Discovery and characterization of ORM‐11372, a novel inhibitor of the sodium‐calcium exchanger with positive inotropic activity

Background and Purpose The lack of selective sodium–calcium exchanger (NCX) inhibitors has hampered the exploration of physiological and pathophysiological roles of cardiac NCX 1.1. We aimed to discover more potent and selective drug like NCX 1.1 inhibitor. Experimental Approach A flavan series‐based pharmacophore model was constructed. Virtual screening helped us identify a novel scaffold for NCX inhibition. A distinctively different NCX 1.1 inhibitor, ORM‐11372, was discovered after lead optimization. Its potency against human and rat NCX 1.1 and selectivity against other ion channels was assessed. The cardiovascular effects of ORM‐11372 were studied in normal and infarcted rats and rabbits. Human cardiac safety was studied ex vivo using human ventricular trabeculae. Key Results ORM‐11372 inhibited human NCX 1.1 reverse and forward currents; IC50 values were 5 and 6 nM respectively. ORM‐11372 inhibited human cardiac sodium 1.5 (I Na) and hERG KV11.1 currents (I hERG) in a concentration‐dependent manner; IC50 values were 23.2 and 10.0 μM. ORM‐11372 caused no changes in action potential duration; short‐term variability and triangulation were observed for concentrations of up to 10 μM. ORM‐11372 induced positive inotropic effects of 18 ± 6% and 35 ± 8% in anaesthetized rats with myocardial infarctions and in healthy rabbits respectively; no other haemodynamic effects were observed, except improved relaxation at the lowest dose. Conclusion and Implications ORM‐11372, a unique, novel, and potent inhibitor of human and rat NCX 1.1, is a positive inotropic compound. NCX inhibition can induce clinically relevant improvements in left ventricular contractions without affecting relaxation, heart rate, or BP, without pro‐arrhythmic risk.

sodium 1.5 (I Na ) and hERG K V 11.1 currents (I hERG ) in a concentration-dependent manner; IC 50 values were 23.2 and 10.0 μM. ORM-11372 caused no changes in action potential duration; short-term variability and triangulation were observed for concentrations of up to 10 μM. ORM-11372 induced positive inotropic effects of 18 ± 6% and 35 ± 8% in anaesthetized rats with myocardial infarctions and in healthy rabbits respectively; no other haemodynamic effects were observed, except improved relaxation at the lowest dose.
Conclusion and Implications: ORM-11372, a unique, novel, and potent inhibitor of human and rat NCX 1.1, is a positive inotropic compound. NCX inhibition can induce clinically relevant improvements in left ventricular contractions without affecting relaxation, heart rate, or BP, without pro-arrhythmic risk.  Kormos et al., 2014;Oravecz et al., 2018). In addition, both mechanical relaxation and [Ca 2+ ] i decay remained unchanged (Kormos et al., 2014). Selective NCX inhibitors either slightly shorten action potential (AP) duration or have no effect in normal oxygen and ionic conditions. Overall, it seems that selective inhibition of NCX exerts minimal effecst on intracellular Ca 2+ or action potential duration (APD) under normal conditions. In heart failure, the intracellular Ca 2+ concentration balance is changed, and the role of NCX becomes even more important (Bers & Despa, 2006). In addition, NCX and intracellular Ca 2+ regulate each other and affect cardiac remodelling, as recently described (Primessnig et al., 2019). ORM-11035, a selective NCX inhibitor, attenuated cardiac hypertrophic remodelling and prevented cardiac dysfunction in rats exhibiting heart failure. NCX also reportedly controls the heart rate (HR) through its effects on the sinus and atrioventricular nodes (Kaese et al., 2017). In sinus node, the funny current (I f ) and the NCX current (I NCX ) together establish a strong depolarization capacity providing an important safety factor for stable pacemaking (Kohajda et al., 2019). NCX also plays a role in BP control in normotensive and hypertensive individuals (Zhang, 2013).
The solute carrier transporter gene family (SLC8) encodes several Na + /Ca 2+ exchanger subtypes. SLC8A1 gene overexpression and NCX1.1 protein up-regulation are linked to many pathological conditions that lead to reduced contractility and arrhythmias (Khananshvili, 2013). NCX is organized into 10 transmembrane segments and is localized in the sarcolemmal membrane (Jost et al., 2013;Shattock et al., 2015). Two Ca 2+ binding domains are known currently.
• Known NCX modulators are unselective small molecules or peptides.

What this study adds
• The new NCX inhibitor ORM-11372 was the most potent and selective inhibtor described so far.
• ORM-11372 exerted positive inotropic effects in rabbits, without pro-arrhythmic risk in human cardiac tissue.

What is the clinical significance
• ORM-11372 exerted positive inotropic effect without other haemodynamic effects. et al., 2016) and all three were highly selective NCX inhibitors (Jost et al., 2013;Kohajda et al., 2016). Despite their higher selectivity, their use as positive inotropic tool compounds in experiments in vivo was prevented by their poor solubility (solubility of SEA0400 is less than 10 μg ml -1 in pH 7.4 phosphate buffer which places it in the solubility class "insoluble"). ORM-10962 is an exception from previously reported selective NCX inhibitors, as it exhibits reasonable solubility in in vivo studies.
Here, we describe the discovery of a new type of positive inotropic compound, ORM-11372, exhibiting high selectivity for NCX1.1 and provide its pharmacological profile in vitro, in vivo, and ex-vivo in human ventricular trabeculae. ORM-11372 was developed for acute short-term use with fast clearance.

| Discovery of a novel chemical series
A novel and unique chemical series was discovered using ligand-based pharmacophores for virtual screening with Catalyst (Accelrys) ( Figure 1a). Pharmacophore features used for virtual screening were derived from previously discovered NCX1 inhibitor flavan structures, such as ORM-10103 (Koskelainen et al., 2003) and ORM-10962 (Otsomaa et al., 2004). Based on the results of virtual screening, a proposed library with 636 commercially available compounds was selected for testing in the fluorescence-based assay, at concentrations of 10 μM. The original hit ORM-120407 inhibited NCX1.1 by 87% and had an IC 50 value of $200 nM. It inhibited human ether-á-go-gorelated gene (hERG) and L-type Ca 2+ channels, with IC 50 values of 2.1 and 3.1 μM respectively. These results indicated that the scaffold exhibited optimization potential. The structure-activity relationship (SAR) of ORM-120407 was explored further leading to the discovery of ORM-11372 (Table 1 and Figure 2), was discovered during medicinal chemistry optimization.
An analysis of 250 previously discovered NCX inhibitor compounds by Orion (Koskelainen et al., 2003;Otsomaa et al., 2004) resulted in a five-feature pharmacophore, which was used for the virtual screening of Cambridge and Specs compound libraries. In silico hits were further filtered based on predicted activity, calculated logarithm of solubility (clogS), calculated logarithm of partition coefficient (clogP), and diversity. The substructure features were optimized in parallel processes, and beneficial structural features were merged, in order to identify the potential overall synergistic effects for NCX1.1 inhibition. Representative samples of 135 synthesized derivatives are presented in Figure 2.
The hydrogen bond donor property of aniline in the original hit molecule ORM-120407 was proven to be important for NCX1.1 activity in the scaffold. The addition of polar substituents was not tolerated in the A ring, but halogen substitution on the p and m positions was tolerated and resulted in improved NCX1 inhibition. However, neither position was favoured over the other nor showed synergistic or additive effects. Their selectivity towards the hERG channel was found to be its differentiating property. SAR tolerated five-membered B-ring systems better than sixmembered systems. Ring heteroatoms, that is, oxygen or nitrogen molecules at position 1 (of furan) in combination with a 1,5-substitution, were proven to be critical for binding; that is, a carbon at position 1 abolished activity. This finding indicates that the presence of a hydrogen bond acceptor at that position is important. Bring optimization provided three potential ring systems that were tolerated: furans, oxazoles, and thiazoles. Two latter systems tolerated 4,2-substitution and 2,4-substitution, while the three heteroatoms in the B-ring reduced NCX1 inhibition activity. Though ORM-120407 showed good inhibitory activity towards NCX1, its selectivity towards hERG (2.1 μM) and L-type calcium (3.1 μM) channels needed to be improved. ORM-120407 also had poor solubility (<10 μgÁml −1 ). ORM-F I G U R E 1 The upper part of the figure shows the pharmacophore model used for virtual screening. In silico hits were further filtered based on predicted activity, calculated logarithm of solubility (clogS), calculated logarithm of partition coefficient (clogP), and diversity. The middle part displays the hit structure and substructure features optimized in parallel processes. At the bottom of the figure, two routes used for the synthesis of the compounds are presented, namely, (1) reductive amination and (2) alkylation under basic conditions 11298 exhibited better inhibitory activity towards NCX1 and good selectivity towards hERG channels. The oxazole series also exhibited good selectivity towards L-type calcium channels such as ORM-11298, but many derivatives were chemically unstable. The substitution of the C-ring phenyl was unnecessary, but some substituents such as aniline and chlorine were tolerated. The replacement of the phenyl ring with different heteroaromatic ring systems was mainly tolerated.

| Synthesis
Most compounds were prepared via reductive amination (Figure 1c, Route 1), except for ORM-11298 and ORM-11863, which were prepared by aniline alkylation (Figure 1c, Route 2). The reaction between the aniline and carbaldehyde derivatives could occur in the presence of a strong acid at elevated temperatures. The reduction of the imine intermediate could be carried out using a suitable reducing agent, such as sodium borohydride (NaBH 4 ). The alkylation reaction in Route 2 is performed in the presence of a base. The products were isolated from the reaction mixture by extraction with ethyl acetate, followed by evaporation. The starting materials used in the processes were either commercially available or could be prepared via synthetic routes (Parry, Bryce, & Tarbit, 2003;Ye et al., 2010). ORM-11372 was synthesized via the reductive amination of 5-(3-nitrophenyl)furan-2-carbaldehyde with the corresponding fluoroaniline, followed by the hydrogenation of the T A B L E 1 NCX inhibition values, as IC 50 , and the selectivity profile towards hERG and L-type Ca 2+ channels for selected compounds  (Sher, Gotti, Pandiella, Madeddu, & Clementi, 1988

| Housing and husbandry
The rats were maintained in standard rat cages (

| 5-HT 2B receptors
Changes in the intracellular calcium concentration of CHO-5-HT 2B cells in the probenecid-R inger were measured, using Fluo-4 or calcium 3 dyes and the FLEXstation, at 37 C. ORM-11372 was applied at seven concentrations (four replicates), 10 s after the initiating the measurement process in the agonist assay, but was preincubated (at eight concentrations with three replicates) in the antagonist assay, with 10-nM 5-HT added at the 10-s time point.
The measurement of current in the K + -free bath solution using a ramp voltage protocol at 20-s intervals represented the first control, after which values were measured in the presence of one or more increasing concentrations of ORM-11372 and finally upon exposure to 10-mM NiCl 2 . The voltage protocol consisted of voltage ramps (at a rate of 100 mVÁs −1 ) with a holding potential of −40 to 60 mV, which changed to −100 mV and then returned to −40 mV. I NCX was defined as the Ni 2+ -sensitive current value.
However, for rat ventricular myocytes, the magnitude of the inward current (forward mode) was particularly small and variable.
Therefore, to enhance the inward current and enable a solution with a certain concentration of ORM-11372 to be measured, the concentrations of the K + -free and pipette solutions were altered. In the bath solution, the CaCl 2 concentration was reduced to 0.5 mM, while in the pipetted solution, the NaCl and EGTA concentrations were decreased to 5 and 10 mM respectively. All experiments were performed at 35-37 C. The outward human I NCX inhibition current was measured at the following ORM-11372 concentrations: 3 (n = 4), 10 (n = 6), 30 (n = 4), and 100 (n = 3) nM.
The outward I NCX current in rat CMs was measured at concentrations ranging from 1 to 1,000 nM (n = 5). The inward I NCX current in rats was measured in additional experiments under changed conditions, to enhance the inward I NCX current. In these experiments (n = 3), the effect of ORM-11372 was tested at a concentration of 10 nM, at which the IC 50 and reverse I NCX current values were approximately equal.

| L-type calcium current (I CaL )
I CaL was recorded from hiPSC-derived CMs (at room temperature, n = 2) and rat ventricular myocytes at 1 μM (n = 3; 5 cells) and at the pH was adjusted to 7.2 using CsOH. I CaL was evoked by a 400-ms depolarization process to 0 mV from a holding potential of −40 mV every 5 s or, in the case of the rat CMs, by 400 ms; depolarizations to potentials ranging from −35 to 55 mV occurred after a prepulse to −40 mV, from the holding potential of −80 mV.
The pipette solution contained (pH 7.2 with CsOH; osmolarity adjusted to 270 ± 3 mOsm) 130 caesium methane sulfonate, 5 MgCl 2 , 5 EGTA, 0.1 GTP, 4 ATP disodium salt hydrate (Na 2 ATP), and 10 HEPES. The voltage protocol, which was repeated after each second, included of the process of hyperpolarization, from a holding potential of −80 to −120 mV for 200 ms, followed by that of depolarization to −15 mV for 10 ms. The peak current values observed while applying the test pulse at −15 mV were used for analysis.

| K V 11.1 channel current (I hERG )
Values from hERG-expressing cells (n = 4) were recorded with the standard extracellular solution at the physiological temperature and a pipette solution containing (in mM) 130 KCl, 7 NaCl, 5 EGTA, 1 MgCl 2 , 5 Na 2 ATP, and 5 HEPES (pH was set to 7.2 with KOH; osmolarity was adjusted to 290 ± 3 mOsm). I hERG was evoked using a voltage protocol, which was repeatedly performed every 10 s, consisting of a depolarization step from the holding potential of −75 to 10 mV for 500 ms, followed by a repolarization step to The extracellular solution contained (in mM) 145 NaCl, 4 KCl, 2 CaCl 2 , 1 MgCl 2 , 10 glucose, and 10 HEPES (pH 7.4 with NaOH; osmolarity adjusted to 305 mOsm). The intracellular recording solution contained (in mM) 120 KCl, 1.75 MgCl 2 , 5.37 CaCl 2 , 4 Na 2 ATP, 10 EGTA, and 10 HEPES (pH 7.2 with KOH; osmolarity adjusted to 295 mOsm). The voltage protocol, which was repeated every 10 s, included a 200-ms step to change the holding potential from −80 to −50 mV, to measure the leak current, and further depolarization to +20 mV for 2 s, followed by a repolarization to −50 mV for 2 s. The hERG tail current was measured as the difference between the peak tail current amplitude during the repolarization step and leak current measurement step. Subsequently, the voltage dependence of the block was also assessed, by carrying out depolarization for 4 s, from −80 to +60 mV in 10-mV steps, before and after the addition of 20-μM ORM-11372 (n = 7). The peak tail current elicited during the 5 s repolarization step to −50 mV was measured and plotted against the preceding depolarization step voltage values.

| Action potentials
APs were recorded in spontaneously beating Cor.4U CMs in the current clamp using the perforated patch technique and voltage-sensitive

| Donor heart procurement
All the human hearts used for this study were obtained after legal consent and were provided by organ donors in the United States.

| APs in human ventricular trabeculae
Procedures used for tissue dissection and recording were similar to those described previously (Page et al., 2016). Briefly, the human heart was transferred into a dissection vessel containing a cold (4 C), fresh proprietary dissection solution. The heart was completely submerged into the dissection solution. Ventricular trabeculae were dissected and transferred to the recording chamber.
The approach used to record APs is similar to that described by

| Selection of animal species
The rat is the most widely used rodent species in toxicology studies performed for drug development. The Sprague-Dawley rat strain (RRID:RGD_70508) is commonly used in the rat myocardial infarction (MI) model (Fishbein, Maclean, & Maroko, 1978). However, the functional similarity between the Ca 2+ handling proteins (including NCX) in rats (Bassani, Bassani, & Bers, 1994) and human hearts is rather low. The Ca 2+ circulation balance in the guinea pig and rabbit hearts are more similar to that of a human (Milani-Nejad & Janssen, 2014). Rabbits and humans are also known to react similarly to medication; hence, the rabbit was selected as a nonrodent species.

| Housing and husbandry
Animals were monitored daily by laboratory personnel. If the general health status of an animal was significantly worsened, the animal was killed with an overdose of pentobarbital. Human endpoints included no spontaneous movements and inability to drink or eat during the 24-h observation period, massive bleeding, spontaneous inflammation, missing anatomical features or swelling, and breathing difficulties. Specific pathogen-free animals were housed in half-barrier rooms where special protective clothing was required by personnel. Animal rooms were cleaned regularly three times per week, and cages and bottles were changed at regular intervals once a week. The temperature was maintained at 22 ± 2 C and humidity at 55 ± 15%. In the light-dark cycle, lights were kept on from 6:00 a.m. to 8:00 p.m. Body weights of all animals were measured weekly.

| Randomization and blinding
Randomization was used whenever feasible. Guinea pigs were ran- All experiments were carried out at 37 C. period before experiments was at least 5 days. In the MI model, rats (7-8 weeks, 200-250 g) were randomized into either the MI or sham group. In this weight range, a rat is considered a young adult, when its evolution phase is characterized by slow growth and its surgical mortality is lower than that of older animals. Male rats were used to minimize the variability of the cardiac response to several stimuli (Zornoff, Paiva, Minicucci, & Spadaro, 2009).

| Experimental MI model
Rats were anaesthetized with a combination of ketamine immediately. To prevent dehydration, rats were subcutaneously administered with 5 ml of 0.9% NaCl. After surgery, rats were given analgesia with buprenorphine for at least 2 days (0.05 mgÁkg −1 , s.c. twice a day). For sham-operated rats, the same procedure was performed, without ligation. Rats were analysed for 7 days after the MI, for signs of severe acute heart failure such as oedema or breathing difficulties.

| Haemodynamic measurements of anaesthetized animals
All animals were given multiple ascending infusion doses of ORM- At the end of the haemodynamic experiments, rats and rabbits were killed with an overdose of pentobarbital. Small laboratory equipment (forceps, scissors, scalpels, etc.) were sterilized in the glass bead sterilizer at 300 C for 10 seconds.
The BP waveform was measured continuously in milliseconds.
Blinding was not considered relevant, because the absolute value was not sensitive to biased interpretations, and all parameters are derived from of them.

| Haemodynamics in MI rats
Haemodynamics were assessed 7 days after the MI (n = 6) or sham operation (n = 6) as follows. Rats were anaesthetized with isoflurane (2.25-2.5%, Baxter) in carbogen (95% O 2 and 5% CO 2 ) and nitrous oxide (1:1), using a small rodent ventilator (Ugo Basile 7025, $10 mlÁkg −1 , 60 strokesÁmin −1 ). The rats were infused with 0.9% NaCl in the carotid vein at the stabilization and baseline levels. ORM-11372 was administered into MI (n = 6) and sham (n = 6) rats in ascending infusion doses of 1.7, 17, 167, and 417 μgÁkg −1 Ámin −1 . Infusions were administered into the jugular vein of MI rats at the infusion rate of 5 mlÁkg −1 Áh −1 (Terumo TE-311, Belgium). Doses were selected based on a pilot study with healthy rats ( Figure S20). The total number of animals in the study was 12 + 2, due to the replacement of two MI rats. One MI rat died while administering anaesthesia, as infarcted animals are sensitive to anaesthesia, and another MI rat was excluded due to the dosing error indicated by the bioanalysis of plasma samples (no ORM-11372 concentration observed in plasma).
After haemodynamic assessments, the anaesthetized rats were killed with an overdose of pentobarbital (2 ml of Mebunat® vet 60 mgÁml −1 per rat). The entire heart of each animal was fixed in buffered 4% formaldehyde solution, trimmed, processed, and embedded in paraffin. The hearts were cut horizontally at three different levels: apex, mid part, and base. Sections (4 μm) were cut from each level stained with haematoxylin and eosin, for general histopathological analysis. Infarct sizes were determined using Picrosirius Red staining as an indicator of cardiac fibrosis, and the size was measured using AnalysisPro software. The infarct area was calculated as the per cent of fibrotic tissue in the total myocardial area (mean of three levels).

| Blood sampling for bioanalysis
At the end of each infusion process, blood samples (300-500 μl) were added to a chilled EDTA polypropylene tube (CapiJect®, Terumo) and centrifuged (  STV was calculated as the beat-to-beat variability in repolarization from APD90 Poincare plots over a 30 sec duration. STV for APD90 was calculated as follows:

| Data and statistical analysis
where APD (n) and APD(n+1) are the APDs for the nth AP and the following AP respectively.
The effects of ORM-11372 were quantified relative to the data collected during the vehicle control period. Threshold values for changes over the baseline control for APD30, APD50, APD90, triangulation, and STV at 1 and 2 Hz pacing frequencies have been determined in a previous validation study (Page et al., 2016). Results are expressed as mean ± SEM.
The data and statistical analysis of in vitro and in vivo studies comply with the recommendations for experimental design and analysis in pharmacology (Curtis et al., 2018). Data were analysed statistically only when the number of independent samples was 5 or more. A P-value of <0.05 was considered to be significant.

| Other chemicals
The suppliers of other materials used are indicated as follows: Isoflurane

| Nomenclature of targets and ligands
Key protein targets and ligands in this article are hyperlinked to corresponding entries in the IUPHAR/BPS Guide to PHARMACOL-OGY (http://www.guidetopharmacology.org) and are permanently archived in the Concise Guide to PHARMACOLOGY 2019/20 (Alexander, Christopoulos et al., 2019;Alexander, Fabbro et al., 2019;Alexander et al., 2019aAlexander et al., , 2019b.

| ORM-11372 inhibits human and rat NCX activity
ORM-11372 concentrations dependently inhibited the increase in intracellular calcium in the insect cell line expressing human NCX1.1.
The insect cell line was used as the screening assay. The IC 50 for the inhibition of NCX in the reverse mode was 6.2 ± 0.4 nM.
The effects of ORM-11372 for NCX was subsequently studied using the whole-cell patch-clamp technique in human iPS-derived CMs (hiPSC-CMs; Figure 3) and rat ventricular CMs (Figure 4). The solutions and experimental protocol used to measure the bidirectional NCX current (I NCX ) were identical for both preparations. As shown in Figure 3a, the current was recorded in response to repetitive voltage ramp pulses first in a K + -free bath solution after blocking Na + , Ca 2+ , K + , and Na + /K + pump currents, to yield the baseline, and ORM-10103, SEA0400, KB-R7943, and lastly SN-6 whose IC 50 appears >100 μM.

| Selectivity
The functional selectivity of ORM-11372 towards the L-type Ca 2+ channel was first tested with the IMR-32 neuroblastoma cell line, using fluorometric images of intracellular changes in calcium levels.

Depolarization of undifferentiated IMR-32 cells by KCl addition
induced an increase in intracellular calcium levels, which can be suppressed by pre-incubation with verapamil or the 1,4-dihydropyridine nicardipine (Sher et al., 1988). L-type calcium channels mediated changes in intracellular calcium levels, and these

| Human ventricular trabeculae
ORM-11372 caused no significant changes in APD 30 , APD 50 , and APD 90 values up to a concentration of 10 μM, at both 1-and 2-Hz pacing frequencies. However, there was a trend in the reduction in APD, particularly APD 30 , at 1 Hz ( Figure S17); this seemingly contrasts with the significant shortening of hiPSC-CM APD by ORM-11372 (at 0.1 or 0.3 μM; Figure S18). As expected, the effect of the positive control (0.1-μM dofetilide) was observed in trabeculae tested with ORM-11372 ( Figure 7); individual values are shown in Figure S17 ( Page et al., 2016;Qu et al., 2017). Moreover, ORM-11372 showed no effects on STV and triangulation, two pro-arrhythmia markers that are used to identify pro-arrhythmic risk (Page et al., 2016;Qu et al., 2017), at a pacing rate of 1 or 2 Hz, at any tested concentration (Individual values in Figure S19). However, the addition of 0.1-μM dofetilide to the trabeculae tested with ORM-11372 increased STV and triangulation ( Figure S18). Taken together, these data clearly suggest that ORM-11372 can be classified as devoid of proarrhythmic risk.

| In vivo haemodynamics in rat MI model
Infarcted areas were 14.5 ± 1.8% and 1.1 ± 0.6% in the MI and shamoperated groups respectively ( Figure S20)

| DISCUSSION
Although the first NCX inhibitor, exchanger-inhibiting peptide (XIP), was a peptide, it enabled us to study the role of NCX under physiological and pathological conditions in cells and isolated tissues (Hobai et al., 1997). However, it took 20-30 min for NCX to be inhibited; it also inhibited SERCA and PMCA (Enyedi & Penniston, 1993). The next important step was the isolation and characterization of the naturally occurring NCX inhibitor (intrinsic factor-inhibiting NCX [NCX IF ]) (Hiller, Shpak, Shavit, Shpak, & Khananshvili, 2000). It was the first specific, cell membrane-penetrating NCX inhibitor (Shpak, Hiller, Shpak, & Khananshvili, 2003), but the structure of NCX IF was not published.
The first generation compounds (Table S1) that were specifically targeted to inhibit NCX were used as pharmacological tools to assess the inotropic mechanisms of various drugs like cardiac glycosides (Ruch, Nishio, & Wasserstrom, 2003;Tanaka et al., 2007) and insulin (Hsu et al., 2006). The potency of the first generation NCX inhibitors was modest, and selectivity between NCX and other membrane currents was poor, especially towards the L-type Ca 2+ current (Birinyi et al., 2005). SEA0400 was positively inotropic in rats, but not in rabbits , in which the role of NCX is similar to that of humans. For the second-generation NCX inhibitor ORM-10103 (Koskelainen et al., 2003), the selectivity between NCX and other currents was improved (Kormos et al., 2014). The following molecule ORM-10962 (Otsomaa et al., 2004) exhibits even better potency and more improved solubility . Discovery of ORM-11372 created the ground for the third-generation NCX inhibitors with a unique structure combined with improved drug-likeness.
In this paper, we show that ORM-11372 inhibited NCX 1.1 reverse and forward mode currents with a similar potency in both human iPS and rat CMs. Furthermore, the inhibition of human and rat NCX 1.1 currents was shown to be of the same magnitude (IC 50 was 5 vs. 10 nM). NCX 1.1 expression has been demonstrated in iPS-CMs F I G U R E 8 Effects of ORM-11372 on haemodynamics in isoflurane-anaesthetized rats, 7 days after the induction of myocardial infarction (MI, n = 6) or in sham rats (n = 6). Effects on the left ventricular contractility (LV + dP/dt max ) are shown in panel (a), along with values for (b) relaxation (LV − dP/dt max ), (c) heart rate, and (d) systolic arterial BP. Data shown are individual values with means ± SEM; n refers to number of rats. *P<0.05, significantly different from base line (BL); two-way repeated measures ANOVA followed by Dunnett's post hoc test (Fine et al., 2013;Kodama et al., 2019). To our knowledge, the presence or absence of the other NCX subtypes in hiPS has not yet been reported. Furthermore, NCX 1.1 expression is tissue specific (Lee, Yu, & Lytton, 1994); it can be stated that ORM-11372 is a potent and selective NCX 1.1 inhibitor compared with other cardiac ion channels, but NCX subtype selectivity is unknown. NCX has specialized role in ECC. Cell-specific regulation of NCX1 expression is due to a cardiac specific promoter, which leads to very high expression of NCX 1.1 in cardiac tissue (Nicholas et al., 1998). Hence, the effect of ORM-11372 is cardiac specific. In earlier studies, it was speculated that positive inotropic effects could be attributable to the asymmetrical blocking of NCX (Oravecz et al., 2018). ORM-11372 induced positive inotropy in healthy and MI rats, although it symmetrically inhibited both inward and reverse NCX currents. In rat plasma, the effective free concentration range of ORM-11372 was from 4 to 300 nM ( Figure S20), corresponding to the NCX inhibition range of IC 25 to IC 80 in rat CMs (Figure 4c). ORM-11372 induced positive inotropy even in vivo, in the healthy rabbits, in contrast to SEA0400 . The fact that the ORM-11372 is more than 10 times more potent than SEA0400, as a NCX inhibitor, explains its positive inotropic effect in rabbits. This was also confirmed in hNCX1-HEK assay ( Figure 5). ORM-11372 increased twitch tension (Figure 6d), which indicates increased SR load. Furthermore, unlike SEA0400, ORM-11372 inhibited NCX 1.1 current selectively, and other CM currents were negligibly inhibited.
ORM-11372 had no effect on relaxation levels in normal rats and rabbits, but at the lowest dose, it improved relaxation levels in MI rats.
Rat haemodynamics is subject to limitation of the fact that rat CM relaxation depends less on NCX and more on the other Ca 2+ handling mechanisms. An earlier study conducted in vitro using SEA0400 showed that NCX inhibition caused deterioration in transient Ca 2+ levels, but not mechanical relaxation (Szentandrassy et al., 2008). An ex vivo study using rat isolated perfused hearts with artificial ionic modification, demonstrated that NCX inhibition also depressed mechanical relaxation levels (Chen et al., 2012). However, it must be emphasized that the earlier studies were performed using the weaker and less selective inhibitor SEA0400 (Table 1), under artificial conditions. ORM-11372 did not affect relaxation in guinea pig papillary muscle (Figure 6g) or in rabbits in vivo, because of the activation of compensatory Ca 2+ handling mechanisms in CMs, as increased intracellular Ca 2+ levels enhance SERCA and/or PMCA activity. Furthermore, ORM-11372 had no effects on resting tension (Figure 6e) indicating no increase in diastolic calcium. ORM-11372 did not either affect time to peak (Figure 6f), which demonstrated that calcium release from ryanodine receptors is normal. Our findings are in line with the study using NCX IF . Haemodynamic assessments in the present study were made only in male rats and rabbits, which might introduce potential bias due to sex differences. Our results with human trabeculae and guinea pig papillary muscle did not show any differences between sexes, although the number of experiments are low. However, calcium F I G U R E 9 Effects of ORM-11372 on haemodynamics in S-ketamine-anaesthetized rabbits (n = 5). The effects on (a) left ventricular contractility (LV + dP/dt max ), (b) left ventricular relaxation (LV − dP/dt max ), (c) heart rate, and (d) systolic arterial BP are shown. Data shown are individual values with means ± SEM; n refers to the number of rabbits. *P<0.05, significantly different from base line (BL); one-way repeated measures ANOVA followed by Dunnett's post hoc test handling activity and NCX expression level in CMs have been shown to vary with sex, age and regional expression of NCX (Janczewski & Lakatta, 2010). Therefore, the role of NCX in calcium handling between sexes should be systematically further explored.
ORM-11372 selectively inhibited I NCX over the K v 11.1 current and provided a safety margin (ratio ≥2,000) of well beyond 30-fold, which was considered adequate for ensuring an acceptable degree of safety from arrhythmogenesis (Redfern et al., 2003). This selectivity was also confirmed in human ventricular tissue. Thus, ORM-11372 can be classified as being non-pro-arrhythmogenic in humans.
Although the NCX current seems to be crucial for beat generation (Groenke et al., 2013), even a small NCX current is enough to maintain the heart beat (Gao et al., 2013). In the present study, ORM-11372 had no effects on HR in vivo in rats or rabbits, and up to 80% NCX can be inhibited without any effects on the HR ( Figure S20). In vascular arterial cells, NCX contributes to the maintenance of the myogenic tone . In a manner similar to SEA0400 (Yatabe et al., 2015), ORM-11372 had no effects on systolic BP. These pharmacological properties resulted in unique positive inotropic effects without the risk of hypotension, increase in HR, or impairment of left ventricular relaxation.
During the discovery of ORM-11372, multiple iterative DMTA cycles (Design, Make, Test, Analyse) of optimization (Andersson et al., 2009) were performed in compliance with the 3R principles, using the protein-and cell-based in vitro methods, before selecting of the compounds for in vivo studies. Group size was optimized but was large enough to obtain reliable results. Animals were also allowed to become habituated to the procedures, and the necessary analgesia was used. Rats and rabbits were selected to ensure similar responses and improved prediction capabilities from in vitro to in vivo studies, which ultimately would be translated into better efficacy and safety in human studies. To improve pharmacokinetic and pharmacodynamic analyses, the plasma levels of ORM-11372 were monitored at the end of each pharmacodynamic experiment.
A critical feature of the ORM-11372 scaffold was the bridged aniline structure with hydrogen bond formation between the A-and Bring systems (Figures 1 and 2). The first-and the second-generation NCX inhibitors lack this hydrogen bond donor in their scaffold. Furthermore, the activity of the ORM-11372 series exhibited better potency in the screening assay than the known NCX inhibitors (Tables S1 and S2). This indicates that an additional specific binding interaction was identified between the ligand and the NCX protein.
However, the finding remains speculative due to the lack of other experimental evidence, for example, protein-ligand crystal structure.
The crystal structure of the mammalian NCX and its splice variants are still unavailable, though the crystal structure of the NCX protein of archaebacterial Methanococcus jannaschii has been published in 2012 (Liao et al., 2012). There are known structural differences between the prokaryotic and eukaryotic NCX proteins that might lead to differences in the potential binding pockets and selectivity (John, Liao, Jiang, & Ottolia, 2013;Khananshvili, 2014). Thus, evidence regarding the binding pocket for ORM-11372, as well as for the other NCX inhibitors is yet to be provided.
In conclusion, ORM-11372 is the first in a class of the third gen-