Transient expression of type III collagen by odontoblasts: developmental changes in the distribution of pro-alpha 1(III) and pro-alpha 1(I) collagen mRNAs in dental tissues

Evolutionary developmental genetics of teeth and odontodes in jawed vertebrates: a perspective from the study of elasmobranchs

Most extant vertebrates display a high variety of tooth and tooth-like organs (odontodes) that vary in shape, position over the body and nature of composing tissues. The development of these structures is known to involve similar genetic cascades and teeth and odontodes are believed to share a common evolutionary history. Gene expression patterns have previously been compared between mammalian and teleost tooth development but we highlight how the comparative framework was not always properly defined to deal with different tooth types or tooth developmental stages. Larger-scale comparative analyses also included cartilaginous fishes: sharks display oral teeth and dermal scales for which the gene expression during development started to be investigated in the small-spotted catshark Scyliorhinus canicula during the past decade. We report several descriptive approaches to analyse the embryonic tooth and caudal scale gene expressions in S. canicula. We compare these expressions wih the ones reported in mouse molars and teleost oral and pharyngeal teeth and highlight contributions and biases that arise from these interspecific comparisons. We finally discuss the evolutionary processes that can explain the observed intra and interspecific similarities and divergences in the genetic cascades involved in tooth and odontode development in jawed vertebrates.

Unique and shared gene expression patterns in Atlantic salmon (Salmo salar) tooth development

To validate the use of Atlantic salmon (Salmo salar L.) as a model species in research on the mechanism of continuous tooth replacement, we have started to collect data on the molecular control underlying tooth formation in this species. This study reports expression patterns in the lower jaw dentition of a number of key regulatory genes such as bmp2, bmp4, and sox9 and structural genes such as col1alpha 1 and osteocalcin (= bgp, Bone Gla Protein) by means of in situ hybridization using salmon-specific, digoxygenin-labeled antisense riboprobes. We compare expression of these genes to that in other skeletogenic cells in the lower jaw (osteoblasts, chondroblasts, and chondrocytes). Our studies reveal both expression patterns that are in accordance to studies on mammalian tooth development and patterns that are specific to salmon, or teleosts. The epithelial expression of sox9 and a shift of the expression of bmp2 from epithelium to mesenchyme have also been observed during mammalian tooth development. Different from previous reports are the expressions of col1alpha 1 and osteocalcin. In contrast to what has been reported for zebrafish, osteocalcin is not expressed in odontoblasts, nor in the osteoblasts involved in the attachment of the teeth. At the lower jaw, osteocalcin is expressed in mature and/or resting osteoblasts only. As expected, col1alpha 1 is expressed in odontoblasts. Surprisingly, it is also strongly expressed in the inner dental epithelium, representing the first report of ameloblast involvement in collagen type I transcription. Whether the collagen is translated and secreted into the enameloid remains to be demonstrated.

Immunohistochemical identification of type I and type III collagen and chondroitin sulphate in human pre-dentine: a correlative FEI-SEM/TEM study

AIM

To identify type I- (I-CF) and type III-collagen fibrils (III-CF) and chondroitin 4/6 sulphate (CS) within human pre-dentine by means of a correlative analysis under field emission in-lens-scanning electron microscopy (FEI-SEM) and transmission electron microscopy (TEM).

METHODOLOGY

Human-extracted teeth were obtained and submitted to either a pre-embedding or a post-embedding immunolabelling procedure using monoclonal primary antibodies anti-I-CF, anti-III-CF and anti-CS. Gold-conjugated secondary antibodies were coupled to primary antibodies to visualize labelling under the electron beam. Correlative labelling patterns were obtained for I-CF and CS under both FEI-SEM and TEM.

RESULTS

Field emission in lens-SEM analysis revealed an intricate three-dimensional network of I-CF and CS clarifying the intimate relationship between the two main components of the pre-dentine organic matrix. TEM analysis revealed odontoblasts exhibiting intracellular labelling for CS, which became more intense and diffuse over the pre-dentine organic matrix. The same diffuse immunoreaction was revealed for I-CF, whereas a weak immunolocalization of III-CF was found scattered throughout the pre-dentine layer and over the collagen fibrils.

CONCLUSIONS

Both the pre- and post-embedding immunohistochemical approaches have led to the visualization of CF- and CS-labelling distribution within the pre-dentine layer, adding further knowledge on the elucidation of collagen-proteoglycans interaction in the organic matrix of human dental roots.

Expression of collagen XVIII and MMP-20 in developing teeth and odontogenic tumors

Collagen XVIII is a basement membrane (BM) component, whereas MMP-20 (enamelysin) is a matrix metalloproteinase predominantly expressed in teeth. Since MMP-20 was found to degrade collagen XVIII, we studied the co-expression of these proteins in dental tissues. Collagen XVIII surrounded the developing tooth during early and late bell stages and was also present in developing enamel. Western blotting indicated that developing enamel contains collagen XVIII N-terminal fragments of the frizzled variant. Enamelysin was co-localized with collagen XVIII in the developing enamel matrix and stratum intermedium. Electron microscope analysis showed that total mineral, calcium and phosphorus contents of enamel were slightly increased in collagen XVIII null mice but the analysis revealed no visible defects in the enamel or dentin structures. In odontogenic tumors MMP-20 and collagen XVIII were co-localized in the enamel-like tumor matrix. Our results show that collagen XVIII is present in developing teeth, but its absence seems not to be critical for the development of the teeth.

Changes in the localization of type I, III and IV collagen mRNAs in the kidneys of hereditary nephritic (ICGN) mice with renal fibrosis

Renal fibrotic change, extreme accumulation of extracellular matrix (ECM) components in glomeruli and tubulointerstitum, is one of the characteristic features of ICR-derived glomerulonephritis (ICGN) mice. Decreased degradation of ECMs by matrixmetalloproteinases was demonstrated in kidneys of ICGN mice. To determine the balance between production and degradation of ECMs in kidneys of ICGN mice, we examined expression of mRNAs of ECMs in those. To demonstrate the localization of type I, III and IV collagen mRNAs in kidney sections of ICGN and control ICR mice, in situ hybridization using digoxigenin-labeled oligonucleotide antisense probes for procollagen-alpha(1) (I), -alpha(1) (III) and -alpha(1) (IV) mRNAs, respectively, was performed. Negative or trace expressions of type I and III collagen mRNAs were observed in the kidneys of control mice, but stronger expressions of those were seen in glomeruli and injured renal tubules of the kidneys of ICGN mice. Moderate expression of type IV collagen mRNA was demonstrated in a part of glomeruli and renal tubules of both control and ICGN mice, and no remarkable difference was seen between them. Severe renal fibrosis, extreme accumulation of interstitial type I and III collagens is caused by increased production and decreased degradation in the kidneys of ICGN mice. Thus, the profiles of metabolism between interstitial and membranous collagens may be different in the kidneys of ICGN mice, and excessive production of interstitial collagens may be the dominant cause of renal disease in them.

Fibronectin accelerates the growth and differentiation of ameloblast lineage cells in vitro

During tooth development, the growth and differentiation of ameloblast lineage (AL) cells are regulated by epithelial-mesenchymal interactions. To examine the dynamic effects of components of the basement membrane, which is the extracellular matrix (ECM) lying between the epithelium and mesenchyme, we prepared AL cells from the epithelial layer sheet of mandibular incisors of postnatal day 7 rats and cultured them on plates coated with type IV collagen, laminin-1, or fibronectin. The growth of AL cells was supported by type IV collagen and fibronectin but not by laminin-1 in comparison with that on type I collagen as a reference. Clustering and differentiation of AL cells were observed on all matrices examined. AL cells showed normal growth and differentiation at low cell density on fibronectin but not on type I collagen. Furthermore, the population of cytokeratin 14-positive cells on fibronectin was lower than that on other ECM components, suggesting that fibronectin may be a modulator to accelerate the differentiation of AL cells. After the cells had been cultured for 9 days on fibronectin, crystal-like structures were observed. These structures overlaid the cell clusters and were positive for von Kossa staining. These findings indicate that each matrix component has a regulative role in the proliferation and differentiation of AL cells and that fibronectin causes the greatest acceleration of AL cell differentiation.

Human odontoblast culture method: the expression of collagen and matrix metalloproteinases (MMPs)

Studies on mature human odontoblasts have suffered for the lack of in vitro models. We recently introduced a human odontoblast and pulp tissue organ culture method, in which the odontoblasts are cultured in the pulp chamber after removal of the pulp tissue, and the pulp tissue can be cultured separately (Tjäderhane et al., 1998a). With this method, we have studied the effects of growth factors on the expression of collagen and extracellular matrix (ECM)-degrading enzymes, matrix metalloproteinases (MMPs), in mature human odontoblasts. TGF-beta 1 was selected because of its ability to regulate the response of the dentin-pulp complex to external irritation. The effect of TGF-beta 1 (10 ng/mL) on pro alpha 1(I) collagen mRNA was analyzed by quantitative PCR, and type I procollagen propeptide (PINP) was analyzed from conditioned culture media with RIA. Odontoblast media were also assayed for respective type III procollagen propeptide (PIIINP). TGF-beta had a negligible effect on collagen mRNA expression or protein synthesis, indicating that TGF-beta alone does not markedly induce dentin matrix formation per se in the human dentin-pulp complex (Palosaari et al., 2001). However, TGF-beta 1 seems to regulate MMP expression in mature human odontoblasts differentially. A strong down-regulation of MMP-8 (Palosaari et al., 2000), a modest down-regulation of MMP-20 (Tjäderhane et al., 2000), and considerable up-regulation of MMP-9, with no apparent effect on MMP-2 expression (Tjäderhane et al., 1998b), indicate that growth factors may affect the matrix synthesis by controlling the expression and activity of MMPs instead of collagen synthesis. The altered expression of MMPs may result in altered ECM formation, which in turn may contribute to the formation of atubular reparative dentin.

Baseline expression and effect of TGF-β1 on Type I and III collagen mRNA and protein synthesis in human odontoblasts and pulp cellsIn Vitro

Since growth factors have been suggested to regulate dentin collagen formation in response to external irritation, we investigated the effect of TGF-β1 on proα1(I) collagen mRNA expression in cultured mature human odontoblasts and pulpal fibroblasts, as well as cultured human pulp tissue, using quantitative PCR. Cultured gingival fibroblasts (GF) and osteoblasts (OB) served as controls. Also, type I collagen synthesis in cultured odontoblasts and pulp tissue, as well as type III collagen synthesis in odontoblasts, were studied by measuring respective procollagen (PINP and PIIINP) secretion into culture media with radioimmunoassay (RIA). Odontoblasts expressed significantly higher basic level of type I collagen mRNA than pulp tissue or pulp fibroblasts in culture, but markedly lower level than GF and OB cells. TGF-β1 (10 ng/ml) had negligible effects on type I collagen mRNA expression or PINP synthesis in cultured odontoblasts and pulp tissue, and PIIINP synthesis in the odontoblasts. In PF cells, the effect of TGF-β1 depended on culturing conditions; a 6-fold increase in mRNA expression was observed using serum-free medium but no effect was seen in the cells cultured with 10% FBS. In contrast, GF cells serving as controls were not markedly affected by the culture conditions, with 2-3-fold increase in mRNA expression by TGF-β1. These experiments demonstrate that mature human odontoblasts are capable of synthesizing type III collagen protein, and that TGF-β1 has negligible effect on mature human odontoblast and pulp tissue collagen expression.

Aberrant gene expression in epithelial cells of mixed odontogenic tumors

Comparative investigations of odontogenic cells in normally forming teeth and tumors may provide insights into the mechanisms of the differentiation process. The present study is devoted to late phenotypic markers of ameloblast and odontoblast cells, i.e., proteins involved in biomineralization. The in situ expression of amelogenins, keratins, collagens type III and IV, vimentin, fibronectin, osteonectin, and osteocalcin was performed on normal and tumor odontogenic human cells. The pattern of protein expression showed some similarities between ameloblasts and odontoblasts present in normally developing human teeth and cells present in neoplastic tissues of ameloblastic fibroma, ameloblastic fibro-odontomas, and complex odontomas. Amelogenins (for ameloblasts) and osteocalcin (for odontoblasts) were detected in cells with well-organized enamel and dentin, respectively. In contrast, "mixed" cells located in epithelial zones of mixed odontogenic tumors co-expressed amelogenins and osteocalcin, as shown by immunostaining. The presence of osteocalcin transcripts was also demonstrated by in situ hybridization in these cells. Keratins and vimentin were detected in the same epithelial zones. Tumor epithelial cells were associated with various amounts of polymorphic matrix (amelogenin- and osteocalcin-immunoreactive), depending on the types of mixed tumors. No osteocalcin labeling was found in epithelial tumors. This study confirms that the differentiation of normal and tumor odontogenic cells is accompanied by the expression of some common molecules. Furthermore, the gene products present in normal mesenchymal cells were also shown in odontogenic tumor epithelium. These data may be related to a tumor-specific overexpression of the corresponding genes transcribed at an undetectable level during normal development and/or to an epithelial-mesenchymal transition proposed to occur during normal root formation. A plausible explanation for the results is that the odontogenic tumor epithelial cells are recapitulating genetic programs expressed during normal odontogenesis, but the tumor cells demonstrate abnormal expression patterns for these genes.

Collagen mRNA expression during tissue development: the temporospacial order coordinates bone morphogenesis with collagen fiber formation

Bone formation of the maxilla and premaxilla of rats was studied by in situ hybridization, using probes for fibrillar collagen mRNAs. Chondroblasts, osteoblasts, fibroblasts and peripheral bone cells differed in their expression patterns. Prospective nasal chondroblasts expressed collagen alpha1(II) and alpha1(XI) RNA from day 15 post coitum. Bone formation in the adjacent maxilla and premaxilla started around day 17: groups of osteoblasts, representing ossification centers, expressed collagen alpha1(I) RNA strongly, and alpha1(V), alpha2(V) and alpha1(XI) RNA weakly, but they were deficient in collagen alpha1(III) RNA. As the centers expanded, osteoblasts in the resulting bone domains expressed collagen alpha1(I) RNA in abundance, whereas collagen alpha1(III) RNA was absent. Bone domains were surrounded by fibroblasts containing collagens alpha1(I), alpha1(III) and alpha2(V) RNA. Widely separated fibroblasts underwent condensation into densely packed periosteum and sutural soft tissues. Cells at the periphery of fast-growing bone domains also displayed, apart from collagen alpha1(I) RNA, collagens alpha2(V) and alpha1(XI) RNA. Given the continuous recruitment of cells from the periosteum, peripheral bone cells represent differentiating osteoblasts synthetizing collagens alpha2(V) and alpha1 (XI) RNA transiently. Thus, gene expression during osteoblast differentiation reflects synthesis of fiber components during bone growth, since collagen V is located in the center of fibers consisting primarily of collagen I.

Expression of type I and XII collagen during development of the periodontal ligament in the mouse

The purpose (of this study) was to determine the temporal and spatial pattern of type XII collagen expression during murine tooth/root development. Using in situ hybridization techniques, expression of type XII collagen was compared with that of type I collagen, the most abundant collagen in periodontal tissues. Mouse first mandibular molars were examined at the following developmental periods: pre-root formation, early root formation, initial alignment of the periodontal ligament (PDL) fibres, and PDL maturation as the tooth erupted and attained occlusal function. Transcripts for type I collagen were identified in bone cells and odontoblasts at all times but not in the dental follicle before root formation. As root formation progressed, type I collagen expression became apparent within cells of the dental follicle and forming PDL. During early stages of tooth development, signal for type XII collagen was not observed in any cells/tissues. Type XII collagen expression was first detected in the dental follicle/PDL region during tooth eruption and increased in the PDL as the molar tooth erupted into the mouth and achieved occlusal contact. These findings suggest that type XII expression is timed with the alignment and organization of PDL fibres and is limited in tooth development to cells within the periodontal ligament.

The genetic control of early tooth development

Most vertebrate organs begin their initial formation by a common, developmentally conserved pattern of inductive tissue interactions between two tissues. The developing tooth germ is a prototype for such inductive tissue interactions and provides a powerful experimental system for elucidation of the genetic pathways involved in organogenesis. Members of the Msx homeobox gene family are expressed at sites of epithelial-mesenchymal interaction during embryogenesis, including the tooth. The important role that Msx genes play in tooth development is exemplified by mice lacking Msx gene function. Msxl-deficient mice exhibit an arrest in tooth development at the bud stage, while Msx2-deficient mice exhibit late defects in tooth development. The co-expression of Msx, Bmp, Lefl, and Activin beta A genes and the coincidence of tooth phenotypes in the various knockout mice suggest that these genes reside within a common genetic pathway. Results summarized here indicate that Msxl is required for the transmission of Bmp4 expression from dental epithelium to mesenchyme and also for Lefl expression. In addition, we consider the role of other signaling molecules in the epithelial-mesenchymal interactions leading to tooth formation, the role that transcription factors such as Msx play in the propagation of inductive signals, and the role of extracellular matrix. Last, as a unifying mechanism to explain the disparate tooth phenotypes in Msxl- and Msx2-deficient mice, we propose that later steps in tooth morphogenesis molecularly resemble those in early tooth development.

Characterization of cellular responses involved in reparative dentinogenesis in rat molars

During primary dentin formation, differentiating primary odontoblasts secrete an organic matrix, consisting principally of type I collagen and non-collagenous proteins, that is capable of mineralizing at its distal front. In contrast to ameloblasts that form enamel and undergo programmed cell death, primary odontoblasts remain metabolically active in a functional tooth. When dentin is exposed to caries or by operative procedures, and when exposed dentinal tubules are treated with therapeutic dental materials, the original population of odontoblasts is often injured and destroyed. The characteristics of the replacement pool of cells that form reparative dentin and the biologic mechanisms that modulate the formation of this matrix are poorly understood. Based on the hypothesis that events governing primary dentinogenesis are reiterated during dentin repair, the present study was designed to test whether cells that form reparative dentin are odontoblast-like. Cervical cavities were prepared in rat first molars to generate reparative dentin, and animals were killed at various time intervals. In situ hybridization with gene-specific riboprobes for collagen types I and III was used to study de novo synthesis by cells at the injured dentin-pulp interface. Polyclonal antibodies raised against dentin sialoprotein (DSP), a dentin-specific protein that marks the odontoblast phenotype, were used in immunohistochemical experiments. Data from our temporal and spatial analyses indicated that cells forming reparative dentin synthesize type I but not type III collagen and are immunopositive for DSP. Our results suggest that cells that form reparative dentin are odontoblast-like.

Immunoelectron-microscopic study of the localization of fibronectin in the odontoblast layer of human teeth

Indirect immunofluorescence-based studies have shown similarities in the distribution patterns of fibronectin-positive fibrous structures and so-called von Korff fibres. The aim of the present study was to analyse the reactivity of fibronectin in the odontoblast layer of fully developed human teeth by means of immunoelectron microscopy. Between the odontoblasts, discrete and undulatory fibrillar fascicles with peroxidase labelling were observed. They seemed to be in contact with odontoblasts in some areas, while in others they appeared to be intervening between two neighbouring odontoblasts. Higher magnifications of the fibrillar material demonstrated axial periodic staining of about 70 nm. Peroxidase reaction of fibronectin was also recognized along the cell membrane of odontoblasts facing predentine. The fibronectin in fibrillar fascicles observed between odontoblasts would be held in place by the direct molecular interaction with collagen fibrils and contribute to the pulpward migration of these cells and maintenance of their specific morphology. At the distal end of odontoblasts, a tight seal would be maintained by means of odontoblast-fibronectin adhesion.

In situ hybridization of calbindin-D 28 k transcripts in undecalcified sections of the rat continuously erupting incisor

Calbindin-D9k and calbindin-D-28k genes are useful systems to investigate the tissue- and stage-specificity as well as the hormonal control of gene expression. Since they regulate cellular calcium mobilization, their study may be of interest in mineralized tissues. However, thus far, immunocytochemical labelling has been mainly realized in these systems. In order to set up methods for mRNA investigation, in situ hybridization of calbindin-D28k mRNAs was performed in the continuously erupting incisor of Sprague-Dawley rats (15-, 30-, and 56-day-old). 35S UTP labelled antisense and sense riboprobes specific for brain calbindin-D 28k were used for in situ hybridization. Specific and non-specific signals could not be discerned when studying decalcified samples. In contrast, on sections not pretreated with EDTA, calbindin-D 28k transcripts (in tooth and kidney) appeared strongly labelled with antisense probes, while sense probes provided a negligible background. In ameloblasts, the signal (i.e., calbindin-D 28k mRNA levels) increased during the presecretory stage. Different mRNA gradients and subcellular distribution patterns characterized the secretory and maturation stages. A nuclear labelling was observed, associated with the highest levels of transcripts. These data suggest a developmental control of calbindin-D28k mRNA transcription. Calbindin-D28k gene expression appears to be up-regulated during the initiation of both secretory and maturation stages of enamel mineralization.

Dentin matrix proteins and dentinogenesis

The precise mechanisms involved in dentinogenesis are not understood; however, the information to date suggests that a number of highly controlled extracellular events are involved. Mature odontoblasts secrete collagen at the cell border into predentin. They synthesize and secrete other non-collagenous proteins (NCPs) at the mineralization front, possibly through odontoblastic processes. A collagen-NCP complex is formed at the predentin-dentin border and apatite crystal initiation and growth takes place. One of the research needs is to uncover the nature of this dentin collagen-NCP complex and to understand how it controls mineralization. At least three dentin specific NCPs are known: phosphophoryn(s), dentin sialoprotein (DSP) and AG1 (Dmp1). Other macromolecules are commonly made by osteoblasts and odontoblasts and participate in bone and dentin formation. Some progress in understanding dentin mineralization has been gained by focusing upon the role of phosphophoryns. These highly phosphorylated proteins are secreted at the mineralization front, where a small portion binds in the gap region of type I collagen fibrils. This portion of phosphoproteins probably initiates formation of plate-like apatite crystals. Additional phosphoryns in higher concentrations bind to the growing apatite crystals and slow their growth, possibly influencing their size and shape. Other areas which need careful investigations are those involving the mechanisms involved in odontoblast differentiation, how the synthesis of the dentin specific NCPs is controlled and the precise roles of these macromolecules in dentinogenesis. Future experimentation will focus on the gene structures for these NCPs and the mechanisms of tissue specific gene regulation.(ABSTRACT TRUNCATED AT 250 WORDS)

Type III collagen is a major component of interodontoblastic fibers of the developing mouse molar root

BACKGROUND

Recently, collagenous interodontoblastic fibers (IOF) were reported in some particular developmental stages and/or locations of the tooth. However, it remained unclear whether these fibers were identical to so-called von Korff fibers.

METHODS

To clarify this issue, we examined the developing mouse molar by three-dimensional reconstruction of 8 confocal images within a 6 micron-thick section using laser scanning confocal microscopy, and confirmed our findings using immunoelectron microscopy.

RESULTS

In the root pulp during circumpulpal dentin formation, the IOF stained weakly for type I collagen, but stained strongly for type III collagen by a double-staining technique. It could be clearly seen that many immunoreactive fibers ran spirally among the odontoblasts and entered the predentin. This distribution pattern of IOF was similar to that of the classical von Korff fibers. Furthermore, the existence of anti-type III reactive collagen fibrils between odontoblasts was confirmed, whereas IOF were not observed in the coronal pulp during circumpulpal dentin formation.

CONCLUSIONS

This study presents for the first time, immunohistochemical observations which demonstrate the presence of IOF at least during root circumpulpal dentin formation and which reveal that type III collagen is a major component of IOF.

ED-A region-containing isoform of cellular fibronectin is present in dentin matrix in dentinogenesis imperfecta associated with osteogenesis imperfecta

To elucidate the defective dentin formation in osteogenesis imperfecta (OI), we analyzed the expression of selected fibronectin (FN) isoforms in the dentin matrix of a patient with dentinogenesis imperfecta (DI) associated with OI, and in normal teeth. Frozen tooth sections were immunostained with three monoclonal antibodies (MAbs). The MAb recognizing the major cell-binding region (f-33), shared by plasma FN (pFN) and cellular FN (cFN), stained the pulp of normal adult permanent teeth intensely, while no reactivity was present in predentin, (demineralized) dentin, or dental cementum. The periodontal ligament stained unevenly. The dentin matrix of the patient with OI displayed reactive zones, alternating layerwise or concentrically with non-reactive ones. Staining throughout the connective tissue of adult oral mucosa, analyzed for the form of FN present, was intense, and in dermis, which was also studied, it was moderate. Reactivities in dental tissues with the MAb specific for the ED-A region (IST-9), included in cFN but not pFN, were similar to those with MAb f-33. The mucosal connective tissue stained weakly and dermis was negative, except that nerves and endothelia of some large blood vessels stained clearly. The MAb specific for the ED-B segment (BC-1), also included in cFN only, did not stain any of the tissues analyzed. The results suggest that, unlike mucosal and dermal FNs, FNs in the dental tissues are largely cellular, and also that dentin formation in OI may be completed by successive generations of pulpal fibroblasts differentiated into hard-tissue-forming cells.