Abstract
Diabetes mellitus is projected to be the leading cause of death worldwide by the year 2030 and has been linked to many cardiovascular disorders, which together are known as, diabetic heart diseases (DHDs). DHDs include coronary heart diseases (CHDs), heart failure (HF) and diabetic cardiomyopathy. The progression of both CHDs and cardiomyopathy can often lead to arrhythmias (irregular heart rhythm). A protein central to arrhythmias is the cardiac ryanodine receptor (RyR2). In the heart, the bulk of Ca2+ is stored in the sarcoplasmic reticulum (SR) with RyR2 acting as a key Ca2+ release channel of the SR. RyR2 dysregulation can result in spontaneous Ca2+ release in a process known as store-overload-induced Ca2+ release (SOICR). Post translational modifications (PTMs) such as phosphorylation and oxidation, are known to regulate RyR2 directly, both of which increase its activity, thereby triggering arrhythmias. RyR2 is also known to be indirectly regulated by O-GlcNAcylation (O-GlcNAc), a PTM involving the enzymatic attachment of the sugar, O-linked β-N-acetylglucosamine, to the serine/threonine residues of proteins. Increased O-GlcNAcylation has been shown to indirectly increase pathological Ca2+ leak through RyR2 due to activation of Ca2+/calmodulin-dependent kinase II (CaMKII). This research aimed to determine if RyR2 is also directly modified by O-GlcNAcylation and if so, delineate the resulting change in RyR2 function and nanoscale structure.
First, using western blotting, it was found that RyR2 is modified by O-GlcNAcylation in human and rat heart samples. However, densitometry analysis of the blots from rat and human samples showed no difference in RyR2 O-GlcNAcylation levels between control and diabetic samples. This led to the use of thiamet-G (Thm-G) and diazo-6-oxornoleucine (DON) to pharmacologically promote and inhibit O-GlcNAcylation respectively in HEK293 cells inducibly expressing RyR2. It was found that Thm-G (O-GlcNAc enhancer) was able to artificially increase O-GlcNAc-RyR2 levels beyond baseline levels in HEK293 cells inducibly expressing RyR2. Furthermore, by employing a single cell Ca2+ imaging approach, it was also discovered that SOICR was increased by Thm-G and blunted by DON (O-GlcNAc inhibitor). Moreover, when cells were treated with KN93 to inhibit CaMKII activity, pharmacological promotion of O-GlcNAcylation using Thm-G was shown to still increase the level of SOICR, possibly through RyR2. In order to ascertain potential sites which RyR2 could be O-GlcNAcylated, HEK293 cells, lacking three known RyR2 phosphorylation sites (serine 2808, serine 2814 and serine 2030; RyR2 3A mutants) were used. Cells lacking these 3 known phosphorylation sites showed no increased SOICR activity upon Thm-G treatment. Furthermore, it was also observed that cells lacking only one phosphorylation site at serine 2808, had a decreased response to O-GlcNAcylating conditions. This suggests that one or more of these serine residues on RyR2 may be required for O-GlcNAcylation to have a functional effect on RyR2. Using super-resolution microscopy, investigations were carried out as to how O-GlcNAcylation could affect the nanoscale morphology and spatial distribution of RyR2 clusters. The findings from this study suggest that treatment with Thm-G to promote O-GlcNAcylation changed the clustering of RyR2 in HEK293 cells and mice cardiomyocytes lacking CaMKIIδ (cardiac specific isoform of CaMKII). This change in clustering was observed as increased RyR2 cluster area/size, but with a more diffuse packing density. Further highlighting important mechanistic insights as to how O-GlcNAcylation and CaMKIIδ deletion could drive cluster migration and remodelling. left ventricular developed pressure (LVDP) was decreased following perfusion with Thm-G in the wild-type mouse strain but not in CaMKIIδ-KO strain. However, even though LVDP was decreased, heart rate, maximum dP/dT and minimum dP/dT were unchanged following Thm-G perfusion in both mouse strains.
Taken together, these findings suggest that RyR2 can be O-GlcNAcylated, potentially, at sites which are also regulated by phosphorylation (e.g. seine 2808). Furthermore, like phosphorylation, O-GlcNAcylation increased SOICR events. Also, long-term in vitro exposure to O-GlcNAcylating conditions as well as chronic in vivo deficiency in CaMKIIδ both led to nanoscale remodelling of RyR2 clusters. Combined, these could explain why Ca2+ leak and associated arrhythmias, together with impaired contractility are prevalent in chronic hyperglycaemic conditions (diabetes).