Metformin prevented high glucose-induced endothelial reactive oxygen species via OGG1 in an AMPKα-Lin-28 dependent pathway

Metformin improved vascular function in obese type 2 diabetic patients. 8-oxoguanine glycosylase (OGG1), a main DNA glycosylase, was involved in vascular complications in diverse diseases. However, whether metformin suppressed endothelial ROS via OGG1 pathway was unclear. Human umbilical vein endothelial cells (HUVECs) were exposed to HG (high glucose) or HG with metformin. OGG1 and AMPfα levels were measured after metformin treatment while HG-caused ROS was measured by DHE prober. Diabetic mice were induced by daily intraperitoneal injections of streptozotocin (STZ). Metformin reduced Endothelial ROS caused by HG via upregulating OGG1. Additionally, OGG1 protein expression was dependent on its mRNA stability, which was reversed by genetic inhibition of AMPKα and Lin-28. The role of OGG1 on ROS stimulated by HG was partially dependent on NFKB/NOX4 pathway in HUVECs. These results suggested that metformin contacted HG-induced endothelial ROS via AMPKα/Lin-28/OGG1 pathway.


Introduction
Vascular complications of Diabetics severely hampered the quality of life and life expectancy of diabetic patients via causing the end-stage renal disease and blindness, and so on [1,2]. Endothelial reactive oxygen species (ROS) induced by hyperglycemia is the pathological basis of vascular complications and the potential clinical treatment target [3,4].
AMP-activated protein kinase (AMPK), consisting of a catalytic α subunit and regulatory β and γ subunits, is the master regulator of energy metabolism [5,6]. It was activated in response to energy stress such as low glucose, hypoxia, stroke and ischemia and so on [7,8]. Metformin was widely used to control the glucose level in the diabetic patients via AMPK activation. Metformin inhibited mortality in myocardial infarction and heart failure patients, the morbidity of cardiovascular events in heart failure and type II diabetes mellitus patients [9], and the incidence and severity of stroke in patients with type 2 diabetes mellitus [10]. The underlying mechanism of Metformin-mediated vascular function attracted our attention.
8-oxoguanine glycosylase-1 (OGG1) is one of DNA glycosylase enzymes for the repair of the DNA Lesion8-oxoguanine (8-oxoG) induced by oxidative stress [11]. Emerging evidence showed the OGG1 was involved in oxidative stress-related pathological progress [12,13]. OGG1 deficiency exhibited lower levels of DNA strand lesions, PARP1 overactivation and promoted cell death induced by hydrogen peroxide, suggesting the role of OGG1 on genome integrity [12]. Gene deletion of OGG1 promoted cells resistant to oxidative stress-induced DNA demethylation caused by oxidative stress [13]. However, the role of OGG1 on endothelial function in type II diabetes mellitus patients was still unclear.
In this study, we found that AMPK inhibitor metformin remarkably prevented endothelial oxidative stress via upregulating OGG1 in response to high glucose.

Metformin decreased endothelial ROS via OGG1 expression
Metformin was a known and widely used anti-diabetic agent. However, whether the Metformin improved vascular function of diabetic patients via OGG1 was unknown. At first, endothelial ROS was measured after high glucose incubation with metformin or not. The results showed that metformin significantly decreased endothelial ROS ( Figure. 2A-B). Then we tested the OGG1 protein levels and mRNA with metformin (0.5 mM) incubation for 24 hours. Metformin increased OGG1 mRNA and protein levels ( Fig. 2C Metformin increased OGG1 protein levels via mRNA stability in response to high glucose. Subsequently, the underlying mechanism of OGG1 upregulation induced by metformin attracted our attention. Thus, we performed mRNA degradation assays using actinomycin D (Act D, 10 μ g/ml) to inhibit de novo mRNA transcription. qPCR analysis showed OGG1 mRNA half-life time was significantly increased in response to metformin, suggesting metformin mainly increased OGG1 mRNA stability (Fig. 3A). Additionally, we tested protein degradation assays using cycloheximide (CHX, 10 μ g/ml)) to prevent protein translation. Western blot exhibited that metformin did not suppress endothelial OGG1 expression ( Fig. 3B-C). All results exhibited that OGG1 protein levels by metformin were dependent on its mRNA stability. HUVECs were incubated with actinomycin D (Act D, 10 μ g/ml) for 0, 6, and 12 hours, and then OGG1 mRNA was measured by qPCR. (B-C) OGG1 protein levels were tested by WB after CHX-treated HUVECs for 0, 6, and 12 hours. *P<0.05 versus control group (mannitol control group). Values are the means±SD with N = 3.

AMPKα-mediated OGG1 expression was dependent on its mRNA stability.
Concerning that AMPKα is a key downstream target of metformin and reported to regulate mRNA stability[18], we investigated whether OGG1 upregulation by metformin was dependent on AMPKα activation. Metformin significantly upregulated AMPKα in the HUVECs ( Lin28 stabilized OGG1 mRNA levels after high glucose treatment. To further identify the role of AMPKα in OGG1 regulation by metformin, we assessed RNA binding proteins (RBPs), which play significant roles in the regulation of mRNA stability [19]. Lin-28 is a member of RNA-binding proteins, and one potential target of AMPKα [20]. Additionally, the role of Lin-28 mainly depended on its cytoplasmic sublocation. We test the Lin-28 cytoplasmic distribution after metformin exposure in the presence of HG or not. The results showed metformin promoted HG-suppressed Lin-28 cytoplasmic distribution ( Fig. 5A-B). Subsequently, silencing Lin-28 significantly reduced OGG1 mRNA and protein levels induced by metformin, showing that Lin-28 was the key upstream of OGG1 mRNA stability ( Fig. 5C-E). Additionally, the OGG1 mRNA stability also measured after gene deletion of Lin-28 in the presence of actinomycin D (Act D, 10 μ g/ml). Silence of Lin-28 significantly decreased OGG1 mRNA stability ( protein and mRNA levels were tested by WB and qPCR after metformin and high glucose treatment simultaneously with silencing of Lin28 or not in the HUVECs. (F) The half-time hours of OGG1 mRNA was measured after silencing Lin-28 or not with actinomycin D (Act D, 10 μ g/ml). *P<0.05 versus control group. Values are the means±SD with N = 3.

Metformin activated OGG1 via Lin-28 phosphorylation by AMPKα.
Next, we explore the mechanism of metformin on Lin-28 cytoplasmic retention. Since AMPKα was reported to mediate Lin-28 nuclear export through phosphating Lin-28 [20]. According, we speculated metformin induced a dramatic accumulation of Lin-28 in the cytoplasm via AMPKα activity. Silence of AMPKα or AMPKα inhibitor with Compound C (10 μ M) attenuated the Lin-28 cytosolic location and phosphorylation ( Fig. 6A-B), suggesting the role of AMPKα in the Lin-28 distribution and activity. Further rescue experiments indicated that silencing Lin28 significantly reduced AMPKα-induced OGG1 expression induced by AMPKα activation (AICAR) or AMPKα plasmid ( Fig. 6C-D).

Metformin decreased endothelial reactive oxygen species via AMPKα/Lin-28/OGG1 pathway.
Lastly, we assessed the role of AMPKα/Lin-28/OGG1 pathway on metformin-prevented endothelial reactive oxygen species. HUVECs were pretreated with AMPKα inhibitor Compound C or gene deletion of Lin-28. The results exhibited that both Compound C or interference by siRNA of Lin-28 predominantly accelerated endothelial ROS production caused by high glucose (Fig. 7A-B).

OGG1 prevented HG-induced endothelial ROS via NFκB/NOX4 pathway.
We next searched for the possible mechanism linking OGG1 and ROS. Previous study reported that NADPH oxidases are the main endogenous resource of endothelial ROS [21]. So we measured the levels of NOX1, NOX2, and NOX4 after OGG1 overexpression with high glucose treatment or not. Forcing expression of OGG1 dramatically abrogated HG-induced NOX4 expression rather than NOX1 and NOX2 ( Fig. 8A-B). The endothelial mainly proinflammatory transcriptional factor NFκB has been reported involved in 8-oxoG-induced endothelial NOX4 expression [22]. Then we tested whether OGG1-inhibited NOX4 expression was dependent on NFκB. HUVECs were incubated with HG after overexpression of OGG1. It's clear that OGG1 significantly downregulated NFκB activity ( Fig. 8C-D). As expected, silence of NFκB partially reversed HG induction of NOX4 expression ( Fig. 8E-F). Collectively, we demonstrated that OGG1 reduced HG induction of ROS via NFκB/NOX4 pathway. (E-F) NOX4 protein level was assessed after NFκB silence with high glucose incubation. *P<0.05 versus Ctrl group. # P<0.05 versus HG group. NS P means no difference. Values are the means±SD with N = 3.

Discussion
In this study, we found that metformin protected against endothelial reactive oxygen species via upregulating OGG1. Additionally, the underlying mechanism of metformin in OGG1 expression was dependent on AMPKα/Lin-28 pathway (Fig. 9). Additionally, OGG1 significantly reduced endothelial ROS through NFκB/NOX4 pathway. Metformin is the first-line medication for the treatment of diabetes agent, improving the risk of cardiovascular disease among patients with type 2 diabetes, and reducing cardiovascular mortality and morbidity in patients comparison with alternative glucose-lowering drugs [23,24]. Metformin has also been reported to improve endothelium-dependent vasorelaxation in a NO dependent manner [25,26]. Additionally, administration of metformin in drinking water displayed an decrease in blood pressure and vascular damage dependently on AMPK/ER stress pathway [27]. Similarly, we also proved the effect of metformin on vascular damage. Metformin incubation predominantly suppressed endothelial ROS induced by high glucose measured by DHE fluorescent dye. Mechanismly, the actions of metformin were dependent on AMP-activated protein kinase (AMPK) or not. AMPK is considered to be a metabolic fuel gauge activated in response to energy depletion [28]. Metformin activated eNOS phosphorylation and NO production in the wild-type but not in AMPK-knockout mice, and overexpression of AMPK-DN (dominant-negative mutation) abrogated the protective effects of metformin on eNOS activity and endothelial apoptosis [29], suggesting the protective role of metformin was dependent on AMPK activity. Further evidence from mouse models with genetic perturbation of AMPKα documented the metformin preserved withdrawal signs precipitated by nicotine withdrawal corroborated dependently AMPKα expression [30]. However, AMPK might be not necessary for the role of metformin on adipogenesis and BMP6-induced hepcidin gene expression [31,32]. In this study, we assessed whether metformin decreased endothelial ROS in an AMPK pathway. The results exhibited that AMPKα is required for the role of metformin on ROS.
8-Oxoguanine DNA Glycosylase 1 (OGG1), a DNA glycosylase enzyme, was reported to involve in vascular oxidative stress and pathological progression of atherosclerosis [15,33]. OGG1 expression was controlled by diverse mechanisms. The nuclear transcription factor Y subunit alpha (NF-YA) promoted OGG1 mRNA transcription via binding to its promoter in the presence of TSC-2 gene [34]. Additionally, signal transducer and activator of transcription 1 (STAT1) mediated the transcriptional activity of OGG1 through recruiting and binding to the gamma-interferon activation site (GAS) motif of the OGG1 promoter region in endotoxintriggered inflammatory response [35]. Furthermore, Rapid and selective inactivation of OGG1 were degraded via binding to CHIP (C-terminus of HSC70-interacting protein) E3 ligase after cells subjection to mild hyperthermia [36]. Surprisingly, we found that OGG1 expression activated by metformin was dependent on mRNA stability after Act D incubation while gene silence of Lin-28 pronouncedly prevented against the OGG1 expression, suggesting the OGG1 levels stimulated by metformin relied on Lin-28 (RNA binding protein). Previously literature showed that AMPKα inhibitor using dorsomorphin promoted dramatic accumulation of Lin-28 in the nucleus and decreased Lin-28 activity. In our study, we also found that AMPKα regulated Lin-28 distribution and activation. However, the mechanism of AMPKα in Lin-28 activity was warranted to explore. . Additionally, post-hypoxic oxidative stress and programmed cell death in cardiomyocytes exposed to high glucose was attenuated by gene deletion of NOX2 [40]. Downregulation of Nox4 by gene deletion of NOX4 in the rat kidney glomerular mesangial cells inhibited the HG-induced augment in mitochondrial superoxide assessed by MitoSOX Red. In this study, we found that OGG1 reduced endothelial reactive species in the HUVECs administrated high glucose through reducing NOX4 expression rather than NOX1 and NOX2. Consistently with previous study [41], NOX4 downregulation by OGG1 was also dependent on NFκB activity. However, the potential mechanism was still unclear, and further research is warranted. Collectively, our results demonstrated that HG induced endothelial ROS, possibly by a mechanism involving in OGG1/NADPH pathway.
In conclusion. Our results showed metformin suppressed high glucose-triggered endothelial reactive oxygen species via OGG1 in an AMPKα-Lin-28 dependent manner.

Grants
This work was supported by grants from Innovation Program in Nanjing Medical University (SJZZ15_0115) and General project of clinical medicine project of Nantong University (2019JY020).  L  o  w  s  h  e  a  r  s  t  r  e  s  s  i  n  d  u  c  e  s  e  n  d  o  t  h  e  l  i  a  l  r  e  a  c  t  i  v  e  o  x  y  g  e  n  s  p  e  c  i  e  s  v  i  a  t  h  e  A  T  1  R  /  e  N  O  S  /  N  O  p  a  t  h  w  a  y  .   J  C  e  l  l  P  h  y  s  i  o  l  ,  2  0  1  8  .  2  3  3  (  2  )  :  p  .  1  3  8  4  -1  3  9  5  .   1  7  .  L  i  u  ,  Z  .  J  .  ,  e  t  a  l  .  ,   O  G  G  1  I  n  v  o  l  v  e  m  e  n  t  i  n  H  i  g  h  G  l  u  c  o  s  e  -M  e  d  i  a  t  e  d  E  n  h  a  n  c  e  m  e  n  t  o  f  B  u  p  i  v  a  c  a  i  n  e  -I  n  d  u  c  e R  a  p  i  d  i  n  a  c  t  i  v  a  t  i  o  n  a  n  d  p  r  o  t  e  a  s  o  m  e  -m  e  d  i  a  t  e  d  d  e  g  r  a  d  a  t  i  o  n  o  f  O  G  G  1  c  o  n  t  r  i  b  u  t  e  t  o  t  h  e  s  y  n  e  r  g  i  s  t  i  c  e  f  f  e  c  t  o  f  h  y  p  e  r  t  h  e  r  m  i  a  o  n  g  e  n  o  t  o  x  i  c  t  r  e  a  t  m  e  n  t  s  .   D  N  A  R  e  p  a  i  r  (  A  m  s  t  )  ,  2  0  1  3  .  1  2  (  3  )  :  p  .  2  2  7  -3  7  .   3  7  .  Z  h  a  n  g  ,  Y  .  ,  e  t  a  l  .  ,   N  A  D  P  H  o  x  i  d  a  s  e  s  a  n  d  o  x  i  d  a  s  e  c  r  o  s  s  t  a  l  k  i  n  c  a  r  d  i  o  v  a  s  c  u  l  a  r  d  i  s  e  a  s  e  s  :  n  o  v  e  l  t  h  e  r  a  p  e  u  t