Angiogenesis and Microvascular Permeability.

Angiogenesis, the formation of new blood microvessels, is a necessary physiological process for tissue generation and repair. Sufficient blood supply to the tissue is dependent on microvascular density, while the material exchange between the circulating blood and the surrounding tissue is controlled by microvascular permeability. We thus begin this article by reviewing the key signaling factors, particularly vascular endothelial growth factor (VEGF), which regulates both angiogenesis and microvascular permeability. We then review the role of angiogenesis in tissue growth (bone regeneration) and wound healing. Finally, we review angiogenesis as a pathological process in tumorigenesis, intraplaque hemorrhage, cerebral microhemorrhage, pulmonary fibrosis, and hepatic fibrosis. Since the glycocalyx is important for both angiogenesis and microvascular permeability, we highlight the role of the glycocalyx in regulating the interaction between tumor cells and endothelial cells (ECs) and VEGF-containing exosome release and uptake by tumor-associated ECs, all of which contribute to tumorigenesis and metastasis.

Heparan Sulfate Modulation Affects Breast Cancer Cell Adhesion and Transmigration across In Vitro Blood-Brain Barrier.

The disruption of endothelial heparan sulfate (HS) is an early event in tumor cell metastasis across vascular barriers, and the reinforcement of endothelial HS reduces tumor cell adhesion to endothelium. Our recent study showed that while vascular endothelial growth factor (VEGF) greatly reduces HS at an in vitro blood-brain barrier (BBB) formed by human cerebral microvascular endothelial cells (hCMECs), it significantly enhances HS on a breast cancer cell, MDA-MB-231 (MB231). Here, we tested that this differential effect of VEGF on the HS favors MB231 adhesion and transmigration. We also tested if agents that enhance endothelial HS may affect the HS of MB231 and reduce its adhesion and transmigration. To test these hypotheses, we generated an in vitro BBB by culturing hCMECs on either a glass-bottom dish or a Transwell filter. We first quantified the HS of the BBB and MB231 after treatment with VEGF and endothelial HS-enhancing agents and then quantified the adhesion and transmigration of MB231 across the BBB after pretreatment with these agents. Our results demonstrated that the reduced/enhanced BBB HS and enhanced/reduced MB231 HS increase/decrease MB231 adhesion to and transmigration across the BBB. Our findings suggest a therapeutic intervention by targeting the HS-mediated breast cancer brain metastasis.

Targeting FoxO1 in traditional Chinese medicine for the treatment of diabetes.

Traditional Chinese medicine (TCM) encompasses treatment strategies for diabetes, which is referred to as "Consumptive Thirsty" syndrome. Recently, there has been discovery regarding the mapping between TCM and signaling molecules, which has revealed a remarkable consistency between TCM and modern medicine from a molecular perspective. In this manuscript, we have summarized the etiology and treatment strategies for diabetes in TCM and have examined these strategies in the context of molecular mechanisms. Our review demonstrates that the targeting molecule of TCM for the treatment of diabetes is FoxO1, a transcription factor that plays a pivotal role in regulating gluconeogenesis and glycogenolysis. TCM ranks the development of diabetes into three stages and utilizes different herbal formulas to control FoxO1 accordingly. At Stage 1, TCM inhibits FoxO1 by lowering its expression in the lung. At Stage 2, TCM increases the expression of FoxO1 by suppressing its activity in the stomach. At Stage 3, TCM utilizes the famous herbal formula Liuwei Dihuang Pill to amplify the expression of FoxO1, and to enhance the concentrations of potassium, phosphorus, and Wnt, but to reduce the concentration of calcium. These TCM treatment strategies are in accordance with corresponding mechanisms in modern medicine.

Inhibition of Abl Kinase by Imatinib Can Rescue the Compromised Barrier Function of 22q11.2DS Patient-iPSC-Derived Blood-Brain Barriers.

We have previously established that the integrity of the induced blood-brain barrier (iBBB) formed by brain microvascular endothelial cells derived from the iPSC of 22q11.2 DS (22q11.2 Deletion Syndrome, also called DiGeorge Syndrome) patients is compromised. We tested the possibility that the haploinsufficiency of CRKL, a gene within the 22q11.2 DS deletion region, contributes to the deficit. The CRKL is a major substrate of the Abl tyrosine kinase, and the Abl/CRKL signaling pathway is critical for endothelial barrier functions. Imatinib, an FDA-approved drug, inhibits Abl kinase and has been used to treat various disorders involving vascular leakages. To test if imatinib can restore the compromised iBBB, we treated the patient's iBBB with imatinib. After treatment, both trans-endothelial electrical resistance and solute permeability returned to comparable levels of the control iBBB. Correspondingly, changes in tight junctions and endothelial glycocalyx of the iBBB were also restored. Western blotting showed that imatinib increased the level of active forms of the CRKL protein. A transcriptome study revealed that imatinib up-regulated genes in the signaling pathways responsible for the protein modification process and down-regulated those for cell cycling. The KEGG pathway analysis further suggested that imatinib improved the gene expression of the CRKL signaling pathway and tight junctions, which agrees with our expectations and the observations at protein levels. Our results indicate that the 22q11.2DS iBBB is at least partially caused by the haploinsufficiency of CRKL, which can be rescued by imatinib via its effects on the Abl/CRKL signaling pathway. Our findings uncover a novel disease mechanism associated with 22q11.2DS.

Glycocalyx Acts as a Central Player in the Development of Tumor Microenvironment by Extracellular Vesicles for Angiogenesis and Metastasis.

Angiogenesis in tumor growth and progression involves a series of complex changes in the tumor microenvironment. Extracellular vesicles (EVs) are important components of the tumor microenvironment, which can be classified as exosomes, apoptotic vesicles, and matrix vesicles according to their origins and properties. The EVs that share many common biological properties are important factors for the microenvironmental modification and play a vital role in tumor growth and progression. For example, vascular endothelial growth factor (VEGF) exosomes, which carry VEGF, participate in the tolerance of anti-angiogenic therapy (AAT). The glycocalyx is a mucopolysaccharide structure consisting of glycoproteins, proteoglycans, and glycosaminoglycans. Both endothelial and tumor cells have glycocalyx at their surfaces. Glycocalyx at both cells mediates the secretion and uptake of EVs. On the other hand, many components carried by EVs can modify the glycocalyx, which finally facilitates the development of the tumor microenvironment. In this short review, we first summarize the role of EVs in the development of the tumor microenvironment. Then we review how the glycocalyx is associated with the tumor microenvironment and how it is modulated by the EVs, and finally, we review the role of the glycocalyx in the synthesis, release, and uptake of EVs that affect tumor microenvironments. This review aims to provide a basis for the mechanistic study of AAT and new clues to address the challenges in AAT tolerance, tumor angiogenesis and metastasis.

Differential effects of vascular endothelial growth factor on glycocalyx of endothelial and tumor cells and potential targets for tumor metastasis.

On the surface of every mammalian cell, there is a matrix-like glycocalyx (GCX) consisting of proteoglycans and glycosaminoglycans (GAGs). Disruption of endothelial cell (EC) GCX by a vascular endothelial growth factor (VEGF, VEGF-A165), a tumor secretion, was found to be an early event in tumor cell (TC) metastasis across vascular barriers. However, how the TC secretion VEGF affects its own GCX is unknown. To investigate the VEGF effect on TC GCX and to elucidate the ultrastructural organization of EC and TC GCX and their alteration by VEGF, we employed super-resolution stochastic optical reconstruction microscopy to observe the spatio-chemical organizations of the heparan sulfate (HS) and hyaluronic acid (HA), two representative GAGs of GCX, on human cerebral microvascular endothelial cells (hCMEC) and malignant breast cancer cells MDA-MB-231 (MB231). We found that HS and HA have distinct organizations on hCMEC and MB231. Only HS of hCMEC is perpendicular to the cell surface, while HA of hCMEC as well as HS and HA of MB231 all lie in the same plane as the cell surface where they appear to weave into a 2D network covering the cell. We also found that VEGF significantly reduces the length and coverage of HS on hCMEC but does not change the thickness and coverage of HA on hCMEC. On the contrary, VEGF significantly enhances the coverage of HS and HA on MB231 although it does not alter the thickness. The differential effects of VEGF on the GCX of TC and that of EC may favor TC adhesion and transmigration across EC barriers for their metastasis.

Quantification of In Vitro Blood-Brain Barrier Permeability.

In vitro blood-brain barrier (BBB) models have been widely used to simulate in vivo models due to their low cost, feasibility, and repeatability. To serve as a valid substitute, the in vitro BBB should have the similar barrier function as that in vivo. This chapter summarizes the detailed methods for quantifying the barrier function, e.g., the permeability of the BBB to water, ions, and solutes for an in vitro BBB generated on the Transwell filter.

Direct Current Stimulation Disrupts Endothelial Glycocalyx and Tight Junctions of the Blood-Brain Barrier in vitro.

Transcranial direct current stimulation (tDCS) is a non-invasive physical therapy to treat many psychiatric disorders and to enhance memory and cognition in healthy individuals. Our recent studies showed that tDCS with the proper dosage and duration can transiently enhance the permeability (P) of the blood-brain barrier (BBB) in rat brain to various sized solutes. Based on the in vivo permeability data, a transport model for the paracellular pathway of the BBB also predicted that tDCS can transiently disrupt the endothelial glycocalyx (EG) and the tight junction between endothelial cells. To confirm these predictions and to investigate the structural mechanisms by which tDCS modulates P of the BBB, we directly quantified the EG and tight junctions of in vitro BBB models after DCS treatment. Human cerebral microvascular endothelial cells (hCMECs) and mouse brain microvascular endothelial cells (bEnd3) were cultured on the Transwell filter with 3 µm pores to generate in vitro BBBs. After confluence, 0.1-1 mA/cm2 DCS was applied for 5 and 10 min. TEER and P to dextran-70k of the in vitro BBB were measured, HS (heparan sulfate) and hyaluronic acid (HA) of EG was immuno-stained and quantified, as well as the tight junction ZO-1. We found disrupted EG and ZO-1 when P to dextran-70k was increased and TEER was decreased by the DCS. To further investigate the cellular signaling mechanism of DCS on the BBB permeability, we pretreated the in vitro BBB with a nitric oxide synthase (NOS) inhibitor, L-NMMA. L-NMMA diminished the effect of DCS on the BBB permeability by protecting the EG and reinforcing tight junctions. These in vitro results conform to the in vivo observations and confirm the model prediction that DCS can disrupt the EG and tight junction of the BBB. Nevertheless, the in vivo effects of DCS are transient which backup its safety in the clinical application. In conclusion, our current study directly elucidates the structural and signaling mechanisms by which DCS modulates the BBB permeability.

Investigation of Neurodevelopmental Deficits of 22 q11.2 Deletion Syndrome with a Patient-iPSC-Derived Blood-Brain Barrier Model.

The blood-brain barrier (BBB) is important in the normal functioning of the central nervous system. An altered BBB has been described in various neuropsychiatric disorders, including schizophrenia. However, the cellular and molecular mechanisms of such alterations remain unclear. Here, we investigate if BBB integrity is compromised in 22q11.2 deletion syndrome (also called DiGeorge syndrome), which is one of the validated genetic risk factors for schizophrenia. We utilized a set of human brain microvascular endothelial cells (HBMECs) derived from the induced pluripotent stem cell (iPSC) lines of patients with 22q11.2-deletion-syndrome-associated schizophrenia. We found that the solute permeability of the BBB formed from patient HBMECs increases by ~1.3-1.4-fold, while the trans-endothelial electrical resistance decreases to ~62% of the control values. Correspondingly, tight junction proteins and the endothelial glycocalyx that determine the integrity of the BBB are significantly disrupted. A transcriptome study also suggests that the transcriptional network related to the cell-cell junctions in the compromised BBB is substantially altered. An enrichment analysis further suggests that the genes within the altered gene expression network also contribute to neurodevelopmental disorders. Our findings suggest that neurovascular coupling can be targeted in developing novel therapeutical strategies for the treatment of 22q11.2 deletion syndrome.

The Glycocalyx and Its Role in Vascular Physiology and Vascular Related Diseases.

PURPOSE: In 2007 the two senior authors wrote a review on the structure and function of the endothelial glycocalyx layer (Weinbaum in Annu Rev Biomed Eng 9:121-167, 2007). Since then there has been an explosion of interest in this hydrated gel-like structure that coats the luminal surface of endothelial cells that line our vasculature due to its important functions in (A) basic vascular physiology and (B) vascular related diseases. This review will highlight the major advances that have occurred since our 2007 paper. METHODS: A literature search mainly focusing on the role of the glycocalyx in the two major areas described above was performed using electronic databases. RESULTS: In part (A) of this review, the new formulation of the century old Starling principle, now referred to as the Michel-Weinbaum glycoclayx model or revised Starling hypothesis, is described including new subtleties and physiological ramifications. New insights into mechanotransduction and release of nitric oxide due to fluid shear stress sensed by the glycocalyx are elaborated. Major advances in understanding the organization and function of glycocalyx components, and new techniques for measuring both its thickness and spatio-chemical organization based on super resolution, stochastic optical reconstruction microscopy (STORM) are presented. As discussed in part (B) of this review, it is now recognized that artery wall stiffness associated with hypertension and aging induces glycocalyx degradation, endothelial dysfunction and vascular disease. In addition to atherosclerosis and cardiovascular diseases, the glycocalyx plays an important role in lifestyle related diseases (e.g., diabetes) and cancer. Infectious diseases including sepsis, Dengue, Zika and Corona viruses, and malaria also involve the glycocalyx. Because of increasing recognition of the role of the glycocalyx in a wide range of diseases, there has been a vigorous search for methods to protect the glycocalyx from degradation or to enhance its synthesis in disease environments. CONCLUSION: As we have seen in this review, many important developments in our basic understanding of GCX structure, function and role in diseases have been described since the 2007 paper. The future is wide open for continued GCX research.

Blood-Brain Barrier (BBB) Permeability and Transport Measurement In Vitro and In Vivo.

Quantification of the blood-brain barrier (BBB) permeability and transport in brain tissue is crucial in understanding brain disorders and developing systemic and non-systemic drug delivery strategies to the brain. This chapter summarizes BBB permeability measurement in vitro (Part I) and the in vivo non-invasive methods for quantifying the BBB permeability to solutes and brain tissue transport in rat brain by employing intravital multiphoton microscopy and a curving fitting method by using an unsteady mass transfer mathematical model (Part II).

Numerical study on the adhesion of a circulating tumor cell in a curved microvessel.

The adhesion of a circulating tumor cell (CTC) in a three-dimensional curved microvessel was numerically investigated. Simulations were first performed to characterize the differences in the dynamics and adhesion of a CTC in the straight and curved vessels. After that, a parametric study was performed to investigate the effects of the applied driven force density f (or the flow Reynolds number Re) and the CTC membrane bending modulus Kb on the CTC adhesion. Our simulation results show that the CTC prefers to adhere to the curved vessel as more bonds are formed around the transition region of the curved part due to the increased cell-wall contact by the centrifugal force. The parametric study also indicates that when the flow driven force f (or Re) increases or when the CTC becomes softer (Kb decreases), the bond formation probability increases and the bonds will be formed at more sites of a curved vessel. The increased f (or Re) brings a larger centrifugal force, while the decreased Kb generates more contact areas at the cell-wall interface, both of which are beneficial to the bond formation. In the curved vessel, it is found that the site where bonds are formed the most (hotspot) varies with the applied f and the Kb. For our vessel geometry, when f is small, the hotspot tends to be within the first bend of the vessel, while as f increases or Kb decreases, the hotspot may shift to the second bend of the vessel.

Resistance Mechanisms of Anti-angiogenic Therapy and Exosomes-Mediated Revascularization in Cancer.

Anti-angiogenic therapies (AATs) have been widely used for cancer treatment. But the beneficial effects of AATs are short, because AAT-induced tumor revascularization facilitates the tumor relapse. In this mini-review, we described different forms of tumor neovascularization and revascularization including sprouting angiogenesis, vessel co-option, intussusceptive angiogenesis, and vasculogenic mimicry, all of which are closely mediated by vascular endothelial growth factor (VEGF), angiopoietins, matrix metalloproteinases, and exosomes. We also summarized the current findings for the resistance mechanisms of AATs including enhancement in pro-angiogenic cytokines, heterogeneity in tumor-associated endothelial cells (ECs), crosstalk between tumor cells and ECs, masking of extracellular vesicles, matrix stiffness and contributions from fibroblasts, macrophages and adipocytes in the tumor microenvironment. We highlighted the revascularization following AATs, particularly the role of exosome stimulating factors such as hypoxia and miRNA, and that of exosomal cargos such as cytokines, miRNAs, lncRNAs, and circRNAs from the tumor ECs in angiogenesis and revascularization. Finally, we proposed that renormalization of tumor ECs would be a more efficient cancer therapy than the current AATs.

Modulation of solute diffusivity in brain tissue as a novel mechanism of transcranial direct current stimulation (tDCS).

The breadth of brain disorders and functions reported responsive to transcranial direct current stimulation (tDCS) suggests a generalizable mechanism of action. Prior efforts characterized its cellular targets including neuron, glia and endothelial cells. We propose tDCS also modulates the substance transport in brain tissue. High resolution multiphoton microscopy imaged the spread across rat brain tissue of fluorescently-labeled solutes injected through the carotid artery after tDCS. The effective solute diffusion coefficient of brain tissue (Deff) was determined from the spatio-temporal solute concentration profiles using an unsteady diffusion transport model. 5-10 min post 20 min-1 mA tDCS, Deff increased by ~ 10% for a small solute, sodium fluorescein, and ~ 120% for larger solutes, BSA and Dex-70k. All increases in Deff returned to the control level 25-30 min post tDCS. A mathematical model for Deff in the extracelluar space (ECS) further predicts that this dose of tDCS increases Deff by transiently enhancing the brain ECS gap spacing by ~ 1.5-fold and accordingly reducing the extracellular matrix density. The cascades leading ECS modulation and its impact on excitability, synaptic function, plasticity, and brain clearance require further study. Modulation of solute diffusivity and ECS could explain diverse outcomes of tDCS and suggest novel therapeutic strategies.

Transcellular Model for Neutral and Charged Nanoparticles Across an In Vitro Blood-Brain Barrier.

PURPOSE: The therapeutic drug-loaded nanoparticles (NPs, 20-100 nm) have been widely used to treat brain disorders. To improve systemic brain delivery efficacy of these NPs, it is necessary to quantify their transport parameters across the blood-brain barrier (BBB) and understand the underlying transport mechanism. METHODS: Permeability of an in vitro BBB, bEnd3 (mouse brain microvascular endothelial cells) monolayer, to three neutral NPs with the representative diameters was measured using an automated fluorometer system. To elucidate the transport mechanism of the neutral NPs across the in vitro BBB, and that of positively charged NPs whose BBB permeability was measured in a previous study, we developed a novel transcellular model, which incorporates the charge of the in vitro BBB, the mechanical property of the cell membrane, the ion concentrations of the surrounding salt solution and the size and charge of the NPs. RESULTS: Our model indicates that the negative charge of the surface glycocalyx and basement membrane of the BBB plays a pivotal role in the transcelluar transport of NPs with diameter 20-100 nm across the BBB. The electrostatic force between the negative charge at the in vitro BBB and the positive charge at NPs greatly enhances NP permeability. The predictions from our transcellular model fit very well with the measured BBB permeability for both neutral and charged NPs. CONCLUSION: Our model can be used to predict the optimal size and charge of the NPs and the optimal charge of the BBB for an optimal systemic drug delivery strategy to the brain.

High-Throughput Tumor-on-a-Chip Platform to Study Tumor-Stroma Interactions and Drug Pharmacokinetics.

Drug screening in oncology, especially for triple-negative breast cancer (TNBC), has high demand but remains unsatisfactory. Currently available models are either nonrepresentative of the complex tumor microenvironment or only suitable for low throughput screening, resulting in a low-yield success for drug development. To tackle these issues, the L-TumorChip system is developed in this study. It is a three-layered microfluidic tumor-on-a-chip platform integrating tumor microvasculature and tumor-stromal microenvironment with high throughput screening capability. Its layered and modular design is readily scalable through simple integration of multiple units. Here, L-TumorChip is validated with a TNBC model. The L-TumorChip system emulates certain tumor-stroma complexities and tumor-endothelium interactions, including TNBC invasion through the leaky microvasculature and angiogenesis. Additionally, with this L-TumorChip, the influence of different stromal cells, including normal fibroblasts, mesenchymal stem cells, and cancer-associated fibroblasts (CAF), on cancer cell growth as well as the stromal effects on drug responses to doxorubicin treatment is investigated. The presence of CAF delays drug pharmacokinetics, while apoptotic responses indicated by caspase-3 activities are higher in coculture with normal fibroblasts. Collectively, the L-TumorChip system represents a translational high-throughput screening toolkit that enables drug screening with a scenario closer to the in vivo conditions. This potential use may therefore facilitate development of new cancer drugs.

In Vivo Modulation of the Blood-Brain Barrier Permeability by Transcranial Direct Current Stimulation (tDCS).

tDCS has been used to treat various brain disorders and its mechanism of action (MoA) was found to be neuronal polarization. Since the blood-brain barrier (BBB) tightly regulates the neuronal microenvironment, we hypothesized that another MoA of tDCS is direct vascular activation by modulating the BBB structures to increase its permeability (P). To test this hypothesis, we used high resolution multiphoton microscopy to determine P of the cerebral microvessels in rat brain. We found that 20 min 0.1-1 mA tDCS transiently increases P to a small solute, sodium fluorescein (MW 376) and to a large solute, Dextran-70k, with a much higher increase in P to the large solute. By pretreating the vessel with a nitric oxide synthase inhibitor, we revealed that the tDCS-induced increase in P is NO dependent. A transport model for the BBB was further employed to predict the structural changes by the tDCS. Comparing model predictions with the measured data suggests that tDCS increases P by temporarily disrupting the structural components forming the paracellular pathway of the BBB. That the transient and reversible increase in the BBB permeability also suggests new applications of tDCS such as a non-invasive approach for brain drug delivery through the BBB.

Numerical study on the dynamics of primary cilium in pulsatile flows by the immersed boundary-lattice Boltzmann method.

An explicit immersed boundary-lattice Boltzmann method is applied to numerically investigate the dynamics of primary cilium in pulsatile blood flows with two-way fluid-structure interaction considered. To well characterize the effect of cilium basal body on cilium dynamics, the cilium base is modeled as a nonlinear rotational spring attached to the cilium's basal end as proposed by Resnick (Biophys J 109:18-25, 2015. https://doi.org/10.1016/j.bpj.2015.05.031). After several careful validations, the fluid-cilium interaction system is investigated in detail at various pulsatile flow conditions that are characterized by peak Reynolds numbers ([Formula: see text]) and Womersley numbers ([Formula: see text]). The periodic flapping of primary cilium observed in our simulations is very similar to the in vivo ciliary oscillation captured by O'Connor et al. (Cilia 2:8, 2013. https://doi.org/10.1186/2046-2530-2-8). The cilium's dynamics is found to be closely related to the [Formula: see text] and [Formula: see text]. Increase the [Formula: see text] or decrease the [Formula: see text] bring to an increase in the cilium's flapping amplitude, tip angular speed, basal rotation, and maximum tensile stress. It is also demonstrated that by reducing the [Formula: see text] or enhancing the [Formula: see text] to a certain level, one can shift the flapping pattern of cilium from its original two-side one to a one-side one, making the stretch only happen on one particular side. During the flapping process, the location of the maximum tensile stress is not always found at the basal region; instead, it is able to propagate from time to time within a certain distance to the base. Due to the obstruction of the primary cilium, the distribution of wall shear stress no longer remains uniform as in the absence of cilia. It oscillates in space with the minimum magnitude which is always found near where the cilium is located. The presence of cilium also reduces the overall level of wall shear stress, especially at the region near the cilium's anchor point.

Anti-angiogenesis triggers exosomes release from endothelial cells to promote tumor vasculogenesis.

Although anti-angiogenic therapies (AATs) have some effects against multiple malignancies, they are limited by subsequent tumor vasculogenesis and progression. To investigate the mechanisms by which tumor vasculogenesis and progression following AATs, we transfected microRNA (miR)-9 into human umbilical vein endothelial cells (HUVECs) to mimic the tumor-associated endothelial cells in hepatocellular carcinoma and simulated the AATs in vitro and in vivo. We found that administration of the angiogenesis inhibitor vandetanib completely abolished miR-9-induced angiogenesis and promoted autophagy in HUVECs, but induced the release of vascular endothelial growth factor (VEGF)-enriched exosomes. These VEGF-enriched exosomes significantly promoted the formation of endothelial vessels and vasculogenic mimicry in hepatocellular carcinoma and its progression in mice. Anti-autophagic therapy is proposed to improve the efficacy of AATs. However, similar effects by AATs were observed with the application of anti-autophagy by 3-methyladenine. Our results revealed that tumor vasculogenesis and progression after AATs and anti-autophagic therapies were due to the cross-talk between endothelial and tumor cells via VEGF-enriched exosomes.