| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
REVIEW |
University of Michigan School of Dentistry, 1011 N. University Ave., Ann Arbor, MI 48109-1078, USA; rennyf{at}umich.edu
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTION(I) LIMITATIONS OF CURRENT...(II) LESSONS FROM BIOLOGY:...(III) GENE THERAPY VECTORS...(IV) GENE THERAPY APPROACHES...(V) SECOND-GENERATION GENE...(VI) FUTURE DIRECTIONSREFERENCES |
|---|
KEY WORDS: adenovirus • BMP • gene expression • tissue engineering
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTION(I) LIMITATIONS OF CURRENT...(II) LESSONS FROM BIOLOGY:...(III) GENE THERAPY VECTORS...(IV) GENE THERAPY APPROACHES...(V) SECOND-GENERATION GENE...(VI) FUTURE DIRECTIONSREFERENCES |
|---|
As will be developed in this article, it is likely that future improvements in bone regeneration therapy will require new molecular biology and tissue-engineering-based technologies. While this review is not intended to be comprehensive, it will summarize major developments in gene therapy and tissue engineering related to bone regeneration, and will propose a new paradigm where gene therapy will be used to mimic natural processes occurring during bone development and repair. This will be accomplished by the use of regulated gene expression systems for the controlled delivery of regenerative factors, coupled with the development of appropriate scaffolds to serve as a platform for gene and cell delivery.
This article is divided into the following sections: (i) a brief discussion of the limitations of current approaches for bone regeneration; (ii) a review of our current understanding of bone development and fracture repair, and a discussion of how these natural processes might be mimicked by the use of gene therapy; (iii) an overview of gene therapy vectors and tissue-engineering scaffolds; (iv) a summary of recent gene therapy approaches for bone regeneration involving the expression of single regenerative molecules; (v) a description of recent work from this and other laboratories where gene therapy was used to express unique combinations of regenerative molecules having enhanced osteogenic activity; and (vi) some possible future directions for this exciting field.
|
(I) LIMITATIONS OF CURRENT METHODS FOR BONE REGENERATION |
|---|
|
TOPABSTRACTINTRODUCTION(I) LIMITATIONS OF CURRENT...(II) LESSONS FROM BIOLOGY:...(III) GENE THERAPY VECTORS...(IV) GENE THERAPY APPROACHES...(V) SECOND-GENERATION GENE...(VI) FUTURE DIRECTIONSREFERENCES |
|---|
For a bone regeneration therapy to be successful, inhibitory influences of the inflamed wound site must be overcome, appropriate precursor cells must be either recruited to or implanted at the site targeted for regeneration, and these cells must be given the appropriate signals and/or environmental cues to grow and differentiate in a controlled temporo-spatial manner. Current regenerative therapies include bone grafts, allogenic and xenograft bone matrix, root-conditioning agents and cell-occlusive barrier membranes (in periodontal disease), and, most recently, recombinant growth/differentiation factors. Each approach attempts to provide different regenerative signals. Bone grafts, still considered the "gold standard" for bone regeneration, contain all the components necessary for regeneration, including viable bone cells and progenitors, as well as the appropriate growth/differentiation factors. However, bone regeneration after grafting is quite variable, probably because of differences in the quality of the grafted bone (Albertson et al., 1991; Enneking and Mindell, 1991). In addition, severe morbidity/trauma occur at both donor and graft sites. Allogenic bone matrix provides a bone-like extracellular matrix and is a crude source of bone-associated growth factors and morphogens, including bone morphogenetic proteins (BMPs). These factors should, in theory, be able to attract appropriate precursor cells to the regeneration site and stimulate their differentiation to bone. However, this material has inconsistent osteoinductive activity, principally because it contains low levels of regenerative molecules that are further inactivated during processing. There is also a potential risk of disease transmission if this material is not appropriately processed (for review, see Cook et al., 1994).
Recombinant bone morphogenetic proteins (rBMPs) are the latest emerging therapeutic agents for bone regeneration. As will be discussed later in this article, these molecules have the unique ability to stimulate the differentiation of mesenchymal cells to chondrocytes and osteoblasts, and can induce formation of new bone at both ectopic and orthotopic sites. BMPs 2 and 7 can stimulate healing of controlled segmental defects in several organisms, including non-human primates (Cook et al., 1995), and have been examined in clinical trials for the treatment of fibular and tibial non-union fractures, maxillofacial reconstructions, and spinal fusion (for reviews, see Einhorn, 2003; Groeneveld and Burger, 2000). Limited FDA approval was recently obtained for the use of BMPs in selected applications, such as spinal fusions and non-unions. In spite of their promise of revolutionizing the field of bone regeneration, BMP devices as currently formulated must be used at very high concentrations to be effective (Govender et al., 2002). Furthermore, these devices fail in a certain percentage of cases (Winn et al., 2000).
In summary, the science of bone regeneration is clearly in its infancy, with all current or emerging therapies having serious limitations. While there are numerous partial explanations for this—including inconsistency of graft materials (bone grafts or allogenic bone matrix) or, in the case of BMPs, problems with protein stability or mode of delivery—an underlying problem is that no current regenerative strategies attempt to mimic events occurring during normal bone regeneration, where multiple regenerative factors interact in a defined temporal and spatial sequence. An appreciation for the elegance of these natural processes can be gained from a brief review of the mechanisms of bone development and fracture repair.
|
(II) LESSONS FROM BIOLOGY: NORMAL BONE DEVELOPMENT AND FRACTURE REPAIR ARE CHARACTERIZED BY COORDINATED EXPRESSION OF MULTIPLE SOLUBLE AND INTRACELLULAR FACTORS |
|---|
|
TOPABSTRACTINTRODUCTION(I) LIMITATIONS OF CURRENT...(II) LESSONS FROM BIOLOGY:...(III) GENE THERAPY VECTORS...(IV) GENE THERAPY APPROACHES...(V) SECOND-GENERATION GENE...(VI) FUTURE DIRECTIONSREFERENCES |
|---|
Specific transcription factors control the differentiation of mesenchymal precursors down chondrocytic and osteoblastic lineages. RUNX2, the product of the Cbfa1 gene and an essential transcription factor for osteoblast and hypertrophic chondrocyte differentiation, is expressed at early times in limb development coincident with the formation of mesenchymal condensations (Ducy et al., 1997; Komori et al., 1997; Otto et al., 1997). Two additional factors, SOX9 and Osterix (OSX), have more selective roles in chondrocyte and osteoblast differentiation, respectively. RUNX2 protein levels can be increased by BMPs under certain conditions, although it is not known if its induction in mesenchymal condensations is primarily in response to BMPs. OSX is also induced by BMP treatment but also requires RUNX2, in that it is not present in Cbfa1 –/– mice (Nakashima et al., 2002; Yagi et al., 2003). Details of differentiation events involved in the progression of a mesenchymal stem cell to a chondrocyte and/or osteoblast are not well-understood, but clearly involve restrictions in lineage potential of pluripotent mesenchymal cells by controlling the cellular transcriptional program.
Numerous laboratories have examined RUNX2 regulation and its mechanism of action. This factor controls the transcriptional activity of target genes by binding to specific enhancer regions in promoter or regulatory regions. The overall mechanism of transcriptional activation is not well-understood. However, a common theme is that RUNX2 functions in heterodimeric complexes with other nuclear factors on the chromatin (for review, see Franceschi, 2003). Some of these complexes enhance transcriptional activity (complexes with CBFß, SMADS, AP1 factors), while others are inhibitory (complexes with TLE/GROUCHO, TWIST). Furthermore, post-translational modifications—such as mitogen-activated protein kinase and protein kinase A-mediated phosphorylations—appear to be important for regulating RUNX2 activity (Selvamurugan et al., 2000; Xiao et al., 2000).
Development of the Craniofacial Skeleton
The flat bones of the cranial vault, certain facial bones, and parts of the mandible form through an intramembranous process involving the direct differentiation of mesenchymal precursors into osteoblasts (Karaplis, 2002). Although the mesenchymal osteogenic precursors in the skull are derived from the neural crest, these cells use many of the same signaling molecules found in the limb bud, suggesting that the developmental controls are similar. Thus, major patterning, proliferation, and differentiation functions in the skull have been attributed to FGFs, SHH, BMPs, and various homeotic genes (Msx, Dlx). FGFs in particular have profound effects on craniofacial bone formation. Antibody blocking of endogenous FGF2 prevents cranial suture osteogenesis (Moore et al., 2002). In addition, several activating mutations in the FGF receptors, FGFR1, FGFR2, and FGFR3, cause premature fusion of cranial sutures (craniosynostosis), due to enhanced osteoblast activity (Chen et al., 1999; Zhou et al., 2000; Britto et al., 2001). Roles for BMPs have been inferred from experiments showing that ectopic application of BMP 2 or 4 alters the patterning of facial bones (Barlow and Francis-West, 1997). Also, many of the same transcription factors identified in long bones also function during cranial development, including RUNX2, TWIST, DLX1-3, DLX5/6, MSX1/2, and AP-1 factors (Merlo et al., 2000; Robledo et al., 2002).
Fracture Healing is Orchestrated by Many of the Same Regenerative Molecules Identified in the Limb
Fracture repair is characterized by the involvement of multiple factors that are expressed in a defined temporal sequence (Khan et al., 2000; Gerstenfeld et al., 2003), and, in some ways, can be considered a recapitulation of endochondral bone development. After a fracture is sustained, the initial inflammatory response recruits activated macrophages and polymorphonuclear neutrophils (PMNs) that together endocytose microdebris and micro-organisms. An initial hematoma is formed under the control of platelet-derived growth factor (PDGF), insulin-like growth factors (IGFs), tumor-derived growth factor ß (TGF-ß ), and FGF2 produced by macrophages (Radomsky et al., 1998; Nakajima et al., 2001). This is followed by blastema formation, with proliferation of granulation tissue fibroblasts. Subsequent BMP, TGF-ß, and FGF2 secretion induces osteoprogenitors to differentiate into chondrocytes (endochondral bone only) and osteoblasts. Generally, FGFs, IGFs, and PDGFs act as mitogenic factors that are widely distributed in the soft callus early in fracture repair, while BMPs are more associated with the chondrocytes and osteoblasts present later in the healing process. However, individual BMPs also exhibit unique temporal expression patterns, with BMP 2 mRNA being highest during the initial inflammatory phase, and BMPs 3a, 4, 7, and 8 highest during late chondrogenic and osteogenic phases of fracture repair (Cho et al., 2002).
Cooperative Interactions between and among Osteogenic Factors
A common thread linking all the biological systems discussed above is the involvement of multiple bioactive factors in bone induction. In some cases, each factor makes a separate contribution to the osteogenic response. For example, Hox genes specify early patterning events, growth factors like FGFs and PDGF stimulate angiogenesis and osteogenic precursor proliferation, while BMPs initiate overt bone formation. In some cases, combinations of bioactive factors can synergistically stimulate bone regeneration. For example, the combined application of FGF4 and BMP 2 synergistically interacts to promote bone formation when implanted into a suitable matrix (Nakajima et al., 2001; Kubota et al., 2002; Lisignoli et al., 2002). In some cases, osteogenic factors form oligomeric complexes with enhanced biological activity. This concept is best illustrated for the BMPs. Although homodimers of BMPs 2, 4, and 7 can induce ectopic bone formation, there is now strong evidence that these factors act in combination. Thus, BMPs 2, 4, and 7 are expressed in overlapping patterns during limb development (Lyons et al., 1995; Nishimatsu and Thomsen, 1998). Similarly, overlapping expression of BMPs 2, 3a, 4, 7, and 8 is observed at various times during fracture healing (Cho et al., 2002). Although most bone induction studies with BMPs used homodimeric molecules, BMP 2/7 and 4/7 heterodimers can be detected when cDNAs encoding these molecules are co-expressed in cell culture. Furthermore, these BMP heterodimers are reported to have greater biological activity than their constituent homodimers (Lyons et al., 1995; Israel et al., 1996; Nishimatsu and Thomsen, 1998; Tsuji et al., 1998).
Mimicking the Natural Process of Bone Formation in the Clinic: Why Gene Therapy?
Three general conclusions can be drawn from the above discussion:
Given the complexity of these natural processes, it is reasonable to predict that current regenerative therapies involving single bioactive factors or the implantation of single cell types will never be able to realize the overall goal of bone tissue engineering, which is to reconstruct entire bones with associated joints, sutures, and precise three-dimensional morphology. Such complex types of regeneration can be achieved only if systems are developed that allow for control over the types of factors produced and their temporal sequence of release, while at the same time providing the appropriate target precursor cell populations, all within a precisely defined three-dimensional lattice.
Scaffold systems are currently under development that allow for the controlled release of proteins from artificial matrices (Richardson et al., 2001; Tabata, 2003). For example, factors have been encapsulated in microspheres and/or incorporated into tissue-engineering scaffolds (Sheridan et al., 2000; Lutolf et al., 2003). Manipulation of microsphere composition can be used to control the rate of factor release. However, as currently devised, these systems control only rate of release and cannot be turned on or off at specific times. Furthermore, they normally release only a single factor at a time. Although none of these limitations is necessarily insurmountable, protein-based delivery devices do not, at present, provide the best means of examining the interplay between and among multiple factors at a regeneration site.
Gene therapy approaches, in contrast, have the potential to provide control over the timing, distribution, and level of multiple regenerative factors that can be either simultaneously or sequentially expressed in a tissue-specific manner. With the completion of the human genome, sequence data are now available for all known regenerative factors. Furthermore, a wide range of viral and non-viral vectors is now available to allow for efficient gene transfer into numerous cell types, including osteogenic precursors and stem cells. Considerable progress has also been made in the design of tissue-engineering scaffolds that can be fabricated into precise three-dimensional configurations to support the growth of genetically modified cells. Last, several regulated gene expression systems have been developed for the efficient activation and inhibition of gene expression. The remainder of this article will explore each of these areas, review recent progress in the use of gene therapy for bone regeneration, and outline future directions for this exciting field.
|
(III) GENE THERAPY VECTORS AND TISSUE-ENGINEERING SCAFFOLDS |
|---|
|
TOPABSTRACTINTRODUCTION(I) LIMITATIONS OF CURRENT...(II) LESSONS FROM BIOLOGY:...(III) GENE THERAPY VECTORS...(IV) GENE THERAPY APPROACHES...(V) SECOND-GENERATION GENE...(VI) FUTURE DIRECTIONSREFERENCES |
|---|
Adenoviruses have highly evolved mechanisms for delivery of DNA to cells and, unlike retroviruses, are not dependent on cell replication for infection. Infection involves the initial cell binding of specific viral proteins, such as the fiber capsid protein by coxsackievirus and adenovirus receptor (CAR), and binding of the viral penton base by 
v integrins on the cell surface. The broad distribution of these receptors explains why adenoviruses can be used to infect such a wide range of cell types (Neumann et al., 1988; Bergelson et al., 1997). After infection, adenoviruses do not normally integrate into the host genome; instead, they remain in the nucleus as an episome that is gradually degraded as cells divide (Oligino et al., 2000). Most gene therapy studies conducted to date used so-called "first generation" adenoviruses. Although these vectors have been genetically modified to be replication-incompetent, in the absence of a helper cell line containing the missing genes, they contain most of the viral genome, including genes encoding the major coat proteins. For this reason, cells infected with first-generation adenovirus vectors will secrete viral proteins and elicit an immune response that will eventually result in their clearance from the body. The combined effects of episomal localization and immunogenicity cause transgene expression from first-generation adenoviruses to be quite brief. As an example, we recently determined that the in vivo duration of BMP expression from fibroblasts transduced with Ad-BMP2 was less than 2 weeks (Zhao et al., 2005a). This short period of BMP production has advantages and disadvantages; it can prevent bone formation from exceeding the boundaries of the desired regeneration site, but it can also restrict osteogenesis so much that it is no longer therapeutically useful. Because of these problems with immunogenicity, second-generation adenovirus-based vectors have been developed that lack genes for most or all viral proteins (Armentano et al., 1997; Hartigan-O’Connor et al., 1999). These "gutted" vectors have the advantage of being able to package up to 30 kb of foreign DNA. However, they can be propagated only in the presence of helper viruses that contain the missing viral genes necessary to form a viable capsid. In spite of these limitations, first-generation adenoviruses continue to be extremely useful for defining which regenerative factors or groups of factors can best stimulate bone regeneration.
Retroviruses are the most extensively used vectors for gene therapy applications. Unlike adenoviruses, these RNA viruses use reverse-transcriptase to make a double-stranded copy of their genome that is randomly integrated into the host cell genome and then replicated as the cell divides (Oligino et al., 2000). Most clinical trials to correct genetic diseases used replication-incompetent murine leukemia virus (MLV), which is non-pathogenic in humans. MLV shares minimal homology with human retroviruses, and is therefore unlikely to generate a pathogenic virus by recombining with other human viruses (Danos and Heard, 1992). After integration, MLV exists in the host cell only as a DNA copy that is unable to transcribe genes encoding viral coat proteins. For this reason, the integrated virus cannot elicit an immune response. However, there are two negative aspects to viral integration. First, the integrated retrovirus can disrupt normal cell function by insertional mutagenesis, an event that, unfortunately, was recently observed in some patients (Hacein-Bey-Abina et al., 2003; Noguchi, 2003). Second, since retroviral genomes are replicated as cells divide, they remain a permanent part of the host cell, and any transgenes they carry must be under the control of a regulated promoter if they are to be turned off after regeneration has been achieved (see "Future Directions" below). Also, as noted above, retroviruses infect only replicating cells. For this reason, MLV vectors are most suitable for ex vivo gene therapy applications.
Lentiviruses are a specialized family of retroviruses, the most common example being human immunodeficiency virus 1 (HIV-1). Unlike the oncogenic retroviruses described in the preceding paragraph, lentiviral vectors can infect non-dividing cells, including hematopoietic progenitor cells and marrow-derived stromal cells (Miyoshi et al., 1999; Zhang et al., 2002). Although lentiviruses integrate into the host cell genome, integration sites may be more limited than is the case for traditional retroviruses. This feature may explain why lentiviral vectors give more stable gene expression after integration into cells, and also reduces the probability of causing disease through insertional mutagenesis (Vigna and Naldini, 2000). These features have prompted considerable interest in this family of viruses as vectors for gene therapy. However, because of the lethal nature of many of the parent viruses used to develop lentiviral vectors (i.e., HIV-1 and related viruses), a considerable effort has been made to produce modified viruses that are safe for clinical use. Strategies for improving the safety of lentiviral vectors include altering the viral genome to minimize the likelihood of recombinations that could produce replication-competent viruses, and elimination of all viral genes that are not necessary for gene transfer. In a recent report, a lentivirus vector was developed to express BMP 2. Marrow stromal cells infected with this vector were shown to stably express BMP 2 for at least 8 weeks in vitro. Furthermore, virally infected cells formed bone after implantation into muscle pouches of immunodeficient mice (Sugiyama et al., 2005).
Like gutted adenovirus vectors, adeno-associated virus (AAV) gene therapy vectors do not express viral proteins and, therefore, are non-immunogenic. These vectors have the further advantage of being non-pathogenic in humans, yet they are able to infect a wide range of dividing and non-dividing cells without integrating into the genome (Oligino et al., 2000). Recently, efficient methods for growing recombinant AAV were developed that avoid the use of helper viruses, which previously made purification difficult. Although only a few studies have examined AAV in gene therapy applications for bone regeneration, Luk and co-workers recently showed that AAV encoding the BMP 4 gene could stimulate bone formation after injection into an intramuscular site (Luk et al., 2003). AAV vectors show considerable promise for eventual use in the clinic. However, because they are still quite difficult to construct, it is probably more appropriate, for research purposes, that critical genetic interactions first be studied with more traditional adeno- or retrovirus vectors.
Non-viral vectors that contain only naked DNA and some type of carrier to facilitate cell uptake have several advantages over viral vectors, including ease of manufacture, stability, low immunogenicity, and low likelihood of being inserted into the host cell genome (Oligino et al., 2000). This class of vector can be injected directly into tissues as naked DNA, adsorbed to liposomes, or attached to micro-projectiles of gold or tungsten that are injected into cells by high-pressure gas or by an electrical discharge (Klein et al., 1992). Although recent advances involving condensation of DNA with liposomes or other carriers have the potential to enhance the uptake of nonviral DNA by cells (Kwok et al., 2001), as currently formulated, cellular uptake of non-viral vectors is an extremely inefficient process, estimated to be 10–9 that of viral vectors (Franceschi et al., 2000).
Tissue-engineering Scaffolds
Although not the emphasis of this article, the design of tissue-engineering scaffolds is critical for the success of any gene therapy strategy. Such scaffolds have the potential to control release rates of gene therapy vectors and/or provide a suitable three-dimensional environment for the growth and differentiation of osteoprogenitor and mesenchymal stem cells (for excellent reviews on various aspects of scaffold design, see Alsberg et al., 2001b; Richardson et al., 2001; Liu and Ma, 2004). Through the use of computer-aided design and three-dimensional printing technologies, scaffolds can also be fabricated into precise geometries. This is particularly important for craniofacial applications, where it is critical for the morphology of regenerated bone to be precisely controlled. With this technology, a three-dimensional reconstruction of the specific craniofacial region targeted for regeneration can be generated from computer-aided tomography (CAT) scans (for an example of this approach, see Chang et al., 1999). Furthermore, recent developments in polymer chemistry have made it possible for both the surface properties and the micro-porosity of polymer scaffolds to be extensively modified, to provide a wide range of structures and bioactive surfaces. Such scaffolds have the potential to provide environments conducive to the growth of specific cell types. For example, surface modification of alginate scaffolds has been shown to provide a supportive environment for the growth of osteoblasts (Alsberg et al., 2001a). Ultimately, the combined use of precisely engineered scaffolds and the appropriate combinations of cells expressing regenerative factors will be necessary if entire tissues or organs are to be engineered.
|
(IV) GENE THERAPY APPROACHES INVOLVING EXPRESSION OF SINGLE REGENERATIVE FACTORS |
|---|
|
TOPABSTRACTINTRODUCTION(I) LIMITATIONS OF CURRENT...(II) LESSONS FROM BIOLOGY:...(III) GENE THERAPY VECTORS...(IV) GENE THERAPY APPROACHES...(V) SECOND-GENERATION GENE...(VI) FUTURE DIRECTIONSREFERENCES |
|---|
.): Vectors were either directly delivered to in vivo sites (In Vivo Gene Therapy) or used to transduce, in tissue culture, cells that were subsequently implanted into animals (Ex Vivo Gene Therapy). The principal advantage of this basic gene therapy approach is that cells can be engineered to produce the molecule of interest for sustained periods by transduction with an appropriate vector, thereby avoiding problems associated with BMP protein degradation and delivery from implanted matrices.
|
In studies of greater therapeutic significance from the Center for Craniofacial Regeneration at the University of Michigan, the biological activity of Ad-BMP7 was examined in two separate orthotopic regeneration models involving critical-sized defects of calvaria and long bones (Krebsbach et al., 2000; Rutherford et al., 2002) and a periodontal alveolar defect (Jin et al., 2003). In all cases, studies used syngeneic dermal fibroblasts cultured from Lewis or Fisher rats that were transduced with Ad-BMP7 ex vivo. For the cranial model, a 9-mm calvarial defect was created in Lewis rats by means of a trephine (Krebsbach et al., 2000). This size of defect will not heal spontaneously in the lifetime of the animal. Cells transduced with either lacZ control adenovirus or Ad-BMP7 were adsorbed to gelatin sponges before being placed in defects. The Ad-BMP7-transduced cells induced sufficient bone formation in 4 weeks to close the calvarial defect almost completely. In contrast, only minimal bone formation was detected in defects receiving control virus. For the long-bone model, a 2- to 3-mm osteotomy was created in femurs of Fisher rats previously immobilized by an external fixator (Rutherford et al., 2002). Transduced dermal fibroblasts were suspended in a type I collagen hydrogel and adsorbed to a gelatin sponge before being placeed in the defect. Histological and radiological examination of defects revealed significant bone healing by Ad-BMP7-transduced cells after 6 weeks. Both bone and cartilage formation was seen. In contrast, lacZ-transduced cells formed only fibrous connective tissue. The periodontal alveolar bone defect model involved removal of bone overlying the mandibular first molar, and the periodontal ligament and cementum from the first and second molars, followed by implantation of virally transduced fibroblasts (Jin et al., 2003). Although both cartilage and bone formation was observed in this model after 10 days, complete bridging of the defect with new bone was observed after 35 days. Furthermore, the denuded tooth root surface in Ad-BMP7-treated animals was covered with a thin layer of new cementum and showed evidence of fiber attachment.
The studies described above are a small part of what has become an extensive body of literature describing the use of gene therapy vectors encoding individual regenerative molecules in bone regeneration. Since much of this work was recently reviewed (Wu et al., 2003; Baltzer and Lieberman, 2004), only a few illustrative examples will be discussed.
Direct injection of an adenovirus encoding BMP 2 (Ad-BMP2) was shown to heal segmental femoral defects partially in both rabbits and rats (Baltzer et al., 2000a,b). For these experiments, muscle surrounding the defect was used to create a closed chamber between the cut ends of the bone, and adenovirus was directly injected into this site without the use of a carrier. In vivo gene therapy approaches have also been developed using direct transfer of plasmid DNAs encoding BMP 4 and a parathyroid hormone fragment that were mixed with a type I collagen carrier (Fang et al., 1996; Bonadio et al., 1999). This so-called "gene-activated matrix", or GAM, was shown to heal a canine femur segmental defect partially after 52 weeks. A common feature of both these in vivo gene therapy strategies was the need for very large amounts of the gene therapy vector (2 x 1010 adenovirus particles or 100 mg plasmid DNA). This requirement is likely explained by the degradation and clearance of adenovirus vectors by the immune system, and the inefficient uptake of plasmid DNA by cells.
Ex vivo approaches have also been extensively examined. As noted above, viral transduction of cells in tissue culture is a much more efficient process than that observed in vivo. Furthermore, specific cell types, including marrow stromal cells (MSCs), can be selected for gene transfer. MSCs are derived from the tissue-culture plastic adherent, non-hematopoietic fraction of marrow. A small subfraction of MSCs has stem-cell-like properties and the capability to differentiate along all 4 mesenchymal lineages (bone, cartilage, muscle, fat) (Jiang et al., 2002). In addition to being a source of BMPs after transduction, these cells directly respond to BMPs and participate in osteogenesis after implantation (Musgrave et al., 2000). This may be particularly important at regenerative sites, where the supply of endogenous osteogenic precursors is limiting. MSCs transduced with adenoviruses encoding BMPs have been shown to stimulate bone regeneration in several experimental models. For example, Ad-BMP2-transduced cells were shown to heal femoral segmental defects in rats (Lieberman et al., 1998, 1999), bilateral maxillary defects and calvarial defects in swine (Chang et al., 2003a,b), and to induce spine fusion in rabbits and rats (Riew et al., 1998; Wang et al., 2003). Cell populations with mesenchymal stem-cell-like properties have also been isolated from muscle and adipose tissue (Lee JY et al., 2000; Morizono et al., 2003) and can induce bone formation after transduction with BMP 2 or BMP 4 vectors (Peng et al., 2002; Gimble and Guilak, 2003).
Taken together, these studies demonstrate the power of single-factor gene therapy to induce bone formation in clinically relevant models of bone regeneration and provide strong evidence for the feasibility of this approach as a clinical alternative to BMP protein therapy.
|
(V) SECOND-GENERATION GENE THERAPY APPROACHES: EXPLOITING THE POTENTIAL OF GENE THERAPY TO EXPRESS COMBINATIONS OF INTERACTING GENES |
|---|
|
TOPABSTRACTINTRODUCTION(I) LIMITATIONS OF CURRENT...(II) LESSONS FROM BIOLOGY:...(III) GENE THERAPY VECTORS...(IV) GENE THERAPY APPROACHES...(V) SECOND-GENERATION GENE...(VI) FUTURE DIRECTIONSREFERENCES |
|---|
Cooperative Interactions between and among BMPs
As noted above (see "Cooperative Interactions between and among Osteogenic Factors"), multiple BMPs are co-expressed during development and fracture healing and may exist as heterodimers. Furthermore, BMP 2/7 and 4/7 heterodimers have greater biological activity than do their constituent homodimers. Taken together, these studies suggest that certain advantages may be accrued by the use of combinations of BMPs for bone regeneration. To begin addressing this possibility, we examined the ability of combinations of adenoviruses expressing BMPs 2, 4, and 7 to induce in vitro osteoblast differentiation and in vivo bone formation (Franceschi et al., 2004; Zhao et al., 2005a). For in vitro studies, mesenchymal cell lines were transduced with individual adenoviruses containing BMP 2, 4, or 7 cDNA under the control of a CMV promoter (Ad-BMP2, Ad-BMP4, Ad-BMP7) or virus combinations. Significantly, combined transduction of Ad-BMP2 plus Ad-BMP7, or Ad-BMP4 plus Ad-BMP7, resulted in a synergistic stimulation of osteoblast differentiation. This synergy is best explained by the formation of BMP 2/7 and BMP 4/7 heterodimers. To test in vivo biological activity, we transduced fibroblasts with specific virus combinations and subcutaneously implanted them into C57BL6 mice. Consistent with in vitro results, strong synergy was observed with the combined Ad-BMP2/BMP7 treatment, which induced two- to three-fold more bone than would be predicted based on the activity of individual Ad-BMPs. This study shows that dramatic enhancement of osteogenesis can be achieved with gene therapy to express specific combinations of interacting regenerative molecules. Because of their increased biological activity, such vector formulations can achieve bone regeneration at much lower viral titers, thereby minimizing possible toxicity and/or immune responses. Recently, Zhu and co-workers used a similar approach to induce in vitro osteoblast differentiation and in vivo spinal fusion with combinations of Ad-BMP2 and Ad-BMP7 (Zhu et al., 2004).
Synergies between Angiogenic and Osteogenic Signals
Normal bone development will not occur in the absence of blood vessel formation. During development, vascular infiltration of the cartilage anlagen is the earliest step in formation of the bony collar (Karaplis, 2002). The importance of angiogenesis in bone formation can also be inferred from experiments where specific inhibitors of this process were shown to interfere with BMP2-induced bone formation (Mori et al., 1998). To examine possible interactions between BMP and angiogenic signals, Peng and co-workers used retroviral gene transfer to establish stable muscle-derived stem cell (MDSC) lines expressing BMP 4, vascular endothelial growth factor (VEGF), and the VEGF antagonist, soluble Flt1, that were subsequently examined, alone or in combination, for osteogenic activity (Peng et al., 2002). VEGF by itself did not enhance the osteogenic activity of MDSC. However, it was found to act synergistically with BMP 4 to increase mesenchymal stem cell recruitment and survival and to stimulate bone formation and repair of a calvarial defect. Effects of VEGF on bone healing were shown to be critically dependent on the ratio of VEGF to BMP, with excessive VEGF/BMP ratios actually shown to inhibit osteogenesis. Furthermore, VEGF effects were totally blocked by the soluble Flt1 antagonist. In related studies, Huang and co-workers recently demonstrated enhanced osteogenesis by human marrow stromal cells adsorbed to poly(lactic-co-glycolic acid) scaffolds containing combinations of condensed plasmid DNA encoding BMP 4 and VEGF (Huang et al., 2005). These studies emphasize the importance of vascularization in the overall process of bone regeneration and show how the co-expression of angiogenic and osteoinductive factors can enhance bone formation.
Use of an Osteogenic Transcription Factor to Enhance BMP Responsiveness of Marrow Stem Cells
RUNX2 is the bone-related product of Cbfa1, a master gene controlling bone and hypertrophic cartilage differentiation (Ducy et al., 1997). RUNX2 is essential for the differentiation of osteoblasts and hypertrophic chondrocytes, and, for this reason, the skeletons of Cbfa1 –/– mice are totally devoid of mineral (Komori et al., 1997; Otto et al., 1997). Overexpression of RUNX2 in mesenchymal cells can increase osteoblast-specific gene expression, at least in part, by directly binding to enhancer sequences in target genes, although it is likely that some of its actions are indirect and involve induction of other downstream factors (Franceschi, 2003). Since BMPs can up-regulate RUNX2 in certain systems (Ducy et al., 1997; Lee KS et al., 2000), it is likely that some of the actions of BMPs may be mediated by this transcription factor. However, other BMP activities are clearly RUNX2-independent. For example, in response to BMP treatment and receptor activation, R-Smad-Smad4 complexes can directly interact with and activate certain target genes in the absence of RUNX2 (Jonk et al., 1998; Hanai et al., 1999). Consistent with these results, BMP 2 can induce the expression of osteocalcin and alkaline phosphatase in cells from Cbfa1–/– mice (Komori et al., 1997). Taken together, these studies suggest that BMPs and RUNX2 may stimulate bone formation through distinct, but complementary, pathways and, furthermore, that it may be possible to enhance the osteogenic activity of mesenchymal precursor cells by transduction with a RUNX2 expression vector.
To begin exploring possible interactions between BMPs and RUNX2, we conducted studies to determine the extent to which a RUNX2 adenovirus can induce functional osteoblast differentiation in the C3H10T1/2 mesenchymal cell line, and then to examine functional interactions between RUNX2 and BMP 2 in stimulation of osteoblast differentiation in vitro and in vivo (Yang et al., 2003). Ad-RUNX2 induced osteoblast markers, such as alkaline phosphatase and osteocalcin, in C3H10T1/2 cells in vitro, and weakly stimulated extracellular matrix mineralization. In contrast, cells transduced with Ad-BMP2 exhibited higher levels of mineralization, but expressed barely detectable levels of alkaline phosphatase, RUNX2, and osteocalcin. Significantly, when cells were transduced with optimal titers of both Ad-RUNX2 and Ad-BMP2, osteoblast differentiation was stimulated to levels that were up to 10-fold greater than those seen with either Ad-RUNX2 or Ad-BMP2 alone. These results showed that Ad-RUNX2 transduction increased the sensitivity of cells to Ad-BMP2 treatment. The responsiveness of C3H10T1/2 cells to recombinant BMP protein (as opposed to Ad-BMP2 virus) was also enhanced by Ad-RUNX2 transduction (Phimphilai et al., 2005). For the measurement of in vivo osteogenic activity, virally transduced cells were subcutaneously implanted into immunodeficient mice. Cells transduced with control virus produced only fibrous tissue, while those transduced with Ad-RUNX2 produced limited amounts of both cartilage and bone, as would be predicted from the known effects of RUNX2 in the formation of both bone and cartilage, while cells transduced with either Ad-BMP2 alone or Ad-BMP2 plus Ad-RUNX2 induced prominent ossicles containing cartilage, bone, and a marrow cavity. However, ossification in the Ad-BMP2 plus Ad-RUNX2 group was more extensive. Both mineral content and fractional bone area were greater than that seen with Ad-BMP2 alone. The increase in bone area was mainly due to enhanced deposition of cortical bone in the Ad-BMP2 plus Ad-RUNX2 group.
More recently, we examined the ability of Ad-RUNX2 to enhance the osteogenic activity of murine MSCs in vitro and in vivo (Zhao et al., 2005b). As noted above, a fraction of MSCs have mesenchymal stem cell properties. For the power of MSCs to be harnessed for bone regeneration, methods must be developed to direct the differentiation of these cells selectively toward osteoblasts and chondrocytes. Transduction of primary MSC cultures with Ad-RUNX2 was shown to increase RUNX2 protein expression in a dose-dependent manner and to stimulate osteoblast differentiation, as measured by induction of several osteoblast marker genes and mineralization. For the assessment of in vivo osteogenic activity, Ad-RUNX2 and control (Ad-LacZ) cells were adsorbed to 2 different carrier scaffolds and subcutaneously implanted into C57BL6 mice. In both cases, MSCs expressing RUNX2 formed substantially more bone than did cells transduced with control vector. Ad-RUNX2-transduced MSCs were also shown to exhibit enhanced responsiveness to Ad-BMP2 or purified recombinant BMP2.
In related studies, stable transduction of primary skeletal myoblasts with a retrovirus encoding RUNX2 has been shown to stimulate in vitro osteoblast differentiation and osteogenic activity (Gersbach et al., 2004a,b). In these studies, RUNX2-expressing cells also exhibited enhanced responsiveness to BMP 2. Last, Zheng and co-workers recently showed that Ad-RUNX2-transduced marrow stromal cells could partially heal a calvarial defect in BALB/c mice under conditions where non-transduced MSCs failed to form bone (Zheng et al., 2004).
Taken together, these results show that the responsiveness of osteoprogenitor cell populations to BMPs can be enhanced in vitro and in vivo by factors like RUNX2 that are major regulators of the osteoprogenitor lineage. In addition, they suggest possible therapeutic benefits that may be derived from the use of bone-related transcription factors to enhance BMP responsiveness in osteoprogenitor populations.
|
(VI) FUTURE DIRECTIONS |
|---|
|
TOPABSTRACTINTRODUCTION(I) LIMITATIONS OF CURRENT...(II) LESSONS FROM BIOLOGY:...(III) GENE THERAPY VECTORS...(IV) GENE THERAPY APPROACHES...(V) SECOND-GENERATION GENE...(VI) FUTURE DIRECTIONSREFERENCES |
|---|
One possible way to overcome this problem would be to control the temporal sequence of factor expression. For example, a regeneration strategy might be envisioned where early FGF2 expression could be used to expand precursor cells in a fracture callous, followed by induction of a BMP to stimulate chondrocyte and osteoblast differentiation. Recently described gene therapy vectors that incorporate inducible promoters may be particularly useful in this regard. For example, Rivera and co-workers described a retroviral vector that contains an inducible promoter controlled by rapamycin (Rivera et al., 1999). Because of its high oral bioavailability, rapamycin can be easily administered in the diet. On being absorbed, it interacts with a two-component engineered transcription factor in virally transduced cells to activate the target gene of choice. With such a system, regenerative strategies can be envisioned in which an early proliferative signal (e.g., FGF2) is followed by a differentiation signal (e.g., a BMP) under the control of an inducible promoter. We recently developed a rapamycin-inducible BMP 2 expression system and showed that it has osteogenic activity in an in vivo cranial defect model (Koh and Franceschi, 2005). In related studies, Moutsatsos and co-workers described a tetracycline-regulated system for the controlled expression of BMP 2 that was engineered in the C3H10T1/2 cell line and was shown to induce repair of a segmental long-bone defect in vivo (Moutsatsos et al., 2001). With such systems, it may be possible to control the temporal sequence of release for groups of interacting factors, thereby gaining greater control over the regenerative response.
A second very interesting, but thus far unexplored, area has to do with the role of cell orientation in bone regeneration. As noted in the section on tissue-engineering scaffolds, technologies now exist to manufacture scaffolds with precisely controlled three-dimensional shapes, as well as with variable porosity and surface modifications that may encourage the growth of certain cell types. Might it be possible to position genetically modified cells in these scaffolds precisely, to produce appropriate gradients in space and time that will guide the development of complex bone-related structures, including cortical and trabecular bone, joints, and ligaments? For example, it might be possible to seed a scaffold with mesenchymal stem cells and, with gene therapy, to guide these cells along the appropriate lineages to form all bone-related tissues and tissue interfaces. In the future, these and similar approaches may allow for control of the amount, type, and shape of bone formed at specific regenerative sites.
|
ACKNOWLEDGMENTS |
|---|
Received January 7, 2005; Accepted June 10, 2005
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTION(I) LIMITATIONS OF CURRENT...(II) LESSONS FROM BIOLOGY:...(III) GENE THERAPY VECTORS...(IV) GENE THERAPY APPROACHES...(V) SECOND-GENERATION GENE...(VI) FUTURE DIRECTIONSREFERENCES |
|---|
Alsberg E, Anderson KW, Albeiruti A, Franceschi RT, Mooney DJ (2001a). Cell-interactive alginate hydrogels for bone tissue engineering. J Dent Res 80:2025–2029.
Alsberg E, Hill EE, Mooney DJ (2001b). Craniofacial tissue engineering. Crit Rev Oral Biol Med 12:64–75.[Abstract]
Anusaksathien O, Giannobile WV (2002). Growth factor delivery to re-engineer periodontal tissues. Curr Pharm Biotechnol 3:129–139.[Medline]
Armentano D, Zabner J, Sacks C, Sookdeo CC, Smith MP, St George JA, et al. (1997). Effect of the E4 region on the persistence of transgene expression from adenovirus vectors. J Virol 71:2408–2416.[Abstract]
Baltzer AW, Lieberman JR (2004). Regional gene therapy to enhance bone repair. Gene Ther 11:344–350.[ISI][Medline]
Baltzer AW, Lattermann C, Whalen JD, Ghivizzani S, Wooley P, Krauspe R, et al. (2000a). Potential role of direct adenoviral gene transfer in enhancing fracture repair. Clin Orthop Rel Res 379(Suppl):S120–S125.
Baltzer AW, Lattermann C, Whalen JD, Wooley P, Weiss K, Grimm M, et al. (2000b). Genetic enhancement of fracture repair: healing of an experimental segmental defect by adenoviral transfer of the BMP-2 gene. Gene Ther 7:734–739.[ISI][Medline]
Barlow AJ, Francis-West PH (1997). Ectopic application of recombinant BMP-2 and BMP-4 can change patterning of developing chick facial primordia. Development 124:391–398.[Abstract]
Bergelson JM, Cunningham JA, Droguett G, Kurt-Jones EA, Krithivas A, Hong JS, et al. (1997). Isolation of a common receptor for Coxsackie B viruses and adenoviruses 2 and 5. Science 275:1320–1323.
Bitgood MJ, McMahon AP (1995). Hedgehog and Bmp genes are coexpressed at many diverse sites of cell-cell interaction in the mouse embryo. Dev Biol 172:126–138.[ISI][Medline]
Bonadio J, Smiley E, Patil P, Goldstein S (1999). Localized, direct plasmid gene delivery in vivo: prolonged therapy results in reproducible tissue regeneration. Nat Med 5:753–759.[ISI][Medline]
Bostrom MP, Saleh KJ, Einhorn TA (1999). Osteoinductive growth factors in preclinical fracture and long bone defects models. Orthop Clin North Am 30:647–658.[ISI][Medline]
Britto JA, Evans RD, Hayward RD, Jones BM (2001). From genotype to phenotype: the differential expression of FGF, FGFR, and TGFbeta genes characterizes human cranioskeletal development and reflects clinical presentation in FGFR syndromes. Plast Reconstr Surg 108:2026–2039.[ISI][Medline]
Chang SC, Liao YF, Hung LM, Tseng CS, Hsu JH, Chen JK (1999). Prefabricated implants or grafts with reverse models of three-dimensional mirror-image templates for reconstruction of craniofacial abnormalities. Plast Reconstr Surg 104:1413–1418.[ISI][Medline]
Chang SC, Chuang HL, Chen YR, Chen JK, Chung HY, Lu YL, et al. (2003a). Ex vivo gene therapy in autologous bone marrow stromal stem cells for tissue-engineered maxillofacial bone regeneration. Gene Ther 10:2013–2019.[ISI][Medline]
Chang SC, Wei FC, Chuang H, Chen YR, Chen JK, Lee KC, et al. (2003b). Ex vivo gene therapy in autologous critical-size craniofacial bone regeneration. Plast Reconstr Surg 112:1841–1850.[ISI][Medline]
Chen L, Adar R, Yang X, Monsonego EO, Li C, Hauschka PV, et al. (1999). Gly369Cys mutation in mouse FGFR3 causes achondroplasia by affecting both chondrogenesis and osteogenesis. J Clin Invest 104:1517–1525.[ISI][Medline]
Cho TJ, Gerstenfeld LC, Einhorn TA (2002). Differential temporal expression of members of the transforming growth factor beta superfamily during murine fracture healing. J Bone Miner Res 17:513–520.[ISI][Medline]
Cook SD, Baffes GC, Wolfe MW, Sampath TK, Rueger DC, Whitecloud TS 3rd (1994). The effect of recombinant human osteogenic protein-1 on healing of large segmental bone defects. J Bone Joint Surg Am 76:827–838.
Cook SD, Wolfe MW, Salkeld SL, Rueger DC (1995). Effect of recombinant human osteogenic protein-1 on healing of segmental defects in non-human primates. J Bone Joint Surg Am 77:734–750.
Crombleholme TM (2000). Adenoviral-mediated gene transfer in wound healing. Wound Repair Regen 8:460–472.[ISI][Medline]
Dahn RD, Fallon JF (2000). Interdigital regulation of digit identity and homeotic transformation by modulated BMP signaling. Science 289:438–441.
Danos O, Heard JM (1992). Recombinant retroviruses as tools for gene transfer to somatic cells. Bone Marrow Transplant 9(Suppl 1):131–138.
Ducy P, Zhang R, Geoffroy V, Ridall AL, Karsenty G (1997). Osf2/Cbfa1: a transcriptional activator of osteoblast differentiation. Cell 89:747–754.[ISI][Medline]
Einhorn TA (2003). Clinical applications of recombinant human BMPs: early experience and future development. J Bone Joint Surg Am 85(A Suppl 3):82–88.
Enneking WF, Mindell ER (1991). Observations on massive retrieved human allografts. J Bone Joint Surg Am 73:1123–1142.
Fallon JF, Lopez A, Ros MA, Savage MP, Olwin BB, Simandl BK (1994). FGF-2: apical ectodermal ridge growth signal for chick limb development. Science 264:104–107.
Fang J, Zhu YY, Smiley E, Bonadio J, Rouleau JP, Goldstein SA, et al. (1996). Stimulation of new bone formation by direct transfer of osteogenic plasmid genes. Proc Natl Acad Sci USA 93:5753–5758.
Franceschi RT (2003). Functional cooperativity between osteoblast transcription factors: evidence for the importance of subnuclear macromolecular complexes? Calcif Tissue Int 72:638–642.[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]
Franceschi RT, Yang S, Rutherford RB, Krebsbach PH, Zhao M, Wang D (2004). Gene therapy approaches for bone regeneration. Cells Tissues Organs 176:95–108.[ISI][Medline]
Francis PH, Richardson MK, Brickell PM, Tickle C (1994). Bone morphogenetic proteins and a signalling pathway that controls patterning in the developing chick limb. Development 120:209–218.[Abstract]
Gersbach CA, Byers BA, Pavlath GK, Garcia AJ (2004a). Runx2/Cbfa1 stimulates transdifferentiation of primary skeletal myoblasts into a mineralizing osteoblastic phenotype. Exp Cell Res 300:406–417.[ISI][Medline]
Gersbach CA, Byers BA, Pavlath GK, Guldberg RE, Garcia AJ (2004b). Runx2/Cbfa1-genetically engineered skeletal myoblasts mineralize collagen scaffolds in vitro. Biotechnol Bioeng 88:369–378.[ISI][Medline]
Gerstenfeld LC, Cullinane DM, Barnes GL, Graves DT, Einhorn TA (2003). Fracture healing as a post-natal developmental process: molecular, spatial, and temporal aspects of its regulation. J Cell Biochem 88:873–884.[ISI][Medline]
Gimble J, Guilak F (2003). Adipose-derived adult stem cells: isolation, characterization, and differentiation potential. Cytotherapy 5:362–369.[ISI][Medline]
Goodman FR (2002). Limb malformations and the human HOX genes. Am J Med Genet 112:256–265.[ISI][Medline]
Govender S, Csimma C, Genant HK, Valentin-Opran A, Amit Y, Arbel R, et al. (2002). Recombinant human bone morphogenetic protein-2 for treatment of open tibial fractures: a prospective, controlled, randomized study of four hundred and fifty patients. J Bone Joint Surg Am 84(A):2123–2134.
Groeneveld EH, Burger EH (2000). Bone morphogenetic proteins in human bone regeneration. Eur J Endocrinol 142:9–21.[Abstract]
Hacein-Bey-Abina S, von Kalle C, Schmidt M, Le Deist F, Wulffraat N, McIntyre E, et al. (2003). A serious adverse event after successful gene therapy for X-linked severe combined immunodeficiency. N Engl J Med 348:255–256.
Hanai J, Chen LF, Kanno T, Ohtani-Fujita N, Kim WY, Guo WH, et al. (1999). Interaction and functional cooperation of PEBP2/CBF with Smads. Synergistic induction of the immunoglobulin germline C alpha promoter. J Biol Chem 274:31577–31582.
Hartigan-O’Connor D, Amalfitano A, Chamberlain JS (1999). Improved production of gutted adenovirus in cells expressing adenovirus preterminal protein and DNA polymerase. J Virol 73:7835–7841.
Huang YC, Kaigler D, Rice KG, Krebsbach PH, Mooney DJ (2005). Combined angiogenic and osteogenic factor delivery enhances bone marrow stromal cell-driven bone regeneration. J Bone Miner Res 20:848–857.[ISI][Medline]
Israel DI, Nove J, Kerns KM, Kaufman RJ, Rosen V, Cox KA, et al. (1996). Heterodimeric bone morphogenetic proteins show enhanced activity in vitro and in vivo. Growth Factors 13:291–300.[ISI][Medline]
Jiang Y, Jahagirdar BN, Reinhardt RL, Schwartz RE, Keene CD, Ortiz-Gonzalez XR, et al. (2002). Pluripotency of mesenchymal stem cells derived from adult marrow. Nature 418:41–49.[Medline]
Jin QM, Anusaksathien O, Webb SA, Rutherford RB, Giannobile WV (2003). Gene therapy of bone morphogenetic protein for periodontal tissue engineering. J Periodontol 74:202–213.[ISI][Medline]
Jonk LJ, Itoh S, Heldin CH, ten Dijke P, Kruijer W (1998). Identification and functional characterization of a Smad binding element (SBE) in the JunB promoter that acts as a transforming growth factor-beta, activin, and bone morphogenetic prot
