| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
a Children’s Hospital Boston,
b Massachusetts General Hospital,
c Harvard Medical School Center for Minimally Invasive Surgery, Boston, Massachusetts, USA
Key Words. Cord blood cells • Differentiation • Fetal stem cells • In vitro differentiation • Stem/progenitor cell • Stromal cells • TGF-ß1 • Tissue regeneration
Correspondence: Dario O. Fauza, M.D., Children’s Hospital Boston, Department of Surgery, Fegan 3, 300 Longwood Avenue, Boston, Massachusetts 02115, USA. Telephone: 617-919-2966; Fax: 617-730-0910; e-mail: dario.fauza{at}childrens.harvard.edu
 |
ABSTRACT |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
INTRODUCTION |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Severe tracheal malformations and chest wall deformities are two examples in which engineered autologous cartilage readily available at birth, or even for prenatal repair, would be extremely beneficial, if not life-saving. We have previously shown in large animal models that tracheoplasty and chest wall reconstruction using cartilage engineered either from fetal chondrocytes or bone marrow mesenchymal cells may be viable alternatives for the treatment of these birth defects [3–5]. One of the limitations of these previous studies is that neither cell source (fetal auricular biopsy or bone marrow aspiration) is easily accessible. This study was aimed at determining whether three-dimensional (3D) cartilage could be engineered from a more accessible cell source, namely umbilical cord blood (CB), and at comparing it with both engineered fetal cartilage and native tissue.
|
MATERIALS AND METHODS |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Cell and Tissue Harvest
Time-dated pregnant ewes at 80–120 days’ gestation were anesthetized with 2%–4% Halothane (Halocarbon Laboratories, River Edge, NJ, http://www.halocarbon.com). They received 1 g of cefazolin (Cefazolin; G.C. Hanford, Syracuse, NY, http://www.hanford.com) intravenously before surgical manipulation. The bicornuate uterus was exposed through a median longitudinal laparotomy and opened with electrocautery. Fetuses underwent either native cartilage or CB harvest. For native cartilage harvest, the fetal head and neck were exposed. Elastic (n = 7) or hyaline (n = 6) cartilage specimens were then harvested from, respectively, an ear or tracheal rings. For CB harvest, the umbilical cord was exposed and the umbilical vein was cannulated and aspirated using a 22-gauge needle. Approximately 10 ml of CB (n = 4 samples) was collected into high-glucose Dulbecco’s modified Eagle’s medium (DMEM) (Gibco, BRL/Life Technologies, Grand Island, NY, http://www.invitrogen.com/content.cfm?pageid=1) containing 5% preservative-free heparin (Elkins Sinn, Cherry Hill, NJ). All samples were transported on ice and processed within 4 hours after harvest. The gestational membranes and uterine wall were closed in one layer with a reusable TA 90-mm titanium surgical stapler (United States Surgical Corporation, Norwalk, CT, http://www.ussurg.com). The mother’s abdomen was closed in layers.
Cell Isolation and Culture
The cartilage specimens were minced and subsequently digested with 0.2% type II collagenase (Worthington Biochemical, Lake View, NJ, http://www.worthington-biochem.com/default.html) for 5 hours in a 37°C water bath. After passage through a 70-µm mesh (Becton, Dickinson and Company, Franklin Lakes, NJ, http://www.bd.com), chondrocytes were plated and expanded in Ham’s F-12 (Gibco), with 10% fetal bovine serum (Sigma, St. Louis, http://www.sigmaaldrich.com), 5 µg/ml ascorbic acid 2-phosphate (Sigma), and an antimicrobial solution containing 10,000 U/ml penicillin, 10 mg/ml streptomycin, and 25 µg/ml amphotericin B (all from Gibco).
Mononuclear cells were isolated from the CB samples by low-density fractionation/density-gradient centrifugation using Hystopaque (Sigma). Briefly, CB samples were filtered through a 100-µm mesh (BD Falcon, Bedford, MA, http://www.bdbiosciences.com/discovery_labware) and then centrifuged at 2,500 rpm for 5 minutes. The resulting pellet was resuspended with 10 ml of phosphate-buffered saline (PBS) (Gibco). The cell suspension was then added to the top of Hystopaque 1077 (Sigma), and the combination was centrifuged at 2,200 rpm for 30 minutes to obtain the mononuclear cell fraction. The interface between the hystopaque and plasma below was carefully recovered with a pipette and placed in a separate centrifuge tube. The remaining hystopaque, plasma, and red blood cell pellet were discarded. Cells were then washed with PBS and centrifuged at 2,500 rpm for 5 minutes twice, after which they were resuspended in growth/expansion medium and plated into a 25-cm2 flask (Corning, Inc., Corning, NY, http://www.corning.com) at a density of approximately 2 µ 106 per cm2. Growth medium consisted of high-glucose DMEM (Gibco), 10% fetal bovine serum (Sigma), and an antibiotic solution containing 10,000 U/ml penicillin G sodium and 10,000 µG/ml streptomycin sulfate (both from Gibco). After 72 hours, non-adherent cells were removed and the medium was replaced. The remaining adherent cells were expanded in vitro.
All cells were maintained in a 95% humidified, 5% CO2 incubator at 37°C and expanded every 4–6 days at a ratio of 1:2 when at least approximately 80% confluent. Cell detachment was achieved using 0.05% trypsin and 0.53 mM EDTA (Cellgro, Herndon, VA, http://www.cellgro.com) for 5 minutes at 37°C. Cells were passaged 5 to 10 times to achieve enough numbers for seeding into multiple constructs obtained from each cell donor (below).
Construct Assembly
Biodegradable scaffolds (0.5–1 cm2) were used, composed of nonwoven, 2-mm-thick, 70-mg/ml polyglycolic acid polymer (Albany International, Mansfield, MA, http://www.albint.com) treated with 3% poly-L-lactic acid solution (Sigma). Before cell seeding, scaffolds were treated with 1 N NaOH for 1 minute, washed with distilled water, and then coated with 1% collagen solution (Vitrogen Angiotech BioMaterials Corp. [formerly Cohesion Technologies, Inc.], Palo Alto, CA, http://www.cohesiontech.com) for 12 hours at 4°C to enhance cell attachment. Scaffolds were then statically or dynamically seeded at 60 million per cm2 density with either fetal chondrocytes or mesenchymal progenitor cells. Static seeding was performed by suspending the cells in 200 µl of medium. In a six-well plate (Becton, Dickinson and Company), one half of the cells were seeded on the upper surface of the scaffold placed and allowed to attach in the incubator at 37°C for 30 minutes. The construct was then inverted, and the remaining cells were seeded in an identical fashion. After 30 minutes, 5 ml of medium was added to the well and changed daily for 3 days. For dynamic seeding, scaffolds were suspended in a magnetically stirred flask containing the cell suspension (Wheaton, Charlotte, NC) for 3 days at 37°C.
All constructs were subsequently maintained in 50-ml polystyrene centrifuge tubes (Becton, Dickinson and Company) in a rotating hybridization oven, or bioreactor (Thermo Electron, Franklin, MA, http://www.thermohybaid.com) at 37°C and 5 rpm. Chondrocyte-based constructs were kept in the same medium used for cell expansion for 6–8 weeks. CB constructs were maintained in serum-free, defined chondrogenic medium for either 4 (n = 8), 8 (n = 8), or 12 (n = 12) weeks, totaling 28 constructs. There were no significant differences in the number of constructs generated from each of the four donors and in their relative distribution among the three different time points. The medium consisted of high-glucose DMEM (Gibco) supplemented with the following: 10–7 M dexamethasone (Sigma), 50 µg/ml ascorbic acid 2-phosphate (Sigma), antibiotic solution containing 10,000 U/ml penicillin G sodium and 10,000 µg/ml streptomycin sulfate (Gibco), 8 ng/ml transforming growth factor-ß1 (TGF-ß1) (R&D Systems, Minneapolis, http://www.rndsystems.com), 100 µg/ml sodium pyruvate (Sigma), 40 µg/ ml L-proline (Sigma), and insulin transferrin selenium-plus (BD Biosciences, Bedford, MA, http://www.bdbiosciences.com/index.shtml) at the final concentrations of 6.25 µg/ml bovine insulin, 6.25 µg/ml transferrin, 6.25 µg/ml selenious acid, 5.33 µg/ml linoleic acid, and 1.25 µg/ml bovine serum albumin, as previously described [6].
Engineered and Native Tissue Analysis
Histology Specimens were fixed in 10% neutral buffered formalin (Sigma), paraffin embedded, sectioned, and stained with hematoxylin and eosin (H&E), safranin O, and toluidine blue. Resulting slides were examined using an Axiophot light microscope (Carl Zeiss, Jena, Germany, http://www.zeiss.com).
Immunohistochemical analysis was performed for types I, II, and X collagen, using the primary monoclonal antibodies I-8H5 and II-4C11 (both from Accurate Chemical & Scientific Corp., Westbury, NY, http://www.accuratechemical.com/accuratechemical) and a polyclonal antibody to deer antler type X collagen (from Dr. Gary Gibson, Henry Ford Hospital, Detroit) [7], respectively. After incubation with the primary antibodies, immunostaining was detected colorimetrically using a labeled streptavidin biotin kit, with 3-amino-9-ethylcarbazole as the chromagen (DakoCytomation, Carpenteria, CA, http://www.dakocytomation.com), according to the manufacturer’s instructions. Sections were counterstained with Mayer’s hematoxylin (Sigma), coverslipped, and analyzed using an Axiophot light microscope (Carl Zeiss).
Quantitative DNA and Matrix Analyses Samples were lyophilized for 8 hours and weighed. Engineered constructs were minced and native cartilage specimens were sliced into small pieces using a cryostat (Leica Microsystems, Wetzlar, Germany, http://www.leica-microsystems.com/website/lms.nsf).
For the DNA assay, samples were incubated in a proteinase K solution (Sigma) at 55°C for 16 hours. Hoechst 33258 dye solution (Sigma) was used to measure total DNA content under a spectrofluorometer (Turner Designs, Sunnyvale, CA, http://www.turnerdesigns.com) as previously described [8].
For quantitation of glycosaminoglycans (GAGs), each sample was mixed with a 4-M guanidine solution (Sigma) and placed on an orbital shaker at 4°C for 48 hours. Digested samples were dialyzed using a 15-kDa cut-off cellulose membrane (Spectrum Laboratories, Inc., Rancho Dominguez, CA, http://www.spectrapor.com) against a bath of distilled water for 12 hours to remove acid. GAG content was measured spectrophotometrically (Hewlett-Packard, Andover, MA, http://www.hp.com) using a dimethylmethylene blue dye [9] and the Blyscan assay kit (Bio-color, Ltd., Belfast, U.K., http://www.biocolor.co.uk), with a chondroitin sulfate standard.
During the GAG extraction process, acid-soluble collagen (mainly type I) was simultaneously extracted [9]. The remaining samples contained type II, or pepsin-soluble, collagen. Therefore, the samples were subsequently digested with 1% (wt/vol) pepsin (Sigma) in 0.5 M acetic acid (Fisher Scientific, Pittsburgh, https://www1.fishersci.com/index.jsp) on an orbital shaker at 4°C for 24 hours. Type II collagen content was measured spectrophotometrically (Hewlett-Packard) using a collagen dye reagent [10] and the Sircol assay kit (Biocolor), with a provided collagen standard containing 1 µg/µl.
To measure elastin (EL) content, samples were incubated in 0.25 M oxalic acid (Sigma) at 95°C for 1 hour and then cooled to room temperature. This process was repeated four times. Digested samples were then dialyzed using a 3.5-kDa cut-off cellulose membrane (Slide-A-Lyzer Mini Dialysis Unit; Pierce, Rockford, IL, http://www.piercenet.com) against a bath of distilled water for 12 hours to remove acid. EL content was measured spectrophotometrically (Hewlett-Packard) using an EL dye reagent [11] and the Fastin assay kit (Biocolor) with a standard containing 
-EL.
Statistical Analysis
Statistical analysis was performed by one-way analysis of variance (ANOVA) and Student’s t-test using commercially available software (InStat; GraphPad Software, San Diego, http://www.graphpad.com). Taking into account the multiple comparisons between groups, p values of less than .01 were considered significant.
|
RESULTS |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
). Sparsely distributed colonies were visible in the original culture after 7–10 days in growth medium. By 14–21 days, these colonies extended toward each other to approximately 80% confluence.
|
Histology
On H&E, CB constructs displayed no evidence of chondrogenic differentiation after 4 weeks in the bioreactor (Fig. 2). After 8 weeks, cell lacunae and basophilic staining were present in the extracellular matrix, in a morphological pattern compatible with ongoing chondrogenesis (Fig. 2). After 12 weeks, CB constructs exhibited evident chondrogenic differentiation by both standard and matrix-specific staining, displaying characteristics of hyaline cartilage, both grossly and microscopically (Figs. 2, 3). At that time point, each construct also contained a multicellular peripheral layer of flattened, elongated, fibroblast-like cells similar to normal perichondrium (Fig. 2).
|
|
). Immunostaining for type II collagen showed its presence throughout the constructs (Fig. 3). On the other hand, immunostaining for type I collagen revealed positivity only on the periphery of the constructs, in the perichondrium-like layer (Fig. 3). Native fetal hyaline cartilage specimens were essentially negative for type I collagen. Type X collagen was not detected in any of the constructs, whereas minimal staining was present at the border of endochondral ossification in the native specimens.
Quantitative Matrix Analysis
There was a significant time-dependent increase in the levels of GAG and type II collagen (C-II) but not of EL in CB constructs (Fig. 4). The following quantitative analyses refer to CB constructs at the 12-week time point. There were no significant differences in GAG and C-II levels between CB and fetal chondrocyte constructs, which, however, had higher EL levels. Compared with native fetal cartilage (both hyaline and elastic), C-II and EL levels were, respectively, similar and lower in the CB constructs. Native hyaline cartilage had higher GAG levels than CB constructs, which showed GAG contents comparable with those of native elastic cartilage. There were no differences in the results from each of the four donors within each time point.
|
|
DISCUSSION |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The surgical treatment of select congenital anomalies, such as tracheal defects, with cartilage engineered from autologous BMSCs, however, would imply the need for a fetal bone marrow biopsy, which would be considerably invasive. Additionally, both the expansion potential and phenotype of BMSCs are known to change with age, even in the fetus [12, 17]. Recent studies have reported the presence of MPCs in CB and even peripheral blood [18–20]. Campagnoli et al. [18] and Lee et al. [19] have shown that, under suitable conditions, CB MPCs can differentiate into chondrocytes, among other cells. To our knowledge, this is the first report of 3D cartilage engineered in vitro from CB MPCs.
In the present study, we successfully isolated MPCs from ovine CB. A limitation of this experiment was the fact that it was not possible to perform a detailed phenotypical analysis of these cells due to the well-known lack of ovine cross reactivity of the commercially available antibodies to identify reported MPC markers through fluorescence-activated cell sorting. However, the fact that these cells were obtained through the techniques previously described for selective MPC isolation from blood, combined with their easy expansion in a dedicated culture medium for mesenchymal cells as well as their ability to produce cartilage, point to the fact that they were of a mesenchymal nature. To our knowledge, no other cell type could have led to these results. Engineered cartilaginous constructs derived from these cells displayed morphological characteristics comparable to those of native fetal hyaline cartilage and of engineered fetal chondrocyte-derived cartilage, both on gross inspection and histologically. Quantitatively, however, the extracellular matrix profile of CB constructs was unique, with select differences compared with fetal chondrocyte-derived constructs and either hyaline or elastic native cartilage. In our system, CB cells were first grown in monolayer culture and then transferred to a 3D polymer scaffold and maintained in a rotating bioreactor. One could speculate that the ability of the CB MPCs to produce large amounts of EL may never have developed during the monolayer culture phase or that it might have eventually surfaced in 3D culture had the study been carried out to further time points. The significance and applicability of such matrix deposition differences are to be assessed in upcoming in vivo experiments.
Somewhat unexpectedly in our study, the CB MPCs required up to 12 weeks to clearly differentiate into cartilaginous tissue. Previously reported differentiation of BMSCs by micromass culture using TGF-ß1 required only 5 days [6], whereas differentiation of these cells on press-coated polylactic acid polymer blocks took not more than 3 weeks [16]. The relatively large size of our constructs (1 cm2) compared with what was achieved in these BMSC studies may have contributed to our findings. The polymer that we used (polyglycolic acid), as well as the unique seeding and bioreactor conditions, may have also played a role. Indeed, the time course of matrix deposition and cartilage formation using CB MPCs correlated with our previous findings using similar seeding and culture techniques with ovine BMSCs [4]. It has been recently shown that TGF-ß2 and TGF-ß3, independently or in combination, are superior to TGF-ß1 as a means to induce chondrogenic response in MSCs [21–23]. It is reasonable to speculate that these other isoforms of TGF-ß could lead to an improved engineered cartilage, should the CB cells be exposed to them as opposed to TGF-ß1, but this remains to be determined.
Our data suggest that CB could be a somewhat easily accessible source of autologous MPCs for cartilage engineering, at least in this ovine model. Although MPCs have been isolated from human CB at different stages of gestation, it remains to be determined whether their density and stage of differentiation are suitable for 3D cartilage engineering, as shown herein [18, 19]. For example, there are well-known interspecies differences in both the cellular and fluid make-up of human and ovine amniotic fluid [24]. Conceivably, analogous differences could also apply to MPCs present in CB.
Although certainly invasive, cordocentesis can be considered a fairly safe procedure in experienced hands [25, 26]. Previous studies on the isolation of MPCs from CB samples have shown that the timing of harvest impacts on the ability to isolate these cells, with a declining cell yield with advancing gestation [18, 27]. Collection of term CB at delivery is, of course, without any risk to the mother or fetus; however, it remains to be determined whether MPCs can be consistently isolated in sizeable amounts from this source. Finally, fetal cells have been isolated from the maternal circulation for diagnostic purposes [28, 29]. Whether fetal MPCs/MSCs can be isolated from peripheral maternal blood and used for tissue engineering is an intriguing question yet to be answered.
Despite the need for further investigation, as discussed above, the present data show that circulating fetal MPCs can be used for the engineering of 3D cartilaginous tissue in vitro, displaying select histological and functional properties of both native and engineered fetal cartilage. Cartilage engineered from CB may prove useful in the surgical treatment of severe congenital anomalies such as tracheal and chest wall defects.
|
ACKNOWLEDGMENTS |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
REFERENCES |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  D Mrugala, C Bony, N Neves, L Caillot, S Fabre, D Moukoko, C Jorgensen, and D Noel Phenotypic and functional characterisation of ovine mesenchymal stem cells: application to a cartilage defect model Ann Rheum Dis, March 1, 2008; 67(3): 288 - 295. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| STEM CELLS | THE ONCOLOGIST | CME | ALPHAMED PRESS JOURNALS |
