Peptides on cellulose membranes were synthesized using a Multipep automated peptide synthesizer (INTAVIS Bioanalytical Instruments AG, Koeln, Germany) (Frank, 1992, Frank, 2002). Peptides in solution were synthesized (listed below) at >80% purity by Genscript (Corporation, Piscataway, NJ, United States). For some of the peptides a cell-permeable TAT (RKKRRQRRR) or a biotin tag was included at either N- or C-terminus with or without an ahx-linker in between to avoid any potential steric hindrance.

PLM_cyt: KRCRCKFNQQQRTGEPDEEEGTFRSSIRRLSSRRR

pSer68-PLM_cyt: KRCRCKFNQQQRTGEPDEEEGTFRSSIRRLpSSRRR

OPT: YHPYKEIEQLIELANYQVLS

Mini-OPT: PYKEIEQLIELANYQV

NOPT: NMP-YKEIEQLIELANYQV

Scrambled control: VYEKNYLLLPAIEQEQSHIY (negative control peptide)

OPT (E6G): YHPYK G IEQLIELANYQVLS (negative control peptide)

OPT (1Y4A): YHPYKEIEQLIELAN A QVLS (negative control peptide)

Anti-NCX1 blocking peptide (amino acids 655–672): CGQPVFRKVHARDHPIPST

Peptide array membranes were blocked in 1% casein for 2–4 h and overlaid with 5 μM of biotinylated pSer68-PLM peptide in 1% casein overnight at 4°C. The membranes were then washed three times for 10 min in TBS-T before incubation with anti-biotin-horseradish peroxidase (HRP) for 1 h (see Antibodies and Recombinant Protein). Peptide-bound spots were detected as described in the immunoblotting section.

Peptide stability was analyzed by incubating the peptide (300 μM) in 30% human serum (diluted in phosphate-buffered saline; PBS) (21792, Innovative research, MI, United States) at 37°C. Aliquots of 50 μl were collected at different time points (0–48 h) and thereafter trichloroacetic acid (TCA) was added to a final concentration of 5% to stop the reaction. The aliquots were kept at −20°C before centrifugation at 13,000 g for 15 min to remove the serum proteins. The supernatants were stored at −20°C until the mass spectrometry analysis.

The peptides were purified using μC18 tips (Pierce C18 Tips, 100 μl) according to manufacturer’s instructions. 142 fmol purified peptide sample was analyzed by a nEASY-LC coupled to QExactive/QExactive Plus with EASY Spray PepMap®RSLC, C18, 2 μl, 100 Å, 75 μm × 25 column using a 60 min separation gradient (0–30% acetonitrile). MS raw files were submitted to MaxQuant software version 1.5.3.8 (Cox and Mann, 2008) for protein and peptide identification and intensity-based quantification. The search parameters were: First search error window of 20 ppm and main search error of 4.5 ppm; Unspecific enzyme; Minimal unique peptides 1; and FDR 0.01 (1%) for peptide and protein identification. Minimum and maximum peptide length for unspecific search were set to 8 and 25, respectively. The Uniprot human database was used (download from UniProt Sept. 2015) including the sequence for NOPT as the search database.

Left ventricle lysates from 3–6 rats were prepared by using the TissueLyser II from Qiagen, Germantown, MD, United States. Briefly, frozen left ventricle tissue from rat heart (∼30 mg) was homogenized with ice-cold RH lysis buffer (PBS with 0.5% Triton and 0.1% Tween) supplemented with PhosSTOP tablets (4906837001, Roche Diagnostics, Indianapolis, IN, United States) and protease inhibitors (11836170001, Complete Mini EDTA-free tablets, Roche Diagnostics). The tissue samples were homogenized twice for 100 s at 30 Hz and then incubated on ice for 30 min. Supernatants were collected by centrifugation at 4,000 rpm for 10 min at 4°C and thereafter stored at -80°C. For IP experiments (Figures 3G, 4B), left ventricular tissue was pulverized in a mortar with liquid nitrogen, then added into lysis buffer (20 mM Hepes, pH 7.5, 150 mM NaCl, 1 mM EDTA, 0.5% Triton) with protease inhibitors and homogenized 3 × 1 min at 0°C. Supernatants were collected by centrifugation at 32,000 rpm for 60 min at 4°C and thereafter stored at -80°C.

Biotinylated peptide (8 μM final concentration) was incubated with 25 µl of monoclonal anti-biotin conjugated beads (A-1559, Sigma-Aldrich, St. Louis, MO, United States) in 100 µl of 1 × PBS for 2 h at 4°C with rotation. The beads were then washed one time with PBS, centrifuged at 1,000 rcf for 30 s at 4°C, and 200 µl of left ventricle lysate was added. The tubes were incubated for 2 h at 4°C with gentle rotation, and thereafter washed once in IP buffer (20 mM HEPES, pH 7.5, 150 mM NaCl, 1 mM EDTA and 1% Triton X-100). Finally, the samples were boiled in SDS loading buffer for 5 min at 96°C and analyzed by immunoblotting.

Left ventricle lysate was incubated with the specific antibodies (2 μg/IP), 50 μl protein PLUS A/G agarose beads (sc-2003, Santa Cruz) and with or without presence of 100 μM NOPT-TAT or OPT overnight at 4°C. Immunocomplexes were washed twice in IP buffer (20 mM Hepes, pH 7.5, 150 mM NaCl, 1 mM EDTA, 1% Triton), centrifuged at 3,000 g for 1 min at 4°C and boiled in SDS loading buffer before the immunoblot analyses were undertaken. Equal amount of mouse IgG (sc-2025, Santa Cruz) or a custom made anti-NCX1 blocking peptide (CGQPVFRKVHARDHPIPST) incubated with anti-NCX1 prior to immunoprecipitation (Hafver et al., 2016) was used as negative control.

The pull-down samples were separated on 4–15 or 15% SDS-PAGE Criterion Tris-HCl Gel (#3450027 and #3450020, BioRad Laboratories, Inc., Hercules, CA, United States) and transferred onto a PVDF membrane (1704157, Trans-Blot Turbo Mini PVDF Transfer Packs, Biorad Laboratories) using the Trans-Blot Turbo Transfer System (Biorad Laboratories). The PVDF was blocked in 1% casein or 5% milk in TBS-T for 60 min at room temperature (RT), and thereafter incubated with gentle agitation with primary antibody overnight at 4°C. After incubation with the primary antibody, the membrane was washed three times for 10 min in TBS-T and further incubated with the secondary antibody (see antibody section) and washed again. The chemiluminescence signals was developed using ECL Plus (RPN2232, GE HealthCare, United States) and detected by a LAS 4000 imager (Fujifilm, Tokyo, Japan). The densitometry analyses were performed using ImageJ (NIH).

Immunoblotting and IPs were carried out using anti-biotin-HRP (1: 5,000, A0185, Sigma-Aldrich, St. Louis, MO, United States), anti-PLM (anti-FXYD1, ab76597, 1:4,000, Cambridge, MA, United States), anti-NCX1 (1:1,000, custom made by Genscript Corp., Piscataway, NJ, United States, epitope: GQPVFRKVHARDHPIPST, mapped in (Wanichawan et al., 2011), anti-NKA (1:2,500, 05-369, Sigma-Aldrich) and anti-serine 68 phosphorylated PLM [pSer68-PLM, 1:1,000 (Fuller et al., 2009)]. Anti-rabbit IgG HRP (1:2,500 or 1:3,000, NA934, GE HealthCare, United States) and anti-mouse IgG HRP (1:3,000, NA931, GE HealthCare, United States) were used as secondary antibodies. Recombinant protein that covered the NCX1 cytoplasmic part added N-terminal histidine (HIS) and trigger factor (TF) tags to stabilize the protein (HIS-TF-NCX1_cyt) was custom made by Genscript Corp., (Piscataway, NJ, United States).

A 96-well ELISA microplate was coated with 10 μg of PLM_cyt or pSer68-PLM_cyt (without biotin tag) (Figures 1C,D) in PBS and agitated gently overnight at 4°C. After this, each well was washed with PBS supplemented with 0.05% Tween (PBS-T), blocked with 0.5% gelatine (G-1890, Sigma) for 1 h at RT and incubated with 5 μl of the given peptides (final concentration: 50 μM) for 2 h at 37°C with gentle agitation. Each well was then washed five times with PBS-T, incubated with anti-biotin-HRP for 30 min at RT, and washed again five times with PBS-T, before incubation with 100 μl of the Ultra TMB substrate solution (34028, Thermo Fisher Scientific, Waltham, MA, United States) for 15–30 min at RT with gentle agitation (blue color development). In order to stop the reaction, 2N hydrochloric acid (HCl) was added into each well (yellow color). The absorbance of each sample was measured at 450 nm (Hidex Sense multimodal microplate reader, Åbo, Finland). For the competition experiment (Figure 1F), the microplate was coated with 10 μg/ml recombinant HIS-TF-NCX1_cyt protein before 0.5 μl biotin-ahx-PLM peptide (final concentration: 5 μM) was incubated with or without 10 μl NOPT (100 μM, without any biotin tag) for 2 h at 37°C with gentle agitation.

The SPR analyses were performed using Biacore X100 analyser (Biacore Inc., Uppsala, Sweden). The biotinylated pSer68-PLM_cyt or PLM_cyt peptides (ligand) were immobilized on streptavidin (SA) chips (BR100032, GE Healthcare Life Sciences) at ∼50–200 resonance units (RU) after the chip was conditioned through application of three consecutive 1 min injections of 1 x Biacore running buffer (BR100826, GE Healthcare Life Sciences). NOPT-TAT or recombinant His-TF-NCX1_cyt (analyte) dialyzed into the running buffer was diluted over a range of concentrations (NOPT-pSer68-PLM_cyt: 4.5–175 nM; His-TF-NCX1_cyt-pSer68-PLM_cyt: 0.6–50 nM, 6.25–100 nM, and 3.12–50 nM; His-TF-NCX1_cyt-PLM_cyt: 9.38–150 nM, 12.5–200 nM, 18.75–300 nM, 4.69–75 nM and 9.38–150 nM), and injected over the sensor surface at a flow rate of 30 or 100 μl/min (to avoid mass transport limitations) for 180 s. The dissociation time was set at 600 s. The obtained sensorgrams were analyzed through use of Biacore X100 evaluation software and curve fitting was performed with the assumption of one-to-one binding (Langmuir) or steady state affinity.

8 weeks old C57BL6/J mice were randomized to SHAM or an aortic banding (AB) operation using o-rings (ORAB) as previously described (Melleby et al., 2018). Mice were anesthetized in a combination of 5% isoflurane and 95% O₂, before endotracheal intubation and ventilation using a MiniVent rodent ventilator (Harvard Apparatus, United States). Left sided hemothoracotomy in the third intercostal space was performed, before dissection and placement of the o-ring around the ascending aorta. Animals were allowed to recover, and post-operative analgesia was achieved by injections of buprenorphine in accordance with the institutional guidelines. Animal experiments was approved by the Norwegian National Animal Research Committee (FDU 7040).

Cardiomyocytes were isolated based on a similar protocol as recently described (Ackers-Johnson et al., 2016). Mice were anesthetized and mask ventilated with a combination of 3% isoflurane and 97% O₂. Deep surgical anesthesia was confirmed by abolished pain reflexes. The chest was opened, before the descending aorta and inferior caval vein were cut. Then 7 ml of buffer A (in mM: HEPES 25, NaCl 130, KCl 5.4, NaH₂PO₄ 0.4, MgCl₂ 0.5, D-glucose 22, pH 7.4) with 5 mM EDTA was injected into the right ventricle. Thereafter, the aorta was clamped and the heart was excised. 10 ml of the buffer solution and thereafter 3 ml of the buffer solution without EDTA were injected into the left ventricle over 2–5 min. Then preheated solution A containing 0.8 mg/ml collagenase II was injected into the left ventricle over ∼20 min. The atria and right ventricle were removed, and the cardiomyocytes were isolated according to the same procedure as described for the Langendorff-based isolation of rat cardiomyocytes (described further below).

Isolated cardiomyocytes were loaded for 10 min with 5 μM fluo-4 AM ester (Molecular Probes, Eugene, OR, United States) at RT. The superfusate contained (mM): NaCl 140, HEPES 5, KCl 5.4, CaCl₂ 1, MgCl₂ 0.5, D-glucose 5.5 and NaH₂PO₄ 0.4. The pH was adjusted to 7.4. The cells were loaded with either the TAT-conjugated NOPT (1 μM) or the TAT-conjugated scrambled control peptide (1 μM) for 20 min at RT. The peptides were also added to the superfusate. Whole-cell Ca²⁺ transients were recorded in field stimulated cardiomyocytes at 0.5 Hz (2.5 ms bipolar pulses) and 37°C with Cairn Research Optoscan Monochromator (excitation 485 nm, emission 515 nm long pass) (Cairn Research Ltd., United Kingdom). Regular Ca²⁺ transients were recorded for at least 3 min per cell, before the field stimulation was stopped and 10 mM caffeine was added. Background fluorescence was recorded for each cell. Ca²⁺ transients amplitude and SR Ca²⁺ load was calculated as maximal fluorescence (F) divided on resting fluorescence (F₀) on field stimulated and caffeine evoked Ca²⁺ transients, respectively (See Figure 4I for illustration of protocol details). Tau values were obtained by mono-exponential fitting of the Ca²⁺ extrusion phase from regular transients (τ) and caffeine transients (τ_caff) after rapid applications of 10 mM caffeine. The NCX rate constant was calculated as:   1 / τ caff (Díaz et al., 1997), as the non SERCA2, non-NCX mediated Ca²⁺ removal was considered negligible.

Male Wistar rats were anesthetized in a chamber using inhalable isoflurane (Abbott Scandinavia Ab, Solna, Sweden), and 150 IE heparin was administrated intravenously for post-excision thrombosis prophylaxis. The heart was excised and immediately cooled in 0.9% NaCl at 4°C. The aorta was cannulated and the coronary arteries were retrogradely perfused in a modified Langendorff setup with buffer A (in mM: HEPES 25, NaCl 130, KCl 5.4, NaH₂PO₄ 0.4, MgCl₂ 0.5, D-glucose 22, pH 7.4) at 37°C for 2–4 min, then with buffer A containing 0.8 g/L collagenase II (Worthington Biochemical Corporation, United States) and 6.7 μM CaCl₂ for 18–22 min. Atria and the right ventricular free wall were removed before the left ventricle was cut into small pieces in 8–10 ml of the perfusate containing 500 μl 2% BSA, and mechanically isolated by careful pipetting with a Pasteur pipette for 1 min. The cardiomyocyte suspension was filtered through a nylon mesh (200 μm) (Burmeister), and left at RT for sedimentation. Immediately following sedimentation (∼5 min), the supernatant was removed. The cardiomyocytes were washed three times in buffer A containing 1) 0.1% BSA and 0.1 mM CaCl₂, 2) 0.1% BSA and 0.2 mM CaCl₂, and 3) 0.05% BSA and 0.5 mM CaCl₂.

Ventricular rat myocytes were voltage clamped using continuous mode on an Axoclamp 2A amplifier and pCLAMP software (Axon Instruments, CA, United States). We used pipettes with moderately low resistance (2–3 MΩ) for a series resistance of 5–9 MΩ. The signal was sampled at 10 kHz and filtered with a low-pass filter before analysis. All amplifier and program settings were held constant during and between experiments. The cells were perfused at 37°C, and the perfusion system was placed very close to the cells, allowing rapid changes of solutions. We used an internal solution containing (in mM) (modified from (Despa and Bers, 2003): NaCl 20, KCl 30, K-Aspartate 80, TEA-Cl 20, HEPES 10, MgATP 5, MgCl₂ 0.7 (free Mg²⁺ 1 mM using Maxchelator, Stanford), BAPTA 3, CaCl₂ 1.15 (free Ca²⁺ 150 nM), pH = 7.2 (adjusted with KOH). 30 μM of OPT or the scrambled control peptide were added to the internal solution. After reaching whole-cell access, the cells were dialyzed for at least 4 min at −20 mV. Myocytes were patched in the following solution (in mM): NaCl 140, HEPES 5, KCl 5.0, CaCl₂ 1.0, MgCl₂ 0.5, D-glucose 5.5, NaH₂PO₄ 0.4, pH adjusted to 7.4 with NaOH. Solutions were switched after reaching whole-cell access to a solution containing (in mM): NaCl 20, NMDG 110, D-glucose 10, HEPES 5, TrisCl (or KCl) 15, NiCl₂ 5, BaCl₂ 2, MgCl₂ 1, pH adjusted to 7.4 with HCl. Symmetrical Na⁺ solutions were used to minimize the effects of intracellular Na⁺ gradients. I_NKA was elicited by rapidly switching from a solution containing 0 mM K⁺ to a solution containing 15 mM K⁺; I_NKA was interpreted as the peak K⁺-sensitive current. Voltage ramp protocols were applied before and immediately after activation of I_NKA. Cardiomyocytes were depolarized from −20 mV to + 70mV for 50 ms, then hyperpolarized to −120 mV (dV/dt = 380 mV s⁻¹), and back to −20 mV. The difference between the ramp after activation of I_NKA and the baseline ramp defined the I_NKA–voltage relationship. All currents were standardized by relating the absolute current (pA) to the cell size (pF).

All animal work was performed at the Institute for Experimental Medical Research, Oslo University Hospital, Norway. Animal handling was preapproved by the Norwegian National Animal Research Committee (FOTS ID 3820 for rats, and FOTS ID 7040 for mice) and conformed to the Guide for the Care and Use of Laboratory Animals (National Institutes of Health Publication 85-23, revised 2011).

To our knowledge, the binding affinities and kinetics of the possible PLM-NCX1 and pSer68-PLM-NCX1 interactions have not been investigated before. The cytoplasmic tails of PLM (PLM_cyt) and pSer68-PLM (pSer68-PLM_cyt) are reported to bind directly to two PASKT- and QKHPD-containing regions in the cytoplasmic loop of NCX1 (NCX1_cyt) (illustrated in Figure 1A) (Zhang et al., 2011; Wanichawan et al., 2016).

To measure the binding affinities and kinetics of the pSer68-PLM-NCX1 interaction, a biotinylated peptide that covered pSer68-PLM_cyt (illustrated in Figure 1B, upper panel) was immobilized on a SA chip, and a range of concentrations of a recombinant polypeptide that covered NCX1_cyt was injected over the chip. The result was analyzed by fitting with a 1:1 interaction model (Langmuir). The K_D for the pSer68-PLM_cyt-NCX1_cyt interaction was 4.3 ± 1.9 nM, with an association rate constant (k_a) = (3.8 ± 1.7) × 10⁵ M⁻¹s⁻¹ and a dissociation rate constant (k_d) = (1.1 ± 0.1) × 10⁻³ s⁻¹ (Figure 1C). These results indicated a strong pSer68-PLM-NCX1 interaction.

NCX1 binding to PLM is thought to be independent of the phosphorylation status of the serine 68 (Wanichawan et al., 2016) Therefore, we also immobilized biotinylated PLM_cyt (illustrated in Figure 1B, lower panel) and injected a range of recombinant NCX1_cyt over the SA chip. Biotin-ahx-PLM_cyt was found to bind with similar affinity to NCX1_cyt (4.1 ± 1.0 nM) as the pSer68-PLM_cyt did, but showed higher association and dissociation rate constants (k_a = (1.1 ± 0.3) × 10⁶ M⁻¹s⁻¹ and k_d = (4.0 ± 1.1) × 10⁻³ s⁻¹) (Figure 1D).

Taken together, our data indicated that both PLM_cyt and pSer68-PLM_cyt bound strongly to NCX1_cyt, but the non-inhibitory PLM_cyt-NCX1_cyt interaction exhibited higher association and dissociation rate constants (higher on/off rates).

To develop a proteolytically stable NCX1 activator peptide, we used a previously designed high affinity PLM binding peptide (OPT: YHPYKEIEQLIELANYQVLS) that was derived from amino acids 301–320 in NCX1 that contained most of the QKHPD motif (Figure 1A) (Wanichawan et al., 2016). OPT has previously been shown to bind to the flexible region that is found upstream of serine 68 in PLM_cyt and to have a somewhat broader binding motif compared with that of the NCX1 region 301–320 (illustrated in Figure 2A, 41-RCKFNQQQRTGEPDEEEGTFRSSIRR-63, in red) (Wanichawan et al., 2016). OPT and NCX1 thus bind PLM independently of the phosphorylation status of serine 68.

To identify the minimal PLM binding sequence, we overlaid the biotinylated pSer68-PLM_cyt sequence (illustrated in Figure 1B, upper panel) onto N- or C-terminal truncations of OPT that was spot-synthesized on membranes. Immunoblotting with HRP-conjugated anti-biotin showed that OPT retained most of its pSer68-PLM_cyt binding capacity, although three or two amino acids were deleted at its N- (Figure 2B) or C-terminus (Figure 2C), respectively. We also performed N- and C-terminal truncations of OPT concomitantly (Figure 2D). Taken together, we identified the 16-amino acids sequence PYKEIEQLIELANYQV (mini-OPT) as the minimum sequence that was required to bind pSer68-PLM_cyt.

To examine the substitution tolerance of this amino acid sequence, mini-OPT was analyzed by a two-dimensional peptide array, in which each amino acid (Figure 2E, sequence given above array) was systematically replaced with every possible other amino acid (Figure 2E, left of the array). When the membrane was overlaid with biotin-pSer68-PLM_cyt and HRP-conjugated anti-biotin was immunodetected, it was observed that positions 1–5 (P1, Y2, K3, E4, and I5), 7 (Q7), 12 (A12) and 15 (Q15) were tolerant of amino acid substitutions, whereas position 6 (E6), 8–11 (L8, I9, E10 and L11), 13–14 (N13 and Y14) and 16 (V16) were less tolerant (Figure 2E, yellow columns). Some of the substitutions in positions 1–5, 7, 12 and 15 appeared also to have a higher affinity for pSer68-PLM_cyt (Figure 2E, green circles) compared with the internal control peptides (Figure 2E, white circles), and were therefore tested in a second array. However, closer inspection revealed that these amino acid substitutions did not increase the pSer68-PLM binding further (Supplementary Figure S1A).

Conclusively, we identified PYKEIEQLIELANYQV (named mini-OPT) to hold the maximum pSer68-PLM_cyt binding capacity and to be substitution-tolerant at the N-terminus.

N-methyl-amino acid substitutions have been shown to improve metabolic stability of peptides (Chatterjee et al., 2008; Ding et al., 2020). Therefore the N-terminal proline in the minimal pSer68-PLM_cyt binding sequence (mini-OPT) was substituted with an N-methyl-proline (NMP). This novel synthetic peptide was named NOPT. To analyze the stability of the peptide, we employed human serum which is the first compartment to which the peptides would be exposed during intravenous administration. Both the NOPT and OPT peptides (Wanichawan et al., 2016) (illustrated in the upper panel of Figure 3A) were incubated in 30% human serum and their degradation patterns were analyzed by mass spectrometry. Whereas the half-life of OPT was very short (0.86 h), the half-life of the smaller NOPT peptide was found to be 17 times higher than this (14.9 h) (Figure 3A, lower panel; MS data in Supplementary Tables S1–S6).

To confirm that the NOPT retained the PLM and pSer68-PLM binding capacities, several binding analyses were performed. First, a pull down experiment was performed with biotinylated peptides (which are illustrated in Figure 3B) in rat left ventricle lysate. Immunoblotting with PLM antibodies showed that NOPT-biotin precipitated endogenous PLM, although the binding capacity appeared somewhat reduced compared with that of biotin-ahx-OPT and the smaller biotin-ahx-mini-OPT (without the NMP substitution) (Figure 3C). Moreover, when peptides were used with a C-terminal biotin tag in an ELISA assay, it was confirmed that NOPT bound to pSer68-PLM nearly as well as the OPT and mini-OPT did (Figure 3D). Next, we analyzed the binding affinity of NOPT to pSer68-PLM_cyt by SPR. A range of concentrations of NOPT was injected over immobilized biotin-ahx-pSer68-PLM_cyt on a SA chip. Using the steady state affinity model, the dissociation equilibrium constant (K_D) for the NOPT-pSer68-PLM_cyt interaction was found to be 129 nM (Figure 3E, the blank subtracted sensorgrams are shown in Supplementary Figure S1B). The dissociation rate (k_d) and association rate constants (k_a) were too fast to be determined. Despite its high on/off rates and lower binding affinity, excess NOPT reduced the biotin-ahx-PLM_cyt-NCX1_cyt interaction with approximately 50% in an ELISA-based competition assay (Figure 3F). However, when applied to left ventricle lysate, NOPT was not able to outcompete the physical interaction between the two full length proteins (Figure 3G). This finding is in line with that NCX1 and PLM might be in larger protein complexes consisting of NCX1 dimers and higher order PLM oligomers (Beevers and Kukol, 2006; Bossuyt et al., 2006; Song et al., 2011).

Altogether, NOPT exhibited improved peptide stability, retained most of its PLM/pSer68-PLM binding capacities, and reduced the PLM_cyt-NCX1_cyt interaction, but was not able to outcompete the physical interaction between the endogenous full length proteins.

Initially and before we developed NOPT, we established that the OPT peptide did not have any effect on NKA activity in isolated adult rat cardiomyocytes (Figure 4A). Consistently, we found that presence of high concentrations (100 μM) of OPT or the smaller proteolytically stable NOPT did not affect the endogenous PLM/pSer68-PLM-NKA interaction in left ventricular lysate (Figure 4B). These findings are in line with that NKA binds to the transmembrane domain of the PLM monomer (Shattock et al., 2015), and not the cytoplasmic part (PLM_cyt and pSer68-PLM_cyt) of the PLM dimer as NCX1.

We further analyzed the effect of NOPT on endogenous NCX1 activity in adult mouse cardiomyocytes. NOPT was made cell permeable by the addition of a TAT tag to its C-terminus. Compared with a TAT-tagged, scrambled control peptide, NOPT-TAT increased the activity of NCX1, as measured as the NCX1 rate constant, in adult cardiomyocytes that were isolated from both SHAM-operated (Figure 4C, approximately 40%) and aorta banded HF mice (Figure 4D, approximately 60%). NOPT-TAT did not change the Ca²⁺ transient amplitude or sarcoplasmic reticulum (SR) Ca²⁺ load in either SHAM (Figures 4E,G) or HF cardiomyocytes (Figures 4F,H). Representative tracings for NCX1 activity in SHAM and HF cardiomyocytes treated with the scrambled control peptide or NOPT-TAT are shown in Figures 4I–L.

Altogether, cell-permeable NOPT was found to increase endogenous NCX1 activity in adult cardiomyocytes that were isolated from both SHAM-operated and aorta banded HF mice. This finding suggests that NOPT is able to disrupt the inhibitory pSer68-PLM-NCX1 interaction.