Troponin destabilization impairs sarcomere-cytoskeleton interactions in iPSC-derived cardiomyocytes from dilated cardiomyopathy patients

The sarcomeric troponin-tropomyosin complex is a critical mediator of excitation-contraction coupling, sarcomeric stability and force generation. We previously reported that induced pluripotent stem cell-derived cardiomyocytes (iPSC-CMs) from patients with a dilated cardiomyopathy (DCM) mutation, troponin T (TnT)-R173W, display sarcomere protein misalignment and impaired contractility. Yet it is not known how TnT mutation causes dysfunction of sarcomere microdomains and how these events contribute to misalignment of sarcomeric proteins in presence of DCM TnT-R173W. Using a human iPSC-CM model combined with CRISPR/Cas9-engineered isogenic controls, we uncovered that TnT-R173W destabilizes molecular interactions of troponin with tropomyosin, and limits binding of PKA to local sarcomere microdomains. This attenuates troponin phosphorylation and dysregulates local sarcomeric microdomains in DCM iPSC-CMs. Disrupted microdomain signaling impairs MYH7-mediated, AMPK-dependent sarcomere-cytoskeleton filament interactions and plasma membrane attachment. Small molecule-based activation of AMPK can restore TnT microdomain interactions, and partially recovers sarcomere protein misalignment as well as impaired contractility in DCM TnT-R173W iPSC-CMs. Our findings suggest a novel therapeutic direction targeting sarcomere- cytoskeleton interactions to induce sarcomere re-organization and contractile recovery in DCM.

3 anti-human antibodies for cardiac troponin T (Thermo Scientific and Abcam) and sarcomeric alpha-actinin (Sigma). DAPI was used for staining of nuclei. Coverslips were mounted on glass slides using Fluoromount-G. Pictures were taken with 10x, 40x (plan apochromat), and 63x (plan apochromat oil) objectives using an inverted confocal microscope (Carl Zeiss, LSM 710 Meta, Göttingen, Germany) and ZEN software (Carl Zeiss).
Quantification of protein co-localization. Confocal images generated as described above were analyzed for protein co-localization analysis using the ImageJ function Coloc2. Background correction was performed using Fiji-ImageJ. For individual cells, ROIs were selected. Coloc2 was used for background subtraction, and analysis for obtaining Manders´ correlation was performed as described before (7)(8)(9).
Co-immunoprecipitation with DYK-tagged TnT and immunoblot. HEK 293T cells overexpressing in a pcDNA 3.1 backbone either TnT-WT-DYK, TnT-R173W-DYK or DYK tag alone as a negative control were lysed and input was kept for analysis. Cell lysates were bound with DYK-antibody-decorated magnetic beads and subsequently, immobilized TnT-DYK was incubated for co-immunoprecipitation (10) with cell lysate from healthy control iPSC-CMs. The bound fraction was analyzed by immunoblot. The following antibodies were purchased from Abcam: Tropomyosin (ab7785), troponin C (ab137130), troponin I (ab52862), phospholamban (ab2865). Myosin heavy chain antibody was purchased from DSHB. PDE activity assay. Human iPSC-CM lysates were prepared as described (5). Sarcomerecontaining fractions were obtained by pelleting pre-cleared lysates (10). The cytosolic supernatant 4 was transferred to a new tube and analyzed separately. Both cytosolic and sarcomere-containing fractions were subsequently analyzed using a PDE activity assay kit (Abcam, ab139460) according to the manufacturer´s instructions.
Analysis of AMPK activity via phosphorylation of T172. Human iPSC-CM lysates were prepared as described (5) from healthy control (n=2 cell lines) and DCM (n=3 cell lines) groups.
To determine AMPK activity based on phosphorylation of Thr-172 in the AMPK catalytic subunit, cell lysates were incubated with A-769662 or control vehicle (DMSO) and subsequently subjected to immunoblot analysis.

Fluorescent resonance energy transfer (FRET) measurements and analysis.
All FRETmeasurements were performed with a Nikon Eclipse FN-1 microscope. An Opto-Led fluorescent light source (Cairn-Research) was used for excitation at 436 ± 25 nm, the excitation / emission dichroic was 455 nm (long pass). Emitted light was split with a Dual-View beam splitter (Optical Insights) and recorded with a CoolSnap HQ 2 camera (Photometrix). Emission light was split with a 505 nm longpass dichroic and was then filtered at 480 ± 15 nm for CFP emission and 535 ± 20 nm for YFP emission. Extracellular buffer for the measurements was a modified and CO2 supplemented Ringer-solution with NaCl (140 mM), KCl (3 mM), MgCl2 (2 mM), CaCl2 (2 mM), Glucose (15 mM), HEPES (10 mM), pre-adjusted to pH 7.2 with NaOH, then supplemented with NaHCO3 (10 mM) and finally pH-adjusted to pH 7.2. Acquisition and analysis were performed using Optofluor software (Cairn Research). All ratios were calculated as emission at 535 nm / emission at 480 nm. The TPNI-CUTie FRET-sensor was developed before (11). Adenoviruses were used to infect cells 24-48h before the measurements. Immunohistochemistry for sarcomeric alpha-actinin and cardiac troponin T, followed by confocal imaging was performed (shown in Fig 1A) and sarcomeric alpha-actinin and TnT signals were analyzed by Pearson´s correlation and Fast Fourier transformation, to obtain correlation coefficients in DCM iPSC-CMs (94 cells) than in healthy controls (n=78 cells). The difference between the groups is not statistically significant as calculated by Mann-Whitney test. TnT KO iPSC-CMs could not be reliably analyzed due to lack of signal for TnT. Data are expressed as mean ± sem.