Angiogenesis in Adipose Tissue: The Interplay Between Adipose and Endothelial Cells

1Recent findings have led us to reconsider the notion of adipose tissue (AT) being a mere storage depot for body energy. Instead, ATs are emerging as endocrine and immunologically active organs with multiple effects on the regulation of systemic energy homeostasis. AT is classified as either white (WAT), brown (BAT), or beige/brite based on whether it functions as energy storage or thermogenic organ. Adipocytes in BAT are rich in mitochondria and generate chemical energy in the form of heat. BAT expands when an organism is exposed to prolonged cold conditions and is linked with increased insulin sensitivity. WAT stores energy in the form of triglycerides in lipid droplets and is an endocrine organ that releases hormones, growth factors, and adipokines such as leptin and adiponectin. AT consists of not only adipocytes (40–50%) but also connective tissue matrix, vascular and neural cells, and non-adipocyte cells called stromal vascular fraction (SVF). SVF includes preadipocytes, immune cells (macrophages, natural killer cells, B-lymphocytes, and T-lymphocytes), endothelial cells (ECs), vascular progenitors, fibroblasts, and mesenchymal stem cells. The WAT is mainly divided into two types based on its distribution site, the subcutaneous (SAT) and visceral AT (VAT). During overnutrition, the surplus energy is stored as triglycerides in the SAT leading to AT expansion. AT generates new adipocytes (hyperplasia) and enlarges existing adipocytes (hypertrophy; ). These processes change the structure and function of AT, and these events are defined as “AT remodeling”.

2The prevalence of obesity has increased all over the world during the last 50 years and more than 650 million obese adults were reported in 2016. According to the World Health Organization (WHO), obesity is an “abnormal or excessive fat accumulation that may impair health.” The general measurement of obesity is the body mass index (BMI). Subjects with a BMI ≥ 25 kg/m2 are classified to be overweight, while those with a BMI ≥ 30 kg/m2 are considered obese. A major cause of obesity is the imbalance between energy intake and expenditure, resulting in the storage of triglycerides in AT. Many factors, including genetics, epigenetics, ethnicity, and environmental factors are involved in the complex pathogenesis of obesity. Obesity is a critical risk factor for many diseases, including type 2 diabetes, non-alcoholic fatty liver disease, hypertension, and cardiovascular disease (CVD); it is also linked to several cancers. In obesity, alteration in AT remodeling may induce the dysregulation of AT-secreted adipokines and cytokines, and the increased secretion of pro-inflammatory molecules may promote systemic low-grade inflammations. This results in macrophage and T-cell infiltration into AT. However, AT remodeling does not necessarily lead to these obesity-related diseases; several studies have reported that a subgroup of obese individuals are metabolically “healthy” and display normal physiology and biomarker profiles. This suggests the presence of differences in pathways that regulate AT dysfunction, which further leads to dysfunction in other organs.

3The vascular circulatory system forms an extensive network of arteries, veins, and capillaries and is important for continuous supply and delivery of nutrients and oxygen to all tissues. The blood vessels are lined with a thin layer of ECs, the main drivers of sprouting angiogenesis. In healthy adult tissue, ECs remain quiescent. The switch from quiescent to proliferative/angiogenic ECs is mediated by changes in metabolism and angiogenic factors. Angiogenesis requires a balance between pro-angiogenic and anti-angiogenic molecules. Overexpression of pro-angiogenic molecules can also lead to EC dysfunction. Dysfunctional ECs cannot induce normal angiogenesis, which leads to lesser blood vessel formation. Lower vascular density is a high-risk factor for hypoxia and inflammation and is associated with various diseases, such as CVDs, obesity, and cancer.

4Obesity is accompanied by EC dysfunction and decreased vascular density (Figure 1). This is mainly due to disbalance and overexpression of pro-angiogenic and pro-inflammatory stimulus. Therefore, understanding the precise relationship between adipocytes and ECs are of significant importance. In this review, we have discussed the function of ECs in AT with respect to the influence of obesity on angiogenesis, the crosstalk between adipocytes and ECs, and how obesity can influence ECs (dys)function.

5Adipose tissue is a highly vascularized tissue, and the density of blood vessels is important for the regulation of adipocyte function and adipogenesis. In addition to nutrients and oxygen, the vessels also transport growth factors, cytokines, and hormones that are required for adipocyte function, growth, and survival. Furthermore, the vessels regulate the transport of adipokines, cytokines, and growth factors from the AT to other organs and thus promote the endocrine function of the AT. The vascular system controls alteration in the AT microenvironment, including acidosis and hypoxia, which influences the adipocyte function, preadipocyte differentiation, and the AT mass (; Figure 1). There are different possible triggers for angiogenesis in AT, the angiogenic expansion may be due to signals emanating from proliferating and enlarging adipocytes or it may be triggered by metabolic signals with the enlargement of adipocytes as a parallel phenomenon. These processes are not mutually exclusive, and a combination of the two is most probably responsible for angiogenesis. The expansion of AT is angiogenesis-dependent. With a high nutrient availability, the adipocytes store lipids in lipid droplets, and as their size expands oxygen availability is reduced. This mild hypoxic status can induce angiogenesis and remodeling of extracellular matrix to reduce hypoxia. In healthy AT, the vessels are lined by a single monolayer of quiescent ECs, which can rapidly switch to the angiogenic/proliferative state in the presence of angiogenic and metabolic signals to form new blood vessels. AT produces and secretes various angiogenic factors such as angiopoietin-2 (Angpt2) and vascular endothelial growth factor (VEGF) as well as adipokines such as leptin and adiponectin, which influence and modulate angiogenesis and the vascular structure. This suggests an autoregulatory function for angiogenesis in AT.

6Interestingly, studies show the presence of precursor cells in the vessel wall, which have the capacity to differentiate into ECs and/or adipocytes in WAT and BAT depots. Tang et al. identified the presence of white adipocyte progenitors in the mural compartment of adipose vasculature, but not in the vasculature of other tissues. These mural cells had high adipogenic potential. These results indicate a link between ECs and adipocytes in terms of their interchangeability in the presence of a possible “switch” and the cell–cell interaction.

7Obesity is characterized by a rapid expansion of AT, which affects tissue vascularization. The lack of vessels leads to decreased oxygen supply in adipocytes, which leads to hypoxia that promotes inflammation, and inadequate vessel maintenance and growth. During hypoxia, the hypoxia-inducible factors (HIFs) signaling is activated. Once translocated to the nucleus, HIF1α dimerizes with HIF1β and forms the functional transcription factor HIF1. HIF1 can bind to hypoxia response elements of target genes, including VEGFA and Angpt2, which induce angiogenic response. HIF1α is reportedly increased in the AT of obese patients and its expression was found to be reduced after surgery-induced weight loss. Genetic deletion of HIF1α in adipocytes decreases the risk of obesity-induced inflammation and insulin resistance. Hypoxia stimulates inflammation and results in the accumulation of macrophages and other immune cells. Activation of HIF-signaling pathways in macrophages in obese mice leads to induction of platelet-derived growth factor (PDGF) expression, which is likely to induce the tube formation of ECs to improve vascular density. During inflammation, several factors released from immune cells, such as tumor necrosis factor alpha (TNFα; ), act on ECs activating the signaling pathway of factor nuclear kappa B (NFκB; ). Additionally, with an increase in angiogenesis, more blood vessels are formed to provide oxygen and nutrients for the metabolic needs of the cells present at the inflammatory sites.

8During obesity, there is a higher release in free fatty acids (FFAs) from adipocytes into the bloodstream, owing to a saturated storage capacity and dysfunctional adipocyte metabolism (; Figure 1). The FFAs can be taken up by ECs through either fatty acid transport proteins 3 and 4 (FATP3/FATP4) or the scavenger receptor cluster of differentiation CD36 (CD36) and can be converted into hydrophilic acyl-CoAs. These are involved in several processes, including the synthesis of ceramides, which are membrane components. Furthermore, FFAs can activate the NF-κB through Toll-like receptor (TLR) signaling, which in turn can activate the process of inflammation.

9The VEGF/VEGF-receptor (VEGFR) system is the main regulator of the angiogenic activity in AT and is expressed in the SVF and mature adipocytes. In a murine model with a VEGF deletion in AT, the reduced vascular density accompanied by enhanced hypoxia, inflammation, and apoptosis was observed. VEGF binds to its tyrosine kinase receptors VEGFR1 and VEGFR2 in order to perform various biological functions. VEGFA plays the role in AT angiogenesis and its expression is increased during the adipocyte differentiation. VEGF secreted by AT also stimulates the proliferation of vascular smooth muscle cells (VSMCs). demonstrated that VSMC cultivated in the adipocyte-conditioned medium had a higher VEGFR1 and VEGFR2 expression, as well as higher VEGF secretion. Although VEGF binds to both receptors, VEGFR2 mediates most of the cellular responses to induce migration, survival, and proliferation of ECs. Blocking VEGFR2 limits diet-induced AT expansion by decreasing angiogenesis and adipogenesis. However, VEGFR2 has an opposite effect in lymphatic vessels. Mice with a genetic deletion of neuropilin 1 (Nrp1) and Vegfr1 are resistant to diet-induced obesity by reduced lacteal chylomicron uptake. The inhibition of Vegfr2 restores the permeable junction and rescued chylomicron transport and the mice are not resistant to diet-induced obesity anymore.

10Overexpression of VEGF in WAT and BAT in mice led to increased number and size of blood vessels, increased insulin sensitivity, and improved glucose tolerance. The transgenic mice were protected from diet-induced obesity and local hypoxia, which was indicated by decreased expression of HIF1. reported that VEGF repression in mice also leads to resistance of diet-induced obesity and surprisingly higher expression of BAT markers, including uncoupling protein 1 (UCP1) and Cell Death Inducing DFFA Like Effector A (CIDEA; ). VEGFB was found to be expressed in ECs of the skeletal muscles, heart, and BAT. It binds to VEGFR1 and Nrp1 and increases the expression of FATP3/FATP4 to induce the FFA uptake. Another study demonstrated that the binding of VEGFB to VEGFR1 results in the activation of VEGF/VEGFR2 pathway and improves the capillary density and insulin signaling in AT.

11Genetic deletion, as well as pharmacological inhibition of endothelial VEGFR1, increased adipose angiogenesis and browning of SAT, leading to elevated thermogenesis. Additionally, the anti-VEGFR1 treatment led to higher expression levels of UCP1 and smaller adipocyte size in WAT. This implied the beginning of browning in WAT, which effected the energy expenditure. A high-fat diet in transgenic mice with a deletion of VEGFR1 in ECs resulted in reduced body weight and body fat mass. Furthermore, the VEGFR1 knockout significantly ameliorated obesity-induced dysfunction by lowering the levels of FFAs, glycerol, triglyceride, glucose, and insulin in the blood of these mice. These results demonstrate the regulatory role and crosstalk between the different VEGF-members and their receptors that are involved in various mechanisms. Also, the potential of anti-angiogenic compounds for the treatment of obesity was investigated. Angiostatin and endostatin, two inhibitors of angiogenesis, inhibit weight gain in ob/ob mice and diet-induced obesity. Endostatin inhibits dietary-induced obesity and adipogenesis via decreased expression of mTOR. Furthermore, treatment with endostatin had a preventive effect on obesity-induced complication such as glucose intolerance.

12In addition to VEGF, other angiogenic factors are expressed in AT. For example, PDGF is expressed in all cell types of AT (preadipocytes, adipocytes, macrophages, and ECs), however, its expression levels may differ. Preadipocytes express more PDGF than mature adipocytes. However, since in obesity, most of the preadipocytes are differentiated into mature adipocytes, the number of preadipocytes is reduced and thus the local PDGF level decreases. To meet the demand for PDGF, AT macrophages increase the PDGF production. In obese subject, in response to the reduced vascular density in AT and increased hypoxia, macrophages express PDGF to facilitate capillary formation. It is mandatory that the balance between these two angiogenic factors is maintained in order to form functional new capillaries. Angiogenesis is coordinated by VEGF and PDGF through their related receptors on ECs and VSMCs, respectively. The expression of another angiogenic factor, fibroblast growth factor-2 (FGF-2), is increased during adipocyte differentiation and during the induction of obesity by high-fat diet in mice. In AT, FGF2 enhanced the inflammation response through NLRP3 inflammasome activation.

13Peroxisome proliferator-activated receptor gamma (PPARγ), the master regulator of adipocyte differentiation, can influence the angiogenesis. Multiple studies demonstrate that PPARγ inhibits proliferation of ECs, however, in AT angiogenesis was enhanced by PPARγ activators. AT obtained from mice and humans treated with PPARγ agonist rosiglitazone were found to exhibit increased capillary density and capillary sprouting. Co-culture experiments of adipocytes and ECs demonstrated that the PPARγ expression levels were lower in EC–adipocyte coculture than that in the adipocyte-only control, and smaller lipid droplets per adipocyte were found in the latter. Other studies described anti-inflammatory effects of PPARγ in ECs including inhibition of NF-κB signaling, decreased expression of chemokines, and proinflammatory adhesion molecules, including intercellular adhesion molecule (ICAM)-1 and vascular cell adhesion molecule (VCAM)-1. Mice with a knockdown of PPARγ in ECs that were fed a high-fat diet were found to have decreased SAT and VAT mass but increased spleen and liver weights compared to those in control mice, in spite of same body weight in both groups. Moreover, the adipocyte size was 25% lower in VAT from PPARγ knockdown mice compared to control mice. Furthermore, the knockdown mice had lower glucose and insulin levels and were more insulin sensitive. However, these mice showed a higher concentration of circulating FFA, increased hepatic CD36 expression, and increased very low-density lipoprotein (vLDL) production. Taken together, these studies demonstrate that PPARγ in ECs contributes to metabolic response in various organs.

14With the development of obesity, AT becomes dysfunctional. In obesity, a higher release of FFAs from the adipocytes into the bloodstream is observed. FFAs from the VAT can drain directly to the liver via the portal circulation, where they affect the hepatocytes. The FFAs reduce insulin degradation, resulting in hyperinsulinemia, and induce insulin resistance thereby increasing glucose production in the liver. Furthermore, the higher amount of circulating FFAs provides more substrate for triglyceride synthesis leading to a higher production of vLDL, which is atherogenic, thus contributing to hyperlipidemia, hepatic steatosis, and nonalcoholic fatty liver diseases. The higher production of vLDL and low-density lipoproteins (LDL) results in a higher release of these molecules to the circulation, which in turn leads to their retention in the intima within the artery wall; further, the oxidation of LDL contributes to EC activation. The adhesion of leukocytes in the artery wall and a superficial erosion of ECs, leads to the formation of a platelet thrombus. The microvessels become more and more fragile and are potential locations for hemorrhage and thrombosis. This eventually leads to atherosclerosis. As previously mentioned, absorbed FFAs in ECs are converted to acyl-CoAs that are involved in the synthesis of ceramides and sphingolipids. These lipid classes were reported to be positively correlated with cardiovascular disease and cardiovascular-related death in a subcohort of the Long-Term Intervention with Pravastatin in Ischemic Disease (LIPID) study, which included individuals with a myocardial infarction or hospital admission for unstable angina, for a randomized trial of pravastatin. Fatty acid uptake by the ECs involves crosstalk between the ECs and AT. showed that Angpt2, which is highly expressed in SAT, can regulate the fatty acid uptake in ECs; ECs treated with Angpt2 showed increased fatty acid uptake via the scavenger receptor CD36 and FATP3 in SAT. Angpt2 activates the integrin α5β1 signaling pathway, which results in a translocation of CD36 and FATP3, thereby improving the fatty acid uptake in SAT. Moreover, Angpt2 overexpression in AT results in an increased expression of angiogenic and endothelial markers, VEGFA and CD31, and an increased vascular density in SAT. Furthermore, under a high-fat diet, mice with endothelial-specific knockout of Angpt2 exhibited in improved glucose clearance and insulin sensitivity, as shown by oral glucose tolerance test (oGTT) and insulin tolerance test (ITT). The Notch signaling via Rbp-jκ activation plays a role in fatty acid transport in the heart. The inhibition of Notch signaling results in decreased expression of endothelial lipase, CD36, and FATP4. This alteration leads to a switch from fatty acids to glucose as sources for energy production, which is known from several animal models of heart failure.

15Obesity is also associated with diabetes, and the expression of p53 in ECs was found to be increased in the diabetic state. Under the high-calorie diet, the EC-specific p53 knockout mice showed improved insulin sensitivity and glucose tolerance along with lower plasma insulin levels as well as VAT and SAT. The deletion of p53 results in higher oxygen consumption and higher glucose uptake in skeletal muscles and BAT through the upregulated endothelial expression of glucose transporter (Glut)1. Moreover, the deletion of p53 resulted in higher expression of genes related to mitochondrial biogenesis, such as PPARγ coactivator-1α (PGC1α), via higher phosphorylation of endothelial nitric oxide synthase (eNOS) in skeletal muscles. Conversely, the upregulation of endothelial p53 caused metabolic abnormalities. PGC1α has widespread functions in different tissues and cell types, and diabetes induces the expression of PGC1α in ECs. Endothelial-specific PGC1α-overexpression resulted in significantly repressed EC migration, a hallmark of EC function, as measured by transwell migration assays. This inhibition of EC migration is induced by the activation of Notch-signaling and inhibition of Rac/Akt/eNOS signaling, suggesting that PGC1α mediates, in part, the vascular dysfunction caused by diabetes.

16Under the high-fat diet, the insulin signaling in ECs is impaired, showing a reduction in insulin receptor substrate-2 (Irs2) and insulin-induced eNOS phosphorylation. This leads to a further reduction in insulin-induced glucose uptake by the skeletal muscles triggered by decreased capillary recruitment and interstitial insulin concentration in the skeletal muscles. The reduction of Irs2 in ECs also influences the insulin secretion in the pancreatic β-cells by impairing islet blood flow. The insulin increase NO synthesis, via Phosphoinositide 3-kinase (PI3-K), in ECs, and platelets and in VSMCs. Mediated by nitric oxide, insulin increases the protein expression and secretion of VEGF. demonstrated that obesity leads to higher activation of Notch signaling in ECs. This results in decreased insulin sensitivity and lower glucose uptake in the muscles. These effects are mediated by inhibition of genes involved in caveolae formation, such as caveolar protein caveolin 1 (Cav1), and results in a diminished insulin transport to the muscles.

17Extracellular vesicles (EVs) can be involved in cell-to-cell and organ-to-organ communication via metabolic signals. This has been demonstrated for e.g., adipose-derived stem cells EVs by. Obese mice treated with these exosomes showed improved insulin sensitivity. EVs of AT macrophages from lean mice injected to obese mice caused whole-body insulin sensitivity and glucose tolerance, as determined by GTT and ITT. Conversely, EVs from obese mice injected to lean mice led to insulin resistance, lower glucose tolerance, and increased inflammation. This outcome was induced by miRNA155, which targets PPARγ and impairs the expression of Glut4. demonstrated crosstalk between ECs and adipocytes, via Cav1-containing adipocyte EVs that are taken up by ECs (Figure 2). This transfer is dependent on metabolic status and increases during fasting. Proteomic analysis of the EVs showed an increase in the expression of proteins involved in polyamine metabolism, antioxidant response, and transport of small molecules during the fasted state. However, a decrease in molecules involved in lipid and amino acid metabolism was observed. There is further evidence of direct crosstalk between ECs and adipocytes/AT. The endothelial-specific knockdown of the transcription factor Forkhead Box O1 (FoxO1) led to a higher vascular density in AT. Under the high-fat diet, VAT displays a higher expression of Pecam1, higher capillary number per adipocyte, reduced triglycerides in blood and less hepatic lipid accumulation. FoxO1 depletion increases the ECs proliferation by upregulating glycolytic markers, including Glut1, hexokinase 2, and phosphofructokinase. The suppression of argonaute (AGO1), a regulator of the endothelial hypoxia response, leads to a desuppression of VEGFA and thereby contributing to hypoxia-induced angiogenesis. The EC-specific AGO1 knockout in mice resulted in lower body weight gain under a high-fat high-sucrose diet, which could be explained by the lower SAT and BAT mass (; Figure 2). The knockout mice had lower fasting glucose level and higher insulin sensitivity, as confirmed by oGTT, ITT, and higher phosphorylation levels of AKT and AMPK, hallmarks of insulin sensitivity, in SAT and BAT. Furthermore, browning in SAT was observed, as indicated by increased expression levels of UCP1 and its regulator PGC1α. Moreover, AGO1 deletion led to increased levels of VEGFA and higher CD31+ staining in SAT.

18Obesity can lead to many complications and comorbidities. It affects not only the AT but also several organs such as the liver or cardiovascular system. EC dysfunction is a hallmark of obesity. For a long time, it was thought that only the AT affects the EC function. However, several studies have demonstrated that changes in EC functions and signaling pathways influence adipocyte metabolism and can improve the obesity-induced dysfunction (Figure 2). These first insights and differences in the metabolism of healthy and dysfunctional ECs could provide a basis for further studies that may target ECs for therapeutic benefits. The EC-specific treatment can potentially have direct effects on the metabolism of adipocytes and/or AT. However, further studies are necessary to understand the direct crosstalk between ECs and adipocytes as well as the organ-to-organ communication that may help develop strategies to reduce the obesity-induced complications and alteration to whole body metabolism.

19JH and JK selected the literature, wrote the manuscripts, and prepared the figures. All authors contributed to the article and approved the submitted version.

20The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.