Apelin/APJ-Manipulated CaMKK/AMPK/GSK3β Signaling Works as an Endogenous Counterinjury Mechanism in Promoting the Vitality of Random-Pattern Skin Flaps

A random-pattern skin flap plays an important role in the field of wound repair; the mechanisms that influence the survival of random-pattern skin flaps have been extensively studied but little attention has been paid to endogenous counterinjury substances and mechanism. Previous reports reveal that the apelin-APJ axis is an endogenous counterinjury mechanism that has considerable function in protecting against infection, inflammation, oxidative stress, necrosis, and apoptosis in various organs. As an in vivo study, our study proved that the apelin/APJ axis protected the skin flap by alleviating vascular oxidative stress and the apelin/APJ axis works as an antioxidant stress factor dependent on CaMKK/AMPK/GSK3β signaling. In addition, the apelin/APJ-manipulated CaMKK/AMPK/GSK3β-dependent mechanism improves HUVECs' resistance to oxygen and glucose deprivation/reperfusion (OGD/R), reduces ROS production and accumulation, maintained the normal mitochondrial membrane potential, and suppresses oxidative stress in vitro. Besides, activation of the apelin/APJ axis promotes vascular migration and angiogenesis under relative hypoxia condition through CaMKK/AMPK/GSK3β signaling. In a word, we provide new evidence that the apelin/APJ axis is an effective antioxidant and can significantly improve the vitality of random flaps, so it has potential be a promising clinical treatment.


Introduction
Wound repair is a common clinical problem in surgery. A skin flap has been proven to be one of the most common and effective methods for wound repair in surgical practice. Random-pattern skin flap replantation is increasingly being selected to close large surficial tissue defects caused by various reasons such as trauma, cancer excisions, and congenital deformity corrective surgery [1][2][3]. Obviously, wound heal-ing is a coordinated, complex process involving oxidative stress, inflammation, cell proliferation, cell migration, and angiogenesis [4]. During skin flap replantation, especially with the purpose of treating large-area wounds, blood supply interruption, oxidative stress, inflammation, and postoperative infection make the distal flap vulnerable to necrosis and to a great extent limit the clinical application of a random skin flap [5][6][7][8]. Blood supply recovery is a consequence of vascular stump's initiation, growth, and recanalization from the pedicle to the distal area [3,9]. During neovascularization, the ischemia-reperfusion injury (IRI) process is inevitable and necessary, in which effective and reasonable artificial intervention can greatly improve the survival of the skin flap by diminishing the accumulation of acute inflammatory cells [10] and suppressing oxidative stress [11]. As it is a main factor resulting from necrosis of grafts [12][13][14], IRI is frequently targeted by a series of medicine, treatments, and even biomaterials [15][16][17][18].
Considering all of the above, discovery of the substances with both proangiogenesis and antioxidant stress may play positive roles in the clinical application of a skin flap. Apelin, a group of peptides initially isolated from the bovine stomach, is an endogenous ligand of the G protein-coupled APJ receptor [19][20][21]. Only by binding to the APJ receptor could apelin exert their biologic function [19,22,23]; while apelin and the APJ receptor are widely distributed in different organs and tissues [22,23], the apelin/APJ signal axis may play a role in diseases of different organs including pulmonary fibrosis, kidney inflammation [24], acute lung injury (ALI) [25], ischemic heart failure [26], and even osteoporosis [27,28]. What we are interested in is that the apelin/APJ signal axis is especially related to vascular pathophysiology [24,29,30]. In addition, the apelin/APJ axis regulates IRI and antioxidant stress through different mechanisms such as regulating FoxO3 trafficking [31], maintaining the Ca2+ transient [32], attenuating excessive mitochondrial ROS production [21], and restoring the metabolic defects of lipid oxidation [33]. Existing researches lay the foundation for our hypothesis that apelin can be selected as an adjuvant therapy of skin flap transplantation by regulating angiogenesis and treating oxidative stress.
AMP-activated protein kinase (AMPK) is described as the "energy sensor" or "gauge" and expressed in all cell types and is closely related to IRI and oxidative stress. Since a previous study has proven that apelin protected against IRI-involved ROS-mediated inflammation via the AMPK/GSK3β pathway activated by apelin receptor/Gα/PL-C/IP3/CaMKK signaling and further upregulated downstream Nrf2-related antioxidant components in the central nervous system [34], skin vessels act as peripheral terminal vessels, and whether a similar mechanism can be established and further improve the survival of a random skin flap is still not clear. Considering this, we carry out our study to verify and increase the feasibility for clinical application. The apelin family is a peptide family, including apelin12, apelin13, apelin17, and apelin36. Our study preliminarily explored the antioxidant stress response and mechanism of the apelin/APJ axis. As a signal axis is damaged under pathological conditions, the apelin/APJ axis has potential therapeutic target value for further research.

Flap Macroscopic Evaluation.
After operation, the viability of skin flaps was tracked macroscopically by necrotic area, edema, colour, and hair conditions on day 1, day 3, and day 7.
Then, on day 7, high-resolution photos of the random flaps were obtained to evaluate the viability. All photos were processed by using the Image-Pro Plus imaging software (ver 6.0; Media Cybernetics) to determine the survival area, and the survival percentage of the skin flap was determined as follows: extent of the survival area × 100%/total area (survival and necrotic).
2.8. Laser Doppler Blood Flow (LDBF) Measurement. LDBF measurement was performed to evaluate vascular flow and blood supply in whole skin flaps. On day 3, day 5, and day 7, all mice were anesthetized and placed in a warm and quiet room, and the laser Doppler instrument (Moor Instruments, Axminster, UK) was used to scan mice in each group. The infrared Doppler scanning probe scanned the area of the skin flap to record the blood flow distribution as described in the previous literature [3,9]. Blood flow of the skin flap was visualized with the strong signal (green, yellow, and red shown in pictures) and quantified with Moor LDI Review software (ver 6.1; Moor Instruments). Measurement of each mouse was performed three times at least, and the mean value was used.
2.9. Tissue Edema Measurement. Water content reflects edema of the skin flap. As previously described [3,9], on day 7, skin flap tissue specimens were obtained and weighed and wet weight was recorded. Then, all specimens were dehydrated in an autoclave at a temperature of 50°C and weighed every day; finally, the stabilized weight was recorded as dry weight. The content of water was calculated as follows: percentage of tissue water content = ð½wet weight − dry weight/wet weightÞ × 100%. We calculated the number of microvessels per unit area (1 mm 2 ) under a light microscope (Olympus Corp, Tokyo, Japan), which was described as the microvascular density.
2.11. Immunohistochemistry (IHC). Skin flap sections in each group were baked at 65°C for 2 hours, then dewaxed twice in xylene, and hydrated with gradient ethanol baths. After washing, we blocked tissue sections with 3% hydrogen peroxide solution and performed antigen retrieval (20 min, 95°C) with 10.2 mM sodium citrate buffer (pH 6.0). Then, after blocking tissue sections with 10% goat serum albumin phosphate-buffered saline (10 min, 37°C), the sections were incubated with primary antibodies against CD34 (1 : 200), apelin (1 : 200), and APJ (1 : 200) overnight at 4°C. The second day, these sections were incubated with an HRPconjugated secondary antibody and counterstained with hematoxylin. Finally, these slices were sealed with neutral resin and placed in a ventilation cupboard overnight. Stained flap sections were pictured at ×200 magnification using a DP2-TWAN image-acquisition system (Olympus Corp, Tokyo, Japan). IHC images were evaluated with Image-Pro 3 Oxidative Medicine and Cellular Longevity Plus software (Media Cybernetics) for integral absorbance quantitation of apelin-, APJ-, and CD34-positive blood vessel counting.

Reactive Oxygen Species (ROS) Measurement.
HUVECs were cultured on glass creep plates at a seeding density of 1 × 10 5 . After all stimulations, a reactive oxygen species (DCFH-DA) assay kit (Beyotime, China) was used to measure the cellular ROS level according to the manufacturer's instructions. Images were pictured by using an Olympus microscope. The average fluorescence intensity of the fluorescence emitted by DCF is measured by using Image-Pro Plus software, which reflects the accumulation level of ROS in a positive correlation.
2.13. Immunofluorescence Staining. HUVECs were cultured on 12-well culture plates containing glass creep plates at a seeding density of 1 × 10 5 . After all stimulations, remove medium and wash creep plates with aseptic phosphate buffer; then, cells were fixed in 4% paraformaldehyde for 20 min and permeabilized with 0.1% Triton X-100 for 20 min. 10% goat serum albumin was used to block plates (10 min, 37°C), and cell creep plates were incubated with primary antibodies against Nrf2 (1 : 200), HO-1 (1 : 200), and SOD2 (1 : 200) overnight at 4°C. For tissue sections, procedures before primary antibody incubation are the same as described in IHC; then, sections were incubated with primary antibodies against apelin (1 : 200) and APJ (1 : 200) overnight at 4°C. The second day, all sections and plates were washed three times with aseptic phosphate buffer, incubated with a Cora-Lite488/594-conjugated second antibody for 1 hour at 37°C, and finally stained with DAPI. All images were evaluated under a fluorescence microscope (Olympus, Tokyo, Japan). The percentage of Nrf2-, HO-1-, and SOD2-positive cells in the dermal layer was calculated, and apelin-and APJ-positive areas were colocated.
2.14. Western Blotting. After being euthanized, samples (0:5 × 0:5 cm) from the middle of area II flaps in each group were removed from mice and stored at −80°C for Western blotting. In vitro, discard the culture medium of each group of cells, wash it with aseptic phosphate buffer, and harvest the cells with cell scrapers on ice. After extracting total protein from the flap tissues and cells with RIPA lysis buffer, protein concentration was measured by using the BCA protein assay kit. An equal amount of 60 μg tissue protein (30 μg cellular protein) was separated by 10-15% gel electrophoresis and transferred to polyvinylidene difluoride membranes (Roche Applied Science, Indianapolis, IN, USA). After blocking with 5% nonfat milk for 2 hours at room temperature, the membranes were incubated with the subsequent primary antibodies at 2.16. MDA and Carbonylated Protein Measurements. Cellular lipid peroxidation in each group was measured by using the Lipid Peroxidation MDA Assay Kit (Beyotime, China). Carbonylated protein level was measured by using the Micro Protein Carbonyl Assay Kit (Solarbio, China). All measurements were performed according to the manufacturer's instructions.

Catalase (CAT) and Glutathione (GSH) Measurements.
Catalase was measured by using the CAT Assay Kit (Solarbio, China), and GSH was measured by using the GSH Assay Kit (Solarbio, China) according to the manufacturer's instructions.

IL-1 and TNFα Measurements.
HUVECs were homogenized by using RIPA buffer, and proteins were collected. IL-1 and TNFα levels inside HUVEC cells were measured by using, respectively, the IL-1 ELISA kit (Boyun, China) and TNFα ELISA kit (Boyun, China) according to the manufacturer's instructions.

Cell Migration and Tube Formation.
Transwell assays (8 μm) (Corning, USA) were used to evaluate the HUVECs' migration ability. Briefly, 2 × 10 5 cells were seeded into the upper chamber in a volume of 200 μl, while the lower chamber contained glucose-free DMEM culture medium (Gibco) with 1% fetal bovine serum (FBS, Gibco) and different components (saline, apelin13, ML221, STO-609, DM, and BX795); then, cells were placed in a hypoxia chamber. Cells treated with RPMI 1640 medium (Meilunbio) with 1% FBS and placed in a normoxic incubator were used as a normal control group. 4 hours after incubation, cells on the upper surface of the upper chamber membrane were carefully wiped off with cotton swabs, fixed in 4% paraformaldehyde for 30 min at room temperature, and then stained with crystal violet to quantify the migrated cells. The migrated cells were pictured using an inverted microscope (Nikon, Japan). A tube formation assay on Matrigel (BD Biosciences, USA) was performed to evaluate the HUVECs' morphogenesis and tube formation capacity. Briefly, Matrigel solution was thawed at 4°C overnight and placed in a μ-Slide (10 μl per well, IBIDI, Germany) in a cell incubator for 30 min to solidify. A total of 5000 cells, which were treated with serum-free and glucose-free DMEM culture medium containing different components (saline, apelin13, ML221, STO-609, DM, and BX795), were seeded onto the Matrigel surface and placed in a hypoxia chamber, and cells treated with RPMI 1640 medium with 10% FBS and placed in a normoxic incubator were used as the normal control. Tube formation was 4 Oxidative Medicine and Cellular Longevity   2.20. Statistical Analyses. Statistical analyses and statistical graphs were implemented using GraphPad Prism 8.0 (GraphPad Software, La Jolla, CA, USA). All data are presented as mean ± SD. Comparisons of mean values between two groups were performed using the independent-sample t -test. Mean value comparisons among multigroups were performed using one-way ANOVA followed by the Tukey test. p < 0:05 was considered significant.

Apelin and APJ Were Downregulated in Random Skin
Flap Tissue. At first, we examined whether the expression of endogenous apelin and its receptor APJ would change in random skin flap tissue. Western blot analysis and immunohistochemistry staining were used to evaluate the expression of apelin and APJ in normal skin tissue and random skin flap tissue. The corresponding results indicated that both apelin and APJ were significantly decreased in random skin flap tissue at day 7 as shown in Figures 1(a) and 1(b). Similarly, compared with normal skin tissue, VEGF, CD34, SOD2, and HO-1 were downregulated synchronously as shown in Figure 1 Antioxidant enzymes such as CAT and GSH and oxidative marker carbonylated protein levels in all groups of random flaps were measured (k). n = 4 independent experiments. " * " means compared with the normal control group; "#" means compared with the random flap group; "@" means compared with the random flap+apelin13 group. * p < 0:05, * * p < 0:01.   experiments. " * " means compared with the normal control group; "#" means compared with the OGD/R group; "@" means compared with the OGD/R+apelin13 group. * p < 0:05, * * p < 0:01.

11
Oxidative Medicine and Cellular Longevity it worst. In addition, apelin protected the random skin flaps from oxidative stress caused by ischemia-reperfusion to a great extent by increasing the level of GSH and CAT; then, oxidative stress marker carbonylated protein level was relatively few (Figure 2(k)).

Apelin13
Protected the HUVECs from OGD/R. As mentioned before, we adopt OGD/R to simulate the IRI after skin flap transplantation. Immunofluorescence and Western blot showed that OGD/R successfully reproduced oxidative stress and apoptosis of the vascular endothelium in vitro. The protective effect of apelin13 on HUVECs was consistent with that in vivo. Apelin13 upregulated Nrf2, SOD2, HO1, and Bcl2, suppressed the expression of Bax, and improved the ability of antioxidant stress and antiapoptosis of HUVECs (Figures 3(a), 3(b), 3(e), and 3(g)). The DCFH-DA fluorescence probe detected a large amount of ROS production and accumulation inside vascular endothelial cells after OGD/R, and treatment with apelin13 helped suppress it and reduced the cell damage caused by ROS during OGD/R (Figures 3(c) and 3(d)). The accumulation of ROS inside cells damages mitochondrial membrane potential and hinders cell energy metabolism. The changes of mitochondrial membrane potential indirectly reflect the state of energy metabolism under oxidative stress. The red and green fluorescence directly revealed the proportion of normal mitochondria. After OGD/R stimulation, mitochondrial membrane potential dropped but apelin13 treatment maintained relatively normal membrane potential (Figures 3(c) and 3(d)). The decrease in mitochondrial membrane potential can be easily detected by the transition of JC-1 from red fluorescence to green fluorescence, which can also be used as an early detection index of apoptosis. Similar to the results of experiments in vivo, GSH and CAT levels were upregulated while carbonylated protein levels were suppressed after apelin13 treatment (Figure 3(f)); thus, the treatment of apelin13 also increased the cellular CAT and GSH and enhanced the ability of vascular endothelial cells to resist OGD/R-induced injury; on the contrary, ML221 reversed the antioxidant function of apelin13.

Apelin13 Activated AMPK/GSK3β Signaling In Vivo and
In Vitro. In vivo, compared with normal skin tissue, the content of phosphorylated and activated AMPK together with its downstream GSK3β phosphorylation ratio decreased significantly in the random flap (Figures 4(a) and 4(b)). We can see that phosphorylated AMPK remains at a high level in the apelin13-treated group, and apelin13 effectively resists the impairment of AMPK phosphorylation involved in IRI of the random flap. While suppressing APJ by ML221, the activation effect of apelin13 on AMPK and GSK3β disappeared (Figures 4(c)-4(e)). HUVECs were stimulated by OGD/R in vitro and treated with apelin13 and apelin13+ML221, and we can see that the changes of phosphorylated AMPK and phosphorylated GSK3β are consistent with those in vivo (Figures 4(f)-4(h)). AMPK and GSK3β played a significant role in protecting the random flap from IRI. This proved once again that there is good consistency between the HUVECs' OGD/R model and the skin flap's IRI.

15
Oxidative Medicine and Cellular Longevity 3.8. Apelin13 Promoted the Migration of HUVECs and Angiogenesis by CaMKK/AMPK/GSK3β Signaling. The migration of vascular endothelial cells is the foundation of angiogenesis. In our study, the Transwell experiment was used to quantify the migrated HUVECs. The tube formation experiment was used to assess the capacity of angiogenesis. When HUVECs are exposed to glucose and oxygen deprivation, both the number of migrated cells and the number of tubes formed decreased visibly. Apelin13 promoted cell migration and tube formation by combining with APJ; contrary to this, ML221 blocked APJ activation and further inhibited apelin13's promotion (Figures 8(a)-8(c)). When we introduced STO-609 into the treatment, apelin13mediated cell migration and angiogenesis were reversed (Figures 8(d)-8(f)). Similarly, DM inhibiting AMPK phosphorylation (Figures 8(g)-8(i)) and BX795 inhibiting GSK3β Ser-9 phosphorylation (Figures 8(j)-8(l)) inhibited cell migration and reduced the tube number in fields. In addition to the antioxidant stress effect mediated by apelin13, CaMK-K/AMPK/GSK3β signaling also played an important role in promoting HUVEC migration and angiogenic capacity retroactively.

Discussion
Large-area skin defects caused by severe trauma, burns, tumor resection, and ulcers often need to be repaired by a skin flap, but the occurrence of potential area necrosis limits the scope of the skin flap. Although delayed operation, pressurization, or drug intervention can improve the survival of the skin flap, the survival of the potential area is still different under different interventions and it is still difficult to guarantee the survival of potential areas. So far, there are many studies on promoting the survival of random skin flap trans-plantation. In order to enhance skin flap viability, a variety of attempts have been made in the postoperation medicine treatment [3], the biomaterial development [7], herbal extract treatment [9], and even acupuncture treatment [35]. However, in clinical application, the curative effect of these methods is not accurate due to potential toxic side effects and unclear therapeutic targets, and many theoretical treatment schemes are not feasible yet. So, it is imperative to further explore its pathogenesis and seek effective therapeutic targets. A random skin flap, lacking a vessel pedicle, and its blood supply mainly come from the regenerated microvessels 21 Oxidative Medicine and Cellular Longevity and small vessel stumps around the wound. It makes ischemia-reperfusion a definite and inevitable pathological process for a random skin flap. Our preexperiments show that the expression abundance of apelin and its receptor APJ decreased in the skin flap after operation. Our study focused on the significant loss of both the orphan receptor APJ and its ligand apelin in the random pattern skin flap. In order to clarify whether apelin plays a protective role in the survival of the skin flap, we try a short-term exogenous apelin supplementation after skin flap operation. The expected result is that the ischemic necrosis of the random skin flap will decrease after apelin treatment. In fact, after apelin treatment, the degree of edema of the flap is reduced, the necrotic area is reduced, and the blood flow signal intensity of LDBF is increased. Our results suggest the administration of exogenous apelin can improve the viability of the skin flap by enhancing the growth of vascular stumps and microangiogenesis in the ischemic area, attenuating oxidative stress and subsequent apoptosis. Fortunately, as we expected and proved, apelin does have both antioxidant stress and angiogenic effects by activating its receptor APJ. We believe that postoperative exogenous supplementation of apelin to " * " means compared with the normal control group; "#" means compared with the OGD/R group; "@" means compared with the OGD/R+apelin13 group; "&" means compared with the OGD/R+apelin13+BX795 group. * p < 0:05, * * p < 0:01.

22
Oxidative Medicine and Cellular Longevity

24
Oxidative Medicine and Cellular Longevity awaken the apelin/APJ axis of vessels is a convenient way to ensure the survival of a random flap and apelin is an effective accelerator for flap survival. It has been proven that apelin attenuates IRI in organs by a variety of mechanisms such as reducing acute injury through suppression of TGF-β1 [36], protecting against myocardial IRI by inactivating GSK3β [37]. In these studies, the levels of inflammation and oxidative stress are suppressed by apelin. Similarly, our experimental results in vivo showed that the expression of Nrf2, SOD2, and HO-1 decreased, which indirectly reflected the high oxidative stress operation, and the treatment of apelin increased the expression of Nrf2, SOD2, and HO-1 to some extent. In vitro, we stimulated HUVEC cells with OGD/R which was consistent with the ischemia/reperfusion process in vivo to simulate the pathophysiological changes in vessels. Free from any harmful stimulations, the expression of Nrf2 inside cells remained quite satisfying. Immunofluorescence showed a strong signal of Nrf2 in the nucleus. As far as we know, translocation of Nrf2 into the nucleus promotes the expression of AREdependent antioxidant genes [38]. As a result, the expression levels of SOD2 and HO-1 increased; in the meantime, the production and accumulation of ROS inside cells were not active and JC-1 staining showed that the mitochondrial membrane potential was not impaired. After OGD/R stimulation, production and accumulation of ROS increased, the content of Nrf2 decreased either at the overall level or in the nucleus, the expression of antioxidant proteins SOD2 and HO-1 was inhibited, and the mitochondrial membrane potential was damaged resulting in the energy metabolic disorder. By treating with apelin13 alone, stimulated HUVECs appeared to have incredible tolerance to OGD/R stimulation, which was manifested in many aspects such as reduction of intracellular ROS, maintenance of 25 Oxidative Medicine and Cellular Longevity intranuclear Nrf2, and a milder mitochondrial membrane potential damage. Given apelin13 treatment together with the inhibitor ML221, apelin13 can no longer activate the APJ receptor as usual, the protective effects were nearly reversed completely, and HUVECs showed a fragile side to OGD/R again.
APJ is an orphan receptor and a G protein-coupled receptor. Apelin13 binds to the APJ receptor and activates it, and the activation of its Gα subunits (including Gαi and Gαq) can activate downstream CaMKK/AMPK/GSK3β signaling, then promotes the translocation of Nrf2 into the nucleus, upregulates ARE-dependent antioxidants SOD, NQO1, and HO-1, suppresses oxidative stress, and protects brain and PC12 cells [39]. When the skin tissue is in a normal and healthy state, normal activation of apelin/APJ ensures a proper phosphorylation ratio of AMPK and GSK3β which declined in a random flap. As shown in Figure 9, similar to the protective mechanism of the apelin/APJ axis in the brain, treatment with apelin13 restores the phosphorylation ratio of AMPK and GSK3β greatly, which means that AMPK/GSK3β signaling also exists as a potential protective mechanism in the skin flap in the same way in the brain.
As we all know, the regeneration of vessels and the recovery of blood supply are the essential problems in wound healing. In order to clarify the angiogenesis mechanism, we focus on the activation of apelin/APJ-dependent CaMK-K/AMPK/GSK3β signaling and the further effect including angiogenesis and antioxidant stress on HUVECs. The selective CaMKK inhibitor STO-609, selective AMPK inhibitor dorsomorphin, and selective GSK3β phosphorylation inhib-itor BX795 were given to cells separately. BX795 can inhibit Ser9-phosphorylation of GSK3β while it cannot change the expression of total GSK3β [40]. It is worth noting that phosphorylation of GSK3β inactivates GSK3β and the maintenance of GSK3β activity increases glycogen degradation, glycolysis substrate oxidation, and lactate production, indicating decreased resistance of cells to oxidative stress [41][42][43][44]. All these three inhibitors inhibited apelin13's callback of Nrf2, SOD2, and HO-1. The similar changes of metabolite of lipid oxidation MDA and inflammatory mediators IL-1 and TNFα were observed, thus reversing apelin13's protective effect on HUVECs. These suggested that apelin13 protected HUVECs from oxidative stress through a CaMK-K/AMPK/GSK3β signaling-dependent mechanism during OGD/R.
As for the mechanism of angiogenesis, apelin, AMPK, and GSK3β have been separately reported to be associated with angiogenesis [45][46][47][48]. However, the relationship among apelin, AMPK, and GSK3β in angiogenesis has rarely been reported, especially in the environment of ischemiareperfusion. In view of the fact that apelin13 can protect HUVECs from oxidative stress in vitro, in a CaMK-K/AMPK/GSK3β-dependent manner after activating the receptor APJ, we are also interested in whether apelin13 promotes angiogenesis in this CaMKK/AMPK/GSK3βdependent manner. Therefore, the model of short-term glucose and oxygen deprivation was established in vitro, and the quantity of migrated cells in Transwell migration assays and tubules' number in tube formation assays were used to comprehensively evaluate angiogenesis capacity. When cells

26
Oxidative Medicine and Cellular Longevity were exposed to OGD, both migrated cells and tubes formed sharply decreased. Apelin13 increased migrated cells and tubes formed. Administration of inhibitors STO-609, dorsomorphin, and BX795 reversed apelin13's angiogenesis promotion; thus, inhibition of CaMKK, AMPK-phosphorylation, and Ser9-phosphorylation of GSK3β declined angiogenesis, indicating that apelin13/APJ promoted angiogenesis capacity at least through a CaMKK/AMPK/GSK3β signaling-dependent mechanism.
Regarding prospect and deficiency, our study is precisely focused on skin flap vessels. Through this study, we prove that the apelin/APJ signal axis is suppressed in postoperative skin flap vessels, and postoperative supplementation of apelin can not only help skin flap resist the IRI but also promote angiogenesis, ultimately enhancing random skin flap viability through a CaMKK/AMPK/GSK3β-dependent mechanism. Apelin is also a kind of widespread inherent polypeptides in animals, which has good compatibility with animals and can effectively reduce the risk of toxicity and side effects. With increasing clinical trials over time, apelin treatment can be expected to be a convenient and efficient clinical application in the field of wound healing. Besides, we also realize that there are still many limitations in our research, and we have yet to show whether apelin/APJ-manipulated CaMK-K/AMPK/GSK3β-dependent mechanism is the only mechanism of apelin's therapeutic work. Other therapeutic targets of apelin are not known yet. The study focuses on vessels but ignores exploring the function of other tissue components in the skin flap.

Conclusion
In conclusion, in vivo, our study proved that the apelin/APJ axis protected the skin flap by alleviating vascular oxidative stress and the apelin/APJ axis works as an antioxidant stress factor dependent on CaMKK/AMPK/GSK3β signaling. Besides, an apelin/APJ-manipulated CaMK-K/AMPK/GSK3β-dependent mechanism improves the tolerance of HUVECs to OGD/R, reduces ROS production and accumulation, maintained the normal mitochondrial membrane potential, and suppresses oxidative stress in vitro. It is worth mentioning that activation of the apelin/APJ axis promotes vascular migration and angiogenesis under relative hypoxia condition through CaMK-K/AMPK/GSK3β signaling. The awakening of the apelin/APJ axis may play a dual role in both antioxidant stress and angiogenesis in wound healing.

Data Availability
The data used to support the findings of this study are available from the corresponding author upon request.

Conflicts of Interest
The authors have no conflicts of interest.