J Dent Res 82(4): 293-297, 2003
© 2003 International and American Associations for Dental Research
RESEARCH REPORT
Biological |
Bone Formation by BMP-7-transduced Human Gingival Keratinocytes
R.B. Rutherford*,
P. Racenis1,
S. Fatherazi, and
K. Izutsu1
Center for Biorestoration of Oral Health, School of Dentistry, University of Michigan, 1011 N. University, Ann Arbor, MI 48109-1078; and
1 Department of Oral Biology, School of Dentistry, University of Washington;
*corresponding author, rbruth{at}umich.edu
 |
ABSTRACT
|
|---|
BMPs are a family of pleiotropic signaling molecules involved at various stages in the formation of bones and teeth. In addition, recombinant BMP can induce bone and dentin regeneration when applied directly to adult tissues. We have shown that fibroblasts transduced ex vivo by BMP cDNA delivered by recombinant adenoviruses induce bone formation and convert to osteoblasts upon implantation in vivo. To determine if this osteogenic capacity was limited to fibroblasts, we found that BMP-7-transduced human oral keratinocyte cells (HOKC) also formed ectopic bone. The ossicles formed by the BMP-7-transduced HOKC were smaller and more dense than those formed by BMP-7-transduced human gingival fibroblasts (HGF). Implanted HOKC were localized adjacent to the developing bone by immunocytochemical detection of keratin expression. However, no bone-like tissue formed when HOKC were implanted into diffusion chambers in vivo. We conclude that BMP-transduced HOKC secrete BMP and form bone in vivo but, unlike BMP-transduced HGF, do not transdifferentiate to osteoblasts.
KEY WORDS: bone morphogenetic protein-7 • keratinocytes • bone formation • gene therapy • cell differentiation
|
INTRODUCTION
|
|---|
Craniofacial defects present a common yet complex problem in oral and craniofacial surgery. Current methods utilizing alloplastic materials, allografts, or autografts have significant limitations. We are developing novel methods for skeletal repair that address these limitations. One is a gene transfer approach for osteogenesis in which autologous fibroblasts from a simple skin biopsy can be cultured, infected ex vivo with an adenovirus containing a bone morphogenic protein-7 (BMP-7) full-length cDNA, combined with a carrier, and implanted as an autograft to form bone (Franceschi et al., 2000; Krebsbach et al., 2000; Rutherford et al., 2002a,b). The transduced, non-osteogenic cells secrete biologically active bone morphogenetic protein and convert to osteoblasts in vivo (Rutherford et al., 2002b). Hence the new bone is produced by both genetically altered engrafted and normal host cells stimulated by secreted BMP (Krebsbach et al., 2000). Typically, this bone is comprised of cortical, trabecular, and marrow elements that completely replace the graft. Importantly, this bone-regenerating strategy does not require that graft cells be cultured from bone. Instead, autologous dermal or gingival fibroblasts can be easily biopsied and utilized in bony defect repair. This approach promises to advance complex, localized skeletal repair while eliminating the morbidity of bony autograft harvest and reducing side-effects such as infection and tissue rejection associated with allografts. In addition, problems associated with alloplastic materials—such as failure to integrate fully or material degradation—are eliminated. Ex vivo gene therapy appears likely to be more effective than protein therapy in cases where a single exogenous application of a protein is insufficient.
To determine if this property was unique to fibroblasts or a general property of differentiated non-osteogenic cells, we compared the capacity of cultured human gingival keratinocytes and fibroblasts transduced by BMP-7 to form bone in vivo. Here we report experiments testing the hypothesis that BMP-7-transduced human keratinocyte cells (HOKC) lack the capacity to form bone in vivo and transdifferentiate to osteoblasts in vivo. Analysis of the data reveals that BMP-7-transduced HOKC induce bone formation in vivo, but fail to develop an osteoblast phenotype. The ossicles formed by BMP-7-transduced HOKC differed in size and density from those formed by fibroblasts.
|
METHODS
|
|---|
Cell Culture, Transduction, and Implantation
Human gingival fibroblasts (HGF) were obtained as surgical waste grown from explant cultures, passaged in vitro, stored in liquid nitrogen, and infected ex vivo with a recombinant adenovirus (AdBMP7) as previously described (Franceschi et al., 2000; Krebsbach et al., 2000; Rutherford et al., 1992). HGF (passages 4-10) were infected at a multiplicity of infection (MOI) of 20,000 virus particles/cell in 8 mL of DMEM with 10% fetal calf serum in 75 cm2 flasks for 20-24 hrs. Primary cultures of gingival epithelial cells were generated as previously described (Lamont et al., 1995). Briefly, healthy gingival tissues were collected from patients undergoing surgical removal of impacted third molars or crown lengthening. Specimens were cut into small pieces and incubated in dispase (50 caseinolytic units/mL) (Collaborative Biomedical Research Products, Bedford, MA, USA) overnight at 4°C. The epithelium was carefully separated from the underlying dermis and placed in sterile 0.05% trypsin/0.53 mM EDTA (GIBCO, Grand Island, NY, USA) to dissociate the intact epithelium into single cell suspensions. Cells were collected by centrifugation, suspended in serum-free keratinocyte growth medium (Clonetics, San Diego, CA, USA) containing 0.03 mM CaCl2, which allows keratinocytes to proliferate but not differentiate, seeded into tissue culture flasks (Corning, Corning, NY, USA) at 2500 cells/cm2, and incubated at 37 C in 5% CO2/95% air. HOKC were transduced ex vivo exactly as described for HGF. Ectopic ossicles were produced in NIH-bg-nu-xid br immunocompromised mice by the suspension of 2 x 106 BMP-7-transduced HGF or HOK cells in 200 µL of a thermoset hydrogel (type I collagen, RD Biosciences, Bedford, MA, USA), injected subcutaneously, and allowed to develop in vivo for 3 wks. Equal numbers of AdCMVpLpA(-), empty vector with CMV promoter, pUC 19 polylinker, SV40 early splice/polyA linker (AdMT) transduced HOKC, and HGF were implanted into contralateral sites.
Diffusion Chambers
We prepared diffusion chambers by luting nitrocellulose membranes (0.45 µM, Millipore, Inc., Bedford, MA, USA) onto sterile 14 x 2-mm custom plastic discs in which 4-mm-diameter chambers had been prepared. Thirty microliters of 105 BMP-7-transduced or non-transduced HOKC or HGF type 1 collagen hydrogel suspensions of single cells were injected into the chambers. The sealed diffusion chambers were implanted subcutaneously into male C57Bl6 mice for 6 wks.
Tissue Analysis
After 3 wks in vivo, the ossicles were harvested by gross dissection and fixed by immersion overnight in approximately 5 vol of fresh 4% paraformaldehyde (in PBS, pH 7.2). The ossicles were decalcified by immersion in 5 vol of 10% formic acid, refreshed every 48 hrs for 1 wk. The demineralized ossicles were then processed for light microscopic analyses. Some sections were stained with hematoxylin and eosin, while others were prepared for immunohistochemical analysis of BMP-7 secretion and keratin production. Immunocytochemistry was performed as described (Krebsbach et al., 2000) with antisera purchased from Santa Cruz Biochemical (Santa Cruz, CA, USA). Stained sections were examined by light microscopy (BX60, Olympus Optical Co., Tokyo, Japan) and photographed by means of a mounted digital camera (Spot RT Color, Diagnostic Instruments Inc., Sterling Heights, MI, USA). The photomicrographs were prepared in Adobe Photoshop (v.5 Adobe, San Jose, CA, USA). Animal experiments were conducted under approved protocols at the University of Michigan. Human surgical waste was obtained under informed-consent protocols approved by the appropriate Institutional Review Boards.
Computer-assisted histomorphometry was performed on multiple sections by means of a Spot Diagnostic model 230 digital camera and Image-Pro Plus ver. 4.1 software (Media Cybernetics, Siver Spring, MD, USA). Group means were compared by a Student’s t test for unpaired samples.
|
RESULTS
|
|---|
Ossicles formed in all BMP-7-transduced HOKC and HGF (n = 6/group), yet the size and bone matrix density differed significantly between the two. The total mass of the ossicles produced by BMP-7-transduced HOKC was less than half that produced by BMP-7-transduced HGF, while the proportion of bony matrix to overall size in the BMP-7-transduced HOKC-derived ossicles was greater than 2.8 times that contained in the HGF-derived ossicles (Table
). This difference in the amount of bone matrix between the two types of ossicles is also evident in histological specimens (Fig. 1). Interestingly, an abundant marrow with both hematopoietic and fatty elements is routinely observed in the BMP-7/HGF ossicles (Fig. 1B) (Krebsbach et al., 2000; Rutherford et al., 2002b) but is absent in the BMP-7/HOKC ossicles (Fig. 1A).
View this table:
[in this window]
[in a new window]
|
Table. BMP-7-transduced Human Oral Keratinocyte Cells Induced Ossicles Smaller and More Dense Than Those Induced by BMP-7-transduced Human Gingival Fibroblasts.
|
|
View larger version (106K):
[in this window]
[in a new window]
|
Figure 1. BMP-7-transduced human oral keratinocytes (A) and fibroblasts (B) produce ossicles after subcutaneous implantation. 106 ex vivo BMP-7-transduced human oral (gingival) keratinocyte cells and gingival fibroblasts were harvested, suspended in a collagen thermoset hydrogel, and injected subcutaneously into immunocompromised mice. The ossicles were harvested after 3 wks in vivo and prepared for histological analyses. The overall bone mass is greater in the keratinocyte-produced ossicles, while bony trabeculae (arrow) and marrow (double arrow) are prominent in the fibroblast-produced ossicles. The keratinocyte-induced ossicles contained keratin-producing cells (C, arrow), while the gingival fibroblast ossicles did not (E). Developing the specimen with non-immune IgG instead of anti-keratin antibody demonstrates the specificity of the immuncytochemical assay for keratin production (D). The preparations in panels A and B were stained with H&E. Bar = 700 µm.
|
|
Studies of the fate of the implanted BMP-7/HOKC reveal that they produce keratin, a differentiation marker for oral keratinocytes, after implantation in vivo (Fig. 1C
). Similarly, non-transduced HOKC produced keratin in vivo (data not shown). As expected, the BMP-7/HGF ossicles failed to display keratin (Fig. 1E). Additionally, both types of cells produced immunoreactive BMP-7 up to 21 days post-implantation (Fig. 2), revealing that the transduced cells secrete BMP-7 in vivo. Hence both types of cells survive. BMP-7 expression cannot be detected after 1 mo, yet ossicles remain essentially unchanged for 6 mos (Rutherford, unpublished data).
View larger version (108K):
[in this window]
[in a new window]
|
Figure 2. BMP-7-transduced oral human keratinocyte (A,B) and gingival fibroblast produced ossicles (C,D) and secreted BMP-7 in vivo. Specimens were developed for immunohistochemical analyses with anti-BMP-7 (A,C), control non-immune IgG (B,D), and counterstained with hematoxylin. The ossicles were harvested and prepared for analysis 3 wks after implantation. Bar = 700 µm.
|
|
We next sought to determine if BMP-7/HOKC, like BMP-7/HGF (Rutherford et al., 2002b), acquired an osteoblastic phenotype in vivo in the absence of the direct influence of host tissue. For these experiments, BMP-7/HOKC, MT/HOKC, and BMP-7/HGF (n = 6/group) were suspended in a thermoset collagen hydrogel, injected into diffusion chambers, and implanted in vivo as described (Rutherford et al., 2002b). Unlike BMP-7/HGF, BMP/HOKC suspensions failed to produce a mineralized matrix indicative of osteoblastic activity after 6 wks in vivo (Fig. 3). To control for the possibility that leakage of the diffusion chambers resulted in the lack of mineralized matrix forming in the BMP-7/HOKC, some chambers were perforated prior to implantation into immunocompetent mice (Fig. 3C). These chambers were filled with an inflammatory infiltrate with evidence of cell necrosis as expected, whereas no such changes were observed in the intact chambers.
View larger version (77K):
[in this window]
[in a new window]
|
Figure 3. BMP-7-transduced keratinocytes fail to form a bone-like matrix in diffusion chambers implanted for 6 wks into c57B/6 immunocompetent BC57B/6 mice. 105 BMP-7/HOKC (A), MT/HOKC (B), and BMP-7/HGF (D) were suspended in a thermoset collagen hydrogel and loaded into diffusion chambers as described (METHODS). As a control against the possibility of diffusion chamber leakage, some of the diffusion chambers were breached prior to implantation (C). The diffusion chambers were surgically implanted into subcutaneous tissue pouches in C57Bl6 mice, harvested after 6 wks, prepared for histological examination, and stained with H & E. Bar = 700 µm.
|
|
|
DISCUSSION
|
|---|
Several studies suggest that ex vivo gene-therapy-mediated delivery of BMP is highly effective at inducing localized skeletal repair in animal models of orthotopic osteogenesis (Breitbart et al., 1998, 1999; Lieberman et al., 1998, 1999; Krebsbach et al., 2000; Lieberman, 2000; Lee et al., 2001; Rutherford et al., 2002b). Typically, the bone produced by this technique, as in BMP protein therapy (Reddi, 1998), is comprised of cortical, trabecular, and marrow elements. However, few studies have addressed the amount or quality of the bone produced. We have demonstrated that, within the limits tested (103 to 2 x 106 cells), the size of the ossicles produced in an ectopic assay was dependent on the number of BMP-transduced cells implanted (Rutherford et al., 2002b). However, the fundamental structure of the bone produced did not vary with cell number. All the ossicles displayed a similar composition of cortex and bony trabeculae interspersed with marrow (Rutherford et al., 2002b). Surprisingly, when BMP-transduced human oral keratinocytes were compared with gingival fibroblasts, the structure of bone constituting the ossicles produced was very different. All the ossicles were smaller with greater bone mass (Table) and contained neither marrow nor clearly demarcated trabecular and cortical elements (Fig. 1).
BMP-transduced fibroblasts secrete biologically active BMP in vitro (Krebsbach et al., 2000) and immunoreactive BMP in vivo (Rutherford et al., 2002b). Additionally, these transduced fibroblasts differentiate into osteoblast-like cells in vivo (Rutherford et al., 2002b), producing bone that is a chimera of donor and host tissue. In vitro studies of the transduced HGF reveal the expression of cartilage glycosaminoglycans and type II collagen (Rutherford et al., 2002b). The fate of the transduced HOKC appears to be different. Whereas both HOKC (Fig. 1) and HGF (Buurma et al., 1999) continue to express respective differentiation markers in vivo, only the BMP-7-transduced HGF transdifferentiate to osteoblasts (Fig. 3) (Rutherford et al., 2002b). Hence the bone produced by the BMP-7/HOKC appears to be composed entirely of host-cell-produced matrix. However, this finding does not explain the difference in the quality of the bone produced by the different BMP-7-transduced cells. Since they are derived from different germ lines and are clearly distinct morphologically and functionally, the proteomes at the moment of infection with AdBMP-7 are different (e.g., keratin expression). The temporal changes in the transcriptomes and proteomes of each transduced differentiated cell type subsequent to BMP-7 transduction are unknown. Preliminary data from genome-wide RNA and protein expression studies of BMP-7/HGF reveal substantial and potentially interesting differences when compared with non-transduced and control-transduced HGF (RBR, unpublished data). The mechanism underlying these differences is being investigated. Analysis of these data and data from other concurrent studies will reveal new information regarding the mechanisms by which BMP signals induce fibroblast transdifferentiation.
|
ACKNOWLEDGMENTS
|
|---|
The authors acknowledge the excellent technical assistance of Brent Accurso. This work was supported by USPHS research grant DE12466 from the National Institute of Dental and Craniofacial Research, and by AR 46254 from the National Institute of Arthritis and Musculo-Skeletal and Skin Diseases, National Institutes of Health, Bethesda, MD 20892.
Received April 24, 2002;
Last revision December 6, 2002;
Accepted January 2, 2003
|
REFERENCES
|
|---|
Breitbart AS, Grande DA, Kessler R, Ryaby JT, Fitzsimmons RJ, Grant RT (1998). Tissue engineered bone repair of calvarial defects using cultured periosteal cells. Plast Reconstr Surg 101:567–574.[ISI][Medline]
Breitbart AS, Grande DA, Mason JM, Barcia M, James T, Grant RT (1999). Gene-enhanced tissue engineering: applications for bone healing using cultured periosteal cells transduced retrovirally with the BMP-7 gene. Ann Plast Surg 42:488–495.[Medline]
Buurma B, Gu K, Rutherford RB (1999). Transplantation of human pulpal and gingival fibroblasts attached to synthetic scaffolds. Eur J Oral Sci 107:282–289.[ISI][Medline]
Franceschi RT, Wang D, Krebsbach PH, Rutherford RB (2000). Gene therapy for bone formation: in vitro and in vivo osteogenic activity of an adenovirus expressing BMP7. J Cell Biochem 78:476–486.[ISI][Medline]
Krebsbach PH, Gu K, Franceschi RT, Rutherford RB (2000). Gene therapy directed osteogenesis: BMP-7-transduced human fibroblasts form bone in vivo. Hum Gene Ther 11:1201–1210.[ISI][Medline]
Lamont RJ, Chan A, Belton CM, Izutsu KT, Vasel D, Weinberg A (1995). Porphyromonas gingivalis invasion of gingival epithelial cells. Infect Immun 63:3878–3885.[Abstract]
Lee JY, Musgrave D, Pelinkovic D, Fukushima K, Cummins J, Usas A, et al. (2001). Effect of bone morphogenetic protein-2-expressing muscle-derived cells on healing of critical-sized bone defects in mice. J Bone Joint Surg Am 83(A):1032–1039.[Abstract/Free Full Text]
Lieberman JR (2000). Orthopaedic gene therapy. Fracture healing and other nongenetic problems of bone. Clin Orthop 379(Suppl):156–158.
Lieberman JR, Le LQ, Wu L, Finerman GA, Berk A, Witte ON, et al. (1998). Regional gene therapy with a BMP-2-producing murine stromal cell line induces heterotopic and orthotopic bone formation in rodents. J Orthop Res 16:330–339.[ISI][Medline]
Lieberman JR, Daluiski A, Stevenson S, Wu L, McAllister P, Lee YP, et al. (1999). The effect of regional gene therapy with bone morphogenetic protein-2-producing bone-marrow cells on the repair of segmental femoral defects in rats. J Bone Joint Surg Am 81:905–917.[Abstract/Free Full Text]
Reddi AH (1998). Role of morphogenetic proteins in skeletal tissue engineering and regeneration. Nat Biotechnol 16:247–252.[ISI][Medline]
Rutherford RB, TrailSmith MD, Ryan ME, Charette MF (1992). Synergistic effects of dexamethasone on platelet-derived growth factor mitogenesis in vitro. Arch Oral Biol 37:139–145.[ISI][Medline]
Rutherford RB, Moalli M, Franceschi RT, Wang D, Gu K, Krebsbach PH (2002a). Bone morphogenetic protein-transduced human fibroblasts convert to osteoblasts and form bone in vivo. Tissue Eng 8:441–452.[ISI][Medline]
Rutherford RB, Gu K, Racenis P, Krebsbach PH (2002b). Osteoblast conversion of BMP transduced fibroblasts: in vitro studies. Connect Tissue Res (in press).
TOP
ABSTRACT
INTRODUCTION
