, mouse metabolic stability was enhanced, showing MLM clearance rates: 1.9–3.3%/min/mg protein. However, these compounds showed CL > 1000 mL/h/kg, and consequently systemic exposure AUC 306.9 ng*h/mL at most when assessed in mouse cassette PK using ICR mice.

3. SAR Exploration of the Aromatic Ring .

^aSee notes in Table .

^bMouse liver microsomal metabolism (MLM) was measured.

^cCassette dosing data via po (1 mg/kg) and iv (0.1 mg/kg) administration. Institute of Cancer Research (ICR) mice (male, n = 3) were used.

^d0.1 mg/kg, iv.

^e1 mg/kg, po.

^fBioavailability. NT = not tested.

Through parallel and extensive chemical modifications of each heteroaromatic ring, we identified substituted oxadiazoles as promising candidates to enhance in vitro mouse metabolic stability and improve mouse PK profiles, as highlighted in Table . Compound 16, an ethyl derivative of parent compound 15, exhibited decreased potency with a CLK2 IC₅₀ of 10 nM. However, further derivatization involving the incorporation of fluorine atoms successfully restored the potency and improved metabolic stability. Notably, the difluoro-methyl derivative 17 demonstrated a remarkable 10-fold increase in potency, with a CLK2 IC₅₀ of 0.56 nM, along with a reduced MLM clearance rate (0.3%/min/mg protein) and in vivo clearance value (CL 995 mL/h/kg) compared to its parent 15. Subsequently, we synthesized compounds 18 and rac-19 as fluorinated derivatives of the parent ethyl compound 16. This endeavor led to the discovery of compound 18, which was characterized by balanced potency, physicochemical properties, mouse metabolic stability, and mouse PK profiles. Notably, the monofluoroethyl derivative rac-19 also exhibited excellent in vitro profiles comparable to those of compound 18 and demonstrated even better solubility. We further conducted chiral separation of rac-19, yielding two enantiomers (R)-19 and (S)-19, with their stereochemistry confirmed through the single X-ray crystal structure analysis of (R)-19 (for details, see the Supporting Information). Interestingly, these two enantiomers showed no significant differences in terms of potency, ADME-Tox, and in vivo mouse studies. Although both compounds were eligible for inclusion, (R)-19 demonstrated slight superiority in HCT116 cell growth inhibition in vitro and better systemic exposure in mice. Consequently, (R)-19 (CTX-712) was selected as a candidate for further development.

4. Optimization of Oxadiazole Moiety .

^aSee notes in Table .

The detailed profiles of CTX-712 are presented in Table . The kinase inhibition profiles of CTX-712 within the CLK and DYRK families were determined by measuring the enzymatic inhibition of each protein. The data revealed that CTX-712 was a highly potent CLK inhibitor, exhibiting the greatest potency against CLK2 and CLK1 (the IC₅₀ values for CLK1, 2, 3, and 4 were 0.69, 0.46, 3.4, and 8.1 nM, respectively), along with the inhibition of the DYRK family (the IC₅₀ values for DYRK1A and 1B were 1.1 and 1.3 nM, respectively). When tested against a panel of 468 kinases at a single concentration of 100 nM by Eurofins Discovery, only 10 kinases, including the primary targets CLK1–4 and DYRK1A and 1B, showed >65% inhibition (see the Supporting Information). Furthermore, CTX-712 was not an inhibitor of CYP3A4 or hERG potassium channels. CTX-712 demonstrated high liver microsomal stability in mice, rats, and humans, indicating favorable PK profiles in animals, as demonstrated via mouse cassette PK studies.

5. Detailed Profiles of CTX-712.

Cmpd	CTX-712
Molecular weight	408.40
Enzymatic inhibition of kinase families
CLK1/CLK2/CLK3/CLK4	0.69/0.46/3.4/8.1
DYRK1A/DYRK1B	1.1/1.3
KinomeScan: Number of kinases with >65% inhibition at 100 nM	10/468
Metabolic clearance rate: mouse/rat/human (%/min/mg protein)	0.1/1.7/–0.6
CYP3A4 inhibition at 10 uM	2.9%
hERG manual patch clamp %inh. at 10 uM with/without FBS	0/21.9
Plasma binding fu: mouse/rat/human	0.23/0.13/0.16

To examine the cellular potency and target engagement of CTX-712, we analyzed the phosphorylation levels of substrate SR proteins in NCI-H1048 cells, a small cell lung cancer cell line. We used an antipan-phospho-SR antibody, which recognizes multiple classical members of the SR protein family (Figure A). Our findings indicated that CTX-712 significantly reduced the phosphorylation levels of SR proteins in a concentration-dependent manner. Specifically, the IC₅₀ values for p-SRSF4 (SRp75) and p-SRSF6 (SRp55), which were the major substrates in this Western blotting analysis, were 45.5 and 116.1 nM, respectively. The inhibition of these phosphorylations was accompanied by a mobility shift, consistent with previous observations. Importantly, these data correlated with the GI₅₀ value of 75 nM in NCI-H1048 cells (Figure B). Subsequently, CTX-712 was evaluated in the NCI-H1048 xenograft model (Figure C). In this efficacy study, the compound was administrated orally at doses of 25, 37.5, or 50 mg/kg, twice-weekly, twice daily (BIW, BID) for 14 days. Notably, CTX-712 demonstrated significant, dose-dependent antitumor effects, with the greatest efficacy observed at 50 mg/kg, achieving a T/C value of 6.2% on day 14. Furthermore, no severe body weight loss was observed throughout the study (see the Supporting Information). These results suggest that the CLK inhibitor, CTX-712, holds promise as an anticancer drug, particularly in patients with splicing factor mutations. Due to its promising pharmacological and safely profile, CTX-712 was selected as the development candidate and entered phase I clinical trials in Japan (jRCT2080224127) and in the US (NCT05732103).

2.

In vitro and in vivo data of CTX-712. (A) In vitro cellular biomarker potency of CTX-712 on NCI-H1048 cell line. (B) In vitro growth inhibition of CTX-712 on the same cell line. (C) In vivo efficacy study of CTX-712 in NCI-H1048 xenograft model. Tumor-bearing mice were dosed with vehicle or CTX-712 at 25, 37.5, or 50 mg/kg orally (PO) BID BIW for 14 days (Day 0, 3, 7, and 10). Data are presented as means ± SEM (n = 5).

In summary, we discovered a potent and bioavailable CLK inhibitor CTX-712, which has been assigned the international nonproprietary name Rogocekib, through novel chemotype generation guided by SBDD, followed by lead optimization. CTX-712 effectively inhibited CLKs and the phosphorylation of SR proteins, leading to significant in vitro and in vivo antiproliferative effects, as described in this manuscript. This finding underscores the potential of CTX-712 as a targeted therapy for malignancies with RNA splicing alterations, such as splicing factor mutations.

Glossary

Abbreviations

CLKs

cdc2-like kinases

DYRKs

dual-specificity tyrosine-regulated kinases

PK

pharmacokinetics

SAR

structure–activity relationship

SBDD

structure-based drug design

SRPKs

serine-arginine protein kinases.

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acsmedchemlett.5c00412.

• Detailed descriptions of general chemistry methods, compound synthesis and characterization, single-crystal X-ray crystallography protocols, biochemical and cellular assays, in vivo studies, as well as supporting data, figures, and tables (PDF)

The manuscript was written through contributions of all authors.

No unexpected or unusually high safety hazards were encountered.

The authors declare the following competing financial interest(s): Midori Sugiyama, Akio Mizutani, Yoshihiko Satoh, Yasuyoshi Arikawa, Daisuke Morishita, and Kenichiro Shimokawa are current employees and stockholders of Chordia Therapeutics Inc.