Xenotransplantation of Human Mesenchymal Stem Cells into Immunocompetent Rats for Calvarial Bone Repair

Xenotransplantation of Human Mesenchymal Stem Cellsinto Immunocompetent Rats for Calvarial Bone Repair

Ching-Kuang Chuang,1 Kun-Ju Lin, M.D., Ph.D.,2,3 Chin-Yu Lin,1

Yu-Han Chang, M.D., Ph.D.,4,5 Tzu-Chen Yen, M.D., Ph.D.,2,3 Shiaw-Min Hwang, Ph.D.,6

Li-Yu Sung,1 Huang-Chi Chen,1 and Yu-Chen Hu, Ph.D.1

Baculovirus efficiently transduces human mesenchymal stem cells (hMSCs) and transplantation of hMSCstransduced with a bone morphogenetic protein 2–expressing baculovirus (Bac-CB) into nude mice results inectopic bone formation. To attest the clinical potential of baculovirus in bone regeneration, hereby we exploredwhether the hMSCs genetically modified by Bac-CB were tolerant in immunocompetent rats and further healedthe critical-sized calvarial bone defect. The histological and computed tomographic studies demonstrated thatBac-CB–engineered hMSCs promoted the cell differentiation and new bone formation in the immunocompetentrats. Immunohistochemical staining revealed that the transplanted human cells remained detectable at 1 and4 weeks posttransplantation, attesting the immunoprivileged properties of hMSCs. In the recipients, the donorcells aggregated and appeared osteoblast like at later stages, which paralleled the infiltration of macrophages,CD3þ, and CD8þ T cells into the graft. Administration of immunosuppressive drugs prolonged the cell survivaland improved the bone regeneration, yet it failed to entirely abolish the immune response and complete the bonehealing. Our data altogether implicate the potential of Bac-CB for hMSCs engineering and calvarial bone repair,but the use of hMSCs cannot overcome the immunological barrier.

Introduction

Reconstruction of osseous defects for congenitalanomalies or after bone loss remains a significant

problem in craniofacial surgery.1 To heal the bone defects,mesenchymal stem cells (MSCs) are promising, thanks totheir capability of self-renewal and multilineage differentia-tion into adipocytes, chondrocytes, and osteoblasts underappropriate environmental cues.2 MSCs are also immuno-modulatory as they can inhibit the maturation and functionsof various immune cells (e.g., dendritic cells, natural killer, T,and B cells).3 It has been suggested that MSCs can escape therecognition by alloreactive T cells and natural killer cells, andhence are immunoprivileged in the allogeneic setting.4 Theseproperties have prompted the use of MSCs as a cell therapyvector for bone repair in animal experiments and clinicaltrials.5,6 In addition, MSCs can be genetically engineered bywhich the local production of osteoinductive growth factorcan promote the cellular differentiation and accelerate tissue=organ regeneration in vivo.6–8 However, the common vectors

employed for MSCs modification (e.g., lentivirus, adenovi-rus, and adeno-associated virus) possess various drawbacksof their own.8

Aside from these vectors, baculovirus (Autographa cali-fornica multiple nucleopolyhedrovirus) is an insect virus, butit also efficiently transduces a wide variety of mammaliancells without replication and appreciable cytotoxicity (forreview see Refs.9–11). These features have inspired the devel-opment of baculovirus vectors carrying mammalian expres-sion cassettes for in vitro and in vivo gene therapy studies,development of cell-based assays, surface display of eucar-yotic proteins, study of gene functions, production of viralvectors, production of virus-like particles, delivery of vaccineimmunogens (for review see Refs.10,12), and genetic modifi-cation of chondrocytes for cartilage regeneration.13,14 Fur-ther, baculovirus transduces human bone marrow–derivedMSCs (i.e., hMSCs) at efficiencies up to 95% under optimizedconditions,15 and the transduced hMSCs remain capable ofdifferentiation into adipogenic, osteogenic, and chondro-genic lineages.16 Given these findings, we constructed a

1Department of Chemical Engineering, National Tsing Hua University, Hsinchu, Taiwan.2Department of Nuclear Medicine, Chang Gung Memorial Hospital, Taoyuan, Taiwan.Departments of 3Medical Imaging and 4Orthopaedic, Chang Gung University, Taoyuan, Taiwan.5College of Medicine, Chang Gung University, Taoyuan, Taiwan.6Bioresource Collection and Research Center, Food Industry Research and Development Institute, Hsinchu, Taiwan.

TISSUE ENGINEERING: Part AVolume 16, Number 2, 2010ª Mary Ann Liebert, Inc.DOI: 10.1089=ten.tea.2009.0401

479

recombinant baculovirus (Bac-CB) expressing bone mor-phogenetic protein-2 (BMP-2) under the control of cyto-megalovirus immediate-early promoter. Bac-CB transductionof hMSCs ex vivo triggered in vitro differentiation of hMSCsinto osteoblasts, and xenotransplantation of the transducedcells into the back subcutis of immunodeficient nude miceresulted in ectopic bone formation.17

Many studies exploiting hMSCs for tissue repair havetransplanted the cells into immunodeficient or allogeneicanimals to circumvent the immune attack. Although hMSCsare immunosuppressive and can be tolerated after allo-transplantation, whether hMSCs are suitable for xeno-transplantation remains to be established. For instance,hMSCs were depleted rapidly after direct injection into themyocardium of rats,18,19 while xenotransplantation of ratMSCs into the quadriceps of immunocompetent mice led toelevated immune responses that obstructed bone formation,although some donor cells remained detectable at week 11.20

In contrast to these results, injection of hMSCs into the liverof Sprague–Dawley rats led to the engraftment and differ-entiation of cells into human hepatocytes with the aid ofimmunosuppression.21 Xenotransplantation of hMSCs intothe rat spinal cord led to long-term survival of cells under,but not without, appropriate immunosuppression.22 Strik-ingly, intravenous injection of hMSCs into the rats led to thesurvival and migration of hMSCs to the spinal cord injurysite and improved the functional recovery even withoutimmunosuppression.23 A recent study also reported thelong-term survival and chondrogenic differentiation ofhMSCs after xenotransplantation into the rat intervertebraldiscs without immunosuppression.24

To attest the clinical potential of baculovirus in hMSCsengineering and bone regeneration, and to address the ques-tion regarding whether the immunoprivileged hMSCs canescape the immune surveillance in the xenogeneic setting, theprimary objective of the present study was to evaluate thecalvarial bone repair mediated by the Bac-CB–transducedhMSCs in the immunocompetent Fisher 344 rats, in thepresence and absence of immunosuppressive drugs.

Materials and Methods

Preparation and cultureof bone marrow–derived hMSCs

Bone marrow–derived human mononuclear cells were ob-tained from Cambrex (Walkersville, MD) and the subsequenthMSCs selection, enrichment, and immunotyping were per-formed as described.25 The resultant hMSCs were culturedusing a-modified minimal essential medium (Hyclone,Ogden, UT) containing 20% fetal bovine serum (Invitrogen,Carlsbad, CA), 4 ng=mL basic fibroblast growth factor(R&D System, Minneapolis, MN), 100 U=mL penicillin, and100 mg=mL streptomycin in a 378C, 5% CO2 incubator. ThehMSCs were expanded to passage 10 for all subsequentexperiments.

Baculovirus preparation and transduction

The recombinant baculovirus (Bac-CB) expressing BMP-2was constructed earlier.17 The virus titers (pfu=mL) weredetermined by end-point dilution method,26 and the bacu-lovirus transduction was performed as described17 with

minor modifications. In brief, hMSCs were cultured over-night in the T-150 flasks (6�106 cells per flask) and washedwith Dulbecco’s phosphate-buffered saline (PBS, pH 7.4)before transduction. For each flask, a certain volume of viruswas diluted to 2 mL with modified Grace’s Insect medium(TNM-FH) to adjust the multiplicity of infection (MOI) to 40,followed by mixing with 8 mL PBS. Transduction was initi-ated by directly adding the virus–PBS solution to the cellsand continued by gentle shaking on a rocking plate at roomtemperature for 4 h. After the incubation period, the cellswere washed, replenished with 20 mL a-modified minimalessential medium, and incubated at 378C. For mock trans-duction, the cells were incubated with 2 mL TNM-FH me-dium plus 8 mL PBS for 4 h at 378C. Supertransduction wasperformed at 6 days after initial transduction in a similarmanner.

Poly (L-lactide-co-glycolide) scaffold preparationand cell seeding

To fabricate the porous poly (L-lactide-co-glycolide)(PLGA) scaffolds, the PLGA (Purac, Gorinchem, The Neth-erlands) dissolved in chloroform (0.1 g=mL) was mixed withsodium chloride (200–300 mm in diameter) and the solutionwas compressed into the Teflon mold (9 mm in diameter).After the chloroform vaporized, the scaffolds were immersedin double-distilled water with three changes to dissolve thesodium chloride, and gently removed from the Teflon moldwith a fin-tip spatula. The resultant scaffolds (porosity&90%) were disinfected by immersion in 50%, 60%, and 70%ethanol for 30 min each, rinsed with double-distilled waterand air dried in a laminar flow hood.

One day after supertransduction, 50 mL hMSCs wereevenly pipetted onto the scaffold (5�105 cells per scaffold),allowed to adhere for 2 h and then supplemented with freshmedium for culture. One day later, the cell=scaffold con-structs were transplanted into the calvarial defects.

Transplantation and immunosuppression

All animal experiments were performed in compliancewith the Guide for the Care and Use of Laboratory Animals(Institute of Laboratory Animals Resources, National ScienceCouncil, Taiwan). Female Fisher 344 rats (10 weeks of age)were randomly assigned to designated groups. After an-esthetization, 8-mm-diameter critical defects1 were created inthe calvaria, implanted with the blank PLGA scaffolds or thescaffolds impregnated with hMSCs, and the wounds wereclosed with suture. The immunosuppression scheme con-sisted of subcutaneous injection of 1 mg=kg of antithymocyteglobulin (Genzyme Polyclonals S.A.S., Marcy l’Etoile,France) from 2 days before transplantation to 7 days aftertransplantation; daily oral administration of mycophenolatemofetil (Roche, Mannheim, Germany) dissolved in thedrinking water (1 mg=mL); and daily subcutaneous injectionof 10 mg=kg of cyclosporine (Roche). Tissue specimens wereharvested at 1, 4, 8, and 12 weeks (n¼ 5 for each group ateach time point) after transplantation.

Microcomputed tomography analysis

The animal computed tomography (CT) scan was per-formed using NanoSPECT=CT (Bioscan, Washington, DC) at

480 CHUANG ET AL.

the Molecular Imaging Center of Chang Gung MemorialHospital. The CT system contained a power-adjustable X-raysource of 65 kVp and microfocus (<9mm) tube, and the res-olution was less than 100mm. The CT image intensity wascalibrated into Hounsfield scale using water (0) and air(�1000) as references. The CT data of the rat skull were ac-quired using a high-resolution frame as setup in the system,with tube voltage of 65 KeV, exposure time of 2000 ms, and180 projections. The CT images were reconstructed using thesoftware provided by the system using an ultrafine recon-struction mode with 512�512�480 voxels and 0.1 mm voxelsize.

The new bone formation within the defect was quantifiedon a PMOD image workstation (PMOD Technologies,Zurich, Switzerland). The skull CT images were loaded, andthe center of scaffold was identified. A volume of interest(VOI) in rectangular shape was defined to cover the defectand the scaffold. The VOI was then segmented to eliminatethe unwanted part of the skull. The maximum intensityproject of the segmented VOI was generated, and the per-centage of new bone formation was calculated from the max-imum intensity project images using the following formula:

% new bone formation¼ (area of new bone formation)=(area of original defect)�100%

The contour of the new bone formation was defined by theabsolute threshold above 200 Hounsfield units.

Hematoxylin and eosinand immunohistochemical staining

The tissue specimens were fixed, decalcified, and sectioned.The sections were deparaffinized and rehydrated usingxylene and a gradient of ethanol (100%, 90%, 70%, 50%, and0%), followed by hematoxylin and eosin staining to detectosteogenesis. For immunohistochemical staining, the sec-tions were treated with the Target Retrieval solution (Dako,Hamburg, Germany) for 10 min at 908C and then blockedwith PBS containing 10% fetal bovine serum for 30 min atroom temperature. The sections were incubated at 48C over-night with the primary mouse monoclonal antibody specificfor human mitochondria (1:100 dilution; Bio Genex, SanRamon, CA), rat macrophage (1:100 dilution; Abcam, Cam-bridge, United Kigdom), rat CD3 (1:10 dilution; AbD Serotec,Oxford, United Kingdom), or rat CD8 (1:100 dilution; GeneTex, San Antonio, TX). After washing, the sections were in-cubated with goat anti-mouse IgG conjugated with AlexaFluor 488 (Invitrogen) for 1 h in the dark. Finally, the sectionswere mounted with the mounting medium containing 40,6-diamidino-2-phenylindole (Vector Labs, Burlingame, CA)and observed using a fluorescence microscope.

Results

Effects of Bac-CB transductionand immunosuppression on calvarial bone formation

The calvarial bone formation starts by the migration andaggregation of mesenchymal cells (for review see Ref.27). Theprogenitor cells differentiate into osteoblasts and secrete thebone matrix (osteoid) that initially appears as small, irregu-larly shaped spicules. The spicules undergo calcification withtime, enlarge, and become joined in trabecular network.During the process, the osteoblasts actively secreting matrix

molecules surround the growing spicules whereas someosteoblasts are enclosed within the matrix and differentiateinto mature osteocytes.

To explore whether the baculovirus-transduced hMSCswere tolerated in immunocompetent rats and repaired thecalvarial bone defect, the hMSCs were transduced by Bac-CBat an MOI of 40 and were supertransduced (MOI 40) 6 dayslater. The transduced and mock-transduced hMSCs wereseeded to PLGA scaffolds and transplanted into the criticaldefects (8 mm) at the rat calvaria (n¼ 5 for each group ateach time point). As controls, blank PLGA scaffolds weretransplanted in a similar fashion.

The hematoxylin and eosin staining (Fig. 1A) revealed thatthe control groups (scaffold only and mock-transducedhMSCs) barely led to cell aggregation at 4 weeks aftertransduction, and only gave rise to a few osteoids at week 12.In sharp contrast, the transduction group resulted in con-spicuous formation of numerous calcified bone matriceswith osteoblast-like cells lining the matrix (indicated by thearrows) at week 4. However, at week 12 a considerablefraction of the calcified bone matrix disappeared and nofurther bone formation was observed.

It is well established that xenotransplantation causes re-jection and failed engraftment. Therefore, the experimentswere repeated as in Figure 1A, but the rats were administeredwith immunosuppressive drugs (antithymocyte globulin,mycophenolate mofetil, and cyclosporine) to repress theimmune responses. With immunosuppression (Fig. 1B), themock transduction group still merely gave rise to a fewosteoids at weeks 4 and 12, indicating a low degree of spon-taneous osteogenic differentiation and bone formation with-out Bac-CB transduction. With the aid of immunosuppressivedrugs, the transduction group led to evident bone spiculeformation at weeks 4 and 12. Notably, the osteoblasts werelocalized to the surface of bone spicules (arrowheads in theinsets) and the osteocytes resided within the lacuna (arrows inthe insets), indicating the cell differentiation.

CT monitoring of calvarial bone repair

The mineralization and calvarial bone repair process weremonitored by CT imaging (Fig. 2) and compared with thearea of new bone within the defects (Table 1). Without im-munosuppression (Fig. 2A), the control groups (scaffold onlyand mock transduction) only manifested bone formationnear the periphery of the defect at week 4, and the boneformation barely improved at week 12. The transductiongroup resulted in the formation of bone islands in the centraland peripheral areas of the defect at week 4, whose bone area(4.6� 1.2%) was approximately two times that of the controlgroups (2.4� 0.6% and 2.6� 0.7%). At week 12 more bonewas formed near the edge of the defect, yet the bone area(7.3� 2.6%) was not significantly larger than that of thecontrol groups (6.2� 2.3% and 7.3� 3.2%).

With immunosuppression (Fig. 2B), the mock-transducedhMSCs still failed to stimulate manifest new bone forma-tion at weeks 4 and 12. Nevertheless, the transduced hMSCsgave rise to very evident formation of new bone thatfilled& 14.7% and& 28.1% of the defects at weeks 4 and 12,respectively (Fig. 2B and Table 1). The new bone was notcompletely bridged but the repair was significantly superiorto those mediated by the mock transduction group with

XENOTRANSPLANTATION OF BACULOVIRUS-ENGINEERED HMSCS 481

FIG. 1. Effects of Bac-CB transduction on calvarial bone formation in the absence (A) or presence (B) of immunosup-pression. The Fisher 344 rats were transplanted with blank scaffolds (scaffold only), or scaffolds impregnated with mock-transduced or transduced hMSCs. The tissue specimens were removed at 4 or 12 weeks after transplantation for hematoxylinand eosin staining (n¼ 5 for each group at each time point). The arrowheads in the insets of (B) indicate the osteoblasts whilethe arrows indicate the osteocytes. Magnification, �200. hMSCs, human mesenchymal stem cells. Color images availableonline at www.liebertonline.com=ten.

482

FIG. 2. Computed tomography monitoring of bone repair in the absence (A) or presence (B) of immunosuppression. Thedata are representative of samples from five animals for each group at each time point. Color images available online atwww.liebertonline.com=ten.

XENOTRANSPLANTATION OF BACULOVIRUS-ENGINEERED HMSCS 483

immunosuppression and the transduction group withoutimmunosuppression.

Assessment of immune responses

Figures 1 and 2 collectively demonstrated that xeno-transplantation of Bac-CB–engineered hMSCs into the de-fects was able to promote the calvarial bone repair in theimmunocompetent rats, but the inferior bone repair in theabsence of immunosuppression suggested that rejection re-sponses retarded the bone-healing process. To confirm theroles of immune responses, the mock-transduced andtransduced hMSCs were seeded to scaffolds and trans-

planted as in Figure 1, and the rats were bred without im-munosuppressive drugs. The survival of the donor cells andthe infiltration of host cells representative of innate immunity(macrophages) and adaptive immunity (CD3þ and CD8þ

cells) were assessed by immunohistochemical staining atdifferent time points.

Figure 3A (left column) illustrates the persistence anddispersion of numerous mock-transduced human cellswithin the graft at week 1, as stained by human mitochon-drial antibody. In accord with this was that only a fewmacrophages and no T cells infiltrated into the graft, sug-gesting weak immune responses evoked by the mock-transduced, undifferentiated hMSCs. The transduced donor

Table 1. Quantification of Bone Formation from Computed Tomography Images

Time Scaffold only Mock-transduced hMSCs Transduced hMSCs

�immunosuppression 4 weeks 2.4� 0.6% 2.6� 0.7% 4.6� 1.2%12 weeks 6.2� 2.3% 7.3� 3.2% 7.3� 2.6%

þimmunosuppression 4 weeks N=D 3.2� 0.6% 14.7� 3.5%12 weeks N=D 7.0� 2.4% 28.1� 4.4%

n¼ 5 for each group at each time point.hMSCs, human mesenchymal stem cells; N=D, not determined.

FIG. 3. Survival of the transplanted cells and infiltration of macrophage and T cells at weeks 1 (A) and 4 (B). The grafts wereremoved, sectioned, and subjected to immunohistochemical staining using primary antibodies specific for human mito-chondria, rat macrophages, rat CD3þ or rat CD8þ, and the secondary antibody conjugated with Alexa Fluor 488 (green). Allnucleated cells were stained by 40,6-diamidino-2-phenylindole (red), thus the transplanted human cells (arrows indicate)appeared orange after superimposing the images. The immune cells were observed with Alexa Fluor 488–conjugated sec-ondary antibody on the cell membrane and appeared green. n¼ 5 for each group at each time point. Magnification,�200.Color images available online at www.liebertonline.com=ten.

484 CHUANG ET AL.

cells (right column, Fig. 3A) exhibited signs of osteogenicdifferentiation and formation of osteoids at week 1, as evi-denced by the cell aggregation. Concomitant with the dif-ferentiation, markedly more macrophages and T cells (CD3þ

and CD8þ) invaded into the graft.At week 4 (Fig. 3B), a number of mock-transduced human

cells (left column) continued to persist, and some of themaggregated as clusters, indicating the onset of cell differen-tiation in the absence of Bac-CB transduction. Concurrently,macrophages, CD3þ T cells, and some CD8þ T cells infil-trated into the graft. Meanwhile, some of the transducedhuman cells remained detectable and colocalized with thebone matrix (indicated by arrows, Fig. 3B), suggesting theparticipation of the transplanted human cells in the newbone formation. When compared with the mock transduc-tion group, the transduction was concomitant with morepronounced infiltration of macrophages and T cells (CD3þ

and CD8þ) into the bone matrix. These data suggested thatthe mock-transduced human cells were less susceptible toimmunosurveillance than the transduced cells in the xeno-geneic recipients.

hMSCs survival after xenotransplantation

To track the fate of the donor hMSCs in the long term, weperformed the immunohistochemical staining specific forhuman cells at 8 and 12 weeks after transplantation. Withoutimmunosuppression, significantly fewer mock-transducedcells were detectable at weeks 8 and 12 (Fig. 4A, left column)than at weeks 1 and 4 (Fig. 3). An even more dramatic de-crease in cell number was observed in the transductiongroup at week 8, and the human cells completely vanished atweek 12 (Fig. 4A, right column).

In the animals receiving the immunosuppressive drugs(Fig. 4B), considerably more mock-transduced cells remainedresident within the graft at weeks 8 and 12 than theircounterparts without immunosuppression (Fig. 4A), provingthe effect of the immunosuppressive drugs in prolonging thecell survival. Tremendously fewer cells in the transductiongroup remained viable at weeks 8 and 12, but the survivedhuman cells surrounded the newly formed bone spicules (asmarked by the dashed lines) in a way similar to the osteo-blasts do upon normal bone formation.27

Discussion

In this study we demonstrated that the mock-transducedhMSCs remained largely undifferentiated and stimulatedpoor bone matrix deposition and ossification, with or with-out immunosuppression (Figs. 1 and 2). In contrast, Bac-CBtransduction considerably boosted the hMSCs aggregation(Figs. 1A and 3), ameliorated the accumulation of mineral-ized bone matrix (Fig. 1A), and initiated the bone islandformation (Fig. 2A) at week 4 even without immunosup-pression. With the aid of immunosuppression, the trans-duction group led to the regeneration of trabecular bone(Fig. 1B) that filled 28.1� 4.4% of the defect area at week 12(Fig. 2B and Table 1), which significantly excelled as com-pared with the mock transduction group. Additionally, theembedding of differentiated donor cells within the bonematrix (Fig. 3) and cell lining the bone spicule (Fig. 4B) sug-gest that the Bac-CB–engineered hMSCs also responded toBMP-2 in an autocrine and=or paracrine fashion and partic-

ipated in the new bone formation, a phenomenon that is alsoobserved for bone formation from the muscle-derived stemcells.28 These data confirmed that Bac-CB transduction di-rected the commitment of hMSCs along the osteogenic line-age and promoted the calvarial bone repair. This stimulatoryeffect is at least partly attributed to the efficient baculovirustransduction of hMSCs15,25 and ensuing BMP-2 expression toa level sufficient to induce bone regeneration.17

Meanwhile, we detected the survival of numerous mock-transduced cells at 1 and 4 weeks after transplantation evenwithout immunosuppression (Fig. 3). Although the numberof engrafted cells remarkably decreased at weeks 8 and 12(Fig. 4A), this suggested that the xenotransplanted hMSCswere fairly tolerated in the immunocompetent rats presum-ably, thanks to the immunosuppressive and immunopri-vileged properties of the hMSCs. Alternatively, the survivalmight be ascribed to the protection of donor cells in thescaffold, which avoided prompt confrontation with the hostimmune cells. Moreover, many Bac-CB–transduced cells re-mained detectable in the graft at 1 week after transplanta-tion, despite the onset of innate and adaptive immuneresponses (Fig. 3A). The mounting of cell-mediated immuneresponses concurred with the elimination of donor cellsand inferior bone repair in case of no immunosuppression(Figs. 3B and 4A). The administration of immunosuppressivedrugs considerably prolonged the cell survival to week 12(Fig. 4B) and augmented the bone repair, which, however,was incomplete (Table 1).

The incomplete bone regeneration can be attributable to(1) the ineffectiveness of baculovirus-engineered hMSCs inorthotopic bone regeneration or (2) the immune responses. Ithas been shown that short-term BMP-2 expression is suffi-cient to irreversibly trigger bone formation in vivo.17,29 Fur-ther, Bac-CB transduction of MSCs derived from the bonemarrow of New Zealand White (NZW) rabbits and subse-quent allogeneic transplantation into a large critical-sizeddefect (8 mm in diameter and 10 mm in length) at the femoraof NZW rabbits led to the segmental bone repair (manuscriptin preparation). Since the femoral segmental model repre-sents a more difficult bone-healing scenario, the success ofBac-CB in femoral bone healing confirmed that Bac-CB–engineered MSCs are capable of healing large bone defectsand ruled out the first hypothesis. As such, the immune re-sponses were primarily responsible for the incomplete bonehealing.

Interestingly, the magnitude of immune responses ap-peared to correlate with Bac-CB transduction (Fig. 3), whichcan be accounted for by several factors. First, baculovirustransduction itself might mitigate the immunoprivilegedproperties of hMSCs, thereby exposing the transduced donorcells to immune surveillance. This assumption is unlikelybecause after xenotransplantation into Fisher 344 rats, thehMSCs that were transduced with a baculovirus expressingno transgene did not elicit considerably stronger macro-phage and T-cell responses than the mock-transduced hMSCs(data not shown). Further, we recently demonstrated thathMSCs transduced with a baculovirus expressing no trans-gene retained the capability to suppress lymphocyte prolif-eration, and allotransplantation of baculovirus-transducedrat MSCs into Fisher 344 rats did not provoke apparent mac-rophage and T-cell responses. These data indicate that thebaculovirus-transduced MSCs remained immunoprivileged

XENOTRANSPLANTATION OF BACULOVIRUS-ENGINEERED HMSCS 485

FIG. 4. Survival of the transplanted cells in the absence (A) or presence (B) of immunosuppression. The cells were detectedby immunohistochemical staining specific for human mitochondria at 8 and 12 weeks after transplantation. n¼ 5 for eachgroup at each time point. The dashed lines in (B) indicate the bone spicule and the arrows indicate the osteoblasts sur-rounding the bone matrix. Magnification,�200. Color images available online at www.liebertonline.com=ten.

486 CHUANG ET AL.

in the allogeneic host30 and excluded the first assumption.Second, the expression of human BMP-2 in the rats resultedin anti-BMP-2 antibodies and associated immune responsesthat eliminated the BMP-2–expressing cells. This is possible,but less likely to play a major role because secretion of BMPsalone does not adversely affect the cell survival in vivo.31

Further, the successful segmental bone healing mediated bythe Bac-CB–engineered rabbit MSCs in the immunocompe-tent NZW rabbits suggests that the anti-BMP-2 immunity(if any) does not explicitly impede the tissue regeneration.

The last and most likely possibility is that Bac-CB trans-duction accelerated the hMSCs differentiation, leading to theloss of immunoprivileged properties and hence depletionby the intrinsic immune systems of xenogeneic rats. Thisassumption is in line with the recent finding that the im-munosuppressive potential of mouse MSCs is lost aftertransplantation, and hence the cells are rejected.32 Moreover,Bac-CB transduction accelerated the bone formation, which isin nature accompanied by neovascularization. The formationof vasculature might increase the flux of blood, facilitate therecruitment of immune cells into the transplantation site,20

and hence lead to a more serious graft rejection response thanthe mock transduction group. Further experiments to scruti-nize these hypotheses are ongoing.

In conclusion, in this study we demonstrated that Bac-CB–engineered hMSCs remarkably promoted the calvarial boneregeneration in the immunocompetent rats, implicating thepotential of baculovirus in MSCs engineering and tissuerepair. Our data supported the notion that the xenotrans-planted hMSCs are able to evade the immune surveillanceinitially (e.g., at 1 week after transplantation), but are even-tually rejected even with immunosuppression. Althoughadministration of immunosuppressive drugs prolonged thesurvival and augmented the bone healing, the immunosup-pression was unable to entirely abolish the immune re-sponse, leading to the incomplete bone healing. These resultspartly concurred with recent reports that stem cells xeno-transplanted into the femoral defect,33 intervertebral discs,34

or posterolateral lumbar spine35 are able to survive for a longterm (up to 6 months) and contribute to new tissue formation(e.g., bone). However, these studies did not report cell clear-ance as a result of immunological rejection, probably becausethe animal models were immunodeficient33 or the transplan-tation sites (e.g., intervertebral discs) are relatively immuno-privileged.24 Despite cell survival, no definitive, completetissue healing was reported in these studies. Therefore, theuse of hMSCs appears to be restricted to autologous or allo-geneic transplantation.

Acknowledgments

The authors acknowledge the financial support from theNational Tsing Hua University Booster Program(97N2511E1, 98N2901E1), National Tsing Hua University-Chang Gung Memorial Hospital Joint Research Program(CMRPG380101), National Science Council (NSC 97-2627-B-007-014), and National Health Research Institutes (NHRI-EX97-9412EI), Taiwan.

Disclosure Statement

No competing financial interests exist.

References

1. Yoon, E., Dhar, S., Chun, D.E., Gharibjanian, N.A., andEvans, G.R.D. In vivo osteogenic potential of humanadipose-derived stem cells=poly lactide-co-glycolic acidconstructs for bone regeneration in a rat critical-sized cal-varial defect model. Tissue Eng 13, 619, 2007.

2. Caplan, A.I., and Bruder, S.P. Mesenchymal stem cells:building blocks for molecular medicine in the 21st century.Trends Mol Med 7, 259, 2001.

3. Uccelli, A., Moretta, L., and Pistoia, V. Mesenchymal stemcells in health and disease. Nat Rev Immunol 8, 726, 2008.

4. Le Blanc, K., Rasmusson, I., Sundberg, B., Gotherstrom, C.,Hassan, M., Uzunel, M., and Ringden, O. Treatment of severeacute graft-versus-host disease with third party haploiden-tical mesenchymal stem cells. Lancet 363, 1439, 2004.

5. Mack, G.S. Osiris seals billion-dollar deal with Genzyme forcell therapy. Nat Biotechnol 27, 106, 2009.

6. Kimelman, N., Pelled, G., Helm, G.A., Huard, J., Schwarz, E.M.,and Gazit, D. Review: gene- and stem cell-based therapeuticsfor bone regeneration and repair. Tissue Eng 13, 1135, 2007.

7. Evans, C.H., Ghivizzani, S.C., and Robbins, P.D. Orthopedicgene therapy in 2008. Mol Ther 17, 231, 2009.

8. Phillips, J.E., Gersbach, C.A., and Garcia, A.J. Virus-basedgene therapy strategies for bone regeneration. Biomaterials28, 211, 2007.

9. Tani, H., Abe, T., Matsunaga, T.M., Moiihi, K., and Mat-suura, Y. Baculovirus vector for gene delivery and vaccinedevelopment. Future Virol 3, 35, 2008.

10. Hu, Y.-C. Baculoviral vectors for gene delivery: a review.Curr Gene Ther 8, 54, 2008.

11. Condreay, J.P., and Kost, T.A. Baculovirus expression vectorsfor insect and mammalian cells. Curr Drug Targets 8, 1126,2007.

12. Hu, Y.-C., Yao, K., and Wu, T.-Y. Baculovirus as an ex-pression and=or delivery vehicle for vaccine antigens. ExpertRev Vaccines 7, 363, 2008.

13. Chen, H.-C., Chang, Y.-H., Chuang, C.-K., Lin, C.-Y., Sung,L.-Y., Wang, Y.-H., and Hu, Y.-C. The repair of osteochon-dral defects using baculovirus-mediated gene transfer withde-differentiated chondrocytes in bioreactor culture. Bio-materials 30, 674, 2009.

14. Chen, H.-C., Sung, L.-Y., Lo, W.-H., Chuang, C.-K., Wang,Y.-H., Lin, J.-L., and Hu, Y.-C. Combination of baculovirus-mediated BMP-2 expression and rotating-shaft bioreactorculture synergistically enhances cartilage formation. GeneTher 15, 309, 2008.

15. Lo, W.-H., Hwang, S.-M., Chuang, C.-K., Chen, C.-Y., andHu, Y.-C. Development of a hybrid baculoviral vector forsustained transgene expression. Mol Ther 17, 658, 2009.

16. Ho, Y.-C., Lee, H.-P., Hwang, S.-M., Lo, W.-H., Chen, H.-C.,Chung, C.-K., and Hu, Y.-C. Baculovirus transduction ofhuman mesenchymal stem cell-derived progenitor cells:variation of transgene expression with cellular differentia-tion states. Gene Ther 13, 1471, 2006.

17. Chuang, C.-K., Sung, L.-Y., Hwang, S.-M., Lo, W.-H., Chen,H.-C., and Hu, Y.-C. Baculovirus as a new gene deliveryvector for stem cells engineering and bone tissue engineer-ing. Gene Ther 14, 1417, 2007.

18. Grinnemo, K.H., Mansson, A., Dellgren, G., Klingberg, D.,Wardell, E., Drvota, V., Tammik, C., Holgersson, J., Ringden,O., Sylven, C., and Le Blanc, K. Xenoreactivity and engraft-ment of human mesenchymal stem cells transplanted intoinfarcted rat myocardium. J Thorac Cardiovasc Surg 127,

1293, 2004.

XENOTRANSPLANTATION OF BACULOVIRUS-ENGINEERED HMSCS 487

19. Terrovitis, J., Stuber, M., Youssef, A., Preece, S., Leppo, M.,Kizana, E., Schar, M., Gerstenblith, G., Weiss, R.G., Marban,E., and Abraham, M.R. Magnetic resonance imaging over-estimates ferumoxide-labeled stem cell survival after trans-plantation in the heart. Circulation 117, 1555, 2008.

20. Wang, Y., Chen, X., Armstrong, M.A., and Li, G. Survival ofbone marrow-derived mesenchymal stem cells in a xeno-transplantation model. J Orthop Res 25, 926, 2007.

21. Sato, Y., Araki, H., Kato, J., Nakamura, K., Kawano, Y.,Kobune, M., Sato, T., Miyanishi, K., Takayama, T., Takaha-shi, M., Takimoto, R., Iyama, S., Matsunaga, T., Ohtani, S.,Matsuura, A., Hamada, H., and Niitsu, Y. Human mesen-chymal stem cells xenografted directly to rat liver are dif-ferentiated into human hepatocytes without fusion. Blood106, 756, 2005.

22. Ronsyn, M.W., Daans, J., Spaepen, G., Chatterjee, S., Ver-meulen, K., D’Haese, P., Van Tendeloo, V.F., Van Marck, E.,Ysebaert, D., Berneman, Z.N., Jorens, P.G., and Ponsaerts, P.Plasmid-based genetic modification of human bone marrow-derived stromal cells: analysis of cell survival and transgeneexpression after transplantation in rat spinal cord. BMCBiotechnol 7, 1, 2007.

23. Cizkova, D., Rosocha, J., Vanicky, I., Jergova, S., and Cizek,M. Transplants of human mesenchymal stem cells improvefunctional recovery after spinal cord injury in the rat. CellMol Neurobiol 26, 1167, 2006.

24. Wei, A., Tao, H., Chung, S.A., Brisby, H., Ma, D.D., andDiwan, A.D. The fate of transplanted xenogeneic bonemarrow-derived stem cells in rat intervertebral discs. J Or-thop Res 27, 374, 2009.

25. Ho, Y.-C., Chung, Y.-C., Hwang, S.-M., Wang, K.-C., andHu, Y.-C. Transgene expression and differentiation ofbaculovirus-transduced human mesenchymal stem cells.J Gene Med 7, 860, 2005.

26. O’Reilly, D., Miller, L., and Luckow, V. Baculovirus Ex-pression Vectors: A Laboratory Manual. New York: W.H.Freeman and Co., 1992.

27. Ross, M.H., Reith, E.J., and Romrell, L.J. Histology: A Textand Atlas. Baltimore, MD: Williams & Wilkins, 1989.

28. Wright, V.J., Peng, H.R., Usas, A., Young, B., Gearhart, B.,Cummins, J., and Huard, J. BMP4-expressing muscle-derived stem cells differentiate into osteogenic lineage andimprove bone healing in immunocompetent mice. Mol Ther6, 169, 2002.

29. Noel, D., Gazit, D., Bouquet, C., Apparailly, F., Bony, C.,Plence, P., Millet, V., Turgeman, G., Perricaudet, M., Sany, J.,and Jorgensen, C. Short-term BMP-2 expression is sufficientfor in vivo osteochondral differentiation of mesenchymalstem cells. Stem Cells 22, 74, 2004.

30. Chuang, C.-K., Wong, T.-H., Hwang, S.-M., Chang, Y.-H.,Chen, Y.-H., Chiu, Y.-C., Huang, S.-F., and Hu, Y.-C. Bacu-lovirus transduction of mesenchymal stem cells: in vitroresponses and in vivo immune responses after cell trans-plantation. Mol Ther 17, 889, 2009.

31. Musgrave, D.S., Pruchnic, R., Wright, V., Bosch, P., Ghi-vizzani, S.C., Robbins, P.D., and Huard, J. The effect of bonemorphogenetic protein-2 expression on the early fate ofskeletal muscle-derived cells. Bone 28, 499, 2001.

32. Prigozhina, T.B., Khitrin, S., Elkin, G., Eizik, O., Morecki, S.,and Slavin, S. Mesenchymal stromal cells lose their immuno-suppressive potential after allotransplantation. Exp Hematol36, 1370, 2008.

33. Jager, M., Degistirici, O., Knipper, A., Fischer, J., Sager, M.,and Krauspe, R. Bone healing and migration of cord blood-derived stem cells into a critical size femoral defect afterxenotransplantation. J Bone Miner Res 22, 1224, 2007.

34. Henriksson, H.B., Svanvik, T., Jonsson, M., Hagman, M.,Horn, M., Lindahl, A., and Brisby, H. Transplantation ofhuman mesenchymal stems cells into intervertebral discs ina xenogeneic porcine model. Spine 34, 141, 2009.

35. Kim, H.J., Park, J.B., Lee, J.K., Park, E.Y., Park, E.A., Riew,K.D., and Rhee, S.K. Transplanted xenogenic bone marrowstem cells survive and generate new bone formation in theposterolateral lumbar spine of non-immunosuppressedrabbits. Eur Spine J 17, 1515, 2008.

Address correspondence to:Yu-Chen Hu, Ph.D.

Department of Chemical EngineeringNational Tsing Hua University

Hsinchu 300Taiwan

E-mail: [email protected]

Received: June 14, 2009Accepted: August 24, 2009

Online Publication Date: September 28, 2009

488 CHUANG ET AL.