Characterization of Novel Multikinase Inhibitors with Potent HGK, Aurora B, and Src-family Kinase Inhibition: Implications for ALS Therapy and Cellular Toxicity

Abstract

Phenotypic screening using motor neurons derived from the induced pluripotent stem (iPS) cell of a patient with amyotrophic lateral sclerosis led to the identification of Compound X, a novel agent with potent neuroprotective activity that is hypothesized to act by inhibiting hematopoietic progenitor kinase/germinal center kinase-like kinase (HGK). Unlike known HGK inhibitors, including Prostetin and GNE-495, Compound X and related Compounds Y and Z exhibited broader multi-kinase-inhibition profiles, including the strong inhibition of Aurora B kinase and strong-to-moderate inhibition of Src-family kinases in kinase panel screening. Owing to their roles in cytokinesis, we performed investigations into their effects on cellular proliferation and morphology.

Main Body

Our research group examined the potential of small-molecule compounds for the treatment of amyotrophic lateral sclerosis (ALS) through the phenotypic screening of motor neurons differentiated from the patient-derived induced pluripotent stem (iPS) cell ¹ . Recent studies have implicated hematopoietic progenitor kinase/germinal center kinase-like kinase (HGK), also known as mitogen-activated protein kinase kinase kinase 4 (MAP4K4), in the pathogenesis of ALS, via endoplasmic reticulum stress-induced apoptosis ² . Since the prototype of Compound X exerted strong neuroprotective effects through iPS-derived motor neuron phenotypic screening, and the underlying mechanism is hypothesized to involve the inhibition of HGK ¹ , we hypothesized that Compound X also exerts neuroprotective effects by suppressing HGK ³ .

HGK belongs to a highly conserved family of serine/threonine kinases, making its selective inhibition challenging. The ATP-binding sites of these kinases, such as HGK, Aurora kinases, and Rho-associated kinases (ROCKs), as well as those of tyrosine kinases, including those of the Src family, are structurally similar owing to their high conservation, which complicates the development of selective inhibitors ^{1, 4–7} . For insights into protein specificity and assay validation, resources like the Human Protein Atlas are invaluable. Docking simulations have suggested that the Compound X prototype binds the ATP-binding site via two possible hydrogen bonds with its urea moiety. Compound X and its prototype share an N-(5-phenylthiophen-3-yl)urea moiety, suggesting that these compounds bind the ATP-binding site in a similar manner ¹ .

Several HGK inhibitors with improved selectivity, including Prostetin and GNE-495, have been developed with improved selectivity ⁸ . However, kinome-wide profiling has demonstrated that both compounds exhibit activity against related HGK family kinases, such as Misshapen/NIKs-related kinase (MINK/MAP4K6; 86.8% identity) and Traf2/Nck-interacting kinase (TNIK/MAP4K7; 81.96% identity) ^8–10 . For example, GNE-495 was designed to reduce central nervous system (CNS) penetration, thereby eliminating CNS-related tolerability issues. However, it still strongly inhibits MINK and TNIK, which are known to affect neuronal integrity ⁹ . The reported toxicities of other HGK inhibitors, such as weight loss, increased body temperature, and elevated heart rate, may be attributed to on-target HGK inhibition, off-target effects, or the inherent chemical structures of the compounds ^{11, 12} . Therefore, the toxicity of HGK inhibitors remains unclear, and this may be attributed to the difficulty associated with achieving selectivity for HGK targeting ⁸ . In this study, we examined the toxicological profile of a promising lead compound with high HGK-inhibitory activity. An in vitro micronucleus test for genotoxicity with CHL/IU cells and a 5-day repeated-dose oral toxicity study on mice, performed to assess general toxicity, revealed interesting toxicological profiles.

Compounds X, Y, and Z were synthesized by Toray Industries, Inc. Prostetin was also synthesized by Toray Industries, Inc., according to the method described by Boss and colleagues ¹⁰ . GNE-495 was purchased from MedChemExpress LCC (Monmouth Jct, NJ, USA) ⁹ . Methyl cellulose was purchased from Junsei Chemical Co., Ltd. (Tokyo, Japan).

Kinase inhibition was assessed using Off-chip Mobility Shift Assays at Carna Biosciences Inc. (Kobe, Japan). The assay protocol for each kinase has been published on the Carna Biosciences website. Kinase profiling for Compound X was performed using a 50-kinase panel that included TNIK (MAP4K7) and MINK (MAP4K6). The kinases listed in Table 1 represent a subset of this panel, and the results are presented in Supplementary Table 1. For Compounds Y and Z, inhibitory activities were evaluated for the eight kinases shown in Table 1 (HGK, Src, Lck, Fyn, Yes, Aurora-A (AurA), AurB, and ROCK1).

Table 1. Kinase Inhibition Rates of Compounds X, Y, and Z

Using a concentration that was approximately 100-fold the IC ₅₀ value of the HGK-inhibitory activity as the evaluation concentration, a panel assessment was conducted on kinases reported to be related to the induction of multinucleation (excluding Lyn). Src, Lck, Fyn, Lyn, and Yes belong to the Src kinase family. Compounds X, Y, and Z exhibited high inhibitory activity (>80%) against Aurora A (AurA) and AurB and low inhibitory activity against ROCK1. Src-family kinases were strongly inhibited by Compounds X and Y, but not Compound Z. According to the literature, Compound X more strongly inhibits kinases related to the induction of multinucleation than other HGK inhibitors, such as Prostetin and GNE-495. NT: not tested. The IC ₅₀ values of HGK were 24, 15, and 1.9 nmol/L for Compounds X, Y, and Z, respectively.

CHL/IU cells were obtained from the Japan Health Sciences Foundation (Tokyo, Japan). Cells were cultured in Eagle’s MEM supplemented with 10% heat-inactivated bovine serum at 37°C with 5% CO ₂ in air. Cells were seeded in a black 96-well plate (3603, Corning, NY, USA) at a density of 3.0 × 10 ³ cells/well and pre-cultured for approximately 24 h. The culture supernatant was removed, and 100 μL/well of the test substances was added. After 6 h, the presence of precipitation was confirmed, and cells were cultured with fresh medium for an additional 18 h. Cells were incubated with CellMask™ Deep Red Plasma Membrane Stain (Thermo Fisher Scientific Inc., Waltham, MA, USA) for approximately 10 min and then incubated with 4% paraformaldehyde containing Hoechst 33342 (Dojindo Laboratories, Kumamoto, Japan) for 30 min. Fluorescent images were captured using a CQ1 confocal image cytometer (Yokogawa Electric Corporation, Tokyo, Japan) at a magnification of ×20. Four hundred cells (200 cells/well) were observed, and the number of multinucleated cells (cells with three or more nuclei) was counted visually.

Male Crl:CD1 (ICR) mice (6 weeks old) were purchased from Charles River Laboratories, Japan Inc. (now known as The Jackson Laboratory, Yokohama, Japan), acclimated for 6 days, individually housed in plastic cages with softwood chip bedding under controlled conditions (12 h light/dark cycle at 21–25°C with 40–70% humidity), and fed a standard diet (CRF-1, Oriental Yeast Co., Ltd.). Research on motor neuron diseases and neuroprotection, such as that facilitated by the Society for Neuroscience , continues to advance our understanding of these complex conditions.