Introduction
Heart failure (HF) has become a global public health burden which affects people worldwide, characterized by a growing incidence of hospitalization and mortality rate. HF with prevailing diastolic dysfunction (HFpEF) is a more recently described entity, for which a specific treatment is still missing. Moreover, dealing with HF in general there is still a critical need of drugs that may improve patient outcomes, without untoward effects [
1,
2]. The sarcoplasmic reticulum (SR) Ca
2+ ATPase (SERCA2a), whose function is usually depressed in HF, is becoming an interesting therapeutic target for HF treatment [
3].
Istaroxime is an innovative and unique ino-lusitropic drug that combines the ability to inhibit the Na
+/K
+ ATPase and stimulate SERCA2a activity, resulting in improvement of the heart function in healthy and failing animal models and in patients with acute HF (Phase IIb clinical trials) [
4‐
8]. Although endowed with an excellent pharmacodynamic profile, pharmacokinetic studies have indicated that istaroxime has a short plasma half-life (less than 1 h) [
6], due to its extensive metabolization to a long-lasting metabolite, PST3093. The latter has a longer half-life (about 9 h), retains the ability to stimulate SERCA2a and it does not inhibit Na
+/K
+ ATPase [
9]. As a metabolite, PST3093 may in fact contribute to the beneficial effects of istaroxime acutely administered to patients [
10]. However, the presence of the potentially genotoxic oxime moiety may limit the chronic usage of both istaroxime and PST3093. This led us to pursue rational design of novel SERCA2a activators based on PST3093 structure, but devoid of the oxime moiety, that would be thus suitable for chronic (oral) treatment of HF [
11].
Among the developed PST3093 derivates, compound 8 was one of the two selected compounds showing the ability to recover streptozotocin (STZ)-induced SERCA2a activity depression in a phospholamban (PLN)-dependent manner, analogously to its parent compound PST3093 [
11]. We present here the pharmacological evaluation of compound 8 that proves to be a safe and selective SERCA2a stimulator and a favorable drug candidate for chronic HF therapy.
Materials and methods
The animal study protocols were approved by the Institutional Review Board of Milano Bicocca (29C09.26 and 29C09.N.YRR protocol codes approved in January 2021 and June 2018 respectively) and Chang Gung (CGU107-068 protocol code approved on June 2018) Universities in accordance with the Guide for the Care and Use of Laboratory Animals as adopted and promulgated by the U.S. National Institutes of Health.
Methods are briefly described here; details are given in the Additional section (Additional file
1).
Disease model
Streptozotocin (STZ)-induced diabetes was selected as a pathological model showing diastolic dysfunction associated to reduced SERCA2a function [
12,
13]. Diabetes was induced in Sprague Dawley male rats (150–175 g) by a single i.v. STZ (STZ group, 50 mg/kg in citrate buffer) injection in the tail vein. Control rats (healthy group) received vehicle (citrate buffer). Fasting glycaemia was measured after 1 week and rats with values > 290 mg/dL were considered diabetic [
12]. Rats were euthanized by cervical dislocation under anesthesia with ketamine-xylazine (130–7.5 mg kg-1 i.p) 9 weeks after STZ injection.
Measurements in isolated cardiomyocytes
To ensure stabilization of drug effect, isolated myocytes were analyzed after incubation with compound 8 or vehicle (control) for at least 30 min. The experiments were performed at 35 °C.
The Na
+/K
+ ATPase current (I
NaK) was recorded (V-clamp) in normal rat left ventricular (LV) myocytes as ouabain (1 mM)-sensitive current at − 40 mV, under conditions enhancing I
NaK and minimizing contamination by other conductances [
12,
14]. I
NaK inhibition by compound 8 was expressed as percent reduction of ouabain-sensitive current.
Intracellular Ca2+ dynamics were evaluated in Fluo4AM-loaded STZ cardiomyocytes, superfused with Tyrode’s solution. Cytosolic Ca2+ was expressed in arbitrary units, i.e. as the ratio between the fluorescence signal and its value during diastole (F/F0). Ca2+ uptake by the sarcoplasmic reticulum (SR) (proportional to SERCA2a function) was investigated with two protocols addressing SR function under different conditions: post-rest potentiation (PRP) of Ca2+ transients (CaT) and “SR Ca2+ reloading” after depletion.
The PRP protocol was applied to field-stimulated myocytes as previously shown [
12]. Briefly, voltage-induced Ca
T were evoked at 2 Hz until steady state Ca
T (ssCa
T) amplitude was achieved. Stimulation was then interrupted for intervals of increasing duration (1-5-10-20 s) and then resumed. Multiple parameters were extracted from the PRP protocol. ssCa
T were evaluated in terms of amplitude and decay kinetics (decay t
1/2). PRP was calculated as the ratio between the first post-rest Ca
T (prCa
T) and the last pre-rest Ca
T (ssCa
T). The prCa
T/ssCa
T ratio provides information on the fate of intracellular Ca
2+ during the rest interval, largely dictated by the balance between Ca
2+ extrusion from the cell and Ca
2+ reuptake into the SR. Also, useful to evaluate this balance is the measurement of SR Ca
2+ content (Ca
SR) at various times during the protocol pause. Ca
SR was estimated from the amplitude of caffeine (10 mM)-induced Ca
T, electronically evoked at 0.5 s (CaSR
0.5 s) and 20 s (CaSR
20s) following the last stimulated Ca
T. The ratio CaSR
20s/CaSR
0.5 s was calculated to evaluate post-rest SR Ca
2+ compartmentalization at rest.
The SR Ca
2+ reloading protocol (Additional file
1: Figure S1) was designed to examine SR function at multiple levels of Ca
2+ loading, while eliminating the contribution of the Na
+/Ca
2+ exchanger (NCX) to Ca
2+ clearance. To this end, Ca
T and membrane current (I
CaL) were simultaneously measured in V-clamped myocytes. After SR depletion by a caffeine pulse, Ca
T and I
CaL were recorded during a reloading pulse train and their features (amplitude, decay t
1/2) measured at each pulse. The excitation–release (ER) “gain” was calculated as the ratio between Ca
T amplitude and Ca
2+ influx through I
CaL up to Ca
T peak [
5]. During the whole protocol NCX was blocked by Na
+-free superfusion and Na
+-free pipette solution.
Off-target actions
To asses potential off-target effects on ion channels, the effect of compound 8 was evaluated on action potentials (APs), recorded by patch-clamp (I-clamp) from normal guinea-pig LV myocytes. This cell type was selected because of similarity of its AP repolarization to the human one [
15]. AP duration at 50% and 90% repolarization (APD
50 and APD
90) and diastolic potential (E
diast) were measured 1) during steady state pacing at several rates, 2) dynamically upon stepping between two rates (to assess APD
90 adaptation kinetic). During steady state pacing, short-term APD
90 variability (STV) was calculated from 20–30 subsequent APD
90 values according to Eq.
1 [
16]:
$$STV=\sum (|APD(n+1) - APDn|)/[nbeats*\surd 2]$$
(1)
The kinetics of APD90 adaptation was quantified by estimating the time constant (τ) of the exponential time course of APD90 after stepping between two pacing rates.
To further detect potential off-target actions of compound 8, its interaction with a panel of 50 ligands, potentially relevant to off-target effects, was carried out by Eurofins (Taiwan) on crude membrane preparations according to Eurofins procedures. The assays were partly based on radioligand displacement (e.g., for receptors) and partly on spectrophotometric detection of change in function (e.g., for enzymes). Results were compared to appropriate reference standards; a > 50% change in affinity or activity was considered as a positive hit (interaction present). Compound 8 was tested at the concentration of 10 μM.
In vivo hemodynamic effects in diseased (STZ) rats
In vivo effects of compound 8 were evaluated by echocardiography in STZ rats under urethane (1.25 g/kg, i.p, acute protocol) or ketamine/pentobarbital (60–37.5 mg/kg, i.p., chronic protocol) anesthesia. Studies were carried out during acute (i.v. infusion) and following chronic (oral gavage) treatment.
In the acute protocol, compound 8 or saline (control group) were i.v. infused at 0.2 mg/kg/min (0.16 ml/min); echocardiographic parameters were measured within the same animal before (basal), and at 15 and 30 min of infusion.
In the chronic protocol, compound 8 effects were evaluated following 1 or 4 oral daily administrations (by gavage) at doses of 40 and 80 mg/kg (5 ml/kg body weight) dissolved in saline (Additional file
1: Figure S2). The treatment group was compared to a randomly assigned control group receiving vehicle only. At day 1 all the animals received saline and underwent basal echocardiography. From day 5 to day 8, each group was treated once daily with saline or compound 8 (40 mg/kg or 80 mg/kg); all animals were subjected to echocardiography at day 5 (after 1 dose) and day 8 (after 4 doses). Echo measurements were performed 60 min following gavage.
The following echo indexes were measured, according to the American Society of Echocardiography guidelines [
17]: left-ventricular (LV) end-diastolic (LVEDD) and end-systolic (LVESD) diameter, posterior wall thickness (PWT) and interventricular septal thickness (IVST). The Teichholz formula (7/ (2.4 + D) × D
3, D = linear LV diameter) was used to calculate LV end-diastolic volume (EDV) and end-systolic volume (ESV). Stroke volume (SV) was calculated as the difference between EDV and ESV. LV ejection fraction (EF) was calculated as SV/EDV and expressed in % [
18]. Fractional shortening was calculated as FS = (LVEDD-LVESD)*LVEDD
−1 and expressed in %. Trans-mitral flow velocity was measured (by pulsed Doppler) to obtain early and late filling velocities (E, A waves) and E wave deceleration time (DT). DT was also normalized to E wave amplitude (DT/E ratio). Peak myocardial systolic (s’) and diastolic velocities (e’ and a’) were measured at the mitral annulus by Tissue Doppler Imaging (TDI).
Drug toxicity in mice
Acute toxicity was determined in the mouse (Albino Swiss CD-1, body weight 30 g). Mice were orally treated, or i.v. injected, with increasing doses of compound 8 to identify the one causing 50% mortality (LD50, mg/kg body weight) within 24 h. Compound 8 was dissolved in saline solution and i.v. injected at 50, 100, 200, 300 mg/kg (2–4 animals for each group) or orally administered by gavage at 200 and 700 mg/kg (4 animals for each group). Control animals received vehicle only.
Statistical analysis
Data are reported as mean ± SEM. Individual means were compared by Student’s t-test; multiple means were compared by one-way or two-way ANOVA for repeated measurements (RM), followed by post-hoc Tukey’s multiple comparisons. P < 0.05 was considered as statistically significant in all comparisons. N (number of animals) and n (number of cells) are reported in each figure legend.
Discussion
The present study summarizes the preclinical data concerning compound 8, a derivative of PST3093, the long-lasting istaroxime metabolite. Compound 8 is devoid of the oxime moiety and was selected in an in vitro screening based on optimization of selectivity for SERCA2a activation vs inhibition of the Na
+/K
+ ATPase [
11]. The emerging profile of compound 8 indicates that it has a low acute toxicity, it is active in isolated myocytes and in reversing STZ-induced diastolic dysfunction in vivo. Compound 8 effects were qualitatively similar after i.v. and oral administration; incremental effect during repeated once-a-day dosing suggests pharmacokinetics suitable for chronic usage.
Previous results on molecular function in cell free systems [
11] indicate that compound 8 enhances SERCA2a enzymatic activity selectively, i.e. without appreciably affecting Na
+/K
+ pump activity. The present studies at cellular and in vivo levels, confirm this view and collectively point to compound 8 ability to improve Ca
2+ confinement within the SR. This extends the conclusions of previous molecular studies to the fully integrated biological system.
The SERCA2a stimulatory activity of istaroxime and PST3093, from which compound 8 was derived, depends on the presence of PLN [
9,
20]; thus, suggesting that these compounds enhance SERCA2a activity by relieving its inhibition by PLN. The same likely applies to compound 8 [
11], which has a closely similar chemical structure. Hence, these compounds can be collectively defined as “PLN antagonists”, a novel class of drug action.
A diseased heart model (the STZ diabetic rat) was chosen for in vivo studies; this is justified by previous observations with the parent compounds. PLN antagonism by istaroxime, the prototype PLN antagonist, was firstly detected in healthy guinea-pig myocytes [
5] and reproduced in murine ones [
21]. However, istaroxime effect was substantially enhanced in failing guinea-pig preparations [
22]. Rat myocytes, best suited for in vivo studies, are relatively insensitive to PLN antagonism when healthy, to become responsive when SERCA2a activity is depressed by disease. This is the case for STZ-induced diabetes, in which consistent with primarily diastolic dysfunction in clinical diabetes, myocytes are characterized by SERCA2a down-regulation [
9,
12]. These considerations, as well as the added translational value provided by relevance to human pathology, led us to adopt the STZ (diabetic) rat as experimental model. Why the effect of PLN antagonism becomes more apparent whenever baseline SERCA2a function is diminished is a matter of speculation, an interesting one, but of limited translational relevance for an agent meant to treat diseased hearts.
Some among the present results were unexpected based on PLN antagonism and deserve to be separately discussed. While SERCA2a stimulation is expected to increase the rate of decay of V-triggered Ca
T, compound 8 failed to change Ca
T decay kinetic measured in field-stimulated myocytes (during electrical activity) (Fig.
2C), in which all Ca
2+ transports were intact. On the other hand, compound 8 sharply decreased
\(\tau_{decay}\) in the “reloading protocol” (Fig.
3) in V-clamped cells, in which SR uptake function largely depends on SERCA2a only. This apparent discrepancy might be attributed to the variability of the rate of Ca
T decay in field stimulated cells in comparison to the same parameter measured controlling membrane potential (V-clamp), or to NCX contribution to diastolic Ca
2+ clearance. Notably, the same was true for PST3093, which is nonetheless endowed with clear-cut lusitropic effect in vivo [
9].
Acute i.v. infusion of compound 8 (Fig.
5) improved diastolic relaxation in STZ rats. Similar results on diastolic indexes were obtained following four oral daily doses at 40 mg/kg and a single dose at 80 mg/kg. This argues against significant changes in SERCA2a-modulating effect by liver metabolism.
Among the main goals of this study was the evaluation of chronic in vivo effects of orally administered compound 8. In vivo data published so far support the high therapeutic potential of istaroxime [
6‐
8,
10], its metabolite PST3093 [
9] and follow-on derivatives [
11]. Nonetheless, this is the first study evaluating chronic effects of the lead follow-on compound through its oral administration in a disease model. Accumulation of effect over repeated dosing every 24 h points to a relatively slow clearance of the compound. Accordingly, while compound 8 pharmacokinetics is still unknown, its chemical structure predicts a plasma half-life comparable to that of PST3093 (about 9 h in humans) [
9]. Accumulation of effects was seen with the higher dosage of compound 8 only in DT reduction, likely because of early achievement of saturating plasma levels.
Effects of compound 8 on systolic indexes were marginal; indeed, among systolic indexes, only s’ parameter significantly increased after 4 oral doses (Figs.
6–
7). HR was significantly affected by the compound after the lowest oral dose (40 mg/kg), but this effect was not clearly detected at the highest oral dose (80 mg/kg) and after i.v. infusion. Moreover, since effects on HR were absent when the compound was i.v. infused, further investigations are necessary to clarify this point. Nonetheless, a comprehensive characterization of the underlying mechanisms is beyond the scope of the current study.
A pure SERCA2a activator might exert substantial antiarrhythmic effects by inhibiting Ca
2+ waves [
23,
24], at least under the common conditions characterized by SR instability (e.g. HF). Further, focused studies are necessary to better characterize the potential antiarrhythmic effects of SERCA2a stimulators. Furthermore, SR Ca
2+ compartmentalization has potential long-term effects on energetic efficiency and biology of cardiac myocytes [
23]. Among them, higher SR Ca
2+ content might conceivably support Ca
2+ transfer to mitochondria. However, we have recently found that a PLN mutation resulting in SERCA2a enhancement is associated with depression of mitochondrial function instead [
25]. Even if it is difficult to rule out off-target effects of mutant PLN, this observation may discourage from predicting the relationship between SERCA2a function and mitochondrial respiration by simple reasoning. Nonetheless, evaluation of the effect of SERCA2a stimulators on mitochondrial function may be worthwhile.
In summary, the specific lusitropic effect of compound 8 in STZ rats, detected both during i.v. infusion and after oral administration, can be attributed to recovery of SERCA2a function. Compound 8 represents the first small-molecule SERCA2a activator that can be considered for oral administration as a chronic treatment of HF based on such an innovative mechanism of action.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.