Notch Signaling Through Jagged-1 Is ... - Wiley Online Library

TISSUE-SPECIFIC STEM CELLS Notch Signaling Through Jagged-1 Is Necessary to Initiate Chondrogenesis in Human Bone Marrow Stromal Cells but Must Be Switched off to Complete Chondrogenesis RACHEL A. OLDERSHAW,a,b,c SIMON R. TEW,a,b AMANDA M. RUSSELL,a KATE MEADE,a ROBERT HAWKINS,a TRISTAN R. MCKAY,a,c KEITH R. BRENNAN,b TIMOTHY E. HARDINGHAMa,b,c a

UK Centre for Tissue Engineering, bWellcome Trust Centre for Cell-Matrix Research, and cNorth West Embryonic Stem Cell Centre, Faculty of Life Sciences, University of Manchester, Manchester, United Kingdom

Key Words. Notch signaling • Jagged-1 • Chondrogenesis • Bone marrow stem cells

ABSTRACT We investigated Notch signaling during chondrogenesis in human bone marrow stromal cells (hMSC) in three-dimensional cell aggregate culture. Expression analysis of Notch pathway genes in 14-day chondrogenic cultures showed that the Notch ligand Jagged-1 (Jag-1) sharply increased in expression, peaking at day 2, and then declined. A Notch target gene, HEY-1, was also expressed, with a temporal profile that closely followed the expression of Jag-1, and this preceded the rise in type II collagen expression that characterized chondrogenesis. We demonstrated that the shut-down in Notch signaling was critical for full chondrogenesis, as adenoviral human Jag-1 transduction of hMSC, which caused continuous elevated expression of Jag-1 and sustained Notch

signaling over 14 days, completely blocked chondrogenesis. In these cultures, there was inhibited production of extracellular matrix, and the gene expression of aggrecan and type II collagen were strongly suppressed; this may reflect the retention of a prechondrogenic state. The JAG-1-mediated Notch signaling was also shown to be necessary for chondrogenesis, as N-[N-(3,5-difluorophenacetyl-L-alanyl)](S)-phenylglycine t-butyl ester (DAPT) added to cultures on days 0 –14 or just days 0 –5 inhibited chondrogenesis, but DAPT added from day 5 did not. The results thus showed that Jag-1-mediated Notch signaling in hMSC was necessary to initiate chondrogenesis, but it must be switched off for chondrogenesis to proceed. STEM CELLS 2008;26:666 – 674

Disclosure of potential conflicts of interest is found at the end of this article.

INTRODUCTION Cartilage is formed from a unique and specialized extracellular matrix (ECM) composed primarily of the large aggregating proteoglycan aggrecan and an extensive, predominantly type II, collagen network. Cartilage is produced and maintained by chondrocytes embedded within the ECM, and it is an important tissue in vertebrate embryogenesis, as it forms the template for skeletal development and has specific mechanical functions at maturity, including forming the compressive load-bearing surface of articular joints [1– 6]. Mature articular cartilage is avascular and aneural, has a limited capacity for self-repair, and is susceptible to degenerative diseases, such as osteoarthritis [2, 6, 7]. At present there, is no effective long-term treatment for the restoration of degenerate articular cartilage, and cell-based approaches to initiating chondrogenesis offer a potential therapeutic strategy for tissue repair [8 –10]. Adult human bone marrow stromal cells (hMSC) provide a potential source of cells, as they can differentiate into chondrocytes [11–15]. However, they have also been reported to differentiate along other mesenchymal lineages, including adipocyte and osteocyte lineages [15, 16], and have been suggested to generate nonmesenchymal lineages, such as pericytes [17–19]. The cells thus have the capacity to respond to a range of composite signals to be directed into different cell fates.

Chondrogenesis involves differentiation into cells that secrete and assemble the specific expanded ECM characteristic of cartilage, with proteins such as collagen type II and aggrecan [1]. The conditions used to drive chondrogenic differentiation of hMSC typically involve a three-dimensional culture, such as a cell aggregate, with differentiation factors such as transforming growth factor-␤3 (TGF-␤3) and dexamethasone [20, 21]. To understand more about chondrogenesis of hMSC, we have investigated the signaling mechanisms active in cell aggregates. The Notch pathway is a highly conserved signaling mechanism involved in many processes determining cell fate during development [22–24]. Activation of Notch occurs following Notch receptor/ligand engagement upon cell-cell contact, which initiates translocation of the Notch intracellular domain to the nucleus and activation of target genes [25–33]. The cell-cell contact encouraged in chondrogenic three-dimensional (3D) cell aggregates provides the ideal environment for Notch signaling. There is previous evidence that the Notch pathway is active during the transition of prehypertrophic to hypertrophic chondrocytes in growth plate in murine and chick limb development [34, 35], and immunolocalization studies have revealed the presence of Notch receptors and ligands in both murine and bovine cartilage [36, 37]. The ␥-secretase inhibitor N-[N-(3,5difluorophenacetyl-L-alanyl)]-(S)-phenylglycine t-butyl ester (DAPT), which blocks Notch activation, has also recently been reported to inhibit chondrogenesis in hMSC [38, 39], although

Correspondence: Timothy Hardingham, Ph.D., DSc., UK Centre for Tissue Engineering, Faculty of Life Sciences, University of Manchester, Oxford Road, Manchester, M13 9PT, United Kingdom. Telephone: 44-161-275-5511; Fax: 44-161-275-5082; e-mail: timothy.e.hardingham@ manchester.ac.uk Received October 4, 2007; accepted for publication December 11, 2007; first published online in STEM CELLS EXPRESS January 10, 2008. ©AlphaMed Press 1066-5099/2008/$30.00/0 doi: 10.1634/stemcells.2007-0806

STEM CELLS 2008;26:666 – 674 www.StemCells.com

Oldershaw, Tew, Russell et al.

667

an earlier report suggested that DAPT increased chondrogenesis in murine limb bud cells [40]. In this study, we have analyzed the gene expression patterns of the four Notch receptors and their five ligands during chondrogenic differentiation of hMSC in 3D cell aggregate culture. The results suggest an important role for switching on and switching off the Jagged-1 (Jag-1)-mediated Notch signaling in the early events of chondrogenesis.

MATERIALS

AND

METHODS

Cell Culture Human bone marrow mononuclear cells (Cambrex, Wokingham, U.K., http://www.cambrex.com) were seeded at a density of 1.66 ⫻ 105 cells per cm2 using medium from the mesenchymal stem cell growth Bullet Kit (Cambrex) supplemented with 5 ng/ml fibroblast growth factor-2 (FGF-2) (Sigma-Aldrich, Poole, U.K., http://www. sigmaaldrich.com). After 24 hours, nonadherent cells were removed, and cells were confluent at between 14 and 21 days. This method of isolating the plastic-adherent population of cells has previously been shown to enrich for multipotent stromal progenitor cells [15]. Cells at passage 2 or 3 were formed into cell aggregates (500,000 cells per aggregate, centrifuged at 150 RCF for 5 minutes) and cultured for up to 14 days in a serum-free chondrogenic medium (high-glucose Dulbecco’s modified Eagle’s medium [DMEM], 2 mM L-glutamine, 110 ␮g/ml sodium pyruvate, 100U/ml penicillin, 100 ␮g/ml streptomycin, 50 ␮g/ml ascorbic acid 2-phosphate, 40 ␮g/ml L-proline, 100 nM dexamethasone, 100 ng/ml TGF-␤3, 1% [vol/vol] ITS⫹1 supplement [13]). Cells were centrifuged into aggregate cultures at day 0. The wet mass (wet weight) of cell aggregates was recorded at the end of each culture period. For ␥-secretase inhibition studies, medium was supplemented with 50 nM DAPT or 0.05% dimethyl sulfoxide vehicle control [36, 41]. Cell cultures were incubated in 5% CO2, atmospheric O2 at 37°C.

Quantitative Gene Expression Analysis Total RNA was extracted using Tri-reagent (Sigma-Aldrich) from cells ground using Molecular Grinding Resin (Geno Technology Inc., St. Louis, http://www.gbiosciences.com). Total RNA (1 ␮g in a 25-␮l reaction) was reversed-transcribed using Moloney murine leukemia virus reverse transcriptase (Promega, Southampton, U.K., http://www.promega.com) primed with random hexamer oligonucleotides. Real-time polymerase chain reaction was with MJ Research Opticon 2 using SYBR Green detection (Eurogentec, Seraing, Belgium, http://www.eurogentec.be). Gene-specific primers (supplemental online Table 1) were either taken from previously described work [42, 43] or designed using Applied Biosystems Primer Express software (Foster City, CA, http://www.appliedbiosystems. com). Relative expression levels were normalized using glyceraldehyde-3-phosphate dehydrogenase (GAPDH) and calculated using the 2⫺⌬CT method as described previously [44]. The amplification efficiencies of primer pairs were validated as described previously [43] to enable quantitative comparison of gene expression. All primers were from Invitrogen (Paisley, U.K., http://www. invitrogen.com).

Histology and Immunolocalization Cell aggregates (paraffin wax 5-␮m sections fixed in 4% formaldehyde on pretreated slides; Super Frost Plus; Menzel Gla¨ser, Braunschweig, Germany, http://www.menzel.de) were stained for glycosaminoglycan with Safranin O (Sigma-Aldrich) (0.1% [wt/ vol]) and counterstained with Harris’s hematoxylin (R.A. Lamb Ltd., Eastborne, U.K., http://www.ralamb.co.uk) and fast green (Sigma-Aldrich). Immunolocalization was carried out as described previously [45], with antibodies against Jag-1 (C-20; sc-6011), collagen type I (C-18; sc-8786), and collagen type II (N-19; sc7764) used at 1:100 (all three antibodies were from Santa Cruz Biotechnology Inc., Santa Cruz, CA, http://www.scbt.com). Histology images were from the Axioplan 2 imaging system with Axio-

www.StemCells.com

Cam HC and AxioVision Release (11-2004) program, version 4.3 (Carl Zeiss, Welwyn Garden City, U.K., http://www.zeiss.com).

SDS-Polyacrylamide Gel Electrophoresis and Immunoblotting Total cell lysates were prepared from monolayer cells, and cell aggregates were ground up using Molecular Grinding Resin in cell lysis buffer (50 mM Tris-Cl, 150 mM NaCl, 1% Triton X-100 (pH 8.0). Protein lysates were analyzed by SDS-polyacrylamide gel electrophoresis under denaturing conditions on 4%–12% Bis-Tris Gels (Invitrogen) followed by immunoblotting with the Jagged-1 (c-20, sc-6011) antibody (1:100). Protein extract loading was normalized with an anti-GAPDH antibody (MAB374; 1:500; Chemicon, Chandlers Ford, U.K., http://www.chemicon.com). Protein bands were visualized with horseradish conjugated donkey anti-goat (AB324P; 1:100; Chemicon) or goat anti-mouse (A0412; 1:2,000; Sigma-Aldrich) secondary antibodies using a Western Lightning Chemiluminescent Reagent Plus Kit (PerkinElmer Life and Analytical Sciences, Boston, http://www.perkinelmer.com). Normalization was carried out using Molecular Analyst software (Bio-Rad, Hemel Hempstead, U.K., http://www.bio-rad.com).

Cloning of Full-Length Human Jag-1 for Adenoviral Transduction of hMSC Jag-1 cDNA was prepared by cloning Jag-1 from human placental RNA using primers based on published sequences [46] and cloned into the pCR2.1 vector using the Original TA Cloning Kit (Invitrogen) and sequenced using Big Dye terminator cycle sequencing-ready reaction with Amplitaq DNA polymerase FS kit (PerkinElmer). For cell transduction experiments, the hJag-1 cDNA was subcloned into the pAdlox vector [42, 47] and the pcDNA3.1/Hygro(⫹) vector (Invitrogen) at the HindIII/XbaI sites. Recombinant adenovirus was created using Cre-lox recombination as described previously [47]. hMSC were transfected at passage 3 with either Adlox-hJag-1 virus or Adlox-green fluorescent protein (GFP) virus as a control at a multiplicity of infection of 50 in 10% DMEM for 2 hours, washed, and incubated for 24 hours in 10% DMEM. This gave 65%–70% transfection as assessed by flow cytometry of GFP fluorescent cells, and cells were used without further passage for chondrogenic culture.

Assay of Recombinant hJag-1 Activity Assays were performed using the Dual-Luciferase Reporter Assay System (Promega). CHO/N2-luc cells carrying a stably transfected full-length Notch-2 receptor and a CBF-1-driven luciferase reporter construct [48] were transfected in triplicate using Lipofectamine 2000 (Invitrogen) with either 800 ng of pcDNA3.1/Hygro(⫹) or pcDNA3.1/Hygro(⫹)-Jag-1 (1, 10, and 100 ng, made up to 800 ng with the pcDNA3.1/Hygro(⫹) vector). After 48 hours, the cell layer was lysed and assayed using Luciferase II reagent (LAR II) (Promega) followed by Stop & Glo reagent (Promega). Luciferase activity was measured using a MicroLumatPlus LB 96V Luminometer (Berthold Technologies, Redbourn, U.K., http://www. bertholdtech.com) and the WinGlow program, version 1.25 (Berthold Technologies). Data were normalized against background Renilla luciferase activity.

Statistical Analysis The distribution of experimental data sets was assessed by quantilequantile plot analysis using the statistical package S-Plus, version 6.1. Where appropriate, data were transformed to obtain a normal distribution and then analyzed by two-sample Student’s t test. Data sets that were not normally distributed were analyzed by KruskalWallis H test using SPSS, version 11.5 (SPSS, Chicago, http:// www.spss.com).

Notch Signaling in Chondrogenesis

668

Figure 1. Gene expression analysis during chondrogenesis of human bone marrow stromal cells (hMSC) in three-dimensional cell aggregate cultures. (A): hMSC were cultured in cell aggregates for up to 14 days. RNA extracts at day 0, 1, 7, and 14 were analyzed for gene expression of SOX-9, aggrecan, and collagens I and II. Gene expression was normalized to GAPDH, and values represent mean ⫾ SEM (n ⫽ 3). Statistical analysis was carried out for comparison of gene expression values with day 0 results. ⴱ, p ⬍ .05. Note that collagen II gene expression was upregulated by 22,000-fold between days 0 and 7. (B): The wet mass of cell aggregates was determined at days 1, 7, and 14. Values represent mean ⫾ SEM (n ⫽ 10). Statistical analysis was carried out for comparison of mass of cell aggregates between day 1 and days 7 and 14. ⴱ, p ⬍ .05; ⴱⴱ, p ⬍ .001. Abbreviation: GAPDH, glyceraldehyde-3-phosphate dehydrogenase.

RESULTS Chondrogenesis in Cell Aggregate Cultures of hMSC Human bone marrow stromal cells were expanded in the presence of FGF-2 [49, 50], and chondrogenesis was initiated in 3D cell aggregate cultures over 14 days with dexamethasone and TGF-␤3 [12, 13, 20]. Chondrogenesis in the 3D cell aggregate cultures was validated by measuring gene expression changes of characteristic cartilage matrix proteins. Expression of collagen type II was massively increased by 22,000-fold at day 7 and 550,000-fold at day 14 compared with that at day 0 (p ⬍ .05), and aggrecan was also increased by 30-fold at day 14 (p ⬍ .05). In contrast, collagen type I gene expression, more characteristic of the undifferentiated state, did not change during 14 days of culture (Fig. 1A). SOX-9, a transcription factor essential for the expression of many cartilage ECM proteins [51–54], showed a twofold increase in gene expression during the 14-day culture (p ⬍ .05). These gene expression changes were accompanied by extensive ECM deposition and a large rise in the wet mass of the

pellets: 170% increase at day 7 (p ⬍ .05) and 390% increase after 14 days (p ⬍ .001) (Fig. 1B). To investigate Notch signaling, we analyzed the gene expression pattern of the four Notch receptors (Notch-1– 4) and their five ligands (Jag-1, Jag-2, DLL-1, DLL-3, and DLL-4) during the chondrogenic cell aggregate culture (Fig. 2). There was an increase in Notch-1 expression on day 0 of culture, which may have been caused by the change in culture conditions, including lifting from monolayer and/or the effects of TGF-␤3 in the culture medium [55], but its expression decreased by the next day (day 1) and then remained low. Notch-2 was the principal Notch receptor expressed in monolayer, whereas Notch-3 was expressed most highly in cell aggregates. However, during chondrogenesis, there was a general decline in Notch receptor expression (Fig. 2). Jag-1 was the only Notch ligand expressed significantly by hMSC, and in cell aggregates it showed a most interesting pattern of gene expression, as it increased by 3.5-fold (p ⬍ .05) transiently to a peak at day 2 before falling to the baseline level at day 4 (Fig. 2). There was no detectable change in the expression of the other Notch ligands, which were all expressed at a low level in the cell aggregate cultures.


669

Jag-1 Protein Was Localized in Regions of Future Chondrocyte Differentiation in hMSC Aggregate Cultures

Figure 2. Gene expression of the Notch receptors and Notch ligands in human bone marrow stromal cells during cell aggregate culture. RNA extracted from cell aggregates at days 0 –12 was reverse-transcribed and analyzed by quantitative real-time polymerase chain reaction (gray columns, monolayers; black columns, cell aggregates normalized to GAPDH). Values represent mean ⫾ SEM (n ⫽ 3). Statistical analysis was carried out for comparison of gene expression values between monolayer and cell aggregate cultures. ⴱ, p ⬍ .05; ⴱⴱ, p ⬍ .001. Abbreviations: GAPDH, glyceraldehyde-3-phosphate dehydrogenase; Jag-1, Jagged-1.

The transient change in gene expression of Jag-1 observed early in culture was also reflected in the expression of Jag-1 protein. Immunolocalization (Fig. 3A) showed only weak staining for Jag-1 at day 1, but staining was much stronger at day 2 within central regions of the pellet. By day 4, the Jag-1 staining was reduced. The transient increase in Jag-1 immunolocalization was supported by a quantitative increase in Jag-1 protein determined in total protein lysates of cell aggregates, which peaked on day 2 of culture (Fig. 3B). Histology of cell aggregates at the end of 14 days showed that Jag-1 localization at day 2 was in areas that were subsequently strongly chondrogenic (Fig. 3C). In these regions, the cells had a rounded morphology and a glycosaminoglycan-rich, Safranin O-stained matrix with strong immunolocalization of type II collagen. In contrast, the outer edge of the cell aggregates was not chondrogenic, as the cells had a flattened morphology and the ECM was less developed, and it was negative for collagen type II and glycosaminoglycan. Therefore, Jag-1 protein expression at day 2 was most active in the central region of the aggregates, where there was subsequent differentiation and deposition of a cartilage-like matrix. The transient rise in Jag-1 expression during the first days of cell aggregate culture suggested that it would provide signals by interaction with Notch receptors on adjacent cells. To seek evidence of active signaling, we carried out gene expression analysis of downstream target genes HES-1, HES-5, HEY-1, and HEY-2 [26] (Fig. 4A). The expression of HES-1 was increased 3.7-fold (p ⬍ .01) on the day 0 of cell aggregate culture, which may have been correlated with the increase in Notch-1 at day 0. As with Notch-1, within 24 hours, the expression of HES-1 had returned to that found in monolayer. HES-5 showed some increased expression on days 2 and 4, but in subsequent experi-

Figure 3. Jag-1 protein expression in human bone marrow stromal cell aggregate cultures. (A): Immunolocalization of Jag-1 protein in cell aggregates at days 1–5. Jag-1 immunostaining was strongest at days 2 and 3, concurrent with raised gene expression, and was present within the central region. (B): Immunoblotting for Jag-1 in extracts of cell aggregates from day 0 to day 5 of culture. (C): Chondrogenic cell aggregate cultures at day 14 were stained with Safranin O and immunostained for the extracellular matrix (ECM) proteins collagen I and collagen II. Cartilage-like ECM was in central areas, which earlier in culture stained strongly for Jag-1 protein. Abbreviations: GAPDH, glyceraldehyde-3-phosphate dehydrogenase; Jag-1, Jagged-1.

www.StemCells.com

670


Figure 4. Gene expression of downstream target genes of Notch signaling. (A): The gene expression of HES-1, HES-5, HEY-1, and HEY-2 was analyzed by quantitative real-time polymerase chain reaction (PCR) analysis from monolayer human bone marrow stromal cells and from cell aggregates at days 0 – 6 (gray columns, monolayers; black columns, cell aggregates). Gene expression was normalized to GAPDH and values represent mean ⫾ SEM (n ⫽ 3). Statistical analysis was carried out for comparison of gene expression values between monolayer and cell aggregate cultures. ⴱ, p ⬍ .05; ⴱⴱ, p ⬍ .01. (B): The gene expression of HES-1 and HEY-1 was analyzed by quantitative real-time PCR analysis of cell aggregates that had been cultured for up to 5 days in chondrogenic medium with either DAPT (50 nM) (gray columns) or dimethyl sulfoxide (DMSO) vehicle control (black columns). Gene expression was normalized to GAPDH, and values represent mean ⫾ SEM (n ⫽ 3). Statistical analysis was carried out for comparison of gene expression between DAPT-treated and control cell aggregates at each time point. ⴱ, p ⬍ .05. (C): The wet mass of cell aggregates was determined at days 1, 7, and 14. DMSO controls (black columns), DAPT 0 –14-day treatment (gray columns), DAPT 0 –5-day treatment (stippled columns), and DAPT 5–14-day treatment (hatched columns). Values represent mean ⫾ SEM (n ⫽ 4). Statistical analysis was carried out for comparison of mass of cell aggregates between control and DAPTtreated aggregates at each time point. ⴱ, p ⬍ .05; ⴱⴱ, p ⬍ .001. Abbreviations: DAPT, N-[N-(3,5-difluorophenacetyl-L-alanyl)]-(S)phenylglycine t-butyl ester; GAPDH, glyceraldehyde-3-phosphate dehydrogenase.

ments, this was not consistently present. However, the expression of HEY-1 showed a more interesting pattern, as it increased almost sixfold from day 0 to day 2 (p ⬍ .05) and subsequently fell. This transient increase in HEY-1 thus occurred at the same time as the transient increase of Jag-1, and it was thus likely to be a target of Jag-1 activation of Notch signaling. These results provided evidence that Notch signaling was active within the first 1–3 days of chondrogenic cell aggregate culture and prior to the upregulation of cartilage matrix genes, such as collagen type II and aggrecan. The decrease in Jag-1 and HEY-1 following the transient activation also suggested that the early Notch signaling ceased before chondrogenesis proceeded.

DAPT Inhibition of Notch Signaling To test whether the transient Notch signaling was required for chondrogenesis to occur, cell aggregates were cultured in the presence of DAPT (50 nM) for the initial 5 days, or from day 5,

or throughout the 14 days of culture. The results showed that chondrogenesis was inhibited by DAPT even if it was present for only the first 5 days of culture, as the cell aggregates were only 37% of the wet mass of controls at 14 days (p ⬍ .05) and were similar to those cultured with DAPT continuously for 14 days. This inhibition of chondrogenesis was confirmed by gene expression analysis (Fig. 4B), which showed that at day 2, HEY-1 gene expression was only 4% (p ⬍ .05) of the control, and day 0 HES-1 expression was 7% (p ⬍ .05) of the control. In contrast, DAPT added on days 5–14 had little effect on chondrogenesis (Fig. 4C). The DAPT, if present in early culture, thus inhibited the activation of the downstream targets of Notch signaling. These results implied that the early transient Notch activation in hMSC provided important signals, which were necessary before full chondrogenesis could proceed, but that once activated, further Notch signaling was not required.


Figure 5. Jag-1 transduction of hMSC and the effect on chondrogenesis. (A): Jag-1 gene expression (quantitative reverse transcriptionpolymerase chain reaction analysis) in transduced hMSC (gray column, Jag-1; black column, GFP control). Gene expression was normalized to GAPDH, and values represent mean ⫾ SEM (n ⫽ 3). ⴱ, p ⬍ .05. (B): Jag-1 protein expression analyzed in cell lysates at 24 hours posttransduction by immunoblot analysis. Total protein for each sample was normalized to GAPDH protein by densitometric analysis and then immunoblotted for Jag-1 to confirm expression and correct expressed size of the full-length Jag-1 protein. (C): The mass (wet weight) of control and Jag-1-transduced cells (passage 3) in cell aggregate culture was determined at days 1, 7, and 14 (black columns, control cells; gray columns, Jag-1 cells). Values represent mean ⫾ SEM (n ⫽ 7). Statistical analysis was carried out for comparison of wet mass of cell aggregates between control and Jag-1-transduced cells at each time point. ⴱ, p ⬍ .001. Abbreviations: GAPDH, glyceraldehyde-3-phosphate dehydrogenase; GFP, green fluorescent protein; hMSC, human bone marrow stromal cells; Jag-1, Jagged-1.

Continued Expression of Jag-1 Inhibited Chondrogenesis in hMSC As Jag-1 was identified as the most actively expressed Notch ligand, we investigated the effect of its continued expression on the chondrogenesis in hMSC. To obtain extended expression over at least 14 days, a human Jag-1 adenoviral expression vector was constructed, and hMSC transduced with pAdloxhJag-1 showed increased expression of hJag-1, by 45-fold (p ⬍ .05) compared with the pAdlox-GFP-transduced controls (Fig. 5A). Expression of Jag-1 protein was confirmed in transduced cells (Fig. 5B), and the bioactivity of the hJag-1 encoded in the pAdlox-hJag-1 was confirmed in additional control experiments demonstrating Notch activation in a cell line (CHO/N2-luc) that expressed a full-length Notch-2 receptor and a CBF-1-driven luciferase reporter construct (results not shown) [48]. The continuous expression of hJag-1 in transduced hMSC over 14 days in cell aggregate culture strongly inhibited chondrogenesis. The aggregates formed were dramatically smaller than controls at day 14. The cells failed to acquire a rounded morphology and remained as an aggregated mass of cells with little ECM. There was no significant Safranin O staining of the ECM, and immunostaining for collagen type II and aggrecan was weak (Fig. 6). The GFP-transduced control cells in aggregates formed a cartilage-like matrix that was comparable to that formed by nontransduced hMSC. The absence of histological www.StemCells.com

671

Figure 6. Histology of Jag cell aggregates. Control and Jag-transduced cells were grown in chondrogenic cell aggregate culture. At 14 days, tissue sections were stained for glycosaminoglycan and extracellular matrix protein deposition with Safranin O and by immunolocalization with antibodies to collagen I and collagen II. Jag human bone marrow stromal cells showed poor chondrogenesis compared with control cells. Original magnification, ⫻40. Scale bar ⫽ 50 ␮m. Abbreviations: GFP, green fluorescent protein; Jag, Jagged-1.

evidence of chondrogenesis with Jag-1-expressing cells was supported by the small size and the low wet mass of the cell aggregates, which was one-sixth that of the GFP controls. Gene expression analysis confirmed that Jag-1 expression in transduced cells was maintained throughout the time course up to 14 days (p ⬍ .05) (Fig. 7), and expression of the Notch target HEY-1 was eight times higher than in the control at the end of 14 days in culture. In the Jag-1-transduced cells, there was thus direct evidence of continued Notch signaling throughout the 14 days and the results showed that this inhibited chondrogenesis. Expression analysis of the genes associated with chondrogenesis showed that in GFP-transduced control cells, the gene expression for cartilage matrix genes collagen II and aggrecan increased; together with the histochemical data, this increase demonstrated that adenoviral transduction did not prevent active chondrogenesis. Although the level of SOX9 in the controls was clearly sufficient to support chondrogenesis, in these control cultures, SOX-9 gene expression declined during the 14-day culture period, in contrast to results with previous controls, which showed a small increase (Fig. 1). The expression of SOX-9 in Jag-1-transduced cell pellets was significantly reduced early in culture (p ⬍ .05) (Fig. 7). In the Jag-1-transduced cells, the poor development of cartilage ECM was correlated with reduced collagen type II and aggrecan expression, which were 1.5% and 15% of controls, respectively (p ⬍ .05). In contrast, there was an increase in collagen type I gene expression from only 5% of the control at day 0 (p ⬍ .05) to 10-fold higher than controls at day 14 (p ⬍ .05). The activation of continuous Notch signaling through Jag-1 expression in hMSC


672

Figure 7. Gene expression analysis of Jagged-1 (Jag-1)-transduced human bone marrow stromal cells in monolayer and in cell aggregates. Control and Jag-1-transduced cells were grown in chondrogenic cell aggregate culture. RNA extracted from transduced cells in monolayer and from cell aggregates at days 0, 1, 7, and 14 was analyzed by quantitative reverse transcription-polymerase chain reaction Gene expression was normalized to GAPDH, and values represent mean ⫾ SEM (n ⫽ 3). Statistical analysis was carried out for comparison of gene expression values between control and Jag-1 cell aggregates at each time point. ⴱ, p ⬍ .083; ⴱⴱ, p ⬍ .05; ⴱⴱⴱ, p ⬍ .01; ⴱⴱⴱⴱ, p ⬍ .001. Abbreviations: GAPDH, glyceraldehyde-3phosphate dehydrogenase; GFP, green fluorescent protein.

thus strongly inhibited chondrogenic differentiation and the production and assembly of a cartilage-like ECM.

DISCUSSION The characterization of the 3D cell aggregate system provided a more quantitative analysis of chondrogenesis in hMSC [20] and emphasized the massive upregulation of collagen type II during the early phase of culture. The system also completed within 14 days the deposition and assembly of an extensive ECM, which resulted in a large increase in size and mass of the cell aggregates and was a direct reflection of the biological function of chondrocytes. The system thus permitted the analysis of fully functional chondrocyte differentiation. Although there were large increases in collagen type II and aggrecan expression during chondrogenesis, SOX9 showed only a modest (twofold) increase in expression, and in one instance, it showed a decrease. As SOX9 is reported to be essential for the function of chondrocytes [56 – 61], the lack of a consistent rise was surprising. However, this may imply that SOX9 expression in the hMSC in monolayer was already close to the level required for chondrogenesis and that activation of other factors, such as SOX13 [62] and PGC-1␣ [63], may be important in the control of SOX9 activity. Investigation of the temporal gene expression profiles of the Notch receptors during the 14-day program of chondrogenesis

showed Notch-3 to have the highest expression, but the expression of all Notch receptors was less in cell aggregates than in monolayer culture. Notch-1 showed a small transient rise in expression on day 0 of aggregate formation, and although there was no corresponding rise in Notch ligand expression, there was increased HES-1 expression day 0, which suggested that there was Notch activation. It is unclear whether these day 0 changes were related to subsequent chondrogenesis or were a consequence of lifting cells from monolayer into 3D aggregate cultures. Previously, Notch-1 has been localized to chondroprogenitor cells in the surface zone of articular cartilage in bovine and murine tissue [36, 37], but Notch-1 was not significantly expressed after day 0 in the chondrogenic cultures of the hMSC. The transient rise in Jag-1 in hMSC between days 1 and 4, with a corresponding rise in Hey-1, showed clear evidence of active Notch signaling early in chondrogenesis. There are known links of mutations in the Jag-1 gene to syndromes with skeletal abnormalities. Jag-1 mutations are responsible for Alagille syndrome, which is an autosomal dominant disorder that causes multiorgan defects in the eyes, kidneys, heart, and liver and skeletal abnormalities in the spine and skull [64 –73]. This suggests that Jag-1-mediated Notch signaling is necessary for many differentiation processes, including events in skeletal formation, and in this study, the inhibition of chondrogenesis by DAPT confirmed other studies suggesting that Notch activation is important in early chondrogenesis.

Oldershaw, Tew, Russell et al. In embryonic development, Jag-1-mediated Notch signaling is involved in many differentiation processes, including bone formation by both intramembranous and endochondral ossification routes. Jag-1 was shown to be negatively regulated by Gli3, such that its gene expression was restricted along the anterior limb margin during anterior-posterior limb development [74]. From these results, Jag-1-mediated Notch signaling was proposed to maintain the chondroprogenitor phenotype, which implicated Jag-1 in controlling chondrogenic differentiation at this key point in differentiation. This possibility was tested in the differentiation of hMSC in this study, which showed that by infecting cells with a Jag-1 adenovirus to maintain Jag-1 expression, it caused sustained active Notch signaling, which strongly inhibited chondrogenic differentiation. Increased Notch signaling has been shown to block other differentiation pathways, including in the nervous system and in muscle, and consequently has been suggested to provide a checkpoint in differentiation [75, 76]. Furthermore, in an analysis of the interactions of stem cells in human bone marrow [77], it was demonstrated that parathyroid-related protein-mediated signaling upregulated Jag-1 expression on the surface of osteoblasts, activating Notch signaling to hematopoietic stem cells (HSCs), and this resulted in increased HSC numbers. It was therefore suggested that Jag-1-mediated Notch signaling between different cell populations within the bone marrow niche maintained the self-renewal of HSCs and prevented further differentiation. Therefore, in this study, the prevention of chondrogenic differentiation in hMSC mediated by sustained Jag-1 expression may have been caused by Notch signaling retaining a prechondrogenic state.

673

showed that Notch signaling through Jag-1 was active only transiently during the initial stages of chondrogenic differentiation and that continued active Notch signaling inhibited chondrogenesis. Although sustained Notch signaling appeared to inhibit hMSC chondrogenesis, the inhibition of early Notch signaling by DAPT showed that it was necessary for full chondrogenesis to proceed. The transient Notch signaling initiated when the cells form a cell aggregate thus appeared to function as a trigger required to prime the cells for chondrogenesis.

ACKNOWLEDGMENTS We acknowledge Prof. Cay Kielty (UK Centre for Tissue Engineering, University of Manchester, Manchester, U.K.) for kindly providing the human placental RNA, Dr. Grahame McKenzie (Lorantis, Ltd., Cambridge, U.K.) for the kind gift of the CHO/N2-luc cells, Prof. Peter Clegg (University of Liverpool, Liverpool, U.K.) for advice on statistical analysis, and Dr. Giles Hardingham (University of Edinburgh, Edinburgh, U.K.) for critical review of the manuscript. We acknowledge the support from Biotechnology and Biological Sciences Research Council, Medical Research Council, and Engineering and Physical Sciences Research Council (U.K.) to the U.K. Centre for Tissue Engineering and the Wellcome Trust support for the Wellcome Trust Centre for Cell-Matrix Research, University of Manchester. Support for the NW Embryonic Stem Cell Centre, funded by NW Science Fund, is also acknowledged.

DISCLOSURE

CONCLUSION The present work, focusing on chondrogenesis by human bone marrow stromal cells in a defined in vitro system,

14

2 3 4 5 6 7 8 9 10 11 12 13

DeLise AM, Fischer L, Tuan RS. Cellular interactions and signaling in cartilage development. Osteoarthritis Cartilage 2000;8:309 –334. Hardingham T. Articular cartilage. In: Isenberg D, Maddison P, Woo P et al., ed. Oxford Textbook of Rheumatology. Oxford, U.K.: Oxford University Press, 2004:325–334. Hunziker EB. Articular cartilage repair: Basic science and clinical progress. A review of the current status and prospects. Osteoarthritis Cartilage 2002;10:432– 463. Hunziker EB, Michel M, Studer D. Ultrastructure of adult human articular cartilage matrix after cryotechnical processing. Microsc Res Tech 1997;37:271–284. Knudson CB, Knudson W. Cartilage proteoglycans. Semin Cell Dev Biol 2001;12:69 –78. Kuettner KE. Biochemistry of articular cartilage in health and disease. Clin Biochem 1992;25:155–163. Sandell LJ, Aigner T. Articular cartilage and changes in arthritis. An introduction: Cell biology of osteoarthritis. Arthritis Res 2001;3: 107–113. Luyten FP. Mesenchymal stem cells in osteoarthritis. Curr Opin Rheumatol 2004;16:599 – 603. Sittinger M, Hutmacher DW, Risbud MV. Current strategies for cell delivery in cartilage and bone regeneration. Curr Opin Biotechnol 2004; 15:411– 418. Tuan RS, Boland G, Tuli R. Adult mesenchymal stem cells and cellbased tissue engineering. Arthritis Res Ther 2003;5:32– 45. Heng BC, Cao T, Lee EH. Directing stem cell differentiation into the chondrogenic lineage in vitro. STEM CELLS 2004;22:1152–1167. Johnstone B, Hering TM, Caplan AI et al. In vitro chondrogenesis of bone marrow-derived mesenchymal progenitor cells. Exp Cell Res 1998; 238:265–272. Mackay AM, Beck SC, Murphy JM et al. Chondrogenic differentiation

www.StemCells.com

CONFLICTS

The authors indicate no potential conflicts of interest.

REFERENCES 1

OF POTENTIAL OF INTEREST

15 16 17 18 19 20 21 22 23 24 25

26

of cultured human mesenchymal stem cells from marrow. Tissue Eng 1998;4:415– 428. Murdoch AD, Grady LM, Ablett MP et al. Chondrogenic differentiation of human bone marrow stem cells in transwell cultures: Generation of scaffold-free cartilage. STEM CELLS 2007;25:2786 –2796. Pittenger MF, Mackay AM, Beck SC et al. Multilineage potential of adult human mesenchymal stem cells. Science 1999;284:143–147. Li X, Jin L, Cui Q et al. Steroid effects on osteogenesis through mesenchymal cell gene expression. Osteoporos Int 2005;16:101–108. Fukuda K. Use of adult marrow mesenchymal stem cells for regeneration of cardiomyocytes. Bone Marrow Transplant 2003;32:S25–S27. Gojo S, Gojo N, Takeda Y et al. In vivo cardiovasculogenesis by direct injection of isolated adult mesenchymal stem cells. Exp Cell Res 2003; 288:51–59. Gojo S, Umezawa A. Plasticity of mesenchymal stem cells—Regenerative medicine for diseased hearts. Hum Cell 2003;16:23–30. Barry F, Boynton RE, Liu B et al., Chondrogenic differentiation of mesenchymal stem cells from bone marrow: Differentiation-dependent gene expression of matrix components. Exp Cell Res 2001;268:189 –200. Sekiya I, Koopman P, Tsuji K et al. Dexamethasone enhances SOX9 expression in chondrocytes. J Endocrinol 2001;169:573–579. Artavanis-Tsakonas S, Rand MD, Lake RJ. Notch signaling: Cell fate control and signal integration in development. Science 1999;284: 770 –776. Brennan K, Gardner P. Notching up another pathway. Bioessays 2002; 24:405– 410. Schweisguth F. Regulation of notch signaling activity. Curr Biol 2004; 14:R129 –R138. Buxbaum JD, Liu KN, Luo Y et al. Evidence that tumor necrosis factor alpha converting enzyme is involved in regulated alpha-secretase cleavage of the Alzheimer amyloid protein precursor. J Biol Chem 1998;273: 27765–27767. Iso T, Kedes L, Hamamori Y. HES and HERP families: Multiple effectors of the Notch signaling pathway. J Cell Physiol 2003;194:237–255.


674

27 Jarriault S, Brou C, Logeat F et al. Signalling downstream of activated mammalian Notch. Nature 1995;377:355–358. 28 Mumm JS, Schroeter EH, Saxena MT et al. A ligand-induced extracellular cleavage regulates gamma-secretase-like proteolytic activation of Notch1. Mol Cell 2000;5:197–206. 29 Nakagawa O, McFadden DG, Nakagawa M et al. Members of the HRT family of basic helix-loop-helix proteins act as transcriptional repressors downstream of Notch signaling. Proc Natl Acad Sci U S A 2000;97: 13655–13660. 30 Qi H, Rand MD, Wu X et al. Processing of the notch ligand delta by the metalloprotease Kuzbanian. Science 1999;283:91–94. 31 Schroeter EH, Kisslinger JA, Kopan R. Notch-1 signalling requires ligand-induced proteolytic release of intracellular domain. Nature 1998; 393:382–386. 32 Shimizu K, Chiba S, Hosoya N et al. Binding of Delta1, Jagged1, and Jagged2 to Notch2 rapidly induces cleavage, nuclear translocation, and hyperphosphorylation of Notch2. Mol Cell Biol 2000;20:6913– 6922. 33 Struhl G, Adachi A. Requirements for presenilin-dependent cleavage of notch and other transmembrane proteins. Mol Cell 2000;6:625– 636. 34 Crowe R, Zikherman J, Niswander L. Delta-1 negatively regulates the transition from prehypertrophic to hypertrophic chondrocytes during cartilage formation. Development 1999;126:987–998. 35 Watanabe N, Tezuka Y, Matsuno K et al. Suppression of differentiation and proliferation of early chondrogenic cells by Notch. J Bone Miner Metab 2003;21:344 –352. 36 Dowthwaite GP, Bishop JC, Redman SN et al. The surface of articular cartilage contains a progenitor cell population. J Cell Sci 2004;117: 889 – 897. 37 Hayes AJ, Dowthwaite GP, Webster SV et al. Archer CW. The distribution of Notch receptors and their ligands during articular cartilage development. J Anat 2003;202:495–502. 38 Karlsson C, Jonsson M, Asp J et al. Notch and HES5 are regulated during human cartilage differentiation. Cell Tissue Res 2007;327:539 –551. 39 Vujovic S, Henderson SR, Flanagan AM et al. Inhibition of gammasecretases alters both proliferation and differentiation of mesenchymal stem cells. Cell Prolif 2007;40:185–195. 40 Fujimaki R, Toyama Y, Hozumi N et al. Involvement of Notch signaling in initiation of prechondrogenic condensation and nodule formation in limb bud micromass cultures. J Bone Miner Metab 2006;24:191–198. 41 Dovey HF, John V, Anderson JP et al. Functional gamma-secretase inhibitors reduce beta-amyloid peptide levels in brain. J Neurochem 2001;76:173–181. 42 Li Y, Tew SR, Russell AM et al. Transduction of passaged human articular chondrocytes with adenoviral, retroviral, and lentiviral vectors and the effects of enhanced expression of SOX9. Tissue Eng 2004;10: 575–584. 43 Martin I, Jakob M, Schafer D et al. Quantitative analysis of gene expression in human articular cartilage from normal and osteoarthritic joints. Osteoarthritis Cartilage 2001;9:112–118. 44 Livak KJ, Schmittgen TD. Analysis of relative gene expression data using real-time quantitative PCR and the 2(-Delta Delta C(T)) method. Methods 2001;25:402– 408. 45 Tew SR, Li Y, Pothacharoen P et al. Retroviral transduction with SOX9 enhances re-expression of the chondrocyte phenotype in passaged osteoarthritic human articular chondrocytes. Osteoarthritis Cartilage 2005;13: 80 – 89. 46 Oda T, Elkahloun AG, Meltzer PS et al. Identification and cloning of the human homolog (JAG1) of the rat Jagged1 gene from the Alagille syndrome critical region at 20p12. Genomics 1997;43:376 –379. 47 Hardy S, Kitamura M, Harris-Stansil T et al. Construction of adenovirus vectors through Cre-lox recombination. J Virol 1997;71:1842–1849. 48 Wong KK, Carpenter MJ, Young LL et al. Notch ligation by Delta1 inhibits peripheral immune responses to transplantation antigens by a CD8⫹ cell-dependent mechanism. J Clin Invest 2003;112:1741–1750. 49 Mastrogiacomo M, Cancedda R, Quarto R. Effect of different growth factors on the chondrogenic potential of human bone marrow stromal cells. Osteoarthritis Cartilage 2001;9:S36 –S40. 50 Solchaga LA, Penick K, Porter JD et al. FGF-2 enhances the mitotic and chondrogenic potentials of human adult bone marrow-derived mesenchymal stem cells. J Cell Physiol 2005;203:398 – 409. 51 Bell DM, Leung KK, Wheatley SC et al. SOX9 directly regulates the type-II collagen gene. Nat Genet 1997;16:174 –178. 52 Huang W, Zhou X, Lefebvre V et al. Phosphorylation of SOX9 by cyclic AMP-dependent protein kinase A enhances SOX9’s ability to transactivate a Col2a1 chondrocyte-specific enhancer. Mol Cell Biol 2000;20: 4149 – 4158.

53 Liu Y, Li H, Tanaka K et al. Identification of an enhancer sequence within the first intron required for cartilage-specific transcription of the alpha2(XI) collagen gene. J Biol Chem 2000;275:12712–12718. 54 Zhang P, Jimenez SA, Stokes DG. Regulation of human COL9A1 gene expression. Activation of the proximal promoter region by SOX9. J Biol Chem 2003;278:117–123. 55 Blokzijl A, Dahlqvist C, Reissmann E et al. Cross-talk between the Notch and TGF-beta signaling pathways mediated by interaction of the Notch intracellular domain with Smad3. J Cell Biol 2003;163:723–728. 56 Akiyama H, Chaboissier MC, Martin JF et al. The transcription factor Sox9 has essential roles in successive steps of the chondrocyte differentiation pathway and is required for expression of Sox5 and Sox6. Genes Dev 2002;16:2813–2828. 57 Bi W, Deng JM, Zhang Z et al. Sox9 is required for cartilage formation. Nat Genet 1999;22:85– 89. 58 Kolettas E, Muir HI, Barrett JC et al. Chondrocyte phenotype and cell survival are regulated by culture conditions and by specific cytokines through the expression of Sox-9 transcription factor. Rheumatology (Oxford) 2001;40:1146 –1156. 59 Kou I, Ikegawa S. SOX9-dependent and -independent transcriptional regulation of human cartilage link protein. J Biol Chem 2004;279: 50942–50948. 60 Lefebvre V, Behringer RR, de Crombrugghe B. L-Sox5, Sox6 and Sox9 control essential steps of the chondrocyte differentiation pathway. Osteoarthritis Cartilage 2001;9:S69 –S75. 61 Stokes DG, Liu G, Dharmavaram R et al. Regulation of type-II collagen gene expression during human chondrocyte de-differentiation and recovery of chondrocyte-specific phenotype in culture involves Sry-type highmobility-group box (SOX) transcription factors. Biochem J 2001;360: 461– 470. 62 Wang Y, Ristevski S, Harley VR. SOX13 exhibits a distinct spatial and temporal expression pattern during chondrogenesis, neurogenesis and limb development. J Histochem Cytochem 2006;54:1327–1333. 63 Kawakami Y, Tsuda M, Takahashi S et al. Transcriptional coactivator PGC-1alpha regulates chondrogenesis via association with Sox9. Proc Natl Acad Sci U S A 2005;102:2414 –2419. 64 Alagille D, Odievre M, Gautier M et al. Hepatic ductular hypoplasia associated with characteristic facies, vertebral malformations, retarded physical, mental, and sexual development, and cardiac murmur. J Pediatr 1975;86:63–71. 65 Crosnier C, Attie-Bitach T, Encha-Razavi F et al. JAGGED1 gene expression during human embryogenesis elucidates the wide phenotypic spectrum of Alagille syndrome. Hepatology 2000;32:574 –581. 66 Jones EA, Clement-Jones M, Wilson DI. JAGGED1 expression in human embryos: Correlation with the Alagille syndrome phenotype. J Med Genet 2000;37:658 – 662. 67 Joutel A, Tournier-Lasserve E. Notch signalling pathway and human diseases. Semin Cell Dev Biol 1998;9:619 – 625. 68 Kamath BM, Bason L, Piccoli DA et al. Consequences of JAG1 mutations. J Med Genet 2003;40:891– 895. 69 Kamath BM, Loomes KM, Oakey RJ et al. Facial features in Alagille syndrome: Specific or cholestasis facies? Am J Med Genet 2002;112: 163–170. 70 Kamath BM, Stolle C, Bason L et al. Craniosynostosis in Alagille syndrome. Am J Med Genet 2002;112:176 –180. 71 Krantz ID. Alagille syndrome: Chipping away at the tip of the iceberg. Am J Med Genet 2002;112:160 –162. 72 Krantz ID, Piccoli DA, Spinner NB. Clinical and molecular genetics of Alagille syndrome. Curr Opin Pediatr 1999;11:558 –564. 73 Loomes KM, Underkoffler LA, Morabito J et al. The expression of Jagged1 in the developing mammalian heart correlates with cardiovascular disease in Alagille syndrome. Hum Mol Genet 1999;8:2443–2449. 74 McGlinn E, van Bueren KL, Fiorenza S et al. Pax9 and Jagged1 act downstream of Gli3 in vertebrate limb development. Mech Dev 2005; 122:1218 –1233. 75 Martinez Arias A, Zecchini V, Brennan K. CSL-independent Notch signalling: A checkpoint in cell fate decisions during development? Curr Opin Genet Dev 2002;12:524 –533. 76 Nofziger D, Miyamoto A, Lyons KM et al. Notch signaling imposes two distinct blocks in the differentiation of C2C12 myoblasts. Development 1999;126:1689 –1702. 77 Calvi LM, Adams GB, Weibrecht KW et al. Osteoblastic cells regulate the haematopoietic stem cell niche. Nature 2003;425:841– 846.

See www.StemCells.com for supplemental material available online.