| BACKGROUNDIn1971Folkman first proposed that tumor growth depends on tumor vessel supply.Afterwards killing tumor cells by inhibiting tumor angiogenesis has long been the majorstrategy in the field of tumor vascularization, triggering the invention of severalVEGF-VEGFR blocking agents. In2004, FDA approved anti-VEGF monoclonalantibody Bevacizumab as the first vascular targeted anti-tumor drug, and laterSorafenib and Sunitinib were also approved. However, over the past10years, theseVEGF-VEGFR blocking agents appeared to only effect on certain patients with coloncancer, breast cancer and some other specific tumors, embodied with drug resistance andobvious side effects. In addition, the in-depth study on tumor vessel implied that the abnormal structure and function of tumor vessel, especially its leakiness, not only preventsthe delivery of chemo-therapy drugs into tumor but also facilitates the metastasis of tumorcells. Then the strategy of normalization of tumor vessel in combination withchemo-therapy drugs to kill tumor cells was proposed, further supported by the study ofPHD2heteto-deficiency on tumor vascularization. PHD2is a sensor of oxgen in thecells and can degrade HIF family members upon sufficient oxgen, thus negativelyregulating tumor angiogenesis. In PHD2heterozygote mice, tumor vessel was normalizedmanifested by the strengthened vessel endothelial cell adhesion and enhanced mural cellcoverage as well as improved vessel perfusion, which facilitated the delivery andeffectiveness of chemo-therapy drugs. Thus, the strategy of tumor vessel normalizationgained more and more attention in this field, calling for the emergence of new drugtargets.In2011, one review entitled with Tumor angiogenesis: molecular pathways andtherapeutic targets” on Nature Medicine proposed that new drug targets in tumorangiogenesis should have the capability to modulate the growth or function of differentcell types involved in tumor angiogenesis. More importantly, it should be selectivelytargeted. Notch pathway has a dual and vital role in the regulation of endothelial cells andpericytes, two major cell types participating in tumor vascularization. Then can it be a newdrug target in tumor vessel normalization? Should Notch signal be blocked or activated intumor vessel to kill tumor cells? How to make it specific targeted to endothelial cells?Previous study showed that blocking Notch signaling such as anti-Dll4antibodycould promote tumor angiogenesis. However, the newly generated tumor vessel wasdisfunctional and leaky, leading to lower perfusion and increased hypoxia in tumor andfinally the retarded tumor growth. In clinical research, the cocktail of Dll4siRNA andBevacizumab appeared to be effective in treating overn cancer. However, later preclinicaltrials on mouse, rat and monkey demonstated long-term administration of anti-Dll4drovethe development of hemagoniama, which greatly hindered the clinical usage of Dll4blocking agents. On the other hand, activating Notch signal also inhibits the tumorangiogenesis and tumor growth. Our group has shown that overexpressing Notch ligand Dll1in tumor cells slows down its growth, partially due to the decrease in tumorvascularization. Thus, activating Notch signal in tumor vessel appears to be morepromising that blocking Notch for the long-term effect. However, the major obstacle ininventing Notch signal activator is that the activation of this pathway is endocytosisdependent and the soluble ligand could not activate the Notch signal well. Based onthis, we intented to design a novel activating ligand for Notch signal: hD1R fusion protein.This protein consists of DSL domain of the ligand of hDll1and RGD nonapeptide. RGDpeptide could binds to integrin αVβ3, which is specifically expressed on vascularendothelial cells, especially on proliferative endothelial cells. More importantly, thebinding of RGD to integrin αVβ3on endothelial cells can promote its endocytosis andactivate Notch signal more effectively.In summary, we will further clarify the role of Notch signal in tumor vesselnormalization with inducible gain of function and loss of function mouse models. Furtherwe will construct the new Notch activator hD1R fusion protein and verify its activity andin vivo function with xenograft tumor model. Finally, we will explore the underlyingmechanism of Notch pathway by whole genome expression microarray and up-to-datemolecular approaches, hoping our study may shed some light on the intention of noveltumor vessel targeted therapy.OBJECTIVESTo clarify the role of Notch signal in tumor vessel normalization with inducible gain offunction and loss of function mouse models; to design and synthesis of a novel ligandhD1R consisting of DSL domain and RGD peptide; to investigate whether it couldtarget endothelial cells and specifically activate Notch signaling; to verify its function ontumor growth and tumor vascularization with xenograft tumor model; to explore theunderlying molecular mechanism with gene microarray.METHODS1. Loss of Notch signaling in endothelial cells effects on tumor vascular structure and function Intraperitoneal injection of poly:I-C induces the transcription factorRBP-J knockout in MxCre; RBP-Jf/fmouse. Tumor bearing mice model was made afterfour times of inductions. To observe the effect of Notch signal loss on tumor vascularstructure and function by immunofluorescence staining of tumor sections.2. Inducible Notch activation in tumor vessel endothelial cells effects on tumor vascularstructure and function Intraperitoneal injection of tamoxifen induces the NICDexpression in Cdh5CreERT; ROSAN1-ICmouse. Tumor bearing mice model was madeafter seven times of inductions. To observe the effect of Notch signal overexpression ontumor vascular structure and function by immunofluorescence staining of tumor sections.3. Construction, expression and purification of hD1R PCR was used to clone targetgene and DNA recombination was used to construct the prokaryotic expression vector.Escherichia coli expression system was employed to express the recombinant protein.Affinity chromatography with nickel chelating beads was used to purify the fusion protein.4. Detection of the biological activity of hD1R Immunofluorescence, flow cytometryand cell adhesion assay was used to observe the binding of hD1R to endothelial cells.Immunofluorescence, Western Blot and real-time quantitative PCR was used to examineNotch signaling activation of hD1R. Real-time quantitative PCR was used to detect theendocytosis of Notch receptor extracellular domain.5. Function experiment of hD1R The effect of hD1R on lumen formation in vitro wasdetected with endothelial network formation assay and fibrin bead assay. The effect ofhD1R on retinal neovascularization and the formation of Tip cells was tested withwhole-mount immunofluorescence staining of retina vessel. The effect of hD1R on theproliferation of endothelial cells was detected by immunofluorescence staining of Ki67.6. Effect of hD1R on tumor growth and tumor angiogenesis The effect of hD1R ontumor growth was monitored with tumor bearing mice model. The effect of hD1Ron tumor necrosis and hypoxia was detected by H&E and immunofluorescence staining.The effect of hD1R on the proliferation of tumor cells was tested with MTT. The effect ofhD1R on the apoptosis of tumor cells was examined by flow cytometry withAnnexin V and PI staining. The effect of hD1R on tumor vascular pericytes recruitment and tumor vascular ultrastructure was examined with immunofluorescencestaining and scanning electron microscope technique.7. Toxicity and side effects of hD1R The toxicity of hD1R was detected with H&Estaining for mouse main organ sections; The effect of D1R on the differentiation of mouseimmune cells were analyzed by flow cytometry for bone marrow cells, thymus cells andspleen cells.8. Molecular mechanisms of Notch signaling on vascular structure and function RNAwas extracted from separate retinal vascular tissue and subjected to the mouse geneexpression profile microarray analysis. MeV-TM4and GSEA software was used forthe cluster analysis and to find the difference of genes; DAVID online software wasused for functional classification and real-time quantitative PCR was employed toverify different expressed genes.RESULTS1. Notch signaling knockout and ovexpression in endothelial cells effects on tumorvascular structure and function1) RBP-J deficiency in endothelial cells hampered thetumor expansion revealed by tumor growth curve and disrupted the structure of tumorvessel showed by broken basemembrane. Tumor vessel was much less perfused in RBP-Jdeficient mice, leading to an increase in tumor hypoxia.2) Inducible Notch activation intumor vessel endothelial cells inhibit tumor angiogenesis and normalize tumor vasculatureto inhibit the growth of tumor cells.2. We constructed, expressed and purified a novel vascular targeting activator forNotch signaling--hD1R and proved it could activate Notch signaling efficientlyby targeting vascular endothelial. We made Notch ligand Dll1(Delta-like1) DSLdomain and RGD nonapeptide specific ligand for integrin αVβ3fuse together, expressedTrx-hD1R successfully in Escherichia coli, and we alsoexpressed Trx-hD1S (Stop codon) and Trx-hD1D (DGR nonapeptide) as control. Wepurified the protein with Nickel ion chelating beads and purified again after the protein digestion with thrombin. We demonstrated hD1R could bind to endothelial cell membranesurface with immunofluorescence of cell climbing glass, flow cytometry and celladhesion assay. We also proved hD1R could activate the Notch signaling efficiently withimmunofluorescence of cell climbing glass, Western Blot and real-time quantitative PCR.Endocytosis inhibitor Dynasore could inhibit the activation of Notch signaling activatedby hD1R. All these results suggest that hD1R activate Notch signaling mainly throughbinding to the endothelial cell membrane surface to promote endocytosis ofNotch receptor extracellular domain.3. hD1R inhibited network formation and angiogenesis both in vitro and in vivo.We found that hD1R could obviously inhibit the lumen branches and lumen length withthe formation of network formation assay in our in vitro experiments. fibrin bead assayshowed that hD1R could inhibit endothelial cell sprouting and tube formation.hD1R could obviously inhibit the vessel network formation of the ganglion cell layer within vivo retinal neovascularization in neonatal mouse model. And these results were mainlyrealized through the inhibition of endothelial cell proliferation and tip cell formation.4. hD1R could inhibit tumor growth through inhibiting tumor angiogenesis. Weconfirmed that hD1R could inhibit the growth of U87, LLC and MCF-7with the nudemice tumor bearing mice modle. Tumor section staining showed that the tumornecrosis region and tumor hypoxia significantly increased. Immunofluorescence stainingof CD31showed significantly lower angiogenesis in hD1R group. Resluts from MTTshowed that hD1R had no effet on tumor cell proliferation. Results from Annexin V and PIdouble staining flow cytometry showed that hD1R had no effet on tumor cell apoptosis.These results demonstrated that hD1R inhibit tumor growth through inhibiting tumorangiogenesis, which led increased necrosis hypoxia region.5. hD1R could normalize both tumor vascular structure and function. Results fromCD31and NG2double staining immunofluorescence of tumour sections showed thattumor vascular pericytes coverage increased significantly in hD1R treated group. Resultsfrom SEM showed that there were fibrous materials and leakage-like materials within andout of microvessels in the control, and these changes were significantly attenuated in the hD1R-treated group. Results from immunofluorescence of Lectin, which was injectedthrough mouse tail vein, showed that Lectin positive signals increased significantly inhD1R-treated group, suggested that hD1R promoted tumor tissue blood perfusion. Allthese result demonstrated that Notch signaling activated by hD1R could normalizetumor vascular structure and function.6. D1R has no toxicity and side effect in mice. Results from H&E staining of mousemain organs including brain, heart, liver, lung and kidney showed that there are nosignificant pathological changes in these organs. Flow cytometry showed that D1R had noeffect on mouse bone marrow B cell, thymus T cell, spleen B cell differentiation.7. Notch signaling regulates vascular structure and function in physiologicalcondition. Immunofluorescence and Western Blot of HD1R or GSI treated HUVECsshowed that hD1R up-regulated the expression of endothelial cell adhesion moleculesVE-Cadherin and β-Catenin while GSI down-regulated the expression of these twomolecules. Furthermore translocation of β-Catenin from cell membrane to cytoplasm wasalso increased in GSI treated HUVECs. Results from mIgG immunofluorescen staining ofbrain sections and whole-mount staining of retina showed that transcription factor RBP-Jdeletion led to leakage of blood in retina and epithalamus in Mx-Cre;RBP-Jf/fmice.α-SMA positive pericytes increased significantly in hD1R-treated group compared withthe control. Three dimensional reconstruction of confocal scanning showed clearly thatα-SMAaround the vessel in hD1R-treated group was higher than that of control group.8. Notch signaling may regulate vascular structure and function through LAMB1,PCDH18, ANGPTL4and HSPG2. RNA was extracted from DMSO or GSI treatedretinal vascular tissue and subjected to mouse gene expression profile analysis.231geneswere found to be differentially expressed showed by the cluster analysis with MeV-TM4software.113genes with the same expression pattern were clustered together with GSEAsoftware. These differentially expressed genes were subjected to DAVID online softwarefor functional classification.17genes related to cell adhesion and9different genesrelated to extracellular matrix were selected for further analysis. Real time quantitativePCR verified the differentially expression of LAMB1, PCDH18, ANGPTL4and HSPG2, implying that they act downstream of Notch signaling.CONCLUSIONRBP-J deficiency in endothelial cells hampered the tumor expansion and disrupted thestructure and function of tumor vessel. Inducible Notch activation in tumor vesselendothelial cells inhibit tumor angiogenesis and normalize tumor vasculature to inhibit thegrowth of tumor cells. These results verified the crucial role of Notch signal in tumorvascularization. Further, we constructed, expressed and purified a novel vasculartargeting Notch ligand--hD1R. It could bind to endothelial cell membrane surface bothin vitro and in vivo, and activate Notch signaling efficiently. Functional study resultsshowed that hD1R could inhibit network formation and angiogenesis both in vitro and invivo through inhibiting endothelial cell proliferation and tip cell formation. hD1R couldinhibit tumor growth through inhibiting tumor angiogenesis, which increased tumornecrosis and hypoxia. hD1R also promoted the maturation and stabilization of normalvessel by strengthening the cell adhesion of endothelial cells and promoting thecoverage of mural cells. hD1R also normalized tumor vasculature and improved itsfunction. LAMB1, PCDH18, ANGPTL4and HSPG2may function downstream of Notchsignal, screened by gene expression profile analysis and verified by quantitative PCRanalysis. Our study may shed some light on the intention of novel tumor vessel targetedtherapy. |
PDF Full Text Request