Skip to main content
NCBI home page
Oxford University Press logoLink to Oxford University Press
. 2020 Mar 26;12(4):81–89. doi: 10.1093/intbio/zyaa006

Endothelial cell crosstalk improves browning but hinders white adipocyte maturation in 3D engineered adipose tissue

Jennifer H Hammel 1, Evangelia Bellas 1,✉
PMCID: PMC7167464  PMID: 32219324

Abstract

Central to the development of adipose tissue (AT) engineered models is the supporting vasculature. It is a key part of AT function and long-term maintenance, but the crosstalk between adipocytes and endothelial cells is not well understood. Here, we directly co-culture the two cell types at varying ratios in a 3D Type I collagen gel. Constructs were evaluated for adipocyte maturation and function and vascular network organization. Further, these constructs were treated with forskolin, a beta-adrenergic agonist, to stimulate lipolysis and browning. Adipocytes in co-cultures were found to be less mature than an adipocyte-only control, shown by smaller lipid droplets and downregulation of key adipocyte-related genes. The most extensive vascular network formation was found in the 1:1 co-culture, supported by vascular endothelial growth factor (VEGF) upregulation. After forskolin treatment, the presence of endothelial cells was shown to upregulate PPAR coactivator 1 alpha (PGC-1α) and leptin, but not uncoupling protein 1 (UCP1), suggesting a specific crosstalk that enhances early stages of browning.

Keywords: adipose tissue engineering, white adipose tissue, endothelial cells, vasculature, adipocyte, co-culture

Insight

Adipose tissue (AT) is highly vascularized in order to keep up with metabolic demands. By integrating vascular networks into a 3D model of AT, the crosstalk between the two can be studied in a physiologically relevant environment. We found that the presence of endothelial cells delays white adipocyte maturation but plays a role in the early stages of browning of AT. As browning has been identified as a potential therapeutic target, this model provides the ability to further develop strategies to combat obesity and other metabolic disorders.

INTRODUCTION

Adipose tissue (AT) engineering aims to develop in vitro models that functionally reflect the native AT properties. These in vitro models then can be applied to disease modeling, drug screening and for regenerative medicine applications, such as soft tissue reconstruction and fat grafting. AT is the master regulator of whole body metabolism and is highly vascularized to keep up with those metabolic demands. In normal, healthy AT, each adipocyte is in contact with the vascular network. In obesity, adipocytes undergo hypertrophy as they take on more nutritional load, expanding faster than angiogenesis, the formation of new blood vessels, can occur. This inadequate vascularization leads to hypoxia, inflammation and necrosis, which are known to induce insulin resistance, and eventually, Type II diabetes. Other comorbidities associated with obesity include hypertension, cardiovascular disease and certain forms of cancer. However, if angiogenesis occurs at an adequate rate as AT expands, the tissue will still function properly, reducing comorbidities. This allows a metabolically healthy, but still obese, phenotype. This phenomenon demonstrates the importance and effect of vascularization on metabolic function in AT. Vascularization is not only relevant for reducing the deleterious effects of obesity, but also for maintenance and long-term outcome success of engineered AT and grafted fat for filling of soft tissue defects [1].

Adipocytes and endothelial cells (ECs), the cells that comprise blood vessels, are in direct contact, engaging in cellular crosstalk. Existing studies have focused on viability and proliferation of both cell types, and differentiation of preadipocytes but do not characterize both EC and adipocyte function in a co-culture system. To adequately model AT, the system must be well vascularized, with ECs not only being present but organized into networks. Ratios of 1:1 to 1:7 adipose derived stem cells (ASCs) or adipocytes with ECs have been shown to be capable of forming robust networks [2–4]. Other studies have demonstrated that a lower number of ECs compared to adipocytes is also capable of preliminary network formation [5–7]. Preadipocytes are known to guide EC migration and aid in their proliferation [8]. As the preadipocytes differentiate, peroxisome proliferator-activated receptor gamma (PPAR𝛾) and leptin are upregulated, which stimulate angiogenesis [9, 10]. Adiponectin, is also upregulated, which has been shown to both inhibit [11–13] and enhance angiogenesis [5, 14]. Further, adipocyte/endothelial co-culture media has also been demonstrated to enhance the differentiation of ASCs as seen by greatest amount of lipid accumulation [6]. Both adipocytes and endothelial cells secrete vascular endothelial growth factor (VEGF), which aids angiogenesis by promoting EC proliferation, migration and survival. VEGF has also been shown to cause browning of white adipose tissue (WAT) by upregulating PPAR coactivator 1 alpha (PGC-1α) and uncoupling protein 1 (UCP1) expression and enhancing angiogenesis [15]. These results demonstrate a complicated relationship between these two cell types, with a fine balance to support both adipocyte function and vasculature. It has been established that an EC-preadipocyte co-culture allows for the proliferation of both cell types, and shows enhanced differentiation of preadipocytes [5]. However, direct contact between mature adipocytes and ECs can lead to dedifferentiation of the adipocytes into preadipocytes [16]. In the early stage of differentiation, preadipocytes, which initially resemble fibroblasts, increase their PPAR𝛾 expression and begin to take up lipids. After induction of differentiation, the adipocyte will continue to mature, eventually reaching a unilocular phenotype. Many studies [5, 17] have observed how EC presence can aid the differentiation of preadipocytes but have not evaluated their maturity. A maturing adipocyte has notable morphological features, it has accumulated larger lipid droplets (LDs) (in contrast to many smaller LDs), it is round (not spread), upregulates key adipogenic genes and can respond to induction of lipolysis. In order to engineer a model of AT that is physiologically relevant, it is important to establish what EC density best supports angiogenesis and adipocyte function and maturation.

Even more vascularized than WAT is brown adipose tissue (BAT). BAT is composed mainly of brown adipocytes and functions as a thermogenic organ. In humans, BAT was thought to exist primarily within infants to regulate non-shivering thermogenesis, but recently has been found in adults [18, 19]. Since that discovery, activating BAT and more recently, browning of existing WAT have become new targets for potential obesity therapies [20–22]. Brown adipocytes differ from white adipocytes in that they have many mitochondria, the reason for their brown color, and multiple smaller LDs. BAT’s thermogenic nature requires smaller LDs in order to facilitate rapid intracellular lipolysis, providing free fatty acids as fuel for adjacent mitochondria [23]. Early brown adipogenesis is initiated by the upregulation of (PGC-1α), where PGC-1α stimulates oxidative metabolism and mitochondrial biogenesis [24]. A later marker of brown adipogenesis is UCP1. UCP1 allows chemical energy from free fatty acids to be dissipated as heat and thus is a major factor in thermogenic function [25]. Browning of white adipocytes can be induced via stimulation of beta-3 adrenergic receptors. Beta-3 agonists, such as norepinephrine, forskolin and isoprenaline, activate BAT thermogenesis, increase UCP1 expression, increase mitochondrial biogenesis and induce lipolysis. Beta-3 agonists have also been shown to increase angiogenesis via increased VEGF expression [26].

To study metabolic disease in AT, understanding how the various cell types within WAT and BAT interact and aid or inhibit one another is critical. It is not well understood how these cell types interact to contribute to adipocyte maturation and function, as well as engineered AT structure as a whole. Several studies in adipocyte biology have been performed with 2D cultures [27–29]; however in these cases, adipocytes lack the appropriate microenvironment. In vivo, adipocytes have a spherical morphology in order to maximize lipid storage. In 2D culture, adipocytes cannot achieve a spherical shape leading to increased cytoskeletal stress. Encapsulation within a 3D collagen Type I hydrogel not only allows for vascular network formation [5, 30–32] but also optimizes for adipocyte function [33, 34]. To study the crosstalk between these two cell types in engineered WAT, we performed a co-culture of human umbilical vein endothelial cells (HUVECs) and adipocytes at varying ratios in 3D collagen gels. With adipocyte number kept constant, the ratios included 1:1 adipocytes to HUVECs, 1:3, 9:1 and a 1:0 control, allowing us to look at both EC presence and density and how these two parameters affect EC organization and adipocyte maturation and function.

MATERIALS AND METHODS

Cell culture

Human mesenchymal stem cells (hMSCs) at passage 4 are cultured to confluence and then treated with adipogenic induction media 1 day post-confluence for 1 week. Adipogenic basal media consists of DMEM/F-12 (Corning Cellgro) with 3% Fetal Bovine Serum (Sigma) and 1% Penicillin-Streptomycin (Sigma). Adipogenic induction media consists of rosiglitazone (2 μM), isobutylmethylxanthine (IBMX) (500 μM), dexamethasone (1 μM), D-calcium pantothenate (17 μM), biotin (33 μM) and insulin (20 nM) added to the basal media. HUVECs at passage 8 are cultured to confluence in Lonza EBM-2 Endothelial Cell Basal Medium-2 supplemented with Lonza EGM-2 Singlequots kit. Cells are maintained at 37°C and 5% CO2.

PDMS device preparation

Polydimethylsiloxane (PDMS, Sylgard) well devices (Fig. 1C) are employed to provide a consistent shape and dimension to each construct by constraining the space in which the construct is maintained [35–38]. To form the devices, PDMS and crosslinker are mixed at a 10:1 ratio by weight and degassed. This mixture is cured at 60°C for 4 h to form a PDMS sheet of 1 mm thickness. The PDMS sheet is cut into 1.5 cm discs with a biopsy punch. A 5 mm biopsy punch (AcuPunch) is used to puncture the PDMS discs so that each disc contains three 5 mm wells. The punched PDMS discs are bonded to 22 × 22 mm glass coverslips by plasma ashing and then placed into 100°C oven for 2 h. The assembled devices are then treated with 0.1% Poly-L-lysine for 1 h, rinsed thoroughly in DI water, and then treated 1% glutaraldehyde for 30 min to enhance collagen gel adhesion. Devices are then rinsed thoroughly again and placed under ultraviolet light for 15 min to sterilize.

Figure 1.

Figure 1

Open in a new tab

Experimental schematic and workflow. (A) hMSCs are grown to confluence and differentiated into adipocytes. (B) Adipocytes and endothelial cells (HUVECs) are combined in the designated ratios. (C) Cells are suspended in Type I collagen gel and pipetted into PDMS well devices. (D) On day 8, gels are either exposed to forskolin or (E) sacrificed for endpoint analysis. (F) Representative confocal images with nuclei in blue, lipid droplets in green and vascular networks in red. Scale bars are 100 microns in length.

3D hydrogel fabrication

Type I collagen gels are prepared according to established protocol [39]. Neutralization solution is prepared on ice, containing 10 mM HEPES, 28.4 μM Sodium Hydroxide (NaOH), ×1 Medium 199, 0.4 μM Sodium Bicarbonate (NaHCO3) and water. Type I collagen from rat tail stored in acetic acid (Corning) is added to the neutralization solution to obtain a final concentration of 2 mg/mL.

Adipocytes were pre-differentiated from hMSCs in 2D for 1 week (see ‘Cell Culture’ section) prior to hydrogel encapsulation. This step ensures that adipocytes will have their classic rounded phenotype, in contrast to when hMSCs are directly differentiated in collagen hydrogels, where the differentiating cells remain spread and do not adopt their rounded phenotype.

Adipocytes and HUVECs are detached with trypsin-EDTA, counted, combined in the proper ratios and pelleted. Adipocyte density is kept constant at 4 million cells/ml of collagen. HUVEC number is varied to account for the ratios (Fig. 1B). Neutralized collagen solution is added to cell pellets. Thirty μl gels are added to each PDMS well (Fig. 1C), then devices are placed into 37°C incubator for 30 min to complete gelation. Then, co-culture media is added. Media is changed on days 3, 5 and 8. Constructs are sacrificed for endpoint analyses on day 8 (Fig. 1E).

Maintenance of co-cultures

Co-culture media is a 1:1 mixture of adipogenic induction media and endothelial cell media. Media is changed every 3rd, 5th and 8th day. During media changes, 1 ml spent medium is banked and placed in −80°C for lipolysis assays. On day 8, a subset of the constructs are treated with 10 μM forskolin (Fig. 1D) in adipogenic basal media for 4 h to induce lipolysis and browning.

Lipolysis assays

To test the basal lipolysis levels of the cells, spent medium is banked to be evaluated via Serum Triglyceride Determination Kit (Sigma-Aldrich). Lipolysis was induced at day 8 via the addition of forskolin as described above and spent medium was banked. The samples were read on an Infinite M200 Pro plate-reader (TECAN). Lipolytic capacity is defined as induced lipolysis glycerol content/basal lipolysis glycerol content.

2-NBDG uptake assay

In order to evaluate glucose uptake ability of these co-cultures, constructs are treated with a fluorescent analog of glucose, 2-Nitrobenz Oxadiazol Amino Deoxyglucose (2-NBDG) (Invitrogen), at 20 μM in the cell culture media for 30 min at the experimental endpoint. Then, gels are treated with a 1:1000 dilution of Hoescht 33342 (Thermo Scientific) in phosphate buffered saline (PBS) for 1 h. They are then fixed with 4% formaldehyde in PBS and stored at 4°C until imaged. Quantification is performed via ImageJ. Corrected total cell fluorescence is calculated by taking the integrated density and subtracting the area multiplied by the background fluorescence.

Immunofluorescence

In order to visualize endothelial cell organization, immunofluorescence staining for cluster of differentiation 31 (CD31) is performed. Collagen gels are fixed in 4% formaldehyde for 1 h, and then permeabilized for 1 h with Triton X-100. Bovine serum albumin is utilized as a blocking agent. The constructs are incubated with anti-mouse CD31 antibody (Invitrogen) at a 1:200 dilution overnight followed by the secondary antibody, CF547 anti-mouse IgG (H + L) (Sigma) at a 1:1000 dilution, for 2 h. After CD31 staining, constructs were stained with Hoescht 33342 (Thermo Scientific) and Bodipy 493/503 (Life Technologies) and stored at 4°C until imaged.

Confocal imaging, LD quantification and vascular network analysis

Constructs are imaged on an Olympus FluoView FV1200 confocal microscope. ImageJ is used to convert original imaging format (OIF) files to z stacks and to perform all image analyses. LDs are quantified in diameter and number by an ImageJ Plugin, ‘Lipid Droplets Tool’ (http://dev.mri.cnrs.fr/projects/imagej-macros/wiki/Lipid_Droplets_Tool), with a minimum size of 25 microns. Endothelial cell networks are manually quantified in terms of branches and nodes, with nodes being defined as a junction of branches, and branches being defined as a linear portion of the vascular networks with distinct CD31 staining.

Gene expression

RNA was collected by pooling 2–3 constructs per condition into Trizol (Ambion) for RNA isolation. RNA is extracted and converted to cDNA via the High Capacity cDNA Reverse Transcription Kit (Applied Biosystems). Quantitative Polymerase Chain Reaction (qPCR) is performed with gene-specific primers and PowerUp SYBR Green Master Mix (Applied Biosystems) with a RealPlex Mastercycler (Eppendorf).

Statistics

All experiments had an n = 3 unless otherwise stated. ANOVA and student’s t-test were performed with an alpha value of 0.05.

RESULTS

ECs hinder adipocyte maturation, do not impact function

After 8 days in co-culture, constructs were sacrificed for gene expression analysis via qPCR and fixed for confocal imaging (Fig. 1E). Co-cultures demonstrate vascular networking and LD accumulation (Fig. 1F). PPAR𝛾, an inductor of adipogenic differentiation, was significantly downregulated in the 1:1 and 1:3 ratios, but not the 9:1 ratio (Fig. 2A). Adiponectin, a key marker of adipocyte maturation, was significantly downregulated in all co-cultures compared to the control (Fig. 2B). Leptin was significantly upregulated in the 1:1 ratio (Fig. 2C).

Figure 2.

Figure 2

Open in a new tab

ECs hinder adipocyte maturation but do not impact function. Key adipocyte-related genes involved in maturation and differentiation are downregulated in co-culture. (A) PPARG is significantly downregulated in the 1:1 and 1:3 co-culture groups. (B) Adiponectin is significantly downregulated in all co-cultures. (C) Leptin is significantly upregulated in the 1:1 co-culture [n = 5]. Lipid droplet accumulation was inhibited in co-cultures. (D) The number of lipid positive (+) cells was found to be significantly lower in the 1:3 ratio. (E) Lipid droplet area was decreased in co-cultures compared to the adipocyte-only control. (F) The 1:1 ratio had a significantly higher number of Lipid droplets per adipocyte [n = 3]. (G) Representative images of 2-NBDG uptake are shown. Scale bars are 100 microns in length. Quantification of 2-NBDG fluorescence shows no differences in glucose uptake between groups (H) [n = 3]. Basal lipolysis, quantified by glycerol content, shows no difference between groups (I) [n = 5]. ANOVA and student’s t-test were performed, with an asterisk denoting P < 0.05 from the control, a hashtag denoting P < 0.05 from the 9:1 group, and a dollar sign denoting P < 0.05 from the 1:1 group. Error bars represent standard deviation.

Lipid droplets (LDs) stained with Bodipy were analyzed via ImageJ plug-in. The number of LD positive cells, the area of LDs and the number of LDs per adipocyte were quantified. The number of LD positive cells was found to be significantly lower in the 1:3 ratio, but consistent among the adipocyte-only control, the 9:1, and the 1:1 ratio (Fig. 2D). LD area quantification demonstrated decreased LD accumulation in co-cultures compared to the adipocyte-only control (Fig. 2E). The 1:1 ratio had a significantly higher number of LDs per adipocyte (Fig. 2F).

After 8 days in culture, constructs are treated with 2-NBDG, a fluorescent glucose analog, at 20 μM for 30 min and then fixed for confocal imaging. Representative images of glucose uptake are shown in Fig. 2G. No significant differences in glucose uptake between groups were found (Fig. 2H). Basal lipolysis is quantified via glycerol content in spent medium at day 8. Glycerol content in spent medium shows no difference in basal lipolysis between co-cultures and the adipocyte-only control, with basal glycerol release between 50 and 75 μg/ml of spent media (Fig. 2I).

There is an optimal ratio to support network formation

Quantitative analysis of network formation was performed after 8 days in co-culture. CD31 immunofluorescence staining is performed to visualize EC network formation (Fig. 3A). Network formation is quantified via branch and node analysis, as well as branch and total network length measurements and vessel diameter measurements (Fig. 3B). There were no significant differences in branch lengths between ratios (Fig. 3C); however, the 1:1 ratio had significantly higher vessel diameters (Fig. 3D) and more branches (Fig. 3E) than the 9:1 ratio. Total network length was found to be consistent among all three ratios (Fig. 3E). VEGF expression was evaluated via qPCR, and found to be significantly upregulated in the 1:1 ratio compared to the other co-cultures and the adipocyte-only control (Fig. 3F).

Figure 3.

Figure 3

Open in a new tab

There is an optimal ratio to support EC network formation (1:1). Quantitative analysis of network formation was performed after 8 days in co-culture. (A) Vascular network formation is visualized via CD31 immunofluorescence staining. Scale bars are 100 microns in length. (B) Branches, nodes, network length and vessel diameter are quantified. (C) The 1:1 ratio is found to have the most varied branch lengths, (D) higher vessel diameter and (E) the most network branching [n = 3]. (F) VEGF is significantly upregulated in the 1:1 ratio, compared to the other co-cultures and the adipocyte-only control [n = 5]. ANOVA and student’s t-test were performed, with an asterisk denoting P < 0.05 from the control and a hashtag denoting P < 0.05 from the 9:1 group. Error bars represent standard deviation.

Increased EC presence enhances browning of adipocytes

In order to induce lipolysis and browning of adipocytes, constructs were treated with 10 μM forskolin for 4 h on day 8. Stimulated lipolysis was quantified via glycerol content in spent medium. Lipolytic capacity demonstrates that adipocytes are responsive to forskolin, with glycerol release increasing by between 1.5 and 2 fold, but that endothelial cell presence has no statistically significant effect on the lipolytic capacity of adipocytes (Fig. 4D). After forskolin treatment, constructs are sacrificed for gene expression analysis and confocal imaging. Forskolin treatment was found to upregulate key browning genes in all groups (not shown). PGC1-α, the initiator of brown adipogenesis, was upregulated in the 1:1 and 1:3 groups compared to the adipocyte-only control (Fig. 4A). Adipocyte-related gene, leptin, was also found to be upregulated in co-cultures compared to the control (Fig. 4C). UCP1 was found to remain constant throughout groups (Fig. 4B).

Figure 4.

Figure 4

Open in a new tab

Browning is enhanced by EC presence through PGC1-α. Co-cultures are treated with forskolin for 4 h on day 8 of culture. (A) PGC1-α is upregulated in the 1:1 and 1:3 ratios compared to both the adipocyte only control and the 9:1 ratio. (B) UCP1 remains constant among all groups. (C) Leptin is significantly upregulated in the 1:1 ratio. (D) Lipolytic capacity remains constant among all groups. ANOVA and student’s t-test were performed, with an asterisk denoting P < 0.05 from the control and a hashtag denoting P < 0.05 from the 9:1 group. For all groups and figures, n = 4. Error bars represent standard deviation.

DISCUSSION

The aims of this study were to investigate adipocyte-EC crosstalk by assessing the function of both cell types and to determine if there is an optimal EC density for supporting function. The maturation and function of the adipocytes in co-culture was characterized via gene expression, LD analysis, lipolysis and glucose uptake. In vivo, every adipocyte is in contact with the vascular network. Therefore, a successful co-culture construct must have adequate vascularization in order to model AT structure. Overall, the co-culture constructs exhibited successful vascular networking by endothelial cells and LD accumulation in the adipocytes but had distinct differences between groups.

PPAR𝛾 is a key initiator of adipocyte differentiation and is known to upregulate other adipocyte-specific genes, such as adiponectin, which has anti-diabetic effects and is a hallmark of mature WAT [40]. These adipogenic genes were found to be downregulated within co-cultures compared to the adipocyte-only control, suggesting a delay in maturation of the adipocytes when exposed to EC crosstalk. These results were further supported by the LD analysis demonstrating more and smaller LDs per cell. As adipocytes mature, smaller LDs undergo fusion to form larger and larger droplets until the cell eventually reaches a unilocular phenotype. The 1:3 ratio was found to have the least lipid positive cells per field of view, suggesting that higher numbers of endothelial cells can further inhibit maturation. Angiogenesis has been shown to precede adipogenesis in various models in vivo, with the establishment of a vascular network aiding the differentiation of preadipocytes [17, 41]. Taken together, this suggests that an in vitro model of AT that includes endothelial cells potentially matures differently than an adipocyte-only model, with a focus first on angiogenesis, and later on adipogenesis. This is supported by the downregulation of adiponectin (an anti-angiogenic gene) and the upregulation of VEGF and leptin (angiogenic genes).

Functionally, adipocytes in co-culture, compared to mono-culture, did not show changes in lipolysis or glucose uptake, suggesting basal metabolism is not impacted, regardless of the delay in maturation. Previous studies have also demonstrated no difference in basal lipolysis between adipocyte-endothelial cell co-cultures and adipocyte mono-cultures [42].

ASCs have been established as supportive of vascular network formation [8]. In a study that co-cultured ASCs and human microvascular ECs at a 1:3 ratio on Matrigel, networks were formed successfully and found to be stable for 5 days [2]. In a 1:2 co-culture of adipocytes and ECs in a 3D silk scaffold, continuous lumen-like structures of 100 μm or more in length formed after 6 months in culture [3]. In a 1:1 co-culture of ASCs and ECs, after 1 week the average vessel length was found to be around 150 μm, with that value reaching 200 at 2 weeks [4]. In another 1:1 co-culture of ASCs and brain microvascular ECs in Type I collagen gels, after 1 week of co-culture, capillary branches and vessel like structures averaged around 200 μm in length and 16 μm in width [43]. In a 9:1 ratio, preliminary network formation was found after two weeks but not quantified [5]. For the purposes of our study, we examined a 9:1, 1:1 and 1:3 ratio. Interestingly, the 9:1 ratio is found to often have no differences from the adipocyte-only control, suggesting that there are not enough ECs present to sufficiently initiate crosstalk throughout the co-culture construct. The 1:3 ratio has the lowest levels of adipocyte maturation. The 1:1 ratio, in between the other two ratios, provides a balance of vascular network formation and adipocyte maturation. The 1:1 ratio has the greatest vascular network formation, with high numbers of branches and nodes, the longest total networks, and greatest diameter vessels of all the ratios tested. This network organization is supported by an upregulation of VEGF and leptin.

Leptin has dual roles; one is pro-angiogenic by inducing proliferation of HUVECs [44]. Leptin’s pro-angiogenic effects include stimulating HUVEC proliferation for more rapid networking in Matrigel, causing increased VEGF release from HUVECs in a dose-dependent manner, and increasing MMP-2 activity [44]. Similar findings were found in bovine corneal endothelial cells [45], wherein leptin induced proliferation and worked synergistically with VEGF to enhance angiogenesis in rat corneas. In our study, we found leptin gene expression to be significantly greater in the 1:1 ratio than the other groups, which correlated with the improved vascular outcomes (increased VEGF expression, vessel diameter, branching/density) in the 1:1 group. In adipocytes, leptin is a hallmark of mature adipocytes and is increased in obesity. However, obese adipocytes lose lipolytic capacity, and in turn do not respond to beta-adrenergic agonists (such as forskolin) for stimulation of lipolysis [46]. Therefore, these co-culture constructs are unlikely to be representative of an obese AT model, as all groups remained responsive to lipolytic stimulation with no significant differences between groups.

The constructs were exposed to forskolin to evaluate lipolytic capacity as well as browning capacity. Higher rates of lipolysis provide substrates for the increased mitochondria density in BAT to produce energy [23]. BAT is densely vascularized, suggesting important crosstalk between brown adipocytes and ECs. Forskolin has also been shown to increase angiogenesis in HUVECs [26]. No significant differences in lipolytic capacity were found between co-cultures and the adipocyte-only control. Previous studies have also demonstrated no difference in stimulated lipolysis by isoproterenol, a similar catecholamine, between co-cultures of adipocytes and endothelial cells and mono-cultures [42]. Our gene expression results potentially suggest a PGC-1α-related enhancement of browning via endothelial cell presence. PGC-1α is produced by both endothelial cells and adipocytes [24]. It has been identified in vascular regulation and proven to be protective of endothelial cells in cases of atherosclerosis and hypoxia [47, 48]. PGC-1α also plays a role in angiogenesis by upregulating VEGF and enhancing the density of capillaries in skeletal muscle [49, 50]. Further exploration of the role PGC-1α plays in the endothelial cells of BAT and browning of WAT specifically is of great interest.

CONCLUSION

From these results, a 1:1 ratio of adipocytes to endothelial cells is recommended as a co-culture model with robust vascular network formation. This work has also established an important connection between vascular density and adipocyte maturation—suggesting that EC-adipocyte crosstalk plays a vital role in AT development. Interestingly, EC presence did not affect white adipocyte function, demonstrating that the crosstalk between these two cell types affects maturation differently than metabolism. EC presence also plays a role in the browning of adipocytes, through PGC1-α. This is especially of interest, as browning is considered a potential therapeutic target for obesity. Future studies need to be performed to understand whether these effects can be attributed toward soluble factors or direct contact between these two cell types. By understanding the function and development of white and brown AT and the role vasculature plays, future in vitro and in vivo studies can be better informed. These results identify new targets for combatting the epidemic of obesity, beginning to tease apart the complexities of AT.

Author Contributions

EB and JH conceived the concept and designed experiments. All experiments were performed by JH. Data analysis and processing of results were performed by EB and JH. The manuscript was prepared jointly by EB and JH.

ACKNOWLEDGEMENTS

This work was supported by the MARC-U*Star Program, National Institutes of Health [T34 GM 087239] to [JH], and the National Institute of Diabetes and Digestive and Kidney Diseases, Diabetic Complications Consortium (www.diacomp.org) [DK076169, DK115255] to [EB].

References

  • 1. Bellas E, Panilaitis BJB, Glettig DL et al. Sustained volume retention in vivo with adipocyte and lipoaspirate seeded silk scaffolds. Biomaterials 2013;34:2960–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2. Traktuev DO, Prater DN, Merfeld-Clauss S et al. Robust functional vascular network formation in vivo by cooperation of adipose progenitor and endothelial cells. Circ Res 2009;104:1410–U320. [DOI] [PubMed] [Google Scholar]
  • 3. Bellas E, Marra KG, Kaplan DL. Sustainable three-dimensional tissue model of human adipose tissue. Tissue Eng Part C-Methods 2013;19:745–54. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4. Cai XX, Xie J, Yao Y et al. Angiogenesis in a 3D model containing adipose tissue stem cells and endothelial cells is mediated by canonical Wnt signaling. Bone Res 2017;5:13. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5. Lai N, Jayaraman A, Lee K. Enhanced proliferation of human umbilical vein endothelial cells and differentiation of 3T3-L1 adipocytes in Coculture. Tissue Eng Part A 2009;15:1053–61. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6. Volz AC, Hack L, Atzinger FB et al. Completely defined co-culture of Adipogenic differentiated ASCs and microvascular endothelial cells. Altex-Alternatives to Animal Experimentation 2018;35:464–76. [DOI] [PubMed] [Google Scholar]
  • 7. Song YH, Shon SH, Shan MR et al. Adipose-derived stem cells increase angiogenesis through matrix metalloproteinase-dependent collagen remodeling. Integr Biol 2016;8:205–15. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8. Borges J, Muller MC, Momeni A et al. In vitro analysis of the interactions between preadipocytes and endothelial cells in a 3D fibrin matrix. Minim Invasive Ther Allied Technol 2007;16:141–8. [DOI] [PubMed] [Google Scholar]
  • 9. Gealekman O, Burkart A, Chouinard M et al. Enhanced angiogenesis in obesity and in response to PPAR gamma activators through adipocyte VEGF and ANGPTL4 production. Am J Physiol-Endocrinol Metab 2008;295:E1056–E64. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10. Cao YH. Angiogenesis and vascular functions in modulation of obesity, adipose metabolism, and insulin sensitivity. Cell Metab 2013;18:478–89. [DOI] [PubMed] [Google Scholar]
  • 11. Man K, Ng KTP, Xu AM et al. Suppression of liver tumor growth and metastasis by Adiponectin in nude mice through inhibition of tumor angiogenesis and Downregulation of rho kinase/IFN-inducible protein 10/matrix metalloproteinase 9 Signaling. Clin Cancer Res 2010;16:967–77. [DOI] [PubMed] [Google Scholar]
  • 12. Brakenhielm E, Veitonmaki N, Cao RH et al. Adiponectin-induced antiangiogenesis and antitumor activity involve caspase-mediated endothelial cell apoptosis. Proc Natl Acad Sci USA 2004;101:2476–81. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13. Palanisamy K, Nareshkumar RN, Sivagurunathan S et al. Anti-angiogenic effect of adiponectin in human primary microvascular and macrovascular endothelial cells. Microvasc Res 2019;122:136–45. [DOI] [PubMed] [Google Scholar]
  • 14. Ouchi N, Kobayashi H, Kihara S et al. Adiponectin stimulates angiogenesis by promoting cross-talk between AMP-activated protein kinase and Akt signaling in endothelial cells. J Biol Chem 2004;279:1304–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15. Sun K, Asterholm IW, Kusminski CM et al. Dichotomous effects of VEGF-A on adipose tissue dysfunction. Proc Natl Acad Sci USA 2012;109:5874–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16. Aoki S, Toda S, Sakemi T et al. Coculture of endothelial cells and mature adipocytes actively promotes immature preadipocyte development in vitro. Cell Struct Funct 2003;28:55–60. [DOI] [PubMed] [Google Scholar]
  • 17. Fukumura D, Ushiyama A, Duda DG et al. Paracrine regulation of angiogenesis and adipocyte differentiation during in vivo adipogenesis. Circ Res 2003;93:E88–97. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18. Cypess AM, Lehman S, Williams G et al. Identification and importance of Brown adipose tissue in adult humans. N Engl J Med 2009;360:1509–17. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19. Leitner BP, Huang S, Brychta RJ et al. Mapping of human brown adipose tissue in lean and obese young men. Proc Natl Acad Sci USA 2017;114:8649–54. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20. Wu J, Bostrom P, Sparks LM et al. Beige adipocytes are a distinct type of thermogenic fat cell in mouse and human. Cell 2012;150:366–76. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21. Seki T, Hosaka K, Fischer C et al. Ablation of endothelial VEGFR1 improves metabolic dysfunction by inducing adipose tissue browning. J Exp Med 2018;215:611–26. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22. Schulz TJ, Huang TL, Tran TT et al. Identification of inducible brown adipocyte progenitors residing in skeletal muscle and white fat. Proc Natl Acad Sci USA 2011;108:143–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23. Nishimoto Y, Tamori Y. CIDE family-mediated unique lipid droplet morphology in white adipose tissue and brown adipose tissue determines the adipocyte energy metabolism. J Atheroscler Thromb 2017;24:989–98. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24. Kadlec AO, Chabowski DS, Ait-Aissa K et al. Role of PGC-1 in vascular regulation: Implications for atherosclerosis. Arterioscler Thromb Vasc Biol 2016;36:1467–74. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25. Fedorenko A, Lishko PV, Kirichok Y. Mechanism of fatty-acid-dependent UCP1 uncoupling in Brown fat mitochondria. Cell 2012;151:400–13. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26. Namkoong S, Kim CK, Cho YL et al. Forskolin increases angiogenesis through the coordinated cross-talk of PKA-dependent VEGF expression and Epac-mediated PI3K/Akt/eNOS signaling. Cell Signal 2009;21:906–15. [DOI] [PubMed] [Google Scholar]
  • 27. Wang XY, Gao L, Han Y et al. Silicon-enhanced Adipogenesis and angiogenesis for vascularized adipose tissue engineering. Adv Sci 2018;5:15. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28. Huttala O, Palmroth M, Hemminki P et al. Development of versatile human in vitro vascularized adipose tissue model with serum-free angiogenesis and natural Adipogenesis induction. Basic Clin Pharmacol Toxicol 2018;123:62–71. [DOI] [PubMed] [Google Scholar]
  • 29. Gogg S, Nerstedt A, Boren J et al. Human adipose tissue microvascular endothelial cells secrete PPAR gamma ligands and regulate adipose tissue lipid uptake. Jci Insight 2019;4:14. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30. Feng X, Shaker M, Clark R et al. Fibrin and type I collagen 3D matrix differentially regulate sprout angiogenesis of human dermal microvascular endothelial cells. Ejc Supplements 2010;8:90–1. [Google Scholar]
  • 31. Ilan N, Mahooti S, Madri JA. Distinct signal transduction pathways are utilized during the tube formation and survival phases of in vitro angiogenesis. J Cell Sci 1998;111:3621–31. [DOI] [PubMed] [Google Scholar]
  • 32. McCoy MG, Seo BR, Choi SY et al. Collagen I hydrogel microstructure and composition conjointly regulate vascular network formation. Acta Biomater 2016;44:200–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 33. Chun TH, Inoue M. 3-D adipocyte differentiation and peri-adipocyte collagen turnover. Methods Adipose Tissue Biol, Pt B 2014;538:15–34. [DOI] [PubMed] [Google Scholar]
  • 34. Stacey DH, Hanson SE, Lahvis G et al. In vitro Adipogenic differentiation of preadipocytes varies with differentiation stimulus, culture dimensionality, and scaffold composition. Tissue Eng Part A 2009;15:3389–99. [DOI] [PubMed] [Google Scholar]
  • 35. Nguyen DHT, Stapleton SC, Yang MT et al. Biomimetic model to reconstitute angiogenic sprouting morphogenesis in vitro. Proc Natl Acad Sci USA 2013;110:6712–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 36. Du Y, Khandekar G, Llewellyn J et al. A bile duct-on-a-chip with organ-level functions. Hepatology 2019. doi: 10.1002/hep.30918. [Epub ahead of print] [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 37. Mu X, Zheng WF, Xiao L et al. Engineering a 3D vascular network in hydrogel for mimicking a nephron. Lab Chip 2013;13:1612–8. [DOI] [PubMed] [Google Scholar]
  • 38. Cross VL, Zheng Y, Choi NW et al. Dense type I collagen matrices that support cellular remodeling and microfabrication for studies of tumor angiogenesis and vasculogenesis in vitro. Biomaterials 2010;31:8596–607. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 39. Parkhurst MR, Saltzman WM. Quantification of human neutrophil motility in 3-dimensional collagen gels - effect of collagen concentration. Biophys J 1992;61:306–15. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 40. Liu ML, Liu F. Up- and down-regulation of adiponectin expression and multimerization: Mechanisms and therapeutic implication. Biochimie 2012;94:2126–30. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 41. Cho CH, Koh YJ, Han J et al. Angiogenic role of LYVE-1-positive macrophages in adipose tissue. Circ Res 2007;100:E47–57. [DOI] [PubMed] [Google Scholar]
  • 42. Choi JH, Bellas E, Gimble JM et al. Lipolytic function of adipocyte/endothelial cocultures. Tissue Eng Part A 2011;17:1437–44. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 43. Lin SY, Xie J, Gong T et al. TGF beta signalling pathway regulates angiogenesis by endothelial cells, in an adipose-derived stromal cell/endothelial cell co-culture 3D gel model. Cell Prolif 2015;48:729–37. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 44. Park HY, Kwon HM, Lim HJ et al. Potential role of leptin in angiogenesis: leptin induces endothelial cell proliferation and expression of matrix metalloproteinases in vivo and in vitro. Exp Mol Med 2001;33:95–102. [DOI] [PubMed] [Google Scholar]
  • 45. Cao RH, Brakenhielm E, Wahlestedt C et al. Leptin induces vascular permeability and synergistically stimulates angiogenesis with FGF-2 and VEGF. Proc Natl Acad Sci USA 2001;98:6390–5. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 46. Dicker A, Astrom G, Sjolin E et al. The influence of preadipocyte differentiation capacity on lipolysis in human mature adipocytes. Horm Metab Res 2007;39:282–7. [DOI] [PubMed] [Google Scholar]
  • 47. Craige SM, Kroller-Schon S, Li CY et al. PGC-1 alpha dictates endothelial function through regulation of eNOS expression. Sci Rep 2016;6:9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 48. Ye JX, Wang SS, Ge M et al. Suppression of endothelial PGC-1 alpha is associated with hypoxia-induced endothelial dysfunction and provides a new therapeutic target in pulmonary arterial hypertension. Am J Physiol-Lung Cell Mol Physiol 2016;310:L1233–L42. [DOI] [PubMed] [Google Scholar]
  • 49. Arany Z, Foo SY, Ma YH et al. HIF-independent regulation of VEGF and angiogenesis by the transcriptional coactivator PGC-1 alpha. Nature 2008;451:1008–U8. [DOI] [PubMed] [Google Scholar]
  • 50. Saint-Geniez M, Jiang AH, Abend S et al. PGC-1 alpha regulates normal and pathological angiogenesis in the retina. Am J Pathol 2013;182:255–65. [DOI] [PMC free article] [PubMed] [Google Scholar]

Articles from Integrative Biology are provided here courtesy of Oxford University Press

RESOURCES