New Diarylamine K_V10.1 Inhibitors and Their Anticancer Potential

Abstract

Expression of the voltage-gated potassium channel K_V10.1 (Eag1) has been detected in over 70% of human cancers, making the channel a promising new target for new anticancer drug discovery. A new structural class of K_V10.1 inhibitors was prepared by structural optimisation and exploration of the structure–activity relationship of the previously published hit compound ZVS-08 (1) and its optimised analogue 2. The potency and selectivity of the new inhibitors between K_V10.1 and hERG were investigated using whole-cell patch-clamp experiments. We obtained two new optimised K_V10.1 inhibitors, 17a and 18b, with improved nanomolar IC₅₀ values of 568 nM and 214 nM, respectively. Compound 17a exhibited better ratio between IC₅₀ values for hEAG1 and hERG than previously published diarylamine inhibitors. Compounds 17a and 18b moderately inhibited the growth of the K_V10.1-expressing cell line MCF-7 in two independent assays. In addition, 17a and 18b also inhibited the growth of hERG-expressing Panc-1 cells with higher potency compared with MCF-7 cells. The main obstacle for newly developed diarylamine K_V10.1 inhibitors remains the selectivity toward the hERG channel, which needs to be addressed with targeted drug design strategies in the future.

1. Introduction

The voltage-gated potassium channel K_V10.1 (Eag1) is one of the eight members of the ether-à-go-go family, which is itself subdivided into three subfamilies (Eag, Erg, and Elk) [1,2]. Analogous to other K_V channels, Eag1 is formed by four subunits, with each monomer consisting of six transmembrane segments (S1–S6). The voltage-sensor domain (VSD), which consists of segments S1–S4, detects changes in membrane potential and triggers conformational changes, leading to pore opening. Segments S5 and S6 (pore domain, PD) of each monomer join together to form the ion-conducting pore, which includes the water-filled cavity and the selectivity filter (SF). The second part represents the narrowest part of the pore, which is lined with oxygen atoms that coordinate the movement of potassium ions during permeation [3].

The channel K_V10.1 has emerged as a very attractive oncological target in the last two decades. The channel is mainly expressed in healthy tissues of the central nervous system, while it is barely detectable in peripheral tissues, with the exception of testis and adrenal gland [4,5]. Aberrant K_V10.1 expression has been detected in over 70% of human cancers, regardless of tumour type, which is why the channel is considered a nearly universal tumour marker [6]. K_V10.1 activity is involved in important processes for cancer cell growth and spread, such as cell cycle control, proliferation, migration, angiogenesis, and resistance to hypoxia [7]. The mechanisms of the channel’s involvement in cancer proliferation appear to be related to alterations in membrane potential control and non-canonical effects of overexpressed K_V10.1, affecting Ca²⁺ signalling and microtubule dynamics [8]. Previous in vitro and in vivo studies have shown that specific inhibition of K_V10.1 can successfully reduce tumour progression [9,10,11,12,13]. On this basis, the channel is a promising target for new anticancer drug discovery. Novel selective K_V10.1 inhibitors could, therefore, prolong the survival of patients with fibrosarcoma [14], ovarian cancer [15], glioblastoma [16], acute lymphoid leukaemia [17], gastric [18], head and neck [19], and colon cancer [20], in which the expression of the channel is significantly higher than in normal tissues. Moreover, in a retrospective study, K_V10.1 inhibitors have already shown an effective improvement in survival in a group of patients with brain metastases compared to an untreated control group [16]. To date, no specific and selective small-molecule K_V10.1 inhibitors have been reported because they all exhibited significant inhibition of the cardiac hERG channel, which raised major concerns about adverse cardiac side effects [21]. Recent studies have pointed to a particular structural difference between K_V10.1 and hERG channels that has profound functional implications and suggests the possibility of selective inhibition of K_V10.1 in cancer therapies [22].

To date, only a very small number of compounds has been described to act on K_V10.1 [21]. Most small molecules were identified by studying K_V10.1 inhibition of the previously described antitumour agents and hERG inhibitors, as these channels share structural similarities. K_V10.1 inhibitors can interact with the channel at multiple binding sites, including VSD and PD. Since K_V10.1 and hERG share a high structural similarity in the central cavities, designing a selective small molecule that blocks the pore is a major challenge [23]. In contrast, a more promising approach is to target the VSD domain. Small molecules, such as mibefradil, purpurealidin I analogues, amiodarone, and 20(S)-ginsenoside Rg3 (Figure 1), bind to VSD and act as gating modifiers [24,25]. The exact binding sites of these inhibitors are not yet known, but new cryo-electron microscopy structures of K_V10.1 and hERG may contribute to the development of new specific inhibitors that bind to VSD [3,23].

Our research group recently published a novel structural class of K_V10.1 inhibitors discovered by a unique ligand-based drug design strategy [26]. Compound ZVS-08 (1, Figure 2) with a diarylamine structure was identified as a hit compound in a computational 3D pharmacophore model. In biological assays, it inhibited the K_V10.1 channel with the effect of the gating modifier (IC₅₀ = 3.70 µM). The even more potent K_V10.1 inhibitor 2 (IC₅₀ = 740 nM, Figure 2) was obtained by structure–activity relationship studies. Compound 2 was appropriately selective for some K_V and Na_V channels, but also showed significant hERG inhibition (IC₅₀ = 207 nM). Growth of the breast cancer cell line MCF-7, which has high K_V10.1 levels and low hERG levels, was more strongly inhibited by compound 2 than the growth of Panc-1 cells, which have low K_V10.1 levels and high hERG levels. Compound 2 also induced significant apoptosis in tumour spheroids of Colo357 cells. Here, we present a new series of potent nanomolar K_V10.1 inhibitors obtained by structural optimisation of previously published compounds 1 and 2 We also demonstrated the potential of this new structural class of Kv10.1 inhibitors for antiproliferative activity in MCF-7 cancer cell models.

2. Materials and Methods

2.1. Synthetic Procedures and Analytical Data

Reagents and solvents for the syntheses were from Fluorochem Ltd. (Derbyshire, UK), Apollo Scientific Ltd. (Stockport, UK), Sigma-Aldrich (St. Louis, MO, USA), Enamine Ltd. (Kiev, Ukraine), and TCI (Tokyo, Japan), and were not further purified. Analytical thin-layer chromatography was performed on silica gel aluminium plates (0.20 mm; 60 F254; Merck, Darmstadt, Germany). Silica gel 60 (particle size 230–400 mesh) was used for column chromatography. Reversed-phase column chromatography (RP-CC) was performed on the Isolera Biotage One Flash Chromatography system (SNAP Biotage KP-C18-HS column, 12 g, Biotage, Uppsala, Sweden) using a gradient of 0.1% TFA in deionised water and MeCN as eluent (gradient 10–100% MeCN in 15 column volumes (300 mL); 100% MeCN for 5 column volumes (100 mL)). ¹H and ¹³C NMR spectra were recorded using a 400 MHz NMR spectrometer (Bruker Advance 3, Bruker, Billerica, MA, USA). The splitting patterns were designated as follows: s, singlet; d, doublet; dd, double doublet; td, triple doublet; t, triplet; dt, double triplet; ddd, double of doublet of doublet; ddt, doublet of doublet of triplets; q, quartet; and m, multiplet. The purity of the prepared compounds was verified by liquid chromatography-mass spectrometry according to method A (see below) on a Thermo Scientific Dionex UltiMate 3000 modular system (Thermo Fisher Scientific Inc., Waltham, MA, USA). HRMS measurements were performed on an LC–MS/MS system (Exactive Plus Orbitrap mass spectrometer; Thermo Scientific Inc., Waltham, MA, USA). The high-resolution mass spectra (HRMS) used optical rotation detection at λ = 589 nm (Polarimeter model 241; Perkin Elmer, Boston, MA, USA).

Analytical reversed-phase UPLC analyses were performed using a modular system (Thermo Scientific Dionex UltiMate 3000 modular system; Thermo Fisher Scientific Inc., MA, USA). Method: Waters Acquity UPLC^® HSS C18 SB column (2.1 × 50 mm, 1.8 μm), T = 40 °C; injection volume = 5 µL; flow rate = 0.4 mL/min; detector λ = 254 nm; mobile phase A (0.1% TFA [v/v] in water), mobile phase B (MeCN). Gradient: 0–2 min, 10% B; 2–10 min, 10–90% B; 10–12 min, 90% B. The purity of the tested compounds was determined to be ≥95% by UPLC at 254 nm.

Detailed chemical synthesis procedures and chemical analysis results of all intermediates and final compounds 3a–b, 4a–g, 5a–f, 6a–d, 7, 8a–b, 9a–b, 10a–b, 11, 12a–d, 13, 14, 15, 16a–d, 17a–c, 18a–c, 19, 20, and 21 are described in Appendix A and Supplementary Materials.

2.2. Electrophysiological Recordings on HEK-293 Cells

For whole-cell patch-clamp measurements, stably transfected HEK-293 cells expressing K_V10.1 [26,27,28] were grown in DMEM with 10% FCS and 300 µg/mL Zeocin on glass coverslips coated with fibronectin. All experiments were conducted at room temperature under perfusion by gravity with extracellular solution (160 mM NaCl, 2.5 mM KCl, 2 mM CaCl₂, 1 mM MgCl₂, 10 mM Hepes pH 7.4 with NaOH, 8 mM glucose). Pipettes were made from #1 glass capillaries (WPI) and had 1–3 MΩ resistance. The intracellular solution contained 100 mM KCl, 45 mM N-methyl-D-glucamine, 10 mM 1,2-bis(o-aminophenoxy)ethane-N,N,N′,N′-tetraacetic acid (BAPTA), 10 mM Hepes pH 7.35. Stimulation was applied and data were acquired with an EPC10 Plus amplifier (HEKA Elektronik) controlled with Patch Master software. Compounds were applied with an 8-channel fast solution exchanger (ALA). Compounds were prepared to a stock concentration of 10 mM in DMSO. The control solution contained 0.5% DMSO. Stimuli consisted of 500 ms depolarizations to +40 mV.

2.3. Electrophysiological Recordings on Xenopus laevis Oocytes

Stage V–VI oocytes were isolated by partial ovariectomy from Xenopus laevis frogs (African clawed frog) [26]. Mature female frogs were purchased from CRB Xénopes (Rennes, France) and were housed in the Aquatic Facility (KU Leuven) in compliance with the regulations of the European Union (EU) concerning the welfare of laboratory animals as declared in Directive 2010/63/EU. The use of Xenopus laevis was approved by the Animal Ethics Committee of the KU Leuven (Project nr. P186/2019). After anesthetising the frogs by a 15 min submersion in 0.1% tricaine methanesulphonate (pH 7.0), the oocytes were collected. The isolated oocytes were then washed with a 1.5 mg/mL collagenase solution for 2 h to remove the follicle layer.

Ion channels (K_V10.1 and hERG) were expressed in Xenopus laevis oocytes, by linearisation of the plasmids and subsequent in vitro transcription using a commercial T7 or SP6 mMESSAGE mMACHINE transcription kit (Ambion, Carlsbad, CA, USA). Defolliculated Xenopus oocytes were then injected with 20–50 nL of the cRNA at a concentration of 1 ng/nL using a micro-injector (Drummond Scientific1, Broomall, PA, USA). The oocytes were incubated in a solution containing (in mM): NaCl, 96; KCl, 2; CaCl₂, 1.8; MgCl₂, 2 and HEPES, 5 (pH 7.4), supplemented with 50 mg/L gentamycin sulphate and 90 mg theophylline. After ex vivo translation, the ion channels are correctly inserted in the cell membrane of the oocytes.

Two-electrode voltage-clamp recordings were performed at room temperature (18–22 °C) using a Geneclamp 500 amplifier (Molecular Devices, San Jose, CA, USA) controlled by a pClamp data acquisition system (Axon Instruments1, Union City, CA, USA) and using an integrated digital TEVC amplifier controlled by HiClamp, an automated Voltage-Clamp Screening System (Multi Channel Systems MCS GmbH, Reutlingen, Germany). Whole-cell currents from oocytes were recorded 1–7 days after injection. The bath solution composition was ND96 (in mM): NaCl, 96; KCl, 2; CaCl₂, 1.8; MgCl₂, 2 and HEPES, 5 (pH 7.5). Voltage and current electrodes were filled with a 3 M solution of KCl in H₂O. Resistances of both electrodes were kept between 0.5 and 1.5 MΩ. The elicited Kv10.1 and hERG currents were filtered at 0.5 kHz and sampled at 2 kHz using a four-pole low-pass Bessel filter. Leak subtraction was performed using a p/4 protocol. Starting from a holding potential of −90 mV, currents for Kv10.1 were evoked by 1 s depolarizing pulses to 0 mV and currents for hERG were evoked by 2.5 s depolarizing prepulses to +40 mV, followed by hyperpolarizing pulses to −120 mV for 2.5 s. The concentration dependency of all compounds was assessed by measuring the current inhibition in the presence of increasing compound concentrations. To this end, a stock solution of the compounds was prepared in 100% DMSO for the sake of solubility. From this stock solution, adequate dilutions with a maximum of 0.5% DMSO were made for testing. The data of the concentration–response curves were fitted with the Hill equation: y = 100/{1 + (IC₅₀/[compound])^h}, where y is the amplitude of the compound- induced effect, IC₅₀ is the compound concentration at half-maximal efficacy, (compound) is the compound concentration, and h is the Hill coefficient.

2.4. Statistical Analysis

All electrophysiological data are presented as means ± SEM of n ≥ 3 independent experiments unless otherwise indicated. Oocyte data were analysed using pClamp Clampfit 10.4 (Molecular Devices, Downingtown, PA, USA) and OriginPro 9 (Originlab, Northampton, MA, USA) or GraphPad Prism 8 software (GraphPad Software, Inc., San Diego, CA, USA). Patch-clamp data analysis was performed using FitMaster (HEKA Electronics), IgorPro (Wavemetrics, Inc., Lake Oswego, OR, USA) and GraphPad Prism 9 (GraphPad Software, Inc., San Diego, CA, USA). The Dunnett test and one-way ANOVA were performed to calculate the significance of the induced inhibition compared to the control (p < 0.05).

2.5. Cell Culture

MCF-7 (DSMZ ACC 115) and PANC-1 (DSMZ ACC 783) cells were grown as recommended by the cell supplier at 5% CO₂ and 37 °C in a humidified atmosphere. For determining cell growth, cells were seeded in 96-well culture plates (Corning, Corning, NY, USA) at a density of 10⁴ cells/well. After 24 h, the compounds were added diluted in medium. Six wells per condition were used. Four images of each well were collected at 1 h intervals for 48–72 h. Two biological replicates were performed. Images were acquired using an Incucyte device (Sartorius/Essen biosciences). Cell confluence was calculated using the Incucyte software on the phase contrast images. The results are expressed as percent confluence per well normalised to the value with no compounds added.

2.6. MTS Assay

The antiproliferative activities of the compounds were evaluated on the breast cancer cell line MCF-7 (ATCC HTB-22, adherent cells isolated from a 69-year-old white woman), using an MTS (Promega, Madison, WI, USA) assay according to the manufacturer’s instructions [29]. The MCF-7 cell line was obtained from the American Type Culture Collection (ATCC, Manassas, VA, USA). Cells were cultured in low-glucose Dulbecco’s Modified Eagle’s Medium (DMEM) (Sigma-Aldrich, St. Louis, MO, USA), supplemented with 10% fetal bovine serum (Gibco, Thermo Fisher Scientific, Waltham, MA, USA), 100 U/mL penicillin (Sigma-Aldrich, St. Louis, MO, USA), 100 µg/mL streptomycin (Sigma-Aldrich, St. Louis, MO, USA), and 2 mM l-glutamine (Sigma-Aldrich, St. Louis, MO, USA). The cell line was incubated in a 5% CO₂ atmosphere at 37 °C. Cells were plated out in 96-well plates at a density of 2000 cells per well and incubated overnight. The next day, cells were treated with the indicated compounds, positive control (17-DMAG), or vehicle control (0.5% DMSO) and incubated for 72 h. Then CellTiter96 Aqueous One Solution Reagent (10 µL; Promega, Madison, WI, USA) was added to each well and cells were incubated for an additional 3 h. Absorbance was measured at 492 nm using a microplate reader (Synergy 4 Hybrid; BioTek, Winooski, VT, USA). Independent experiments were repeated twice, each time in triplicate. Statistical significance (p < 0.05) was calculated using a two-tailed Welch’s t-test between treated groups and DMSO. IC₅₀ values are given as averages of the independent measurements and were calculated using GraphPad Prism 8.0 software (San Diego, CA, USA). IC₅₀ values represent the concentration at which a compound elicits a half-maximal response.

3. Results

3.1. Design

New K_V10.1 inhibitors were designed and synthesised to increase the potency of published compounds 1 (ZVS-08) and 2 against K_V10.1 and selectivity against hERG, and to investigate structure–activity relationships (SAR) important for K_V10.1 inhibition. Compound 1 (Figure 2) with diarylamine structure and 1-(dimethylamino)-propan-2-ol segment attached to aromatic moiety B was modified in the different parts (R¹-R⁶) of the molecule to obtain a focused library of compounds screened for K_V10.1 inhibition (Figure 2, strategies I-V). In strategy I, the basic dimethylamino centre of ZVS-08 (1) was extended to different cyclic amines bearing aromatic rings (R⁴, Figure 2) to obtain new analogues with three aromatic scaffolds. The aim of strategy II (Figure 2) was to investigate the importance of the amine linker (R¹) and the electron-withdrawing nitro (R²) and trifluoromethyl (R³) groups of the hit compound ZVS-08 (1) and compound 2 for K_V10.1 inhibition. As a result, several new analogues with modifications at four different positions (R¹-R⁴) were developed. The new analogues in strategy II contained the cationic dimethylamino centre of hit compound 1 or was replaced by a morpholino group (R⁴) to maintain the basic character of the compound but further increase its polarity. Using strategy III (Figure 2), we tested whether the basic dimethylamino centre was required for the inhibition of K_V10.1. Therefore, two new analogues with modifications of this segment were developed. Since the hydroxyl group of ZVS-08 (1) was found not to be an important structural component for the inhibition of K_V10.1 (based on the comparison between 1 and 2, Figure 2), the new analogues of strategy IV contained different aliphatic tertiary amine basic centres connected to the aromatic part B via a shorter ethoxy linker. The new dihalogenated analogues (strategy V, Figure 2) contained additional chloride or bromide atoms in aromatic part B and dimethylamino- or diethylamino-ethoxy groups.

3.2. Synthesis

The syntheses of the new analogues are shown in Scheme 1, Scheme 2, Scheme 3, Scheme 4 and Scheme 5. The differently substituted diarylphenols 3a–b, 7, 8a–b, and 12a–d were prepared through four different procedures. The synthesis of 4-hydroxyphenoxy analogues 3a–b was carried out with the addition of CuO as a catalyst and a base at high temperature (Scheme 1) [30]. For the preparation of benzamide 7, 2-nitro-4-(trifluoromethyl)benzoic acid was first converted with oxalyl chloride to the corresponding benzoyl chloride and the product was then refluxed in the presence of base and p-aminophenol to give benzamide 7 (Scheme 2). For the synthesis of compounds 8a–b without the electron-withdrawing nitro group, CuO and Cu were used as catalysts (Scheme 3) [30]. The carboxylic acids 8a–b were converted to the methyl esters 9a–b by addition of thionyl chloride in anhydrous methanol (Scheme 3). o-Nitro-p-trifluoromethyl-substituted diarylphenols 12a and 12c–d were synthesised in two separate steps (Scheme 4 and Scheme 5). First, the nitro group was introduced by nitration of 1-chloro-4-(trifluoromethyl)benzene with concentrated nitric and sulfuric acids to give 11, followed by the addition of the nonhalogenated or dihalogenated p-aminophenols at high temperature in the presence of a base to give 12a and 12c–d [31]. The dinitro-substituted compound 12b was prepared by coupling p-aminophenol and 1-fluoro-2,4-dinitrobenzene with the addition of a base at reflux (Scheme 4). Epoxides 4a–g were prepared using (±)-epichlorohydrin and K₂CO₃ as base (Scheme 1, Scheme 2, Scheme 3 and Scheme 4). For the epoxide opening, morpholine was used to give a racemic mixture of compounds 5a–f, dimethylamine to give a racemic mixture of compounds 6a–d, methylamine to give a racemic mixture of compound 15, and the corresponding amines with aromatic moieties to give a racemic mixture of analogues 16a–d. The methyl esters 5d and 6c were hydrolysed with 1 M sodium hydroxide to form the corresponding benzoic acids 10a–b (Scheme 3). Compounds 13, 19, 20, and 21 were prepared from o-nitro-p-trifluoromethyl-substituted diarylamine phenol 12a by addition of 4-(2-chloroethyl)morpholine for 13, 1-(2-chloroethyl)pyrrolidine for 19, 1-(2-chloroethyl)piperidine for 20, or 1-(2-chloroethyl)azepane for 21 and base at reflux (Scheme 4 and Scheme 5). Compound 14 was synthesised from 12a by addition of 4-benzyloxyaniline and base at high temperature (Scheme 4). Differently substituted diethylamino and dimethylamino analogues were formed at reflux from nonhalogenated or dihalogenated diarylamine phenols 12a or 12c–d by addition of base and 2-chloro-N,N-diethylethan-1-amine for 17a–c or 2-chloro-N,N-dimethylethan-1-amine for 18a–c.

3.3. Analogues of Compounds 1 and 2 and Structure–Activity Relationships

Five different types of new K_V10.1 inhibitors were designed and synthesised to increase the potency on K_V10.1. Based on the study of the structure–activity relationships (SAR) of the analogues of ZVS-08 (1), we identified the structural components in this structural type of K_V10.1 inhibitor that are important for K_V10.1 inhibition.

The new analogues 16a–d, based on three aromatic scaffolds (

Table 1. Structures and potencies of compounds from Strategy I.

Compound ID	R4	% of KV10.1 Inhibition at 50 μM (HEK-293 Cells)
16a		44.6 ± 8.6%
16b		60.7 ± 9.6%
16c		26.8 ± 10.1%
16d		21.8 ± 4.4%

, strategy I), were an unsuccessful attempt to increase the potency of previously published compounds 1 and 2. Similarly, almost all compounds from strategy II (

Table 2. Structures and potencies of compounds from Strategy II.

Compound ID	R1	R2	R3	R4	% of Kv10.1 Inhibition at 10 μM (HEK-293 Cells)
5a	O	COOCH3	H		4.1 ± 2.6%
5b	O	NO2	CF3		2.9 ± 5.3%
5c	NHCO	NO2	CF3		7.5 ± 3.5%
5d	NH	COOCH3	H		2.7 ± 1.2%
5e	NH	COOCH3	CF3		12.6 ± 6.4%
5f	NH	NO2	NO2		61.7± 9.1%
6a	O	NO2	CF3		18.8 ± 4.0%
6b	NHCO	NO2	CF3		1.7 ± 6%
6c	NH	COOCH3	CF3		4.2 ± 5.5%
6d	NH	NO2	NO2		61.9 ± 9.0%
10a	NH	COOH	H		5.8 ± 5.3%
10b	NH	COOH	CF3		−2.1 ± 5.7%

) were inactive at 10 μM unless an amine linker (R¹), two nitro groups (R² and R³), and morpholino (5f, R⁴) or dimethylamino (6d, R⁴) basic centres were present. Compound 14 from strategy III without a tertiary amine (R⁴) was inactive as a K_V10.1 inhibitor (

Table 3. Structures and potencies of compounds from Strategy III.

Compound ID	Structure	% of KV10.1 Inhibition at 10 μM (HEK-293 Cells)
14		3.3 ± 1.7%
15		66.6 ± 4.3%

), but analogue 15 with a secondary aliphatic methylamine showed significant inhibition of the Kv10.1 current. Because the results for the new analogues of strategies II and III showed that the amine linker (R¹), two electron-withdrawing groups (-NO₂ as R² and -CF₃ as R³), and the aliphatic tertiary amine (R⁴) were very important structural elements for the inhibitory activity, these structural elements were included in the new compounds of strategies IV and V, but the alkoxy segments of 1 and 2 (strategies IV and V) and their second aromatic part (strategy V) were modified. The new compounds in strategy IV (

Table 4. Structures and potencies of compounds from Strategy IV.

Compound ID	R5	% of Kv10.1 Inhibition at 10 μM (HEK-293 Cells)
13		62.4 ± 3.9%
17a		86.2 ± 6.8%
18a		81.4 ± 5.6%
19		68.4 ± 14.8%
20		65 ± 6.0%
21		71.0 ± 4.12%

), which included shorter ethoxy linkers and different aliphatic tertiary amines, such as diethylamine (17a), dimethylamine (18a), piperidin-1-yl (20) and azepan-1-yl (21) were the most potent K_V10.1 inhibitors. Some of the new dihalogenated diarylamines in strategy V (

Table 5. Structures and potencies of compounds from Strategy V.

Compound ID	R5	R6	% of Kv10.1 Inhibition at 1 μM (HEK-293 Cells)
17b		Br	38.2 ± 1.4%
17c		Cl	66.5 ± 5.5%
18b		Br	73.5 ± 0.5%
18c		Cl	28.4 ± 12.1%

) with basic diethylamine or dimethylamine centres caused a high-percentage inhibition of K_V10.1 already at 1 μM.

It can be concluded that the most potent new diarylamine inhibitors contain a combination of a disubstituted phenyl ring with para-trifluoromethyl and ortho-nitro groups and a phenyl ring with two meta-halogen groups or an unsubstituted phenyl ring. By modifications of the aliphatic chain between the basic centre and the aromatic ring of ZVS-08 (1), we succeeded in increasing the potency and the most potent K_V10.1 inhibitors contained a combination of a shorter ethyloxy linker and various aliphatic tertiary amines, of which diethylamine or methylamine were optimal.

Diethylamine-containing compound 17a and dimethylamine-containing compound 18a from strategy IV and novel 3,5-dibromo-(dimethylamino)ethoxy) analogue 18b from strategy V were further evaluated for K_V10.1 and hERG inhibition (dose–response curves, Figure 3). Compound 17a inhibited K_V10.1 with an IC₅₀ value of 568 nM and was more potent than the previously most active compound 2 (IC₅₀ = 740 nM). Compound 17a was also similarly potent against hERG (IC₅₀ = 232 nM) than 2 (IC₅₀ =207 nM), allowing us to slightly increase selectivity index against hERG with the new analogue. Compound 18b was the most potent new K_V10.1 inhibitor (IC₅₀ = 216 nM), but also inhibited hERG very strongly with an IC₅₀ value of 4.9 nM. Increasing the lipophilicity of the molecule by adding two meta-halogen groups to the unsubstituted phenyl ring, as in the case of 18b, increased the potency for both channels but impaired the affinity for hERG considerably more. It appears that the diarylamine core with a disubstituted phenyl ring containing ortho-nitro and para-trifluoromethyl groups and an unsubstituted phenyl ring is essential for K_V10.1 inhibition, but that with modifications of the alkoxy linker lengths and the basic dimethylamino centre, selectivity toward hERG can be increased.

The new analogues were also tested for their K_V10.1 and hERG inhibition in another independent assay ex vivo in Xenopus laevis oocytes expressing K_V10.1 or hERG (Tables S1–S5, Supplementary Materials). Some of the most active compounds (17a, 18a, 18c, 20, and 21) on HEK-293 cells also achieved the highest percentage of K_V10.1 in Xenopus laevis oocytes. In addition, we determined their IC₅₀ values for K_V10.1 and hERG and the most potent compound was 17a, which inhibited K_V10.1 currents with an IC₅₀ value of 3.6 µM and hERG currents with an IC₅₀ value of 2.0 µM (Table S4).

3.4. Antiproliferative Activities of Novel K_V10.1 Inhibitors 17a and 18b

K_V10.1 inhibitors are promising candidates that have antiproliferative activity against cancer cell lines expressing this channel. Our new, highly potent K_V10.1 inhibitors 17a and 18b were tested for their antiproliferative activity against the MCF-7 breast cancer cell line, which has high K_V10.1 and low hERG expression. Different concentrations of the compounds were added to the culture medium and the MTS assay was used to determine the number of viable cells after 72 h of incubation. The results obtained are shown in

Table 6. IC₅₀ values for K_V10.1, hERG, and the antiproliferative activities in the MCF-7 breast cancer cell line of the synthesized compounds 17a and 18b.

Compound ID	IC50 [nM] on KV10.1	IC50 [nM] on hERG1	IC50 [μM] on MCF-7
17a	568 nM	232 nM	18.5 ± 1.5
18b	216 nM	4.9 nM	11.9 ± 1.0

. Both compounds moderately inhibited the growth of MCF-7 cancer cells. Compound 18b, which was the most potent K_V10.1 inhibitor, inhibited the proliferation of MCF-7 cancer cells with an IC₅₀ value of 11.9 μM. Compound 17a, whose IC₅₀ value for K_V10.1 was approximately twice that of 18b, inhibited proliferation of MCF-7 cancer cells with an IC₅₀ value of 18.5 μM.

The effect of different concentrations of inhibitors 17a and 18b was tested on a cancer cell line with low hERG and high K_V10.1 (MCF-7, breast) expression and another cancer cell line with high hERG and very low K_V10.1 expression (Panc-1, from pancreas). Cell growth was then monitored over 48–72 h (Figure 4). Both compounds inhibited cancer cell growth to a very similar extent and also with similar potency, reducing proliferation of both cell lines by approximately half at 25 µM. Higher concentrations up to 100 µM only slightly increased growth inhibition. The MCF-7 cells were less sensitive to both compounds. The higher sensitivity of Panc-1 cells may be due to the high potency of both compounds in inhibiting hERG (Table 6), which is highly expressed in this cell line and is one of the major ion-channel-encoding genes involved in the initiation and maintenance of neoplastic growth [32,33]. In summary, 17a and 18b were able to reduce the growth of cancer cells expressing K_V10.1 or hERG, suggesting that both channels contribute to cancer cell proliferation. Therefore, our new inhibitors have a potential anticancer effect.

4. Conclusions

Two new potent K_V10.1 inhibitors were discovered by structural optimisation of previously published compounds based on the diarylamine scaffold. Important structural elements for the activity of these inhibitors were identified by SAR studies. We learned that the aliphatic chain between the basic centre and the diarylamine scaffold can be modified to increase potency at K_V10.1 and improve selectivity toward hERG. The new inhibitor 17a, with a shorter ethoxy linker and a basic diethylamine centre, showed improved potency (IC₅₀ = 568 nM) and a better ratio between IC₅₀ values for hEAG1 and hERG than previously published diarylamine inhibitors. The most potent analogue of the series, 18b, inhibited K_V10.1 with an IC₅₀ value of 216 nM, but also caused significant hERG inhibition. Compounds 17a and 18b moderately inhibited the growth of MCF-7 breast cancer cells, which have low hERG and high K_V10.1 expression. The new diarylamine K_V10.1 inhibitors provided new structural information required for K_V10.1 inhibition, but selectivity toward hERG remains a challenge.