Phase-contrast photomicrograph of the cloning of human bone marrow mesenchymal stem cells 4 days after plating in the primary culture (original magnification ×100).
Gross specimen of tissue-engineered cartilage using human bone marrow mesenchymal stem cells, which was harvested 10 weeks after implantation and measured about 1.3 cm in length.
Hematoxylin-eosin staining of engineered cartilage using human bone marrow mesenchymal stem cells after 10 weeks (original magnification ×400).
Safranin O staining of engineered cartilage using human bone marrow mesenchymal stem cells after 10 weeks. Abundant sulfated glycosaminoglycan was shown in the matrix (original magnification ×400).
Abundant collagen was detected by Masson trichrome staining of engineered cartilage using human bone marrow mesenchymal stem cells after 10 weeks (original magnification ×400).
Immunochemical staining of type II collagen of a cross section of engineered cartilage using human bone marrow mesenchymal stem cells after 6 weeks. Type II collagen was mostly localized inside the cells at this stage (original magnification ×400).
Pang Y, Cui P, Chen W, Gao P, Zhang H. Quantitative Study of Tissue-Engineered Cartilage With Human Bone Marrow Mesenchymal Stem Cells. Arch Facial Plast Surg. 2005;7(1):7-11. doi:10.1001/archfaci.7.1.7
Author Affiliations: Department of Otorhinolaryngology (Drs Pang, Cui, Chen, and Gao) and Central Laboratory (Dr Zhang), TangDu Hospital, The Fourth Military Medical University, Xi’an, People’s Republic of China.
Correspondence: Yonggang Pang, MD, Department of Otorhinolaryngology, TangDu Hospital, The Fourth Military Medical University, Xi’@an 710038, People’s Republic of China (firstname.lastname@example.org).
Objectives To assess the possibility of cartilage tissue engineering using human mesenchymal stem cells (hMSCs) and to investigate the quantitative relationship between hMSCs and engineered cartilage.
Design Human mesenchymal stem cells were cultured, cryopreserved, and expanded in vitro. Surface antigens were detected by flow cytometry. In vitro chondrogenesis of hMSCs and cryopreserved hMSCs was performed. The chondrogenesis-induced hMSCs were seeded onto polyglycolic acid scaffolds, cultured in vitro for 3 weeks in chondrogenic medium, and then implanted into nude mice. The implants were harvested after 10 weeks and examined with histologic and immunochemical staining.
Results The construction of cartilages was identified grossly and histologically: 1.9 to 2.5 × 107 nucleated cells were obtained from 1 mL of bone marrow, and about 1 to 2 × 106 hMSCs were obtained from the primary culture. The number of hMSCs tripled at every passage and reached 1.4 to 2.8 × 1012 at passage 15. The purity of hMSCs was 95% and 98% at the primary and the fourth passages, respectively. Twenty-one days was the optimal (induction rate, 95%) induction time, with no apparent differences in induction rates among different passages. Based on our findings, hMSCs from 0.07 to 0.14 mL of bone marrow, expanded during 4 passages and induced for 21 days, would be sufficient to engineer 1 cm2 of cartilage, 3-mm thick.
Conclusion Quantitative standards of hMSCs as seed cells for cartilage tissue engineering were established and may have value for later clinical work.
Cartilage tissue engineering, established 10 years ago, was one of the first research fields of tissue engineering study, and great progress has been made since then. However, most investigations used chondrocytes isolated from the cartilage as seed cells, which have poor expanding ability and are unable to excrete matrix after only a few passages.1 Various growth factors and dynamic culture methods have been used to try to improve the expansion and matrix secretion of chondrocytes.2 Nevertheless, the results are not encouraging.
Cells and their sources are the current biological challenges of tissue engineering.3 Bone marrow contains stemlike cells that are precursors of nonhematopoietic tissue. The precursors are referred to as mesenchymal stem cells.4 With the ability to differentiate into chondrocytes, osteoblasts, adipocytes, and myoblasts,5 these cells represent a promising option for future tissue engineering strategies. One objective of our study was to assess the possibility of forming cartilage using chondrogenesis-induced human mesenchymal stem cells (hMSCs) as seed cells for cartilage tissue engineering.
Human mesenchymal stem cells are easily isolated from a small aspirate of bone marrow and generate single cell–derived colonies. In about 10 weeks, the colonies can be expanded through 50 doublings. Although the differentiation6 and expansion7 have been documented, the quantification of hMSCs used as seed cells for cartilage tissue engineering has not yet been well studied. The other objective of our study was to investigate the quantitative relationship between hMSCs and engineered cartilage.
The institutional review board of The Fourth Military Medical University approved the harvest and use of hMSCs, as well as the animal use protocol. Human mesenchymal stem cells were obtained from 5 to 10 mL of aspirate of the iliac crest of normal donors and isolated as described previously.8 Briefly, bone marrow was diluted 1:1 with Dulbecco modified Eagle medium (DMEM) (Invitrogen Corp, Carlsbad, Calif) and centrifuged to pellet the cells; the fat layer was removed. Cell pellets were then layered over about 10 mL of 70% Percoll (Sigma-Aldrich Corp, St Louis, Mo). After centrifuging at 3000g for 30 minutes, an hMSC-enriched low-density fraction was collected, rinsed with DMEM twice, and plated at 2.5 × 105 nucleated cells/cm2 in control medium (DMEM, 10% fetal bovine serum [FBS] lot-selected for rapid growth of hMSCs [Invitrogen Corp], 100 U/mL of penicillin [Sigma-Aldrich Corp], 100 μg/mL of streptomycin [Sigma-Aldrich], and 2mM l-glutamine [Invitrogen Corp]). All of the cells were incubated at 37ºC with 5% humidified carbon dioxide. After 5 days, nonadherent cells were discarded, and the medium was changed twice a week.
When the attached cells reached confluence after 12 to 15 days of primary culture, they were harvested with 0.25% trypsin and replaced at a 1:3 dilution for the first subculture and the following passages.
At passage 4, confluent cells were released with trypsin, and cells were slowly frozen (using a temperature decrease of 1ºC/min) in DMEM with 20% FBS and 10% dimethyl sulfoxide. When a temperature of −80ºC was reached, they were cryopreserved in liquid nitrogen.
Aliquots of frozen cells were thawed and resuspended in DMEM with 10% FBS after variable times (range, 1-10 months). The cellular suspension was centrifuged, the supernatant was discarded, and the pellet was resuspended in DMEM with 10% FBS. The viability of the cells was determined by the ability to exclude trypan blue dye.
Cells were trypsinized and stained with anti-CD14, -CD29, -CD34, -CD44, and -CD45 fluorescein isothiocyanate (Beckman Coulter, Inc, Miami, Fla) and were analyzed by flow cytometry.
We modified the chondrogenesis assays as described previously.9 In brief, 2 × 105 cells in 15 mL of medium were centrifuged at 500g in 15-mL polypropylene conical tubes. The pelleted cells were incubated at 37ºC with humidified carbon dioxide in FBS-free media containing high-glucose DMEM (GIBCO BRL, Gaithersburg, Md), 10 ng/mL of transforming growth factor-β1 (Sigma-Aldrich Corp), 100 nM of dexamethasone, 50 μg/mL of ascorbic acid, 10 U/mL of insulin (Sigma-Aldrich), 10mM β-glycerophosphate, 50 ng/mL of thyroxine, 1.25 mg/mL of bovine serum albumin, and 6.25 μg/mL of transferring medium. The chondrogenic differentiation was evaluated by reverse transcription–polymerase chain reaction. RNA was extracted from the pellets with TRIZOL reagent (GIBCO BRL) and used for oligo dT–primed complementary DNA synthesis. The complementary DNA was used as the template for polymerase chain reaction amplification per 25-μL reaction volume using primer pairs designed as follows: (forward)-5′-ACGGCGAGAAGGGAGAAGTTG-3′ and (reverse) 5′-GGGGGTCCAGGGTTGCCATTG-3′. Denaturation at 94ºC was performed for 30 seconds, followed by 10-second annealing at 60ºC, a 1-minute extension at 72ºC, and amplification for 30 cycles. The expected product size was 352 base pairs.
The cell aggregates were frozen in embedding medium, and 5-μm-thick cryosections were collected onto poly-l-lysine–coated slides (Sigma-Aldrich Corp). For immunohistochemistry, an LSAB Kit (Dako Corp, Glostrup, Denmark) was used. Type II collagen was immunolocalized with monoclonal antibody to type II collagen (1:100; Santa Cruz Biotechnology Inc, Santa Cruz, Calif). Primary antibody was omitted in the negative control specimens. The number of positive cells as indicated by type II collagen staining was determined by scoring at least 300 cells from 5 different microscopic fields. The percentage of positive cells was considered as the induction rate. The induction rates of passage 4 on days 7, 14, 21, and 28 were determined. The rates of passages 1, 8, and 15 on day 21 were also determined.
Nonwoven mesh of fibers of polyglycolic acid about 2-mm thick (Albany International Research Co, Mansfield, Mass), coated in advance with poly-l-lysine, was cut into 1.2 × 1-cm pieces. Cells were suspended in DMEM culture media and concentrated to a density of 5 × 107 cells/mL, and 200 μL/cm2 of the cell suspension was seeded onto the mesh. The cell-polymer constructs (n = 10) were incubated in vitro for 3 weeks in chondrogenic medium and implanted into subcutaneous pockets in nude mice. The constructs were kept in vivo for 10 weeks. Cryopreserved hMSCs were also used as seed cells to construct tissue-engineered cartilage in a similar manner. Nonwoven mesh of fibers of polyglycolic acid, similarly treated with the chondrogenic medium but not seeded with cells, was implanted and served as control specimens (n = 4).
Specimens were harvested after 6 and 10 weeks and then fixed in buffered formalin. Using standard histochemical techniques, serial sections of 5 μm were cut and stained with hematoxylin-eosin, Masson trichrome, and safranin O. Immunohistochemistry was performed in the same way as already described.
The data were exported to SPSS version 11.0 (SPSS Inc, Chicago, Ill). Analysis of variance models were used to assess differences among groups. Differences at P<.05 were considered significant.
About 1.9 to 2.5 × 107 nucleated cells were isolated per milliliter of bone marrow after density gradient isolation. The cells were plated at an optimal seeding density of 2.5 × 105 cells/cm2 according to previous findings10 (Figure 1). About 1 to 2 × 106 hMSCs were obtained from the primary culture. With the 1:3 dilution, the yields for passages 1, 4, and 15 were 3 to 6 × 106, 8.1 to 16.2 × 107, and 1.4 to 2.8 × 1012 cells, respectively. The time from the introduction of hMSCs into culture until their harvest for subcultivation in passage 1 was 12.8 ± 1.0 days, and hMSCs could be expanded by successive cycles of trypsinization, seeding, and culture for 4 days.
In vitro chondrogenesis of hMSCs was proved by reverse transcription–polymerase chain reaction and immunohistochemical staining. The mean ± SD induction rate of passage 4 on days 7, 14, 21, and 28 was 8.47% ± 1.17%, 49.93% ± 7.68%, 95.13% ± 3.92%, and 95.07% ± 3.95%, respectively. There was no apparent difference between the induction rates on day 21 and day 28 (P>.05). The mean ± SD day-21 induction rates among different passages (95.13% ± 3.92%, 96.46% ± 1.99%, and 94.86%±4.48% for passages 4, 8, and 15, respectively) showed no apparent differences (P>.05).
To construct a piece of engineered cartilage of about 1×1 cm × 0.3 mm, 1.1 × 107 hMSCs were induced to form 1×107 chondrocytes. When hMSCs were expanded for 1, 4, and 15 passages, the amount of bone marrow needed was 5.5 to 11 mL, 0.07 to 0.14 mL, and 0.4 to 0.8 mL, respectively.
About 95% of the cells in the primary culture were positive for CD44 and CD29 and negative for CD14, CD34, and CD45. From passage 4, almost-pure hMSCs could be obtained, and more than 98% of the cells expressed the phenotype already described. The induction rate reached the highest point at day 21 and day 28, and there was no significant difference between them (P>.05). No apparent differences in induction rates were detected between different passages.
Cartilages were identified grossly after 10 weeks’ implantation (Figure 2) and were further proved by histologic evaluation with hematoxylin-eosin, safranin O, Masson trichrome, and immunohistochemical staining (Figure 3, Figure 4, Figure 5, and Figure 6). Tissue-engineered cartilages were obtained using the cryopreserved hMSCs after chondrogenesis induction, as the seed cells proved to be consistent with human cartilage. In the control group, the polyglycolic acid scaffolds were completely absorbed after 10 weeks, without any cartilage forming. The thickness of the cartilages was about 3 mm.
The aims of this study were to investigate the possibility of using hMSCs as seed cells for cartilage tissue engineering and to perform a quantitative study. The field of tissue engineering uses living cells in different ways to restore, maintain, or enhance tissues and organs.11 Bone marrow mesenchymal stem cells provide a new alternative source for tissue engineering. Previous studies1-2,12 on cartilage tissue engineering used chondrocytes isolated from the cartilage from various sites. One study12 showed that only immediately harvested chondrocytes would be suitable as seed cells for cartilage tissue engineering and that chondrocytes in the following passages would contain decreased matrix sulfated glycosaminoglycan and type II collagen, which would limit the mechanical performance of neocartilage constructs. This is one of the obstacles to cartilage tissue engineering. In this study, we used expanded hMSCs after chondrogenesis induction as seed cells. We previously determined10 that hMSCs would expand over 15 passages, providing multilineage potential, in contrast to the limited expanding ability of chondrocytes. Engineered cartilage of about 1×1 cm×3 mm can be obtained from less than 0.2 mL of bone marrow aspiration when the isolated hMSCs are expanded for 4 passages.
Previous work has demonstrated that chondrocyte induction could be successfully carried out when hMSCs were cultured in a 3-dimensional state in aggregates in chondrogenic medium.13 However, only small spheres of about 1 mm or less in diameter were formed. The small 3-dimensional state limits the engineering of larger cartilage. Other authors also demonstrated that hMSCs could be successfully induced into chondrocytes when encapsulated in alginate beads in chondrogenic medium.14 In this study, we extended the experiment above13-14 by using polyglycolic acid scaffolds to provide the larger 3-dimensional state, and neocartilages were constructed with 3 weeks of 3-dimensional induction and 10 weeks of in vivo culture.
It is clear that chondrogenesis-induced hMSCs have advantages over chondrocytes isolated from cartilage. Nevertheless, clinical work needs strict quantitative standards. In this study, we presented the relationship between specific-sized tissue-engineered cartilage and hMSCs. The quantitative study addresses problems such as how to construct an engineered cartilage in a determined volume, how much bone marrow needs to be aspirated, how many hMSCs can be isolated from the aspirate, how many times hMSCs should be passaged, how long hMSCs should be induced, and if there are any differences among different passages.
The purity of hMSCs is a primary issue in quantitative research. Flow cytometry is one of the best means to analyze the phenotype and purity of hMSCs.15 The flow cytometry results of our investigations showed that, from passage 4, the purity of hMSCs was greater than 98%, similar to previous findings.5 The quantitative relationship results are reliable as we used almost pure hMSCs.
Measuring the expansion of hMSCs was the core of the quantitative study. Aspiration of bone marrow is a limited source of hMSCs, as we found that only 1 to 2 × 106 hMSCs could be obtained from 1 mL of bone marrow. The cell seeding concentration would be as high as 5 × 107 cells/mL, and 200 μL of the cell suspension would only support about 1 cm2 of cartilage construction. Therefore, hMSCs obtained from bone marrow in a primary culture will not be sufficient for clinical use, and the expansion of hMSCs is necessary. The results of our study showed that hMSCs isolated from 1 mL would expand to as many as 1.4 to 2.8 × 1012 at passage 15 and that theoretically there would be enough cells to construct a piece of cartilage of about 13 to 26 m2.
The induction rate was another important issue in the quantitative study. The expansion ability greatly declined when hMSCs were induced into chondrocytes. A previous study16 showed that hMSCs, when loaded on porous calcium phosphate ceramics and implanted subcutaneously into nude mice, will spontaneously give rise to bone in the absence of differentiation stimuli medium. Therefore, hMSCs should be induced into chondrocytes before implanting them to avoid wasting of hMSCs and the differentiating of hMSCs into unwanted types of cells in an uncontrolled way. Another study17 showed that type II collagen, the marker of chondrocytes, could be detected on the third day of chondrogenesis induction and throughout the cell aggregates on the following days. However, the induction rate has not been well studied before. The induction rate of osteoblasts in rat mesenchymal stem cells could be as high as 98.4%.18 In our study, on days 21 and 28, the chondrogenesis induction rate of hMSCs was greater than 95%, with sufficient purity to support cartilage tissue engineering. Nevertheless, there was no significant difference between the day-21 and day-28 rates; therefore, 21-day inducing is preferable.
This study demonstrates that tissue-engineered cartilage can be constructed using hMSCs after chondrogenesis induction. Because of their high-expansion ability, hMSCs can serve as alternative seed cells for cartilage tissue engineering. We also established quantitative standards between hMSCs and tissue-engineered cartilage.
Correspondence: Yonggang Pang, MD, Department of Otorhinolaryngology, TangDu Hospital, The Fourth Military Medical University, Xi’an 710038, People’s Republic of China (email@example.com).
Accepted for Publication: June 17, 2004.
Funding/Support: This project was supported by grant 30171007 from the National Natural Science Foundation of People’s Republic of China, Beijing.
Previous Presentation: This study was presented at the second International Symposium on Stem Cell Research; December 18, 2003; Beijing, China.