Leptin-induced transphosphorylation of vascular endothelial growth factor receptor increases Notch and stimulates endothelial cell angiogenic transformation

Leptin increases vascular endothelial growth factor (VEGF), VEGF receptor-2 (VEGFR-2), and Notch expression in cancer cells, and transphosphorylates VEGFR-2 in endothelial cells. However, the mecha- nisms involved in leptin’s actions in endothelial cells are not completely known. Here we investigated whether a leptin-VEGFR-Notch axis is involved in these leptin’s actions. To this end, human umbilical vein and porcine aortic endothelial cells (wild type and genetically modified to overexpress VEGFR-1 or -2) were cultured in the absence of VEGF and treated with leptin and inhibitors of Notch (gamma-secretase inhibitors: DAPT and S2188, and silencing RNA), VEGFR (kinase inhibitor: SU5416, and silencing RNA) and leptin receptor, OB-R (pegylated leptin peptide receptor antagonist 2: PEG-LPrA2). Interestingly, in the absence of VEGF, leptin induced the expression of several components of Notch signaling pathway in endothelial cells. Inhibition of VEGFR and Notch signaling significantly decreased leptin-induced S- phase progression, proliferation, and tube formation in endothelial cells. Moreover, leptin/OB-R induced transphosphorylation of VEGFR-1 and VEGFR-2 was essential for leptin’s effects. These results unveil for the first time a novel mechanism by which leptin could induce angiogenic features via upregulation/trans- activation of VEGFR and downstream expression/activation of Notch in endothelial cells. Thus, high levels of leptin found in overweight and obese patients might lead to increased angiogenesis by activating VEGFR-Notch signaling crosstalk in endothelial cells. These observations might be highly relevant for obese patients with cancer, where leptin/VEGFR/Notch crosstalk could play an important role in cancer growth, and could be a new target for the control of tumor angiogenesis.

1.Introduction
Endothelial cells (EC) play a major role in normal biological processes such as blood-tissue exchange, blood-cell activation, vasculogenesis, and angiogenesis. EC line the blood vasculature, regulate vascular homeostasis, and form a barrier to create the endothelium; which acts as a vessel for the circulating blood. Under the actions of several circulating factors EC can modify the pheno- type of the vessel wall (Aird, 2007). Vascular endothelial growth factor (VEGF) and its receptor type 2 (VEGFR-2) play major roles in these processes (Hoeben et al., 2004). VEGFR-2 is a kinase insert domain receptor encoded by the human KDR gene, which cor- responds to the FLK-1 mouse gene. It plays an essential role in the angiogenic process in physiologic and pathological scenarios (Fischer et al., 2006). VEGFR-2 is a major mitogenic and chemo- tactic receptor expressed by EC and other cells, which can induce several cellular processes common to many growth factor recep- tors, such as, cell proliferation, migration, survival, and vascular development (Ferrara et al., 2003; Garonna et al., 2011). In addi- tion, VEGF/VEGFR-2 autocrine/paracrine signaling is essential for survival of breast cancer cells (Guo et al., 2010). Notch is also an important factor involved in angiogenesis. When activated, Notch intracellular domain (NICD) induces proliferation, differentiation, and survival of EC that in turn increases angiogenesis. Abnormal expression/activation of Notch correlates to cancer development and poor outcomes (Guo et al., 2011). The Notch receptor family is composed of four members, Notch1, Notch2, Notch3, and Notch4. Notch1 is considered a hall- mark of breast cancer, and is generally overexpressed in tumor tissues. In contrast, Notch receptors are modestly expressed in normal breast tissue (Guo et al., 2011). Several ligands of Notch: Jagged1, Jagged2, and Delta-like ligands (DLL1, DLL3, and DLL4) are also abnormally expressed in cancer tissues. Notch1 and DLL4 have been shown to be expressed in human umbilical vein endothe- lial cells (HUVEC) and upregulated by VEGF/VEGFR signaling (Liu et al., 2003).

Ligand binding to Notch receptors on adjacent cells leads to the activation of signaling pathways. The protease, gamma- secretase, is an essential enzyme involved in NICD formation. A series of proteolysis events involving gamma-secretase, cleave and activate Notch receptors leading to the formation of NICD; which binds CSL/RBP-Jk, a transcription factor in the nucleus and leads to activation of several genes (e.g., survivin and Hey2).The molecular signature of tip EC forming capillaries is charac-terized by the expression of VEGFR-2, DLL4, and other receptors (Gerhardt et al., 2003). Notch and VEGF/VEGFR-2 signaling path- ways have been shown to promote a switch from stalk to tip EC; thus promoting differentiation and migration of tip EC that become highly polarized and extend their filopodia to form cap- illaries (Gerhardt et al., 2003). Notch signaling in stalk EC impairs filopodia extension, promotes tube length by cell proliferation, and inhibits vessel branching by lowering expression of VEGFR-2 and other receptors (Tammela et al., 2008; Williams et al., 2006). Notch induces lateral inhibition of EC during tip-stalk cell fate switch- ing, which involves DLL4/Notch signaling (Hellstrom et al., 2007a; Noguera-Troise et al., 2006; Suchting et al., 2007).Leptin is a small (16 kD) and non-glycosylated cytokine secretedby adipocytes and cancer cells that promotes angiogenesis in physiological and pathological contexts (Halaas et al., 1995; Sierra- Honigmann et al., 1998). The leptin receptor, OB-R, is mainly expressed in hypothalamic cells, but also in peripheral cells includ- ing EC (Guo et al., 2012; Schroeter et al., 2007). Obese individuals exhibit the highest levels of leptin and usually develop leptin- resistance; where leptin cannot control appetite or energy balance (Halaas et al., 1995; Sierra-Honigmann et al., 1998). Increased leptin signaling correlates to an enhanced expression of proteins, and is involved in cancer progression and tumor angiogenesis (Gonzalez- Perez et al., 2013). Leptin signaling has been shown to regulate VEGF/VEGFR-2 and Notch and its targets in breast cancer (Guo and Gonzalez-Perez, 2011). Additionally, a crosstalk between Notch, leptin, and IL-1 signaling (NILCO) affects the expression of pro- angiogenic molecules, leading to cell proliferation and migration of breast cancer cells (Guo and Gonzalez-Perez, 2011). Therefore, leptin secreted by adipocytes and cancer cells could be an impor- tant factor contributing to tumor angiogenesis by acting directly on EC. In addition, leptin can increase tumor angiogenesis by inducing VEGF secretion (Garonna et al., 2011). However, the mechanisms involved in leptin’s actions on EC are not completely understood. Thus, the aim of this investigation is to determine whether leptin could induce the angiogenic transformation in EC via activation of VEGFR and Notch independently of VEGF.

2.Materials and methods
Human recombinant leptin, human VEGFR-2 Quantikine ELISA Kits, and human VEGF cytokine (293-VE) was purchased from R&D Systems, Minneapolis, MN. Notch1 (sc-373891), Notch4(sc-56594) and Jagged1 (JAG1, sc-8303) polyclonal antibodies were obtained from Santa Cruz Biotechnology, Santa Cruz, CA. DLL4 (ab7280), Notch2 (ab8926), and Notch3 (ab23426) poly- clonal antibodies were purchased from Abcam, Cambridge, MA. VEGFR-2 monoclonal antibody (55B11) was purchased from Cell Signaling, Danvers, MA. Anti-mouse and anti-rabbit poly- clonal antibodies conjugated to horseradish peroxidase were from Bio-Rad Laboratories, Hercules, CA. SU5416 was purchased from Sellekchem, Houston TX. Enhanced chemiluminescence (ECL)-western blot stripping buffer was from Thermo Scientific, Rockford, IL. Pegylated leptin receptor antagonist 2 (PEG-LPrA2) was prepared by us as previously described (Gonzalez and Leavis, 2003). Beta-actin (A5316) and GAPDH (glycerol aldehyde phosphate dehydrogenase; G8795) monoclonal antibodies, pro- tease and phosphatase inhibitor cocktails 1 and 2, fetal bovine serum (FBS), DAPT [N-[N-(3,5-difluorophenacetyl)-l-alanyl]-S- phenylglycine t-butyl ester], 4r,6-diamidino-2-phenylindole] and S2188 gamma-secretase inhibitors, and other chemicals were pur- chased from Sigma-Aldrich, St. Louis, MO.HUVEC (Invitrogen, Grand Island, NY) were cultured in Medium 200 (Invitrogen) containing 10% fetal bovine serum (FBS, Med Sup- ply Partners, Inc.) plus P/S [1% penicillin/streptomycin (Gibco, Inc.)], 20 µg/ml of EC growth supplement (Millipore Inc.), and 0.4 µg/ml Geneticin (Life Technologies). Wild type porcine aortic EC (PAEC, that do not express VEGFR-1 or VEGFR-2) and PAEC transfected with VEGFR-1 or VEGFR-2 cultured in Hams F-12 Medium (Invitro- gen) containing 10% FBS and P/S, were used as negative and positive control models, respectively, to determine the impact of leptin and the mechanism associated with VEGFR signaling (Waltenberger et al., 1994). EC cultures were carried out at 37◦ C in 5% CO2/95% air in 25 mm2 tissue culture flasks with cells in maximum passage equal to five.Semi confluent EC were further cultured in FBS deprived medium for 24 h.

Then, EC were cultured in medium containing human recombinant leptin (0.6, 1.2, and 6.2 nM, which are equiv- alent to blood levels of leptin in lean and obese individuals) and inhibitors of OB-R (pegylated LPrA-2; 1.2–6.2 nM), Notch (gamma- secretase inhibitors: DAPT 5 µM and S2188 10 nM in 0.1% DMSO), and VEGFR-2 (Semaxinib/SU5416; 5 µM). Notch expression in cell lysates was investigated using western blot (WB) and real time polymerase chain reaction (RT- PCR).HUVEC were cultured in serum-supplemented medium at a density of 10,000 cells/well in 96-well culture plates for 24 h; and then cultured in FBS-deprived medium for an additional 16 h. Cells were treated with leptin and inhibitors as described above. Cell proliferation was determined using MTT (3-(4,5-dimethylthiazol- 2-yl)-2,5-diphenyl tetrazolium bromide) assay. Optical density (OD) values corresponding to formazan formation were recorded at 540 nm on a Microplate Reader (Spectramax). The results were expressed relative to the OD value of untreated HUVEC.HUVEC were plated in 12-well cell culture plates at a den- sity of 100,000 cells/well in culture medium for 24 h; and then in FBS deprived-medium for an additional 24 h. Cells were treatedas described above. Propidium iodide staining was used to quan- tify DNA content and cell cycle progression was measured using a Cellometer device (Nexcelom; Lawrence, MA). Gated cells were counted and expressed as percentage of untreated HUVEC.Total proteins were extracted from EC on ice using radioim- munoprecipitation assay (RIPA) buffer and a protease/phosphatase inhibitor cocktail (Sigma). Protein concentrations of tissue lysates were determined using the Bradford method (Bio-Rad).

Fifty micro- grams of protein lysates were used for WB analysis as previously described (Gonzalez-Perez et al., 2010). After electrophoresis, protein bands were transferred onto nitrocellulose membranes (BioRad), which were incubated with specific primary antibodies for Notch1, Notch2, Notch3, Notch4, Jagged1, DLL4, and non- phosphorylated and phosphorylated (p)VEGFR-2 at 4 ◦C overnight. Beta-Actin or GAPDH were used as loading controls. Image J soft- ware program (The National Institute of Health, NIH) was used for protein quantitative analysis.VEGF levels in culture supernatants were determined by enzyme-linked immunoassay (VEGF-ELISA, R&D system; sensitiv- ity: 15 pg/ml). Recombinant VEGF was used as a positive control. Additionally, pVEGFR-2 levels were determined in EC lysates (VEGFR-2 ELISA, R&D System; sensitivity: 5 pg/ml). All determina- tions and calibrations were carried out in triplicate.To determine mRNA Notch expression, total RNA from EC was extracted and purified using RNeasy Mini Kit (Qiagen) following the manufacturer’s protocol. Complementary deoxyribonucleic acid (cDNA) was synthesized from total RNA using SuperScript First-Strand Synthesis System with SuperScript II reverse tran- scriptase according to the manufacturer’s protocols (Bio-Rad). cDNA was used as a template for RT-PCR with SYBR-Green PCR master-mix, and iQ5 RT-PCR detection System. RT-PCR reac- tions consisted of 1x SybrGreen Supermix, 0.20 nmol/l forward and reverse primers, and 1 µg of cDNA. To generate a standard curve, amplified cDNA from the reference sample was obtained using a five-fold dilution series of cDNA per reaction. Relative gene expression was calculated by dividing the specific expression value by the corresponding expression of GAPDH. Primers include Notch1 forward primer: 5r-GTCAACGCCGTAGATGACC- 3r, reverse primer: 5r-TTGTTAGCCCCGTTCTTCAG-3r, Notch2 forward primer: 5r-TCCACTTCATACTCACAGTTGA- 3r, reverse primer: 5r-TGGTTCAGAGAA AACATACA-3r, Notch3 forward primer: 5r-GGGAA AAAGGCAATAGGC- 3r, reverse primer: 5r-GGAGGGAGAAGCCAAGTC-3r, Notch4 forward primer: 5r-AACTCCTCCCCAGGAATCTG-3r, reverse primer: 5r-CCTCCATCCAGCAGAGGTT-3r, DLL4 for- ward primer: 5r-TCCAACTGCCCTTCAATTTCAC-3r, reverse primer: 5r-CTGGATGGCGATCTTGCTGA-3r, Jagged-1 forward primer: 5r-GGAGGCGTGGGATTCCA-3r; reverse primer: 5r- CCGAGTGAGAAGCCTTTCAATAAT-3r and GAPDH, forwardprimer: 5r-ATGGGGAAGGTGAAGGTCG-3r and reverse primer:5r-GGGGTCATTGATGGCAACAATA-3r (Qiagen).

PCR conditions consisted of: 1 cycle, 95 ◦C for 3 min (45 cycles), 95 ◦C for 30 s; 52 ◦C for 30 s and 72 ◦C for 30 s. Annealing temperatures varied around 61 ◦C. RT-PCR determinations were performed in triplicate.Knock down of VEGFR-1/FLT-1, VEGFR-2/FLK-1, Notch1, Notch3, and CSL/RBP-Jk gene expression in EC was achieved using specific small interfering RNA (SiRNA, Qiagen). EC were seeded at a den- sity of 2 × 105/ml and cultured until 60% confluence in growth medium without antibiotics in 12 well plates. Cells were then transfected with VEGFR-1, VEGFR-2, Notch1, Notch3, or CSL/RBP-Jk oligonucleotides (an essential transcription factor for Notch sig- naling; 10 nM) that were composed of three to five specific siRNA targets of 19–25 nucleotide length (Qiagen). Negative controls included scramble SiRNAs (SiControl). EC were incubated at 37 ◦C for 6 h in serum-free transfection medium (Qiagen). After SiRNA transfection, EC were cultured at 37 ◦C for an additional 24 h in normal growth medium with antibiotics, 10% FBS, and transfection medium. Cell lysates were analyzed by WB for VEGFR-1, VEGFR-2, Notch1, Notch3, and CSL protein expression.EC were cultured in M200 medium containing 5% FBS and 5 mg/ml of EC growth factor (ECGF, Millipore). The effects of leptin and inhibitors of VEGFRs and Notch on EC ability to form tube- like structures in vitro were determined by Matrigel Matrix assays (BD Sciences). VEGF was used as positive a control. Growth factor- reduced Matrigel was plated onto 96-well plates (200 µl/well) and incubated at 37 ◦C for 30 min. HUVEC were seeded in the Matrigel- coated plates at 10,000 cells/well, and treated with leptin (1.2 nM), SU5416 (5 µmol/l), and DAPT (5 µmol/l) in 1% reduced FBS medium. After 8 h, the cells were washed twice in PBS and fixed in 4% paraformaldehyde. Images of tube formation were captured at 10× using an inverted microscope. Quantitative determination of HUVEC tube formation was carried out using Image Pro Plus soft- ware. The visible tubes were counted using the software, and the analyses of vessel average length and area (mm2) were performed. Statistical analysis was performed using one-way ANOVA and student t-test to compare results between treatment groups. All experiments were repeated 3 or more times. Data is presented as an average ± standard deviation. P < 0.05 was considered to be statistically significant.

3.Results
The ability of leptin to induce EC-derived tube-like structures in the absence of VEGF was assessed. DAPT (an inhibitor of gamma secretase that activates Notch) and SU5416 inhibitors were used to determine whether activity of Notch and VEGFR-2, respectively, are required for leptin-induction of HUVEC proliferation. Leptin was shown to increase tube-like structure formation in HUVEC (Fig. 1A and B). The addition of DAPT to HUVEC cultures treated with leptin significantly decreased tube-like structure formation (Fig. 1A and B). Moreover, the inhibition of VEGFR-2 signaling via SU5416 (a selective tyrosine kinase inhibitor of VEGFR-2) also reduced leptin- induced tube-like formation (Fig. 1A and B). Leptin induced EC proliferation in the absence of VEGF as was shown from MTT cell proliferation assay data (Fig. 1C). However, leptin failed to promote HUVEC proliferation when Notch or VEGFR-2 was inhibited via DAPT and SU5416, respectively (Fig. 1C). Furthermore, data from HUVEC treated with leptin and an additional gamma-secretaseinhibitor, S2188, assessed that inhibition of Notch activation nega- tively impaired leptin-induced EC proliferation (Fig. 1D). Moreover, inhibition of VEGFR-2 and Notch blocked leptin-induced S-phase progression in HUVEC as determined by Cellometer (Fig. 1E). These results suggest that leptin’s pro-angiogenic effects on EC involve Notch and VEGFR-2 signaling.To test specificity of leptin effects on HUVEC, PEG-LPrA-2 was used. Addition of PEG-LPrA-2 blocked leptin-induced tube forma- tion, proliferation, and S-phase progression in HUVEC (Fig. 1A, B and E). The addition of VEGF to HUVEC induced proliferation and S-phase progression to a similar degree than leptin (Fig. 1C and E). However, VEGF (25 ng/ml equivalent to 1.3 nM) showed supe- rior ability than leptin (1.2 nM) to induce tube formation in HUVEC (Fig. 1A and B).To determine whether leptin affects Notch expression and its targeted molecules in EC, WB and RT-PCR were performed. 1.2 nM of leptin (in the absence of VEGF) induced Notch in EC (Fig. 2A). HUVEC treated with the lower dose of leptin assayed (0.6 nM) showed increased protein expression of Notch1, Notch2, and DLL4 (Fig. 2A and B).

Leptin (1.2 nM) induced all Notch receptors and ligands assayed and mRNA expression (Fig. 2B). Higher leptinconcentration (6.2 nM) increased several proteins and Notch2, DLL4, and Jagged1 mRNA in HUVEC (see Fig. 2B).Time response experiments showed that leptin 1.2 nM increased the levels of Notch2 and Notch4 proteins in HUVEC after 12 h. Moreover, HUVEC increased all Notch molecules (Fig. 3C and D) investigated after incubation with leptin for 24 and 48 h (Fig. 3B). To investigate whether leptin-induced VEGFR transactivation regulates Notch, wild type PAEC (that lack VEGFR-1 or VEGFR-2 expression) (Kroll and Waltenberger, 1997) and VEGFR-transfected PAEC (that overexpress VEGFR-1 or VEGFR-2) were challenged with leptin (0–6.2 nM). Firstly, PAEC (wild type and VEGFR-transfected) were analyzed by WB for expression of VEGFR-1 and VEGFR-2. As expected, wild-type PAEC did not express detectable levels of VEGFR-1 or VEGFR-2 proteins (Fig. 4A). In contrast, VEGFR- transfected PAEC express VEGFR-1 or VEGFR-2 (Fig. 4A). It was also determined that the PAEC used express leptin receptor, OB-R (Fig. 4B). Wild type PAEC (lacking VEGFR-1 and VEGFR- 2) cultured in basal conditions or cultured with leptin did not express detectable quantities of Notch proteins (data not shown). In contrast, PAEC VEGFR-1 and VEGFR-2 expressed some Notch proteins in basal conditions (Fig. 4C and D). PAEC VEGFR-1 chal- lenged with leptin showed increased levels of Notch3, Notch4, and Jagged1 proteins (Fig. 4C). Similar results were found in PAEC VEGFR-2 cells treated with leptin, but Notch1 protein was alsoupregulated (Fig. 4D). In comparison, leptin induced higher levels of Notch proteins in PAEC overexpressing VEGFR-2 (see Fig. 4C and D). Additionally, HUVEC (that express VEGFR-1 and VEGFR-2) treated with leptin showed increased protein levels of Notch receptors and ligands (see Fig. 1A).Additionally, leptin induced Notch3 mRNA in PAEC VEGFR-1 (Fig. 5B), and Notch3, Notch4, and DLL4 mRNA in PAEC VEGFR-2 (Fig. 5C).

Moreover, HUVEC treated with leptin showed increased mRNA levels of all Notch molecules investigated (see Fig. 2B). This data further suggest that leptin induces Notch in EC, and VEGFR activity is likely involved in these effects of leptin.To further explore the mechanisms underlying the role of VEGFR in leptin induction on Notch expression in EC, SU5416 was added to HUVEC cultures that were also treated with 1.2 nM of leptin. Addi- tionally, VEGF was used as a positive control. Inhibition of VEGFR-2 kinase activity via SU5416 blocked leptin-induced expression ofNotch1, Notch2, Notch3 and Notch4 proteins in HUVEC (Fig. 6A). These findings further suggest that leptin’s induction of Notch in EC involves VEGFR kinase activity.To additionally investigate the role of VEGFR in leptin-induced Notch expression, knockdown of VEGFR-2 was carried out via siRNA. Remarkably, VEGFR-2 SiRNA transfection blocked leptin- induced expression of Notch2, Notch3, Notch4, and Jagged1 proteins in HUVEC (Fig. 6B). However, silencing VEGFR-2 did not alter the levels of leptin-induced Notch1 and DLL4 proteins (see Fig. 6B). Additionally, double silencing of VEGFR-1 and VEGFR-2 was carried out in HUVEC treated with leptin. Double knockdown of VEGFR-1 and VEGFR-2 inhibited leptin-induced expression of Notch2, Notch3, and Notch4 proteins (Fig. 6C). However, Notch1 protein was upregulated in HUVEC after knockdown of VEGFR-1 and VEGFR-2, which might suggest other mechanisms are involved in leptin’s induction of Notch1 in HUVEC.In order to further investigate the role of VEGFR in leptin-induced effects, Notch mRNA levels were determined in HUVEC treated with leptin and transfected with VEGFR-2 SiRNA ordouble VEGFR-1 and VEGFR-2 SiRNA. mRNA silencing of VEGFR-1 and VEGFR-2 effectively decreased mRNA levels of these receptors (Fig. 6D).

Notch1, Notch3, and Notch4 mRNA levels were signifi- cantly decreased in HUVEC transfected with VEGFR-1 and VEGFR-2 siRNA and treated with leptin (Fig. 6D). Additionally, the knock- down of VEGFR-2 significantly reduced Notch2 mRNA in HUVEC treated with leptin. However, Notch1 mRNA was increased by VEGFR-2 siRNA transfection (Fig. 6D).We further determined that HUVEC and PAEC cultured in basal conditions or medium containing leptin do not secrete detectable levels of VEGF (Fig. 7A) (Garonna et al., 2011; Waltenberger et al., 1996). However, leptin enhanced the expression of VEGFR- 2 protein in HUVEC (Fig. 7B). Moreover, it was found that leptin upregulation of VEGFR-2 protein in HUVEC involves Notch activity, as DAPT inhibition of Notch-activating enzyme (gamma-secretase)reduced leptin’s effects (Fig. 7C). To further investigate whether Notch signaling is involved in leptin-induced expression of VEGFR- 2 in HUVEC, Notch1, Notch3, and CSL/RBP-Jk were knocked down via siRNA. HUVEC treated with leptin and transfected with Notch1, Notch3, and CSL/RBP-Jk SiRNA showed decreased levels of leptin- induced VEGFR-2 protein (Fig. 7C) and mRNA (Fig. 7D). It was assessed that Notch silencing treatment significantly decreased the levels of Notch1 and Notch3 (see Fig. 7D).It was further assessed that leptin/OB-R signaling transphos- phorylates VEGFR-2 in HUVEC in the absence of VEGF (Fig. 7E and F) as was earlier reported (Garonna et al., 2011). Phosphory- lated VEGFR-2 (pVEGFR-2) levels were increased in HUVEC treated with leptin, as determined by ELISA (Fig. 7E). These effects of leptin were diminished by inhibition of VEGFR-2 kinase activity (see Fig. 7E). To test if the effects of leptin were specific, HUVEC were treated with leptin and OB-R antagonist, PEG-LPrA-2, which decreased the levels of leptin-induced phosphorylated VEGFR-2 (see Fig. 7E). Moreover, WB analysis showed that leptin signifi- cantly increased transphosphorylation of several tyrosine sites inthe intra-cytoplasmic VEGFR-2 tail in HUVEC and PAEC transfected with VEGFR-2 (Fig. 7F). Leptin enhanced the levels of phosphory- lated VEGFR-2 at residues, Y951, Y996, and Y1175 in HUVEC, and Y966, Y1059, and Y1175 in PAEC VEGFR-2 (see Fig. 7F). Incubationof HUVEC with VEGF (positive control) induced similar VEGFR-2 phosphorylation as with leptin, but also increased phosphorylation of Y1059 (Fig. 7F).

4.Discussion
Leptin is a known regulator of normal and pathological angio- genesis (Bouloumie et al., 1998; Gonzalez-Perez et al., 2013). However, the specific mechanisms involved in leptin’s actions are not completely understood. We earlier described that leptin upregulates the expression of VEGF in breast cancer cells through mechanisms involving several canonical signaling pathways and specific transcription factors (e.g., HIF and SP1) (Gonzalez-Perez et al., 2010). Furthermore, we found that leptin/OB-R signal- ing was linked to increased levels of VEGF/VEGFR-2 in breast cancer (Battle et al., 2014; Rene et al., 2009). Moreover, we pre- viously found that leptin upregulates Notch in breast cancer, and that a complex crosstalk between Notch, IL-1, and leptin (NILCO) regulates the expression of angiogenic molecules (VEGF/VEGFR) in breast cancer cells (Battle et al., 2014; Gillespie et al., 2012; Guo and Gonzalez-Perez, 2011). Wheeler-Jones and collaborators have previously shown that leptin induces the phosphorylation of VEGFR-2 (Y1175) in HUVEC independently of VEGF signaling (Garonna et al., 2011). Moreover, leptin was found to induce the angiogenic transformation of EC via COX-2 (cyclooxygenase 2) and activation of specific kinases, which was linked to increased VEGFR-2 phosphorylation in the absence of VEGF (Garonna et al., 2011). The inhibition of VEGFR-2 tyrosine kinase activity in HUVEC reduced leptin-stimulated pro-angiogenic responses. In addition, the blockade of VEGFR-2 or COX-2 activities abolished leptin-driven neo-angiogenesis in a chick chorioallantoic membrane vascularization assay in vivo (Garonna et al., 2011).

In the present work, we expanded these investigations by show- ing that leptin-induced proliferation and pro-angiogenic features of EC are related to leptin-induced expression, transphosphoryla- tion, and activation of VEGFR-2, which mediates the upregulation of Notch expression and signaling; in the absence of VEGF. These effects of leptin involved the VEGF-independent transactivation of VEGFR-1 and/or VEGFR-2.Here for the first time we show that leptin pro-angiogenic actions in EC are dependent on a functional OB-R/VEGFR/Notch signaling axis (Fig. 8). Indeed, the inactivation of OB-R (via PEG- LPrA2), VEGFR kinase activity (via SU5416) or interfering with gene expression by SiRNA, and the inhibition of Notch activity via gamma-secretase inhibitors (DAPT and S2188), and Notch and CSL/RBP-Jk SiRNA impaired leptin-induced HUVEC proliferation, S-phase progression, and tube-like structure formation.It is known that Notch signaling plays critical roles in cell fate determination and angiogenesis (Shawber et al., 2003; Takeshita et al., 2007). Notch1 and Notch4 receptors, DLL1, DLL4, and Jagged1 ligands are predominantly found in the endothelium. Positive or negative modulation of Notch results in vascular pathologies. The Notch pathway is tightly regulated in EC by VEGF signals, but not by basic fibroblast growth factor (bFGF). VEGF signaling induces Notch1 and DLL4 gene expression in EC. Moreover, the Notch path- way is involved in a feedback loop with VEGF (Hellstrom et al., 2007a).The regulation of Notch/DLL gene expression by VEGF was found to be critical for arteriogenesis and angiogenesis. VEGF reg- ulates fate and phenotypic changes of EC during the formation of capillaries and the formation of tip cells during angiogenesis via Notch1/DLL4 activity (Hellstrom et al., 2007b). Additionally, Notch1 induced signals have been implicated in other angiogenic functions (e.g., formation of hematopoietic stem cells from EC) (Kumano et al., 2003).

However, the signaling mechanisms con- trolling Notch receptor and ligand gene expression in EC are not completely known.Previous findings show that VEGFR-1 and VEGFR-2 receptor sig- naling are necessary for tumor angiogenesis and EC migration and proliferation (Holmqvist et al., 2004; Lee et al., 2010; Matsumoto et al., 2005). Present data suggest that leptin is a novel inducer of Notch in EC via transactivation of VEGFR. Our results indi- cate that leptin regulates specific Notch receptors and ligands inEC either through transactivation of VEGFR-2 or the heterodimer pair (VEGFR-1 and VEGFR-2). Leptin upregulation of Notch was linked to both VEGFR-1 and -2 activities, as the simultaneous knock down of these receptors abolished the majority of leptin effects on Notch. Therefore, leptin‘s actions could involve VEGFR-1/VEGFR-2 heterodimer signaling. The simultaneous knockdown of VEGFR- 1 and VEGFR-2 inhibited leptin-induced Notch1 mRNA but was unable to block leptin upregulation of Notch1 protein, which sug- gest additional mechanisms independent of VEGFR signaling could be involved in leptin-induced Notch1 expression in EC. The specific biological relevance of these findings needs to be further investi- gated. Dissimilar transcription and translation rates could indicate that changes in gene expression level are not reflected at the protein level (Vogel and Marcotte, 2012).In this investigation we assessed that leptin induced thetransphosphorylation of Y1175, Y951, and Y996 of VEGFR-2 in HUVEC in the absence of VEGF. Moreover, leptin’s effects were more evident in PAEC-VEGFR-2 transfected cells. However, present data further show that leptin-induced trans-phosphorylation of VEGFR- 2 is not equivalent to VEGF-induced VEGFR-2 phosphorylation, at least in terms of activated Y residues in the intracytoplasmatic tail of the receptor. We found that VEGF also induces the phosphoryla- tion of Y1059 VEGFR-2 in HUVEC.Of note, present results show that the blockade of Notch acti- vation impaired leptin-induced EC angiogenic features, but also reduced leptin-induced VEGFR-2 phosphorylation. Taken together, the data supports the notion that leptin induction of Notch requires VEGFR-2 activity, which is in turn regulated by a Notch feedback in EC.

Remarkably, the actions of leptin on EC were independent of VEGF signaling. Therefore, leptin could re-enforce VEGF actions by stimulating Notch expression and activity, which can be linked to the progression of cell cycle, proliferation, and angiogenic differ- entiation of EC. However, the fact that VEGF and leptin actions induced different patterns of VEGFR-2 phosphorylation might sug- gest that additional undetermined mechanisms occur in EC, which may be linked to Notch expression, and leptin-induced angiogenic features.Present data suggest that leptin could have some redundant actions to VEGF on EC, which may profoundly affect angiogenesis; especially in obesity contexts, which are characterized by abnor- mally high levels of leptin. Previous research has shown that obesity and overweight conditions are strongly related to increased tumor angiogenesis (Battle et al., 2014; Gillespie et al., 2012), and the inci- dence and poor prognosis of several cancer types (Calle and Kaaks, 2004; Guo et al., 2012; Lipsey et al., 2016). Whether leptin induces Notch through VEGFR-2 activation during tumor angiogenesis has yet to be determined.

5.Conclusion
Present data shows for the first time that leptin can induce Notch expression in EC through the VEGF-independent transactivation of VEGFR-1 and VEGFR-2. Moreover, leptin/OB-R signals induce a novel functional signaling axis in EC, leptin/VEGFR/Notch axis, which is essential for leptin-stimulated proliferation and angio- genic transformation of EC (see Fig. 8). Present findings further suggest that leptin/OB-R signaling could be a potential target for anti-angiogenesis therapies. Combination therapies targeting the leptin/VEGFR/Notch axis could be beneficial for cancer patients, particularly those showing high levels of leptin due to obesity or overweight conditions. However, SU5416 present findings need to be fur- ther validated using in vivo models.