Acellular adipose matrix as a natural scaffold for tissue engineering

+ MODEL

Journal of Plastic, Reconstructive & Aesthetic Surgery (2013) xx, 1e8

Acellular adipose matrix as a natural scaffoldfor tissue engineering*

Hitomi Sano a,b,*, Hakan Orbay a, Hiroto Terashi c,Hiko Hyakusoku a, Rei Ogawa a

aDepartment of Plastic, Reconstructive and Aesthetic Surgery, Nippon Medical School, Tokyo, JapanbDepartment of Surgical Science, Graduate School of Medicine, The University of Tokyo, Tokyo, JapancDepartment of Plastic and Reconstructive Surgery, Kobe University, Kobe, Japan

Received 13 February 2013; accepted 6 August 2013

KEYWORDSAcellular biologicalmatrices;Adipose tissue;Decellularisation;Scaffold;Tissue engineering

* (1) This work was carried out at NBunkyo-ku, Tokyo 113-0022, Japan. (Venice, Italy, 1e3 December 2011.* Corresponding author. Department

Tokyo 113-8603, Japan. Tel./fax: þ81E-mail address: sasasa116sasasa@h

Please cite this article in press as: SaReconstructive & Aesthetic Surgery (

1748-6815/$-seefrontmatterª2013Brihttp://dx.doi.org/10.1016/j.bjps.2013.0

Summary Background: In conventional tissue-regeneration technologies, stem cells and/orother cells are injected into or incubated on scaffolds. In general, scaffolds can be classifiedinto synthetic and natural polymers and natural matrices. Polymers are generally less suitablethan natural matrices in terms of biocompatibility and biodegradability. A highly promisingalternative may be the acellular adipose matrix (AAM), which is a natural scaffold that couldmediate tissue regeneration without any artefacts. The optimal method for adipose-tissue de-cellularisation is described in this article.Methods: Discarded human adipose tissues harvested from routine operations were used. Inexperiment 1, four different adipose-tissue-decellularisation methods were compared andmodified. In experiment 2, the most effective method was tested by using adipose-tissueblocks from various donor sites (the abdomen, chest and forearm) and of different weights(0.8, 25 and 80 g). Haematoxylin and eosin (H &E) staining, immunohistochemistry (IHC) andscanning electron microscopy were used to determine the efficacy of decellularisation.Results: In experiment 1, a method using an enzymatic digestion solution yielded com-plete decellularisation after some modifications. In experiment 2, the 0.8-g specimenswere completely decellularised by the modified method. However, cell components re-mained in the 25- and 80-g specimens. The donor site had no effect on the degree of de-cellularisation.Conclusions: An optimal method for adipose-tissue decellularisation is reported. BecauseAAM is a natural collagen scaffold that is of human origin, this report describes an

ippon Medical School Department of Plastic, Reconstructive and Aesthetic Surgery, 1-1-5, Sendagi,2) This work was previously presented at ICAT e 1st International Conference on Adipose Tissue,

of Plastic, Reconstructive and Aesthetic Surgery, Nippon Medical School, Sendagi 1-1-5, Bunkyo-ku,3 3822 2131.otmail.com (H. Sano).

no H, et al., Acellular adipose matrix as a natural scaffold for tissue engineering, Journal of Plastic,2013), http://dx.doi.org/10.1016/j.bjps.2013.08.006

tishAssociationofPlastic,ReconstructiveandAestheticSurgeons.PublishedbyElsevierLtd.All rightsreserved.8.006

mailto:[email protected]

http://dx.doi.org/10.1016/j.bjps.2013.08.006



2 H. Sano et al.

+ MODEL

Please cite this article in press as: SaReconstructive & Aesthetic Surgery (

important first step in a tissue-engineering innovation that may be suitable for the regen-eration of various tissues.ª 2013 British Association of Plastic, Reconstructive and Aesthetic Surgeons. Published byElsevier Ltd. All rights reserved.

Autologous tissue transfers or synthetic materials are gener-ally used to repair tissue defects.1 Despite improvements inoperative techniques, invasiveness and complications such asnecrosis, infections, shrinkage and oil-cyst formation canresult in serious problems.2 Moreover, the long-term survivaland functionality of the transplanted materials can belimited. Synthetic and non-human biologically derived ma-terials that have been used as injectable tissue fillers, alongwith volume-filling constructs that are used for soft-tissuerepair, have the advantage of not requiring donor-site sacri-fice. However, transfer of these materials often results ininfection, resorption or foreign-body responses.3 An idealalternativewould be a regenerativemethod that involves theharvesting of human tissue with low invasiveness and thatpermits the reuse of these tissues.

A highly attractive regenerative technique may be to useacellular adipose tissue matrix (AAM). Since the AAM iscomposed of collagen, it may be suitable for the regener-ation of various organs. Moreover, like adipose-stem-celltransfer, which has attracted a great deal of attention inthe tissue regeneration field recently, the great advantageof this matrix is that its harvest requires little donor-sitesacrifice. In the present article, several different decellu-larisation methods that have been described previouslywere compared in terms of their ability to decellulariseadipose tissue. After certain modifications, a highly effec-tive method was obtained. This method was then testedwith adipose-tissue specimens of different sizes and fromdifferent parts of the body.

Table 1 The original acellularization protocols and theirmodifications.

Method A Originalmethod

ModifiedA-1

ModifiedA-2

ModifiedA-3

Freezeand thaw

3 cycles 6 cycles 12 cycles 18 cycles

Method B Originalmethod

ModifiedB-1

ModifiedB-2

ModifiedB-3

1. DNase 1. 1 h 1. 2 h 1. 4 h 1. 8 h2. 4% Sodiumdeoxycholate

2. 1 h 2. 2 h 2. 4 h 2. 8 h

3. 1% Sodiumazide

3. 2 h 3. 4 h 3. 8 h 3. 16 h

Method C Originalmethod

ModifiedC-1

ModifiedC-2

1. 1% TritonX-100

1. 1 h 1. 4 h 1. 16 h

2. DNase 2. 1 h 2. 8 h 2. 24 hMethod D Original

methodModifiedD-1

1. Solution 1 1. 16 h 1. 20 h2. Solution 1 2. 6 h 2. 16 h3. Solution 2 3. 16 h 3. 30 h

Materials and methods

In accordance with the regulations set forth by the Insti-tutional Review Board (IRB) of Nippon Medical School Hos-pital, informed consent was obtained from all patients andtheir families. Discarded human adipose tissues harvestedfrom routine operations were used. The samples weredelivered in sterile physiological saline to the lab for pro-cessing within 2 h.

Experiment 1: comparison and modification ofdifferent decellularisation methods

Prior to processing, the adipose tissue samples were cutinto blocks whose masses ranged from 0.7 to 0.8 g. Thisweight was selected according to the previously publishedpapers. Four decellularisation methods that have beendescribed previously were selected.4e8 These methodswere originally established to decellularise different tis-sues. Various steps in each method were modified, as shownin Table 1, and the abilities of these modified methods todecellularise adipose tissue were analysed. The mosteffective method was selected as follows.

no H, et al., Acellular adipose ma2013), http://dx.doi.org/10.1016

Method A was developed for muscle decellularisation.4

This method was selected because of its feasibility andcost-effectiveness. The original method required that thetissue be subjected to three cycles of freeze-thawing usingliquid nitrogen. The cycle number was modified from 3 to 6(Modified A-1), 12 (Modified A-2) and 18 (Modified A-3)(Table 1).

Method B was developed to isolate intact basementmembranes from bovine retinal and brain blood vessels,rabbit renal tubules and rat renal glomeruli.5 This methodwas selected because it is a classical method that has beenwidely adapted to many other organs with several modifi-cations.6 Thus, the tissues were first placed in distilledwater containing 0.1% sodium azide for 1 h (step 1). Thetissue suspension was then centrifuged and the pellet wassuspended and incubated for 1 h in 40 ml of 1 M NaCl (WakoChemical, Tokyo, Japan) and 2000 Kunitz units of DNAse(Sigma, Deoxyribonuclease 1) (step 2). The mixture wascentrifuged again and the pellet was suspended and incu-bated for 2 h in 40 ml of 4% sodium deoxycholate containing0.1% sodium azide (step 3). Finally, the mixture wascentrifuged and rinsed. Method B was modified by length-ening incubation steps 1, 2 and 3 by two- (Modified B-1),four- (Modified B-2) and eightfold (Modified B-3) (Table 1),respectively.

Method C was developed to decellularise embryoidbodies.7 The method was selected because of its feasibility.Thus, the tissue was kept in a serum-free media for 3 h,

trix as a natural scaffold for tissue engineering, Journal of Plastic,/j.bjps.2013.08.006

Usage of AAM in tissue engineering 3

+ MODEL

after which the tissue was centrifuged and placed in TritonX-10 (VWR) for 1 h (step 1). The mixture was centrifugedagain and the pellet was suspended in 2000 Kunitz units ofDNAse (Sigma, Deoxyribonuclease 1) and stirred for 1 h(step 2). At the end of the treatment, the mixture wascentrifuged and rinsed. Method C was modified by length-ening incubation steps 1 and 2, from 1 h and 1 h to 4 h and8 h (Modified C-1) and 16 h and 24 h, respectively (ModifiedC-2) (Table 1).

Method D was developed for adipose-tissue decellular-isation.8 The tissues were first subjected to three cycles offreeze-thawing in a freezing buffer solution, which is ahypotonic Tris buffer (pH 8.0) containing 10 mM Tris baseand 5 mM ethylenediaminetetraacetic acid (EDTA) (Gibco,Burlington, Canada). Next, the tissues were incubated for16 h in enzymatic digestion solution 1, which consisted of0.25% trypsin/0.1% EDTA (step 1). The samples then un-derwent a 48-h polar solvent extraction in 99.9% iso-propanol (Wako Chemical, Tokyo, Japan). Thereafter, theprocessed tissues were rinsed and incubated for another 6 hin enzymatic digestion solution 1 (step 2). After three morewashes, the samples were incubated for 16 h in enzymaticdigestion solution 2, which consisted of 55 mM Na2HPO4,17 mM KH2PO4, 4.9 mM MgSO4$7H2O, 15,000 U DNAse type II(from bovine pancreas),12.5 mg RNase type III A (ribonu-clease; from bovine pancreas), and 2000 U lipase type VI-S(from porcine pancreas)(step 3). The tissues were thenrinsed and subjected to a final polar solvent extraction in99.9% isopropanol for 8 h. At the end, the tissues wererinsed again. This protocol was modified by lengtheningsteps 1, 2 and 3 from 16 h, 6 h and 16 h to 16 h, 20 h and30 h, respectively (Modified D-1) (Table 1).

Experiment 2: optimisation of the selected method

The most efficacious method for decellularisation of theadipose tissue was selected at the end of experiment 1. Theaim of experiment 2 was to determine the optimum weightthat can be sufficiently decellularised using the selectedmethod and also to find out if there is any difference be-tween the adipose tissues from different donor sites.Therefore, the selected method was tested with 0.8 g ofadipose tissue from abdomen, chest and forearm and withdifferent weights of adipose tissue from the abdomen (0.8,25 and 80 g).

Histology

Haematoxylin and eosin and immunohistochemicalstainingRepresentative samples of unprocessed human adiposetissue and tissue after each procedure were fixed informalin, embedded in paraffin, sectioned into 5-mm slicesand affixed to glass slides. Haematoxylin and eosin (H&E)staining was used to detect the presence of residual cellsor cell fragments in experiments 1 and 2. Immunohisto-chemical (IHC) staining for collagen type IV and lamininwas conducted to assess the architecture of the basement-membrane components in experiment 2. The antibodiesthat were used were rabbit polyclonal anti-collagen typeIV antibody (1:100; AbCam, Cambridge, MA, USA) and

Please cite this article in press as: Sano H, et al., Acellular adipose maReconstructive & Aesthetic Surgery (2013), http://dx.doi.org/10.1016

rabbit polyclonal anti-laminin antibody (1:100; AbCam,Cambridge, MA, USA). Antibody binding was detected byusing the VECTASTAIN Elite ABC rabbit IgG kit (Vectorlaboratories, Burlingame, CA, USA). The VECTOR�

NovaRED� substrate kit (for peroxidase) (Vector labora-tories, Burlingame, CA, USA) served as the enzymesubstrate.

Transmission electron microscopeTransmission electron microscopy was conducted on AAMsamples isolated from various regions and on differentsizes of AAM isolated from the abdominal region inexperiment 2 to confirm decellularisation and to assessthe architecture of the matrix. Oolong tea extract (OTE)staining was used for the ultrastructural observation ofconnective tissues.9 Thus, the tissue was cut into smallpieces and fixed in 2.5% glutaraldehyde in 0.1 M phos-phate buffer at room temperature for 2 h. The specimenswere then postfixed in 1% osmium tetroxide, dehydratedin a graded series of alcohol and embedded in Epok 812.Serial ultrathin sections were cut by a Leica ultracut Fultramicrotome (Leica, Wetzlar, Germany) with a dia-mond knife and picked up on copper grids. The grids werestained with 0.2% OTE (OTE powder; Suntory, Osaka,Japan) in 0.1 M phosphate buffer. The sections weresoaked in single drops of the OTE solution for 20e30 minat room temperature. The sections were then submergedin single drops of distilled water for 5 min, post-stainedwith uranyl acetate and lead citrate, and examined at80 kV under a H-7500 electron microscope (Hitachi,Tokyo, Japan).

Results

Experiment 1

All the results were compared to unprocessed adipose tis-sue histologically (Figure 1a). After method A, the cells andcell components remained, even when the freezeethawcycles were increased by sixfold from 3 to 18 (Figure 1b).Moreover, the macroscopic view revealed that the matrixstructures were destroyed. After methods B and C, cells orcell fragments continued to be observed, especially in thevessel walls. This was true even when the incubation timesof the various steps were lengthened by at least eightfold(Figure 1c, d). When the original protocol of method D wasfollowed, cells and cellular components also remained(Figure 2a). However, complete decellularisation was ach-ieved when some modifications were instituted (Figure 2b).As shown in Figure 3, a significant volume of loose, whitematrix (designated AAM) remained after the 0.8-g tissuewas processed by modified method D-1. The AAM hadsimilar dimensions to the tissue block at the beginning ofthe processing, which suggested that the architecture ofthe matrix was well preserved. The hydrated mass of theAAM represented between 30% and 40% of the original tis-sue mass.

H&E staining confirmed the absence of cells and celldebris in the AAM.

To assess the architecture of the basement componentsin the AAM, IHC staining of collagen type IV and laminin


Figure 1 H&E staining of 0.8 g adipose tissue specimens that were processed by the original and modified methods AeC. (a)Unprocessed normal human adipose tissue from abdominal region. The adipose tissue after processing with (b) original method A,smaller pane shows modified method A-3, (c) original method B, smaller pane shows modified method B-3, (d) original method C,smaller pane shows modified method C-2. The arrows indicate the cells and cell remnants. Scale bars Z 100 mm.

4 H. Sano et al.

+ MODEL

was conducted. In human adipose tissue, laminin andcollagen IV are expressed in the basement membrane thatseparates individual adipocytes and the lining of the sup-porting vascular structures. The IHC-staining analysisrevealed that the laminin and collagen IV contents werepreserved during the tissue processing (Figure 4).Regarding the staining patterns, laminin and collagen IVlined the decellularised, empty lumens of former bloodvessels. Thus, the vascular structures were conservedduring the processing.

TEM analysis revealed the absence of cells or cell frag-ments and that the ultrastructure of the collagen fibrils(CFs) of the processed matrix were well preserved(Figure 5).

The pore size of the resulting AAM was found to be be-tween 20 and 100 mm in diameter. The modified method D-1was thus selected for further testing.

Figure 2 (a) Original method D, (b) and modified method D-1. Tabsent in the matrix that had been generated by modified method


Experiment 2

Subsequently, 0.8-g adipose-tissue blocks from variousdonor sites, namely, the abdomen, chest and forearm,were processed by modified method D-1. H&E stainingrevealed that all samples lacked cells and cell debris(Figure 6). Thus, no difference was found between theadipose tissues from different donor sites in terms of theefficacy of modified method D-1.

Adipose tissues with different weights, namely, 0.8, 25and 80 g, were then processed by modified method D-1. Asexpected, H&E staining revealed that the 0.8-g specimenswere completely decellularised (Figure 7a). Although theperiphery of the 25-g specimens was decellularised(Figure 7b), cell components remained in the central area(Figure 7c). Moreover, in 80-g specimens, cell componentsremained both in the periphery (Figure 7d) and in the

he H&E staining revealed that cells and cell debris were onlyD-1. Scale bars Z 100 mm.


Figure 3 Macroscopic findings of a 0.8 g adipose tissue block before (left) and after (right) processing by modified method D-1.


+ MODEL

central area (Figure 7e). Therefore, the optimal weight ofadipose tissue for preparation of AAM using the selectedmethod was determined to be 0.8 g.

Discussion

Tissue-engineered adipose tissue may be suitable forimproving soft-tissue volume and shape in both recon-structive and cosmetic applications.10,11 Since the adiposescaffold that we describe in this paper is acellular, it doesnot contain any immunogenic components; therefore, itcan easily be transplanted to other patients without therisk of immune rejection. Significant quantities of humanadipose tissue are routinely discarded as medical wasteduring routine operations. Discarded fat tissue can bestored and used as a source for acellular adipose scaffold on

Figure 4 IHC staining of normal human adipose tissue and the AAshowed that both basement membrane components localised athroughout the decellularisation process. Scale bars Z 100 mm.


demand. Even though the initial operations on the donorsmay involve large donor-site scars, the insertion of thepreserved, non-immunogenic acellular adipose scaffoldinto the body of other patients will prevent the creation ofnew donor-site scars in the body of recipients. Conventionaltissue-regeneration technologies involve the injection orincubation of stem cells and/or other cells in scaffolds thatrange from synthetic or natural polymers to naturalmatrices.12 Since synthetic and natural polymers havelimitations in terms of their biocompatibility and biode-gradability, natural matrices are a more promising optionfor tissue regeneration. Most soft- and hard-connectivetissues, such as bone, cartilage, tendon, cornea, bloodvessels and skin, contain CF that are arranged around thecells in a three-dimensional network called the extracel-lular matrix (ECM). Collagen, the main component of ECM,is the most abundant protein in mammals,13 making up

M. Immunostaining for laminin (a&c) and collagen type IV (b&d)long the lumens of the vascular structures were preserved


Figure 5 TEM findings of the AAM after decellularisation withmodified method D-1. Cells or cell fragments could not bevisualised. The ultrastructure of the ECM appeared to be wellpreserved.

6 H. Sano et al.

+ MODEL

approximately 25e35% of the whole-body protein content.The ECM not only offers structural support for cells but alsoprofoundly influences the major cellular programs ofgrowth, differentiation and apoptosis.14,15 These propertiesmean that animal-derived collagens are one of the mostuseful biomaterials available. Indeed, they are widely usedfor tissue engineering.16 However, the use of these animal-derived collagens, including bovine and porcine collagen, inhumans can be compromised by immune reactions/sensi-tisation and the possibility of xenogenic disease trans-mission.17,18 Alternative biomaterials should mimic boththe physical and biochemical properties of native colla-gens. It is a well-known fact that the tissues from the samespecies reduce the risk of an immune reaction, as well asxenogenic disease transmission, typically associated withthe use of animal products, such as bovine or porcinecollagen.11 This means that AAM may be ideal: as anautologous collagenous matrix, it can be used for adipose-tissue regeneration with minimal concerns about allergicreactions and pathogen transmission. Furthermore, thepreserved vascular architecture can be of benefit in termsof promoting the organisation of infiltrating endothelialcells to vascularise the construct and allow for vascularingrowth, subsequently improving the viability of thetransplanted grafts. Thus, the structure of human originscaffold may provide an ideal environment for human cellsto grow with its preserved three-dimensional structure andvascular network.

Figure 6 Comparison of the AAMs derived from various donorabdomen, (b) the chest, and (c) the forearm, were not observed.


When 0.8-g adipose-tissue specimens were processed byusing modified method D-1, between 30% and 40% of theoriginal adipose-tissue weight remained. The pore size wasthen estimated because biomaterials that must replacenative collagen-based ECM should have adequate scaffoldporosity. This is needed to facilitate cell migration and thegrowth of blood vessels across the scaffold and to ensurethe effective exchange of nutrients and waste productsbetween the cells and their microenvironment. It has beensuggested that the ideal pore size is between 10 and100 mm.18 AAM has pore sizes between 20 and 100 mm,which means that it can function as a scaffold. The cellu-larity and cell size of adipose tissue is known to differdepending on the region of the body,19 sex,20 age21 andbody conditions, such as metabolic abnormalities andobesity.21 However, when 0.8-g adipose-tissue blocks fromdifferent donor sites were processed in the present study,differences between donor sites were not observed.Adipose-tissue blocks that differed in weight were alsoprocessed. While the 0.8-g specimens were completelyacellular, the 25- and 80-g blocks continued to bear cells.Thus, modified method D-1 is suitable for adipose tissuefrom various regions of the body as long as the specimensare about 0.8 g in weight. Further research is needed todevelop a protocol that will completely decellularise largertissue specimens. To date, 28 collagen types have beenidentified. Types I, II, III and V are the main ones that makeup the essential collagen in bone, cartilage, tendon, skinand muscle.22e24 Adipose-tissue ECM also contains multipletypes of collagen, including IeVI.25 Adipose tissue is histo-logically categorised as a type of loose connective tissue,and collagen contributes considerably to the non-cell massof this tissue. Moreover, each adipocyte is surrounded by athick ECM, referred to as basal lamina, of which the maincomponent is collagen.26 Since osteoblasts and chon-drocytes, like adipocytes, are of mesenchymal origin, thebasal lamina characteristic of adipocytes is shared by thesebone and cartilage cells. It would be ideal to use humanorigin scaffolds for bone and cartilage regeneration for thesame reasons that human adipose-tissue-derived scaffoldswould be ideal for adipose-tissue regeneration. However,the use of human origin scaffolds in bone and cartilageregeneration has not been reported previously. This mayreflect the significant difficulties that are faced in gener-ating sufficient amounts of bone or cartilage-derived scaf-fold material. By contrast, significant quantities of humanadipose tissue are routinely discarded after abdomi-noplasty, liposuction or panniculectomy. We assume thatAAM could be stored in a freezer of freeze and dry

sites. Differences between the donor sites, namely, (a) theScale bars Z 100 mm.


Figure 7 Ability of original method to decellularize adipose tissue specimens of different weights is shown in this figure. H&Estaining revealed that the 0.8 g specimens were completely decellularized (a). However, although the periphery of the 25 gspecimens was decellularized (b), cell components remained in the central area (c). In the 80 g specimens, the cell componentsremained in both the periphery (d) and the central area (e). Scale bars Z 100 mm.


+ MODEL

technology for extended periods of time for use as an off-the-shelf biomaterial.

Since the structure of adipose tissue is similar to boneand cartilage tissue, AAM may also be suitable as a scaffoldfor bone or cartilage regeneration. It is also possible thatinjections with minced AAM could serve as a non-absorbingnatural filler in the cosmetic field. Further studies on thefunction and utility of this material are needed to deter-mine its suitability in the clinic.

Conclusions

An optimal method for decellularising adipose tissue isdescribed here. Since the AAM scaffold is a natural collagenscaffold and is derived from humans, it can be used forclinical application with minimal risks. Its preserved three-dimensional structure and the intact vascular networkshould greatly enhance the cellecell and cellematrix in-teractions within the scaffold upon cell seeding and in vivotransplantation, subsequently yielding a better viability.This report describes an important first step in a tissue-engineering innovation that may be suitable for theregeneration of different tissues.

Conflict of interest/funding statement

Authors have no competing financial interests or funding todeclare.

References

1. Erba P, Ogawa R, Vyas R, Orgill DP. The reconstructive matrix:a new paradigm in reconstructive plastic surgery. PlastReconstr Surg 2010;126:492e8.


2. de Blacam C, Momoh AO, Colakoglu S, Tobias AM, Lee BT.Evaluation of clinical outcomes and aesthetic results afterautologous fat grafting for contour deformities of the recon-structed breast. Plast Reconstr Surg 2011;128:411e8.

3. Friedman RJ. Silicone breast prostheses implantation andexplanation. Semin Rheum 1994;24:8e10.

4. Fansa H, Keilhoff G, Wolf G, Schneider W. Tissue engineering ofperipheral nerves: a comparison of venous and acellular mus-cle grafts with cultured Schwann cells. Plast Reconstr Surg2001;107:485e96.

5. Sutherland RS, Baskin LS, Hayward SW, Cunha GR. Regenera-tion of bladder urothelium, smooth muscle, blood vessels andnerves into an acellular tissue matrix. J Urol 1996;156:571e7.

6. Schultheiss D, Gabouev AI, Cebotari S, et al. Biological vascu-larized matrix for bladder tissue engineering: matrix prepara-tion, reseeding technique and short-term implantation in aporcine model. J Urol 2005;173:276e80.

7. Nair R, Ngangan AV, McDevitt TC. Efficacy of solvent extractionmethods for acellularization of embryoid bodies. J BiomaterSci Polym Ed 2008;19:801e19.

8. Flynn LE. The use of decellularized adipose tissue to provide aninductive microenvironment for the adipogenic differentiation ofhumanadipose-derivedstemcells.Biomaterials2010;31:4715e24.

9. Sato Y, Sasaki A, Adachi W, Dai XL, Liu S, Namimatsu S. Use ofoolong tea extract (OTE) for elastin staining and enhancementin ultrathin sections. Med Electron Microsc 2003;36:179e82.

10. Katz AJ, Llull R, HedrickMH, Futrell JW. Emerging approaches tothe tissue engineering of fat. Clin Plast Surg 1999;26:587e603.

11. Patrick Jr CW. Adipose tissue engineering: the future of breastand soft tissue reconstruction following tumor resection.Semin Surg Oncol 2000;19:302e11.

12. Itoi Y, Takatori M, Hyakusoku H, Mizuno H. Comparison of readilyavailable scaffolds for adipose tissue engineering using adipose-derived stemcells. J Plast Reconstr Aesthet Surg 2010;63:858e64.

13. Di Lullo GA, Sweeney SM, Korkko J, Ala-Kokko L, SanAntonio JD. Mapping the ligand-binding sites and disease-associated mutations on the most abundant protein in thehuman, type I collagen. J Biol Chem 2002;277:4223e31.

14. Nakajima I, Muroya S, Tanabe R, Chikuni K. Positive effect ofcollagen V and VI on triglyceride accumulation during


http://refhub.elsevier.com/S1748-6815(13)00470-1/sref1
























































8 H. Sano et al.

+ MODEL

differentiation in cultures of bovine intramuscular adipocytes.Differentiation 2002;70:84e91.

15. Adams JC, Watt FM. Regulation of development and differentia-tionby theextracellularmatrix.Development1993;117:1183e98.

16. Oliveira SM, Ringshia RA, Legeros RZ, et al. An improvedcollagen scaffold for skeletal regeneration. J Biomed MaterRes A 2010;94:371e9.

17. Koide T. Designed triple-helical peptides as tools for collagenbiochemistry and matrix engineering. Philos Trans R Soc Lond BBiol Sci 2007;29:1281e91.

18. Yannas IV, Tzeranis DS, Harley BA, So PT. Biologically activecollagen-based scaffolds: advances in processing and charac-terization.PhilosTransactAMathPhysEng Sc2010;28:2123e39.

19. Tchoukalova YD, Koutsari C, Karpyak MV, et al. Subcutaneousadipocyte size and body fat distribution. Am J Clin Nutr 2008;87:56e63.

20. Sjostrom L, Smith U, Krotkiewski M, Bjorntorp P. Cellularity indifferent regions of adipose tissue in young men and women.Metabolism 1972;21:1143e53.


21. Ktotkiewski M, Sjostrom L, Bjorntorp P, Smith U. Regionaladipose tissue cellularity in relation to metabolism inyoung and middle-aged women. Metabolism 1975;24:703e10.

22. Veit G, Kobbe B, Keene DR, Paulsson M, Koch M, Wagener R.Collagen XXVIII, a novel von Willebrand factor A domain-containing protein with many imperfections in the collage-nous domain. J Biol Chem 2006;281:3494e504.

23. Canty EG, Kadler KE. Procollagen trafficking, processing andfibrillogenesis. J Cell Sci 2005;118:1341e53.

24. Smith K, Rennie MJ. New approaches and recent results con-cerning human-tissue collagen synthesis. Curr Opin Clin NutrMetab Care 2007;10:582e90.

25. Divoux A, Clement K. Architecture and the extracellular ma-trix: the still unappreciated components of the adipose tissue.Obes Rev 2011;12:494e503.

26. Pierleoni C, Verdenelli F, Castellucci M, Cinti S. Fibronectinsand basal lamina molecules expression in human subcutaneouswhite adipose tissue. Eur J Histochem 1998;42:183e8.





















































Documents

Acellular adipose matrix as a natural scaffold for tissue engineering