QUESTION: The crucial contribution of the giant myofilament protein titin to diastolic stiffness and cardiomyocyte passive force (F_passive) is dependent, in part, on titin isoform composition and phosphorylation. Phosphorylation of titin by cyclic guanosine monophosphate (cGMP)-dependent protein kinase G (PKG) lowers titin-based stiffness, thus mediating a mechanical signaling process that is impaired in heart failure. In this project we elucidate which elements of the nitric oxide (NO) cGMP-PKG signaling network are critical for titin phosphorylation and stiffness in vivo.

METHODS: We used genetic knockout (KO) mouse models deficient for enzymes of the cGMP-PKG pathway, including cardiomyocyte-specific deletion of the guanylyl cyclase (GC)-A receptor and cGMP-dependent PKG (cGKI), as well as global deletion of soluble GC (sGC), as well as wildtype controls. We assessed titin phosphorylation in the heart by immunoblotting using phosphoserine-specific titin antibodies and by mass spectrometry quantification using stable isotope labeling of amino acids in mixed cultures of heart tissue from either wild-type (WT) or KO mice. The F_passive of single permeabilized cardiomyocytes was recorded before and after administration of PKG. In addition, we explored the effects of oxidative stress on cGMP-dependent PKG signalling in myocardial biopsies of heart failure with preserved ejection fraction (HFpEF) patients and the regulatory action of PKG on CaMKII activity.

RESULTS In all three genetic models, all-titin phosphorylation was reduced compared to WT hearts. The important PKG-dependent phospho-S4080 site within the N2-Bus region of mouse titin was hypophosphorylated in all three KO models. Unexpectedly, mass spectrometry analysis revealed that most class 1 titin phospho-sites within the molecular spring segment of titin, including those present in the Ig-domain regions, were hyperphosphorylated in the cGKI KO hearts compared to WT hearts. Only a few sites showed a phosphorylation deficit or remaining unchanged. Particularly in the cGKI model many class 1 phospho-sites were hyperphosphorylated compared to WT hearts indicative of the presence of compensatory processes following loss of PKG; indeed, this was associated with upregulation of CaMKII and MAPK and a clear rise in F_passive in KO vs. WT cardiomyocytes. While administration of PKG lowered F_passive of WT and KO cardiomyocytes in all models, this effect was more pronounced in the cGKI KO. In both HFpEF biopsies and cGKI KO mice oxidative stress correlated significantly with increased CaMKII activity and reduced PKG activity.