Research ArticleSynergistic Effects of FGF-18 and TGF-β3 on theChondrogenesis of Human Adipose-Derived Mesenchymal StemCells in the Pellet Culture

Liangjie Huang,1,2 Lingxian Yi,3 Chunli Zhang,1 Ying He,1 Liangliang Zhou,1 Yan Liu,1

Long Qian,1 Shuxun Hou ,1 and Tujun Weng 1

1Department of Orthopaedics, First Affiliated Hospital of PLA General Hospital, No. 51 Fucheng Road, Beijing, China2Department of Orthopaedics, The 41st Hospital of PLA, No. 80 Naidong Road, Shannan, China3Department of ICU, The 306th Hospital of PLA, No. 9 Anxiangbeili Road, Beijing, China

Correspondence should be addressed to Shuxun Hou; hsxortho@hotmail.com and Tujun Weng; wengtujun@163.com

Liangjie Huang, Lingxian Yi, and Chunli Zhang contributed equally to this work.

Received 18 December 2017; Accepted 20 February 2018; Published 9 May 2018

Academic Editor: Heng Zhu

Copyright © 2018 Liangjie Huang et al. This is an open access article distributed under the Creative Commons Attribution License,which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Cell-based therapy serves as an effective way for cartilage repair. Compared with a limited source of autologous chondrocytes,adipose-derived stem cells (ADSCs) are proposed as an attractive cell source for cartilage regeneration. How to drivechondrogenic differentiation of ADSCs efficiently remains to be further investigated. TGF-β3 has shown a strong chondrogenicaction on ADSCs. Recently, fibroblast growth factor 18 (FGF-18) has gained marked attention due to its anabolic effects oncartilage metabolism, but existing data regarding the role of FGF-18 on the chondrogenic potential of mesenchymal stem cells(MSCs) are conflicting. In addition, whether the combined application of FGF-18 and TGF-β3 would improve the efficiency ofthe chondrogenic potential of ADSCs has not been thoroughly studied. In the current study, we isolated human ADSCs andcharacterized the expression of their surface antigens. Also, we evaluated the chondrogenic potential of FGF-18 on ADSCs usingan in vitro pellet model by measuring glycosaminoglycan (GAG) content, collagen level, histologic appearance, and expressionof cartilage-related genes. We found that FGF-18, similarly to TGF-β3, had a positive impact on chondrogenic differentiationand matrix deposition when presented throughout the culture period. More importantly, we observed synergistic effects of FGF-18 and TGF-β3 on the chondrogenic differentiation of ADSCs in the in vitro pellet model. Our results provide criticalinformation on the therapeutic use of ADSCs with the help of FGF-18 and TGF-β3 for cartilage regeneration.

1. Introduction

Articular cartilage is an avascular tissue with a poor capacityfor regeneration and repair [1]. Cartilage defects often inducethe development of osteoarthritis and consequent disability.Autologous chondrocyte implantation has proven beneficialfor treating articular defects in patients during the last twodecades [2, 3]. However, difficulties in the expansion of chon-drocytes and proneness to dedifferentiate during extendedcultures have led to the use of mesenchymal stem cells(MSCs) as an alternative for cartilage repair [4].

Adult MSCs display high potential for proliferation andmultipotency and can be isolated from various tissues suchas bone marrow and adipose tissue [4]. The capacity of self-renewal and the chondrogenic potential of human bonemarrow-derived MSCs (BMSCs) have been well documented[5]. Compared with BMSCs, human adipose-derived stemcells (ADSCs) contain similar characteristics as BMSCs andhave their own advantages in abundance, ease of access,and better maintenance of cell phenotype during passage[6]. Growing evidence has suggested that bone marrow andadipose-derived stem cells could promote the regeneration

HindawiStem Cells InternationalVolume 2018, Article ID 7139485, 10 pageshttps://doi.org/10.1155/2018/7139485

of cartilage; however, underlying mechanisms that governtheir chondrogenic potential remain largely unknown [4].Therefore, it is crucial to identify key growth factors that willselectively promote chondrogenesis and cartilage productionin the microenvironment.

Since mesenchymal condensation is critical in cartilagedevelopment, high-density cell pellets that mimic the chon-drogenic program were developed and used to examine thechondrogenic differentiation of MSCs [7]. Specific growthfactors or bioactive molecules were added into the pelletculture system to determine the effects in the chondro-genic differentiation for MSCs. Presently, members of thetransforming growth factor-β (TGF-β) superfamily arewell recognized as the main chondroinductive growth fac-tors [8]. There are three members of TGF-βs, includingTGF-β1, TGF-β2, and TGF-β3, which bind with the TGF-β receptor then activate downstream molecules. Amongthese, TGF-β3 has the strongest chondrogenic effects onADSCs, showing an increased matrix deposition and chon-drogenic differentiation [9, 10].

Recently, fibroblast growth factors (FGFs) and theirreceptors have been found to play a critical role in articularcartilage repair and the maintenance of cartilage [11].Among these members, FGF-18 has acquired marked atten-tion because of its anabolic effects on cartilage [12]. It hasbeen found that FGF-18 can promote the development ofcartilage, delay the degeneration of articular cartilage, andstimulate the potency for the regeneration of hyaline articularcartilage [13–16]. However, existing data regarding theeffects of FGF-18 on the chondrogenic potential of MSCsare conflicting. In limb bud mesenchymal cells, it has beensuggested that FGF-18 through selectively binding withFGFR3 promotes the differentiation of prechondrogenicmesenchymal cells to cartilage-producing chondrocytes[13]. On the other hand, FGF-18 exhibited a negative roleon the chondrogenesis of human BMSCs when added atthe beginning of the program, while it displayed an anaboliceffect when added later in the presence of TGF-β1 [17]. Thus,the chondrogenic effects of FGF-18 on ADSCs remain to befurther investigated. In addition, whether the combinedapplication of FGF-18 and TGF-β3 would improve theefficiency of chondrogenic potential of ADSCs has not beenthoroughly studied.

In the current study, we isolated human ADSCs andexamined the main expression markers of adult MSCs. Weevaluated the chondrogenic potential of FGF-18 on ADSCsby measuring glycosaminoglycan (GAG) content andollagen amount, histologic appearance, and expression ofcartilage-related genes. We also determined whether FGF-18 in combination with TGF-β3 could enhance chondro-genic differentiation of ADSCs in an in vitro pellet model.

2. Materials and Methods

2.1. Cell Isolation and Cultivation. All adipose tissues wereobtained from three different donors with informed con-sent (age range 55–75 years). Ethical approval for thiswork was granted by the ethics committee of the No. 1Affiliated Hospital of PLA General Hospital. The tissue

was washed extensively with PBS, completely diced, andthen digested with 0.1% collagenase A (Sigma, USA) solu-tion at 37°C for 60min. Then the digested tissue was fil-tered using a 75μm filter mesh (BD Biosciences, USA)and centrifuged at 1200 rpm for 5min, and the superna-tant was removed along with the mature adipocytes. Sub-sequently, cell pellets were resuspended in DMEM LGwith 10% FBS (Corning, USA) and cultured on flasks inan incubator with 5% CO2, at 37°C. The medium waschanged on the following day and thereafter every 3 days.When the cells reached 90% confluence, the cultures weretrypsinized and passaged a further two times in T75 cul-ture flasks (Corning, USA). Passage 3 to 5 cells were usedfor chondrogenic differentiation experiment.

2.2. Cell Characterization by Flow Cytometry.Human ADSCsfrom passage 3 were characterized by flow cytometry usingthe following monoclonal antibodies: anti-human CD14-PE, CD34-PE, CD45-PE, CD90-FITC, CD105-PerCP, andCD166-PE (BD Biosciences). Cells were digested with0.25% trypsin/EDTA, resuspended with PBS/FBS(1%),counted, and adjusted to a density of 1× 107/mL. In each flowcytometric tube, aliquots containing 1× 106 ADSCs wereincubated with antibodies for 1 h at 4°C away from light,according to the instructions. For negative control, the bufferwas added with a mouse IgG1 antibody. The suspendedADSCs were washed and then analyzed with a FACSCaliburdevice (BD Biosciences). For each sample, 20,000 events wereacquired and analyzed with the CellQuest software (BD Bio-sciences). Cells were considered MSCs if they were positivefor CD90, CD105, and CD166 and negative for CD14,CD34, and CD45.

2.3. Chondrogenic Differentiation in Pellet Culture. High-density cell pellet cultures are used to examine the chon-drogenic differentiation of MSCs. For the preparation ofeach pellet, passage 3~5 human ADSCs were harvestedand counted and aliquots of 5× 105 cells in the definedmedium were spun down at 300 g for 10minutes in a15mL polypropylene conical tube (Corning, USA). Thebasic chondrogenic culture medium consists of DMEMsupplemented with 10% FBS (Corning, USA), 100nM dexa-methasone, 1% ITS+ Premix, 100μg/mL sodium pyruvate,50μg/mL L-ascorbate-2-phosphate, 40μg/mL proline, and1% penicillin–streptomycin. All reagents were obtained fromSigma-Aldrich unless otherwise noted. Pellets were dividedinto 4 experimental groups, including control group, TGF-β3 group (10ng/mL), FGF-18 group (100 ng/mL), and thecombination of TGF-β3 and FGF-18 group. Pellets were cul-tured with the basic chondrogenic medium (see above) sup-plemented with TGF-β3 (PeproTech, USA) and/or FGF-18(PeproTech, USA) for indicated periods. The control groupwas treated with the basic chondrogenic medium which isthe same with that in the other three groups. The mediumwas changed every three days.

2.4. Glycosaminoglycan (GAG) and DNA Quantification. Atthe indicated time points, ADSC pellets were collected anddigested for 18h at 60°C using 125μg/mL papain solution

2 Stem Cells International

(1mL/sample papain solution) and then clarified by centrifu-gation. Aliquots of the supernatant were assayed separatelyfor glycosaminoglycan (GAG) and DNA content. GAG inpellets was evaluated using the dimethylmethylene blue(DMMB) binding assay described before [18]. Two hundredmicroliters of DMMB reagent (16mg of dye in 1 L of watercontaining 3.04 g glycine, 2.37 g NaCl, and 95mL 0.1MHCl) was added to 50μL of supernatant of papain digest,and the absorbance at 525nm was immediately measured.The standard curve for the analysis was made using a chon-droitin 4-sulfate standard. DNA was extracted from thepapain-digested sample by the standard protocol usingproteinase K and phenol, then the concentration was deter-mined directly using NanoDropND 2000 (Thermo Scientific,USA). The DNA content per scaffold was calculated while

taking into consideration sample dilution. For glycosami-noglycan synthetic activity, the ratio of GAG/DNA in eachsample was normalized.

2.5. Hydroxyproline (HYP) Assay. Total collagen content wasmeasured by first quantifying the hydroxyproline content inpellets after a 5-week chondrogenic induction. Hydroxypro-line (HYP) assay was performed according to the instructionsof the commercial kit (Nanjing Jiancheng BioengineeringInstitute, China). Briefly, samples were hydrolyzed with anequal amount of 6M HCl at 95°C for 5 h. Then, hydrolyzedsamples were cooled to room temperature and the pH valuewas adjusted to 6.0–6.8, followed by an orderly addition ofthree reagents in the kit. After that, samples were incubatedat 60°C for 15min and centrifuged at 3500 rpm for 10min.

1000

800

600

400

200

0

SSC-

heig

ht

0 200 400 600 800 1000FSC-height

(a)

100 101 102 103 104100

101

102

103

104

Mou

se lg

G1

PE

Mouse lgG1 FITC

UL0.63%

UR0.68%

LL98.40%

LR0.29%

(b)

100 101 102 103 104100

101

102

103

104

CD16

6 PE

CD45 FITC

UL92.39%

UR0.62%

LL6.37%

LR0.62%

(c)

CD90 FITC100 101 102 103 104

100

101

102

103

104

CD14

PE

UL0.00%

UR2.64%

LL9.67%

LR87.7%

(d)

100 101 102 103 104100

101

102

103

104

CD10

5 Pe

rCP

CD90 FITC

UL1.25%

UR79.44%

LL5.12%

LR13.92%

(e)

100 101 102 103 104100

101

102

103

104

CD10

5 Pe

rCP

CD34 FITC

UL73.9%

UR1.73%

LL24.05%

LR0.32%

(f)

Figure 1: FACS analysis of ADSC characteristics. (a) The representative images of gating. (b) The negative results with the isotype-matchedcontrol antibodies. ADSC surface markers (CD90, CD105, CD166, CD14, CD 34, and CD45) were determined (c–f).

3Stem Cells International

Supernatants were used to determine the absorbance at550nm. The hydroxyproline content was calculated accord-ing to the formula described in the instructions. The totalamount of collagen per scaffold was converted by the contentof hydroxyproline using a previously reported hydroxypro-line to collagen ratio of 1 : 7.69 [19].

2.6. RNA Extraction, cDNA Synthesis, and Real-Time PCR.Total RNA was extracted from pellets using the TRIzolreagent according to the manufacturer’s instructions (Invi-trogen, USA). Before RNA isolation, human ADSC pelletswere preserved in RNAlater RNA stabilization reagent afterchondrogenic induction for 5 weeks (Qiagen, USA). Then,RNA was reverse-transcribed to cDNA using All-in-OnecDNA Synthesis SuperMix (Bimake, USA). Real-time PCR

was performed to measure the relative mRNA levels usingQuantStudio 5 (Applied Biosystems, USA) with UltraSYBRMixture (CWBIO, China). All samples were measured intriplicate and normalized to internal control GAPDH. Theexpression level of each target gene was calculated by Quant-Studio Design and Analysis desktop software. The annealtemperature is 57°C, and primer sequences were describedas follows: collagen II (forward-5′AAG TCC CTC AACAAC CAG AT3′ and reverse-5′CCA GTA GTC TCC ACTCTT CCA3′), collagen X (forward-5′GGC AAC AGC ATTATG ACC C3′ and reverse-5′GAT GAT GGC ACT CCCTGA A3′), Sox-9 (forward-5′GTA CCC GCA CTT GCACAA C3′ and reverse-5′TTC TTC ACC GAC TTC CTCC3′), aggrecan (forward-5′TAC AAA CGC AGA CTA

⁎

⁎

⁎

⁎

⁎

⁎⁎

⁎

⁎

⁎⁎

⁎

⁎

500

400

300

200

100

0

GAG

(�휇g/

pelle

t)

1 week 2 weeks 3 weeks 4 weeks

GAG content

ControlTGF-�훽3

FGF-18TGF�훽3 + FGF18

(a)

15

10

5

0

DN

A (�휇

g/pe

llet)

1 week 2 weeks 3 weeks 4 weeks

DNA content

ControlTGF-�훽3

FGF-18TGF�훽3 + FGF18

(b)

⁎⁎

⁎

⁎⁎

⁎

⁎

⁎⁎

⁎

⁎

⁎⁎

⁎

⁎

50

40

30

20

10

0

�휇g

GAG

/ �휇

g D

NA

1 week 2 weeks 3 weeks 4 weeks

GAG/DNA

ControlTGF-�훽3

FGF-18TGF�훽3 + FGF18

(c)

Figure 2: The quantification of GAG, DNA, and GAG/DNA in ADSC pellet induced by TGF-β3 alone, FGF-18 alone, and theircombination at indicated time points. (a) Synthesis of sulfated GAG (sGAG) from hADSC pellet. (b) Change in hADSC DNAcontents in pellet culture at different time points. (c) The change of GAG/DNA in ADSC pellet at different groups and indicatedtime points. Values are mean ± SEM, n = 5. The asterisk ∗ (p < 0 05) and asterisks ∗∗ (p < 0 01) indicate significant difference comparedto the control group.

4 Stem Cells International

CAG AAG C3′ and reverse-5′AAA GCG ACA AGA AGAGGA CA3′), COMP (forward-5′CCC AGA AGA ACGACG ACC AA3′ and reverse-5′CCC TAT ACC ATC GCCATC ACT3′), Prg4 (forward-5′TGA CGG CTA TGA TTACTA TGC3′ and reverse-5′AGT TGA CTC CTC CTT TGCTC3′), and GAPDH (forward-5′TGG CTA CAG CAACAG GGT G3′ and reverse-5′ATG GCA ACT GTG AGGAGG G3′).

2.7. Histological and Immunohistochemical Analysis. After a5-week chondrogenic induction, human ADSC pellets werefixed overnight in 4% paraformaldehyde (PFA), followed bydehydration in a graded ethanol series, and embedded in par-affin. The specimens were sliced into 5μm sections. Thesections were stained with hematoxylin and eosin (H&E) orAlcian blue and Sirius red to visualize the presence and distri-bution of glycosaminoglycan (GAG) and total collagen asdescribed [20]. Immunohistochemical examination was per-formed using a monoclonal antibody against type II collagen(Abcam, USA) as described previously [20], and cell nucleiwere counterstained with hematoxylin. All images wereobtained with a light microscope (Olympus, Japan) and adigital camera attachment (Nikon, Japan).

2.8. Statistical Analysis. Results are reported as means± stan-dard deviation. Statistical analysis was performed usingSPSS 17.0 software. One-way ANOVA and Dunn’s mul-tiple comparison test were used to examine significantdifference (p < 0 05).

3. Results

3.1. Immunophenotypic Characterization of Cultured ADSCs.Human adipose-derived stem cells (ADSCs) were isolatedfrom adipose tissue, and subsequently passages 3 to 5 wereused in our study. Morphologically, cultures showed thetypical fibroblast-like features as primary MSC (data notshown). Since mesenchymal stem cells were known to

express typical surface markers, the phenotypes of ADSCsat passage 3 were characterized by flow cytometry analysis.In our study, mesenchymal characteristics were evaluatedusing the positive MSC markers, CD90, CD105, andCD166, and the negative MSCs markers, CD14, CD34, andCD45. As expected, the MSC markers CD90, CD105, andCD166 were observed in more than 75% of the cells, whilethe MSC negative markers CD14, CD34, and CD45 wereobserved in less than 5% of the cells (Figures 1(a)–1(f)).These results suggested that ADSCs we cultured maintainedtypical phenotypes of MSCs and could be used to investigatethe chondrogenic differentiation.

3.2. Biochemical Analysis of Extracellular Matrix on ADSCPellet. The 3D high-density pellet culture is a commonlyused model for chondrogenesis in vitro. To investigatethe chondrogenic potential of FGF-18 alone and incombination with TGF-β3, ADSCs were induced towardschondrocytes in the pellet culture using the basic chondro-genic medium supplemented with TGF-β3 only, FGF-18only, or the combined application of FGF-18 and TGF-β3 for the indicated periods. ADSCs cultured in the basicchondrogenic medium were considered as negative con-trol, while exposed TGF-β3 was considered as a positivegroup. For a quantitative comparison of matrix productionbetween the different groups over time, glycosaminoglycan(GAG) levels and collagen content in ADSC pellets wereexamined by DMMB assay at 1, 2, 3, and 4 weeks andhydroxyproline quantification at 5 weeks, respectively.After a 2-week chondrogenic induction, TGF-β3 signifi-cantly increased the production of GAG as in previousreports (Figure 2(a)). Notably, pellets cultured with FGF-18 showed a remarkable increase in GAG content com-pared with the negative control at four different timepoints, suggesting that addition of FGF-18 throughoutthe culture period stimulated chondrogenesis of ADSCsas well as TGF-β3 (Figure 2(a)). In addition, highestGAG levels were observed in the combined treatment ofTGF-β3 and FGF-18 group at the indicated time points(Figures 2(a) and 2(c)), suggesting that there were syner-gistic effects of FGF-18 and TGF-β3 on the chondrogenicdifferentiation of ADSCs. No significant difference in DNAcontent per pellet was observed among these groups(Figure 2(b)). GAG normalized to DNA was higher inthe chondrogenic group containing the growth factor thanin the negative control, regardless of the presence ofTGF-β3 or FGF-18 (Figure 2(c)). There was nearly noincrease in GAG/DNA in the control group over time(Figure 2(c)). Also, synergistic effects of FGF-18 andTGF-β3 on the production of GAG were observed in fourdifferent time points (Figure 2(c)). Quantification ofcollagen content in ADSC pellet induced for 5 weeksshowed that TGF-β3 or FGF-18 alone enhanced collagenproduction compared with the control, respectively(Figure 3). In addition, TGF-β3 in combination withFGF-18 largely increased the average amounts of collagenper pellets and showed synergistic effects on ADSC chon-drogenesis (Figure 3). A biochemical evaluation of theGAG and collagen content suggests that FGF-18 worked

50

40

30

20

10

0

Col

lage

n co

nten

t (�휇

g/pe

llet)

⁎

⁎

⁎

Total collagen content

ControlTGF-�훽3

FGF-18TGF�훽3 + FGF18

Figure 3: The quantification of collagen content per ADSC pelletafter a 5-week chondrogenic induction. The asterisk ∗ (p < 0 05)indicates significant difference compared to the control group.

5Stem Cells International

as a chondrogenic agent as well as TGF-β3, and a combi-nation of FGF-18 and TGF-β3 would probably improvethe chondrogenic differentiation potential of ADSCs.

3.3. Histological and Immunohistochemical Analysis of ADSCPellet. In order to evaluate the action of FGF-18 or combina-tion of FGF-18 and TGF-β3 on ADSC chondrogenicdifferentiation, a histological and immunohistochemicalevaluation was performed to determine the extracellularmatrix synthesis and deposition in chondrogenic differenti-ated ADSCs in pellet culture. H&E staining showed nodifference in pellet size and histological appearance in thesepellets at 5 weeks (Figure 4(a)). Next, we performed Alcianblue and Sirius red staining for cartilage proteoglycancollagen, respectively. Alcian blue staining demonstrated thatthe blue coloration was increased in the extracellular matrixof ADSC cultures’ exposure to FGF-18 and TGF-β3compared to that of other cultures (Figure 4(b)). It alsoshowed more GAG deposition in FGF-18-treated pelletscompared with the negative control group (Figure 4(b)).After a 5-week chondrogenic induction, Sirius Red stainingshowed that a more intense staining signal for collagen wasobserved in the ADSC pellets treated with FGF-18 andTGF-β3 together (Figure 4(c)). We further determined thetype II collagen expression level in ADSC pellets culturedfor 5 weeks by immunohistochemical analysis. Consistentwith the histological results, a higher type II collagen expres-sion was found in the group of the combination of FGF-18and TGF-β3 compared to that of other groups(Figures 5(a)–5(d)). There was also an increased expressionof type II collagen in pellet cultures treated with FGF-18 orTGF-β3 alone compared with the negative control(Figures 5(a)–5(c)). These results further confirmed thatFGF-18 has chondrogenic property as well as TGF-β3 andthat FGF-18 in combination with TGF-β3 greatly potentiated

a chondrogenic differentiation of human ADSCs in thein vitro pellet model.

3.4. Gene Expression Changes on ADSC Pellet. To uncoverthe underlying gene expression pattern, the extent ofchondrogenesis was evaluated by quantifying the chondro-cytic cell-specific genes, including type II collagen (Col2a1),aggrecan, SOX-9, proteoglycan 4 (Prg4), cartilage oligomericmatrix protein (COMP), and type X collagen (Col10). After 5weeks of chondrogenic induction, real-time PCR analysisfound that all of these chondrocyte-related genes weremuch more highly expressed in ADSC pellets cultured inthe presence of growth factors (FGF-18 alone, TGF-β3alone, or FGF-18+TGF-β3) than that in the negative con-trol (Figures 6(a)–6(f)). Treatment with FGF-18 alone sub-stantially upregulated chondrogenic gene expression relativeto the negative control, similar to the effects of the well-known chondrogenic agent TGF-β3 (Figures 6(a)–6(f)). Nosignificant difference in COMP expression was foundbetween FGF-18+TGF-β3 and FGF-18 or TGF-β3 alone(Figure 6(e)). However, the results revealed that the Col2a1,aggrecan, SOX-9, Prg4, and Col10 mRNA expression in theFGF-18- and TGF-β3-cotreated pellets was significantlyhigher than that in pellets treated with FGF-18 or TGF-β3alone, suggesting that there were synergistic effects of FGF-18 and TGF-β3 on chondrogenic gene expression in thein vitro pellet model (Figures 6(a)–6(d) and 6(f)).

4. Discussion

In this study, we had successfully isolated human ADSCs andcharacterized the expression of their surface antigens. Wefound that FGF-18, similarly to TGF-β3, has a positiveimpact on chondrogenic differentiation as well as matrixdeposition when present throughout the culture time. In

H&

E

Control TGF-�훽3 + FGF-18TGF-�훽3 FGF-18

Alc

ian

blue

Siru

s red

Figure 4: Histological appearance of hMSC pellets: (a) H&E staining; (b) Alcian blue staining; (c) Sirius red staining. Images are fromrepresentative cell cultures at 200 times magnification. Bar, 100 μm.

6 Stem Cells International

addition, we observed synergistic effects of FGF-18 and TGF-β3 on the chondrogenic differentiation of ADSCs.

HumanADSCswere considered as an optional alternativefor cell-based articular cartilage repair in the clinic becausethey were abundant in supply and possessed chondrogenicpotential and easy accessibility [21]. ADSCs show similarcharacteristics with BMSCs, but they also exhibit a numberof different features, that is, cell surface markers, differentia-tion potential, and abundance in the body [4]. In the currentstudy,we isolatedADSCs fromhuman adipose tissue and ana-lyzed the cell surface markers for undifferentiated ADSCs.Consistent with the results described previously [22], weobserved that our cultured ADSC populations were positivefor CD90, CD105, and CD166 and negative for CD14, CD34,andCD45, suggesting that these ADSCs fittedwith the criteriaofMSCs and could be further used for chondrogenic differen-tiation analysis using an in vitro pellet culture model.

A major challenge in ADSC application for cartilagerepair is on how to control and facilitate their chondrogenicdifferentiation. Despite growing information regardingMSCs, the mechanisms contributing to the chondrogenesisof ADSCs are not well defined. One rational strategy is toidentify key growth factors and recapitulate the in vivo envi-ronment by using these growth factors to improve chondro-genesis of ADSCs. Previous studies have suggested thatvarious growth factors such as TGF-βs, IGF-1, insulin, andBMPs were capable of stimulating in vitro chondrogenesisof ADSCs [8, 23–25]. In our pellet chondrogenic culture sys-tem, as a positive group, TGF-β3-treated ADSC pellets alsoshowed an increased total amount of GAG at different times,a rise in collagen expression, and an upregulated expressionof chondrocyte-related genes after a 5-week chondrogenicinduction. Importantly, our results revealed that FGF-18greatly upregulated the production of the extracellular matrixof GAG and collagen by biochemical evaluation and histo-logical analysis, suggesting that FGF-18 alone supplemented

with the basic chondrogenic medium is sufficient to directchondrogenic commitment of ADSCs as well as TGF-β3when present throughout the entire differentiation program.It was consistent with the results that FGF-18 promotedchondrogenic differentiation and cartilage production inlimb bud mesenchymal cells [13].

There are two different opinions about the role of FGF-18in MSC chondrogenic differentiation: one thinks thatFGF-18 positively regulates chondrogenic differentiation[12, 13], and another suggests that FGF-18 negativelyregulates chondrogenesis [26]. Our results showed thatFGF-18 promotes chondrogenic differentiation of ADSCsby increasing the expression of SOX-9 and also promotesthe expression of chondrocytic matrix gene Col2a1 andaggrecan during chondrogenesis of ADSCs. Besides that,the timing point of FGF-18 action is also controversial. Somereports suggested that FGF-9 and FGF-18 inhibited extracel-lular production or had no significant effects at the early stageof chondrogenesis, but played an anabolic effect when addedat the late stage [17, 27]. In contrast with these reports, wefound that FGF-18 had potent chondroinductive actions onMSCs when present in the chondrogenic medium at theentire culture period. The contradictory results might berelated to different chondrogenic induction media (with orwithout 10% FBS), the concentration of FGF-18 (100 ng/mLor 10 ng/mL), inherent differences of ADSCs, and so on.The chondrogenic culture medium used in our study con-tained 10% FBS, and that might be enough to stimulate theproliferation and survival of ADSCs in the pellet culturemodel and cover or exceed the effects of growth factors onADSC proliferation. In addition, we inferred that higher con-centration of FGF-18 (100 ng/mL) may also have strongerchondrogenic potency for ADSCs than the previously used10 ng/mL FGF-18 [17].

To our knowledge, this study firstly showed the enhancedand synergistic effects of FGF-18 in combination with TGF-

(a) (b)

(c) (d)

Figure 5: Immunohistochemical staining for type II collagen (ColII) expression in ADSC pellet induced to chondrogenic differentiation for 5weeks: (a) control; (b) TGF-β3; (c) FGF-18; (d) FGF-18 +TGF-β3. Increased levels of type II collagen expression in pellet culture containingFGF-18 and TGF-β3 were detected by immunohistochemistry.

7Stem Cells International

β3 on chondrogenesis of ADSCs. When ADSC pellets weresubjected to a combination of FGF-18 and TGF-β3 through-out the culture period, more chondrocytic markers andextracellular matrix were produced. ADSC pellets cotreatedwith these two growth factors displayed increased GAGaccumulation, improved histologic appearance with intenseAlcian blue and Sirius red staining signal, and increasedtranscription of cartilage-related genes, such as aggrecan,Sox-9, Col2a1, and Prg4. Previously, some results alsoshowed that insulin-like growth factor 1 (IGF-1) enhancedthe chondrogenesis effects of TGF-β1 and the combinedapplication of TGF-β3 and BMP-2 has a synergistic effecton chondrogenic differentiation [23, 25]. In addition, italso showed that BMP-2 and IGF-1 were capable of mo-dulating the proliferation and chondrogenic differentiationof ADSCs [28]. In the current study, we found that FGF-18 and TGF-β3 had combined effects on inducing differ-entiation of ADSCs into chondrocytes with a histological

appearance and chondrogenic gene expression profiles inthe in vitro pellet model. The specific molecular mecha-nisms of synergism need to be further elucidated in thenear future study. In addition, whether these two growthfactors can promote chondrogenesis in vivo remains tobe further investigated.

5. Conclusions

In conclusion, our results from ADSC pellet culture de-monstrated that chondroinductive actions of FGF-18 werenearly equal to those of TGF-β3, and the two growthfactors had synergistic effects on ADSC chondrogenesis.Therefore, simultaneous stimulation of ADSCs with TGF-β3 and FGF-18 may serve as a useful approach for assist-ing translational research for cell-based therapies targetingcartilage tissue regeneration.

15

10

5

0Re

lativ

e qua

ntity

Col2a1

⁎ ⁎

⁎⁎

ControlTGF-�훽3

FGF-18TGF�훽3 + FGF18

(a)

6

4

2

0

Rela

tive q

uant

ity

Aggrecan

⁎⁎

⁎

ControlTGF-�훽3

FGF-18TGF�훽3 + FGF18

(b)

6

8

4

2

0

Rela

tive q

uant

ity

SOX-9

⁎

⁎

⁎

ControlTGF-�훽3

FGF-18TGF�훽3 + FGF18

(c)

10

8

6

4

2

0Re

lativ

e qua

ntity

Prg-4

⁎

⁎

⁎

ControlTGF-�훽3

FGF-18TGF�훽3 + FGF18

(d)

10

8

6

4

2

0

Rela

tive q

uant

ity

COMP

⁎ ⁎ ⁎

ControlTGF-�훽3

FGF-18TGF�훽3 + FGF18

(e)

8

6

4

2

0

Rela

tive q

uant

ity

Col10

⁎ ⁎

⁎

ControlTGF-�훽3

FGF-18TGF�훽3 + FGF18

(f)

Figure 6: Gene expression analysis of the chondrogenic genes in ADSC pellets. Chondrocytic-related genes of Col2a1 (a), Aggrecan (b), SOX-9 (c), Prg4 (d), COMP (e), and Col10 (f) were quantified by real-time PCR in ADSC pellet, which were stimulated with TGF-β3 alone, FGF-18alone, and their combination for 5 weeks. The asterisk ∗ (p < 0 05) and asterisks ∗∗ (p < 0 01) indicate significant difference compared to thecontrol group.

8 Stem Cells International

Conflicts of Interest

The authors declare that there is no conflict of interestsregarding the publication of this paper.

Acknowledgments

This project was funded in part by grants from the NationalKey Research and Development Program of China (Grantno. 2017ZYC1103300) and the Beijing Municipal NaturalScience Foundation (Grant no. 7162187). The authors thankProfessor Xiushan Yin from Shengyang University of Chem-ical Technology for revising and proofreading themanuscript.

References

[1] J. Becerra, J. A. Andrades, E. Guerado, P. Zamora-Navas, J. M.Lopez-Puertas, and A. H. Reddi, “Articular cartilage: structureand regeneration,” Tissue Engineering. Part B, Reviews, vol. 16,no. 6, pp. 617–627, 2010.

[2] M. Brittberg, A. Lindahl, A. Nilsson, C. Ohlsson, O. Isaksson,andL. Peterson, “Treatment of deep cartilage defects in the kneewith autologous chondrocyte transplantation,” The NewEngland Journal ofMedicine, vol. 331, no. 14, pp. 889–395, 1994.

[3] L. C. Biant, G. Bentley, S. Vijayan, J. A. Skinner, and R.W. Car-rington, “Long-term results of autologous chondrocyteimplantation in the knee for chronic chondral and osteochon-dral defects,” The American Journal of Sports Medicine, vol. 42,no. 9, pp. 2178–2183, 2014.

[4] C. R. Fellows, C. Matta, R. Zakany, I. M. Khan, andA. Mobasheri, “Adipose, bone marrow and synovial joint-derived mesenchymal stem cells for cartilage repair,” Frontiersin Genetics, vol. 7, p. 213, 2016.

[5] M. F. Pittenger, A. M. Mackay, S. C. Beck et al., “Multilineagepotential of adult human mesenchymal stem cells,” Science,vol. 284, no. 5411, pp. 143–147, 1999.

[6] G. I. Im, “Bone marrow-derived stem/stromal cells andadipose tissue-derived stem/stromal cells: their comparativeefficacies and synergistic effects,” Journal of Biomedical Mate-rials Research. Part A, vol. 105, no. 9, pp. 2640–2648, 2017.

[7] B. T. Estes and F. Guilak, “Three-dimensional culture systemsto induce chondrogenesis of adipose-derived stem cells,”Methods in Molecular Biology, vol. 702, pp. 201–217, 2011.

[8] I. Grafe, S. Alexander, J. R. Peterson et al., “TGF-β family sig-naling in mesenchymal differentiation,” Cold Spring HarborPerspectives in Biology, 2017.

[9] Q. O. Tang, K. Shakib, M. Heliotis et al., “TGF-β3: a potentialbiological therapy for enhancing chondrogenesis,” ExpertOpinion on Biological Therapy, vol. 9, no. 6, pp. 689–701, 2009.

[10] C. Cicione, E. Muinos-Lopez, T. Hermida-Gomez, I. Fuentes-Boquete, S. Diaz-Prado, and F. J. Blanco, “Alternative protocolsto induce chondrogenic differentiation: transforming growthfactor-β superfamily,” Cell and Tissue Banking, vol. 16, no. 2,pp. 195–207, 2015.

[11] M. B. Ellman, D. Yan, K. Ahmadinia, D. Chen, H. S. An, andH. J. Im, “Fibroblast growth factor control of cartilage homeo-stasis,” Journal of Cellular Biochemistry, vol. 114, no. 4,pp. 735–742, 2013.

[12] E. E. Moore, A. M. Bendele, D. L. Thompson et al., “Fibroblastgrowth factor-18 stimulates chondrogenesis and cartilage

repair in a rat model of injury-induced osteoarthritis,” Osteo-arthritis and Cartilage, vol. 13, no. 7, pp. 623–631, 2005.

[13] D. Davidson, A. Blanc, D. Filion et al., “Fibroblast growth fac-tor (FGF) 18 signals through FGF receptor 3 to promote chon-drogenesis,” The Journal of Biological Chemistry, vol. 280,no. 21, pp. 20509–20515, 2005.

[14] J. Power, P. Hernandez, H. Guehring, A. Getgood, andF. Henson, “Intra-articular injection of rhFGF-18 improvesthe healing in microfracture treated chondral defects in anovine model,” Journal of Orthopaedic Research, vol. 32, no. 5,pp. 669–676, 2014.

[15] F. Eckstein, W. Wirth, A. Guermazi, S. Maschek, andA. Aydemir, “Brief report: intraarticular sprifermin not onlyincreases cartilage thickness, but also reduces cartilage loss:location-independent post hoc analysis using magnetic reso-nance imaging,” Arthritis & Rhematology, vol. 67, no. 11,pp. 2916–2922, 2015.

[16] D. Howard, J. Wardale, H. Guehring, and F. Henson, “Deliver-ing rhFGF-18 via a bilayer collagen membrane to enhancemicrofracture treatment of chondral defects in a large animalmodel,” Journal of Orthopaedic Research, vol. 33, no. 8,pp. 1120–1127, 2015.

[17] D. Correa, R. A. Somoza, P. Lin et al., “Sequential exposure tofibroblast growth factors (FGF) 2, 9 and 18 enhances hMSCchondrogenic differentiation,” Osteoarthritis and Cartilage,vol. 23, no. 3, pp. 443–453, 2015.

[18] J. S. Mort and P. J. Roughley, “Measurement of glycosamino-glycan release from cartilage explants,” Methods in MolecularMedicine, vol. 135, pp. 201–209, 2007.

[19] H. V. Almeida, R. Eswaramoorthy, G. M. Cunniffe, C. T.Buckley, F. J. O'Brien, and D. J. Kelly, “Fibrin hydrogelsfunctionalized with cartilage extracellular matrix and incor-porating freshly isolated stromal cells as an injectable for car-tilage regeneration,” Acta Biomaterialia, vol. 36, pp. 55–62,2016.

[20] S. Zhou, Z. Wang, J. Tang et al., “Exogenous fibroblast growthfactor 9 attenuates cartilage degradation and aggravates osteo-phyte formation in post-traumatic osteoarthritis,”Osteoarthri-tis and Cartilage, vol. 24, no. 12, pp. 2181–2192, 2016.

[21] F. Perdisa, N. Gostynska, A. Roffi, G. Filardo, M.Marcacci, andE. Kon, “Adipose-derived mesenchymal stem cells for thetreatment of articular cartilage: a systematic review on preclin-ical and clinical evidence,” Stem Cells International, vol. 2015,Article ID 597652, 13 pages, 2015.

[22] M. Dominici, K. Le Blanc, I. Mueller et al., “Minimal criteriafor defining multipotent mesenchymal stromal cells. TheInternational Society for Cellular Therapy position statement,”Cytotherapy, vol. 8, no. 4, pp. 315–317, 2006.

[23] L. Longobardi, L. O'Rear, S. Aakula et al., “Effect of IGF-I inthe chondrogenesis of bone marrow mesenchymal stem cellsin the presence or absence of TGF-β signaling,” Journal ofBone and Mineral Research, vol. 21, no. 4, pp. 626–636, 2006.

[24] M. B. Mueller, T. Blunk, B. Appel et al., “Insulin is essential forin vitro chondrogenesis of mesenchymal progenitor cells andinfluences chondrogenesis in a dose-dependent manner,”International Orthopaedics, vol. 37, no. 1, pp. 153–158, 2013.

[25] B. Shen, A. Wei, H. Tao, A. D. Diwan, and D. D. F. Ma, “BMP-2 enhances TGF-β3-mediated chondrogenic differentiation ofhuman bone marrow multipotent mesenchymal stromal cellsin alginate bead culture,” Tissue Engineering. Part A, vol. 15,no. 6, pp. 1311–1320, 2009.

9Stem Cells International

[26] N. Ohbayashi, M. Shibayama, Y. Kurotaki et al., “FGF18 isrequired for normal cell proliferation and differentiation dur-ing osteogenesis and chondrogenesis,” Genes & Development,vol. 16, no. 7, pp. 870–879, 2002.

[27] C. A. Hellingman, W. Koevoet, N. Kops et al., “Fibroblastgrowth factor receptors in in vitro and in vivo chondrogenesis:relating tissue engineering using adult mesenchymal stem cellsto embryonic development,” Tissue Engineering. Part A,vol. 16, no. 2, pp. 545–556, 2010.

[28] C. An, Y. Cheng, Q. Yuan, and J. Li, “IGF-1 and BMP-2induces differentiation of adipose-derived mesenchymal stemcells into chondrocytes-like cells,” Annals of Biomedical Engi-neering, vol. 38, no. 4, pp. 1647–1654, 2010.

10 Stem Cells International

Hindawiwww.hindawi.com

International Journal of

Volume 2018

Zoology

Hindawiwww.hindawi.com Volume 2018

Anatomy Research International

PeptidesInternational Journal of

Hindawiwww.hindawi.com Volume 2018

Hindawiwww.hindawi.com Volume 2018

Journal of Parasitology Research

GenomicsInternational Journal of

Hindawiwww.hindawi.com Volume 2018

Hindawi Publishing Corporation http://www.hindawi.com Volume 2013Hindawiwww.hindawi.com

The Scientific World Journal

Volume 2018

Hindawiwww.hindawi.com Volume 2018

BioinformaticsAdvances in

Marine BiologyJournal of

Hindawiwww.hindawi.com Volume 2018

Hindawiwww.hindawi.com Volume 2018

Neuroscience Journal

Hindawiwww.hindawi.com Volume 2018

BioMed Research International

Cell BiologyInternational Journal of

Hindawiwww.hindawi.com Volume 2018

Hindawiwww.hindawi.com Volume 2018

Biochemistry Research International

ArchaeaHindawiwww.hindawi.com Volume 2018

Hindawiwww.hindawi.com Volume 2018

Genetics Research International

Hindawiwww.hindawi.com Volume 2018

Advances in

Virolog y Stem Cells International

Hindawiwww.hindawi.com Volume 2018

Hindawiwww.hindawi.com Volume 2018

Enzyme Research

Hindawiwww.hindawi.com Volume 2018

International Journal of

MicrobiologyHindawiwww.hindawi.com

Nucleic AcidsJournal of

Volume 2018

Submit your manuscripts atwww.hindawi.com