ORIGINAL ARTICLE Journal of
Multipotential Human Adipose-Derived Stromal Stem Cells Exhibit a Perivascular Phenotype In Vitro and In Vivo
Cellular Physiology
A.C.W. ZANNETTINO,1 S. PATON,2 A. ARTHUR,2 F. KHOR,2 S. ITESCU,3 J.M. GIMBLE,4 AND S. GRONTHOS2* 1
Myeloma Research Program, Bone and Cancer Laboratories, Division of Haematology, Institute of Medical and Veterinary Science,
Hanson Institute and University of Adelaide, Adelaide, Australia 2
Mesenchymal Stem Cell Group, Bone and Cancer Laboratories, Division of Haematology, Institute of Medical and
Veterinary Science, Hanson Institute and University of Adelaide, Adelaide, Australia 3
Department of Medicine, University of Melbourne, Melbourne, Australia and Department of Transplantation Immunology,
Columbia University, New York—Presbyterian Hospital, New York 4
Pennington Biomedical Research Center, Louisiana State University, Baton Rouge, Louisiana
Mesenchymal stem-like cells identified in different tissues reside in a perivascular niche. In the present study, we investigated the putative niche of adipose-derived stromal/stem cells (ASCs) using markers, associated with mesenchymal and perivascular cells, including STRO-1, CD146, and 3G5. Immunofluorescence staining of human adipose tissue sections, revealed that STRO-1 and 3G5 co-localized with CD146 to the perivascular regions of blood vessels. FACS was used to determine the capacity of the CD146, 3G5, and STRO-1 specific monoclonal antibodies to isolate clonogenic ASCs from disassociated human adipose tissue. Clonogenic fibroblastic colonies (CFU-F) were found to be enriched in those cell fractions selected with either STRO-1, CD146, or 3G5. Flow cytometric analysis revealed that cultured ASCs exhibited similar phenotypic profiles in relation to their expression of cell surface markers associated with stromal cells (CD44, CD90, CD105, CD106, CD146, CD166, STRO-1, alkaline phosphatase), endothelial cells (CD31, CD105, CD106, CD146, CD166), haematopoietic cells (CD14, CD31, CD45), and perivascular cells (3G5, STRO-1, CD146). The immunoselected ASCs populations maintained their characteristic multipotential properties as shown by their capacity to form Alizarin Red positive mineralized deposits, Oil Red O positive lipid droplets, and Alcian Blue positive proteoglycan-rich matrix in vitro. Furthermore, ASCs cultures established from either STRO-1, 3G5, or CD146 selected cell populations, were all capable of forming ectopic bone when transplanted subcutaneously into NOD/SCID mice. The findings presented here, describe a multipotential stem cell population within adult human adipose tissue, which appear to be intimately associated with perivascular cells surrounding the blood vessels. J. Cell. Physiol. 214: 413–421, 2008. ß 2007 Wiley-Liss, Inc.
A stromal stem cell population (CFU-F: colony forming units-fibroblastic), distinct from the resident haematopoietic stem/progenitor cell populations, responsible for the maintenance of the stromal tissue of adult bone marrow (BM) was first described by Friedenstein and colleagues (Friedenstein et al., 1970, 1976; Friedenstein, 1980). This precursor compartment, was found to originate from a small stromal stem cell pool, distinct from the resident haematopoietic stem/ progenitor cell populations (Friedenstein et al., 1978; Friedenstein, 1980). These seminal findings have led to a gradual escalation of research activity focused on the cellular and molecular characteristics of bone marrow stromal stem cells (BMSSC) or mesenchymal stem cells (MSC) (Owen and Friedenstein, 1988). BMSSC have been proposed as a novel cellular therapy for regenerative medicine due to their capacity to differentiate into multiple cells types including myelosupportive stroma, smooth muscle cells, osteoblasts, adipocytes, chondrocytes, cardiomyocytes, and astrocytes (Kuznetsov et al., 1997; Azizi et al., 1998; Pittenger et al., 1999; Liechty et al., 2000; Toma et al., 2002; Gronthos et al., 2003; Airey et al., 2004; Martens et al., 2006). Whilst, these multipotential BMSSC hold enormous promise for future cellular-based tissue engineering applications, their rarity and limited ex vivo growth potential has restricted the wide spread clinical use. For these reasons, a number of investigators have sought alternative sources of ß 2 0 0 7 W I L E Y - L I S S , I N C .
adult MSC-like cells outside of the bone marrow cavity including bone trabeculae, muscle, peripheral blood, synovium, dental pulp, periodontal ligament and adipose tissue (Gronthos et al., 2000; Kuznetsov et al., 2001; Young et al., 2001; Gimble and Guilak, 2003; Miura et al., 2003; Tuli et al., 2003; Seo et al., 2004; Sakaguchi et al., 2005). To date, extramedullary adipose tissue represents the most readily accessible site for harvesting the large quantities of MSC-like cells thought to be required for cellular based therapies. Previous studies have shown that adipose-derived stromal/stem cells (ASCs), like their BMSSC counterparts, possess the capacity to differentiate into different cell lineages
Contract grant sponsor: Australian National Health and Research Council Project; Contract grant number: 242804. *Correspondence to: S. Gronthos, Division of Haematology, Institute of Medical and Veterinary Science, Frome Road, Adelaide 5000, Australia. E-mail:
[email protected] Received 20 April 2007; Accepted 8 June 2007 DOI: 10.1002/jcp.21210
413
414
ZANNETTINO ET AL.
including osteogenic, adipogenic, chondrogenic, myogenic, and cells of neuronal phenotype (Zuk et al., 2001; Gimble and Guilak, 2003; Wickham et al., 2003; Safford et al., 2004; Strem et al., 2005). Whilst, the accessibility and quantity of ASCs offers much appeal, the practical application and safety of using these cells for clinical application remains to be determined. A greater understanding of the microenvironmental cues that regulate this newly identified population may provide important insights into their properties to help manage the cultivation of these cells in vitro and to help prepare safe therapeutic cell preparations for regenerative medicine applications (Nathan et al., 2003). The environmental hub or niche that nurtures and maintains stromal stem cells at different anatomical sites is one area that is poorly understood. We have recently demonstrated that different MSC-like populations derived from various adult tissues such as bone marrow (BMSSC), periodontal (PDLSC: periodontal ligament stem cells), and dental pulp tissue (DPSC: dental pulp stem cells) can be identified and isolated based on their expression of a panel of markers associated with smooth muscle cells, pericytes and endothelial cells (Miura et al., 2003; Shi and Gronthos, 2003; Seo et al., 2004). Collectively, these studies show that MSC-like populations within bone marrow, dental pulp and periodontal tissues share a common perivascular stem cell niche within the microvascular networks of their respective tissues. These findings are supported by a number of reports demonstrating that MSC-like cells from different postnatal tissues exhibit the characteristics of pericyte-like cells (Brighton et al., 1992; Galmiche et al., 1993; Nehls and Drenckhahn, 1993; Schor et al., 1995; Schor and Canfield, 1998; Doherty et al., 1998; Carlile et al., 2000; Helmbold et al., 2001; Zuk et al., 2001; Shi and Gronthos, 2003). In an attempt to characterize the stem cell niche that harbors ASCs, we have utilized a number of well characterized markers of smooth muscle cells, endothelial cells, and pericytes to identify and selectively isolate ASCs directly from human adipose tissue. The findings presented here, describe a multipotent stem cell population within adult human adipose tissue, which appear to be intimately associated with perivascular cells surrounding the blood vessels. Materials and Methods Tissue samples
Abdominal wall derived adipose-tissue was obtained during routine adominoplasty following informed patient consent and according to the guidelines set by the Royal Adelaide Hospital Human Ethics Committee. The adipose tissue (three female and two male donors between 25 and 45 years old) was separated from the skin and dermis crown, cut into small pieces and digested in a solution of a solution of 3 mg/ml collagenase type I (Worthington Biochem, Freehold, NJ) and 4 mg/ml dispase (Boehringer Mannheim, GMBH, Germany) for 2–3 h at 378C. The enzyme-digested cell tissue preparation was subsequently washed twice in HHF (HBSS supplemented with 10 mM HEPES and 5% FCS) by centrifugation at 400g. A single cell suspension (0.01–1 105/well) of adipose-derived cells was recovered by sequential separation through a 70 mm and then a 40 mm cell strainer (Falcon BD Labware, Franklin Lakes, NJ). ASCs were then isolated by FACS as described below. Purified preparation of ASCs were cultured in a-MEM supplemented with 20% FCS, 2 mM L-glutamine, and 100 mM L-ascorbate-2-phosphate as previously described (Gronthos et al., 2003; Shi and Gronthos, 2003). Colony efficiency was assessed on day 14 of culture after 4% formalin fixation and staining with 0.1% toluidine blue as previously described (Gronthos et al., 2000, 2003). The unfractionated cells and different imunoselected populations were plated in triplicate cultures at 1.0 105 cells per 10 cm2 dish. Aggregates of 50 cells were scored as colonies. JOURNAL OF CELLULAR PHYSIOLOGY DOI 10.1002/JCP
Fluorescence activated cell sorting (FACS)
Single cell suspensions of adipose tissue, were incubated with either STRO-1 (mouse IgM anti-human stromal stem cells (Simmons and Torok-Storb, 1991), 3G5 (mouse IgM anti-human islet cells; ATCC), or CC9 (mouse IgG2a anti-human CD146/MUC18/Mel-CAM) (Shi and Gronthos, 2003), antibody supernatant (1/2) for 45 min on ice. Isotype-matched control mouse monoclonal antibodies used in this study: 1A6.12 (IgM), and 1A6.11 (IgG2b) (kindly provided by Prof. L.K. Ashman, University of Newcastle, NSW, Australia). After washing with HHF, the cells were incubated with either goat anti-mouse IgM or IgG conjugated to PE (Southern Biotechnology Associates, Inc., Birmingham, AL) for 45 min on ice. Single-color FACS was performed using a FACStarPLUS flow cytometer (Becton Dickinson, Sunnyvale, CA). Positive reactivity for each antibody was defined as the level of fluorescence greater than 99% of the isotype matched control antibodies. Flow cytometric analysis
Ex vivo expanded populations were used at between passage 4 of culture for the immunophenotypic analysis (approximately 20 population doublings). Ex vivo expanded ASCs were treated with trypsin/EDTA and resuspended in blocking buffer for 30 min. Individual tubes containing 1 105 cells were incubated with murine monoclonal antibodies reactive to either CD14, CD31, CD45, CD90, CD105, CD166 (BD Biosciences, San Jose, CA), CD44 (Clone H9H11; Dr. Andrew Zannettino, Division of Haematology, IMVS, Adelaide, SA, Australia), CD106 (Clone QE4G9; kindly provided by Dr. Ravi Krishnan, Transplantation Immunology Laboratory, The Queen Elizabeth Hospital, Adelaide SA, Australia), 3G5 (American Tissue Culture Collection, Manassas, VA), alkaline phosphatase (clone B4-78; Developmental Studies Hybridoma Bank, Iowa, IA) or isotype matched controls, 1B5 (IgG1), 1A6.11 (IgG2b), and 1A6.12 (IgM) (kindly provided by Prof. L.K. Ashman, The University of Newcastle, Newcastle, NSW, Australia) at a concentration of 10 mg/ml for 1 h on ice. After washing, cells were incubated with secondary detection reagents, goat anti-mouse IgG-FITC or IgM-FITC conjugated antibodies (1/50; Southern Biotechnology Associates, Inc., Birmingham, AL) for 45 min on ice. Following washing, samples were analyzed using an Epics-XL-MCL flow cytometer (Beckman Coulter, Hialeah, FL). Immunohistochemistry
Frozen tissue sections (5 mm) and cytospin preparations were fixed with cold acetone at 208C for 15 min then washed in PBS. The samples were subsequently incubated with PBS containing 1.5% of hydrogen peroxide for 30 min, washed with PBS and blocked with 5% non-immune goat serum for 1 h at room temperature. Samples were incubated with primary antibodies for 1 h at room temperature in the following combinations: (i) 1A6.12/1B5; (ii) STRO-1/CD146; (iii) 3G5/CD146. Dual-color fluorescence labeling was achieved by adding the secondary antibodies, goat anti-mouse IgM-Texas Red and IgG-FITC at 1/50 dilution (CALTAG Laboratories), for 45 min at room temperature. After washing, the samples were mounted in VECTASHIELDTM fluorescence mountant. Confocal imaging was performed using a Nikon Eclipse TE 2000-E microscope and analyzed by E2-C1 Viewer 3.20 Nikon Software. In vivo transplantation studies
Approximately 5.0 106 ex vivo expanded ASCs (passage 4) derived from either STRO-1, 3G5, or CD146 selected populations were mixed with 40 mg of hydroxyapatite/tricalcium phosphate (HA/TCP) ceramic powder (Zimmer, Inc., Warsaw, IN) and then transplanted subcutaneously into both left and right side pockets formed in the dorsal surface of 10-week-old immunocompromised NOD/SCID mice (ARC, Perth, WA, Australia) for 8 weeks as
PERIVASCULAR ORIGIN OF ADIPOSE-DERIVED STROMAL STEM CELLS
previously described (Gronthos et al., 2003). These procedures were performed in accordance to specifications of an approved animal protocol (University of Adelaide Ethics Number M19/2005). Implants (two per mouse) were recovered after 8 weeks, fixed in 4% paraformaldehyde for 2 days, then decalcified for a further 10 days in 10% EDTA prior to embedding in paraffin. Each transplant (three replicates per cell line) was cut into two pieces, then placed cut-surface down for paraffin embedding. Five micrometer sections of the implants were prepared and stained with H&E representative of the middle and either end of each transplant approximately 3–4 mm in length. The amount of new bone formation was calculated as a percentage of the total surface area present in twelve tissue sections using Scion Imaging Software. The unpaired t-test was used to compare the percent new bone formation per unit area between different samples. Reverse transcription-polymerase chain reaction (RT-PCR)
Total RNA was prepared from passage 4 ex vivo expanded ASCs using RNA STAT-60 (TEL-TEST, Inc. Friendswood, TX). First-strand cDNA synthesis was performed with a first-strand cDNA synthesis kit (Invitrogen Life Technologies, Carlsbad, CA) using an oligo-dT primer as recommended by the manufacturer. First strand cDNA (2 ml) was added to 46 ml of a 1 PCR master reaction mix (Roche Diagnostics Gmbh, Mannheim, Germany) and 10 pmol of each human specific primer sets: CBFA1 (632 bp, and three smaller alternative splice variants) sense, 50 -CTATGGAGAGGACGCCACGCCTGG-30 ; antisense, 50 -CATAGCCATCGTAGCCTTGTCCT-30 ; osteocalcin (310 bp) sense, 50 -CATGAGAGCCCTCACA-30 ; antisense, 50 -AGAGCGACACCCTAGAC-30 ; GAPDH (800 bp) sense, 50 -AGCCGCATCTTCTTTTGCGTC-30 ; antisense, 50 -TCATATTTGGCAGGTTTTTCT-30 (Gronthos et al., 1999, 2003). The reactions were incubated in a PCR Express Hybaid thermal cycler (Hybaid, Franklin, MA) at 958C for 2 min for 1 cycle then 948C for 30 sec, 608C for 30 sec, 728C for 45 sec for 35 cycles, with a final 7 min extension at 728C. Following amplification, each reaction was analyzed by 1.5% agarose gel electrophoresis, and visualized by ethidium bromide staining.
Differentiation of adipose-derived colony forming units
We have previously reported the in vitro conditions to induce human BM stromal cells to develop a mineralized bone matrix, adipose cells, and chondrocytes. To induce osteogenesis, passage 4 ASCs were cultured in a-MEM supplemented with 10% FCS, 100 mM L-ascorbate-2-phosphate, dexamethasone 107 M and 3 mM inorganic phosphate for 4 weeks, where mineralized deposits were identified by Alizarin Red staining (Gronthos et al., 1994, 2003). Adipogenesis was induced in the presence of 0.5 mM methylisobutylmethylxanthine, 0.5 mM hydrocortisone, and 60 mM indomethacin for 3 weeks, where Oil Red O staining was used to identify lipid-laden fat cells as previously described (Gimble and Guilak, 2003; Gronthos et al., 2003). Chondrogenic differentiation was assessed in aggregate cultures treated with 10 ng/ml TGF-b3 for 3 weeks, and assessed by Alcian blue staining for proteoglycan synthesis as previously described (Gronthos et al., 2003; Yoo et al., 1998). Results Adult stromal cells derived from peripheral adipose tissue exhibit a perivascular-like phenotype
Mounting evidence suggests that MSC-like cells identified in different tissues reside in a perivascular niche (Brighton et al., 1992; Canfield et al., 2000; Bianco et al., 2001). Consistent with this notion, we determined the putative niche of ASCs by co-staining sections of human peripheral adipose tissue with antibodies to markers associated with mesenchymal and perivascular cells, including STRO-1, CD146, and 3G5 (Shi and Gronthos, 2003). As shown in Figure 1, dual immunofluorescence staining of frozen sections of human adipose tissue, revealed that CD146 co-localized with a minor proportion of 3G5-positive pericytes at the outer wall of large blood vessels (Fig. 1G). Similarly, the STRO-1 and CD146 antigens co-localized in perivascular regions surrounding large blood vessels as analyzed by confocal microscopy (Fig. 1H). Our previous studies (Gronthos et al., 2003; Shi and Gronthos, 2003; Seo et al., 2004) have shown that monoclonal
Fig. 1. Reactivity of perivascular makers in human adipose tissue. Dual-color immunofluorescence staining in situ demonstrating reactivity of (A,B) 1A6.12 and 1B5, (C,D) CD146, (E) 3G5, (F) STRO-1, (G) 3G5 and CD146, and (H) STRO-1 and CD146, reactivity around large blood vessels in frozen sections of human adipose tissue (bar U 20mm). The 1A6.12, STRO-1, and 3G5 antibodies were identified using a goat anti-murine IgM-Texas Red while 1B5 and CD146 were identified using a goat anti-murine IgG-fluorescein isothiocyanate. Co-localization of (G) 3G5 and CD146 or (H) STRO-1 and CD146 is indicated by overlaping areas of yellow and orange fluorescence (white arrows) was analyzed using a Nikon Eclipse TE 2000-E confocal microscope. Cell nuclei were stained blue with DAPI. Phase contrast images of the same blood vessels are shown (I) and (J).
JOURNAL OF CELLULAR PHYSIOLOGY DOI 10.1002/JCP
415
416
ZANNETTINO ET AL.
antibodies to the cell surface markers CD146, 3G5, STRO-1 isolate the majority of CFU-F present within bone marrow, periodontal and dental pulp tissues. In the present study, we examined the capacity of these antibodies to isolate colonyforming cells (CFU-F) derived from single cell suspensions of enzymatic digested human adipose tissue, using single-color FACS. Representative examples of the staining patterns observed following incubation of this cell population with monoclonal antibodies to CD146 (29% mean positive population 6.3 SE, n ¼ 4 donors), 3G5 (24% mean positive population 11.0 SE, n ¼ 4 donors) and STRO-1 (41.5% mean positive population 7.8 SE, n ¼ 4 donors) are shown (Fig. 2). Different cell populations were sorted based on their expression (region R3) or lack of expression (region R2) of each marker. The sorted populations were plated into regular
Fig. 3. Isolation of ASCs by single-color FACS. Cell populations were sorted by FACS, based on their negative (region R2) or positive (region R3) reactivity to either STRO-1, CD146, and 3G5 as shown in Figure 2. The different cell fractions were then plated (1 T 105 cells per 10 cm2 dish) into regular growth medium in triplicate to assess the incidence of adherent colony-forming cells in each cell fraction. The bar graph depicts the number of clonogenic colonies retrieved following fluorescence activated cell sorting, based on their expression of either STRO-1, CD146, or 3G5. The data are expressed as the number of colony-forming units obtained per 105 cells plated in the positive and negative cell fractions averaged from three separate experiments.
growth medium to assess the colony-forming efficiency (Fig. 3). The incidence of clonogenic colonies was found to be highest in the STRO-1 and CD146 and 3G5 positive cell fractions, with 3G5-selection exhibiting the greatest enrichment for CFU-F (Fig. 3). Ex vivo expanded ASCs share a similar immunophenotype to BMSSC
Fig. 2. Expression patterns of STRO-1, CD146, and 3G5 by adipose. Representative histograms are shown, depicting the expression of STRO-1, CD146, and 3G5 in fresh preparations of human adiposederived single-cell suspensions generated following collagenase/ dispase digestion as previously described in Section Materials and Methods. The antibodies were identified using either a goat antimurine IgM- or IgG-phycoerythrin using flow cytometric analysis.
JOURNAL OF CELLULAR PHYSIOLOGY DOI 10.1002/JCP
Studies were designed to characterize the progeny of ex vivo expanded ASCs cells derived from 3G5þ immunoselected cells. Representative flow cytometric histograms are shown in Figure 4, and depict the phenotypic profile of ex vivo expanded 3G5þ-selected ASCs, based on their expression of a panel of cell surface markers associated with stromal cells (CD44, CD90, CD105, CD106, CD146, CD166, STRO-1, alkaline phosphatase), endothelial cells (CD31, CD105, CD106, CD146, CD166), haematopoietic cells (CD14, CD31, CD45), and perivascular cells (3G5, STRO-1, CD146). When placed in culture, the ASCs were found to downregulate their expression of 3G5 and acquire the expression of a range of stromal and vascular cell related markers. The cultured ASCs lacked expression of the haematopoietic stem cell/endothelial cell marker, CD31, the leukocyte marker CD45 and the monocyte/ macrophage marker CD14. Conversely, the minor population of colony forming 3G5 cells were found have a limited growth potential of less than twelve population doublings. Therefore, there were insufficient numbers of cells in the 3G5 population for performing in vitro and in vivo differentiation studies and flow cytometric analysis. Comparable staining patterns were observed in similar experiments performed with
PERIVASCULAR ORIGIN OF ADIPOSE-DERIVED STROMAL STEM CELLS
Fig. 4. Immunophenotypic analysis of adipose-derived MPC. Representative flow cytometric histograms depicting the expression of CD14, CD31,CD44, CD45, CD90, CD105, CD106, CD146, CD166, STRO-1, 3G5, alkaline phosphatase (thin solid line) and isotype control antibodies (thick solid line) in fresh preparations of peripheral adipose-derived single-cell suspensions generated following trypsin/EDTA digestion. The primary antibodies were identified using either a goat anti-murine IgM-fluorescein isothiocyanate or a goat anti-murine IgG-fluorescein isothiocyanate. The mean percent positive fluorescence values W SE are shown relative to the isotype control antibody as determined from four different donors.
JOURNAL OF CELLULAR PHYSIOLOGY DOI 10.1002/JCP
417
418
ZANNETTINO ET AL.
cells isolated with STRO-1 and CD146 from adipose tissue derived from three normal donors (data not shown). Multi-differentiation potential of ASCs in vitro and in vivo
Previous studies have demonstrated that ex vivo expanded ASCs isolated by plastic adherence have the potential to differentiate into multiple cell types (Gimble and Guilak, 2003). We therefore investigated whether the immunoselected ASCs populations maintained their characteristic multipotential properties when cultured under defined in vitro conditions. Monoclonal antibody 3G5þ-isolated ASCs were capable of forming Alizarin Red positive mineralized deposits, Oil Red O positive lipid droplets, and Alcian blue positive proteoglycan-rich matrix, when cultured under osteogenic, adipogenic, and chondrogenic inductive conditions in vitro, respectively (Fig. 5A–C). No measurable differences in the developmental potential of ex vivo expanded STRO-1, CD146, or 3G5 positive selected cells were observed (data not shown). The conversion of ASCs into functional osteoblast, adipocytes, and chondrocytes, correlated with the upregulation of markers associated with bone (CBFA-1, osterix, osteocalcin), fat (PPARg2, leptin/obese gene product, lipoprotein lipase), and cartilage (SOX9, collagen type II and X) development as assessed by RT-PCR analysis (Fig. 6). Studies demonstrating the multipotentiality of ASCs in vitro were also complemented with experiments that examined the potential of purified STRO-1, CD146, or 3G5 positive ASCs populations to develop ectopic bone in vivo. Following selection and ex vivo expansion, the ASCs were transplanted subcutaneously into NOD/SCID mice in combination with osteoconductive ceramic carrier particles comprised of hydroxyapatite-tricalcium phosphate. The implants were
harvested 8 weeks posttransplantation and prepared for histological examination. The data demonstrated the presence of new bone formation (Fig. 5D) with no measurable differences in ectopic bone formation between ASCs cultures established from STRO-1, 3G5, or CD146 selected cell populations. Furthermore, the different imunoselected ASC did not demonstrate the capacity to support local haematopoiesis in vivo for the different transplants examined.
Discussion
The present study demonstrates the existence of a multipotent ASCs population associated with the microvasculature of adult human adipose tissue extracted from extramedullary sites. Until now, efforts to characterize the nature and properties of adipose-derived MSC-like cells have been limited to cultures established following plastic adherence. In order to avoid the presence of various accessory cell types and mature stromal elements that can influence the growth, phenotype, and functional characteristics of ASCs, an immunoselection strategy was employed. The immunoselection of ASCs from freshly digested adipose tissue provides a more accurate assessment of the phenotypic profile of ASCs, where many cell markers are often differentially expressed following ex vivo expansion. Previous studies from our laboratory have made use of antibody reagents reactive to either STRO-1, CD146, or 3G5 to prospectively isolate and enrich for MSC-like populations from various postnatal human tissues including bone marrow, dental pulp and periodontal ligament (Gronthos et al., 2003; Shi and Gronthos, 2003; Seo et al., 2004). The present study builds on these findings and reiterates the utility of immunoselection as a reliable methodology for identifying and enriching different MSC-like populations.
Fig. 5. Developmental potential of purified ASCs in vitro and in vivo. Preparations of primary ASCs cultures derived from 3G5R adipose cells were cultured either in, osteogenic (A), adipogenic (B), or chondrogenic (C), inductive conditions as described in Section Materials and Methods. Following 2 weeks of multi-differentiation induction, the ASCs demonstrated the capacity to form bone (A; Alizarin positive mineral deposits), fat (B; Oil Red O positive lipid), and cartilage (C: collagen type II matrix) in vitro. Primary 3G5R derived ASCs cultures were transplanted with HA/TCP particles subcutaneously into NOD/SCID mice. A cross-section of a representative 8-week-old transplant is shown stained with H&E (D). The arrows indicate new bone formation has occurred at the surfaces of the HA/TCP carrier.
JOURNAL OF CELLULAR PHYSIOLOGY DOI 10.1002/JCP
PERIVASCULAR ORIGIN OF ADIPOSE-DERIVED STROMAL STEM CELLS
Fig. 6. Expression of lineage associated genes in vitro. Preparations of primary ASCs cultures derived from 3G5R adipose cells were recultured either in, (A) chondrogenic, (B) osteogenic, or (C) adipogenic inductive medium for 2 weeks. Total RNA was isolated from cells treated under the different conditions (white) and control cultures (black) grown in regular growth media. Semi-quantitative RT-PCR analysis was performed to detect RNA transcripts relative to GAPDH expression for: (A) cartilage markers, Sox9, collagen type II (COL II), collagen type X (COL X); (B) bone markers, CBFA-1, Osterix (OSX), osteocalcin (OCN); (C) adipose markers, PPARg2, leptin/obese gene product (OGP), lipoprotein lipase (LPL). The values represent the mean W standard error bars from three different experiments.
In situ localization studies demonstrated that the STRO-1 antigen preferentially identified perivascular cells within adipose tissue, whereas CD146 was expressed predominantly by endothelial cells. In contrast, the marker 3G5, was found to be restricted to areas within the perivascular regions of human adipose tissue. These findings are in accord with JOURNAL OF CELLULAR PHYSIOLOGY DOI 10.1002/JCP
previous reports that described the reactivity of the 3G5 antibody to an o-acetylated disialoganglioside epitope and identifies capillary perycites (Nayak et al., 1988; Andreeva et al., 1998). Moreover, the notion that ASCs are derived from a pericyte-like population, supports previous studies which showed that other MSC-like populations derived from postnatal human bone marrow, retina, skeletal muscle and dermis, dental pulp and periodontal ligament reside in a perivascular niche (Doherty et al., 1998; Young et al., 2001; Gronthos et al., 2003; Shi and Gronthos, 2003; Seo et al., 2004). Similarly, other investigators have hypothesized that these progenitors originate from perivascular cell populations which emigrate from the capillary walls into the surrounding fibrous tissues for the purpose of tissue regeneration in both embryonic and postnatal development (Brighton et al., 1992; Canfield et al., 2000; Bianco et al., 2001; Tavian et al., 2005). Overall, these data demonstrate that ASCs exhibit a phenotype consistent with that previously described for pericytes. While the exact developmental relationship between pericytes, smooth muscle cells and endothelial cells remains to be determined, it has been proposed that pericytes may represent a precursor pool for generating endothelial and/or smooth muscle cells (Nehls et al., 1992; Schor et al., 1995; Schor and Canfield, 1998). Moreover, accumulating evidence suggests that, in addition to participating in the maintenance of blood vessel wall integrity, a subset of pericytes may also represent multipotential MSCs (Diaz-Flores et al., 1991; Brighton et al., 1992; Nehls et al., 1992; Galmiche et al., 1993; Nehls and Drenckhahn, 1993; Andreeva et al., 1998; Doherty et al., 1998; Zuk et al., 2001; Shi and Gronthos, 2003). The most compelling evidence for this, comes from studies of cultured bovine retinal pericytes that were found to express both the STRO-1 and 3G5 markers and exhibited properties characteristic of multipotential MSCs (Doherty et al., 1998; Canfield et al., 2000). However, while the majority of ASCs continued to express CD146 and STRO-1 following ex vivo expansion, the 3G5 antigen was observed to be rapidly downregulated by the majority of ASCs in vitro. This loss of 3G5 expression, which is consistently observed for both BMSSC and DPSC following cell proliferation and differentiation in vitro, is thought to reflect the increasing maturation of cultured MSC over time when grown in the presence of high levels of serum (Gronthos et al., 1999; Stewart et al., 1999; Ahdjoudj et al., 2001). A major issue concerning the properties of different immunoselected cell populations is the possibility that these procedures are selecting for different progenitor pools that may or may not be derived from a common mesenchymal precursor cell population. To help address this issue, we performed a comprehensive flow cytometric analysis of cultured STRO-1, CD146, and 3G5-immunoselected ASCs. The data revealed an immunophenotype (STRO-1þ/CD44þ/ CD90þ/CD105þ/CD146þ/CD166þ/CD14/CD31/CD45) consistent with that previously described for different MSC-like populations following ex vivo expansion. While our previous analysis of plastic adherent, culture expanded human ASCs failed to detect expression of the STRO-1 antigen (Gronthos et al., 2001), this may have resulted from our reliance on immunohistochemical methods and the use of unselected crude preparations of ASCs. In contrast with our previous findings and consistent with the current work, Zuk et al. (2002) identified ASCs as STRO-1 positive based on flow cytometry. In addition, irrespective of the antibody used to select the ASCs, we could not find any discernible difference in the expression of this panel of cell surface markers. However, while several markers were observed to be highly expressed by all of the culture expanded ASCs (CD44, CD90, CD105, CD146, CD166), it should be noted that antibodies to these cell surface
419
420
ZANNETTINO ET AL.
molecules are unable to be used as individual reagents to discriminate between multi-, bi- and uni-potential mesenchymal cell populations. Therefore, the current lack of specific mesenchymal cell markers is one major limitation for investigating the exact physical location of this stem cell pool in adipose or indeed other connective tissues. To overcome the paucity of specific markers to MSC, our study demonstrated the use of three vascular/MCS-associated markers to distinguish between different stromal and vascular cell populations in adipose tissue. Importantly, unlike primary cultures derived by plastic adherence, the ASCs selected by either STRO-1, CD146, or 3G5 lacked expression of the CD14, CD31, and CD45, markers commonly associated with cells of macrophage, endothelial cells and leukocyte lineages, respectively. Further characterization of the different immunoselected ASCs populations showed that cultured expanded ASCs established following either STRO-1, CD146, or 3G5 selection appeared to possess a similar capacity to differentiate into functionally active adipocytes, osteoblasts, and chondrocytes in vitro. The development of the different selected ASCs was also correlated with an upregulation of genes sets with important roles in the development and function of bone (Cbfa1, Osterix, osteocalcin), fat (PPARg2, leptin, lipoprotein lipase) and cartilage (Sox9, collagen type II and X) tissues. Moreover, all three immunoselected ASCs populations demonstrated the capacity to generate mature lamella bone in vivo when transplanted into immunocompromised mice in the presence of osteo-conductive HA/TCP carrier particles. However, unlike their bone marrow counterparts, ASCs demonstrated a poor capacity to support haematopoiesis in vivo. This in accord with noted differences in the proliferation, phenotype, and developmental potential of different tissue specific stem/progenitor cell populations that share similar, but non-identical characteristics with bone marrow derived MSC. Nevertheless, the present study clearly shows that multi-potential ASCs arise from the perivascular regions within human adipose tissue, analogous to what has previously been described for BMSSC and DPSC (Shi and Gronthos, 2003). Interestingly, a common feature between different MSC-like populations is that only a small proportion of clonal derived lines exhibit the hallmarks of a ‘‘true stem cell’’ population (Kuznetsov et al., 1997; Gronthos et al., 2002, 2003; Zuk et al., 2002; Guilak et al., 2006). Further characterization of multipotential MSC-like cells derived from different tissues, could lead to the development of novel reagents or selection criteria that may help us discriminate between committed progenitor cells and multipotential stem cell populations with the capacity for self renewal. However, despite the developmental similarities between ASCs and BMSSC, a recent study by Rubio et al. (2005) reported that human ASCs have the potential to undergo spontaneous transformation following extensive ex vivo expansion and subculture. In contrast, the transformation potential of human BMSSC has only been reported in clonal cell lines derived from genetically modified BMSSC (Burns et al., 2005). These studies highlight the necessity for more rigorous assessment of the nature of different MSC-like populations intended for cellular therapies. In the present study, the demonstration that ASCs are derived from a perivascular stem cell niche, suggests that they share a similar microenvironmental ‘‘cradle’’ that harbors and regulates different MSC-like populations derived from bone marrow, adipose tissue, dental pulp, and periodontal ligament. It is anticipated that continued studies in this area will help elucidate the fundamental conditions necessary to facilitate the development of viable and safer tissue engineering applications brought about by an increased understanding of the precise factors or conditions that regulate the growth and development of ASCs. JOURNAL OF CELLULAR PHYSIOLOGY DOI 10.1002/JCP
Acknowledgments
We would like to thank Dr. John Greenwood, and Ms. Amy Li (Burns Unit, Royal Adelaide Hospital, Adelaide South Australia) for kindly providing us with the adipose tissue specimens. This study was supported largely by the Australian National Health and Research Council Project Grant (242804).
Literature Cited Ahdjoudj S, Lasmoles F, Oyajobi BO, Lomri A, Delannoy P, Marie PJ. 2001. Reciprocal control of osteoblast/chondroblast and osteoblast/adipocyte differentiation of multipotential clonal human marrow stromal F/STRO- 1(þ) cells. J Cell Biochem 81:23–38. Airey JA, Almeida-Porada G, Colletti EJ, Porada CD, Chamberlain J, Movsesian M, Sutko JL, Zanjani ED. 2004. Human mesenchymal stem cells form Purkinje fibers in fetal sheep heart. Circulation 109:1401–1407. Andreeva ER, Pugach IM, Gordon D, Orekhov AN. 1998. Continuous subendothelial network formed by pericyte-like cells in human vascular bed. Tissue Cell 30:127–135. Azizi SA, Stokes D, Augelli BJ, DiGirolamo C, Prockop DJ. 1998. Engraftment and migration of human bone marrow stromal cells implanted in the brains of albino rats—Similarities to astrocyte grafts. Proc Natl Acad Sci U S A 95:3908–3913. Bianco P, Riminucci M, Gronthos S, Robey PG. 2001. Bone marrow stromal stem cells: Nature, biology, and potential applications. Stem Cells 19:180–192. Brighton CT, Lorich DG, Kupcha R, Reilly TM, Jones AR, Woodbury RA. II. 1992. The pericyte as a possible osteoblast progenitor cell. Clin Orthop Relat Res 275:287–299. Burns JS, Abdallah BM, Guldberg P, Rygaard J, Schroder HD, Kassem M. 2005. Tumorigenic heterogeneity in cancer stem cells evolved from long-term cultures of telomeraseimmortalized human mesenchymal stem cells. Cancer Res 65:3126–3135. Canfield AE, Doherty MJ, Wood AC, Farrington C, Ashton B, Begum N, Harvey B, Poole A, Grant ME, Boot-Handford RP. 2000. Role of pericytes in vascular calcification: A review. Z Kardiol 89:20–27. Schor AM, Canfield AE. 1998. Osteogenic potential of vascular pericytes. In: Beresford JN, Owen ME, editor. Marrow stromal cell culture. Cambridge, UK: Cambridge University Press. p 128–148. Carlile MJ, Sturrock MG, Chisholm DM, Ogden GR, Schor AM. 2000. The presence of pericytes and transitional cells in the vasculature of the human dental pulp: An ultrastructural study. Histochem J 32:239–245. Diaz-Flores L, Gutierrez R, Varela H, Rancel N, Valladares F. 1991. Microvascular pericytes: A review of their morphological and functional characteristics. Histol Histopathol 6:269– 286. Doherty MJ, Ashton BA, Walsh S, Beresford JN, Grant ME, Canfield AE. 1998. Vascular pericytes express osteogenic potential in vitro and in vivo. J Bone Miner Res 13:828–838. Friedenstein AJ. 1980. Stromal mechanisms of bone marrow: Cloning in vitro and retransplantation in vivo. Hamatol Bluttransfus 25:19–29. Friedenstein AJ, Chailakhjan RK, Lalykina KS. 1970. The development of fibroblast colonies in monolayer cultures of guinea- pig bone marrow and spleen cells. Cell Tissue Kinet 3:393– 403. Friedenstein AJ, Gorskaja JF, Kulagina NN. 1976. Fibroblast precursors in normal and irradiated mouse hematopoietic organs. Exp Hematol 4:267–274. Friedenstein AJ, Ivanov-Smolenski AA, Chajlakjan RK, Gorskaya UF, Kuralesova AI, Latzinik NW, Gerasimow UW. 1978. Origin of bone marrow stromal mechanocytes in radiochimeras and heterotopic transplants. Exp Hematol 6:440–444. Galmiche MC, Koteliansky VE, Briere J, Herve P, Charbord P. 1993. Stromal cells from human long-term marrow cultures are mesenchymal cells that differentiate following a vascular smooth muscle differentiation pathway. Blood 82:66–76. Gimble JM, Guilak F. 2003. Differentiation potential of adipose derived adult stem (ADAS) cells. Curr Top Dev Biol 58:137–160. Gronthos S, Brahim J, Li W, Fisher LW, Cherman N, Boyde A, DenBesten P, Robey PG, Shi S. 2002. Stem cell properties of human dental pulp stem cells. J Dent Res 81:531–535. Gronthos S, Franklin DM, Leddy HA, Robey PG, Storms RW, Gimble JM. 2001. Surface protein characterization of human adipose tissue-derived stromal cells. J Cell Physiol 189:54–63. Gronthos S, Graves SE, Ohta S, Simmons PJ. 1994. The STRO-1þ fraction of adult human bone marrow contains the osteogenic precursors. Blood 84:4164–4173. Gronthos S, Mankani M, Brahim J, Robey PG, Shi S. 2000. Postnatal human dental pulp stem cells (DPSCs) in vitro and invivo. Proc Natl Acad Sci U S A 97:13625–13630. Gronthos S, Zannettino AC, Graves SE, Ohta S, Hay SJ, Simmons PJ. 1999. Differential cell surface expression of the STRO-1 and alkaline phosphatase antigens on discrete developmental stages in primary cultures of human bone cells. J Bone Miner Res 14:47–56. Gronthos S, Zannettino AC, Hay SJ, Shi S, Graves SE, Kortesidis A, Simmons PJ. 2003. Molecular and cellular characterisation of highly purified stromal stem cells derived from human bone marrow. J Cell Sci 116:1827–1835. Guilak F, Lott KE, Awad HA, Cao Q, Hicok KC, Fermor B, Gimble JM. 2006. Clonal analysis of the differentiation potential of human adipose-derived adult stem cells. J Cell Physiol 206:229–237. Helmbold P, Nayak RC, Marsch WC, Herman IM. 2001. Isolation and in vitro characterization of human dermal microvascular pericytes. Microvasc Res 61:160–165. Kuznetsov SA, Krebsbach PH, Satomura K, Kerr J, Riminucci M, Benayahu D, Robey PG. 1997. Single-colony derived strains of human marrow stromal fibroblasts form bone after transplantation in vivo. J Bone Miner Res 12:1335–1347. Kuznetsov SA, Mankani MH, Gronthos S, Satomura K, Bianco P, Robey PG. 2001. Circulating skeletal stem cells. J Cell Biol 153:1133–1140. Liechty KW, MacKenzie TC, Shaaban AF, Radu A, Moseley AM, Deans R, Marshak DR, Flake AW. 2000. Human mesenchymal stem cells engraft and demonstrate site-specific differentiation after in utero transplantation in sheep. Nat Med 6:1282–1286. Martens TP, See F, Schuster MD, Sondermeijer HP, Hefti MM, Zannettino A, Gronthos S, Seki T, Itescu S. 2006. Mesenchymal lineage precursor cells induce vascular network formation in ischemic myocardium. Nat Clin Pract Cardiovasc Med 3:S18–22. Miura M, Gronthos S, Zhao M, Lu B, Fisher LW, Robey PG, Shi S. 2003. SHED: Stem cells from human exfoliated deciduous teeth. Proc Natl Acad Sci U S A 100:5807–5812. Nathan S, Das De S, Thambyah A, Fen C, Goh J, Lee EH. 2003. Cell-based therapy in the repair of osteochondral defects: A novel use for adipose tissue. Tissue Eng 9:733–744.
PERIVASCULAR ORIGIN OF ADIPOSE-DERIVED STROMAL STEM CELLS
Nayak RC, Berman AB, George KL, Eisenbarth GS, King GL. 1988. A monoclonal antibody (3G5)-defined ganglioside antigen is expressed on the cell surface of microvascular pericytes. J Exp Med 167:1003–1015. Nehls V, Denzer K, Drenckhahn D. 1992. Pericyte involvement in capillary sprouting during angiogenesis in situ. Cell Tissue Res 270:469–474. Nehls V, Drenckhahn D. 1993. The versatility of microvascular pericytes: From mesenchyme to smooth muscle? Histochemistry 99:1–12. Owen M, Friedenstein AJ. 1988. Stromal stem cells: Marrow-derived osteogenic precursors. Ciba Found Symp 136:42–60. Pittenger MF, Mackay AM, Beck SC, Jaiswal RK, Douglas R, Mosca JD, Moorman MA, Simonetti DW, Craig S, Marshak DR. 1999. Multilineage potential of adult human mesenchymal stem cells. Science 284:143–147. Rubio D, Garcia-Castro J, Martin MC, de la Fuente R, Cigudosa JC, Lloyd AC, Bernad A. 2005. Spontaneous human adult stem cell transformation. Cancer Res 65:3035–3039. Safford KM, Safford SD, Gimble JM, Shetty AK, Rice HE. 2004. Characterization of neuronal/ glial differentiation of murine adipose-derived adult stromal cells. Exp Neurol 187:319– 328. Sakaguchi Y, Sekiya I, Yagishita K, Muneta T. 2005. Comparison of human stem cells derived from various mesenchymal tissues: Superiority of synovium as a cell source. Arthritis Rheum 52:2521–2529. Schor AM, Canfield AE, Sutton AB, Arciniegas E, Allen TD. 1995. Pericyte differentiation. Clin Orthop Relat Res 313:81–91. Seo BM, Miura M, Gronthos S, Bartold PM, Batouli S, Brahim J, Young M, Robey PG, Wang CY, Shi S. 2004. Investigation of multipotent postnatal stem cells from human periodontal ligament. Lancet 364:149–155. Shi S, Gronthos S. 2003. Perivascular niche of postnatal mesenchymal stem cells in human bone marrow and dental pulp. J Bone Miner Res 18:696–704. Simmons PJ, Torok-Storb B. 1991. Identification of stromal cell precursors in human bone marrow by a novel monoclonal antibody, STRO-1. Blood 78:55–62.
JOURNAL OF CELLULAR PHYSIOLOGY DOI 10.1002/JCP
Stewart K, Walsh S, Screen J, Jefferiss CM, Chainey J, Jordan GR, Beresford JN. 1999. Further characterization of cells expressing STRO-1 in cultures of adult human bone marrow stromal cells. J Bone Miner Res 14:1345–1356. Strem BM, Hicok KC, Zhu M, Wulur I, Alfonso Z, Schreiber RE, Fraser JK, Hedrick MH. 2005. Multipotential differentiation of adipose tissue-derived stem cells. Keio J Med 54:132–141. Tavian M, Zheng B, Oberlin E, Crisan M, Sun B, Huard J, Peault B. 2005. The vascular wall as a source of stem cells. Ann N Y Acad Sci 1044:41–50. Toma C, Pittenger MF, Cahill KS, Byrne BJ, Kessler PD. 2002. Human mesenchymal stem cells differentiate to a cardiomyocyte phenotype in the adult murine heart. Circulation 105:93– 98. Tuli R, Tuli S, Nandi S, Wang ML, Alexander PG, Haleem-Smith H, Hozack WJ, Manner PA, Danielson KG, Tuan RS. 2003. Characterization of multipotential mesenchymal progenitor cells derived from human trabecular bone. Stem Cells 21:681–693. Wickham MQ, Erickson GR, Gimble JM, Vail TP, Guilak F. 2003. Multipotent stromal cells derived from the infrapatellar fat pad of the knee. Clin Orthop Relat Res 412:196–212. Yoo JU, Barthel TS, Nishimura K, Solchaga L, Caplan AI, Goldberg VM, Johnstone B. 1998. The chondrogenic potential of human bone-marrow-derived mesenchymal progenitor cells. J Bone Joint Surg Am 80:1745–1757. Young HE, Steele TA, Bray RA, Hudson J, Floyd JA, Hawkins K, Thomas K, Austin T, Edwards C, Cuzzourt J, Duenzl M, Lucas PA, Black AC, Jr. 2001. Human reserve pluripotent mesenchymal stem cells are present in the connective tissues of skeletal muscle and dermis derived from fetal, adult, and geriatric donors. Anat Rec 264:51–62. Zuk PA, Zhu M, Ashjian P, De Ugarte DA, Huang JI, Mizuno H, Alfonso ZC, Fraser JK, Benhaim P, Hedrick MH. 2002. Human adipose tissue is a source of multipotent stem cells. Mol Biol Cell 13:4279–4295. Zuk PA, Zhu M, Mizuno H, Huang J, Futrell JW, Katz AJ, Benhaim P, Lorenz HP, Hedrick MH. 2001. Multilineage cells from human adipose tissue: Implications for cell-based therapies. Tissue Eng 7:211–228.
421
