Periodontal regeneration in the perspective of new advances in the cell biology of human cementum

Introduction

In spite of substantial advances in the development of regenerative procedures, the regeneration of the periodontium is still unpredictable. In contrary, implant dentistry proved to have a relative high rate of predictable success. Thus, the clinician frequently faces the dilemma whether to replace periodontically compromised teeth that cannot be further treated by resective surgery by predictable implants, or to attempt rescuing them by considerably less predictable regenerative procedures.

In contrary to dentistry where dental tissues are being routinely replaced by artificial materials, in medicine the paradigm is that a diseased organ is replaced by an artificial one only when there is no doubt that the affected organ is beyond salvage or that it jeopardizes the general health and function of the patient. New concepts in reconstructive medicine and tissue engineering advocate the idea of regenerating specific affected tissues within the failing organs rather than the entire organ. For example, cartilage regeneration is proposed as an alternative to total or partial joint replacement by an artificial prosthesis. If to extrapolate from medicine to dentistry, the general paradigm should be to finding new approaches that will enable the regeneration of those specific tissues required for predictable periodontal regeneration.

In this context, cementum can be viewed as such a specific tissue. Cementum mediates the anchorage of the tooth to the periodontal tissues and thereby it enables the tooth function. Periodontal disease destroys the attachment of the periodontal Sharpey's fibers to cementum and exposes the outer layer of the cementum to the bacterial environment of the oral cavity. Thus, the regeneration of a functional cementum with inserted Sharpey's fibers is considered a prerequisite for predictable periodontal regeneration to occur.

The purpose of this review is to analyze the regenerative potential of cementum in the perspective of recent developments in its biology.

Biological events in the de novo formation of functional cementum

Periodontal regeneration implies the formation of new functional cementum on root surfaces previously exposed to the oral cavity. For the purpose of this review, functional cementum is determined as cementum firmly attached to the root surface on the one side and comprising Sharpey's fibers extending into the periodontal ligament on the opposite side. Our knowledge regarding the cellular and molecular events required for de novo formation of functional cementum on previously exposed root surfaces is extremely limited. However, based on evidence from wound healing studies in animals (^{Gould et al., 1980} ; ^{Lekic et al., 1996} ; ^{Boyko et al., 1981} ; ^{Nyman et al., 1982} ; ^{Pitaru et al., 1987} ; ^{Pitaru et al., 1988}) and guided tissue regeneration studies in humans (^{Gottlow et al., 1986}) as well as from general biological patterns of wound healing, it is proposed that the following basic biological events take place during the process of new cementogenesis :

- activation of the cementoblastic progenitor pool ;

- recruitment of cementoblastic progenitors to the root surface which implies directed cell migration towards the root surface and concomitant progenitors proliferation and differentiation into precementoblasts ;

- attachment of cementoblastic progenitors and/or precementoblasts to the root surface ;

- terminal differentiation of these cells into cementoblasts ;

- secretion of the cementum matrix onto the exposed root surface ;

- mineralization of the secreted matrix in order to form a stable and firm connection between the exposed root surface and the new formed cementum ;

- formation and insertion of new Sharpey's fibers into the new formed cementum.

Each of the above-mentioned events are supposed to require a myriad of cell-matrix and cell-cell interactions, which are regulated by insoluble and soluble mediators that activate signal transduction pathways whereby membranal and cytoplasmatic receptors.

A number of extensive reviews discuss new information regarding the molecular events and cellular events related to developmental cementogenesis (^{Schroeder, 1992} ; ^{Pitaru et al., 1994} ; ^{Chung and Amar, 1995} ; ^{Bosshardt and Schroeder, 1996} ; ^{Bosshardt and Selvig, 1997} ; ^{Saigin et al., 2000}). However, less attention has been paid to the adult cementoblastic lineage in general and to the human-derived adult cementoblastic lineage in particular. Since this lineage is supposed to be responsible for the regeneration of cementum during periodontal wound healing, the purpose of this review is to address questions related to location, origin and development of the cementoblastic lineage in the adult organism.

Questions

The issue of new cementogenesis in the context of periodontal regeneration in the adult poses a number of questions :

- where is the location of the cementoblastic progenitor pool in the adult periodontium ?

- is this pool derived from ancestor pluripotential cells that are also the source of other periodontal lineages ?

- are there specific markers for the cementoblastic lineage ?

- what is the mechanism that controls the recruitment of cementoblastic progenitors to the root surface ?

The above questions are further complicated by the fact that cementum is a complex tissue that includes two major subtypes : acellular extrinsic fiber cementum that contains the functional Sharpey's fibers. Thus, this subtype is viewed as the functional type of cementum ; and cellular intrinsic fiber cementum which lacks Sharpey's fibers and is associated frequently with cementum repair (^{Bosshardt and Selvig, 1997}).

Is it possible that each type of cementum is derived from two separate sublineages ? Do these lineages have a common cementoblastic progenitor and if so where is the divergence point ?

Most of the work that has addressed these questions utilized rodent models (reviewed by ^{Saigin et al., 2000}). Even though the homeostasis and repair of cementum in the adult might be similar in rodent and humans, recent studies indicate that during development the molecular and cellular biology of cementogenesis differ between the two species (reviewed by ^{Bosshardt and Schroeder, 1996}). More recent evidence demonstrates that mature human cementum and developing and mature bovine cementum contain a derived cementum attachment protein that is not found in rodents (^{Arzate et al., 1992a} and ^b ; ^{Saito et al., 2001} ; Pitaru, unpublished data). These studies further substantiate the possible differences between rodent and higher mammalian species. Since this review focuses on new cementogenesis in the perspective of periodontal regeneration in humans, whenever available, it will focus on data that is not derived from rodents.

Location of the cementoblastic progenitor pool

Activation of the progenitor pool is one of the first processes that have to occur during the wound healing process of any tissue. In order to trigger the stem cells and progenitors of the pool to proliferate and provide new cells for the regenerating tissues, it is important to identify its location. Identification of proliferating cells in situ requires the utilization of vital labelling with radioactive or chemical markers. This infers that such studies can be undertaken only on small animals, usually rodents. Thus, to date, virtually nothing is known on the location of progenitors pools in the human periodontium. Most of our knowledge is based on the work of ^{McCulloch and Melcher (1983a} ; ^1983b) and ^{McCulloch (1985)} performed in the eighties on mice. Kinetic studies by these investigators, in which proliferating cells were labelled with 3H-thymidine, indicate that cohorts of proliferating cells are located in the paravascular regions of the periodontal ligament of adult mice (^{McCulloch and Melcher, 1983a} ; ^{McCulloch and Melcher, 1983b}). These proliferating cells were shown to migrate towards the cementum, periodontal ligament tissue proper and the bone. In a latter study, ^{McCulloch (1985)} demonstrated that part of the 3H-thymidine labelled cells were slowly cycling, indicating that they belong to a possible stem cell population located within the 10 µm distance from the blood vessels. ^{Gould et al. (1980)} demonstrated that paravascular cells respond to the wounding of the periodontal ligament.

^{Melcher et al. (1986)} showed that stromal bone marrow cells are capable of producing a cementum-like material on denuded root surfaces in vitro. ^{McCulloch et al. (1987)} identified another progenitor pool in the endosteal spaces of the alveolar bone in mice that contributed to the periodontal ligament cell population. These two studies point to the possible existence of an additional progenitor pool that might contribute to the cementoblastic lineage. Thus, there is supportive evidence that, at least in rodents, there are two progenitor pools that are capable of contributing to the cementoblastic lineages.

^{Brunette et al. (1976)} located proliferating cells in the vicinity of blood vessels in the pig periodontal ligament and showed that these cells contribute to the expansion of periodontal ligament cell cultures, suggesting that the paravascular progenitor pool is not an exclusive property of rodent species. Furthermore, canine periodontal ligament cells grown in culture on denuded root surfaces and implanted into edentulous mandibles were capable of supporting the formation of new functional cementum (^{Boyko et al., 1981}). In spite of this circumstantial evidence the migration of cells from these pools has never been followed to the root surface and therefore these studies provide only supporting evidence to the belief that the progenitor pool for the cementoblastic lineage is located in the paravascular zone of the periodontal ligament and endosteal alveolar spaces. Moreover, there has been no evidence that similar pools are found in the human periodontium.

To determine whether the human periodontal ligament contains progenitors that can express the cementoblastic-like phenotype, our group cloned primary cultures derived from adult periodontal ligament tissue and found that part of these clones produced a mineralized-like tissue in culture (^{Pitaru et al., 1997} ; ^{Liu et al., 1997}). Some of these clones expressed markers characteristic to cells of the cementoblastic lineage (^{Pitaru et al., 2002} ; Bar-Kana and Pitaru, in preparation). These data and the fact that the periodontal tissue was obtained from extracted teeth and therefore consisted mainly of the tooth related part of the periodontal ligament strongly suggest that the human periodontal ligament comprises a cementoblastic progenitor pool. Whether the adult human periodontal ligament comprises a paravascular pool and whether the cementoblastic progenitor pool is part of this pool remains to be determined.

Recent studies (^{Grzesik et al., 1998} ; ^{Grzesik and Narayanan, 2002}) indicate that cementocytes obtained from premolars of young human donors retain their potential to proliferate in culture and produce a cementum like material when transplanted in vivo. These authors allude to the possibility that cellular cementum might retain a progenitor population that might be the target of future regenerative therapies. However, since cells in culture tend to dedifferenciate, it is doubtful whether these cementocytic populations are capable of proliferating in vivo and whether the ability of proliferating is maintained with aging.

Origin of the cementoblastic lineage

The adult periodontal ligament cell population is heterogeneous comprising fibroblastic lineages and the osteoblastic and cementoblastic lineages, the last two belonging to the mineralized tissue forming cell populations (^{McCulloch and Borodin, 1991} ; ^{Pitaru et al., 1994}). The question of interest is whether a pluripotential stem cell gives rise to the main cell lineages of the periodontium or there is a stem cell for each of the lineages or alternatively, there are two stem cell populations - one for the fibroblastic lineage and the second for the mineralized tissue forming cell lineages (cementoblastic and osteoblastic) (^{fig. 1}). The elucidation of this question has tremendous importance for designing new therapeutic models for periodontal regeneration in general and cementum regeneration in particular. As mentioned above, a prerequisite for tissue regeneration is the activation of the stem cells pool. If there is a specific cementoblastic pool, then it is important to determine what are the factors and mechanisms that specifically activate this pool. If the cementoblastic lineage is derived from a pluripotential pool, the activation of this pool in order to induce new cementogenesis implies the activation of the fibroblastic and osteoblastic lineages as well. In this case the target is to differentially induce the new young multipotential progenitors towards the cementoblastic lineage.

The lack of knowledge regarding the ontogeny of periodontal cell lineages in the adult has been presented 15 years ago (^{Melcher, 1988}). Since then very little progress has been made and there is very limited experimental evidence to support each of the above mentioned hypothesizes. The difficulty to identify stem cell populations in situ and particularly the lack of specific markers for the cementoblastic lineage have been the main causes for this absence of information.

However, recent studies in the biology of stem cells derived from adult organisms point to their considerable plasticity (^{Watt and Hogan, 2000}) supporting the idea that a pluripotential stem cell is the origin of the periodontal ligament cell populations. In an attempt to test this hypothesis our group (^{Liu et al., 1997} ; Bar-Kana and Pitaru, in preparation) undertook an in vitro approach in which primary cultures of adult periodontal ligament were cloned. This approach makes possible to generate cell colonies derived from one single cell. We observed that part of the clones that were growing more slowly could be extensively propagated, whereas the other ceased growing after a few cell doublings. Since stem cells are characterized by a slow cycle and an unlimited number of doublings (^{Potten and Loeffler, 1990}), we assumed that the first group of clones could represent a stem cell population of the periodontal ligament. When these clones were treated with dexamethasone, which is known to induce progenitors to differentiate into mineralized-tissue-forming (MTF) cells, part of them differentiated and produced a mineralized like-tissue in culture (^{Liu et al., 1997}). We termed these clones mineralized tissue-forming clones (MTF clones). The clones that did not produce mineralized-like tissue in culture could either represent a differentiated fibroblastic cloned population non-responsive to dexamathasone, or more conceivably, clones comprising very early progenitors that require more than dexamethasone in order to express the MTF phenotype. Indeed, when these dexamethasone-unresponsive clones are stimulated with bone morphogenetic protein 2 (BMP2), part of them differentiate into MTF clones (^{Pitaru et al., 2002}). Taken together, these data indicate that the periodontal ligament cell population comprises pluripotential progenitors, possible stem cells, which under the appropriate conditions can differentiate into MTF cells. However, it has been unclear whether the MTF clones comprise the cementoblastic phenotype. The answer to this question was made possible by the discovery of specific markers for the cementoblastic lineage.

Specific cementoblastic proteins

Examination of the biochemical profile of the osteoblastic and cementoblastic phenotypes reveals a very close resemblance between the two. Thus, the question is whether the cementoblastic phenotype differs from the osteoblastic one. The evidence accumulating during the past 10 years points towards a positive answer. Two specific cementum derived proteins have been identified : a cementum derived growth factor (CGF) (^{Nakae et al., 1991} ; ^{Yonemura et al., 1993a} ; ^{Yonemura et al., 1993b} ; ^{Ikezawa et al., 1997} ; ^{Ikezawa et al., 1998}) and a cementum derived attachment protein (CAP) (^{Olson et al., 1991} ; ^{Pitaru et al., 1992} ; ^{Arzate et al., 1992a} ; ^{Arzate et al., 1992b} ; ^{Arzate et al., 1998} ; ^{Narayanan et al., 1995} ; ^{Wu et al., 1996}). Both proteins could be purified only from cementum suggesting that they are characteristic for this tissue.

CGF is an insulin-like growth factor that was found to be a poor mitogen for fibroblastic cells. However, in the presence of epidermal growth factor or low concentrations of serum, its activity is considerable increased. CGF is mainly active in inducing the early stages of cell divisions. The periodontal ligament and cementum contain a variety of growth factors (^{Saigin et al., 2000} ; ^{Cochran and Wozney, 1999}). Thus, a relevant question would be what is the function of an additional growth factor. One possibility is that it might specifically stimulate the proliferation of the cementoblastic lineage. Alternatively, CGF could have been trapped within the cementum matrix during its development and its main function could be related to the growth of Hertwig's epithelial root sheath. One also must consider that some growth factors as for example factors of the transforming growth beta family have a differentiation effect rather than a mitogenic one (^{Hogan, 1996}). In this case CGF could act as differentiating factor for the cementoblastic lineage. Moreover, nothing is known about the cell type capable of secreting CGF during development and homeostasis. Thus, the source and function of this protein and whether it might be instrumental in new cementum formation remain to be elucidated.

CAP is a collagenous protein (^{Wu et al., 1996}) that supports cell attachment of periodontal cells (^{Olson et al., 1991}). CAP was found to bind with high avidity to hydroxyapatite and fibronectin, but not to collagen type I (^{Pitaru et al., 1992}). Anti-CAP monoclonal antibodies localized the protein to the cementum of human teeth and to some paravascular cells in the endosteum of the human alveolar bone (^{Arzate et al., 1992a} ; ^{Arzate et al., 1992b}). CAP has been first found to be expressed by a human cementoma derived cell line. These cells were alkaline phosphatase positive, expressed bone sialoprotein and collagens type I and V and produced a mineralized-like tissue in culture (^{Arzate et al., 1992a} ; ^{Arzate et al., 1992b}). Since these proteins are characteristic to the cementoblastic lineage the works of Arzate et al. were the first indication that CAP is specifically related to the cementoblastic lineage.

The interaction between extracellular matrix ligands and cells via membranal integrin receptors is important for activating signal pathways that support cell proliferation, differentiation and motility (^{Hynes, 1992}). CAP activates the MAP-kinase pathways and this effect is related to process of cell spreading (^{Yokokoji and Narayana, 2001}). The effect is probably mediated through a RGD sequence characteristic to attachment proteins (^{Hynes and Yamada, 1982}), as synthetic RGD peptides and antibodies against integrin components block CAP-mediated cell attachment (^{Pitaru et al., 1995} ; ^{Ivanovski et al., 1999}). Cementum contains other attachment proteins as osteopontin, bone sialoprotein and fibronectin (reviewed by ^{Grzesik and Narayanan, 2002}), which are found also in bone and other mineralized tissues. Thus, the existence of a new functional attachment protein in cementum suggests that CAP may play a specific role in the growth and development of the cementoblastic lineage. It was found that CAP has differential effect on the attachment and migration of human alveolar bone cells, periodontal ligament fibroblasts and gingival fibroblasts (^{Pitaru et al., 1995} ; ^{Metzger et al., 1998}). Bone cells bound CAP, attached to and migrated better onto CAP-coated root surfaces than periodontal ligament cells and these in turn did better than gingival fibroblasts (^{Pitaru et al., 1995} ; ^{Bar-Kana et al., 1998a} ; ^{Bar-Kana et al., 1998b}). Interestingly, CAP had no effect on the attachment of bone cells derived from rat and anti-CAP antibodies did not stain rat cementum. CAP was expressed in vitro by human periodontal ligament cells and to a lesser extend by human alveolar bone cells, whereas human gingival fibroblasts and human stromal bone marrow cells were CAP negative (^{Bar-Kana et al., 1998b}). Considering the fact that CAP was identified in developing cementum of bovine roots (^{Saito et al., 2001}), these results suggested that CAP may have a preferential effect on human and other higher mammalian-derived MTF cell lineages in general and on the cementoblastic lineage in particular and that it might serve as a marker for the cementoblastic cell population.

Is CAP a marker for the cementoblastic lineage ?

^{Liu et al. (1997)} found that MTF clones exhibit a high capacity to bind CAP relative to their parent cell population, that is, the primary periodontal ligament cultures from which they were derived. Clones that did not produce mineralized tissue in culture did not bind CAP or had a relative very low binding capacity. The non or low binding clones, which represented 60 % out of 1 280 progenitor clones obtained from periodontal ligament tissue of 10 donors were termed fibroblastic clones (Fb clones). A direct correlation could be established between the capacity of human periodontal ligament-derived clones to bind CAP and their ability to form mineralized tissue in culture - the higher the CAP-binding capacity of specific clone the higher its ability to form mineralized tissue. MTF clones with a high capacity to bind CAP exhibited two patterns of mineralization. Fifteen percent of the clones produced mineralized tissue having a ridge-like appearance that was confined to a restricted area of the culture dish (^{fig. 2a}). This pattern of mineralization was identical to that exhibited by the cementoma-derived cell line mentioned above. These clones were termed MTF1 clones. The remaining 85 % of the clones (MTF2) formed mineralized nodules that were spread evenly in the culture dish (^{fig. 2b}). This pattern of mineralized tissue formation in culture is characteristic to the osteoblastic lineage derived either from the rat calvaria (^{Bellows et al., 1990}) or the bone marrow of rodents and humans (^{Maniatopoulos et al., 1988} ; ^{Pitaru et al., 1993} ; ^{Pri-Chen et al., 1998}). When these two MTF cloned populations, cementoma cells, human alveolar bone cells, Fb clones and human gingival fibroblast derived clones were tested for their capacity to express alkaline phosphatase it was found that the MTF1 clones and the cementoma cells expressed low levels of CAP, the MTF2 clones and the bone cells had a high alkaline phosphatase activity whereas the gingival clones and the Fb clones were alkaline phosphatase negative. Collectively these data indicated that CAP binding can serve as a marker for selecting MTF clones from the fibroblastic clones at early stages of the culture and demonstrated that the MTF clones are heterogeneous.

Differential expression of alkaline phosphatase in the periodontium in vivo was studied only in the rat model. It was found that cementoblasts have a lower expression than osteoblasts (^{Tenorio et al., 1993}) and that cementoblasts of the cellular cementum are more active than those of the acellular cementum. Even though these data was not obtained from humans, it points to the possibility that the MTF1 clones represent the cementoblastic lineage, and the MTF2 clones represent the osteoblastic one. Considering the possibility that two sublineages - for acellular and cellular cementum - can diverge from a main cementoblastic lineage, then one may suspect that the MTF1 clones stand for the cell lineage forming acellular cementum whereas the MTF2 clones stand for that forming cellular cementum and/or for the osteoblastic lineage. The fact that MTF1 clones express higher levels of CAP in culture than the MTF2 ones further points to the differences between these two cloned populations.

Is there a pluripotential stem cell ?

The constitutive phenotypic expression of the MTF clones points to the existence of early progenitors of the cementoblastic and osteoblastic lineages that can be stimulated to differentiate in culture by dexamethasone which is known to induce osteoblastic differentiation in various bone cell cultures. This together with the existence of Fb clones that are non-responsive to dexamethasone would raise the suspicion that there are two stem cells pools - one for the fibroblastic lineage and the other for the mineralized tissue forming lineages. However, a recent study done in our laboratory demonstrates that when the Fb clones are stimulated with BMP2, 75 % of them undertake the MTF pathway of differentiation as reflected by their capacity to express CAP, alkaline phosphatase, bone sialoprotein and form mineralized tissue (^{Pitaru et al., 2002}). This finding is similar to that observed in the bone marrow system where BMP2 stimulates stem cells to prefer the osteoblastic pathways of differentiation over the fibroblastic or adipocytic ones (^{Gazit et al., 1999}). Thus, one may suspect that part of the Fb clones that responded to BMP2 stimulation represented pluripotential progenitor clones derived from a stem cell population. The migration of cells from the endosteal spaces into the periodontal ligament and the findings that alveolar bone cells in culture express CAP and stromal bone marrow cells are pluripotential raise the question whether the real stem cell pool of the adult periodontium is not located within the endosteal spaces of the alveolar bone.

If this might be the case, then paravascular pool serves only as reservoir for the development of the periodontal cell lineages during homeostasis and wound healing. It might also be that the environmental cues established during the development of the periodontium serve to direct the directed migration and differentiation of these lineages. Since during development cementogenesis and Sharpey's fiber formation is strongly related to the function of the developmental Hertwig's epithelial root sheath (^{McNail and Thomas, 1993}), one may wonder whether the network of epithelial rests of Malassez might serve as one of these cues.

Since during wound healing the stem and progenitor cells migrate through a provisional matrix (^{Grzesik and Narayanan, 2002}) that lack the homeostatic cues of the normal tissue, their path-way of differentiation may be altered. Programmed spatial and temporal delivering of cell lineage specific proteins, such as CGF and CAP may be instrumental in achieving predictable regeneration of the periodontium in general and of cementum in particular. The recent work of ^{Bar-Kana et al. (2000)}, showing that CAP enriches the cementoblastic population on root surfaces in vitro, is the first attempt to test this concept.

BIBLIOGRAPHIE

• Arzate H, Olson SW, Page RC, Gown AM, Narayanan AS. Production of a monoclonal antibody to an attachment protein derived from human cementum. FASEB J 1992a;6:2990-2995.
• Arzate H, Olson SW, Page RC, Narayanan AS. Isolation of tumor cells that produce cementum protein in culture. Bone Miner 1992b;18:15-30.
• Arzate H, Alveraz-Perez MA, Aguilar-Mendoza ME, Alvarez-Fregoso O. Human cementum tumor cells have different features from human osteoblastic cells in vitro. J Periodont Res 1998;33:249-258.
• Bar-Kana I, Savion N, Narayanan AS, Pitaru S. Cementum attachment protein manifestation is restricted to the mineralized tissue forming cells of the periodontium. Eur J Oral Sci 1998a;106:357-364.
• Bar-Kana I, Savion N, Narayanan AS, Pitaru S. Cementum attachment protein identifies cementoblastic progenitors in the adult human periodontium. J Dent Res 1998b;77:929(abstract).
• Bar-Kana I, Narayanan AS, Grosskopf A, Savion N, Pitaru S. Cementum attachment protein enriches putative cementoblastic populations on root surfaces. J Dent Res 2000;79:1482-1488.
• Bellows CG, Heersche JNM, Aubin JE. Determination of the capacity for proliferation and differentiation of osteoprogenitor cells in the presence and absence of dexamethasone. Dev Biol 1990;140:132-138.
• Bosshardt DD, Selvig KA. Dental cementum : the dynamic tissue covering the root. Periodontol 2000 1997;13:41-75.
• Bosshardt DD, Schroeder HE. Cementogenesis reviewed : a comparison between human premolars and rodent molars. Anat Rec 1996;245:267-292.
• Boyko G, Melcher AH, Brunette DM. Formation of new periodontal ligament by periodontal ligament cells implanted in vivo after culture in vitro. J Periodont Res 1981;16:73-88.
• Brunette DM, Melcher AH, Moe HK. Culture and origin of epithelium-like and fibroblastic-like cells from porcine periodontal ligament explants and cell suspensions. Arch Oral Biol 1976;21:393-400.
• Chung KM, Amar S. Molecular and cellular biology research in periodontal regeneration and its clinical implication. Ann Acad Med Singapore 1995;24:58-67.
• Cochran DL, Wozney JM. Biological mediators for periodontal regeneration. Periodontol 2000 1999;19:40-58.
• Gazit D, Jurgeman G, Kelley P, Wang E, Jalenak M, Zilberman Y et al. Engineered pluripotent mesenchymal cells integrate and differentiate in regenerating bone : a novel cell-mediated gene therapy. J Gene Med 1999;1:121-133.
• Gould TRL, Melcher AH, Brunette DM. Migration and division of progenitor cell populations in periodontal ligament after wounding. J Periodont Res 1980;15:20-42.
• Gottlow J, Nymans, Karring T, Lindhe J. New attachment formation in the human periodontium by guided tissue regeneration. Case reports. J Clin Periodontol 1986;13:604-616.
• Grzesik WJ, Kuzentsov SA, Uzawa K, Mankani M, Robey PG, Yamauchi M. Normal human cementum-derived cells : isolation, clonal expansion, and in vitro and in vivo characterization. J Bone Min Res 1998;13:1547-1554.
• Grzesik WJ, Narayanan AS. Cementum and periodontal wound healing and regeneration. Crit Rev Oral Biol Med 2002;13:474-484.
• Hogan BLM. Bone morphogenetic proteins : multifunctional regulators of vertebrate development. Genes Develop 1996;10:1580-1594.
• Hynes RO. Integrin : versatility, modulation and signaling in cell adhesion. Cell 1992;69:11-25.
• Hynes RO, Yamada KM. Fibronectins : multifunctional modular glycoproteins. J Cell Biol 1982;95:369-377.
• Ikezawa K, Hart CE, Williams DS, Narayanan AS. Characterization of a cementum derived growth factor as an insulin-like growth factor-I like molecule. Connect Tissue Res 1997;36:309-319.
• Ikezawa K, Ohtsubo M, Norwood TH, Narayanan AS. Role of cycline and cycline-dependent kinase in mitogenic stimulation of by cementum-derived growth factors in human fibroblasts. FASEB J 1998;12:1233-1239.
• Ivanovski S, Komaki M, Bartold PM, Narayanan AS. Periodontal-derived cells attach to cementum attachment protein via alpha 5 beta 1 integrin. J Periodont Res 1999;34:154-159.
• Lekic P, Sodek J, McCulloch CAG. Relationship of cellular proliferation to expression of osteopontin and bone sialoprotein in regeneration rat periodontium. Cell Tissue Res 1996;28:491-500.
• Liu HW, Yacobi R, Savion N, Narayanan AS, Pitaru S. A collagenous cementum-derived attachment protein is a marker for progenitors of the mineralized tissue-forming cell lineage of the periodontal ligament. J Bone Miner Res 1997;12:1691-1699.
• Maniatopoulos C, Sodek J, Melcher AH. Bone formation in vitro by stromal cells obtained from bone marrow of young adult rats. Cell Tissue Res 1988;254:317-330.
• McCulloch CAG. Progenitor cell populations in the periodontal ligament of mice. Anat Rec 1985;211:258-262.
• McCulloch CAG, Melcher AH. Continuous labeling of the periodontal ligament of mice. J Periodont Res 1983a;18:231-241.
• McCulloch CAG, Melcher AH. Cell density and cell generation in the periodontal ligament of mice. Am J Anat 1983b;167:43-58.
• McCulloch CAG, Nemeth E, Lowenberg B, Melcher AH. Paravascular cells in endosteal spaces of alveolar bone contribute to periodontal ligament cell populations. Anat Rec 1987;219:233-242.
• McCulloch CAG, Borodin S. Role of fibroblast subpopulation in periodontal physiology and pathology. J Periodont Res 1991;26:144-154.
• McNeil RL, Thomas HF. Development of the murine periodontium. II. Role of the epithelial root sheath information of the periodontal attachment. J Periodontol 1993;64:285-291.
• Melcher AH. Does the developmental origin of cementum periodontal ligament and bone predetermine their behavior in the adults ? Zurich : Periodontology Today International Congress, 1988:6-14.
• Melcher AH, Cheong T, Cox J, Nemeth E, Shiga A. Synthesis of cementum-like tissue in vitro by cells cultured from bone : a light and electron microscopic and electron microscopic study. J Periodont Res 1986;21:592-612.
• Metzger Z, Weinstock B, Dotan M, Narayanan AS, Pitaru S. Differential chemotactic effect of cementum attachment protein on periodontal cells. J Periodont Res 1998;33:126-129.
• Nakae H, Narayanan AS, Raines E, Page RC. Isolation and partial characterization of mitogenic factors from cementum. Biochemistry 1991;30:7047-7052.
• Narayanan AS, Ikezawa K, Wu D, Pitaru S. Cementum specific components which influence periodontal connective tissue cells. Connect Tissue Res 1995;33:341-343.
• Nyman S, Gottlow J, Karring T, Lindhe J. The regenerative potential of the periodontal ligament. An experimental study in monkey. J Clin Periodontol 1982;9:146-148.
• Olson S, Arzate H, Narayanan AS, Page RC. Cell attachment activity of cementum proteins and mechanism of endotoxin inhibition. J Dent Res 1991;70:1272-1277.
• Pitaru S, Tal H, Soldinger M, Grosskopf A, Noff M. Collagen membranes prevent the apical migration of epithelium during periodontal wound healing. J Periodont Res 1987;22:331-333.
• Pitaru S, Tal H, Soldinger M, Grosskopf A, Noff M. Partial regeneration of collagen tissues using collagen barriers : initial observation in canine. J Periodontol 1988;59:458-467.
• Pitaru S, Savion N, Hekmati H, Olson S, Narayanan AS. Binding of a cementum attachment protein to extracellular matrix components and to dental surfaces. J Periodont Res 1992;27:640-646.
• Pitaru S, Kotev-Emeth S, Noff D, Kapfulerr S, Savion N. Basic fibroblast growth factor stimulates mineralized bone-like tissue formation in vitro by enhancing the proliferation, differentiation and protein synthesis of stromal bone marrow cells. J Bone Miner Res 1993;8:919-929.
• Pitaru S, McCulloch CAG, Narayanan SA. Cellular origin and differentiation control mechanisms during periodontal development and wound healing. J Periodont Res 1994;29:81-94.
• Pitaru S, Narayanan AS, Olson S, Savion N, Hekmati H, Alt I et al. Specific cementum attachment protein enhances selectively the attachment and migration of periodontal cells to root surfaces. J Periodont Res 1995;30:360-368.
• Pitaru S, Narayanan AS, Kotev-Emeth S, Liu HW, Savion N. The effect of age on the expression of mineralized tissue progenitors in the periodontium - The effect of bFGF. J Periodont Res 1997;32:179-182.
• Pitaru S, Pritzki A, Bar-Kana I, Grosskofp A, Savion N, Narayanan AS et al. Bone morphogenetic protein 2 induces the expression of cementum attachment protein in human periodontal ligament clones. Connect Tiss Res 2002;43:257-264.
• Potten CS, Loeffler M. Stem cells : attributes, cycles, spirals, pitfalls and uncertainties. Lessons for and from the crypt. Development 1990;110:1001-1020.
• Pri-Chen S, Pitaru S, Lokiec F, Savion N. Basic fibroblast growth factor enhances the growth and expression of the osteogenic phenotype of dexamethasone-treated human bone marrow derived bone-like cells in culture. Bone 1998;23:111-117.
• Saygin NE, Gianobille WV, Somerman MJ. Molecular and cell biology of cementum. Periodontol 2000 2000;24:73-98.
• Saito M, Iwase M, Maslan S, Nozaki N, Yamauchi M, Handa K et al. Expression of cementum derived attachment protein in bovine tooth germ during cementogenesis. Bone 2001;29:242-248.
• Schroeder HE. Biological problems of regenerative cementogenesis : synthesis and attachment of collagenous matrices on growing and established root surfaces. Int Rev Cytol 1992;141:1-59.
• Tenorio D, Crushley A, Hughes FJ. Immunocytochemical investigation of the rat cementoblast phenotype. J Periodont Res 1993;28:411-419.
• Watt FM, Hogan BLM. Out of Eden : stem cells and their niches. Science 2000;287:1427-1433.
• Wu D, Ikezawa K, Parker T, Saito M, Narayanan AS. Characterization of a collagenous cementum-derived attachment protein. J Bone Miner Res 1996;11:686-692.
• Yokokoji T, Narayanan AS. Role of D1 and E cyclins in cell cycle progression of human fibroblasts adhering to cementum attachment protein. J Bone Min Res 2001;16:1062-1067.
• Yonemura K, Narayanan AS, Miki Y, Page RC, Okada H. Isolation and partial characterization of a growth factor from human cementum. Bone and Mineral 1993a;18:187-198.
• Yonemura K, Raines EW, Ahn NG, Narayanan AS. Mitogenic signaling mechanisms of human cementum-derived growth factor. J Biol Chem 1993b;35:26120-26126.