Study On The Modulation Of Macroautophagy And The Autophagy-promoting Role Of ANXA7 In Vascular Endothelial Cells

Macroautophagy (hereafter referred to as autophagy) is an evolutionarily conserved lysosomal degradation process in eukaryotes. Cytoplastic constituents, including long-lived proteins and superfluous or aberrant organelles, are enwrapped by double-membrane vacuolar structures called autophagosomes which may fuse with lysosomes or endosomes. The product of autophagosome and lysosome fusion is autophagolysosome, in which the cargo and the inner membrane are degraded by acid lysosomal hydrolase enzymes. The degradation products of aminal acids, monosaccharides and nucleotides could be reused by cells for nutrients recycling and organelles turnover. This process not only provides nutrients for vital cellular functions, but confers survival advantage to cells undergoing microorganism invasion, protein misfolding, endoplasmic reticulum (ER) stress or oxidative stress. In contrast, autophagy has been more recently implicated in cell death; termed type II programmed cell death (autophagic cell death). It facilitates to protect the cellular inner surrounding to eliminate the harmful elements, and to maintain and/or refresh the organisms. A growing body of evidence proved that autophagy is involved in various diseases, such as neurodegenerative disease, cancer, atherosclerosis, and aging. Thus, exploiting the molecular mechanisms of autophagy may undoubtedly have pathophysiological implications for development of novel clinical strategy.Vascular endothelial cells (VECs) are a subset of cells which connect with each other tightly to form a single cell layer, covering the intima surface of the vessel wall. This interface validly separates blood from the vascular wall. VECs play pivotal roles in signal transduction, endocrine, coagulation, inflammation, and angiogenesis. Maintenance of normal functions of endothelial cells may have clinical significance and implications in that various fetal diseases which are threatening the public health, such as hypertension, atherosclerosis, heart failure, and cerebral stroke, attribute to endothelial dysfunction. Up to now, accumulating studies focus on the role of autophagy and its modulation in normal endothelial functions. Thus, what is the underlying molecular mechanism of endothelial autophagy? Is there molecular switch of endothelial autophagy? If so, whether it participates in canonical signal pathways of autophagy? All these need further investigations.Annexins are Ca2+-dependent, negatively charged phospholipid-binding proteins that form an evolutionarily conserved family universally expressed throughout animal and plant kingdoms. Annexins participate in membrane traffic and a series of calcium dependant activities, including vesicle transport, membrane fusion in exocytosis, signal transduction and calcium channel formation, inflammation, cell differentiation, and cell skeleton protein interaction. There are approximately 20 members in annexin family, dividing into 5 groups:A, B, C, D, and E. Annexins harbor a conserved Core domain and a variable N-terminal domain that appears to perform its unique funcitons. The annexin core domain comprises four (in annexin A6 eight) consensus sequences of homology. It spontaneously forms a curvature of a-helix disk of 70～80 amino acids. NH2-terminal domain of annexins has the calcium-binding site and phosphorylation site generally occurred in the calcium-binding protein S100 family. Tumor suppresser gene annexin A7 (ANXA7) is located on chromosome 10q21, revealing two isoforms (47 and 51 kD) of ANXA7 after alternative splicing. Under normal conditions, ANXA7 can bind with lipid vesicles and at low Ca2+ levels phosphorylated ANXA7 provokes fusion of lipid vesicles. Lamellar body and plasma membrane fusion are necessary for lung surfactant secretion in alveolar type II cells. The signaling lipids, diacylglycerol (DAG) and phosphatidylinositol-4,5-biphosphate (PIP2), play key roles in regulating the membrane fusion activity of ANXA7. Co-incubation of isolated lung lamellar bodies and phosphatidylinositol phospholipase C (PI-PLC) or phosphatidylcholine phospholipase C (PC-PLC) may enhance the ANXA7 fusion activity. This study aims to exploit the modulation of autophagy in vascular endothelial cells using chemical genetics and molecular biology methods, to discuss the role of ANXA7 in endothelial autophagy, and to uncover the molecular mechanisms of ANXA7-induced autophagy. Thereby it may deepen the knowledge of endothelial autophagy, provide novel effective tools and even scientific theory basis for development of clinical strategies.STUDY CONTENTS1. Cadmium (Cd2+) induced autophagy and alleviated apoptosis in vascular endothelial cells under bovine calf serum and bFGF withdrawal condition.2. The molecular mechanisms of cadmium-induced autophagy.3. Benzoxazine derivative ABO evoked autophagy in normally cultured vascular endothelial cells.4. Annexin A7 (ANXA7) participates in ABO-evoked autophagy. 5. The underlying molecular mechanisms of ABO-induced autophagy.METHODS1. The isolation and culture of HUVECs:referred to Jaffe et al. (Jaffe et al.,1973).2. Measurement of autophagy and autophagic flux:2.1 The quantity and intracellular distribution of acidic vesicles were assessed by acridine orange (AO) staining with fluorescent microscope.2.2 The relative quantity of acid vesicles in cells was quantified by FACS (Fluorescence Activated Cell Sorter).2.3 The intracellular quantity and distribution of autophagy marker MAPI LC3 were analyzed by the method of immunohistochemical staining.2.4 The protein levels of MAPI LC3-Ⅱand p62 were evaluated to estimate the number of autophagosomes and autophagic flux.2.5 To detect autophagy and autophagic flux with rapamycin.2.6 To detect autophagy and autophagic flux with autophagy inhibitors 3-MA and bafilomycin A1.2.7 Exploring the phosphorylation state of p70S6K and 4EBP1 (mTOR substrates) using Western blot method.3. Measurement of apoptotic cell death:3.1 The changes of cell morphology were observed under phase contrast microscope.3.2 Nuclear fragmentation was analyzed by acridine orange (AO) staining with fluorescent microscope.3.3 Cell apoptotic rate was evaluated by TUNEL assay.4. Phosphatidylcholine phospholipase C (PC-PLC):PC-PLC activity assay was performed as described before (Wu et al.,1997).5. RNA interference (RNAi):Specific siRNAs against ANXA7 (ANXA7-HSS100516, HSS100517, HSS100518 [3 RNAi], 5'-UUGAUAGAGACGCUGAGCAUCUUCC-3', 5'-GGAAGAUGCUCAGCGUCUCUAUC AA-3', 5'-UGCCUCUGAUCAUUGGAACGGUUGG-3') were synthesized by Invitrogen. Cells at 80% confluence were transfected with scramble RNA (negative control) or siRNA against ANXA7 (10-40 nM) by RNAiFect Transfection Reagent according to the manufacturer's instructions (QIAGEN,1022076). After 8 h, siRNA in the medium was substituted for normal M199 medium with bFGF. Cells were cultured to 24 h for the desired assays. siRNAs from 10 to 40 nM downregulated ANXA7 in a concentration-dependent manner. In subsequent experiments,40 nM siRNA (ANXA7-HSS100516) for 24 h was employed. Western blot analysis was used to estimate the effect of gene silencing.6. Protein over-expression in eukaryotes:Transient transfection was performed with LipofectamineTM 2000 transfection reagent according to the manufacturer's instructions. Cells were transiently transfected with pCMV6-XL5 or pCMV6-XL5-ANXA7 cDNA for 4-6 h and mormally cultured to 24 h. Then total proteins were extracted for western blotting assay. The transfection effect of pCMV6-XL5-ANXA7 plasmid into HUVECs was confirmed by western blotting or immunohistochemical assay.7. Measurement of ROS level:ROS level was analyzed by the fluorescent probe DCHF. 8. Measurement of mitochondrial membrane potential (MMP):MMP was measured by fluorescence probe JC-1.9. Analysis of expression and distribution of proteins:9.1 The intracellualr distribution of MAP1 LC3, ANXA7, integrinβ4, or caveolin-1 was analyzed by immunocytochemistry.9.2 The mRNA levels of ANXA7 and GAPDH were analyzed by RT-PCR method.9.3 The protein levels of LC3, p62, p70s6k, p-p70s6k, ANXA7,4EBP1, p-4EBP1 GAPDH andβ-actin were evaluated by western blot assay.RESULTS:1. Low concentrations of cadmium stimulated endothelial autophagy and inhibited cell apoptosis under serum and bFGF deprivation condition.1.1 A novel Cd2+-selective sensor suitable for living cells was recently synthesized. It can distinguish Cd2+ from Zn2+ and many other metal ions with both emission shift and fluorescent intensity. To detect the transportation of Cd2+ ions into cytoplasm, we used a novel Cd2+-selective fluorescent chemosensor to image Cd2+ in living HUVECs. After cells were exposed to 1,5 and 10μmol/L Cd2+ for 30 min, markedly enhanced fluorescence intensity was visualized compared with sensor group in which HUVECs were treated with sensor alone. Although certain amount of fluorescence presented in sensor group, the variation of fluorescence intensity exhibits in a dose dependent manner (Figure 1-1).1.2 Phase contrast graphs showed that low concentrations of Cd2+ reduced cell detachment which was obviously visualized in control group (Figure 1-2).1.3 TUNEL assay is a reliable test for detecting apoptosis. We observed that Cd2+ at low concentrations could inhibit apoptosis induced by deprivation of serum and bFGF. Cadmium treatment reduced the TUNEL-positive cells, but 10μM Cd2+ showed lower effect (Figure 1-3).1.4 Acridine orange staining was performed to detect autophagy. The cells in control group after AO staining displayed green fluorescence in cytoplasm and nucleolus, but displayed considerable red fluorescent dots in cytoplasm of Cd2+-treated cells, suggesting the formation of acidic autophagic vacuoles (Figure 1-4).1.5 Microtubule-associated protein 1 light chain 3-1 (MAPI LC3-Ⅰ) localizes in cytoplasm and phosphatidylethanolamine (PE)-conjugated LC3-Ⅱon the membrane of autophagic vacuoles. Conversion of soluble 18 kD LC3-Ⅰto membrane-bound,16 kD LC3-Ⅱ, assessed by Western blot assay, represents hallmark of autophagy. Western blotting showed that LC3 processing, namely increased ratio of LC3-Ⅱ/β-actin was markedly enhanced after Cd2+ treatment (Figure 1-5).2. Molecular mechanisms of cadmium-induced autophagy.2.1 HUVECs were treated with indicated concentrations of Cd2+ for 4 h. Cells were stained with fluorescent probe DCHF for ROS detection. Data showed that ROS levels were evidently increased compared with control (Figure 1-6).2.2 HUVECs were treated with indicated concentrations of Cd2+ for 4 h. Then integrinβ4, which localizes in glycosphingolipid (GSL)-enriched domains, was imaged by the method of immunofluorescence staining. Our data indicated that level of integrinβ4 was markedly decreased in HUVECs in a dose dependent manner. The relative levels of integrinβ4 were 72.9%,62.7% and 49.0% in 1,5 and 10μM Cd2+ groups, respectively (presumed as 100% in the control group, Figure 1-7).2.3 Caveolin-1 is marker protein of caveolae. To elucidate the involvement of caveolae after Cd2+ exposure, caveolin-1 was visualized with immunohistochemical method. Different concentrations of Cd2+ depressed caveolin-1 level. Caveolin-1 showed a minimum value of 37.2% in 5μM Cd2+-treated group and it was 55.7% in 10μM Cd2+ group (Figure 1-8).2.4 After 4 h of Cd2+ treatment, HUVECs were harvested and cell lysates were employed for PC-PLC assay. PC-PLC activities, which appeared in similar variation tendency to caveolin-1 level, were sharply attenuated after cadmium exposure,In 5μM group PC-PLC activity was 53.8% of that in control group and it was 64.7% in 10μM Cd2+-treated group (Figure 1-9).3. Benzoxazine derivative ABO promoted autophagy in normally cultured vascular endothelial cells. 3.1 VECs were challenged with ABO (25-100μmol/1) for 24 h. AO staining assay showed that the number of acidic vacuoles increased.3-Methyladenine (3-MA) is an inhibitor of autophagy.3-MA inhibited the effect of ABO. Flow cytometry (FACS) was employed to detect AO flurescence intensity, which was enhanced from about 3% in control to above 50% in ABO treated group (Figure 2-2).3.2 Electronic microscopy was used for visualization of autophagic vacuoles. Double membrane-bound structures of autophagic vacuoles were detected in 50μmol/1 ABO group, whereas autophagic vacuole profiles per cell area decreased in ABO/3 MA group (Figure 2-3).3.3 Distribution of autophagic protein LC3 was detected with immunostaining assay. When challenged with ABO (25-100μmol/1), cells stained with Anti-LC3 antibody showed distinctively punctate compared to the diffuse staining pattern in non-treated cells. The class III PI3K inhibitor 3-Methyladenine (3-MA) attenuated the formation of LC3-positive structures evoked by ABO (Figure 2-4).3.4 LC3 processing and LC3-II accumulation in cytoplasm represent a hallmark of autophagy. The PE-conjugated, autophagic LC3-II levels were determined by Western blot analysis. ABO (25-100μmol/1) improved a prominent LC3-I to LC3-II shift, which was restrained by the specific autophagy inhibitor 3-MA. P62 degradation is commonly used for characterizing the increased autophagic flux. After ABO challenge, the p62 levels were decreased but restored by 3-MA (Figure 2-5).3.5 ABO (50μmol/1) treatment for 6,12, and 24 h stimulated LC3 processing and LC3-II accumulation and rapamycin was used as positive control. LC3-II level increased in a time-dependent manner (Figure 2-6).3.6 After challenged with ABO (25-100μmol/1) for 24 h, cell lysates were used for Western blot assay. Data indicated that beclinl protein level was upregulated by ABO, which could be reversed by autophagy inhibitor 3-MA compared with control (Figure 2-7).3.7 Bafilomycin A1, an inhibitor of vacuolar H+ ATPase (V-ATPase), prevents fusion of autophagosomes with lysosomes. In our study 100 nmol/1 bafilomycin A1 (saturated concentration for HUVECs) increased LC3-Ⅱlevel and cotreatment of bafilomycin A1 and ABO accumulated LC3-II more remarkably, implying that ABO induces autophagosome formation instead of blockade of autophagosome and lysosome fusion (Figure 2-8).4. The role of annexin A7 in ABO-induced autophagy.4.1 After challenged with ABO (25-100μmol/1) for 24 h, cell lysates were used for Western blot assay. The protein level of ANXA7 was dose-dependently heightened after ABO treatment compared with vehicle control (Figure 2-9A). RT-PCR data showed that ANXA7 mRNA level was elevated compared with control (Figure 2-9B). ABO (50μmol/1) treatment for 6,12 and 24 h elevated ANXA7 in a time dependent manner (Figure 2-9C). Immunostaining assay showed that the ABO-treated cells also exhibited higher fluorescence intensity of ANXA7 per cell in a noticeable punctate pattern (Figure 2-9D).4.2 VECs were incubated with scramble RNA, or 10,20,40 nmol/1 siRNA against ANXA7 for 6 h, and cultured to 24,48, or 72 h. Immunostaining results illustrated that ANXA7 level was evidently decreased (Figure 2-10A). HUVECs were incubated with scramble RNA, or 10,20,40 noml/1 siRNA for ANXA7 for 6 h, and cultured to 24 h. Cell lysates were used for immunoblotting.40 nmol/1 siRNA for 24 h was the most effective and chosen for the further experiments (Figure 2-10B).4.3 To further evaluate whether ANXA7 is necessary in autophagy, we detected the localization of ANXA7 and LC3 by double immunohistochemical staining. The micrographs showed that they co-located in a punctate pattern in the cytoplasm of ABO-treated cells but were ubiquitous in control cells, suggesting a putative role for ANXA7 in ABO-induced autophagy. Immunostaining of LC3 demonstrated that autophagy was impaired in ANXA7-deficient cells. Wild-type HUVECs were faintly labeled with anti-LC3 antibody in a diffuse pattern and showed a clear induction of autophagic vacuole formation after ABO treatment (Figure 2-11).4.4 After ANXA7 was downregulated, cells were treated with different concentrations of ABO for 24 h and stained with Acridine Orange.40nmol/l ANXA7 siRNA attenuated accumulation of orange-red compartments (Figure 2-12).4.5 Western blot analysis combined RNA interference technique was applied in ABO-induced autophagy. In cells treated with non-silencing (scramble, Scr) small interfering RNA, ABO promoted the accumulation of autophagic LC3-Ⅱ,whereas in ANXA7-knockdown cells the effect was markedly suppressed. Western blot analysis showed lowered level of LC3-II in ANXA7-deficient (siA7) cells compared with scramble siRNA (Scr) group even under ABO stimulus (Figure 2-13).4.6 We overexpressed ANXA7 in HUVECs to examine whether it is sufficient to induce autophagy. The plasmid pCMV6-XL5-ANXA7, containing the full length ANXA7 cDNA, was transfected into HUVECs using lipofectamineTM 2000 transfection reagent. Compared with control and the empty vector pCMV6-XL5, HUVECs transfected with pCMV6-XL5-A7 showed higher level of ANXA7. Interestingly, overexpression of ANXA7 is available to stimulate autophagy according to LC3-II accumulation (Figure 2-14).5. ANXA7-regulated autophagy is NOT involved in the mTOR signal pathway.5.1 To understand the underlying molecular mechanisms, we investigated the phosphorylation of p70S6K and 4EBP1. ABO (25-100μmol/1) did not affect the phosphorylation of p70S6K and 4EBP1. In the meantime, we forwarded these assays with 50μmol/1 ABO for 6,12, and 24 h. The phosphorylation of p70S6K and 4EBP1 also remained unchanged (Figure 2-15). These data revealed that ABO induced autophagy in mTOR-independent manner.5.2 We then evaluated the role of ANXA7 in rapamycin-induced autophagy. siRNA targeted to ANXA7 effectively reduced the level of ANXA7. Inhibition of mTOR kinase by rapamycin induced autophagy in HUVECs with increased level of LC3-II and decreased level of p62. Knockdown of ANXA7 (40 nM siANXA7 for 24 h) could not suppress the rapamycin-induced autophagy, indicating that ANXA7 is not involved in mTOR signaling (Figure 2-16).6. ROS and MMP levels remained unchanged in ABO-stimulated autophagy.6.1 After challenged with ABO (25-100μmol/1) for 24 h, cells were stained with DCHF fluorescent probe for ROS detection. ABO had no effect on ROS level (Figure 2-17A). In contrast, benzoxazine derivative DBO elevated the intracellular ROS. Then ROS-scavenger NAC was employed for DBO-induced autophagy. Data showed that NAC effectively reduced DBO-elevated ROS level (Figure 2-17B). Western blot data indicated that NAC may inhibit DBO-evoked LC3-II accumulation. These results illustrated the DBO stimulates autophagy through ROS instead of ABO (Figure 2-17C).6.2 After challenged with ABO (25-100μmol/1) for 24 h, cells were stained with JC-1 fluorescent probe for mitochondrial membrane potential (MMP) detection. Benzoxazine derivative ABO had no effect on MMP level (Figure 2-18).CONCLUSIONS:1. Low concentrations of cadmium stimulated endothelial autophagy and inhibited cell apoptosis under bovine calf serum and bFGF deprivation condition. During the process, ROS was upregulated;integrinβ4,caveolinel and PC-PLC activity were decreased.2. Benzoxazine derivative ABO stimulated autophagy in VECs.3. ABO stimulated endothelial autophagy via upregulating ANXA7.4. ABO-stimulated autophagy is mTOR-independent. ANXA7 is NOT involved in rapamycin-induced autophagy.5. ABO did not affect ROS level, but DBO-elevated ROS conferred DBO the autophagy-promoting ability.6. ABO did not affect MMP level in vascular endothelial cells.