Prevascular Structures Promote Vascularization in Engineered Human Adipose Tissue Constructs upon Implantation

Abstract

Vascularization is still one of the most important limitations for the survival of engineered tissues after implantation. In this study, we aim to improve the in vivo vascularization of engineered adipose tissue by preforming vascular structures within in vitro-engineered adipose tissue constructs that can integrate with the host vascular system upon implantation. Different cell culture media were tested and different amounts of human adipose tissue-derived mesenchymal stromal cells (ASC) and human umbilical vein endothelial cells (HUVEC) were combined in spheroid cocultures to obtain optimal conditions for the generation of prevascularized adipose tissue constructs. Immunohistochemistry revealed that prevascular structures were formed in the constructs only when 20% ASC and 80% HUVEC were combined and cultured in a 1:1 mixture of endothelial cell medium and adipogenic medium. Moreover, the ASC in these constructs accumulated lipid and expressed the adipocyte-specific gene fatty acid binding protein-4. Implantation of prevascularized ASC/HUVEC constructs in nude mice resulted in a significantly higher amount of vessels (37 ± 17 vessels/mm²) within the constructs compared to non-prevascularized constructs composed only of ASC (3 ± 4 vessels/mm²). Moreover, a subset of the preformed human vascular structures (3.6 ± 4.2 structures/mm²) anastomosed with the mouse vasculature as indicated by the presence of intravascular red blood cells. Our results indicate that preformed vascular structures within in vitro-engineered adipose tissue constructs can integrate with the host vascular system and improve the vascularization upon implantation.

Keywords

Angiogenesis Tissue engineering Adipose tissue Mesenchymal stem cells Endothelial cells

Introduction

Despite various levels of clinical success, currently used surgical procedures to repair soft tissue defects are not optimal. Autologous adipose tissue grafts suffer from donor site morbidity (26) and have a tendency to lose volume over time (24,25,29). Synthetic implants also have significant drawbacks such as infection, rejection, leakage, and/or dislocation (29).

The in vitro development of autologous adipose tissue substitutes using human adipose tissue-derived mesenchymal stromal cells (ASC) may offer a future alternative to these current treatments (1,4,7,8,11). However, limited survival of in vitro-engineered adipose tissue after in vivo implantation is still a major hurdle for clinical use (12,26). The reason for this limited survival is that ingrowth of blood vessels from the host into the engineered adipose tissue construct can take several days (10), which is too slow and fatal for many of the cells in the construct. A sufficiently rapid in vivo vascularization, preventing cell death in the core of the implanted engineered adipose tissue constructs, has not been realized yet.

We and others have shown that endothelial cells (EC) cocultured with stromal cells derived from bone marrow, skin, cardiac muscle, skeletal muscle, or adipose tissue can form primitive vascular-like structures in vitro (2,3,5,9,14,23,27,31,34). Few studies also indicate that such in vitro preformed vascular-like structures (or prevascular structures) are able to connect to the host vessels after implantation (21,23,27,31). The ability of prevascular structures to promote the in vivo vascularization of engineered adipose tissue constructs, however, has not been tested yet.

In this study, we generate prevascularized adipose tissue constructs using three-dimensional (3D) spheroids (about 0.6 mm in diameter) containing cocultured ASC and endothelial cells (EC) and test this hypothesis. Specifically, we investigated the following questions: (1) Can we induce in vitro prevascularization of the 3D spheroid HUVEC/ASC constructs while simultaneously enabling adipogenic differentiation of the ASC? (2) Can the in vitro-generated prevascular structures integrate with the host vasculature?

Materials and Methods

Cell Culture

Subcutaneous abdominal adipose tissue was obtained from donors with approval of the Medical Ethical Committee (# MEC-2005-157). ASC from three different donors of 47, 54, and 59 years of age were isolated from the adipose tissue as previously described (33) and grown on basal medium [Dulbecco's modified Eagle medium 1 g/L glucose (Invitrogen, Carlsbad, CA), 10% FCS (PAA Laboratories, Pasching, Austria), 10⁻¹² M dexamethasone, 10⁻⁵ M ascorbic acid (both from Sigma-Aldrich, St. Louis, MO), 1% penicillin/streptomycin, 0.5% gentamycin (both from Invitrogen)].

Commercially derived pooled human umbilical vein endothelial cells (HUVEC) (Lonza, Verviers, Belgium) were cultured in endothelial growth medium consisting of human endothelial serum-free medium (Invitrogen), supplemented with 20% FCS, 10% human serum (PAA Laboratories), 20 ng/ml fibroblast growth factor 2 (FGF-2), and 100 ng/ml epidermal growth factor (EGF) (both from Peprotech EC, London, UK). Cells were maintained at 37°C in a humidified atmosphere with 5% CO₂, and the medium was replenished every 3 days. Passage 2 ASC and passage 5 or less HUVEC were used in this study.

Medium Selection: Proliferation of ASC and HUVEC in Different Media

The proliferation of ASC and HUVEC in different media was tested in six-well (ASC) and 24-well (HUVEC) plates (Costar®, Corning Inc., Corning, NY) seeded with 1042 and 5000 cells/cm², respectively. ASC and HUVEC were seeded onto the culture plates in basal medium and endothelial growth medium respectively and incubated overnight at 37°C in a humid atmosphere with 5% CO₂ to ensure attachment of the cells. The next day, these media were removed and the cells were washed with PBS and replenished with three different test media: 1) endothelial cell medium consisting of endothelial serum-free medium supplemented with 5% FCS, 20 ng/ml FGF-2, and 100 ng/ml EGF; 2) adipogenic medium consisting of Dulbecco's modified Eagle medium 4.5 g/L glucose (Invitrogen), 10% FCS, 1 μM dexamethasone, 0.01 mg/ml insulin (Eli Lilly, Houten, The Netherlands), 0.2 mM indomethacin (Sigma-Aldrich), 0.5 mM 3-isobutyl-I-methyl-xanthine (Sigma-Aldrich), 1% penicillin/streptomycin, and 0.5% gentamycin; 3) 1:1 mixture of endothelial cell medium and adipogenic medium (endothelial/adipogenic mix medium). The culture media were refreshed every 3 days and cell numbers of duplicate wells were counted twice with a Casy® cell counter (Innovatis, Bielefeld, Germany) at 0, 3, 7, and 10 days of culture.

Formation of Spheroids

Spheroids containing only ASC or ASC with different amounts of HUVEC (5% HUVEC/95% ASC, 20% HUVEC/80% ASC, 40% HUVEC/60% ASC, and 80% HUVEC/20% ASC) were prepared in 10-ml polypropyl-ene tubes (Techno Plastic Products, Trasadingen, Switzerland) in 0.5-ml endothelial cell medium. ASC or HUVEC/ASC suspensions containing a total of 2.5 × 10⁵ cells were added to the tubes and subsequently centrifuged at 150 x g for 5 min. The tubes with the resulting cell pellets were then incubated overnight at 37°C with 5% CO₂ in a humidified atmosphere to allow the formation of spheroids. After overnight incubation, the endothelial cell medium of the ASC/HUVEC coculture spheroids was replaced by endothelial/adipogenic mix medium, and in ASC spheroids by endothelial/adipogenic mix medium or adipogenic medium. Medium was replenished every 2 days.

Flow Cytometric Analysis of Coculture Spheroids

Following 2, 7, and 14 days of culture, HUVEC/ASC coculture spheroids were harvested for flow cytometric analysis. To obtain a single cell suspension, spheroids were washed twice with PBS and digested in a sterile-filtered collagenase type I solution (1000 U/ml in PBS) for 20 min at 37°C. After digestion, cell suspensions were washed with FACS buffer [Hanks' balanced salt solution (Invitrogen), 2% BSA, 0.05% sodium azide (both from Sigma-Aldrich), and 2% heat-inactivated human serum (PAA Laboratories)] and labeled for 30 min on ice in the dark with mouse anti-human CD31 fluorescently labeled antibody [dilution 1:1 in FACS buffer, BD Biosciences Pharmingen (Cat. No. 555446), Erembodegem, Belgium]. Cells were washed again, resuspended in 300 μl of FACS buffer, and directly analyzed on a FACS-Calibur instrument (Becton-Dickinson Biosciences, Franklin Lakes, NJ).

RNA Isolation, Complementary DNA (cDNA) Synthesis, and Quantitative Polymerase Chain Reaction (Q-PCR)

ASC spheroids were cultured in endothelial/adipogenic mix medium or adipogenic medium for 7 days in vitro. Six spheroids were prepared from each ASC donor, and three pools of two spheroids from each donor were used for RNA isolation. The spheroids were fragmented with a pestle on ice and sheared using an insulin syringe.

Total RNA was extracted from ASC spheroids using Qiazol Lysis Reagent (Qiagen, Venlo, The Netherlands). RNA was further purified using the RNeasy Micro Kit (Qiagen) with on-column DNA digestion. Total RNA was quantified using a NanoDrop™ 1000 spectrophotometer (Thermo Scientific, Wilmington, DE) according to the manufacturer's instructions and 250 ng RNA was reverse transcribed into cDNA using RevertAid™ First Strand cDNA Synthesis Kit (Fermentas, St. Leon-Rot, Germany).

The mRNA levels of the adipogenic marker fatty acid binding protein 4 (FABP4), were analyzed using the Taqman® Gene Expression Assay for FABP4 (Hs0060 9791_m1) (ABI) according to the manufacturer's instructions. β-2-Microglobulin (B2M) mRNA levels were analyzed for normalization with the Q-PCR MasterMix Plus for SYBR® Green I dTTP (Eurogentec, San Diego, CA) and the following, gene-specific primers set: forward 5′-TGCTCGCGCTACTCTCTCTTT-3′, reverse 5′-TCTGCTGGATGACGTGAGTAAAC-3′. Q-PCR was performed with an ABI PRISM® 7000 Sequence Detection System and analyzed using 7000 System SDS software (ABI, Foster City, CA).

In Vivo Implantation

Four spheroids consisting of ASC and four spheroids consisting of 80% HUVEC/20% ASC were created from each of the three donors, resulting in a total of 12 ASC spheroids and 12 80% HUVEC/20% ASC spheroids. After 7 days of in vitro culture, the spheroids were implanted subcutaneously in the left and right scapular area of six 9-week-old athymic male nude mice (NMRI-nu/nu, Taconic, Hudson, NY). The mice were placed under general anesthesia with 2.5% isoflurane after which two separate 0.5-cm incisions were made through the dorsal skin. Next, two separate subcutaneous pockets were prepared by blunt dissection of the subcutaneous tissue. One pocket was filled with two ASC spheroids and the other pocket was filled with two 80% HUVEC/20% ASC spheroids. Pockets were closed with discontinuous sutures using Mersilk 5–0 (Ethicon, Somerville, NJ). The content of each pocket was regarded as one construct.

Seven days after implantation, the mice were sacrificed and the constructs retrieved. All procedures were approved by the animal ethical committee (EUR. 1292).

Histology

Constructs harvested 7 days after culture or 7 days after implantation were fixed in 10% formalin in PBS and embedded in paraffin, or directly embedded in Tissue-Tek (Sakura, Finetek Europe, Zoeterwoude, The Netherlands) and snap frozen.

Paraffin-embedded sections (5 μm) were deparaffinized and rehydrated. Cryosections (5 μm) were fixed with 3.7% formalin in deionised water for 1 h.

CD31 Staining of In Vitro Cultured Spheroids

To determine HUVEC organization in spheroid cultures, a monoclonal mouse anti-human CD31 antibody [Clone JC70A, Dako (Cat. No. M0823), Glostrup, Den-mark] was used. Antigen retrieval was achieved by heating at 95°C for 15 min in Dako Cytomation Target Retrieval Solution high pH. After overnight incubation with CD31 antibody (1:40 dilution in PBS/1% BSA) at 4°C, a secondary biotin-conjugated goat anti-mouse antibody [1:200 dilution in PBS/1% BSA, Dako (Cat. No. E0433)] was used for 30 min at room temperature followed by incubation with streptavidin-horseradish peroxidase [1:300 dilution in PBS/1% BSA, Dako (Cat. No. P0397)] for 30 min. Diaminobenzidine (Sigma-Aldrich) was used to visualize CD31 expression. The slides were weakly counterstained with hematoxylin, dehydrated through graded alcohols, and mounted with Permount (VWR International B.V., Amsterdam, The Netherlands).

CD31 and Vimentin Staining in Spheroid Constructs Postimplantation

To distinguish human tissue from mouse tissue and to determine the number and relative area occupied by human vascular structures in construct sections postimplantation, monoclonal mouse anti-human vimentin antibody [Clone V9, 1:40 dilution in PBS/1% BSA, Sigma-Aldrich (Cat. No. V6630)], and monoclonal mouse anti-human CD31 antibody (Dako) were used, respectively. To reduce unspecific binding of the secondary goat anti-mouse antibody (Dako) to mouse IgGs, the mouse-on mouse HRP-Polymer Kit (Biocare Medical, Concord, CA) was used, according to the manufacturer's instructions with some slight modifications. In short, antigen retrieval was performed through incubation in Rodent Decloaker® (Biocare Medical) for 60 min at 95°C. Nonspecific binding sites were blocked with Rodent Block M® (Biocare Medical) and sections were stained overnight with CD31 or vimentin at 4°C. The MM-polymer-HRP® secondary antibody (Biocare Medical) was used, followed by incubation in diaminobenzidine (Sigma-Aldrich) to visualize CD31 or vimentin expression, respectively. The slides were weakly counterstained with hematoxylin, dehydrated through graded alcohols, and mounted with Permount (VWR International B.V., Amsterdam, The Netherlands).

Oil Red O Staining

Accumulated lipid in ASC in spheroids was detected after 7 days of in vitro culture and after 7 days of implantation. In short, a 0.5% (w/v) stock solution of Oil Red O in triethyl-phosphate (Sigma-Aldrich) was diluted 3:2 with deionized water to prepare an Oil Red O working solution. Subsequently, cryosections were fixed, rinsed, and immersed in Oil Red O working solution for 30 min. Hereafter, sections were washed with deionized water and counterstained with hematoxylin for 1 min. Finally, sections were rinsed with running tap water for 10 min, and covered with Imsolmount (Klinipath, Zevenaar, The Netherlands).

Histomorphometry

High resolution (0.23 μm/pixel), low magnification (40x) digital micrographs covering two to three complete vimentin/CD31 immunostained cross sections of each construct were made with a Nanozoomer HT (Hamamatsu Photonics, Hamamatsu City, Japan) for analysis. Subsequently, NIH ImageJ software (http://rsb.info.nih.gov/ij/) was used to count the number of CD31⁺ prevascular structures, the area occupied by these structures, and the complete area of human tissue in the constructs.

Statistical Analysis

All data are expressed as mean ± SD, except a part of the in vivo vascularization data, which are expressed as median (IQR). The FACS data were analyzed with the mixed model analysis of variance using SAS software (SAS Insitute, Cary, NC). The FABP4 expression data were analyzed with the unpaired t-test with Welch correction using Graphpad Prism 5.01 software (Graphpad Software, San Diego, CA). Differences in the number of (mouse) vascular lumens per cross-sectional area between prevascularized and nonprevascularized constructs were analyzed with the nonparametric Mann-Whitney test using Graphpad Prism 5.01 software. Values with p ≤ 0.05 were considered significant.

Results

HUVEC and ASC Proliferation in Different Media

To find the appropriate culture conditions for our HUVEC/ASC spheroid cocultures we determined the proliferation potential of both ASC and HUVEC in three different media: endothelial cell medium, adipogenic medium, and endothelial/adipogenic mix medium.

ASC and HUVEC proliferated on both endothelial cell medium and endothelial/adipogenic mix medium, while neither cell type proliferated in adipogenic medium (Fig. 1). ASC and HUVEC proliferation was higher on endothelial cell medium than on endothelial/adipogenic mix medium. However, ASC had a more spindle-like morphology and accumulated no lipid on endothelial cell medium when compared to the ASC cultured in endothelial/adipogenic mix medium (data not shown), indicating that endothelial/adipogenic mix medium had a more positive effect on the adipogenic differentiation of ASC. Because endothelial/adipogenic mix medium supported both ASC and HUVEC proliferation and was more likely to support ASC adipogenic differentiation than endothelial cell medium, subsequent coculture experiments were conducted in endothelial/adipogenic mix medium.

Figure 1.

Proliferation of ASC and HUVEC in tissue culture flasks in different media. (A) ASC and (B) HUVEC were cultured in tissue culture flasks for 10 days in three different media: endothelial cell medium, adipogenic medium, and a 1:1 mixture of endothelial cell medium and adipogenic medium (endothelial/adipogenic mix medium). The number of cells was determined at days 3, 7, and 10 of culture. Results are shown as mean ± SD. Proliferation of ASC was determined for three different donors. HUVEC and ASC numbers were assayed in duplicate wells.

Formation of Spheroids

Following the described method, ASC and HUVEC/ASC cocultures formed solid 3D spheroids with a diameter of 0.63 ± 0.12 mm. These spheroids were stable during the total subsequent culture period of 14 days.

Flow Cytometric Analyses of CD31-Positive Cells in HUVEC/ASC Spheroid Cocultures

To determine whether the amount of HUVEC seeded in the spheroid cocultures remained stable during culture we determined the percentage of CD31⁺ cells in the coculture spheroids after 2, 7, and 14 days of culture using flow cytometric analysis. The percentage of CD31⁺ cells decreased in all spheroids during culture (Fig. 2A). Especially between days 2 and 7 there was a significant decline in the percentage of CD31⁺ cells in almost all spheroids. However, after 14 days of culture all spheroids still contained CD31⁺ cells, with the spheroids seeded with 80% of HUVEC still containing almost 25% of CD31⁺ cells. This percentage decreased with decreasing HUVEC seeding densities (Fig. 2A–E). Spheroids containing no HUVEC also contained a small percentage (0.77–3.21%) of CD31⁺ cells.

Figure 2.

Flow cytometric analysis of CD31⁺ cells in HUVEC/ASC spheroid cocultures. The percentage of CD31⁺ cells in coculture spheroids seeded with different percentages of HUVEC and ASC was determined using flow cytometric analysis after 2, 7, and 14 days of culture. (A) Graph depicting the percentage of CD31⁺ cells in coculture spheroids seeded with 0%, 5%, 20%, 40%, and 80% HUVEC after 2, 7, and 14 days of culture. Values are shown as mean ± SD. For each of the three ASC donors two spheroid cocultures seeded with 0%, 5%, 20%, 40%, and 80% HUVEC were assayed. Statistical comparisons were performed with the mixed model analysis of variance. ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001 (compared to day 7 and 14 of culture). (B–E) Representative dot-plots, depicting the expression of CD31 in coculture spheroids seeded with (B) 5% HUVEC, (C) 20% HUVEC, (D) 40% HUVEC, and (E) 80% HUVEC after 7 days of culture. The percentages in the gated areas represent the percentage of CD31⁺ cells.

HUVEC Organization in HUVEC/ASC Spheroid Cocultures

After 7 days of culture in endothelial/adipogenic mix medium, cross sections of spheroid cocultures were stained with CD31 to determine endothelial organization. CD31⁺ cells in ASC spheroids seeded with 5%, 20%, and 40% HUVEC were present as round cell clusters dispersed throughout the spheroids (Fig. 3A–C).

Figure 3.

HUVEC elongate and form organized prevascular structures in 80% HUVEC/20% ASC spheroid cocultures. Spheroid cocultures seeded with different percentages of HUVEC were cultured for 7 days in vitro, and immunostained with anti-human CD31 (black), showing organization of HUVEC. Cross sections were counterstained with hematoxylin. (A) 5% HUVEC seeded, (B) 20% HUVEC seeded, (C) 40% HUVEC seeded, and (D) 80% HUVEC seeded (white arrows point at strand and loop formation). Scale bars: 20 μm. Insets show general overview of cross sections. Inset scale bars: 200 μm.

In ASC spheroids seeded with 80% HUVEC, however, the CD31⁺ endothelial cells aligned and formed prevascular structures amidst the ASC (Fig. 3D). This endothelial organization was observed in spheroids of all three donors. Although HUVEC aligned and formed loops, no lumen formation was observed.

Adipogenic Differentiation of ASC in Spheroid Cultures In Vitro

To assess adipogenic differentiation of the ASC in the spheroids, we measured the expression of the adipogenic marker FABP4 in ASC spheroids after 7 days of culture. Q-PCR analysis showed that FABP4 is expressed by ASC in spheroids cultured in endothelial/adipogenic mix medium, albeit significantly lower than in the ASC cultured in adipogenic medium (Fig. 4A).

Figure 4.

ASC in spheroid cultures express the adipocyte-specific gene fatty acid binding protein 4 (FABP4) and stain positive for Oil Red O. (A) ASC spheroid cultures were cultured for 7 days in vitro in endothelial/adipogenic mix medium or in adipogenic medium. Q-PCR was used to measure the expression levels of FABP4. Expression levels are relative to β-2-microglobulin-positive control housekeeping gene (dCt). Values are means ± SDs. Six ASC spheroid cultures were prepared from each of the three ASC donors, and three pools of two spheroids from each donor were assayed. Statistical comparisons were performed with the unpaired t-test with Welch correction. ∗p < 0.05 (compared to adipogenic medium). (B) Cross sections were stained with Oil Red O (red) and counterstained with hematoxylin (blue). Scal bar: 50 μm.

To further characterize adipogenic differentiation, we evaluated the ability of ASC to accumulate lipid after 7 days of culture in endothelial/adipogenic mix medium. Oil Red O staining showed that Oil Red O-positive cells were evident in these ASC spheroid cultures (Fig. 4B).

Assessment of Human Vascular Structures in HUVEC/ASC Spheroids Following Implantation in Nude Mice

Following 7 days of culture, the prevascularized 80% HUVEC/20% ASC spheroids and non-prevascularized ASC spheroids were implanted subcutaneously in nude mice, in order to determine if preformed human vascular structures were able to promote in vivo vascularization. After 7 days the spheroid constructs were retrieved. Gross examination revealed that the constructs still had a spherical shape and were invaded by vessels (Fig. 5).

Figure 5.

Macroscopic view of spheroid constructs 7 days after implantation. Representative images of (A) ASC spheroid construct and (B) 80% HUVEC/20% ASC spheroid construct taken at day 7 postimplantation. Scale bar: 1 mm.

To determine whether human vascular structures were present inside the retrieved constructs, we performed immunostaining with anti-human CD31. Cross sections of the retrieved spheroid constructs showed that none of the ASC spheroid constructs and four out of the six 80% HUVEC /20% ASC spheroid constructs contained human CD31⁺ structures. To provide insight in the vascularization of the constructs, we measured three parameters: 1) the number and 2) density of vascular lumens of human origin and 3) the percentage of human CD31⁺ cells per cross section. The number of human vascular lumens per cross section ranged from 13 to 45 lumens/mm² (average 23 ± 15 lumens/mm²) with a lumen density that ranged from 0.1% to 4.5% (average 1.8 ± 2.0%) and a human CD31⁺ cell density of 6.9 ± 4.5%. The number of human vascular lumens and the lumen density were variable in cross sections of different spheroid constructs of the same ASC donor as well as between ASC donors.

Importantly, the human CD31⁺ vascular structures showed an increased organization when compared to the constructs prior to implantation (compare Fig. 3D with Fig. 6A) and frequently contained lumen. A subset of human vascular lumens (3.6 ± 4.2 lumens/mm²) with a lumen density of 1.0 ± 1.1% was filled with red blood cells, indicating that these lumens were perfused (Fig. 6A, B).

Figure 6.

In vivo organization of prevascular structures in spheroid constructs. Spheroid cocultures seeded with 80% HUVEC /20% ASC were cultured for 7 days in vitro, and then implanted subcutaneously in the right and left scapular region of nude mice for 7 days. Cross sections were immunostained with human-specific CD31 antibody (black) and counterstained with hematoxylin. (A) Overview of a cross section, showing human CD31⁺ vascular structures (spheroid located in white oval). Scale bar: 50 μm. (B) High-magnification image of human CD31⁺ lumina (see black arrows) and unstained mouse vessels (see white arrows). Scale bar: 20 μm. Note the presence of red blood cells in human CD31⁺ lumina.

The number of vascular lumens of mouse origin (lumens not stained with CD31) and the total number of vascular lumens (lumens not stained with CD31 plus the CD31⁺ lumens) per tissue section were also evaluated for the ASC and 80% HUVEC/20% ASC constructs. Both the median values of the number of mouse vascular lumens/mm² and the total number of vascular lumens were significantly higher in 80% HUVEC/20% ASC constructs when compared to ASC constructs [number of mouse vascular lumens/mm² median: 17 (IQR: 14–17) vs. median: 0 (IQR: 0–6), p < 0.05; total number of vascular lumens median 38 (IQR: 35–47) vs. median: 0 (IQR: 0–6), p < 0.05].

Assessment of Lipid Accumulation in Spheroids Following Implantation in Nude Mice

To evaluate whether ASC in spheroid constructs also contained lipids after implantation, cross sections of retrieved constructs were stained with Oil Red O. Similar to the in vitro cultures, Oil Red O-positive cells were evident in spheroid constructs (Fig. 7A).

Figure 7.

Accumulated lipid in HUVEC/ASC spheroid constructs 7 days after implantation. Spheroid cocultures seeded with 80% HUVEC and 20% ASC were cultured for 7 days in vitro, and then implanted subcutaneously in the right and left scapular region of nude mice for 7 days. (A) Cross sections were stained with Oil Red O (red) or (B) immunostained with vimentin (brown) and counterstained with hematoxylin (blue). Scale bar: 50 μm. Note that Oil Red O-positive cells and vimentin-positive cells coincide in the same area.

Vimentin staining of consecutive cross sections of the same spheroid constructs showed that Oil Red O-positive cells and vimentin-positive cells coincide in the same area, indicating that the Oil Red O-positive cells were from human origin (Fig. 7B).

Discussion

The present study demonstrates that our HUVEC/ASC spheroid coculture system leads to the formation of prevascular structures while simultaneously enabling adipogenic differentiation of ASC in vitro, as indicated by the accumulation of lipid and the expression of FABP4. Moreover, our study shows that the prevascular structures within the implanted constructs are able to continue their development in vivo and can be incorporated into the host vasculature, thereby contributing to the vascularization of the constructs.

After 7 days of implantation of the spheroids, we found that most of the 80% HUVEC/20% ASC spheroid constructs contained human CD31⁺ vascular structures. These vascular structures regularly included lumen, indicating stability and advanced development of the vascular structures in vivo. In contrast, ASC spheroid constructs did not contain any human CD31⁺ vascular structures, demonstrating that the addition of HUVEC to ASC spheroids is necessary for the formation of human vascular structures in vivo.

In vitro-formed vascular structures have been shown to be able to connect to host vessels upon implantation (21,23,27,31). In our study, we observed red blood cells in a subset of human CD31⁺ stained lumina, indicating a connection of the HUVEC-derived vascular structures with the mouse vasculature. In addition, we found a higher total number of vascular lumens in cross sections of 80% HUVEC/20% ASC constructs when compared to the total number of vascular lumens in ASC constructs. This higher total number of vascular lumens in 80% HUVEC/20% ASC constructs is partly due to the presence of HUVEC-derived lumens, but also the result of a higher number of mouse-derived lumens, indicating augmented ingrowth of mouse vessels. This increased ingrowth of host (mouse) vessels likely results from paracrine signaling induced by the implanted HUVEC/ASC coculture spheroids. Taken together, these findings indicate that the in vitro formation of prevascular structures in spheroid cultures can contribute to the development of vascular structures in vivo. Such a contribution of in vitro prevascularization on in vivo vascularization has also recently been shown in HUVEC/fibroblast fibrin-based cocultures by Chen et al. (6).

Spheroids composed of combinations of tissue-derived stromal cells and endothelial cells for the purpose of prevascularization have been used before (16, 17,27). To our knowledge, however, this is the first study in which ASC are combined with HUVEC in a spheroid coculture. We found that only spheroids seeded with a high (80%) percentage of HUVEC lead to the formation of prevascular structures. In contrast, Rouwkema et al. (27) showed that the formation of prevascular structures in HUVEC/bone marrow-derived stromal cell (BMSC)/spheroid cocultures was positively affected by low (2%) percentages of HUVEC seeding. It is uncertain whether this difference is due to the use of ASC instead of BMSC, or the use of a different culture medium (endothelial/adipogenic mix medium instead of osteogenic medium).

Although HUVEC are an easily accessible source of endothelial cells, they will not be able to be used in a clinical setting due to their incompatibility with the recipient's immune system. A potential alternative is the use of microvascular endothelial cells derived from adipose tissue (13). Microvascular endothelial cells have already been demonstrated to be able to form vascular structures in cultured skin substitutes after implantation (30). In addition, endothelial progenitor cells derived from blood (22,32) are a promising alternative.

An optimal engineered tissue is assembled from multiple tissue-specific cells, has a typical tissue architecture and function, and supports a vasculature compatible with the native vascular system (15,20,28). Endothelial cell/ASC spheroid constructs may be appropriate tissue units for engineering vascularized adipose tissue equivalents. We have shown that the ASC in our spheroid constructs obtain adipocyte-specific characteristics whereas the HUVEC remain able to form prevascular structures that can integrate with the host (mouse) vasculature in vivo. Spheroid constructs have also been shown to be able to form coherent macrotissue patches (mm³ scale), indicating that spheroid constructs can be scaled-up and used for the generation of larger sized tissue constructs (16). Further studies should determine the in vivo survival of prevascularized adipose tissue constructs in the long term, as well as attempt to engineer thicker prevascularized adipose tissue equivalents. Potential strategies to produce prevascularized adipose tissue equivalents with clinically relevant dimensions could be the integration of spheroid constructs with biomaterials such as photocross-linkable gels and advanced techniques such as microfluidics (18,19).

Conclusion

Based on our results, we conclude that a spheroid coculture of ASC with endothelial cells is suitable for the engineering of prevascularized adipose tissue constructs and that prevascularization improves the vascularization of engineered adipose tissue constructs in vivo. Integration of these prevascular structures in the host vasculature might make the difference between cell survival and cell death in especially thicker (>0.6 mm in diameter) adipose tissue constructs following the initial period after implantation.

Footnotes

Acknowledgments

This work was supported by the NutsOhra Foundation (contract # SNO-T-07-75). The authors wish to thank Suzanne Reneman, Vincent Vaes, and Corinna de Ridder for their assistance with the in vivo experiments, the Department of Pathology for the use of the Nanozoomer HT, Bastiaan Tuk for the support with the histomorphometric analyses and Ed Hull for careful review of this manuscript.

References

Alhadlaq

; Tang

; Mao

J. J.

Engineered adipose tissue from human mesenchymal stem cells maintains pre-defined shape and dimension: Implications in soft tissue augmentation and reconstruction. Tissue Eng. 11(3-4): 556–566; 2005.

Black

A. F.

; Berthod

; L'Heureux

; Germain

; Auger

F. A.

In vitro reconstruction of a human capillary-like network in a tissue-engineered skin equivalent. FASEB J. 12(13): 1331–1340; 1998.

Borges

; Mueller

M. C.

; Padron

N. T.

; Tegtmeier

; Lang

E. M.

; Stark

G. B.

Engineered adipose tissue sup-plied by functional microvessels. Tissue Eng. 9(6): 1263–1270; 2003.

Brey

E. M.

; Patrick

C. W.

Jr. Tissue engineering applied to reconstructive surgery. IEEE Eng. Med. Biol. Mag. 19(5): 122–125; 2000.

Caspi

; Lesman

; Basevitch

; Gepstein

; Arbel

; Habib

I. H.

; Gepstein

; Levenberg

Tissue engineering of vascularized cardiac muscle from human embryonic stem cells. Circ. Res. 100(2): 263–272; 2007.

Chen

; Aledia

A. S.

; Ghajar

C. M.

; Griffith

C. K.

; Putnam

A. J.

; Hughes

C. C.

; George

S. C.

Prevascularization of a fibrin-based tissue construct accelerates the formation of functional anastomosis with host vasculature. Tissue Eng. Part A 15(6): 1363–1371; 2009.

Cho

S. W.

; Song

K. W.

; Rhie

J. W.

; Park

M. H.

; Choi

C. Y.

; Kim

B. S.

Engineered adipose tissue formation enhanced by basic fibroblast growth factor and a mechanically stable environment. Cell Transplant. 16(4): 421–434; 2007.

Fischbach

; Spruss

; Weiser

; Neubauer

; Becker

; Hacker

; Gopferich

; Blunk

Generation of mature fat pads in vitro and in vivo utilizing 3-D long-term culture of 3T3-L1 preadipocytes. Exp. Cell Res. 300(1): 54–64; 2004.

Frerich

; Zuckmantel

; Winter

; Muller-Durwald

; Hemprich

Maturation of capillary-like structures in a tube-like construct in perfusion and rotation culture. Int. J. Oral Maxillofac. Surg. 37(5): 459–466; 2008.

10.

Griffith

L. G.

; Naughton

Tissue engineering—current challenges and expanding opportunities. Science 295(5557): 1009–1014; 2002.

11.

Halberstadt

; Austin

; Rowley

; Culberson

; Loebsack

; Wyatt

; Coleman

; Blacksten

; Burg

; Mooney

; Holder

Jr. A hydrogel material for plastic and reconstructive applications injected into the subcutaneous space of a sheep. Tissue Eng. 8(2): 309–319; 2002.

12.

Hemmrich

; von Heimburg

; Rendchen

; Di Bartolo

; Milella

; Pallua

Implantation of preadipocyte-loaded hyaluronic acid-based scaffolds into nude mice to evaluate potential for soft tissue engineering. Biomaterials 26(34): 7025–7037; 2005.

13.

Hewett

P. W.

; Murray

J. C.

; Price

E. A.

; Watts

M. E.

; Woodcock

Isolation and characterization of microvessel endothelial cells from human mammary adipose tissue. In Vitro Cell. Dev. Biol. Anim. 29A(4): 325–331; 1993.

14.

Kang

J. H.

; Gimble

J. M.

; Kaplan

D. L.

In vitro 3D model for human vascularized adipose tissue. Tissue Eng. Part A 15(8): 2227–2236; 2009.

15.

Kelm

J. M.

; Diaz Sanchez-Bustamante

; Ehler

; Hoerstrup

S. P.

; Djonov

; Ittner

; Fussenegger

VEGF profiling and angiogenesis in human microtissues. J. Biotechnol. 118(2): 213–229; 2005.

16.

Kelm

J. M.

; Djonov

; Ittner

L. M.

; Fluri

; Born

; Hoerstrup

S. P.

; Fussenegger

Design of custom-shaped vascularized tissues using microtissue spheroids as minimal building units. Tissue Eng. 12(8): 2151–2160; 2006.

17.

Kelm

J. M.

; Fussenegger

Microscale tissue engineering using gravity-enforced cell assembly. Trends Biotechnol. 22(4): 195–202; 2004.

18.

Khademhosseini

; Langer

; Borenstein

; Vacanti

J. P.

Microscale technologies for tissue engineering and biology. Proc. Nat. Acad. Sci. USA 103(8): 2480–2487; 2006.

19.

Khademhosseini

; Yeh

; Jon

; Eng

; Suh

K. Y.

; Burdick

J. A.

; Langer

Molded polyethylene glycol microstructures for capturing cells within microfluidic channels. Lab Chip 4(5): 425–430; 2004.

20.

Lee

E. J.

; Vunjak-Novakovic

; Wang

; Niklason

L. E.

A biocompatible endothelial cell delivery system for in vitro tissue engineering. Cell Transplant. 18(7): 731–743; 2009.

21.

Lesman

; Habib

; Caspi

; Gepstein

; Arbel

; Levenberg

; Gepstein

Transplantation of a tissue-engineered human vascularized cardiac muscle. Tissue Eng. Part A 16(1): 115–125; 2010.

22.

Levenberg

Engineering blood vessels from stem cells: Recent advances and applications. Curr. Opin. Biotechnol. 16(5): 516–523; 2005.

23.

Levenberg

; Rouwkema

; Macdonald

; Garfein

E. S.

; Kohane

D. S.

; Darland

D. C.

; Marini

; van Blitterswijk

C. A.

; Mulligan

R. C.

; D'Amore

P. A.

; Langer

Engineering vascularized skeletal muscle tissue. Nat. Biotechnol. 23(7): 879–884; 2005.

24.

Lorenz

H. P.

; Hedrick

M. H.

; Chang

; Mehrara

B. J.

; Longaker

M. T.

The impact of biomolecular medicine and tissue engineering on plastic surgery in the 21st century. Plast. Reconstr. Surg. 105(7): 2467–2481; 2000.

25.

Niechajev

; Sevcuk

Long-term results of fat transplantation: Clinical and histologic studies. Plast. Reconstr. Surg. 94(3): 496–506; 1994.

26.

Patrick

C. W.

Jr. ; Zheng

; Johnston

; Reece

G. P.

Long-term implantation of preadipocyte-seeded PLGA scaffolds. Tissue Eng. 8(2): 283–293; 2002.

27.

Rouwkema

; de Boer

; Van Blitterswijk

C. A.

Endothelial cells assemble into a 3-dimensional prevascular network in a bone tissue engineering construct. Tissue Eng. 12(9): 2685–2693; 2006.

28.

Rouwkema

; Rivron

N. C.

; van Blitterswijk

C. A.

Vascularization in tissue engineering. Trends Biotechnol. 26(8): 434–441; 2008.

29.

Stosich

M. S.

; Mao

J. J.

Adipose tissue engineering from human adult stem cells: Clinical implications in plastic and reconstructive surgery. Plast. Reconstr. Surg. 119(1): 71–83; 2007.

30.

Supp

D. M.

; Wilson-Landy

; Boyce

S. T.

Human dermal microvascular endothelial cells form vascular analogs in cultured skin substitutes after grafting to athymic mice. FASEB J. 16(8): 797–804; 2002.

31.

Tremblay

P. L.

; Hudon

; Berthod

; Germain

; Auger

F. A.

Inosculation of tissue-engineered capillaries with the host's vasculature in a reconstructed skin transplanted on mice. Am. J. Transplant. 5(5): 1002–1010; 2005.

32.

Urbich

; Dimmeler

Endothelial progenitor cells: Characterization and role in vascular biology. Circ. Res. 95(4): 343–353; 2004.

33.

Verseijden

; Jahr

; Posthumus-van Sluijs

S. J.

; Ten Hagen

T. L.

; Hovius

S. E.

; Seynhaeve

A. L.

; van Neck

J. W.

; van Osch

G. J.

; Hofer

S. O.

Angiogenic capacity of human adipose-derived stromal cells during adipogenic differentiation: An in vitro study. Tissue Eng. Part A 15(2): 445–452; 2009.

34.

Verseijden

; Posthumus-van Sluijs

; Pavljasevic

; Hofer

; van Osch

; Farrell

Adult human bone marrow- and adipose tissue-derived stromal cells support the formation of prevascular-like structures from endothelial cells in vitro. Tissue Eng. Part A 16(1): 101–114; 2010.