Presence of Vasculature Results in Faster Insulin Response in Adipocytes in Novel In Vitro Vascularized Adipose Tissue Model

ALTEX 36(3), 2019 419 Received November 27, 2018; Accepted March 13, 2019; Epub March 14, 2019; © The Authors, 2019. ALTEX 36(3), 419-434. doi:10.14573/altex.1811271 Correspondence: Outi Huttala, FICAM, Faculty of Medicine and Health Technology, Tampere University, Arvo Ylpön katu 34, 33520 Tampere, Finland ([email protected]) ture transports oxygen, nutrients, and waste, and also produces local growth factors and cytokines that communicate with adipose tissue (Cao, 2014). In turn, adipose tissue modulates the vasculature by secreting both proand anti-angiogenic factors (Cao, 2014; Christiaens and Lijnen, 2010). Due to these characteristics of adipose tissue, chemicals affecting it may cause a wide variety of adverse health effects throughout the body. A wide variety of animal models, including fruit flies, dogs, cats, pigs, rabbits, hamsters, squirrels, mice, rats, and primates, have been used to study adipose tissue (Lai et al., 2014). Although animal models contain all components of the in vivo adipose tissue, the results obtained with animal models translate poorly to humans (Lai et al., 2014; Bergen and Mersmann, 2005; Chandrasekera and Pippin, 2014), especially as the complex pathways of lipid metabolism are mostly species-specif1 Introduction As obesity, both in adults and children, has reached epidemic proportions globally, the prevalence of associated diseases has increased correspondingly. In addition to type 2 diabetes, obesity and overweight are associated with cancer, sleep apnea, asthma, degenerative joint disease, hypertension, renal failure, stroke, and cardiovascular disease (Switzer et al., 2013; van Baak, 2013; Pi-Sunyer, 2009). In addition to storing energy, adipose tissue is a dynamic endocrine organ secreting many different bioactive factors that control insulin sensitivity, energy metabolism, immune responses, and cardiovascular homeostasis (Gu and Xu, 2013; Kershaw and Flier, 2004). Adipose tissue is also one of the most vascularized tissues in the body (Cao, 2013). The vasculaResearch Article Presence of Vasculature Results in Faster Insulin Response in Adipocytes in Novel In Vitro Vascularized Adipose Tissue Model Outi Huttala1, Jertta-Riina Sarkanen2,3, Tuula Heinonen1 and Timo Ylikomi2,3 1 FICAM, Faculty of Medicine and Health Technology, Tampere University, Tampere, Finland; 2 Cell Biology, Faculty of Medicine and Health Technology, Tampere University, Tampere, Finland; 3 Science Centre, Pirkanmaa Hospital District, Tampere, Finland Abstract Besides being an energy store, adipose tissue is an endocrine organ closely associated with the vascular system. Human relevant in vitro models are needed to study adipose tissue and related diseases. Vasculature plays a central role in the development and inhibition of adipose tissue-related diseases. Here, an adipocyte culture was established from hASC (human adipose stromal cells), and a vascularized adipose tissue model was established from hASC and HUVEC (human umbilical cord vein endothelial cells) co-culture, utilizing the same differentiation procedure. Comparing these models allowed analysis of the effect of vascularization on adipocytes. Both models were characterized on gene (adipocyte and vasculature-related), protein (von Willebrand factor, collagen IV, CD140b and CD144, secretion of leptin, adiponectin and FABP4), and functional (triglyceride accumulation, glucose uptake, and lipolysis) levels. Additionally, the vascularized adipose tissue model was exposed to chemicals with known effects on adipogenesis and angiogenesis (rosiglitazone, chlorpyrifos, prochloraz, mancozeb, butylparaben, 15-deoxy- ∆12,14-prostaglandin J2, bisphenol A, bis-(2-ethylhexyl) phthalate, tributyltin chloride) to compare their effects to the literature. The in vitro vascularized adipose tissue model demonstrated the presence of functional adipocytes and an extensive vascular network, displaying relevant gene and protein markers. Insulin induced glucose uptake, inhibited lipolysis, and influenced vasculature-related genes. The presence of vasculature led to a faster lipolysis inhibition by insulin and modulated responses to chemicals. This novel, thoroughly characterized, vascularized adipose tissue model is a promising new tool for studying adipose tissue as well as the effects of chemicals on adipogenesis and angiogenesis in adipose tissue. This is an Open Access article distributed under the terms of the Creative Commons Attribution 4.0 International license (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution and reproduction in any medium, provided the original work is appropriately cited. Huttala et al. ALTEX 36(3), 2019 420 2 Materials and methods∗ Ethical considerations This in vitro study conforms to the ethical principles outlined in the Declaration of Helsinki. Human adipose tissue samples were obtained from excess material of surgical operations, and human umbilical cords were received from caesarean sections with written informed consent at Tampere University Hospital, Tampere, Finland. The use of human adipose stromal cells (hASC) and human umbilical cord endothelial cells (HUVEC) was approved by the Ethics Committee of the Pirkanmaa Hospital District, Tampere, Finland (permit numbers: R15161 and R15033). Cells hASC used in the study were heterogenous cell populations obtained by isolating the stromal vascular fraction cells from human adipose tissue. HUVEC used in the study were isolated from umbilical cords. hASC and HUVEC were isolated as described by Sarkanen et al. (2012b). Before cryopreservation, all cell batches were tested for mycoplasma using Mycoplasma kit (MycoAlert® Detection Kit, Lonza Group LTD, Basel, Switzerland). Only Mycoplasma-free cell batches were used. hASC were characterized for markers CD73, CD90, and CD105 (BD biosciences, Franklin Lakes, NJ, USA) using a flow cytometer FACSCanto II (BD biosciences) as published previously (Huttala et al., 2015). hASC were expanded in hASC medium (Tab. 1) and HUVEC in Endothelial Cell Growth Medium-2 BulletKit (EGM-2, Lonza). In the co-cultures, hASC were at passage 2 and HUVEC at passage 4. Adipose tissue extract Adipose tissue extract (ATE) was produced as described previously (Huttala et al., 2018). Briefly, a human adipose tissue sample was manually cut into small pieces and incubated in DMEM/ F12 (Gibco, Carlsbad, CA, USA) for 24 h in 37°C. The resulting extract was filtered through a 0.2 µm filter. The protein content of the extract was determined with Pierce™ BCA Protein Assay Kit (Thermo Fisher Scientific, Waltham, MA, USA) according to the manufacturer’s instructions using bovine serum albumin (BSA) as a standard. Results were measured after 30 min incubation at 37°C at 562 nm with Varioskan™ Flash Multimode Reader (Thermo Fisher Scientific). Vascularized adipose tissue model The culturing scheme of the vascularized adipose tissue model can be seen in Table 2 and the different media compositions in Table 1. The undifferentiated control contained hASC and HUVEC cells cultured in basic media corresponding to the media used for differentiation (hASC medium corresponding to ATE medium, SFM corresponding to SM and MM). The adipocyte culture and the vascularized adipose tissue model were produced using a protocol published previously as “protocol 7” (Huttaic (Lai et al., 2014; Bergen and Mersmann, 2005; Chandrasekera and Pippin, 2014). Thus, new human cell-based models are needed to replace animal models to support 3R principles and to offer biologically relevant tools for research of human adipose tissue-related diseases. Adipogenic cocktail treatments, which include insulin, isobutylmethylxanthine, biotin, pantothenate, dexamethasone, and thiazolidinediones, have been problematic in developing vascularized adipose tissue as they have multiple negative effects on endothelial cells, including delayed growth (Kang et al., 2009). Volz et al. (2018) developed an elegant, serum-free defined medium for in vitro vascularized adipose tissue, but the need to pre-differentiate the adipocytes prior to combining the endothelial cells into the culture hampers its usability and lengthens the culture time. In the present study, a novel vascularized human adipose tissue model was established. This model is based on protocols of our previously established insulin-sensitive adipocytes (Huttala et al., 2016) and vascularized adipose tissue model (Huttala et al., 2018). All these models utilize adipose tissue extract (ATE), which is a natural adipogenesis inducer that also enables angiogenesis (Sarkanen et al., 2012a). The novel vascularized adipose tissue model, established from human adipose stromal cells (hASC) and human umbilical cord vein endothelial cell (HUVEC) co-culture, was compared to adipocyte culture established from hASC, using the same differentiation procedure. The use of the same differentiation procedure for both models provides a reliable model pair to study the effect of vasculature on adipocytes. Using the vascularized adipose tissue model, the effects of chemicals on both angiogenesis and adipogenesis can be investigated simultaneously. In order to ensure that the vascularized adipose tissue model is suitable for human adipose tissue research, it was characterized comprehensively. The gene expression of markers of adipocyte tissue and vasculature, secretion of adipokines known to be produced in in vivo white adipose tissue (leptin, FABP4, and adiponectin), accumulation of triglycerides, insulin responses (inhibition of lipolysis and induction of glucose uptake), and vascular formation were investigated. In addition, the function of the vascularized adipose tissue model was confirmed by challenging it with reference chemicals that have known effects on adipogenesis and adipose tissue angiogenesis. Nine chemicals (rosiglitazone, chlorpyrifos, prochloraz, mancozeb, butylparaben, 15-deoxy-∆12,14-prostaglandin J2, bisphenol A, bis-(2-ethylhexyl) phthalate, and tributyltin chloride) were used to test the adipogenic and angiogenic responses of the vascularized adipose tissue. Three of the chemicals (rosiglitazone, chlorpyrifos, and prochloraz) were employed on both the adipocyte culture and the vascularized adipose tissue model to determine the contribution of the vasculature to the chemicals’ effects. Abbreviations hASC, human adipose stromal cells; HUVEC, human umbilical cord vein endothelial cells; EC50, half maximal effective concentration; ATE, adipose tissue extract; BSA, bovine serum albumin; ECM, extracellular matrix; ins, insulin; SFM, serum free medium; ATEm, ATE medium; SM, stimulation medium; MM, maintenance medium (SM without insulin and troglitazone) Huttala et al. ALTEX 36(3), 2019 421 vascularized adipose tissue model, the HUVEC (4400 cells/cm2) were plated on day 7 by first removing 100 µl of culture media from the wells and then adding the HUVEC in 100 µl of EGM-2 (Lonza). Analyses were performed on day 14. Effect of chemicals on the properties and functions of the vascularized adipose tissue model Table 3 shows the chemicals used and their final concentrations in the test. The vascularized adipose tissue model was exposed to test chemicals on day 8 to investigate the effect of study chemla et al., 2016) with some modifications: one additional medium change and one additional macromolecule component (Ficoll-Paque Plus) were included. Day 0, hASC were plated in hASC medium at a density of 20,000 cells/cm2. TrypLE™ Express (Gibco) was used for the detachment of cells. On days 1, 4, 8 and 11 the medium was changed according to Table 2. Total medium volume used per well in a 48 well plate was 500 µl. After each medium change macromolecule Ficoll-Paque Plus (GE Health care, Buckinghamshire, UK) was added into each well at a volume of 25 µl per 500 µl of culture medium. For the Tab. 1: List of cell culture media and their components and manufacturers Medium hASC medium Serum free medium ATE medium Stimulation medium Maintenance medium (stimulation medium without insulin and troglitazone) Abbreviation - SFM ATEm SM MM Content DMEM/F12 10% human serum 2 mM L-glutamine DMEM/F12 2.56 mM L-glutamine 0.1 nM 3,3’,5-triiodo-L-thyronine sodium salt ITSTM Premix: - 1.15 µM: 6.65µg/ml insulin - 6.65 µg/ml transferrin - 6.65 ng/ml selenious acid 1% bovine serum albumin (BSA) 2.8 mM sodium pyruvate 1800 µg/ml adipose tissue extract (ATE) DMEM/F12 10% human serum 2 mM L-glutamine 50 IU/ml penicillin/ 50 µg/ml streptomycin Serum free medium supplemented with 200 µg/ml ascorbic acid 0.5 µg/ml heparin 5.5 µM: 2 µg/ml hydrocortisone/ cortisol 10 ng/ml vascular endothelial growth factor 1 ng/ml fibroblast growth factor β 9 µM troglitazone DMEM/F12 2.56 mM L-glutamine 0.1 nM 3,3’,5-triiodo-L-thyronine sodium salt 6.65 µg/ml transferrin 6.65 ng/ml selenious acid 1% BSA 2.8 mM sodium pyruvate 200 µg/ml ascorbic acid 0.5 µg/ml heparin 2 µg/ml hydrocortisone 10 ng/ml VEGF 1 ng/ml FGF2 Manufacturer Gibco PAA laboratories Gibco Gibco Gibco Sigma BD biosciences PAA laboratories Gibco - Gibco PAA laboratories Gibco Gibco Sigma Sigma Sigma R&D Systems R&D Systems Sigma Gibco Gibco Sigma Sigma Sigma PAA laboratories Gibco Sigma Sigma Sigma R&D Systems R&D Systems Huttala et al. ALTEX 36(3), 2019 422 of vasculature were assessed. Two independent experiments with two replicates were performed. Half maximal effective concentration (EC50) values were calculated using GraphPad Prism 6.05 software. Mitochondrial activity, triglyceride accumulation, and protein secretion Mitochondrial activity, as an index of the number of living cells, was analyzed by WST-1 (Roche Diagnostics, Basel, Switzerland) after one-hour incubation. Absorbance was measured at 450 nm with a Varioskan Flash Multimode Reader (Thermo Fischer Scientific). The WST-1 results are shown without normalization to depict the cell numbers as well as mitochondrial activity. Triglyceride accumulation in the cell cultures was measured with AdipoRed assay reagent (Lonza) after 10 min incubation at room temperature subsequent to the WST-1 analysis. Fluorescence was measured with a Varioskan Flash Multimode Reader (Thermo Fischer Scientific), with excitation at 485 nm and emission at 572 nm. The amount of accumulated triglyceride per cell icals on both adipogenesis and angiogenesis. After one week of exposure, the analyses (viability/mitochondrial activity, lipid accumulation, amount of vasculature, and with three chemicals also the adipocyte-related gene expression) were performed. The actual purity of the chemicals was used in the dilution calculations. Vehicle controls for all chemicals were exposed to the same concentration of solvent that was present in the chemical dilutions. Both the adipocyte culture and the vascularized adipose tissue model were exposed to three chemicals (rosiglitazone, chlorpyrifos, and prochloraz) on day 8 to investigate the impact of vascularization on the effects of the chemicals on adipocyte lipid accumulation and adipocyte related gene expression. The impact of chemicals on angiogenesis was also investigated. Three independent experiments with two replicates were performed. To further evaluate the performance of the vascularized adipose tissue model, it was challenged with mancozeb, butylparaben, 15-deoxy-Δ12,14-prostaglandin J2, bisphenol A, bis-(2-ethylhexyl) phthalate, and tributyltin chloride and their effects on lipid accumulation, viability/mitochondrial activity, and amount Tab. 2: Culture schedule for the undifferentiated control, the adipocyte culture, and the vascularized adipose tissue model Name Day 0 Day 1 Day 4 Day 7 Day 8 Day 11 Day 14 Undifferentiated cells hASC seeded hASC medium SFM + 25 µl HUVEC in SFM + 25 µl SFM + 25 µl in hASC + 25 µl Ficoll Ficoll EGM-2: Ficoll Ficoll medium on 4400 cells/cm2 48-well plate: Adipocyte culture 20,000 cells/ ATEm SM + 25 µl SM + 25 µl MM + 25 µl Analysis cm2 + 25 µl Ficoll Ficoll Ficoll Ficoll Vascularized adipose ATEm + 25 µl SM + 25 µl HUVEC in SM + 25 µl MM + 25 µl tissue model Ficoll Ficoll EGM-2: 4400 Ficoll Ficoll cells/cm2 Tab. 3: Chemicals used in this study Chemical Product number/ Purity Concentrations tested CAS number Diluent manufacturer Rosiglitazone 71740/Cayman Chemical 99% 139.9; 44; 14; 4.4; 1.4; 0.44; 0.14; 0.044 µM 122320-73-4 0.5% DMSO Chlorpyrifos 45395/SIGMA 99.7% 100; 31; 10; 3.1; 1; 0.3; 0.1; 0.03 µM 2921-88-2 0.5% ethanol Prochloraz 45631/SIGMA 98.6% 5000; 1000; 500; 250; 50; 5; 0.5; 0.05 µM 67747-09-5 0.5% ethanol Mancozeb 45553/SIGMA 97.5% 0.5; 0.16; 0.05; 0.016; 0.005; 0.0015; 0.0005; 8018-01-7 medium 0.00015 µg/ml Butylparaben PHR1022/SIGMA 99.7% 500; 158; 50; 16; 5; 1.5; 0.5; 0.15 µg/ml 94-26-8 medium 15-Deoxy-Δ12, D8440/SIGMA 98.2% 3.15; 1; 0.315; 0.1; 0.0315; 0.01; 0.003; 87893-55-8 0.5% DMSO 14-prostaglandin 0.001 µg/ml J2 Bisphenol A 239658/SIGMA 99.9% 99.4; 31.45; 9.95; 3.15; 0.5; 0.3; 0.0995; 80-05-7 0.5% DMSO 0.03 µg/ml Bis-(2-Ethylhexyl) 36735/SIGMA 99.7% 100; 31.6; 10; 31.7; 1; 0.32; 0.1; 0.0032 µg/ml 117-81-7 0.5% DMSO phthalate Tributyltin 45713/SIGMA 97.3% 325.5; 103; 32.5; 10.3; 3.25; 1.04; 0.33; 1461-22-9 0.5% DMSO chloride 0.11 ng/ml Huttala et al. ALTEX 36(3), 2019 423 (Carl Zeiss), and images were further processed with ZEN 2012 software (Carl Zeiss). All image types were further processed using Adobe Photoshop CS3 software (Adobe Systems Incorporated, San Jose, CA, USA). In order to investigate the area of tubular networks, cell cultures were imaged with Cell-IQ (CM Technologies Oy, Tampere, Finland) as described earlier (Huttala et al., 2018). Briefly, culture plates were imaged with a 10x objective and 5x5 grid per well. Grids were stitched together with Cell-IQ Analyzer (CM Technologies Oy) and further analyzed with ImageJ software (National Institutes of Health (NIH), Bethesda, MD, USA). Images were converted to an 8-bit gray scale, the background was subtracted and binary threshold was adjusted to determine the total tubule area in pixels. The fluorescence from adipocytes was removed from the count by size and circularity restrictions. This analysis method does not take length, number, or branch points of structures into consideration. Hence, the analysis focuses only on changes in the area covered by the vascular network. Gene expression studies Expression of adipose tissue-related genes Glut1, PPARγ, leptin, Glut4, FABP4, PPARα, adiponectin, and PPARγ2 was studied as follows. Gene expression studies of insulin exposed cultures were performed by adding 500 µM insulin for 24 h into the culture prior to RNA isolation. RNA isolation was performed using the Purelink RNA mini kit (Invitrogen, Carlsbad, CA, USA) and PCR was performed with ITaq universal SYBR green one-step kit (Biorad, Hercules, CA, USA), both according to the manufacturer’s instructions. Each reaction contained 30 ng of the template and 300 nM primer concentration. Primer sequences and annealing temperatures are shown in Table S11. A melt curve analysis was performed in each run. PCR amplification was performed with CFX96 Real-Time System (Biorad). Results were calculated with the ΔΔCt method 36B4 as housekeeping gene. Unexposed and insulin exposed samples were analyzed using undifferentiated cells as control. The chemical exposed cultures were compared to a corresponding vehicle control. Fluidigm analysis of vessel related genes The developmental stage of vasculature in the model was further analyzed with a panel of 30 genes by Fluidigm system. Genes analyzed were VEGFA, VEGFR-2/FLK-1, VEGFR-1/FLT-1, FGF2, FGFR2, EGFL7, EGFR, DLL4, Notch 1, Notch 4, AGGF1, PDGF, PDGFRb, ANGPT1, ANGPT2, TIE2, Dkk1, SMO, TGFB, TGFBR1, CD34, VE-cadherin 5, GJA1, occludin, fibronectin, ITGA5, CSPG4 /NG2, S1PR1, eNOS, and HPRT. Results were produced as concentrations based on the standard curve and normalized to housekeeping gene GAPDH. Integrity of the RNA samples was confirmed by using a 2100 Bioanalyzer (Agilent, Santa Clara, CA, USA) and the concentration of the samples was measured with a NanoDrop 1000 Spectrophotometer (Thermo Fisher Scientific). cDNA synthesis was performed using Reverse Transcriptase Master mix (Fluidigm, South San Francisco, CA, USA) according to the manufacturer’s was determined by dividing AdipoRed values by WST-1 values. To calculate the amount of lipid per cell, absorbance values were first normalized to the average of the absorbance values of negative samples of each run. Secretion of proteins, i.e., leptin, adiponectin and FABP4, was compared between the adipocyte culture and the vascularized adipose tissue model. Medium samples were analyzed with ELISA kits for adiponectin, leptin and FABP4 (cat#DRP300, DLP00 and DFBP40, R&D Systems, Abingdon, UK) according to the manufacturer’s instructions. Glucose uptake and lipolysis The glucose uptake test was performed as follows: On day 14, DMEM/F12 (Gibco) was changed on the cells and incubated for 2 h at 37°C with 5% CO2. Then cells were exposed to 100 nM or 500 nM insulin and incubated at 37°C with 5% CO2 for 30 min, and [3H]-2-deoxy-D-glucose (0.2 µCi/well, Perkin Elmer, Waltham, MA, USA) was added for another 20 min. The cells were lysed using 0.1% sodium dodecyl sulfate. The radioactivity of the samples was measured by liquid scintillation counter Wallac1410 (Perkin Elmer) using OptiPhase HiSafe 2 oscillation liquid (Perkin Elmer). Lipolysis, i.e., glycerol release, was analyzed using the EnzyChrom Adipolysis Assay Kit (BioAssay Systems, Hayward, CA, USA) according to the manufacturer’s instructions. On day 14, DMEM/F12 (Gibco) was changed. After one hour, 2 µM isoproterenol was added and incubated for two hours after which 100 nM or 500 nM insulin (Sigma, Saint Louis, MO, USA) was added. After 15 min or 30 min incubation, medium was collected and fluorescence was measured with a Varioskan™ Flash Multimode Reader (Thermo Fisher Scientific) at I ex 530 nm and I em 585nm. Visualization of vasculature by immunocytochemical staining The immunocytochemical staining was performed as described earlier (Huttala et al., 2015) except that the fixative used here was 4% formaldehyde at room temperature for 20 min. Antibodies used were anti-von Willebrand factor IgG (produced in rabbit, Sigma), anti-collagen IV IgG (produced in mouse, Sigma), FITC-labeled goat polyclonal anti-mouse IgG (Sigma), and TRITC-labeled goat polyclonal anti-rabbit IgG (Sigma). In addition, for confocal imaging, anti-collagen IV (produced in rabbit, ab6586, Abcam, Cambridge, UK) used with secondary antibody TRITC-labeled goat polyclonal anti-rabbit IgG (Sigma), CD140b-PE (BD Biosciences, 558821), and CD144-FITC (BD Biosciences, 560411) were used. Nuclei were stained using Fluoroshield™ with DAPI mounting medium (Sigma). Microscopy and quantification of the vascular structures Microscopic imaging was done with a Nikon Eclipse Ti-s inverted fluorescence microscope (Nikon, Tokyo, Japan) and Nikon digital sight DS-U2-camera (Nikon), and images were further processed with NIS Elements (Nikon), ZEN 2012 software (Carl Zeiss, Oberkochen, Germany). Confocal imaging was done with LSM710 and with Zeiss Axio Observer Z1 inverted microscope 1 doi:10.14573/altex.1811271s Huttala et al. ALTEX 36(3), 2019 424 culture and the vascularized adipose tissue model contained lipid accumulating cells, which the undifferentiated cells did not contain (Fig. 1). Upon morphological inspection, the lipid accumulation in the vascularized adipose tissue model was more evenly spread out than in adipocyte culture, where there were fewer adipocytes that contained larger lipid storages. This morphology was also described in our previous study (Huttala et al., 2018) but could not be confirmed by the quantitative results of lipid per cell measurements (Fig. 1). Of the secreted proteins studied, adiponectin was significantly increased in the vascularized adipose tissue model (Fig. 1), but the secretion of leptin and FABP4 was similar in both models. 3.2 Faster lipolysis response to insulin in the vascularized adipose tissue model The responses to insulin were studied in both culture models to determine the influence of the vasculature. Inhibition of lipolysis by insulin was found to be faster in the vascularized adipose tissue model, starting within 15 min in response to 500 nM (Fig. 2), while it only started after 30 min in the adipocyte culture. Both cultures equally increased glucose uptake in response to insulin (Fig. 2). When comparing the vascularized adipose tissue model and adipocyte culture at gene expression level, Glut1 and leptin were significantly upregulated in the vascularized adipose tissue model (Fig. 3). However, there was no difference in secreted leptin between the cultures (Fig. 1). In gene expression analysis of the effect of insulin on both models, Glut1 was shown to be significantly down regulated in response to insulin in the vascularized adipose tissue model (Fig. 3). In addition, a trend towards an upregulation of Glut4 and FABP4 in response to insulin can be seen for both cultures. Leptin and adiponectin expression increased slightly but not significantly in the vascularized adipose tissue model in response to insulin. PPARα, PPARγ2 and general PPARγ expression did not respond to insulin. 3.3 Markers of different stages of vascular formation present in vascularized adipose tissue model The vascularization was analyzed in the vascularized adipose tissue model with immunostaining and expression analysis of a panel of angiogenesis related genes. The gene panel was also analyzed in insulin-challenged samples. The vascular network in the vascularized adipose tissue model was well formed as seen in the collagen IV staining in Figure 4. Both CD144, i.e., ve-cadherin, and CD140b, i.e., PDGFRb, were expressed as expected in well-developed vasculature (Fig. 4). CD144 junctions are located in the tubules and CD140b on the surface of the tubules. CD140b indicates the presence of pericytes that have been shown to be present in the vascular structures previously (Huttala et al., 2015). Gene expression results show the presence of ongoing angiogenesis as well as mature tubules (Tab. S3 1 ). The individual genes that were upregulated (2-fold or more compared to the non-insulin treated cultures) in response to insulin were PDGF, FGFR2, and ve-cadherin 5 (Tab. S3 1 ). In the vascularized adipose tissue model prior to insulin exposure, PDGF and ve-cadherin were exinstructions on a T100 thermal cycler (Biorad). The “no template” control contained water instead of total RNA. In the “no amplification” controls, the Reverse Transcriptase Master Mix was omitted. The cDNA synthesis reaction was performed with 50 ng of total RNA. Pre-amplification was performed using PreAmp Master Mix (Fluidigm) and the instructions provided from Fluidigm for “Fast Gene Expression Analysis Using EvaGreen” using 60°C as annealing temperature on a T100 thermal cycler (Biorad). Primers were designed with Primer3 software. Amplicons were designed to overlap intron sequences where possible. Amplicons were designed to be between 100-250 bp in length. Exonuclease I treatment was performed with Exonuclease I (E. coli) (Lot no 0201507, New England Biolabs, Ipswich, MA, USA) on a T100 thermal cycler (Biorad). After the exonuclease treatment, 10 mM Tris0.1 mM EDTA was added to give a 10-fold dilution of the cDNA, “no template” and “no amplification” control samples. For gene expression analysis, 96.96 Dynamic Array Chip for Gene Expression (Fluidigm) was utilized according to the manufacturer’s instruction using Fluidigm IFC Controller HX (Fluidigm) and BioMark HD (Fluidigm). Sample pre-mixes were prepared by mixing 2x SsoFast EvaGreen Supermix with low ROX (Biorad), Supermix with low ROX (Biorad 172-5211), 20X DNA Binding Dye (Fluidigm 100-7609), preamplified and Exo 1-treated samples. Sample pre-mixes were then added to sample inlets of the primed 96.96 IFC. Assay mixes contained 2X Assay Loading Reagent (Fluidigm, 100-7611), 10 mM Tris-0.1 mM EDTA buffer, and 100 µM forward and reverse primer mix (primer sequences Tab. S2 1 ). Assay mixes were added to assay inlets of the primed 96.96 IFC. The 96.96 IFC was then loaded to the Fluidigm Biomark HD selecting the Gene Expression, ROX, single probe, EvaGreen, and GE 96x96 PCR+Melt v1 Thermal protocol. Results were analyzed using Fluidigm’s BioMark Data Collection Analysis program (version 4.1.3). Statistical analyses All calculations were done in Microsoft Excel 2010 (Microsoft Corporation, Redmond, WA, USA) and statistical analyses were done in GraphPad Prism (GraphPad Software Inc., La Jolla, CA, USA). Results are depicted as mean ± standard deviation; statistical significance is expressed as *, p < 0.05 ; **, p < 0.01; and ***, p < 0.001. Results from triglyceride accumulation (n = 5), ELISA (n = 6), glucose uptake (n = 5), lipolysis (n = 3), and onestep RT-qPCR (n = 6, in insulin exposure n = 3) were subjected to one-way analysis of variance (ANOVA) followed by Fisher’s LSD test. 3 Results 3.1 The vascularized adipose tissue model shows increased secretion of adiponectin but similar lipid accumulation and secretion of leptin and FABP4 compared to the adipocyte culture The characteristics of the novel vascularized adipose tissue model were compared to those of adipocyte culture. Both were differentiated with a modified protocol of the previously published in vitro adipocyte model (Huttala et al., 2016). Both the adipocyte Huttala et al. ALTEX 36(3), 2019 425 drial activity/viability determined by WST-1 increased in adipocyte culture (EC50 7.97 µM) but no effect was seen in the vascularized adipose tissue model in the tested concentration range of 0.0100-140 µM. Only a slight effect of rosiglitazone was seen on lipid accumulation in both models, with lipids increasing at a higher concentration in vascularized adipose tissue model than in adipocyte culture. There was an increase in tubules in response to rosiglitazone at the lower concentrations up to 1.4 µM, but no pressed at lower levels than the housekeeping gene. FGFR2 was already upregulated before insulin exposure. 3.4 The effect of the adipogenesis stimulator rosiglitazone on gene expression Rosiglitazone was used to test the response of the vascularized adipose tissue model to adipogenesis stimulators (Fig. 5). The model was exposed to rosiglitazone for 7 days. The mitochonFig. 1: Morphology, lipid accumulation and secretion of adiponectin, leptin and FABP4 in the vascularized adipose tissue model at day 14 A) Morphology of the different cultures stained with AdipoRed. Top: undifferentiated hASC-HUVEC co-culture; middle: adipocyte culture differentiated from hASC; bottom: vascularized adipose tissue model; scale bar 100 µm. B) Lipids accumulated per cell. Secretion of C) adiponectin, D) leptin, and E) FABP4. Fig. 2: Response of adipocyte culture and vascularized adipose tissue model to insulin at day 14 A) Change in glucose uptake in response to 100 nM and 500 nM insulin. B) Inhibition of lipolysis 15 min and 30 min after exposure to 100 and 500 nM insulin. Ins, insulin Huttala et al. ALTEX 36(3), 2019 426 3.5 Effects of prochloraz and chlorpyrifos on the vascularized adipose tissue model The effects of prochloraz and chlorpyrifos were studied in a similar manner to rosiglitazone after 7 days exposure time. Prochloraz was toxic with an EC50 of 182.2 µM in adipocyte culture and 181.6 µM in the vascularized adipose tissue mod el (Fig. 6). Inhibition of lipid accumulation in adipocyte culture reached EC50 at 164 µM and for the vascularized adipose tissue model at 187.2 µM. No effect on the vasculature was observed at the tested constatistically significant angiogenic effect was seen at the studied concentration range of 0.0100-140 µM. The vascular network was intact and well branched in all exposed cultures. Gene expression of FABP4 and adiponectin greatly increased in response to 4.4 µM rosiglitazone compared to the corresponding unexposed cultures in both the adipocyte culture and the vascularized adipose tissue model. Glut4 was also slightly upregulated in both cultures. The expression of leptin was increased in the vascularized adipose tissue model. Fig. 3: Expression of adipose tissue related genes (Glut1, PPARγ, leptin, Glut4, FABP4, PPARα, adiponectin, and PPARγ2) with or without 500 nM insulin in the adipocyte culture compared to the vascularized adipose tissue model at day 14 ins, insulin Huttala et al. ALTEX 36(3), 2019 427 Fig. 4: Confocal images of vascularization in the vascularized adipose tissue model at day 14 Top row: CD144 (green), i.e., ve-cadherin junctions can be found in the tubules surrounded by collagen IV stained basement membrane (red). Nuclei stained with DAPI (blue). Bottom row: CD140b (green), i.e., PDGFRb staining aligns along the tubules together with collagen IV staining (red). Scale bars 100 µm. Fig. 5: Response of the adipocyte culture and the vascularized adipose tissue model to rosiglitazone A) Mitochondrial activity/viability measured with WST-1. B) Lipid accumulation analyzed by AdipoRed. C) The area of vasculature shown for example concentrations of rosiglitazone. D) Gene expression in the adipocyte culture and the vascularized adipose tissue model after exposure to 4.4 µM rosiglitazone. The red line indicates the level of unexposed corresponding cultures. Huttala et al. ALTEX 36(3), 2019 434 Varzaneh, F. E., Shillabeer, G., Wong, K. L. and Lau, D. C. (1994). Extracellular matrix components secreted by microvascular endothelial cells stimulate preadipocyte differentiation in vitro. Metabolism 43, 906-912. doi:00260495(94)90275-5 Volz, A. C., Hack, L., Atzinger, F. B. and Kluger, P. J. (2018). Completely defined co-culture of adipogenic differentiated ASCs and microvascular endothelial cells. ALTEX 35, 464476. doi:10.14573/altex.1802191 Watson, R. T. and Pessin, J. E. (2007). GLUT4 translocation: The last 200 nanometers. Cell Signal 19, 2209-2217. doi:10.1016/j.cellsig.2007.06.003 Winer, J. P., Janmey, P. A., McCormick, M. E. and Funaki, M. (2009). Bone marrow-derived human mesenchymal stem cells become quiescent on soft substrates but remain responsive to chemical or mechanical stimuli. Tissue Eng Part A 15, 147154. doi:10.1089/ten.tea.2007.0388 Zeiger, A. S., Loe, F. C., Li, R. et al. (2012). Macromolecular crowding directs extracellular matrix organization and mesenchymal stem cell behavior. PLoS One 7, e37904. doi:10.1371/ journal.pone.0037904 Conflict of interest Patent issued in USA (9056084B2), pending elsewhere (WO2010026299A1). Acknowledgements We thank the staff and donors at Tampere University Hospital for collaboration concerning donation of adipose tissue and umbilical cord samples. At the University of Tampere, we want to thank Ms Sari Leinonen, Ms Mirja Hyppönen, Ms Paula Helpiölä, Ms Maaret Vaani and Ms Hilkka Mäkinen for their excellent technical assistance. We would also like to thank Biomedicum Functional Genomics Unit (FuGU) at the University of Helsinki, especially Hanna Ala-Hongisto, for providing the Fluidigm analysis as a service. Funding for the project was provided by The Diabetes Research Foundation, Finnish Funding Agency for Technology and Innovation, and City of Tampere Science Foundation. Robciuc, M. R., Kivela, R., Williams, I. M. et al. (2016). VEGFB/VEGFR1-induced expansion of adipose vasculature counteracts obesity and related metabolic complications. Cell Metab 23, 712-724. doi:10.1016/j.cmet.2016.03.004 Ruiz-Ojeda, F. J., Ruperez, A. I., Gomez-Llorente, C. et al. (2016). Cell models and their application for studying adipogenic differentiation in relation to obesity: A review. Int J Mol Sci 17, 1040. doi:10.3390/ijms17071040 Sargeant, R. J. and Paquet, M. R. (1993). Effect of insulin on the rates of synthesis and degradation of GLUT1 and GLUT4 glucose transporters in 3T3-L1 adipocytes. Biochem J 290, 913919. doi:10.1042/bj2900913 Sarkanen, J. R., Kaila, V., Mannerstrom, B. et al. (2012a). Human adipose tissue extract induces angiogenesis and adipogenesis in vitro. Tissue Eng Part A 18, 17-25. doi:10.1089/ten. TEA.2010.0712 Sarkanen, J. R., Vuorenpaa, H., Huttala, O. et al. (2012b). Adipose stromal cell tubule network model provides a versatile tool for vascular research and tissue engineering. Cells Tissues Organs 196, 385-397. doi:10.1159/000336679 Schwalie, P. C., Dong, H., Zachara, M. et al. (2018). A stromal cell population that inhibits adipogenesis in mammalian fat depots. Nature 559, 103-108. doi:10.1038/s41586-018-0226-8 Sheu, W. H., Ou, H. C., Chou, F. P. et al. (2006). Rosiglitazone inhibits endothelial proliferation and angiogenesis. Life Sci 78, 1520-1528. doi:10.1016/j.lfs.2005.07.046 Switzer, N. J., Mangat, H. S. and Karmali, S. (2013). Current trends in obesity: Body composition assessment, weight regulation, and emerging techniques in managing severe obesity. J Interv Gastroenterol 3, 34-36. doi:10.1111/vaa.12219 Tait, S., Tassinari, R., Maranghi, F. and Mantovani, A. (2015). Bisphenol A affects placental layers morphology and angiogenesis during early pregnancy phase in mice. J Appl Toxicol 35, 1278-1291. doi:10.1002/jat.3176 Taxvig, C., Dreisig, K., Boberg, J. et al. (2012). Differential effects of environmental chemicals and food contaminants on adipogenesis, biomarker release and PPARgamma activation. Mol Cell Endocrinol 361, 106-115. doi:10.1016/j. mce.2012.03.021 van Baak, M. A. (2013). Nutrition as a link between obesity and cardiovascular disease: How can we stop the obesity epidemic? Thromb Haemost 110, 689-696. doi:10.1160/TH13-010045