1,4-Benzothiazepines with Cyclopropanol Groups and Their Structural Analogues Exhibit Both RyR2-Stabilizing and SERCA2a-Stimulating Activities

To discover new multifunctional agents for the treatment of cardiovascular diseases, we designed and synthesized a series of compounds with a cyclopropyl alcohol moiety and evaluated them in biochemical assays. Biological screening identified derivatives with dual activity: preventing Ca2+ leak through ryanodine receptor 2 (RyR2) and enhancing cardiac sarco-endoplasmic reticulum (SR) Ca2+ load by activation of Ca2+-dependent ATPase 2a (SERCA2a). The compounds that stabilize RyR2 at micro- and nanomolar concentrations are either structurally related to RyR-stabilizing drugs or Rycals or have structures similar to them. The novel compounds also demonstrate a good ability to increase ATP hydrolysis mediated by SERCA2a activity in cardiac microsomes, e.g., the half-maximal effective concentration (EC50) was as low as 383 nM for compound 12a, which is 1,4-benzothiazepine with two cyclopropanol groups. Our findings indicate that these derivatives can be considered as new lead compounds to improve cardiac function in heart failure.


■ INTRODUCTION
−3 Its main function is to initiate Ca 2+ release from the SR into the cytoplasm, which activates the contraction of the heart muscle by a process termed calcium-induced Ca 2+ release. 4,5The sarco/endoplasmic reticulum Ca 2+ -ATPase 2a (SERCA2a) is another cardiac Ca 2+ -transport SR protein, a subtype of SERCA transmembrane P-type ATPase, whose primary function is to pump calcium ions back into the SR after they are released by RyR2 during cardiac muscle contraction.This decreases the concentration of cytosolic Ca 2+ and initiates the diastolic relaxation phase of the myocardium during which a muscle relieves its tension.The Ca 2+ ions are then stored in the SR, until they are released again by activated RyR2 triggering muscle fiber contraction during the next heartbeat.−9 RyR2 channels in the healthy heart remain mostly closed during the relaxation phase of the cardiac cycle. 7owever, a sustained increased level of Ca 2+ leak from the SR in the diastolic phase has been reported in heart disease, which has been monitored, for example, in the form of Ca 2+ sparks representing the spontaneous opening of individual RyR2 clusters. 7Under HF conditions, the opening of the receptor is promoted by sustained stress signaling mainly by its hyperphosphorylation by protein kinase A (PKA) and/or calcium/ calmodulin-dependent kinase II (CAMKII), as well as oxidation resulting in chronically increased RyR2 channel activity and increased Ca 2+ leak from the SR. 2,3,5,10,11Increased RyR2 leak leads to Ca 2+ depletion from the SR reducing cardiac contractility and overactivates sodium−calcium exchanger (NCX) that induces abnormal cardiomyocyte depolarization, promoting cardiac arrhythmias. 12−15 Spontaneous Ca 2+ leak from the SR through the RyR2 macromolecular complex can be prevented by increasing its affinity to small stabilizing proteins, e.g., the peptidyl-propylcis-trans isomerase calstabin2 (FKBP12.6).−18 Restoring calstabin-RyR interaction is a therapeutic strategy that relies on a number of agents belonging to 1,4-benzothiazepines known as Rycals. 2,3,5,16,19,20−19 However, JTV-519 is also a nonspecific blocker of Na + , K + , and Ca 2+ channels and acts as a Ca 2+ -dependent SERCA blocker. 17,21,22In recent years, there has been considerable interest in ARM210, a second-generation Rycal compound, which binds to disease-associated ryanodine receptor RyR channels not only near the caffeine binding site but also in the RY1&2 domain, thereby stabilizing the RyR closed state by inhibiting the formation of the primed state and thus preventing pathological pore opening (Figure 1a,b). 20,23−27 Recently, Luraghi et al. reported on selective SERCA2a activators, derivatives of istaroxime metabolite, with unknown mechanism of action. 28There are published data on the N-aryl-N-alkyl-thiophene-2-carboxamide compound, which increases endoplasmic reticulum Ca 2+ load by enhancing SERCA2a-mediated Ca 2+ transport. 29The authors mentioned that the mechanism of action of this compound may involve synergistic effects on both SERCA2a and RyR2.
In our search for new chemical agents that improve maintenance of diastolic Ca 2+ levels, we aimed on the development of dual-acting drugs that both inhibit RyR2 Ca 2+ leak and simultaneously activate Ca 2+ uptake via SERCA2a.To this end, we developed novel 1,4-benzothiazepines, 3-[(4-methoxyphenyl)oxy]-and 3-[(4-methoxyphenyl)thio]propane-1-amine derivatives containing a cyclopropyl alcohol fragment as a pharmacophore (Scheme 1).The rationale was that the structural similarity between these molecules may result in similar physical properties and biological functions.Indeed, not only 1,4-benzothiazepines but also their "open" analogs, 3-[(4-methoxyphenyl)oxy]-and 3-[(4-methoxyphenyl)thio]propane-1-amine derivatives, reduced Ca 2+ leak from the ER in our assay on inducible RyR2-expressing HEK-293 cells.In addition, these compounds enhanced ATP hydrolysis mediated by SERCA2a activity in cardiac SR microsomes and increased the SR Ca 2+ content in HL-1 heart cells in the caffeine-induced Ca 2+ release assay.Therefore, the presence of cyclopropyl alcohol groups in 1,4benzothiazepines and their structural analogs results in compounds with dual RyR2-stabilizing and SERCA2a-activating properties.We report here the synthesis and structure−activity relationship of these compounds, some of which exhibit EC 50 in the nanomolar range for modulating either RyR2 or SERCA2a activity, suggesting their potential therapeutic application.
Characterization of Target Compounds: RyR2 Activity.To assess the functional potency of novel agents for modulating RyR2 channel activity, we measured changes in Ca 2+ levels in ER of HEK-293 cells with inducible expression of wild-type RyR2 (WT RyR2). 33,34Because the expression of WT RyR2 introduces an additional spontaneous Ca 2+ leak from the ER, these cells are suitable for our purpose of testing potential RyR2 stabilizers. 35The cells additionally stably express the Ca 2+ fluorescence indicator R-CEPIA1er, which is a genetically encoded sensor protein of Ca 2+ in the ER (the cell line was a gift from T. Murayama, Department of Pharmacology, Juntendo University School of Medicine, Tokyo, Japan). 36luorescence microscopy of live HEK-293 RyR2 R-CEPIA1er cells revealed a highly dynamic branching tubule network of ER structures, indicating the correct position of the calcium sensor in the ER (Figure 2a).Additionally, we used stimulated emission depletion (STED) microscopy and found out that the R-CEPIA1er indicator is compatible with superresolution live cell imaging. 37In STED microscopy, the stained ER structures appear about three times thinner than in the confocal one (for the line profiles, see Figure S1).
Using time-resolved fluorescence measurements (λ ex = 560 nm, λ em = 610 nm), we observed changes in the concentration of Ca 2+ in the ER ([Ca 2+ ] ER ) in HEK-293 RyR2 R-CEPIA1er cells upon addition of tested compounds that correspond to RyR2 channel activity (Figure S2a). 36We used a RyR2 inhibitor dantrolene, 38 as well as RyR stabilizers S36 8 and S107 3 as positive controls to test how their addition would affect the calcium concentration in ER in this model system.The addition of these compounds, which are known to inhibit excessive cardiac RyR2 activity, should reduce Ca 2+ leak from the ER, thus increasing the R-CEPIA1er fluorescence.
For the evaluations of our results, we calculated fluorescence ratio F/F 0 , where F 0 is the average fluorescence for the first 90 s and F is the average fluorescence for the last 100 s, and normalized it to the effect obtained from the control solution (Figure S2a).Indeed, the addition of 25 μM dantrolene or S36 or 1 mM S107 resulted in an increase in R-CEPIA1er fluorescence (Figure 2b).ARM210, 23 which belongs to the second generation of Rycals, significantly increased [Ca 2+ ] ER , proving its stabilizing activity in HEK-293 RyR2 R-CEPIA1er cells (Figure 2b).To avoid false-positive results due to autofluorescence of the tested compounds, we repeated experiments with HEK-293 cells endogenously expressing wild-type RyR2 without the Ca 2+ indicator (Juntendo University School of Medicine, Tokyo, Japan) and detected no fluorescence in this spectral region (data not shown).
Compounds 11 and 12, with the exception of 11d, significantly increased [Ca 2+ ] ER , in the RyR2 R-CEPIA1er assay (Figure 2b).Compounds 11a−c and 12 were tested in a concentration range from 10 −11 to 10 −4 M.These compounds increased F/F 0 in a dose-dependent manner (Figure S2b−d).At concentrations above 25 μM, compound 12a demonstrated the most pronounced effect of all of the other compounds.
Encouraged by these results, we moved on to a caffeine assay to investigate how the new compounds might modulate the SERCA2a activity.
SERCA2a Activity in HL-1 Cells.It is known that caffeine stimulates the release of calcium from intracellular stores in cells by acting as a RyR activator. 39,40The binding site for caffeine is located on the cytosolic side of the protein between the C-terminal and the S2S3 domain. 41When caffeine binds to RyR2, it interacts with specific amino acids on the protein, which causes conformational changes in the protein's structure. 42This opens the RyR2 channel pore, enabling the mass release of calcium ions from the SR.We investigated the effect of caffeine-induced Ca 2+ release in HL-1 cells treated with cytosolic Ca 2+ indicator FLIPR Calcium 6 (Molecular Devices).The HL-1 cell line is an immortalized mouse cardiomyocyte cell line commonly used to study the effects of various substances on cell function. 43If the tested compound modulates SERCA2a activation, it will cause an increase in Ca 2+ concentration in the SR and the effect of a fixed dose of caffeine on SR Ca 2+ release will be potentiated by increasing doses of a proposed SERCA activator. 26As a positive control, we used CDN1163, a small-molecule SERCA activator, which is at present in a clinical trial due to its potential therapeutic use for certain cardiac conditions. 27ndeed, the addition of 10 mM caffeine solution to HL-1 cells incubated for 2 h in CDN1163 solutions of various concentrations together with the cytoplasmic Ca 2+ indicator showed a clear dependence of the magnitude of the caffeine effect from the concentration of CDN1163 (Figure S3a,b).The half decay time (T 1/2 ) of caffeine-induced Ca 2+ release gradually increases from 3 to 50 s with increasing CDN1163 concentration (Figures S3a).We assumed that the decay of cytosolic [Ca 2+ ] in the presence of caffeine is an indicator of NCX activity, as postulated by Chen et al. 44 NCX is a transmembrane transporter in the cytoplasmic membrane that removes one Ca 2+ ion out of the cell in exchange for three sodium ions. 45Such a concentration-dependent change in T 1/2 may indicate that NCX activity slows down as the concentration of CDN1163 increases.CytoTox-Glo assay in 100 μM solutions after 24 h incubation.d "+++" a strong effect, comparable to or greater than that of a known compound at a given concentration (25 μM in RyR2 R-CEPIA1er assay, 10 μM in caffeine assay and on microsomal membrane vesicles), "++" moderate effect, "+" observable effect, "−" no effect.
Similarly to CDN1163, compounds 11 and 12 improved SERCA2a activity in the micromolar concentration range, with 11d being the least efficacious and 12c being the most efficacious of all other compounds including CDN1163 (Figures 3 and 4).Compound 11d showed no saturation effect at concentrations up to 50 μM, which did not allow us to calculate its EC 50 (Figure 4).The EC 50 values for 11a−c and 12a,c were found to be 2.2, 8.6, 7.6, 9.2, and 10.9 μM, correspondingly (Table 1).Compound 12b displayed a bellshaped dose response; it has shown a moderate to good stimulatory effect at 3 and 10 μM (up to 150% compared to the control) that decreased at higher doses, suggesting that the compound may exhibit inhibitory activity at higher concentrations (Figure 4).T 1/2 values demonstrated only a slight difference among samples with various concentrations of tested compounds (from 3 to maximum 5 s when the concentrations changed from 0 to 50 μM), suggesting no significant change of NCX activity within this set of compounds.
S36 has shown no concentration-dependent effect in the caffeine-induced Ca 2+ release assay (Figure 5 and Figure S4).This suggests that S36 does not increase SERCA2a activity, which agrees with our previous experiments on wild-type (WT) and disease model cardiomyocytes. 46Addition of caffeine to HL-1 cells treated with different concentrations of S107 did not show any effects on SR Ca 2+ content as well (Figure 5 and Figure S4).Surprisingly, the second-generation Rycal ARM210 drug demonstrated a clear concentrationdependent effect in this assay (Figure S3b). Figure 5 shows the data obtained after the incubation of HL-1 cells in 10 μM (25 μM for 11d) solutions of compounds.
SERCA2 Activity in Microsomal Membrane Vesicles.Cyclopropanol derivatives have a variety of biological activities including enzyme inhibition and antibacterial and anticancer properties. 47−50 However, there have been no published reports on cyclopropanol-containing molecules specifically acting as modulators of RyRs and/or activators of SERCA.To investigate more directly whether the compounds influence the SERCA2 activity, we performed nicotinamide adenine dinucleotide (NADH) fluorescence-coupled ATPase assay according to Radnai et al. 51 In this assay, the change in NADH coenzyme fluorescence reflects the level of the  consumption of ATP, which we took as indicator of SERCA2 activity. 51,52ATP hydrolysis by SERCA2 is required to pump calcium ions into the SR.We assume that a higher rate of ATP hydrolysis reflects an increase in SERCA2 ATPase activity that leads to a greater pumping of calcium ions into the SR.
We observed the rate of NADH oxidation, which is coupled to ATP hydrolysis, by measuring changes in NADH intrinsic fluorescence over time after ATP injection into microsomal membranes enriched in SR vesicles isolated from mouse heart ventricles (mouse SR) and HEK-293T cells (HEK-293 ER).The presence of SERCA2 in mouse SR and HEK-293 ER of isolated microsomal vesicles was confirmed by Western blotting (Figures S5 and S6; for details see Experimental Section).The NADH fluorescence decrease was recorded on a multiwell plate reader TECAN Spark 20M (λ ex = 380 nm, λ em = 470 nm). Figure 6a shows a typical decrease in NADH fluorescence after the injection of 2 mM ATP using mouse heart SR vesicles pretreated with compound 12a.Consistent with our previous experiment on HL-1 cells, we observed a strong decrease in NADH fluorescence reflecting an increase in SERCA2a activity in SR microsomes derived from mouse heart ventricles (n = 3−5) in solutions containing the tested compounds compared to the 0.1 v/v% DMSO sample (Figure 6).In microsomes from HEK-293 ER, the treatment with compounds 11a and 11b did not evoke a significant increase in ATPase activity while other compounds, 11c,d, 12a−c, and ARM210, showed significant activation of SERCA2 ATPase (n = 3−5) (Figure S7).Similar results were obtained in cardiac SR vesicles from the mouse heart.In this case, all new cyclopropanol compounds, as well as ARM210, caused the improvement of ATPase activity (Figure 6c).For compound 12a, ATPase activity increased in a concentration-dependent manner with an EC 50 of 16 and 383 nM for HEK-293 ER and mouse SR, respectively (Table 1, Figure 6b).This difference may indicate isoform-specific effects of our compound in HEK-293 cells expressing mostly SERCA2b 53 versus mouse SR vesicles containing mostly SERCA2a. 54In both experiments, we used CDN1163 as a positive control.
Drug-Related Cell Toxicity and Membrane Permeability.The cyclopropyl alcohol-substituted compounds showed no significant cell death at concentrations up to 50 μM in CytoTox-Glo Cytotoxicity Assay (Promega GmbH, Figure S8).The membrane permeability of the derivatives was determined using cell-free transport models (PermeaPad, innoME GmbH), and permeability rates (P e ), which are a measure of how fast a compound can cross a membrane, were established (Table 1).Albeit poor permeability shown in the PAMPA assay, compounds S36, 11a, and 12a acted on live RyR2-HEK-293 R-CEPIA1er cells and the latter two had an effect on HL-1 cells in the caffeine-induced Ca 2+ release assay.
The properties of the new compounds together with the properties of the known substances S36, S107, and ARM210 are given in Table 1.The values of calculated partition coefficients (log P) for cyclopropanol derivatives range from 2.2 to 4.4, which is in line with Lipinski's "rule of five" claiming that the calculated octanol−water partition coefficient should not exceed 5. 55 ■ DISCUSSION AND CONCLUSIONS Almost all cyclopropyl alcohols, except for 11d, which is a 1,4benzothiazepine derivative with the shortest C 4 H 8 linker, showed dual RyR2/SERCA2a activity.1,4-Benzothiazepine 11d had no significant RyR2-stabilizing effect on HEK-293 RyR2 R-CEPIAer cells but demonstrated slight (HL-1 and HEK ER) and moderate (mouse SR) SERCA2 stimulating activity.The RyR2 tests clearly point out that derivatives of 3-[(4-methoxyphenyl)oxy]-and 3-[(4-methoxyphenyl)thio]propane-1-amine, which are the mimetic analogues of 1,4benzothiazepine, can activate RyR2 almost to the same extent as known Rycals.Compound 11a, which is a 1,4benzothiazepine derivative with one cyclopropanol group and C 4 H 8 NHC 4 H 8 linker, was most potent in the caffeine-induced Ca 2+ release assay in HL-1 cells.Although we could not detect SERCA2 activation by this compound (10 μM) on ER vesicles derived from HEK-293T cells, we observed a significant effect on mouse heart microsomes (Figure 6c).
Under certain experimental conditions, compounds 11b,c and 12 demonstrated activities comparable to or superior to those of the known SERCA activator CDN1163.Compound 11c, which is the 3-[(4-methoxyphenyl)thio]propane-1-amine derivative with one cyclopropanol group, demonstrated nanomolar EC 50 in the RyR2 assay.The 1,4-benzothiazepine  and 12c), compound 12c had a stronger activating effect on caffeine-induced Ca 2+ release at concentrations above 10 μM, whereas 12b showed a slight increase in SERCA2astimulating activity up to 10 μM and a presumable inhibitory effect at higher concentrations, 25 and 50 μM (Figure 4).Results arising from physiological target validation of this group of compounds will be published in a separate manuscript.
The 1,4-benzothiazepine derivative ARM210, despite its known RyR2-modulating activity, was found to act as a SERCA2a activator.According to single-particle cryo-electron microscopy data, this compound modulates PKA-phosphorylated RyR1 and RyR2, binding to the secondary adenosine triphosphate (ATP) site in the RY1&2 domain. 20,23The authors proposed that the similarity of the ATP and ARM210 structures allows this Rycal to occupy the ATP site.It would be interesting for further studies to determine whether ARM210 may also interact with the ATP site of SERCA2a.
Figure 7 shows a schematic representation of the intracellular mechanism of dual RyR2-SERCA2a action of the novel compounds.The entry of a small amount of Ca 2+ through the L-type Ca channel (LTCC) activates RyR2 and initiates Ca 2+ release from the SR, a process called Ca 2+ -induced Ca 2+ -release or CICR, which leads to cardiomyocyte contraction. 56During the relaxation, Ca 2+ is removed from the cytoplasm and pumped back into the SR by SERCA2a.This requires that the RyR channels remain closed.In the heart cells at the pathological stage, one or both proteins might be dysfunctional, leading to abnormal distribution of intracellular Ca 2+ .PKA and CAMKII can phosphorylate RyR2 increasing its sensitivity toward activating Ca 2+ leading to hyper-active RyR2 that decrease SR Ca 2+ load due to Ca 2+ leak, which is not compensated by the reduced Ca 2+ uptake.Diminished SR Ca 2+ load is involved in impaired contractility and relaxation observed in HF.
The new cyclopropyl alcohol containing 1,4-benzothiazepines and their structural analogs act on RyR2 stabilizing its closed state while they activate SERCA2a, increasing SR Ca 2+ load.We showed that the relatively small changes in the structure of the new compounds lead to different biochemical properties with respect to RyR2 and SERCA2a.Both proteins, RyR2 and SERCA2a, are key players in cardiac Ca 2+ cycling.Therefore, the ability to obtain drugs that can be flexibly adjusted to individual needs (for example, with stronger RyR2 stabilization and moderate SERCA2a activation, or vice versa) could be used in future studies as to achieve dual-acting tools for personalized treatment, improving the contractility of heart muscle to prevent or slow HF and to prevent Ca 2+ triggered heart rhythm disorders on the individual level.
The dose−response curves were calculated using the software package GraphPad Prism version 8.3.1.(GraphPad Software, Inc.).All data are presented as mean ± SD in three independent experiments.Y = bottom + (top − bottom)/(1 + 10 ((LogEC 50 − X) × HillSlope)) was used, where HillSlope describes the steepness of the curve, and top and bottom are plateaus in the units of the Y-axis.
Changes in Ca 2+ concentration were measured on a multiwell plate reader TECAN Spark 20 M using a FLIPR 6 Calcium Assay Kit from Molecular Devices (Molecular Devices LLC, Munchen, Germany) at 37 °C.The 1000 × stock solutions of the test compounds were prepared in DMSO.A 100 μL solution of test compound in Tyrode's solution containing FLIPR Calcium 6 dye was incubated for 2 h at 37 °C and 5% CO 2 in a humidified incubator.The time courses were recorded in each well.Fluorescence was evoked at 485 nm excitation wavelength and read out in a bottom-read mode at 525 nm.Data were recorded every 2 s, exposure −20 flashes, excitation bandwidth −10 nm, emission bandwidth −15 nm.To initiate Ca 2+ influx, 25 μL of 50 mM caffeine solution in Tyrode's solution containing 2 mM CaCl 2 was injected in each well (speed 100 μL/s) at the 10 s time point, resulting in 10 mM final concentration in the well.The differences between the caffeine-induced peak minus basal fluorescence, ΔF Caff , were taken into account for the analysis.The Tyrode's solution +0.1 v/v% DMSO was used as a control.The dose−response curves were calculated using the software package GraphPad Prism version 8.3.1.All data are presented as mean ± SD from eight independent experiments.Y = bottom + (top − bottom)/(1 + 10 ((LogEC 50 − X) × HillSlope)) was used, where HillSlope describes the steepness of the curve and top and bottom are plateaus in the units of the Y axis.
Isolation of Microsomal Membrane Vesicles from HEK-293 Cells.Isolation of microsomal membrane vesicles from HEK-293T cells (HEK-293 ER) was performed according to Stewart et al. 61 HEK-293T cells were harvested with Trypsin/EDTA, washed twice with ice-cold PBS containing protease inhibitor cocktail (cOmplete, Roche, Mannheim, Germany), and centrifuged 800 × g for 5 min.The pellet was resuspended in ice-cold lysis buffer (1 mM EDTA, 10 mM HEPES, cOmplete, pH 7.4) and incubated on ice for 20 min.After the cells were homogenized with a tight-fitting Dounce homogenizer (10−20 strokes), an equal volume of restoration buffer (500 mM sucrose, 10 mM HEPES, cOmplete, pH 7.2) was added and the cells were homogenized again (10−20 strokes).The homogenate was centrifuged at 10,000 × g for 20 min at 4 °C.The supernatant was collected and centrifuged again for another 2 h (10 5 × g at 4 °C).Then, the pellet was carefully resuspended in resuspension buffer (250 mM sucrose, 10 mM HEPES, cOmplete, pH 7.2) on ice by pipetting the solution up and down until it was fully dissolved.The obtained solution was diluted to a protein concentration of 15−20 μg/μL with the resuspension buffer; the aliquots were snap frozen in liquid nitrogen and stored at −80 °C until usage.
Isolation of Microsomal Membrane Vesicles from Mouse Hearts.All animal procedures were performed in accordance with Directive 2010/63/EU of the European Parliament and the Council on the protection of animals used in research as well as Animal Welfare Law of the Federal Republic of Germany (Tierschutzgesetz der Bundesrepublik Deutschland, TierSchG).Sacrificing rodents for subsequent preparation of tissue did not require specific authorization or notification ( §7 Abs.Two Satz 3 TierSchG).All mice were housed with a 12 h light/dark cycle with free access to food and water.Adult C57BL/6N mice of both genders were sedated with isoflurane in a sealed container and quickly euthanized by cervical dislocation.Heart ventricles were taken out instantly, washed with ice-cold PBS plus complete, cut into smaller pieces, and snap frozen in liquid nitrogen straight away.
Lysis buffer thawed on ice (10 mM EDTA, 10 mM HEPES, cOmplete, pH 7.4) was added, and the ventricles were homogenized with a Qiagen Cellruptor for 20 s.On ice, an equal volume of restoration buffer (500 mM sucrose, 10 mM HEPES, cOmplete, pH7.2) was added and briefly mixed.The homogenate was centrifuged at 10 4 × g for 20 min at 4 °C.The supernatant was collected and centrifuged again for another hour (10 5 × g at 4 °C).Then, the pellet was carefully resuspended in resuspension buffer (250 mM sucrose, 10 mM HEPES, cOmplete, pH 7.2) on ice by pipetting the solution up and down until fully dissolved.The obtained solution was diluted to a protein concentration of 5 μg/μL with the resuspension buffer, and the aliquots were snap frozen in liquid nitrogen and then stored at −80 °C until usage.
SERCA2 Activity Measurements.To determine whether new compounds could increase the SERCA2 activity, HEK-293 microsomes (40−150 μg of total protein) or mouse ventricular microsomes (1−20 μg of total protein) were diluted in 100 μL of the assay buffer used for measurement of SERCA activity (described below) and incubated at room temperature for 5 min prior to the experiments.The measurements of SERCA2 activity were made at 28 °C by using a NADH-coupled ATPase assay as described in Radnai et al. with slight modifications. 51Briefly, the assay buffer contained 60 mM MOPS, 120 mM KCl, 6 mM MgCl 2 , 1 mM EGTA, 5 mM NaN 3 , 0.5 mM phosphoenolpyruvate, and 1.5 mM CaCl 2 , pH 7.0.Before the reaction was started by the injection of 5 mL of ATP (resulting in 2 mM final concentration in the well), five units/mL of both lactate dehydrogenase and pyruvate kinase, 0.5 mM NADH, and 1 μM Ca 2+ ionophore A-23187 (Sigma, C-7522) were added to the diluted sample with or without the tested compound.The 1.5 mM Ca/EGTA buffer has a free Ca 2+ concentration of about 200 μM.The NADH fluorescence decrease was recorded each 2 s for a 10 min course on a multiwell plate reader TECAN Spark 20 M in 96-well glass bottom plates (MatTek Corporation; Cat.No. PBK96G-1.5-5-F).Fluorescence was evoked by a 380 nm excitation wavelength and collected in a bottomread mode at 425 nm.Exposure −30 flashes, excitation bandwidth: 10 nm, emission bandwidth: 10 nm.ATP was injected into the well at 30 s.The amount of added protein, 7−13 μg/mL for mouse SR and 133−270 μg/mL for HEK ER, did not affect the slope of the fluorescence decrease in the control samples (0.1% DMSO).The data were analyzed with the computer software GraphPad Prism version 8.3.1 (GraphPad Software, Inc.).The difference in signal decrease with and without microsomes was taken as an indicator of the additional ATPase activity performed by SERCA2.The obtained values for maximal SERCA2 activity were normalized for the total protein concentration.Maximal SERCA2 activity values are reported as percentages of the control (0.1% DMSO) values.Microsomal samples were kept on ice until usage and set to 100%.
Statistical Analysis.Normality and log normality tests were performed using the Shapiro−Wilk test.The significance between means was tested using unpaired Student's t-test and one-way ANOVA.The Dunnett test was performed to compare every mean to a control (0.1% DMSO) mean.Results are represented as mean ± SD.A value of P < 0.05 was considered significant.
Western Blot Analysis.SERCA2a protein levels in microsomal membrane vesicles (HEK-293T cells and mouse hearts, 8−11-weekold mice, n = 10 males and 2 females, C57Bl/6N genotype) were determined by Western blot analysis.20 μL of microsomes were mixed with 2 μL of NuPAGE LDS Sample buffer (Thermo Fisher) and heated at 70 °C for 10 min.The samples (100−150 μg protein) were loaded onto a 3−8% gradient Tris−acetate gel or 4−20% SDS− Tris−glycine gel, and the gel was run for 1 h at 150 V in running buffer.After that, the gel was equilibrated for 10 min in 20% EtOH, transferred onto a nitrocellulose membrane, and blotted with the iBlot 2 system (dry blot) for 10 min at 25 V. Membranes were blocked with 5% skim milk for 30 min at room temperature and incubated with primary antibodies against SERCA2a (Badrilla A010-20 rabbit polyclonal, 1:1000) overnight at 4 °C with moderate shaking.Binding of the primary antibody was detected by a horseradish peroxidase (HRP)-conjugated secondary antibody (1 h, RT).The detection was done with the Western Lightning Plus ECL Kit (PerkinElmer LAS GmbH, Rodgau) and an Amersham Imager 600.
Cytotoxicity.Cell viability was assessed using the CytoTox-Glo Cytotoxicity Assay according to the manufacturing procedure (Promega GmbH).Briefly, HL-1 cells were cultivated for 24 h in black-walled, clear-bottom 96-well microplates (Corning) covered with fibronectin in DMEM, containing different concentrations of the tested compound (10,000 cells/well).Luminescence of a luminogenic peptide substrate was measured before and after the addition of the lysis reagent (digitonin).The viable cell contribution was determined by a subtractive method.0.1% DMSO-treated cells were taken as a control.
Membrane Permeability.Membrane permeability was assessed using PermeaPad Plate according to the manufacturing procedure (innoME GmbH, Espelkamp, Germany).Briefly, 200 μL of water was added into the acceptor plate.200 μL of 100−500 μM test compound in water was added directly to the well membranes of the donor plate.The donor plate contains a biomimetic membrane for simulating passive mass transfer through different barriers in the body.Then, the donor plate was placed into the acceptor plate wells and incubated at room temperature for 24 h in the dark.To determine the peak absorbance of test compounds, absorbance spectra of acceptor solutions and initial solutions for each test compound were read out.The permeability rate (P e ) was calculated using the formula: , where OD A is the absorbance of acceptor solution and OD E is the absorbance of the standard solution; if the compound is able to permeabilize the membrane and fully reach equilibrium, 250 μM will be the final concentration of solution in the donor and acceptor wells.The coefficient C was calculated using the formula: , where V A is the acceptor volume (cm 3 ), V D is the donor volume (cm 3 ), the membrane area is 0.24 cm 2 , and the time is 86,400 s.
The Supporting Information is available free of charge via the Internet at http://pubs.acs.org." 1 H and 13 C NMR spectra, LC-MS and HPLC spectra, and supplementary methods (resolution measurements, time-lapse fluorescence measurements, functional measurements in isolated ER vesicles, Western blot analysis, and cell viability measurements) (PDF) Molecular formula strings (CSV)

■ AUTHOR INFORMATION
Corresponding Authors

Figure 3 .
Figure 3. Caffeine-induced Ca 2+ release assay.HL-1 cells were incubated with the FLIPR Calcium 6 Ca 2+ indicator and then treated with various concentrations of compound 12c.Ca 2+ influx was initiated by the addition of 10 mM caffeine.The differences between the caffeine-induced peak minus basal fluorescence, ΔF Caff , were taken for the analysis.The inset shows the dose−response curve of 12c.The response on caffeine addition was normalized to the control (Cnt, 0.1 v/v% DMSO).Data represent as mean ± S.D., n = 8 independent measurements.

Figure 6 .
Figure 6.NADH-coupled ATPase assay.(a) Mouse ventricular microsomes were incubated in the assay buffer with and without 100 μM 12a.After ATP injection, the intrinsic fluorescence of NADH decreases due to ATP consumption.The reaction rate of NADH oxidation is greater in the mouse SR sample containing 12a.(b) Compound 12a evokes an enhancement in the kinetic rate of NADH−NAD + conversion in a dosedependent manner.The effects of compound 12a are expressed as a percentage of the control effect.Data represented as mean ± S.D., n = 3−15 independent measurements.(c) Effect of 10 μM CDN1163, 11, 12 (3 μM for 12b), and ARM210 on SERCA2a activity in mouse SR.The data were normalized to the control (Cnt, 0.1 v/v% DMSO).Values represent the mean ± S.D., * P < 0.05, ** P < 0.005, *** P < 0.001, **** P < 0.0001 vs control by unpaired t test; n = 9−24/3−5 mice.

Figure 7 .
Figure 7. Schematic representation of the modulating effect of cyclopropanol compounds 11 and 12 (modified from Niggli et al.56 ).LTCC = Ltype calcium channel; NCX = sodium−calcium exchanger that removes Ca 2+ from cells.Phosphorylated RyR2 is depicted by added beige stars with a "P" sign.The binding of Calstabin2 (FKBP12.6) to RyR2 is shown by brown circles.