α-Lipoic Acid-Plus Ameliorates Endothelial Injury by Inhibiting the Apoptosis Pathway Mediated by Intralysosomal Cathepsins in an In Vivo and In Vitro Endothelial Injury Model

α-Lipoic acid-plus (LAP), an amine derivative of α-lipoic acid, has been reported to protect cells from oxidative stress damage by reacting with lysosomal iron and is more powerful than desferrioxamine (DFO). However, the role of LAP in experimental carotid artery intimal injury (CAII) has not yet been well investigated. Therefore, we sought to uncover the role and potential endovascular protective mechanisms of LAP in endothelial injury. In vitro, oxyhemoglobin (OxyHb) stimulation of cultured human umbilical vein endothelial cells (HUVECs) simulated intimal injury. In vivo, balloon compression injury of the carotid artery was used to establish a rat CAII model. We found that the protein levels of cathepsin B/D, ferritin, transferrin receptor (TfR), cleaved caspase-3, and Bax increased in the injured endothelium and HUVECs but were rectified by DFO and LAP treatments, as revealed by western blotting and immunofluorescence staining. Additionally, DFO and LAP decreased oxidative stress levels and endothelial cell necrosis of the damaged endothelium. Moreover, DFO and LAP significantly ameliorated the increased oxidative stress, iron level, and lactic dehydrogenase activity of HUVECs and improved the reduced HUVEC viability induced by OxyHb. More importantly, DFO and LAP significantly reduced mitochondrial damage and were beneficial for maintaining lysosomal integrity, as indicated by acridine orange (AO), Lyso-Tracker Red, JC-1, and ATPB staining in HUVECs. Finally, LAP might offer more significant endovascular protective effects than DFO. Our data suggested that LAP exerted endovascular protective effects by inhibiting the apoptosis signaling pathway mediated by intralysosomal cathepsins by reacting with excessive iron in endothelial lysosomes after intimal injury.


Introduction
Carotid artery (CA) stenosis is a familiar and recognized risk factor for ischemic stroke, accounting for approximately 10-20% of strokes or transient ischemic attacks [1]. Although precautionary medical treatment has increased, the proportion of patients requiring surgical intervention remains high [2]. Both open operation and interventional surgery can induce different degrees of vascular injury. In brief, the repair process of blood vessels generally includes injured site reendothelialization, vascular remodeling, and intimal hyperplasia (IH) [3,4]. Excessive IH is the primary cause of failure of vascular interventional therapy and hema-dostenosis [5]. Endothelial cell (EC) injury has been identified as the first step toward postoperative IH [6] because the injured endothelial area is not capable of generating antiproliferative factors. The damaged area increases the risk of thrombus formation because of the coverage of platelets, fibrinogen, circulating red blood cells, and macrophages that release thrombotic factors and growth factors. Erythrocyte metabolites, especially the release of catalytically redoxactive iron, have toxic effects, which are often neglected by investigators [7]. Reendothelialization relying on the proliferation and migration of ECs is a key process in vascular healing. Hence, the stimulation of EC proliferation and migration is essential to accelerate endothelial healing and restore vascular function in response to damaged ECs resulting from intimal injury [8]. Delayed reendothelialization induced by EC apoptosis and degeneration is one of the principal causes of excessive IH.
Iron is essential in the synthesis of hemoglobin and participates in mitochondrial respiration. Nevertheless, excessive iron intake can lead to damage to cells, organs, and the entire body [9]. Furthermore, iron increases reactive oxygen species (ROS) production and promotes calcium influx into mitochondria, which can disturb mitochondrial respiratory function and ultimately result in a decline in mitochondrial membrane potential (ΔΨm) and mitochondrial damage. A previous study found that iron treatment significantly increased apoptotic cells and endothelial microparticles (EMPs) [10]. Iron overload is also involved in a number of cardiovascular and endothelial injury events [11]. Excessive free radical generation mediated by iron overload induces the loss of tight junction proteins and EC degeneration, leading to the opening of the blood-brain barrier (BBB) after transient forebrain ischemia [12]. Because iron induces EMP generation and EC apoptosis in association with increased oxidative stress, iron overload has been identified as a risk factor for cardiovascular events [13]. Hence, we designed this experiment to investigate the effects of redox-active iron on the endothelium in vivo, EC status in vitro, and the relevant damage mechanisms.
As a radical scavenger and an iron-chelating agent, αlipoic acid-plus (LAP) has already been confirmed to have neuroprotective effects in rat subarachnoid hemorrhage (SAH) models resulting from playing a stronger role in resisting oxidative stress by chelating iron than desferrioxamine (DFO). It has been demonstrated in the literature that LAP with a weak base (pKa = 8:0) can more easily enter lysosomes (pH = 4:6-5.0) through the differences in pH between the cytoplasm and lysozymes. Furthermore, the reduced forms of LAP (DHLAP) can provide sulfhydryl to react with redox-active iron [14,15]. The structures of DFO, LAP, and DHLAP are shown in Figures 1(a) and 1(b). Although LAP has been shown to protect neuronal cells from oxidative stress injury triggered by labile iron in vivo and alleviate early brain injury after SAH, it has not been fully investigated whether LAP could inhibit endovascular injury induced by balloon compression, and the underlying mechanisms of action are unknown. Therefore, in this study, we explored the role of LAP in the reduction of EC apoptosis through the regulation of oxidative stress to identify novel medication for treating carotid artery intimal injury (CAII).

Methods
2.1. Experimental Animals. The experiment was approved by the Ethics Committee of the First Affiliated Hospital of the University of Science and Technology of China and was conducted in accordance with the National Institutes of Health regulations on animal feeding and application. Adult male Sprague-Dawley (SD) rats weighing approximately 300 g were supplied by Zhaoyan New Drug Research Center (Suzhou, China) Co., Ltd. The animals were housed in a constant humidity and temperature environment and were fed food and water regularly.

Experimental Designs and Drug
Interventions. Experiment 1 was designed to demonstrate the role of cathepsin B/D in the endothelium following CAII. In experiment 1, 54 male adult rats were stochastically divided into nine point-in-time groups (n = 6): sham group; 6-hour, 12-hour, 24-hour, 48-hour, and 72-hour group; and 1-week, 2-week, and 4-week CAII groups. The CAs were collected from sham and CAII rats at various time points for western blotting and immunofluorescence (IF) assays. The experimental procedure is outlined in Figure 1(c). Experiment 2 was designed to explore the mechanisms of LAP in alleviating endothelial injury induced by CAII. In experiment 2, 84 male adult SD rats were stochastically divided into seven groups (n = 12): sham, CAII, CAII+vehicle, CAII+DFO (25 mg/kg), and CAII+LAP (100 mg/kg, 150 mg/kg, and 200 mg/kg). The animals were treated with LAP mixed with 0.5% methylcellulose through oral administration. DFO was administered through an intraperitoneal injection. DFO and LAP were administered 4 h after CAII induction, and both treatments were continued two more times. The doses of DFO and LAP were based on previous studies [15,16]. Six rat CAs in every group were cut into slices and used for IF and Fluoro-Jade B (FJB) staining analysis. The remaining rats were sacrificed and perfused, and then CA samples were cut for western blotting and oxidative stress evaluation. The experimental procedure is displayed in Figure 1(d). Experiment 3 was designed to explore the roles and mechanisms of LAP in alleviating human umbilical vein endothelial cell (HUVEC) injury induced by oxyhemoglobin (OxyHb) in vitro. In experiment 3, logarithmically growing HUVECs were divided into the control, OxyHb, and OxyHb (Iron, DFO, LAP-L, LAP-M, and LAP-H) groups. HUVECs were exposed to OxyHb (10 μM) and treated with DFO (1 mM) or three concentrations of LAP (0.2 μM, 0.3 μM, and 0.4 μM) for 24 h before performing the subsequent assays [14]. Following the treatments, living HUVECs were first collected for cell viability assays, lactate dehydrogenase assay (LDH), acridine orange (AO) staining, live-dead cell staining, Lyso-Tracker Red, measurement of mitochondrial membrane potential (MMP), and oxidative stress evaluation. Second, HUVECs were fixed with paraformaldehyde for IF differential staining. Total protein from HUVECs was collected for western blotting assays. The specific experimental procedures are displayed in Figure 1(e).
2.3. Establishment of Rat CA Balloon Injury Models. Rat CAII models were established based on the approach described in our previous study [17]. In brief, the rats were primarily fixed in a specific head frame, and the CAs were thoroughly dissected. Then, a catheter with a balloon (Medtronic Inc., Minneapolis, MN, USA) was inserted from the gap of the external CA and slipped into the common CA. The balloon was inflated to approximately 2 atm, and the injury was induced by rubbing the inner surface of the common CA back and forth three times. Then, the catheter and balloon were withdrawn, and the injured external CA was   Oxidative Medicine and Cellular Longevity sutured tightly under the operating microscope. The 2 cm common CA injured region was cut for analysis.

Cell Culture and Treatment.
We performed the procedures for HUVEC extraction and culture in accordance with the previous literature [18]. To assess the impact of LAP on injured HUVECs in vitro, the HUVECs were exposed to OxyHb, DFO, and LAP at a gradient concentration for 24 h before performing the subsequent experiments.
2.9. FJB Staining. FJB is a hypersensitive and specific fluorescent dye used to indicate HUVEC degradation, and the staining process was based on a previous literature [23]. Briefly, the CA sections were immersed in 0.06% potassium permanganate (KMnO 4 ) solution shielded from light at room temperature for 15 min after dewaxing. Next, the sections were added and reacted with FJB working solution (in 0.1% acetic acid solvent) (AG310, Sigma-Aldrich, St. Louis, MO, USA) for 1 h after being washed. Then, the sections were air-dried in a fume hood. Finally, the stained sections were sealed with neutral balsam mounting medium and observed under a fluorescence microscope, and images were taken in parallel to count FJB-positive cells.
2.10. Cell Viability Assays. First, the treated HUVECs were fixed in 50% trichloroacetic acid and stained with a   Oxidative Medicine and Cellular Longevity sulforhodamine B (SRB) solution (230162, Sigma-Aldrich) as described previously [24]. Next, the absorbance of SRB was measured at a wavelength of 565 nm using a microplate reader. The assay was performed in triplicate and repeated at least three times independently. Finally, we conducted fluorescence detection after discarding the staining solution and washing it with culture medium two or three times. When cell apoptosis occurred, the nuclei of apoptotic cells were observed with dense staining or fragmented dense staining under a fluorescence microscope.
2.14. AO Staining. HUVECs at the logarithmic growth stage were plated into 24-well plates and cultured for 24 h. Then, the original culture medium was removed from each well, and 1 mL of culture medium containing OxyHb, DFO, and LAP was added to each well at the corresponding    2.15. Lyso-Tracker Red. For lysosomal staining, 1 μL of Lyso-Tracker Red solution was added to 20 mL of cell culture medium and mixed to form Lyso-Tracker Red working solution. Then, the cell culture medium was removed, and Lyso-Tracker Red working solution preincubated at 37°C was added. Next, the cells were incubated for 30-120 min at 37°C. Finally, the Lyso-Tracker Red working solution was removed, and a new cell culture solution was added. The lysosomes in the cytoplasm were stained with bright and intense fluorescence under fluorescence microscopy.
2.16. MMP Measurement. The MMP was detected by a JC-1 kit (Beyotime) based on the instruction manual [26]. First, the treated HUVECs cultured in 12-well plates were washed with PBS; then, JC-1 working solution was added and reacted in a dark room for 20 min. Finally, the cultured HUVECs were observed under a fluorescence microscope. The JC-1 fluorescence intensity ratio (red/green) was calculated for the MMP in HUVECs.
2.17. Statistical Analysis. GraphPad Prism 7.0 software (San Diego, CA, USA) was applied for data processing and analysis. The data are shown as the mean ± standard error of the mean (SEM). One-way or two-way analysis of variance (ANOVA) was applied for multiple comparisons. Tukey's post hoc tests were used for comparisons between two pairs in multiple groups. P < 0:05 was considered statistically significant.

Cathepsin B/D Is Activated following CAII.
To detect changes in cathepsin B/D expression levels after CAII, we performed western blotting and IF staining. Western blotting results showed that cathepsin B/D expression levels increased after CAII, attained the highest peak at 48 h and then gradually recovered within 4 weeks (Figures 2(a)-2(d)) ( * * P = 0:0055, * * * P = 0:0007, and * P = 0:0374). IF also showed that the immunopositivity of cathepsin B/D increased at each time point after SAH in comparison with the sham group (Figures 2(e)-2(h)) ( * * P = 0:0021, * * P = 0:0020). The previous results indicated that cathepsins B and D may participate in the pathological process of CAII and are activated following CAII. Furthermore, 48 h was regarded as an optimal intervention point in further studies.

LAP Rescues Damaged HUVECs In Vitro.
Live-dead cellular staining was applied to assess the survival rate of   (Figures 6(a) and 6(b)). The staining for live (green) and dead (red) cells indicated that OxyHb intervention significantly decreased the survival rate of HUVECs ( * * * P = 0:0004). However, the combined iron and OxyHb treatment group had a lower survival rate than the OxyHb group ( ## P = 0:0019). In contrast, the OxyHb+LAP-treated group had a higher survival rate than the OxyHb group ( ## P = 0:0023, ### P = 0:0007, and ### P = 0:0005). Meanwhile, the medium and high concentrations of LAP showed a more significant protective effect than DFO ( $ P = 0:0133, $$ P = 0:0015). The SRB assay was used to measure the viability of cells (Figure 6(c)). The results showed that OxyHb significantly decreased the viability of HUVECs ( * * * P = 0:0003), and iron treatment further decreased the viability of HUVECs after OxyHb treatment ( ## P = 0:0018). The cell viability in the OxyHb+LAP-treated group was markedly higher than that in the OxyHb group ( ### P = 0:0003, ### P = 0:0002). It was apparent that a high concentration of LAP further increased the viability of HUVECs compared with DFO ( $ P = 0:0230). Consistently, the LDH assay was used to measure the damage to HUVECs (Figure 6(d)). OxyHb increased the activity of LDH in comparison with the control group ( * * * P = 0:0005), and iron treatment further decreased the activity of LDH after OxyHb treatment ( ### P = 0:0003). The activity of LDH in the OxyHb+LAPtreated group markedly decreased ( # P = 0:0476, ## P = 0:0026, and ### P = 0:0001) compared with that in the OxyHb group. The high concentration of LAP further decreased the activity of LDH compared with DFO ( $$ P = 0:0052). Moreover, Hoechst staining was used to measure HUVEC apoptosis (Figure 6(e)). OxyHb increased HUVEC apoptosis in comparison with the control group, and iron treatment further increased HUVEC apoptosis under OxyHb treatment, whereas DFO and different concentrations of LAP obviously decreased HUVEC apoptosis. In addition, FJB staining showed almost no FJB-positive cells in the sham group (Figure 6(f)), whereas the number of FJBpositive cells in the CAII group was significantly higher than that in the sham group. Compared with the CAII+vehicle group, the number of FJB-positive cells was significantly attenuated by DFO and LAP. Together, these results suggested that LAP had more significant vascular protective effects against DFO.
3.6. LAP Promotes Mitochondrial Transport and Distribution in Damaged HUVECs. The IF intensity of the mitochondrial marker ATPB was used to evaluate mitochondrial survival in HUVECs damaged by OxyHb stimulation (Figures 7(a) and 7(c)). The results showed that OxyHb significantly decreased the IF intensity of ATPB ( * * * P = 0:0005). Iron treatment further decreased the IF intensity of ATPB after OxyHb treatment ( # P = 0:0468). Compared with that in the OxyHb group, the immunofluorescence intensity of ATPB in the OxyHb+LAP-treated group was markedly increased ( # P = 0:0149, ### P = 0:0009). The high concentration of LAP further increased ATPB expression in HUVECs compared with DFO ( $ P = 0:0395). To study the effect of LAP on mitochondrial damage, we stained HUVECs in vitro with JC-1. In healthy mitochondria of HUVECs, JC-1 was shown by red fluorescence, whereas JC-1 transformed into green fluorescence in damaged mitochondria of HUVECs (Figures 7(b) and 7(d)). The results revealed that the percentage of green fluorescence-positive signals in the OxyHb group was brighter than that in the control group ( * * * P = 0:0002), whereas iron treatment further increased the percentage of green fluorescence-positive signals ( # P = 0:0409). In contrast, in the OxyHb+DFO-and OxyHb+LAP-treated groups, the percentage of green-fluorescent-positive signal was lower than that in the OxyHb group ( # P = 0:0134, ### P = 0:0005, ### P = 0:0003, and ### P = 0:0002). We also found that medium and high concentrations of LAP decreased the percentage of green-fluorescent-positive signals compared with DFO ( $$ P = 0:0046, $$$ P = 0:0004).

DFO and Different Concentrations of LAP Alleviate
Ferritin and TfR Expression Levels and Iron Deposition following Endothelial and HUVEC Injury. Western blotting was applied to assess ferritin and TfR expression in HUVECs after treatment with DFO and different concentrations of LAP (Figures 8(a)-8(f)). The results showed that ferritin ( * * * P = 0:0006, * * * P = 0:0004, and * P = 0:0362) and TfR ( * * * P = 0:0002, * * * P = 0:0002, and * P = 0:0423) expression levels in the CAII and OxyHb groups increased relative to the sham and control groups, respectively. Iron treatment further increased the levels of ferritin ( # P = 0.0153) and TfR ( # P = 0:0392) in HUVECs. However, DFO and LAP decreased the levels of ferritin ( ### P = 0:0008, ### P = 0:0005; . We performed Prussian blue reactions to assess the iron content in HUVECs (Figure 8(k)). In the control group, little iron deposition was observed in the HUVECs, whereas in the OxyHb group, significant iron deposition was found. Compared with the OxyHb group, iron deposition in HUVECs was markedly increased by iron addition and reduced by DFO and oral administration of LAP. In addition, the high concentration of LAP presumably   3.8. LAP Protects Lysosomes from Rupture. To evaluate the quantity and state of lysosomes in HUVECs, we performed IF staining with lysosomal associated membrane protein 1 (LAMP-1) attached to different groups (Figure 9(a)). We observed a distinct reduction in LAMP-1 expression in the OxyHb group, and iron treatment further reduced LAMP-1 expression. In contrast, DFO and different concentrations of LAP remarkably inhibited lysosome rupture based on the alteration of LAMP-1 fluorescence intensity in HUVECs. However, the high concentration of LAP possibly exerted a more positive impact on stabilizing the lysosomal membrane than low concentrations of LAP and DFO. As a preliminary indicator of the lysosomal state, the acidic compartments in HUVECs were observed by Lyso-Tracker Red staining and AO staining. As shown in Figures 9(b) and 9(c), the accumulation of acidic compartments in HUVECs in the control group was normal. The accumulation of acidic compartments in HUVECs significantly decreased compared with that in the control group following OxyHb stimulation and iron treatment. The accumulation of acidic compartments in HUVECs, however, significantly increased after DFO and LAP treatments.

Discussion
In this study, we found that LAP exerts an endovascular protective effect by reacting with excess iron in endothelial lysosomes to inhibit the intimal injury-mediated apoptotic signaling pathway induced by intralysosomal cathepsins (Figure 10). As a lysosomotropic and iron-chelating agent, LAP can target and gather in lysosomes and inhibit the Fenton reaction by reacting with active iron. Hence, LAP inhibited the apoptosis pathway mediated by mitochondria by reducing the generation of hydroxyl radicals, stabilizing the lysosome membrane, and decreasing the release of cathepsins. Ultimately, LAP reduced EC apoptosis and injury, promoted reendothelialization of the damaged area, inhibited excessive proliferation of the vascular intima, and reduced the rate of vascular restenosis. Vascular endothelium damage caused by iron overload in association with oxidative stress and the underlying mechanisms have not yet been clarified. Prolonged exposure to iron enhances endothelial nicotinamide adenine dinucleotide