Journal of Endodontics Research - http://endodonticsjournal.com
Location, arrangement and possible function of interodontoblastic collagen fibres in association with calcium hydroxide-induced hard tissue bridges
http://endodonticsjournal.com/articles/126/1/Location-arrangement-and-possible-function-of-interodontoblastic-collagen-fibres-in-association-with-calcium-hydroxide-induced-hard-tissue-bridges/Page1.html
By JofER editor
Published on 12/3/2008
 
Y. Kitasako, S. Shibata, C. F. Cox & J. Tagami
Cariology and Operative Dentistry, Department of Restorative Sciences and Maxilofacial Anatomy, Department of Maxillofacial/ Neck Reconstruction Maxillofacial Biology, Graduate School, Tokyo Medical and Dental University, Tokyo, Japan.
Departments of Restorative Dentistry and Biomaterials, University of Alabama, Birmingham, AL, USA.


Aim.
To assess the location, arrangement and possible function of interodontoblastic collagen fibres in association with calcium hydroxide-induced hard tissue bridges by using light and transmission electron microscopy techniques and immunohistochemical staining localization.

Conclusions.
Interodontoblastic collagen fibres were routinely detected throughout early dentine bridges. Interodontoblastic collagen fibres are thought to be important for initial dentine bridging to induce and support a dentinogenesis framework.

Introduction - Materials and methods.
Y. Kitasako, S. Shibata, C. F. Cox & J. Tagami
Cariology and Operative Dentistry, Department of Restorative Sciences and Maxilofacial Anatomy, Department of Maxillofacial/ Neck Reconstruction Maxillofacial Biology, Graduate School, Tokyo Medical and Dental University, Tokyo, Japan.
Departments of Restorative Dentistry and Biomaterials, University of Alabama, Birmingham, AL, USA.


Introduction.
A biological and/or mechanically tight bacteriometic seal is essential for an orderly progression of cell proliferation, contact recognition against the agent interface and eventual bridge formation between the restoration interface and the capping agent (Cox et al. 1982, Tziafas et al.1992b, Kitasako et al. 2000a). Fitzgerald (1979) followed cellular reorganization in mechanically exposed and directly capped monkey pulps. He confirmed that dentine bridge formation occurred when the capping agent was placed indirect contact with the pulp, the cells following a pattern of cell proliferation, stratification and polarization to form new odontoblastoid cells and matrix formation, without being preceded by a zone of necrosis. Routine use of Ca(OH)2 and certain other ‘bacteriometic’ agents permits cell organization and differentiation of new odontoblastoid cell formation directly adjacent to the agent interface (Cox et al.1982,1987).
In von Korff’s original paper (1906), an argyrophilic interodontoblastic fibre (von Korff fibre) arose from the dental pulp, passing between the odontoblasts and fanning out to form the fibrous matrix of the dentine. In general, von Korff fibres may help to guide the odontoblasts in their pulp ward migration during dentinogenesis and to bind the soft tissue pulp and odontoblasts to the dentine. Lester & Boyde (1968) have cited evidence that von Korff fibres are present during secondary dentinogenesis. The micrographs of Silva & Kalis (1972) and Reith (1968) also suggested that the fibres might be present in early circumpulpal dentine formation, whilst Furseth (1971) reported that they might be present later in the roots of human teeth.
In his original paper, von Korff (1906) described argyrophilic von Korff fibres as radiating between the primary odontoblasts and fanning out near their apical neck to take part in the formation of the collagen matrix of the developing intertubular substrate. Silva & Kalis (1972) suggested that von Korff fibres were present in early circumpulpal dentine formation. Hayward & Webb (1984) demonstrated collagen bundles coursing from the pulp into the predentine substrate suggesting that von Korff fibres are present in all stages of dentinogenesis.
Recent transmission electron microscopy and scanning electron microscopic studies show the existence of interodontoblastic collagen fibres in human dentine (Sogaard-Pedersen et al.1990), in the root of cat dentine (Bishop et al.1991) and in mechanically exposed monkey dental pulp (Kitasako et al. 2000a). A parallel study by Salmon et al. (1991) demonstrated bundles of fibrils that course between odontoblasts through the predentine to become incorporated into the mineralized dentine. Following experimental pulpotomy capping with a Ca(OH)2 slurry in dogs, Higashi & Okamoto (1996)postulated that these interodontoblastic fibrils provide support to the cells before their formation of odontoblasts.
However, some controversy remains over these fibres as Ten Cate (1998) suggests the term von Korff fibre to be incorrect. Consequently, the current mechanical or biological nature of von Korff fibres or von Korff-like fibres (VKF) remains unclear; however, due to their frequent occurrence, they appear an important source of collagen for dentine substrate. The aim of this study was to assess the location, arrangement and possible function of interodontoblastic collagen fibres in association with calciumhydroxide-induced hard tissue bridges by using light and transmission electron microscopy techniques and immunohistochemical staining localization.

Materials and methods.

Animals.
Two monkeys (Macaca fuscata), both aged 7 years and showing no evidence of dental or periodontal disease, were housed in facilities approved by the Tokyo Medical and Dental University. The animal use protocol form was reviewed and approved by the Screening Committee for Animal Research of the Tokyo Medical and Dental University, prior to the study. Because five different observation periods (3, 14, 21, 30 and 90 days) were used, two primates were divided into two groups: one was tested at 3, 21and 90 days and the other one was tested at 3, 14 and 30 days after operation, respectively (Table 1).

Table 1. Numbers of tested teeth for each animal.

Numbers of tested teeth for each animal

Experimental procedures.
Each monkey was subjected to general anaesthesia by an intramuscular injection of 20 mg kg_1 ketamine (Ketaral; Sankyo Co., Tokyo, Japan), followed by an intravenous injection of 10 mg kg_1 pentobarbital sodium (Nembutal; Sodium Solution, Abbott Laboratories, Abbott Park, IL, USA). Class V cavities were prepared in descending order of the length of observation periods (10 teeth per time period) with a high-speed tapered diamond bur (number 170; GC Co., Tokyo, Japan) under water-spray coolant. Each pulp was intentionally exposed on the cavity floor with a round carbide bur (number1; Shofu Inc., Kyoto, Japan), 0.8 mm in diameter. Infiltration anaesthesia was performed in the adjacent gingiva of the exposed pulps with lidocaine hydrochloride containing 1: 80000 adrenaline (Xylocaine; Fujisawa Co., Osaka, Japan) to control haemorrhage and exudate from the exposure site. Each exposed pulp was immediately capped with a hard-set Ca(OH)2 (Dycal; L. D. Caulk Co., Milford, DE, USA). The base and catalyst were mixed and applied directly onto the exposed pulps. Effort was made to keep all dentine and cavity walls free of the Ca(OH)2 agent. After the Ca(OH)2 had set, each cavity was then sealed with an adhesive resin system (Clearl Liner Bond II V; Kuraray Co. Ltd, Osaka, Japan). Each cavity was first conditioned with LB Primer for 30 s, air-dried and then coated with LB Bond and photo-cured for 20 s. Protect Liner F, a low-viscosity resin composite, was then applied to the cavity floor, covering the Ca(OH)2 agent and photo-cured for 20 s. All cavities were restored to the cavo surface margin with a hybrid resin composite (Clearl AP-X; Kuraray Co. Ltd, Osaka, Japan) and photo-cured for 30 s.

Tissue preparation.
At the end of the experimental observation periods, both monkeys were killed with an intravenous injection of 250 mg kg_1 thiopental sodium (Ravonal, Tanabe Pharmaceutical Co., Osaka, Japan). The teeth were immediately removed by surgical extraction and immersed in 4% paraformaldehyde (PA) solution (0.1m phosphate buffer, pH 7.4) for light microscopy or 5%glutaraldehyde (GA)_4% paraformaldehyde (PA) solution (0.1m phosphate buffer, pH 7.4) for transmission electron microscopy at room temperature for 2 days. Twenty-five PA-fixed specimens were demineralized in10%EDTA (Ethylenediamine- N,N,N0,N0-tetraacetic acid) at 4 8C for 2 weeks and embedded in parafin. Twenty GA-PA-fixed specimens were demineralized with EDTA, if required, and post-fixed in1% osmium tetroxide (0.1m phosphate buffer, pH 7.4, 4 8C) for 3 h, rinsed and embedded in Rigolac resin (Oken,Tokyo, Japan).

Silver staining and immunohistochemistry.
Serial parafin sections (4 mm) were cut and the tissues silver impregnated by Gomori’s reticulin method (Gomori1937) to identify collagen fibres. Immunohistochemistry was performed as previously described (Shibata et al. 1997). Briey, after deparafinization, the sections were immersed in PBS and digested with testicular hyaluronidase (25 mg mL_1; Sigma Chemicals) in PBS for 30 min at 37 8C. After several washings in PBS, the sections were immersed in methanol containing 1% hydrogen peroxide and 1% bovine serum albumin to block endogenous peroxidase activity and nonspecific reactions. Sections were then incubated with commercial rabbit polyclonal antibodies (LSL, Tokyo, Japan) against collagen type I (antibovine type I collagen), III (antiratmultiple antigen peptide), or fibronectin (antirat multiple antigen peptide), diluted 1:1000 with PBS. The streptavidin-biotin method was applied to the sections using HISTOFINE SAB kits (Nichirei, Tokyo, Japan). The sections were treated with 3-amino-9-ethylcarbazole to reveal any reaction. Negative control sections for immunohistochemistry were incubated with normal rabbit IgG instead of the primary antibodies. Sections were observed after counterstaining with haematoxylin.

Transmission electron microscopy.
One micrometer mm resin sections were cut and stained with toluidine blue for light microscopy to confirm the orientation. Ultrathin sections were then cut and contrasted with phosphotungstic acid, uranyl acetate and lead citrate and examined in a Hitachi H-800 transmission electron microscope.

Results.
Light microscopy with silver staining.
At 3 days, classic argyrophilic von Korff fibres were observed at some distance from the exposure area (Fig.1 A,B).At14 days, thick darkgreyish-purple stained VKF were observed extending from the original dentine, passing through the odontoblasts and projecting out to the central pulp fibres (Fig.1 C, D). The VKF were different from the classic von Korff fibres in their thickness as well as their silver staining intensity. At 21days, VKF extended from the expanded predentine (Fig. 1 E, F). At 30 days, the exposure site was almost completely occluded with a thin layer of a new dentine bridge (Fig.1G). At some distance from the exposure area, argyrophilic VKF were observed with in the new dentine bridge, reaching the material interface (Fig.1 H). At the periphery of the exposure, VKF were observed embedded within the dentine bridge (Fig.1 I). New tubular dentine without VKF was seen following initial dentine bridge formation. VKF were seen extending perpendicularly from the remaining dentine chips (Fig.1 J) at the exposure site. In the absence of dentine chips at the surface of the exposure area, VKF were randomly arranged (Fig.1K). In this area, some defects were observed with in the dentine bridge. A new dentine bridge generally occluded the exposure with increasing time.

Figure 1. (A-K) LM observations by silver staining. (A and B) At 3 days post capping. B is a higher magnification of the rectangular area in'A'. There is no evidence of a zone of necrotic tissue. As light inflammatory cell infiltrate is observed just below the Ca(OH)2 agent (arrow in'A'). Classic von Korff fibres (arrows in B) are observed. (C and D) At14 days after capping. D is a higher magnification of the rectangular area in C. Interodontoblastic VKFs (arrows in D) extend from the original dentine and connect to central pulpal fibres (CPF). (E and F) At 21 days after capping. F is a higher magnification of the rectangular area in E. The distinct expansion of predentine is observed from the periphery of the exposed area (arrow in E). VKFs (arrow in F) extend from the expanded predentine (PD) and connect to the central pulp fibres. (G-K)At 30 days post capping .H-Kare higher magnifications of the rectangular area h-k in G, respectively. The exposure site is almost completely occluded with a thin layer of a new dentine bridge (arrow in G). At the distant site from the exposure area, VKFs are observed within the newly formed dentine bridge, and reach up to its surface (arrow in H). Although classical von Korff fibres (arrowhead in H) are observed, no VKFs are seen in the pulp. At the periphery of the exposed area, VKFs are observed within the dentine bridge (arrow in I). New tubular dentine (_) without VKF was formed following the initially formed dentine bridge. VKFs extend perpendicularly from the remaining dentine chips (_in J) at the surface of the exposed area (arrow in J), whilst VKFs are randomly arranged beneath the chip-free surface (arrow in K). In this area, some defects were also observed (arrowheads in K). A, C, E, G: _40. B, D, F, H-K: x400.

There is no evidence of a zone of necrotic tissue

Immunohistochemistry of type I and III collagen and fibronectin.
At 21 days, immunohistochemical staining for type I collagen and fibronectin were detected in the VKF and fibres of the central pulp (Fig. 2A, C). Immunohistochemical staining for type III collagen was only lightly detected in the VKF and fibres of the central pulp (Fig. 2B). No positive immunohistochemical staining reaction was observed in any negative control sections incubated with normal rabbit IgG (Fig. 2D).

Figure 2. Immunostaining for type I collagen is detected in the VKF and the central pulp fibres (arrows in A). Immunostaining for type III collagen is hardly detected in the VKF, but slightly detected in the central pulp fibres (arrow in B). Immunostaining for fibronectin is detected in the VKF and the central pulp fibres (arrow in C). No positive reaction was observed in negative control sections reacted with normal rabbit IgG (D). x400.

Immunostaining for type I collagen is detected in the VKF and the central pulp fibres

Transmission electron microscopy.
At 14 days, the VKF were seen running in bundles towards the central pulp, passing through the newly formed odontoblastoid cells (Fig. 3A), consisting of two portions; a thick fibril (_240 nm in diameter; Fig. 3B) and a thin fibril (_80 nm in diameter; Fig. 3C). Both fibrils were connected to each other at an angle. VKF presented close morphological relationships with the central pulp broblasts (Fig. 4A). At 21 days, VKF were embedded in the new predentine, presenting a close relationship with broblasts of the central pulp (Fig. 4B). At 30 days, as odontoblastoid cells differentiated, they were seen attaching to each other, and hence VKF lost their close relationship with the central pulp (Fig. 5A) and as the dentine bridge developed, VKF were no longer observed near the odontoblastoid cells (Fig. 5B).

Figure 3. (A-C) TEM of the periphery of the exposure area at 14 days post capping.
(A) AVKF passing through odontoblastoid cells (OB) consists of two portions: the thick fibrillar portion (Tc) and the thin fibrillar portion (Tn). Both portions are connected to each other at an angle (arrow).
(B) Higher magnification of the thick fibrillar portion. These fibrils are ~240 nm in diameter.
(C) Higher magnification of the thinner fibrillar portion. These fibrils are ~80 nm in diameter.
A: x3000. B and C: x20000.

Size of the image files

Figure 4. TEM of the periphery of the exposure area.
(A) At 14 days post capping. Inset shows LM of the corresponding area and VKFs (arrows in inset) are found to pass through odontoblastoid cells (OB) and reach up to the central pulp. VKFs consist of thick fibrillar (Tc) and thin fibrillar (Tn) portions and show a close relationship with central pulp fibroblasts (F) (see arrow).
(B) At 21 days post capping. Inset shows LM of the corresponding area and VKFs (arrows in inset) are embedded in the expanded predentine. AVKF is embedded in the predentine (PD), passing through the odontoblastoid cells (OB), and shows a close relationship with the central pulp fibroblasts (F).
A-B: x2800. All insets x400.

Compression ratios with and without loss of information

Figure 5. (A) TEM of a site distant from the exposure area at 30 days after capping. Inset shows LM of the corresponding area. Odontoblastoid cells (OB) have well-organized cell processes (_and arrowheads in inset) and begin attaching to each other. VKFs (VKF and arrow in inset) reach up to the cell process of the odontoblastoid cells but do not continue to the central pulp.
(B) TEM of the periphery of the exposure area at 30 days after capping. Inset shows LM of the corresponding area and VKFs (arrows in inset) do not reach up to the odontoblastoid cells (OB). Although matrix fibres of predentine are observed (arrowheads), VKFs are no longer observed near the odontoblastoid cells (OB), A-B: Decalcified sections, x2800.All insets x400.

Mean grey values corresponding to the images that resulted from digital subtraction


Discussion - References.
Discussion.
During root dentinogenesis in the rat molar, the occurrence of collagen is irregular and does not contribute to predentine formation (Hayward&Webb1984, Salmon et al.1991). During secondary physiologic dentine formation, the pattern of interodontoblastic collagen fibre occurrence is irregular in cats (Bishop et al.1991). In this study, argyrophilic stained VKF within the dentine matrix of early dentine bridges were routinely detected. Classic von Korff fibres were postulated to help guide the odontoblasts in their pulp ward migration during dentinogenesis and to bind the soft tissue pulp and odontoblasts to the dentine (von Korff 1906). In the case of a pulp exposure, although the VKF is different from classical von Korff fibres in some ways, they are important for initial dentine bridging to induce and support a dentinogenesis framework, helping to guide the ‘preodontoblastoid’ cells in their migration, adhesion and arrangement, and dentine bridge formation. Furthermore, the VKF may provide a biological/mechanical tie that binds the fibres of the central pulp and newly formed odontoblastoid cells, become embedded in the predentine and are finally incorporated to serve asmatrix fibres in the early formed dentine bridge (Fig. 4).
VKF were quite different from classic von Korff fibres in their thickness and their silver staining intensity. Moreover, they consist of two types of collagen fibrils, a thick fibril (_240 nm in diameter) and a thin fibril portion (_80 nm in diameter). Since no similar findings have been described in the literature, this may be a unique structural feature beneath a hard-setting Ca(OH)2 agent as the capping material. Yoshiba et al. (1994) have suggested that the fibronectin-positive fibres may correspond to von Korff fibres. Fibronectin is an extracellular matrix glycoprotein distributed in the tissues and blood and has been demonstrated to induce reparative dentinogenesis (Seux et al. 1991, Tziafas et al.1992b). Yoshiba et al. (1996) suggested a fibronectin-rich matrix served as a reservoir for cell migration and attachment following direct pulp capping in human teeth. The results of this study shows the VKF stained positive for fibronectin, and this most likely corresponds with VKF fibres. Additional matrix components such as proteoglycans, fibronectin, or other collagen types may be important in the control of collagen-fibril structure (Birk & Trelstad 1984). Further immunohistochemical staining studies are needed to more precisely define, distinguish and differentiate VKF from classic von Korff fibres.
Initially, we observed VKF extending from the original dentine, passing through the odontoblastoid cells with a close relationship to the central pulp. Therefore, both odontoblastoid and pulp broblasts appear to be involved in the formation of VKF. Bishop et al. (1991) suggested a similar possibility with interodontoblastic fibres during secondary physiologic dentine formation in the root dentine. However, as the odontoblastoid cells begin organizing and attaching to each other, the VKF appear to lose their close relationship with fibres of the central pulp. After that, the VKF are no longer observed near odontoblastoid cells. These data strongly suggest that pulp broblasts are mainly involved in the formation of the thin fibril portions of VKF, and the thick fibril portion extends directly from the original dentine, possibly being produced by odontoblastoid cells.
Dentine and predentine contain only type I collagen (Takita et al.1987) whereas the pulp contains mostly type I, with up to 45% type III (Lechner & Kalnitsky 1981) and smaller amounts of type V collagen (Tsuzaki et al. 1990). Based on the thickness of fibrils, Bishop et al. (1991) speculated the interodontoblastic fibre was type I collagen. The thickness (80-240 nm in diameter) and immunohistochemical staining indicates the VKF in this study are type I collagen. Shroff & Thomas (1992) showed that the distribution of type I and III collagen is related to the degree of odontoblast differentiation and the reactivity of type I collagen observed between the cells. Thus, the reactivity of type I collagen is possibly related to initial dentine bridge formation.

Figure 6. Composite drawing of the sequence of VKF during early dentine bridge formation. This figure is made from (A) 14-, (B) 21-,(C) 30-day section, at some distance from the exposure area and (D) 30-day section, at the periphery of the exposure area.
(A) The VKF consisting of a thick and a thin fibril portion extends from the original dentine, passes through the odontoblastoid cells, and has a close relationship with the central pulp.
(B) As the predentine expands, the VKF becomes embedded in it.
(C) As odontoblastoid cells differentiate, they begin displaying well-organized cell processes and attaching to each other, and hence the VKF loses its close relationship with the central pulp.
(D) As the formation of dentine bridge advances, the VKF is embedded in the calcified dentine bridge and no longer observed near the odontoblastoid cells.

Immunostaining for type I collagen is detected in the VKF and the central pulp fibres

Tziafas et al. (1992a) suggested that the physicochemical properties of a surface to which pulp cells attach is a critical requirement for expression of their odontoblastic potential. The presence of residual dentine chips at the wound surface might be effective to promote a surface to which pulp cells will attach and differentiate into new odontoblastoid cells (Kitasako et al. 2000b). In this study, dentine chips remained at the wound surface and VKF extended perpendicularly from the remaining dentine chips. On the other hand, in the absence of dentine chips at the wound surface, VKF were randomly arranged with some defects observed within the new dentine bridge. At the periphery of the exposure, VKF seemed to serve as a guide for cell arrangement following tubular dentine bridge formation. In the absence of a basement membrane, adhesion of pulp cells onto an appropriate surface may be the critical requirement for the appearance of elongated, polarized, odontoblastoid cells (Veis1985).The small surface recesses like a tubular structure within dentine may support a favorable surface to which VKF can attach and so enhance pulp cell adhesion, arrangement and differentiation. Conclusions Interodontoblastic collagen fibres were routinely detected throughout early dentine bridges. Interodontoblastic collagen fibres may be important for initial dentine bridging to induce and support a dentinogenesis framework.

References.

Birk DE, Trelstad RL (1984) Extracellular compartments in   matrix morphogesis: collagen fibril, bundle and lamellar formation by corneal   broblasts. Journal of Cell Biology 99, 2024-33.
Bishop MN, Malhotra M, Yoshida S (1991) Interodontoblastic collagen (von   Korff fibres) and circumpulpal dentin formation: an ultrathin section study   in the cat. American Journal of Anatomy191, 67-73.
Cox CF, Bergenholtz G, FitzgeraldM, Heys DR, Heys RJ, Avery JK (1982) Capping   of the dental pulp mechanically exposed to the oral microflora - a 5-week observation   of wound healing in the monkey. Journal of Oral Pathology 11, 327-39.
Cox CF, Keall CL, Keall HJ, Ostro E, Bergenholtz G (1987) Biocompatibility   of surface-sealed dental materials against exposed pulps. Journal of Prosthetic   Dentistry 57,1-8.
Fitzgerald M (1979) Cellular mechanics of dentinal bridge repair using 3H-Thymidine.   Journal of Dental Research 58,2198-206.
Furseth R (1971) The fine structure of the odontoblast predentin area in   the root. Scandinavian Journal of Dental Research 79, 141-50.
Gomori G (1937) Silver impregnation of reticulum in parafin sections.   American Journal of Pathology 13,993-1002.
Hayward AF, Webb BW(1984) Interodontoblastic collagen fibres in the roots   of molar teeth of rats. Journal of Anatomy 138, 581-2 (abstract).
Higashi T, Okamoto H (1996) Electron microscopic study on interodontoblastic   collagen fibrils in amputated canine dental pulp. Journal of Endodontics   22,116-9.
Kitasako Y, Shibata S, Arakawa M, Cox CF, Tagami J (2000a) A light and transmission   microscopic study of mechanically exposed monkey pulps: dynamics of fibre elements   during early dentin bridge formation. Oral Surgery, Oral Medicine, Oral   Pathology, Oral Radiology and Endodontics 89,945-53.
Kitasako Y, Shibata S, Pereira PNR, Tagami J (2000b) Short-term dentin bridging   of mechanically exposed pulps capped with adhesive resin systems. Operative   Dentistry 25,155-62.
von Korff K (1906) Die entwicklung der zahnbeingrundsubstanz der saugetiere.   Archiv Fur Mikroskopischen Anatomie 67,1-17.
Lechner JH, Kalnitsky G (1981) The presence of large amounts of type III   collagen in bovine dental pulp and its significance with regard to the mechanism   of dentinogenesis. Archives of Oral Biology 26, 265-73.
Lester KS, Boyde A (1968) The question of von Korff fibres in mammalian dentine.   Calcified Tissue Research1, 273-87.
Reith EJ (1968) Collagen formation in developing molar teeth of rats.   Journal of Ultrastructure Research 21, 383-414.
Salmon JP, Septier D, Goldberg M (1991) Ultrastructure of interodontoblastic   fibres in the rat molar. Archives of Oral Biology 36,171-6.
Seux D, Couble ML, Hartmann DJ, Gauthier JP, Magoloire H (1991) Odontoblast-like   cytodifferentiation of human dental pulp cells in vitro in the presence of a   calcium hydroxide-containing cement. Archives of Oral Biology 36,117-28.
Shibata S, Fukuda K, Suzuki S, Yamashita Y (1997) Immunohistochemistry of   collagen types II and X, and enzyme-histochemistry of alkaline phosphatase in   the developing condylar cartilage of the fetal mouse mandible. Journal of   Anatomy191,561-70.
Shroff B, Thomas HF (1992) Investigation of the role of von Korff fibres   during murine dentinogenesis. Journal de Biologie Buccale 20,139-44.
Silva DG, Kalis DG (1972) Ultrastructural studies on the cervical loop and   the development of the amelodentinal junction in the cat. Archive of Oral   Biology17, 279-89.
Sogaard-Pedersen B, Boye H, Matthiessen ME (1990) Scanning electron microscope   observations on collagen fibres in human dentin and pulp. Scandinavian Journal   of Dental Research 98,89-95.
Takita K, OhsakiY, Nakata M, Kurisu K (1987) Immunofluorescence localization   of type I and type III collagen and fibronectin in mouse dental tissue in late   development and during molar eruption. Archives of Oral Biology 32, 273-9.
Tsuzaki M, Yamauchi M, Mechanic GL (1990) Bovine dental pulp collagens: characterization   of type III and V collagen. Archives of Oral Biology 35,195-200.
Tziafas D, Alvanou A, Kaidoglou K (1992a) Dentinogenic activity of allogenic   plasma fibronectin on dog dental pulp. Journal of Dental Research 71,1189-95.
Tziafas D, Kolokuris I, Alvanou A, Kaidoglou K (1992b) Shortterm dentinogenic   response of dog dental pulp tissue after its induction by demineralized or native   dentine, or predentine. Archives of Oral Biology 37,119-28.
Veis A (1985) The role of dental pulp - thoughts on the session on pulp repair   processes. Journal of Dental Research 64,552-4.
Yoshiba N, Yoshiba K, Iwaku M, Nakamura H, Ozawa H (1994) A confocal laser   scanning microscopic study of the immunofluorescent localization of fibronectinin   the odontoblast layer of human teeth. Archives of Oral Biology 39,395-400.
Yoshiba K, Yoshiba N, Nakamura H, Iwaku M, Ozawa H (1996) Immunolocalization   of fibronectin during reparative dentinogenesis in human teeth after pulp capping   with Ca(OH)2. Journal of Dental Research 75,1590-7.