S100A1ct: A Synthetic Peptide Derived From S100A1 Protein Improves Cardiac Performance and Survival in Preclinical Heart Failure Models

Abstract

Background: The EF-hand Ca²⁺ sensor protein S100A1 has been identified as a molecular regulator and enhancer of cardiac performance. The ability of S100A1 to recognize and modulate the activity of targets such as SERCA2a (sarcoplasmic reticulum Ca²⁺ ATPase) and RyR2 (ryanodine receptor 2) in cardiomyocytes has mostly been ascribed to its hydrophobic C-terminal α-helix (residues 75-94). We hypothesized that a synthetic peptide consisting of residues 75 through 94 of S100A1 and an N-terminal solubilization tag (S100A1ct) could mimic the performance-enhancing effects of S100A1 and may be suitable as a peptide therapeutic to improve the function of diseased hearts.

Methods: We applied an integrative translational research pipeline ranging from in silico computational molecular modeling and in vitro biochemical molecular assays as well as isolated rodent and human cardiomyocyte performance assessments to in vivo safety and efficacy studies in small and large animal cardiac disease models.

Results: We characterize S100A1ct as a cell-penetrating peptide with positive inotropic and antiarrhythmic properties in normal and failing myocardium in vitro and in vivo. This activity translates into improved contractile performance and survival in preclinical heart failure models with reduced ejection fraction after S100A1ct systemic administration. S100A1ct exerts a fast and sustained dose-dependent enhancement of cardiomyocyte Ca²⁺ cycling and prevents β-adrenergic receptor-triggered Ca²⁺ imbalances by targeting SERCA2a and RyR2 activity. In line with the S100A1ct-mediated enhancement of SERCA2a activity, modeling suggests an interaction of the peptide with the transmembrane segments of the sarcoplasmic Ca²⁺ pump. Incorporation of a cardiomyocyte-targeting peptide tag into S100A1ct (cor-S100A1ct) further enhanced its biological and therapeutic potency in vitro and in vivo.

Conclusions: S100A1ct is a promising lead for the development of novel peptide-based therapeutics against heart failure with reduced ejection fraction.

Keywords: heart failure; peptides; therapeutics.

Figure 1.

S100A1ct stimulates SERCA2a activity and sarcoplasmic reticulum Ca2+ uptake. A, Primary sequences of the parent human S100A1 protein (National Center for Biotechnology Information reference NP_006262; top) and the synthetic S100A1ct peptide with the synthesis tag (bottom). The amino acid residues 75 through 94 (helix IV) are highlighted in red. Dotted lines indicate the EF-hand Ca²⁺ binding sites of the S100A1 monomeric subunit. B, Left: 3-dimensional structure of Ca²⁺-bound S100A1 (holo, Protein Data Bank ID 2LP3, NMR model 1; Ca²⁺ ions are shown as cyan spheres, and the protein is shown in cartoon representation in gray, indicating the 2 subunits, with the exposed C-terminal helix IV (residues 75–94) shown in red. Right: a computationally predicted, representative 3-dimensional conformation of the synthetic S100A1ct peptide (red) with the synthesis aid highlighted in blue. C, Dose-dependent increase of SERCA2a (sarcoplasmic reticulum Ca2⁺ ATPase) activity by the S100A1ct peptide in cardiac sarcoplasmic reticulum (SR) vesicle preparations (n=3). D, Dose-dependent increase of SERCA2a-mediated SR Ca²⁺ uptake by the S100A1ct peptide in β-escin–permeabilized rabbit ventricular cardiomyocytes. Measurements were carried out at a free Ca²⁺ concentration of 1 μM (n=3). Vehicle: aqueous carrier used for scrambled and S100A1ct peptide dilution. Control: scrambled peptide. Data are given as mean±SEM. Statistical comparisons were performed by 1-way ANOVA with Sidak multiple comparison test (C and D). *P<0.05, ****P<0.0001.

Figure 2.

S100A1ct as a modeled SERCA2a binding partner is a cell-penetrating peptide. A, Putative binding mode of S100A1ct to SERCA2a (sarcoplasmic reticulum Ca2⁺ ATPase) obtained by the computational docking workflow (rank 2). SERCA2a (E1 state; gray; docking site transmembrane [TM] helices colored) and S100A1ct (hydrophilic synthesis aid tag, blue; apolar amino acid stretch VVLVAALTVA, yellow; remaining peptide, red) are shown in cartoon representation. The sarcoplasmic reticulum (SR) membrane is indicated by dashed lines. The black inset compares the docking mode with the experimentally determined SERCA1a binding mode of the transmembrane part of phospholamban (PLB; black cartoon, N-terminal end indicated by a black sphere). Other docking modes to SERCA2a are shown in Figure S5. B, Parallel artificial membrane permeability assay showing the ability of S100A1ct (1 μM) to cross the 4% lecithin artificial membrane in comparison with low and high permeability controls (n=3). C, Representative confocal microscopic images of isolated adult ventricular cardiomyocytes incubated with FITC-coupled S100A1ct (1 μM; left; transmission and fluorescent image) and fluorescent dye only for 20 minutes show intracellular accumulation of S100A1ct in cardiomyocytes. Black bar=50 μm. D, Quantification of FITC-S100A1ct uptake into adult mouse cardiomyocytes under different conditions (n=8). Data are given as mean±SEM. Statistical comparisons was performed by 1-way ANOVA with Sidak multiple comparison test (C) or Dunnett multiple comparison test (D). *P<0.05, ****P<0.0001.

Figure 3.

Biological effect of S100A1ct on normal cardiomyocyte performance. A, Dose-dependent increase of the Ca²⁺ transient amplitude in 2-Hz electrically stimulated and FURA2-AM–loaded normal rat adult ventricular cardiomyocytes (AVCMs) by S100A1ct (n=20). B, Preserved gain-in-function of the contraction amplitude in rat AVCMs by 1 μM S100A1ct in the presence of isoproterenol (0.1 μM) at 2 Hz stimulation (n=20). C, Increased sarcoplasmic reticulum (SR) Ca²⁺ load after 1 μM S100A1ct stimulation in rat AVCMs, as assessed by the caffeine-pulse protocol (n=20). D, Decreased rate of Ca²⁺ spark frequency in quiescent normal rat AVCMs in response to 0.1 μM S100A1ct (n=40). E, Diminished rate of diastolic aftercontractions by 1 μM S100A1ct in 2 Hz electrically stimulated rat AVCMs under basal conditions and in response to a proarrhythmogenic isoproterenol/caffeine treatment protocol (n=40). Numbers above each bar graph present portion of cardiomyocytes that responded with aftercontractions from total number of cardiomyocytes analyzed. F, Abrogated isoproterenol/caffeine-induced SR Ca²⁺ leak by 1 μM S100A1ct in quiescent rat AVCMs, as assessed by Ca²⁺ spark frequency rate (n=60). Vehicle: aqueous carrier used for scrambled and S100A1ct peptide dilution. Control: scrambled peptide. Data are given as mean±SEM. Statistical comparisons were performed by nested unpaired t test (D), nested 1-way ANOVA with Sidak multiple comparisons test (A through C) or Tukey multiple comparisons test (F), or χ² test/Fisher exact test (E). *P<0.05, **P<0.01, ***P<0.0005, ****P<0.0001.

Figure 4.

Therapeutic effect of S100A1ct in failing cardiomyocytes. A and B, Dose-dependent enhancement of the Ca²⁺ transient amplitude in failing rat (A) and human (B) adult ventricular cardiomyocytes (AVCMs) by S100A1ct stimulation. Because the control peptide had no effect at any concentration, further use was limited to a 1-μM control group in failing human left ventricular AVCMs. Rat n=20, human n=12. C and E, Effect of 1 μM S100A1ct on the sarcoplasmic reticulum (SR) Ca²⁺ load in failing rat (C) and human (E) AVCMs subjected to the caffeine-pulse protocol. Rat n=20, human n=12 cells. D and F, Effect of 1 μM S100A1ct on fractional shortening in failing rat (C) and human (H) AVCMs. Rat n=24, human n=18. G and H, Decreased rate of diastolic aftercontractions by 1 μM S100A1ct in failing rat (G) and human (H) AVCMs at 2 Hz subjected to a proarrhythmogenic isoproterenol/caffeine protocol. Rat n=40, human n=30. Numbers above each bar graph present portion of cardiomyocytes that responded with aftercontractions from total number of cardiomyocytes analyzed. Vehicle: aqueous carrier used for scrambled and S100A1ct peptide dilution. Control: scrambled peptide. Data are given as mean±SEM. Statistical comparisons was performed by nested unpaired t test (E and F), nested 1-way ANOVA with Sidak multiple comparisons test (A through D), or χ² test/Fisher exact test (G and H). *P<0.05,**P<0.01, ***P<0.0005, ****P<0.0001.

Figure 5.

Effect of cor-S100A1ct on normal and failing cardiomyocytes. A, Primary sequence of the cor-S100A1ct peptide (cardiomyocyte targeting peptide [CTP] coupled to the 20 C-terminal residues of S100A1ct) and the cor-scrambled peptide (CTP coupled to scrambled peptide). An N-terminal synthesis aid (KKKKKP) and a spacer (GG) were required for scalable synthesis and solubility of the peptides. B and C, Dose-dependent increase of the Ca²⁺ transient amplitude in normal (B) and failing (C) rat adult ventricular cardiomyocytes (AVCMs) by cor-S100A1ct treatment. Normal n=20, failing n=20. D and E, Effect of 1 μM cor-S100A1ct on the sarcoplasmic reticulum (SR) Ca²⁺ load in normal (D) and failing (E) rat AVCMs subjected to caffeine-pulse protocol. Normal n=20, failing n=15. F and G, Effect of 1 μM cor-S100A1ct on the fractional shortening in normal (F) and failing (G) rat AVCMs subjected to caffeine-pulse protocol. Normal n=24, failing n=24. Vehicle: aqueous carrier used for scrambled and S100A1ct peptide dilution. Control: cor-scrambled peptide. Data are given as mean±SEM. Statistical comparisons was performed by nested 1-way ANOVA with Sidak multiple comparisons test (B through G). *P<0.05, **P<0.01, ***P<0.0005, ****P<0.0001.

Figure 6.

Short-term effect of cor-S100A1ct on cardiac performance in vivo. A, Study design to determine the short-term effect of cor-S100A1ct on cardiac performance in male C57Bl/6 healthy mice. B and C, Effect of a single intravenous dosage range of cor-S100A1ct on LV+dp/dt_max (B) and heart rate (HR; C) 60 minutes after administration (n=7). D and E, Preserved gain in function in LV+dp/dt_max 60 minutes after a 3 μg/kg body weight cor-S100A1ct intravenous dose in response to a single intravenous dose of isoproterenol (ISO; 300 pg; D) or metoprolol (Meto; 90 μg; E; n=7). F and G, The rise and decline in HR by isoproterenol or metoprolol was not changed by cor-S100A1ct (n=7). Average body weight of enrolled healthy mice was 30.66±1.41 g in the control and 30.58±2.03 g in the cor-S100A1ct group. Vehicle: aqueous carrier used for scrambled and S100A1ct peptide dilution. Control: cor-scrambled peptide. Data are given as mean±SEM. Statistical comparison was performed by 1-way ANOVA with Sidak multiple comparisons test (B and C) or 2-way ANOVA with Tukey multiple comparisons test (D through G). **P<0.005, ****P<0.0001.

Figure 7.

Safety study of long-term cor-S100A1ct administration in healthy mice. A, Study design to assess the safety profile of cor-S100A1ct in healthy C57Bl/6 mice. Female (n=4) and male (n=4) mice were treated every 48 hrs IP with 6 μg/kg body weight (BW) cor-control or cor-S100A1ct for up to 4 weeks. B through D, left ventricular ejection fraction (LVEF; B), heart rate (HR; C), and left ventricular end-diastolic diameter (LVEDD; D) before (baseline [BL]) and 4 weeks after cor-S100A1ct treatment (n=8). E, high-sensitive troponin T (hsTnT) concentrations in plasma of cor-S100A1ct–treated mice after 2 weeks and 4 weeks (n=8). F, Unchanged brain natriuretic peptide expression in left ventricles after 4 weeks cor-S100A1ct stimulation (n=8). Control: cor-scrambled peptide. Data are given as mean±SEM. Statistical comparisons were performed by unpaired t test (F) or 2-way ANOVA with uncorrected Fisher least significant difference (B through E). ****P<0.0001.

Figure 8.

Therapeutic effects of cor-S100A1ct on cardiac performance in post–myocardial infarction models. A, Study design to assess the therapeutic effect of cor-S100A1ct on cardiac dysfunction in the post–myocardial infarction (MI) C57Bl/6 mouse model. Two days after MI, surviving animals were treated for 12 days with cor-S100A1ct (3 μg/kg per body weight [BW]; n=20; M/F 10/10) by daily IP injections and compared with cor-control (n=10; M/F 5/5) and vehicle (n=10; M/F 5/5) treatment. Average BW of enrolled post-MI mice was 30.44±1.01 g in the cor-S100A1ct, 30.12±1.82 g in the control, and 31.32±1.43 g in the vehicle group. B, Post-treatment effect of cor-S100A1ct on the decline in left ventricular ejection fraction (LVEF) in surviving animals (vehicle n=4, cor-control n=5, cor-S100A1ct n=14). C, Significantly fewer animals had died after the 12-day cor-S100A1ct treatment than in the combined control/vehicle group (control vs vehicle showed no significant difference and were pooled; n=20). D, Acute intravenous epinephrine stress (2 mg/kg BW) in the 12 days after treatment showed a clear trend toward fewer deaths in cor-S100A1ct–treated mice than controls (combined cor-control/vehicle group; cor-control n=9, cor-S100A1ct n=14). E, Gene expression analysis in left ventricles of surviving animals (n=3). F, Study design to study the short-term therapeutic effect of cor-S100A1ct on cardiac dysfunction in a post-MI pig model. Three weeks after MI, cardiac function was monitored for 3 hours in anesthetized and ventilated post-MI animals either after a catheter-based intracoronary single dosage administration of cor-S100A1ct or control (vehicle). Average BW was 40.86±1.82 kg in the controls (n=7) and 40±1.37 kg in the cor-S100A1ct (n=8) group. G through I, Effect of 150 μg/kg BW cor-S100A1ct and control treatment on left ventricular stroke volume (SV; G), heart rate (HR; H), and LVEF (I) over a course of 3 hours, shown as relative hourly changes (%) compared with baseline (BL) hemodynamics of each animal (control n=7, cor-S100A1ct n=8). J through L, Unchanged cardiac electrical conductivity and activity of porcine hearts after cor-S100A1ct administration compared with control, as assessed by PQ interval (J), QRS duration (K), and corrected QT interval (L; control n=7, cor-S100A1ct n=8). Vehicle: aqueous carrier used for cor-S100A1ct peptide dilution. Control: cor-scrambled peptide. Data are given as mean±SEM. Statistical comparisons was performed by 1-way ANOVA with Tukey multiple comparisons test (B and E), χ² test/Fisher exact test (C and D), unpaired t test (G through I and K), or Mann-Whitney test (J and L). *P<0.05, **P<0.005 ***P<0.003, ****P<0.0001.