Introduction
Angiogenesis, defined as the formation of new capillaries from pre-existing blood vessels, is an essential physiological process involved in wound healing, ovulation, vascular and embryonic development [
1]. It is a complex process characterized by the migration of endothelial tip and stalk cells in response to a variety of stimuli including chemokines and growth factors such as VEGFA, and is regulated by a variety of signaling pathways [
2]. Disordered angiogenesis stimulates a wide range of pathophysiological conditions such as ischemic circulatory disease, cancer, chronic inflammation, atherosclerosis, arthritis and diabetic retinopathy [
3]. The development of angiogenesis inhibitors and immunotherapies represents an active area of investigation to combat cancer and other sequelae of dysregulated angiogenesis [
4‐
6]. Despite some advances in this field, angiogenesis inhibitors have, in general, only extended the survival of chemotherapy patients for a few months and resistance to treatment often develops rapidly [
7]. Thus, further investigation of underlying molecular mechanisms of angiogenesis is critical to understanding these common disease processes.
The Wnt proteins are a family containing 19 secreted cysteine-rich glycoproteins that accumulate in the extracellular matrix to activate pathways in adjacent cells; they are highly conserved among species [
8]. The Wnt/β-catenin pathway is involved in cellular proliferation, survival, differentiation, migration and apoptosis [
9,
10]. Moreover, this pathway also appears to be critical in vascular endothelial cells and acts through a variety of regulators, including Bach1 [
11], Rspo1 [
12], Endostar [
13], and ERG [
14]. In general, the Wnt/β -catenin pathway is activated when a Wnt ligand binds to its coreceptor complex, which contains a Frizzled family member and low-density lipoprotein receptor-related proteins 5 and 6 (LRP5/6) [
15]. As a result, β-catenin binds to the T cell factor/lymphoid enhancer-binding factor (TCF/LEF) family in the nucleus, and alters the expression of key regulators of angiogenesis, such as VEGF, IL-8, Cyclin D1 and MMP-2 [
16].
The glucocorticoid receptor (GR) is encoded by the gene
NR3C1, which is located on chromosome 5q31-32 in humans, and acts as a ligand-inducible transcription factor belonging to the nuclear receptor superfamily [
17]. It controls many distinct gene networks, governing various aspects of development, metabolism, inflammation and the stress response, as well as other key biological processes in the cardiovascular system [
18]. Glucocorticoids are often components of chemotherapeutic regimens due their recognized angiostatic properties [
19]. In our previous work, we have demonstrated that GR is involved in the pathogenesis of atherosclerosis, sepsis and proteinuria [
20‐
22]. Recently, using a genomic approach (GSE119093),as well as in vitro and in vivo studies, we demonstrated that the loss of endothelial GR (eGR) resulted in upregulation of Wnt signaling [
23]. These genomic data also revealed that there was enhanced binding of some autophagy-related genes in the absence of GR. Given the interest in better determining the links between autophagy, angiogenesis and cancer [
24], we aimed to investigate whether GR could regulate angiogenesis via the Wnt signaling pathway and additionally whether GR could regulate Wnt signaling through autophagy flux.
Materials and methods
Reagents
AllStars negative control siRNA and Stealth RNAi for NR3C1 were from Qiagen. CellTiter-Glo® Luminescent Cell Viability Assay (G7570) and CellTiter 96® AQueous One Solution Cell Proliferation Assay (G3582) were purchased from Promega. Matrigel® Basement Membrane Matrix, Growth Factor Reduced (GFR) was purchased from Corning. Antibodies against LRP5 (5731s), LRP6 (3395s), p-β-catenin (4176s), Cyclin D1 (2978s), LC3A/B (4108s) and p62 (5114 s) were purchased from Cell Signaling Technology. Antibodies against α-smooth muscle Actin (α-SMA) (ab5694), CD68 (ab53444), GAPDH (ab8245) and recombinant mouse Wnt3a protein (ab81484) were purchased from Abcam. Antibodies against CD31 (553370) and CD102 (553326) were purchased from BD Bioscience. Anti-GR (sc-393232) was purchased from Santa Cruz Biotechnology. Recombinant mouse sFRP3 protein (592-FR/CF) was purchased from Bio-techne. Rapamycin (tlrl-rap) was purchased from InvivoGen. Premo™ Autophagy Tandem Sensor RFP-GFP-LC3B Kit including chloroquine (P36239) was purchased from Thermo Fisher Scientific.
Quantitative PCR
MLECs were treated with control or GR siRNA for 48 h. Total RNA was isolated from cells using the RNeasy Mini kit (Qiagen) according to the manufacturer’s instructions. RNA was reverse transcribed using the iScript cDNA Synthesis kit (Bio-Rad) and qPCR was performed on a Bio-Rad C1000 Touch thermal cycler using the resultant cDNA, qPCR Master mix and gene specific primers. The following primers were used:
β-catenin: Forward 5′ TGACACCTCCCAAGTCCTTT 3′.
and Reverse 5′ TTGCATACTGCCCGTCAAT 3′.
TCF4: Forward 5′ GGTGGCCGAATGCACATTGAAAGA 3′.
and Reverse 5′ TTTGCCTGTTCTTCCCTGGACA 3′.
GR: Forward 5′ TGGAGAGGACAACCTGACTTCC 3′.
and Reverse 5′ ACGGAGGAGAACTCACATCTGG 3′.
LRP5: Forward 5′ CTTTCCCCACGAGTATGTTGGT 3′.
and Reverse 5′ AAGGGACCGTGCTGTGAGC 3′.
LRP6: Forward 5′ GGCTGGCATGTGATTGGCT 3′.
and Reverse 5′ GCTCTGGGTTGATCCAACTCT 3′.
Gene expression was normalized to the housekeeping gene 18 s and is presented as fold change.
Western Blot
Cells were lysed on ice with lysis buffer (50 mM Tris·HCl pH 7.4, 0.1 mM EDTA, 0.1 mM EGTA, 1% Nonidet P-40, 0.1% sodium deoxycholate, 0.1% SDS, 100 mM NaCl, 10 mM NaF, 1 mM sodium pyrophosphate, 1 mM sodium orthovanadate, 1 mM Pefabloc SC, and 2 mg/ml protease inhibitor mixture; Roche Diagnostics). Protein concentrations were determined with the DC Protein assay kit (Bio-Rad Laboratories). Lysates were analyzed by SDS/PAGE and immunoblotted. Primary antibodies used include the following: LRP5, LRP6, phospho-β-catenin, Cyclin D1, GR, LC3A/B, P62 and GAPDH as the loading control. Secondary antibodies were fluorescence-labeled antibodies (LI-COR Biotechnology). Bands were visualized with the Odyssey Infrared LI-COR system (LI-COR Biotechnology).
Isolation and culture of ECs from murine aorta
For primary mouse aortic endothelial cell (MAOEC) isolation and culture, 10-day-old C57BL/6 (WT), GR fl/fl, Tie-1 Cre + (eGR KO) or LRP5 fl/fl, LRP6 fl/fl, Tie-1 Cre + (eLRP5/6 KO) mice (n = 3/group) were sacrificed by an overdose of isoflurane. The aortas were removed, minced into 1 mm-sized pieces, washed in DMEM medium, and filtered through a 40 µm nylon-mesh cell strainer. The tissue was digested in 15 mL collagenase A for 45 min at 37 °C and the homogenate was further filtered through a 70 µm cell-strainer and centrifuged at 800 rpm for 8 min. The resulting pellet was then washed with phosphate-buffered saline supplemented with 0.1% bovine serum albumin (0.1%BSA/PBS) and the cell suspension was incubated with rat anti-mouse CD31 antibody (553370, BD Bioscience)-coated magnetic beads (sheep anti-rat; 11035, Thermo Fisher Scientific) for 15 min on a Dynalmixer. The magnetic beads were washed four times with PBS/BSA on Dynal-MPC and resuspended in EBM-2 culture medium containing 20% fetal bovine serum (FBS, Cambrex), endothelial cell growth factor with penicillin (100U/ml) and streptomycin (100 g/ml; Invitrogen) and plated on gelatin (Sigma Aldrich)-coated dishes. Cells were re-purified with CD102 antibody (553326, BD Bioscience)-coated magnetic beads during the first two passages and cells from the third passage were used for experiments.
Cell culture
The MLECs and MAOECs were routinely cultured at 37 °C with 5% CO2 in EBM-2 endothelial cell (EC) basal medium with SingleQuot Kits (Lonza, Walkersville, MD) as well as 10% FBS. Cells were used in experiments at passages three to five. For experiments involving Wnt3a or sFRP3 administration, cells were cultured and starved for 4 h in 0.5% FBS. Then they were treated with either recombinant Wnt3a (200 ng/ml) or recombinant sFRP3 (200 ng/ml) for 4 h.
Cell viability assay
Cell viability was analyzed by CellTiter-Glo® Luminescent Cell Viability Assay (Promega) according to the manufacturer’s instructions. Briefly, 5 × 103 cells/well were seeded in a 96-well plate in 100 μl EBM-2 with 10% FBS/well and we waited 4 h until all the cells became attached. At 4 h and 24 h, 100ul CellTiter-Glo® Reagents were added per well. After mixing contents for 2 min on an orbital shaker and incubating at room temperature for 10 min, luminescence was recorded using a spectrophotometer microplate reader. The assays were carried out in triplicate. Each experiment was repeated three independent times.
Cell proliferation assay
EC proliferation was analyzed by CellTiter 96® AQueous One Solution Cell Proliferation Assay (Promega) according to the manufacturer’s instructions. Briefly, 5 × 103 cells/well were seeded in a 96-well plate in 100 μl EBM-2 with 10% FBS/well and we waited 4 h until all the cells became attached. At 4 h and 24 h, 20ul CellTiter 96® AQueous One Solution Reagents were added per well. Then, the plate was incubated at 37 °C for 4 h in a humidified, 5% CO2 atmosphere. The absorbance was measured at 490 nm using a spectrophotometer microplate reader. The assays were carried out in triplicate. Each experiment was repeated three independent times.
7.5 × 104 cells were cultured in a 24-well plate coated with 250 μl of growth factor-reduced Matrigel (Corning). At 9 h, tubes were photographed in five random microscopic fields with an inverted phase-contrast microscope. Then, the original images were analyzed by the software ImageJ-Angiogenesis Analyzer. In this process a map of the tubes in each imaged was generated and the number of nodes, junctions, branches and length of branches were calculated by ImageJ software. Each experiment was repeated three independent times.
Mice
GR fl/fl, Tie-1 Cre + (eGR KO) mice were generated as previously described [
25].
LRP5 fl/fl, LRP6 fl/fl mice were the kind gift of Dr. Dianqing Wu at Yale University. These mice were bred to obtain endothelial-specific LRP5/6-deficient (
LRP5 fl/fl, LRP6 fl/fl, Tie-1 Cre +) (eLRP5/6 KO) mice as well as controls (
LRP5 fl/fl, LRP6 fl/fl, Tie-1 Cre-). In vivo experiments were performed with eGR KO and eLRP5/6 KO male mice and their wild-type (WT) littermates aged 8–12 weeks. All mice were housed in a standard, pathogen-free facility. All studies were approved by the IACUC at Yale University School of Medicine, New Haven, Connecticut, USA, and were consistent with the
Guide for the Care and Use of Laboratory Animals (National Academies Press, 2011).
Hind limb ischemia model
The severe hind limb ischemia (HLI) model was conducted under anesthesia with isoflurane delivered via a precision vaporizer at 2%. First, the right femoral artery was exposed under a dissection microscope. The proximal femoral artery and the distal portion of the saphenous artery were ligated and arteriectomy was performed. No operation was performed on the left femoral artery, which was regarded as the healthy control. Blood flow recovery to the ischemic foot was sequentially monitored by a MoorLDI2™ laser Doppler imaging system (Moor Instruments, Devon, UK) for up to 21 days after ischemia. Mice were anesthetized and maintained at 37 °C on a heating plate to minimize temperature variation. Ischemic and non-ischemic limb perfusion was measured pre- (day 0), and post-surgery at days 3,7,14 and 21. Final blood flow values were presented as the ratios of ischemic to non-ischemic hindlimb perfusion. Mice were sacrificed on day 21 post-surgery and adductor muscles were harvested, hematoxylin and eosin (H&E) staining was performed to demonstrate the morphology of the muscle tissue, and immunostaining was performed to investigate capillary density, arteriolar density and macrophage density. Capillaries in tissue sections were stained using a rat anti-mouse monoclonal CD31 antibody and arterioles were visualized using a rabbit anti-mouse monoclonal α-smooth muscle actin (α-SMA) antibody. Macrophages were also recognized by a rat anti-mouse monoclonal CD68 antibody. Donkey anti-rat IgG (Alexa Fluor® 488) and donkey anti-rabbit IgG (Alexa Fluor® 488) were used as secondary antibodies for visualization. CD31-positive, α-SMA-positive and CD68-positive areas were quantified in randomly acquired images using ImageJ. Three mice were used in each group.
Sponge implantation assay
For sponge implantation experiments, animals were anesthetized via isoflurane delivered from a precision vaporizer at 2%. A single midline incision was made and two sub-dermis implantation sites were created by blunt dissection on each side of the incision. 9 mm-dimension polyvinyl alcohol (PVA) sponges (Henry Schein) were implanted in the corresponding sites. Ten days later, sponges were excised and made into cryosections, then they were stained with anti-CD31 antibody, anti-α-SMA antibody or anti-CD68 antibody and visualized. Six mice were used in each group.
Autophagy assessment
MLECs were cultured as mentioned above. For autophagy induction, cells were washed three times with PBS and incubated for 4 h in Hanks balanced salt solution (HBSS; Invitrogen) or complete medium containing 2 μM rapamycin (Rap) at 37 °C. Chloroquine (CQ) was used as an autophagy inhibitor at a concentration of 30 μM for 16 h in the complete medium. In some experiments, Wnt3a was administrated for 4 h after treatment with Rap or CQ. Autophagy was assessed by Premo™ Autophagy Tandem Sensor RFP-GFP-LC3B Kit, LC3II/I conversion and p62 degradation. For visualization of autophagosomes and auto phagolysosomes, cells were fixed with 4% formaldehyde and permeated with 0.1% Triton X-100 after incubation with the RFP-GFP-LC3 Kit, then an immunofluorescent microscope was used to visualize the images.
Statistical analysis
To investigate the influence of gene defect, binary comparisons (siCT vs siGR, WT vs eGRKO, eGRKO vs eLRP5/6KO, WT vs eLRP5/6KO) were analyzed using Student’s 2-tailed t tests. To study the effect of certain signaling pathway by using agonists and antagonists, multiple comparisons were analyzed using One-way ANOVA with Tukey’s post hoc test (Control vs Wnt3a vs sFRP3, Control vs Dex vs RU486, Control vs HBSS vs CQ, Control vs Rap vs CQ). Data are expressed as mean ± SEM. A p value less than 0.05 was considered significant. GraphPad Prism 8 was used to generate graphs and perform statistical analysis.
Discussion
The major finding of this study is that absence of endothelial GR can augment angiogenesis through activation of the Wnt/β-catenin signaling pathway and that, mechanistically, this occurs via induction of autophagy flux.
Angiogenesis is involved in the development of many diseases. Most importantly, it is regarded as one of the hallmarks of cancer [
27]. Dysregulation of angiogenesis in cancer results in an abnormal vascular network, which not only reduces the therapeutic efficacy of conventional anticancer therapies but stimulates tumor progression as well [
28]. It is also involved in sequelae of diabetes in that pathological angiogenesis is the main cause of proliferative diabetic retinopathy (PDR), which is the advanced form of diabetic retinopathy (DR). In order to interfere with pathological angiogenesis, multiple treatment approaches have been established, such as gene therapy, monoclonal antibodies (Bevacizumab), miRNAs, small molecules (Sorafenib), Angiostatin and Endostatin [
1].
In the current study, we demonstrate that endothelial GR inhibits angiogenesis via suppression of canonical Wnt signaling. When canonical Wnt signaling is up regulated, due to loss of endothelial GR, cell viability, cell proliferation and tube formation ability are all augmented significantly in GR- deficient MLECs and MAOECs, indicating that endothelial GR can tonically suppress both angiogenic functions and the Wnt/β-catenin pathway under homeostatic conditions. According to the results from both the hind limb ischemia model and sponge implantation model, endothelial GR suppresses angiogenesis as well as inflammation in vivo; we have previously demonstrated that endothelial GR acts as a negative regulator of inflammation [
21].
Previously, we showed that endothelial GR could bind to TCF/LEF directly [
23]; however, here we investigated whether GR could regulate Wnt signaling via other indirect mechanisms. We investigated the expression of p62 between GR-replete and GR-deficient endothelial cells. Intriguingly, we observed an apparent decrease of p62 in GR knockdown cells. As p62 is regarded as a very critical marker of autophagy [
29], further analysis focused on whether GR could regulate autophagy flux. By assessing the LC3 II/I ratio in vitro, which was consistent with decreased expression of p62 in GR knockdown cells, we concluded that endothelial GR could regulate autophagy flux. These results were further confirmed by using rapamycin to demonstrate increased formation of both autophagosomes and auto phagolysosomes in the absence of GR. GR has been reported to modulate autophagy in other cell types such as hepatic cells and prostate cancer cells [
30,
31], but has not previously been investigated in endothelial cells.
In the absence of GR, rapamycin could also potentiate the protein levels of some key components of Wnt signaling, while CQ had an inhibitory effect. These results indicated that loss of endothelial GR could trigger autophagy flux and augmentation of the Wnt/β-catenin signaling pathway. Administration of Wnt3a seemed to have a positive feedback effect on autophagy according to the expression of p62 and LC3. We speculate there might be a GR-mediated autophagy-Wnt feedback loop which promotes angiogenesis.
The relationship between autophagy and Wnt signaling is somewhat controversial. In some cancer cells, the Wnt pathway component DVL2 maintains glioblastoma multiforme (GBM) cell proliferation through canonical and noncanonical Wnt signaling[
32,
33]. In turn, autophagy negatively regulates Wnt signaling by promoting DVL2 and β-catenin degradation[
34‐
36]. However, some other studies reveal different effects. In the study by Ma et al.[
37], the authors found that inhibiting autophagy by downregulating Atg5 gene expression impaired hepatic differentiation of hepatic progenitor cells (HPCs) and inhibited activation of the Wnt/β-catenin pathway, which could be rescued by overexpression of the β-catenin gene. Collectively, these results demonstrated that autophagy promotes HPCs differentiation by up-regulating Wnt/β-catenin signaling. In another study[
38], researchers examined β-catenin phosphorylation at Ser33 in hepatocellular carcinoma cells cultured with Earle's Balanced Salt Solution (EBSS) or 3- Methyladenine (3-MA) for 6 h. The starved cells had decreased β-catenin phosphorylation, while the autophagy-inhibited cells had increased β-catenin phosphorylation. Immunofluorescence examination of the cellular localization of β-catenin showed that autophagy increased β-catenin nuclear accumulation after 6 h starvation. These results indicated that starvation-induced autophagy could activate β-catenin and were consistent with our current findings in endothelial cells. Additionally, according to the study of Jing et al.[
39], in squamous cell carcinoma of the head and neck (SCCHN), Wnt3a binding to its receptor on the membrane activated the canonical Wnt signaling pathway, accelerated the nuclear translocation of β‐catenin and then enhanced the expression of Beclin1, which promoted autophagy. In our study, we also demonstrated that Wnt3a had a positive effect on autophagy.
Our current study also has some limitations. First, in our HLI model, the blood flow in the injured limb recovered to just 20–30% after 21 days, which was lower than that reported in some other studies[
40,
41]. However, we noted via direct observation that the ambulation of the mice became normal within 10 days after surgery and the staining of CD31 and α-SMA in the adductor muscles also revealed adequate vascular supply at day 21 after surgery. Second, although we verified GR-deficiency induced autophagy, we didn’t fully investigate the intrinsic mechanism, which will be the subject of a future study.
In conclusion, endothelial GR impedes autophagy flux, leads to suppression of Wnt/β-catenin signaling and, in turn, suppresses angiogenesis. Our data might prove useful in the clinical arena and suggest that autophagy inhibitors or canonical Wnt signaling inhibitors might be used in suppressing angiogenesis and may be as efficacious as steroids, which are fraught with poorly tolerated side effects. With endothelial cell heterogeneity now recognized as a key driver of vascular physiology, both homeostatic and pathological, endothelial GR may be an important therapeutic target with profound effects on specific vascular beds, an innovative concept that has yet to be studied.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.