| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
a UK Centre for Tissue Engineering, Division of Laboratory and Regenerative Medicine, University of Manchester, Manchester, United Kingdom;
b UK Centre for Tissue Engineering, Clinical Engineering, University of Liverpool, Liverpool, United Kingdom
Key Words. Mesenchymal stem cell • Nucleus pulposus • Chondrocyte-like • Coculture • Differentiation
Correspondence: Judith A. Hoyland, Ph.D., Division of Laboratory and Regenerative Medicine, Stopford Building, University of Manchester, Oxford Road, Manchester, M13 9PT, U.K. Telephone: +44(0)161-275-5425; Fax: +44(0)161-275-5289; e-mail: judith.hoyland{at}manchester.ac.uk
Received May 5, 2005;
accepted for publication October 5, 2005.
 |
ABSTRACT |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionConclusionDisclosuresReferences |
|---|
|
INTRODUCTION |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionConclusionDisclosuresReferences |
|---|
The IVD is comprised of a central gelatinous nucleus pulposus (NP) surrounded by a more highly organized ring of predominantly type I collagen fibrils (the annulus fibrosus [AF]). Within the NP, rounded chondrocyte-like cells are embedded in a disorganized matrix of mainly type II collagen and aggrecan, supplemented with other proteoglycans, including versican [2, 3], and collagens, including type VI collagen. Aggrecan is a highly hydrophilic molecule. The amount of aggrecan in the NP is such that by imbibing water it generates sufficient swelling pressure to force even loaded vertebrae apart, a force resisted by the AF. Vertebral stability relies, in part, on the balance of these two opposing forces. During degeneration in the NP, there is enzymatic breakdown and decreased production of extracellular matrix components, in particular proteoglycans. This leads to loss of the hydrogel properties of the NP, including an increased fibrillar collagen content (particularly that of type I collagen), reduced swelling pressure, instability, and traumatic damage to the disc and surrounding structures [2, 4, 5]. This change in matrix composition is driven by an altered cell physiology. Our group has shown that this is probably caused by a relative increase in expression of the catabolic cytokine IL (interleukin)-1
[4] and the degradative enzymes MMP (matrix metalloproteinase)-1, -3, and -13 and ADAMTS (a disintegrin and metalloprotease with thrombospondin motifs)-4 [6] as the severity of degeneration increases.
Although there have been recent developments in therapies to treat IVD degeneration, such as those involving laser beam therapy or needle aspiration, the majority of preferred treatments involve either removal of tissue or fusion of the affected disc level; however, these treatments remove the symptoms rather than repair the tissue. Furthermore, they are often short-term solutions that frequently lead to complications due to alterations in the biomechanics of the spine. An ideal solution to managing disc degeneration would be to repair (or regenerate) the NP, producing a matrix with similar or improved biological and biomechanical properties compared with the original.
One such strategy is to use autologous cell implantation similar to that used in articular cartilage repair. As with articular cartilage, however, the IVD is relatively acellular, with only approximately 4,000 cells per mm3 in the NP [7], making the source of cells one of the main problems for any tissue repair strategy.
Recent studies using a number of species have suggested that reimplantation of extracted NP cells into the disc can retard degenerative changes [8–10]. However, the numbers of cells required for a successful treatment are more than can be reasonably acquired from a single disc. Furthermore, evidence suggests that removal of tissue from a degenerate disc can accelerate degeneration within that disc, whereas removal of cells from an adjacent normal disc could induce degeneration in that disc level [11]. In addition to this, cells from degenerate discs have an altered phenotype, showing increased senescence [12], increased expression of both catabolic cytokines and degradative enzymes [4, 6], and decreased expression of matrix components [2, 5].
One potential method to overcome these problems is to coculture NP cells with other cells types either to increase the number of NP-like cells or to increase the activity and matrix production rates of the NP cells. Aguiar et al. [13] cocultured canine and bovine NP cells with notochordal cells derived from adult canine discs and showed an increase in proteoglycan levels expressed by the NP cells. Although the NP is derived from the notochord during embryogenesis, the extent to which notochordal cells persist is species-dependant. Thus, whereas in certain breeds of dog, notochordal cells persist throughout life, in humans the NP is usually devoid of these cells by the age of 10 [14]. This lack of notochordal cells in adult human discs makes this strategy unworkable for treatment of human IVD degeneration.
We are examining a different strategy. As NP cells are essentially of the chondroid lineage [2], it should be possible to produce them from mesenchymal stem cells (MSCs). Previous studies have shown that these undifferentiated cells have a high proliferative capacity and have the ability to differentiate into a number of lineages, including both chondrocytes [15] and NP-like cells [16], although this study was performed using rat MSCs. This differentiation also required the use of a three-dimensional matrix such as alginate, although for use in a tissue engineering strategy differentiation in monolayer culture would be preferable. However, although studies have shown that adult human MSCs can be differentiated into chondrocytes for repair of articular cartilage defects [17, 18], the only data showing adult human MSC differentiation to NP-like cells were achieved using three-dimensional spheroid culture [19], which is not applicable to tissue engineering. A recent study by Yamamoto et al. [20] using rabbit NP cells and MSCs showed an increase in cell proliferation and proteoglycan synthesis during coculture where there was direct cell–cell contact. However, although the authors suggest that MSCs were stimulating NP cell proliferation and matrix production, the cells were not separated after coculture, and so the fundamental question of whether NP cells can induce MSC differentiation, or whether MSCs stimulate NP cells, has not been answered.
The present study aimed to investigate the effect of human MSC and NP cell coculture on the differentiation of MSCs to NP-like cells in monolayer culture. Cells were cocultured with and without cell–cell contact and at a variety of different ratios to investigate whether these factors affected cell differentiation. MSCs were labeled with a cell-permanent inert green-fluorescent dye, and this allowed NP cells and MSCs to be separated by high-speed cell sorting subsequent to the coculture with cell–cell contact. Finally, gene expression of the transcription factor SOX-9, which is involved in differentiation of MSCs to chondrocyte-like NP cells, the proteoglycans aggrecan and versican, and the collagens types I, II, and VI, was measured by real-time polymerase chain reaction (PCR) to examine the differentiation states of both the MSC and NP cells.
|
MATERIALS AND METHODS |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionConclusionDisclosuresReferences |
|---|
MSC Source and Culture
To obtain human mesenchymal stromal stem cells, bone marrow was obtained from proximal femur samples removed during hip replacement surgery on a 62-year-old male. The patient gave informed consent, and local Ethical Committee approval was obtained for the use of the sample for research. Mononucleated cells were isolated using a Histopaque-1077 (Sigma, St. Louis, http:// www.sigmaaldrich.com) density gradient method [21]. These cells were cultured in a 75-cm2 flask with Minimum Essential Medium, -modification (-MEM), supplemented with 10% heat-inactivated FCS, 100 U/ml penicillin, 100 µg/ml streptomycin, 2.5 µg/ml amphotericin B, and 2 mM L-glutamine at 37°C in a humid atmosphere containing 5% CO2. After 7 days, nonadherent cells were discarded, and adherent cells were cultured to confluence, with medium changed every 3 days.
Fluorescent Labeling of MSCs
At confluence, MSCs were trypsinised, counted, and resuspended at a concentration of 1 x 106 in -MEM. A sufficient number of cells for the experiment were incubated for 30 minutes at 37°C with a final concentration of 10 µM 5,6-carboxyfluorescein diacetate, succinimidyl ester (CFDA; Sigma). After incubation, the reaction was stopped by adding an excess of medium, centrifuging at 400g for 5 minutes and resuspending in DMEM-HG to a concentration suitable for seeding.
Coculture of MSCs and NP Cells
All cocultures were conducted in triplicate in 24-well plates (Becton, Dickinson and Company, Franklin Lakes, NJ, http:// www.bd.com) and using 0.4-µm pore size, high pore density, polyethylene terephthalate (PET) track-etched tissue culture inserts (Becton, Dickinson and Company) where appropriate. CFDA-labeled MSCs and unlabeled NP cell control wells were prepared using 6 x 103 cells per cm2. For coculture wells without contact, CFDA-labeled MSCs were seeded on tissue culture plastic, whereas NP cells were seeded on the upper surface of the membrane of the tissue culture inserts. For co-cultures with direct cell-cell contact, both NP cells and CFDA-labeled MSCs were seeded in the base of the appropriate wells of a 24-well plate. Cells were seeded at ratios of 75:25% NP cells/MSCs, 50:50% NP cells/MSCs, and 25:75% NP cells/ MSCs. The actual number of cells seeded in each well is summarized in Table 1
. Cocultured cells were maintained for 7 days in DMEM-HG at 37°C and 5% CO2 in a humidified atmosphere with medium being changed every 2 days.
|
Real-Time PCR Analysis of Gene Expression
Control cells were rinsed in PBS, and cells were lysed in TRIzol (Invitrogen, Carlsbad, CA, http://www.invitrogen.com). For cells cocultured without contact, tissue culture inserts were removed to clean wells, cells were rinsed in PBS, and TRIzol was added to both the original wells (MSCs) and the inserts (NP cells) to lyse the cells.
Total RNA was then extracted according to manufacturer instructions. Briefly, chloroform was added to each sample, and sample tubes were centrifuged to enable phase separation. RNA was precipitated by addition of isopropanol to the aqueous phase, followed by centrifugation. Precipitated RNA pellets were washed in 75% ethanol and then resuspended in distilled RNAse-free water.
cDNA was prepared from RNA using Superscript II RT (Invitrogen). RNA (<1 µg) was mixed with random prime hexamers (250 ng) and dNTPs (500 µM each dNTP) and then incubated at 65°C for 5 minutes. Tubes were cooled on ice, and then 5x first strand buffer, 0.1 M dithiothreitol (DTT), 40 U RNAseOUT RNAse-inhibitor (Invitrogen), and 200 U reverse transcriptase were added, giving a final volume of 20 µl. Samples were then incubated at 42°C for 90 minutes and finally heated to 70°C for 15 minutes.
Gene expression was analyzed by real-time PCR using an ABI Prism 7000 sequence detection system (Applied BioSystems, Foster City, CA, http://www.appliedbiosystems.com). Primer/probe sets of housekeeping genes glyceraldehyde-3-phosphate dehydrogenase (GAPDH) and 18S were predesigned and purchased from Applied BioSystems. Primers for SOX-9; collagen types I, II and VI; proteoglycans aggrecan and versican; and elastin were designed using the Applied BioSystems Primer Express 2 software or derived from Martin et al. [22] (Table 2).
|
The PCR reaction consisted of an initial enzyme activation step at 95°C for 10 minutes, followed by 40 cycles of 95°C for 15 seconds and 60°C for 1 minute. A cycle threshold (Ct) value was obtained for each sample, and duplicate sample values were averaged. The 2–
Ct method was then used to calculate relative expression of each target gene [23]. Briefly, mean Ct value of target genes in each sample was normalized to its averaged housekeeping gene Ct value (the average of the mean 18S and mean GAPDH values [24]) to give a Ct value. This was then normalized to control samples (Ct), which in all cases was MSC or NP cells cultured alone, and finally the 2–Ct value was obtained.
Coculture of MSCs with Human Dermal Fibroblasts
To ensure that the changes seen in MSC and NP gene expression after coculture with contact were caused by the NP cells, MSCs were cultured in contact with human dermal fibroblasts (HDFs). The same experimental setup and cell ratios were used as for the NP-MSC coculture. CFDA-labeled MSCs were separated from HDFs using MoFlo high-speed cell sorting as described earlier. Cells were lysed in TRIzol, and RNA was extracted and reverse-transcribed to cDNA. This was then used in real-time PCR for housekeeping genes GAPDH and 18S as well as collagen types I and II, SOX-9, aggrecan, and elastin as described earlier.
Statistical Analysis
Statistical significance was determined using the Student’s t test, in which p < .05 was considered significant.
|
RESULTS |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionConclusionDisclosuresReferences |
|---|
As the CFDA molecule is divided evenly between daughter cells with each cell division, it was important to isolate cells that were distinctly positive or negative after culture (Fig. 1). Therefore the region 1 (R1) gate was placed over a small percentage of "negative" NP cells to ensure no contamination with MSCs that had undergone several divisions. Similarly, the region 2 (R2) gate only covered cells that were definitely positive. After 7 days, the HDF-MSC control coculture samples were sorted using the same method, with unlabeled HDFs gated on region 2 and labeled MSCs gated on region 3 (Fig. 2). A sample of each cell population was reanalyzed on the MoFlo high-speed cell sorter to check cell purity. This demonstrated whether there were contaminating labeled cells (MSCs) in the unlabeled cell (NP or HDF) population and vice versa, and it was established that each population had a purity of more than 97% (i.e., insignificant numbers of contaminating cells).
|
|
).
|
Where cells were cocultured without contact, there were only minor changes in the genes studied (Fig. 4). Although a number of these changes were significant, none of the increases or decreases was larger than 10-fold over controls. It was also evident from the data that cell ratio had no effect on gene expression.
|
), with the largest increase being seen at a ratio of 75:25 NP/MSC (700-fold over controls compared with 63-fold and 51-fold in 50:50 NP/MSC and 25:75 NP/MSC, respectively). Although SOX-9 mRNA levels in NP cells were raised significantly, these increases were below 10-fold over controls.
Type II collagen expression after 7 days in coculture with contact (Fig. 4B) was increased significantly in MSCs at all ratios, with the largest increase of almost 6,000-fold again being seen at a ratio of 75:25 NP/MSC. A significant increase was seen in type II collagen expression in NP cells at the same ratio; however, there was a small but not significant increase in expression in the 50:50 sample and a significant decrease in the 25:75 NP/MSC sample. Although after 7 days MSCs at all cell ratios with contact showed significant increases in expression of type I collagen, the increases were relatively low, being below a 10-fold increase in every case (Fig. 4C).
Type VI collagen showed statistically significant increases in expression in MSCs at all ratios, with the largest increase being seen with the most NP cells present (75:25 NP/MSC) (Fig. 4D). These increases were small, with the largest increase being only fivefold over controls. NP cells showed significant increases at 75% and 50% NP ratios, although again these were very small increases (both 1.6-fold).
Examination of the changes in expression of the proteoglycans aggrecan and versican showed very different results. Although MSCs showed large and significant increases in aggrecan expression (Fig. 4E), there was no significant change in versican mRNA expression (Fig. 4F). Aggrecan expression in MSCs showed the largest increase in samples with the highest number of NP cells (5876-fold) and decreased as NP cell number decreased (485-fold at 50% and 11-fold at 25%). NP cells, however, showed small but significant increases in aggrecan at 75% and 50% NP but a decrease at 25% NP, although again there was no significant change in expression of versican.
Gene Expression after HDF and MSC Coculture with Contact
HDFs and MSCs were cocultured with contact for 7 days, and changes in expression of a number of NP and HDF marker genes were measured by real-time PCR. At a ratio of 75:25 HDF/MSC, both cell types showed small (2.8-fold in HDF and 2.9-fold in MSCs) but significant increases in SOX-9 (Fig. 5A), whereas there was also a significant increase in SOX-9 expression (6.3-fold) in MSCs at a ratio of 75:25 NP/MSC. Type II collagen could not be reproducibly detected in any of the samples (data not shown), and although detectable, aggrecan did not show any significant changes (Fig. 5B). Similarly, there was very little change in expression of type I collagen, with only MSCs at a ratio of 75% showing a small but significant decrease (Fig. 5C). The largest changes were seen in the expression of elastin, with both HDFs and MSCs showing significant increases in expression after 7 days in coculture (Fig. 5D).
|
|
DISCUSSION |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionConclusionDisclosuresReferences |
|---|
Although a number of different methodologies can be employed to induce and maintain a differentiated phenotype in MSCs, these usually involve addition of growth factors or culture in a three-dimensional environment that makes subsequent usage difficult [16, 19]. The current study aimed to elucidate the effects that coculture had on the differentiation state of both human NP cells and MSCs and whether this gene expression was modulated by culture with or without cell-cell contact or the alteration of cell ratio.
Our findings suggest that rather than inducing matrix production by NP cells, as suggested by Yamamoto and co-workers [20], the NP cells induce differentiation and matrix component gene expression by MSCs. We were able to study this by using a novel fluorescent cell-labeling coculture method, combined with high-speed cell sorting to isolate both NP cells and MSCs after coculture with cell–cell contact.
The Effect of Cell–Cell Contact
Yamamoto et al. [20] and others have suggested that coculture with cell–cell contact induces growth factor expression, in particular transforming growth factor-ß (TGF-ß), insulin-like growth factor-1 (IGF-1), epidermal growth factor (EGF), and platelet-derived growth factor (PDGF), that is not seen in co-culture without contact, and Thompson et al. [28] showed that these growth factors are responsible for matrix production by NP cells. Our work using coculture with and without contact suggests that cell–cell contact itself is required for MSC differentiation and matrix production by NP cells and MSCs, since coculture without contact did not show any significant changes in matrix gene expression by either NP cells or MSCs. This differs from other studies using different cell types cocultured with MSCs where no cell-cell contact has been shown to have an effect, although not to the same extent as when cells were cultured with cell-cell contact [29]. Our results may therefore be due to the specific cell type (i.e., the NP cell) used for coculture.
Cells cultured with cell–cell contact showed large increases in gene expression of SOX-9 and a number of matrix molecules, in particular aggrecan and type II collagen, indicating that direct cell-cell contact is essential for MSC differentiation to NP cells. The changes in cell proliferation and gene expression profiles of cells cocultured with cell-cell contact demonstrated in both this study and that of Yamamoto et al. [20] could also be due to expression of a number of cell adhesion molecules or changes to cell signaling pathways, and further studies are needed to clarify this.
Our results have shown that the differentiation pathway of MSCs is dependant on the cell type with which they are cocultured. In our study, control experiments, in which MSCs were cultured with HDFs with cell–cell contact, showed increases in elastin that were not seen after MSC coculture with NP cells (data not shown). This is important because elastin is a key component of dermal tissue and has been shown to be expressed by cultured dermal fibroblast [30]. There were small, yet significant increases in SOX-9 mRNA expression, which could possibly be due to release of factors by the HDFs that affect SOX-9 expression, such as members of the fibroblast growth factor (FGF) family [31]. FGFs have been shown to stimulate SOX-9 expression in both chondrocytes and MSCs [32], and this could account for the increased expression of SOX-9 during MSC coculture with HDFs. There was, however, no increase in either type II collagen or aggrecan, which showed the largest increases in MSCs after coculture with NP cells, suggesting that this increase in SOX-9 mRNA expression is not causing differentiation of either the MSCs or HDFs to chondrocyte-like cells.
These findings are consistent with those of Ball et al. [33], who demonstrated that coculture of MSCs and vascular endothelial cells with cell–cell contact produced a different effect on MSC differentiation than did coculture of MSCs with HDFs. It is therefore clear that the cells with which MSCs are cocultured determine their fate and that coculture of MSCs with NP cells with cell–cell contact produces a population of differentiated NP-like cells from the stem cells.
Gene Expression and the Effect of Cell Ratio
Most previous coculture studies have used only a 50:50 cell ratio, and although these studies have shown that coculture affects cell phenotype, no studies have investigated whether this ratio is the optimum for directing differentiation. It is for this reason that we investigated a range of cell ratios. Importantly, we have shown that in most cases a ratio of 75:25 NP cells/MSCs was the optimal ratio for stimulating MSC differentiation, possibly due to the increase in cellular signals received by the MSCs from the greater NP cell population. This was evidenced by the relative increases in SOX-9, type II collagen, and aggrecan gene expression after 7 days, particularly at a ratio of 75:25 NP cells/MSCs. Whereas type I and VI collagen showed only minor, yet still significant, changes in expression, results suggest that type I collagen was constitutively expressed at high levels by both NP cells and MSCs in monolayer culture. Type VI collagen has been shown to account for as much as 5% of the IVD tissue dry weight (compared with only approximately 1% of the dry weight of articular cartilage [34]). Type VI collagen is involved in proteoglycan complex interactions, which combined with its abundance in the NP of the disc (approximately 20% of total collagen content [34]) makes it a potentially important collagen in the proteoglycan-rich NP. Our studies have shown that coculture leads to significant increases in expression by both NP cells and MSCs, which should be a key factor in production of a integrated tissue-engineered matrix.
Preliminary real-time PCR data using primers specific for the type IX collagen isoforms (data not shown) showed low levels of expression of the short-form of type IX collagen in NP cells rather than the long-form, although this expression was unchanged by coculture. MSCs did not show detectable levels of either form in control samples or samples cocultured without contact. However, after coculture with NP cells with cell–cell contact, MSCs did show detectable levels of the short-form, but not the long-form. As NP cells have previously been shown to express only the short-form of type IX collagen, whereas articular chondrocytes express the long-form [35], this molecule may be a usable marker of an NP-specific phenotype, rather than a chondrocytic phenotype.
One of the largest and potentially most important changes was the increase in expression of aggrecan. A recent study comparing proteoglycan and collagen ratios in NP tissue and cartilage showed a 27:1 ratio of proteoglycan/collagen in the NP, but only a 2:1 ratio in cartilage, demonstrating the importance of proteoglycans, in particular aggrecan, in the structure and function of the IVD [36]. Therefore the increase in aggrecan, particularly in MSCs after 7 days in coculture with cell–cell contact with NP cells, suggests that these cells have differentiated to cells with an NP-like phenotype and could potentially be used in therapeutic strategies.
The Application of CFDA Labeling and High-Speed Cell Sorting
MSCs were labeled with CFDA to allow separation of MSCs and NP cells after coculture with direct cell–cell contact. This inert, non-fluorescent molecule is cell-permeable and once within a cell is cleaved by intracellular esterases to a cell-permanent green fluorescent dye that binds covalently to intracellular proteins. This dye is shared evenly between daughter cells and has been used extensively to track cell division in a number of cell types, including NP cells and stem cells, and has been shown not to affect cell phenotype, function, or proliferation rate [37–41]. Results comparing CFDA-labeled and -unlabeled MSCs showed no significant change in gene expression after culture for 7 days, suggesting that CFDA is a safe and reliable molecule for labeling cells.
High-speed cell sorting allows each population of cocultured cells to be examined individually. The MoFlo high-speed cell sorter has the theoretical ability to identify, classify, and sort individual cells at rate of up to 100,000 cells per second at above 95% purity, allowing exact subpopulations to be identified and separated [42, 43]. This means that CFDA-labeled MSCs can be separated from unlabeled NP cells with high purity and specificity and that these cells can be used for further culture or application in tissue engineering therapies to treat IVD degeneration.
|
CONCLUSION |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionConclusionDisclosuresReferences |
|---|
Although only one cell population has been used, our results suggest that this novel methodology has major implications for the future of cell-based, tissue engineering strategies for treatment of the degenerate IVD, because it allows rapid production of NP-like cells from MSCs without the dangers of removing IVD tissue.
|
DISCLOSURES |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionConclusionDisclosuresReferences |
|---|
|
ACKNOWLEDGMENTS |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionConclusionDisclosuresReferences |
|---|
|
REFERENCES |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionConclusionDisclosuresReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  M. G. Hubert, G. Vadala, G. Sowa, R. K. Studer, and J. D. Kang Gene Therapy for the Treatment of Degenerative Disk Disease J. Am. Acad. Ortho. Surg., June 1, 2008; 16(6): 312 - 319. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| STEM CELLS | THE ONCOLOGIST | CME | ALPHAMED PRESS JOURNALS |
