Prohibiting MG53 Phosphorylation Optimizes its therapeutic Potential in Diabetes

Organ damage from diabetes drastically reduces patient quality of life and places an enormous burden on healthcare systems across the globe. The heart is particularly affected in type 2 diabetic patients with most deaths attributed to cardiovascular complications. During diabetes, heart function is impaired with compromised cellular signaling, making it more susceptible to additional insults such as myocardial infarction (MI) and ischemia/reperfusion (I/R) injury.¹ Diabetes is a risk factor for ischemic heart disease as evidenced by a two- to four-fold increase in the incidence of MI in diabetic patients compared to non-diabetics. Understanding how diabetes sets the stage for increased myocardial damage has important clinical implications.

The heart is an insulin responsive organ.¹ Cardiac insulin signaling mediates cellular homeostasis via control of protein synthesis, substrate utilization, and cell survival. In a healthy heart, insulin activates phosphoinositide-3-kinase (PI3K)-Akt signaling and promotes GLUT4 translocation to the cell surface to facilitate glucose transport. Under normal conditions, the heart primarily uses lipid as a fuel via aerobic metabolism. However, during ischemia the heart shifts its energy production from aerobic metabolism to anaerobic metabolism, in which anaerobic glycolysis becomes important in the preservation of myocardial viability. The impaired ability to utilize glucose due to blunted insulin responsiveness, i.e., insulin resistance, contributes to the pathology of the diabetic heart.

MG53, also known as TRIM72, has been shown to regulate insulin sensitivity in skeletal and cardiac myocytes. MG53 is highly enriched in striated muscle and promotes the ubiquitination and degradation of IRα/β and IRS1.^2,3 The E3 ubiquitin ligase function of MG53 localizes to its N-terminus and within its tripartite motif (TRIM) consisting of RING, B-box, and coiled-coil sequences. A cystine to alanine mutation at residue 14 (C14A) in the RING domain of MG53 blocks its E3 ligase activity and prevents ubiquitination of its substrates including insulin receptor (IR) α/β and insulin receptor substrate 1 (IRS1).^2,3 The function of the central and C-terminal regions of MG53, on the other hand, acts to repair damaged membranes. Adult differentiated myocytes have limited regenerative capacity and therefore rely on effective membrane repair mechanisms for cell survival. Upon membrane rupture and exposure to the oxidizing extracellular environment, MG53 oligomerizes through interchain Cystine 242 disulfide bond formation and translocates to membrane rupture sites with bound vesicles to facilitate membrane repair using its PRY/SPRY domain. ⁴ How the separate degradation versus reparative activities of MG53 contribute to its overall function has yet to be fully understood. Genetic manipulation of MG53 suggests its membrane repair function dominates under basal conditions. MG53 null mice have a progressive skeletal myopathy and exhibit diminished myocyte membrane repair after exercise.⁴ However, cardiomyocyte transverse-tubule membrane integrity remains intact in the absence of MG53 at baseline and over a long period of time. Under mechanic pressure overload stress, MG53 deletion leads to more severe cardiac dysfunction as a result of exacerbated T-tubule degeneration, likely due to the loss of MG53-mediated membrane repair function.⁵

Interestingly, MG53 is not locally restrained to the damaged myocyte from where it is produced but is also effectively released into the circulation during exercise and pathological stress including myocardial infarction and ischemia/reperfusion (I/R) injury.^6,7 Several studies have now demonstrated that circulating MG53 can act as an important systemic myokine for recruitment to sites of organ damage. Intravenous injection of recombinant human MG53 (rhMG53) reduces the pathological consequences from muscular dystrophy and kidney injury, among other models.^8,9 In a similar fashion, transgenic expression of serum-targeted MG53 in MG53 null mice demonstrates its systemic myokine function is distinct from its intracellular role ¹⁰ and can even cross the blood-brain barrier to facilitate recovery from neuronal I/R injury.¹¹

Given its ability to ubiquitinate insulin signaling molecules for their degradation, there is growing interest into whether intracellular and/or systemic MG53 plays a role in the pathogenesis of diabetes. MG53 deleted mice are protected against high fat diet induced hyperglycemia, hyperlipidemia, and insulin insensitivity.^2,3 In parallel, insulin stimulation of MG53 knockout skeletal myocytes results in higher levels of total and phosho-IRS1 and phospho-Akt compared to wildtype myocytes suggesting insulin signaling is elevated in MG53 null muscle through reduced MG53-dependent ubiquitination. On the other hand, insulin triggers MG53 secretion from skeletal and cardiac tissue both in normal and diabetic animals.^7,12 The consequences of circulating MG53 on diabetic outcomes apart from its intracellular role has begun to be explored but remains a matter of debate.¹²

Mechanistic studies that differentiate between the positive actions of MG53 on membrane repair from its negative effects on insulin signaling would be expected to shed light into the function of MG53 in the context of diabetes and diabetic cardiomyopathy. Xiao and colleagues wondered if the E3 ligase function of MG53 could be manifested more clearly in diabetic animals subjected to myocardial ischemic events. Earlier this year, the Xiao group found that E3 ligase deficient knock-in mutation of MG53 (C14A), or systemic administration of rhMG53-C14A protein is more effective at protecting mice from MI and I/R damage than wildtype (WT) MG53 controls,¹³ supporting the idea MG53 ligase activity is detrimental in diabetic animals upon organ damage.

In this issue, Lv, et al. extended their findings into MG53 ligase function by exploring whether endogenous kinases act to shift the balance between MG53 substrate ubiquitination and membrane healing activities. See Figure 1.¹⁴ By performing mass spectrometry analysis on MG53 immunoprecipitated from skeletal myocytes and heterologously expressing HEK cells, the residue Serine 255 was found to be phosphorylated in both groups and among five phosphosites in the muscle expressed protein. Of the candidate phosphorylated residues, only mutation of Serine 255 to alanine (S255A), which prevents phosphorylation of the residue, blocked the ubiquitination and downregulation of IRS1 and IRβ in cultured neonatal rat ventricular myocytes (NRVMs) and immortalized C2C12 myotube cells. Mimicking phosphorylation of Serine 255 by aspartate and glutamate substitution had no effect on IRS1 and IRβ downregulation suggesting Serine 255 is basally phosphorylated. Serine 255 appears to mediate MG53 recognition of its substrates as MG53 S255A mutants fail to interact with either IRS1 or IRα.

Figure 1.

Schematic diagram showing the dual paradoxical functions of MG53 protein (left), and how modifications of key residues transform MG53 into a single purposed therapeutic (right).

Prediction analyses suggests Serine 255 is a GSK3β consensus site. Using skeletal muscle lysates, GSK3β was found to reciprocally co-immunoprecipitate with MG53. Furthermore, inhibition of GSK3 blunted MG53 phosphorylation at Serine 255 and MG53-S255A mutation blocked the ubiquitination of IRS1 in response to GSK3 overexpression. These results establish that a mechanistic feedback loop exists between insulin sensitivity, GSK3β, and MG53. Examination of cardiac tissue from several rodent and non-human primate models of metabolic syndrome revealed that IRS1, IRβ, and inhibitory GSK3β phosphorylation levels are decreased in parallel with increased MG53 phosphorylation at Serine 255. Furthermore, the S255A mutation does not affect the membrane repair function of MG53. Overexpression of MG53-WT and MG53-S255A constructs equally protected NRVMs against hypoxia-induced cell death. Exogenously applied rhMG53-WT and rhMG53-S255A protein both attenuated hypoxic LDH release and cell death as well as accelerated the healing of detergent-ruptured myocyte membranes.

Cumulatively, these findings suggested MG53 phosphorylation at Serine 255 is a pathologically relevant modification. To investigate the translational potential of MG53, the authors studied short and long-term consequences from MI and I/R in spontaneously diabetic db/db animals treated with rhMG53-WT or rhMG53-S255A. In each scenario, rhMG53-S255A was found to be more effective than rhMG53-WT at improving cardiac outcomes from ischemic and reperfusion injury.

In summary, the study by Lv et al.¹⁴ provides new insights into how MG53 targets protein substrates for ubiquitination. In diabetic heart, GSK3-dependent phosphorylation of MG53 at Serine 255 increases MG53 recognition, ubiquitination, and degradation of critical insulin signaling pathway proteins, contributing to insulin resistance and increased susceptibility to ischemic injury. Preventing Serine 255 phosphorylation by GSK3 potentiates the beneficial effect of MG53 in limiting the degree of cardiac damage from ischemic events in the context of metabolic disease. It will be important for future studies to determine whether other MG53 ubiquitinated substrates in addition to IRs and IRS1 contribute to end-organ damage. Also important to address is whether Serine 255 directly affects the E3 ligase activity of MG53 or only prevents MG53 from finding its substrates. Another open question is whether Serine 255 phosphorylation also contributes to self-dimerization and/or membrane fusion in addition to its effects on substrate ubiquitination.