MG53 inhibits angiogenesis through regulating focal adhesion kinase signalling

Abstract Mitsugumin 53 (MG53), which is expressed predominantly in striated muscle, has been demonstrated to be a myokine/cardiokine secreted from striated muscle under specific conditions. The important roles of MG53 in non‐striated muscle tissues have also been examined in multiple disease models. However, no previous study has implicated MG53 in the control of endothelial cell function. In order to explore the effects of MG53 on endothelial cells, human umbilical vein endothelial cells (HUVECs) were stimulated with recombinant human MG53 (rhMG53). Then, rhMG53 uptake, focal adhesion kinase (FAK)/Src/Akt/ERK1/2 signalling pathway activation, cell migration and tube formation were determined in vitro. The efficacy of rhMG53 in regulating angiogenesis was also detected in postnatal mouse retinas. The results demonstrated that rhMG53 directly entered into endothelial cells in a cholesterol‐dependent manner. The uptake of rhMG53 directly bound to FAK in endothelial cells, which resulted in a significant decrease in FAK phosphorylation at Y397. Accompanied by the dephosphorylation of FAK, rhMG53 uncoupled FAK‐Src interaction and reduced the phosphorylation of Src at Y416. Consequently, the activation of FAK/Src downstream signalling pathways, such as Akt and ERK1/2, was also significantly inhibited by rhMG53. Furthermore, rhMG53 remarkably decreased HUVEC migration and tube formation in vitro and postnatal mouse retinal angiogenesis in vivo. Taken together, these data indicate that rhMG53 inhibits angiogenesis through regulating FAK/Src/Akt/ERK1/2 signalling pathways. This may provide a novel molecular mechanism for the impaired angiogenesis in ischaemic diseases.


| INTRODUC TI ON
Mitsugumin 53 (MG53), also known as TRIM72, belongs to tripartite motif (TRIM) protein family, which is composed of a prototypical TRIM domain (including a RING domain, a B-box domain and a coiledcoil domain) at its N terminus and a SPRY domain at its C terminus. 1,2 MG53 is expressed predominantly in striated muscle (cardiac and skeletal muscle) 3 and has been reported to be involved in the regulation of a variety of physiological and pathological functions in striated muscle. One of the first discovered physiological functions of MG53 is involved in the repair of membrane damage. 3 It is now well established that both endogenous and exogenous MG53 play important roles in the repair process of skeletal muscle, cardiac muscle and even other non-striated muscle cell plasma membrane. 2,4,5 Subsequently, it has been demonstrated that MG53 prevents myocardium from ischaemia/reperfusion (IR) injury through participating in ischaemic preconditioning and ischaemic postconditioning. 6,7 Meanwhile, the effects of MG53 on system metabolism regulation have been gradually elucidated, but are still controversial. The RING domain of MG53 is identified as a muscle-specific E3 ubiquitin ligase that catalyses ubiquitin conjugation and mediates protein degradation. [8][9][10][11] It has been previously shown that endogenous MG53 negatively regulates insulin signalling via ubiquitination and degradation of insulin receptorβ (IRβ) and insulin receptor substrate-1 (IRS-1) in skeletal muscle. 9,10 Recombinant human MG53 (rhMG53) has also been demonstrated to allosterically inhibit insulin signalling through binding to the extracellular domain of IRβ. 12 These published reports indicate a causative role of MG53 in promoting the pathogenesis of diabetes. On the contrary, Ma and colleagues emphasize that neither sustained elevation of MG53 in circulation nor whole-body ablation of MG53 has significant influences on glucose handling and insulin sensitivity in db/db mice, 13,14 which argues against the proposed function of MG53 as a causative factor for diabetes development.
Furthermore, growing evidence has also indicated the widely protective effects of rhMG53 on non-striated muscle tissues in multiple disease models. Systemic administration of rhMG53 could reduce symptoms of lung injury via protecting impaired lung epithelial cells, 15 ameliorate I/R-, nephrotoxin-and cisplatin-induced acute kidney injury via repairing injured renal proximal tubular epithelium cells, 16 alleviate I/R-induced brain injury via suppression of apoptotic neuronal cell death, 17 decrease I/R-induced liver injury via inhibiting oxidative stress and hepatocyte apoptosis, 18 and promote the efficacy of human umbilical cord-derived mesenchymal stem cells in the recovery of traumatic brain injury via promoting hUC-MSC proliferation and migration. 19 Moreover, the toxicology studies have certified that repetitive intravenous administration of rhMG53 does not result in a measurable impact on metabolic function or vital organ toxicity. 14 These data highlight rhMG53 as a potentially safe biological reagent available for therapeutic application with no to limited effects on diabetes.
More importantly, previous studies have demonstrated that MG53 is myokine/cardiokine secreted from striated muscle in response to high glucose and high insulin, which may lead to a remarkable increase in serum MG53 content in rodents and humans with obesity and diabetes. 9,12,20 As well as known, angiogenesis disorders, including both excessive and defective angiogenesis, generally exist in patients with diabetes, which contribute to the development of cardiovascular complications. 21,22 However, no previous studies have investigated whether the elevation of serum MG53 is involved in the impaired angiogenesis. Therefore, the aim of this study was to determine the effects of rhMG53 on angiogenesis and to understand the molecular mechanism involved.
Angiogenesis is a physiological process including the development of new blood vessels from pre-existing vessels and the subsequent formation of a vascular network. The functional responses of endothelial cell, such as proliferation, migration and tube formation, are of great importance during angiogenesis. Focal adhesion kinase (FAK) is a cytoplasmic tyrosine kinase which is involved in integrin-and other cell surface receptor-mediated signal transductions. Previous studies have shown that the activation of FAK and its downstream signalling pathways, such as Src, Akt and ERK1/2, plays essential roles in cell adhesion, migration, tube formation and angiogenesis. 23
Photomicrographs were taken immediately after the scratch and after rhMG53 treatment. ImageJ software was used to measure the change in scratch area over time.

| Cell migration assay
HUVEC migration was also performed with transwell chambers with an 8.0 μm-sized porous membrane. HUVECs (2 × 10 4 /well) were added to the upper chambers and treated with rhMG53 (0, 5, 10 and 20 μg/ mL). As a chemoattractant, the endothelial cell medium containing 2% FBS was added to the lower chambers. After 24 hours, cells remaining in the upper chambers were removed and the cells that have migrated to the lower chambers were fixed with 4% paraformaldehyde, stained with 0.5% crystal violet, photographed with a microscope and counted.

| Immunofluorescent staining
In order to investigate the time course of rhMG53 uptake, HUVECs were stimulated with rhMG53 (10 μg/mL) for 0, 15, 30 and 120 minutes. In order to determine the effects of MβCD on rhMG53 uptake, HUVECs were pre-treated with MβCD (5 mM) for 1 hour, followed by stimulation with rhMG53 (10 μg/mL) for 3 hours. Then, the cells were fixed with 4% paraformaldehyde for 15 minutes at room temperature, permeabilized with 0.5% Triton X-100 for 15 minutes at 4℃, blocked with 5% bovine serum albumin at 37℃ and incubated with a primary anti-MG53 antibody (1:50) at 4℃ overnight. After washing with PBS, the cells were incubated with rhodamine-conjugated secondary antibody (1:100) against the primary antibody applied and the nuclei were stained with 4', 6-diamidino-2-phenylindole (DAPI).

| Immunoprecipitation
HUVECs were stimulated with rhMG53 (10 μg/mL) for 24 hours, and the protein samples were prepared. Classic IP kit was used to perform protein immunoprecipitation. In brief, equal amounts of protein were incubated with protein A/G plus agarose beads and anti-FAK antibody or IgG overnight at 4℃, with gentle rotation. According to the IP kit's instruction, the immune complexes were captured and solubilized by 50 μL SDS sample buffer. Then, the immune complexes were subjected to SDS-PAGE and immunoblotted with primary antibodies against MG53 and Src.

| Animal models
Postnatal retinal angiogenesis model was used in this study. All animal care and experimental procedures were approved by the Animal Care and Use Committee of Southern Medical University. Postnatal C57BL/6J mice were injected intraperitoneally with vehicle control or rhMG53 at 6 mg/kg bodyweight from postnatal day 5 (P5) to P12.

| Retina whole-mount immunostaining
Mouse retina whole-mount immunostaining was performed as previously described. 27 In brief, the eyes were harvested and immediately fixed with 4% paraformaldehyde for 2 hours at room temperature. Then, the eyes were washed 3 times with ice-cold PBS and the retinas were dissected and cut into four radial incisions under a stereomicroscope. After being fixed with 4% paraformaldehyde, permeabilized with 0.5% Triton X-100 and blocked with 5% bovine serum albumin, the retinas were washed and incubated with endothelial marker IB4 (1:100) at 4℃ overnight. Following thorough washing, the retinas were transferred to the glass slide, flat-mounted and sealed under a glass coverslip in mounting media.

| Statistical analysis
One-way ANOVA followed by post hoc comparison was performed to analyse the results, and all of the data are shown as mean ± standard deviation (SD). A P value of <.05 was considered as statistically significant.

| rhMG53 directly enters endothelial cells
MG53 is mainly expressed in striated muscle, and we fail to detect endogenous MG53 protein in HUVECs ( Figure S1). To explore whether rhMG53 could act on endothelial cells, we firstly determined whether rhMG53, when added to the extracellular space, could directly enter the HUVECs. HUVECs were treated with rhMG53 (0, 5, 10 and 20 μg/ mL) for 24 hours. Then, the MG53 protein in endothelial cells was detected with Western blotting. The results demonstrated that the MG53 protein bands clearly appeared after rhMG53 stimulation, indicating the direct entry of rhMG53 into endothelial cells ( Figure 1A). To determine the time course of rhMG53 uptake, HUVECs were exposed to rhMG53 (10 μg/mL) for 0, 0.25, 0.5, 1, 2, 4, 8, 12 and 24 hours and the results showed that the entry of rhMG53 into endothelial cells occurred 15 minutes after it was added to the extracellular space ( Figure 1B). Consistent with Western blotting results, the experiments involving immunofluorescence staining also confirmed that 15-min incubation of HUVECs with rhMG53 resulted in protein uptake ( Figure 1C,D). Taken together, these results indicate rhMG53 can directly enter into endothelial cells.
Furthermore, to investigate the specific mechanism of rhMG53 uptake, HUVECs were pre-treated with MβCD (5 mM) for 1 hour to deplete membrane cholesterol, 28,29 followed by stimulation with rhMG53 (10 μg/mL) for 3 hours. The results showed that MβCD significantly decreased the uptake of rhMG53 by endothelial cells (Figure 2), suggesting that rhMG53 entered endothelial cells in a cholesterol-dependent manner.

| rhMG53 binds to FAK and inhibits FAK phosphorylation
Previous studies have reported that endogenous MG53 can interact with FAK in skeletal muscle cells. 10 In order to investigate the role of F I G U R E 1 rhMG53 enters endothelial cells. A, HUVECs were exposed to rhMG53 (0, 5, 10 and 20 μg/mL) for 24 h. Then, the entry of rhMG53 into the endothelial cells was detected by Western blotting in total cell lysates. Representative images of 3 independent experiments are shown. B, HUVECs were exposed to rhMG53 (10 μg/mL) for 0, 0.25, 0.5, 1, 2, 4, 8, 12 and 24 h. Then, the rhMG53 entry into the endothelial cells was detected by Western blotting in total cell lysates. Representative images of 3 independent experiments are shown. C, HUVECs were stimulated with rhMG53 (10 μg/mL) for 0, 15, 30 and 120 min. Then, the uptake of rhMG53 by endothelial cells was identified by incubating with a primary anti-MG53 antibody at 4℃ overnight, followed by immunostaining with a rhodamine-conjugated secondary antibody (red). DAPI was used to label the nuclei (blue) and representative images of at least 3 independent experiments are shown. Scale bar is 50 μm. D, The enlarged images of the single cell in the white square are shown To explore the functional significance of rhMG53-FAK interaction, HUVECs were stimulated with rhMG53 (0, 5, 10 and 20 μg/mL) for 24 hours and the phosphorylated and total FAK were detected by Western blotting. The results demonstrated that rhMG53 prominently decreased FAK phosphorylation at Y397 in a concentrationdependent manner ( Figure 3B). However, no significant difference in total FAK expression was seen ( Figure S2). Furthermore, to determine the time course of rhMG53-induced dephosphorylation of FAK, HUVECs were exposed to rhMG53 (10 μg/mL) for 0, 2, 4, 6, 8, 12 and 24 hours. The results displayed that rhMG53 remarkably decreased the phosphorylation of FAK at Y397 as early as in 4 hours and rhMG53 gradually induced FAK dephosphorylation in a time-dependent manner ( Figure 3C). All of the results suggest that rhMG53 directly binds to FAK and simultaneously inhibits FAK phosphorylation.

| rhMG53 inhibits the activation of Src, Akt and ERK1/2
Src is one of the important downstream signals of FAK. FAK autophosphorylation at Y397 induces the binding of Src to FAK, which not only phosphorylates additional sites on FAK leading to its full activation, but also activates Src. 30 The mutually activation of FAK/ Src complex further phosphorylates multiple proteins to regulate different cell functions. 24,31 In order to investigate the effects of rhMG53 on the downstream signalling pathways of FAK, the interaction between FAK and Src was firstly determined by immunoprecipitation. The results showed that rhMG53 (10 μg/mL) treatment for 24 hours significantly inhibited the FAK-Src complex formation ( Figure 4A). The activation of Src was also detected and the results demonstrated that rhMG53 remarkably reduced the Src phosphorylation at Y416 ( Figure 4B). Moreover, the effects of rhMG53 on FAK/Src downstream signalling pathways, such as Akt and ERK1/2, were subsequently investigated and rhMG53 also significantly suppressed the phosphorylation of Akt and ERK1/2 ( Figure 4C). As a whole, all of these results indicate that rhMG53 inhibits the activation of FAK/Src/Akt/ERK1/2 signalling pathways. HUVECs were pre-treated with MβCD (5 mM) at 37℃ for 1 h, followed by exposure to rhMG53 (10 μg/mL) for 3 h. A, Then, the expression of MG53 was detected by Western blotting in total cell lysates. Representative images of 3 independent experiments are shown. **P < .01. B, The rhMG53 entry into cells was identified by incubating with a primary anti-MG53 antibody at 4℃ overnight, followed by immunostaining with a rhodamine-conjugated secondary antibody (red). DAPI was used to label the nuclei (blue) and representative images of at least 3 independent experiments are shown. Scale bar is 50 μm. C, The enlarged images of the single cell in the white square are shown

| rhMG53 inhibits angiogenesis in mouse retina
To examine the significance of our findings in vivo, the effects of rhMG53 on retinal angiogenesis in postnatal mice were detected.
Postnatal mice were injected intraperitoneally with rhMG53 (6 mg/ kg) every day for 1 week. Then, the whole retina was cut into four radial incisions and the vessels were labelled by IB4 staining. The results demonstrated that the vessel density in rhMG53-treated mouse retina significantly decreased ( Figure 6). These data are consistent with our results in vitro and clarify the significance of antiangiogenesis effects of rhMG53 in vivo.  Akt signalling pathway activation, 8,9,12 which is at least partly consistent with the findings in the present study. However, in cardiac ischaemia/reperfusion model, endogenous MG53 has been shown to participate in cardioprotection via activating Akt and ERK1/2, 6,7 which are opposite to our results. But, when rhMG53 was applied, In summary, the findings in the present study advance our understanding for rhMG53 in the regulation of endothelial cell function. We show that rhMG53 inhibits angiogenesis with mechanisms involving entry into HUVECs in a cholesterol-dependent manner, directly binding to FAK, uncoupling FAK-Src interaction and consequently inactivation of Src/Akt/ERK1/2 signalling pathways. These data indicate that the elevation of serum MG53 may contribute to the underlying molecular mechanisms of diabetic microangiopathy.

| D ISCUSS I ON
In addition, our results shall provide a base for the potential application of rhMG53 to treat diseases with excessive angiogenesis.

CO N FLI C T S O F I NTE R E S T
The authors declare no conflict of interest.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data that support the findings of this study are available from the corresponding author upon reasonable request.