Modifying Vascular Calcification in Diabetes: Contribution of O-GlcNAcylation

Vascular calcification is increased in patients with diabetes,¹ and is associated with increased morbidity and mortality rates compared to diabetics without calcification.² Despite its considerable clinical significance, little is known about the molecular pathways through which vascular calcification is triggered by diabetes pathology, although several diabetes associated factors, including high glucose, could conceivably play important roles in the pathogenesis. Therefore a greater understanding of the mechanisms through which diabetes may induce calcification is required in order to develop effective strategies to interrupt this process.

Morphologically, vascular calcification is observed in the vessel intima or media layers and frequently overlaps in both regions. Intimal calcification has been associated with atherosclerotic vascular disease, wherein it appears as spotty calcifications detected by computed tomography. These spotty calcifications exist as small hydroxyapatite mineral clefs, referred to as microcalcifications, and appear to associate with cholesterol crystals in early lesions.³ A pathological consequence of these microcalcifications could be ruptures occurring in vulnerable plaques where they are located. Medial calcification on the other hand is mostly found in patients with diabetes along with those with chronic kidney disease. With the occurrence of medial calcification being independent of hypercholesterolemia, additional mechanisms including factors associated with diabetes pathogenesis may help drive its formation. Medial calcification develops as sheet-like calcifications of the tunica media, which lead to increased vascular stiffness associated with reduced compliance of vessels. These occurrences allow for both medial and intimal calcification to be observed directly by imaging the calcification of the vasculature,⁴ along with medial calcification to be observed indirectly through measuring vascular stiffness.⁵

The pathogenesis of vascular calcification is a highly complex, active, cell-regulated process, with several cell types including smooth muscle cells,⁶ macrophages,⁷ and circulating bone marrow-derived cells playing a role.³ Mechanistic studies of vascular calcification have largely focused on calcium and phosphate homeostasis, the release of calcifying vesicles, changes in the extracellular matrix, loss of inhibition, and cellular differentiation, along with several other molecular processes.⁸ Each of these processes have been explored further in detail. For example, the switching of vascular smooth muscle cells to osteoblast-like cells capable of generating a mineralized matrix, has been associated with the release of calcifying vesicles,^3,9 bone morphogenetic proteins,¹⁰ inflammation,^11,12 and oxidative stress associated with Runx2 activity and AKT signaling.¹³ Understanding the connections between these pathways, along with those to diabetes associated complications could aid in the development of anti-calcification therapeutics.

Hyperglycemia is a major complication associated with diabetes, with chronic hyperglycemia being tied to vascular complications in both type 1 and type 2 diabetic patients.¹⁴ However, how hyperglycemia is related to the development of vascular calcification in diabetic patients is currently unknown. One potential connection between hyperglycemia and vascular calcification may lie in the posttranslational modification, O-GlcNAcylation. O-GlcNAcylation is the glycosylation process through which N-acetylglucosamine (O-GlcNAc) gets added to serine and threonine residues of proteins. Recently this process has been associated with signaling, transcription and chronic disease pathologies, including diabetes,¹⁵ as such O-GlcNAcylation shares similarities to the process of protein phosphorylation. Although through its regulation it differs from the process of phosphorylation, which is regulated by many kinases and phosphatases. Instead O-GlcNAcylation involves a tight regulation controlled by two molecules: one that adds O-GlcNAc onto proteins, N-acetylglucosaminyl-transferase (OGT), and one that removes the modification, N-acetylglucosaminidase (OGA). Of interest O-GlcNAcylation has been shown to stimulate chondrogenesis, osteogenesis and correlates with the transcriptional activity of the osteogenesis regulator, Runx2. ^{16, 17} Given the association between osteogenesis and vascular calcification,¹⁸ and the involvement of O-GlcNAcylation in diabetes,¹⁹ exploring the role of O-GlcNAcylation in the development of diabetes-induced calcification may provide important mechanistic findings.

In this issue of Circulation Research Heath et al.²⁰ build on previous studies exploring the development of vascular calcification in diabetes. Specifically, they provide novel mechanistic insight that identifies O-GlcNAcylation of AKT on its T430 and T479 amino acids as a potential regulator of diabetes induced calcification. The in vitro analysis of this posttranslational modification hints at a pathway in which the phosphorylation of AKT on S473 may be regulated by AKT T430/T479 O-GlcNAcylation. This work concludes that phosphorylated AKT may result in a signaling cascade that leads to increased Runx2 transcription and activity, thereby inducing vascular calcification (Figure 1).

Figure 1. Working hypothesis generated from the findings of the Heath et al. study.

Diabetes associated hyperglycemia may lead to increased O-GlcNAcylation of AKT at the T430/T479 amino acids, causing increased AKT binding to the mTOR complex 2, and subsequent phosphorylation of AKT at S473, which may lead to increased Runx2 activity and the induction of vascular calcification.

Using streptozotocin (STZ) injected mice, a drug induced mouse model of type 1 diabetes, Heath et al. found a strong increase in vascular O-GlcNAcylation following the onset of increased blood glucose levels, along with a significant increase in aortic calcium content and vascular stiffness. In an effort to gain mechanistic understanding of the connection between O-GlcNAcylation and the development of vascular calcification they used a mouse vascular smooth muscle cell culture model. When these cells were cultured in osteogenic differentiation media along with an inhibitor of OGA, Thiamet-G, or with shRNA knockdown of the OGA enzyme, they developed a mineralized matrix. Suggesting blocking the removal of O-GlcNAc may promote a mineralization process. Supporting this in vitro finding in vivo, Thiamet-G administration to STZ diabetic mice was found to further enhance the level of vascular calcium accumulation. Taken together these results suggest that O-GlcNAcylation may promote vascular calcification in smooth muscle cells and diabetic mice.

On a mechanistic level, when the AKT T430/T479 amino acids were mutated such that they could no longer be posttranslationally modified by O-GlcNAcylation, constitutively active AKT phosphorylation at S473 and cellular calcium content were reduced to control levels. In addition, AKT binding to a component of the mTOR complex 2, Rictor, was substantially reduced by these mutations. With these findings it can be postulated that O-GlcNAcylation may aid in the binding of AKT to the mTOR complex 2, leading to AKT phosphorylation at S473. The role of mTOR in this pathway was further confirmed by blocking O-GlcNAcylation driven calcification in smooth muscle cells through rapamycin treatment. Altogether these findings point to the AKT T430/T479 amino acids as crucial regulators of an O-GlcNAcylation driven AKT signaling pathway that leads to calcium accumulation. While the full mechanistic pathway connecting O-GlcNAcylation to the development of vascular calcification remains to be found, this work helps elucidate a few potential key regulators and signals involved. It also suggests O-GlcNAcylation of AKT as a potential target in the treatment of diabetes-induced calcification.

Further analysis including in vivo imaging of vascular calcification and the use of additional model systems could strengthen the interpretations of this study. While the in vitro analysis in the Heath et al. work confirmed calcification through the use of Alizarin Red staining, the in vivo analysis measured it indirectly. Increased calcium content and vascular stiffness measurements suggest medial calcification; however some form of imaging would be needed to confirm this. For example, vascular calcifications in diabetic animals can be clearly seen through the use of a near-infrared fluorescent imaging agent (Figure 2). Imaging could also provide a means to more closely follow and evaluate changes in the development of diabetes-induced calcification when used in combination with modulators of O-GlcNAcylation of AKT.

Figure 2. Vascular calcification in diabetic mice.

Aortic valve calcification imaging using a calcium sensitive near-infrared fluorescence (NIRF) probe (red) in apoE−/− diabetic (STZ+) mice and controls (STZ−). Aorta (Ao), aortic valve (AV), fibrous trigone (FT).

Exploring this pathway in additional diabetic animals models, including those of type 2 diabetes, could be of importance. Especially as type 2 diabetes is a known risk factor for cardiovascular disease, with insulin resistance correlating with increased vascular calcification.²¹ Studying O-GlcNAcylation of AKT in animal models of type 2 diabetes such as ldlr^−/− diabetic mice could help expand the clinical relevance of the Heath et al. findings. It could also potentially discriminate mechanisms involved in the development of medial versus intimal calcification, helping to identify therapies most likely to benefit specific patient populations.