In Vivo Matrix Production by Bone Marrow Stromal Cells Seeded on PLGA Scaffolds for Ligament Tissue Engineering

Original Article

In Vivo Matrix Production by Bone Marrow Stromal CellsSeeded on PLGA Scaffolds for Ligament Tissue Engineering

Floor van Eijk, M.D.,1 Daniel B.F. Saris, M.D., Ph.D.,1 Natalja E. Fedorovich, M.D.,1

Moyo C. Kruyt, M.D., Ph.D.,1 W. Jaap Willems, M.D., Ph.D.,2 Abraham J. Verbout, M.D., Ph.D.,1

Anton C. Martens, Ph.D.,3 Wouter J.A. Dhert, M.D., Ph.D.,1 and Laura Creemers, Ph.D.1

Ligament tissue engineering based on cell-seeded biomechanically functional constructs is a commonly studiedstrategy toward native anterior cruciate ligament replacement. Little is known about the survival and differ-entiation of the seeded cells after the transplantation. We applied retroviral genetic marking to trace implantedcells and studied their differentiation by species-specific immunolabeling of the extracellular matrix produced.Goat bone marrow stromal cells were transduced with a MoMuLV-based vector encoding the DLNGFR gene.Transduced cells were seeded onto poly(lactic-co-glycolic acid) (PLGA) fibers and implanted subcutaneously intonude mice and left for various periods up to 6 weeks. Immunohistochemistry for LNGFR expression showedsurvival of the seeded cells after transplantation for up to 6 weeks. Immunohistochemistry for collagen type Iand III showed the production of fibrous tissue inside the scaffolds. Moreover, using a goat-specific anti-collagentype III, donor-derived matrix could be demonstrated. We conclude that bone marrow stromal cells survived invivo and at least partially differentiated after implantation.

Introduction

One of the challenges in the field of regenerativemedicine is engineering a ligament with the character-

istics of a native anterior cruciate ligament (ACL).1–8 TheACL is an intraarticular structure that has poor healing ca-pacities and requires surgical replacement in patients withpersistent knee instability. Current surgical strategies such asthe use of auto- or allografts have several drawbacks, such asdonor-site morbidity and risk of disease transmission.9–11

Engineering ligament grafts by combining autologous cellswith a biodegradable and biomechanically functional scaf-fold is an interesting strategy to bypass these drawbacks.Differentiated6,12 as well as undifferentiated13,14 cell typeshave been used in tissue engineering of tendon and liga-ments. Promising candidates are bone marrow stromal cells(BMSCs), which have been studied extensively and can becultured and directed toward several lineages of differenti-ation, including tendon and ligament.15 Previously, it wasshown that BMSCs seeded and cultured on poly(lactic-co-glycolic acid) (PLGA) scaffolds proliferate more readily andproduce more matrix as compared to more differentiated celltypes, including ACL fibroblasts, and therefore may be moresuitable for tissue engineering of the ACL.16

Although many in vitro and in vivo studies have focusedon tissue engineering of ligaments, little is known about thefate of the seeded cells after the transplantation. By use ofphysical separation from donor and host tissue and selectivedestruction of native cells, it was shown that fibroblasts in areconstructed patellar tendon autograft for ACL reconstruc-tion in a rabbit model died after reconstruction in the kneejoint and the graft was rapidly repopulated by host cells4 weeks after transplantation.17,18 Rabbit skin and ACL fi-broblasts labeled with PKH26-GL and seeded on collagenscaffolds were found to be present in the intraarticular en-vironment after 6 weeks.19 However, as is a common prob-lem with fluorescent membrane labels, label transfer andsignal dilution hamper proper data interpretation.

A means to unequivocally study the presence of donorcells in an implanted construct is retroviral transduction withnonnative marker genes. LacZ (b-galactosidase) labeling hasbeen applied to trace allografted and autografted cells in tworabbit models of ligament repair and autografting,20,21 buttill now, no tissue-engineered ligament constructs have beenevaluated for the presence of genetically labeled donor cellsafter implantation.

Another issue in addition to the actual presence of viabledonor cells in vivo is their functionality. It is not known

1Department of Orthopaedics, University Medical Center Utrecht, Utrecht, The Netherlands.2OLVG, Department of Orthopaedics, Amsterdam, The Netherlands.3Department of Haematology, University Medical Center Utrecht, Utrecht, The Netherlands.

TISSUE ENGINEERING: Part AVolume 15, Number 10, 2009ª Mary Ann Liebert, Inc.DOI: 10.1089=ten.tea.2008.0541

3109

whether implanted cells actively take part in tissue forma-tion, let alone whether successful tissue engineering actuallyrequires the implanted cells to produce the desired matrixin vivo. To demonstrate that transplanted cells do not onlysurvive but also produce extracellular matrix, the origin ofthe deposited tissue should be identified. One option toaddress this problem is to use species-specific recognitionof proteins produced by the donor cells.

The objective of current study was to trace BMSC cellsseeded onto braided PLGA scaffolds and to demonstratetheir viability and functionality in terms of matrix produc-tion in vivo. Therefore, as a first step to demonstrate theapplicability of this method, goat BMSCs were transducedwith the DLNGFR gene, seeded on braided PLGA scaffolds,and implanted in an ectopic nude mice model. The origin ofthe produced matrix inside the scaffolds was qualitativelystudied by species-specific immunostaining of collagen.

Materials and Methods

Cell culture

Goat BMSCs from bone marrow aspirates of a femaleDutch milk goat (weight 60 kg, 19–24 months old) wereharvested and expanded as described earlier.22 The cellswere cryopreserved at passage 1. Within 6 months, the goatBMSCs (passage 2) were thawed and replated in standardculture medium consisting of aMEM supplemented with30% fetal bovine serum (FBS; Gibco, Paisley, Scotland, lot#3030960S), L-glutamine (2 mM; Invitrogen, San Diego, CA),ascorbic acid (0.2 mM; Sigma, St Louis, MO), and penicillin=streptomycin (each 100 U=mL; Gibco). When confluent, thecells were detached and replated at 5000 cells=cm2 in culturemedium containing 15% FBS.

The Phoenix-Ampho retrovirus packaging cell line wascultured in Dulbecco’s modified Eagle’s medium (Gibco)supplemented with 10% FBS, penicillin (100 U=mL), strep-tomycin (100 mg=mL), and 2 mM L-glutamine. Cultures werepassaged twice a week and selected for gag, pol, env ex-pression every 8 weeks using hygromycin B (300 mg=mL;Roche diagnostics, Mannheim, Germany) and diphtheriatoxin A (1mg=mL; Sigma-Aldrich, Zwijndrecht, The Nether-lands). The construction of the retroviral vector pLZRS-TKhas been described previously.23 Twenty micrograms ofDNA of a retroviral plasmid construct was transfected into70% confluent Phoenix-Ampho packaging cells by calciumphosphate precipitation.24 Twenty-four hours after transfec-tion, medium was replaced with fresh culture medium. Thefollowing day, retroviral supernatant was collected, filteredthrough a 0.45 mm filter, and stored at �808C. For additionalharvest of retroviral supernatant, transfected Phoenix-Ampho cells were cultured for 3 days in the presence ofpuromycin (1 mg=mL; Sigma) followed by 2 days in culturemedium without puromycin.

Retroviral transduction of goat BMSCs

Transduction was performed by replacing the standardcultured medium of the BMSCs (P3) by a threefold dilutionof the retroviral supernatant in culture medium supple-mented with 6 mg=mL polybrene (Sigma). Culture mediumsupplemented with polybrene alone was used for mock-transduction (control). The BMSCs were cultured for another

24 h, after which the medium was renewed. Two days laterthe cells were harvested and analyzed.

Transduction efficiency was determined by flow cytome-try analysis. The BMSCs were labeled with primary mouseanti-LNGFR monoclonal antibody (20.4 culture supernatant,1:20) for 20 min at 48C, washed, and then incubated withgoat anti-mouse phycoerythrine-conjugated IgG1 (SouthernBiotechnologies, Birmingham, AL). Cells were washed inPBS–1% FBS and resuspended in standard culture medium,immediately before analysis of 10,000 events on a flow cyt-ometer (FACS Calibur; Becton and Dickinson, San Jose,CA).23,25 In vitro culture was conducted to investigate long-term label expression. Two separately produced batches oftransduced cells were subcultured further in 25 cm2 flasks for6 weeks and analyzed by FACS at each passage. The re-mainder of the labeled BMSCs were cultured until conflu-ence and subsequently seeded.

Transduction efficiency and long-term stability of cells

In a separate study in vitro culture was conducted to in-vestigate long-term label expression.25 Two batches oftransduced cells were cultured in 25 cm2 flasks for 6 weeksand analyzed by FACS at each passage. To assess prolifer-ative capacity, transduced and mock-transduced cells werecocultured in a 50%=50% mixture in a separate study.25

Scaffolds

The degradable scaffolds were made of Panacryl (size 2=0;Ethicon, Somerville, NJ), which is composed of poly(L-lactide=glycolide) 11 mm bundles, containing approximately 40monofilaments.16 Nine bundles were organized in parallel,and both ends of the structure were sutured together (1 cm inlength). The scaffolds were sterilized by repeated cycles ofincubation in ethanol 70%, and after air-drying washed withsterile PBS. Before seeding, the scaffolds were preincubatedovernight in standard culture medium.

Cell seeding

After trypsinization, both transduced and control BMSCs(P4) were resuspended at 2.5�106 cells per mL in standardculture medium. The scaffolds were seeded with 250,000cells=scaffold and allowed to attach for 2 h, after which thescaffolds were submerged in culture medium consisting ofaMEM supplemented with 15% FBS, L-glutamine (2 mM),ascorbic acid (0.2 mM), and penicillin=streptomycin (each100 U=mL). All scaffolds were cultured for 7 days to allowfor proliferation of seeded cells in 25-well bacterial cultureplates (Sterilin, Stone, UK) before implantation.

Surgical procedure

All animal experiments were conducted with approval ofthe ethics committee of the University of Utrecht. Ten femaleNMRI nude mice (age, 4–6 weeks; weight, 20–25 g; Harlan,Horst, The Netherlands) were anaesthetized with an intra-peritoneal ketamine (100 mg=mL), xylazine (20 mg=mL), andatropine (0.5 mg=mL) solution. After disinfection of the backwith 70% alcohol, separate subcutaneous pockets were cre-ated for implantation of the scaffolds on the back of the mice.Implantations included two mice per evaluation period andtwo scaffolds with transduced cells and one scaffold with

3110 VAN EIJK ET AL.

mock-transduced cells in each mouse (in addition to threetissue-engineered bone constructs that were implanted in thecourse of a separate study). The mice were sacrificed aftertime periods of 2, 4, and 10 days, and after 4 and 6 weeks.The implants were retrieved and processed for histologicalevaluation.

Histology

Samples were fixed in 10% buffered formalin. Dehydra-tion was performed through a graded alcohol series beforeparaffin embedding. Four-micrometer-thick longitudinal sec-tions of the scaffolds were deparaffinized, rehydrated, stainedwith hematoxylin and eosin, and analyzed with light mi-croscopy (E600 Nikon Eclipse, Nikon, Japan). In addition,parts of the sections were stained for 60 min in PicrosiriusRed (Direct Red 80; Aldrich, Milwaukee, WI). The presenceof a collagenous matrix was viewed using polarization mi-croscopy (E600 Nikon Eclipse).26–28 Collagen type I appearsas thick, bright (strongly birefringent) yellow or red fila-mentous structures shining against a dark background,whereas collagen type III highlights as thin filamentousstructures with a weak birefringence of a greenish color.28

Immunohistochemistry

Paraffin-embedded scaffolds were deparaffinized and re-hydrated, and endogenous peroxidase activity was blockedwith 1.5% H2O2 in phosphate-citrate buffer for all samples.The samples were stained for LNGFR-positive cells and thepresence of collagen type I and III to asses survival andfunctionality of the transplanted cells.

For the LNGFR immunolocalization, the Dako Ark� kit wasapplied according to the manufacturers’ recommendations(Dako Corporation, Carpinteria, CA) to prevent nonspecificstaining due to reactivity of the secondary anti-mouse IgGantibodies with the surrounding murine tissue. For the pri-mary antibody a mouse anti-LNGFR (2.56mg=mL; DakoCorporation) was used.

For collagen immunodetection the sections were firsttreated with pepsin for 15 min at 378C. Subsequently, thesamples were incubated with rabbit anti-human collagentype I (26.7 mg=mL) or collagen type III (1:200; both Biogen-esis, Dusseldorf, Germany) for 1 h at room temperature (RT).As secondary antibody, ready-to-use PowerVision (poly-HRP-anti-rabbit IgG; Immunologic, Duiven, The Netherlands) wasapplied according to the manufacturers’ recommendations.Diaminobenzidine (Liquid DABþ Substrate; Dako Corpora-tion) was used to obtain a signal on the sections. To verifywhether the primary antibodies for collagen type I and IIIcould be used for species-specific immunolabeling, controlsections containing goat muscle and skin and mouse skin andtendon tissue were labeled with each antibody.

An additional type III immunolocalization with an anti-body known to detect both goat and mouse tissue was per-formed to demonstrate the presence of collagen type III inmouse tissue. Sections of mouse tendon were blocked with1.5% H2O2 in phosphate-citrate buffer and subsequentlytreated with pepsin for 15 min at 378C. Subsequently, thesamples were incubated with rabbit anti-rat collagen type III(1:80; Division of Morwell Diagnostics, Zurich, Switzerland)for 1 h at RT. As secondary antibody, a biotinylated goatanti-rabbit (3mg=mL; Vector laboratories) was applied for 1 h

at RT. Subsequently, streptavidin horse radish peroxidase(SA-HRP; Beckman Coulter, Fullerton, CA) was applied for1 h at RT. Finally, diaminobenzidine was used to obtain asignal in matrix.

All samples were counterstained with hematoxylin andanalyzed by light microscopy (E600 Nikon Eclipse).

Results

Transduction efficiency and stability of labeled cells

The genetic labeling procedure of goat BMSCs resulted ina 40–60% transduction efficiency, measured 3 days aftertransduction. In vitro long-term culture showed a slight de-crease of the percentage of labeled cells in the first week, butit stabilized at 30–40% positive cells for up to 6 weeks. In thecoculture experiment the actual percentage of labeled cellsshowed to be initially lower during the first week, indicat-ing a relative decreased proliferation compared to the unla-beled cell fraction. However, during the subsequent weeks,the percentage transduced cells in the coculture reachedthe expected percentage, indicating a minimal influence oncell proliferation and viability at the long term.25 Viabilityimmediately before seeding was in the range for normalBMSC P3 cells. Bone formation of the transduced BMSCswas described in a separate study,25 showing their multi-potency.

Histology

All mice survived the implantation periods and all sampleswere retrieved without signs of infection. As demonstrated bylight microscopy of the paraffin-embedded scaffolds, cells andextracellular matrix were present throughout the wholescaffold (Fig. 1). At day 10 the cells and tissue in and aroundthe fibers of the scaffold showed alignment along the fibers.At this time also small blood vessels were seen in the center ofthe scaffold. With time, the tissue inside the scaffold appearedmuch denser as demonstrated by the hematoxylin and eosinstaining. Although no quantitative analysis was done, the celldensity in this formed tissue appeared comparable to theearlier time points. Picrosirius Red staining likewise demon-strated a comparable increase in collagen matrix inside thescaffolds in time.

Polarization analysis indicated that both collagen type Iand III could be seen inside the scaffolds, with collagen typeIII also seen adjacent to the scaffolds (Fig. 2).

Immunohistochemistry for detectionof LNGFR expression

Labeled cells could be traced unequivocally from day 2 upto 6 weeks after implantation. The labeled cells were spreadhomogenously throughout the scaffold, without a specificpreference for localization (Fig. 3). Due to the small samplesize, no statements can be made about differences in celldensity in time. Control scaffolds with mock-transduced cellsand sections incubated with secondary antibody only werealways negative (data not shown).

Collagen I labeling

Both mouse tendon and skin and goat skin tissue showedimmunoreactivity for collagen type I. Inside the scaffolds

IN VIVO MATRIX PRODUCTION BY BMSCS FOR LIGAMENT TE 3111

there was a very dark positive staining that increased fromday 10 to day 42. Around the scaffolds the tissue stainedlightly (Fig. 4).

Collagen III labeling

Immunohistochemistry for collagen type III using therabbit anti-human antibody demonstrated a goat tissue-specific staining in goat muscle and skin sections (Fig. 5A),with mouse skin and tendon tissue sections consistentlynegative (Fig. 5B). Using the rabbit anti-rat antibody, colla-gen type III could be demonstrated in both the above-mentioned mouse tissues, showing collagen type III is indeedpresent in the mouse skin and tendon (Fig. 5C). Staining withthe goat-specific rabbit anti-human antibody collagen typeIII could be specifically identified inside the scaffold (Fig.5D–G). Collagen type III could be identified from day 10 upto 6 weeks after implantation. The control tissue (PBS, mousetissue) was always negative.

Discussion

In the current study we demonstrate the capacity ofBMSCs to survive in vivo in a tissue-engineered construct forthe replacement of ligamentous structures. Moreover, to ourknowledge, this is the first study showing that this contri-bution represents more than just stimulating invading hostcells by the release of factors, known to be produced byBMSCs, but that the cells actually also produce matrix. Afterseeding on PLGA scaffolds, goat BMSCs survived for up to 6weeks when transplanted subcutaneously in nude mice. Atleast part of the extracellular matrix was indeed produced bythe transplanted goat BMSCs as shown by species-specificlabeling of type III collagen, thus indicating some form ofdifferentiation. Conventional and Picrosirius Red histologicalstaining suggested an increase in tissue density within thescaffold in time.

The first aim of current study was to follow the fate ofseeded DLNGFR-transduced goat BMSCs implanted in an

FIG. 2. Polarization micros-copy of scaffolds cultured for10 days (A) and 6 weeks (B),stained with Picrosirius Red.Red staining denotes collagentype I and green staining de-notes type III. The scaffoldsappear as white fibers. Colorimages available online atwww.liebertonline.com=ten.

FIG. 1. Histology of scaffolds stained with hematoxylin and eosin (B–F) or hematoxylin with Picrosirius Red (A and D).Small blood vessels were seen in the center of the scaffold (closed arrow, B). From day 10 the cells and tissue around the fibersshowed alignment (C). In time, the tissue inside the scaffold appeared much denser as demonstrated by the hematoxylin andeosin staining (B–C, 10 days, in contrast with E–F, 6 weeks). A fiber of the scaffold is indicated with an open arrow (B–D). Thedense tissue around the fibers of the scaffold showed alignment after 6 weeks (F). Color images available online atwww.liebertonline.com=ten.


ectopic nude mice model. LNGFR has been applied suc-cessfully for tracing many cell types, by both flow cytometryand immunohistochemistry.23,25,29–31 Transplantation of ge-netically modified autologous cells may be more challengingin more clinically relevant models of immunocompetent,larger animal models such as in a goat. However, severalin vivo studies have already been conducted introducingforeign genes, such as luciferase and GFP in vivo without anyserious immunological reaction being reported.32,33 Previousresearch aiming at tracing transplanted and transduced cellsshowed variable survival periods for such cells. The densityof GFP-transduced allogeneic chondrocytes embedded inalginate and transplanted into osteochondral defects in rab-bits decreased to 15% within 4 weeks.33 EGFP-labeled allo-geneic chondrocytes transplanted into a similar rabbit modelcould not be detected after 6 weeks anymore.34 It was un-clear whether a decrease in density of labeled cells was dueto cell death, through rejection by the host cells, or due to lossof expression. Also, the process of transduction itself couldhave had an effect on the behavior of transplanted cells. Both

studies showed a down-regulation of cartilage gene expres-sion of the transfected cells in vitro, but neither study in-vestigated to what extent this down-regulation changed theactual functionality of the cells after transplantation. In thecurrent study, the effect of transduction on in vitro cell via-bility and proliferative capacity in current study was mini-mal, as shown previously with coculture experiments onproliferation of labeled and unlabeled cells.25 Still it cannotbe excluded that transduction had some effects on in vivodifferentiation. The method used to isolate the BMSCs hasbeen shown to generate BMSCs capable of differentiationinto the chondrogenic, osteogenic, and adipogenic cell linein vitro.22,35,36 Also, the multipotency of the same transducedpopulation used in this study has been shown by their in-duction of bone formation in vivo.25

Despite unequivocal demonstration of donor cells in agiven in vivo situation, nothing can be concluded concerningthe actual functionality of the implanted cells. The presenceof labeled cells inside osteocyte lacunae may be regarded asproof of the bone-forming capacity of these cells25,37;

FIG. 3. Immunohisto-chemistry for LNGFR-labeledcells at day 2 (A) and 6 weeks(B, C). Brown staining denotesthe presence of LNGFR-positive cells. Controls werealways negative (D). Colorimages available online atwww.liebertonline.com=ten.

FIG. 4. Immunohisto-chemistry for collagen type Iinside the scaffolds at 6 weeks(A). Brown staining denotesthe presence of collagen typeI. Controls were always neg-ative (B). Color imagesavailable online at www.liebertonline.com=ten.


however, in other tissues such as ligament, it is more difficultto pinpoint the origin of the produced matrix. The above-mentioned study on the fate of transgenic transplantedchondrocytes into articular defects did not show integrationof these cells into the repair tissue, and therefore the authorsconcluded that the cells did not form the repair tissue.33 Inthe current study, in addition to showing the presence of thetransplanted cells, the second aim was to prove their differ-entiation by species-specific immunostaining of collagentype III produced after xenotransplantation. This is an im-portant step toward demonstrating the functionality andthus the need for implanted cells in orthotopic locations. Asto the application of the biomaterials used, PLGA-basedconstructs have been evaluated extensively before. Theywere found to allow for efficient attachment and prolif-eration of and matrix production by seeded goat BMSCsin vitro.16,38 Maximum loads for braided constructs were

shown to be around 60 (PLLAþPLGA)39 to 100 N (PLGA orPLLA),40 with ultimate tensile strength decreasing to 50%after 16 weeks in vitro. However, the maximal tensile load ofPLA=PLGA constructs after 20 weeks in vivo had decreasedto about 10% of native ligaments, indicating biomechanicalfunctionality at later time-points is only partial and stressingthe need for efficient stimulation of matrix deposition afterimplantation.41

In addition to collagen type III, production of collagentype I inside the scaffolds was clearly demonstrated, al-though we were not able to show that this was formed by thetransplanted cells. Collagen type I and III are found in boneand in soft connective tissues like skin and are major extra-cellular matrix components of tendon and ligaments.42

Wound healing in general is associated with relatively highcontents of type III collagen, which has also been shown forregenerating tendon tissue,43 and therefore the tissue formed

FIG. 5. Immunohistochemistry for collagen type III for goat skin (A) and mouse skin (B) using the rabbit anti-humanantibody. Mouse tissue showed negative staining (B), indicating that this antibody was specific for collagen type III formedby goat cells. Immunohistochemistry for collagen type III for mouse ligament using the rabbit anti-rat antibody (C) indicatingthat collagen type III was indeed present in these mouse tissues. Immunohistochemistry using the species specific rabbit anti-human antibody showing collagen type III inside the scaffolds at 4 days (D, 1:100), 10 days (E, 1:200), 4 weeks (F, 1:200) and 6weeks (G, 1:200) indicating that transplanted goat cells produced collagen type III at all these time points. Controls werealways negative (H) Color images available online at www.liebertonline.com=ten.


by the labeled cells may partly reflect scar tissue. As themajor goal of this study was to investigate the in vivo be-havior of BMSCs implanted on biomaterial scaffolds forligament reconstruction, rather than the production of liga-ment tissue, no efforts were undertaken to further specify thenature of the tissue. Putative markers further identifying thetissue as ligament-like would have been fibronectin, tenascin-C, scleraxis, and tenomodulin, but in the end none of theseare truly specific, as they are also found in other tissues andcells such as cartilage, dermis, testis, and sclerae.44–50 Forfuture in vivo studies on the functionality of transplantedcells, a combination of the markers mentioned above, as wellas the ratio of collagen type I versus III, may prove someinsight into the nature of the tissue formed. Also, the uniqueorientation of the collagen fibers in ligaments and tendonsresulting in its high mechanical properties are importantdeterminants to investigate. However, as has been shownextensively that under unstrained conditions extracellularmatrix is deposited without any particular orientation,4,13 thelatter parameter is only meaningful when analyzed on con-structs that actually have been biomechanically loaded in vivo,in contrast to the current study.

Still the lack of a subset of ligament markers in any givenconstruct at an ectopic location may not preclude function-ality at orthotopic locations. Exposure to the proper biome-chanical cues in vivo may be sufficient to direct differentiationinto the tenogenic=ligament lineage, as has been shownfor fibroblasts differentiating into functional tendon tissuein vivo.51,52

In conclusion, we demonstrate that BMSCs survive in vivotransplantation on PLGA scaffolds and are able to producecollagenous matrix inside these scaffolds designed for tissueengineering of the ACL. Future studies are required to showwhether this also occurs in seeded constructs implanted atorthotopic locations and will need to address the final bio-mechanical functionality of these constructs.

Acknowledgments

The authors acknowledge Mirjam Stijns for her help withthe cell labeling and The Netherlands Technology Founda-tion (STW; grant UGN.4966) for financial support. Dr. Sarisis supported by the Netherlands Organisation for HealthResearch and Development (NWO) and Dr. Creemers by theDutch Arthritis Association.

Disclosure Statement

No competing financial interests exist.

References

1. Altman, G.H., Horan, R.L., Lu, H.H., Moreau, J., Martin, I.,Richmond, J.C., and Kaplan, D.L. Silk matrix for tissue en-gineered anterior cruciate ligaments. Biomaterials 23, 4131,2002.

2. Awad, H.A., Butler, D.L., Harris, M.T., Ibrahim, R.E., Wu,Y., Young, R.G., Kadiyala, S., and Boivin, G.P. In vitrocharacterization of mesenchymal stem cell-seeded collagenscaffolds for tendon repair: effects of initial seeding densityon contraction kinetics. J Biomed Mater Res 51, 233, 2000.

3. Bourke, S.L., Kohn, J., and Dunn, M.G. Preliminary devel-opment of a novel resorbable synthetic polymer fiber scaf-

fold for anterior cruciate ligament reconstruction. Tissue Eng10, 43, 2004.

4. Cao, D., Liu, W., Wei, X., Xu, F., Cui, L., and Cao, Y. In vitrotendon engineering with avian tenocytes and polyglycolicacids: a preliminary report. Tissue Eng 12, 1369, 2006.

5. Cooper, J.A., Lu, H.H., Ko, F.K., Freeman, J.W., and Laur-encin, C.T. Fiber-based tissue-engineered scaffold for ligamentreplacement: design considerations and in vitro evaluation.Biomaterials 26, 1523, 2005.

6. Lin, V.S., Lee, M.C., O’Neal, S., McKean, J., and Sung, K.L.Ligament tissue engineering using synthetic biodegradablefiber scaffolds. Tissue Eng 5, 443, 1999.

7. Lu, H.H., Cooper, J.A., Jr., Manuel, S., Freeman, J.W., Atta-wia, M.A., Ko, F.K., and Laurencin, C.T. Anterior cruciateligament regeneration using braided biodegradable scaf-folds: in vitro optimization studies. Biomaterials 26, 4805,2005.

8. Sahoo, S., Ouyang, H., Goh, J.C., Tay, T.E., and Toh, S.L.Characterization of a novel polymeric scaffold for potentialapplication in tendon=ligament tissue engineering. TissueEng 12, 91, 2006.

9. Crawford, C., Kainer, M., Jernigan, D., Banerjee, S., Fried-man, C., Ahmed, F., and Archibald, L.K. Investigation ofpostoperative allograft-associated infections in patientswho underwent musculoskeletal allograft implantation. ClinInfect Dis 41, 195, 2005.

10. Jackson, D.W., Corsetti, J., and Simon, T.M. Biologic incor-poration of allograft anterior cruciate ligament replacements.Clin Orthop 324, 126, 1996.

11. Kartus, J., Movin, T., and Karlsson, J. Donor-site morbidityand anterior knee problems after anterior cruciate liga-ment reconstruction using autografts. Arthroscopy 17, 971,2001.

12. Cooper, J.A., Jr., Bailey, L.O., Carter, J.N., Castiglioni, C.E.,Kofron, M.D., Ko, F.K., and Laurencin, C.T. Evaluation ofthe anterior cruciate ligament, medial collateral ligament,Achilles tendon and patellar tendon as cell sources for tissue-engineered ligament. Biomaterials 27, 2747, 2006.

13. Altman, G.H., Horan, R.L., Martin, I., Farhadi, J., Stark, P.R.,Volloch, V., Richmond, J.C., Vunjak-Novakovic, G., andKaplan, D.L. Cell differentiation by mechanical stress.FASEB J 16, 270, 2002.

14. Moreau, J.E., Chen, J., Horan, R.L., Kaplan, D.L., and Alt-man, G.H. Sequential growth factor application in bonemarrow stromal cell ligament engineering. Tissue Eng 11,

1887, 2005.15. Hildebrand, K.A., Jia, F., and Woo, S.L. Response of donor

and recipient cells after transplantation of cells to the liga-ment and tendon. Microsc Res Tech 58, 34, 2002.

16. Eijk van, F., Saris, D.B., Riesle, J., Willems, W.J., van Blit-terswijk, C.A., Verbout, A.J., and Dhert, W.J. Tissue engi-neering of ligaments: a comparison of bone marrow stromalcells, anterior cruciate ligament, and skin fibroblasts as cellsource. Tissue Eng 10, 893, 2004.

17. Kleiner, J.B., Amiel, D., Roux, R.D., and Akeson, W.H. Ori-gin of replacement cells for the anterior cruciate ligamentautograft. J Orthop Res 4, 466, 1986.

18. Kleiner, J.B., Amiel, D., Harwood, F.L., and Akeson, W.H.Early histologic, metabolic, and vascular assessment of an-terior cruciate ligament autografts. J Orthop Res 7, 235, 1989.

19. Bellincampi, L.D., Closkey, R.F., Prasad, R., Zawadsky, J.P.,and Dunn, M.G. Viability of fibroblast-seeded ligamentanalogs after autogenous implantation. J Orthop Res 16,

414, 1998.


20. Hildebrand, K.A., Deie, M., Allen, C.R., Smith, D.W.,Georgescu, H.I., Evans, C.H., Robbins, P.D., and Woo, S.L.Early expression of marker genes in the rabbit medialcollateral and anterior cruciate ligaments: the use of dif-ferent viral vectors and the effects of injury. J Orthop Res17, 37, 1999.

21. Martinek, V., Latterman, C., Usas, A., Abramowitch, S.,Woo, S.L., Fu, F.H., and Huard, J. Enhancement of tendon-bone integration of anterior cruciate ligament grafts withbone morphogenetic protein-2 gene transfer: a histologicaland biomechanical study. J Bone Joint Surg Am 84-A, 1123,2002.

22. Kruyt, M.C., de Bruijn, J.D., Wilson, C.E., Oner, F.C., vanBlitterswijk, C.A., Verbout, A.J., and Dhert, W.J. Viable os-teogenic cells are obligatory for tissue-engineered ectopicbone formation in goats. Tissue Eng 9, 327, 2003.

23. Weijtens, M., van, S.A., Hagenbeek, A., Braakman, E., andMartens, A. Reduced graft-versus-host disease-inducing ca-pacity of T cells after activation, culturing, and magnetic cellsorting selection in an allogeneic bone marrow transplanta-tion model in rats. Hum Gene Ther 13, 187, 2002.

24. Kinsella, T.M., and Nolan, G.P. Episomal vectors rapidlyand stably produce high-titer recombinant retrovirus. HumGene Ther 7, 1405, 1996.

25. Kruyt, M.C., Stijns, M.M., Fedorovich, N.E., de Bruijn, J.D.,van Blitterswijk, C.A., Verbout, A.J., Rozemuller, H., Ha-genbeek, A., Dhert, W.J., and Martens, A.C. Genetic markingwith the DeltaLNGFR-gene for tracing goat cells in bonetissue engineering. J Orthop Res 22, 697, 2004.

26. Junqueira, L.C., Cossermelli, W., and Brentani, R. Differ-ential staining of collagens type I, II and III by Sirius Redand polarization microscopy. Arch Histol Jpn 41, 267,1978.

27. Junqueira, L.C., Bignolas, G., and Brentani, R.R. Picrosiriusstaining plus polarization microscopy, a specific method forcollagen detection in tissue sections. Histochem J 11, 447,1979.

28. Montes, G.S. Structural biology of the fibres of the collage-nous and elastic systems. Cell Biol Int 20, 15, 1996.

29. Arts, C.H., Hedeman Joosten, P.P., Blankensteijn, J.D., Staal,F.J., Ng, P.Y., Heijnen-Snyder, G.J., Sixma, J.J., Verhagen,H.J., de Groot, P.G., and Eikelboom, B.C. Contaminantsfrom the transplant contribute to intimal hyperplasia asso-ciated with microvascular endothelial cell seeding. Eur JVasc Endovasc Surg 23, 29, 2002.

30. Bonini, C., Ferrari, G., Verzeletti, S., Servida, P., Zappone, E.,Ruggieri, L., Ponzoni, M., Rossini, S., Mavilio, F., Traversari,C., and Bordignon, C. HSV-TK gene transfer into donorlymphocytes for control of allogeneic graft-versus-leukemia.Science 276, 1719, 1997.

31. Kolen, S., Weijtens, M., Hagenbeek, A., van, S.A., Smulders,S., de, W.R., de, W.T., Dolstra, H., van de Wiel vanKemenade, E., and Martens, A. Monitoring of developinggraft-versus-host disease mediated by herpes simplex virusthymidine kinase gene-transduced T cells. Hum Gene Ther14, 341, 2003.

32. Martinek, V., Usas, A., Pelinkovic, D., Robbins, P., Fu, F.H.,and Huard, J. Genetic engineering of meniscal allografts.Tissue Eng 8, 107, 2002.

33. Mierisch, C.M., Wilson, H.A., Turner, M.A., Milbrandt, T.A.,Berthoux, L., Hammarskjold, M.L., Rekosh, D., Balian, G.,and Diduch, D.R. Chondrocyte transplantation into articularcartilage defects with use of calcium alginate: the fate of thecells. J Bone Joint Surg Am 85-A, 1757, 2003.

34. Milbrandt, T., Berthoux, L., Christenson, V., Baumbusch, C.,Rekosh, D., Balian, G., and Diduch, D. Tracing transducedcells in osteochondral defects. J Pediatr Orthop 23, 430, 2003.

35. Bosnakovski, D., Mizuno, M., Kim, G., Takagi, S., Okumura,M., and Fujinaga, T. Chondrogenic differentiation of bovinebone marrow mesenchymal stem cells (MSCs) in differenthydrogels: influence of collagen type II extracellular matrixon MSC chondrogenesis. Biotechnol Bioeng 93, 1152, 2006.

36. Petrenko, Y.A., Petrenko, A.Y., Damshkaln, L.G., Volkova,N.A., and Lozinsky, V.I. Growth and adipogenic differenti-ation of mesenchymal stromal bone marrow cells duringculturing in 3D macroporous agarose cryogel sponges. BullExp Biol Med 146, 129, 2008.

37. Cui, Q., Ming, X.Z., Balian, G., and Wang, G.J. Comparisonof lumbar spine fusion using mixed and cloned marrowcells. Spine 26, 2305, 2001.

38. Jenner, J.M., van Eijk, F., Saris, D.B., Willems, W.J., Dhert,W.J., and Creemers, L.B. Effect of transforming growth fac-tor-Beta and growth differentiation factor-5 on proliferationand matrix production by human bone marrow stromal cellscultured on braided poly lactic-co-glycolic acid scaffolds forligament tissue engineering. Tissue Eng 13, 1573, 2007.

39. Ge, Z., Goh, J.C., Wang, L., Tan, E.P., and Lee, E.H. Char-acterization of knitted polymeric scaffolds for potential usein ligament tissue engineering. J Biomater Sci Polym Ed 16,

1179, 2005.40. Lu, H.H., Cooper, J.A., Jr., Manuel, S., Freeman, J.W., Atta-

wia, M.A., Ko, F.K., and Laurencin, C.T. Anterior cruciateligament regeneration using braided biodegradable scaffolds:in vitro optimization studies. Biomaterials 26, 4805, 2005.

41. Ge, Z., Goh, J.C., and Lee, E.H. The effects of bone marrow-derived mesenchymal stem cells and fascia wrap applicationto anterior cruciate ligament tissue engineering. Cell Trans-plant 14, 763, 2005.

42. Liu, S.H., Yang, R.S., al-Shaikh, R., and Lane, J.M. Collagenin tendon, ligament, and bone healing. A current review.Clin Orthop 318, 265, 1995.

43. Amiel, D., Frank, C.B., Harwood, F.L., Akeson, W.H., andKleiner, J.B. Collagen alteration in medial collateral ligamenthealing in a rabbit model. Connect Tissue Res 16, 357, 1987.

44. Hargus, G., Kist, R., Kramer, J., Gerstel, D., Neitz, A.,Scherer, G., and Rohwedel, J. Loss of Sox9 function results indefective chondrocyte differentiation of mouse embryonicstem cells in vitro. Int J Dev Biol 52, 323, 2008.

45. Shukunami, C., Kondo, J., Wakai, H., Takahashi, K., Inoue,H., Kamizono, A., and Hiraki, Y. Molecular cloning ofmouse and bovine chondromodulin-II cDNAs and thegrowth-promoting actions of bovine recombinant protein.J Biochem 125, 436, 1999.

46. Zhao, B., Etter, L., Hinton, R.B., Jr., and Benson, D.W. BMPand FGF regulatory pathways in semilunar valve precursorcells. Dev Dyn 236, 971, 2007.

47. zur Nieden, N.I., Kempka, G., Rancourt, D.E., and Ahr, H.J.Induction of chondro-, osteo- and adipogenesis in embry-onic stem cells by bone morphogenetic protein-2: effect ofcofactors on differentiating lineages. BMC Dev Biol 5, 1, 2005.

48. Kuo, C.K., and Tuan, R.S. Mechanoactive tenogenic differ-entiation of human mesenchymal stem cells. Tissue Eng PartA 14, 1615, 2008.

49. Murchison, N.D., Price, B.A., Conner, D.A., Keene, D.R.,Olson, E.N., Tabin, C.J., and Schweitzer, R. Regula-tion of tendon differentiation by scleraxis distinguishesforce-transmitting tendons from muscle-anchoring tendons.Development 134, 2697, 2007.


50. Tallheden, T., Karlsson, C., Brunner, A., van der, L.J., Hagg,R., Tommasini, R., and Lindahl, A. Gene expression duringredifferentiation of human articular chondrocytes. Osteoar-thritis Cartilage 12, 525, 2004.

51. Liu, W., Chen, B., Deng, D., Xu, F., Cui, L., and Cao, Y.Repair of tendon defect with dermal fibroblast engineeredtendon in a porcine model. Tissue Eng 12, 775, 2006.

52. Juncosa-Melvin, N., Shearn, J.T., Boivin, G.P., Gooch, C.,Galloway, M.T., West, J.R., Nirmalanandhan, V.S., Bradica,G., and Butler, D.L. Effects of mechanical stimulation on thebiomechanics and histology of stem cell-collagen spongeconstructs for rabbit patellar tendon repair. Tissue Eng 12,

2291, 2006.

Address correspondence to:Laura B. Creemers, Ph.D.

Department of OrthopaedicsUniversity Medical Center Utrecht

Heidelberglaan 100Utrecht 3508 GAThe Netherlands

E-mail: [email protected]

Received: September 28, 2008Accepted: April 1, 2009

Online Publication Date: May 19, 2009


In Vivo Matrix Production by Bone Marrow Stromal Cells Seeded on PLGA Scaffolds for Ligament Tissue Engineering

Documents