Human Adipose-Derived Mesenchymal Stem Cells: Direction to a Phenotype Sharing Similarities with the Disc,Gene Expression Profiling,and Coculture with Human Annulus Cells

Abstract

Biologic therapies for disc degeneration hold great promise as an emerging concept. Due to ease of harvest and abundance, adipose derived-mesenchymal stem cells (AD-MSC) are a readily available cell source for such therapies. Our objectives in this study were (1) to develop/validate methods to harvest AD-MSC and direct them to a disc-like phenotype by three-dimensional (3D) culture and transforming growth factor (TGF)-β3 exposure, (2) to assess cell phenotypes with gene expression profiling for these human AD-MSC and annulus cells, and (3) to test whether disc cell-AD-MSC coculture could augment glycosaminoglycan (GAG) production. When AD-MSC were exposed to TGF-β3, greater extracellular matrix was formed containing types I and II collagen, keratan sulfate, and decorin. Biochemical GAG measurement showed that production was significantly greater in TGF-β3-treated AD-MSC in 3D culture versus untreated controls (p < 0.05). Gene expression patterns in AD-MSC were compared to annulus cells; 4424 genes were significantly upregulated, and 2290 genes downregulated. Coculture resulted in a 44% greater GAG content compared with AD-MSC or annulus culture alone (p = 0.04). Data indicated that human AD-MSC can successfully be manipulated in 3D culture to express gene products important in the disc, and that coculture of annulus cells with AD-MSC enhances total GAG production.

Introduction

Low back pain is a primary cause of disability and plays a major role in this country's medical, social, and economic structure. As shown in studies by Boden et al.,^1,2 even healthy, asymptomatic individuals are at risk. During degeneration and aging of the intervertebral disc, cell numbers decrease, apoptosis and senescence occur, the extracellular matrix (ECM) is not appropriately remodeled, and the ECM fails; dehydration and matrix fraying culminate in the formation of tears within the annulus during functional compressive and torsional loading.^3–5 The nucleus and disc material rupture through these tears, impinging on nerves and causing pain.

Recent advances have directed attention toward the exciting potential of biologic therapies for the treatment of disc degeneration. One of the major, well-recognized events in disc aging and degeneration is the marked reduction in cell number,⁶ and the failure of this dwindling cell population to appropriately maintain the disc ECM. With other researchers today, we share enthusiasm for the potentials that cell-based tissue engineering, including employment of MSC, offers toward the augmentation of disc cell numbers and prevention of deleterious events in the aging/degenerating disc, such as apoptosis and cell senescence. In this realm, our lab's focus is upon adult, adipose derived-mesenchymal stem cells (AD-MSC) because of their ease of harvest and abundance from humans, and their potential autologous cell applications.

AD-MSC have the ability to differentiate into adipocytes, chondrocytes, and osteoblasts, and have been shown to differentiate into insulin-, somatostatin-, and glucagon-expressing cells.^7–9 Several bioactive agents function in these differentiation processes, including transforming growth factor (TGF)-β, fibroblast growth factor, and insulin-like growth factor.^10–12

The structural integrity of the disc requires a complex interaction between disc cells and the ECM they produce and remodel. In this study, our objectives were (1) to develop and validate methods to harvest AD-MSC, to direct them to a disc-like phenotype in three-dimensional (3D) culture, and test the effect of added TGF-β3, (2) to test how well this direction succeeded by comparing gene expression profiling patterns of AD-MSC in 3D culture to patterns seen in human annulus cells in 3D culture, and (3) to test whether annulus cell coculture with AD-MSC could be used to augment proteoglycan production.

Materials and Methods

Clinical study population

Experimental study of disc specimens was approved prospectively by the authors' Human Subjects Institutional Review Board. Since disc surgeries are routinely performed in our institution, and surgically removed tissue discarded, informed consent was not required.

Scoring of disc degeneration utilized a modification of the Thompson scoring system¹³ incorporating author ENH's radiologic, MRI, and surgical findings. Disc degeneration was graded over the spectrum from a healthy disc (Thompson grade I) to discs with advanced degeneration (grade V, the most advanced stage of degeneration). Patient specimens were derived from surgical disc procedures performed on individuals with herniated discs and degenerative disc disease. Surgical specimens were transported to the laboratory in a sterile tissue culture medium. Care was taken to remove all granulation tissue and to sample only disc tissue. Nonsurgical control donor disc specimens were obtained via the National Cancer Institute Cooperative Human Tissue Network (CHTN); they were shipped overnight to the laboratory in a sterile tissue culture medium and processed as described below. Specimen procurement from the CHTN was included in our approved protocol by our human subjects Institutional Review board.

Adipose specimens were obtained after IRB approval from our medical center cosmetic and plastic surgery physicians; panniculectomy specimens were immediately collected from the surgical suite, taken to the laboratory, and processed as described below for AD-MSC isolation and characterization.

Clinical demographic features of both the adipose tissue donors and subjects from whom annulus cells were derived are presented in Table 1. Annulus cells were P2 or P3 when used in experiments; AD-MSC were P1 when used in experiments.

Table 1.

Demographic Features of Subjects

Subject no.	Site	Thompson score	Age/gender	Other information
Disc cell cultures
1^a	L5-S1 Annulus	I	23/Male	Disc surgical specimen
2^a	T5-T6 Annulus	II	45/Female	Disc surgical specimen
3^a	L2-L3 Annulus	III	52/Female	CHTN Disc–Cardiac Dysfunction
4^a	L5-S1 Annulus	III	36/Female	Disc surgical specimen
5^a	L5-S1 Annulus	III	31/Male	Disc surgical specimen
6^a	L2-L3 Annulus	III	46/Female	Disc surgical specimen
7^a	L5-S1 Annulus	III	28/Female	Disc surgical specimen
8^b	L5-S1 Annulus	II	21/Male	Disc surgical specimen
9^b	L1-2	IV	55/Female	Disc surgical specimen
10^b	L5-S1 Annulus	IV	40/Female	Disc surgical specimen
11^b	L5-S1 Annulus	III	47/Female	Disc surgical specimen
12^b	L5-S1 Annulus	III	25/Female	L5-S1 Annulus
AD-MSC cultures
1	—	—	59/Female	Panniculectomy
2	—	—	65/Female	Panniculectomy^c
3	—	—	49/Female	Panniculectomy
4	—	—	69/Female	Panniculectomy

Used in three-dimensional culture microarray analyses.

Used in coculture experiments with adipose derived-mesenchymal stem cells (AD-MSC).

Cells used in two coculture experiments.

Specific methodologies for human disc cell culture

Methods used here for disc cell culture have previously been reported from our laboratory.¹⁴ Briefly, disc tissue was rinsed twice with sterile modified minimal essential medium with Earle's salts (MEM; Gibco) with 1% (v/v) L-glutamine (Irvine Scientific), 1% (v/v/) nonessential amino acids (Irvine Scientific), and 1% (v/v) penicillin–streptomycin (Irvine Scientific), with the addition of 20% (v/v) fetal bovine serum (Gibco). Regions of annulus were visually identified and minced with a scalpel into small portions (1–2 mm²). Fragments were cultured in T25 flasks (Costar) and anchored by placement of a sterile nylon mesh (Spectra Mesh; Spectra Laboratory Products) over the fragments. Cells then grew out from the fragments, and were fed three times a week. When cultures showed a confluent out-growth of cells under the mesh, cultures were trypsinized (1:250; trypsin 0.5g/L and EDTA 0.2 g/L; Irvine Scientific) and split with a ratio of 1:4 for subculturing.

Specific methodologies for AD-MSC

Methods used here for AD-MSC isolation, monolayer culture, 3D culture in collagen sponge matrix, and validation of the stem cell phenotype have previously been reported from our laboratory.¹⁵ Major details are summarized briefly below.

AD-MSC isolation and plating followed the method of Cowan et al.,¹⁶ and usually took ∼2 h depending upon the amount of fat tissue being processed. Briefly, human adipose tissue was digested with collagenase type II (1 mg/mL; Sigma) at 37°C in a water-bath shaker for 30–40 min (180–200 rpm) with a brief vortex every 10 min. Undigested tissue was removed by filtration, and AD-MSC harvested by centrifugation at 42 g for 5 min at RT. AD-MSC were resuspended in Hanks balanced salt solution (HBSS), filtered through a 40 μm cell strainer, counted, and plated as the primary culture (P0) on 100 × 20 mm round plastic tissue culture dishes (Primera; Falcon, BD Biosciences) at a density of 1000 cells/mm². In our laboratory, the average yield of AD-MSC is 471,000 cells/mL. Plated cells were fed every 48–72 h with 10 mL of the medium (mesenchymal stem cell basal medium [MSCBM]; Cambrex Bio Science). When confluent, cells were trypsinized, centrifuged at 42 g for 5 min, and re-plated at a density of 1000 cells/mm².

Verification of stem cell isolation

Isolated presumptive AD-MSC were verified as being stem cells by CD marker analysis, osteogenic differentiation, and chondrogenic differentiation as described below.

CD marker analysis of AD-MSC

AD-MSC were characterized by immunolocalization of the multipotent mesenchymal stromal cell markers as follows: positive localizations for CD105, CD90, CD44, and CD29, and negative localization for CD45 and CD34. This CD marker MSC characterization followed patterns as listed by the International Society for Cellular Therapy.¹⁷

Osteogenic differentiation

Osteogenic differentiation of stem cells was performed using an Osteogenesis Kit (Chemicon International)¹⁸ and employed positive alizarin red (Osteogenesis Kit) staining of mineralized matrix after 21 days of culture. Control cultures were fed MSCBM medium only.

Chondrogenic differentiation

Chondrogenic differentiation used micromass culture with cells grown for 7–10 days in the chondrogenic induction medium (Cambrex Bio Science) supplemented with 5% fetal calf serum. Proteoglycan production in the ECM was observed by toluidine blue staining (0.1% in dH₂O; Sigma) in paraffin-embedded specimens.

Stimulation of AD-MSC with TGF-β3

Growth and differentiation of stem cells in Gelfoam^®

Sterile Gelfoam (referred to here as collagen sponge; Pharmacia & Upjohn Co.) was used as a 3D scaffold.¹⁹ AD-MSC were suspended in MSCBM at 1 × 10⁷ cells/mL concentration. Ten microliters containing 1 × 10⁵ cells was injected into collagen sponges trimmed into 0.5 cm³ cubes. Replicate collagen sponges were placed on Costar Transwell Clear Inserts (Corning Incorporated-Life Sciences) in 24-well plates and fed three times per week with 2.0 mL of MSCBM with TGF-β3 (10 ng/mL; Lonza) or without TGF-β3 (control). About 10 ng/mL TGF-β3 is the typical dose used in the literature for chondrogenic differentiation.¹¹ Cells were grown for 2–6 weeks and assayed for proteoglycan production in the presence or absence of TGF-β3 using the 1,9-dimethylmethylene blue (DMB) assay (see below). Additional constructs containing cells were fixed in 10% neutral-buffered formalin for 1 h and embedded in paraffin. Gelfoam was sectioned for immunohistochemical analysis and stained for ECM proteoglycan production using toluidine blue (0.1% in dH₂O; Sigma). Proteoglycan production was quantitatively measured using the DMB assay (see below).

Assay of total sulfated glycosaminoglycan production

Cells were grown in 3D culture for 14 days in the presence or absence of TGF-β3 and assayed for concentrations of total sulfated proteoglycan using the DMB assay.²⁰ DMB (Aldrich) solution was prepared and stored at room temperature in an amber bottle. Collagen sponge constructs containing cells were rinsed two times in HBSS (Gibco) and transferred to separate 1.5 mL microcentrifuge tube. Collagenase type V (Sigma) in HBSS (Gibco) was added to produce a final concentration of 3.0 units/mL. Microcentrifuge tubes were incubated at 37°C until the sponge was completely digested (15–30 min). Samples were assayed in duplicate by mixing 0.1 mL of sample with 1.25 mL DMB solution was added to all samples and standards (0–7.5 μg chondroitin sulfate; Sigma). All tubes were incubated at room temperature in the dark for 30 min followed by centrifugation at room temperature for 15 min at 10,000 rpm. One milliliter of the supernatant was removed and the absorbency read at 595 nm using a Beckman DU 600 spectrophotometer (Fullerton). Formate buffer (pH 3.1) is used as a blank. Sample concentrations are determined from the standard curve, and results expressed as micrograms glycosaminoglycan (GAG)/construct.¹⁹

Immunohistochemistry

Immunohistochemical localizations were carried out on sections of paraffin-embedded 3D cultures as previously described^14,21 to identify type I collagen (20 μg/mL; Biodesign International), type II collagen (20 μg/mL; Biodesign International), decorin (25 μg/mL; R&D Systems), keratan sulfate (5 μg/mL; Seikagaku Corporation), and chondroitin sulfate (20 μg/mL; ICN Biomedicals). Negative controls consisted of Rabbit IgG (for collagen I and II; Dako Cytomation) or mouse IgG (for all other antibodies; Dako Cytomation) used at the same concentration as each antibody. The secondary antibody employed the LSAB2 Biotinylated Secondary kit (Dako Cytomation) (10 min.) Localizations utilized peroxidase-conjugated streptavidin (Dako Cytomation) for 10 min and DAB (Dako Cytomation) for 5 min for observation. Slides were counterstained with light green, dehydrated, cleared, and mounted with a resinous mounting medium. Negative controls were processed minus primary antibodies.

Verification of AD-MSC isolation using cell surface markers¹⁷ was performed on paraffin-embedded specimens using the following antibodies: positive localization for stemness was verified by positive localization of CD105 (8 μg/mL; Thermo Scientific), CD90 (3.125 μg/mL; BD Biosciences), CD44 (8 μg/mL; Thermo Scientific), and CD29 (9 μg/mL; Thermo Scientific), and negative localization of CD45 (3.5 μg/mL; Dako Cytomation) and of CD34 (1.5 μg/mL; Dako Cytomation). Immunolocalizations were performed using the LSAB2 Dako detection system as described above.

Gene expression studies and analyses

Three AD-MSC 3D cultures and seven 3D cultures of annulus cells were utilized (Table 1). AD-MSC were cultured using our basal conditions with addition of TGF-β3; disc cells were cultured in 3D as previously described.¹⁹ Cells were cultured for 2 weeks, RNA harvested according to instructions with the Trizol isolation method, checked for quality using the 2100 Bioanalyzer (Agilent Technologies, Inc.), reverse-transcribed to double-stranded cDNA, subjected to two rounds of transcription, and hybridized to the DNA microarray in the Affymetrix Fluidics Station 400. Affymetrix human U133 X3P arrays were used.

Coculture experiments

Disc and AD-MSC grown in 3D scaffold coculture were characterized by cell labeling for separation and observation. Disc cells were labeled before coculture by incubation (10 min) with carboxyfluorescein succinimidyl ester (CFSE; 10 μM) fluorescent-label as previously described.²² AD-MSC were identified by immunolocalization of the CD44 cell surface marker (as described above) since this marker was not present on disc cells. Replicate samples of AD-MSC, CFSE-labeled disc cells, or premixed AD-MSC and CFSE-labeled disc cells were injected into absorbable collagen scaffold (50,000 AD-MSC and 50,000 annulus cells [to total 100,000 cells] were injected into 0.5 cm³ cubes). Three-dimensional cultures and cocultures were cultured on Costar Transwell Clear Inserts (Corning) for 2 weeks. Concentrations of total sulfated proteoglycans were measured using the DMB assay as described above.

Predicted proteoglycan production in the coculture experiments was calculated using the method of LeVisage et al.²³ For coculture experiments, data are means ± s.e.m. based upon five coculture experiments run in triplicate.

Statistical analysis

Statistical analysis of data other than the microarray studies utilized standard methods using SAS software (version 8.2; SAS Institute). Methods used included unpaired t-tests and Spearman's correlation statistics. A p-value of <0.05 was considered statistically significant. Data are expressed as means ± SEM (n) or ±SD (n) as noted in text and figures. Statistical analysis of gene expression: GeneSifter™ Web-based software (VizX Labs) was used to analyze all microarray data. Using GC-RMA (Robust multiarray average), Affymetrix “.cel” files were uploaded to the GeneSifter Web site and normalized. Using the Student's t-test (two tailed, unpaired), statistical significance was determined (p < 0.05). A correction factor for false discovery rate was applied using the Benjamini and Hochberg method.²⁴ For filtering purposes, the fold change was set at 2.0 or greater. Gene ontologies (GO) were generated by GeneSifter based on the Gene Ontology Consortium.²⁵ z-scores were also generated by GeneSifter and indicate any biological significance. A z-score of ≥2 or ≤−2 suggest that genes in a particular GO were altered in expression compared to what would be expected by chance.²⁶

Results

Verification of stem cell isolation

Human AD-MSC attached well to the standard plastic culture substrate and proliferated well (Fig. 1). Cells exhibited a spindle-shaped morphology. An important part of our experimental design for this work included verification of isolation of MSC. Traditional accepted methods were utilized for this: direction of stem cells to osteogenic and chondrogenic phenotypes (Fig. 2) and immunocytochemical marker localization according to the International Society for Cellular Therapy guidelines¹⁷ (see Materials and Methods section) (Fig. 3). Figure 2 illustrates proteoglycan production in micromass culture, confirming direction to a chondrogenic phenotype (Fig. 2A, B), and CD marker verification of the presence of AD-MSC (positive localizations for CD105, CD90, CD44, and CD29, and negative localization for CD45 and CD34) (Fig. 3).

FIG. 1.

Human adipose derived-mesenchymal stem cells (AD-MSC) in monolayer culture show a spindle-shaped morphology. Original magnification: ×155.

FIG. 2.

Verification of stem cell status of isolated presumptive AD-MSC was confirmed by culture under specific conditions that directed cells to a chondrogenic phenotype (A, B; pink matrix denotes high proteoglycan content of the extracellular matrix [toluidine blue staining]) and to an osteogenic phenotype (C, shown by production of a mineralized matrix, stained red with alizarin red). Control AD-MSC cultured under normal conditions show no matrix mineralization with alizarin red staining (D). Original magnifications: A ×95, B ×240, and C and D ×105. Color images available online at www.liebertonline.com/ten.

FIG. 3.

AD-MSC cultures showed the appropriate immunocytologic localizations of stem cell CD markers characteristic of MSC.¹⁷ Isolated AD-MSC showed the appropriate positive localizations for CD105 (A), CD44 (B), CD29 (C), and CD90 (D), and the appropriate negative localization of CD45 (E) and CD34 (F). (G) A representative negative control processed without the primary antibody. Original magnification: ×145. Color images available online at www.liebertonline.com/ten.

AD-MSC can be directed to a phenotype that shares similarities with disc cells by 3D culture in the presence of TGF-β3

When cultured in the 3D microenvironment in the presence of TGF-β3, AD-MSC attached well and grew well. ECM was also abundant (Fig. 4A). ECM components included type II collagen (Fig. 4B) and chondroitin sulfate (Fig. 4C). To further quantitatively assess proteoglycan production in the presence of TGF-β3, the DMB method was used to assess total sulfated proteoglycans. TGF-β3 culture significantly enhanced proteoglycan production (Fig. 5; p < 0.05).

FIG. 4.

(A) Toluidine-blue-stained section of AD-MSC cultured in three-dimensional (3D) in the presence of transforming growth factor (TGF)-β3. Note the presence of pink staining indicating proteoglycan content in the extracellular matrix. (B, C) Immunohistochemical localization of type II collagen and chondroitin sulfate in 3D cultured cells without (control) or with TGF-β3. (The “Control” image shows cultures grown in the absence of TGF-β. The “Negative control” image shows a representative section processed without the primary antibody.) Original magnifications: A × 240, B and C × 400. Color images available online at www.liebertonline.com/ten.

FIG. 5.

TGF-β3-treated AD-MSC show significantly greater proteoglycan content in 3D cultures compared with controls (p < 0.05) Data are means ± SD, n = 5. Color images available online at www.liebertonline.com/ten.

Comparison of gene profiles of AD-MSC with TGF-β3 versus human annulus cells in 3D culture

Gene expression patterns of AD-MSC in 3D culture in the presence of TGF-β3 were determined and compared to profiles of human annulus cells in 3D culture. GeneSifter was used with application of a correction factor to account for false discovery rates, and for this analysis the fold change level was set at 2.0 or greater. In this comparison of AD-MSC in the presence of TGF-β3 to annulus cells, 4424 genes were found to be significantly upregulated, and 2290 genes downregulated. Table 2 summarizes the gene findings, presenting the gene identified ID, the fold change, and adjusted p-value.

Table 2.

Gene Findings for Human Adipose Derived-Mesenchymal Stem Cells (AD-MSC) in the Presence of Transforming Growth Factor-β Compared to Genes Expressed by Human Disc Cells from Discs of Grades I, II, and III

Gene name	Gene identifier	Ratio/fold change	Direction	Adjusted p-value^a	Ontology ^b
Genes involved in extracellular matrix production, cell adhesion and binding
Aggrecan	NM_013227	69.33	Down	0.00028	EM; BM; CCA
Decorin	AI336924	3.1	Down	0.047	EM
Biglycan	AA845258	8.18	Up	2.69E-05	EM
Epiphycan	NM_004950	8.85	Up	0.041	EM
Cartilage intermediate layer protein 2	BC034926	9.17	Up	0.0014	EM
Cartilage oligomeric matrix protein	NM_000095	10.41	Up	0.024	EM
Cartilage intermediate layer protein, nucleotide pyrophosphohydrolase	NM_003613	37.39	Up	0.014	EM
Collagen, type I, alpha 1	K01228	11.29	Up	0.0036	EM; C
Collagen, type I, alpha 2	AA628535	4.42	Up	0.0080	EM; C
Collagen, type III, alpha 1	AU146808	11.93	Up	0.038	EM; C; CSA
Collagen, type V, alpha 1	AI983428	4.47	Up	0.021	EM; C
Collagen, type V, alpha 3	NM_015719	6.28	Up	0.038	EM; C; CSA
Collagen, type VIII, alpha 2	AI806793	4.36	Up	0.0013	EM; CCA; BM
Collagen, type X, alpha 1	X98568	91.66	Up	4.82E-06	EM; C
Collagen, type XII, alpha 1	AU146651	9.57	Up	0.011	EM; C
Collagen, type XIV, alpha 1	BF449063	15.19	Down	0.0083	EM; C; CCA
Collagen, type XVI, alpha 1	NM_001856	3.35	Up	0.0064	EM; C
Collagen, type XVIII, alpha 1	NM_030582	12.13	Up	0.035	EM; C
Collagen, type XXI, alpha 1	NM_030820	23.13	Down	0.0013	EM
Collagen triple helix repeat containing protein 1	AA584310	3.42	Up	3.82E-04	EM
Versican	R94644	10.52	Up	0.00029	EM; CA
Brevican	AI205180	2.15	Down	0.0070	EM: CA
Laminin, alpha 1	AI990816	7.57	Up	8.20E-04	EM; LC; BL; BM; CA
Laminin, alpha 2 (merosin)	NM_000426	14.23	Up	0.041	EM; LC; BL; BM; CA
Laminin, alpha 4	BC004241	2.68	Down	0.022	EM; LC; BL; BM; CA
Laminin, alpha 5	BC003355	8.36	Down	0.0065	EM; LC; BL; BM; CA
Fibronectin 1	AJ276395	2.12	Up	0.0096	EM; CA
Fibronectin leucine-rich transmembrane protein 2	AI188161	3.76	Down	0.0089	EM; CA
Hyaluronan and proteoglycan link protein 1	U43328	17.96	Down	8.35E-04	EM; CA
Matrilin 2	NM_002380	41.59	Down	0.00016	EM
Matrilin 3	NM_002381	11.7	Up	0.0023	EM
Nidogen 1	NM_002508	6.18	Up	0.0057	EM; BM; CSA
Heparan sulfate proteoglycan 2	M85289	2.78	Up	0.049	EM; BM; CA
Elastin	AA479278	4.58	Up	6.56E-04	EM
Elastin microfibril interfacer 1	NM_007046	6.8	Up	0.030	EM
Microfibrillar-associated protein 2	NM_017459	3.65	Up	0.0065	EM; F
Microfibrillar-associated protein 4	R72286	9.8	Up	0.00016	EM; F; CA
Spondin 1, extracellular matrix protein	AB018305	4.32	Up	0.027	EM
Spondin 2, extracellular matrix protein	NM_012445	289.27	Up	1.82E-04	EM
Fibulin 2	NM_001998	8.3	Up	0.0012	EM
Fibulin 5	NM_006329	29.03	Up	0.0013	EM; CSA
Fibulin 6 (Hemicentin 1)+ A35	BF446673	5.37	Up	0.0092	EM
Lectin, galactoside-binding, soluble, 3 binding protein	NM_005567	15.64	Up	0.023	EM; ES; CA
SPARC-like 1 (mast9, hevin)	NM_004684	19.73	Up	4.89E-05	EM
SPARC-related modular calcium binding 2	AB014737	38.73	Up	0.0027	EM
SPARC-related modular calcium binding 1	AJ249900	69.46	Down	0.0016	EM; BM
Proline/arginine-rich end leucine-rich repeat protein	AA573140	22.13	Down	0.0086	EM
Surfactant, pulmonary-associated protein C	AI831055	2.2	Down	6.69E-04	EM
Dermatopontin	NM_001937	47.86	Up	1.23E-04	EM
Lysyl oxidase	NM_002317	3.8	Down	0.0042	EM; C
Lysyl oxidase-like 1	NM_005576	2.05	Up	0.040	EM
Sarcoglycan, gamma (35 kDa dystrophin-associated glycoprotein)	NM_000231	10.13	Up	0.020	SC
Sarcoglycan, beta (43 kDa dystrophin-associated glycoprotein)	NM_000232	6.44	Down	0.021	SC
Sarcoglycan, epsilon	NM_003919	3.94	Down	0.0051	SC; CSA
Sarcospan (Kras oncogene-associated gene)	AL136756	2.37	Up	0.016	CA
Protocadherin 7	NM_002589	4.79	Up	0.000024	CCA
Protocadherin 9	NM_020403	13.49	Up	0.013	CCA
Protocadherin 12	NM_016580	5.14	Up	0.027	CCA
Protocadherin 19	AB037734	11.21	Up	0.0058	CCA
Cadherin 2, type 1, N-cadherin (neuronal)	NM_001792	4.54	Up	0.047	CCA
Cadherin 4, type 1, R-cadherin (retinal)	NM_001794	3.81	Up	0.009	CCA
Cadherin 5, type 2, VE-cadherin (vascular epithelium)	NM_001795	23.69	Up	0.0016	CCA
Cadherin 13, H-cadherin (heart)	NM_001257	4.68	Down	0.0008	CSA; CCA; ES
Integrin, alpha 3 (antigen CD49C, alpha 3 subunit of VLA-3 receptor)	NM_002204	4.22	Down	0.016	CSA
Integrin, alpha 7	AK022548	6.34	Down	0.040	CSA
Integrin, beta 8	BF513121	13.41	Down	0.0065	CSA
Integrin, alpha 10	AF112345	24.54	Down	0.0002	CSA
integrin, alpha 11	AF109681	3.54	Up	0.026	CSA
Integrin-linked kinase-2	NM_004517	6.51	Up	0.019	CSA
Thrombospondin 3	L38969	4.66	Up	0.005	CSA
Podocan-like 1	NM_024825	7.67	Up	0.0025	EM
Podocan	AW015571	7.81	Up	0.0033	EM
Chordin	AF283325	3.17	Up	1.75E-02	ES
Genes involved in the breakdown or inhibition of the extracellular matrix
Matrix metallopeptidase 9 (gelatinase B, 92 kDa gelatinase, 92 kDa type IV collagenase)	NM_004994	558.13	Up	5.79E-05	EM; ES
Matrix metallopeptidase 11 (stromelysin 3)	NM_005940	64.05	Up	6.93E-06	EM
Matrix metallopeptidase 16 (membrane-inserted)	U79292	13.57	Up	0.028	EM
Matrix metallopeptidase 1 (interstitial collagenase)	NM_002421	21.54	Up	0.037	EM
Matrix metallopeptidase 2 (gelatinase A, 72 kDa gelatinase, 72 kDa type IV collagenase)	NM_004530	7.43	Up	0.012	EM; ES
Matrix metallopeptidase 12 (macrophage elastase)	NM_002426	6.8	Up	0.0080	EM
Matrix metallopeptidase 14 (membrane-inserted)	Z48481	5.03	Up	0.024	EM; CSA
Hyaluronoglucosaminidase 1	AF173154	6.05	Down	3.67E-02	ES
ADAM metallopeptidase with thrombospondin type 1 motif, 7	AW007289	3.3	Up	0.0096	EM
ADAM metallopeptidase with thrombospondin type 1 motif, 1	AF060152	29.23	Down	0.0020	EM
ADAMTS-like 1	BF111214	4.41	Down	0.0076	EM
ADAM metallopeptidase with thrombospondin type 1 motif, 9	AI431730	10.05	Up	5.38E-04	EM
ADAM metallopeptidase with thrombospondin type 1 motif, 12	W74476	30.47	Up	0.0066	EM
ADAM metallopeptidase with thrombospondin type 1 motif, 16	AI740544	2.66	Up	0.020	EM
Matrix Gla protein	AW512787	2.02	Up	0.021	EM; CCA
Endomucin	AF205940	8.97	Up	0.0006	CCA
Genes involved in inhibiting extracellular matrix metalloproteinases
TIMP metallopeptidase inhibitor 2	NM_003255	2.07	Up	0.0029	EM; BM
Genes involved in cell proliferation and growth regulation
Glypican 4	AF064826	8.4	Up	0.013	EM
Glypican 6	AF111178	2.01	Up	0.033	EM
Transforming growth factor, beta 1	BC000125	7.61	Up	0.0063	EM; CCA; ES
Transforming growth factor, beta 2	M19154	22.34	Down	0.0015	CCA
Transforming growth factor, beta 3	J03241	8.22	Up	0.0041	CCA
Transforming growth factor, beta receptor III	AW193698	7.54	Down	2.68E-04	ES
Genes encoding growth factors
Bone morphogenetic protein 1	NM_006132	8.33	Up	0.0013	CCA; ES
Bone morphogenetic protein 2	AA583044	3.84	Down	0.031	CCA; ES
Bone morphogenetic protein 4	D30751	6.83	Down	0.0058	EM; ES
Bone morphogenetic protein receptor, type IB	AA935461	6.08	Up	0.0032	CCA
EGF-containing fibulin-like extracellular matrix protein 1	NM_004105	3.05	Down	0.029	EM
EGF-containing fibulin-like extracellular matrix protein 2	AB030655	2.42	Up	0.027	EM; BM
Fibroblast growth factor binding protein 1	NM_005130	10.53	Down	1.28E-02	ES
Fibroblast growth factor 2 (basic)	NM_002006	5.24	Down	1.66E-03	ES
Fibroblast growth factor 5	AB016517	6.35	Down	8.38E-03	ES
Insulin-like growth factor binding protein 1	NM_000596	21.47	Down	1.59E-02	ES
Insulin-like growth factor binding protein 3	M31159	2.31	Down	8.89E-04	ES
Epidermal growth factor receptor (erythroblastic leukemia viral (v-erb-b) oncogene homolog, avian)	AF277897	11.84	Down	0.013	CCA; ES
Latent transforming growth factor beta binding protein 1	NM_000627	3.33	Down	0.0044	EM
Brain-derived neurotrophic factor	NM_001709	8.25	Down	0.00223	ER
Genes involved in angiogenesis/vascular development
Angiogenin, ribonuclease, RNase A family, 5	AI761728	6.78	Down	0.0083	EM; BL; BM; ES
Vascular endothelial growth factor A	AF022375	3.09	Down	0.0024	EM; ES
Vascular endothelial growth factor C	U58111	2.02	Down	4.15E-02	ES
Angiopoietin-like 2	NM_012098	7.55	Up	1.46E-02	ES
Angiopoietin 2	NM_001147	18.12	Up	2.47E-03	ES
Angiopoietin-like 4	NM_016109	19.62	Down	3.29E-04	EM; ES
Fms-related tyrosine kinase 1 (vascular endothelial growth factor/vascular permeability factor receptor)	AA058828	2.84	Up	4.42E-02	ES
Proprotein convertase subtilisin/kexin type 5	AU152579	6.03	Up	2.16E-02	ES
Genes involved in inflammatory and immune response
Chitinase 3-like 2	U58515	158.31	Down	1.16E-03	ES
Chitinase 1 (chitotriosidase)	NM_003465	19.80	Up	4.60E-03	ES
Interleukin 8	NM_000584	7.68	Down	1.11E-02	ES
Interleukin 15	NM_000585	6.70	Down	2.84E-03	ES
Chemokine (C-X-C motif ) ligand 6 (granulocyte chemotactic protein 2)	NM_002993	199.18	Down	8.90E-05	ES
Chemokine (C-X-C motif ) ligand 5	AK026546	2.06	Up	1.69E-02	ES
Chemokine (C-X-C motif ) ligand 3	NM_002090	20.20	Down	6.14E-03	ES
Chemokine (C-X-C motif ) ligand 2	M57731	11.38	Down	7.22E-03	ES
Chemokine (C-X-C motif ) ligand 16	AF275260	5.24	Up	1.57E-02	ES
Chemokine (C-X-C motif ) ligand 14	AF144103	21.48	Up	3.57E-03	ES
Chemokine (C-X-C motif ) ligand 1 (melanoma growth stimulating activity, alpha)	NM_001511	46.78	Down	7.70E-05	ES
Chemokine (C-C motif ) ligand 4	NM_002984	9.24	Up	5.35E-05	ES
Chemokine (C-C motif ) ligand 3	NM_002983	47.11	Up	4.18E-06	ES
Chemokine (C-C motif ) ligand 18 (pulmonary and activation-regulated)	AB000221	188.17	Up	1.00E-06	ES
Chemokine (C-C motif ) ligand 11	D49372	2.87	Up	3.70E-02	ES
Small inducible cytokine subfamily E, member 1 (endothelial monocyte-activating)	NM_004757	2.42	Down	1.00E-02	ES
Tumor necrosis factor receptor superfamily, member 11b (osteoprotegerin)	NM_002546	201.5	Down	8.44E-04	EM
Tumor necrosis factor (ligand) superfamily, member 4 (tax-transcriptionally activated glycoprotein 1, 34 kDa)	NM_003326	16.9	Up	2.77E-04	ES

p-Value is adjusted for false discovery rate.

Classification by GeneSifter™: ES, extracellular space; EM, extracellular matrix; CSA, cell–substrate adhesion; CCA, cell–cell adhesion; C, collagen; LC, laminin complex; BL, basal lamina; BM, basement membrane; F, fibril; CA, cell adhesion; SC, sarcoglycan complex.

GO analyses with GeneSifter were performed to classify and group gene changes in Table 2. Major classifications were chosen that encompassed genes associated with (1) ECM, cell adhesion, and cell binding; (2) ECM catabolism; (3) genes inhibiting ECM metalloproteinases; (4) cell proliferation and growth regulation; (5) growth factors; (6) angiogenesis and vascular development; and (7) inflammation and the immune response.

For ECM-related genes, GeneSifter identified 85 genes that were upregulated, and 48 downregulated. Relevant upregulated genes included biglycan, versican, alpha 1, fibronectin 1, elastin, chitinase 3-like 1 (cartilage glycoprotein-39), hevin, SPARC, thrombospondin 3, bone morphogenetic protein (BMP)-1 and BMP receptor, vascular endothelial growth factor A, and alpha 1 chains of collagens type I, III, V, XVIII, VIII, X, XII, matrix metalloproteinase (MMP)-9, -11, -16, -1, -2, -14, several ADAMTS, lysyl oxidase, and protocadherin. Important downregulated genes were integrin alpha-7, -8, -10, -11, TNF receptor superfamily 11b, BMP-2 and -4, Matrilin 2, matrix Gla protein, aggrecan, and decorin.

GeneSifter software was used to generate the z-scores for the Table 2 GO and results are shown in Table 3. z-scores are used to determine any biological significance; z-scores ≥2 or ≤−2 indicate an alteration of gene expression compared to what would be expected by chance. As shown in Table 3, ontologies with a significant finding of overexpressed upregulated genes were ECM (z-score, 4.30), collagen (z-score, 2.23), and cell adhesion (z-score, 5.96). The cell adhesion ontology contained the following subterms that were also significant: cell–cell adhesion (z-score, 2.63) and cell–substrate adhesion (z-score, 2.32). Within the downregulated genes, there was also overexpression representation in ontology groups of extracellular space (z-score, 2.88), basal lamina (z-score, 2.06), and sarcoglycan complex (z-score, 2.19).

Table 3.

Z-scores^a for Gene Ontologies Represented in Table 2 (Stem Cells Exposed to Transforming Growth Factor-β3 Compared to Disc Cells)

Ontology	Genes	Up	Down	Array genes ^b	z-up	z-down
Cellular Component
Extracellular Space	114	58	56	462	−0.57	2.88
Extracellular Matrix	93	65	28	297	4.30	0.61
Basement Membrane	16	10	6	52	1.22	0.80
Collagen	11	9	2	34	2.23	−0.54
Basal Lamina	5	2	3	12	0.33	2.06
Laminin Complex	4	2	2	7	1.17	1.91
Fibril	2	2	0	6	1.43	−0.74
Sarcoglycan Complex	3	1	2	6	0.23	2.19
Biological Process
Cell Adhesion	209	146	63	687	5.96	0.83
Cell–Cell Adhesion	72	48	24	248	2.63	0.78
Cell–Substrate Adhesion	33	20	13	91	2.32	2.07

Gene ontologies having a z-score ≥2 or ≤−2 are in bold, and signify that more genes in those ontology groups were altered in expression compared to what would be expected.

z-Scores were calculated using GeneSifter software.

Array genes indicate the number of genes on the array for each ontology.

Comparison of AD-MSC gene expression profiles in the presence or absence of TGF-β3

We also were interested in determining the extent to which TGF-β3 exposure in 3D culture modified the gene expression patterns of AD-MSC. Affymetrix gene analyses identified 69 genes that were upregulated and 81 that were downregulated. Table 4 presents the major GeneSifter findings for all genes with organization according to ontology classification by genes involved in development and differentiation; ECM production, cell adhesion, and binding; ECM catabolism; pro- and antiapoptotic genes, cell motility and migration, cell proliferation, and inflammatory/immune responses; coagulation; energy production and ion exchange; and signaling pathways, vesicle trafficking, and oxidative stress.

Table 4.

Gene Findings for Human AD-MSC in the Presence of Transforming Growth Factor-β Compared to Genes Expressed in the Absence of Transforming Growth Factor-β

Gene name	Gene identifier	Ratio/fold change	Direction	Adjusted p-value^a	Ontology ^b
Genes involved in development and differentiation
AF4/FMR2 family, member 3	AW085505	3.54	Up	0.038677	DP; IO: CMP
Basic helix-loop-helix domain containing, class B, 3	BE857425	2.05	Up	0.0395	DP; IO
Cytokine receptor-like factor 1	NM_004750	5.37	Up	0.041253	ES
Distal-less homeobox 2	NM_004405	8.59	Up	0.035968	DP; IO; CMP
FAT tumor suppressor homolog 3 (Drosophila)	AI283093	6.59	Up	0.041947	DP; CCA; M
Forkhead box O1	AW117498	3	Up	0.034003	DP; IO; C; CMP; ST
Frizzled homolog 8 (Drosophila)	AB043703	3.51	Up	0.048668	DP; M; ST
Glia maturation factor, gamma	NM_004877	5.8	Down	0.035968	IO; C; CMP
Glycoprotein, alpha-galactosyltransferase 1	AI972498	5.2	Down	0.048302	IO; C
Laminin, beta 3	L25541	2.14	Down	0.033929	DP; EM; BM; CA
LIM domain only 2 (rhombotin-like 1)	NM_005574	4.57	Down	0.045048	DP; IO
LIM homeobox 6	NM_014368	7.31	Down	0.035968	DP; CMP
Lymphoid-restricted membrane protein	NM_006152	4.04	Down	0.033929	DP; IO; C; M
Neuregulin 1	NM_013957	4.59	Up	0.048668	DP; IO; CD; CMP; M; ST
SHC SH2-domain binding protein 1	NM_024745	3.97	Down	0.037231	SDB
Paired-like homeodomain 2	NM_000325	5.53	Up	0.033929	DP; IO; CMP
Pentraxin-related gene, rapidly induced by IL-1 beta	NM_002852	3.08	Down	0.041253	DP; CMP
Protein tyrosine phosphatase, receptor type, N polypeptide 2	AF007555	3.99	Down	0.049025	IO; C; CMP
Genes involved in extracellular matrix production, cell adhesion, and binding
Carbohydrate (N-acetylglucosamine 6-O) sulfotransferase 6	AF280086	4.19	Up	0.048668	IO; C; CMP; M
EGF-like repeats and discoidin I-like domains 3	AA053711	4.62	Up	0.037231	DP; CA
Epiphycan	NM_004950	30.34	Up	0.044051	EM
Spectrin repeat containing, nuclear envelope 1	AB033088	2.71	Up	0.048668	DP; C
Hemicentin 1	BF446673	3.25	Up	0.046523	EM; BM; CA; CMP; M
Solute carrier family 26 (sulfate transporter), member 2	AI025519	3.46	Up	0.041253	M
Syndecan 1	NM_002997	9.77	Up	0.045048	CS; M
Tumor necrosis factor receptor superfamily, member 11b (osteoprotegerin)	BF433902	5.81	Down	0.045048	DP; EM; ST
Erythrocyte membrane protein band 4.1-like 4A	AU144565	3.76	Down	0.041253	IO; C; M
Sarcospan (Kras oncogene-associated gene)	NM_005086	2.38	Up	0.034003	IO; CA
Genes involved in the breakdown or inhibition of extracellular matrix
Matrix metallopeptidase 13 (collagenase 3)	NM_002427	56.52	Up	0.047022	DP; EM; CMP
Genes involved in inducing apoptosis
NLR family, pyrin domain containing 3	NM_004895	5.05	Down	0.035968	DP; C; CMP; ST
Nucleotide-binding oligomerization domain containing 1	AF126484	2.39	Down	0.037231	DP; C; CMP
PRKC, apoptosis, WT1, regulator	AA632147	2.25	Up	0.037231	DP; IO; C; CMP
Solute carrier family 22, member 17	NM_020372	2.5	Up	0.041253	IO; C
Genes involved in inhibiting apoptosis
Fas apoptotic inhibitory molecule 3	AI084226	18.53	Down	0.041947	DP; IO; C; M
NLR family, apoptosis inhibitory protein	AI817801	2.29	Down	0.048542	DP
Genes involved in cell motility, migration, and actin remodeling
Actin, gamma 2, smooth muscle, enteric	NM_001615	5.66	Down	0.033929	IO; C
Rho GTPase activating protein 25	NM_014882	4.41	Down	0.0395	ST
Rho GTPase activating protein 6	NM_001174	5.86	Down	0.041253	IO; C; ST
Cells involved in cell proliferation and growth regulation
Anillin, actin binding protein	NM_018685	5.12	Down	0.048668	DP; IO; C
Cancer susceptibility candidate 5	BF248364	6.16	Down	0.041253	DP; IO; C; CD
Cell division cycle 2, G1 to S and G2 to M	AL524035	4.97	Down	0.037231	DP; IO; C; CMP
denticleless homolog (Drosophila)	AK001261	3.13	Down	0.028881	IO; C; CMP
E2F transcription factor 1	NM_005225	2.81	Down	0.035968	DP; IO; C; CMP
E2F transcription factor 8	NM_024680	3.56	Down	0.041253	IO; CMP
hepatocyte growth factor (hepapoietin A; scatter factor)	X16323	10.15	Down	0.0395	DP; IO; C; CCA; CD; CMP; ST
PR domain containing 6	AF272898	3.2	Up	0.045423	DP; IO; CMP
Cells involved in inflammatory and immune response
CD209 molecule	AF290886	11.05	Down	0.0395	DP; C; CCA; M; ST
Chemokine (C-C motif ) ligand 2	S69738	3.52	Down	0.047894	DP; ES; CA; ST; CMP; M
Chemokine (C-C motif ) ligand 3	NM_002983	7.64	Down	0.041253	DP; ES
Spleen tyrosine kinase	BF593625	3.94	Down	0.048668	DP; IO; C; CCA; CMP; M; ST
Fc fragment of IgG binding protein	NM_003890	13.77	Down	0.041253	C; CA; M
Lymphocyte cytosolic protein 2 (SH2 domain containing leukocyte protein of 76 kDa)	NM_005565	4.06	Down	0.041253	C; ST
Family with sequence similarity 19 (chemokine (C-C motif )-like), member A5	AV723914	5.45	Up	0.035968	M
Hydrogen voltage-gated channel 1	AW406569	2.4	Down	0.035968	M
Immunoglobulin superfamily containing leucine-rich repeat	NM_005545	2.18	Up	0.033929	CA
X-prolyl aminopeptidase (aminopeptidase P) 2, membrane-bound	NM_003399	6.88	Down	0.0395	CMP; M
Genes involved in blood coagulation and iron transport
Tissue factor pathway inhibitor (lipoprotein-associated coagulation inhibitor)	J03225	6.64	Down	0.041253	IO; C
Hephaestin	NM_014799	2.34	Up	0.028881	M
Genes involved with energy production, metabolism, and ion exchange
ATP/GTP binding protein-like 5	AA045527	3.39	Up	0.041253	IO; C; CMP
ATPase, class I, type 8B, member 1	BG290908	3.48	Down	0.047022	CMP; M
Glycerol-3-phosphate acyltransferase, mitochondrial	AB046780	3.65	Up	0.048668	IO; C; CMP; M
ST3 beta-galactoside alpha-2,3-sialyltransferase 6	AI989567	3.09	Down	0.047022	IO; C; CMP; M
ST6 beta-galactosamide alpha-2,6-sialyltranferase 2	AW004016	30.52	Up	0.031078	DP; IO; C; CMP; M
Rho GDP dissociation inhibitor (GDI) beta	NM_001175	4.7	Down	0.035968	DP; IO; C; CA
Cytochrome P450, family 26, subfamily B, polypeptide 1	NM_019885	4.6	Down	0.047022	DP; IO; C; CMP; M; ST
Homogentisate 1,2-dioxygenase (homogentisate oxidase)	NM_000187	2.1	Up	0.028881	CMP
Oxysterol binding protein-like 5	AW271225	2.18	Up	0.041253	C; CMP
Potassium channel, subfamily K, member 2	AF004711	3.34	Down	0.047022	M; ST
Potassium intermediate/small conductance calcium-activated channel, subfamily N, member 2	NM_021614	2.86	Up	0.048668	M
Potassium large conductance calcium-activated channel, subfamily M, beta member 1	U61536	9.3	Down	0.035968	M
Male sterility domain containing 1	NM_018099	3.9	Down	0.037961	IO; C; M
Frequenin homolog (Drosophila)	AI079596	2.4	Up	0.028881	IO; C; M
Genes involved in signaling pathways
Stathmin 1/oncoprotein 18	NM_005563	2.04	Down	0.041253	IO; C; ST
Contactin-associated protein 1	NM_003632	4.95	Down	0.041253	CA; M; ST
Cytokine inducible SH2-containing protein	D83532	4.24	Down	0.038677	M; ST
Neuronal cell adhesion molecule	NM_005010	2.35	Down	0.041253	DP; CCA; CS; M
FERM domain containing 4B	AU145019	4.85	Down	0.0395	IO; C; M
Genes involved in vesicle trafficking and exocytosis
Synaptotagmin-like 2	N21426	3.24	Down	0.0395	IO; C; M
Exocyst complex component 6	AL137438	2.49	Down	0.048668	C
Genes involved in oxidative stress
Cytoglobin	AL513673	3.06	Down	0.035968	C

p-Value adjusted for false discovery rate.

Classification by GeneSifter; DP, developmental process; IO, intracellular organelle; EM, extracellular matrix; ES, extracellular space; BM, basement membrane; C, cytoplasm; CD, cell development; CA, cell adhesion; CCA, cell–cell adhesion; CMP, cellular metabolic process; SDB, SH2 domain binding; CS, cell surface; ST, signal transduction; M, membrane.

z-Scores that were generated by GeneSifter for the GO represented in Table 4 appear in Table 5. There were two ontologies with a significant finding of overexpressed upregulated genes: ECM (z-score, 2.32) and basement membrane (z-score, 2.18). There were other ontologies with significant overexpression of downregulated genes, including cytoplasm (z-score, 2.30), cell adhesion (z-score, 4.37), developmental process (z-score, 2.37), and SH3 domain binding (z-score, 3.09). The cell adhesion ontology also contained the following subterm, which was significant: cell–cell adhesion (z-score, 2.97).

Table 5.

Z-scores^* for Gene Ontologies Represented in Table 4 (Stem Cells Exposed to Transforming Growth Factor-β (TGF-β3) Compared to Stem Cells without TGF-β3)

Ontology	Genes	Up	Down	Array Genes ^**	z-up	z-down
Cellular Component
Extracellular Space	4	2	2	462	−0.57	0.03
Extracellular Matrix	5	3	2	297	2.32	0.67
Basement Membrane	2	1	1	52	2.18	1.67
Cytoplasm	49	12	37	6613	−2.27	2.30
Intracellular Organelle	51	20	31	8122	−1.16	−0.90
Cell Surface	2	1	1	193	0.59	0.20
Membrane	52	24	28	6129	1.90	0.52
Biological Process
Cell Adhesion	16	6	10	687	3.27	4.37
Cell-Cell Adhesion	5	1	4	248	0.46	2.97
Developmental Process	39	14	25	4101	1.28	2.37
Signal Transduction	23	6	17	3115	−0.82	1.32
Cellular Metabolic Process	41	18	23	6972	0.01	−1.62
Cell Development	4	1	3	473	−0.20	0.76
Molecular Function
SH3 Domain Binding	2	0	2	75	−0.44	3.09

Z- scores were calculated using GeneSifter^TM software. Gene ontologies having a z-score ≥ 2 or ≤ −2 are in bold and signify more genes in those ontology groups were altered in expression than compared to what would be expected.

Array Genes indicates the number of genes on the array for each ontology.

Coculture of AD-MSC with human annulus cells enhances proteoglycan production

Our final experimental study tested whether 3D coculture of human intervertebral disc cells with human AD-MSC would stimulate expression and production of ECM components important in the disc, including proteoglycans. Before coculture, annulus cells were labeled with CFSE (Fig. 6A), which was internalized into the cell cytoplasm and could thus be used as a marker with immunohistochemistry in harvested coculture preparations. Postharvest, AD-MSC could also identified by anti-CD44 immunolocalization (Fig. 6B). After the 2 week coculture of annulus and AD-MSC, both cells types were still present at harvest as evidenced by immunocytochemical localizations (Fig. 6C shows dark immunolocalization of CFSE-containing annulus cells; arrows mark AD-MSC that were not labeled with CFSE before coculture).

FIG. 6.

(A) Carboxyfluorescein succinimidyl ester (CFSE)–labeled human annulus cells in monolayer; (B) anti-CD44 immunolocalization in AD-MSC in 3D culture; (C) in this coculture experiment, annulus cells were labeled with CFSE before coculture so that they could be distinguished from the MSC. CFSE-containing annulus cells in the coculture harvest were then identified with anti-CFSE immunohistochemistry and show a dark brown localization product, whereas AD-MSC are unlabeled and marked with arrows. This microscopic field was chosen so that both types of cells could be clearly seen. Note that in other regions, cells of both types have proliferated and filled in the majority of the sponge cavities. (D) Significantly greater proteoglycan content is seen in 3D in cocultures compared to separate AD-MSC or annulus cultures (*p < 0.05). Original magnifications: A ×155, B ×145, and C ×145. Color images available online at www.liebertonline.com/ten.

Quantitative analysis of proteoglycan production showed significantly greater proteoglycan concentrations in 3D cocultures than in the AD-MSC cultures (p = 0.04, Fig. 6D). As previously described by LeVisage et al., to assess proteoglycan production in coculture, data were expressed as an increase in sulfated proteoglycans compared with the predicted value taken as the average of the individual control stem and disc cell cultures.²³ Three-dimensional coculture of annulus and AD-MSC resulted in a 44% increase in proteoglycan production compared to the predicted value taken as the average of the individual control stem and disc cultures (predicted proteoglycan content [μg/construct] 2.34 ± 1.62 (5), mean ± SEM (n) vs. coculture mean of 3.21 ± 1.57 (5), p < 0.05).

Discussion

In this study, we utilized cells from the annulus fibrosus of the human intervertebral disc. Although maintenance and regulation of proteoglycan synthesis is also critical within the nucleus pulposus, we have chosen annulus cells here because of our specific interest in development of clinically appropriate future biologic therapies for disc degeneration, especially those that might employ autologous disc and MSC or AD-MSC therapies.^27–34 We concur with An et al. and others that for clinical applications, research on the annulus is a logical first step since no therapeutic methods for the nucleus would be likely to be successful in the environment of a compromised annulus tissue.^35,36 Ultimately, it is the ECM of the annulus that fails in disc degeneration; dehydration and matrix fraying culminate in the formation of tears within the annulus during biomechanical loading and torsion. This is believed by many to be the major source of incapacitating low back pain conditions. Additionally, nucleus and annulus disc material may rupture through these tears, causing radicular pain. Annular tears are present in >50% of patients in early adulthood and are almost always present in the elderly.³⁷

Data presented here have shown that AD-MSC were isolated from surgical fat specimens in a manner that retained their multiipotency for osteogenic and chondrogenic differentiation, and their stem cell markers.²³ The AD-MSC grew well in monolayer, and supplementation with TGF-β3 increased GAG production in the 3D microenvironment. Future studies would benefit from utilization of flow cytometry to quantify CD marker profiles of the isolated stem cells, data that the present study did not collect.

Considerable recent research has focused upon AD-MSC, with specific focus upon their ability to be subcultured for considerable periods while retaining their stem cell potentials,³⁸ their capability for rapid expansion,³⁹ and their biologic and functional properties.^40–42

Gene expression profiling showed patterns that characterized the AD-MSC in 3D culture that involved ECM genes, MMPs, BMPs, and integrins (Table 2). Our study of expression profiles for AD-MSC was designed to utilize seven annulus cell cultures derived from Grades I, II, and III discs, thus excluding the patterns that would be present from cells from the most degenerated Thompson grade IV and V discs. These novel data showed differences in the expression patterns of AD-MSC and annulus fibrosus cells in 3D culture—these changes are important as we continue to explore the potential of AD-MSC in biology therapies for disc degeneration.

It has long been recognized that characterization of disc cell phenotypes, and identification of phenotypic changes with aging/degeneration is a high priority research topic,⁴³ and although researchers have not yet derived a list of specific gene expression profiles that fully characterize human disc cells, good progress is being made on this topic.^44–46 As such research continues to advance, we hope that data presented here will be of use as investigators identify key regulatory events that direct AD-MSC to a disc-like phenotype.

In terms of a general overview of data presented here, AD-MSC cultured in TGF-β3 in collagen sponges showed upregulation of several key disc proteoglycans (biglycan, cartilage oligomeric matrix protein, versican, and brevican) and several key collagens, laminin, fibronectin and elastin) (Table 2); these findings speak to the appropriate direction of the AD-MSC cells to a disc-like phenotype. Associated with these findings were concomitant changes in genes involved in ECM turnover (MMPs and ADAMs) that tell us that the in vitro microenvironment was suitable for matrix remodeling (Table 2, second section). Genes encoding growth factors, such as BMP-1, showed upregulation, and a downregulation of the neurotrophin brain-derived neurotrophic factor (Table 2). Additional expression of other important genes regulating other growth factors and their receptors, ECM components, and chemokines was also identified; their investigation will continue as we further study how AD-MSC can be directed in vitro. Our findings in general support the value of TGF-β as a culture component in the direction of AD-MSC (Table 4 and Table 2) (see additional comments below related to TGF-β).

As might be expected, there were differences in gene expression profiles for human annulus cells in 3D compared to profiles of human AD-MSC. The multipotency of the AD-MSC genome prepares for a wide spectrum of gene expression capabilities. Differences were also present in the gene expression profiles for AD-MSC exposed to TGF-β3 compared to profiles of control AD-MSC.

Our 3D culture per se also has provided a specialized microenvironment that also acted in conjunction with TGF-β3 to direct gene expression.^47–49 Understanding how these microenvironments influence the AD-MSC is essential for potential use of these cells in autologous tissue-engineered constructs for biologic therapy of disc degeneration. Researchers hope to be able to utilize specific microenvironments in the future to direct stem cells in a manner that makes them capable of coping with in vivo environments with inherent problems,⁵⁰ such as the avascular intervertebral disc.

The use of specific cytokines is now known to be especially important when trying to direct MSC along specific lines of differentiation so as to maximize gene expression characteristics. MSC appear to require such modulation for differentiation as recently shown by Ng et al. in their work with MSC directed to chondrogenic, adipogenic, and osteogenic lineages.⁵¹ Similar previous work by Sales et al.⁵² and Mehlhorn et al.⁵³ have also illustrated the complexity of transcriptional regulation of MSC direction to specific phenotypes.

Data identified in the gene expression profiles presented here are important first steps in exploring the potential of AD-MSC in tissue engineering applications for the intervertebral disc. Elucidation of the resulting cell signaling networks involved in these culture conditions is important for future research studies. Our literature review showed relatively few gene expression characterization studies for AD-MSC, including the earlier studies of Lee et al.⁵⁴ and the recent work of Peroni et al.⁵⁵ Wagner et al. compared bone marrow MSC with AD-MSC⁵⁶ and found many differentially expressed genes.

Guilak et al. have carried out an interesting clonal analysis of the differentiation potential of human AD-MSC that confirmed that they are a type of multipotent adult stem cell (and not a mixed population of unipotential progenitor cells).⁵⁷

Relatively little is known or understood about the direct cell contact coculture of AD-MSC and human disc cells, and the majority of this work has utilized cells from the nucleus pulposus, not the annulus as reported here. In the present work, we showed that coculture with annulus cells was feasible with retention of both types of cells in a viable state. Coculture resulted in significantly increased GAG content compared to single culture of either cell type. These novel findings are important for future biology disc therapies that may utilize an autologous cell approach. Li et al. cocultured AD-MSC and rabbit nucleus pulposus cells.⁵⁸ Their favorable results showed an increase in type II collagen and aggrecan expression. Richardson et al. utilized human bone-marrow-derived MSC, and cocultured these cells with human cells from the nucleus pulposus.⁵⁹ Direct cell contact coculture showed increased expression in marker genes for the nucleus pulposus by MSC. This result was also favored only if the cell ratio favored the nucleus pulposus cells in culture. The authors suggested that if differentiated MSC were implanted into a disc, the retention of the disc phenotype could be restorative to the disc. Lu et al. used the transwell system for coculture of AD-MSC and nucleus pulposus cells in micromass conditions.⁶⁰ This technique induced stem cell changes toward a chondrogenic phenotype with type II collagen and aggrecan expression. A final comment about our coculture studies deserves consideration. One should remember that at present we are quite naïve about how other proteins important to disc cell function may be influenced by such coculture; this merits additional study in future experiments.

Work by LeVisage et al., which tested both annulus and nucleus cells, found that only the annulus cells were associated with greater GAG content during MSC coculture.²³ Gaetani et al. have recently studied coculture of nucleus pulposus cells with AD-MSC.⁶¹ They also show improved ECM production. In our study, the annulus cell was chosen because it is the annulus that shows fraying, dehydration, and rupture during disc degeneration and herniation.

We chose to use the Gelfoam 3D sponge system over alginate because our previous extensive studies have shown that annulus cells form only small colonies in alginate with minimal ECM production.^21,62 In the collagen sponge microenvironment, ECM formation is superior, and cells proliferate well. Additional evidence supporting the use of the Gelfoam 3D sponge system and annulus cells came from our earlier studies of AD-MSC harvested from fat in a small animal model; these cells were associated with enhanced GAG production and ECM deposition when cocultured with human annulus cells.¹⁵

In summary, findings presented here showed that human AD-MSC were enhanced by 3D culture with the addition of TGF-β3, that AD-MSC showed differences in their gene expression profiles compared to human annulus during 3D culture, and that coculture of AD-MSC and annulus cells significantly increased production of total sulfated proteoglycans. These novel findings have potential application in future cell-based therapies for disc degeneration.

Footnotes

Acknowledgments

The authors wish to thank the Carolinas Back Pain Research Endowment for general lab support, and internal funding from the Health Services Foundation of Carolinas Medical Center. The authors also thank Stanley Getz, M.D., and the Carolinas Cosmetic and Plastic Surgery group for assistance with collection of adipose tissue. This research was performed at Carolinas Medical Center, Charlotte, NC.

Disclosure Statement

No competing financial interests exist.

References

Boden

S.D.

, Davis

D.O.

, Dina

T.S.

et al. Abnormal magnetic-resonance scans of the lumbar spine in asymptomatic subjects: a prospective investigation. J Bone Joint Surg, 72A:403. 1990.

Boden

S.D.

, McCowin

P.R.

, Davis

D.O.

et al. Abnormal magnetic-resonance scans of the cervical spine in asymptomatic subjects. A prospective investigation. J Bone Joint Surg, 72A:1178. 1990.

Kakitsubata

, Theodorou

D.J.

, Theodorou

S.J.

et al. Magnetic resonance discography in cadavers: tears of the annulus fibrosus. Clin Orthop Res, 1:228. 2003.

Hilton

R.C.

, Ball

, Benn

R.T.

Annular tears in the dorsolumbar spine. Ann Rheum Dis, 39:533. 1980.

Gordon

S.J.

, Yang

K.H.

, Mayer

P.J.

et al. Mechanism of disc rupture. A preliminary report. Spine, 16:450. 1991.

Buckwalter

J.A.

Aging and degeneration of the human intervertebral disc. Spine, 20:1307. 1995.

Awad

H.A.

, Wickham

, Leddy

H.A.

et al. Chondrogenic differentiation of adipose-derived adult stem cells in agarose, alginate, and gelatin scaffolds. Biomaterials, 25:3211. 2004.

Erickson

G.R.

, Gimble

J.M.

, Franklin

D.M.

et al. Chondrogenic potential of adipose tissue-derived stromal cells. In vitro and in vivo. Biochem Biophys Res Commun, 290:763. 2002.

Gimble

J.M.

, Guilak

Adipose-derived adult stem cells: isolation, characterization, and differentiation potential. Cytotherapy, 5:362. 2003.

10.

Fukomoto

T.S.J.W.

, Sanyal

, Fitzsimmons

J.S.

et al. Combined effects of insulin-like growth factor-1 and transforming growth factor-beta1 on periosteal mesenchymal cells during chondrogenesis in vitro. Osteoarthritis Cartilage, 11:55. 2003.

11.

Mehlhorn

A.T.

, Schmal

, Saiser

et al. Mesenchymal stem cells maintain TGF-beta-mediated chondrogenic phenotype in alginate bead culture. Tissue Eng, 12:1393. 2006.

12.

Tuan

R.S.

, Boland

, Tuli

Adult mesenchymal stem cells and cell-based tissue engineering. Arthritis Res Ther, 5:32. 2002.

13.

Thompson

J.P.

, Pearce

R.H.

, Schechter

M.T.

et al. Preliminary evaluation of a scheme for grading the gross morphology of the human intervertebral disc. Spine, 15:411. 1990.

14.

Gruber

H.E.

, Fisher

E.C.

Jr. , Desai

et al. Human intervertebral disc cells from the annulus: three-dimensional culture in agarose or alginate and responsiveness to TGF-β1. Exp Cell Res, 235:13. 1997.

15.

Tapp

, Deepe

, Ingram

J.A.

et al. Adipose-derived mesenchymal stem cells from the sand rat: transforming growth factor beta and 3D co-culture with human disc cells stimulate proteoglycan and collagen type 1 rich extracellular matrix. Arthritis Res Ther, 10:R89. 2008.

16.

Cowan

C.M.

, Shi

Y.-Y.

, Aalami

O.O.

et al. Adipose-derived adult stromal cells heal critical-size mouse calvarial defects. Nat Biotechnol, 22:560. 2004.

17.

Dominici

, LeBlanc

, Mueller

et al. Minimal criteria for defining multipotent mesenchymal stromal cells. The International Society for Cellular Therapy position statement. Cytotherapy, 8:315. 2006.

18.

Pittenger

M.F.

, Mackay

A.M.

, Beck

S.C.

et al. Multilineage potential of adult human mesenchymal stem cells. Science, 284:143. 1999.

19.

Gruber

H.E.

, Hoelscher

G.L.

, Leslie

et al. Three-dimensional culture of human disc cells within agarose or a collagen sponge: assessment of proteoglycan production. Biomaterials, 27:371. 2006.

20.

Muller

, Hanschke

Quantitative and qualitative analyses of proteoglycans in cartilage extracts by precipitation with 1,9-dimethylmethylene blue. Connect Tissue Res, 33:243. 1996.

21.

Gruber

H.E.

, Leslie

, Ingram

et al. Cell-based tissue engineering for the intervertebral disc: in vitro studies of human disc cell gene expression and matrix production within selected cell carriers. Spine J, 4:44. 2004.

22.

Gruber

H.E.

, Leslie

K.P.

, Ingram

J.A.

et al. Optimization of 5-(and-6)-carboxyfluorescein diacetate succinimidyl ester for labeling human intervertebral disc cells in vitro. Biotech Histochem, 75:118. 2000.

23.

LeVisage

, Kim

S.W.

, Tateno

et al. Interaction of human mesenchymal stem cells with disc cells: changes in extracellular matrix biosynthesis. Spine, 31:2036. 2006.

24.

Reiner

, Yekutieli

, Benjamini

Identifying differentially expressed genes using false discovery rate controlling procedures. Bioinformatics, 19:368. 2003.

25.

Ashburner

, Ball

C.A.

, Blake

J.A.

et al. Gene ontology: tool for the unification of biology. Nat Genet, 25:25. 2000.

26.

Cheadle

, Vawter

M.P.

, Freed

W.J.

et al. Analysis of microarray data using Z score transformation. J Mol Diagn, 5:73. 2003.

27.

Sobajima

, Vadala

, Shimer

et al. Feasibility of a stem cell therapy for intervertebral disc degeneration. Spine J, 8:888. 2008.

28.

Ganey

, Hutton

W.C.

, Moseley

et al. Intervertebral disc repair using adipose tissue-derived stem and regenerative cells. Experiments in a canine model. Spine, 34:2297. 2009.

29.

Sakai

, Mochida

, Iwashina

et al. Regenerative effects of transplanting mesenchymal stem cells embedded in atelocollagen to the degenerated intervertebral disc. Biomaterials, 27:335. 2006.

30.

Leung

V.Y.L.

, Chan

, Cheung

K.M.C.

Regeneration of intervertebral disc by mesenchymal stem cells: potentials, limitations, and future direction. Eur Spine J, 15,Suppl. 3:S406. 2006.

31.

LeMaitre

C.L.

, Baird

, Freemont

A.J.

et al. An in vitro study investigating the survival and phenotype of mesenchymal stem cells following injection into nucleus pulposus tissue. Arthritis Res Ther, 11:R20. 2009.

32.

Jandial

, Aryan

H.E.

, Park

et al. Stem cell-mediated regeneration of the intervertebral disc: cellular and molecular challenges. Neurosurg Focus, 24:E20. 2008.

33.

Hoogendoorn

R.J.W.

, Lu

Z.F.

, Kroeze

R.J.

et al. Adipose stem cells for intervertebral disc regeneration: current status and concepts for the future. J Cell Mol Med, 12:2205. 2008.

34.

Acosta

F.L.

Jr. , Lotz

, Ames

C.P.

The potential role of mesenchymal stem cell therapy for intervertebral disc degeneration: a critical overview. Neurosurg Focus, 19:E4. 2005.

35.

H.S.

, Thonar

E.J.M.A.

, Masuda

Biological repair of intervertebral disc. Spine, 28:S86. 2003.

36.

Bron

J.L.

, Helder

M.N.

, Meisel

H.-J.

et al. Repair, regeneration and supportive therapies of the annulus fibrosus: achievements and challenges. Eur Spine J, 18:301. 2009.

37.

Videman

, Nurminen

The occurrence of anular tears and their relation to lifetime back pain history: a cadaveric study using barium sulfate discography. Spine, 29:2668. 2004.

38.

Zhu

, Liu

, Song

et al. Adipose-derived stem cell: a better stem cell than BMSC. Cell Biochem Funct, 26:664. 2008.

39.

Suga

, Shigeura

, Matsumoto

et al. Rapid expansion of human adipose-derived stromal cells preserving multipotency. Cytotherapy, 9:738. 2007.

40.

Izadpanah

, Trygg

, Patel

et al. Biologic properties of mesenchymal stem cells derived from bone marrow and adipose tissue. J Cell Biochem, 99:1285. 2006.

41.

Varma

M.J.O.

, Breuls

R.G.M.

, Schouten

T.E.

et al. Phenotypical and functional characterization of freshly isolated adipose tissue-derived stem cells. Stem Cells Dev, 16:91. 2007.

42.

Helder

M.N.

, Knippenberg

, Klein-Nulend

et al. Stem cells from adipose tissue allow challenging new concepts for regenerative medicine. Tissue Eng, 13:1799. 2007.

43.

Andersson

G.B.J.

, An

H.S.

, Oegema

T.R.

et al. Directions for future research. J Bone Joint Surg [Am], 88A,Suppl. 2:110. 2006.

44.

Gilson

, Dreger

, Urban

J.P.B.

Differential expression levels of cytokeratin 8 in cells of the bovine nucleus pulposus complicates the search for specific intervertebral disc cells markers. Arthritis Res Ther, 12:R24. 2010.

45.

Sive

J.I.

, Baird

, Jeziorsk

et al. Expression of chondrocyte markers by cells of normal and degenerate intervertebral discs. J Clin Pathol Mol Pathol, 55:91. 2002.

46.

Minogue

B.M.

, Richardson

S.M.

, Zeef

L.A.H.

et al. Transcriptional profiling of bovine intervertebral disc cells: implications for identification of normal and degenerate human intervertebral disc cell phenotypes. Arthritis Res Ther, 12:R22. 2010.

47.

Evans

N.D.

, Gentleman

, Polak

J.M.

Scaffolds for stem cells. Materials Today, 9:26. 2006.

48.

Kolf

C.M.

, Cho

, Tuan

R.S.

Mesenchymal stromal cells. Biology of adult mesenchymal stem cells: regulation of niche, self-renewal and differentiation. Arthritis Res Ther, 9:204. 2007.

49.

Philp

, Chen

S.S.

, Fitzgerald

et al. Complex extracellular matrices promote tissue-specific stem cell differentiation. Stem Cells, 23:288. 2005.

50.

Engler

A.J.

, Sen

, Sweeney

H.L.

et al. Matrix elasticity directs stem cell lineage specification. Cell, 126:677. 2006.

51.

, Boucher

, Koh

et al. PDGF, TGF-β, and FGF signaling is important for differentiation and growth of mesenchymal stem cells (MSCs): transcriptional profiling can identify markers and signaling pathways important in differentiation of MSCs into adipogenic, chondrogenic, and osteogenic lineages. Blood, 112:295. 2009.

52.

Sales

V.L.

, Engelmayr

G.C.

Jr. , Mettler

B.A.

et al. Transforming growth factor-β1 modulates extracellular matrix production, proliferation, and apoptosis of endothelial progenitor cells in tissue-engineering scaffolds. Circulation, 114,Suppl 1:I 193. 2006.

53.

Mehlhorn

A.T.

, Niemeyer

, Kaschte

et al. Differential effects of BMP-2 and TGF-β1 on chondrogenic differentiation of adipose derived stem cells. Cell Prolif, 40:809. 2007.

54.

Lee

H.R.

, Kim

B.C.

, Choi

I.S.

et al. Characterization and expression analysis of mesenchymal stem cells from human bone marrow and adipose tissue. Cell Physiol Biochem, 14:311. 2004.

55.

Peroni

, Scambi

, Pasini

et al. Stem molecular signature of adipose-derived stromal cells. Exp Cell Res, 314:603. 2008.

56.

Wagner

, Wein

, Seckinger

et al. Comparative characteristics of mesenchymal stem cells from human bone marrow, adipose tissue, and umbilical cord blood. Exp Hematol, 33:1402. 2005.

57.

Guilak

, Lott

K.E.

, Awad

H.A.

et al. Clonal analysis of the differentiation potential of human adipose-derived adult stem cells. J Cell Physiol, 206:229. 2006.

58.

X.D.

, Lee

J.P.

, Balian

et al. Modulation of chondrocytic properties of fat-derived mesenchymal cells in co-cultures with nucleus pulposus. Connect Tissue Res, 46:75. 2005.

59.

Richardon

S.M.

, Walker

R.V.

, Parker

et al. Intervertebral disc-cell-mediated mesenchymal stem cell differentiation. Stem Cells, 24:707. 2006.

60.

Z.F.

, Doulabi

B.Z.

, Wuisman

P.I.

et al. Differentiation of adipose stem cells by nucleus pulposus cells: configuration effect. Biochem Biophys Res Commun, 359:991. 2007.

61.

Gaetani

, Torre

M.L.

, Klink

et al. Adipose-derived stem cell therapy for intervertebral disc regeneration: an in vitro reconstructed tissue in alginate capsules. Tissue Eng, 14:1415. 2008.

62.

Gruber

H.E.

, Ingram

J.A.

, Leslie

et al. Cell shape and gene expression in human intervertebral disc cells: in vitro tissue engineering studies. Biotech Histochem, 78:109. 2003.

Human Adipose-Derived Mesenchymal Stem Cells: Direction to a Phenotype Sharing Similarities with the Disc,Gene Expression Profiling,and Coculture with Human Annulus Cells

Abstract

Introduction

Materials and Methods

Clinical study population

Specific methodologies for human disc cell culture

Specific methodologies for AD-MSC

Verification of stem cell isolation

CD marker analysis of AD-MSC

Osteogenic differentiation

Chondrogenic differentiation

Stimulation of AD-MSC with TGF-β3

Growth and differentiation of stem cells in Gelfoam®

Assay of total sulfated glycosaminoglycan production

Immunohistochemistry

Gene expression studies and analyses

Coculture experiments

Statistical analysis

Results

Verification of stem cell isolation

AD-MSC can be directed to a phenotype that shares similarities with disc cells by 3D culture in the presence of TGF-β3

Comparison of gene profiles of AD-MSC with TGF-β3 versus human annulus cells in 3D culture

Comparison of AD-MSC gene expression profiles in the presence or absence of TGF-β3

Coculture of AD-MSC with human annulus cells enhances proteoglycan production

Discussion

Footnotes

Acknowledgments

Disclosure Statement

References

Growth and differentiation of stem cells in Gelfoam^®