P. E. Scherer, Touchstone Diabetes Center, Department of Internal Medicine, University of Texas Southwestern Medical Center, 5323 Harry Hines Boulevard, Dallas, TX 75390 8549, USA Fax: +1 214 648 8720 Tel: +1 214 648 8715 E-mail: firstname.lastname@example.org
Adipose tissue is an endocrine organ made up of adipocytes, various stromal cells, resident and infiltrating immune cells, and an extensive endothelial network. Adipose secretory products, collectively referred to as adipokines, have been identified as contributors to the negative consequences of adipose tissue expansion that include cardiovascular disease, diabetes and cancer. Systemic blood circulation provides transport capabilities for adipokines and fuels for proper adipose tissue function. Adipose tissue microcirculation is heavily impacted by adipose tissue expansion, some adipokines can induce endothelial dysfunction, and angiogenesis is necessary to counter hypoxia arising as a result of tissue expansion. Tumors, such as invasive lesions in the mammary gland, co-opt the adipose tissue microvasculature for local growth and metastatic spread. Lymphatic circulation, an area that has received little metabolic attention, provides an important route for dietary and peripheral lipid transport. We review adipose circulation as a whole and focus on the established and potential interplay between adipose tissue and the microvascular endothelium.
With overconsumption and decreased physical activity combining to propagate an epidemic of obesity in Western cultures, the pathophysiological aspects of adipose tissue expansion are becoming increasingly appreciated. There has been a steady increase in research focusing on adipose tissue contributions towards diabetes, cardiovascular disease and cancer. Some years ago, we outlined some key areas that we proposed would be essential in elucidating the key systemic and local effects of adipose tissue . Many of these topics are now areas of intense research and have further supported the concept of adipose tissue as an endocrine organ. In this minireview, we focus on the importance of the vasculature in adipose tissue function and related pathologies. Rather than discussing the greater relationship between cardiovascular diseases and obesity – an area of significant importance in its own right – we focus on the circulation of adipose tissue itself and the relevance of the circulatory microenvironment to pathologies and changes associated with adipose tissue, including adipocyte differentiation and adipose tissue expansion, hypoxia-induced neovascularization, and the relationship of adipose tissue with lymphatic circulation.
Adipose tissue depots and obesity
Obesity is a potent risk factor for metabolic and cardiovascular disease at the population level. At the individual patient level; however, correlations between body mass index and cardiovascular disease are not always straightforward as a result, in part, of differences among adipose tissue depots with respect to the overall rate of adipocyte dysfunction, local degree of inflammation and tissue vascularization . Adipose tissue is a heterogeneous mix of adipocytes, stromal preadipocytes, immune cells and endothelium . Combined, adipose tissue functions as a complex endocrine organ, secreting a host of factors collectively referred to as adipokines . The adipocyte ‘secretome’ ranges from molecules of direct metabolic relevance to those with effects unrelated to metabolism. These include the adipocyte-specific proteins adiponectin and leptin, the inflammatory chemokines tumor necrosis factor-α (TNFα) and an array of interleukins, and angiogenic and vasoactive molecules such as vascular endothelial growth factor (VEGF) and angiotensin II .
Adipose tissue develops in several distinct anatomical depots within the body, and the differential expansion of these depots is of great importance. Expansion of visceral or abdominal white adipose tissue (WAT) has been most strongly correlated with insulin resistance and cardiovascular disease in humans and animals . Conversely, the expansion of subcutaneous WAT does not appear to have the same negative systemic consequences on metabolism . At the other end of the spectrum is the condition of lipodystrophy, wherein the dramatic loss of adipose tissue triggers a high degree of insulin resistance and signs of other metabolic dysregulation similar to visceral WAT expansion. The importance of maintaining at least remnants of WAT was demonstrated by injecting adipocyte progenitors into the residual adipose depots of lipodystrophic mice: the depots expanded and the systemic metabolic profile was properly restored . Brown adipose tissue (BAT) is in an entirely different metabolic category as a result of its primary function in generating body heat in infants and rodents . BAT is rich in mitochondria, highly vascularized and, because it affords none of the ill effects of visceral WAT, serves as an ideal paradigm for ‘good’ adipose despite its limited presence in adult humans [8,9]. Combined, these disparities in the metabolic effects of distinct fat deposits not only dispel the generalized notion that adipose tissue exerts negative metabolic consequences under all conditions, but also beg the question as to what distinguishes these individual depots with respect to their ability to expand? Recent results suggest that the balance between angiogenesis and hypoxia has a significant impact on the modulation of ‘good’ versus ‘bad’ tissue expansion, thereby implicating the local microvasculature as a key modulator of adipose depot physiologies and their systemic impacts [6,10,11].
Adipose tissue vasculature
Adipose tissue possesses a relatively dense network of blood capillaries, ensuring adequate exposure to nutrients and oxygen. WAT varies in its vascularity both between depots and within the tissue itself. For example, the expanding tip of the epididymal WAT fat pad contains a high vessel density compared to the rest of the depot . This vessel network must be considered in the diverse roles that adipose tissue performs. Metabolically, the adipose vasculature serves to transport systemic lipids to their storage depot in the adipocytes. On the other hand, the vasculature also transports factors (adipokines) and nutrients (such as free fatty acids) from these cells in time of metabolic need. Expansion and reduction of the fat mass thus relies on the adipose tissue circulation. Insufficient circulation results in local hypoxia (whose effects are discussed below). The microvasculature of adipose tissue is necessary for the expansion of adipose mass not only as a result of its ability to prevent hypoxia, but also as a potential source of the adipocyte progenitors in WAT because these progenitor cells can derive from the microvasculature of the tissue . In addition to its necessity in metabolite transport, the blood capillary network also contributes to immunity and inflammation. Adipose tissue macrophages serve multiple functions, including the removal of necrotic adipocytes leading to lipid-engulfed foam cells, acting as proinflammatory mediators, and serving as angiogenic precursors . Often implicated in the adverse effects of adipose tissues because of their inflammatory impact, adipose-associated macrophages utilize the microcirculation to rapidly reach their targets . The microcirculation is itself modulated by locally produced chemokines from macrophages, stromal cells and adipocytes that encompass the tissue . Changes in local and systemic endothelial permeability or endothelial dysfunction induced by adipokines alter transendothelial transport and exclusion, and also control immune cell migration (Fig. 1). Leptin may impair nitric oxide production and sensitivity and induce angiogenesis . TNFα increases endothelial-immune cell adhesion molecules and immune trafficking. Adiponectin, in turn, down-regulates each of these responses . High concentrations of free fatty acids may directly impair endothelial function, leading to further local metabolic instability . Hypoxia inducible factor (HIF)-1α induces fibrosis in response to hypoxia in WAT . Overall, tissue function and homeostasis are therefore intimately tied to a properly functioning microcirculation.
The lymphatic circulation also likely contributes to adipose tissue maintenance. Despite their anatomical proximity and noted roles in lipid metabolism, storage and transport, lymphatics and adipose tissue are rarely discussed in the same context. We would like to propose that the lymphatic vasculature should be considered as an important player in adipose tissue circulation and will discuss the interactions between the lymphatic circulation and adipose tissue later in this minireview.
Adipose tissue angiogenesis
Studies that describe the quality of adipose tissue consistently point to the microvasculature and angiogenesis within adipose tissue as critical role players in adipose tissue health and expansion . Expansion of adult adipose tissue is not unlike tumor propagation: rapid growth induces hypoxia that induces angiogenesis, which in turn fuels more growth, etc. . In what has become increasingly indicative of the metabolic disease potential, the expansion of WAT results in hypoxia and increased levels of HIF-1α that, in turn, lead to an up-regulation of the inflammatory adipokines interleukin-6, TNFα and monocyte chemotactic protein-1, amongst others  (Fig. 2). These proinflammatory secretory products have been implicated in many aspects of insulin resistance . Hypoxia also induces adipose tissue fibrosis that leads to further adipose dysfunction [10,17]. Hypoxia may also block the differentiation of preadipocytes and stimulate glucose transport by adipocytes , although additional in vivo studies are required to validate this concept. Angiogenesis within adipose tissues is necessary to counteract hypoxia and WAT is rich in angiogenic factors, as well as endothelial cells, macrophages and circulating progenitors, that contribute to this process . The propensity for angiogenesis in the various adipose depots is likely reflected in their expansion potential . BAT, as a model adipose depot, exhibits an increased expression of VEGF, angiogenesis and vascular density expansion in response to hypoxia during exposure to cold . The angiogenic potential of adipose tissue may also vary from individual to individual. For example, it was recently demonstrated that, with increasing age and the progression of insulin resistance in obese db/db mice, the tissue stroma of WAT had a decreased capacity to induce the necessary pro-angiogenic effectors for healthy adipose tissue expansion ; the implications for human disease are that individuals in the diabetic state are even further at risk.
Increasing angiogenesis in normally hypoxic adipose tissue may improve some of the negative systemic effects associated with dysfunctional WAT. The overexpression of adiponectin, which is normally reduced in expanding WAT, may potently mediate angiogenesis within hypoxic adipose tissue [6,11]. The overexpression of adiponectin in wild-type mice results in highly vascularized subcutaneous adipose tissue. More importantly, in the morbidly obese ob/ob mouse line, the overexpression of adiponectin results in better overall health, despite an even further expansion of the subcutaneous WAT. The increased subcutaneous WAT in this mouse is highly vascularized . VEGF secreted in both subcutaneous and visceral adipose tissues is potently angiogenic . Blocking angiogenesis via the VEGF pathway in young ob/ob mice prevented the expansion of adipose tissue, resulting in mice with normal phenotypes and a return to normal metabolic function in adulthood [21,22]. Currently prescribed anti-diabetic drug therapies also present differential effects on adipose angiogenesis, despite the mutually positive effects on insulin sensitivity. The drug metformin, for example, reduces adipose tissue angiogenesis , whereas the thiazolidinedione class of drugs result in more vascularized adipose with increased adiponectin secretion . The concept of healthy adipose expansion is an apparent contradiction in the context of excess caloric intake and a potential increase in other detrimental health effects arising from overnutrition. This effect can only be rationalized if either food intake is repressed and/or energy expenditure is increased, and it has yet to be extensively studied in these circumstances. This also complicates potential anti-obesity therapies targeted at angiogenic processes because blocking the vascularization of existing adipose tissue may result in increased levels of inflammation.
Adipose tissue and tumor growth
Although adipose tissue vascularization functions in a delicate balance in tissue homeostasis, the perturbations initiated by tumor growth dysregulate all involved cell types. The most extreme example of tumor infiltration into an area rich in adipose tissue can be observed in the context of breast cancer. After filling the lumen of mammary ducts, transformed ductal epithelial cells break through the basal lamina and invade the mammary stromal compartment, which is highly enriched in adipose tissue. Here, the local pro-angiogenic machinery, such as VEGF and the adipose-specific leptin and monobutyrin , are co-opted to function in conjunction with autonomous tumor-derived factors to meet the circulatory demands of the invading lesion. The adipokine leptin is strongly angiogenic  and may increase tumor angiogenesis either by directly acting on the endothelium or by increasing local VEGF secretion [27,28]. We recently reported our findings on the relative contributions adipocyte-derived adiponectin on tumor growth in the murine mammary gland . Mice lacking adiponectin crossed into the MMTV-PyMT mammary tumor model initially exhibited a smaller lesion size compared to tumor growth in MMTV-PyMT adiponectin-normal mice. Lesions in the adiponectin null mice had impaired vascularization and displayed increased intratumoral necrosis. However, similar to tumors grown in the presence of pharmacological angiogenesis inhibitors, these tumors adapted to the chronic hypoxic conditions and eventually assumed a much more aggressive growth phenotype . Whether or not adiponectin serves as a direct angiogenic factor or tumor promoter remains to be clarified. Tumor cell entry into lymphatic capillaries en route to lymph node metastases may also be adipokine mediated. Adipose tissue expresses a host of lymphangiogenic growth factors  that, in combination with tumor, stromal and vascular derived factors, present an environment that apparently all but ensures metastasis . There are unquestionably consequences for the local paracrine crosstalk between the tumor cells, adipocytes and the adipose microvascualture, and the marked similarities in tumor growth and hypoxia compared with those of adipose tissue expansion remain of great interest.
Adipose tissue and lymphatic circulation
There has been a rapidly increasing interest in lymphatic circulation, particularly with respect to tumor progression and immunological responses. As an important part of the circulatory system with roles in lipid absorption and transport, and as an emerging interest area, it is therefore necessary to examine what is known regarding the lymphatic vasculature and its potential interplay with adipose tissue.
The lymphatic system
Fluid transport through the lymphatic vasculature forms an integral part of the body’s circulation. Throughout almost all tissues of the body, lymphatic capillaries transport fluid, macromolecules, and cells collected from the interstitial space via larger conducting lymphatic vessels and the lymph nodes back to systemic blood circulation (Fig. 3A) . In doing so, the lymphatic vasculature serves three critical roles. Firstly, as interstitial fluid is sourced from fluid extravasated from the blood vasculature, the lymphatics maintain tissue homeostasis and complete the body’s circulatory loop . Secondly, lymphatic collection of interstitial fluid permits downstream immune scavenging by sentinel lymph nodes, as well as providing the initial entry point for antigen-presenting cells en route to propagating required immune responses . Thirdly, lymphatic capillaries serve as the entry point of all dietary lipids into circulation . Although all of these roles are certainly interconnected, here we focus on the role of lymphatics in lipid absorption, the consequences of lymphatic dysfunction and the potential symbiotic relationship between the lymphatic system and adipose tissue.
Lymphatic capillaries differ from blood capillaries not only in their gene and molecular expression, but also in their strikingly different morphology . Lymphatic vessels exist in the tissue as a collapsed network of overlapping lymphatic endothelial cells, are not surrounded by pericytes, possess minimal interrupted basement membrane, and are directly anchored to the extracellular matrix by anchoring filaments where basement membrane is lacking . These properties permit open fluid flow from the interstitial space through the overlapping lymphatic endothelial cells through unique cell–cell junctions . These primary valves permit macromolecules and particles of up to 1 μm in size to freely enter lymphatic circulation . It is this transport potential that allows the lymphatics to star in the role of lipid transporter.
Lymphatic function and lipid absorption
In the jejunum, dietary lipids are absorbed by enterocytes lining the luminal wall, which then ‘package’ the lipids into large lipoprotein particles called chylomicrons. These particles are exocytosed and taken up by intestinal lacteals (specialized lymphatic capillaries found within each intestinal villus) (Fig. 3B) . Chylomicrons are then transported through the lymphatic network and enter the venous circulation at the thoracic duct. Proper lymphatic function is clearly necessary for this process because changes in intestinal hydration, and thus lymphatic clearance rate, modulate the rate of chylomicron transport . High concentrations of chylomicrons give the lymph a milky white appearance and the mixture is referred to as chyle. The presence of free chyle in the peritoneum or thoracic cavity, chylous ascites and chylothorax, respectively, may indicate dysfunctional lymphatic transport. Indeed, mice lacking or possessing mutations in the important lymphatic genes Ang-2, Foxc2, Prox1, Sox18, VEGF-C and VEGFR-3 (among others) possess poorly developed lymphatic networks and exhibit high infant or embryonic mortality and/or notable chylous accumulation as pups . Improper intestinal lymphatic function may also be present and be propagated by intestinal inflammation, such as in inflammatory bowel disease and Crohn’s disease . In these instances, flux of dietary lipids into lymphatics, downstream lymphatic vessel drainage and contractility, and mesenteric lymph node immune surveillance are all significantly reduced . Failures in intestinal lymphatic transport most likely result in cyclic worsening of these inflammatory conditions: lymphatic immune function is compromised, leading to increased inflammation and increased inflammatory mediators that further impede the ability of lymphatics to function, and so on .
Lymphatic dysfunction and adipose tissue
Lymphatic physiology also provides for peripheral lipid transport. Failures in lymphatic transport can result in marked lipid accumulation throughout the body. Lymphedema is a pathology of deficient lymphatic transport, either inherited or acquired through some inflammatory or surgical intervention, that results in the significant accumulation of fluid, matrix remodeling, and adipose expansion in the affected limb . Adipose expansion is also present in mouse models of secondary (induced) lymphedema . VEGFR-3 heterozygote mice are used as a model for inherited lymphedema because of their lack of dermal lymphatics. These adult mice exhibit substantial thickening of the subcutaneous adipose tissue [44,45]. Most notable of the lymphatic deficient mouse models in respect to adipose tissue are the Prox1 heterozygous mice. Few pups of this model survive to adulthood, although those that do demonstrate adult-onset obesity with significant expansion of all fat pads . When Prox1 was specifically deleted in the lymphatic vasculature and adult adipose expansion still occurred, lymphatic dysfunction was thereby directly implicated in obesity (Fig. 3C) . Collected lymph has also been demonstrated to induce adipocyte differentiation, further supporting this hypothesis [46,47]. Although no treatment has been successful in providing for, or restoring, lymphatic function in these tissues (compression and massage can manage, but not cure the disease), liposuction has been prescribed as a potential therapeutic intervention to diminish limb volume with varying success . Lymphatic dysfunction has also been noted in lipidema, a pathology of regionalized excessive lipid accumulation and adipose expansion. Malformed lymphatic vessels and improper lymphatic drainage function have been observed in these patients [48,49]. Classification of this pathology is thus difficult because it remains unknown whether adipose expansion or lymphatic dysfunction occurs first. Adipose tissue, itself a secretory organ, can provide a source of molecules that directly affect the lymphatic endothelium by changing the capillary permeability and collecting vessel tone, as well as effects similar to those observed in blood endothelium as described above (e.g., changes in adhesion molecule expression). Increased production of vascular endothelial growth factor-C, for example, with adipose expansion  may further reduce lymphatic function by inducing hyperplasia . A reduction in lymphatic drainage and degeneration of collecting lymphatic vessel smooth muscle was recently reported in a hypercholesterolemic mouse model . Lymphatic dysfunction would lead to further adiposity, and so the condition worsens. Peripheral lymphatic management, uptake and transport of adipocyte secretions and reverse transport of lipids and lipophyllic molecules from the interstitium is therefore of great importance not only to interstitial homeostasis, but also potentially to systemic metabolism.
Perivascular and perinodal adipose tissue
Although lymphatic capillaries have not been identified within the bulk of adipose tissues, adipose tissue does surround all collecting lymphatic vessels and lymph nodes. These larger lymphoid vessels and structures are morphologically different from the lymphatic capillaries discussed thus far, although their anatomical proximity demands attention and suggests synergistic potential. Indeed, a study by Pond et al.  has defined this perilymphatic adipose tissue as being metabolically essential for proper immune responses and as a source of energy for immune activation and proliferation. Expansion of these adipose deposits appears to occur with localized chronic inflammation , supporting the energy source hypothesis. Additionally, antigen-presenting cells may migrate between the contiguous tissues. Appreciation for the relevance of perinodal adipose tissue is increasing within the immunology community. It should be noted that it is the intimacy of the lymphatic and perinodal adipose that provides these benefits, and that dyslipidemia and obesity as a whole result in decreased immune trafficking . A hypothesis has also been put forward in which the immune system and leukocytes may directly buffer the increase in circulating glucose after a meal . By adding this additional metabolic function, this interesting concept would further strengthen the importance of the lymphatic network. When combined, these observations highlight the important roles of the lymphatic system: fluid and macromolecular transport, immune modulation and lipid uptake. All of these processes are tightly interconnected. Because adipose tissue is an organ that requires macromolecular transport, impacts inflammation and immunity, and provides a metabolic depot, the symbiosis of the lymphatic circulation with adipose tissue is certainly worthy of further study.
The role of adipose tissue as an endocrine organ critically depends on its circulation for metabolic function and transport. Variations in the vascularization of different types of adipose tissue and between WAT depots likely contribute to the metabolic dysfunction, or lack thereof, associated with adipose expansion and obesity. Rich in vasculogenic and proinflammatory adipokines, adipose tissue serves as an intriguing model system for understanding the contributory molecules in angiogenesis and tumor progression. An increased study of modulating adipose tissue expansion and the emerging interplay between adipose tissue pathophysiology and lymphatic circulation should provide a strong basis for future research into this complex tissue.
This work was supported by NIH grants R01-DK55758, R24-DK071030-01 and R01-CA112023 (P.E.S.) and by T32-HL007360-31A1 (to J.M.R.) and F32-DK081279 (to K. E. D.).