New N-aryl-N-alkyl-thiophene-2-carboxamide compound enhances intracellular Ca²⁺ dynamics by increasing SERCA2a Ca²⁺ pumping

Abstract

The type 2a sarco/endoplasmic reticulum Ca²⁺-ATPase (SERCA2a) plays a central role in the intracellular Ca²⁺ homeostasis of cardiac myocytes, pumping Ca²⁺ from the cytoplasm into the sarcoplasmic reticulum (SR) lumen to maintain relaxation (diastole) and prepare for contraction (systole). Diminished SERCA2a function has been reported in several pathological conditions, including heart failure. Therefore, development of new drugs that improve SERCA2a Ca²⁺ transport is of great clinical significance. In this study, we characterized the effect of a recently identified N-aryl-N-alkyl-thiophene-2-carboxamide (or compound 1) on SERCA2a Ca²⁺-ATPase and Ca²⁺ transport activities in cardiac SR vesicles, and on Ca²⁺ regulation in a HEK293 cell expression system and in mouse ventricular myocytes. We found that compound 1 enhances SERCA2a Ca²⁺-ATPase and Ca²⁺ transport in SR vesicles. Fluorescence lifetime measurements of fluorescence resonance energy transfer between SERCA2a and phospholamban indicated that compound 1 interacts with the SERCA-phospholamban complex. Measurement of endoplasmic reticulum Ca²⁺ dynamics in HEK293 cells expressing human SERCA2a showed that compound 1 increases endoplasmic reticulum Ca²⁺ load by enhancing SERCA2a-mediated Ca²⁺ transport. Analysis of cytosolic Ca²⁺ dynamics in mouse ventricular myocytes revealed that compound 1 increases the action potential-induced Ca²⁺ transients and SR Ca²⁺ load, with negligible effects on L-type Ca²⁺ channels and Na⁺/Ca²⁺ exchanger. However, during adrenergic receptor activation, compound 1 did not further increase Ca²⁺ transients and SR Ca²⁺ load, but it decreased the propensity toward Ca²⁺ waves. Suggestive of concurrent desirable effects of compound 1 on RyR2, [³H]-ryanodine binding to cardiac SR vesicles shows a small decrease in nM Ca²⁺ and a small increase in μM Ca²⁺. Accordingly, compound 1 slightly decreased Ca²⁺ sparks in permeabilized myocytes. Thus, this novel compound shows promising characteristics to improve intracellular Ca²⁺ dynamics in cardiomyocytes that exhibit reduced SERCA2a Ca²⁺ uptake, as found in failing hearts.

Significance

SERCA2a plays a central role in the intracellular Ca²⁺ homeostasis of cardiac myocytes, pumping Ca²⁺ from the cytosol into the sarcoplasmic reticulum (SR) lumen to maintain SR Ca²⁺ load. In this study, we characterized the effect N-aryl-N-alkyl-thiophene-2-carboxamide or compound 1 on SERCA2a function. Compound 1 enhances SERCA2a-mediated Ca²⁺ transport in SR vesicles, increases endoplasmic reticulum Ca²⁺ load and endoplasmic reticulum Ca²⁺ uptake in HEK293 cells expressing human SERCA2a, and enhances SR Ca²⁺ load and Ca²⁺ transients in cardiomyocytes. Thus, this novel compound can be used to improve intracellular Ca²⁺ dynamics and contraction in cardiomyocytes that exhibit reduced SERCA2a function, as found in failing hearts.

Introduction

The sarco/endoplasmic reticulum Ca²⁺-ATPase (SERCA) plays a central role in Ca²⁺ signaling by removing cytosolic Ca²⁺ and maintaining the sarco/endoplasmic reticulum (SR/ER) Ca²⁺ load. This pump uses the energy of ATP hydrolysis to translocate two Ca²⁺ ions from the cytosol into the SR/ER lumen, creating a steep Ca²⁺ gradient between the ER/SR lumen and the cytosol. This gradient is critical for Ca²⁺ signaling, as activation of SR/ER Ca²⁺ release channels produces a robust cytosolic Ca²⁺ transient that is essential for regulation of many cellular processes, including growth, migration, secretion, gene regulation, metabolism, neurotransmission, and contraction (1). Due to its critical role in Ca²⁺ homeostasis, SERCA expression and function are tightly regulated by posttranslational modifications, hormones, regulatory peptides, and microRNAs (2).

There are 10 isoforms of SERCA expressed in different tissues of mammalian species (3). The type 2a isoform or SERCA2a plays a central role in the cardiac contractile cycle. The rate of SERCA2a Ca²⁺ transport determines the heart’s relaxation rate during diastole when the heart refills with blood. SERCA2a activity also plays a major role in setting SR Ca²⁺ load, which determines the magnitude of SR Ca²⁺ release and the strength of cardiac contraction during systole (4). In ventricular myocardium, SERCA2a activity is predominantly regulated by the small transmembrane protein phospholamban (PLB). Under resting conditions, PLB binds to SERCA2a and inhibits SR Ca²⁺ transport by decreasing the pump’s apparent affinity for cytosolic Ca²⁺ (5). During adrenergic stress, phosphorylation of PLB by cAMP-dependent protein kinase (PKA) relieves SERCA2a inhibition by PLB, enhancing SR Ca²⁺ uptake (6). SERCA2a regulation by PKA is the major mechanism behind accelerated relaxation and increased contraction strength during adrenergic receptor activation. Diminished SERCA function and altered regulation have been reported in several pathological conditions, such as cardiac and skeletal myopathies (7,8,9). Consequently, the Ca²⁺ pump has become a promising therapeutic target for treating muscle diseases associated with disrupted Ca²⁺ homeostasis. Improving SERCA function has been also shown to be a viable therapeutic strategy in animal models of type 2 diabetes and Parkinson disease (10,11).

Existing strategies to improve SERCA function during pathological conditions include in vivo gene delivery (12,13,14,15) and small-molecule allosteric SERCA activators (9,10,11,16,17,18). Although SERCA2a gene therapy showed promising results in initial clinical studies (19,20), it failed to reach primary and secondary endpoints in a subsequent clinical trial on heart failure (CUPID 2) (21). Although the exact reasons are unclear, the probable culprits were protocol changes versus CUPID 1. The pharmacological approach has been shown to be effective in activating SERCA function. However, its clinical application is still limited by the scarcity of available lead compounds. Therefore, there is a high demand for developing new drugs that can effectively restore normal Ca²⁺ homeostasis in the diseased heart by enhancing SERCA2a function. Using both fluorescence resonance energy transfer (FRET) and biochemical measurements of Ca²⁺-ATPase activity, we have recently identified several new SERCA modulators, including compound 1 (22). In the present study, we took advantage of a sequence of complementary approaches that connect the effects of compound 1 on SERCA2a function and on ER/SR Ca²⁺ regulation to arrhythmia mitigation in the cardiomyocyte. Our results indicate that new compound 1 may improve intracellular Ca²⁺ dynamics in cardiomyocytes by increasing SERCA2a-mediated Ca²⁺ transport and reducing SR Ca²⁺ leak.

Materials and methods

Vectors and stable cell lines

pCMV R-CEPIA1er was a gift from Dr. Masamitsu Iino (Addgene, USA). The vector encoding the human type 2 ryanodine receptor (RyR2) cDNA fused to GFP at the N-terminus domain was kindly provided by Dr. Christopher George (University of Cardiff, UK). Human wild-type SERCA2a (SERCA2a^WT) cDNA was cloned into the mCerluean-M1 (mCer) modified plasmid (Addgene, USA) using KpnI and NotI restriction enzymes, yielding SERCA2a fused to mCer at the N-terminus. The sequence for human PLB was synthesized by GenScript and inserted at the 3′ end of mMaroon1 to create a mMaroon1-PLB fusion construct. SERCA2a cDNA was also cloned into the inducible expression vector pcDNA5/FRT/TO for the generation of SERCA2a stable cell line. The sequences were all verified by single pass primer extension analysis (ACGT Inc., USA).

Chemical synthesis

Compound 1 has been identified in a screen of DIVERSet-CL (ChemBridge, San Diego, CA), and initially tested as repurchased from the ChemBridge Hit2Lead online chemical store (compound ID, DS81260867) and validated via mass spectrometry (MS), ¹H and ¹³C NMR. To obtain sufficient quantities for this project, we have re-synthesized compound 1 (Fig. 1 a). Briefly, compound 1 was successfully synthesized from 2-methoxybenzylamine, 1-bromopropane, and 4,5-dimethylthiophene-2-carboxylic acid in a two-step synthetic procedure. First, the propyl group was transferred from 1-bromopropane to 2-methoxybenzylamine by N-alkylation in DMSO. The obtained N-(2-methoxybenzyl)propan-1-amine was then coupled to 4,5-dimethylthiophene-2-carboxylic acid employing HBTU/HOBt under microwave irradiation to obtain N-(2-methoxybenzyl)-4,5-dimethyl-N-propylthiophene-2-carboxamide (i.e., compound 1) in 65% yield.

Figure 1.

Compound 1 chemical structure and concentration response curves of Ca²⁺-transport and Ca²⁺-ATPase activity and FRET of SERCA-PLB interaction. (A) Chemical structure of compound 1. (B) Effect of compound 1 on Ca²⁺-transport and Ca²⁺-ATPase activity of SR vesicles isolated from porcine heart tissue. Measurements were carried out in the presence of compound 1 at the indicated concentrations. Data are represented as means ± SEM (n = 3) and were fitted using the Hill function, yielding compound potency (EC₅₀) and maximum effect (Δ_max) as indicated for ATPase (open symbols) and uptake (solid symbols), at V_max [Ca²⁺] (black) and intermediate [Ca²⁺] (red). (C) Fluorescence lifetime (FLT) response of the SERCA-PLB FRET biosensor to a range of [compound 1]. Measurements were conducted in 384-well microplates read in an FLT plate reader. Samples were obtained from a stable HEK cell line expressing the FRET biosensor (solid symbols) or a donor-alone control (open symbols). Null controls containing the corresponding volume of DMSO were read at the same time. Data are represented as means ± SEM (n = 3). To see this figure in color, go online.

SERCA2a Ca-ATPase and Ca²⁺-uptake measurements using porcine cardiac SR vesicles

Cardiac SR vesicles were isolated from porcine ventricular tissue, as described (23). To determine compound effect on the Ca²⁺-dependent ATP hydrolysis activity of SERCA2a, we used an NADH-coupled enzyme-linked assay in 384-well plate format, as previously described (16). Briefly, each well contained a solution (termed “assay mix”) consisting of 50 mM MOPS (pH 7.0), 100 mM KCl, 5 mM MgCl₂, 1 mM EGTA, 0.2 mM NADH, 1 mM phosphoenol pyruvate, 10 IU/mL of pyruvate kinase, 10 IU/mL of lactate dehydrogenase, 1 μM of the calcium ionophore A23187, and added CaCl₂ to obtain pCa 6.2 (intermediate Ca) and pCa 5.4 (V_max Ca). Next, 10 μg/mL of cardiac SR, CaCl₂, the tested small-molecule, and the assay mix were incubated for 20 min at 21°C. The assay was started upon the addition of ATP, to a final concentration of 5 mM (80 μL total assay volume). The time-dependence of 340-nm absorbance was measured in a SpectraMax Plus³⁸⁴ microplate spectrophotometer (Molecular Devices, Sunnyvale, CA).

To determine compound effect on the Ca²⁺-uptake activity of SERCA2a, we used an oxalate-supported assay, as previously described (24). A solution consisting of 50 mM MOPS (pH 7.0), 100 mM KCl, 30 mg/mL sucrose, 1 mM EGTA, 10 mM potassium oxalate, 2μM Fluo-4, 30 μg/mL cardiac SR vesicles, and CaCl₂ calculated to reach the desired free [Ca²⁺] (intermediate or V_max Ca, as above) was dispensed into 384-well plates containing the tested small molecule and incubated at 21°C for 20 min while protected from light. To start the reaction, MgATP was added to a final concentration of 5 mM, and the decrease in 485-nm excited fluorescence was monitored at 520 nm for 15 min using an FLIPR Tetra (Molecular Devices, San Jose, CA).

[³H]ryanodine binding to SR vesicles

Cardiac SR vesicles were isolated from porcine cardiac ventricle using previously described protocols (25). Using 96-well microplates, SR samples (3 mg/mL) were preincubated 30 min with DMSO or 10 μM compound 1 (dissolved in DMSO, final DMSO 0.02% in the well) in media containing 20 mM PIPES (pH 7.0), 150 mM KCl, 5 mM GSH, 2 mM DTT, 1 μg/mL aprotinin/leupeptin, 0.1 mg/mL BSA, 5 mM ATP, 1 mM EGTA plus CaCl₂ (calculated using MaxChelator to yield 0.1 or 30 μM free Ca²⁺). Binding of [³H]ryanodine was determined after a 3-h incubation at 37°C. Nonspecific and maximal [³H]ryanodine binding to SR were separately assessed by addition of 40 μM nonradioactively labeled ryanodine or 5 mM adenylyl-imidodiphosphate, respectively. These controls were each present in four wells/plate. Samples were filtered through grade GF/B Glass Microfiber filters (Brandel, Gaithersburg, MD) using a 96-channel Brandel harvester, and filters placed in 4 mL of Ecolite scintillation cocktail (MP Biomedicals, Solon, OH). [³H] retained on the filter was determined using a PerkinElmer Tri-Carb 4810 scintillation counter.

Isolation of ventricular myocytes

Male and female C57Bl6/J mice, (13 animals, Jackson Laboratories) were housed according to approved IACUC guidelines. Mice aged between 2 and 5 months were anesthetized using isoflurane (1%). After thoracotomy, hearts were quickly excised, immersed in Ca²⁺ free buffer, mounted on a Langendorff apparatus, and retrogradely perfused with a solution (37°C) containing Liberase H (Roche), according to a procedure described previously (26). The left ventricle was excised from the digested heart, placed in stop buffer containing BSA 1 mg/mL, cut into several pieces (average size 1 mm), and gently triturated into single cells. Myocytes were pelleted by gravity (∼0.1 mL) and resuspended in low-Ca²⁺ Tyrode buffer (in mM: NaCl 140; KCl 4; CaCl₂ 0.1; MgCl₂ 1; glucose 10; Hepes 10; pH 7.4). [Ca²⁺] was gradually adjusted to 1 mM. Isolated cardiomyocytes were stored at room temperature (20°C). All chemicals and reagents were purchased from Sigma-Aldrich (St Louis, MO, USA).

Culturing and transfection of HEK293 cells

HEK293 cells at 60%–80% confluency were transiently co-transfected with plasmids containing the cDNA of RyR2, SERCA2a, and R-CEPIA1er. Experiments were conducted 48 h after transfection to obtain the optimal level of recombinant proteins expression. Flp-In T-Rex-293 stably expressing SERCA2a were co-transfected with plasmids containing the cDNA of RyR2 and R-CEPIA1er, using the same conditions for the HEK293 cells. The surface membrane was permeabilized with saponin (0.005%). Experiments were conducted after washing out saponin with an experimental solution (in mM): K-aspartate 100; KCl 15; KH₂PO₄ 5; MgATP 5; EGTA 0.35; CaCl₂ 0.22; MgCl₂ 0.75; HEPES 10; dextran (MW: 40,000) 2% and pH 7.2. Free [Ca²⁺] and [Mg²⁺] of this solution were 200 nM and 1 mM, respectively.

Confocal microscopy

Changes in the cytosolic [Ca²⁺] ([Ca²⁺]_i) and the luminal ER [Ca²⁺] ([Ca²⁺]_ER) were measured with laser scanning confocal microscopy (Radiance 2000 MP, Bio-Rad, UK) equipped with a ×40 oil-immersion objective lens (N.A. = 1.3).

Measurement of [Ca²⁺]_i

To record [Ca²⁺]_i, we used the high-affinity Ca²⁺ indicator Fluo-4 (Molecular Probes/Invitrogen, Carlsbad, CA, USA). To load the cytosol with Fluo-4, ventricular myocytes were incubated at room temperature with 10 μM Fluo-4 AM for 15 min in Tyrode solution (in mM: NaCl 140; KCl 4; CaCl₂ 2; MgCl₂ 1; glucose 10; Hepes 10; pH 7.4), followed by a 20-min wash. Fluo-4 was excited with the 488-nm line of an argon laser, and the emission signal was collected at wavelengths above 515 nm. Changes in [Ca²⁺]_i were expressed as changes in F/F₀, where F₀ is the Fluo-4 signal at the resting condition. Ca²⁺ sparks were measured with Fluo-4 in permeabilized ventricular myocytes as described previously (27,28). Permeabilized myocytes were studied in an internal solution composed of (in mM): K-aspartate 100; KCl 15; KH₂PO₄ 5; MgATP 5; EGTA 0.35; CaCl₂ 0.1; MgCl₂ 0.75; phosphocreatine 10; HEPES 10; Fluo-4 pentapotassium salt 0.04 mM; dextran (MW: 40,000) 8%, and pH 7.2 (KOH). Free [Ca²⁺] and [Mg²⁺] in this solution were 100 nM and 1 mM, respectively. Sparks were detected and analyzed using the SparkMaster algorithm (29).

Measurement of [Ca²⁺]_ER

[Ca²⁺]_ER was recorded as changes in fluorescence intensity of the genetically encoded ER-targeted Ca²⁺ sensor R-CEPIA1er in HEK293 cells (30,31). R-CEPIA1er was excited with a 514-nm line of the argon laser, and signal was collected at >560 nm. Line scan images were collected at a speed of 20 ms/line. The R-CEPIA1er signal (F) was converted to [Ca²⁺]_ER by the following formula: [Ca²⁺]_SE = K_d × [(F – F_min)/(F_max – F)]. F_max was recorded in 5 mM Ca²⁺ and 5 μM ionomycin, and F_min was recorded after ER Ca²⁺ depletion with 5 mM caffeine. The K_d (Ca²⁺ dissociation constant) was 564 μM (32). SERCA-mediated Ca²⁺ uptake was calculated as the first derivative of [Ca²⁺]_ER refilling (d[Ca²⁺]_ER/dt) after RyR2 inhibition with ruthenium red (15 μM) and tetracaine (1 mM). ER Ca²⁺ uptake rate was plotted as a function of [Ca²⁺]_ER to analyze maximum ER Ca²⁺ uptake rate and maximum ER Ca²⁺ load. All 2D images and line scan measurements for [Ca²⁺]_SR were analyzed with ImageJ software (NIH, USA).

Microplate FLT-FRET measurements in live HEK293-6E cells

Using a previously reported FRET biosensor that concatenates (fuses) RFP (acceptor), SERCA, GFP (donor), and PLB in a single polypeptide chain (24), we have determined the profile of fluorescence lifetime FRET (FLT-FRET) response to a range of compound 1 concentrations (concentration response curves, CRCs). This SERCA2a-PLB biosensor was generated starting from a previously developed RFP-SERCA2a construct (33), which was fused to a flexible linker (47 residues) followed by GFP-PLB. A donor-only construct was created using the same approach. Each biosensor DNA was subcloned into the puromycin-resistant expression vector pTT22, and stable clones expressing the SERCA2a-PLB fusion biosensor at high levels were generated as described (24). These clones were used to acquire the FRET CRC measurements in live HEK293-6E cells. Stable cell lines over-expressing either donor-acceptor or donor-only biosensor were cultured in F-17 media (Thermo Fisher) with 200 nM/mL GlutaMAX (Invitrogen), 25 μg/mL Geneticin (Thermo Fisher), and 2.0 μg/mL puromycin (Invitrogen). On the day of testing, cells were harvested and centrifuged at 300 g for 5 min. Cell pellets were resuspended and washed in PBS (Thermo Fisher) three times before being passed through a 70-μm cell strainer. Cell concentration and viability were evaluated by a Countess automated cell counter (Invitrogen) using the Trypan Blue method. Samples for CRCs were prepared in 384-well format using a Mosquito HV liquid dispenser (SPT Labtech, Melbourne, UK). Next, 1 μL of DMSO or compound dissolved in DMSO was dispensed into 384-well black polystyrene plates (Greiner Bio-One) yielding a final assay concentration range of 0–100 μM. Finally, 50 μL of cells at a density of 2.0 × 10⁶ cells/mL were dispensed into each well using a Multidrop Combi (Thermo Fisher Scientific). Plates were sealed, incubated at room temperature, and FLT measurements were taken 120 min using a high-throughput FLT plate reader (Fluorescence Innovations, Minnesota, USA) provided by Photonic Pharma, as previously described (24).

Statistics

Data are presented as mean ± standard error of the mean (SEM) of n measurements. When only two groups were compared, statistical significance was determined by Student's t-test. Significance between multiple groups was determined by one-way ANOVA followed by a Newman-Keuls post hoc test. p < 0.05 was considered statistically significant. Statistical analysis and graphical representation of averaged data were carried out on OriginPro7.5 software (OriginLab, USA).

Results

Effects of compound 1 on SERCA ATPase activity and Ca²⁺ transport in SR vesicles

Using an enzyme-linked NADH-coupled Ca²⁺-ATPase assay of rabbit fast skeletal muscle SR, we have carried out a high-throughput screen of 40,000 drug-like small-molecules in the DIVERSet-CL library (ChemBridge, San Diego, CA) (22). In that screen, compound 1 (10 μM) was identified as an enhancer of the Ca²⁺-ATPase activity as measured at pCa 5.4 ([Ca²⁺] that produces the maximum ATP hydrolysis rate; V_max) in SR from rabbit fast skeletal muscle. Here, we used the same enzyme-linked assay to measure the response of Ca²⁺-ATPase activity of porcine cardiac SR over a range of compound 1 concentrations, thus constructing CRCs (Fig. 1 b). To gain further insight into this compound’s mode of action, we compared CRCs acquired at the V_max Ca²⁺ (pCa 5.4) and at intermediate Ca²⁺ (pCa 6.2). We observed robust (>50%) ATPase activation at V_max Ca²⁺, with a low-μM EC₅₀ (5.4 ± 0.3 μM). However, when ATPase was measured at pCa 6.2, maximum activation by compound 1 reached only ∼20%. This behavior was not surprising for a compound that was discovered using a V_max screen. A similar pattern was found in the CRCs of Ca²⁺-uptake, where we observed ∼20% activation at V_max Ca²⁺ and a much smaller effect at intermediate Ca²⁺. At the V_max Ca²⁺, the uptake CRC had a lower EC₅₀ than the ATPase CRC, which placed the uptake curve above the ATPase curve in the range of 0.3–3 μM compound 1. This suggests an enhanced coupling ratio (transported Ca²⁺ per hydrolyzed ATP) at V_max Ca²⁺. This was not observed for the intermediate-Ca²⁺ profiles, where the enhancement of Ca²⁺ uptake was too small to bring the curve above the ATPase CRC. This indicates a decreased coupling ratio at the intermediate Ca²⁺. For all CRCs, we observed biphasic profiles, with the rollover occurring between 1 and 10 μM of [compound 1]. This behavior (hormesis) is frequently encountered in biology and especially in pharmacology (34).

To determine whether compound 1 directly interacts with the SERCA-PLB complex, we measured its effect on the fluorescence lifetime of the concatenated SERCA-PLB FRET biosensor. We observed a decrease in FLT (increase in FRET), with EC₅₀ ≈ 10 μM (Fig. 1c). The effect of compound 1 on the FLT of the donor-only control was much smaller than on the donor-acceptor construct, thus suggesting that compound 1 directly binds to the targeted complex where it causes a structural change that shortens the distance between the FRET probes.

Effect of compound 1 on ER Ca²⁺ load and uptake in HEK293 cells expressing human SERCA2a

A recently developed assay (31) was used to determine the effect of compound 1 on ER Ca²⁺ dynamics in a human SERCA2a stable cell line. Changes in [Ca²⁺]_ER were recorded with the ER-targeted Ca²⁺ sensor R-CEPIA1er (32). To manipulate [Ca²⁺]_ER, cells were transfected with the Ca²⁺ release channel ryanodine receptor (RyR2). Experiments were conducted in permeabilized cells to control the cytosolic environment and to exclude sarcolemmal Ca²⁺ transporters from ER Ca²⁺ regulation. The R-CEPIA1er fluorescence signal was calibrated with ionomycin to convert the fluorescent signal to [Ca²⁺]_ER. Due to differences in RyR2 expression level and therefore in ER Ca²⁺ leak, resting [Ca²⁺]_ER (or ER Ca²⁺ load) varied significantly among the studied cells, allowing us to observe the effects of compound 1 on cells with a range of ER Ca²⁺ loads. Application of compound 1 (10 μM) increased [Ca²⁺]_ER particularly in cells with depleted ER Ca²⁺ load (black trace, Fig. 2 a). Compound 1 increased [Ca²⁺]_ER by 194% ± 25% (n = 18) in cells with depleted ER Ca²⁺ load (0.17 ± 0.04 mM; Fig. 2 b). In cells with high ER Ca²⁺ load, compound 1 was less effective to increase [Ca²⁺]_ER (gray trace, Fig. 2 a). At high ER Ca²⁺ load (0.45 ± 0.04 mM; Fig. 2 b), compound 1 increased [Ca²⁺]_ER by 55% ± 7% (n = 17).

Figure 2.

Effect of compound 1 on ER Ca²⁺ load and ER Ca²⁺ uptake in HEK293 cells expressing human SERCA2a. (A) Effect of compound 1 (10 μM) on [Ca²⁺]_ER in two HEK293 cells with different ER Ca²⁺ load. Compound 1 increased [Ca²⁺]_ER in cells with depleted ER Ca²⁺ load (black trace). However, compound 1 did not change [Ca²⁺]_ER in cells with high ER Ca²⁺ load (gray trace). Caffeine (Caff) application was used to completely deplete [Ca²⁺]_ER. (B) A relative increase in ER Ca²⁺ load induced by compound 1 in cells with initially depleted ER Ca²⁺ load (n = 18 cells) and cells with high ER Ca²⁺ load (n = 17 cells). ^∗p < 0.05 versus high load. (C) Effect of compound 1 (10 μM) on SERCA2a-mediated Ca²⁺ uptake. ER Ca²⁺ uptake was analyzed as the first derivative (d[Ca²⁺]_ER/dt) and plotted as a function of ER Ca²⁺ load (control 16 cells and DS 17 cells). ^∗p < 0.05 versus control.

Next, we studied the effect of compound 1 on SERCA2a function throughout a wide range of [Ca²⁺]_ER: from completely depleted to maximal ER Ca²⁺ loads. SERCA2a-mediated Ca²⁺ transport was analyzed from the rate of [Ca²⁺]_ER reuptake after full Ca²⁺ depletion induced by RyR2 activation with caffeine followed by RyR2 blockage with ruthenium red and tetracaine, as described before (31). To quantify the rate of SERCA Ca²⁺ transport, the first derivative of ER Ca²⁺ uptake (d[Ca²⁺]_ER/dt) was plotted against the corresponding [Ca²⁺]_ER. Compound 1 (10 μM) significantly increased ER Ca²⁺ uptake rate, particularly at low [Ca²⁺]_ER (between 0.05 and 0.26 mM; Fig. 2 c). These results illustrate that compound 1 can increase ER Ca²⁺ load by enhancing SERCA2a-mediated Ca²⁺ uptake.

Effect of compound 1 on Ca²⁺ signaling in ventricular myocytes

To assess the effect of compound 1 on cardiac Ca²⁺ regulation, action potential (AP)- and caffeine-induced Ca²⁺ transients were measured in mouse ventricular myocytes (Fig. 3 a). To evoke AP-induced Ca²⁺ transients, myocytes were electrically stimulated at 0.5 Hz. The amplitude of Ca²⁺ transient during caffeine (5 mM) application was used as an index of SR Ca²⁺ load. Compound 1 (10 μM) increased AP-induced Ca²⁺ transients to 175% ± 12% (n = 12 cells; Fig. 3b) and SR Ca²⁺ load to 153% ± 13% (n = 10 cells; Fig. 3 c). Moreover, compound 1 increased the rate of AP-induced Ca²⁺ transient decay by 14% ± 2% (n = 12 cells; Fig. 3 d). The decay rate was measured as the time constant of exponential decay, which changed from 308 ± 17 to 267 ± 19 ms (n = 12). These effects developed within 5 min and were reversible after washout. However, compound 1 did not change the rate of [Ca²⁺]_i decay during the caffeine application (Fig. 3 e). As this decay is mainly mediated by NCX-dependent Ca²⁺ extrusion, these results suggest that 1 does not affect NCX function. The first AP-induced Ca²⁺ transient after complete depletion of [Ca²⁺]_SR with caffeine is mainly mediated by Ca²⁺ current via LTCC (35). Analysis of LTCC-mediated Ca²⁺ transients suggests that compound 1 does not affect LTCC activity (Fig. 3 f). Therefore, the increase of AP-induced Ca²⁺ transients and SR Ca²⁺ load during the compound 1 application is likely to be mediated by enhanced SERCA2a-mediated Ca²⁺ uptake.

Figure 3.

Effect of compound 1 on Ca²⁺ signaling in ventricular myocytes. (A) Confocal line scan images of [Ca²⁺]_i with corresponding F/F₀ profiles of AP-induced (marked by triangles) and caffeine-induced Ca²⁺ transients (or SR Ca²⁺ load) in control conditions and during compound 1 (10 μM) application. Effect of compound 1 (Cpd 1; n = 12 myocytes) on AP-induced Ca²⁺ transient amplitude (B), SR Ca²⁺ load (C), AP-induced Ca²⁺ transient decay (D), NCX-mediated [Ca²⁺]_i decay rate (E), and LTCC-mediated Ca²⁺ transient amplitude (F). ^∗p < 0.05 versus control.

Effect of compound 1 on Ca²⁺ regulation under adrenergic activation and increased SR Ca²⁺ leak

Beta adrenergic receptor activation with isoproterenol (ISO; 0.1 μM) increased the AP-induced Ca²⁺ transient amplitude to 197% ± 16% of control (n = 22 cells; Fig. 4 a and b). This effect is largely mediated by an increase in SR Ca²⁺ uptake and sarcolemmal Ca²⁺ current due to PLB and LTCC phosphorylation by PKA (36). As expected, ISO increased both SR Ca²⁺ load to 139% ± 8% of control (n = 22 cells; Fig. 4 c) and LTCC-mediated Ca²⁺ transients to 220% ± 45% of control (n = 22 cells; Fig. 4 A). We then tested whether compound 1 could further increase SR Ca²⁺ load during adrenergic receptor stimulation. This analysis revealed that compound 1 (10 μM) had a minimal additional effect on the AP-induced Ca²⁺ transient amplitude (Fig. 4 b) and SR Ca²⁺ load (Fig. 4 c) already increased by ISO.

Figure 4.

Effect of compound 1 on Ca²⁺ signaling during adrenergic receptor activation in ventricular myocytes. (A) F/F₀ profiles of changes in [Ca²⁺]_i during AP- and caffeine-induced Ca²⁺ transients in control conditions, during adrenergic receptor activation with ISO (0.1 μM) and during the consecutive application of compound 1 (10 μM). Effect of ISO and compound 1 in the presence of ISO on AP-induced Ca²⁺ transient amplitude (B) and SR Ca²⁺ load (C). A total of 22 myocytes were studied in these experiments. Among the studied cells, five myocytes exhibited Ca²⁺ waves during ISO application (Fig. 5).

In ventricular myocytes, SR Ca²⁺ overload during adrenergic receptor activation is frequently associated with spontaneous Ca²⁺ waves (Fig. 5 a) (37,38,39). We found that 5 myocytes out of 22 studied (or ∼22%) exhibited Ca²⁺ waves during ISO application. Vehicle time-control ISO experiments showed that these Ca²⁺ waves appeared 5 min after ISO application and remained consistent during adrenergic stimulation. Application of compound 1 did not increase the fraction of myocytes with Ca²⁺ waves during adrenergic stimulation (Fig. 5 b). Importantly, in cells that produced Ca²⁺ waves during ISO application, compound 1 significantly reduced Ca²⁺ wave frequency (Fig. 5 c). This effect was associated with a decrease of the Ca²⁺ wave propagation velocity from 253.6 ± 38.7 μm/s (ISO alone) to 155.3 ± 27.3 μm/s (ISO and compound 1; n = 5 cells).

Figure 5.

Effect of compound 1 on Ca²⁺ waves during adrenergic receptor activation in ventricular myocytes. (A) Confocal line scan images of [Ca²⁺]_i with corresponding F/F₀ profiles of AP-induced (marked by triangles) and spontaneous Ca²⁺ waves during ISO (0.1 μM) and during the consecutive application of compound 1 (10 μM). (B) Effect of compound 1 (Cpd 1) on the Ca²⁺ wave occurrence during adrenergic receptor activation with ISO. (C) Effect of compound 1 (Cpd 1) on the Ca²⁺ wave frequency during adrenergic receptor activation. ^∗p < 0.05 versus control.

It has been shown that several heart pathologies, including heart failure and myocardial infarction, are associated with increased RyR2-mediated Ca²⁺ leak (40,41,42,43). Thus, we analyzed an effect of Compound 1 on Ca²⁺ regulation under the conditions of increased SR Ca²⁺ leak. In these experiments, RyR2-mediated Ca²⁺ leak was selectively increased by treating cells with a low dose of the RyR agonist caffeine. Application of 250 μM caffeine to myocytes paced at 0.5 Hz produced an immediate increase in Ca²⁺ transient amplitude, with afterward a decline of SR Ca²⁺ release to a new steady-state level. Under these conditions, compound 1 (10 μM) was still effective to increase Ca²⁺ transient amplitude by 38.2% ± 5.7% and SR Ca²⁺ load by 8.6 ± 2.3 (n = 8 cells; Fig. 6).

Figure 6.

Effect of compound 1 on Ca²⁺ signaling under conditions of increased SR Ca²⁺ leak. (A) F/F₀ profiles of changes in [Ca²⁺]_i during AP- and caffeine-induced Ca²⁺ transients in control conditions, during RyR2 activation with caffeine (250 μM) and during the consecutive application of compound 1 (10 μM). Effect of caffeine and compound 1 in the presence of caffeine on AP-induced Ca²⁺ transient amplitude (B) and SR Ca²⁺ load (C).

Effect of compound 1 on RyR2 function

The observation that compound 1 inhibited myocyte Ca²⁺ waves led to the hypothesis that effects on RyR2-mediated Ca²⁺ leak might contribute to its mechanism of action. To test this, we first used an established index of RyR channel open state, i.e., [³H]ryanodine binding to SR vesicles. We performed this assay using isolated cardiac SR (as described in materials and methods), and we observed that compound 1 is slightly inhibitory in 0.1 μM Ca²⁺ and slightly activatory in 30 μM Ca²⁺ (Fig. 7 a). These results suggest that RyR2 leak might also be inhibited in cells.

Figure 7.

Effect of compound 1 on measures of RyR2 function. (A) Concentration response curves of [³H]ryanodine binding to RyR2. Using SR isolated from porcine heart in the presence of compound 1 concentrations ranging 0–100 μM, we measured [³H]ryanodine binding to RyR2 in 0.1 μM Ca²⁺ (solid symbol) or 30 μM Ca²⁺ (open symbol). Data are shown normalized relative to the values sans-drug DMSO control (dotted line), means ± SD, n = 4–6. ^∗p < 0.05, ^∗∗p < .01, ^∗∗∗p < .001 for samples versus DMSO control using Student’s unpaired t-test. (B) Ca²⁺ sparks in permeabilized ventricular myocytes. Confocal line scan images of [Ca²⁺]_i with corresponding F/F₀ profiles of Ca²⁺ sparks in control conditions and during compound 1 (10 μM) application. F/F₀ profiles were obtained by averaging the fluorescence over the 1-μm-wide regions indicated by the black arrows. (C) Effect of compound 1 on Ca²⁺ spark frequency. 14 myocytes were studied in these experiments. ^∗p < 0.05 versus control.

Therefore, we determined whether compound 1 affects a cellular measure of RyR2-mediated Ca²⁺ leak, by studying its effect on Ca²⁺ sparks in permeabilized ventricular myocytes. After sarcolemma permeabilization, Ca²⁺ sparks were observed at a stable frequency of 11.5 ± 3.1 sparks×s⁻¹×(100 μm)⁻¹. Fig. 7 b shows Fluo-4 line scan images and corresponding F/F₀ profiles of Ca²⁺ sparks recorded in control conditions and in the presence of 1 (10 μM). Compound 1 produced a small decrease in Ca²⁺ spark frequency to 90% ± 5% (n = 14 cells; Fig. 7 c). This inhibitory effect on Ca²⁺ sparks developed within 3 min, remained stable during compound 1 application, and was reversible after washout.

Discussion

SERCA is an essential component of Ca²⁺ signaling in virtually all eukaryotic cells. Consequently, defects in SERCA function can cause significant health problems. For example, mutations in the skeletal muscle SERCA1a isoform cause Brody myopathy, a disease characterized by impaired muscle relaxation (44). Alterations in the cardiac SERCA2a regulation contribute to heart pathologies, such as hypertrophy and heart failure (45,46,47,48). Mutations of the nonmuscle SERCA2b lead to Darier’s disease, an autosomal dominant inherited disorder of skin cells (49). It is not surprising that the Ca²⁺ pump has attracted attention as an important target for therapeutic approaches for numerous diseases (8,50).

The small molecule evaluated here, compound 1 (Fig. 1 a), has been identified as an enhancer of both the Ca²⁺-ATPase and Ca²⁺-transport activity of SERCA at physiologically relevant Ca²⁺ concentrations (Fig. 1 b). For these reasons, compound 1 has the potential to correct myocyte dysfunction, when that is attributable to diminished SR Ca²⁺ uptake capacity, as found in various forms of heart failure (48,51,52,53). Examination of compound 1 also reveals favorable physicochemical properties and Ca²⁺-transport activities. In general, very polar compounds (calculated partition coefficient, clogP <1 and total polar surface area, tPSA >130 Å²) are likely to have poor permeability that restricts their oral bioavailability and tissue distribution, whereas highly lipophilic molecules (clogP >5) are typically poorly soluble, have reduced receptor selectivity, and are rapidly metabolized. Compound 1 has desirable physicochemical properties (MW = 317.4, clogP = 4.5, tPSA = 30 Å²) and does not possess any structural alerts for toxicity or promiscuity (e.g., reactive quinones or metal chelating motifs). For these reasons, compound 1 is very attractive for further development. Exploration of structure-activity relationships (SARs) of compound 1 by modification of the thiophene and methoxybenzyl moieties could further improve the potency and physicochemical properties. This will be the focus of a future study.

To evaluate effects of compound 1 on SERCA2a function in the cellular environment, we used a newly developed approach that directly measures Ca²⁺ dynamics within the ER lumen (31). Pump function was studied in a human SERCA2a HEK stable cell line transfected with the human Ca²⁺ release channel, RyR2. Transient RyR2 DNA transfection led to a variability in the RyR2 expression level among cells, which we exploited in this study. Since resting ER Ca²⁺ load is set by a balance between ER Ca²⁺ uptake and ER Ca²⁺ leak, this caused significant variability in ER Ca²⁺ load between cells. Compound 1 increased ER Ca²⁺ load only in cells with initially depleted Ca²⁺ load (Fig. 2 a and b). In cells with ER Ca²⁺ load approaching the maximal level, compound 1 was less effective to increase [Ca²⁺]_ER. The maximum ER Ca²⁺ load is determined by the pump’s catalytic efficiency (i.e., ATPase-Ca²⁺ transport coupling) and the pump’s thermodynamic limit set by the ATP/ADP ratio. Our results suggest that compound 1 increases ER Ca²⁺ uptake primarily by enhancing the SERCA2a Ca²⁺ transport kinetics, but not its catalytic efficiency. As a result, a more active pump in the presence of compound 1 can more efficiently compensate ER Ca²⁺ leak. The direct measurements of SERCA2a-mediated Ca²⁺ transport further confirmed that compound 1 stimulates ER Ca²⁺ uptake, particularly at low [Ca²⁺]_ER (Fig. 2 c).

Analysis of intracellular Ca²⁺ dynamics in ventricular myocytes revealed that compound 1 increases the Ca²⁺ transients induced by AP. This effect was associated with the increased SR Ca²⁺ load and the rate of AP-induced Ca²⁺ transient decay, indicating that compound 1 can stimulate SERCA Ca²⁺ transport in cardiomyocytes (Fig 3). Further analysis also revealed that small molecule 1 has an insignificant effect on Ca²⁺ extrusion by NCX and Ca²⁺ entry via LTCC. Since myocardial contraction is directly proportional to cytosolic [Ca²⁺], this effect of compound 1 on SERCA2a function should cause a positive inotropic effect, which should increase cardiac output without changes in the heart rate. This effect on cardiac Ca²⁺ regulation would be particularly beneficial during pathological heart conditions associated with reduced SERCA2a Ca²⁺ uptake, as found in heart failure with reduced ejection fraction. Importantly, compound 1 did not further increase AP-induced Ca²⁺ transients and SR Ca²⁺ load during adrenergic receptor activation (Fig. 4). Results in Fig. 2 c suggest that this lack of additional stimulatory effect on Ca²⁺ release is due to the SR Ca²⁺ load approaching its maximal level (i.e., thermodynamic limit) during adrenergic receptor stimulation and PKA activation. This was similar to the Ca²⁺ load dependence of compound 1 action observed in the HEK cells experiments (Fig. 2). In cardiomyocytes, SR Ca²⁺ overload is frequently associated with spontaneous SR Ca²⁺ release in a form of propagating Ca²⁺ waves (37,38,39). In myocytes with diastolic Ca²⁺ waves, compound 1 reduced Ca²⁺ wave frequency. It appears that in these myocytes, SR Ca²⁺ load could not reach maximal level, due to elevated diastolic SR Ca²⁺ release (during spontaneous Ca²⁺ waves) that offsets SERCA Ca²⁺ transport (37). This suggests that in cells with arrhythmogenic behavior, compound 1 was effective to stimulate SERCA Ca²⁺ uptake, which prevented wave Ca²⁺ formation by breaking the positive feed-forward of increased spontaneous Ca²⁺-induced Ca²⁺ release.

In cardiomyocytes, an increase in SR Ca²⁺ load is typically accompanied by increased Ca²⁺ spark activity (54). Here, we found that compound 1 has the opposite effect, slightly inhibiting Ca²⁺ sparks (Fig. 7 b), and this is consistent with the small inhibition of [³H]-ryanodine binding to isolated cardiac SR in diastolic (nM) Ca²⁺ (Fig. 7 a). Moreover, compound 1 causes a small activation of [³H]-ryanodine binding in systolic (μM) Ca²⁺ (Fig 7 a), which should contribute to the higher amplitude of AP-induced Ca²⁺-transients. This represents an ideal profile for a therapeutic agent targeting RyR2 (i.e., inhibition of Ca²⁺ leak and unchanged or higher AP-triggered Ca²⁺ release). By serendipity, the compound 1 mechanism of action seems to involve synergistic effects on both SERCA2a and RyR2.

In conclusion, the novel compound 1 shows properties consistent with a potential to improve intracellular Ca²⁺ dynamics in cardiomyocytes with reduced SERCA2a Ca²⁺ uptake, as found in failing hearts. Future studies will focus on evaluating the ability of compound 1 to improve Ca²⁺ cycling in cardiomyocytes from animal models of heart failure or derived from iPSCs procured from patients with relevant cardiomyopathies. SAR studies will follow to improve potency and efficacy. The biochemical assays of SERCA and RyR2 function featured here (Figs. 1 and 7, respectively) have the throughput necessary to drive the SAR work (where we expect that hundreds of compounds will be tested). The downstream cellular assays (Figs. 2, 3, 4, 5, and 6) will be essential to establish compound safety and efficacy in the targeted cardiomyopathies.

Author contributions

C.C.A., R.L.C., D.D.T., and A.V.Z. conceived and supervised the study. R.N., E.B., S.L.Y., L.M.T., K.B., R.L.C, D.D.T., and A.V.Z designed and performed experiments. R.N., E.B., S.L.Y., L.M.T., and K.B. performed analysis. R.N., E.B., C.C.A., R.L.C., D.D.T., and A.V.Z. wrote the manuscript. All the authors read and approved the manuscript version to be published.

Editor: Eleonora Grandi