Stem Cell Reviews and Reports

, Volume 8, Issue 4, pp 1211–1222

Effect of Anatomical Origin and Cell Passage Number on the Stemness and Osteogenic Differentiation Potential of Canine Adipose-Derived Stem Cells


  • J. F. Requicha
    • 3B’s Research Group—Biomaterials, Biodegradables and Biomimetics, Department of Polymer EngineeringUniversity of Minho
    • Department of Veterinary Sciences, School of Agrarian and Veterinary SciencesUniversity of Trás-os-Montes e Alto Douro
    • ICVS/3B’s—PT Government Associated Laboratory
  • C. A. Viegas
    • 3B’s Research Group—Biomaterials, Biodegradables and Biomimetics, Department of Polymer EngineeringUniversity of Minho
    • Department of Veterinary Sciences, School of Agrarian and Veterinary SciencesUniversity of Trás-os-Montes e Alto Douro
    • ICVS/3B’s—PT Government Associated Laboratory
  • C. M. Albuquerque
    • Department of Veterinary Sciences, School of Agrarian and Veterinary SciencesUniversity of Trás-os-Montes e Alto Douro
  • J. M. Azevedo
    • CECAV—Centre for Studies in Animal and Veterinary Sciences, Department of Animal Sciences, School of Agrarian and Veterinary SciencesUniversity of Trás-os-Montes e Alto Douro
  • R. L. Reis
    • 3B’s Research Group—Biomaterials, Biodegradables and Biomimetics, Department of Polymer EngineeringUniversity of Minho
    • ICVS/3B’s—PT Government Associated Laboratory
    • 3B’s Research Group—Biomaterials, Biodegradables and Biomimetics, Department of Polymer EngineeringUniversity of Minho
    • ICVS/3B’s—PT Government Associated Laboratory

DOI: 10.1007/s12015-012-9397-0

Cite this article as:
Requicha, J.F., Viegas, C.A., Albuquerque, C.M. et al. Stem Cell Rev and Rep (2012) 8: 1211. doi:10.1007/s12015-012-9397-0


Mesenchymal stem cells have a great potential for application in cell based therapies, such as tissue engineering. Adipose derived stem cells have shown the capacity to differentiate into several lineages, and have been isolated in many animal species. Dog is a very relevant animal model to study several human diseases and simultaneously an important subject in veterinary medicine. Thus, in this study we assessed the potential of canine adipose tissue derived stem cells (cASCs) to differentiate into the osteogenic and chondrogenic lineages by performing specific histological stainings, and studied the cell passaging effect on the cASCs stemness and osteogenic potential. We also evaluated the effect of the anatomical origin of the adipose tissue, namely from abdominal subcutaneous layer and from greater omentum. The stemness and osteogenic differentiation was followed by real time RT-PCR analysis of typical markers of mesenchymal stem cells (MSCs) and osteoblasts. The results obtained revealed that cASCs exhibit a progressively decreased expression of the MSCs markers along passages and also a decreased osteogenic differentiation potential. In the author’s knowledge, this work presents the first data about the MSCs markers profile and osteogenic potential of cASCs along cellular expansion. Moreover, the obtained data showed that the anatomical origin of the adipose tissue has an evident effect in the differentiation potential of the ASCs. Due to the observed resemblances with the human ASCs, we conclude that canine ASCs can be used as a model cells in tissue engineering research envisioning human applications.


Adipose-derived stem cellsCell based therapiesTissue engineeringCanine modelStemnessOsteogenic potential


Stem cells play a vital role in regenerative medicine approaches due to their self-renew capacity and differentiation potential when cultured under specific biochemical and/or mechanical stimulus [13]. Several adult tissues derived from mesodermal layer of the embryo are known sources of mesenchymal stem cells (MSCs), such as, the bone marrow [1], adipose tissue [24], umbilical cord blood [5], muscle tissue [6], neuronal tissue [7], bone [8], periosteum [9], periodontal ligament [10], dental pulp [11], exfoliated deciduous teeth [12], dental follicle [13] and pancreatic or hepatic tissues [14].

Human adipose-derived stem cells (hASCs) were firstly isolated by Zuk and colleagues in 2001 [4] and have gained importance in the last years, mainly due to the wide availability of the adipose tissue [15, 16], enabling to collect larger quantities and due to the easiness of its harvesting procedures, minimizing site morbidity. ASCs can be isolated by enzymatic digestion [16] and, in culture, they are characterized by their adherence to the plastic substrate of culture plates, in opposition to hematopoietic stem cells [2, 17]. Furthermore, ASCs are able to differentiate in different cell types [3, 15, 18], such as, osteoblasts [4, 15, 19, 20], chondroblasts [3, 4, 15, 20], adipocytes [4, 15, 20, 21], skeletal, smooth and cardiac myocytes [4, 20, 22], neuronal cells [20, 2326], tendon cells [27] and intervertebral disc cells [28].

This source of multipotent cells has been mostly studied in humans, but also in other species, such as mouse [29], rat [30], rabbit [31], sheep [32], horse [27, 33], non-human primate [25, 34] and dog [24, 3537].

Canine ASCs have been isolated from visceral [24, 35, 38] and subcutaneous [35, 36, 3840] adipose tissue and it has been proved their potential to differentiated, in vitro, into the osteogenic [35], chondrogenic [35], adipocytic [35, 38], neuronal [24] and myogenic [37] lineages. Canine animal models are very important to study a variety of human diseases; simultaneously, dog is a very important subject in veterinary medicine, therefore it is essential to know more about canine ASCs. These cells can also be used in autologous tissue engineering approaches, as it has already been reported in previous works related to the regeneration of periodontal defects [36], cranial defects [41] and in osteoarthritis clinical trials [39], in all cases demonstrating promising results. In fact, stem cells based therapies have gained increasing importance in veterinary medicine and have already went into the clinical routine in some countries.

The aim of this study was to compare canine adipose-derived stem cells derived from different anatomical sites, namely, subcutaneous and omental fat depots, and to characterize their stemness along its proliferative process by analyzing the expression of typical MSC genes at different passages. The multipotency of these cells was assessed by inducing their differentiation into different cellular lineages, including osteogenic and chondrogenic and analysing the production of specific extra cellular matrix, as well as the expression of specific osteoblastic genes.

The results obtained provide essential information for the future application of canine ASCs in new stem cells based therapies in veterinary medicine, as well as for the use of the dog as a model [42] for developing new human regenerative medicine approaches.

Materials and Methods

Harvesting of Canine Adipose Tissue

Adipose tissue was collected from 6 adult dogs (3 males and 3 females), between 1 and 5 years of age, subject of convenience surgery in the Veterinary Hospital of University of Trás-os-Montes e Alto Douro in accordance with Portuguese legislation (Portaria no1005/92) and international standards on animal welfare as defined by the European Communities Council Directive (86/609/EEC), and with previous informed consent of the owners.

The dogs were sedated by an intramuscular administration of 0.2 mg/kg IM butorphanol (Torbugesic 1 %; Fort Dodge, the Netherlands) and 30 μg/kg IM acepromazine (Vetranquil; CEVA Sante Animal, France). Anesthesia was achieved by an intravenous administration of 0.25 mg/kg IV diazepam (Diazepam MG; Labesfal, Portugal), 4 mg/kg IV ketamine (Imalgene 1000; Merial, France) and 4 mg/kg IV propofol (Lipuro 2 %; Braun, Germany) and was maintained using 1 % isoflurane (IsoFlo; Abbott Animal Health, USA), administered in oxygen through an endotracheal tube.

After anesthetized, the animals were placed in dorsal recumbence and it was made a careful tricothomy and asepsis of the abdominal region. A small surgical incision was made through the skin, subcutaneous tissue and muscular wall of the abdominal region.

In all the animals of the study, a small amount of abdominal subcutaneous and omental adipose tissue was resected using scalpel and surgical scissors and placed into labeled sterile containers with PBS (Phosphate Buffer Solution; Sigma Aldrich, Germany) with 10 % antibiotic/antimycotic (100 IU/ml penicillin and 100 IU streptomycin; Fluka, UK). Samples were refrigerated at 4 °C and transported to the laboratory for further processing.

Isolation and Expansion of cASCs

All the samples were processed within 12 h upon harvesting. Adipose tissue samples were washed with PBS containing 10 % antibiotics/antimycotic and minced into small fragments using a scissor. The fragments were digested with a solution of 0.1 % collagenase type IA (Sigma Aldrich, Germany) in PBS at 37 °C under shacking at 200 rpm for 40 min (10 ml/10 cm3 Tissue). Collagenase was inactivated with an equal volume of culture medium containing 10 % FBS (Foetal Bovine Serum; Invitrogen, USA).

The digested tissue was filtered with a 100 μm nylon meshe and centrifuged at 1,250 rpm for 5 min at 20 °C and the supernatant removed. The obtained cells were resuspended in culture medium (basal medium, BM) composed of DMEM (Dulbecco’s Minimum Essential Medium Eagle; Sigma Aldrich, Germany), 10 % FBS, sodium bicarbonate (Sigma Aldrich, Germany) and 1 % antibiotic/antimycotic and then seeded in 25 cm2 culture flasks. After 48 h, the cells were rinsed with PBS and the medium was changed to remove non-adherent cells, such as hematopoietic cells or dead cells, as performed in other studies [32].

cASCs Expansion and Differentiation

The cells were cultured in plastic adherent culture flasks, using basal medium, which was changed three times a week until reaching a confluence of around 80 %, along four passages (P1, P2, P3 and P4).

Osteogenic Differentiation

Canine ASCs, at P1, P2, P3 and P4, were seeded at a density of 5 × 104 cells per culture flask (75 cm2 T-flask), and further cultured for 14, 21, 28 and 35 days in osteogenic medium (OM) composed of αMEM (Alpha Minimum Essential Medium Eagle; Sigma Aldrich, Germany) supplemented with 10 % FBS, sodium bicarbonate, 1 % antibiotic/antimycotic, 50 ug/ml ascorbic acid (Sigma, USA), 10−8 M dexamethasone (Sigma, USA) and 10 mM β-glycerophosphate (Sigma, USA); culture medium was changed twice a week.

Chondrogenic Differentiation

Canine ASCs (about 5 × 105 cells), at P1, P2, P3 and P4, were resuspended in chondrogenic medium and centrifuged at 800 G for 5 min, in order to obtain cell pellets. Pellets were cultured in the same medium for up to 21 days at 37 °C and 5 % CO2, in 15 ml falcon tubes. Chondrogenic medium was composed of DMEM supplemented with 5 % FBS, sodium bicarbonate, 1 % antibiotic/antimycotic, 1 mM dexamethasone, 17 mM ascorbic acid, 10 ng/ml TGF-β1 (Human Transforming Growth Factor Beta 1; eBioscience, USA), 10 % Insulin-Transferrin-Selenium (ITS) (Sigma, USA), 35 mM L-proline (Sigma, USA) and 0.1 M sodium pyruvate (Sigma, USA) and was changed twice a week.

Canine ASCs Characterization

Gene Expression Analysis of MSCs Markers

Real time RT-PCR analysis was used to characterize the phenotype of cASCs, obtained from tissue collected from different anatomical sites (subcutaneous and omental) at different passages, namely, through the relative expression of CD73, CD90 and CD105 genes, which are the widely accepted markers of the mesenchymal stem cells[31]. The fact that is difficult to find in the market specific antibodies for cells of canine origin was the main reason to select real time RT-PCR analysis and not flow cytometry or other immune-based assays, for the assessment of the MSCs and osteoblasts markers

For this purpose, cASCs were detached using trypsin, and samples consisting of 1 × 106 cells at P0, P1, P2, P3 and P4 cultured in basal media, were retrieved and kept in 800 μl of TRIzol Reagent (Invitrogen, USA). The mRNA was extracted with TRIzol following the procedure provide by the supplier. Briefly, after an incubation of 5 min, additional 160 μl of chloroform (Sigma Aldrich, Germany) were added; the samples were then incubated for 15 min at 4 °C and centrifuged at the same temperature and 13,000 rpm for 15 min. After the centrifugation, the aqueous part was collected and an equal part of isopropanol (VWR, USA) was added. After an incubation of 2 h at −20 °C, the samples were washed in ethanol, centrifuged at 4 °C and 9,000 rpm for 5 min and resuspended in 12 μl of RNase/DNase free water (Gibco, UK). The samples were quantified using a NanoDrop ND1000 Spectrophotometer (NanoDrop Technologies, USA). For the cDNA synthesis, it were used the samples with a 260/280 ratio between 1.7 and 2.0. The cDNA synthesis was performed in the Mastercycler real time PCR equipment (Eppendorf, USA) using the iScript cDNA Synthesis Kit (Quanta Biosciences, USA) and an initial amount of mRNA of 2 μg and a total volume of 20 μl RNAse free water (Gibco, UK) was used as a negative control.

After the synthesis of the cDNA, real time PCR analysis was carried in the Mastercycler real time PCR equipment (Eppendorf, USA) using the PerfeCta Sybr-Green FastMix (Quanta Biosciences, USA) to analyze the relative expression of the genes CD73, CD90 and CD105 in each sample, using GAPDH as housekeeping gene. The primers were previously designed using the Primer 3 Plus v0.4.0 (MWG Biotech, Germany) (see Table 1).
Table 1

Primer sequences for targeted cDNAs



Product length (bp)

5′-3 sequence (F, forward; R, reverse)




































Delta Delta Ct method, according to Livak and Schmittgen (2011) [43], was performed, using the results corresponding to cells at P0 as calibrator of data obtained from cell at P1, P2, P3 and P4.

Canine ASCs Osteogenic Potential Assessment

Alizarin red Staining

Cells cultured in osteogenic medium for 14, 21, 28 and 35 days, were fixed with 4 % formalin (Sigma Aldrich, Germany) and washed with PBS and then with distilled water. Afterwards, the cells were stained with a 2 % Alizarin Red solution (Merk, Germany) in distilled water with a pH of 4.1–4.3, for 10 min, and finally washed with distilled water. Stained cells were observed under an optical microscope (Axivert 40 CFL; Zeiss; Germany) and photographed using an Axio Cam MRm (Zeiss, Germany) camera.

Gene Expression Analysis of Specific Osteoblasts Markers

Cells at different passages (P1-4) and further cultured for up 14, 21, 28 and 35 days in osteogenic medium were trypsinized and samples (of 1 × 106 cells) were kept in 800 μl of TRIzol. After extraction of the mRNA, the samples were quantified using the NanoDrop ND1000 Spectrophotometer. The cDNA synthesis was performed in the Mastercycler real time PCR equipment using the iScript cDNA Synthesis Kit. Real time PCR analysis was carried in the Mastercycler real time PCR equipment using the PerfeCta Sybr-Green FastMix to assess the relative expression of collagen I alpha 1 (COLIA1), runt-related transcription factor 2 (RUNX2) and Osteocalcin genes, using GAPDH as housekeeping gene (see Table 1).

Delta Delta Ct method was performed, using the data corresponding to undifferentiated cells (cultured in basal medium) at the respective passage (P1-4) as calibrator of the results obtained for cells after 14, 21, 28 and 35 days in osteogenic medium [43].

Canine ASCs Chondrogenic Potential Assessment

After 21 days of culture in chondrogenic medium, cASCs pellets were fixed with 4 % formalin, washed with PBS, embedded in paraffin and cut into sections of 3 μm; finally section were fixed on a slide and deparaffinised.

In order to observe cells and matrix formed, sections were stained with hematoxylin & eosin (H&E; Sigma, USA). Cartilage-like matrix deposition (mucopolysaccharides and glycosaminoglycans) was assessed by staining the samples sections with 1 % Toluidine Blue (Sigma, USA), 0.1 % Safranin O (Sigma, USA) and 1 % Alcian Blue (Sigma, USA).

In all cases, the stained sections were subsequently rinsed with distilled water and dehydrated using 100 % ethylene alcohol and allowed to dry overnight. Afterwards, the slides were observed under an optical microscope (Axivert 40 CFL; Zeiss; Germany), and photographed using an Axio Cam MRm (Zeiss, Germany) camera.

Statistical Analysis

The results obtained from the real time RT PCR analysis were elaborated using the software JMP® v9.0.1 (SAS Institute Inc., USA) and the correlation between the dataset from the same gene analyzed was investigated with the ANOVA single factor method. Data was presented as mean ± standard error of the mean (SEM) and differences between data groups were considered to be statistically significant for p < 0.05 (Student’s t-test).


Canine adipose tissue derived stem cells were successfully isolated from adipose tissue harvested from subcutaneous abdominal region, as well as from the omental fat depots. After isolation through enzymatic digestion, the obtained stromal vascular fraction adhered to the plastic culture flaks, and revealed a fibroblastic morphology in a monolayer growth (Fig. 1). The morphology of these cells does not seem to be affected by the number of passages, maintaining their characteristic spindle like shape and showing signals of cell viability and integrity along all studied passages.
Fig. 1

Representative light microscopy images of cASCs obtained by enzymatic digestion from different anatomical sites, namely subcutaneous (a) and omental (b), after 48 h of culturing in basal medium

Effect of the Passage Number in the Undifferentiated Canine ASCs

The isolated cells were subsequently cultured along four passages in basal medium without differentiation factors, and characterized through the expression of typical MSCs markers, namely CD73 (NT5E), CD90 (Thy1) and CD105 (Endoglin) [17], assessed by real time RT-PCR analysis. In general, the results revealed a decrease in the expression levels of these markers in the canine ASCs cultured in basal medium along passages, from P1 until P4 (Fig. 2).
Fig. 2

Real time RT-PCR analysis results of the various genes, namely a CD73, b CD90 and c CD105 in canine ASCs isolated by enzymatic digestion and cultured in basal medium along four passages (mean ± SEM). Considering each marker independently, levels not connected by the same letter are significantly different (p < 0.05, Student’s t-test). Table shown in d refers to the means. Abbreviations: P passage; Sc subcutaneous; Om omental

The registered down-regulation of the above mentioned MSCs markers was found always statistically significant from P1 to P2 and from P2 to P3, being more evident in the CD90 and CD105 genes, independently of the anatomical origin. Changes on the expression fold were not observed between P3 and P4.

Effect of the Anatomical Site in the Undifferentiated Canine ASCs

Regarding the influence of the anatomical site of origin on the cASCs stemness, in general, it was found a higher expression of the three markers analyzed in the cASCs obtained from omental fat depots. No differences were observed in the CD73 expression on cASCs from both anatomical origins. Significant differences were registered in the CD90 and CD105 expression at P1. As this work did not concern the evaluation of the proliferative capacity of the cASCs, no data regarding cell or colonies count were obtained; nevertheless, no evident microscopic differences in terms of cellular proliferation were observed between subcutaneous and omental cASCs during expansion, and they usually reached the confluence at the same time, after seeding with the same cell density.

Osteogenic Differentiation Assessment

The osteogenic differentiation was induced by supplementing culture medium with well known osteogenic factors, such as ascorbic acid, β-glycerophosphate and dexamethasone [44]. The osteogenic differentiation was firstly demonstrated as observed by Alizarin Red staining (Fig. 3) of the cells after 21 days of culture, which showed an increase of matrix mineralization along culturing time. Analysis of cells stained with H&E revealed no evident differences in terms of cellular morphology and matrix mineralization between cASCs obtained from subcutaneous or from omental adipose tissue.
Fig. 3

Light microscopy images of cASCs from subcutaneous, (Sc, ad) and omental (Om, eh) origins and expanded up to passage 2, cultured for 14 (a, e), 21 (b, f), 28 (c, g) and 35 (d, h) days in osteogenic conditions, stained with Alizarin Red

The osteogenic differentiation was further evaluated by real time RT-PCR analysis of the expression profile of three different osteogenic markers, namely COLIA1, RUNX2 and Osteocalcin along culturing. In the Fig. 4, it is possible to observe the different expression profiles of the genes analyzed at different culturing time points. A down-regulation of COLIA1 is evident along time and is significantly different between the first three culturing times in all conditions. Regarding the RUNX2 gene, an early marker for osteoblastic differentiation, it was found a significant up-regulation from day 14 to 21 followed by a decrease in all conditions. Osteocalcin, a late marker, increased its expression along culturing time, but more evidently between day 21 and 28.
Fig. 4

Real time RT-PCR analysis results of the various osteoblastic genes: a COLIA1, b RUNX2 and c Osteocalcin analysed in canine ASCs, obtained from subcutaneous and omental origin and expanded up to four passages, cultured in osteogenic conditions during 14, 21, 28 and 35 days (mean ± SEM). Considering each marker independently, levels not connected by same letter are significantly different (p < 0.05, Student’s t-test). Tables shown in d, e and f mention the means. Abbreviations: P passage; Sc subcutaneous; Om omental

The number of passages has also shown to have a strong influence in the osteogenic differentiation potential of cASCs, since it was observed a higher expression of the studied markers in the cells from lower passages. Concerning COLIA1, the differences were significant in the cells from subcutaneous origin. In the case of the RUNX2, significant differences were identified in the cells from subcutaneous origin at day 14 and 21 and in cells from omental origin at the 21th day. Moreover, the passaging effect was observed in almost all conditions when evaluated the Osteocalcin gene, particularly, when comparing the P2 and P4.

Taking in account the anatomical origin of the cells, it was observed a higher expression of COLIA1 in those from subcutaneous depots, drastically different in the first culturing time, and also in other some punctual conditions. In the RUNX2 profile, it was not observed significant differences, except in the 21th day of culture. Osteocalcin gene was up-regulated in the omental cASCs, mostly in the 14th and 21th culturing times.

Comparing the effect of anatomical origin of cASCs in the osteogenic potential, it was observed a different expression profile of the osteogenic markers analyzed for ASCs of omental and subcutaneous origin. Specifically, early markers (COLIA1 and RUNX2) expression was higher in subcutaneous ASCs, except at 21 days of culture, while the late marker analyzed (Osteocalcin) exhibit a higher expression level in omental ASCs, but this result was only statistically significant for the cells in passage 2.

Chondrogenic Differentiation Assessment

The chondrogenic differentiation of cASCs was induced by culturing cell pellets with medium supplemented with well established chondrogenic factors, namely, TGF-β1, insulin-transferrin-selenium, ascorbic acid, dexamethasone, L-proline and sodium pyruvate [45]. The Fig. 5 shows the positive staining of cartilage matrix components, demonstrating the ability of cASCs to differentiate into the chondrogenic lineage. No visible differences were observed comparing samples from the various passages and from different anatomical origins, all of them revealing round cell morphology, low cell density and extensive cartilaginous matrix deposition.
Fig. 5

Representative light microscopy images of subcutaneous cASCs cultured for 21 days in chondrogenic conditions stained with Toluidine Blue (a), Safranin O (b), Alcian Blue (c) and H&E (d)


Adipose tissue is a promising source for adult mesenchymal stem cells due to its abundance and easiness of harvesting. Adipose-derived stem cells from canine origin (cASCs) have been poorly studied as patented in the relatively low number of scientific publications [24, 35, 37, 40, 46]. In this work, we characterized the stemness of these cells obtained from different anatomical sites, assessing the expression of three typical markers of MSCs [17] along successive passages, from passage 0 (P0) to passage 4 (P4). Additionally, we evaluated the cASCs potential to differentiate into two distinct cellular lineages, namely the osteogenic and chondrogenic. The osteogenic potential was further analysed, considering the influence of passaging, assessing mineralization by alizarin red staining and also by real time RT-PCR analysis of typical osteogenic markers expression.

While in humans it was found a significant variability on the ASCs behavior when obtained from donors of different ages [47], there is no such data on canine ASCs. In this study, we have obtained cASCs from donors of different ages, but as we had only one donor of each age, it was not possible to obtain conclusive data on this issue.

The cASCs obtained by enzymatic digestion of both omental and subcutaneous tissue revealed the fibroblast-like shape typical of MSCs [17]. This morphology is characteristic of the ASCs obtained from human, rodents, rabbit, sheep, horse and dogs [20, 30, 32, 33, 35, 48] from distinct anatomical origins, as subcutaneous abdominal and inguinal regions, omentum, arms and legs, breast, buttocks and footpad [35, 47, 49, 50].

Cell-based therapies often require in vitro expansion of cells collected from small tissues samples, in order to obtain clinically relevant cell numbers, i.e., to enable a specific therapeutic effect in a given application. The obtained results regarding the cASCs stemness, a parameter directly related to the cellular proliferative potential, have shown us that along passages there is a decrease of MSCs markers expression, particularly evident in the cells from subcutaneous origin. The decreasing expression of one or several MSCs markers along passages herein assessed for ASCs, and previously reported for human [34], equines (subcutaneous tissue from tail’s region) [33], rabbit and sheep [32], may compromise their proliferation and/or differentiation potential and thus limit their in vitro expansion. However, this potential limitation can probably be overcome by the large amount of adipose tissue samples and high cell yields which can be safely harvested and isolated, as compared to other stem cell sources.

Previous studies on cASCs, reported the positive expression of CD90 and negative expression of CD73 and CD105 [37], assessed by flow cytometry (FC), which are in disagreement with our data. These differences might be due to the method of collection (liposuction instead of surgical collection) or due to the use of unspecific antibodies for flow cytometry. Oh et al. (2011) reported a positive expression (determined by FC) only of CD90 [40] for cASCs, however, this data is difficult to compare because in the referred studies, human-specific antibodies, and not dog-specific antibodies, were used to assess CD73 and CD105 expression.

Apart from the MSCs markers, it is has been demonstrated that cASCs express pluritotency genes, such as, OCT4, NANOG and SOX2, typical of the embryonic stem cells [26, 35], as it has been also reported for rhesus monkey and human ASCs [34]. Lim et al. (2010) added that CD90 is expressed until the 7th passage of cASCs and that at the same moment, the embryonic markers found previously, are not present anymore [26].

Additionally to the markers above mentioned, many others were identified in ASCs of other species. In human ASCs, was already observed the positive expression of HLA-ABC, CD9, CD10, CD13, CD29, CD34, CD44, CD49d, CD49e, CD54, CD51, CD55, CD71, CD166, SH3, STRO1, OCT4, UTF1 and Snail2, and the negative expression of HLA-DR [3, 20]. The decrease of MSCs and embryonic markers expression, like OCT4 and SOX9, along passaging has also been reported for human ASCs [34]. Baglioni et al. (2009) concluded that, as observed in this work for cASCs, in the early passages of hASCs the expression levels of MSCs markers are higher, adding that there was no evident difference between the visceral and subcutaneous ASCs due to the heterogeneity of the cellular populations [51]. In human and monkey, a strong positive expression of CD90 and a positive expression of CD105, among other markers, was identified by FC remaining stable trough passage 20–30 [34]. In equines, it was proved, by flow cytometry, the negative expression of CD13 and the positive expression of CD44 and CD90 with variations along the first three passages [33]. Also in horses, was referred the expression of the same MSCs markers that we have studied, but variations in the proliferative pattern were observed until P4, being the propagation cessed after P8 [52]. Rabbit and sheep ASCs have been reported to have higher proliferative capacity as compared to humans [32], however, these studies did not address the gene expression of the MSCs markers.

In this study it was shown that the cASCs from both anatomical sites analysed, subcutaneous and omental, expressed the typical MSCs markers. The adipose tissue can be collected from subcutaneous (superficial and deep) and visceral depots and from many internal organs [53]. Other authors have already evaluated the proliferative potential of cASCs obtained from different anatomical sites [35, 38, 54, 55] in dogs. Neupane et al. (2008) observed that subcutaneous cASCs reveal higher proliferation capacity comparing to omental [35]. Wu and colleagues (2000) have registered differences in the adipogenic potential, which was found lower in the cells from omental and perineum as compared to the cells from inguinal origin [38]. In humans it has been also reported that subcutaneous adipose tissue has a higher proliferation rate [3, 31]. Furthermore, Dragoo et al. (2003) stated the number of cells harvested from human patellar adipose tissue was higher comparing to fat pad [55]. Oedayrajsingh-Varma et al. (2006) concluded that human subcutaneous adipose tissue from the abdomen had a higher yield of stromal vascular cells in comparison to mammary fat, but all type of cells demonstrated similar viability. In the same study, was observed that three different collection techniques (surgical resection, tumescent and ultrasound-assisted liposuction) did not affect the number and viability of the hASCs, however, when used the ultrasounds the proliferation was negatively affected [54].

In the present work, all the samples were collected by surgical resection, avoiding contaminating the tissue with blood and vessels, where also exist cells which express MSCs markers. The higher CD105 (endoglin) expression in the omental tissue, where more vessels were observed during the surgeries, can be due to the fact that endoglin has an important role in the vascular remodeling [56].

Recently, a work on ASCs from canine, equine as porcine origin proved that, apart from the passaging and anatomical site effect, the cellular kinetics can be influenced by the way the cells are cultured. They demonstrated that the cASCs, when cultured with 10 % FBS supplementation start to proliferate more quickly in comparison to those cultured with serum free medium supplemented with 2 % UltroserG; however in both cultures the same number of cells was obtained at the end of the 14th culturing day [57]. These findings evidence that the culturing protocols should be taken into account when we compare data between different studies, even when referring to the same cells source.

As mentioned before, canine adipose-derived stem cells have demonstrated ability to differentiate into many lineages, namely osteogenic, chondrogenic, adipogenic [35, 37, 40, 46], myogenic [37, 40] and neurogenic [24, 40]. In the present work, we evaluated the osteogenic potential of these cells, isolated from fat tissue harvested from distinct anatomical sites, and observed that cASCs cultured under osteogenic stimuli, expressed the three osteoblastic markers analysed, namely COLIA1, RUNX2 and Osteocalcin. The later one, which is a late marker of osteogenesis, increased its expression along culturing time, but more evidently between day 21 and 28, as observed in human ASCs by PCR analysis, that showed a higher expression at the 28th day [23], and similarly to a report in mice where the Osteocalcin expression was higher at the 25th day, and earlier (15th day) on the cells cultured with retinoic acid supplementation [29]. In another study on human ASCs, the secretion of Osteocalcin were registered by ELISA at the 18th culturing day [58]. Thus, in terms of their osteogenic differentiation potential, canine, rodent and human ASCs can be considered similar.

Canine ASCs from both anatomical origins expressed the three studied osteogenic markers as previously reported by Neupane et al. (2009) and by Vieira et al. (2010), and exhibit the typical expression profile along the osteogenic process described in the literature. Alonso et al. (2008) observed that Osteocalcin started to be expressed after 2nd week of osteogenic induction as found in the present study [59].

The passaging effect, herein demonstrated by the decrease of expression levels of the typical MSCs markers, was also evident on the osteogenic differentiation; however, cells from later passages maintain the expression of the osteoblastic maskers, leading us to conclude that the negative effect on the stemness is not directly followed by a negative effect on the osteogenic potential, and thus it does not seem to compromise the use of canine subcutaneous or omental adipose tissue in potential applications envisioning bone regeneration, as observed in previous studies in humans or in mice [23, 29, 58].

Izadpanah and colleagues (2006) observed that in primates, including humans, the percentage of colonies that underwent osteogenic differentiation, decrease along time, comparing the lower passages (50–65 %) and the higher passages (20–25 % at P20) [34], contrary to the findings reported by Alonso and colleagues (2008), who concluded that human ASCs reveal similar osteoblastic potential between P1 and P4 [59].

Human ASC have already been isolated from different fat depots, being reported that subcutaneous adipose tissue has a higher proliferation rate while visceral fat has a higher osteogenic differentiation capacity [3, 31]. Aksu and colleagues (2008) revealed that the ASCs from superficial subcutaneous layers had higher osteogenic potential, namely in males, comparing to deeper subcutaneous depots, adding that the ASCs from male donors had better proliferative behavior [47]. In addition, Marcassus et al. (2006) observed that the density of osteogenic clusters was significantly lower in cultures from epidydimal fat pads as compared to inguinal depot [60].

Other osteoblastic genes, such as bone sialoprotein (BSP) [35, 37], Osterix [37] and Osteoprogesterin [46], were previously assessed in canine ASCs cultured in osteogenic medium. In humans, the collagen type II, Osteopontin, Osteonectin [61], parathyroid hormone receptor, bone morphogenic proteins (BMP) 2 and 4 and BMP receptors types IA, IB andII have been reported as markers of ASCs previously exposed to culture medium supplemented with ascorbic acid, β-glycerophosphate and dexamethasone [58]. In murine ASCs, cultured both in 2D and 3D conditions, Osteopontin, RUNX2, and Osteocalcin were also expressed [29]. As observed in studies on ASCs from other animal species, including dog, these cells have the capacity to maintain the chondrogenic potential along passages, even until P10 [32]. Recently, a study on cASC registered the positive expression of SOX9 and collagen type II, after culturing with chondrogenic media. Nevertheless, the expression levels were lower comparing to those registered for MSCs from canine bone marrow [62]. In humans, the collagen type II expression was also found higher in BMSCs than in ASCs, when cultured with chondrogenic medium supplemented with TGF-β3 up to 21 days [49].

In summary, these results obtained in the present work clearly demonstrate the capacity of the canine adipose-derived stem cells to differentiate into the osteogenic and chondrogenic lineage, making them promising candidates for bone and cartilage regeneration approaches and disease management based on veterinary regenerative medicine approaches.

In the authors’ knowledge, this study brought about the first data on the stemness and osteogenic potential of cASCs upon subsequent culturing periods. The results obtained revealed that cASCs exhibit a progressively decreased expression of the typical stem cells markers along passages and also a decreased osteogenic differentiation potential. Moreover, the anatomical origin of the adipose tissue has an evident effect in the differentiation potential of the ASCs. In the literature, many other factors, such as, age, sex, isolation and conservation methods are proved to have effect in the biologic behavior these cells in humans [21, 47] but the effect of such factors in canine ASCs behavior remains unclear.

As observed in other animal species, the cASCs show resemblances to human ASCs profile, regarding stemness and osteogenic potential along culturing time and passages, which allow us to conclude that canine adipose tissue is a source of cells which could be used in tissue engineering research envisioning human application.


Authors acknowledge the support from the Portuguese Foundation for Science and Technology (FCT) project (ref. MIT/ECE/0047/2009) and for João Filipe Requicha PhD scholarship (SFRH/BD/44143/2008).

Conflict of interest statement

None of the authors has any financial or personal relationships that could inappropriately influence or bias the content of the paper.

Copyright information

© Springer Science+Business Media, LLC 2012